Dispersion of positron emitting radioactive gases in a complex urban building array: a comparison of dose modelling approaches

Abstract

A radiological assessment was carried out on the release of positron-emitting radioactive gases from a roof-level stack at a central London site. Different modelling approaches were performed to investigate the range of radiation doses to representative persons. Contributions from plume inhalation, gamma shine and immersion to effective dose were taken into account. Dry and wet surface deposition on the roof, and exposure from contamination on the skin of roof-workers, added only a mean 4.7% to effective dose and were neglected. A 1:200 scale model, consisting of the stack and surrounding buildings, was tested in a wind tunnel to simulate pollutant dispersion in the near-field region i.e. rooftop. Concentration field measurements in the wind tunnel were converted into effective dose, including for roof-workers installing glass cladding to the stack building. Changes in the building shape, from addition of the cladding layer, were investigated in terms of the near-field flow pattern and significant differences found between the two cases. Pollutant concentrations were also modelled using Air Dispersion Modelling System (ADMS) and the results used to calculate the effective dose using the same meteorological data set and source release terms. Sector averaged wind tunnel dose estimates were greater than the ADMS figure by approximately a factor of two to three. Different stack release heights were investigated in the wind tunnel and ADMS simulations in order to determine the best height for the replacement flue stack for the building. Other techniques were investigated: building wake models, modified Gaussian plume methods and uniform dilution into a hemispherical volume to show the wide variation in predicted dose possible with different approaches. Large differences found between simpler analytic approaches indicated that more robust radiological assessments, based on more complex modelling approaches, were required to achieve satisfactory estimates of radiation dose to representative groups in adjacent buildings and on the building rooftop.

Conversion of simulated radioactive pollutant gas concentrations for a complex building array into radiation dose

“…The concentration is then calculated on the basis of the entrained fraction into the recirculation zone on the leeward side of the building. This is the case in the accompanying paper (Gallacher et al 2016) where the length of the building in the direction of flow in the primary wind direction was greater than W/2. In the converse case, where the building is shorter, the flow is taken to separate from the leading edge of the building and reattach to the ground downwind of the building.…”

“…The WT directions were chosen on the basis of (a) an investigation of the features of the flow dynamics and pollutant dispersion for both unclad and clad building cases, and (b) to sample the concentration distribution over the full angular range, with sufficient samples, to enable a comparison with ADMS which uses the meteorological data set with wind from all directions. Further details are given in Gallacher et al (2016) the accompanying paper.…”

“…Since the doses at elevated positions on the roof of the building were a prime consideration, a second set of plume gamma-shine doses was calculated using an effective raised ground plane at the building height of H = 50 m for the unclad (existing) building and H = 53 m for the clad building. Dose tallies were then obtained for elevated receptor points at the EW roof height (Gallacher et al 2016), whereas standard ADMS output was used at the ground level. The following formulation was applied to estimate doses at intermediate height receptor points.…”

“…Building-wake model results are presented in Gallacher et al (2016) for the same release term as the other models. The wind direction is assumed to be the primary direction over the roof for this study.…”

“…The methods used to convert a number of concentration distributions into effective doses are described. The results of that analysis for a particular site are presented in an accompanying paper (Gallacher et al 2016).…”