The high-speed mode of an autonomous underwater vehicle (AUV) with a wide speed range determines that its shaftless propulsion system must have overload capability. In order to efficiently and accurately estimate the overload capability of the system, this paper builds the lumped parameter thermal network (LPTN) model created by the thermal properties and dimensions of materials. With the powertrain simulation of the system, the losses of the system are calculated and used as input to the LPTN model. Furthermore, the system is not only limited by temperature alone but may also be limited by power output as well as battery capacity. Finally, the LPTN model is used to predict the runtime of the shaftless propulsion system at different overload powers.