A statistical approach to the air pollution problem of chimney emission

Abstract

A method is described that not only gives one single concentration value but the expected concentration frequency distribution at an arbitrary point in the vicinity of the emitting chimney. Unlike the method of gifford (1959) the concentration frequency distribution obtained comprises the full range of meteorological conditions at a given place and not only the short term variations in concentration during one particular meteorological situation. In making the principal mathematical formulation of the method no assumptions whatsoever are needed about the mechanism of dispersion. In the practical application on the other hand, full use is made of the detailed knowledge of the dispersion process. First a “basic case” is treated, the chief characteristics of which are: isolated stack, situated on a flat surface of uniform roughness; sampling time about an hour. The following information is found necessary for solving a concrete problem of this kind: 1. (1) plant data, viz. rate of emission, gas volume and temperature, chimney height and exit diameter; 2. (2) geographical site data, viz. roughness length; 3. (3) meteorological data, viz. wind direction frequencies and statistics of “dispersion categories” (stability and wind speed). Applicable formulae are discussed and also how statistics of “dispersion categories” can be obtained by evaluating data from radiosonde stations. Certain deviations from the basic case are also treated in some detail; limited mixing height, the effect of large heat content on the dispersion parameters, the effects of buildings and topography, ground level release and sampling times other than an hour. A thorough test of the described method is presented. Nine months statistics of 2-hr SO 2 concentration (2602 values) measured at a point situated 750 m from a sulphuric acid plant are compared with theoretically obtained statistics. The result strongly supports the method presented.