Uncertainty Analysis Of Solute Transport Simulation In Heterogeneous Porous Media

Groundwater resource is an important component of water resource management strategy.In order to ensure the availability of water in sufficient quantity and quality to meet the growing demand from industrial,agricultural,and municipal uses,it is necessary to know as accurately as possible the distributions of groundwater and the transport properties of potential contaminants.Groundwater numerical simulation provides strong support in groundwater resource management and contamination risk evaluation by modeling and predicting groundwater movement and solute transport.However,the subsurface environment is highly heterogeneous,the hydrogeological structure and the heterogeneity of geological properties control the groundwater movement patterns and strongly influence the solute transport processes.At the same time,this heterogeneity is often directly related to scaling issues.From the local scale to the regional scale,the controlling roles of water flow and solute transport are quite different,and the model parameters also exhibit strong scale effects.Limited by the measurement method and measurement accuracy,the uncertainty relating to an incomplete representation of heterogeneous sediment structures and geological property parameters often propagates through the modeling processes,producing uncertain hydrogeologic predictions.Based on this,this thesis focuses on three main features in the modeling processes in heterogeneous porous media: the multi-scale property,heterogeneity and uncertainty of the simulation results.Under a hierarchical sedimentary structure framework,a set of high-resolution three-dimensional heterogeneous sedimentary structure models are constructed through the transition probability-Markov chain-based indicator kriging method by using site borehole data and trench data: the multiscale models(Multiscale model)and the large scale models(Scale II model).On this basis,the transport simulations of non-reactive solute are performed.The simulation results identified the sediment structure scale that controls solute transport.The sedimentary structure parameters and hydraulic conductivity univariate statistical parameters that are most sensitive to the solute transport processes are selected through the global sensitivity analysis.With the consideration of other influencing parameters,the impact of changes in these parameters on the solute dispersion and mixing progress is discussed in detail.Through the above research,the following conclusions are mainly obtained:(1)Under the framework of the hierarchical sedimentary architecture,the stratal architecture in the Qiqihar area can be defined as a hierarchy of textural based strata types,with larger-scale(level Ⅱ)strata(including a total of three kinds of lithofacies),each comprising assemblages of smaller-scale(level I)strata types(including a total of eight sub-lithofacies).(2)The strata information on small scales(scale Ⅰ)contributes insignificantly to solute transport,and the strata types actually controlling the plume dispersion and mixing is the larger-scale level II units.Overall,smaller solute plume may lead to a significant increase in the uncertainty of the simulation results.When a smaller solute plume is used,it requires a longer transport time for the macrodispersivity to reach its asymptotically stable value,and the final asymptotically value is reduced.Different from the solute dispersion process,the smaller plume promotes solute mixing and shortens the time required to reach a fully mixed state by increasing the dilution rate.(3)The ensemble dispersivity which is usually evaluated in the literature considerably overestimates the dispersion typically found in one given realization of the medium.Especially,when the small solute plume is used,this overestimation error is more pronounced.Therefore,it is more accurate to use the effective dispersivity.The dilution index calculated based on the improved method can effectively avoid the overestimation error caused by the concentration interpolation,so as to achieve an accurate characterization of the solute mixing evolution.(4)The solute dispersion is sensitive to sedimentary architectures and the corresponding fluctuations of the conductivity field,while the solute mixing is mainly influenced by molecular diffusion and local-scale dispersion.In addition,incorporating local-scale dispersion and molecular diffusion into transport modeling can effectively avoid overestimation of solute dispersion and underestimation of solute mixing,and reduce the uncertainty of simulation results.(5)Larger volume proportion of higher permeability facies(which means higher connectivity),higher heterogeneity(which means higher conductivity variance),and higher hydraulic conductivity contrast all lead to significant enhancement of longitudinal dispersion and transverse mixing.Among them,the interactions between the volume proportion and other parameters has the largest effect on the longitudinal dispersion,and the mean conductivities is the second sensitive parameters.The larger the mean conductivity,the more sensitive it is to the longitudinal dispersion.Valuable information on geological structures can be obtained from borehole measurements,outcrops,laboratory analyses of field samples,and geophysical and hydrogeological experiments,however,relevant information about the aquifer heterogeneous structure parameters along the flow direction is most important for interpreting the solute plume dynamics.The above research results will enhance our scientific understanding of the scale effect mechanisms of groundwater flow and solute transport parameters in heterogeneous porous media,provide an insightful and informative platform for groundwater contaminant assessment and remediation at various spatial scales,and thus achieve sustainable development and scientific management of groundwater resources.