| dc.description.abstract | In the present study, a monitoring facility for hazardous gases was set up at Texas
Tech University's Wind Engineering Research Field Site. The major objective of this study
was to use the downwind experimental data acquired at the field site to develop a
methodology for estimating point source emissions rates. Experiments were carried out by
releasing controlled quantities of hydrogen sulfide (H2S) and ammonia (NH3) in the fields
adjoining the field site. The downwind concentration of gases released during the studies
was measured using a Single Point Monitor. The meteorological tower at the site was
used to record meteorological data such as wind speed and direction, ambient temperature
and relative humidity.
An empirical correction to the Pasquill-Gifford model was determined which
conservatively estimated emission rates. The estimation of atmospheric stability was
studied using various methods published in literature. Terrain specific parameters such as
friction velocity and surface roughness were determined. This approach was validated
only for direct downwind sampling and neutral atmospheric stability. The dispersion
parameters in the Gaussian model were then modified using experimental data.
Finally, a novel method based on the use of artificial neural networks was also
developed to model gaseous atmospheric dispersion. This approach offers significant
potential because it does not require the accurate determination of the numerous variables
which traditional models require. However, the neural network must be trained over the
span of variables of interest. | en_US |
