Abstract
Urban mining increasingly concentrates on secondary raw materials derived from waste streams, among which municipal solid waste incineration (MSWI) residues are of growing importance. Fly ash, in particular, contains a wide range of trace and critical elements, yet its variability and internal associations remain insufficiently characterised. In the present study, 30 fly ash samples were collected in 2021 on a weekly basis from an MSWI facility in southern Poland. Eighteen elements (Ag, Al, Au, Ba, Co, Cr, Cu, Fe, Li, Mn, Mo, Ni, Pb, Pt, Sb, Sr, V and Zn) were quantified using ICPMS and ICP-OES, producing a robust dataset suitable for multivariate analysis. Data exploration comprised descriptive statistics, normality assessment with the Shapiro-Wilk test, outlier detection via Rosner’s test, correlation analysis using Pearson’s and Spearman’s coefficients, and hierarchical cluster analysis (HCA). Results show that most elements display moderate concentrations (10–1000 ppm), while Al, Fe, and Zn exceed 1000 ppm, and noble metals remain below 10 ppm. Strong positive correlations were observed between Sr and Li, as well as Fe, Ni, and Mo, while HCA consistently grouped Cr, Fe, Mo, and Ni into a stable cluster across methods. The most accurate dendrogram structure was achieved with average linkage (Euclidean or Manhattan), whereas Pearson-based distances produced sharper cluster boundaries. Importantly, the elemental concentrations determined in fly ash were systematically compared with both the geochemical background of the Earth’s crust and typical grades in natural ore deposits. This comparison revealed substantial enrichment in Zn, Pb, Sb, Ag, Au, and Pt relative to crustal averages, while only Zn (and occasionally Cu and Ag) reached concentrations approaching the lower thresholds of economically exploited ore deposits. These findings demonstrate the internal geochemical structure of MSWI fly ash and underscore its significance as a potential source of valuable elements within the framework of urban mining.