University of Western Macedonia
Πρόγνωση ατομικής έκθεσης με μοντέλα υπολογιστικής ρευστομηχανικής
Abstract
dc:descriptionOne of the key problems, which is directly related to the safety of human life, is the emissions of hazardous substances into the atmosphere from point sources, caused mainly by deliberate or accidental atmospheric releases. In many cases, the releases are short and the concentrations are high and there is a need to reliably predict the maximum expected individual exposure at short time intervals. The classic method that appears in the literature for the prediction of the maximum expected concentration/exposure is based on statistical distributions. According to this method, a known probability density function is used to model the concentration. The maximum expected concentration/exposure is estimated from the probability distribution by using predefined confidence intervals. This methodology presents certain weaknesses which are analyzed in this thesis. An alternative method for the prediction of the maximum individual exposure is the use of empirical models. A sophisticated empirical model that appears in the literature relates the expected maximum exposure with parameters such as the exposure time, the mean concentration, the fluctuation intensity and the turbulent autocorrelation time scale of concentration. The present empirical model is simple and overcomes the limitations of probabilistic models. A key advantage of this model is that it can expand the capability of computational, atmospheric dispersion models to estimate the expected maximum exposure for a wide range of time intervals. A review of the available dispersion models based on the open literature indicates that the Computational Fluid Dynamics (CFD), Reynolds Averaged Navier Stokes (RANS) models are ideal for the incorporation of this empirical model, due to the reliable prediction of the wind flow and pollutant dispersion in complex configurations such as urban environments. The need for the reliable prediction of the dispersion parameters of the empirical model (concentration mean, variance and time scale) poses new challenges in the RANS methodology. In the present thesis, a new CFD-RANS approach is introduced for the atmospheric dispersion from point sources. The key feature of the new approach is the capability to predict the concentration time scales not only as a function of the hydrodynamic time scale but also as a function of the pollutant travel time. The new approach is applied for the estimation of the dissipation time scale of concentration variance and the maximum individual exposure for short time intervals. The pollutant travel time is estimated in the CFD grid by introducing the new 'radioactive tracer method'. The new methodology for the prediction of the maximum individual exposure is evaluated in two steps. Firstly, the empirical model is evaluated by using the extensive concentration database of the MUST (Mock Urban Setting Test, moderate urban complexity) field experiment that includes data of all atmospheric stability classes. Furthermore the uncertainty of the model’s parameters is estimated. The results of this step strengthen the reliability of the empirical model. The reliability of the overall methodology is examined by performing CFD simulations on selected trials of the field experiments MUST and FLADIS (Research on the dispersion of two-phase flashing releases) under neutral atmospheric conditions. The new CFD-RANS approaches and the empirical model for the prediction of the maximum individual exposure are incorporated into the computational code ADREA, a three-dimensional CFD code appropriate for simulations of atmospheric flow and dispersion in complex geometries. The solution of the RANS equations is based on the finite volume method. In this thesis, the closure of the equations is obtained by using a new turbulence model that appears in the literature, which is appropriate for simulations requiring the extension of the grid at heights where the Coriolis force is not negligible. The variables that need to be evaluated include the concentration mean and standard deviation and the maximum individual exposure for various time intervals. The new methodology is evaluated mainly by using validation metrics and scatter plots. Overall, the numerical results show good agreement with the experimental data according to the state-of-the-art values presented in the literature for the validation metrics. Moreover, the results indicate that the present methodology is capable of dealing properly with complex transient dispersion phenomena.
Degree
thesis:*- Grantor dc:publisher
- University of Western Macedonia
- Year dc:date
- 2013
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Efthimiou, Georgios
- Ευθυμίου, Γεώργιος
Subjects
dc:subject × 20- Υπολογιστική ρευστομηχανική
- Μέθοδος ραδιενεργούς ανίχνευσης
- Χρόνος μεταφοράς ρύπου
- Τυρβώδεις χρονικές κλίμακες
- Ατομική έκθεση
- Διακυμάνσεις συγκέντρωσης
- CFD modeling
- Radioactive tracer method
- Pollution travel time
- Turbulent time scales
- Individual exposure
- Concentration fluctuations
- Φυσικές Επιστήμες
- Άλλες Φυσικές Επιστήμες
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Μηχανολόγου Μηχανικού
- Natural Sciences
- Other Natural Sciences
- Engineering and Technology
- Mechanical Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/34624