Proposal of procedure to determine metals and metalloids background values in contaminated soils. Case study of a national interest site in South Italy

The present work concerns the development of a procedure to determine background values of metals and metalloids in soils of sites of national interest that are subjected to characterization and remediation. The proposed calculation procedure allows to select in metal-metalloid data-set, the representative parameters of site-specific pollution sources. It allows to identify anthropogenic concentration of substances in the soil resulting both natural pedo-geochemical content and moderate widespread immission in soil. The statistical analysis was performed in samples set of topsoil, soil and underground soil. Chemical tests were carried out on all parameters of metals and metalloids present in samples. This work has analized parameters, as As and Be, which have passed statistical test Shapiro e Wilk (W test) with contamination hot spot of anthropogenic, respect to a large area of substantially homogeneous values. Specifically, this study focuses on area adjacent to perimeter of site of national interest of Brindisi (South Italy). It is characterized by the presence of many process industries and of large power plants. Due to many human activities in this site and difficulty of access to sampling points, analysis was carried out starting from No. 30 sampling sites located in the province of Brindisi. This study was carried out to define a suitable criterion for determination of background concentrations in industrial area. It was found that cumulative frequency curve deviates greatly from straight trend which represents the trend of other measures. In all cases examined, inflection point was coincident with 90 degrees percentile. On the other hand, value distribution of arsenic concentration in subsoil shows that there are hot spots of contamination probably of anthropogenic origin, different from those found in a other large site that are result substantially homogeneous. However, it is not possible to exclude these points from sample distribution and, therefore, they are estimated. Having to provide a representative background value of concentrations of metals present in area was used a conservative approach based on analysis of discontinuity points in cumulative frequency curve. In this way, background value about contamination anthropogenic source corresponds to previous value at point of discontinuity in cumulative frequency curve (90 degrees percentile). Thus, background conditions would be identified by graphic of concentrations distribution, from the zero point to the point of inflection.