Comparison of different cleaning methods on photovoltaic module efficiency under controlled laterite dust deposition in Kampala Uganda

Dust accumulation on photovoltaic (PV) modules is a major source of performance degradation in dusty environments, leading to reduced energy yield and increased operation and maintenance requirements. This study presents a comprehensive experimental evaluation of different PV cleaning methods under controlled and field-representative dusty conditions. A field-based experiment was conducted at a solar test site in Kampala, Uganda, using 15 identical 100 W monocrystalline silicon PV modules arranged in a triplicate design. Controlled laterite-based dust was applied at standardized densities of 10, 20, and 30 g/m 2 to simulate light, moderate, and heavy soiling conditions. Five cleaning strategies were investigated: no cleaning (control), manual brushing with water, mechanical wiping, automated robotic cleaning, and electrostatic cleaning. Module performance was monitored at 15-min intervals between 10:00 and 15:00 using calibrated irradiance, temperature, and current–voltage measurement systems. Each dust-level tests was repeated on 5 clear-sky days. Dust was characterized in terms of particle size distribution, morphology, and composition to support the interpretation of soiling behavior and cleaning effectiveness. Statistical analysis was conducted using two-way analysis of variance (ANOVA), Tukey HSD post-hoc testing, and regression modeling, complemented by uncertainty propagation and effect size evaluation. In addition to instantaneous efficiency, cumulative energy yield was calculated to assess practical performance implications. The results show that PV efficiency decreases systematically with increasing dust density, with uncleaned modules exhibiting substantial performance losses. All cleaning methods significantly improved module efficiency relative to the control ( p < 0.001). Automated robotic cleaning achieved the highest efficiency and energy yield across all dust levels, while electrostatic and mechanical cleaning demonstrated comparable intermediate performance. Manual cleaning provided moderate improvement but was limited by water and labor requirements. Regression analysis revealed a strong linear relationship between dust density and efficiency loss ( R 2 > 0.99), confirming the predictability of soiling effects. Techno-economic analysis further showed that while robotic cleaning offers the best technical performance, its economic viability depends on system scale and local cost conditions. This study provides a high-resolution, controlled, and statistically robust comparison of PV cleaning strategies in a dust-intensive urban environment. The findings offer practical and context-aware guidance for optimizing PV maintenance strategies, highlighting the importance of combining technical performance, resource efficiency, and economic considerations to enhance the reliability and sustainability of solar energy systems in dusty regions.