84th AMS Annual Meeting

Tuesday, 13 January 2004: 4:00 PM
Dispersion Modeling of Accidental Releases in Russia
Room 611
Eugene Genikhovich, Voeikov Main Geophysical Observatory, St. Petersburg, Russia
Accidental releases of noxious chemicals in the atmosphere unfortunately happen frequently enough to be considered as a serious problem for the modern civilization. The main sources of accidents are industrial enterprises and storage facilities, where poisonous chemicals are generated, stored or used as a part of technological processes, as well as transportation of the noxious chemicals. For enterprises and facilities, the potential sources of accidental releases are usually well localized, and the possible amount of pollutants, which could be discharged in the case of accidents, is known in advance. For the emergency preparation, it is required in Russia to estimate the size of the zone (also called "the scale of contamination") around an enterprise, which could be affected by harmful pollutants in the case of accidental release at such level that certain precautions are to be taken (for example, people could be evacuated from this zone in the case of emergency).

A corresponding regulatory model was introduced in Russia in the beginning of 1990th and is still in effect. It is based on the dose (exposure) approach, i.e., on calculation of the doses, D, of different noxious pollutants, corresponding to 1-hour exposure time, Ċ, and comparison those with criteria limiting the permissible level of these doses. An equation for doses is derived by integration of the advection-diffusion equation over the time. It is assumed also that its non-stationarity resulted from the non-stationary accidental discharge of pollutants into the atmosphere. This equation, similar to those for the stationary concentration distribution, has been solved numerically and the results, analytically approximated, are used in computations.

The model considered is a majorant (upper-limit) one, i.e., it predicts an upper percentile of the probability distribution function of doses corresponding to a given set of governing meteorological parameters. In addition, the upper limit is taken over the location and orientation of the emitted plume/jet. Such an approach helps in reducing the level of the noise in the model predictions.

The model could be applied to different kind of sources (discharge of gases from tanks and pipes, evaporation of liquid spills and so on) either developing the emergency preparedness scenarios ("prognostic mode") or calculating the consequences in actual emergency response situations ("diagnostic mode"). Theoretical foundations of the model, its structure and results of its validation are discussed in this paper. A special attention is given to discussion of the methodology of modeling of atmospheric phenomena with a high level of stochasticity.

Supplementary URL: