Monte Carlo simulation for real-world energy yield analysis of car park solar PV system installations in harsh environments
Abdulrhman Klifa Al-Hanoot, Hazlie Mokhlis, Saad Mekhilef, M.A. Alghoul, Muhammad Aqil, Manal Alhanut
Abstract
Maintaining a reliable power supply despite uncertainties from failures, renewable fluctuations, and load growth is vital for electricity infrastructure. Assessing reliability helps identify operational weaknesses, yet evolving system variability demands advanced analytical tools. This study presents a comprehensive performance evaluation of a 1,179 kWp grid-tied car park solar photovoltaic (PV) installation in Abqaiq, Saudi Arabia, designed to meet daytime consumption for industrial buildings under harsh desert-coastal environmental conditions. Using PVsyst simulation software, real-time monitoring data, and Monte Carlo simulation techniques, the research investigates discrepancies between modelled and actual system performance during the plant’s first operational year. The actual annual energy yield was measured at 1,506.17 MWh approximately 24.4% below the PVsyst predicted P50 value of 1,992 MWh. Key performance indicators, including Performance Ratio (PR) and Capacity Utilization Factor (CUF), were also significantly lower than expected, at 64.5% and 14.5%, respectively, compared to simulated values of 81.64% and 18.9%. Monte Carlo analysis, using 10,000 iterations, confirmed a low probability of the actual output occurring under the simulation’s standard variability assumptions, indicating severe underperformance. The study identifies real-world challenges such as excessive dust accumulation, elevated ambient temperatures, and system degradation as primary contributors to the energy shortfall. The integration of deterministic simulation with probabilistic modelling provides a robust methodology for yield prediction and risk assessment. The findings underscore the need for improved forecasting accuracy, routine performance diagnostics, and adaptive system design tailored to extreme climates. The integration of confidence interval analysis and probabilistic modelling provides a deeper understanding of yield deviations and system underperformance, offering a replicable methodology for performance validation in extreme environments. This approach supports improved planning, investment evaluation, and maintenance strategies for large-scale PV systems exposed to challenging climatic conditions.