Splitting 10mm cable for multiple ovens

[winston1]

She has further ordained that all of the above will be routed through a grid switch array; one switch per appliance.

You should point out to her that this is completely unnecessary and is just the latest fad.

Also why five ways. There only seem to be 2 ovens, warming drawer, and igniter. That's four ways.

 

[imroberts]

Your appliances cannot cause an overload.

Short circuit (L to N) and Earth fault (L to E) currents might be hundreds of amps and cause instant disconnection.

Interestingly this is something I debated with myself when recently rewiring my own kitchen. I went to the other extreme and now pretty much everything has a dedicated feed and MCB of 16A or less. My own thought process was the oven itself can't cause an overload, but a short circuit, let's say to the oven casing, could. I hadn't considered that this latter event is likely to caue instant disconnection so the cable wouldn't be overloaded anyway.

Not that I would want to, but from what you're suggesting I could technically wire up a 13A oven using 1.5mm T&E and a 50A MCB/RCBO, and that would be fine?

 

[flameport]

wire up a 13A oven using 1.5mm T&E and a 50A MCB/RCBO, and that would be fine?

Perhaps.
Depends on the temperature rise of the cable when a short circuit occurs.
Calculations required.

 

[EFLImpudence]

Not that I would want to, but from what you're suggesting I could technically wire up a 13A oven using 1.5mm T&E and a 50A MCB/RCBO, and that would be fine?

The only consideration is whether that 1.5mm² can withstand the fault current without damage. There are calculations which can be done using the actual prospective fault current.

Other than that then yes - and its not my suggestion as such but following regulations and physics which most do not bother with.

 

[JohnW2]

The only consideration is whether that 1.5mm² can withstand the fault current without damage. There are calculations which can be done using the actual prospective fault current. Other than that then yes - and its not my suggestion as such but following regulations and physics which most do not bother with.

All true, but I think that the powers wot be might be able to make it more likely that more people would 'bother' to undertake the calculations ....

... the information sources commonly used by electricians (BS7671 and OSG) do not contain immediately usable information on the actual disconnection time (or I²t) of an MCB at the PFC (the best one can do is make an 'intelligent guess' from the t/I curves presented in BS7671) that is needed for an adiabatic calculation, and I suspect that may well put some people off attempting such a calculation.

Whilst I've seen it suggested that a reason for this 'omission/absence' is that I²t will vary between makes/models of device, that seems to be a fairly lame excuse, since it presumably also applies to some extent to all the 'generic' curves in Appendix 4 of BS7671 - and they could presumably present at least 'maximum' figures.

Kind Regards, John

 

[flameport]

When calculations are required, manufacturer's data for the specific protective device should always be used.
Generic info as contained in BS7671 is only there as a guide, and using that will typically give very conservative results, such as oversized conductors, maximum impedance values that are rather low, etc.

 

[JohnW2]

When calculations are required, manufacturer's data for the specific protective device should always be used. Generic info as contained in BS7671 is only there as a guide, and using that will typically give very conservative results, such as oversized conductors, maximum impedance values that are rather low, etc.

Well, yes, as I hinted in my last paragraph, I assumed that was the reason - and I obviously agree that what you say represents the ideal.

However, it remains the case that this is probably one of the reasons why adiabatic calculations are not undertaken as frequently as they might otherwise be. If they had easy access to at least generic data, I imagine that many more would undertake the calculations, even though, as you say, the results they got would quite possibly be conservative.

I also do wonder a little about the extent to which manufacturer's I²t data really is empirically determined for each and every of the protective devices they manufacture. The below comes from the technical data for MK Sentry MCBs. For genuinely empirically-determined product-specific data, the curves are incredibly linear throughout the range 1kA - 6kA, implying a precisely linear inverse relationship between current and disconnection time within that current range - which I find a little 'suspicious'!

In passing, I would also note that the figures as 6kA imply disconnection times ranging from 0.39 ms (for B6) to 1.9 ms (for B63) - whilst I have to assume that those figures are probably roughly correct, I have to say that I'm actually surprised that (given inertia of parts etc.) anything electro-mechanical can operate quite that fast! I would also add (since it is a 'pet hate' of mine!) that MK (who should know better!) describe I²t as "energy" - which it isn't!

However, of most practical importance (and relevance to this discussion) is that even if an electrician took the trouble to seek out that information, the curves only start at 1 kA, whereas I presume that an awful lot of installations (like mine) will have a PFC less than that - so back to 'guesswork' (or, at least, {not definitely valid} extrapolation).

Kind Regards, John

 

[EFLImpudence]

Isn't the graph only linear because it has been drawn like that by adjusting the axes divisions?

 

[JohnW2]

Isn't the graph only linear because it has been drawn like that by adjusting the axes divisions?

In a sense, that is true (even though both axes have log scales). However, my point was that, no matter what axes one uses, to get a series of 'perfect straight lines' (with whatever axes one is using) is not what one really expects of empirically-measured data. However, maybe I'm too cynical!

Kind Regards, John

 

[JohnW2]

... also, there is surprising variation between mnufacturers., The corresponding curves below from Wylex indicate an I²t of about 40,000 A².sec at 6 kA for a B40, whereas the MK curves give a figure of a bit over 50,000 A².sec.

Is it really the case that there is such a big difference in I²t (hence also in the results of adiabatic calculations) for B40s of two different makes? If that is true, I would regard it as a bit worrying.

[ Note that although (unlike MK) Wylex don't (incorrectly) call I²t "energy", they have the units totally wrong ("amp²/sec", rather than the correct "amp².sec")! ... and I'm not too sure what "Ip/A" in the x-axis labelling is all about, either ]

I would also repeat a comment I've made in the past that, at least in mathematical terms, it's rather odd to plot I²t again I - it would be far more logical to just plot t against I (as in the BS7671 MCB curves).

