Experimental and numerical study of bimolecular reactive transport in a single rough-wall fracture

Abstract In order to test the performance of the incomplete mixing model including the advection-dispersion-reaction equation (IM-ADRE) in simulating bimolecular reactive transport in fractured media, the laboratory-controlled experiments and numerical modeling were conducted in a single rough-wall fracture. Bimolecular reaction of aniline (AN) + 1, 2-napthoquinone-4-sulfonic acid (NQS) → 1, 2-naphthoquinone-4-aminobenzen (NQAB) was employed. Three different flow rates (0.15, 0.825 and 1.5 mL/s) and three fracture apertures (2, 3 and 4 mm) were considered in the experiment. The numerical simulation was analyzed and discussed based on the Peclet (Pe) number and Damkohler (Da) number. The results showed that although the spatial distribution of product concentration varied with the Pe number, the IM-ADRE model could simulate it well. The discrepancy between the simulated and measured peak product concentrations was less than 0.7%, which was far less than the value of 5% in previous study (Qian et al., 2015) produced by the IM-ADRE model used to modeling the reactive transport in porous media. This meant that the IM-ADRE model was effective for simulating and predicting the experimental results of bimolecular reactive transport in a single rough-wall fracture. Due to the instability of the product, the simulated value of the peak product concentration was always higher than the experimental value. Through further analysis of the model parameters, it was found that the hydrodynamic dispersion coefficient (D) changed significantly with the change of Pe number. The parameters m and β0 were linearly correlated with the Pe number. In addition, the IM-ADRE model was more sensitive to the parameters m and β0 but less sensitive to D.