Mathematical simulation of a dipole delivery system for in-situ remediation
Abstract In-situ remediation using reactive zones is a promising groundwater contaminant treatment technology that involves the injection of a reagent(s) into the subsurface to destruct harmful target chemicals. For efficient and effective treatment the reagent has to be delivered into a specific contaminated zone for the desired chemical reaction(s) to occur. The most commonly used delivery method is a conventional well where the distribution of injected reagent is mainly controlled by the surrounding hydraulic conductivity field. In this case, the reagent is easily delivered into the higher hydraulic conductivity zones but the lower hydraulic conductivity zones are missed. The goal of this research effort is to investigate a novel delivery method involving a single well vertical recirculation system or a dipole well. The configuration of this single dipole well is that injection and extraction occurs from two chambers separated by an impermeable central packer. Thus, this dipole well system can induce predominantly vertical flow across bedding plane features and it is therefore hypothesised that this delivery system can overcome physical heterogeneities creating a more uniform reactive zone. The objective of this research was to demonstrate that the dipole well is a useful delivery tool compared to the commonly used single injection well. Mathematical simulations were used to investigate the delivery performance of a dipole well using steady-state and transient approaches. A simple analytical model was used to determine the steady-state dipole flow field and observe the impact of system parameters on reagent delivery behaviour. The size of coverage area (the area swept by the injected reagent) was used as the performance metric to assess the impact of each system parameter on the dipole well performance. Numerical simulations were used to extend this investigation to homogeneous and heterogeneous (structured or randomly correlated hydraulic conductivity) aquifers under pulsed operation to identify those situations where the dipole delivery system is more efficient or effective. Both forward and backward particle path lines were used to identify reagent coverage areas around the injection well and down gradient. The impact of each system parameters on the dipole well performance was studied. The shoulder length and the injection cost are characteristic parameters that affect dipole delivery performance. A relationship between the down gradient coverage area vs. characteristic system parameters was developed and can be used to predict the dipole well performance in homogenous aquifers. The impact of the hydraulic conductivity distribution on dipole well performance is consistent with either a structured hydraulic conductivity field or randomly correlated hydraulic conductivity fields. Regions of lower hydraulic conductivity can be swept by the dipole well and the dipole well outperforms a single injection well, which is analyzed as a base case in terms of the shape of down gradient coverage area. However, the advantage of dipole well over a single well delivery is small if the degree of heterogeneity is large or the horizontal extent of the bedding plane is small.