Multi-scale aquifer characterization and groundwater flow model parameterization using direct push technologies

Groundwater flow and contaminant transport models are used to support decision making regarding waste disposal options, sites contaminated by surface or subsurface sources, or to develop and test cost-effective groundwater remediation schemes. Such models are influenced by different sources of uncertainty, including those due to spatial variability in aquifer and aquitard properties including hydraulic conductivity (K). However, quantifying spatial variability in K remains challenging. Classical drilling techniques for shallow heterogeneous unconsolidated sedimentary deposits involving continuous coring are expensive and time-consuming, especially when the area of interest exceeds several tens of km².
Alternative techniques such as direct push technologies use hydraulic rams, supplemented with vehicle weight, or high-frequency hammering, to advance small-diameter tools into the subsurface. These tools are typically used for cost-effective geotechnical characterization of unconsolidated deposits; recent developments also allow for hydraulic characterization. The depth of investigation is up to ~40 m, depending on the tools used (i.e applied load) and sediment properties (friction). Up to now, only a limited number of studies document using this type of data to parameterize regional groundwater flow models. To fill this gap, this study aims at parameterizing a regional groundwater flow model using data from various types of direct push technologies.
We discuss the characterization of an area (~60 km²) near the nuclear zone of Mol/Dessel (Belgium), using various direct push technologies. Most of the measurements are concentrated in an area of 200×400 m². The data include 265 cone penetration tests (CPTs), 113 pore pressure dissipation tests (PPDTs), 17 direct push injection logs (DPIL), 6 hydraulic profiling tool (HPT) logs and 19 direct push slug tests (DPST). Resulting K values, either calculated or estimated, and the corresponding spatial variability are compared with that of borehole and outcrop studies.
The benefit of using standard CPT data for the parameterization of an aquitard at the study site has previously been shown. The approach is now applied to the aquifer units and incorporates new direct push data for the entire upper ~40 m of the hydrogeological domain. The effect of the 3D heterogeneous hydraulic conductivity field on the performance of the groundwater flow model is discussed; the value of the different direct push technologies is equally addressed.