Terminating SWA at remote metal enclosure with TT earthing

[nothanks]

I have a sub-main to a remote socket with a TT earth and, currently, the SWA braid is terminated on a plastic enclosure. I need to change the enclosure to metal so have to find a way to terminate the SWA braid without connecting it to the metal enclosure. This must be a standard problem with a standard solution, but it's not one I've had to deal with before so I don't know the "standard solution" :-( Answers I can think of: 1 terminate on a plastic box attached to the metal enclosure (not ideal in this application), 2 cut-off and insulate the braid and use a standard nylon gland (loss of some mechanical protection, but I could clip the cable to the metal enclosure), 3 isolate the head end of the SWA and connect the remote end to the TT earth (easier to do because it comes through a plastic enclosure, but it doesn't "feel" like the right approach). I'm favouring #2 at the moment but would welcome input from those who have solved this problem before.

[spuorgelgoog]

On Tuesday, 30 June 2020 22:27:52 UTC+1, [e-mail] wrote:

This is the technique illustrated here

Note that the comments include site applied insulation (heatshrink) so the armouring cannot be touched inside the Class 2 enclosure.

The text in grey reads:

This article addresses the questions which typically arise where such a supply is provided by an SWA cable buried in the ground, in particular: ? Can the armouring be used as a circuit protective conductor? ? What are the equipotential bonding requirements? ? What are the particular requirements where PME conditions apply? Armouring for use as a circuit protective conductor Unless provided with adequate mechanical protection, cables buried directly in the ground should incorporate an earthed armour or metal sheath or both which is suitable for use as a protective conductor (Regulation 522.8.10). In view of this, a SWA cable is often specified for the supply to an outbuilding with the armouring serving as the cpc. However, where this is the case, the relevant disconnection times of Chapter 41 must be met. For a distribution circuit connected to a TN system a maximum disconnection time of 5 s is permitted (Regulation 411.3.2.3) whilst for a TT system the maximum disconnection time is 1 s (Regulation 411.3.2.4). To confirm the adequacy of the armouring, the size of the cpc may be determined by either of the following methods: ? selection in accordance with Regulation 543.1.4 (Table 54.7 of BS 7671),or ? calculation in accordance with Regulation 543.1.3 (adiabatic equation). Where the csa of the protective conductor is determined using Table 54.7, reference should be made to the column highlighted in Table 1. For most cable sizes the armouring will generally be suitable for use as a cpc nonetheless the designer should ensure compliance with BS 7671. Using the armouring as a protective equipotential bonding conductor (Where PME conditions do not apply) Where the outbuilding contains extraneous-conductive-parts, such as metallic water, gas or oil pipework, the designer may decide to use the armouring of the SWA cable to also perform the function of an equipotential bonding conductor. In such circumstances, the armouring should afford the ?equivalent conductance? to that of copper (Regulation 544.1.1). It should be noted that the resistivity of steel is approximately 8 times higher than that of copper (8:1 ratio), so in order to achieve the ?equivalent conductance?, the size of the steel armouring will need to be 8 times larger than that required for copper. For a TN-S or TT earthing arrangement, the csa of a copper protective bonding conductor should be not less than half the size of the protective earthing conductor for the installation and not less than 6 mm². This would require the armouring of the cable to have a minimum csa of 48 mm2, which should be confirmed by reference to the particular cable manufacturer?s data.

