We’re seeing an issue with batteries in cold conditions, and wanted to check a few points please.
(3015 bat firmware on 9.5kWh battery in this example. The battery was as low as 0 deg C this morning)
1 - The only reference in the bat firmware changelogs to bat calibration being adjusted for temperature is in point 8 of the firmware log for BMS 3012. This indicates different calibration thresholds for batteries under 20degC and over 25degC, but nothing about lower temperatures.
Could I please have confirmation of whether or not there are any differences in the calibration for batteries between 0degC upto 20 deg C?
2 - Once a battery is calibrated, does the state of charge calculation then adjust for temperature changes within the battery? Or is the SOC calculation static regardless of battery temperature in the 0-20degC range?
3 - Is the battery temperature data drawn from a single thermocouple within the battery pack, or an average of multiple thermocouples across the battery pack? And does Giv Energy have access to cell level temperature data?
4 - Point 9 of vs 3014 changelog indicates that 1 cell at 3.6V or 2 cells at 3.55V triggers the SOC calibration to jump to 100%. Could you confirm what happens at this point with other cells that are not at the same voltage, particularly cells on the outside of the battery that are colder than the cells in the middle? Does the BMS then continue running a balancing charge to balance the cells and bring them all up to the same voltage, or does the SOC calibration just cut out once 1 cell hits 3.6V?
5 - Are those 3.55V and 3.6V calibration cut offs varied by cell temperature?
Cells that have dropped to close to 0 degrees C are never going to hit 3.6V / would get damaged if they did / would have to heat up considerably to achieve 3.6V.
6 - Is there any form of internal heater within the battery packs to balance the temperature of the cells in low temperature conditions?
7 - I’ve seen a statement by a GE employee on a facebook forum that the maximum charge rate is reduced with battery temperature to 0.33C rate betwen 10-20degC and 0.25C rate between 0-10degC. Could this be confirmed please, and also confirm which BMS version this was introduced in?
Ultimately I’m concerned that the voltage / SOC curve for batteries close to 0 deg C is significantly different to the curve at 20 deg C, with overall capacity around 8% lower, but also cell voltages generally being much lower, with eg 3V in the cell representing around 60% DoD at 0degC, 85% at 10degC and 90% at 20degC.
The issues I’m seeing seem to indicate to me that the Giv system isn’t adjusting sufficiently (or maybe not at all) for the differences in SoC for cell voltages at temperatures significantly lower than the 20degC calibration temperature, and probably also not for the temperature differences between cells.
Thanks, Gavin, Leeds Solar
