Applications of fuzzy faulty tree analysis and expert elicitation for evaluation of risks in LPG refuelling station

A method is presented for analysis of reliability of complex engineering systems using information from fault tree analysis and uncertainty/imprecision of data. Fuzzy logic is a mathematical tool to model inaccuracy and uncertainty of the real world and human thinking. The method can address subjective, qualitative, and quantitative uncertainties involving risk analysis. Risk analysis with all the inherent uncertainties is a prime candidate for Fuzzy Logic application. Fuzzy logic combined with expert elicitation is employed in order to deal with vagueness of the data, to effectively generate basic event failure probabilities without reliance on quantitative historical failure data through qualitative data processing.The proposed model is able to quantify the fault tree of LPG refuelling facility in the absence or existence of data. This paper also illustrates the use of importance measures in sensitivity analysis. The result demonstrates that the approach is an apposite for the probabilistic reliability approach when quantitative historical failure data are unavailable. The research results can help professionals to decide whether and where to take preventive or corrective actions and help informed decision-making in the risk management process.     

Uncertainty reduction for improved mishap probability prediction: Application to level control of distillation unit

At all levels, the understanding of uncertainty associated with risk of major chemical industrial hazards should be enhanced. In this study, a quantitative risk assessment (QRA) was performed for a knockout drum in the distillation unit of a refinery process and then probabilistic uncertainty analysis was applied for this QRA. A fault tree was developed to analyze the probability distribution of flammable liquid released from the overfilling of a knockout drum. Bayesian theory was used to update failure rates of the equipment so that generic information from databases and plant equipment real life data are combined to gain all available knowledge on component reliability. Using Monte Carlo simulation, the distribution of top event probability was obtained to characterize the uncertainty of the result. It was found that the uncertainty of basic event probabilities has a significant impact on the top event probability distribution. The top event probability prediction uncertainty profile showed that the risk estimation is improved by reducing uncertainty through Bayesian updating on the basic event probability distributions. The whole distribution of top event probability replaces point value in a risk matrix to guide decisions employing all of the available information rather than only point mean values as in the conventional approach. The resulting uncertainty guides where more information or uncertainty reduction is needed to avoid overlap with intolerable risk levels.     

Uncertainty analysis framework for tubular connection sealability of underground gas storage wells

The seal failure of tubing and casing connections compromises underground gas storage well safety. This work proposes a systematic uncertainty analysis framework for connection sealability assessment. The framework covers reliability analysis and reliability sensitivity analysis and attempts to provide more effective support for the reliability design of connection seals. The reliability analysis introduces an adaptive Kriging with stopping criterion P-Monte Carlo simulation (AKP-MCS) method, which can provide a satisfactory estimate of failure probability with a small number of performance function evaluations. This metamodeling technology can effectively reduce the numerical efforts required for the reliability assessment of connection sealability. In the reliability sensitivity analysis, the refined metamodel obtained from the reliability analysis is coupled into a single-loop Monte Carlo simulation (MCS) method. The classifier attribute of this metamodel can meet the requirement of the single-loop MCS method to classify the signs of sampling points. This attribute enables sample matrices to be evaluated on this metamodel instead of the performance function, making the reliability sensitivity assessment more feasible. The proposed method is first demonstrated with four academic examples with promising results. Next, an illustrative tubular connection case is provided. The proposed scheme gives estimates of the failure probability and reliability sensitivity close to the classical model but requires less computational cost. The results of the analysis can provide useful information for the scheme decision-making and reliability optimization of connection seal design.     

System failure probability evaluation using fault tree analysis and expert opinions in intuitionistic fuzzy environment

In quantitative fault tree analysis of a system, exact failure probability values of components are utilized to calculate the failure probability of the system. However, in many real world problems, it is problematic to get precise and sufficient failure data of system components due to insufficient or imprecise information about components, changing environment or new components. A methodology has already been developed by employing fuzzy set theory for the system reliability evaluation by utilizing qualitative failure data of system components when quantitative failure data of components are inaccessible or insufficient. This paper extends the concept of fuzzy set to intuitionistic fuzzy set and proposes a novel approach to evaluate system failure probability using intuitionistic fuzzy fault tree analysis with qualitative failure data of system components. The qualitative failure data such as expert opinions are collected as linguistic terms. These linguistic terms are then quantified by triangular intuitionistic fuzzy numbers in form of membership function and non-membership function. Additionally, a method is developed for combining the different opinions of experts. To illustrate the applicability of proposed approach, a case study of the crude oil tank fire and explosion accident is performed. The obtained results are very close to the results from pre-existing approaches which confirm that the proposed approach is a more realistic alternative for the study of system reliability in intuitionistic fuzzy environment when quantitative failure data of system components are not known. To help decision makers for improving the security execution of the crude oil tank system, importance measures including Fussell-Vesely importance and cut sets importance are also executed.     

