Voltage on CPC from PV Array.

[TaffyDuck]

Hi Guys

I wonder if someone with more experience of PV might be able to give me some advise.

Was called out today to a domestic building with a remote barn on which a 4kW PV array is sited. The Inverter is in the barn and a long (50m) 2.5mm twin and earth in pipe (historic) feed back to the house.

The problem I have is that there is a full short (L-E) on the cable between the house and the barn. So far no problem.

Just to prove the diagnosis I disconnected the earth at both ends of the cable between the house and the barn and re-energised the system

However I was then getting an Off Grid notification on the inverter and more worryingly 240V present when testing between N-E on from the PV inverter.

Is this energised CPC normal? And would the lack of an earth be picked up by the PV system and cause the OFF GRID notification.

The supply cable I can sort, just need to know if I need to call in a PV company to sort the PV Array out.

Thanks for any help

 

[Vegelen]

The supply cable I can sort, just need to know if I need to call in a PV company to sort the PV Array out.

Solar PV arrays approximate location?

 

[SolarCity]

What kind of earthing set up do you have? Is there an earth rod at the barn?

Incidentally, if the cable to the barn needs replacing then it needs to be upgraded to a larger cable.

 

[Earthstore]

Do your calcs on the cable for sizing, as Solarcity said it needs upgrading.
You need to work on a 1% loss not the normal 3% or the inverter will go overvoltage.

Why not diconnect the cable altogether then do your tests, live to earth, neutral to earth etc on the cable.
I had a fault once on an outside light but only when it was wet, so one rainy day, coat on pinpointed the fault, vermin had chewed the wire...

 

[Gavin A]

that's not normal. I'd be wanting to rule out the potential for the earth being live through a fault elsewhere in the AC circuit before considering that it might be the inverter... though they do have capacitors etc inside them so it's possible these are discharging through the earth or something.

eta - ie is the pipe itself that the t&e is running through bonded at both ends, so the system is still actually connected to the MET through that?

eta 2 - hmm, actually in some countries I believe they sometimes bond the negative pole of the panels to the negative pole of the inverter for some reasons that I forget (we don't do it here), I wonder if the installers maybe done that here for some reason, as well as bonding the actual frame to an earth spike (as was the guidance for transformerless inverters), then if the pipe is bonded in any way it will also be in the ground, so forming a bit of a very low power DC circuit, which just happens to be at 240V. It'd be a bit unusual, but not impossible, as you can get a slight leakage current flowing through the frames... do you know what current is flowing?

 

[TaffyDuck]

Not my cable, first time I`ve set foot in the place. The supply cable looks like it was put in at least a decade ago by a DIYer as its standard 2.5mm T/E in an alcathene pipe.

I have isolated it completely as suggested above so there is no earth path at all. However there is a voltage running through the cpc from the inverter when it is powered up.

 

[TaffyDuck]

PME exported from the house by the way. Panels on roof of barn, isolators invertor and FIT meter all in the barn.

 

[TaffyDuck]

Do your calcs on the cable for sizing, as Solarcity said it needs upgrading.
You need to work on a 1% loss not the normal 3% or the inverter will go overvoltage.

Why not diconnect the cable altogether then do your tests, live to earth, neutral to earth etc on the cable.
I had a fault once on an outside light but only when it was wet, so one rainy day, coat on pinpointed the fault, vermin had chewed the wire...

Yeah did all that before I posted on here. I know the supply cable has a fault, that can be fixed, its this odd energised cpc and the question about OFF GRID above.

 

[SolarCity]

If there is a live to earth fault on the cable then by disconnecting the earth from both ends you are effectively making the earth cable live. This would explain the reading between neutral and (live) cpc.

Edit:

Actually, I may have misunderstood what you were saying. Are you testing neutral to earth AT the inverter? If so, is there any cable connected to the earth connection of the inverter?

 

[Gavin A]

well, I'm confused, so I think I'll just have to wish you luck with this. Though I'd definitely double check for the negative DC > neutral bonding on the inverter as I have heard of installers getting confused by the bonding point on the inverter and asking if they're meant to bond to it (not many mind, but it's possible). If it's a TL inverter could be that there's a patch back through the inverter somehow.

Maybe try it with the DC isolator switched off as well, which will tell you if it's panel related or capacitor related.

erm right, I'm out, keep us posted though please, as it's always good to know how quirks like this occur for future reference.

 

[TaffyDuck]

With the cpc from the inverter disconnected from CU in the barn and the cpc on the supply cable disconnected and L+N connected there is no earth connection to the inverter at all. The disconnected cpc from the inverter shows 240V when tested between the cpc and Neutral. . Glad I avoided these things when so many others were getting involved.

 

[TaffyDuck]

.........the cpc is live