![]() |
ИСТИНА |
Войти в систему Регистрация |
ИПМех РАН |
||
Radioactive waste has been injected in deep artesian aquifers of Cretaceous terrigenous deposits in Western Siberia since 1963.It is well known that for unconsolidated formations geologic heterogeneity strongly affects contaminant transport. The goal of this work is estimation of effective hydraulic and macrodispersion parameters on the basis of 3D high spatially resolution lithological heterogeneity model of the injection site. The data set includes detailed geologic core description from 295 wells. A 3D Markov chain model in combination with a conditional stochastic simulations were used to simulate 3D four hydrofacies distributions at a resolution of 25 m x 25 m x 0.5 m for 18.5 km2study area. Hydraulic conductivity is assigned to each cell corresponding to the simulated hydrofacies. For the modeled domain the numerical estimation of effective hydraulic and dispersion properties was performed. A three-dimensional steady state modeling approach was used to simulate groundwater flow with MODFLOW 2005. 3D advective solute transport is simulated with a particle tracking technique using PMPATH. The apparent average velocity, the effective longitudinal macrodispersion and the effective longitudinal macrodispersivity from temporal moments of observed breakthrough curves for conservative tracer were calculated. The results of this study show that for the studied medium: (1) the horizontal mean lengths exceeds the vertical lengths more than 30 times and as the result the vertical effective conductivity is in two order magnitude less than the horizontal effective hydraulic conductivity; (2) BTCs exhibit longer late tails and they are looks like non-Fickian; (3) the estimated effective longitudinal macrodispersivity in the vertical direction is in one order magnitude less than longitudinal macrodispersivity in the horizontal direction, that means for using Fickian framework for dispersion modeling in regional transport simulation oneneed to apply at least anisotropic-media dispersion model.