Impactos da sujidade na geração e temperatura de uma planta fotovoltaica em área urbana do semiárido brasileiro

Abstract

The energy transition requires the advancement of low-emission technologies, such as electricity generation from photovoltaic (PV) systems, whose performance is influenced by environmental, climatic, and anthropogenic factors, such as soiling. In this context, this study experimentally investigates the impacts of soiling on the electrical performance and thermal behavior of a PV plant, named LEA2, installed in an urban area of the Brazilian semiarid region, considering different rainfall periods and manual cleaning intervals. Initially, the climatic characterization of the site is presented, followed by an assessment showing that the analytical models Fixed Fill Factor Model (FFFM), Variable Fill Factor Model (VFFM), and Approximate Maximum Power Point Model (AMPPM) can be combined to estimate soiling losses using a limited set of input data. The generation analysis reveals a weekly relative percentage difference (%ELEA2) of up to 2.59% when construction work occurs on an avenue near LEA2, highlighting the relevance of anthropogenic impacts. During the rainy and post-rainy periods of 2024, the increasing trend of %ELEA2 indicates that natural cleaning, although reducing soiling, is not sufficient to fully restore the performance of urban PV systems. In the dry period, %ELEA2 reaches a peak of 5.25%. Performance metrics show that the combination of manual cleaning and regular precipitation enhances the performance of clean modules. Regarding thermal behavior, clean modules exhibit slightly higher average temperatures than dirty ones, with a statistically significant difference, though without substantial impact on LEA2’s operation. Additionally, as the interval without cleaning increases, the thermal gradient between the center and inferior edge tends to decrease, while the gradient between the superior and inferior edges shows a nonlinear behavior, suggesting distinct phases of soiling deposition. Thus, the innovative nature of this thesis is evident, which resides in the investigation of the seasonal effects of manual cleaning, with different frequencies, on the generation and operating temperature of a PV plant installed in an urban area in the semi-arid region, including the impacts of short-term anthropogenic activities; additionally, the applicability of analytical models for estimating electrical parameters to calculate losses due to dirt in the analyzed PV plant is verified.