On the evaluation of power density models for oscillatory baffled reactors using CFD

• Published in: Chemical engineering and processing Vol. 134; pp. 153 - 162
• Main Authors: Jimeno, Guillermo, Lee, Yeaw Chu, Ni, Xiong-Wei
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.12.2018
• Subjects: Computational fluid dynamics (CFD); Oscillatory baffled reactors; Power density; Pressure drop propagation; Power density; Pressure drop propagation; Computational fluid dynamics (CFD); Oscillatory baffled reactors
• ISSN: 0255-2701; 1873-3204
• DOI: 10.1016/j.cep.2018.11.002

[Display omitted] •CFD validation of power density estimation models in oscillatory baffled devices is for the first time investigated.•We have updated the orifice discharge coefficient and the dependency on the number-of-baffles term for the Quasi-steady model.•We have updated the dependency on the number of baffles & proposed empirical model to estimate mixing length for the Eddy enhancement model.•Both models are validated for a much wider application range than originally stated and for both batch and continuous.•Both revised models can be used interchangeably with high confidence. While continuous oscillatory baffled reactors (COBR) have been proven a viable alternative to traditional batch reactors for organic synthesis and crystallization, research into the estimation of power density for this type of device has largely been stagnated for the past 25 years. This work reports, for the first time, detailed analysis and examination of the applicability, capability and deficiencies of two existing models using CFD methodology. The “quasi-steady” model (QSM) over-estimates power dissipation rates due to the inaccurate formulation of two of its geometric parameters for modern COBRs. By using a revised power law dependency on the number-of-baffles term (nx) and an appropriate orifice discharge coefficient (CD), we demonstrate that the updated QSM can not only be used for a much wider application range than previously outlined, but also for both batch and continuous operations. The “eddy enhancement” model (EEM) generally provides better predictions of power density for the conditions tested; however, its accuracy can substantially be enhanced by applying the aforementioned power law dependency on n and an empirical correlation proposed in this work to estimate EEM’s “mixing length”. After full validation, both models give very similar power density estimations and can be used interchangeably with high confidence.