Disconnection time formula?

[JohnW2]

In response to your last question ELFI, excel fits in my pocket, the regs book does not and I find it quicker to enter things into excel than flick through the regs book to remind myself of formulas, rewrite them, and work things out with a pencil and calculator, but each to their own!

As I said, there's no reason why you could not transcribe data from the graphs into 'tabular form' in Excel or whatever else fits in your pocket, if you really want to. It would not have to be particularly 'high resolution' data since, frankly, it's difficult to read particularly precise figures off those log-log graphs, anyway!

The thing I continue to find difficult to understand is how often you feel need to look at such information. I strongly suspect that most electricians almost never look at performancer graphs for OPDs!

Kind Regards, John

 

[skenk]

John ^ I don't think formulae are magic, and I'm not that arsed, just asked what I thought was a simple question. Seems the way breakers work and the physics of fuses blowing mean that it's not too simple or reliable to have a formula, I got that from the original thread I linked to and the first few replies here, ta

Yeah I suspect most/many electricians don't think about such things very much at all, they just install by rule-of-thumb, do fake £50 eicrs and suggest a rewire if they can't figure out what's going on!

 

[JohnW2]

John ^ I don't think formulae are magic, and I'm not that arsed, just asked what I thought was a simple question.

You did, and the initial replies you got were, I would have thought, simple answers to that simple question. As you go on to say ...

Seems the way breakers work and the physics of fuses blowing mean that it's not too simple or reliable to have a formula, I got that from the original thread I linked to and the first few replies here, ta

Exactly. Particularly with the 'breakers', the behaviour is really too difficult to 'predict' by theoretical calculations, since so many factors are involved.

Yeah I suspect most/many electricians don't think about such things very much at all, they just install by rule-of-thumb,....

There's not very much they need to think about in relation to these mattes, particularly in the case of MCBS and RCBOs. All they need to do is to ensure that (if designing/installing), or determine whether (if inspecting/testing), a circuit is such there is adequate overload protection (i.e. that cable, as installed, is appropriate for the device which is protecting it) and that an L-E or L-N fault will result in magnetic tripping of the OPD - and the spec of those devices tells one what the requirements are for that, without the need to use any graphs or formulae. That's just 'how it is', rather than 'rule of thumb'. Maybe I'm missing something, but I really can't think of any situation in which an electrician would need to look at the curves for a 'breaker', since the 'thermal' part of the curve is never going to be adequate for fault protection (i.e. result in adequately fast disconnection times), and the performance of the magnetic part is defined by the spec/Standard.

Kind Regards, John

 

[ban-all-sheds]

The relatively common melted cable reel extension ...

... has absolutely NOTHING to do with overcurrent.

 

[flameport]

on the end of a 4-way, at the end of 100m+ of extensions

Extension leads are intended to be used connected to a fixed outlet and not connected to other extension leads.

There is a reason extension leads are only available up to a certain length, usually about 30m, or in a few cases 50m.
It's the same reason the longer ones typically have 10A fuses in them.

If the impedance is high enough that the L-N fault current is too low to ensure disconnection within the required time, the voltage drop on the extension lead will also be so large that appliances connected at the end won't work properly.

 

[JohnW2]

If the impedance is high enough that the L-N fault current is too low to ensure disconnection within the required time, the voltage drop on the extension lead will also be so large that appliances connected at the end won't work properly.

Whilst I agree with the spirt of what you have written, I'm not sure that bit is necessarily true as a generalisation. Particularly if the Zs at the socket was already pretty marginal (for the circuit's OPD), an extra ohm or so could make a lot of difference to disconnection time but, at least in houses like mine, the resultant VD would still leave over 230V being supplied to whatever was on the end of the extension.

Kind Regards, John

 

[skenk]

John ^ I'm sure you'll correct me if I'm wrong but there does not appear to be any requirement that a L-N fault is disconnected within a given time or using the "magnetic part" of the breaker but rather only that the disconnection will occur before the cable overheats.

In practical terms with open circuit voltages which can be higher than 250v and equipment which will work fine at 220v volt drop is not always going to cause problems.

Aside from extension cables plugged together (I saw this recently for a boat house supplied from a house 100m away) there's things like festoon lighting running for 100's of metres and supplies to remote parts of music festivals as examples of where you need to look at L-N loop impedances/fault current limits.

 

[EFLImpudence]

there does not appear to be any requirement that a L-N fault is disconnected within a given time

411.3 and Table 41.1
It is for fault protection but if that is met then L-N will be covered.

or using the "magnetic part" of the breaker but rather only that the disconnection will occur before the cable overheats.

