Atmospheric Diffusion of Particulate Matter Released from an Elevated Continuous Source

Robert E. Stewart University of Florida, Gainesville

Search for other papers by Robert E. Stewart in
Current site
Google Scholar
PubMed
Close
Restricted access

Abstract

Dustfall simulation experiments were conducted to estimate the ground-level deposition pattern. Batches of different size glass microspheres (50–200 μ) were continuously released into the atmosphere from an elevated point source. Samplers were placed at ground level along arcs downwind from the stack and the particles were subsequently identified and counted. Measurements of wind speed, gustiness and temperature gradient were made. The data and data obtained by previous investigators were compared with theory.

Two mathematical models were used in the comparison: a three-dimensional classical model by Denisov (comparable to Rounds's two-dimensional solution), and a three-dimensional statistical model by Csanady. Both models assume steady-state conditions, horizontal homogeneity, linear growth relationships for the plume dimensions, and complete particle retention at the ground. It was found that the deposit pattern can be predicted reasonably well by the statistical model but should be restricted to values of f/u and f/σw (u is the mean wind speed, f the particle terminal velocity, and σw the rms value of the vertical velocity fluctuations of the turbulent field) greater than 0.05 and 1.0, respectively. At lower values of these ratios, the statistical model overestimates the maximum deposition and underestimates its location. The classical model, in general, underestimates the maximum deposition by a factor of approximately 1.5.

Abstract

Dustfall simulation experiments were conducted to estimate the ground-level deposition pattern. Batches of different size glass microspheres (50–200 μ) were continuously released into the atmosphere from an elevated point source. Samplers were placed at ground level along arcs downwind from the stack and the particles were subsequently identified and counted. Measurements of wind speed, gustiness and temperature gradient were made. The data and data obtained by previous investigators were compared with theory.

Two mathematical models were used in the comparison: a three-dimensional classical model by Denisov (comparable to Rounds's two-dimensional solution), and a three-dimensional statistical model by Csanady. Both models assume steady-state conditions, horizontal homogeneity, linear growth relationships for the plume dimensions, and complete particle retention at the ground. It was found that the deposit pattern can be predicted reasonably well by the statistical model but should be restricted to values of f/u and f/σw (u is the mean wind speed, f the particle terminal velocity, and σw the rms value of the vertical velocity fluctuations of the turbulent field) greater than 0.05 and 1.0, respectively. At lower values of these ratios, the statistical model overestimates the maximum deposition and underestimates its location. The classical model, in general, underestimates the maximum deposition by a factor of approximately 1.5.

Save