Voltage drop again

[JohnW2]

I’m sure we’ve been through this before, I but I remain a bit uncertain as to the basis one is meant (or allowed) to use when determining voltage drop in socket circuits to confirm compliance with 525 of the regs.

For a start, 525 is badly written. 525.1 and 525.100 (BGB) both relate specifically to fixed equipment, yet the next item (515.101 in BGB) introduces the concept of sockets, yet talks about what is deemed to satisfy “the above requirements” (which relate only to fixed equipment).

That aside, the acceptable voltage drops (with which we are all familiar) are given in Table 4Ab in 6.4 of Appendix 4 (BGB). A footnote to that table reads “The voltage drop is determined from the demand of the current-using equipment, applying diversity factors where applicable, or from the value of the design current of the circuit”.

If one has fixed loads, that’s straightforward enough. If one uses the ‘design current of the circuit’ for socket circuits, it’s also pretty straightforward – although here I have to ask whether one has to calculate on the basis of the (theoretically impossible) scenario of the entire design current (e.g. 32A for a standard RFC or 4mm² radial circuit) being drawn at the furthest point in the circuit.

However, my main question is this. If one has a 32A (or 20A) socket circuit, does the designer have to calculate voltage drop on the basis of the maximum demand which that circuit could serve (32A or 20A), or is (s)he allowed to exercise judgment, based on the location of the sockets and the nature of their probable use, to calculate voltage drop on the basis of the maximum probable demand? Situations in which this could be relevant would include supplies to distant outhouses or ‘distant’ upper floor bedrooms where demand was anticipated to be very low.

I suppose the official answer might be that if the designer believed that the demand was going to be low, (s)he should perhaps protect the circuit with a 6A, 10A or 16A OPD, as appropriate, rather than 32A or 20A one – in which case I presume that it would be acceptable to regard the rating of that OPD as the ‘design current of the circuit’ and use that for calculation of VD. However, what do people feel about the (more likely) situation in which there is a 32A or 20A OPD?

Of course, in the case of an 'outhouse', it's quite likely that a single supply will serve lighting as well as sockets, in which case the limiting factor will be the permissible VD for lighting (although perhaps one could cheat by having plug-in lighting!) - but it would still be the VD due to socket use which would be the primary determinant of the actual VD.

Kind Regards, John

 

[Sherbertlion]

I haven't looked at the reg you have mentioned but if it talks about applying diversity then surely you could use this as your basis from which to calculate voltage drop (as you have said using what you think will be the probable loadings).

Are you calcs demanding you out in a large sub-main cable?!

 

[JohnW2]

I haven't looked at the reg you have mentioned but if it talks about applying diversity then surely you could use this as your basis from which to calculate voltage drop (as you have said using what you think will be the probable loadings).

As was probably implicit in what I said, that would be my view - but I was wondering whether others agree. The problem with 'diversity' is that it's a pretty vague concept, open to considerable variation in interpretation - so I guess that's what I'm asking for views about.

Are you calcs demanding you out in a large sub-main cable?!

I don't fully understand what you're saying here, but .... my question is hypothetical but, yes, it would most commonly arise when there was a long sub-main or long cable run to a circuit in outbuilding.

Kind Regards, John/

 

[Sherbertlion]

I think using diversity with some judgment and common sense is perfectly acceptable but for ease and simplicity then using In as Ib for a ring or radial circuit with multiple socket outlets is probably the way forward...unless it means you end up having to run a huge cable!

I must say that I've never calculated VD using diversity, only In or actual load.

 

[EFLImpudence]

Edit - pressed button by mistake

 

[EFLImpudence]

Surely, the actual load - or probable load on a socket circuit - has to be used if there is any doubt.

As the current is the only variable in the circuit/equation, were In the required value to be used it would merely equate to a maximum length for any given circuit thus rendering the equation unnecessary.

I read on another forum of a circuit for four high powered lights. The advice was to use the amperage of all four lights on the first leg, that of three lights on the second leg and so on.

 

[JohnW2]

Surely, the actual load - or probable load on a socket circuit - has to be used if there is any doubt.

Sure. However, as you imply, with socket circuits, where the actual load will not be known, one would have to work with 'probable load' and my question related to how much design judgement is allowed here (give that Joe public could plug high loads into every socket he could find as soon as the electrician wqalked away).

In practical terms, I suppose what I'm saying/asking/thinking is this. Whilst I'm living in my own home, where I have a very clear idea about 'probable loads', I would be happy to have socket circuits which only just satisfied VD requirements with those (low) probable loads. However, does one have to consider the fact that a new owner might plug in fan heaters in several of those 'distant bedrooms' (something which is not anticipated in the 'probable load' whilst I'm living here)?

