An Evaluation of Municipal Solid Waste Composition Bias Sources

The University of Central Florida (UCF) was contracted by the Florida Center for Solid and Hazardous Waste Management (FCSHWM) to develop a well-defined methodology for conducting municipal solid waste composition studies. This methodology must account for the statistical variations in waste composition, be economical and practical in implementation, and build on a consensus of waste management professionals. This paper identifies possible sources of bias in waste composition study results and provides guidance for future planning of local waste stream composition analysis. To accomplish this objective, a composition study was designed and implemented for Marion County, FL, in fall 1996. The potential sources of concern investigated in detail were sample weight and contamination. The methodology developed by UCF is statistically valid and if widely implemented would provide a better representation of the waste stream. Lack of contamination adjustment is a major contributor to error in the waste stream analysis and should be accounted for in the methodology. For sample sorts using a large number of categories, sample size may be a contributor to bias. This likelihood for bias can be reduced by increasing the sample weight to at least200 kg, particularly when sorting commercial loads or reducing the number of categories.