Comparison of saturated hydraulic conductivity estimated by hydraulic and indirect methods in a coastal sand aquifer at Magilligan (Northern Ireland)

Hydraulic conductivity or coefficient of permeability (K) is one of the most difficult soil properties to measure with its accuracy is highly dependent on the method used. Knowing realistic values of K is essential in many civil engineering projects, in which groundwater flow can play an important role. Several approaches can be taken to estimate K, which can be grouped into two general categories, namely hydraulic and indirect methods. Presented here is a comparison of the saturated hydraulic conductivities, estimated using six different methods applied to the relatively uniform coastal sand aquifer underlying Benone Strand (Magilligan, Northern Ireland). K has been indirectly estimated from: 1) Cone Penetration Test (CPT), 2) Hydraulic Profiling Tool (HPT) system, and 3) tide aquifer interaction techniques, and directly using 1) grain size analysis, 2) pumping test data, and 3) groundwater flow modelling. Estimated K values lie within the typical ranges for sands in all the cases, but there are significant differences between the K values estimated using the six methods (K ranges between 1.2 and 87 m/d). CPT data and grain size analysis provided order of magnitude estimates of K. However, CPT data overestimated K calculated from HPT data. Different empirical equations commonly used to estimate K from grain size analysis showed variations of one order of magnitude. HPT data indicated that K decreases with depth due to an increase in soil compaction. K estimated from tide aquifer interaction techniques depend greatly on the model selected, with hydraulic conductivities estimated using the time lag model are between two to eight times larger than those estimated using the tidal efficiency model. While analysis of pumping test data offered a representative average K of the aquifer, numerical models provided an estimate of the spatial distribution of K.