More on diversity (electrical)

[Andy Wade]

Interesting...

[js.b1]

100A 80A or 60A supply is protected by a DNO cutout fuse which permits considerably more to be drawn continually and intermittently. Likewise the supply cable is designed usually with this in mind (135A nomimal real capacity, although the dreadful farm growers may have disproved that at some poor devils expense?).

Not considering the real world fuse characteristics is why we can "meltdown" a 4-way strip adapter and "do a right number" on a BS1363 plug supplying a 2-way adapter with twin 2kW fan heaters.

Interesting how far they derate the 32A ring final tho, but I find derating an electric shower final circuit quite bizarre.

[John Rumm]

seem to have lifted that from NICEIC Connectons Mag Issue 160, Pg 51

(which was what I used as a source for the last section we added to the wiki on the judgemental approach ;-) )

[ARWadsworth]

> They seem to have lifted that from NICEIC Connectons Mag Issue 160, > Pg 51

see that the ESC version managed to get the correct figures into the chart!

[Andy Wade]

That figures. The ESC is the campaigning charitable wing of the NICEIC. Is /connections/ mag still on-line anywhere? I can't find it anywhere at niceic.com and my downloaded copies stop at no. 156.

[Andy Wade]

They're not suggesting that you derate the final circuit. The 80% diversity factor is applied to the circuit feeding the distribution board - in this case the domestic CU. The argument rests on the longer thermal time constant in the supply wiring (DNO's works) giving an averaging effect. This won't necessarily hold true for the final circuit, especially if someone's used 6 where they should have used 10...

Mind you I think they've miscalculated. The equivalent continuous current giving the same dissipation in the cables as a load with a duty cycle of 12/15 is sqrt(12/15), not 12/15. Therefore shouldn't the diversity factor be 89.4%, not 80%? Make it 90% say.

The other thing I don't quite buy is their use of a 40% factor for both ring ccts. The circuit supplying sockets for kitchen appliances surely needs a higher factor.

[ARWadsworth]

the later mags were some Flash based horror story. I sent John details of how to access them, maybe he still has that link.

[ARWadsworth]

You could probably reduce non kitchen circuits to well below 40%. Especially if you have upstairs, downstairs and kitchen rings.

[<me9]

Why? Most kitchen appliances are used very intermittently and/or are of low rating. A ring serving several othr rooms could have a fan heater on in each room for consifderable times in poorly insulated houses.

[John Rumm]

Any use:

on the other publications link at the top yields more)

[ARWadsworth]

That's the one. Ta.

[Tim Watts]

[e-mail] ( [e-mail]) wibbled on Wednesday 16 February 2011

17:17:

Also, a washing machine and dishwasher may be plate-rated at 3kW but the heater will come on for a small fraction of the time - same with the kettle.

[Andy Wade]

Fair point, although what we're concerned with is estimating the likely total maximum loading across all socket outlets, so it doesn't matter which particular circuit sees most load. Mention of kitchen appliances is natural because washing m/c + tumble drier + dishwasher could represent a significant sustained load, whereas widespread use of portable heaters is not so common nowadays. Extensive permanently installed electric space and/or water heating should of course be on its own circuits, with no diversity allowed.

[<me9]

Dish washer and washing machine unlikely to be on full load for more than 15 minutes, and unlikely to be on simultaniously om full load. Tumble drier perhaps, but 3kw for that, and 1½kw (averaged over an hour) for the others only amounts to 18A.