Characteristics and Health Risk Assessment of Groundwater Metal(loid)s in Different Geomorphological Units of the Coal-Fired Power Cluster in the “Ji”-Shaped Bend of the Yellow River

Groundwater metal(loid)s and associated health risks across geomorphic units in the Yellow River Ji-bend coal-fired power cluster are of growing concern. We profiled metal(loid)s in groundwater from 34 wells spanning hilly, piedmont-alluvial, and Yellow River alluvial units; applied self-organizing map (SOM) clustering to delineate pollution patterns; apportioned sources with APCS–MLR; and quantified source-specific human health risk using Monte Carlo simulation (MCS). The aquifer is generally weakly alkaline and overall metal(loid)s concentrations are low; however, the alluvial plain shows localized enrichment of arsenic (As) and manganese (Mn), with levels exceeding those in the hilly unit. Using the Heavy Metal Pollution Index (HPI), Nemerow Index (NI), and Contamination Degree (CD), 35.3% and 8.8% of samples were classified as moderate and high pollution, respectively, with pronounced spatial heterogeneity. APCS–MLR attributes approximately 45% of the total mass to geogenic sources, followed by industrial, traffic, and agricultural inputs. Monte Carlo simulation (MCS) identified As as the dominant health-risk driver, with the carcinogenic risk for adults (2.32 × 10^-4) exceeding the acceptable threshold (1 × 10^-4), while the non-carcinogenic risk is generally higher in children (HQ = 3.31) than in adults (HQ = 2.36). Source-specific risk apportionment indicates that >80% of total risk is attributable to geogenic sources, dominated by As. The framework supports risk-based zoning and management by linking spatial patterns, sources, and health risk across geomorphic units.