HR: 0800h
AN: H21C-1357    [Abstracts]
TI: Experimental And Modelling Investigation Of Direct And Indirect Hydromechanical Coupling Effects In A Multi-permeability Fractured Aquifer
AU: Tsang, C
EM: CFTsang@lbl.gov
AF: Lawrence Berkeley National Laboratory, 1 Cyclotron road, Berkeley, CA 94720 United States
AU: * Cappa, F
EM: Frederic.Cappa@geoazur.unice.fr
AF: Geosciences azur, 250 albert einstein street, valbonne, Fra 06560 France
AU: Rutqvist, J
EM: JRutqvist@lbl.gov
AF: Lawrence Berkeley National Laboratory, 1 Cyclotron road, Berkeley, CA 94720 United States
AU: Guglielmi, Y
EM: yves.guglielmi@geoazur.unice.fr
AF: Geosciences azur, 250 albert einstein street, valbonne, Fra 06560 France
AU: Thoraval, A
EM: Alain.Thoravaal@ineris.fr
AF: INERIS, Parc de Saurupt, Nancy, FRA 54020 France
AB: We investigate poro-elastic effects in a shallow fractured limestone reservoir with three ranges of permeabilities associated, respectively, to vertical faults, 45›X dipping bedding-planes and the rock matrix. Previous in situ data indicate characteristics as follows: permeabilities (khfaults = 100 khbedding-planes = 10e7 khmatrix = 0.001 m/s), normal stiffnesses (knfaults = 0.4 knbedding-planes = 40 GPa/m) and 70GPa matrix Young's modulus. In this study, we analyze a 2 hour pressure drawdown experiment in which pressures and deformations were monitored on the discontinuities and in the matrix using short-base extensometers and pressuremeters and at the land surface with tiltmeters. Results were interpretated using a finite element representation of the reservoir with discontinuities represented by 0.05m thick solid elements. A sensitivity study was then performed. In the faults, a 40 kPa pressure fall coupled with a 2 ŸYm normal closure occurs in 10 minutes. It is twice the pressure fall observed in the bedding planes that lasts 120 minutes. No normal deformation of bedding planes is measured. No pressure variation is observed in the matrix where only a parallel-to-bedding planes contraction of 2.5 micrometers occurs in the first 10 minutes followed by a 0.5 micrometers extension lasting an additional 120 minutes. Land surface tilt towards the fault occurs in the first 10 minutes with a magnitude of about 4 microradians. This is numerically reproduced respectively with knfaults = 0.01 knbedding-planes and khfaults = 1000 khbedding-planes. Fault normal closure induces shear displacement along the bedding planes and also matrix deformation parallel to bedding plane, before any pressure variation in bedding planes and in the matrix occurs. All the rock volume is tilted towards the fault. Then, the pressure decrease in the bedding planes induces their closure and the matrix expansion. Sensitivity study shows that the reservoir behaviour depends on faults and bedding planes properties contrast. For khbedding-planes less than 10 khfaults, faults drainage is influenced by bedding planes that act as internal hydraulic fault boundary. Slower and lower faults pressure-deformation variations modify the reservoir behaviour. If knfaults less than 0.1 knbedding-planes reservoir deformation is increased because of the lowered global stiffness of bedding planes with the matrix volumes and because of bedding planes normal closure correlated to the pressure fall. Matrix anisotropic deformation is coupled to the discontinuities behaviour. Low matrix stiffness or high matrix permeability reduces the reservoir heterogeneous behaviour.
DE: 1822 Geomechanics
SC: Hydrology [H]
MN: Fall Meeting 2005