Comprehensive glacial sediment characterization and correlation with natural gamma log response to identify hydrostratigraphic units in a rotosonic well core

Common borehole geophysical methods use electrical, acoustic, and nuclear measurements to determine physical and chemical characteristics of soils and rocks encountered while drilling. In texturally and mineralogically mature detrital sediments, natural gamma logs are frequently used to differentiate coarse and fine textures that exert primary control on fluid flow in reservoir and aquifer units. Interpretation of natural gamma log response in immature sediments can be complicated by the presence of clays in the fine fraction of poorly sorted units and further confounded in heterogeneous glacial sediments which characteristically display a high degree of vertical and lateral lithologic variability. This study examined natural gamma log response to textural and mineralogical variability in a continuous 70m (230ft) rotosonic borehole drilled through Pleistocene glacial sediments west of Ann Arbor, Michigan, USA. The objective of the investigation is to establish a statistical linkage between natural gamma radiation and glacial sediment texture that can be extrapolated to a set of more than 120 monitoring wells and used as soft conditioning information for stochastic simulations of aquifer characteristics across a 10 km2 area impacted by groundwater contamination. The core was sampled at 0.3m (1ft) intervals and sieved to quantify sediment texture at 225 points. 150 grain mount thin sections from samples selected to correspond to a distribution of low, moderate, and high natural gamma log readings.were point counted to quantify mineralogy of the sand size fraction. 30 samples of clay-sized particles taken from the decanted fine fraction of the sieved material were analyzed by X-ray diffraction for clay component analysis. Weak correlations of gamma response to measured characteristics were observed over the entire dataset. As expected, positive relationships with gamma response were observed in samples with higher K-feldspar content and greater proportions of fine particles. Conversely, negative correlations were observed for samples with higher quartz content and greater proportions of coarse particle sizes. Separation of data corresponding to higher confidence gamma readings (i.e., excluding samples corresponding to gamma log inflection points) improved correlations, especially for K-feldspar content and in samples bearing a greater proportion of decanted fine materials. X-ray diffraction results indicate a relationship between gamma response and chlorite and illite, the dominant clays detected in the decanted fine clay fraction. Comparison to sorting metrics (i.e., inclusive graphic standard deviation, uniformity index, and effective size) confirms that higher gamma responses are associated with more poorly sorted sediments. Results demonstrate that gamma ray log response in these glacial sediments is controlled by a combination of texture and mineralogy. Consequently, a stochastic simulation will be required to accommodate uncertainty when linking texture to gamma response.