Abstract
Industrial accidents can be triggered by natural hazards (as earthquakes, floods, lightning, and extreme temperatures) which can result in fires, explosions, or release of hazardous substances. These high impact accidents are called NaTech (Natural Hazards Triggered Technological Accident). NaTech can determine huge damages and their rapid growth in the latter years is also associated with climate change evolution. In this context, refineries are among the facilities majorly affected by NaTech and the presence of large amounts of hazardous materials makes easier to envisage potentially catastrophic scenarios. Risk assessments should include NaTech scenarios, and this should come with the availability of robust and efficient screening tools for safety engineering applications. Due to the complexity of extreme natural events, no general or reference models are officially available, leaving to the analyst the duty to estimate aspects such as probability of occurrence of an extreme natural event. Extreme events weather databases are also often incomplete, reporting data on past natural events with insufficient information for risk assessment purposes. This is particularly true for extreme winds. This paper applies a modified version of a literature reference model, based on the analysis of the occurrence of extreme winds on a Lat-Long quadrant classification. A grid choice dependence is shown, highlighting the variation of extreme winds likelihood of extreme winds on 3 Italian refineries located close to the sea. By changing the reference area for extreme winds analysis, the screening analysis leads to different results.
Keywords: hazards, NaTech, refineries, risk analysis, natural e