Dynamic Bayesian networks for reliability evaluation of subsea wellhead connector during service life based on Monte Carlo method

The subsea wellhead connector is a critical connection component between subsea Christmas tree and subsea wellhead for preventing the leakage of oil and gas in the subsea production system. Excited by cyclical loadings due to environmental forces and the other support forces, the subsea wellhead connector is prone to the failure, which could lead to the loss of subsea tree or wellhead integrity and even catastrophic accidents. With the Monte Carlo simulation method, this paper presents a reliability analysis approach based on dynamic Bayesian Networks, aiming to assess the failure probability of the subsea wellhead connector during service life. Take the driving ring component of the subsea wellhead connector as an example to demonstrate the reasonability of the proposed model. The generation data is processed by the transform between the numerical value and the state variable. Based on the stress-strength interference theory, the structure reliability of the driving ring with 96.26% is achieved by the proposed model with the consideration the aging of the material strength and the most influential factors are figured out. Meanwhile, the corresponding control measures are proposed effectively reduce the failure risk of the subsea wellhead connector during service life.     

Risk-based reliability assessment under epistemic uncertainty

Existing risk in production systems has a direct relationship with unreliability of these systems. Under such circumstances, the approach to maximize the reliability should be replaced with a risk-based reliability assessment approach. Calculating the absolute reliability for systems and complex processes, when we are not provided with any data on failure, is extremely complex and difficult. Until now, studies of reliability assessment have been based on the probability theory, in which the failure time is anticipated after determining the type of size distributions. However, in this paper, the researchers have developed an approach to apply the possibility theory instead of the probability theory. Instead of using absolutely qualitative methods, this new approach applies the Dempster–Shafer Theory. It is obvious when there are insufficient data; an index is needed to make a decision. Then, a novel method is proposed and used in a real case study in order to determine the reliability of production systems based on risk when the available data are not sufficient, helping us to make decisions. After calculating the failure probability and analyzing the assessment matrix and risk criteria, we may conclude that the failure risk of equipment is reduced while the system reliability is increased.     

Precursor-based hierarchical Bayesian approach for rare event frequency estimation: A case of oil spill accidents

Due to a scarcity of data, the estimate of the frequency of a rare event is a consistently challenging problem in probabilistic risk assessment (PRA). However, the use of precursor data has been shown to help in obtaining more accurate estimates. Moreover, the use of hyper-priors to represent prior parameters in the hierarchical Bayesian approach (HBA) generates more consistent results in comparison to the conventional Bayesian method. This study proposes a framework that uses a precursor-based HBA for rare event frequency estimation. The proposed method is demonstrated using the recent BP Deepwater Horizon accident in the Gulf of Mexico. The conventional Bayesian method is also applied to the same case study. The results show that the proposed approach is more effective with regards to the following perspectives: (a) using the HBA in the proposed framework provides an opportunity to take full advantage of the sparse data available and add information from indirect but relevant data; (b) the HBA is more sensitive to changes in precursor data than the conventional Bayesian method; and (c) using hyper-priors to represent prior parameters, the HBA is able to model the variability that can exist among different sources of data.     

Reliability data acquisition and evaluation in process plants

Based on the practical experience of assessing reliability data in two plants of the explosives industry, the organization and scope of the data collection, the component delimitation, the procedures of quality assurance and the data evaluation for their final use in PSA studies are presented. An example is given for the type of detection and repair, etc. The mathematical background for evaluating observed component lifetimes is discussed. Both frequentist and Bayesian methods are addressed. The resulting failure rate distributions and their approximations by log-normal distributions for several key components are presented. They were obtained using a Bayesian approach with a non-informative prior. Remarks on uncertainty, their treatment and a proposal for transferring data to plants other than the ones investigated conclude the presentation.     

