On the influence of uncertainty in risk analysis based on digital models

In the Oil and Gas (O&G) industry, subsea assets are inspected by remotely operated vehicles, in a complex and time-consuming task. The decision to inspect a subsea asset is surrounded by high costs, consequences of failure and uncertainties on degradation estimates. This scenario drives the use of risk-based maintenance strategies, which aims to identify when and at which points a procedure, like pipe wall thickness inspection, is required. Digital models of the key degradation phenomena that occur in subsea assets can be used for integrity management, enhancing planning assertiveness and increasing the reliability of decisions. However, the output of a digital model contains uncertainties that influence the probability of failure and the risk assessment. These uncertainties are related to measurement uncertainties, process variations and to the representativeness of the model. This paper addresses how these uncertainties influence the risk assessment and impact the reliability of a risk-based decision.