SWA on DC side of PV system

I have been out and about doing my 2346 experienced worker assessments and touched base with some relatively large PV roof installations. Peaked my interest in these systems greatly! I was about to ask about the use of SWA multicore on the DC side when I was directed to GK’s excellent video presentation on the introduction of the current COP which looked like it was given at an IET presentation at some electrical event. The caution about using SWA cables seemed to perplex a couple of guys at the front. Perhaps they were diligent installers who had noted the acknowledgment about SWA being a suitable wiring system in the previous COP.

My take from what GK said and what is says in 7671 is that SWA is not a suitable wiring system for the DC side of PV systems. I have noted that in the several systems I have visited, all of which are no more than a few months old, almost exclusive use of SWA on the roof.

One installation underway at the moment has almost 1Mwp on the roof of a large warehouse. Multicore swa runs from the combiner boxes on the roof to the optimisers are up to 200m in places. Everything very neatly installed. Do you think this is just a technical breach of 7671 or is there a clear worry about insulation breakdown, perhaps as a result of water ingress into the jacket? After all, SWA is often used in the most arduous conditions outdoors.

  • Are there labels afixxed on the outside of the cable every meter or so stating DC like you would have for Solar PV DC cables in a house loft?

  • FYI, we had a related discussion on this here

    IET forum: PV submain cables: 712.521.101

    The short version is that, for PV purposes, SWA is not considered sufficient for use within the protection method of Double or Reinforced Insulation, because the electrical properties of the bedding is not controlled within the standard and therefore is not guaranteed insulation. Note that for PV in particular the regs would suggest we (now) need double insulation to earth as well as between poles (i.e. four faults to danger assuming a truly floating system, notwithstanding leakage capacitance).

    However, I am entirely unsurpised at your finding in practice... Let alone installers even many cable manufacturers aren't even aware of the issue, and I have had  technical departments from reputable vendors tell me at length that SWA is fine.

  • It probably would be fine, if the makers of SWA subjected their bedding and over sheathing materials to an insulation test - which  it would almost certainly pass.

    Sadly rather like the missing braid resistance on SY, without some guarantee, it requires engineering judgement on a cable by cable basis.


  • I won't make a statement regarding suitability of any particular cable in this thread, just point out that:

    • Reference: Reg 712.410.102. Double or reinforced insulation is required on the DC side where the inverter has no protective separation between AC and DC sides (i.e. SELV or PELV can't be met), because, if the inverter has no separation between AC and DC sides, effectively you're "shaking hands" with the grid under some conditions as the DC is switched to the AC (quite literally) by the electronics in the inverter (perhaps though inductances etc.). Hence, it could not be
    • Reference: Reg 712.521.101. Double or reinforced insulation is required on the DC side to help separate L+ and L- conductors to minimize risk of short-circuit, in parts of the DC side without overcurrent protection (i.e. because there are no fuses at the array). Required to protect against fire, and also possibly in line with 712.410.102 electric shock. The methods exemplified in the standard demonstrate how conductors need to be arranged (each insulation, and then each enclosed either in a sheath or another run of non-conductive containment.
    • Regulation 412.2.4.1 is the requirements for wiring systems used where double or reinforced insulation are employed.
    • In an SWA cable, there is no insulation or sheath between the armour and the insulated conductors. If it can [ever] be touched it's therefore an extraneous-conductive-part. In an SWA cable.

    It's necessary to consider all of the above points when looking for a suitable wiring system for the DC side.

  • Doncaster Cables now produce DC rated SWA cable.  Interested to know your view on this new cable?  https://www.doncastercables.com/uploads/PV-Ultra_Datasheet_.pdf  

  • It seems it double insulation (or insulated + sheath if you prefer) around each core - so would seem to meet requirements.

    But it presumably there's no BS (or equivalent) for that cable type - so it's use would be on the head of the designer and would have to be recorded as a departure on the cert. (133.1.3).

      - Andy.

  • Had a few discussions about it already.  Likely to be used where damage to cable may occur, what happens if the live conductors faulted to the armour?  And what happens if the armoured is earthed through to the AC side and you end up connecting the DC side to the AC side.  Lots to consider here.