thermal_theo

52°C for LFP is in the 'degradation accelerating' zone, not 'immediate damage' — you probably lost 8–12% capacity from those cells over that season rather than catastrophic failure. The Arrhenius equation doubled your degradation rate at that temperature. It adds up quietly over years and you don't notice until a capacity test.

The MOSFET failure mode you described is why I spec contactor-based architecture for anything vehicle-related above 72V. Contactors fail open (safe state), MOSFETs fail closed (unsafe). At traction voltages and currents the failure mode matters more than the cost difference.

The heat generation differential is real and the number most builders miss is the peak cell temperature, not the pack average. In a well-balanced 20kWh NMC pack at 2C discharge, peak cell temp runs 8–12°C above average. That delta is what your cooling system must handle — not the average readout on your BMS dashboard.