Analyses of slug tests and hydraulic conductivity variations in the near field of a two-well tracer experiment site

Three-hundred eight slug tests were conducted in a 5×5 m area in a coastal, sandy aquifer at the Georgetown site in South Carolina to characterize three-dimensional aquifer heterogeneity. Methods developed by Hvorslev, Bouwer and Rice, and Cooper et al. were employed to estimate hydraulic conductivity values from the slug test data. These three methods produced similar spatial distributions of the hydraulic conductivity but quite different values. Overall, the method of Cooper et al. produces higher conductivity values in high permeability zones but lower values in low permeability areas than the Hvorslev method. Variances of the natural log of conductivity values derived from Hvorslev's and Bouwer and Rice's methods agree with those in the other aquifers under similar depositional environments. However, the variance calculated for the data based on the method of Cooper et al. appears unreasonably large. Despite these differences, histograms of the three sets of conductivity values exhibit bimodal distributions, reflecting stratification of the aquifer. Geostatistical analyses show that correlation lengths and statistical anisotropy of the hydraulic conductivity spatial structure varies with depth.