REAL-TIME OPTIMAL STOCHASTIC-CONTROL OF POWER-PLANT RIVER HEATING

To meet environmental pollution standards, the operators of industrial plants must include environmental criteria in their operating objectives. This involves dealing with the variability and uncertainty of natural processes, and requires mathematical modeling to quantify operational effects. An example is given in which optimal operation of a once-through-cooling power plant is affected by the uncertainty in the hydrometeorological conditions that control the heat exchange between the river water surface and the atmosphere. Real-time optimal control of the power plant is based on simulations provided by a mathematical model of the fiver thermal regime. It is demonstrated that optimal results are achieved when correct information is provided on the uncertainty in the forecasts of the hydrometeorological conditions. Example study results arc given for the Joliet, Illinois, power station and hydrometeorological data for the drought period of July 1988.