Engineering Applications of Computational Fluid Mechanics | |
Towards a robust CFD model for aeration tanks for sewage treatment – a lab-scale study | |
Anna M. Karpinska1  John Bridgeman1  | |
[1] University of Birmingham; | |
关键词: aeration; CFD; Eulerian two-fluid model; mixing; oxygen transfer; PIV; | |
DOI : 10.1080/19942060.2017.1307282 | |
来源: DOAJ |
【 摘 要 】
Detailed insight into the hydrodynamics of aeration tanks is of crucial importance for improvements in treatment efficiency, optimization of the process design and energy-efficient operation. These factors have triggered increasing interest in the use of Computational Fluid Dynamics (CFD) to evaluate performance of wastewater treatment systems. Whilst factors such as incorrect input assumptions, poor model choice and excessive simplifications have been recognized as potential sources of output errors, there remains a need to identify the most robust strategy to faithfully simulate aeration tank performance. Therefore, the focus of this work was to undertake rigorous transient simulations of the hydrodynamics and oxygen mass transfer in a lab-scale aeration tank in order to work towards the development of robust modeling guidelines for activated sludge systems. Unlike most previous CFD analyses of aeration systems, the work reported here employed the shear stress transport (SST) $ k - \omega $ turbulence model to account for the turbulent interactions between the phases inducing bubble breakup and/or coalescence, and as a consequence, promoting the formation of bubbles of different sizes and shapes. The results obtained were compared with those arising from an analysis using the standard $ k - \varepsilon $ ( $ sk - \varepsilon $ ) model – and assuming fixed bubble diameter- the most common CFD modeling framework used within the wastewater modeling community. Model validation was achieved using acoustic Doppler velocimetry and particle image velocimetry techniques, and experimentally derived oxygen mass transfer data. Limitations of both turbulence models used and modeling assumptions concerning bubbly flow are discussed. The benefits of the SST $ k - \omega $ turbulence model are demonstrated, but the need to balance the increased computational expense of this approach compared to the $ sk - \varepsilon $ model and, indeed, bubble flow modeling are recognized. Thus, this paper presents the first rigorous analysis of turbulence model and bubble flow generation models together for activated sludge system optimization. Abbreviations: ADV: Acoustic Doppler velocimetry; AS: Activated sludge; ASM1: Activated Sludge Model No. 1; BOD: Biochemical oxygen demand; CCD: Charge-coupled device; CFD: Computational fluid dynamics; COD: Chemical oxygen demand; CPU: Central processing unit; DO: Dissolved oxygen; GCI: Grid convergence index; HPC: High performance computing; IAC: Interfacial area concentration; MRF: Multiple reference frame; MLSS: Mixed liquor suspended solids; PBM: Population balance models; PIV: Particle image velocimetry; PST: Phase-space thresholding; RAM: Random access memory; RANS: Reynolds-averaged Navier-Stokes; SNR: Signal-to-noise-ratio; SST: Shear stress transport
【 授权许可】
Unknown