| The objective of this research was to identify physical, chemical, and biological characteristics of landfills bioreactors that may affect the production of methane from this type of facilities. Furthermore, the objective was to develop a mathematical model to relate methane production from bioreactors to measurable parameters, which are routinely available for operating facilities. Finally, the reliability of the models was evaluated by comparing raw field data versus data predicted by the models.; A total of 526 data samples from 12 different landfills were analyzed. Of this total, 17 are from field scale operations, while 509 are from laboratory scale experiments. Raw data from pilot scale studies were not available during the data collection phase of the research. The data was obtained from different bibliographical sources. Papers from journals, magazines, and federal publications were read and studied to verify the reliability of the data acquired. Data sampling was conducted at different geographic locations around the world. For instance, data from Australia, Japan, Korea, Spain and United States are presented in this report.; Data analysis proceeded by applying five different variable selection statistical methods as well as multiple regression analyses to evaluate the simultaneous effects of 40 variables on the landfill gas production. These analyses concluded with models for cumulative gas production and gas production rates for different categories of landfills. These categories covered conventional landfills and bioreactors as well as categories for different geographic zones and different landfill scales.; Results indicate that refuse methane potential, the mass of dry refuse and the cell depth are significantly associated with the gas produced from refuse. In addition, cumulative gas production seems to be a better indicator of gas production when compared versus gas production rate. The model explained 97% of the observed cumulative gas production for 526 different data points according to its R2 value. The reliability of the model was verified by carrying out significant-difference tests for both set of data, observed and predicted, for a 5% of significance level. Verification of the model indicates high precision and accuracy when raw field data is compared with the one predicted by the model. |
