Developing a framework for dynamic risk assessment using Bayesian networks and reliability data

Abstract

Process Safety in the oil and gas industry is managed through a robust Process Safety Management (PSM) system that involves the assessment of the risks associated with a facility in all steps of its life cycle. Risk levels tend to fluctuate throughout the life cycle of many processes due to several time varying risk factors (performances of the safety barriers, equipment conditions, staff competence, incidents history, etc.). While current practices for quantitative risk assessments (e.g. Bow-tie analysis, LOPA, etc.) have brought significant improvements in the management of major hazards, they are static in nature and do not fully take into account the dynamic nature of risk and how it improves risk-based decision making In an attempt to continually enhance the risk management in process facilities, the oil and gas industry has put in very significant efforts over the last decade toward the development of process safety key performance indicators (KPI or parameters to be observed) to continuously measure or gauge the efficiency of safety management systems and reduce the risks of major incidents. This has increased the sources of information that are used to assess risks in real-time. The use of such KPIs has proved to be a major step forward in the improvement of process safety in major hazards facilities. Looking toward the future, there appears to be an opportunity to use the multiple KPIs measured at a process plant to assess the quantitative measure of risk levels at the facility on a time-variant basis.ExxonMobil Research Qatar (EMRQ) has partnered with the Mary Kay O'Connor Process Safety Center -Qatar (MKOPSC-Q) to develop a methodology that establishes a framework for a tool that monitors in real time the potential increases in risk levels as a result of pre-identified risk factors that would include the use of KPIs (leading or lagging) as observations or evidence using Bayesian Belief Networks (BN).In this context, the paper presents a case study of quantitative risk assessment of a process unit using BN. The different steps of the development of the BN are detailed, including: translation of a Bowtie into a skeletal BBN, modification of the skeletal BN to incorporate KPIs (loss of primary containment (LOPC), equipment, management and human related), and testing of the BBN with forward and backward inferences. The outcomes of the dynamic modeling of the BN with real time insertion of evidence are discussed and recommendation for the framework for a dynamic risk assessment tool are made.

Why Do Workers Take Safety Risks?—A Conceptual Model for the Motivation Underpinning Perverse Agency

“…The model provides a general categorization of human errors into slips, lapses, mistakes, and violations. This has been useful and further developed and applied into the crew resource management (CRM) framework e.g., [70], and the barrier or the bowtie method [116][117][118][119]. The limitation of the Swiss Chess model and its derivatives is the inability to explain why the human factors occur in the first place.…”

Bayesian network-based failure analysis of fire safety barriers in floating LNG facility

“…It is obvious that all SBs failure probability decreased from an independent model to the dependent mode. This is due to uncertain causal relationships between MOFs and Regenerator failure X 30 Operation without work permit X 58 Fire sprinkler failure X 3 Reboiler failure X 31 Failure to follow work permit X 59 Inadequate firefighting in given time duration X 4 Pump failure X 32 Hot surface shielding not available X 60 Long delay in fire fighting X 5 Valves and joints failure X 33 Hot surfaces shielding failed X 61 Firefighting not performed X 6 Compressor failure X 34 Lightening protection facility is not installed X 62 Flame alarm failure X 7 Cooler failure X 35 Deflector damaged X 63 Flame detection sensor failure X 8 HEX failure X 36 Lightening rod damaged X 64 Flame detection controller failure X 9 Flash drum failure X 37 Inadvertent burner flare trip failure X 65 Inadequate detection coverage of flame X 10 N-column failure X 38 Flame detector failure X 66 Operator did not detect fire X 11 Pipe failure X 39 Flame detector not available X 67 Manual fire alarm failure X 12 Tank leakage X 40 Inadequate detector coverage X 68 Manual activation alarm switch not available X 13 High pressure X 41 Manual inspection of ignition source failure 20 Material defects X 48 Long delay in operator response X 76 Inadequate maintenance procedures X 21 Welding detects X 49 Manual ESD valve failure X 77 Inadequate inspection and testing procedures X 22 Sensor failure X 50 Operator awareness failure X 78 Inadequate safety programs X 23 Overrun X 51 ESD sensor failure X 79 Poor risk assessment and decisionmaking process X 24 Alarm failure X 52 ESD controller failure X 80 Inadequate leadership at critical time X 25 Controller detection failure X 53 ESD valve failure X 81 Poor organizing and reporting structure X 26 Inadequate detection coverage of gas X 54 Inadequate fire and explosion resistance The occurrence probabilities of possible consequences were also computed using both BN models after setting the failure of RPB as evidence. Table 7 provides both prior and posterior probabilities of consequences from dependent and independent BN.…”

“…Esmaeil et al employed BT-BN to perform a dynamic risk analysis of natural gas stations [12]. Kanes et al [13] utilized a BN and reliability data to develop a dynamic risk assessment approach. Meng et al [14] developed an integrated method by using the decision-making trial and evaluation laboratory (DEMATEL-BN) for assessment of severe accidents induced by oil and gas leakage.…”

Risk Control Technology for Water Inrush during the Construction of Deep, Long Tunnels

“…A Bayesian network, as an important mathematical analysis method, offers the following advantages [32,33]: (i) e causal relationship in the network structure diagram is clear and suitable for complex systems with multiple factors (ii) It can not only predict the changing trend of events by forward reasoning technology, but also allow rapid diagnosis of accident disasters by reverse reasoning to find the causes of accidents (iii) Expert knowledge and empirical data can be combined, which is effective for the risk prediction of large projects (iv) e system can update the analysis results in time according to the changes in root node information and judge the changing trend of the system by predicting changes in the results.…”