Total selenium removal by the activated sludge process, where selenite is reduced to colloidal elemental selenium nanoparticles (BioSeNPs) that remain entrapped in the activated sludge flocs, was studied. Total selenium removal efficiencies with glucose as electron donor (2.0 g chemical oxygen demand (COD) L(-1)) at neutral pH and 30 °C gave 2.9 and 6.8 times higher removal efficiencies as compared to the electron donors lactate and acetate, respectively. Total selenium removal efficiencies of 79 (±3) and 86 (±1) % were achieved in shake flasks and fed batch reactors, respectively, at dissolved oxygen (DO) concentrations above 4.0 mg L(-1) and 30 °C when fed with 172 mg L(-1) (1 mM) Na2SeO3 and 2.0 g L(-1) COD of glucose. Continuously operated reactors operating at neutral pH, 30 °C and a DO >3 mg L(-1) removed 33.98 and 36.65 mg of total selenium per gram of total suspended solids (TSS) at TSS concentrations of 1.3 and 3.0 g L(-1), respectively. However, selenite toxicity to the activated sludge led to failure of a continuously operating activated sludge reactor at the applied loading rates. This suggests that a higher hydraulic retention time (HRT) or different reactor configurations need to be applied for selenium-removing activated sludge processes. Graphical Abstract Scheme representing the possible mechanisms of selenite reduction at high and low DO levels in the activated sludge process.
Keywords: Activated sludge; Dissolved oxygen; Elemental selenium; Selenite; Selenium removal; Toxicity.