Model-free measurements of calorific content and ash content of mixed garden wastes using a bomb calorimeter

Abstract

Two of the most important properties of a material to determine its suitability as a waste-to-energy feedstock are its calorific content and ash content (AC). In the literature, it has been widely suggested that using the results from ultimate and proximate analysis to determine the former through empirical models is an accurate method that saves time and reduces labor. In this work we evaluated the suitability of these models for mixed garden waste using samples collected from two geographical distinct municipal waste disposal sites. It was found that when the models, with parameters given in the literature, are used, the errors in the estimates of higher heating values (HHV) could be higher than acceptable limits. Using linear regression, we determined the model parameters for our specific feedstocks based on the measured elemental, AC and the HHV values from a bomb calorimeter that reduced this error, with AC being the best predictor. However, there were still points with errors over 15%, indicating that these empirical models may not be suitable for the highly varying biomass characteristics typical of mixed green waste feedstocks and that direct measurement through bomb calorimetry still provides the most accurate results. We did find though, that the residues from bomb calorimetry could be used to determine the ash content of mixed garden waste as a time-, labor- and energy-saving measure. Standards for ash content measurements involve the use of a furnace. Here we compared two standard procedures to each other and to the bomb calorimeter residue. Statistical analysis explicitly showed for the first time that we are aware of, that the residue of mixed garden waste from bomb calorimetry tests provided a reliable estimate of ash content over a large range with agreement within 10 % to standards and therefore can be used in place of the standards thereby significantly reducing the time and cost of the tests while getting more accurate results for HHV compared to the models suggested in the literature.