APPLICATION OF COMPUTATIONAL FLUID DYNAMICS TO CLOSED LOOP BIOREACTORS ? Analysis of Macro-Environment Variations in Simultaneous Biological Nutrient Removal Systems

摘要

A full-scale closed loop bioreactors (Orbal? oxidation ditch) previously examined for simultaneous biological nutrient removal (SBNR) was further evaluated using computational fluid dynamics (CFD). The field data, CFD simulation and microbiological observation all suggest that SBNR is occurring in these systems despite the lack of formal anoxic or anaerobic zones. We have previously hypothesized that three mechanisms might be responsible for SBNR: reactor macro-environment, floc micro-environment, and novel microorganisms. The research reported here focuses on characterizing the macro-environment in terms of the mixing pattern in a closed loop bioreactor. A CFD model was developed by imparting the known momentum calculated by tank recirculation velocity and fluid mass to the fluid at the aeration disc. Oxygen source (aeration) and sink (DO consumption) terms were introduced, and statistical analysis applied to the CFD simulation results. The CFD model was calibrated with field data obtained from a test tank and a full-scale tank. The results indicated that CFD could predict the mixing pattern in closed loop bioreactors. This enables visualization of the flow pattern both with regard to flow velocity and oxygen distribution profiles. The velocity and oxygen distribution gradients suggested that the flow patterns produced by directional aeration in closed loop bioreactors created a heterogeneous environment that can result in variations in the oxidation-reduction potential throughout the bioreactor. Distinct anaerobic zones on a macro environment scale were not observed, but it is clear that when flow passed around curves, a secondary flow was generated with an alternate flow speed in a spiral fashion. This second current along with main recirculation flow could create alternating anaerobic and aerobic conditions vertically and horizontally, which would allow SBNR to occur. Reliable SBNR performance in Orbal? oxidation ditches might be due to such a spatially varying environment. This will be further characterized by introducing PAOs into the computational fluid dynamics model.