Process-specific emission fingerprints of volatile organic compounds and source apportionment using Positive Matrix Factorization at a full-scale petroleum refinery

This study presents an integrated application and interpretation of volatile organic compound (VOC) emission fingerprints from major petroleum-refinery processes and evaluates their seasonal influences on fenceline air quality. Process-unit measurements revealed distinctive VOC signatures dominated by light alkanes, alkenes, BTEX aromatics, and oxygenated species such as 2-butanone and acetone, enabling clear differentiation among distillation, catalytic cracking, reforming, and sulfur-recovery operations. Seasonal fenceline observations showed marked spatial gradients shaped by refinery activities and additional inputs from adjacent industrial facilities. Integrated Positive Matrix Factorization (PMF)-based interpretation identified refinery fugitive emissions as the predominant contributor to perimeter VOCs (27.6%), exceeding mobile sources (17.5%) and process-stack emissions (14.2%). The integrated VOC fingerprint–PMF framework provides enhanced interpretation of refinery VOC sources and supports targeted emission control and leak detection in complex industrial environments. • Integrated refinery measurements enabled unified interpretation of process and fenceline VOC emissions. • Dominant light hydrocarbons defined alkane/alkene signatures across refinery units. • Key aromatic and oxygenated VOCs supported source discrimination. • PMF results identified fugitive tank and equipment leaks as the dominant VOC source. • The integrated fingerprint–PMF framework enhances interpretation of refinery VOC sources and supports targeted control.