Droplet size prediction model based on upper limit log normal distribution in venturi scrubber

Abstract

A model for the prediction of the droplet diameter during breakup in a venturi scrubber is proposed. The proposed model simultaneously solves the Lagrangian particle tracking of a droplet and the Reitz-Diwakar breakup model. Results demonstrate that the proposed model for the maximum stable droplet diameter provides better predictions using the existing experimental data, especially under a high Weber number. As a well-known droplet size distribution function, the Upper Limit Log Normal (ULLN) distribution function requires three parameters: maximum droplet diameter (Dmax) and two distribution parameters (a, $\delta$). For a Dmax, the calculation results using the proposed method were used, while the values of the a and $\delta$ were selected using the values that best fit the existing experimental data. The comparison of results against experimental data for the Sauter mean diameter (SMD) in venturi scrubber geometry indicate that the proposed model provides better predictions than the existing empirical correlations suggested by Boll et al., and Nukiyama and Tanasawa do. Based on SMD, one-dimensional calculation methodology is proposed to obtain the filtering efficiency and the pressure drop inside the venturi scrubber. To verify and validate the proposed prediction model, comparative analyses against two experiments are performed. As a first verification and validation case, we perform a comparative analysis for the existing Haller experiment operating at atmospheric pressure. The results show that the Normalized Root Mean Square Differences (NRMSD) of the proposed model show the 6.6% on the filtering efficiency and 18.7% for pressure drop which is a smaller value than those for the previous Yung and Calvert model of 10.3% and 26.8%, respectively. As a second verification and validation case, we perform a comparative analysis for the existing Yung experiment operating at high pressure. The operating condition of the venturi scrubber in the nuclear industry is in a high pressure condition because the filtered venting system operates in the high containment pressure condition. The proposed model predicted all data with the NRMSD of 7.0% for the filtering efficiency and 24.0% for the pressure drop while the Yung model does with those of 15.6% and 37.7%, respectively. In overall, the proposed model shows a quite reasonable prediction capability in the high pressure condition in respect to the droplet diameter, filtering efficiency and pressure drop. Because of this significant improvement in the prediction of venturi scrubber performance, the applicability range of the proposed model can extend to the high pressure condition.