Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Etats Unis → 2020 → PERFORMANCE LOSS RATE ANALYSIS OF 1100 PHOTOVOLTAIC POWER PLANTS

Case Western Reserve University (2020)

PERFORMANCE LOSS RATE ANALYSIS OF 1100 PHOTOVOLTAIC POWER PLANTS

Xin, Arthur S

Titre : PERFORMANCE LOSS RATE ANALYSIS OF 1100 PHOTOVOLTAIC POWER PLANTS

Auteur : Xin, Arthur S

Université de soutenance : Case Western Reserve University

Grade : Master of Sciences (MS) 2020

Résumé
Linear Performance Loss Rate (PLR) has been widely used in the photovoltaic (PV) community as a tool for modeling the degradation of PV modules over time. In the real-world commercial deployment of solar power, PV modules are manufactured by different companies to varying degrees of quality, deployed in a wide variety of locations, and undergo system damage and degradation. Since factors such as these will affect the performance of PV modules in different ways after data is already recorded, a linear model would not be fully able to effectively capture their lifetime performances. Many tools and models have been developed for specific lab tested PV systems, but have failed to generalize to large quantities of commercial powerplants. In this work, we introduce a new solar power plant analysis tool that uses a non-linear PLR method that better models the degradation of PV modules over their lifetime, and we will be testing our model on a population of 1100 commercial PV systems that have been operating for up to 6 years. In this dataset, the median adjusted R2 value of the Linear PLR method is 0.03, while the median adjusted R2 of a Segmented method is increased at 0.21. Decomposition by season can further increase the performance of both linear and segmented methods to median adjusted R2 values of 0.19 and 0.28, respectively. Comparing the metadata factors to the calculated PLR values can lead to insights regarding which factors contribute to the greatest change in PLR and degradation of PV systems.

Présentation

Version intégrale (5,22 Mb)

Page publiée le 26 décembre 2022