Perhaps the title should be Diversity (Electrical).
- I think John needs to make it clear that the first table is used for
estimating the load for a particular final circuit and that the second
table is for the diversity for a group of circuits fed by a distribution
board (or, in the case of a house with a single consumer unit, for the
- "If the Power factor of the lighting units is unknown, then use the VA
rating times 1.8", err, no, use the rated lamp watts times 1.8 (and
divided by 230 V). If you had a VA figure for a luminaire there'd be no
need to multiply it by 1.8 :~)
More suggestions, based on the page as it stood a few mins ago:
I can't see why that shouldn't also apply to mixed tungsten and
discharge loads. (It would err on the safe side re phasorial addition in
in-phase and lagging currents.)
Also the meaning of "discharge" might not be clear to some of the
audience. I suggest "fluorescent" or "discharge (fluorescent)".
Suggest: "for fluorescent and similar fittings use the total VA load
rating (if known). This allows for ballast losses and power factor."
"If no VA load figure is available simply use 1.8 times the total
nominal lamp wattage." (It's lamp, not fitting, since the 1.8 factor
includes an allowance for ballast losses.)
Delete that. The wording suggested above makes it fairly clear that this
is to get a 'presumed' VA figure, I think. Certainly don't talk about
"peak current" - that could have two meanings: peak of waveform as
opposed to RMS, or maximum value of a time varying current. Both are
red herrings in the present context. "Full load current (FLC)" or
"maximum demand" are the pertinent terms, the former for a point of use
or a final circuit and the latter for the dis-board/CU/whole installation.
Agreed. Although of course if you know they've got modern HF ballasts
the PF will be very close to unity and the tubes will be running at
about 50 W, so the FLC is much less - more like 2.2 A. Best to err high
Suggest you add an example for the second table too, say two lighting
circuits, two rings, immersion heater and cooker. This is to get across
the point that you use the design currents of the final circuits (with
diversity where allowed) and not the MCB ratings. Something like this,
--- --- --------------- --- --------- ----
Cct MCB Load type Ib Diversity ADMD*
--- --- --------------- --- --------- ----
1. 32A Cooker (12kW) 23A 100% 23A
2. 32A Ring 1 (kitchen) 32A 100% 32A
3. 32A Ring 2 32A 40% 13A
4. 16A Imm htr (3kW) 13A 100% 13A
5. 6A Lights (1kVA) 4.3A 66% 3A
6. 6A Lights (600VA) 2.6A 66% 2A
. Total ADMD 86A
* ADMD = after diversity maximum demand
Helpful as ever.
One question. Would it be practicable and helpful to say something
about the way the maximum demand often ends up greater than the supply
fuse even after applying the factors in the 2nd table? That did my head
in for a while; and I'm still unclear what the regs. require/permit by
way of design. At the risk of displaying yet again the depths of my
ignorance I had in mind something vaguely along the lines of
"Even after making the allowances for diversity in the second table the
maximum demand on a consumer unit often works out as more than the
supplier's main fuse. This is not necessarily a cause to worry. The
allowances for diversity are conservative. [But it may help to use a
split load, dual RCD consumer unit so the maximum demand on each side is
no more than their rated 80A or 100A; or to use 2 consumer units so the
maximum demand on each is within the (usually) 100A capacity of the main
ISTR finding the advice in square brackets but I'm not sure how the
first works if there is still just one main switch; or how either works
if the whole thing hangs off a 100A REC. In other words, I'm not waving
Yes, perhaps "discharge (Fluorescent tubes, Sodium / Mercury Vapour
lamps, or any lamp technology that requires a current limiting ballast
a bit less chewy, ta.
As the lamp wattage is not so intimately tied to the length of the
fittings these days, that is better anyway.
Its surprising just how many recent off the shelf fittings still have
magnetic ballasts though. Not sure if the new rules on PFC will affect
I wonder which is cheaper to make, HF ballast with built in PFC, or mag
ballast, and separate cap?
Yup, very good idea. I realised he wording allowed for ambiguity as to
which current one uses, and was thinking about adding a note, but an
example will do nicely ;-)
Yup, I think Andy''s point about adding an example to the second table
ought to address that.
There is as written (for the next few minutes anyway ;-) a avenue for
misinterpretation (its in the OSG as well in fact) that does not really
make clear which numbers you use on the second table. If you just apply
the second table rules to a list of MCB values, then you will almost
certainly end up with a number way too large.
So the first table helps assess the design load of a circuit, and then
the second one applies further diversity to a group of circuits.
I can see the logic of spreading circuits out over several RCDs (for
many reasons rather than just load), but the point that they tend to
have smaller maximum switch ratings is also valid. Not sure that the
logic really applies to split load in general though, since the same
main switch typically switches both sides at once.
Have a read again after the next update and see if it is any better...
=A0 =A0 =A0 =A0|
=A0 =A0 =A0 =A0|
I'm assuming when you mention sockets other than 2A lighting and 13A
power, you mean things like 15A round-pin, and 16A commando - where
they're occasionally used in domestic settings?
What about 5A round-pin lighting sockets? Hardly appropriate to rate
the 2A version and 0.5A load (115W) and the 5A at 5A (1150W).
I believe that is the intention, yes.
