Validation of the Canadian Atmospheric Dispersion Model for the CANDU Reactor Compex at Wolsong, Korea

Abstract

AECL is undertaking the validation of ADDAM, an atmospheric dispersion and dose code based on the Canadian Standards Association model CSA N288.2. The key component of the validation program involves comparison of model predicted and measured vertical and lateral dispersion parameters, effective release height and air concentrations. A wind tunnel study of the dispersion of exhaust gases from the CANDU complex at Wolsong, Korea provides test data for dispersion over uniform and complex terrain. The test data are for distances close enough to the release points to evaluate the model for exclusion area boundaries (EAB) as small as 500 m. Lateral and vertical dispersion is described well for releases over uniform terrain but the model tends to over-predict these parameters for complex terrain. Both plume rise and entrainment are modelled conservatively and the way they are combined in the model produces conservative estimates of the effective release height for low and high wind speeds. Estimates for the medium wind speed case (50-m wind speed, 3.8 ms-1) are conservative when the correction for entrainment is made. For the highest ground-level concentrations, those of greatest interest in a safety analysis, 82% of the predictions were within a factor 2 of the observed values. The model can be used with confidence to predict air concentrations of exhaust gases at the Wolsong site for neutral conditions, even for flows over the hills to the west, and is unlikely to substantially under-predict concentrations.