贫数据条件下海底电缆故障概率评估方法*

为解决贫数据引起海底电缆失效概率评估的不确定性影响,实施有效的海底电缆故障风险管理,提出1种耦合模糊集理论、层次贝叶斯分析（HBA）和贝叶斯网络的海底电缆失效概率评估方法,识别海底电缆失效致因因素,梳理各因素之间的关联关系,并采用贝叶斯网络(BN)构建海底电缆失效模型;根据数据源特点将电缆失效因素分为数据完全缺失和具有稀少的先兆数据,采用模糊集理论(FST)计算完全没有可用数据的失效致因发生概率,通过HBA估计有稀少数据失效致因的发生概率;以失效致因发生概率为输入,通过贝叶斯网络实现海底电缆失效概率的动态评估。研究结果表明:FST-HBA-BN方法可以解决基本风险因素的数据稀缺问题,量化评估海底电缆失效概率,研究结果可为贫数据条件下的电缆失效风险管理提供支撑。     

Fuzzy dynamic risk-based maintenance investment optimization for offshore process facilities

A methodology for maintenance planning is developed which helps in improving the reliability of the components and safety performance in process facilities. This methodology helps design an optimum safety maintenance investment plan by integrating the optimization techniques and a fuzzy dynamic risk-based method. Intuitionistic Fuzzy Analytic Hierarchy Process (IFAHP) is applied to deal with uncertain data. The proposed approach employs multi-experts’ knowledge which helps to optimize the maintenance investments. A separator system in an offshore process facility platform is selected as a case study to demonstrate the application of the proposed methodology. A practical example in the separator system is surveyed and potential failures and Basic Events (BEs) are identified. Finally, a risk-based maintenance plan is provided for future safety investment analysis. The results indicate that the developed methodology estimates the risk more accurately, which enhances the reliability of future process operations.     

A novel Dempster-Shafer theory-based approach with weighted average for failure mode and effects analysis under uncertainty

Risk priority number (RPN) is a commonly-used prioritization method in failure mode and effects analysis (FMEA) for systemic reliability and safety study. However, conventional RPN confronts wide criticisms, due to the neglect of the uncertainties of experts' opinions. It is important to handle the conflicts among experts' multi-opinions which could be described in imprecise, incomplete or crisp forms in lack of knowledge. In this paper, a novel method is proposed to alleviate and/or eliminate counter-intuitive behaviors against conflicts among multi-opinions under Dempster-Shafer theory. Firstly, the abnormity test for experts’ opinions is undertaken to identify and discard those distinct judgments. Then, an approach to determining the subjective weights of experts is proposed, which is used in weighted average for multi-opinions before calculating RPNs. Finally, the risk priority evaluation on main engine crankcase explosion failure on-board ship is implemented to verify the feasibility of the proposed approach.     

Importance and uncertainty analysis of flue gas scrubber systems with extreme reliability parameter values

Although reliability analyses have been used to improve the reliability of industrial systems, generic reliability data from publications, such as component failure intensities and repair times, are used to calculate reliability measures instead of the real reliability data collected from the plant. One reason is that the repair history of the components is not well managed in the plant. In this work, the effect of extreme reliability parameter values on system reliability and unavailability is studied. To do this, importance and uncertainty analysis of the components of flue gas scrubber systems is carried out, and results calculated with the extreme reliability parameter values are compared with those with the mean reliability values of the systems. The different reliability parameter values can give us totally different ranks of the components critical to the reliability of the representative scrubber system. Consequently, the effort to establish a reliable database is emphasized to perform accurate reliability analysis of the system.     