The OSG does not mention them explicitly. I expect this is a case rather
like with 15A round pin and 16A commando connectors mentioned above,
where you need to apply actual knowledge of the installation to come up
with a sensible answer. For lighting purposes, 2A sockets would seem
more than adequate, so its harder to think of applications for 5A ones
that you can make blanket assumptions about.
For 2A or 5A on a lighting circuit I'd treat as per any other lighting
point - 100 W per outlet - unless indicated otherwise. No-one's going
to plug any other appliance into them these days.
I saw some guidance on IEC 60309 outlets somewhere recently (might have
been in /Wiring Matters/). One or two 16 A outlets per 16 A circuit,
AFAICR, again in the absence any contra-indication. They're unshuttered
of course so aren't allowed for general use in a domestic installation -
OK if out of reach of younger children in a workshop though, IMHO.
I like the example but it doesn't really deal with the point I had in
Yes thanks - I made that mistake early on but have got beyond it.
It is getting better and better. But I am still left with the problem
that some modern installations - especially but not only where people
cook electric and have an electric shower - seem to have maximum demands
in excess of the fuse/isolator. I am sad enough to have looked at
neighbours and tried to do the sums (making best guesses about numbers
of lights etc). Several end up well over 100A.
I don't think I've I mucked up the sums. Mine look much like the
example in the Wiki. (Indeed, taking the example there, if the owners
asked for an electric shower and there was a spare way in the CU should
they be told it's not practicable because it'd take the maximum demand
Of course they may just be bad installations. But I am still left
wondering about the OSG. I say that in part because I have now tracked
where I read about it being conservative and about using a split board.
It was in "Advanced Electrical Installation Calculations" published by
Newnes. They had an example where the MD after applying all the OSG
factors was 143A. And the text which followed was:
"In this case as in many other cases, the assumed maximum demand is
still greater than the supply fuse.You will find that the supply fuse
has been in place for many years and never been a problem and should not
give rise for concern. This is because the diversity allowed is usually
quite a conservative figure. The greatest problem with this situation
is that the switched disconnector is usually 100A. In these cases
consideration should be given to using a split load board with possibly
the power circuits on an RCD protected side of the board."
I'm sorry if I'm flogging a dead wotsit. Should I just give up worrying
on the basis I'll never learn the guild's secret handshake either :-) ?
I was avoiding the issue, although the type of installation it describes
would often have only a 60 A main fuse.
I think the underlying problem is that the OSG guidance is out of date.
I don't think it's changed since the 14th edition, possibly even the
13th! Modern installations tend to have more final circuits, an
approach encouraged by the 17th edition and RCD-related issues.
To a good approximation domestic MD is probably more related to the size
of the house than anything else. Normal size houses (up to 4 or 5
bedrooms, say, seldom seem to have any problems on a 60 or 100 A supply
(cannabis farms excepted).
It would not be unreasonable to bring down the 40% value on remaining socket
circuits. I don't think that I could load my upstairs sockets to 13A if I
plugged in all my electrical appliances that I keep upstairs.
A 4/5 bedroomed house with teenagers having showers and using hairdryers
whilst dinner is been prepared and the washer and tumble dryer are on may
draw over 100A for a short time, however I doubt the main switch or fuse
will have a problem with that - we are talking a few minutes here not hours.
I suspect Andy chose values deliberately to avoid the question! ;-)
I might go and add another circuit to deliberately provoke the point!
I think this is one of those occasions where you need to apply local
knowledge to override the guideline. For example, the calculations on
socket circuit loadings are pessimistic in two ways... Firstly in terms
of load per circuit; where with the possible exception of kitchen
circuits it is going to be exceedingly rare to reach 32A of load.
Secondly, the calculation for load over the remaining circuits is also
rather restrictive... saying 100% of first, 40% of second - is ok IMO,
but then I would tend to ignore all the remaining socket circuits or
count them at a nominal values based on actual load rather than also add
40% of those.
So in my house for example I currently have 4 ring circuits protected by
32A MCBs. Kitchen / utility room, might be able to get close to that -
but probably only by including very short term loads like kettles etc.
Ground floor ring will typically struggle to reach 10A, and only then
manage it when running the vacuum for example - the rest of the load is
electronic niff naff. Upstairs will be even lower. The workshop one
could potentially get up to about 25A load in the winter with some
heating running, dust extraction, and a power tool - but even then would
have difficulty sustaining it.
I still don't follow the logic of the advice - since split load boards
will typically connect all "sides" back through the main switch.
Splitting across multiple CUs however makes more sense.
No I quite like it when folks do flog dead wotsits, since it tends to
elicit the sort of answers that are not typically provided in the more
formal publications, and yet are known as a matter of lore to some of
the trade. After all the purpose of the wiki is not to save people
buying the OSG etc, but provide the info required by DIYers in a quickly
digestible form with practical advice in there as well as the theory and
I'd be relaxed about that approach for a Scout badge but as the basis
for a BS which has statutory force by virtue of the building regs. I'm
unimpressed by the combined might of the IET and ODPM as was. Anyone
installing a new CU under a building notice and not working within the
very conservative OSG seems to risk a nasty surprise from the appointed
Yes, thanks. (I started worrying about this after splitting a
downstairs ring into 2 rings - one for the kitchen and one for living
room and hall. It seemed bizarre that as a result of that alone my MD
had increased beyond the meter's rated capacity.)
Many thanks. (And if I may say so I like very much the way you bring
the total in just below 100A.) I'll shut up now.