I would say you are, as with other things, looking at it the wrong way round.
The cable is required not to overheat before the disconnection.

Aside from extension cables plugged together (I saw this recently for a boat house supplied from a house 100m away) there's things like festoon lighting running for 100's of metres and supplies to remote parts of music festivals as examples of where you need to look at L-N loop impedances/fault current limits.

People do all sorts of stupid things which instructions have told them not to do.
It is not possible to design to prevent such things.

 

[skenk]

411.3 and Table 41.1
It is for fault protection but if that is met then L-N will be covered.

Those are disconnection times for L-E faults, which can be covered by an RCD, I'm looking at L-N faults where the loop impedance exceeds the given values.

I would say you are, as with other things, looking at it the wrong way round.
The cable is required not to overheat before the disconnection.

I said "...disconnection will occur before the cable overheats" it's another way of saying exactly what you said. Are you just trying to find ways to have a dig at me?

People do all sorts of stupid things which instructions have told them not to do.
It is not possible to design to prevent such things.

It's not design I'm looking at, it's understanding what would happen when installations which already exist have a fault.

 

[JohnW2]

John ^ I'm sure you'll correct me if I'm wrong but there does not appear to be any requirement that a L-N fault is disconnected within a given time or using the "magnetic part" of the breaker but rather only that the disconnection will occur before the cable overheats.

I think you're right but, as EFLI has said, if the requirements (which obviously are there) for disconnection times in relation to L-E faults are provided by an OPD (which essentially has to be the case in a TN installation), then the disconnection time for an L-N fault will be at least as short.

In practical terms with open circuit voltages which can be higher than 250v and equipment which will work fine at 220v volt drop is not always going to cause problems.

Indeed - that's what I just wrote. In fact, unless otherwise indicated by a product Standard (which is very rare), BS7671 assumes that non-lighting loads will work satisfactorily down to at least 204.7V (216.2V - 5%{of 230V}). The catch, of course, is that, although very unusual, it is permitted and not-impossible for the no-load supply voltage to be as low as 216.2V.

Kind Regards, John

 

[JohnW2]

Those are disconnection times for L-E faults, which can be covered by an RCD, I'm looking at L-N faults where the loop impedance exceeds the given values.

You keep saying that, but it's really only acceptable in a TT installation (and that only because there is no choice). In TN installations, RCDs are only considered to be 'additional protection', the requirement being that primary ADS be provided by OPDs. As far as I can make out, only in exceptional circumstances would it be acceptable to rely on RCDs for ADS in a TN installation "because one could not be bothered (or deemed it to be impractical) to make the circuit's Zs low enough for OPD-mediated ADS".

Kind Regards, John

 

[EFLImpudence]

Those are disconnection times for L-E faults, which can be covered by an RCD, I'm looking at L-N faults where the loop impedance exceeds the given values.

Pretend it's not TT and use the Neutral loop as already said.
If the given values are exceeded then repair the cause. It doesn't matter what the resultant time is.

I said "...disconnection will occur before the cable overheats" it's another way of saying exactly what you said. Are you just trying to find ways to have a dig at me?

It's not really the same. The OPD is chosen first in circuit design and then the cable which is sufficient for the rating.

It's not design I'm looking at, it's understanding what would happen when installations which already exist have a fault.

Depending on the fault, the cable might melt.
As you cannot predetermine the exact perameters of a fault there is no point trying to design a programme to work it out.

Reread and believe Post #4.

 

[JohnW2]

I said "...disconnection will occur before the cable overheats" it's another way of saying exactly what you said.

As I keep saying, I don't really understand how, at least in the case of an MCB, you think you could apply your calculations to that.....

If the fault current were sufficiently high to magnetically trip the MCB, then the cable would not overheat. On the other hand, if the current were not high enough for a magnetic trip, then the disconnection time would be at least ~10s, in which case you definitely could not use your adiabatic equation - so how do you think you might undertake your calculations?

Kind Regards, John

 

[JohnW2]

Reread and believe Post #4.

Indeed, and in particular trhe bit which says ...

No. As there are only a certain number of fuse ratings, and only a few disconnection times, the small chart to the right of the graph is really all that is required.

As I recently wrote, and although I may have 'missed something', I can't really think of when an electrician would need to look at the actual curves, whether in the context of design or inspection/testing. One could even wonder why they bother to include the curves in BS 7671.

Kind Regards, John

 

[skenk]

...it's really only acceptable in a TT installation (and that only because there is no choice). In TN installations, RCDs are only considered to be 'additional protection'

411.4.4 says quite clearly that RCD's can be used for (earth) fault protection in TN systems, nothing about only for additional protection or exceptional circumstances, am I missing something?