Statistical Evaluation of an Exhaust Gas Mixture for Photocatalytic Reactor Test-Station Improvement

The paper contains the description of the statistical analysis performed in order to correlate concentration of hazardous volatile organic compounds from the BTX group with non-hazardous constituents of the gas mixture. The gas composition and concentration of particular compounds were determined by numerous test performed in order to evaluate the effectiveness of a photocatalytic reactor in terms of VOCs reduction. Such a reactor is dedicated to the automotive industry and is set to be a member of a ventilation system. The photocatalysis is used to purified polluted air aspirated from the surrounding into the vehicle cabin. Therefore, the composition of a gas mixture was determined in various conditions resembling the operation of a standard ventilation system. The concentration of mixture constituents measure at a particular test condition was correlated with each other with aid of partial component analysis (PCA) in order to evaluate the power with which the individual components affect the model. Subsequently, the partial least squares analysis (PLS) was performed in order to select the non-hazardous compounds of a gas mixture which can represent the concentration of hazardous Benzene Toluene and Xylene (BTX). Lastly a curve fitting analysis was performed in order to find a ready equation with which the concentration of hazardous compounds can be predicted.