In Court

[jason121]

Which regulations would be referred to in a court of law for PV installations.

 

[telectrix]

BS 7671, for starters. also whatever BS covers pv. and building regs.

 

[Gavin A]

All relevant regulations and guidance could be referred to, and the company involved would have to have some very good justification to have not followed any of it.

Where 2 conflicting sets of guidance exist though, probably the most recent would take precedence unless you have good cause to follow the older version. eg MCS vs IET on array frame earthing, my take on it is that there's a major difference between over and under 50kWp systems as it's around that point that capacitance levels can start getting up into the sorts of mA levels that could potentially be dangerous by themselves, so I intend to ignore IET and use my own judgement up to that sort of level as I've seen zero evidence to justify their decision.

 

[BruceB]

Gavin,

Can you direct me to the IET decision you are referring to. I might have missed it. I did not realise there was any particular conflict at the moment.

On second thoughts do you mean the new IET cop? Is it going to say 'earth everything' without a regulation to back it up?

You have made me look at my new digital yellow book. I see type B rcds now required for additional protection as well ADS, unless standard exemption applies. Andy (Sibert) will be pleased.

 

[SibertSolar]

They're coming from the angle that the inverter's RCMUs need to have the frame earthed (with TL inverter's) Bruce......the fallout will be "interesting"

 

[Gavin A]

They're coming from the angle that the inverter's RCMUs need to have the frame earthed (with TL inverter's) Bruce......the fallout will be "interesting"

oddly neither SMA nor ABB are aware of this issue, it's not something that's come from the inverter manufacturers, and if it actually is an issue then presumably it must mean that 500,000 or so dangerous solar PV installs have gone in since MCS changed the rules to remove this requirement, on the recommendation of the same lead author as is now saying the complete opposite.

Also they're attempting to make out this should also apply to galvanically isolated inverters, which is complete nonsense.

 

[SibertSolar]

I know, odd isn't it....maybe there are other reasons but the couple of conversations I've had, with the likes of BRE etc, have tended to refer to RCMUs and DC ground faults etc. I think Martin Cotterell's blog may have touched on this recently too

 

[Gavin A]

I know, odd isn't it....maybe there are other reasons but the couple of conversations I've had, with the likes of BRE etc, have tended to refer to RCMUs and DC ground faults etc. I think Martin Cotterell's blog may have touched on this recently too

yes, however when first discussed the primary reason given was to align the UK with other countries regulations.

I suspect that's the main reason, and Martin and associates simply got brow beaten into submission on it at the international level.

My critique remains though that either there are half a million dangerous installations out there that followed the MCS guidance on frame bonding, in which case surely something would have to be done to rectify that situation.... or it's not dangerous in which case why are we being told that we now need to bond all frames?

If it turned out to be the former, then I expect MCS would be sued by the industry to recover the costs involved in rectifying that situation, if the latter, then I for one intend to ignore the IET guidance on that issue.

Either way I want to see the evidence used by IET to reach their judgement, and won't comply with this rule until they release that evidence.

 

[BruceB]

Following some links I see I looked at the draft cop discussion last year, but had thought no more of it since. Earthing and bonding are never black and white and by protecting against one failure mechanism you can easily introduce vulnerability to another.

I suspect Gavin is right and that commonality is driving this.

As I understand it, an RCMU does not need earthed array frames to work. But if an array frame is earthed then it is more liklely that a leakage to earth will trigger a fault. That at face value is a good thing. But it does introduce a vulnerability that there is now a good earth route so if a body gets between an earth leakage and the array frame then the resulting current is more likely to be fatal than if say standing on dry ground.

It is a bit like insisting the outer metal case of a double insulated electric kettle is earthed. Not required of course.

There was a story, real not apochrophal I believe, of a man doing a job under a bath in his house using a faulty table lamp for illumination. Some exposed metal parts of the lamp were live, but had not electrocted the man because the flooring was sufficiently insulating. However he then touched the metal bath which was supplementary bonded as was required ..... and curtains.

Admittedly the wide use of rcds would probably prevent that today.