2004 Denver Annual Meeting (November 7–10, 2004)

Paper No. 4
Presentation Time: 8:45 AM

USING HIGH RESOLUTION SEISMIC DATA TO CONSTRAIN 2D SEMI-ANALYTICAL GROUNDWATER MODELS:APPLICATION TO THE IMPACT ASSESSMENT OF A PUMPING FIELD ON THE MEDOCAN OLIGOCENE AQUIFER (GIRONDE,FRANCE)


LARROQUE, Francois J.C., EGID, BORDEAUX 3, 1 allee daguin, pessac, 33400 and DUPUY, Alain, EGID, BORDEAUX 3, 1 allee Daguin, pessac, 33607, France, larroque@egid.u-bordeaux.fr

In the South-West of France, a large part of drinking water supply is provided by groundwater pumping. The Oligocene aquifer is particularly solicited. In the study area, the confined Oligocene aquifer, formed by limestones, covers an area of 1500 km2. This area is bounded eastwards by the Atlantic shoreline and westwards by the Gironde estuary. To precise the reservoir shape, six high-resolution seismic reflection lines were made, leading to a total distance of 21 km. Sonic logs were realised to compute the velocity model. The geological interpretation pointed out an important flexural fold, induced by a deep fault, leading to the pinching out of the reservoir. In order to meet an increasing water demand, a new exploitation field is considered. It would be planned 9 wells tapping Oligocene formation and producing approximately 15,000 m3.j-1. The development is planned at a pumping rate of 150 m3.h-1, 12 hours per day for each production well. To estimate the impact of the exploitation on the existing wells, a first homogeneous approach of 2D semi-analytical model is worked based on Voronoi grid. The advantage of the Voronoi model is its flexibility and the possibility to integrate exact solutions for the drawdowns in wells instead of numerical approximations. Calibration of the model parameters is done from the interpretation of pumping tests on existing wells realised in the region. A transient simulation is realised including the 9 planned wells with the exact pumping plan (150 m3.h-1 and 12 hours per day). The total time duration is 12 months. The preliminary results show that the maximum drawdown, observed in the pumping field, reaches the value of 21.8 meters, after 12 months of exploitation. A modified semi-analytical model, taking into account geophysical data, shows that the computed drawdown in the pumping area reaches the value of 32.1 meters after 12 months of exploitation. So the overall depression induced by the pumping is stronger by 49% compared to the one generated using the previous hypothesis. It appears that this project cannot be retained, because of the high interferences induced on the existing wells. The accuracy of high resolution seismic data coupled to the efficiency and the flexibility of hydrodynamic semi-analytical model gives a performing tool for groundwater management, as demonstrated here.