A new physically-based numerical model to simulate flood triggered oil spills

Floods are the most common natural disaster, and recent studies suggests that their frequency and magnitude will increase due to climate change. Factors such as demographic growth, urbanization, and land consumption contribute to heightened vulnerability for structures, infrastructure, and populations, elevating the risk of cascading incidents. In this context, flood events can lead to multiple simultaneous releases of hazardous materials, causing severe harm to both the environment and human health. In these cases, the term Natech accident is used, referring to industrial accidents triggered by natural events, for which a multi-risk approach is required. Natech accidents caused by floods require particular attention, as the high velocities of water can rapidly transport pollutants to areas far from the point of emission. The need to focus on this issue is further justified by the fact that many chemical and petrochemical plants are in flood-prone areas, making them particularly vulnerable to the risk of failure following a flood. In the context of emergency management, having access to rapid-response models for assessing the fate and transport of spills is crucial for evaluating their trajectory and for planning recovery interventions. Additionally, these models are key for generating risk maps for various spill scenarios. Within Natech risk management, particular attention is given to oil spills in water, as they introduce additional complexity due to the unique behaviour of this substance in water and their potential toxicity, as well as the risk of cascading events (i.e. environmental contamination, fires, explosions). The need to develop specific models for simulating oil spills in floodwaters is particularly important, as the existing literature provides numerous models for offshore spills, but knowledge regarding fluvial systems is still limited.

This study presents the initial results from the implementation of oil spill routing within the CAESAR-LISFLOOD flood inundation model, which addresses the challenge of solving a simplified shallow water equations using a straightforward numerical approach. This results in a model that is computationally efficient while still being grounded in a solid physical framework. The model is enhanced with a module that simulates the dispersion of oil in floodwaters, accounting for the key processes that influence oil movement in a river system. This implementation allows the model to track the behaviour of an oil slick after a spill in areas with complex topography. It provides valuable insights into the dynamics of the spill, the changes in the slick’s thickness over time, and the extent of the affected area. The model was tested on a case study in Italy, where several simulations were performed for multiple spill scenarios, demonstrating the model’s effectiveness, its ability to accurately simulate the oil spill propagation, as well as its computational efficiency.