Analysis and synthesis of an innovative, temperature dependent run time model for the performance estimation of batteries for eletric traction applications

The research activity of this Ph.D. Thesis is about study and definition of innovative methodologies and technologies for the estimation State of Charge and State of Health of energy storage systems. The specific target of the activities was to find proper technological and methodological solutions for efficiency and optimization of energy storage systems for what concerns storage capacity, performances in time, number of charge and discharge cycles and prediction and extension of useful life. The solution found and validated in this Ph.D. Thesis, is based on an innovative, dynamic, circuital model which allows to accurately represent the functioning of a battery (in charge and discharge) in function of its state of charge (SoC), state of health (SoH) and temperature (T). This model let to determine the battery internal state and its evolution, if the parameters of the model are known. Model’s characterization was made through tests in laboratory on three different sets of LiFePO4 batteries at the ambient temperature setpoints of 0°C, 25°C and 50°C. Once the tests were completed, the data obtained were analyzed in the frequency domain to obtain the parameters of the capacitance - resistance model of the battery. So, with non linear interpolation techniques between known points, it is found a relation for each parameter, that highlights its dependence on: temperature, state of charge, state of health and number of cells. The model was finally validated by performing tests on new batteries in different ambient temperature conditions and comparing the data provided by the model with the data obtained from the measurements. Finally, an ideal room temperature range was identified for the type of batteries under examination, in order to improve their performance over time in terms of energy efficiency and extension of useful life.