Collaborative Research: Solute Transport in Multimodal, Heterogeneous Geological Formations Combining Sedimentologic and Engineering Approaches

0001125Ritzi 我们将开发一个分层空间随机函数 (HSRF) 模型，该模型表示由粒状含水层系统中的边界表面定义的渗透性、沉积结构和沉积单元。 在最小尺度上，相被定义为单峰渗透率区域，以及与纹层具有共同方向的跨地层组（如果存在）。 相将由连续随机变量表示，该变量具有由层理和/或跨地层组的方向控制的各向异性相关性。 相的组合，例如填充单个河道的相的组合，将由指示随机变量和相关的转变概率来表示，所述转变概率表示每个相的比例、几何形状和并置。 河道带沉积物内的并置河道充填、坝体增生和河漫滩组合等复杂的相组合由更高层的指示随机变量表示，其转移概率代表组合的比例、几何形状和并置。 所有三个水平的连续变量和指示变量将以数学方式表达在单个 HSRF 模型中，该模型适合与溶质扩散的分析表达式集成。 HSRF 模型不同于当前的模型和方法，因为它可以捕获复杂的非高斯空间分布，并且可以结合解释沉积学来约束地质统计模型。在结合现场研究和模型气质中，我们寻求解决以下与（1）颗粒介质中的渗透率分布和（2）溶质扩散相关的问题。 (1) 与渗透率分布相关的问题： o 具有三个层级的 HSRF 模型是否捕获了渗透率和沉积结构方面对于理解典型羽流尺度上的溶质扩散很重要的信息？ o 任一层级的属性是否与另一层级的属性相关？ o HSRF 模型能否用于改进未知位置渗透率的条件估计？ o HSRF 模型是否可以用作开发地质统计模型时纳入地质/沉积学软数据的框架？ (2) 与溶质扩散相关的问题： o 三个层级对纵向和横向溶质扩散的相对贡献是什么？ o 随着单变量统计数据和渗透率的空间相关性在相边界上变得越来越不同，溶质扩散的时刻在什么时候受到显着影响？ o 在什么情况下可以忽略任何层次级别的表示？ o 在什么条件下可以使用有效的、羽流尺度相关的参数对多模式异质性中的运输进行建模？ 为了回答这些问题，我们正在开展一个跨学科合作项目，涉及加州大学伯克利分校和莱特州立大学的研究人员。 与赖特州立大学过去 URI 资助的项目一样，该项目将为本科生提供研究经验。

0001125Ritzi We will develop a hierarchical spatial random function (HSRF) model that represents permeability, sedimentary structures, and sedimentary units defined by bounding surfaces in granular aquifer systems. At the smallest scale, facies are defined as regions of unimodal permeability and, if present, of cross-strata sets with common orientation to the laminae. Facies will be represented by a continuous random variable having anisotropic correlation controlled by bedding and/or the orientation of cross-strata sets. Assemblages of facies, such as the combination of facies filling a single channel, will be represented by an indicator random variable and associated transition probabilities which represent the proportions, geometries, and juxtapositioning of each facies. A complex of facies assemblages, such as juxtaposed channel fill, bar accretion, and overbank assemblages within channel belt deposits, is represented by a higher-tier indicator random variable with transition probabilities representing the proportions, geometries, and juxtapositioning of the assemblages. The continuous and indicator variables of all three levels will be mathematically expressed in a single HSRF model suitable for integration with analytical expressions for solute spreading. The HSRF model departs from current models and approaches in that it can capture complex, non-Gaussian spatial distributions and can incorporate interpretive sedimentology to constrain the geostatistical models. In combined field studies and model temperament, we seek to address the following questions relating to (1) the distribution of permeability in granular media and (2) solute spreading. (1) Questions relating to distribution of permeability: o Does the HSRF model, with its three hierarchical levels, capture the aspects of permeability and sedimentary structure important for understanding solute spreading at the scale of typical plumes? o Are the attributes of any one hierarchical level correlated with those of another level? o Can the HSRF model be used to improve the conditional estimation of permeability at unknown locations? o Can the HSRF model be used as a framework for incorporating geological/sedimentological soft data as geostatisticl models are developed? (2) Questions related to solute spreading: o What are the relative contributions of the three hierarchical levels to both longitudinal and lateral solute spreading? o As the univariate statististics and the spatial correlation of permeability become increasingly different across facies boundaries, at what point are the moments of solute spreading significantly affected? o Under what conditions can representation of any of the hierarchical levels be ignored? o Under what conditions can transport in a multimodal heterogeneity be modeled with effective, plume-scale-dependent parameters? To answer these questions, we are conducting a collaborative, interdisciplinary project involving investigators at both the University of California, Berkeley, and Wright State University. As in past URI-funded projects at Wright State University, the project will provide research experiences for undergraduates.