API PUBL 4491 VOL 1-1989

Numerical and advanced similarity models that are used to estimate the concentrations for heavier-than-air gas (HTAG) releases have only been tested against field data sets collected in flat homogeneous terrain. The models appear to predict reasonbly well for that situation. Most HTAG releases for the petroleum industry will, however, be in industrial-type settings where the roughness elements are large and generally heterogeneous within the plant and then homogeneous or heterogeneous beyond the plant boundary. Because the models have not been evaluated for "real-world" applications of interest, the American Petroleum Institute's Air Modeling Task Force initiated this study to evaluate the effects of homogeneous and heterogeneous surface roughness on HTAG dispersion and to provide a data base for testing and refining numerical and advanced similarity models.

An initial set of tests was conducted to document the validity of the wind tunnel simulation method. One set of these tests demonstrated that the results in the wind tunnel were repeatable and that the concentrations measured in the wind tunnel were steady state averages with a corresponding full scale averaging time from 2 to 20 minutes. Another set of tests demonstrated that the dispersion of wind-tunnel-simulated neutrally buoyant plumes under stable and neutral stratification compared favorably with the dispersion observed to occur in the atmosphere for similar surface roughness conditions. Tests were also conducted to simulate two of the Thorney Island continuous release trials. The two trials simulated did not have any obstructions present, the wind speeds were low and the atmospheric stratification was neutral for one case and stable for the other. The wind tunnel measurements with a neutral and stable atmospheric boundary-layer simulation compared well with respective field observations.