Solar-thermal enhancement and data-driven modeling of turmeric drying using a PCM-integrated solar chimney dryer for low-carbon agro-processing

Post-harvest drying of turmeric is essential to preserve its quality, reduce energy costs, and minimize drying losses. Open sun drying is one of the most commonly used traditional methods; however, it is slow, weather-dependent, and has low energy efficiency. This study investigates the feasibility of a solar chimney drying technique enhanced with phase change materials (PCM) to achieve improved process stability and higher efficiency. For comparison, three drying modes were examined: open sun drying, solar chimney drying, and solar chimney drying assisted by PCM. Among these, the PCM-assisted system exhibited the highest energy efficiency of 56.7% and a drying efficiency of approximately 59%. It required only 24 h to reduce the moisture content of turmeric from 75% to 10% (wet basis). The advantage of the PCM system lies in its ability to store and release latent heat, thereby maintaining consistent airflow within the chimney. Machine learning models, including random forest, support vector machine, and feedforward multilayer perceptron, were developed using experimental data to predict system performance. Among these, the multilayer perceptron model demonstrated the highest prediction accuracy, with an R2 value between 0.92 and 0.94 and a mean absolute percentage error below 3.2%. The findings indicate that the integration of thermal energy storage and machine learning models enhances the reliability and efficiency of solar drying systems.