Is my PV setup safe? Should my Inverter be grounded?

[SolarLover]

Hi

I have a new on-grid system of:

17 x Q Cells Peak Duo Black ML-G9 380w Panels (6.46 kWh array) split over 2 strings 10 & 7

4 x Growatt PowerAmp ML33RTA 3.3 kWh batteries (13.2 kWh)

1 x Growatt SPH3000 6.6 kWh Inverter

My inverter is not grounded & I’ve checked with my installer and they say the Inverter does not need grounding; whereas the Growatt manual says it does!

They say “grounding the PV array this isn’t needed unless we are installing on a erroneous roofing material, we use that earthing point on the inverter to bond the 6.5kw batteries but there is no point available on the 3.3 kw batteries. The battery storage is 50 volts low voltage”

Our panels are on a tiled roof with inverter & batteries inside garage.

Is this correct?

Pics attached

Thanks in advance

 

[binky]

We don't generally ground (earth) the array on the roof, but do generally bond the battery cases to the inverter, which in turn has an earth connection to the circuit it's connected to. I've not fitted Growatt, so maybe what I have fitted is different.

 

[SolarLover]

We don't generally ground (earth) the array on the roof, but do generally bond the battery cases to the inverter, which in turn has an earth connection to the circuit it's connected to. I've not fitted Growatt, so maybe what I have fitted is different.

Thanks for the reply

I found this in the Growatt SPH3000 manual that concerned me

 

[binky]

Yep, big bonding terminal, battery packs should have something similar. I've never been completely convinced of the need to bond, but if the manual says it should be done, it should be done.

 

[Gian]

Yep, big bonding terminal, battery packs should have something similar. I've never been completely convinced of the need to bond, but if the manual says it should be done, it should be done.

Our of interest should the panels and rail be bonded and would it not be good practice to run a bond from arrays to inverter to allow test equipment like the seaward pv150 to connect up to test ground faults issues

 

[binky]

Our of interest should the panels and rail be bonded and would it not be good practice to run a bond from arrays to inverter to allow test equipment like the seaward pv150 to connect up to test ground faults issues

No, not necessary unless you have AC on the roof ie microinverters on a TN-C-S earthing system.

I've always been interested how inverters or the seaward test gear detect earth leakage without a bond, but they do.

 

[Gian]

No, not necessary unless you have AC on the roof ie microinverters on a TN-C-S earthing system.

I've always been interested how inverters or the seaward test gear detect earth leakage without a bond, but they do.

I simply run an earth cable down on all installs now. I was advised by an expert.

 

[binky]

I simply run an earth cable down on all installs now. I was advised by an expert.

but then you are making a good attraction for lightning .

What was his reasoning for this? You could be, potentially, be introducing DC fault of 300/400V onto the earthing for the property without ADS for the DC side, even if the inverter detects the fault and stalls the DC side, 'if it's daylight, it's live'

 

[Gian]

but then you are making a good attraction for lightning .

What was his reasoning for this? You could be, potentially, be introducing DC fault of 300/400V onto the earthing for the property without ADS for the DC side, even if the inverter detects the fault and stalls the DC side, 'if it's daylight, it's live'

It's placed in an IP rated box and only used for test purpose to save getting ladders out to connect when testing

 

[binky]

It's placed in an IP rated box and only used for test purpose to save getting ladders out to connect when testing

Cunning idea! I'm still not completely convinced of a need for this, if the inverters can find earth leakage without a bond, I suspect test gear can too?