Using the armouring as a protective equipotential bonding conductor (Where PME conditions apply) Where the outbuilding is to be connected to a TN-C-S (PME) earthing system the csa of the protective bonding conductor(s) should satisfy the requirements of Table 54.8. As a minimum, this requires the armouring to have the equivalent conductance of a 10 mm² copper conductor and sized in accordance with the largest PEN conductor of the supply. As previously stated, this should be confirmed by referencing the particular cable manufacturer?s data. However, where the installation is connected to a PME terminal, ELECSA recommends that the outbuilding is provided with a separate means of protective bonding. TT earthing arrangement Where the PME terminal is extended to an outbuilding, there is an increased risk that in the event of loss of the supply neutral, say as a result of an open circuit fault on the distributor?s PEN conductor, the cable armouring and all conductive parts connected to the MET are likely to be raised to the supply potential (230 V). In order to minimise this risk, it is recommended that the armouring is connected to the MET at the origin but the outbuilding is divorced (separated) from the PME terminal. Normally, this is achieved by terminating the protective conductor (armouring) into a fully insulated (Class II) enclosure Fig 1, and connecting the outbuilding to an earth electrode to form a TT earthing arrangement. The earth electrode should be a type listed in Regulation 542.2.2 and where practicable, it should be located close to the outbuilding. it is not permitted to use a metallic utility water supply pipe or a pipe containing gas or flammable liquid as an earth electrode (Regulation 542.2.6). Other metallic water supply pipes may be used, but suitable precautions must be taken to prevent their removal, which in practice is difficult to implement and maintain. Cable installation Cables should be sufficiently buried to avoid being damaged by any reasonably foreseeable disturbance of the ground (Regulation 522.8.10). Although BS 7671 doesn?t specify a particular depth for buried cables, ELECSA recommends a minimum depth of 600 mm, but where there is a possibility of the ground being disturbed such as by planting for example, the depth should be increased appropriately or additional mechanical protection should be provided for the cable. Conclusion While it is permitted for the metallic sheath or armouring of a SWA cable to serve as both a circuit protective conductor and a protective equipotential bonding conductor, as outlined in this article, the relevant requirements of BS 7671 must be satisfied.

[spuorgelgoog]

Lateral thinking is use a

to join two lengths of SWA, but *not* use the armour continuity straps. You therefore earth each section of SWA separately, from the appropriate end.

There remains a risk of the separate earth potentials being exposed during maintenance, so it should be appropriately labelled.

Owain

[nothanks]

That's an interesting idea (I love lateral thinking) but I may have come-up with a simpler one: find a regular nylon gland that will clamp on the SWA outer (which may not be simple, I haven't looked yet) and heatshrink the end of the SWA inside the enclosure to insulate the braid. Perhaps adding a clamp on the outer, inside the enclosure, as belt'n'braces.

[nothanks]

Well, that was much easier than I thought. I had both a suitable nylon gland and the heatshrink in "stock" - job now done but there's a supplementary question about MCB ratings. The 2.5mm2 buried SWA sub-main is currently protected by a B16 MCB at the head end but I'd like to increase it to a B25 because I have a couple of B16 RCBOs at the remote end. 25A is inside the rating of the cable so I don't see a problem. Is there any reason not to do this?

BTW, everyone says put in a larger cable than you need when burying cables - I wish I'd done that!

[John Rumm]

The standard in the past would have been what you have - an insulated CU at the TT end. With a move to "non combustible" (e.g. metal) CUs for all things that is more difficult.

Yup that is commonly done - an adjacent plastic box etc

e.g.

It would work.

There are actually "insulated" SWA glands available that maintain the IP and mechanical protection:

(no idea on price or a supplier though)

That is fine if you have RCD protection at the origin of the submain - then a fault to the TT earth, would still disconnect the power. It would not be practical if the submain is fed only from a Fuse/MCB, since you are unlikely to have a low enough earth loop impedance to reliably trip the circuit protection under fault conditions.

[John Rumm]

Chances are it will be fine, although it depends on the length of the submain.

Factors that might come into play are maximum circuit impedance, and possibly voltage drop (if you assume a higher design current)

If you add up the total round trip resistance presented at the end of the submain in the event of a phase to neutral short, it needs to be low enough to achieve a prospective fault current equal or higher than the

160A required to trip the B32 "instantly" on the magnetic part of its response.

You can use the tabulated cMin value of Zs of 1.37 Ohms (down from the

2.73 Ohms for the B16 MCB):

One probably ought to allow for the submain operating at capacity (i.e. conductor temp of 70 deg C) at the time of the fault, which will push the resistance up to 17.78 mOhms/m (L + N). So your length limit is

1.37/ 17.78 x10^-3 = c. 77m

Voltage drop you probably care less about, but if running lighting circuits from the submain, then you are limited to 3% (6.9V) total drop. If you took the design current as 25A, then then your drop would be 18 mV/A/m, so 6.9 / 25 / 0.018 = c. 15m (Vs c.22m for the current setup)

(voltage drop allowed for socket circuits is 5% or 11.5v)

You are not alone!

[nothanks]

It's going to be a B25 (not B32) at the head end so that gives me a max length of around 98m - the actual length is 75-85m, split as 70m'ish of buried SWA and 15m'ish of T&E running through the house from where the SWA terminates to the CU.

