Battery overcharging due to PVKeeper V2.0 Glitch?

[JLBinTN]

I routinely keep both PVKeeper V2.0 and BMS_Test_V1.0 running at the same time. Today I noticed the charged capacity as reported by BMS_Test_V1.0 climb up over the usual 91% cutoff mark all the way to 96% (and continuing to climb) while PVKeeper V2.0 appeared to be "stuck" at 90%. Reported battery voltages were also not in agreement. I exited and restarted the BMS_Test software with no change, so I then exited and restarted PVKeeper V2.0. Upon restart of PVKeeper V2.0 the solar battery charger immediately ceased charging the battery. Both software packages had been up and running continuously for about a week. My system includes a Growatt SPF 3000 TL LVM-48P all-in-one inverter and a LiFePower4 48V 100Ah battery connected to 4200 watts of solar panels.

Anybody else ever see this happen? What would have happened had I not been present to see what was happening? I'm in an offgrid situation, and often leave the system unattended, so this is a bit disturbing. Am I worrying over nothing? TIA

PS: The Growatt BMS is directly connected to the LiFePower4 battery via RJ cable. Using the Li setting for 05. The Growatt is connected to the laptop using the provided USB cable. The battery is connected to the same laptop using the provided modified RJ/USB cable.

 

[JLBinTN]

Another anomaly spotted today. As (apparently) designed, once the LiFePower4 battery reached a State of Charge of 91% (as measured by the Growatt BMS), the Growatt stopped pulling from the solar panels and began powering loads using stored power from the battery. This is normal, but usually, once the battery drops a percent or so below 91% SOC, the system switches back to using solar power again.

When the above described scenario happened today, The battery SOC, as reported by PVKeeper V2.0, remained frozen at 91% even though BMS_Test_V1.0 showed the battery SOC as dropping below 90%. SOC (BMS_Test_V1.0) was at 77% when I noticed the discrepancy. I pushed one of the buttons on the Growatt in order to get the display panel to light up, and suddenly PVKeeper V2.0 began displaying the same SOC as BMS_Test_V1.0, and solar resumed powering loads and charging the battery.

Inspection of PVKeeper V2.0 logs indicates that solar was off for about 15 minutes, which cost me roughly 1 kWh of production.