|
Abstract
|
The paper proposes a generalized model of hazard systems. It seeks to demonstrate the underlying commonality of the elements and structures which make up hazard systems, and to render explicit certain concepts which are implicit in earlier models.
The model comprises a source with a potential to emit harmful energy or matter or some combination of the two, and one or more receptors which may be affected by such harmful emissions. It takes account of the fact that an inanimate receptor may itself, on absorbing energy, then become a source.
Except where the source and the receptor are in direct contact, the model provides for one or more successive transmission paths which are carriers of the harmful emission from source to receptor(s). It also provides for the presence of barriers. The emission, which may be quantified at any point in the system by the integral of its flux (intensity) with respect to time, may be attenuated by the transmitting medium, and/or by barriers. This, after taking account of the relevant impact area of the receptor (typically the projected area), yields the dose incident upon the receptor(s). The prediction of harm from this dose requires the incorporation of a transform, denoted in the paper by TD → H, which cannot be expressed in purely physical terms but which is derived from observation or from experiment. This provides a probabilistic relationship between the level of incident dose and the level of harm inflicted.
The paper subjects the purely physical elements of such hazard systems to dimensional analysis. It demonstrates how, in some cases, emissions are quantified in practice by indices with dimensions which differ from those of an integral of a flux with time.
Though the model is conceived as a perfectly general one which is not limited by the nature of the hazard or by the time-scale of the events, the authors claim that it can be applied with particular relevance to the potentially harmful acute emissions of energy or of matter, or of both in combination, which occur from time to time in the process industries.
Based upon the above concepts, and using existing methodologies for the estimation of the frequency with which the potential of a source may be realized, algorithms are proposed for calculating individual risk and its derivative, societal risk. |