Abstract
The aim of this study is to check the influence of the chemical composition of odorous and non-odorous gases on the odour concentration and on the sensor responses. Design of experiments (DOE) is used to plan mixtures composition for 5 compounds with different levels of concentrations. The synthetic gas experiments consist of preparation of samples based on the factorial design consisting of five compounds i.e., Trimethyl Amine, Hydrogen Sulphide, Methane, Limonene and n-Butanol. Pareto plots are examined to evaluate the main effects of each sensor on the odour concentration. In addition, the response of each individual gas sensor is examined as a function of the chemical composition as well as the odour concentration. The response surface methodology analysis (Peris et al, 2009) is carried out on the experimental design of the synthetic gases. The stepwise procedure was used to analyse the results which is an automated tool used in the exploratory stages of model building to identify a useful subset of predictors. All the combinations are carried out with step wise regression and without stepwise regression. Also BOX-COX transformations with different values of alpha are also considered during the development of the regression model. Afterwards, multivariate analysis is performed to check the performances of a sensors array regarding the fluctuation of chemicals concentration. Finally, real odorous samples from a municipal solid waste site are measured with the same array and the results are compared with the ones obtained with the synthetic odorous gases.
