An Advanced Lagrangian Puff Model, With Emphasis On Low Wind Speed Conditions
Free (open access)
A. Del Borghi, B. Fabiano, V. Dovì, C. Solisio & M. Del Borghi
An advanced Lagrangian puff model, with emphasis on low wind speed conditions A. Del Borghi, B. Fabiano, V. Dovi, C. Solisio & M. Del Borghi Chemical and Process Engineering Department DICheP \“G.B. Bonino” Genoa University, Italy Abstract Two main approaches can be outlined when dealing with the solution of transport and dispersion problems, namely Eulerian and Lagrangian one. The latter approach is becoming more and more widespread due to the availability of cheap computational power, which is a strict requirement of the relevant algorithms involved in the modelling. Lagrangian algorithms make it possible to give a much more detailed description of atmospheric diffusion phenomena, provided that the corresponding amount of information (air and soil properties) are available. Nowadays, a considerable number of well tested models is available and the state-of-the-art of the corresponding modelling is presently well developed . However, there are apparently minor problems that might cause large discrepancies between predicted and measured values especially in conditions of low wind speed. In fact, under these conditions, the meandering of the flow can be quite significant, leading to enhanced horizontal dispersion. An advanced Lagrangian approach is developed in this paper, modelling in particular plume rise and very low wind speed conditions. We developed a rigorous model including mass, energy and momentum conservation equations, coupled with constitutive equations related to physical and atmospheric properties. Special procedures were developed for source extinction (estimation of the final plume rise) and source merging before reaching the final plume rise. The numerical solution of the resulting system of algebraic equations has been carried out using Powell’s dogleg strategy. Comparison with a standard widely used Lagrangian code, as well as with experimental data-set, were carried out and showed good performance of the proposed approach.