Effect of water washing on pyrolysis of Shaerhu coal and prediction of sodium and chloride release using neural network model

Coal consistently plays a major role in global energy consumption. With high-quality coal reserves dwindling, low-cost, abundant high-sodium coal has attracted attention despite challenges posed by Na and Cl release during thermal utilization, which can lead to equipment corrosion. This study investigated the pyrolysis of high-sodium Shaerhu coal (SEH) from Xinjiang in a fluidized bed. Results showed that water washing significantly reduces water-soluble Na and Cl, decreasing their content to 764 μg g⁻¹ and 698 μg g⁻¹, respectively, after 120 min. Compared to unwashed coal, washed coal (WS) demonstrates increased char yield and ash melting point, alongside reduced char reactivity and Na and Cl release. Characterization by Raman, XRD, and FTIR reveals changes in char's graphitization, microcrystalline structure, and functional groups with washing time. BET analysis indicates reduced adsorption and specific surface area in WS char. A neural network model developed for predicting Na and Cl release during pyrolysis achieved an error margin within 5 %. These findings provide critical data and theoretical support for optimizing the pyrolysis process of high-sodium coal.