RI 7237 Application Of Statistical Response Surface Fitting In Predicting Lignite Properties

The Bureau of Mines used data from 31 core samples to develop prediction equations for evaluating chemical and carbonization properties of six seam units in a lignite deposit approximately 2 miles wide and 8 miles long in the Wilcox Formation of central Texas. Statistical response surfaces, based on core analyses, were computed for 14 parameters indicative of lignite quality. These computed surfaces, and their mathematical equations, can be used to predict values of the selected properties of the lignite seams at any desired surface coordinate position between core hole locations. The wide range of application shown by this study indicates that the methods of statistical regression analysis may be useful to evaluate large deposits of lignite or higher rank coal.