| Anaerobic Digestion Model 1 (ADM1) has been widely accepted by the industrial and research wastewater communities because it is capable of predicting the breakdown of complex substrates into methane and carbon dioxide gases, biomass, and inert components. A recent review of ADM1's successful implementations identified three major limitations, and this thesis addresses the most important of these points, which was to find more accurate regulators for the sugar fermentation pathways.;ADM1 fermentation pathways use the Monod equation, in conjunction with inhibition parameters, to predict the growth rate of fermentation microorganisms. Despite the use of these inhibition parameters, ADM1 formulas still predict growth in situations when growth is not thermodynamically possible. The technical consensus is that a thermodynamic regulation would provide the best prediction, and several attempts have been made to implement such control; some of which have been partially successful. However, before a precise tool like thermodynamics can be applied, the fermentation conditions must be known very accurately, so the ADM1 structure and calculation algorithms were examined first, to determine whether any improvements were possible.;ADM1 calculates the fermentation pH using temperature, but not activity corrections for the ionization constants. It was found that activity corrections were more significant than temperature corrections, and that these corrections must be combined for optimal results, so a simple algebraic pH algorithm was developed that combines both corrections. Similarly, the ADM1 gas transfer calculations were problematic in that they were not sufficiently accurate and introduced considerable mathematical stiffness, so a Particle Swarm Optimization algorithm was adapted for ADM1 application and was found to be both accurate and robust. Furthermore, a new type of biogas meter was developed to allow accurate and precise gas flow measurements.;To find better fermentation pathway regulators, a wide range of fermentation pathways were evaluated using thermodynamically based cell yield formulas. Cell yield values were suggested to be good predictors of pathway activity. Cell yields can also be used to construct instructive reaction networks, and as yields appear to correspond with operational experience, cell yields are proposed as regulators for the fermentation pathways of ADM1. |
