Abstract:Photovoltaic (PV) module soiling, i.e., the accumulation of dust on PV module surfaces, poses several challenges to PV system performance. Among these challenges, the nonuniform deposition of soiling across the module surface has received scarce attention. Soiling is directly associated with an overall performance loss, but can also potentially give rise to localised hotspots that can lead to long-term PV module failure. Therefore, addressing the issues arising from this nonuniformity is not only important for optimising energy production, but also for enhancing system reliability, and ensuring the long-term operation of relevant power generation systems. In this study, the impact of nonuniform soiling on PV performance is investigated experimentally by examining soil deposition on the upper surfaces of low-iron glass samples. Samples positioned at four different tilt angles were collected on a monthly basis over a one-year study period. Since the horizontal samples were found to represent the worst-case conditions, the most soiled sample at horizontal tilt was divided into four zones, each housing a single monocrystalline solar cell and examined further. The findings reveal that the soiled sample experiences an average transmittance deterioration of 13% relative to a clean sample, and a maximum (relative) spatial variation of 4% between the four zones. These optical losses affect the amount of sunlight received by the cells, resulting in a power deterioration of ∼6–7% per 5% drop in transmittance. The soiled sample experienced an average temperature rise of 2 °C, and an average power output (and efficiency) reduction of 30% relative to the clean sample, and a maximum (relative) spatial variation of 7% between the zones. The 30% average power loss measured in this nonuniform soiling case is more than double that which would be expected theoretically for a transmittance loss of 13% but from uniform soiling, so these results highlight the importance of addressing PV soiling for optimal PV performance, and of accounting for spatial soiling nonuniformity.

Estimating PV Soiling Loss Using Panel Images and a Feature-Based Regression Model

Deep-Learning-Based Probabilistic Estimation of Solar PV Soiling Loss

Estimation of Soiling Losses in Unlabeled PV Data

Novel and comprehensive approach for power loss estimation of soiled photovoltaic modules

PV Module Soiling Detection Using Visible Spectrum Imaging and Machine Learning

A novel image enhancement algorithm to determine the dust level on photovoltaic (PV) panels

Data-Driven Soiling Detection in PV Modules

Cell-Resolved PV Soiling Measurement Using Drone Images

Two-stage Infrared Images Photovoltaic Panel Extraction Based on Deep Semantic Segmentation.

Deep Learning-based Method for PV Panels Segmentation and Defects Detection with Infrared Images

Modeling of soiled photovoltaic modules with neural networks using particle size composition of soil

Assessing the Features of PV System's Data and the Soiling Effects on PV System's Performance Based on the Field Data

Unveiling the impact of climate diversity on solar PV soiling

Modeling of soiled PV module with neural networks and regression using particle size composition

Deeplab V3+ Based Segmentation Method for PV Panels with Aerial Orthoimages

Water photovoltaic plant contaminant identification using visible light images

Experimental investigation of nonuniform PV soiling

Correlating Photovoltaic Soiling Losses to Waveband and Single-Value Transmittance Measurements

A deep residual neural network identification method for uneven dust accumulation on photovoltaic (PV) panels

A Method for Extracting Photovoltaic Panels from High-Resolution Optical Remote Sensing Images Guided by Prior Knowledge

Electrothermal Modeling of Photovoltaic Modules for the Detection of Hot-Spots Caused by Soiling