Extension of Ring Main & Readings

[loopy66]

I need to extend the ring main upstairs in our house, in two rooms there is only one single socket. I have taken up the floor boards, run 2.5 twin/earth around under the floor, chipped away the brick, installed the backboxes and sockets and then left the new cable waiting to be cut in to the ring main. The amount of cable I have run is about 20 meters.

Our ring main covers the whole house, a usual type 1930s 3 bed terrace and we are still on the older type fuse box consumer unit.

I have taken a reading with a multimeter on the existing cable and get 0.5 ohms on the live and 0.5 ohms on the neutral. Sometimes it jumps to 0.6, ie, flips between 0.5 and 0.6. I also get 6 ohms between the live and neutral. I have done these readings from a socket in honesty and not the consumer unit, which I have read in places is actually ok if absolutely neccessary.

My questions are the following before I actually cut the new extension in...

• If I use what I read as 0.5 ohms / 0.00741 = 67.48 meters. Is this a good estimate of the existing ring main cable length and therefore adding 20 meters is ok.
• Am I getting 6 ohms across the live and neutral because there is a fault somewhere on the ring or because I still have devices plugged in around the ring?

 

[EFLImpudence]

f I use what I read as 0.5 ohms / 0.00741 = 67.48 meters. Is this a good estimate of the existing ring main cable length and therefore adding 20 meters is ok.

Yes.

You should measure the CPC(earth wire) as well.
It should be 1.67(2.5/1.5) x 0.5 = 0.835 if modern cable (single core), or
2.5(2.5/1) x 0.5 = 1.25 if old imperial cable (stranded cores).

• Am I getting 6 ohms across the live and neutral because there is a fault somewhere on the ring or because I still have devices plugged in around the ring?

• Probably. You should measure between L & E and N & E (Main switch OFF) and get an open circuit.

So, if your lengthened circuit is 90 metres, then, if all modern cable -
Line(Live) will be - 90 x 0.00741 = 0.67 (r1) and
CPC - 90 x 0.00741 x 1.67 = 1.11 (r2)

Fitting new sockets and concealed cables does require that there should be an RCD on the circuit.
Do you have one? If not you might require an electrician.
RCDs cannot be fitted in Fuse Boxes so you could have a stand-alone one.

If you are interested -

Therefore the Line and CPC (r1 + r2) = 1.78Ω
Because it is a ring, which is like two cables in parallel, the 1.78 is divided by 4 = 0.445Ω (this is called R1 + R2).

This figure, 0.445 should be multiplied by 1.2 to allow for temperature rise to the cable maximum, and
divided by 0.95 (95% - should actually be 94%) to allow for the supply voltage dropping to the minimum allowed (230 - 6%)

So, 0.445 x 1.2 ÷ 0.95 = 0.56Ω

This figure should be added to the external Line and Earth impedance (Earth Fault Loop Impedance) (Ze) to see if it is within the limit which will blow your 30A fuse instantly in the event of a fault.
That figure for the fuse is 1.04Ω, so you whould normally be alright but obviously I can't tell from here.


Edit - The 1.04 figure already accounts for the 95% so disregard that part.

 

[securespark]

It could be 2.5/1.0.

 

[JohnW2]

It could be 2.5/1.0.

EFLI appears to have at least partially addressed that ...

You should measure the CPC(earth wire) as well.
It should be 1.67(2.5/1.5) x 0.5 = 0.835 if modern cable (single core), or
2.5(2.5/1) x 0.5 = 1.25 if old imperial cable (stranded cores).

However, I have a feeling that early metric (solid core) may also have been 2.5/1.0 - was that the case?

Kind Regards, John

 

[securespark]

Imperial was imperial.
Metric was indeed 2.5/1.0 and later 2.5/1.5.
The 2.5/1.0 was what I was referring to.

 

[ericmark]

Theory above is correct, however practice one has to ask does it matter? We are looking at a number of safety and other items.
1) With a line - neutral fault will the magnetic part of a MCB trip within set time?
2) With a line - earth fault will the magnetic part of MCB or the RCD trip within set time?
3) Will the volt drop exceed the permitted values?
For 1 and 2 with a type B MCB we are looking a 5 times rated valve plus 5% for safety, so loop impedance 1.37Ω or prospective short circuit current 168 amp or with 2 RCD protection, and this is including the supply limitations, so we with a TN-C-S can take 0.35Ω as incoming, but clearly meter needs to read to two decimal points and better if a loop impedance meter is used.

With 3 the calculation assumes 20 amp centre and 12 amp even distributed, so we use 26 amp to calculate and that equates to 106 meters of cable max in a ring final, the loop impedance works out lower than the trip values above, but big question is the volt drop that important?

Fridge/freezer if not inverter type may have a problem with excessive volt drop, and old radios may have mains hum, but most items today use switch mode power supplies and it is unlikely it would cause a problem.

 

[JohnW2]

The 2.5/1.0 was what I was referring to.

I realise that. My point was that EFLI said that imperial was (I presume he meant 'the equivalent of') 2.5/1.0 but implied that all metric was 2.5/1.5 - although, as I suggested and you have confirmed, early metric was also 2.5/1.0.

Kind Regards, John

 

[EFLImpudence]

Sorry I didn't include every possible option.

 

[securespark]

Yeah, it's all your fault!

 

[JohnW2]

... with a type B MCB we are looking a 5 times rated valve plus 5% for safety ...

Where did that come from? - just 'sounds like a good idea', perhaps?

Kind Regards, John

 

[loopy66]

Thanks for the information, some great stuff. I will measure the earth wire and kicking myself why I didn't do that already!

The earth wire in the existing ring cable from a memory is thinner than the wire in my new cable so suspect it is 1mm whereas my new cable is 1.5mm. I will check it with a micrometer just to be super accurate but believe it will be 1mm, the original ring cable is the older red/black colouring but it is all solid core.

I didn't do all this today as I went to take readings from the new ring extension and it was giving me 5ohms on the live and neutral so I guessed that I connected one of the sockets a bit funky so I have taken them all back and reconnected. I will do the readings in full as advised above in the week.

Regards the requirement for an RCD though, that's a good point and I am having a smart meter installed soon (hopefully soon), what would you say the chances are of me giving the guy from the electric company a few quid cash to install me a decent modern consumer unit with RCD protection whilst he is at it, can this kind of job happen these days or has everyone gone super by the book?

 

[flameport]

These are the imperial cable sizes for flat cables with the metric conversion added in red.

 

[EFLImpudence]

Nobody noticed my Zs value of 1.04 (30A BS3036) already accounts for the 95% Cmin.


Is that th 5% Eric is thinking of?

 

[EFLImpudence]

Regards the requirement for an RCD though, that's a good point and I am having a smart meter installed soon (hopefully soon), what would you say the chances are of me giving the guy from the electric company a few quid cash to install me a decent modern consumer unit with RCD protection whilst he is at it, can this kind of job happen these days or has everyone gone super by the book?

NIL. I don't think they are electricians.

Do you really want a smart meter? You don't have to.

 

[loopy66]

NIL. I don't think they are electricians.

Do you really want a smart meter? You don't have to.

Gutted, was hoping for a good deal.

Regards having a smart meter, I asked the company to install one, I liked the idea of being able to physically see the electricity being used and the cost and also, not having to keep doing readings myself as it seems quite regular they slip me onto an estimated bill and the price goes up until I send one in.