Pv on rcd

[SolarCity]

Well not quite ...

You'd need one both ends, and the grid supply end would need to be time delayed, as you want the the Inverter end one to trip first and then the grid end one.

If you only have one at the inverter end, the cable would still be 'live' to the grid'
If the grid end one isn't time delayed, then you can't guarantee to trip the inverter one first, and if you don't it would still be powered by the inverter.

Hence why we design out the requirement

Why is that?

 

[Deleted member 26818]

That's how's it's installed either mistakenly to offer Rcd protection or couldn't be bothered to fit an additional board and earth rods to get the Ra acceptable I imagine

Well it's clearly wrong.
Needs to be changed so the supply is on the supply side of the RCD.

 

[gnselectricals]

Will chuck them an email with my concerns and see what they say then

 

[davesparks]

Why would the PV supply be connected to the load side of an RCD?
PV is a secondary supply and should be connected at or very near to the origin of the installation.

There is one hell of a difference between where it should be connects and where some idiots actually connect it.

The trouble is that the inverter is mounted remotely from the CU, so there is a cable run from the inverter to the CU. if that cable gets chased in then it would need RCD protection.

 

[davesparks]

Ok, having slept on this and thought a bit more about the regulation 551.7.2 (iv) I think this may not be quite as clear cut as I first thought.

The regulation actually allows the PV to be connects on the load side of the RCD provided that the RCD disconnects all live conductors and that the inverter supply is not reference to earth. The theory being that the RCD disconnecting both L and N from the grid will disconnect the earth reference from the inverter supply and therefore make the inverter output floating with respect to earth and so not deliver a shock (in the same way as electrical separation).
This however raises the question of whether the inverter output is fully isolate from the earth? And of course what happens in the case of a N-E fault on the load side of the RCD?

 

[davesparks]

Why is that?

Because the grid connection may be the only earth reference that the inverter output has, so disconnecting the grid supply first may prevent the PV end RCD from operating.
But that leads to the question of whether the RCD at the PV end would then be needed at all.

 

[davesparks]

Well not quite ...

You'd need one both ends, and the grid supply end would need to be time delayed, as you want the the Inverter end one to trip first and then the grid end one.

If you only have one at the inverter end, the cable would still be 'live' to the grid'
If the grid end one isn't time delayed, then you can't guarantee to trip the inverter one first, and if you don't it would still be powered by the inverter.

Hence why we design out the requirement

Which of course wouldn't comply as you can't have a time delay on a 30mA RCD used in this way.

It could be done using a shunt trip on a DP breaker at the PV end triggered by an aux contact on the supply end RCD.

 

[Worcester]

Why isn't time delayed acceptable? That's what was suggested by our NICEIC assessor in a similar situation


Or is it the 30mA time delay that you are saying isn't acceptable? - I didn't quote values...

 

[davesparks]

Why isn't time delayed acceptable? That's what was suggested by our NICEIC assessor in a similar situation


Or is it the 30mA time delay that you are saying isn't acceptable? - I didn't quote values...

A time delayed RCD won't disconnect in the required time.

If the RCD is 100mA for fault protection in a TT then it would have to be time delayed at both ends for the usual reasons. A non-TD RCD at the PV end wouldn't discriminate with the 30mA RCDs used elsewhere in the installation.

If they are 30mA RCDs providing additional protection in the usual manner then they have to disconnect in 300mS at 30mA

 

[GEOFF S]

Hi

Interesting thread, PV is not an area im involved in but i would like to know more.

It seems to me that the PV supply needs to be connected upstream of all final circuits, connection made via a CPD.

With regards to a RCD on the PV ac supply, providing the inverter has simple seperation, where is the earth reference for the RCD?

Cheers

 

[davesparks]

Hi

Interesting thread, PV is not an area im involved in but i would like to know more.

It seems to me that the PV supply needs to be connected upstream of all final circuits, connection made via a CPD.

With regards to a RCD on the PV ac supply, providing the inverter has simple seperation, where is the earth reference for the RCD?

Cheers

I know next to nothing about PV, but sources of supply are all basically the same.

That's one of the things I want to know, do all inverters have outputs which are electrically separate from Earth?

 

[GEOFF S]

I know next to nothing about PV, but sources of supply are all basically the same.

That's one of the things I want to know, do all inverters have outputs which are electrically separate from Earth?

Well you can earth the d.c side of the inverter providing the inverter has seperation.

Now the seperation is there to stop d.c current feeding onto the a.c side. Now where simple seperation is not provided by the inverter then a type b rcd needs to be installed.

Cheers

 

[Deleted member 26818]

No.
In some earth and neutral are connected.

 

[GEOFF S]

No.
In some earth and neutral are connected.

Hi

Are you referring to HV systems, can you expand?

Cheers

 

[Deleted member 26818]

LV 3 phase inverters.
In particular the ABB Trio range.
Although there are separate connections for each phase, neutral and earth, the earth and neutral are connected together internally.
In fact it appears that the neutral is nothing more than another earth conductor, as there is no internal connection from the neutral to any other part of the inverter.