Influence of aeration onto As trapping in ZVI/sand reactor
Résumé
Arsenic toxicity and occurrence in the environment lead to the research of easy to handle and cheap water treatment process. Zero Valent Iron (ZVI) supports meet these requirements. ZVI trapping capacity is due to iron by-products, such as green rusts, lepidocrocite... According to physico-chemical conditions applied, the nature of by-products changes and arsenic trapping capacity is highly impacted. The aim of this work concerns the influence of aeration onto arsenic trapping in ZVI/sand columns. Four up-flow reactors filled with a ZVI/sand mixture were fed with synthetic water spiked with AsV. Two columns were aerated by air diffusers placed at the bottom while the two others were not aerated. Arsenic concentrations were monitored for three months in order to characterize arsenic sorption capacity of both configurations. A nearly complete removal was observed during the first five days into aerated system, whereas arsenic removal was stable, close to 50% during the first 20 days for the non aerated units. After this period an inversion of arsenic trapping capacity was observed, the non-aerated columns showed a higher arsenic removal than the aerated ones. Finally, when all columns reach their saturation, highest As trapping capacities were obtained in the non-aerated systems. According to arsenic monitoring, these retention capacities were estimated at 220 and 140 mg(As)/g(Fe) for non-aerated and aerated columns respectively. Solid support samples were collected along the reactor and four different layers were distinguished. After acid mineralization, analysis of total arsenic highlighted different repartitions inside the reactors: a homogeneous concentration in the non-aerated systems and a decreasing concentration form inlet to outlet for the aerated pilot units. In order to characterize As sorption into ZVI/sand columns, breakthrough curves were modelized using Thomas model. The determined KTh constants were different for aerated and non-aerated columns, which confirmed that the affinity of arsenic for solid support was different in the two systems. Oxygen measurement showed that synthetic water was close to saturation for aerated systems while it increased from 0.5 to 8 within 70 days in the non-aerated units. According to iron corrosion mechanisms, aeration induced a fast ZVI iron by-products generation, and thus a higher arsenic trapping capacity at the beginning of the experiment. Arsenic sorption into non-aerated reactors was lower but lasted longer, due to the regular ZVI-by-products generation. The lower trapping capacity of aerated systems could be due to a conversion over time of the early-created byproducts to less reactive forms.