I've been ironing out the kinks at a site with an MP2 3000kva and a 150/45 MPPT, which uses the top 25% of battery at night.
The SOC reported by the multi seems to drift beneath the actual charge level by approximately 10%.
The current draw in absorption tails off to 1.2A and below which if it was attached to a shunt would be well beneath the tail current to determine 100% SOC.
Trying to understand what is causing this daily deviation.
Just to add, these are 4 x Victron Gel 220AH batteries in series.
I have an identical setup but that uses a smartshunt which behaves fine.
On this system with the multi being used as the monitor (connected via Venus GX to the MPPT) it seems the SOC just does not increase fast enough so it comes to a halt near 90% and never gets near 100%.
Feedback from the local team is that the multi's internal battery monitor is too inaccurate for an ESS/solar environment and that a shunt/monitor is needed.
10-15% per day seems really inaccurate to me but systems with shunts are working fine.
Interestingly I forced a deeper discharge overnight which then hit the limits and pushed charging to the grid when the scheduled charge started.
Bulk charging from grid got the SOC to the preset 85% and when the load moved to PV it has been reporting fine.
It seems that there is an interaction/measurement issue when the MPPT is charging and the multi does not bulk from grid. Seems to be one of the many quirks with ESS.
