Doctoral Dissertations

Date of Award

5-1996

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Civil Engineering

Major Professor

Terry L. Miller

Committee Members

James Smoot, Lawrence Miller, Arijit Pakrasi, Wayne Davis

Abstract

There are several deficiencies in the capabilities of the United States Environmental Protection Agency's (EPA) guideline air dispersion models that limit their effectiveness in predicting concentrations for conditions that can have the greatest effect on plume dispersion. A new model was developed based on the EPA's Industrial Source Complex Short-Term (ISCST2) model. The new model is called ISCST2C to denote its capabilities of predicting refined concentrations in the cavity region of a building wake and on complex terrain according to the EPA's intermediate terrain policy. Three cavity algorithms are contained in the model: the Hosker formulation (from the EPA's screening dispersion model called SCREEN2); the Schulman-Scire formulation; and a Gaussian formulation that was developed as part of this research. The Gaussian formulation is consistent with existing EPA guidance for predicting concentrations due to plume downwash. The complex terrain algorithms are based on the EPA's rural complex terrain dispersion model called COMPLEX I and the EPA's urban complex terrain dispersion model called SHORTZ. The sensitivity of the ISCST2C model predictions to changes in input parameter values was analyzed by preparing curves of the normalized concentration (χu/Q) along the plume centerline versus downwind distance. At the end of the cavity and beginning of the near-wake, there are large discontinuities between the concentration predictions of the Hosker and Schulman-Scire cavity algorithms and the ISCST2 near-wake algorithm. There is continuity between the concentration predictions of the Gaussian cavity algorithm and the ISCST2 near-wake algorithm for plume heights no more than 1.2 times the building height. The three cavity algorithms of the ISCST2C model were tested by comparing predicted concentrations against measured concentrations from a field experiment. Tracer gas was released from a variety of heights for a broad set of meteorological conditions. Concentration measurements were made along an arc traversing the width of the cavity. The resulting cross wind concentration profiles tend to support the hypothesis of the Gaussian formulation that horizontal plume dispersion within the cavity generally follow a Gaussian distribution. The Gaussian cavity algorithm generally predicts measured concentrations in the cavity more accurately and more precisely than either the Hosker or Schulman-Scire algorithms.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS