Twin and earth used as meter tails

[radiohead]

Give the fella a break FFS, all he's going to do is replace the cpc in a T/E with a short bit of 16mm and you are acting like its the crime of the century.
Crack on and replace it mate if that's what you want to do, it's certainly not going to reduce the safety of the installation.

 

[Mark Wright]

While we're on the subject.........

Can you show me with adiabatic equation sizing the Earthing conductor - just make a guess on the Fuse and I can look up the curve etc this will be great for my learning.
Thanks.

 

[Spoon]

While we're on the subject.........

Can you show me with adiabatic equation sizing the Earthing conductor - just make a guess on the Fuse and I can look up the curve etc this will be great for my learning.
Thanks.

Have a look at page 196 of the BBB mate.

 

[Baddegg]

As per spoon mate....also if you are a trainee @Mark Wright PM @Dan and they’ll be able to give you access to the trainee section ?

 

[Mark Wright]

I know that, ok ill need a pen a paper, what's our fault current if we guess Ze =0.8
230/0.08 = 287A

Need a guess from someone on the type of fuse???
we're say 100A
How long will DNO fuse take to Blow with a current of 287A ?

 

[Dustydazzler]

Yes very common , loads of council estates built in the 60s and 70s in my area which have 10mm or 16mm twin and earth from the meter to the fuse box. Usually just poked up through the cavity wall...
then moving to the 80s/90s builds usually find 10mm or 16mm tails...

 

[Snailspark]

For a new installation yes that is standard but for an existing installation you need to assess the suitablity of what is already installed.
Only a cowboy would charge a customer for an 'upgrade' when the existing installation is not unsafe.
[automerge]1592088004[/automerge]


Have you carried out the calculation using the adiabatic equation in BS7671?
Generally speaking the 6mm earthing conductor you have is more than adequate as long as it does not need to perform the function of a main bond (which it doesn't appear to here since the bonding connects directly to the cutout)

Just a small thing I know but page 48 OSG Table 4.4(ii) shows minimum csa of main earthing conductor for tncs - no calculations necessary - simple table and all relates to the line conductor csa - my understanding from this is that it needs to be 10mm as an absolute minimum or am I seeing it differently?

 

[pc1966]

I think these days the DNO fuses would be BS88-3 but hopefully someone who knows more will say?

A quick look at Fig 3A1 in the BBB suggests time to blow is around 3 minutes at 287A! At that sort of time-scale the adiabatic calculation is pessimistic as heat is also being lost from the system.
[automerge]1592126372[/automerge]
They may be on a 60A fuse of course, in which case something like 10 seconds

 

[Mark Wright]

I think these days the DNO fuses would be BS88-3 but hopefully someone who knows more will say?

A quick look at Fig 3A1 in the BBB suggests time to blow is around 3 minutes at 287A! At that sort of time-scale the adiabatic calculation is pessimistic as heat is also being lost from the system.
[automerge]1592126372[/automerge]
They may be on a 60A fuse of course, in which case something like 10 seconds

Sir, excellent ill try and have a look throughout today I think we should stick with 100A as that's what current recommendations are based on.

Now K isn't going to be 115 is it?
[automerge]1592126749[/automerge]

Just a small thing I know but page 48 OSG Table 4.4(ii) shows minimum csa of main earthing conductor for tncs - no calculations necessary - simple table and all relates to the line conductor csa - my understanding from this is that it needs to be 10mm as an absolute minimum or am I seeing it differently?

OSG recommends tails of 25mm so a 16mm earthing conductor would be needed, some people say you can use smaller with adiabatic equation, don't think anyones ever actually done it though lol.

 

[Baddegg]

You wouldn’t do it from a new installation on domestic, what would be the point, but if doing an eicr for example which is now going to be a lot more regular with new legislation, then you may need to confirm older installs....

 

[Mark Wright]

Sorry Fig 3A1 - I'm see about 5 seconds?

 

[pc1966]

Sir, excellent ill try and have a look throughout today I think we should stick with 100A as that's what current recommendations are based on.

Now K isn't going to be 115 is it?

But the DNO fuse is not being used for protecting your circuit!

The OCPD that you should consider is the biggest of the CU as it would set the disconnection time for the biggest likely fault.

Of course if you are using tails > 3m and so are providing a fused-switch for that then you really ought to check it meets the 5s disconnection time which is not happening if you have a 100A fuse and the typical max TN-S Ze of 0.8 ohm...
[automerge]1592127453[/automerge]

Sorry Fig 3A1 - I'm see about 5 seconds?

Yes, for 63A is is probably closer to 5s, the table on the right has 320A for 5s disconnection.

 

[Mark Wright]

But the DNO fuse is not being used for protecting your circuit!

The OCPD that you should consider is the biggest of the CU as it would set the disconnection time for the biggest likely fault.

Of course if you are using tails > 3m and so are providing a fused-switch for that then you really ought to check it meets the 5s disconnection time which is not happening if you have a 100A fuse and the typical max TN-S Ze of 0.8 ohm...
[automerge]1592127453[/automerge]

Yes, for 63A is is probably closer to 5s, the table on the right has 320A for 5s disconnection.

No 100A between 5 and 10seconds ?
[automerge]1592128157[/automerge]
Yeah I'm trying to calculate what was suggested here, that the Earthing conductor is large enough to withstand fault until the DNO fuse blows.

 

[pc1966]

No 100A between 5 and 10seconds ?

If you have Fig 3A1 next to you, look at the table which has 5 secs for 100A needing 580A. So there is no way that half that current is going to be just a little different!

If I try to guesstimate where 270A is on the log scale and go up to the highest (100A) curve, I am seeing somewhere between 200 and 300 seconds.

 

[Mark Wright]

I'm sorry I can't see that? why would 580A be 5 seconds and 270A be 300 seconds?

 

[pc1966]

I'm sorry I can't see that? why would 580A be 5 seconds and 270A be 300 seconds?

Because fuse disconnection time decreases very rapidly with current!

Doubling the fault current leads to 4 times the heat in the fuse from I2R, so it will be at least 4 times faster, but also as the fuse element heats up its resistance increase further increasing the heating of the element so you quickly get thermal runaway and disconnection.

The reason that fuses have such good high-energy fault energy liming comes down to this aspect - they will limit the peak fault current far faster than MCB/MCCB can open their contacts and suppress the arc in the air-brake chute.

 

[Mark Wright]

Because fuse disconnection time decreases very rapidly with current!

Doubling the fault current leads to 4 times the heat in the fuse from I2R, so it will be at least 4 times faster, but also as the fuse element heats up its resistance increase further increasing the heating of the element so you quickly get thermal runaway and disconnection.

The reason that fuses have such good high-energy fault energy liming comes down to this aspect - they will limit the peak fault current far faster than MCB/MCCB can open their contacts and suppress the arc in the air-brake chute.

Ok thanks, I've got some stuff to do now but if we go with 580A for our fault current.
but can I come back to you about the 270A because I'm looking at the curves and not quite seeing so would be good learning?

 

[Baddegg]

On the curves go along the bottem with your selected/calculated fault current then go up until you meet the curve for your fuse size (assumes 100a in this scenario which will be the furthest curve to the right I think I havnt got my regs book with me)....then left from that point and that’s ya time in seconds mate,

 

[pc1966]

Here is my example:

A bit more effort to get log-log interpolation, etc, would make it more accurate but it is around the right value.

 

[Mark Wright]

Damm your right, ill look at the calcs and post later.