Recent major accidents in the offshore oil and gas (O&G) industry have showed inadequate assessment of system risk and demonstrated the need to improve risk analysis. While direct causes often differ, the failure to update risk evaluation on the basis of system changes/modifications has been a recurring problem. Risk is traditionally defined as a measure of the accident likelihood and the magnitude of loss, usually assessed as damage to people, to the environment, and/or economic loss. Recent revisions of such definition include also aspects of uncertainty. However, Quantitative Risk Assessment (QRA) in the offshore O&G industry is based on consolidated procedures and methods, where periodic evaluation and update of risk is not commonly carried out. Several methodologies were recently developed for dynamic risk analysis of the offshore O&G industry. Dynamic fault trees, Markov chain models for the life-cycle analysis, and Weibull failure analysis may be used for dynamic frequency evaluation and risk assessment update. Moreover, dynamic risk assessment methods were developed in order to evaluate the risk by updating initial failure probabilities of events (causes) and safety barriers as new information are made available. However, the mentioned techniques are not widely applied in the common O&G offshore practice due to several reasons, among which their complexity has a primary role. More intuitive approaches focusing on a selected number of critical factors have also been suggested, such as the Risk Barometer or the TEC2O. Such techniques are based on the evaluation of technical, operational and organizational factors. The methodology allows supporting periodic update of QRA by collecting and aggregating a set of indicators. However, their effectiveness relies on continuous monitoring activity and realtime data capturing. For this reason, this contribution focuses on the coupling of such methods with sensors of different nature located in or around and offshore O&G system. The inheritance from the Centre for Integrated Operations in the Petroleum Industries represents the basis of such study. Such approach may be beneficial for several cases in which (quasi) real-time risk evaluation may support critical operations. Two representative cases have been described: i) erosion and corrosion issues due to sand production; and ii) oil production in environmental sensitive areas. In both the cases, dynamic risk analysis may employ real-time data provided by sand, corrosion and leak detectors. A simulation of dynamic risk analysis has demonstrated how the variation of such data can affect the overall risk picture. In fact, this risk assessment approach has not only the capability to continuously iterate and outline improved system risk pictures, but it can also compare its results with sensor-measured data and allow for calibration. This can potentially guarantee progressive improvement of the method reliability for appropriate support to safety-critical decisions.
Fostering real-time climate adaptation: Analyzing past, current, and forecast temperature to understand the dynamic risk to Hawaiian honeycreepers from avian malaria