Iron and arsenic removal rates in a continuous flow reactor treating As-rich acid mine drainage (AMD)
Résumé
Arsenic is one of the most hazardous elements associated to acid mine drainage with concentrations
ranging from ≤ 1 µg L-1 to hundreds of mg L-1 in these leachates. To date, there is no affordable method
to efficiently treat As-rich AMD over the long term. Natural attenuation processes based on oxidation
of iron and arsenic by autochthonous bacteria present in AMD promote precipitation of these elements.
The exploitation of these processes offer a promising method for the treatment of As-rich acid mine
drainage (AMD).
In the present study, the Fe(II) oxidation rates and Fe and As removal rates have been investigated in a
laboratory pilot-scale system with a continuous circulation of AMD water under controlled conditions
(temperature, light, water-height and residence time). The system was fed with AMD from the Carnoulès
mine with a permanent acid character (pH 2-4), an average iron concentration of 0.5 g L-1 in the reduced
form (Fe(II)) and an average arsenic concentration of 40 mg L-1 predominantly under As(III) oxidation
state.
The efficiency and rates of Fe(II) oxidation and of Fe and As removal were determined in the pilot as a
function of experiment duration and for a range of residence times, in presence or absence of a floating
film that develops naturally at the surface of the water. The Fe- and As-rich biofilms formed in the pilot
were characterized for their mineralogy, redox As speciation, and bacterial quantification.
During the early stage of experiment, results evidenced that the gradual coverage of the pilot channel
bottom with an Fe-As precipitate, that contained 6 ± 3 x 106
bacterial cell/g (dry wt.), increased the
Fe(II) oxidation efficiency and the Fe and As removal by two orders of magnitude. At the steady state,
90 % Fe(II) was oxidized and 30 % Fe and 80 % As were precipitated within a residence time of
200 min. Rate values were slightly lower than those measured in the field in the natural AMD stream at
Carnoulès. The development of a thin floating film at the surface of water in the pilot channel affected
oxygen diffusion and retarded these rates. All these results improve our understanding of natural
processes responsible for arsenic attenuation in AMD, and give some indications that might be
considered when designing an Fe and As removal passive treatment in AMDs.