Quantitative sensitivity and reliability analysis of sensor networks for well kick detection based on dynamic Bayesian networks and Markov chain

Kick is considered as an early warning sign to the blowout that is among the most undesired and feared accidents during drilling operations. Kick detection system is commonly used to timely identify the occurrence of a kick. The method commonly used for kick detection relies on the proper selection of monitoring indicators. A kick detection system should not only have very high accuracy but also maintain reliable over a long time. Different from the existing studies focusing on improving the detection accuracy, this paper presents a frame emphasizing on quantitatively analyzing and enhancing the reliability of the kick detection sensor networks. The dynamic Bayesian network (DBN) for the sensor networks is established that employs Markov chain to obtain the reliability degradation of measurement sensors over time. The proposed method is applied and evaluated by case studies to conduct reliability and sensitivity analysis for kick detection sensor networks. The reliability analysis results demonstrate that the proposed method can quantitatively analyze the reliability of a kick detection sensor networks consisting of various sensors over given time periods. The sensitivity analysis results indicate that the proposed method is effective in identifying the critical sensors that have the greatest effect on the reliability of one certain kick detection system. Based on the analysis results, optimized logical combination of sensors of a kick detection system can be achieved. An improved sensor network for the unreliable case was proposed and evaluated.     

Failure Mode and Effect Analysis using an integrated approach of clustering and MCDM under pythagorean fuzzy environment

Failure Mode and Effect Analysis (FMEA) is an effective risk analysis and failure avoidance approach in the design, process, services, and system. With all its benefits, FMEA has three limitations: failure mode risk assessment and prioritization, complex FMEA worksheets, and difficult application of FMEA tables. This paper seeks to overcome the shortcomings of FMEA using an integrated approach based on a developed Pythagorean fuzzy (PF) k-means clustering algorithm and a popular MCDM method called PF-VIKOR. In the first step, Pythagorean fuzzy numbers (PFNs) were used to collect Severity (S), Occurrence (O), and Detection (D) factors for failure modes to incorporate uncertainty and fuzziness into subjective judgments. Afterward, failure modes were clustered by developing a novel k-means clustering algorithm that accepts PFNs as input. Finally, the PF-VIKOR approach was used to analyze the ordering of cluster risks. The proposed approach was implemented in the dehydration unit of an Iranian gas refinery and the results were compared with the traditional FMEA. The findings showed the flexibility and applicability of the proposed approach in addressing real-world problems. This research provides two key contributions: (1) designing a PFN-based k-means clustering algorithm that tackles FMEA limitations and (2) using the PF-VIKOR method for prioritizing and evaluating failure mode clusters.     

Modelling of BP Texas City refinery accident using dynamic risk assessment approach

Process industries involve handling of hazardous substances which on release may potentially cause catastrophic consequences in terms of assets lost, human fatalities or injuries and loss of public confidence of the company. In spite of using endless end-of-the-pipe safety systems, tragic accidents such as BP Texas City refinery still occur. One of the main reasons of such rare but catastrophic events is lack of effective monitoring and modelling approaches that provide early warnings and help to prevent such event. To develop a predictive model one has to rely on past occurrence data, as such events are rare, enough data are usually not available to better understand and model such behavior. In such situations, it is advisable to use near misses and incident data to predict system performance and estimate accident likelihood. This paper is an attempt to demonstrate testing and validation of one such approach, dynamic risk assessment, using data from the BP Texas City refinery incident.Dynamic risk assessment is a novel approach which integrates Bayesian failure updating mechanism with the consequence assessment. The implementation of this methodology to the BP Texas City incident proves that the approach has the ability to learn from near misses, incident, past accidents and predict event occurrence likelihood in the next time interval.     

基于贝叶斯网络的一种事故分析模型

贝叶斯网络被认为是人工智能研究中不确定性知识表示和推理的重要工具。当前在系统安全领域中已开始运用贝叶斯网络技术进行故障诊断分析，然而故障只是诱发事故的因素之一，无法系统的评价事故背后的隐患，对事故后果的预测也甚少涉及。笔者将贝叶斯网络作为一种事故分析手段，在事故致因理论的基础上提出了一种基于危险因素-事故-事故危害的三层贝叶斯网络拓扑模型；阐述了网络模型层次间的因果关联关系、各层次的构成、节点的描述方法以及网络模型的构建方法；最后通过一个天然气球罐的分析案例验证了该模型分析方法的可行性和有效性。     

A risk-based approach to design warning system for processing facilities

Alarm flooding is a major safety issue in today's processing facilities. Important recommendations are available for alarm management; however, they are often violated in practice, especially in the alarm systems implemented through the distributed control system. An effective process alarm prioritization and management system is desired for a safe and effective operation of a process facility.In present work, authors address two main issues related to an alarm system – the reliability and the prioritization of the alarms. The main objective is to deal with the alarm-flooding problem in process facilities. A multi alert voting system based on sensor redundancy approach is proposed to improve the reliability. A quantitative risk-based alarm management approach is proposed to address the flooding issue. In the risk-based approach, an integrated model consisting of the probability (P), the impact (I) of the potential hazards, and the process safety time is proposed to prioritize these raised alarms.The proposed approach is further explained by a reactor system with pressure and temperature variable monitoring and controls, where the hazards associated with two alerts caused by over high pressure and over high temperature are analyzed and integrated with response time for alarms generation and prioritization.     

