Evaluating the effects of data density and upscaling of hydraulic conductivity in the middle and lower Wilcox, South Texas

Hydraulic conductivity is a fundamental property in hydrogeology. Its complexity can make it difficult to accurately model. This study evaluates the effects of data density and grid size on upscaled hydraulic conductivity values assigned to a modeling grid. The middle and lower Paleogene-Eocene age Wilcox Group was mapped through the study area in LaSalle County, Texas using electric log data from 157 wells. These log data were then used to calculate permeability. Vertical upscaling for each well yielded a local estimate of permeability. The geometric mean permeability for each well ranged from 0.0016 to 0.3080 darcy for the middle Wilcox and 0.0045 to 0.2784 darcy for the lower Wilcox. Kriging was then used to upscale the values to different grid sizes. Subsets of 10 and 50 percent of the data were chosen randomly and then permeability was upscaled to 640 acre grids. Estimation error for grid centers was then calculated using all 157 data points. Permeability error using 10 percent of the data appears to have a wider range than that of the 50 percent subset. Permeability errors for grids of different sizes were also compared. Smaller grids appear to more accurately represent the upscaled parameter than larger grids. The representative accuracy of the subset grid is dependent on the subsets' numerical and spatial distribution.