Oil sludge depository assessment using multivariate data analysis

Oil-containing industrial wastes tend to accumulate and present a growing environmental danger. This is of particular concern in certain areas of Russia. For effective processing of depositories, the wastes' physico-chemical properties and depository characteristics should both be taken into account. Representative sample sets were collected from fifty four depositories of different age, origin, and location in Samara region and analyzed using multivariate data analysis: Principal Component Analysis (PCA) and Partial Least-Squares (PLS) regression. PCA results provide a better understanding of the internal data structure, i.e. variable correlations and groupings. Based on the PCA results, a new approach to the classification of oil sludge depositories has been suggested. Another practically important task of site assessment has been solved by PLS regression modeling. The method has been successfully applied to the accurate estimation of the depository processing profitability for a specific site. ► We examined 54 oil sludge storage sites of various type, origin, age and composition. ► We suggest using both sludge and depository attributes for the site assessment. ► The PCA model reveals data structure that aids in the waste treatment selection. ► PLS regression on the suggested descriptors is well-suited for storage evaluation.