EEC Kemapack - Wrong SoC calculation

 

Hi Niklas,

We are glad to hear that the information we sent has been useful for you.

About the remaining topics:

Thank you for all the information.

Looking forward to your feedback.

Best regards,

Adrià Salvadó Bautista System Engineer WATTIUS BATTERY SOLUTIONS, S.L. address: Camí de Can Bassa 6-8 | 08401 Granollers Barcelona, Spain mail: adria.salvado@wattius.com website: www.wattius.com

 

 

Hi Adria,

Thank you for your quick and detailed response.

In principle, the behavior seems plausible, and I agree with your assessment. Your email has certainly helped us gain a better understanding of the BMS and its current behavior. However, I do have a few follow-up questions.

Regarding point 2: Why does the SoC only recalibrate after several hours instead of after 60 seconds, as we configured?

Regarding point 5: I have attached additional screenshots showing the behavior of the battery storage after 0:45. The forced charge remained active for several hours, even after the cell voltages were well above the values for 2% user SoC. The system only calculated the SoC corresponding to the cell voltage after a BMS reset at around 09:15 (see screenshots 1 and 2). I hadn't mentioned this in my first email.

Unfortunately, we do not track the average cell voltage in Grafana. However, I have provided a screenshot of the current cell voltages from your toolkit. At the moment, no cell stands out due to high cell drift. If I understand correctly, the SoC cannot be recalibrated during a forced charge because current is constantly flowing. But why is it not calculated via Coulomb counting? Or is the 1 A current too low for this method?

We also experience issues with incorrect SoC at other times (see screenshots 3-5). In these instances, the SoC sometimes jumps by 20–30% without any corresponding current flow or change in cell voltage. Based on this behavior, we also assumed there might be an issue with the Hall sensor. We have no other system that displays a Current Counter value of -2.4 million, so this really stands out for us.

I hope these new details help to specify the problem more precisely. Thanks again for your response.

Best regards,

Niklas Kranzusch Service Engineer   niklas.kranzusch@voltfang.de | www.voltfang.de
Production Facility & Office: Voltfang GmbH Philipsstraße 8 52068 Aachen
Voltfang GmbH – Sitz: Aachen, Amtsgericht Aachen HRB 25081 | Geschäftsführer: Afshin Doostdar, David Oudsandji, Roman Alberti | USt.-ID.Nr.: DE345407279 | Steuernummer: 201/5997/5675

ACHTUNG: Diese E-Mail stammt von einem externen Absender. Bitte vermeide es, Anhänge oder externe Links zu öffnen.

Hi Niklas,

Thank you for all the information.

After analysing the issue, we understand that the sequence of events was as follows:

Could you confirm if this behaviour matches your understanding?

To contrast our interpretation, could you confirm that after the 2:30, there was a moment that the real SoC was calibrated? Furthermore, could you provide a plot of the average cell voltage?

We hope that this information helps you understanding this issue.

Looking forward to your feedback.

Best regards,

Adrià Salvadó Bautista System Engineer WATTIUS BATTERY SOLUTIONS, S.L. address: Camí de Can Bassa 6-8 | 08401 Granollers Barcelona, Spain mail: adria.salvado@wattius.com website: www.wattius.com

 

 

Hi Wattius Support,

We are currently experiencing issues with the SoC calculation in one of our systems with a BMS v3.

During discharge, the SoC is generally displayed normally, though it appears to be calculated incorrectly at times (as seen in the first section of the SoC curve around 15:00). The storage is configured to discharge to a minimum of 1% SoC (real SoC 9%), at which point power output stops.

However, around 23:00, the BMS recalculates the SoC and begins releasing power again. At this stage, the cells are in a different voltage range than the calculated SoC would suggest. The battery continues to discharge until the voltage limits (approx. 3250 mV) are reached. At this point, the SoC is recalculated and set to approximately 4% real SoC, even though the minimum SoC in the BMS is set to 8% real SoC.

Our EMS then triggers a 1 kW forced charge. While this increases the cell voltage, the BMS fails to recalculate the SoC during this period. A manual reset of the BMS is required before the SoC is recalculated correctly.

We noticed that the Hall Sensor Current Counter value is currently at -2.4 million. Could this specific value be causing the calculation errors?

How can the counter reach such an extreme value?

Are there other known reasons for this behavior (mismatch between voltage-based and Coulomb-counting SoC)?

Thanks a lot for your answer. I hope the attached screenshots will help you, if you need any further information please let me know. 

Best regards,

Niklas Kranzusch Service Engineer   niklas.kranzusch@voltfang.de | www.voltfang.de
Production Facility & Office: Voltfang GmbH Philipsstraße 8 52068 Aachen
Voltfang GmbH – Sitz: Aachen, Amtsgericht Aachen HRB 25081 | Geschäftsführer: Afshin Doostdar, David Oudsandji, Roman Alberti | USt.-ID.Nr.: DE345407279 | Steuernummer: 201/5997/5675