Predicting the technological properties of uranium ores in natural deposits on the basis of a regression model

Abstract

The technological properties of ores are characterized by spatial variability, the level of which depends on the geological-struc tural and mineralogical-ge ochemical conditions of localization of the deposit. For most deposits of nonferrous, rare, trace, and radioactive elements, the variability of these properties is relatively high. The discovery and description of variability in the technological properties of ores (without which optimization of the processes of mtning, refining, and hydrometallurgic al treatment is impossible) requires a closely meshed screen of technological tests, which, because of the complexities and difficulties of obtaining and studying bulk samples, has not been attained by the stages of detailed and commercial prospecting as a rule. Therefore, in practice, the activities of mining concerns raise the problem of quantitatively predicting and mapping the technological properties of ores in natural deposits, without significantly increasing the volume of bulk, technological testing, on the basis of a complex of indirect geological-struc tural and mineralogical-geochemical factors. A description of these factors may be obtained in the course of performing the geological documentation and testing of the mine workings or well center. It should be possible to obtain extensive quantitative (or semiquantitative ) estimates of the technological properties of ores in natural deposits by constructing and investigating mathematical models relating the technological properties of the ores with their chemical and mineralogical composition, the geological-struc tural conditions of localization of the deposit, and the form of the changes in the area surrounding the ore, and then predicting, on this basis, the properties of the ore in the natural deposit according to the data of the primary geological documentation, testing, and mineralogical-te chnological mapping. Examples of various approaches to the solution of this problem are described in [1-9].