LAUSR

The accuracy and reliability of volatile organic compound (VOC) emission data are essential for assessing emission characteristics and their potential impact on air quality and human health. This paper describes a new method for determining VOC emission data by multipoint sampling from various process units inside a large-scale refinery. We found that the emission characteristics of various production units were related to the raw materials, products, and production processes. Saturated alkanes accounted for the largest fraction in the continuous catalytic reforming and wastewater treatment units (48.0% and 59.2%, respectively). In the propene recovery unit and catalytic cracking unit, alkenes were the most dominant compounds, and propene provided the largest contributions (57.8% and 23.0%, respectively). In addition, n-decane (12.6%), m,p-xylene (12.4%), and n-nonane (8.9%) were the main species in the normal production process of the delayed coking unit. Assessments of photochemical reactivity and carcinogenic risk were carried out, and the results indicate that VOC emissions from the propene recovery unit and catalytic cracking unit should be controlled to reduce the ozone formation potential; in addition, alkenes are precedent-controlled pollutants. The cancer risk assessments reveal that 1,2-dibromoethane, benzene, 1,2-dichloroethane, and chloroform were the dominant risk contributors, and their values were much higher than the standard threshold value of 1.0 × 10-6 but lower than the significant risk value defined by the US Supreme Court. Based on the VOC composition and a classification algorithm, the samples were classified into eight main groups that corresponded to different process units in the petroleum refinery. In conclusion, this work provides valuable data for investigating process-specific emission characteristics of VOCs and performing associated assessments of photochemical reactivity and carcinogenic risk in petrochemical refineries.