Pv array earthing why?

[power7171]

I am confused why you have to carry out a bond to the array if you use a inverter such as the Sunnyboy HF>

To me when u install the array on a roof it is out of reach and you will cause problems if you bond the array to the main earth of the house because of lighting strike.

If you bond it with a spike you creating a earth on the roof which is out of the earth zone and most panels are double insulated.

Doesnt make sence to me anybody help? I have read the DTI and there view on earthing.

 

[svm]

You don't need to earth the array when using Sunnyboy HF series as they have an isolation transformer built in.
The DTi publication "Photovoltaics in Buildings - Guide to the installation of PV systems 2nd Edition" mentions this on page 23 (Array Frame Earthing Decision Tree)

 

[power7171]

Sorry yes i got the model wrong sunny boy TL

 

[svm]

The array should be bonded to ensure that if a fault occurs within the inverter and / or module and the frame becomes live or the frame charges up due to capacitive effects that this is shunted to a local earth preferable just below the array location.
This is done to stop someone placing a metal ladder below the array and perhaps cleaning the panels and being electrocuted. The above will ensure the ladder is at the same potential as the array.

 

[power7171]

If a fault occured to the inverter it has a earth connected via the AC feed and if the a fault should occur to the module it should not become live as they are class 2. Well the Sharp ones are not sure if all panels are class 2 most probably are. So still dont get the point of the earth bond.

How can the frame charge up due to capacitive affect?

 

[malcolmsanford]

It is not just fault currents you need to worry about.

All inverters will "leak" and if there is not the separation between AC/DC you could have stray voltages leaking upto the array, and if you had breakdown of insulation for any reason they could induce that voltage onto frame.

The worry then is someone working off a ladder sitting on the earth could make a good conductor for these leaked voltages.

 

[VMan]

Ok i'm installing my 1st TL inverter, Fronius IG 3.6 TL, just going throught the manual and found this:

'NOTE The inverter is designed exclusively to be connected andused with non-grounded solar modules (protection class II). The
solar modules cannot be grounded at either the positive or negative pole'

Im assuming i still need to bond the arrray to an earth rod / MET?

 

[SolarCity]

Yep, you still need to bond it. Don't worry about the positive-negative grounding info, it's nothing to do with the bonding.

 

[VMan]

Yep, you still need to bond it. Don't worry about the positive-negative grounding info, it's nothing to do with the bonding.

Brill and earth loop impedance of the electrode in accordance with a TT system (< 200Ohm)? if on PME?

 

[VMan]

ok got another bonding question,

Scenario : Array on roof, TL inverter and PME supply.

The array on the roof, is this inside the equipotential zone?

Then what if there is a velux window and the array can be touched from this window, now i would say definatly in the equipotential zone?

so i should bond to MET? anyone provide guidance on the resistance of this bonding conductor (<0.05ohm? ) or just use 10mm as in DTI?

Array NOT in equipotential zone:
OK the earth spike cannot be outside EPZ, but i cannot see how this would help someone working on the roof or working off a ladder to clean the array.
My thoughts are to put the spike directly below the array just outside the property, where would you guys suggest?

thanks

 

[malcolmsanford]

ok got another bonding question,

Scenario : Array on roof, TL inverter and PME supply.

The array on the roof, is this inside the equipotential zone? No, unless you have extended the zone by extending the PME to the roof

Then what if there is a velux window and the array can be touched from this window, now i would say definatly in the equipotential zone? Again no you can only extended the zone if extend the earth, all you have is a floating piece of metal on a roof, the roof is not within a zone.

so i should bond to MET? anyone provide guidance on the resistance of this bonding conductor (<0.05ohm? ) or just use 10mm as in DTI? If you bonding to a PME system then you bonding size of the conductor is taken from Table 54.8 of the regs which tell us for a supply neutral not exceeding 35mm the bonding conductor should be 10mm

Array NOT in equipotential zone:
OK the earth spike cannot be outside EPZ, but i cannot see how this would help someone working on the roof or working off a ladder to clean the array. Sorry where would you put the rod?

My thoughts are to put the spike directly below the array just outside the property, where would you guys suggest? I have always fitted rods as close to the property I can that will give me good soil conditions, but not close enough to be near the wall to pick up rubble from the build.


thanks

..........................................

 

[BruceB]

I would disagree with Malcolm on his second point. If you can touch the array from a Velux window. then I would bond the array to the MET. My reasoning is that otherwise you could be touching metal work earthed to the ground locally as well as touching something that was earthed in the house that could be at a different potential.

