• Domino effects are modeled as double-weighted directed graphs. • New relationships are developed for ‘distance’ and ‘closeness centrality’ • Firefighting strategies are identified by minimizing the closeness of burning units. • The method is easy to implement as it is independent of fire spread probabilities. Firefighting at process plants is considered the last line of defence in fire protection. As such, the success and efficiency of firefighting strategies directly impact the safety, integrity, and damage cost of process plants and adjacent on-site and off-site structures and facilities. Suppressing burning units while cooling exposed units is the most intuitive strategy, which is also practiced at process plants, for preventing fire spread and potential domino effects, assuming that firefighting resources (equipment, crew, water, etc.) are sufficient to deal with all the burning and exposed units. However, when firefighting resources are insufficient, models are required to help firefighters identify the most critical units and prioritize viable firefighting strategies accordingly. In the present study, after modelling domino effects as a double-weighted directed graph, i.e., a graph with weighted nodes and edges, it is shown that effective firefighting strategies can be identified by minimizing the closeness centrality of burning units. The developed method not only results in consistent firefighting strategies compared with the risk-based methods in the literature, but also requires fewer input parameters (basically, heat flux and asset monetary values), alleviating the need for calculating the fire escalation probabilities.
Khakzad et al. (Fri,) studied this question.