Single phase sub-main

What is 3a on page 509 then?

Yup - that's the SPD element connected N-PE - typically a spark gap/gas discharge tube type.

Also fault protection as defined in Chap 41 shall remain effective in the protected installation even in the event of SPD failure. 536.4.6.

534.4.6. - and the 2nd indent of that reg describes the permitted connection types for TT systems.  

Also table 534.5 applies. SPD only downstream of R.C.D.

Which is why you'd usually use the right-hand column instead - CT2. S-type RCDs are relatively expensive and contain electronics, so it seems daft to omit them from SPD protection by putting the SPD afterwards and so drawing the full spike current (possibly many kA) through the delicate RCD.

   - Andy.

Even gas discharge tubes as found in the Wylex neutral S.P.Ds?

According to the teardown Wylex use the melt-solder approach to indicate failure on those as well - and I can't see that operating unless the SPD element has gone short.

   - Andy.

no, the gas tubes fail open, and if grossly overloaded the seals fail, gas escapes and air gets in raising the breakdown voltage higher. Hence the use of a GDT for the NE connection.

gkenyon: 
 

Zoomup: 
Hello G A, We are using the 18th edition nowadays. Many times the armour is adequate to use as a C.P.C.

Use SWA as CPC, a guide to the acceptability of steel wire armour use as CPC. (gadsolutions.biz).

Z.

Just to note that the calculation method for loop impedance shown in Table 4 on this site, whilst still demonstrated in the latest (2018) IET Electrical Installation Design Guide, should really be replaced for cables with line conductor csa 16 sq mm and above (where reactance comes into play more) by the method shown in Annex NA of PD IEC/TR 50480:2011 - specifically sections NA 4.4 (armour alone used as cpc) and NA 4.5 (external cpc). 

For the case where the armour is used as cpc, the total loop impedance contribution of the SWA cable for a line to earth fault is given by:
 

R1 is the DC resistance milliohms per metre) of the line conductor, R2 is the DC resistance (milliohms per metre) of the SWA.
 

Don't forget the above is in milliohms per metre, so multiply by length, divide by 1000. 

This leads to a slightly higher overall DC resistance per metre but a slightly lower overall reactance per metre.

 

 

It looks like that involves arithmetic. I am not too good at sums so will rely on Table 54.7 for guidance. Armoured cable to the stable and garage will not cost too much.

Z.

Zoomup: 
It looks like that involves arithmetic. I am not too good at sums so will rely on Table 54.7 for guidance. Armoured cable to the stable and garage will not cost too much.

Z.

Absolutely … with multicore SWA it's usually the case that the armour is OK for cpc for smaller csa conductors and this only comes into play with larger CSA. (Of course, that doesn't take into account the requirements in PME installations to achieve the equivalent of main protective bonding where that applies with very small CSA cables.)

I just wanted to point out that some of the data on the site (e.g. R1+R2) and calculating length of run (for Zs) in TN systems, and also if you did want to do adiabatic for larger CSA cables, you should use an alternative method now.

no, the gas tubes fail open, and if grossly overloaded the seals fail, gas escapes and air gets in raising the breakdown voltage higher. Hence the use of a GDT for the NE connection.

I'd believe GDTs are unlikely to fail to short (certainly a lot less likely than the MOVs) - but if it's really impossible, why all the fuss banning them from L-PE connections on TT systems (CT2 etc)? (From what I can see the performance characteristics can be made pretty similar to MOV ones).

… and how would that out-of-service flagging mechanism work - if the GDT fails would there reliably be enough energy in the surge to melt that solder blob?

    - Andy.

this ought to be in the test your SPD thread but 

you can not  must not  put a GDT between L and N, as once it has struck it will not go out, rather it will sue the energy of the mains to transition from spark to glow discharge to full blown arc with impact ionisation and melting of the electrodes over a millisecond or so, and then fail spectacularly - unlike the MOV type which has a limiting action conducting at high voltage , and then as the volts fall away, the current drops back to 9/10 of very little, at least unless it was damaged. 

Its the same reason you do not short the ballast on a florry tube.

Mike

Chris Pearson: 
 

perspicacious: 
Will the stable occupiers be content if their installation is connected to PME?

The horses??? ?

Long “wheel” base.

Zoomup: 
 

 

 

Thanks. How will my suggested upfront S type 300mA R.C.D. affect the overall situation?

 

Z.

Where do you buy those?

Chint do a 3phase  one

Schneider do a single phase one

as do eaton

Very few places will have stock, some will probably deny they exist.

Mike.