Charakterisierung von Stoffströmen vorbehandelter Siedlungsabfälle in Deponiebioreaktoren (Volume 67) (English shop)

Preview

Extract, PDF (460 KB)
Table of Contents, PDF (100 KB)

Previous investigations into the degradation behaviour of municipal solid waste in landfill bioreactors frequently revealed problems in comparing different reactor sizes as well as in simulations aimed at scaling up to landfill scale. In order to counteract these problems, three reactors of different sizes (1.3 l; 18.8 l; 78.8 l) with geometrically similar reactor design and identical aspect ratio were designed, constructed and operated in the present work for the characterisation of mechanically-biologically pretreated municipal solid waste, its degradation behaviour and the associated biogas production in reaction systems of differing size.

To ensure the greatest possible comparability, a representative waste sample was used, consisting of a fine fraction with a particle size < 5 mm and a coarse fraction with a particle size > 5 mm, which was further classified into biodegradable and poorly biodegradable substances as well as inert materials. Investigations into the degradation behaviour of the waste at 40 °C showed that, once gas formation had stagnated, an artificial increase of the temperature to 60 °C brought about a renewed onset of gas synthesis. Furthermore, the effects of different water contents (40, 60, 70 % w/w) on gas emissions were evaluated, whereby in all experiments carried out and at every reactor size a clear optimum of biogas synthesis was demonstrated at a water content of 60 % w/w.

With the aid of an Artificial Neural Network (ANN), the chemical and biochemical processes responsible for gas emission could be simulated with very high accuracy in the different reactor sizes and under the differing influencing parameters. It was shown that it is possible, using one and the same network, to simulate gas formation at the different scales with the aid of experimentally obtained transient on-line and off-line process data.