IWRA Proceedings

AbstractKarst aquifers represent important water resources, supplying an estimated 25% of the world population with drinking water. Flow in karst aquifers is often focussed in the highly permeable conduit system that comprises only a small percentage of the total aquifer volume. The storage is mainly provided by the low permeability porous fissured rock. Due to the permeability contrast between both systems the hydraulic characterisation of karst aquifers still poses a challenge to hydrogeologists. Typical methods used in karst hydrology are tracer tests for the estimation of transport parameters and conduit volumes as well as small-scale pumping tests and slug tests for the estimation of hydraulic parameters of the porous fissured rock. Spring responses provide information about the entire karst aquifer, influenced however by the respective recharge mechanism that has to be known for hydraulic characterisation. This work aims at the characterisation of karst aquifers with a large-scale pumping test abstracting groundwater directly from the karst conduit system. High pumping rates to several hundred litres per second allow a scale-continuous characterisation of the hydraulic properties of karst aquifers depending on duration of pumping. This contribution will be focussed on modelling studies conducted with a view towards field investigations in two karst groundwater systems located near Montpellier (South France). A hybrid numerical model is employed for the simulation of turbulent flow in a discrete pipe network, hydraulically coupled to a continuum. In order to test the influence of different conduit geometries, parameter studies are performed for a single conduit as well as dendritic and network conduit systems. Parameters for the porous fissured rock as well as for the conduit system are derived from field studies. Well drawdowns are simulated in the pumping and observation wells distributed over the model area. The model results help to identify typical drawdown patterns for the different scenarios, thus facilitating the interpretation of field data and to suggest some predictive water resources management scenarios.