Methodological alternatives for calculation of enrichment factors used for assessment of topsoil contamination

Purpose

The research is aimed at comparison of different enrichment factors (EFs) calculated for potentially harmful chemical elements (PHEs) basing on topsoil data from Klaip?da with geological differences between two areas, choice of the reference elements (REs), consideration of the expediency and efficiency of normalisation, the influence of PHE and RE background or upper threshold values on the estimated contamination level of 50 schoolyards and the arising problems, i.e. which PHE should be normalised.

Materials and methods

Composite topsoil samples taken in 50 schoolyards and 28 football fields from 0 to 10-cm depth were air-dried, sieved to <1 mm, milled and analysed for total contents of major elements and As, Ba, Cr, Cu, Mn, Mo, Ni, Pb, Sn, U, V and Zn by energy-dispersive X-ray fluorescence and of Ag and Co by optical atomic emission spectrophotometry. Simple EFs in schoolyards were calculated normalising by concentration factor (CF) of Fe, while complex EFs normalising by weighted average of Al, K and Ti concentration factors (WACFs).

Results and discussion

The variability of RE is shown with significant differences between sandy western and loamy eastern areas and lower differences between schoolyards and football fields. The highest effectiveness of normalisation, i.e. the number of sites where WACF > 1 or CF > 1, is when RE background values for schoolyards differentiated for each area are selected. The formulas for the ratio of different EFs demonstrate that (1) for Co, Ba, Mn, Mo, Pb and Cu, the upper threshold-type EFs exceed the background-type EFs; (2) there is the tendency to obtain lower EFs for the greater part of PHE when using differentiated background of PHE and RE. The problems of estimation of the background of some PHE and the expediency of their normalisation are discussed.

Conclusions

The recommendations are (1) to explore the variability of RE in the urban territory and to find out if there are areas with significant differences in RE content, in this case try to estimate the differentiated background values of RE and PHE, because respective EFs are usually lower; (2) for more effective normalisation, to use background values of RE which correspond to the object of assessment; (3) not to use upper threshold-type EFs; (4) to be careful with interpretation of EFs for those PHEs which have significant negative correlation with RE.