Stochastic PV model for power system planning applications

• Published in: IET renewable power generation 2019-12, Vol.13 (16), p.3168-3179
• Main Authors: Al-Sumaiti, Ameena Saad, Ahmed, Mohammed Hassan, Rivera, Sergio, El Moursi, Mohammed Shawky, Salama, Mohamed M.A, Alsumaiti, Tareefa
• Format: Article
• Language: English
• Subjects: 31‐bus distribution system; advanced five parameter probabilistic distribution Wakeby; grid losses; incorporating weather temperature; Monte Carlo methods; photovoltaic power systems; power grids; power system planning applications; presented uncertainty modelling approach; PV energy estimation; PV models; PV power data training; PV power generation uncertainty; PV power losses; PV power measurements; PV systems; PV uncertain power generation; renewable power generation; Research Article; stochastic PV model; Wakeby model; Wakeby‐Monte Carlo Simulation
• ISSN: 1752-1416; 1752-1424; 1752-1424
• DOI: 10.1049/iet-rpg.2019.0345

Planning photovoltaic (PV) power systems integration into the grid necessitates accurate modelling of renewable power generation. Global solar irradiance, weather temperature and PV power losses due to overheating specifically in hot regimes are major factors contributing to PV power generation uncertainty. This study targets demonstrating the effectiveness of deploying advanced five parameter probabilistic distribution ‘Wakeby’ for modelling PV uncertain power generation, measured as a function of such factors, in power system planning applications. The impact of different approaches for incorporating weather temperature on PV energy estimation is studied. Wakeby-Monte Carlo Simulation for PV power data training with an emphasis on MCS stopping criteria for such advanced distribution is presented. The model is tested and verified in 31-bus distribution system to demonstrate its effectiveness over other literature uncertainty modelling approaches when planning integration of PV systems' integration into the grid to minimise the grid losses cost. Real PV power measurements are utilised as benchmark verifying the accuracy and suitability of the presented uncertainty modelling approach. Simulation results demonstrate a small error of $4.7 in the expected annual cost of grid losses when deploying Wakeby model compared to the benchmark case and that error can vary significantly when deploying other PV models.