Previously working PV/Battery system starting popping RCBO when SMETS2 meter fitted

[speculatrix]

We had a PV/battery system was fitted in late May, and measurements showed it was working very well, but then a few weeks ago the old meter (electronic box about the size of a large cigarette packet) was replaced with a SMET2 meter and everything went down hill!

When the PV system was working, when the sun rose, the grid draw dropped as the PV took the load, a little later the the battery would charge, and after the battery was full, excess power was exported. If there was a high level of demand, even during the day, battery power would be drawn rather than pull from the grid.


I have a Sofar Hybrid Inverter and a Pylontech 7.2kWh pack, and a Growatt power conditioner (which reduces the 245 to 247V of my local grid down to 225 to 227 ish). It's the Growatt which is the source of my headaches. As far as I can tell, the Growatt is really just a toroidal auto-transformer, with additional RCDs and circuit breakers for protection, and some protective windings in the transformer to detect shielding faults and safely trip things.
Here's my attempt at annotating it:

Everybody was happy. Well, apart from the fact I didn't get paid for exporting energy, because I had no smart meter. Finally, I got a date.

The Landis+Gyre smart meter was fitted on the 19th August.

The day after the smart meter was fitted, I observed that the solar shut down after the battery was charged, there was no export at all, nor was the house being powered by the PV/Battery system. When I looked closely, I found that the RCBO, which is between the PV/Battery system and the grid connection, had failed completely and couldn't be re-engaged. There was no exterior sign of damage - no scorching etc. I contacted the PV/battery installer, and a tech person at Growatt, and after a bit of to-and-fro a replacement RCBO was agreed.


RCBO Replaced On Friday 30th August
An electrician came from the PV installers to replace the RCBO and service was restored. but only for a day, on the 31st, power was being exported until maybe 2pm and then nothing. That evening, when I returned home, the RCBO had popped again.


It sees very suspicious to me that a system that had been working so well for months could suddenly develop a fault when the smartmeter was fitted! Could some sort of earth problem have been created, and this led to the RCBO failing completely? And it's not as if we were exporting more power than before, the graphs I've captured show we hit higher peak electricity generation in July!

I'd be very grateful for any ideas, even if it's just things to inspect. I'm not certified to work on domestic mains, as I'm retired from being an electronics engineer, but I can at least understand the principles, ask the right questions and cautiously take measurement with generic multimeters.

The only hint of progress is that the smart meter people are coming soon to fit the gas meter which they were unable to do last time, and will inspect the wiring on the electricity meter too. I did check the earth bonding when the first RCBO died and it all seemed good, using my limited set of test equipment (Fluke m/m and such) to see if there was any measurable potential between the earth point in the Growatt box and the neutrals, and there was none.

thanks!

 

[Dan]

We had a PV/battery system was fitted in late May, and measurements showed it was working very well, but then a few weeks ago the old meter (electronic box about the size of a large cigarette packet) was replaced with a SMET2 meter and everything went down hill!

When the PV system was working, when the sun rose, the grid draw dropped as the PV took the load, a little later the the battery would charge, and after the battery was full, excess power was exported. If there was a high level of demand, even during the day, battery power would be drawn rather than pull from the grid.


I have a Sofar Hybrid Inverter and a Pylontech 7.2kWh pack, and a Growatt power conditioner (which reduces the 245 to 247V of my local grid down to 225 to 227 ish). It's the Growatt which is the source of my headaches. As far as I can tell, the Growatt is really just a toroidal auto-transformer, with additional RCDs and circuit breakers for protection, and some protective windings in the transformer to detect shielding faults and safely trip things.
Here's my attempt at annotating it: View attachment 52004

Everybody was happy. Well, apart from the fact I didn't get paid for exporting energy, because I had no smart meter. Finally, I got a date.

The Landis+Gyre smart meter was fitted on the 19th August.

The day after the smart meter was fitted, I observed that the solar shut down after the battery was charged, there was no export at all, nor was the house being powered by the PV/Battery system. When I looked closely, I found that the RCBO, which is between the PV/Battery system and the grid connection, had failed completely and couldn't be re-engaged. There was no exterior sign of damage - no scorching etc. I contacted the PV/battery installer, and a tech person at Growatt, and after a bit of to-and-fro a replacement RCBO was agreed.


RCBO Replaced On Friday 30th August
An electrician came from the PV installers to replace the RCBO and service was restored. but only for a day, on the 31st, power was being exported until maybe 2pm and then nothing. That evening, when I returned home, the RCBO had popped again.


It sees very suspicious to me that a system that had been working so well for months could suddenly develop a fault when the smartmeter was fitted! Could some sort of earth problem have been created, and this led to the RCBO failing completely? And it's not as if we were exporting more power than before, the graphs I've captured show we hit higher peak electricity generation in July!

