Sampling method to determine a household waste composition variance

Knowledge of waste composition is of crucial importance for waste management forecasting. Composition is usually specified by average content of glass, paper, organic matter etc. In this paper a sorting method and its application to variance determination is described. A variation coefficient and a confidence interval are then calculated. From these two parameters an appreciation of the dispersion and the uncertainty associated with the mean values can be derived. In the case studied, the variation coefficients calculated were between 0.10 and 0.50 depending on the class of waste. Analysis of confidence intervals shows that reliability is good for low-abundance components such as, for example, aluminium, iron and plastics. The influence of practical constraints on the theoretical guidelines is also discussed.