Neutral earth links in feeder pillars

Are both the neutral and earth cables sized for full potential current? You maybe need just a few to be low in capacity to provide additional reasons for 'fixing' any problems. Just a thought.

Is this part of your on-site distribution, or consumption side ?

I.e are any current consuming loads connected to the TNS part between the links ? 
If not you might be marginally legally OK, even in the UK. Depending how the system was set up, a TNC distro to each building say, but in the building feeding TNS final circuits is permitted - for none consumer circuits only (DNOs do exactly that of course)
Legality aside, it needs some serious thought to get it right, and then does not work well if the earths of different areas are cross linked,  and what you have sounds like it may be if it is diverting current through the earthing and perhaps the building superstructure which is not a good sign.

Don't remove anything yet though - if there are high resistance links masked by the down stream ones it may make it vastly worse. If there are separate neutrals and CPCs in the feeds to the pillars/ cabinets or whatever, then the CPC and neutral continuity should both be verified, as should the presence of upstream link,  before removing the load end link and converting to TN-S, (or TN-s depending on your preference, I tend to follow a former forum contributor's convention of the small 's' load side, and the large one distribution side)

Don't move more than one thing at a time and don't have it energised if/when you do.

Also, related, where is the CPC earthed to terra-firma - where are whatever is acting as the system LV electrodes ?
Mike

Morning, sorry, just trying to decipher the first bit.

Each feeder pillar has TP+N plus CPC supply, so wouldn't be TN-C.  Not sure what is meant by on-site distribution or consumption side though, although it may be asking whether I believe that there is TN-C to each feeder pillar and then that's where the neutral earth link is, which I don't believe it is.

Would definitely be eyeballing and testing each and every part before taking anything out, when it's isolated of course.

With regards to the LV electrodes, we have a combined HV/LV earth electrode as we are not a hot site. 

The neutrals are bigger than the earths, not taking into account the armouring of the supply cable.  These systems have been working without issue for the last 17 years, but that doesn't mean that the earth should be carrying load current in its normal working life.

Well distribution cables only supply boxes doing the  ADS and perhaps metering, that in turn  supply final circuits that may have loads directly connected,. If it has a directly connected load at any point  its not "distro".

I think what you have could perhaps be described as TN-S-C-s, as NE is first both common then separated then connected then separated again, which is not really a recognized combination in the UK unless you are operating your own distribution network ;-)

I'm assuming you are in the UK. In much of Europe, or in AU/NZ this sort of thing would, or at least if done rightl could, be fine.

Mike.

mapj1 said:

I think what you have could perhaps be described as TN-S-C-s, as NE is first both common then separated then connected then separated again, which is not really a recognized combination in the UK unless you are operating your own distribution network ;-)

'reconnection' of N and PE is not permitted by BS 7671, and I don't think it has been at least as far back as prior to 15th Ed of the Wiring Regulations.

'reconnection' of N and PE is not permitted 

quite. But it is very common on the network operator side of the intake fuse, and in what we would now call the "building networks" of some older blocks of flats that used to be maintained by the area boards, neither of which fell/fall under the wiring regs / '7671.

Quite sensibly, since the rule changes that turned some unsuspecting freeholders of flats into unskilled owners and operators of building networks, the distro in new blocks of flats,  is now designed and tested  to '7671, and normally with SWA or trunked cables with no additional NE bonding. 
But, there is a lot of older stuff out there and we don't know who put this in, or when, or who they were working for and private networks can be a bit odd, to say the least.

Mike.

mapj1 said:

"building networks" of some older blocks of flats that used to be maintained by the area boards, neither of which fell/fall under the wiring regs / '7671

Although, BS 7671 is often specified for new/upgrade works on these systems ... BS 7671 does cover TN-C systems, so can be used, but TN-C-S-c-...-s isn't permitted.

TN-C-S-c-...-s

Ignoring any colour coding, I guess you could deem it to be simply TN-C-S with the PEN consisting of two (or more) conductors in parallel for part(s) of the route. As long as they're not designated as separate N and PE conductors, but rather as PENs, BS 7671 wouldn't object I think. (As Mike suggests, common practice in DNO land where old 4-core + armour cables are converted to PME with "N"-armour links scattered here there any everywhere).

   - Andy.

cannot do that, if there is a load part way along, between links, as that makes use of the '-s' nature of things.
Its all splitting hairs, it may or may not be illegal, its not a good approach technically, and as a country we don't like it.

M.

AJJewsbury said:

gnoring any colour coding, I guess you could deem it to be simply TN-C-S with the PEN consisting of two (or more) conductors in parallel for part(s) of the route.

Not really ... especially if exposed-conductive-parts, or extraneous-conductive-parts, connected to different bits of the TN-C-S-c-...-s, are simultaneously-accessible (in any combination). It's back effectively to 411.3.1.1 2nd para.

AJJewsbury said:

gnoring any colour coding, I guess you could deem it to be simply TN-C-S with the PEN consisting of two (or more) conductors in parallel for part(s) of the route.

Not really ... especially if exposed-conductive-parts, or extraneous-conductive-parts, connected to different bits of the TN-C-S-c-...-s, are simultaneously-accessible (in any combination). It's back effectively to 411.3.1.1 2nd para.