Waste as fuel in thermal waste treatment is characterised by a heterogeneous and fluctuating composition. Fractions such as paper, wood or organics have significantly different properties relevant for thermal treatment than plastics, composite materials or electronic waste. If the composition of the fuel mixture is available, however, the plant operation can be adapted. This allows primary and secondary measures to adjust to short-term fluctuations in pollutants in the raw gas in order to reduce emissions [1]. In addition, deposit-oriented cleaning can reduce or eliminate mineral deposits on steam generator heating surface walls, resulting in improved maintenance of the area of corrosion and erosion in the long term [2]. In the context of CO2 pricing, it is also possible to determine the fossil and biogenic content in waste from a known fractional composition.
To be able to react directly to changes in the fuel composition, it is necessary to continuously determine the composition during operation. Therefore, a calculation method has been developed consisting of two sequential models. Firstly, the elementary fuel composition (C, H, O), the water and ash content and the calorific value of the fuel are calculated using extended online balancing. The results are then used in the numerical fraction model to determine the fractional composition.
Theoretical validation and trials at a pilot plant with constant fuel composition have already proven the reliability of these models across a wide range. This paper demonstrates the adaption of the model to more realistic conditions of large-scale waste incineration with constant change in the waste composition on the grate. In the first step synthetic fuel mixtures from several mono-fractions with changing shares during one trial period were combusted in pilot scale. The measured data of the flue gas composition and the volume flows of flue gas and supply air were then used to supply the model with input data. To validate the results, the calculated elemental and fractional fuel compositions were compared with the actual mixtures used. After achieving good traceability of changes in the fuel composition data from a large-scale waste incineration plant were processed and plausible results were obtained.