Support and Information Effect Modeling for Recoverable Reserve Estimation of a Beach Sand Deposit in India

Application of geostatistics in estimating recoverable reserves of beach sand deposit is rare. This paper made an attempt to estimate local recoverable reserves using disjunctive kriging and discrete Gaussian model considering support and information effects for a beach sand deposit located in the eastern part of India. The dependence of different selective mining unit (SMU) sizes and different production sampling strategies on the estimated tonnage, metal quantity, and the ore tonnage versus metal quantity relationships has been examined. The results of the study show that nonlinear geostatistics should be used for more precise assessment of the grade, ore tonnage, and metal quantity and their relationships, which are necessary for recoverable reserve estimation. In selective mining operation, both support and information effects have significant influence on recoverable reserve. Recoverable reserve estimation based on SMU involves estimating grade distributions of mining unit with much bigger support than the available drill core sample data. Information effect comes into picture from the real scenario where the actual grades of the blocks remain unknown even during mining. At the mining stage, discrimination of ore and waste blocks is carried out based on estimated grades of the production samples and it is likely that the blocks might be misclassified as either ore or waste and thus sent to wrong destination. Information effect modeling makes the estimation more reliable by taking care of misclassification.