Stream sediment geochemical surveys for uranium

Abstract

One of the major objectives of the National Uranium Resource Evaluation (NURE) program (sponsored by the U.S. Department of Energy) is to identify areas favorable for commercial recovery of uranium. One phase of this program is Hydrogeochemical and Stream Sediment Reconnaissance (HSSR) of the contiguous forty-eight states and Alaska. Stream sediment is more universally available than ground- and surface waters and comprises the bulk of NURE samples. Orientation studies conducted by the Savannah River Laboratory indicate that several mesh sizes can offer nearly equivalent information. Sediment is normally sieved in the field to pass a 420-micrometer screen (U.S. Std. 40 mesh) and that portion of the dried sediment passing a 149-micrometer screen (U.S. Std. 100 mesh) is recovered for analysis. Sampling densities usually vary with survey objectives and types of deposits anticipated. Principal geologic features that can be portrayed at a scale of 1:250,000, such as major tectonic units, plutons, and pegmatite districts, are readily defined using a sampling density of 1 site per 5 square miles (13 km 2). More detailed studies designed to define individual deposits require greater sampling density. Analyses for elements known to be associated with uranium in a particular mineral host may be used to estimate the relative proportion of uranium in several forms. For example, uranium may be associated with thorium and cerium in monazite, and with zirconium and hafnium in zircon. Readily leachable uranium may be adsorbed or trapped in oxide coatings on mineral particles. Soluble or mobile uranium may indicate an ore source, whereas uranium in monazite or zircon is not likely to be economically attractive. Various schemes may be used to estimate the form of uranium in a sample. Simple elemental ratios are a useful first approach. Multiple ratios and subtractive formulas empirically designed to account for the presence of particular minerals are more useful. Residuals calculated from computer-derived regression equations or factor scores appear to have the greatest potential for locating uranium anomalies.