I'd be very grateful for any ideas, even if it's just things to inspect. I'm not certified to work on domestic mains, as I'm retired from being an electronics engineer, but I can at least understand the principles, ask the right questions and cautiously take measurement with generic multimeters.

The only hint of progress is that the smart meter people are coming soon to fit the gas meter which they were unable to do last time, and will inspect the wiring on the electricity meter too. I did check the earth bonding when the first RCBO died and it all seemed good, using my limited set of test equipment (Fluke m/m and such) to see if there was any measurable potential between the earth point in the Growatt box and the neutrals, and there was none.

thanks!

I've attached the image directly to the thread and approved your post.

 

[speculatrix]

Here's a great example of when the sun comes up, powers the house, charges the battery and then finally starts exporting:

forum says I'm spamming if I actually put the image inline.

blue line is PV power, yellow is battery (dis)charging power (above 0 = charge), green line is battery state of charge, and red is grid export/import (above 0 is export).

we went on holiday so we didn't use any power in the evening, so battery discharge was zero and battery stayed at 100%.

 

[Dan]

Once you have a couple more posts under your username it'll stop tripping the spam checker.

I prefer the images in the threads. Other hosts have have come and gone over the years.

I'll tweak the spam filters tomorrow.

 

[snowhead]

Where is the Smart meter in relation to the Consumer unit /RCBO?

 

[Wilko]

Hi - sorry I maybe missed it - the new RCBO was installed and and some hours later tripped, could it be reset? What is the status now? What rating was the old / is the new rcbo?

 

[speculatrix]

Hi,
so the new RCBO was implanted in the Growatt, and for 24 hours it looked fine, and then popped again the following afternoon.


so now we're running entirely on grid power.

just to add insult to injury, on the first visit when the gas meter was fitted and removed (because of a jurisdiction argument) the boiler pilot light went out as expected. and then we found that the igniter in the boiler had died, so we have no gas-powered hot water, only the small immersion, and no heating. still waiting for the gas fitter to have enough time to buy and then come and fit the igniter :-(

 

[Wilko]

Hi - so was the new rcbo able to be reset at all? What rating was the old and new? (Something like B32 30mA perhaps).

Sorry to hear about your boiler, what a pain.

 

[speculatrix]

the RCBO was a 16A one.

the paddles were floppy, no possibility of re-engaging it, so it was entirely u/s.

Thinking aloud, I think the RCBO was at its limit, the solar array is capable of 4.8kW peak, the inverter is 5kW, so let's say 4.5kW with inverter losses. Our grid is 245V, so that makes 18A. Could it have been that if the domestic load hit over, say, 6kW, and with full battery and strong sunlight, the inverter drew on both battery and PV to hit the full 5kW output, (4.8 with losses) and if the grid voltage drooped, then the current could have exceeded 20A through the RCBO and blew it?

 

[Wilko]

Hi - well a 16A one would seem too low for a 7.2kW array (original post) and it may be running above its rated value each day. If it was a lot over it’s rated value, say 5 times, it would trip in milliseconds. At low level overloads it will likely eventually trip, but it’s not good for the device. Having said that, an overload like this may cause degradation long term but it shouldn’t cause a new RCBO to immediately fail. That’s why I’m asking if you can reset the new RCBO after it tripped please?

 

[speculatrix]

sorry, to clarify, it's a 4.8kW solar array and a 7.2kWh battery capable of about 3kW discharge, with a 5kW inverter, so the max power from the PV/batt system would be about 4.9kW after losses.

the PV installer sent their guy over, managing to fit in a surprise visit, and uprated the RCBO to a B20 from a B16. Hopefully this will resolve the problem. I'm not qualified to say whether the Growatt is designed for the larger breaker. Hopefully all will be ok, fingers crossed!!

 

[Wilko]

No worries
It’s good that they’re on the case, and 20A RCBO would seem a better fit (to me) for a 4.8kW solar array. They should know what’s right - maybe they were thinking “it’s England, it’ll never make full output” ... ha!

 

[speculatrix]

when the RCBO originally blew, I thought it would be an easy matter for them to get a replacement, but there ended up a three-way discussion between the installer, the wholesaler and the OEM.

In an effort to speed up the repair, I thought I would buy an RCBO myself, how hard could it be to find one online? Well, after hours, literally hours, of searching the internet I found an RCBO which was the right trip curve (B), the right orientation (neutral on left) etc and ordered it even though it was 20A not 16A, because nowhere seemed to have a 16A one meeting all the criteria!

I got it from Electrical World, with a discount code and being my first ever order, price wasn't too bad!