question

daza avatar image
daza asked

Ac in on Multiplus 2 tripping RCD

Hi I keep getting RCD tripping on the input side of the multiplus2 I have 50amp MCB that feeds a 30mA RCD to the inverter everything in the house is connected to Ac out 1, the whole house has its own RCBO’s on every circuit which are working normally.

when the inverter is turned on and loaded up with over half of the circuits it trips out, no matter what sequence or starting position on the board I use. My system is a TT system not connected to the DNO at all for earth. The system isn’t overloading as when testing everything I could in the house loaded I only drew 48amps, but my dryer and washing machine are not on my output side now they have been moved to grid side.

Im on the assumption that loading all the circuits the RCD is seeing 30mA on the whole install so tripping. has anyone else had this and fixed it I assume it’s a higher rated RCD is the fix but would like to hear from others that had the issue.

rcd tripping
2 comments
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

ejrossouw avatar image ejrossouw commented ·

I understand you have a TT earthing system but remain am curious as to: Do you have have an earth from the inverter to the main consumer unit that supplies it AND also a separate one from the earthing rod to the MP chassis? I assume the main consumer unit will also be connected to the earth rod? I ask because "earth loops" can also be a trigger.

0 Likes 0 ·
daza avatar image daza ejrossouw commented ·
@ejrossouw didn’t think of the inductive loop good thought process, but earthrod is connected to an MET the earth then spans to all the earthing points from there, it looks like when the inverter does a relay test when it’s connected back to the grid is back ie that earth on input that it breaks for the DNO earth then it trips the earth think it either needs a time delay RCD. No no issues with the earth I had earth reading of 2.15 it’s now down to .16ohms I just had my EV charger put in so had to do a ZE said no issues it’s the best earth he’s seen on a TT ever. Think it’s more I time thing rather than sensitivity now.


0 Likes 0 ·
9 Answers
Craig Chamberlain avatar image
Craig Chamberlain answered ·

Hi @daza , personally I don't have an RCD on my feed to AC-In for exactly this reason. I have RCBOs on all circuits which go out into the house but I only have a 50A MCB feeding the Inverter from my primary consumer unit. I stand to be corrected but the Inverter installation manual only specifies over current protection not RCD protection for the input, and since the wires between primary consumer unit and the AC Isolator and ultimately the Inverter, are all inside enclosures and trunking, they don't require RCD protection. If the Inverter developed a fault where the chassis became live then it would very quickly trip the 50A MCB and therefore would not pose a shock risk to anyone.

That's my take, hope it helps.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

matt1309 avatar image
matt1309 answered ·

Hi @Daza

I'd agree sounds like individual circuit earth leakage is under 30ma, but overall it's greater hence the one 50amp RCD trips but RCBOs don't.

If overall RCD is on ACIN and the others are on ACOUT could you test the same loads in off grid mode. If no trips then that seems like the cause.

Don't suppose you've got a tester that can measure earth leakage just to confirm it, would be the ideal test.

It's odd that any configuration of turning circuits on causes the trip though, even if they're all under 30ma you'd still expect there to be one circuit that was a large culprit that would keep you just under the threshold if off.

May also be worth testing that 50amp RCD if you've got a tester.


2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

daza avatar image
daza answered ·

@matt1309 unfortunately one of the only testers I haven’t got :( yes without the grid the inverter is powering the loads without the RCD on AC 1 in tripping which is perplexing me now as it doesn’t trip with everything loaded on the multiplus 2 in off grid.

could it possibly be my TT system as the earth throughout is all the same Ie I’m not using the DNO’s earth so is it a relay thing as the Multiplus does a relay switchover it looks like it trips was running fine until the grid was turned on. Is there something to do with the ground relay inside the Multiplus as I don’t need it to switch as my earth isn’t connected to the DNO’s earth I have an earth rod that does the whole electrical supply

2 comments
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

matt1309 avatar image matt1309 commented ·

Hmm that's a good point.


I suppose the check for that would be does the acin rcd hold if you manually switch it back on once grid is connected? If it does then I think you've worked it out that it's the transferring from offgrid to grid that's the cause.

I dont think the earth relay switches earths as you mention (and anyway the house board is connected directly to earth rod i imagine so it's not like the earth is switching), However the earth relay i think activates/disables a neutral earth link. That's what i think earth relay does which may be the cause. Not sure what the solution is though sorry.



0 Likes 0 ·
Craig Chamberlain avatar image Craig Chamberlain commented ·

Bear in mind that when running on grid power, the L-N relays are engaged inside the Inverter which essentially connects AC-In to AC-Out1 (and AC-Out2 via a further relay). So any cumulative earth leakage from your house loads will be sensed by the RCD on AC-In. It's not uncommon for house appliances to have significant earth leakage and even 3 or 4 circuits might be enough to cumulatively trip a 30mA RCD. If you did want to retain an RCD in the AC-In circuit then you'd probably need to get a 100mA one and when I looked into this I found them to be quite pricey. I also don't think I could find one that was type tested for my Fusebox branded consumer unit. That would mean I would have to install a different brand which is okay but not ideal. That's when I started to question whether I even needed RCD protection on AC-In and ultimately decided I didn't.

0 Likes 0 ·
daza avatar image
daza answered ·

@matt1309 thanks you and @Craig Chamberlain thank you very much you just saved me a lot of money that Hager will no longer get lol

It’s been doing my head in thank you both will test in a bit

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

daza avatar image
daza answered ·

@Craig Chamberlain @matt1309 brilliant thanks guys all on line now but it’s drawing a lot from the grid every now and again and sending back a lot of solar and battery I’ve set the grid set point to -20w but still up and down

3 comments
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

matt1309 avatar image matt1309 commented ·
Hmmm what's monitoring the grid, ET112 or CT?

