HR: 1340h
AN: H33F-0527 [Abstracts]
TI: A Sample-based Stochastic Approach For Modeling Of Fluid Flow Through Heterogeneous Unsaturated
Fractured Rock
AU: * Zhang, K
EM: kzhang@lbl.gov
AF: Earth Sciences Division, Lawrence Berkeley National Laboratory, 1 Cyclotron Rd, Berkeley, CA 94720
United States
AU: Wu, Y
EM: yswu@lbl.gov
AF: Earth Sciences Division, Lawrence Berkeley National Laboratory, 1 Cyclotron Rd, Berkeley, CA 94720
United States
AU: Pan, L
EM: Lpan@lbl.gov
AF: Earth Sciences Division, Lawrence Berkeley National Laboratory, 1 Cyclotron Rd, Berkeley, CA 94720
United States
AB:
Modeling fluid flow and chemical transport processes in large-scale, three-dimensional, fractured reservoirs is conceptually
difficult and computationally demanding. One key issue in the modeling studies of real field problems is how to represent
heterogeneous rock properties of fractured media in numerical models. For most studies, large-scale spatial and temporal
averaging is employed to represent a heterogeneous fracture and matrix system. An alternative approach is to prescribe the
heterogeneous system stochastically, using measurements and calibration data. However, the stochastic approach, compared with
the traditional deterministic approach, is computationally more demanding. In this study, we present a modeling approach to
examine the effects of rock-property heterogeneity on flow within the unsaturated zone of Yucca Mountain, Nevada. The
heterogeneity of each of the geological layers within the unsaturated zone system is represented using a sample-based
stochastic distribution scheme. This scheme determines the rock properties of each gridblock based on the statistical
information of field-measurement data for the corresponding geological unit. The rock properties are chosen randomly from one
of the measured data groups. The frequency for each rock property is conditioned by statistical information from
field-measurement data. Simulation results are compared to the traditional stochastic approach, which employs a spherical
semivariogram model with empirical log permeability semivariograms to generate a 3-D spatially distributed rock-permeability
field. Geostatistical parameters and cumulative distribution functions of rock permeability for the spherical semivariogram
model are derived using the measured data.
DE: 1829 Groundwater hydrology
DE: 1875 Unsaturated zone
SC: Hydrology [H]
MN: 2004 AGU Fall Meeting