Voltage drop isn't an issue. This supply runs a Well pump (for irrigation) and now also supplies a couple of IP66 sockets for general use at that end of the garden. In the future I may need to provide a further extension from this sub-main to provide electrics at a gate, but that's a long way down the job list!

Thanks for the prompts and sanity check.

I had a momentary brain-off wobble when a socket tester objected to the ELI at the new sockets, but then remembered that the N-E link was a long way away. Measuring from the TT earth to the TN-C-S earth gives about

120 ohms but I'll double-check the actual quality of the TT earth later with an intermediate probe.

[nothanks]

I think the problem with those is that it would still be possible to touch both earths, although they would solve the isolation problem. By insulating the braid and using a regular IP66 nylon gland there is no possibility of touching both earths. The cable isn't restrained as well as when the braid is clamped but in my case that isn't an issue.

Yes, but that would result in an RCD at each end of the sub-main. Much simpler (on reflection) just to isolate the braid at the remote end.

[John Rumm]

That question arises from time to time with SWA and exported TN-C-S earthing, since even with the insulating boot fitted to normal SWA glands, one can in theory still touch the earth without the need of a tool since the boot can be moved by hand.

Indeed, and plenty adequate in the circumstances by the sounds of it.

Yup, there are practical advantages to not having a RCD at the head end. The 18th edition requirements to RCD protect all "unprotected" cables though can make it harder to implement in a compliant way when there is a run of T&E feeding the start of the SWA.

[No Name]

That is the situation I currently find myself in and need to work out the best way forward:

When the house CU was changed, a new 6mm2 T&E was run from the new consumer unit to the garage between a floor and ceiling void. This is fed by a 32A Type B RCBO.

The original intention was for a car charging point.

20 metres of 10mm2 SWA has now been laid from the garage to the outbuilding & greenhouse. A TT rod has been installed and this has 10mm2 protective conductor single core wire attached to it.

A 45A cooker switch has been used to connect the end of the 6mm2 T&E to the start of the 10mm" SWA and this is in the integral garage.

A plastic flame proof IP67 sub-consumer unit will be fitted in the outbuilding to the other end of the SWA. THis will use all RCBOs like the main house consumer unit comprising of 2 off 6A RCBOs and 2 off 20A RCBOs

Now I need to change the 32A RCBO in the main house CU to avoid double tripping of two sets fo RCBO's in teh event of a fault at the outbuildings

So what to change it to:

32A Type C RCBO

Time delayed 32A RCBO

100mA 32A RCBO?

Or to a simple 32A MCB as the 6mm2 run of T&E is only about 7m long and is either in a ceiling/floor void or surface clipped.

S.-

[nothanks]

Is there anything that prevents you putting the surface-clipped section into earthed conduit and using a B32 MCB?

[nothanks]

In my case the T&E runs in a floor void (and terminates to SWA under the floor, accessible by lifting a labelled board) so, I believe, does not need RCD protection.

[No Name]

the route is as follows:

CU unit on hall wall

through double skin wall

travels for 2.5m between ground floor ceiling and first floor floorboards which is then covered by laminate

then through a 2nd double skin wall

then surface clipped along a 2.5m sistered joist

then down the wall to the aforesaid cooker switch on garage wall

then to SWA which then travels on through the concrete floor en ruote underground to the outbuilding

If I was going to use and earth said conduit, I would have to earth it from the cooker switch end otherwise I would have to cut into the 6mm2 T&E cable when it transits from a void to a surface clipping.

(it would not be a trivial job to run a new earth bonding wire from teh CU to the point where cable becomes surface clipped.

S.

[John Rumm]

Yup that would count as more than 50mm from the surface, and so not need additional protection.

[John Rumm]

That gives no discrimination.

That would be ok.

Again, little or no discrimination.

Oddly, the surface clipped section being "visible" does not need additional protection. So as long as the section in the ceiling void is safe from nail penetration to a depth of 50mm or greater, then the plain MCB is fine.

(ceiling void normally treated as reference method B, giving a nominal

38A capacity for 6mm^2 T&E)

[ARW]

A MCB sounds good to me.

[ARW]

That run would suggest that a MCB instead of a RCBO at the head end would be a better option. It might not even be the correct type of RCBO for a car charging point.

[nothanks]

One of John's other replies has just reminded me that I need to de-rate my sub-main because part of the route is under a floor in the house - I need a B20 MCB rather then a B25.