Abstract

Arsenic contamination in water sources, mainly as arsenite (As(III)), poses significant health risks and remains a global environmental challenge. Horizontal subsurface flow constructed wetlands offer a sustainable, costeffective solution for arsenic removal, but their application is limited. Robust predictive models are required to increase their understanding and evaluate their application. This study presents a mechanistic model able to simulate the abiotic removal processes for As(III) in horizontal subsurface flow constructed wetlands. Building on RCB-ARSENIC, the proposed model incorporates the key removal processes: sorption onto granular media and adsorption/coprecipitation onto ferrihydrite, providing a comprehensive tool for predicting arsenic and iron transport and retention dynamics. Relevant precipitation reactions were also considered, but they do not occur under the experimental conditions. The model was validated using experimental data from a laboratory-scale, non-vegetated horizontal subsurface flow constructed wetland prototype, operated with synthetic water containing 100 mu g/L of As(III) and having natural silty sand as supporting media. Simulation results showed a 14.7 % arsenic removal efficiency through media sorption and 50.5 % iron removal via physical ferrihydrite filtration. Parameters fitting yielded a longitudinal dispersivity of 0.903 m and a maximum sorption capacity of 66,604 g/ kg, consistent with the reported values for silty sand; highlighting the roles of dispersivity and sorption capacity in arsenic transport and retention and emphasizing their importance in wetland design. The model presented a high goodness of fit, properly representing the main physical and chemical processes that remove arsenic in horizontal subsurface flow constructed wetlands, with potential for enhancement by including biotic processes.