DIY PV install

10-10-20

https://www.boards.ie/discussion/comment/121215977#Comment_121215977

This is actually a really critical document as it explains what we were asking ourselves about 3 months ago when @micks_address had some trips.

https://www.boards.ie/discussion/comment/120700172#Comment_120700172

I was of the understanding that a PV inverter's AC side would by fully isolated from the DC side and hence the DC loss to ground when it was raining was non-comparable to AC leakage. After all you would think that if DC current was lost off the positive terminal of a connection it was effectively untraceable as the DC supply was coming off the solar panels and then inverse of the DC loss would not be seen on the negative terminal, hence showing an imbalance as you would do with AC. But no, that's not true in an AC connected inverter.

It seems that transformerless inverters oscillate the DC supply at around half the AC input voltage (section 2, page 4) and as a result there is AC current present which is capacitively coupled between the silicon panel element and the frame of the panels and the mount and then the roof. They all add-up to become "Cpe", the parasitic capacitance.

During times of very heavy rain on the panels the Cpe increases significantly due to the surface-area of then panels and it causes a leakage trip on the circuit RCD.

The workaround mentioned is to use a correctly sized compensation capacitor on a quad-pole RCD rather than a dual-pole RCD. IE, one of these:

https://www.meteorelectrical.com/distribution-control/consumer-units-accessories-1/garo-63-amp-30ma-4-pole-rcd.html

So yeah, @allinthehead, thanks for posting that!

allinthehead

https://www.boards.ie/discussion/comment/121219531#Comment_121219531

Glad you found it useful. It's not well understood, even by some installers.

What happens then is that the symptoms are treated rather than the cause, ie 100ma rcd.