diy solar

diy solar

Growatt rcbo or rcd tripping or disconnecting

I have installed many spf5000 (single.phase single inverter) with standard 30mA RCD upstream, and had no problems at all. However, when installing 3 SPF5000 in paralell I had to remove the rcd from the main supply, It kept poping, nor even a 300mA worked...
 
Interesting comment. I should have mentioned my experience on the input side of the inverter.

There are two RCDs needed in the UK: one for the 240V output and one from the mains to the inverter for battery charging/utility supply input when there is insufficient sun.

The Type A mentioned in https://www.4blue.nl/media/wysiwyg/Downloads/Downloads-2020/RCD_Declaration-Growatt_.pdf is for the inverter output.

On the output side Growatt tells us that the RCD needs a 300mA margin, this must be because the line and neutral wires are carrying different currents during startup. But how could that happen? In normal domestic situations this line-neutral difference is almost always leakage to earth.

I have a changeover switch downstream of the inverter so that I can swap from inverter power to mains if necessary. When I close this switch the 300mA output RCD pops. So going from no load to about 500W load in about 100 millisecs causes the line and neutral from the inverter to have more than 300mA difference in current. If I reset the RCD the unit operates perfectly. If the inverter is started under load or no load the RCD does not pop, it is only when a sudden load is applied that it pops.

It is remotely possible that there is an inductive effect of the domestic circuit (fridge motors etc) but this should be a problem for RCDs generally if it happened, even in the absence of an inverter. It is more possible that there is DC leakage that causes the current difference, however this should also happen if the output RCD is reset. My best guess is that the inverter is transiently leaking to earth. This might be checked by, just for test purposes, disconnecting the inverter from earth then operating the changeover switch. That the output RCD can be reset, presumably imposing a sudden load, is a mystery. The effect is transient so I would guess that capacitors are responsible, especially the back up battery which is a giant capacitor. It is possible that the negative side of the battery is earthed (the battery casing is earthed) so disconnecting the battery/battery earth might be interesting.

The input side also has problems. I had a 40 amp RCBO on the input but this popped so I replaced it with a 40 amp mcb because the wiring to the inverter input is not hidden in a wall (an exemption in the UK regulations). If the inverter is inclined to pop both input and output residual current detectors it looks very likely that there is earth leakage inside the inverter at startup or during sudden loads. However, I cant see how an imbalance of line-neutral current on the input would affect the output RCD.

The big question is, having got the system working within UK Regulations*, do I want to spend a few hours finding out what is happening? I probably should "fault find" by flipping the changeover switch in the following cases:

1. disconnecting the inverter and feeding the input mains straight to the output RCD.
2. reconnecting mains to the inverter and running a single inverter on utility only supply with the earth disconnected
3. reconnecting earth
4. connecting batteries

My guess is it will fail at (2) which will be the only fault that I cannot explain

*technically the regulations require Type B RCDs on both the mains input and inverter output but the Growatt assurances that the inverter has DC residual current detection permits a Type A RCD on the output.


:)
 
Last edited:
Back
Top