RCD's with Photovoltaic's?

[supasparxs]

Hi guys
I just want to pick your brains about the basic schematic for a Domestic Photovoltaic system.
If the High Integrity CU has not got a spare way on the RCD side of the board can I bring the circuit out of a Non-RCD controlled circuit?

The circuit would be as follows

I have singles going from an MCB to rotary isolator to a generation meter.
Then a SWA (6mm core) 40 meters to another rotary isolator, to an inverter to a DC isolator and finally to the panels.

The SWA will be cleated up the outside of the premises and then dressed across the loft. The only time the cable will be burried in a wall is when the cable passes horizontally through the wall.

basically does this need an RCBO or would a MCB be ok?

I think an MCB would be sufficient however I have been told that an RCBO is required

Richard

 

[dougalthedog]

Richard,

I know this isn't what you're asking but you should specifically not connect an inverter to an RCD protected CU or section thereof - there's a theoretical possibility that the RCD trips and the inverter continues supplying power for a short time longer which could be enough to injure or kill.

If you are using SWA, I would think there's no reason to RCD protect the cable itself.

The only remaining issue would be dependent on which inverter you are using you may or may not need a DC sensitive RCD. In this case a standard RCBO would not suffice - there are a number of other threads on this board discussing these issues.

I'm sure someone will correct me if any of this is inaccurate!

 

[vernam616]

Completely inaccurate ^^^^^^^

If it was my install id cover my arse with a RCBO, lets face it cant do anymore harm with "Additional Protection"

 

[Markc]

Completely inaccurate ^^^^^^^

If it was my install id cover my arse with a RCBO, lets face it cant do anymore harm with "Additional Protection"

NOT completely inaccurate but completely correct.

The RCBO you suggest is belt and braces in my opinion. Do me a favor? While your inverter is in generation mode time test your RCBO or RCD. I bet it will not trip out in less than 40mS. If you are connecting via a communal RCD device then the other outgoing circuits such as shower, socket outlets will also not be protected for supplementary protection.

The SWA route is a good option as even a RCBO offering protection for a buried cable for example will not trip in under 40mS.

Obviously we are talking RCD's rated at 30mA, cables buried in walls at less than 50mm etc. It pays to check this with all inverters. I know several that we have checked and none lost terminal voltage in <40mS.

In generation mode then the 30mA RCD, RCBO will offer no "Additional Protection"

 

[vernam616]

NOT completely inaccurate but completely correct.

The RCBO you suggest is belt and braces in my opinion. Do me a favor? While your inverter is in generation mode time test your RCBO or RCD. I bet it will not trip out in less than 40mS. If you are connecting via a communal RCD device then the other outgoing circuits such as shower, socket outlets will also not be protected for supplementary protection.

The SWA route is a good option as even a RCBO offering protection for a buried cable for example will not trip in under 40mS.

Obviously we are talking RCD's rated at 30mA, cables buried in walls at less than 50mm etc. It pays to check this with all inverters. I know several that we have checked and none lost terminal voltage in <40mS.

In generation mode then the 30mA RCD, RCBO will offer no "Additional Protection"

Oh well my tried and tested way must be wrong then, just been fitting rcbos to pv systems on a big project, all tested within normal expected perameters

 

[JulianC]

Oh well my tried and tested way must be wrong then, just been fitting rcbos to pv systems on a big project, all tested within normal expected perameters

Test conditions and operating conditions are not the same - as stated above the inverter can continue to generate for up to 5 seconds after the mains is lost. Your RCBO may well physically trip within 40mS but the inverter side will continue to be live.

 

[powervan]

relax lads ur all now unemployed anyhow!

 

[Markc]

Oh well my tried and tested way must be wrong then, just been fitting rcbos to pv systems on a big project, all tested within normal expected perameters

But I bet you did not allow the inverter to go back into generation mode before continuing on the the remaining side of the sin wave for 1x and for each side of the sin wave for 5x tests. If the inverter isn't in generation mode there is no terminal voltage so yes <40mS is achievable, however this is not accurate operating conditions. As a 30mA RCD is for supplementary protection the required <40mS has to be measured while the inverter is generating.

So don't shot the messenger. Your tried and tested way isn't correct and nor should it be posted that it is. Check the G83 & G59 touch voltages degradation times for your inverter's. None have to be <40mS and will be 5s maximum.

 

[moggy1968]

yep, your installations don't comply I'm afraid. lots of previous threads on here about this and also some very good technical advice on siberts online pages.
You are also going against the manufacturers recommendations in many cases as they will often specify a 100ma or even 300ma RCD.not to say we haven't done it, when we have used a garage board for the PV taken on it's own tails from a henley block so it doesn't interfere with the household electrics. In this case the RCD is purely there as a DP switch as we haven't been able to get a non RCD board.
In these cases we have advised the customer to monitor the system and advise us of nuisance tripping so we can obtain and fit a 100ma type B RCD or non RCD DP switch. to date only one has caused nuisance tripping so it's for you to decide how much of an issue this is, but theoretically it is not following the manufacturers instructions.
fitting PV onto an RCD used by other household circuits is a big no no and definately a breach of 1771 as you are making the installation less safe for the reasons entirely accurately given above!
while an RCD may give additional earth fault protection it may also fail to operate at all and will have a longer trip time which places it outside the limits of 1771. most inverters have additioanl earth fault protection built in I beleieve

 

[Markc]

Inverter's on a communal RCD protecting other circuits will also earn you a major NC with your certification body and Mr NICEIC (or similar) will have a fit.

 

[Markc]

relax lads ur all now unemployed anyhow!

Unemployed? Nar, will be removing PV's from communal RCDs by the looks of things for months to come!

 

[vernam616]

Will have to dig a little deeper on this and see what the score is

Just finished installing PV on a housing estate, all pv circuits were put onto the existing 16th edition board, on a MCB on the rcd side

I can see heads rolling for this.....

On another note i have actually completed the installer course and this was never mentioned which i find strange

 

[powervan]

lol i would love to have one of u lads with me next time i get inspected!!

 

[Markc]

Will have to dig a little deeper on this and see what the score is

Just finished installing PV on a housing estate, all pv circuits were put onto the existing 16th edition board, on a MCB on the rcd side

Oh Bugger! May be don't use these for next years inspection.

On another note i have actually completed the installer course and this was never mentioned which i find strange

Nether did they on ours two years ago! In fact the test rig even had a 17th edition board!

Might be worth trying out yourself. Run your RCD tester on one but give the inverter time to go back in generation. (Do it in a warm house rather than a cold garage this time of year as it takes AGES!

 

[vernam616]

How did you find this out?