A multi-indicator predictive maintenance framework for industrial equipment under Sahelian climatic and technical constraints: Case of diesel generator sets in a diesel thermal power plant

Industrial equipment used in agri-food processing, mining, power generation, and agricultural mechanization across Sahelian countries is exposed to severe environmental stressors. In Burkina Faso, the commissioning of new diesel thermal power plants highlights the growing importance of diesel generator sets in ensuring energy supply. However, their operation remains strongly constrained by harsh environmental conditions. High ambient temperatures, heavy dust loads, and strong hygrometric variability - further aggravated by climate change - accelerate equipment degradation, impair performance, and reduce the effectiveness of conventional maintenance strategies.
This study proposes a predictive maintenance framework tailored to the Sahelian context, based on the exploitation of real operational data from the National Electricity Company of Burkina Faso (SONABEL). Statistical methods combined with advanced algorithmic approaches were used to identify failure patterns and develop novel predictive indicators. In particular, a Thermal Stress Index (TSI) was formulated by integrating ambient temperature, machine-room temperature, and applied load. The results indicate that ambient temperature is among the most influential factors associated with the occurrence of failures.
To the best of our knowledge, this study is among the first to investigate the operational reliability of diesel generator sets under Sahelian operating conditions. The findings provide practical insights for implementing predictive maintenance strategies based on composite indicators, thereby strengthening the resilience and long-term sustainability of industrial equipment in tropical developing countries.