ADS and RCDs on PV installations

[SolarCity]

Okay, I know this has been discussed extensively elsewhere but I think this deserves another thread.

Seeing as we have generally come to a consensus that RCDs intended to provide ADS may still be being supplied from the LOAD side by the inverter, perhaps as much as 5s after the RCD has operated, this seems to me to throw up an important problem.

For installations where RCDs are required, we therefore won't be able to provide ADS on a circuit without additional protection. It is my thinking that the only way we could provide such a thing would be by installing a seperate RCD at the inverter. Unless I am missing something?

 

[BruceB]

PV circuits are distribution circuits not final circuits, so do not fall under 411.3.2.2, but 411.3.2.3 and 411.3.2.4. In other words in a TN system have a disconnection time of 5s and in a TT system a time of 1s. Inverters shut off within 5s so ok for TN but that leaves TT to worry about.

So I think I see the problem you are talking about Biggs - a TT installation protected by an rcd where the inverter continues producing voltage for some seconds after the rcd has tripped on a low impedance fault to earth. I have not drawn out all the diagrams, but is that really a problem? The protection we are talking about is against thermal (and other) effects, but not against electric shock (from the note at the bottom of 411.3.2.2). As fault current levels from PV are inherently limited by design and as neither L or N of the PV circuit is by design connected to earth (so no earth loop circuit) I cannot see that you are going to get sufficient I squared t out of an inverter to cause any problems on the thermal front under these fault conditions.

I think this is going to be a bit difficult to discuss authoritively because we do not know exactly what was in the mind of the authors of 7671 as they drafted the various sections. But arguably they have thought through this scenario because section 712.411 of the regs is all about the modifications required to section 411 (which is ADS) for PV systems. It does not overtly raise any particular issue for TT.

The final point that occurs is 411.3.2.6 which says if you cannot meet the times (say the 1s time for TT) then supplementary bonding is the answer (and not difficult to achieve).

Regards
Bruce

 

[SolarCity]

Are you sure that a PV circuit is a distribution circuit and not a final circuit?

 

[BruceB]

Pretty sure, yes.

It has always been discussed in those terms on the IET site. Also if you look at the definition for final circuit, it talks of 'current using equipment'. A PV generator generates rather than uses current (ignoring some milliamps at night).

 

[edexlab]

I would think of it as a distribution circuit also, after reading Bruce's previous posts on this , but I wanted the opinion of the Niceic and they had a discussion and decided it would be a final circuit because in darkness it became a load rather than a generator
so I discussed the idea again with them and they agreed that it was really a Distribution circuit but I would have to record it as a Final circuit. so still none the wiser really but I've read in another guide some time ago, possibly the DTi guide that it should be classed as final

 

[SolarCity]

Interesting stuff. I have to say that I tend to see it as a final circuit but I suppose it depends on which way you see the circuit. It could be argued that it is in fact both.

 

[BruceB]

Even if one were to accept that the circuit became a final circuit at night, then there is surely no problem because without any generation taking place the rcd at the mains supply end will provide all the necessary protection?

Regards
Bruce

 

[SolarCity]

Good point. Distribution circuit it is then and RCD at inverter end not necessary.

 

[edexlab]

The only problem is when you have an inspection by a scheme provider is that they won't agree that it is a distribution circuit on the certificate would this lead to a non compliance being observed or can it be argued as a matter of opinion

 

[SolarCity]

It could easily be argued that it is both in my opinion so I can't see how it could be considered a NC.

 

[edexlab]

I agree but this is what the Niceic have told me ,even though their own engineers have a divided opinion .

 

[moggy1968]

it wouldn't be the first time NICEIC took their own (incorrect) interpretation of the regs