Evaluation of Hydraulic Conductivity Collected by Various Approaches at a Highly Heterogeneous Field Site

Abstract

Significant research efforts have been conducted over the last several decades to better understand the groundwater flow and subsurface contaminant transport. It has been found that building a groundwater model for remediation decision-making requires an accurate delineation of spatial variation in hydraulic conductivity (K) and specific storage (Ss). Currently, numerous methods are available for site characterization. Traditional methods such as grain size analyses, permeameter and slug tests can provide point-scale estimates of K, while large-scale estimates from pumping tests are widely used for water-supply and water-quality investigations. However, when the degree of local heterogeneity increases, the necessary number of K increases dramatically, which presents a challenge to conventional methods. As a consequence, Direct Push (DP) based methods have been developed as efficient alternatives to conventional well-based approaches to provide K variability for shallow, unconsolidated aquifers. Hydraulic Profiling Tool (HPT) is one of the novel DP approaches designed for high-resolution site characterization with a test interval of about 1.5 cm. Various site-dependent formulae can be utilized to convert data collected during the HPT surveys into K estimates over a limited range. More recently, inverse modeling approaches of varying degrees of parametrization have become one of the most promising techniques to map hydrostratigraphic spatial variations between boreholes and identify heterogeneity characteristics with a level of detail never before possible. Many comparisons of diverse approaches have been performed, but there is no consensus on which approach yields parameters that are representative for field sites. The main objective of this study is to evaluate K estimates obtained via various site characterizations methods including: (1) grain size analyses; (2) falling head permeameter tests; (3) slug tests; (4) HPT with three different formulae; (McCall and Christy, 2020; Borden et al., 2021; and Zhao and Illman, 2022b) (5) inverse modeling based on a geological zonation approach, and (6) a highly parametrized transient hydraulic tomography (THT) approach. The performance of each approach is first qualitatively analyzed by comparing it with site geology. A 19-layer geological model and forward groundwater model are employed to further assess various methods by simulating seven independent pumping tests that are not used for model calibration under both steady-state and transient-state conditions. Results reveal that the highly parametrized THT analysis with prior geological information yields the best results in model validation under both steady and transient states, and the generated K field revealed the most salient features of inter- and intra-layer heterogeneity. In contrast, traditional methods yield biased prediction of drawdowns, while HPT methods are primarily constraint by the limited range of estimates, especially for low permeable materials.