A machine learning approach for significant utilization of high-ash Indian coals by metal chloride modification.

A continuous and dependable energy supply is essential for maintaining a nation's economic stability. Globally, coal ranks as the second largest fossil fuel resource after oil and gas, leading to the establishment of coal-fired power infrastructure. Nonetheless, the pyrolysis and "burn-out" reactions of High-ash coal impose fundamental limitations that hinder its efficient use and exacerbate environmental degradation. Coal pyrolysis processes is significantly influenced by numerous experimental factors, including the, chemical concentration, operating temperature, process time. A significant weight loss was seen for periods of up to 30 min at 510 °C; yet, the change in responsiveness reduced after this time. It was found that as an increasing the concentration of SnCl causes a remarkable burn-out increase, up to 9%, whilst at lower concentrations a consistent temperature and pyrolysis time shows a considerable decrease. At 610 and 710 °C, 9% SnCl-impregnated coal. In present investigation Artificial Neural Networks and Response Surface Methodology employed to envisage the percentage of burn-out of High-ash coal. The sensitivity analyses indicated that the pyrolysis temperature stands out as the most significant input parameter, with pyrolysis time and catalyst concentration following closely behind. The ANN and RSM techniques were employed to forecast the burn-out percentage of High-ash coal. The ANN (R = 0.9965) indicates superior predictability compared to RSM.