Modelling landfill degradation behaviour

Landfill gas is a term used to describe the mixture of gases created by the decomposition of waste within a landfill. A landfill gas generation model is a tool that describes in simple terms the complex changes that occur during decomposition of waste in a landfill. The majority of simple models take into account microbial growth and decay only. The more complex models (known as component models) include not only microbial growth and decay, but also liquid, gas and heat transport through the waste, settlement and the chemical reactions that take place within the landfill. The majority of these models have been verified using laboratory scale data only and are highly parameter hungry. Landfill operators are very unlikely to use such complex models; since the majority of the required parameters are expensive and difficult to obtain at full-scale, or the historical data required to obtain them has never been collected. A better model for full-scale landfills is required based on the more sophisticated component models, but with fewer parameters, and able to be used by operators. To work towards developing such a model, an investigation is required examining existing models to determine the effect of the data set scale on the predictions from models and the sensitivity of the input parameters. An explanation of the processes involved in the generation of landfill gas, and a comparison of existing landfill models is presented. This thesis compares the predictions from five landfill models: two state of the art component models, a commonly used simple microbial growth and decay model and two simple Monod degradation models developed by the author. The scope of this thesis is the examination of the degradation component of landfill models and not subsequent transport and settlement components. The five models were selected to examine the parameter sensitivity of the various models at three different scales: laboratory scale: HPM2 challenge, UK (0.15 m3), field test cell: Brogborough test cells, UK (20,000 m3) and full-scale: Narre Warren landfill, Australia (3.5 x106 m3). Detailed information on the three data sets is provided. Relationships are developed between the critical model input parameters and the cumulative gas generation. These developed relationships will aid in the selection of input parameters for other landfills. The most sensitive input parameters required for the prediction of measured values were found to be the growth rates for simple models and the lytic and/or hydrolysis parameters for the more complex models. The selected models were assessed to see if they were able to predict measured gas generation at field-test cell and full-scale; in particular the component models, which had been developed and calibrated only at laboratory scale. Only one of the five models, LDAT, was unable to predict all three data sets.