Importance of Sampling in Mineral Dressing Investigations

The papers in this section provide several examples of the application of Process Mineralogy to metallurgical problems. Any experienced metallurgist could cite numerous other examples when process mineralogy has, or could have, provided data of value in his understanding of metallurgical processes. The data developed by mineralogic techniques may supplement conventional chemical analyses, but cannot be replaced by assays. Mineralogic and chemical analyses each provide different types of information. In order for the process mineralogist to interpret his results in the most useful manner, it is desirable that he should have some knowledge of metallurgical processes. Conversely, the metallurgist should have some knowledge of mineralogic techniques to understand the scope and limitations. Mineralogic examinations are time-consuming and expensive. To ensure the effort is not wasted, and indeed does not generate misleading information, the utmost care must be taken at every stage of sampling and sample preparation. The mineralogist may examine microscopically or by x-ray diffraction a one-gram sample representing a flow of possibly thousands of tons per day through a concentrator. Conditions in a concentrator change continually in response to changes in feed rate, feed characteristics, grind, reagent additions, temperature, circulating loads, etc. Lengthy sampling periods are required to ensure the samples examined are representative of average conditions. The sampling program should be drawn up jointly by the metallurgist and mineralogist. Sampling points should be selected with a view to ease of sample cutting. The sampler should be instructed in the method of cutting the samples, the weight or volume of sample to be taken at each cut, the sampling frequency, and sampling period. The sampling period required depends on the scope of the investigation. For example, if a complete survey is to be made of all final and intermediate products of a flotation concentrator, samples should be cut at not more than two-hour intervals, twenty-four hours per day, for not less than one week. For a more limited investigation of one circuit or one section of a concentrator, 'the sampling frequency could be increased to every hour and the sampling period reduced to one or two days.