The Real-Time Management of Air Quality on a Regional Scale

In the paper the problem of the real-time emission control is discussed, where the air quality forecast is based on the regional-scale model of air pollution transport. The problem formulation utilizes the optimal control technique for distributed parameter systems, which is integrated with the governing equations of air pollution dispersion process. The approach presented in the paper is sulfur-oriented and a dynamic model of sulphur oxides transport is the main forecasting tool used in the optimization algorithm. The control problem is formulated as on-line minimization of an environmental cost function, by the respective modification of emission level in the controlled emission sources, according to the changing meteorological conditions. A set of point-wise emission sources (power plants) with known location and emission characteristics is considered as the controlled object. The minimized objective function includes the air quality damage, as well as the cost of the controlling action. The emission control process is based on a gradient optimization algorithm. The back-trajectory technique (based on the adjoint equation solution) is used to calculate the gradient of the objective function in the main optimization procedure. The test computations have been performed for the set of the major power plants in the selected industrial region of Poland, and some determined meteorological episodes. 1. The forecasting model of air pollution transport The paper presents a possible application of regional-scale models of air pollution transport for supporting decisions related to environmental quality. In the sequel the problem of the real-time emission control on a regional scale is formulated and discussed. The approach presented below is based on utilizing a dynamic forecasting model of air pollution dispersion. Moreover, some general optimization methods are applied in formulation of the basic emission control problem as well as in the construction of the optimization algorithm. The most natural and common application of environmental models is the forecasting of dispersion of pollutants. Some air quality studies are also aimed at optimization, but numerous applications of optimization methods mainly occur in the design of monitoring networks. On the other hand, many important decisions in air pollution and environmental problems, which could be supported by the respective models, are directly made by decision makers. However, some optimization methods and recently developed, integrated environmental models (e.g. Cofa!a et al. 2004) give the possibility of implementation of complex air pollution control strategies. The problem of the regional-scale strategy for emission abatement in a set of the major power plants was discussed in (Holnicki and Ka!uszko 2002; Holnicki 2004). The solution of the last task is searched by the optimal selection of the desulphurization technologies for emission sources considered. From the viewpoint of the mathematical formulation, the above applications relate to the long-term strategies, and the main tasks are stated as the static optimization problems. This paper presents some possibilities of utilizing dynamic pollution transport models in the real-time control of air quality. The optimal control problem is formulated as on-line minimization of an environmental cost function, by modification of emission level in a set of the controlled sources, according to the changing meteorological conditions. The case of linear, single-component transport 1 Systems Research Institute of the Polish Academy of Sciences, Newelska 6, PL-01447 Warsaw, email: [email protected], Internet: http://www.ibspan.waw.pl EnviroInfo 2006 (Graz) Managing Environmental Knowledge Copyright © Shaker Verlag 2006. ISBN: 978-3-8322-5321-9