Entwicklung und Einsatz eines interaktiven Biogas-Echtzeit-Simulators

Ecological and economical efficiency of biogas plants significantly depends on applied process control strategies and implemented automations. The complex biological process of biogas production takes place in a technical system that is mainly responsible for process control and stabilisation. To be able to design and operate such a technical system, one has to have a thorough understanding of biological, biochemical, physicochemical and technical mechanisms occurring herein. Hence, education and training of planning and operative personnel is of great importance. Because, on an industrial level, experiments cannot be carried out without the danger of malfunctions, process modelling and process simulation provide good alternatives. Within the scope of this thesis, the process of anaerobic digestion was modelled with regard to biological, biochemical, physicochemical, reactor and peripheral sub-processes. On the basis of these mathematical models, an interactive training simulator of the process of anaerobic digestion was developed that can be used for a variety of applications. These include designing and testing of process control strategies, application in academic and industrial education and utilisation as a tool for gaining enhanced understanding of the process. From the wide range of models described in the literature a rather reduced model published by Bernard et al. [2001] was chosen and adapted to the requirements of a real-time simulator and the objective of this thesis. The simulator was tested during a variety of process states. It can be run in real-time or in an accelerated mode. In both operation modes the simulator worked stable and reacted robustly to manual and automated control inputs. The modular design of the simulator-i.e. the strict separation into the four sub-models described above-allows for easy replacement and adaptation of every sub-model individually. Furthermore, the influence of technical factors on the biological process could be evaluated comprehensively. By doing so, the importance of correctly interpreting measuring data became evident. Especially during unstable process transients, assessment of the current process state becomes difficult since biological effects are superimposed by metrological effects like sensor dynamics, intermixture and dilution. It is important to consider these effects when designing process control strategies. In industrial biogas plants such effects can hardly be distinguished from biological or chemical effects. Thus, a simulator can contribute to gaining a deeper understanding of these processes. The four sub-models were fitted both individually and together to experimental data. Inhibiting effects such as an accumulation of organic acids were well reproduced by the model. Biogas yield and composition were modelled accurately for all considered substrates. Parameterisation of the biological sub-system was performed using only a small number of state variables as it can often be observed at real plants. With the help of the training simulator developed in this thesis-including the underlying mathematical sub-models-a tool was developed that can be applied in education as well as for process design und optimisation. It can be adapted to industrial-scale plants and substrates with reasonable effort