Detecting bad cell connections in parallel battery configurations using voltage delta measurements

TEXT | Johan Dams
Permalink http://urn.fi/URN:NBN:fi-fe2024092574789
Loistava patteri tummalla taustalla.

With the increased availability of cost-effective lithium battery cells (Salinas, 2021), we see more and more people building their own battery packs for both stationary applications such as full  house back-up systems (Wu & Lindman, 2022) as well a mobile applications in recreational vehicles and marine (Reeves, 2018). One issue that arises when building one’s own batteries, and indeed one found in the industry in general, are the cell interconnections (Offer & et.al., 2012). These interconnections in the DIY sphere are accomplished through bus bars attached to cell terminals using a nut on a threaded stud welded to the terminal. Commercial offerings usually have the bus bars themselves laser welded to the cell terminals.

The stud with nut solution is effective and easy since it does not require specialized tools and offers the builder flexibility in configuration (both when it comes to voltage as well as form factor). However, it does introduce a potential point of failure, since the nut can become loose over time (Gomes, 2024) – especially in mobile applications. When it does, the connection with the terminal becomes sub-optimal. This bad connection can cause issues such as heat generation (both at the terminal and inside the battery) and imbalance in the battery pack (Madani & et. al., 2018). In space constrained applications especially, the heat generated can lead to fire with catastrophic results. In a typical ‘48V’ battery made up of 16 x 3.2V Lithium Iron Phosphate cells, we find 32 such connection points that could potentially lead to issues.

A popular method for expanding battery storage capacity is to parallel multiple batteries with each other (Diao & et. al., 2019) at a central bus bar. This method allows for flexibility (expand as you go) and increased redundancy (a single pack can be taken offline for maintenance without bringing power down altogether). By experimentally introducing a bad cell connection in one of the battery packs in this set-up, we observed an ‘oscillating’ behaviour in the battery cell voltage delta data log during normal operation of charging and discharging. After fixing the connection, the oscillation disappeared again as illustrated in Figure 1.

Cell Delta Oscillations with bad cell connection.
Figure 1: Cell Delta Oscillations with bad cell connection. Red arrow indicates point where connection was fixed.

These oscillations are in the order of single millivolts or less, but can be recorded with a typical battery management system. As far as we know this behaviour has not been described before, but could be used to potentially mitigate problematic issues before they present themselves. The behaviour was not observed in a single battery pack. More research is needed.

References
  • Salinas, F. (2021). Second life assessment of Lithium-ion cells obtained from portable electronics. Technische Universitaet Berlin (Germany).

  • Wu, A., & Lindman, R. (2022). Current and Future State of the European Li-ion Battery Recycling Market.

  • Reeves, B. (2018). Solar power DIY handbook. Revisa Publishing LLC.

  • Offer, G. J., Yufit, V., Howey, D. A., Wu, B., & Brandon, N. P. (2012). Module design and fault diagnosis in electric vehicle batteries. Journal of Power Sources, 206, 383-392.

  • Gomes, V. B. (2024). Development of a novel joining process for prismatic cell-to-busbar connections in electric vehicle batteries.

  • Madani, S. S., Schaltz, E., & Kær, S. K. (2018). Effect of Bad Connection on Surface Temperature of Lithium-Ion Batteries by Using Infrared Thermography. ECS Transactions, 87(1), 39.

  • Diao, W., Pecht, M., & Liu, T. (2019). Management of imbalances in parallel-connected lithium-ion battery packs. Journal of Energy Storage, 24, 100781.