It could just be a spike in load and ESS not quite keeping up, ET112 or similar meter does tend to have a slightly delay. It might just be something like fridge compressor kicking in/off?

0 Likes 0 ·
Craig Chamberlain avatar image Craig Chamberlain commented ·

Yeah, I used to use a CT clamp and it was pretty responsive and minimised grid import and export. But I now use an ET112 and it does have ramp up and ramp down delays. This was improved IIRC in the Cerbo GX v3.0 release so upgrade to that if you haven't already. There are also other supported energy meters which are much faster now.

0 Likes 0 ·
matt1309 avatar image matt1309 Craig Chamberlain commented ·
I thought the speed improvements were revoked before final release for et112. I did see in v3 beta it was.


Im tempted by the faster meters, they're so pricey though. Im debating going ct. Issue I have is I have a long run to the multiplus 90m. If I ever install a new wire I'm going to throw in another premium cat 6 see if a ct signal can reach that far (pretty sure specs are rated way less than that but maybe with a premium wire)
1 Like 1 ·
daza avatar image
daza answered ·

@Craig Chamberlain @matt1309 yeah I have all three but was planing to use that CT for something else but it appears the multi still reads it even if you do external meter anyway that was the cause I have now just done it on monitored loads guess I will have to get some Shelley’s to give me what I want. I set it to 10watts it’s working well now thank you gentlemen much appreciated

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Craig Chamberlain avatar image
Craig Chamberlain answered ·

I originally had the CT on the main meter line tail to sense grid current, and I originally had my ET112 on the input to my maintained loads consumer unit. But that doesn't give me as much information on my touch screen.

So I moved the ET112 to become the grid meter by installing it in my primary (unmaintained loads) consumer unit instead even though I knew my ramp up and ramp down responsiveness would suffer a bit.

In practice though, my grid consumption and unintentional export is virtually nil. If you sit and watch the touchscreen/remote console then you'll tend to obsess about it but even a 2500w unintended import for 3-4 seconds when the kettle turns on doesn't amount to much. In fact 2500w for 4 seconds is 10000 Joules and there are 3.6 million Joules in a kWh. So it would be around 0.0028 kWh which at 30p/kWh is only 0.084 pence worth. And if the ramp up was linear then it'll go from 2500 to 0w so it would only be half this amount.

So IMO it's better to operate with the ET112 or other meter as grid meter rather than the CT unless you have a particular need for optimum responsiveness. ie. in a case where you are forbidden from exporting at all.

Glad you got it working. Well done!

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

daza avatar image
daza answered ·

@Craig Chamberlain @matt1309 So should I set it too 0watts or leave it at 10watts so it’s given me too monitors one for the once at ac1 in and the grid as well, no thanks to everyone hare that helped me, this place is a true wealth of knowledge.

img-4228.png


img-4228.png (512.6 KiB)
2 comments
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Craig Chamberlain avatar image Craig Chamberlain commented ·

That's pretty much what I get as well, although I have a "PV Charger" instead of "PV Inverter" because mine is DC coupled via Victron MPPTs and I presume yours is AC coupled to AC-Out1. That's good because your PV will continue to work even in island mode.

As for grid set point, I've got mine set to 0w but it really just depends on whether you want to minimise grid import or minimise grid export as to where you choose to set this. If you set this negative then it'll be a constant drain on your batteries which might be more of an issue in the winter when you're less likely to filling your batteries from PV. What you don't want to be doing is charging the batteries during cheap rate at say 9.5p and then just slowly draining them back to the grid at 4.1p or even 0p depending on your tariffs.

So I would say set it to 0w initially and run for a full billing month. Then analyse your consumption either by downloading data from VRM or from Octopus if you're with them. You'll be able to see how much grid import you have outside of scheduled charging times and if that is too much for your taste then set grid setpoint at -20w or even -50w. I don't think I'd go much further than that especially as we get into the autumn/winter.

That said, I have on occasion set my grid setpoint massively negative (eg. -3500w) at times when I have more solar than I can possibly absorb and my batteries are already at 100% early in the day. Why I do this is because recently on very sunny days, I was seeing PV power of anything up to 6.4kW (it's a 6.8kWp array facing due south at 14 degree slope) but it varies quite a lot due to clouds etc. So my average PV might be 3.5kW but with peaks of 6.4kW. But since my inverter can only export a maximum of 4.4kW, once the batteries are 100% full, the maximum my MPPTs will capture from the PV array is 4.4kW and I therefore can't absorb the peaks of 6.4kW. So if I can set a large negative setpoint then the batteries go down to maybe 95% and therefore I can capture the full 6.4kW from PV when the sun is strongest. This only works if the average PV output is less than 4.4kW, otherwise I simply can't export fast enough to keep the batteries below 100%.

HTH

Edit: Just to add - if our EV is at home and needing charged, then I prioritise EV charging over export because it's pretty much always better to consume energy in the home than to export it and then import it later. The Octopus Flux tariff might change that dynamic though in certain cases.

1 Like 1 ·
Show more comments
mjs500 avatar image
mjs500 answered ·

30mA RCD witch type AC- A or B ?

1 comment
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

daza avatar image daza commented ·

@mjs500 type A, type B is an arm and a leg through Hager for stronger DC leakage but that’s not what’s tripping it

0 Likes 0 ·

Related Resources

Additional resources still need to be added for this topic