In the context of in situ groundwater remediation, mixing is vital for a successful outcome. A slow mixing rate between the contaminated groundwater and the injected treatment solution can severely weaken the effective degradation rate. Engineered Injection-Extraction (EIE) has been proposed as a means to accelerate dilution within the porous medium. However, existing studies on the subject have not considered the potential impact of connectivity and preferential flow-paths. Neglecting connectivity can lead to an overestimation of EIE's capabilities, since the fluid may in reality be carried mainly through a few high-permeability channels, thus hampering mixing and reaction. Due to the fact that channeling can be found in many actual sites, in this work we aim to evaluate EIE methods in both poorly connected (represented as Multigaussian fields) and well-connected fields (represented as non-Multigaussians). The approach is to identify, for each given medium, a stirring protocol—defined by a specific combination of rotation angle and rotation rate—which maximizes mixing. To that end, metrics are proposed in order to (a) quantify both the mixing and the containment of the treatment solution within a given remediation volume, and (b) characterize the particle trajectories to explicitly evaluate if preferential paths are broken. The results obtained from these metrics are quite similar for both types of fields, proving that the enhancing of mixing by means of EIE is effective regardless of the presence of preferential flow paths. This study demonstrates that EIE via rotating dipoles diminishes the remediation outcome uncertainty induced by medium heterogeneity.
