Improved model and simulation tool for dynamic SOH estimation and life prediction of batteries used in PV systems

Abstract

Accurate estimation of the battery state of health (SOH) is necessary for effective monitoring and prediction of battery performances and useful life in PV systems. This paper proposes an improvement of a previous SOH model to reflect the battery aging state by monitoring the working zone and temperature. It presents an important contribution for simple, easy-to-implement, dynamic and non-destructive estimation of battery SOH. Unlike the former model, the proposed model overcomes the problems associated with the inaccurate SOH estimations found in the aged state of the battery. This is achieved by completely modifying the SOH equation and including several important battery characteristics not considered in the traditional model – mainly by introducing the battery design life parameter in the calculation of the safe working zone factor. These enhancements allow the user to adjust the new model to different solar battery types simply from the datasheet information. Using real measurements, model parameters were computed to reduce the estimation error of the battery SOH. To verify the accuracy of the proposed model, an experimental rig that comprises a solar battery tested for a long time is set up. The promising part of the method is the improvement shown in the SOH estimation results. The mean error, in the aged state of the battery, is reduced from 90% to 5%. Furthermore, the proposed model is integrated into a PV designer software for the prediction of battery lifetime and SOH degradation. Using meteorological data, the software tool shows the ability to predict the endurance of the solar batteries in a designed PV system for any location. It is expected that this work will benefit a large number of BMS designers and simulator developers who require a Battery SOH estimation method with a good compromise between simplicity and accuracy.