Experimental and Numerical Analysis of the Effect of Sediment Bed Thickness and Dune Bed-Form on Contamination Transmission in Rivers

The industrialization of human societies is one of the factors in increasing the injection of pollution into surface waters. Contamination in the river is maintained both in the water stream and in the sediment bed of the river. The hyporheic zone is the saturation zone of the riverbed, which has a very important function in the transmission of pollution. River bed-form is one of the effective factors in creating hyporheic exchanges between surface flow and sediment bed. Transient Storage Model (TSM) is one of the suitable methods in the analysis of advection and dispersion of pollution in rivers with hyporheic zone. The efficiency of the Transient Storage Model (TSM) depends on accurate estimation of the four parameters of the model (Dx, As, A and α). Previous studies have investigated the effect of bed-form on hyporheic exchanges, but the effect of these exchanges on contamination transmission has not been investigated. On the other hand, previous studies have not investigated the effect of hyporheic exchanges caused by the formation of the bed-form on the four parameters of the transient storage model (TSM). In this study, the effect of dune bed-form on the transmission of pollution was investigated. Also, the effect of dune formation on the four parameters of the transient storage model (TSM) with a numerical model (OTIS and OTIS-P) and the temporal moment approach (TM) was discussed.

Experiments of tracer material (NaCl) were performed in a flume with a length of 12 m, a width of 0.5 m and a height of 0.7 m applying four different flow discharges (5, 7.5, 10 and 12.5 l/s). The experiments were performed in the range of average flow velocity (U) from 0.087 to 0.361 and in the range of Froude number (Fr) from 0.069 to 0.290. An ultrasonic flow-meter was used to measure the flow discharge in all experiments. Grain material with an average diameter (D50) of 11.85 mm and the porosity (n) of 0.28 were used to create a sedimentary bed. The first bed (WF1) with a thickness of 32 cm, a width of 0.5 m and a length of 10 m was created in the Flume. In order to investigate the effect of sediment bed thickness on contamination transmission, a second bed (WF2) with a thickness of 8 cm, a width of 0.5 m and a length of 10 m was also created. In this study, the effect of dune bed-form on the transmission of contamination was investigated by creating three dune bed-forms (D1, D2 and D3) with different wavelengths (λ) and different amplitudes (∆) in the second bed (WF2). The length of the flume was divided into three equal reaches. Two sensors were placed to measure the electrical conductivity (ec) of water in each reach aiming to monitor the concentration of contamination. A Pitot tube and an ultrasonic depth-gauge were used to measure the velocity (U) and depth (d) of water flow at each reach, respectively. The laboratory results were simulated by the OTIS-P numerical model and the four parameters of the Transient Storage Model (TSM) were estimated. OTIS-P numerical model estimates the four parameters of the Transient Storage Model (TSM) using the Nonlinear Least Squares (NLS) optimization algorithm and then simulates the breakthrough curves using the Crank-Nicolson implicit finite difference method. The parameters of the transient storage model (TSM) were estimated by optimizing the temporal moment approach (TM) relations using the Genetic Algorithm (GA) method and the breakthrough curves were reproduced using these parameters in OTIS software.

The results showed that increase the thickness of the sediment bed (from 8 cm (WF2) to 32 cm (WF1)) reduces the (Dx). Hyporheic exchanges decrease with increasing d_b/d, so the amount of hyporheic exchanges in WF2 bed is more than WF1 bed. Increasing hyporheic exchanges in the WF2 bed-form reduces the amount of contamination concentration in the main flow area, so the amount of (Dx) in this bed-form increases. The results showed that increasing Froude number (Fr) increases the (Dx) in both cases of bed thickness. The results showed that the storage zone exchange coefficient (α) in WF2 bed was higher than WF1 bed. Decreasing d_b/d increases the hyporheic exchanges at the time scale (t_f^*), so the residence time of contamination in the sediment bed increases. Since the storage zone exchange coefficient (α) indicates the amount of time exchanges of contamination in hyporheic zone, so reducing the bed thickness (db) increases this parameter. The results showed that the formation of the dune bed-form does not cause an absolute increase or an absolute decrease in the longitudinal dispersion coefficients (Dx). The main channel area (A) considering the D2 bed-form case was estimated more than that in the D1 bed-form case. Therefore, increasing the wavelength of the bed-form (λ) increases the share of the main flow region area (A) in the contamination transmission. On the other hand, in the case of less exchanges of contamination with the storage zone, the share of the storage zone area (As) decreases with increasing wavelength of the bed-form (λ). The results showed that the effect of dune bed-form on changes in TSM parameters was in the range of Fr