Resilience-Based Portfolio Optimization for Predictive Maintenance and End-of-Life Battery Replacement in Solar Power Plants Under Uncertainty

Resilience-Based Portfolio Optimization for Predictive Maintenance and End-of-Life Battery Replacement in Solar Power Plants Under Uncertainty

Azadeh.amjazi nejad1

1) Master of Electronics and Integrated Circuit Design

IOI Code : ABDE-ABHGCF
Publication : 6th International Conference on Advanced Research in Management And Humanities(huconf.com)
Abstract :
    Abstract Background and Problem: Solar power plants, as one of the main pillars of renewable energy production, face a serious challenge of managing the life of storage batteries. Batteries lose their capacity over time, and making optimal decisions about when to perform predictive maintenance or replace them is a complex issue under the influence of multiple uncertainties, including energy price fluctuations, consumption patterns, and battery degradation rates. Objective: This study aims to present a resilience-based portfolio optimization model for predictive maintenance and end-of-life battery replacement in solar power plants under uncertain conditions. Method: The present study uses a mixed quantitative-simulation approach. The proposed model is designed based on modern portfolio theory and a robust scenario-based optimization approach. The required data included performance information of fifty lithium-ion batteries in operation in three solar power plants over a twelve-month period. The Monte Carlo method with ten thousand iterations was used to simulate uncertainty scenarios. The optimization model was solved using a hybrid metaheuristic algorithm. Findings: Compared to traditional maintenance approaches, the proposed model was able to reduce operating costs by twenty-three percent, increase the effective battery life by eighteen percent, and improve the system resilience index against price shocks and production reductions by thirty-seven percent. It was also shown that the optimal strategy tends towards gradual replacement based on the health status of each battery, rather than replacing all batteries at once. Another important finding is that the optimal time window for maintenance decisions, under high uncertainty conditions, is between eight and fourteen days before the predicted performance loss. Conclusion: The presented model can be used as an efficient tool for solar power plant managers to reduce costs, increase reliability, and improve the resilience of the energy storage system.

Keywords : portfolio optimization predictive maintenance end of life batteries solar power plant uncertainty resilience