Mitigation of dense gas releases in buildings: use of simple models

• Published in: Journal of hazardous materials 2000-01, Vol.71 (1), p.129-157
• Main Authors: Deaves, D.M., Gilham, S., Spencer, H.
• Format: Article
• Language: English
• Subjects: Air Pollution, Indoor - adverse effects; Air Pollution, Indoor - prevention & control; Gas dispersion; Gases - pharmacokinetics; Humans; Mitigation; Models, Theoretical; Risk Assessment; Safety Management - methods; Ventilation
• ISSN: 0304-3894; 1873-3336
• DOI: 10.1016/S0304-3894(99)00076-X

When an accidental release of a hazardous material is considered within a safety case or risk assessment, its off-site effects are generally assessed by calculating the dispersion of vapour from the site. Although most installations handling flammables will be in the open air, many types of plant, particularly those handling toxics, are enclosed, partly to provide some form of containment and hence to mitigate the effects of any release. When such a release occurs within a building, the gas or vapour will undergo some mixing before emerging from any openings. The degree of mixing will depend upon the building geometry and the nature of the ventilation, which in turn may be modified by the leak. This situation is considered in this paper, with specific application to calculating the rate of release of a dense vapour from a building. All the calculations presented are based upon simple zone modelling, such that the region occupied by the vapour is assumed to be well mixed, and, in the isothermal case, either its concentration or its depth increases as it is fed by the gas leak. Transfer of air or gas/air mixture through the building openings is estimated by use of standard ventilation calculation methods. For the non-isothermal case, a preliminary model is presented in which it is assumed that there is complete mixing throughout the building and no wind-driven ventilation effects. A moderate release of chlorine is used as an example, and results are shown of the effects of various ventilation possibilities on the release rate to the atmosphere. In addition, comparisons are given between model results and experimental data, demonstrating the level of confidence which can be placed in the models, and also identifying areas where there is scope for further improvement.