As the current is the only variable in the circuit/equation, were In the required value to be used it would merely equate to a maximum length for any given circuit thus rendering the equation unnecessary.

That's very true. For example, for a 2.5mm² radial socket (not lighting) circuit, I think the maximum lengths (if one assumes all load at the furthest point) would be around 106m, 64m, 40m and 32m for In of 6A, 10A, 16A and 20A respectively. However, I guess this is the 'no argument' approach, since it results in an acceptable voltage drop provided only that the circuit is not overloaded (by present or future occupants).

I read on another forum of a circuit for four high powered lights. The advice was to use the amperage of all four lights on the first leg, that of three lights on the second leg and so on.

That makes total sense, but only with fixed loads. When one has fixed loads, one can obviously calculate (or measure) VDs precisely, so there's no issue. It's the uncertainty about what will be plugged into socket circuits which causes the 'issues'.

Kind Regards, John.

 

[JohnW2]

I think using diversity with some judgment and common sense is perfectly acceptable but for ease and simplicity then using In as Ib for a ring or radial circuit with multiple socket outlets is probably the way forward...unless it means you end up having to run a huge cable!

Yes, that certainly would seem to be the 'no argument' approach - but see some of the other points/questions in my recent response to EFLI.

Kind Regards, John.

 

[EFLImpudence]

In practical terms, I suppose what I'm saying/asking/thinking is this. Whilst I'm living in my own home, where I have a very clear idea about 'probable loads', I would be happy to have socket circuits which only just satisfied VD requirements with those (low) probable loads. However, does one have to consider the fact that a new owner might plug in fan heaters in several of those 'distant bedrooms' (something which is not anticipated in the 'probable load' whilst I'm living here)?

If you did not install the circuit yourself I suppose whoever did would have done that for any circuit.
This would not be too much of an unexpected occurrence with electric fires in the past.
If In was used to calculate VD then all's well.
I would think that on new-builds conservative estimates would be employed.
As, I think, has been decided from previous threads the regulations do seem to be very conservative to start with.

That's very true. For example, for a 2.5mm² radial socket (not lighting) circuit, I think the maximum lengths (if one assumes all load at the furthest point) would be around 106m, 64m, 40m and 32m for In of 6A, 10A, 16A and 20A respectively. However, I guess this is the 'no argument' approach, since it results in an acceptable voltage drop provided only that the circuit is not overloaded (by present or future occupants).

Are you forgetting it can't be overloaded if In used.

 

[JohnW2]

If In was used to calculate VD then all's well. I would think that on new-builds conservative estimates would be employed.

Indeed - that's why using In is the 'no argument' approach.

Are you forgetting it can't be overloaded if In used.

No, I'm not forgetting anything. Rather, I'm remembering that In is not the same as I1 or I2 - and that the thing about socket circuits (and diversity) is that, although the designer's Ib should not be above In, the actual demand may sometimes actually exceed In.

Kind Regards, John

 

[sparkticus]

I suppose the official answer might be that if the designer believed that the demand was going to be low, (s)he should perhaps protect the circuit with a 6A, 10A or 16A OPD, as appropriate, rather than 32A or 20A one – in which case I presume that it would be acceptable to regard the rating of that OPD as the ‘design current of the circuit’ and use that for calculation of VD. However, what do people feel about the (more likely) situation in
which there is a 32A or 20A OPD?

Yes just for the sake of my own clarity on the subject I am thinking about a 30 meter radial run in 2.5mm. Assuming a regular load close to 20 amps then we are close to the VD limit of 5% (non-lighting)

So it is not difficult to bump into VD issues!

In my opinion diversity is a very tricky subject for sockets these days in the domestic world.

We use so many appliances and on one hand we are moving to "energy saving" and "smart power" appliances but that has only really gained any serious momentum in the lighting world as far as I can see. A single radial feeding two rooms can easily become overloaded these days and appliances at the end of the thing may well not be getting all the volts they rightly deserve.

This is one of several reasons why I like radials and lots of them. Ideally one per room and at least two in a kitchen.

I would not limit the rating of a protective device for VD reasons though. If you protect 5 sockets with a 10amp MCB then you will soon get a telephone call asking why it keep tripping. Protecting/limiting a long 4mm radial with a 20amp MCB seems reasonable.

If the circuit does need to be long (distant bedroom) then I would consider a 4mm radial or a sub-main with local CU.

Of course, in the case of an 'outhouse', it's quite likely that a single supply will serve lighting as well as sockets, in which case the limiting factor will be the permissible VD for lighting (although perhaps one could cheat by having plug-in lighting!) - but it would still be the VD due to socket use which would be the primary determinant of the actual VD.

