Prediction of coal grindability based on petrography, proximate and ultimate analysis using multiple regression and artificial neural network models

Geology of North American Coalbed Methane Reservoirs

“…The strength and brittleness of coal is dependent on rank and coal composition, and the Hardgrove grindability index a commonly employed proxy for these properties (Hower and Wild, 1988;Chelgani et al, 2008). In general, the grindability index of coal is greatest in medium and low volatile bituminous rank (Esterle, 2008) (Figure 3.17).…”

Estimation of gross calorific value based on coal analysis using regression and artificial neural networks

“…(4)-(6). Although neural networks have been successfully applied by many coal studies (Yao, 2005;Bagherieh et al, 2008;Jorjani et al, 2007;Acharya et al, 2006, Chehreh Chelgani et al, 2008, because of uncomplicated relationship between proximate and ultimate analysis parameters and coal GCV, it can be concluded that ANN is not better or much different from regression when the proximate and ultimate analysis are predictors. 6)) can predict the GCV with correlation coefficient, minimum error, maximum error and deviations from experimentally calculated GCVs of 0.995, −5.67 (MJ/kg), 2.85 (MJ/kg) and 21.34%, respectively.…”

Notes on the relationship between microlithotype composition and Hardgrove grindability index for rank suites of Eastern Kentucky (Central Appalachian) coals

“…CAER's extensive HGI + petrology + other coal quality data set has proven to be of interest in the application of neural network techniques to coal characterization data Chelgani et al, 2008;Jorjani et al, 2008;Modarres et al, 2009).…”