Inverse estimation of the release rate of the heated pollutant source with strong convection and its application to the exhaust system design

Abstract

Accurate release rate of the source is a crucial parameter for the refined design of the exhaust system in the industrial buildings. For the heated pollutant sources emitted by strong convection (SCHP source), it is difficult to accurately measure the source release rate with instruments due to inconsistent emission parameters tested at different locations near the source. In this paper, the three-dimensional CFD (Computational Fluid Dynamics) simulation was used to obtain the hourly concentration of pollutants and to study the applicability of four different regularization methods in the inverse estimation of the release rate of the SCHP source. The influence of the denoising filter and the strong convection of the SCHP source on the accuracy of the inverse estimated source release rate (IESR) was analyzed, and an exhaust flow rate calculation method based on the IESR is proposed. The results show that, compared with the zero-order Tikhonov regularization (ZOTR) and the LSQR methods, the second-order Tikhonov regularization (SOTR) and the truncated SVD (TSVD) methods are more suitable for the inverse estimation of the SCHP source. And it is found that, the introduction of the denoising filter can effectively eliminate the high-frequency or the high-amplitude deviations caused by the regularization method, compared with the SOTR method, the RMSE can be reduced by a maximum of 37.04 %. It is also concluded that the strong convection and the measurement error both have the negative impact on the accuracy of the IESR. Finally, compared to the calculation methods in the existing design manuals, the local exhaust system designed by the IESR method can efficiently capture the pollutants with a 46 % reduction in the exhaust flow rate. This study is useful for the accurately determining the SCHP source release rate and the optimal design of the exhaust system.