Array frame earthing

[JulianC]

Have a read of the DTI guide:

"a) Isolating transformer: An isolating transformer is one in which the input
and output windings are electrically separated by double or reinforced insulation
(see BS 3535).
While the hazards presented by an array frame reaching the system d.c. potential
may be significant, the potential fault/shock current is typically much less than
that from a mains fault. Hence it is the electrical separation of the mains from the
d.c. using an isolating transformer that is the key determining factor when
assessing the requirement for array frame earthing."

Array frame bonding required when using a transformerless inverter due to increased liklihood of AC appearing on the frame in the event of a fault, due to lack of electrical separation.

 

[moggy1968]

I like the battery analagy, I can see that. I know what the guide says but, like I say, had a bit of a conceptual problem with it as touching the live on AC will produce a fairly unpleasant effect as you are connecting it to earth, but now I think I see.
I avoid transformerless inverters if I can 'cos of the aggro of running an earth

 

[bigsparks]

It is interesting to read all the different opinions on the array earthing. If you follow the DTI flow chart, they advise that you should earth the frame if it is a transformerless inverter, because there is no galvanic isolation between DC & AC, but the SMA TL's say that have a safety device built in that would ensure that AC does not reach the DC side & vice versa in the event of a fault. The regs say you should install a DC type B RCD which cost £300 each, but the SMA has a device that appears to negate the need for one.
Also, if you follow the 17th Edition (or just good practice) all dedicated PV circuits should be protected by an AC 30mA RCD which again would isolate the supply if the inverter developed a fault to earth over 30mA ie AC onto the array framing, so again the window cleaner should not get a fatal shock if the frame becomes live with AC (he may fall & break his back however) but good sense would tell him to turn the AC side off first or as a minimum hold a volt stick to the frame to see if its live with AC current.

Additionally, and probably most crucially, most of the pv panels on the market are double insulated which, even in the event of a fault, the current that could cause a shock should not be apparent on the frame if they are double insulated.

So taking all those factors into account, PV panels must be the safest pieces of equipment we install!! :lol:

 

[moggy1968]

you don't need to fit an RCD if you have simple seperation according to the big green book (assuming none of the other criteria for RCDs are there).

another reason for using transformer inverters.

Also some inverters (4000TL for example) don't like 30ma RCDs.

If your inverter is next to the board, say if your installing in the garage I really don't see any need for an RCD on a transformer inverter to protect less than a meter of visible AC cable.

If your using an RCD it must be a type B, the usual domestic type A may not operate at all or not operate within the required timeframe.
A common error is to confuse a type B MCB (very common) with a type B RCD (very uncommon!). next time you look at an RCBO see what I mean, it says type B MCB, type A RCD. As above type B RCDs are very expensive and not easy to get hold of.

I would suggest a very large percentage of installations out there don't comply with this.

 

[BruceB]

Also, if you follow the 17th Edition (or just good practice) all dedicated PV circuits should be protected by an AC 30mA RCD ...............

People get blinded by domestic 17th practice, but there is no general requirement for a 30mA rcd in 7671. There are reasons to have a rcd of some sort, for example:

- TT
- agricultural
- cables buried <50mm in walls
- etc

but it will often be better to design out the requirement.

Also the regs do not say "you should install a DC type B RCD which cost £300 each..." for TLs. They say if you have an rcd for a TL inverter then it should be type B, which is subtlely different.

Regards
Bruce

 

[SolarCity]

Big solar
as I understand it, and as I say I do have a conceptual problem with this, if there is a short between the DC and the frame, that frame will hold a potential. it will continue to hold that potential until it is given a path to earth, through a bonding cable if present, or the window cleaner if it isn't!!

Surely there will still be zero potential to earth?

 

[SolarCity]

All elasatictrickery wants to do is get back to it's point of origin.

If a DC pos is touching the array frame, it will do nothing untill it finds the DC neg of the source. It can't flow until that path is made.

This is why bonding DC down to ground is frowned upon, in doing that, you are creating a path.

It isn't frowned upon - it is a necessity on TL inverters.

Beyond taking an earth outside of the equipotential zone, I can't see why it should be a problem.

 

[SolarCity]

I avoid transformerless inverters if I can 'cos of the aggro of running an earth

That surprises me. We rarely use anything other than TL inverters if we can help it. The increase in efficiency is significant and it doesn't take long to bond the frame.

 

[paulsamuel1984]

Biggsolar - as well as bonding the array frames, if required, are you earthing the array frames? i.e down to separate earth spike for TN-C-S or back to main earth for other earthing systems?

 

[SolarCity]

As far as I can see, you are bonding the array and not earthing it. Earthing, unless I'm somehow mistaken, would have no effect whatsoever as the DC voltage will always be present.

 

[paulsamuel1984]

So you are bonding back to a separate earth spike or back to the MET (depending on existing earthing system)? I am sure it is me who is mistaken when referring to earthing!

 

[SolarCity]

back to spike or MET depending on earthing system (much to some electrician's dismay! - Malcomsanford).

The Dti guide mixes up earthing and bonding and to be fair I think most electricians mix the terms up. I'm sure I often say "earthing the array" when what I actually mean is "bonding the array".

 

[paulsamuel1984]

Thanks biggsolar. I will and have been continuing to bond the array frame back to spike or MET as we mainly use transformerless inverters. But as suggested on this post, my concern was that it could potentially (no pun intended!!) do more harm than good.

 

[and]

Reading this thread initially prompted me to unbond my array; read some more of this thread and then went and re-bonded it;now to be on the safe side I have half of it bonded and half un-bonded.

Damn, this thread is hard to follow!
I hope my system is hazard-free:
TL inverter
array bonded (earth wire connects each of the four rails)
earth wire connected all on its own down to a brand new earth spike.

 

[bigsparks]

The upshot of all this is there are pro's and cons to both sides of arguments reference earthing the array with tranformerless inverters. It comes down to everyones own judgement for each different site. some might say earthing array to electrode is good in event ac gets onto frames etc, others would say if u have an ac 30mA rcd protecting the ac pv circuit any earth fault current over 30mA will isolate the supply and protect the 'window cleaner', some would say that earthing the array just increases likelihood of lightning strikes due the earth grounding. The nic tech guys we spoke to said if u have minimised the risk by: installing an ac rcd to pv circuit, using a transformerless inverter which has an in built protection which would not allow ac into the frame in event of fault and the panels are double insulated are you really gaining anything by earthing the array other than creating a higher potential for lightning strikes etc. The chances of the frames becoming live with ac is very slim with all the above things in place. u just need to ensure u have civered yourself in the event of an incident and as said before there are pros and cons for earthing and not