Rewire After 25 Years?

[EFLImpudence]

There are two 1mm² CPCs in the ring?

 

[JohnW2]

There are two 1mm² CPCs in the ring?

Yes, I thought about that. However, if a fault occurs very close to one end of the ring, almost all the fault current will flow through the CPC of the 'short route', so I suppose one has to design on the basis of that 'worst case' scenario?

Kind Regards, John

 

[JohnW2]

If the 1mm² CPC in 2.5mm² T&E had to be increased, why was it there in the first place?

Good question. If the assumptions underlying my recent calculation (hence the answer!) are roughly correct, they should have realised that it was inadequate for a 30A BS3036 fuse.

Kind Regards, John

 

[EFLImpudence]

Yes, I thought about that. However, if a fault occurs very close to one end of the ring, almost all the fault current will flow through the CPC of the 'short route', so I suppose one has to design on the basis of that 'worst case' scenario?

True, but if all that is required is actually 1.18mm² then...

 

[JohnW2]

True, but if all that is required is actually 1.18mm² then...

... "then" (if that calculated figure is actually correct), given that the fault could be 'infinitely close' to one end of the ring, then the CSA of the CPC would have to be at least 1.18mm² - which, as I said, is clearly more than 1.0mm² - which, as I said, begs your question as to why 1.0mm² was ever used (in the days when 30A 3036s were probably 'the norm').

Kind Regards, John

 

[He who knows]

My guess is that nobody noticed. Or if they did they were reluctant to change, like they were with the old reg for ring-final cable current-carrying-capacity when someone noticed that 2.5mm² in conduit in thermal insulation didn't work.

 

[JohnW2]

My guess is that nobody noticed. Or if they did they were reluctant to change ...

Reluctant to change from what? Given that they were producing an 'equivalent' to 7/0.029", they surely must have known/noted/discovered that the CSA of its 3/0.029" CPC was about 1.28 mm² - and one would therefore have thought that they would have done some calculations to check 'adequacy' before 'rounding that down' to to the substantially smaller 1.0 mm², wouldn't one?

Why, for example, didn't they decide upon the 'obvious' 1.25 mm² - which would seemingly probably have been OK for a 30A BS 3036?

Kind Regards, John

 

[He who knows]

As I said last night, the problem I had was in trying to ascertain the appropriate value of I²t for a 30A BS3036 fuse to use for an adiabatic calculation. The best I can currently do is to look at the 'bottom end' of the characteristic curve for a 30A BS3036 in the Appendix of BS7671. That indicates a disconnection time of 0.1 sec at a current of about 430A.

That equates to an I²t of 18,490 A²sec.

....'worst case' scenario?

What if a 3036 never gets any faster as the current increases?

 

[JohnW2]

<paste from Fig 3A2(a) of BS7671>

Yes, I saw that table which suggested 450A for 0.1sec disconnection time with a 30A BS 3036, but I decided to go with the graph instead - not easy with the log scale, but clearly a bit lower than 450A, which is why I estimated ~430A.

So, you can pay your money and take your choice. My 430A gives an I²t of 18,490 A².sec, which leads to a minimum CPC CSA of ~1.18 mm², whereas your (the table's) 450A equates to an I²t of 20,250 A².sec, hence a minimum CPC CSA of about 1.237 mm - still obviously well over 1.0 mm², (and, for what it's worth, close to my suggested 'obvious' figure of 1.25 mm² ).

What if a 3036 never gets any faster as the current increases?
View attachment 224667

That's why, as I said, one needs to know the I²t at the actual PFC of the circuit - and neither the graphs nor the table in BS7671 go below 0.1 sec (hence above a PFC of 430/450 A, as above). If, as you have postulated, disconnection time remains as 0.1 sec when PFC rises above 430A (or 450A!), there's not actually any need to consider the actual value of I²t in the way you have done, since the calculation then simply reduces to:

minimum CSA = ( PFC x √0.1 ) / 115 = PFC / 364 (roughly!)

