Date & Time: 13/09/2025, approximately 22:00 Location: [User’s installation site]
System Overview
The installation consists of the following components:
SMA 5000TL inverter, connected to AC Out1.
MPPT 450/100 DC/DC converter, directly charging the battery bank.
Three Multiplus II units, configured in three-phase, with the entire load connected to AC Out1.
A Cerbo GX for system control and monitoring. SW VR 3.65
Four battery boxes, each containing 16 kWh of LifePO₄ cells, managed by a JK BMS model PB2A16S20P V15.
An EVCS NS Station (Electric Vehicle Charging Station).
DESS is active
Incident Description
During a thunderstorm on the evening of 13 September 2025, the following sequence of events occurred:
A lightning strike was registered in the vicinity.
At the time, the battery bank was discharging at a rate between 220 and 250 Amps.
The residual current device (RCD), located immediately after the grid connection, tripped, resulting in a complete loss of grid power.
The entire home installation went dark, and the backup system failed to activate.
The Multuplus system appeared to continue operating and attempted to feed power into the grid, but no power was supplied to the critical loads.
An attempt was made to reset the RCD, but due to the load from the Multiplus units, it tripped again immediately.
The system was then manually bypassed using a transfer switch, restoring grid connection to the critical loads, which allowed the system to restart.
The three Multiplus units were manually switched off using their hardware switches. After waiting for the battery discharge period to end, the units were restarted, and the critical loads were reconnected to AC Out1 via the transfer switch.
Primary Concern
The backup system did not activate during the grid outage, resulting in a complete loss of power to critical loads, including two large pumps essential for stormwater management. This posed a significant risk, and it was only due to the user’s presence that manual intervention was possible.(3 months ago
Request for Analysis
Please investigate and advise on the following:
Why did the backup system fail to activate despite sufficient battery capacity and system configuration?
Are there any known issues or configuration settings that could prevent the Multiplus units from switching to island mode or supplying power to AC Out1 during grid failure?
What measures can be taken to ensure automatic backup activation in future events?
· Additional Information
· Approximately three months ago, a manual simulation of a grid outage was performed by switching off the main grid breaker. During this test, the Multiplus units successfully took over and continued to supply power to the critical loads without issue. This confirms that the backup functionality was operational under controlled conditions.
I’m not that clued up on 3phase but I would say your RCD is the weak link it’s done its job and tripped possibly the lightning causing live and neutral to imbalance.
How much current was being drawn by the appliances for restart of the RCD and what, also is the RCD input or output is there and earth rod is it a delayed RCD?
I know I’ve just asked a load more questions also are you able to show the schematics so it can be picture layout wise. UK install?
Is it setup for ESS as that should have discounted AC2 and AC in so should not continue to discharge to the grid, is the relay tests being done by the multi, is there a no delay on the RCD or is it RCBO but the latter doesn’t go up high in amperage, assume it’s 30mA
40A RCD isn’t much and I would not be surprised if it has failed so if 230v by 12kW has been going through that then it’s 52amps. The transfer switch on the back end switches all 4 L123 and neutral not just the L123?
Faulty RCD could be the issue now which would explain why it wouldn’t latch for ages and triggered could be over sensitive but if you have that amount of power going through it I wouldn’t think it would last too long, but this is my assumption as to why it worked on the dry run and now doesn’t, assume no one has tried to test grid out again since this?
Yes I’m getting there just working through possibilities so that RCD is working overtime and outside of it’s design window which could be the issue with the tripping. The transfer switch do you have a data sheet for that as if it’s not switching neutral as well that could be why the one down stream is failing critical load side. Again what rating is the critical load RCD and is it 3 phase also 3 phase
The issues are two fold as in a grid down no power should go to the grid that’s a safety issue and the CT’s should have picked this up. Just waiting for the stuff on the transfer switch data sheet or model and what RCD is used critical load side
Is it this as if you are using the neutral bar ie the blue bit then ac in and AC 1 out are linked even in a grid down which would result in the system seeing power the RCD tripping and the system still thinking it’s connected to the grid this is just my assumption, I would check all the switching gear and that ac in and ac1 outs neutrals are not linked I was trying to get a diagram of the switch but can’t find one to see the make and break of the neutral
withe MP2 the ACin, ACout1 and the inverter are directly connected via the internal AC bus bar. The ACin has two relays for grid protection.
If the MP2 inverter tried to push 12 kW to the AC bus bar in any case the energy was available on ACout1 as there is no relay.
If critical loads could be sourced from grid after switching the transfer switch the RCD in front of the loads can be taken out, leaving the 2p automaat at ACout1.
Maybe one of those automaats had an issue with current from the lighting.
I have:
Grid → NH00 → SLS 3p+N 35A → Dehn lightning protect → Breaker 3p+N 50A → ACin → ACout → Breaker 3p+N → transfer switch → RCBOs for each circuit
Yes AC in and AC out are linked normal operation but if the neutral of the transfer switch is linked. As @Erik2 said it was still sending power to the grid when it was down, so my assumption in a grid down the multiplus would fail to operate and the internal relays as for all intense and purpose its been by passed, this would also cause an imbalance between the neutral and the phases causing the RCD to constantly trip at least until the grid came back. But sounds like you are use to the kit that @Erik2 is talking about but I’ll stick to my theory till @Erik2 does some testing
If you’re asking about the grid meter, this is handled by the three Multiplus units themselves. Here’s a representation of my installation. SMA invert comes on when sun start shinning.
Yes wasn’t aware they could do it accurately enough as a 3 phase setup, have you tested the system since ie by using test function of the rcd instead of the switch to simulate the same issue that happened with the lightning?
Have you done an RCD test with a meter to make sure they trigger as manufactured and have not got over sensitive? Faulty
Whilst 3phase isn’t as common in the UK the neutral from AC out to ac in is messing with my I could be down a completely different rabbit hole as I assume 0 is disconnected neutral and net and the other might be linked? As I can’t see any internal diagrams for this switch but please someone that has these switches and knows the operation chime in.
