Detecting High Conductivity Channels in Heterogeneous Porous Media Using an Iterative Sampling Strategy

Transport of chemical species in the subsurface are controlled by the spatial distribution of the hydraulic conductivity field. In many cases, special structures describing the subsurface, such as high conductivity channels, have been shown to control solute transport. However, in real scenarios the exact location of these structures is unknown and subject to a high level of uncertainty. Reducing the uncertainty of high conductivity channels will indirectly reduce the uncertainty of solute transport key features. In this work, we propose an iterative framework to improve the characterization of high conductivity channels. The channels are identified using the least resistance path and minimum hydraulic resistance. Instead of sampling the hydraulic conductivity in different locations altogether, we divide the site characterization campaign into multiple steps. For each step, the sampling locations are chosen to maximize the probability of finding the least resistance path.. Application of the iterative strategy in synthetic conductivity fields have shown more than 40% reduction of first arrival time uncertainty when compared to a regular grid sampling protocol, while keeping the characterization effort (i.e., number of sampling locations) constant.