Kind Regards, John

 

[JohnW2]

When calculations are required, manufacturer's data for the specific protective device should always be used. Generic info as contained in BS7671 is only there as a guide, and using that will typically give very conservative results, such as oversized conductors, maximum impedance values that are rather low, etc.

I also do wonder a little about the extent to which manufacturer's I²t data really is empirically determined for each and every of the protective devices they manufacture. The below comes from the technical data for MK Sentry MCBs. For genuinely empirically-determined product-specific data, the curves are incredibly linear throughout the range 1kA - 6kA

Having thought and looked a little more deeply, I am now even more inclined to think that one should be careful not to over-interpret what a "manufacturer's data for the specific protective device" actually is ...

... if you look at the two t/I curves for MCBs below (one Wylex, the other MK), or the corresponding curves in the Appendix of BS7671, it is fairly clear that what one is seeing does not relate to the "specific protective device". Rather, it is merely showing (at least in terms of magnetic trip) what is required by the product Standard - e.g. magnetic tripping at 3In - 5In for a Type B (BS7671 just shows the 'worst case' upper figure, but MK and Wylex show both). This data clearly has not been derived from testing of the products concerned (beyond confirming that the products comply with the Standard). The "manufacturer's data" is, therefore, essentially 'generic'.

If that is the case for those curves, I wonder if the same may also be true for the manufacturer's I²t ones - i.e. are they perhaps not showing how the actual product performs but, rather, are reflecting some requirement (e.g. for a certain maximum I²t at peak rated current - i.e. 6kA ) in the product Standard (does anyone know if there is such a requirement?). If that were the case, then the "manufacturer's data" would, again, be essentially generic.

WYLEX

MK

Kind Regards, John

 

[John D v2.0]

Johnw2 i think the t that you can calculate is not really the disconnection time, i think your confusion is caused by assuming that. Actually the graph shows the "effective" i2t that you should use even though the mcb can't move that fast, presumably taking into account ac cycles and inductances and all the rest.

 

[Sureitsoff?]

You could use a dual plate cooker connection unit.

10mm in

And 4 or 5 2.5mms out

60A junction box might give more room for the 4 (or5) smaller cables


https://www.googleadservices.com/pa...hUKEwiIh5aQgvTlAhWYiFwKHdBOBjUQ9aACCEk&adurl=

https://www.googleadservices.com/pa...hUKEwiIh5aQgvTlAhWYiFwKHdBOBjUQ9aACCEk&adurl=

https://www.googleadservices.com/pa...hUKEwiIh5aQgvTlAhWYiFwKHdBOBjUQ9aACCEk&adurl=

 

[JohnW2]

Johnw2 i think the t that you can calculate is not really the disconnection time, i think your confusion is caused by assuming that. Actually the graph shows the "effective" i2t that you should use even though the mcb can't move that fast, presumably taking into account ac cycles and inductances and all the rest.

I once thought that, but when I started thinking, it didn't seem to make much sense.

Apart from anything else, adiabatic equations like the one presented in 543.1.3 of BS7671 are written in terms of the PFC ("I") and the disconnection time of the device at that current ("t") (those variables being defined as such) - NOT in terms of "I²t" (as an 'entity'), let alone "effective I²t". Hence if the manufacturer's data refereed to some sort of "effective I²t", calculations based on it (rather than true I and t) would seemingly be 'non-compliant'.

In any event, just from 'basic principles', I²tR (where I is the current, t is the period for which it flows {until disconnection) and R is the resistance of the conductor) undeniably IS the amount of energy dissipated in the conductor as heat and, if adiabatic conditions pertain (i.e. not enough time for appreciable heat loss from the conductor), then all that heat will be reflected in the temp rise. Hence, from adiabatic principles, it IS "PFC squared times disconnection time) that matters.

As I've said, I think it is rather confusing that protective device performance curves generally plot t against I for disconnection times down to 0.1s (sometimes 0.01s), but plot I²t against I for shorter disconnection times (although the 'stylised' MK one I posted yesterday goes down to below 1ms). However, have now done a fair bit of looking around, I'm becoming increasingly inclined to believe that the published I²t/I curves are nothing more than a (convoluted) 'downward continuation' of the more familiar t/I ones.

If the "I²t" plotted in the published curves related to anything other than "I squared times t" (i.e. "PFC squared times disconnection time"), then it surely should say so. If is does mean that, then it is trivial to turn an It plot into a t/I one.

Are there flaws in anything I've written above? One issue is that of disconnection times less than one cycle, but (given that the time of onset of a fault is unpredictable), I would think that one has to assume the 'worst case' - of current flow being centred on the peak of a cycle.

Kind Regards, John

 

[John D v2.0]

Sorry missed your reply there.

One issue is that of disconnection times less than one cycle, but (given that the time of onset of a fault is unpredictable), I would think that one has to assume the 'worst case' - of current flow being centred on the peak of a cycle.

I think that's the issue. PFC is the average over at least one full cycle where there is pretty much only a 50Hz component to the waveform.
I'm not sure when the worse case would be, possibly is the peak of a cycle as you say.
In that case, if a fault occurs then it would theoretically be a square edge increase of thousands of amps (16kA), and it's very hard to get thousands of amps to either start or stop flowing in a millisecond. This would be due to the inductance of the fault path.
Presumably their measurements take into account the expected growth in current on whatever the largest transformer the DNOs would use, and then they just work out the tripping time and take the area under the current squared over time graph.