Assessment of offshore geo-hazard and failure prediction using AI-ML

PhD student

Farid Fazel Mojtahedi

Supervisors

Dr Negin Yousefpour	A/Prof Shiaohuey Chow	Prof Mark Cassidy

Project details

Offshore systems and subsea infrastructures are vulnerable to different natural geo-hazards, including turbidity currents, submarine landslides (and events that trigger tsunami), scour (seabed sediment mobility), and fluid flow. Such geo-hazards are the features that are commonly found in deep-marine settings. It is essential to early characterize, predict, and assess the risk and impacts of geo-hazards, particularly in deep remote fields, for operation maintenance of these systems and minimizing the failure, damage, and environmental risks. Assessment of geo-hazards is traditionally based on site investigation data that are exposed to considerable uncertainties for such factors as variable ground and water (current) conditions, dynamic nature of seafloor condition, lack of resolution, and gaps in survey data. Our understanding of Mass-transport complexes (MTCs) in the submarine and offshore environments have been improved nowadays as a result of emerging technologies. Nevertheless, there are still uncertainties about the way of evolution of the flow and volume behavior of MTCs during their translation, the factors controlling these changes, the relationship with their internal geometry and architecture, and the implications of MTC emplacement processes for the assessment of geo-hazard risk in sedimentary basins. Innovative predictive modelling and probabilistic approaches allow maximizing the interpretations, decreasing the required frequency of seafloor data acquisitions, and capturing the uncertainties. The purpose of the current research is providing innovative solutions for better understanding the offshore geo-hazards triggers and signs and assessing the risk of future geo-hazard events according to real-time, dynamic data obtained from the seafloor.