Accurate prediction of battery energy storage system state of health is very important in renewable energy systems. This paper presents a methodology for state of health estimation of lead acid battery bank by parametric identification. A particle swarm optimization algorithm is used for parameter fitting of a real battery bank. A periodic perturbation is introduced in the population to prevent the algorithm from falling into local minimums. The perturbation will consist of a new population PSjk based on the best global solution allowing the reactivation of the PSO algorithm. The proposed method is validated using experimental data that is obtained from a renewable energy system located at Chocó-Colombia. The capacity, state of health, and internal resistance of the battery bank is estimated and the evolution of the parameters associated with the battery capacity are shown. The voltage and state of charge are estimated with high accuracy confirming the effectiveness and robustness of our method. The results show that the battery bank lost 5% of its nominal capacity, locating in a 95% his state of health. Moreover, it is observed that as of February, the battery bank current presents a significant increase that can lead to a deterioration and premature substitution of the battery energy storage system.