Good day

Let me start off by saying I have a workaround for this issue, but surely it must be a glitch? Not that I am at all qualified in the field of PV, so my comments and deductions might be silly to the experts on this forum.

In my limited technical understanding the setup of my solar system is as follows. 

20x455w panels

2x5kw Deye inverters

5x5kw Greenrich batteries

Setup to feed house 100% at all times - so nothing in the house excluded from getting electricity off the inverters. Only when battery and PV can't supply demand if would use Eskom

My system is big enough to take me off grid about 85% of the time, which means my batteries are fully charged through the PV about 85% of the time. I like to run my system in a way to allow me absolutely minimum Escom dependancy, which obviously also equals very low Escom cost.

Typically I will set all of my daytime zones to run down to 20% if need be, and only then start to use Escom if PV is not enough to handle the load at the time.

So what I have noticed a few times now, and the first time it took me a good week to get the workaround going, is that if the batteries do not charge to 100% for more than 3 days in a row, then it starts to "ignore" my 20% setting. This typically happens when we have the type of weather currently in Gauteng where you have a number of days in a row where it rains and/or there is thick cloud cover.

So what would happen is it would start to add 1%-2% a day (when it does not charge to 100%) to the 20% minimum and then starts to use Escom power. Just to be clear, it does not change the actual setting on the device, it just "dynamically" start to do its own thing and creates its own minimum.

For example the first time this happened I had a week or so where I could not get enough sun to charge to a 100%
SOC minimum set to 20%
From about day 3 onwards this happens (I am fairly sure this does not happen in the first two days, but could be wrong)
Day Setting Escom usage starts

2 20% 22%

3 20% 23.5%

4 20% 26%

And so it increased daily up to around 31%, even though the minimum was set at 20%

And then I got a full sunshine day, the system went back up to 100% and from there on the issue was resolved again. This has happened 3 times now in the last year or so, and there are two ways of "fixing" it, either you charge the batteries to 100% via Escom, or you let the sun take it back to 100% - both takes the issue away. So it seems as if the "self regulating" minimum appears after around 3 days of no full charge, and disappears as soon as the batteries have had a full charge. But as soon as we have not had full sunshine (like for the least couple of days) it does its own thing again.

Maybe the system should work like this? Not sure if anybody else has had the same experience.

The BMS of the batteries can instruct the inverter to force-charge if there's an alternative power supply from which to charge. Not all batteries implement this though.

You'll see this indicated on the LiBMS info screen which will include something like "Request Full Charge" when this is occurring.

The BMS may be doing this to protect the battery since LFP batteries for example need to hit 100% on a regular basis for cell balancing/SOC measurement purposes (my own AM5's risk losing their warranty this doesn't happen once a week at least). This would also explain why getting to 100% 'fixes' it for you.

Here's an example of what it looks like in the LiBMS screen, note the 'Request Force Charge' indicator on the 4th image:

https://powerforum.co.za/topic/23770-batteries-wont-go-under-20-soc-sunsynk-12kw-3ph-hubble-am2s/

Edited March 5, 20251 yr by JayMardern

Thanks for the feedback. I do not see that specific indication on mine (picture 4), but either way the explanation you give make sense thanks.