Reactive transport of arsenic-enriched geothermal spring water into a sedimentary aquifer.

The movement of high-arsenic (As)-concentration Beitou geothermal spring water up to 4600 μg/L from the upstream area may affect the downstream Guandu plain and Guandu wetland. The major As pathway is via the subsurface flow. The study assesses the fate and transport of As-enriched geothermal spring water in the Beitou-Guandu area. The groundwater head and flow field velocity are first simulated by using HYDROGEOCHEM-fluid flow model. The steady state flow field is well-calibrated with root mean square error 1.73 m and R = 0.992. The spatial-temporal distributions of As in the Beitou-Guandu area are simulated by the HYDROGEOCHEM reactive transport model using the calibrated steady state flow field. The results show that high As concentration (250 μg/L) in groundwater of Guandu plain was caused by the movement of high As concentration from the Beitou geothermal spring water. In contrast, the low As concentration (5-50 μg/L) in Guandu wetland was mitigated by the tidal water dilution. The simulated As concentrations increase in the first 3 years, and then gradually decrease due to the adsorption of As on the iron oxide minerals ferrihydrite and iron sulfide minerals pyrite. Furthermore, the hydrogeochemical transport model is applied to assess the effect of bioaccumulation of As by the mangrove plants of Guandu wetland. The dominant mangrove plants, Kandelia obovata, can reduce about 5-30 μg/L As concentration in groundwater. It may be one of sinks of As in Guandu wetland. The inclusion of K. obovata can uptake the aqueous As and allow the simulated As concentration further close to the field measurement in the Guandu wetland. The study successfully models the reactive chemical transport of As by considering both geochemical reactions and biochemical uptakes in the Beitou-Guandu area. The result demonstrates that the complex biogeochemical transport can be quantified by the sophisticated HYDROGEOCHEM model. Moreover, the salient features of the biogeochemical reactions can be recovered and elucidated through a series of systematic simulation.