The conventional reserve/resource estimation methods (triangular, polygonal, and isopach) are briefly described and then compared against a new geostatistical approach. The comparison is based on estimates using real-world data from southern Illinois. The geostatistical procedure employed is a kriging technique called the intrinsic random functions of order k, using spherical covariance models. While the intrinsic random function approach is usually considered to be a local estimation procedure, the introduction of sampling methodology allows its extension to global estimation problems without excessive cost for computation. The major advantages of the geostatistical approach are that it provides a built-in measure of the precision of its estimates, thus allowing the dete mination of confidence limits, and also gives additional insight into the spatial variability of coal seams. The estimates are improved as more geological information of the coal deposits becomes available.

End_of_Article - Last_Page 705------------