A framework to estimate the risk-based shutdown interval for a processing plant

Petrochemical plants and refineries consist of hundreds of pieces of complex equipment and machinery that run under rigorous operating conditions and are subjected to deterioration over time due to aging, wear, corrosion, erosion, fatigue and other reasons. These devices operate under extreme operating pressures and temperatures, and any failure may result in huge financial consequences for the operating company. To minimize the risk and to maintain operational reliability and availability, companies adopt various maintenance strategies. Shutdown or turnaround maintenance is one such strategy. In general, shutdown for inspection and maintenance is based on the original equipment manufacturer's (OEM) suggested recommended periods. However, this may not be the most optimum strategy given that operating conditions may vary significantly from company to company.The framework proposed in this work estimates the risk-based shutdown interval for inspection and maintenance. It provides a tool for maintenance planning and decision making by considering the probability of the equipment or system for failure and the likely consequences that may follow. The novel risk-based approach is compared with the conventional fixed interval approach. This former approach, characterized as it is by optimized inspection, maintenance and risk management, leads to extended intervals between shutdowns. The result is the increase in production and the consequent income of millions of dollars.The proposed framework is a cost effective way to minimize the overall financial risk for asset inspection and maintenance while fulfilling safety and availability requirements.     

Epistemic uncertainty in fault tree analysis approached by the evidence theory

Process plants may be subjected to dangerous events. Different methodologies are nowadays employed to identify failure events, that can lead to severe accidents, and to assess the relative probability of occurrence. As for rare events reliability data are generally poor, leading to a partial or incomplete knowledge of the process, the classical probabilistic approach can not be successfully used. Such an uncertainty, called epistemic uncertainty, can be treated by means of different methodologies, alternative to the probabilistic one. In this work, the Evidence Theory or Dempster–Shafer theory (DST) is proposed to deal with this kind of uncertainty. In particular, the classical Fault Tree Analysis (FTA) is considered when input data are supplied by experts in an interval form. The practical problem of information acquisition from experts is discussed and two realistic scenarios are proposed. A methodology to propagate such an uncertainty through the fault tree up to the Top Event (TE) and to determine the belief measures is supplied. The analysis is illustrated by means of two simple series/parallel systems. An application to a real industrial safety system is finally performed and discussed.     

Mitigation of operational failures via an economic framework of reliability,availability, and maintainability (RAM) during conceptual design

Abnormal process situation may lead to tremendous negative impact on sustainability, wellbeing of workers and adjacent communities, company's profit, and stability of supply chains. Failure of equipment and process subsystems are among the primary causes of abnormal situations. The conventional approach in handling failure-based abnormal situations has usually focused on operational strategies. Such an approach overlooks the critical role of process design in mitigating failure, while simultaneously considering the effects of such failure on process economic performance. The aim of this work is to introduce a systematic methodology that accounts for failure early enough during the conceptual design stages. Once a base-case design is developed, the methodology starts by identifying the sources of failure that are caused by reliability issues including equipment, operational procedures, and human errors for a given process system or subsystem. This allows for the identification of critical process subsystem(s) that are more failure-prone or cause greater downtime than other subsystems. Bayesian updating and Monte Carlo techniques are utilized to determine the appropriate distributions for the failure and repair scenario(s), respectively, in question. Markov analysis is used to determine the system availability. Next, the process revenue is described as a function of inherent availability. The effects of failures are incorporated into profitability calculations to establish an economic framework for trading off failure and profitability. In the proposed framework, the economic potential of alternative design scenarios is evaluated and an optimization formulation with the objective of maximizing incremental return on investment (IROI) is utilized to make a design decision. A case study on an ethylene plant is solved to demonstrate the applicability and value of the proposed approach.