A mathematical model for composting kinetics

Composting plays an important role in waste management schemes and organic farming, as the compost produced enables reuse of organic matter and nutrients. Modern composting plants must comply with strict environmental regulations, including gas emissions such as nuisance odors. Designing composting plants to meet these requirements using current trial-and-error strategies is too costly and time consuming and poor performance and failure are too often the result. Mathematical reactor models can serve as an essential tool for faster and better process designs, system analysis, and operational guidance. However, all reactor models developed so far are based on empirical kinetic formulations, restricting the generality and thus applicability of the results. To achieve greater generality for design and analysis, a mechanistic model for composting kinetics is needed. Any mechanistic model is based on a number of assumptions and must be validated against experiments. To make validation possible, all model parameters must be identifiable. A parameter is identifiable if one can uniquely determine its value from the data at hand. The objective of this thesis is to develop a mechanistic kinetic model of the composting process whose parameters are all identifiable. This thesis has been structured in three main parts. The first part, "dimensional identifiability analysis," is concerned with the use of dimensional analysis of parameter identifiability. Together with a proposed modified deductive modeling strategy, this part of the thesis is a methodological contribution to modeling of relatively complex systems with limited available measurements. The second part, "the single particle model," is focuses on the development and validation of a theoretical model for the aerobic degradation of a single waste particle. This theoretical model gives insight into the processes occurring within a composting waste particle. An analytical solution of this model, containing only identifiable parameters, is both derived and validated. The third part of the thesis, "the distributed model," deals with the development, validation and application of a kinetic model for a waste consisting of a distributed range of waste particle sizes. The model is based on a distribution function describing the particle size distribution and the previously developed analytical solution to the identifiable single particle model. The distributed model is validated and is used to analyze aeration requirements, compost quality and compost quantity for a new composting reactor concept. This model application shows the advantages of the distributed model relative to previous first order models for reactor design and analysis.