usual question - pv array earthing!

[paulsamuel1984]

I am currently carrying out a design for a 4kW system, using class 2 modules, cables and connectors. using sunnyboy 3.8kW (with isolation transformer).

Having read the dti guide i am sure I am to leave this system floating (TNCS earthing system). i.e. I will not connect/bond the rails, frame etc, and will then not connect to new earth spike.

can someone confirm if this is the right thing to do?

 

[BruceB]

That looks correct. Is there any reason you think not?
Regards
Bruce

 

[malcolmsanford]

If the frame is not an extraneous conductive part then there is no need to bond it and no need to earth it if the invertor is a transformer type and therefore has seperation.

 

[paulsamuel1984]

Thanks for this. The reason I am second guessing myself is because I have been speaking to a fellow installer who bonds framing / mounting structure etc. together even though the array is left floating. so, in short everything is bonded together but there is no connection to either earth rod or met.

Am I missing something? there is no merit in carrying this process out unless you are connecting to earth. correct?

 

[malcolmsanford]

Basically he as got a lot of pretty Yellow/Green cables between the frames and the brackets, so totally useless

 

[paulsamuel1984]

Exactly the point I made to him, maybe he finds it aesthetically pleasing!

 

[ihugs]

Earthing of the array is not required,also earthing for lightning protection is not required. you only have to earth the array if the inverter does not have an isolating transformer between AC and DC parts or any conductive part of the array and frame is in the equipotential zone.

 

[malcolmsanford]

Earthing of the array is not required,also earthing for lightning protection is not required. you only have to earth the array if the inverter does not have an isolating transformer between AC and DC parts or any conductive part of the array and frame is in the equipotential zone.

Why would you want to earth a class II installation or bond a non extraneous conductive part, even if it is within an equipotential zone?

 

[ihugs]

If the inverter does not have an isolating transformer between AC and DC parts, like i said in last post.

 

[SolarCity]

Is there no merit in bonding the rails together, without then bonding this to the MET?

 

[paulsamuel1984]

that is my question. i cannot see any merit in bonding the rails together if there is no connection to earth. anyone agree or disagree?

 

[SolarCity]

How large is the risk that one rail could have a negative fault to it and another rail have a positive fault to it?

 

[malcolmsanford]

How would these rails develop a fault, they are not exposed conductive parts because the panels are Class II equipment, and unless there is a chance of the rails being at a different potential and therfore a extraneous conductive part, which is highly unlikely on a floating rail, then earthing and bonding can be more dangerous, as you could transfer a fault from the equipotential zone onto the frame work.

 

[SolarCity]

Perhaps a fault on the cable? Perhaps from damage from vermin or birds?

I just think there is a special risk with PV in that the installation is hidden from view yet not shielded from damage. These systems are likely to be on a roof for a long time.

I'm not saying bonding the rails together is the right thing to do, I'm just asking other opinion.

 

[malcolmsanford]

Perhaps a fault on the cable? Perhaps from damage from vermin or birds?

I just think there is a special risk with PV in that the installation is hidden from view yet not shielded from damage. These systems are likely to be on a roof for a long time.

I'm not saying bonding the rails together is the right thing to do, I'm just asking other opinion.

Not sure if birds would attack it Briggs but I do see your point about vermin or squirrels. Is there in corelation between Panel double insulated tails and vermin attack.

 

[edexlab]

I met an installer from an established firm when we started Pv and he'd been to the same job twice to replace module cables which had been chewed by squirrels ,and were causing some arcing every time the wind blew ,these cables were cable tied but some had also been attacked by the tree rat (squirrel) ,anyway the customer was an animal lover until he got the bill , he then hired a couple of blokes to come and shoot the offenders and has'nt had any problems since.

 

[malcolmsanford]

Do any of the other PV guys on here have experience of vermin attack.

I have to be honest and say though far from experienced in PV installation, every one I have seen or done the panel tails are just basically cable tied to the array frame and left un-protected. As edexlab above pointed out, should the tails now be protected under regulation 522.10. Do any of the PV guys protect the tails from fauna and if they do, it would be interesting in how they do it.

 

[paulsamuel1984]

I do not provide any mechanical protection for the cables, but cable tie to the array frame. I have not had any vermin issues (as yet!) with any installs. This has raised some interesting points regarding bonding. As earlier stated, under the correct circumstances I have not provided any earthing / bonding for the dc, including connection to earth or any bonding of the frame.
But as a result of the above, I would be interested to know how people protect the tails to further justify not bonding the frame.

 

[SolarCity]

Bonding the rails together takes 10 minutes. We've had many discussions on site about it and we do it on all our installs now (although to be fair, most of our installs are on TL inverters now anyway)

 

[malcolmsanford]

IMO as the installation is a class II type then the frame is not an exposed conductive part and therefore does not need earthing, agreed if the invertor does not have AC/DC speration in the form of a transformer that is a differnt matter.

The same for bonding, if the frame is floating it can not introduce a potential and therefore is not an extraneous conductive part and should not be bonded, as there is a possibilty by bonding is making the frame prone to become LIVE during a fault within the equipotential Zone.

So if there is in the designer opinion a chance of fauna attack on the exposed panel tails then you have to protect them by regulation 522.10.1 and either of the 3 options. Well relocating can not be one, so it will have to be mechanical protection either in the form of a sheathed metallic cable which would not be cost effective or conduit/trunking which IMO would be the easiest.

 

[malcolmsanford]

I deal with inverter sales both here in the the UK and also USA. As it stands in the UK if the inverter has a transformer in it you don't need to bond the rails to earth. If it doesn't (TL) you do.
In the USA under UL regulations - ALL PV installs have to earth the frames - regardless of the inverter type. Surely they are doing this for a good reason?

Think of it this way. Faulty panel, engineer on aluminium ladder in the rain climbing on roof and holding the external tv ariel mast to steady himself. What could possibly go wrong?

Ken

There is a very goood reason for this, they have a completely different earth arrangement to the UK ......

As for your scenario.....................nothing as the panel is a class II piece of equipment which relies on double insulation for protection, the TV mast ariel is just a bit of metal and will have no way of being able to introduce a potential .................

 

[Worcester]

TV aerial mast usually (unknowingly) connect to earth.

Screen on TV cable is connected to mast, and also to the TV/Digibox, Digibox/TV chassis connected to earth.

 

[malcolmsanford]

is your digi box and TV a class I piece of equipment??