Sediment-water interaction in a water reservoir affected by acid mine drainageexperimental and modeling

Resumen

The discharge of acid mine drainage into a water reservoir may seriously affect the water quality. In this setting, sediment is commonly thought to act as a sink for pollutants. However, redox oscillations in the bottom water promoted by stratification-turnover events may significantly alter the metal cycling. A new sequential extraction procedure has been developed to study the metal partitioning in the sediment. The new scheme for iron, sulfur and organic carbon rich sediments was evaluated for iron partitioning in synthetic iron monosulfides, natural pyrite, schwertmannite and goethite. This procedure enabled the following fractions to be distinguished: (1) water soluble, (2) exchangeable (monosulfides and carbonates), (3) low crystalline Fe(III)-oxyhydroxide, (4) crystalline Fe(III)-oxide, (5) organic matter, (6) sulfides (pyrite) and (7) residual. In soils or sediments where no carbonates are present step 2 can be used as the monosulfides fraction, which is recovered at 70%. Low crystalline Fe(III)-oxyhydroxides and crystalline Fe(III)-oxides are isolated in steps 3 and 4, respectively (>80%). Pyrite is completely isolated with HNO3 8M in step 7 (>99%). To study the effect of the redox in the sediments, an experiment was performed under controlled laboratory conditions. Sediment cores from the Sancho reservoir (Odiel Basin, SW of Spain), which receives AMD (pH<4), were maintained in a tank with bottom water for approximately 2 months and subjected to alternating oxic-hypoxic conditions. The tank water concentrations were monitored during the whole experiment, and the sediments were analyzed (porewater and solid phase) at the end of each oxic-hypoxic period. The results were then used to calibrate a diffusion-reaction model and to quantify reaction rates and sediment-water fluxes. Under oxic conditions, protons, Fe and As concentrations decreased in the tank due to schwertmannite precipitation, whereas those of Al, Zn, Cu, Ni, and Co increased due to Al(OH)3 and sulfide dissolution. The reverse trends occurred during hypoxia. Under oxic conditions, the fluxes calculated by applying Fick¿s first law on concentration gradients contradicted the ones expected from the evolution of the concentration in the tank water. This discrepancy was attributed to the coarse sediment resolution during sampling (cm) that failed to capture the very narrow concentration in the shallow pore water (mm) due to sulfide and Al(OH)3 dissolution. Finally, a new reactive transport model integrating solid fluxes and sedimentation was developed to extent to realistic field conditions. The model was again calibrated against porewater and solid phase composition from sediment cores collected in several points of the Sancho reservoir during stratification and turnover seasons. Results allowed to quantify that 45% of FeS and 30% of other metal sulfides precipitated under anoxic conditions were re-oxidized during the turnover. Both models agreed in that most of organic matter was degraded by sulfate. However, in the previous laboratory calibrated model oxygen was mainly consumed by sulfide and metal sulfides (around 90%), whereas according to the new model it was mainly consumed by organic matter (76%) and Fe(II) (9%). This difference is because if no accumulation is considered, the metal sulfides are not buried and the oxygen front progresses in depth oxidizing all the precipitated metal sulfides. Both models confirm that sediment acts as a sink for all the elements. Around 10% of the SO4, Al, Zn and Cu and only 2% of Co and Ni dissolved in the reservoir water were accumulated in the sediment. Around 80% of Fe and 70% of As and Mn transported to or precipitated in the water reservoir was accumulated as solid phase in the sediment. Finally, the model predicted that in ten years after the complete cease of the AMD input, the sediment would reach a new steady state with no contaminants released from the sediment to the water.