Reactive transport modeling of subsurface arsenic removal systems in rural Bangladesh

M. M. Rahman, M. Bakker, C. H. L. Patty, Z. Hassan, W. F. M. Roeling, K. M. Ahmed, B. M. van Breukelen

Research output: Contribution to journalArticleScientificpeer-review


Subsurface Arsenic Removal (SAR) is a technique for in-situ removal of arsenic from groundwater. Extracted groundwater is aerated and re-injected into an anoxic aquifer, where the oxygen in the injected water reacts with ferrous iron in the aquifer to form hydrous ferric oxide (HFO). Subsequent extraction of groundwater contains temporarily lower As concentrations, because As sorbs onto the HFO. Injection, storage, and extraction together is called a cycle. A reactive transport model (RTM) was developed in PHREEQC to determine the hydrogeochemical processes responsible for As (im)mobilization during experimental SAR operation performed in Bangladesh. Oxidation of Fe(II) and As(III) were modeled using kinetic-rate expressions. Cation exchange, precipitation of HFO, and surface complexation, were modeled as equilibrium processes. A best set of surface complexation reactions and corresponding equilibrium constants was adopted from previous studies to simulate all 20 cycles of a SAR experiment. The model gives a reasonable match with observed concentrations of different elements in the extracted water (e.g., the r2 value of As was 0.59 or higher). As concentrations in the extracted water are governed by four major processes. First, As concentration decreases in response to the elevated pH of injection water and likewise increases when native neutral pH groundwater flows in. Second, the sorption capacity for As increases due to the gradual buildup of HFO. Third, As sorption is enhanced by preferential removal of As(V). Fourth, competitive sorption of Si limits the capacity of freshly precipitated HFO for As sorption. Transferability of the developed reactive transport model was demonstrated through successful application of the model, without further calibration, to two additional SAR sites in Bangladesh. This gives confidence that the model could be useful to assess potential SAR performance at locations in Bangladesh based on local hydrogeochemical conditions.
Original languageEnglish
Pages (from-to)277-293
Number of pages17
JournalScience of the Total Environment
Publication statusPublished - 2015


  • Arsenic
  • Reactive transport modeling
  • Groundwater
  • Bangladesh
  • Subsurface arsenic removal

Cite this