Improvements Of The Determination Of Block Grades From Blast Hole Data In Porphyry Type Deposits

In most open pit operations mining porphyry type deposits, production planning rests on an are reserve inventory block file which is derived from diamond drill hole data. With such a practice, the way by which block values are generated is critical and in the recent years many studies deal with the comparison of different estimation methods for block grades (Mueller, E. (1), Knudsen et al (2) (3), Barnes, (4), Williamson and Mueller (5). Typically in such studies, one selects a bench or a group of benches where there is a "sufficient" number of blast holes (BH) with assay values in each of the blocks (generally 30.5 x 30.5 m (100 x 100 ft) or 15.2 x15.2 m (50 x50 ft). Then the arithmetic mean of the BH values in each block is computed and it is assumed to represent the real block grade against which different estimated values are compared. However it is sometimes observed that the variation of these BH values within a block is high and that their arithmetic average may significantly differ from the real average grade of the block (Barnes (4). The problem of the precision of the mean BH grade in a block has been examined by Williamson and Langlois (6) in an experimental way. They strove to find a better estimate of the average block grade than the mean BH grade but their results were not conclusive. The purpose of this study is to investigate the same problem within the frame of geostatistical models (David (7). Besides practical results obtained with these models, two general points are emphasized and illustrated in the text. First, "good results" are obtained if "good geostatistics" is done (David et al (8).