Experimental and computational investigations of subsurface heterogeneity and contaminant distributions in complex geologic media

地下在多个尺度上是不均匀的，这是规则，而不是例外。关于流动和运输参数的空间变异性的信息最常见的是通过对小规模测量进行地统计分析而获得的。尽管地质统计学方法普遍使用，但尚不清楚它是否能够保存地质特征、它们的连通性和形态，这些特征可能对流体流动以及更重要的污染物传输产生重大影响。我们最近在利用水力层析成像绘制水力传导性(K)和比储存量(SS)非均质性方面取得了重大进展。水力层析成像包括从封隔层顺序抽水，以及相应地监测附近井眼跨座式封隔器系统中的水头变化信息。以层析方式抽取和收集水头变化数据，使人们能够从不同方向获得关于压力变化在含水层中传播的信息。该信息随后被用于计算K和SS层析图像的随机逆模型中。本课题组在实验室进行的水压层析成像实验表明，它可以比克立格法更好地成像K和SS的分布以及它们的连通性。该方法最近已扩展到包括分割示踪剂传输信息，以成像三维中稠密非水相液体(DNAPL)饱和的位置。初步结果表明，DNAPL饱和度及其水池可以得到准确的成像。震源区DNAPL饱和度的准确成像将大大提高我们修复加拿大及其他地区数千个地点的能力。拟议的研究重点是进行补充的实验和计算调查，这将大大提高我们绘制复杂地质介质中地下非均质性和污染源区域的能力。这些能力的发展还将使我们能够大大提高我们对这种介质中的流动和运输过程的基本理解。

The subsurface is heterogeneous at multiple scales and this is the rule rather than the exception. Information about spatial variability of flow and transport parameters is most commonly obtained through geostatistical analysis of small-scale measurements. Despite its common use, it is unclear whether the geostatistical approach can preserve geological features, their connectivity and morphology which can have large impacts on fluid flow and more importantly, contaminant transport. We have recently made significant strides in mapping hydraulic conductivity (K) and specific storage (Ss) heterogeneity using hydraulic tomography. Hydraulic tomography involves the sequential pumping of water from packed-off intervals and the corresponding monitoring of head change information in straddle-packer systems in nearby boreholes. Pumping and collection of head change data in a tomographic manner allows one to obtain information on the propagation of pressure changes through the aquifer from various directions. This information is then used in a stochastic inverse model that computes K and Ss tomograms. Hydraulic tomography experiments performed in the laboratory by our group show that it can image the distributions of K and Ss as well as their connectivity considerably better than kriging. The approach has been recently extended to include partitioning tracer transport information to image the locations of Dense Nonaqueous Phase Liquid (DNAPL) saturations in 3-D. Preliminary results show that DNAPL saturations and their pools can be imaged accurately. Accurate imaging of DNAPL saturations in source zones will considerably improve our abilities to remediate thousands of sites in Canada and beyond. The proposed research focuses on conducting complementary experimental and computational investigations that should significantly advance our capabilities to map subsurface heterogeneity and contaminant source zones in complex geologic media. Development of these capabilities will also allow us to significantly advance our fundamental understanding of flow and transport processes in such media.