Efficiency Study of a Water-Only Cyclone by Experimental and Data Modeling Techniques when Cleaning Indian Coal Fines

Abstract

Experiments have been conducted on a water-only cyclone in conjunction with washability studies for Patherdih and Munidih coal samples of Eastern India to characterize the cyclone efficiency. Data driven semi-empirical performance models have been developed using in-house experimental data. Washability studies on the Patherdih sample have been undertaken to determine the specific gravity of separation for a targeted coal quality. Size classification analysis has been conducted for the Munidih sample to estimate the cut-size. The Mayer and partition curves have been generated from float-and-sink analysis. Subsequently, reduced efficiency curves have been constructed using specific gravity as well as size classification data. The Rosin-Rammler and logistic distribution functions have been employed to model the data to generate the reduced efficiency curve, which characterizes the classification efficiency. Model parameters have been estimated for these distribution functions. Parametric sensitivity analysis was carried out by changing the operating parameters, namely, apex diameter, and feed inlet pressure and percentage solids in the feed to study the classification behavior. Model predictions were found to be in good agreement with the published literature. Efficiency mapping by the Rosin-Rammler distribution was found to be well suited for the Patherdih coal and both the Rosin-Rammler and logistic distributions are equally appropriate for Munidih coal.