Analysis of Pollution Characteristics and Identification of Driving Factors in Soils at a Typical Petroleum Refinery

Petroleum refining is a high-risk process with significant potential for soil contamination. In this study, we applied Geodetector software to identify the spatial drivers of six contaminants of concern (COC): vinyl chloride (VC), 1,2,3-trichloropropane (TCP), benzo(a)pyrene (BaP), naphthalene (NAP), dibenzo(a,h)anthracene (DBA), and petroleum hydrocarbons (C10-C40) at a representative refinery in China. Soil sampling (n = 496) revealed distinct vertical distributions: VC/TCP penetrated deeper aquifers (1.5 m) than did polycyclic aromatic hydrocarbons (PAHs) (BaP/DBA < 1.0 m). Geographic detector (GD) analysis revealed significant drivers (q > 0.1, p < 0.05) and showed that storage tanks emerged as the dominant driver for VC (q = 0.182), BaP (q = 0.369), NAP (q = 0.653), and C10-C40 (q = 0.680) distributions, while process units predominantly governed DBA (q = 0.389) dispersion. Notably, TCP exhibited no significant association with the assessed factors. Factor interaction analysis revealed synergistic effects between wastewater facilities and storage zones (qmax = 0.912). This study provides critical insights to guide the remediation of retired refineries and implement precision prevention strategies in active facilities across China, while offering a reference for soil environmental management in refineries worldwide.