In this case I would plan for a sub-main and local CU. You have somehow got to "fuse down" for lighting anyway.

 

[EFLImpudence]

No, I'm not forgetting anything. Rather, I'm remembering that In is not the same as I1 or I2 - and that the thing about socket circuits (and diversity) is that, although the designer's Ib should not be above In, the actual demand may sometimes actually exceed In.

If I remember correctly from that thread it was deduced that this must be incorporated in the allowed values and the cable must be capable of handling the current.

Also the centre point of a (2.5) Ring final is actually able to accommodate 54A. I realise that this would worsen the VD but, as above, this must be taken into account.
As at least two regular contributors frequently intimate the regulations are for blind obedience for your no argument scenario.

 

[JohnW2]

Yes just for the sake of my own clarity on the subject I am thinking about a 30 meter radial run in 2.5mm. Assuming a regular load close to 20 amps then we are close to the VD limit of 5% (non-lighting). So it is not difficult to bump into VD issues!

Indeed, as I said, the limit based on VD for a 20A 2.5mm² radial socket circuit with all of the load hypothetical at the furthest point is about 32 metres.

In my opinion diversity is a very tricky subject for sockets these days in the domestic world.

Exactly - hence my question. In fact, I suspect that we are 'over the peak' of demands on socket circuits, both because high load appliances (partiocularly 'fires'/heaters) are probably less common and because there tend to be more socket circuits per house these days.

I would not limit the rating of a protective device for VD reasons though. If you protect 5 sockets with a 10amp MCB then you will soon get a telephone call asking why it keep tripping. Protecting/limiting a long 4mm radial with a 20amp MCB seems reasonable.

I agree - the only reason for a downrated OPD would be to 'achieve compliance' (VD-wise), and I personally think that would be a bit silly. However, that takes me back to the original question - how would you feel about a (very!) long 4mm radial with a 20A MCB which only satisfied VD requirements with, say, the 10A total load which the designed regarded (currently) as the 'probable load'?

If the circuit does need to be long (distant bedroom) then I would consider a 4mm radial or a sub-main with local CU.

That's almost a matter of words, isn't it - since we'd then be talking mainly about the CSA of the 'sub-main' rather than the final circuit?

In this case I would plan for a sub-main and local CU. You have somehow got to "fuse down" for lighting anyway.

As above, this is almost semantic. Whether one has a local CU or an FCU (for the lighting), what we need to talk about is the cable supplying it - whether called a 'sub-main' or a 'radial final'?

Kind Regards, John.

 

[EFLImpudence]

This is getting to be a bit like - I know the answer but what if it isn't like that.

Obviously you cannot have a ring final to a kitchen 100 metres from the CU.

Therefore you have to work out what you can/must/are allowed to do.

 

[JohnW2]

No, I'm not forgetting anything. Rather, I'm remembering that In is not the same as I1 or I2 - and that the thing about socket circuits (and diversity) is that, although the designer's Ib should not be above In, the actual demand may sometimes actually exceed In.

If I remember correctly from that thread it was deduced that this must be incorporated in the allowed values and the cable must be capable of handling the current.

Yes, you may remember that's one thing which had me confused when I first joined this forum. In terms of curren-carring capacity it does, indeed, appear that the tabulated CCCs take into account the fact that I2>In - i.e. it is satisfactory to select cable on the basis of its CCC being at least as great as In, even if higher currents may flow before the OPD operates.

However, I've never seen any suggestion that there is any similar consideration for VD. I have always assumed that the maximum permitted VDs are the maximum permitted VDs - not that they are the maximum permitted VDs which would occur at the In of the OPD, with recognition that it would be acceptable for VD to be as high as 1.45 times that if the circuit were loaded beyond In.

Also the centre point of a (2.5) Ring final is actually able to accommodate 54A. I realise that this would worsen the VD but, as above, this must be taken into account.

Only if clipped direct. As you know, with Ref Method A, 2.5mm² RFCs would not even be compliant if it were only for Table 4D2A (CCC 18.5A, regs require at least 20A for deemed-to-satisfy RFC), so they had to invent Table 4D5 (in which the CCC magically rises from 18.5A to 20A ) to make Method A 2.5mm² RFCs compliant! Mind you, the I2 of a 32A Type B MCB is only about 46A, so 54A would not flow for very long.

Are you suggesting that the VD for a 32A RFC should be calculated on the basis of a 46A load?

Yes As at least two regular contributors frequently intimate the regulations are for blind obedience for your no argument scenario.

Indeed so.

Kind Regards, John.