One of the most important challenges of high-risk industries regards environmental risk management. In all sectors characterized by high-risk processes, failure can lead to catastrophic environmental events, so it is necessary to have a model capable of extracting useful information from the data collected and able to provide company managers with decision-making tools. In this work, a framework has been developed to manage environmental risk in a process industry. In order to analyse adverse environmental events, data provided by different sources and geographically dispersed repositories have been considered. A conceptual model, based on Association Rules (AR), has been developed to investigate the network of influences among data collected. Moreover, a Social Network Analysis has been used to represent the association rules, providing a complete overview of the factors' interaction and to identify communities of nodes in order to define local and global patterns and locate influential entities. To test the proposed approach and assess its impact on environmental management strategies, a medium-sized refinery case study has been analysed. The big data analytics approach proposed in this work, taking into consideration a wide set of objective and predictive variables, allowed the refinery managers to show new cause-effect correlations in refinery processes regarding adverse environmental event typology, immediate and root causes, refinery plant area involved in the adverse event, risk index and corrective actions. (C) 2019 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.