Towards new simplified methodologies on source term estimation and associated uncertainties from accidental airborne releases

Abstract

The estimation of an airborne hazardous release and its associated uncertainty from a point source is a necessity for assessing the associated risk and taking possible protective actions. The problem here is to be able to make such an estimation, based on the availability of a rather limited number of sensors measuring pollutant/agent concentrations with a given time resolution. The challenge is to not only provide reliable outcome but also timely results even in operation level. A radically new concept is under development by (a) making full use of the real detailed inflow data and sensors concentration signals, (b) performing in the customisation – pre-emergency – phase only, a quite limited number of steady-state flow and dispersion simulations under reference inflow and release conditions and (c) projection of the steady state/reference results to real atmospheric and release conditions via appropriate novel scaling approaches based on experimental evidence and theory and (d) generating a low computational time demand tool, that can be handled by the operator on duty, even in the extreme cases of very complex topographies. In this study first experimental evidence is provided from a wind-tunnel flow and dispersion experiment in a typical urban area supporting the validity of the whole approach. In addition, the present approach is simultaneously assessing each sensor's signal suitability to be used for source quantification.