Diversity and the resilience of circuit protection.

But, we can't just keep adding, enlarging or increasing loads without serious engineering design considerations, or there may be serious consequences. Here is a mechanical example.

Well, the installation was never fit for purpose in the first place.

An interesting account of an event which I had never heard of despite living and working in the area for seven years or so.

Evening all.

I thought it was time to show my cards in support of Alan.

I did the period inspection and test on this installation. It is a large house belonging to a friend of mine. All of these large kitchen appliances are on a single radial. That radial is wired in 6/2.5 T&E grouped with other T&E, some of it enclosed and with some parts in insulation. The cable is protected by a BS 3871 Type 2 45A circuit breaker. That cable as installed is not protected against overload so does not meet Regulation 433.1.1. My observation for this was a C2. My report, my signature my liability. You may code the non-compliance differently to me but that is your report, your signature your liability.

I do know what diversity is and I do know it is not an exact science. The connected load is 21.kW which is 91.7A. The previous occupant was a woman living alone. The new occupant is my friend and could very well host a large diner party cooked by professional caterers hired in for the night. A lot of the connected load may be in use at the same time. So in these circumstances how would you design this installation and circuits? Would you put all that kitchen plant on a single 6/2.5 radial?

The installation is about 40 years old, a single 12 way old Wylex board, no RCD protection with many alterations and additions  by muppets. Many non-compliances, bunched circuits including 2 No. Air Con compressors,  low insulation resistance, unearthed Class 1 fittings etc etc. 

The supply was a 80A PME supply now upgraded to 100A by WPD. The supply supplies a large double garage and work shop, where the main switch is and a 70mm SWA to the house. The supply at the moment also supplies a new building with a further 95mm SWA. This building has not been used yet. A flat  over 2 large double garages. This building is going to be extended I have agreed with WPD for a further 100A supply to the new building but this will be TT as directed by WPD.

Alan is working his socks off fault finding, cable tracing and locating unknown circuits (as usual the consumer unit was not marked up) and clearing up the general mass of carp installed by the muppets over the years. I can vouch that he is very well qualified electrician, experienced and diligent person. If he was not he would not be doing this job without my endorsement and approval. Alan will be jointing the 95mm underground cable to extend it to the new external GRP cabinet for the new incoming supply, installing an earth mat in conductive concrete for the TT earth. Beyond the capability of the average electrician nowadays!

PS just looked at my NAPIT Code Breaker book that also say C2 for the absence of overload  protection for a cable.

I see two ways of improving the installation.

Option one, a new sub main from a 50 amp MCB with cable sized to suit, into a local consumer unit with a suitable sub circuit for each appliance. This has the merit that any appliances intended for use with a 13 amp plug may be so used, rather than being directly connected to a much larger circuit.

Option two, a pair of new circuits each 32 amp in 6mm with the appliances divided as well as possible between the two circuits. 

John Peckham: 
I do know what diversity is and I do know it is not an exact science. The connected load is 21.kW which is 91.7A. The previous occupant was a woman living alone. The new occupant is my friend and could very well host a large diner party cooked by professional caterers hired in for the night. A lot of the connected load may be in use at the same time. So in these circumstances how would you design this installation and circuits? Would you put all that kitchen plant on a single 6/2.5 radial?

John, thank you. I can picture it!

If I understand the OP correctly, there are now 2 x 32 A circuits, which should be fine.

Even if you have a couple of cooks, they are not going to use everything at once. We have not been told the age of the appliances, but presumably not ancient. The oven might go on first, but will probably take no more than 15 minutes to warm up. After that, the thermostat will click on and off; and there will be a second burst when food is put in until that gets up to temperature. Similarly, when boiling vegetables, you get the water up to temperature, add them and wait for the water to boil again before turning down to a simmer. Clearly, you do not put your peas on at the same time as your potatoes. If the vegetables are cooked in the steamer, there will be no need for the hob except perhaps for sauces. Similarly, if meat or fish is fried on the frying plate, the cooks will not be using the hob for the same purpose. And so on.

I am curious to know why the sub-mains are so large. Are the buildings so far apart that voltage drop is an issue? And how do you terminate such large cables in an ordinary CU?

Morning Chris

You are correct the large SWAs are for volt drop as the supply comes in to the garage near the entrance gate and then long runs to the houses. 25mm tails supply  Glasgow switch fuses at the supply end then the 70mm and 95mm from the switch fuse. Then junction boxes at the far end down to 25mm tails. 

This has been an interesting thread to watch. From my side, I think it highlights 2, perhaps 3 things.

1. Professional opinions vary - even when strongly expressed, and from longstanding and well respected forum members , it does not mean a consensus. This is not a bad thing, but it may worry those who like a black and white world. ( I'm not sure who I'd want as my ‘expert witness’ Depends perhaps if I wanted something edgy condemning or letting pass… )
2. There may be additional information not given. The cable sizes (assuming that's not a typo, or the other loads are such that it should have been 3 phase) imply a larger site, outbuildings and scope for more visitors/ extended family and more load. This is not a 2 bed flat with a super fancy show kitchen where the whole place may be on a 6mm or 10mm submain anyway.  Also the description of the general state means that it is  wise to assume the original design may be compromised,  or might even have been inadequate from the outset.
3. Diversity  estimation is at least partly an art not a science, and things like the change of use(r) and there fore change of likely load profile may tip something that has been OK for years into the ‘not OK any more’ category.

An illuminating read. 

And as running in a new cable is small beer compared to the rest, while it is all in bits it makes good sense to bring it more inline with regs.

Mike.

My personal opinion about diversity is that it often conflates two issues: safety and convenience/cost tradeoff. The latter is about about making reasonable assumptions about real-word situations: in a non-industrial setting, devices such as cookers will never be consuming full power 24/7. So it saves money to not over-engineer the circuit. The worst that should happen if the assumptions turn out to be faulty is inconvenience - an aspiring Jamie Oliver may discover that if all 5 hob elements are used for long periods, the CB trips. So they'll learn never to use more than 4. If this is an issue then they can get in a sparky to upgrade the circuit.

What should never happen is for the 5 elements in continuous use to not trip the CB but instead allow a cable to significantly overheat and get damaged.

Or to put it another way, getting diversity guesses wrong should at worst only cause inconvenience, not threaten safety.

And of course the thermal element in CBs already allows for reasonable temporary overloads.

Consider this, and I would give the same EICR code as John for the same reason, but I have started an interesting discussion! The best solution for the kitchen is not 2 32A circuits but another 6/2.5mm cable in parallel with the existing one. The reason is that the loads cannot easily be divided over the two circuits, and so tripping would still be a possibility with 32A breakers. Obviously, the existing 45A would still be satisfactory as it does not have a trip history. Cables in parallel for domestics are a bit unusual but certainly not a problem. Given the astronomical price of cable today, it is good to remember, two x 2.5mm in parallel has a higher rating than 6.0mm under most installation methods yet is considerably cheaper. The two cables should be similar in length and take the same route to ensure correct current sharing but may be somewhat spaced for cooling.

I wonder if anyone has tried the cable heating experiment yet?