Contaminant transport in coarse-grained porous medias: a multi-model approach with advanced spectral analysis

This study investigates contaminant transport in coarse-grained porous medium through experimental and modeling exercises. Breakthrough curves were obtained from a one-dimensional gravel bed with a mean particle diameter of 9 mm and porosity of 42%, using an array of electrical conductivity sensors at different depths. Sodium chloride (NaCl) concentrations ranging from 25 to 100 g/L were injected under two flow rates: 0.19 L/s and 0.36 L/s. Three transport models were employed including mobile immobile model (MIM), dual-porosity model (DPM), and fractional advection dispersion model (FADE) to analyze the data. Key estimated parameters included a mobile water fraction (β) of 0.65-0.75, a mass transfer coefficient (ω) of 0.001-0.005 s^-1, and a fractional order (α) of 1.6-1.8. High Péclet numbers (Pe ≈ 270) indicated advection-dominated transport, while retardation factors (1.2-1.4) suggested intermediate sorption. Mass recovery rates ranged from 92 to 97%, while the lowest recovery observed at the highest injection concentration. Vertical velocity profiles revealed potential flow acceleration and preferential pathways, while dispersion coefficients generally decreased with depth. Model evaluation using multiple efficiency criteria indicated that FADE consistently outperformed MIM and DPM. Advanced processing techniques, including continuous wavelet transform, functional data analysis (FDA), and Hilbert-Huang transform, provided further insight into the temporal structures and dominant transport processes. The findings highlight the complexity of contaminant transport in coarse-grained porous media and emphasized the necessity of accounting for non-Fickian behavior, in modelling efforts. Moreover, the findings offer valuable insights for improving the prediction and management of contaminant transport in natural and engineered coarse-grained systems, such as riverbeds and groundwater recharge zones.