Whether the roof is within or without the EPZ is I believe moot, but if you can touch it from within the house then to my mind that makes it within.

Regards
Bruce

 

[malcolmsanford]

Thanks Bruce I worded that bit badly for a TL inverter.

I wanted to say that the roof in general would be beyond EPZ but if you can touch it from a window then you have to identify and assess if you want to extend the Zone onto it or use a rod.

I said on a previous post that I felt the blanket decision of the DTI on PME systems and not extending the EPZ could be more dangerous than the cure, especially if you have other metal work on the roof, which could be a metal flue, or an aerial mast that is bonded at the MET, and your introducing a different potential via a rod.

 

[BruceB]

Agreed, and I look forward to a sensible edition 3 of the DTI Guide........

 

[trebormint]

I read it that a TL inverter will require an earth stake & 10mm conductor. It should not be bonded to the consumer unit as it's outside the equipotential zone. Agree?

 

[SolarCity]

Yes, on a TN-C-S system

 

[Werneroeder]

And how about a TT system using ClassII equipment and mounting on a wooden structure, where the array is 500 meters from the CU? Would I use a localised groundspike, but do not connect it to the earth coming from the CU?

 

[malcolmsanford]

Why would you want to earth/bond wood?

 

[SolarCity]

I'd be careful installing a TL inverter on a TT system anyway. TL inverters don't like high impedances

 

[moggy1968]

personally i always bond the array and always bond to the MET if it is attached to the house.
the registering bodies will allow you to transport the earth to a detached garage as long as it is 'not too far away' so if thats allowed I see no reason why the array cant be treated the same, especially as it's actually attached to the building.
If there was a fault on the array due to a damaged cable it could short on the array presenting a hazard to anyone going up to work on it. given that this also involves working at height this could potentially be very hazardous, thats why I bond it.
If the array can be reached, i.e. through a velux then earthing the array is an absolute.

 

[SolarCity]

personally i always bond the array and always bond to the MET if it is attached to the house.
the registering bodies will allow you to transport the earth to a detached garage as long as it is 'not too far away' so if thats allowed I see no reason why the array cant be treated the same, especially as it's actually attached to the building.
If there was a fault on the array due to a damaged cable it could short on the array presenting a hazard to anyone going up to work on it. given that this also involves working at height this could potentially be very hazardous, thats why I bond it.
If the array can be reached, i.e. through a velux then earthing the array is an absolute.

You're in the minority with that conclusion but I have to say that I fully agree with you.

 

[malcolmsanford]

personally i always bond the array and always bond to the MET if it is attached to the house.
the registering bodies will allow you to transport the earth to a detached garage as long as it is 'not too far away' so if thats allowed I see no reason why the array cant be treated the same, especially as it's actually attached to the building.
If there was a fault on the array due to a damaged cable it could short on the array presenting a hazard to anyone going up to work on it. given that this also involves working at height this could potentially be very hazardous, thats why I bond it.
If the array can be reached, i.e. through a velux then earthing the array is an absolute.

'not too far away' So what is the definition of that 5 metres, 10 metres anything over, not sure why you consider extending the equipotential zone is distant related, unless you have extraneous conductive parts that needed bonding.

If the array can be reached, i.e. through a velux then earthing the array is an absolute is that for any installation regardless of the inverter type??

 

[moggy1968]

'not too far away' is exactly what ELECSA said to me, I pushed them for a definition but they wouldn't be commited. That was on the phone. This was my assessment installation and the assessor didn't querry it. the garage was about 5 metres from the house, so a fair distance. Yes, for any inverter. an Inverter with a transformer gives electrical seperation of DC and AC circuits, it has no bearing on earthing the array. likewise I saw somewhere on here a suggestion that some inverters effectively have their own 'RCD' in them (although different terminology is used) and this negates the need for an earth. this if course won't affect the array bonding because the inverter is 'downstream' of the DC cables and array.

 

[malcolmsanford]

Well shame on ELECSA it seems they are being influenced by the NICEIC and having their take on the regulations, and by an assessor just blanketing giving advice such as "not far away: IMO opinion is bordering on negligence, not far away give me a break!!!

It would only be TL transformers you would earth/bond the array frame, I just wonder wondered as the statement read that any frame that can be touched while in the equipotential zone should be so,

 

[moggy1968]

As I understand it bonding the frame has nothing to do with the transformer. the transformer gives electrical seperation of the DC and AC. we are treating the frame as an extraneous conductive part and I would always be inclined to earth the frame, regardless of what transformer is used.