Implications of the geology and geochemistry of the Maclean Five uranium deposit, Three Rivers, Texas

Petrographic analysis of the barren host rock reveals a composition including 17.3% limestone rock fragments (LRF's), 22.2% volcanic rock fragments (VRF's), and 19.4% calcite cement. The percentage of calcite cement decreases significantly away from the fault, suggesting the bulk of cementing solution was introduced through the fault zone. The preservation of details in calcite fossil detritus, the calcite cement, and the groundwater geochemistry reflect alkaline conditions during the evolution of the Oakville Sandstone. Alkaline leaching of VRF's, especially glassy fragments, led to mobilization of metallic elements such as U. Bulk rock calculations based on U concentrations of volcanics in the source area (Big Bend, Texas) show considerable amounts of uranium may have been leached from the Oakville during diagenesis, providing an internal source for mineralizing fluids.

X-ray diffraction and electron microprobe analyses indicate the ore assemblage to be predominantly uraninite and cofffinite, with subordinate sphalerite and marcasite (rare pyrite). The high molybdenum content of the ore suggests the presence of Mo minerals or mineraloids. Ore textures reflect dissolution, replacement, and precipitation in voids, and indicate drastic local fluctuations in pH. Ore-stage minerals almost invariably replace the calcite cement, proving that much of the calcite was an early or pre-ore phase.

Neutron activation analysis (NAA) data establish that elements fixed in inherent host rock minerals are pseudonormally or normally distributed, whereas elements associated with uranium are lognormally distributed. Depth-concentration plots show three vertical zones of uranium enrichment. Uranium in the core samples displays highly significant positive correlations with Mo and As.

The Maclean Five deposit differs from roll-front and tabular uranium deposits in several important respects: (1) dike-like distribution of mineralization; (2) strong positive correlation of U with associated elements; (3) unusually high concentrations of U and Mo, implying a sharp depositional gradient; (4) rarity of pyrite and abundance of marcasite. These deviations are significant in terms of U geochemistry, and warrant forumlation of a new ore-model.

The proposed "fault-trap" model of ore formation includes three phases: (1) dissolution of pre-existing mineralized bodies in response to weathering and stream erosion; (2) discharge of uraniferous, carbonate solutions to a restricted fault zone; (3) precipitation of calcite within the microenvironments of pre-existing cement and deposition of ore-stage minerals, due to sharp, local reductions in pH. Eh remains relatively constant at the fault zone due to a continual influx of reducing gases from hydrocarbon deposits at depth.

Narrow linear anomalies of U, Mo, and As in geochemical sampling media should serve as potential pathfinders for this deposit-type. Considering the numerous regional normal faults along the Gulf Coast, it seems highly probable that similar deposits exist elsewhere. Application of the fault-trap model of uranium mineralization may lead to exciting new discoveries, even in areas of extensive previous exploration.