A power plant complex of an integrated iron and steel mill is normally driven by burnable by-products of the iron and steel making process and used to supply the steel mill with electricity and steam. The power plant complex cannot be operated at a steady state, due to the fact that the supply of the by-products and the demand of energy is very volatile. Because of the non-steady-state operation, all control variables have to be adjusted continuously. In the control stand, the software system "Leoben/Donawitz Energieeinsatzoptimierer" is used to ensure the operation of the plant complex at the economical optimum. This software system uses detailed thermodynamic models in order to find the economically best operation point of the power plant complex. The data basis for the optimization is provided by the current operation point and forecasts of the supply and demand of blast-furnace gas, converter gas, saturated steam and process steam. The point of operation, which is determined this way, will be shown in the control stand as a precise recommendation of the system operation. This thesis deals with the evaluation of the above-named software system. Therefore, a few, partially not connected, issues will be treated. These include the quality of the software system from the point of view of the stability of the optimization result and possible consequences of a reduction of the optimization period. Furthermore, the quality of the forecasts was analysed, a sensitivity analysis was applied and a concept of a reporting system introduced. Apart from the quality of the software system, the forecasts of the imbalance energy price, the converter gas and the process steam were also evaluated. It was found that the quality of the current forecasts has a big influence on the optimization result. The result of the sensitivity analysis is a prioritization of the control variables relating to the economical optimum. Therefore, the influence of a variation of control variables on the result was evaluated. This analysis showed that the two most important control variables are the fuel's thermal output and the feed water temperature. The introduced concept of a reporting system allows an objective comparison of the optimization result with the actual operation of the power plant complex. It can be used to determine a theoretically possible earnings increase. Consequently, it was possible to show that the optimization system with perfect forecasts and under current circumstances could generate an increase in profit by several million Euro per year.