... hence, if PFC were above 546A and disconnection time was still only 0.1 sec (per your example), even a 1.5 mm² CPC would be inadequate! However, unlike the situation with an MCB, with a fuse I imagine that the disconnection time continues reducing for a fair bit beyond/below the bottom of the displayed curve (0.1 sec).

However, as I said, to actually get a proper answer, one needs to know the actual I²t at the PFC concerned.

Kind Regards, John

 

[He who knows]

neither the graphs nor the table in BS7671 go below 0.1 sec (hence above a PFC of 430/450 A, as above).

One can apply the Mk.1 eyeball curve extrapolator.

Might work better on a piece of paper without the fuse rating legend, but on a screen one can get some kind of idea.

 

[JohnW2]

One can apply the Mk.1 eyeball curve extrapolator. Might work better on a piece of paper without the fuse rating legend, but on a screen one can get some kind of idea.

Sure, but one doesn't have to rely on eyeballs to do the extrapolation - once one has managed to read a few points reasonably accurately off the curve, it can easily be mathematically modelled (in fact, the log-log curve is almost straight from about 150A onwards). However, what we obviously don't know whether the actual behaviour of the fuse beyond what is shown on the curve (i.e. above about 450A) is anything like what we would expect by extrapolation (whether 'by eyeball' or by mathematically modelling) ...

Kind Regards, John

 

[teaboyjim]

Are logarithmic graphs something that an electrician is expected to understand?

 

[Adam_151]

What if a 3036 never gets any faster as the current increases?

Unless you are assuming an arc which fails to clear (which 3ka could result in... IIRC the breaking capacity of most 3036 is 2kA, it can be 1, 2, 4 but I'm breatty sure type s4a is only appicable for ratings over 60A anyway)
then its not generally possible as I²t is related to the let through energy, and to the energy to vapourise a fuse element will be more or less a cosnstant. So i'd expect it gets closer and closer to a specific value of I²t the faster it happens, and infact for BS88 fuses at least, you can look up the manufacturers total I²t when you are looking at clearance times faster than 0.1sec. There is also a pre-arcing I²t if you need to ensure discrimination. (The total I²t of the smaller fuse must be below the pre-arcing I²t of the bigger one)

Its breakers that can exhibit the 'can't go any faster' problem, no quite so much of a problem for modern types of energy limiting class three, but you'll still get a chart from the manufacturer showing a different I²t value depending on fault level (or use the generic ones from BSEN 60898), but early ones did indeed have I²t increasing with the sqaure of the fault level. The 2.5mm/1mm can still be a problem on C curve breakers, especially if teh PEFC is high

 

[SUNRAY]

Is that a graph of logarithmic values? Is this something that a qualified electrician should be expected to understand because I can barely remember them and this type of calculations is Advanced and I don't think you'd cover that at GCSE mathematics level. It's used in A level sciences but not below that I don't believe

Oh that's scary, can I ask your age please?

 

[JohnW2]

Is that a graph of logarithmic values?

The actual 'values' (seconds and amps) are the actual values but, yes (as can be seen from the spacing of the values and of the dashed gridlines) both the vertical and horizontal axes of that graph are logarithmic. The range of values is such that, if that weren't the case, the graph would be impossible to use.

Is this something that a qualified electrician should be expected to understand ...

The blue curve in my graph is nothing more than a magnified version of part of one of the graphs for all types of fuses and MCBs which appear in the Wiring Regulations (BS7671) and are the standard source of reference for such information used by all electricians.

... because I can barely remember them and this type of calculations is Advanced and I don't think you'd cover that at GCSE mathematics level. It's used in A level sciences but not below that I don't believe

No mathematics or calculations, advanced or otherwise, is needed to use the graphs. For 'practical use', all that anyone (e.g. an electrician) needs to be able to do is to 'read figures' from the graph - and I'm sure that I was not alone is being able to do that before I left primary school.

Kind Regards, John