I built myself a lithium-ion battery and bought a MultiPlus II to charge it using surplus PV generation and then discharge it in the evening.
The MultiPlus is connected to the household grid via AC-in. AC-out is not being used, although I may add that later.
Venus OS is running on a Raspberry Pi.
The import/export readings come via Node-RED from the Tibber API.
The BMS (JK-BMS) is not integrated into Venus.
Everything initially works completely normally: the MultiPlus charges the surplus into the battery until it is full. If there is a sudden increase in load, or a cloud passes in front of the sun, it discharges the battery until there is surplus again, then switches back to charging mode.
In the evening, it discharges the battery. At some point during the night, the battery is empty and it stops charging, switching to float charge. Unfortunately, it stays in that state. When electricity starts coming from the roof again in the morning, it remains in float charge. If I switch the MultiPlus off using the switch, wait 10 seconds and then switch it back on, everything works completely normally again.
In the Venus interface, I can switch to “Charge only”, but nothing is charged.
Since this works for other people, I assume I have an incorrect setting, but I have no idea what it could be…
I assume this is worded incorrectly. The MultiPlus discharges the battery and then stops discharging at some point during the night.
This point will be when the BMS disconnects the battery because it is empty. However, the MP is still connected to the mains, so the only thing it can do is keep its output voltage constant, even without a battery connected.
Why was communication between the BMS and GX omitted? This is an important part of safe operation with LFP batteries. Alternatively, at least use the ATC/ATD contacts on the JK BMS to stop the MP, so that the BMS does not have to perform a hard shutdown. This should always be the last resort for protecting the battery, not the only option.
The MP can measure the total voltage. However, the BMS will already shut down if one of the cells is too low, even if the total voltage is still above the limit.
It would be best to check whether the BMS is actually shutting down; it’s only a theory for now.
According to the log, the lowest cell voltage was 3.5 V. The BMS would shut down at 2.85 V. I also can’t find anything in the log indicating that the BMS ever shut down.
How is the ESS configured? Is charging scheduled? Do you see any reason why the MP stops discharging overnight? What exactly is being done in NodeRed—is it just creating the Gridmeter as a virtual device?
It stops at night because the battery is simply empty, or rather drops below 50 V. Node-RED only simulates a GridMeter using the values from the Tibber API.
I think the main problem is the lack of communication. Using the SOC reported by the battery, ESS can stop discharging at a certain SOC in “Optimised” mode (with or without BatteryLife), without the system going into an undervoltage shutdown.
In principle, this should also work with the MP’s internal battery monitor as a replacement, but it isn’t particularly accurate. And of course, it needs to be configured for the battery being used.
When I open the configuration, I get an error message, although that could be down to me. The battery monitor is set to 180Ah—is that correct? No ESS Assistant is loaded, but that may be because the error message is preventing it from being displayed.
A 37.2V undervoltage shutdown is the factory setting, if I remember correctly. Are you sure the MP is only supposed to shut down at that point? That’s rather low for LFP. Or is this also an incorrect value caused by the error message?
It may be that the lack of communication is the problem, but I can hardly imagine that. I’m not the only one who doesn’t have the BMS connected to Venus… Apparently, someone from France even runs the whole system completely without a BMS. In that case, they would all have the same problem, and I wouldn’t be the first person to raise it.
An ESS Assistant is loaded; I don’t know why there is an error message. The file loads normally for me. I’ll attach a screenshot of the ESS Assistant settings.
The undervoltage shutdown at 37.2 V should actually never be triggered, because at two other points 50 V has been set as the lower limit for discharging. Or have I misunderstood that? This is a little confusing for me… According to the Venus GUI, the battery never drops below 50 V either. At 50 V, it shuts down correctly.
