Solar Array - Grounding

Your mileage may vary  on a transformer-less inverter the actual guts of the solar panel take it in turns on alternate half cycles to be connected to the mains one way round or the other. I cannot find it now but a few years ago I posted a query where a colleague was looking at a solar installation where folk working on the roof were complaining of tingles that turned out to be a sort of half mains ripple on the metalwork, Earthing it, rather than letting it float  made the problem go away. I can see a problem with TNCs, but on a roof installation perhaps only folk with really long arms and legs will touch that and the true earth, so then being part of the indoors earth zone is probably OK, At least then if one panel was damaged and the internals faulted to the frame something would then trip.

On the ground, I am less sure - perhaps to earth to an electrode....

Mike.

What would you ground them to? I was looking at a set up at Elex and it would seem that brackets are squeezed between the rows of tiles and screwed to the rafters. In other words, the arrays are electrically isolated from everything else.

Quite what you would do with my garage, which has a corrugated steel roof, is another matter entirely.

Do solar arrays (the frames) need grounding? The inverters in most cases are DC (and isolated from mains) and indeed micro-inverters are class 2 with isolated DC inputs from the array. 

Being d.c. alone doesn't solve the shock risk issue - as even modest arrays can run a several hundred volts - which demands proper protection from shock (as per chapter 41).  A lot of modern inverters (i.e. transformerless designs) don't necessarily reliably isolate the d.c. side from the mains/earth either. Some small systems might run as SELV/PELV on the d.c. side.

The usual solution (as recommended by the updated section 712) is to use double/reinforced insulation as the method of protection against electric shock on the d.c. side (both for wiring systems and other equipment) - with the inverter away from the panels (or class II micro inverters and double insulated a.c. wiring system in the vicinity of the panels) that then usually removes any need to Earth any part of the panels for normal shock protection reasons.

On very large systems you might well have problems of static (or capacitive coupling with a.c. parts of the system) building up a charge on the metalwork so would want to earth them. But then it's just a matter of picking a suitable earthing system as you would for anything else in a similar situation (e.g. probably not PME if it's outdoors and/or easily touched by those in good contact with true Earth).

   - Andy.

What about lightning protection? Does having various conducting pieces around increase the risk of strikes?

Roger B said:

What about lightning protection? Does having various conducting pieces around increase the risk of strikes?

On buildings with LPS, BS EN 62305 series applies, as required by BS 7671.

If the building is tall enough, or close to a building with LPS, then changing it by adding PV array on the roof may well alter the requirements for LPS, and/or need for bonding the frames to LPS..

vantech said:

Do solar arrays (the frames) need grounding?

The answer is "it depends".

Guidance on this is changing in the 2nd Edition of the IET Code of Practice for Grid-Connected Solar PV Systems, which is due to publish on 29 November 2022. The former 1st Edition (2015) said to earth the frames in most cases, but as others have pointed out, there are pro's and con's:

• in some cases, you earth the frames, to create a touch-voltage say to the satellite disk or terrestrial TV antenna (which can also give a mains-type "ripple" as Mike describes, or if installed properly to the relevant standards, earthed in its own right).
• In other cases, if you don't earth the frames, there is a definite shock risk.

Highlights the fact that protective conductors are not always "safe to touch" (the larger the installation, and the further a cable runs through that installation without further bonding of the protective conductor, or armour or screen, the risk could well increase).

Working practices for roof-mounted systems should perhaps assume that PV array frames are not always "safe to touch", especially from a metallic ladder in contact with Earth. Simply because, even small currents (<1 mA, which could occur through "leakage" even in a double or reinforced insulation DC side), whilst not necessarily highly dangerous in themselves from a shock risk perspective, can cause someone to be startled and fall.