Pumping test analysis performed in the Griftpark’s subsurface, Utrecht, the Netherlands

Abstract

In preparation of a tracer test, a pumping test was performed in a heterogenous and highly contaminated site. The porous media is contaminated with DNAPL that infiltrates the water table. The 1st aquifer of the site is sealed from the sides with bentonite walls and from below with an aquitard composed of several clay layers. However, this bounded aquifer is believed to have leakages through the aquitard where the clay layers are not continuous, notably in the central-east side of the site. During the pumping test, the hydraulic heads of the 1st aquifer, the aquitard, and the 2nd aquifer below the site as well as the 1st aquifer outside the bentonite walls were continuously monitored with divers. Manual measurements were also taken from several Multi-Level Sampling wells to create hydraulic head profiles with depth. The pumping test revealed that the connectivity between the 1st and 2nd aquifer is very weak in the central-east site of the site. The hydraulic separation of the 1st and 2nd aquifer in the central-east side is mainly due to a clay layer located between 54-56m below ground level. The leakages are therefore coming from another location in the aquitard or through a hole somewhere in the bentonite wall. An averaged hydraulic conductivity of the 1st aquifer was also estimated with Stallman’s method based on the data collected during the pumping test. It estimated a diagonal hydraulic conductivity of 12.7m/day. A 14-layer model and two 3-layer models were also created to simulate the hydraulic head reaction of the pumping test. These models were not able to accurately match the hydraulic reaction from the site recorded during the pumping test. As such, a sensitivity analysis testing the hydraulic conductivity of the 1st aquifer, the aquitard, and the bentonite walls considering its thickness were performed to increase the understanding of their impact on groundwater flows at the site.