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.     

Technical,human, and organizational factors affecting failures of firefighting systems (FSs) of atmospheric storage tanks: Providing a risk assessment approach using Fuzzy Bayesian Network (FBN) and content validity indicators

Most risk assessment methods have problems such as uncertainty, static structure, and lack of validation. Also, in most of these studies, less attention has been paid to human, managerial, and organizational issues. Therefore, this study proposes a risk assessment method based on the Fuzzy Bayesian Network (FBN) to prevent failure of firefighting systems (FFSs) in the atmospheric Storage Tanks of a Petrochemical Industry. The first stage of the study is the development of a fault tree (FT) and investigation of basic events (BEs). In this study, content validity indices and brainstorming technique were used to validate the FT structure and reduce the uncertainties of Completeness, Modeling, and Parameter. After determining the probability of basic events (BEs) by the expert team opinions and fuzzy logic, events were transmitted to the Bayesian Network (BN) and then analyzed with deductive and inductive reasoning, followed by sensitivity analysis in the GeNIe software. Finally, results of a case study in the Atmospheric Storage Tanks of the Methanol Floating Roof of a Petrochemical Industry showed that FBN simulation and FT validation could provide a practical way to determine FFSs probabilities, identify impactful events, and reduce the above uncertainties. Also, taking account of hidden factors of events, such as organizational and managerial factors, can help managers to prevent FFSs in tanks.     

3-Parameters SPW technique: A new method for evaluation of target safety integrity level

Introduced by IEC-61508 standard, safety integrity levels (SIL) have been used for assessing the reliability of safety instrumented functions (SIF) for protection of the system under control in abnormal conditions. Different qualitative, semi-qualitative and quantitative methods have been proposed by the standard for establishing target safety integrity levels amongst which “Risk Graph” has gained wide attention due to its simplicity and easy-to-apply characteristics. However, this method is subject to many deficiencies that have forced industry men and experts to modify it to fit their demands. In this paper, a new modification to risk graph parameters has been proposed that adds more flexibility to them and reduces their subjective uncertainties but keeps the method as simple as before. Three parameters, namely severity (S), hazard avoidance probability (P), and demand rate (W) are used instead of former four parameters. Hence, the method is named SPW. The outcome results of this method can be directly converted to probability of failure on demand (PFD) or risk reduction factor (RRF). The proposed method has been tested on an example case that has been studied before with conventional risk graph and LOPA techniques. The results show that new method agrees well with LOPA and reduces costs imposed by conservative approximations assumed during application of conventional risk graph.     

Risk analysis of Chongqing urban rail transit network

Urban rail network safety is a critical sector of urban public safety. However, there is no uniform standard for the safety evaluation of the urban rail network. This paper presents a novel methodology by integrating a multilevel decision tree with a fuzzy analytical approach to enhance urban rail network safety. The proposed methodology overcomes serious limitations such as subjectivity in the data and independence of the variables in decision-making processes. The proposed methodology is applied to the risk evaluation of the selected Chongqing rail transit lines and the Expo Line. The risk analysis is considered using the field data collected from these transit lines. The applied case studies confirm the general applicability of the methodology and the multilevel decision tree network. The main risk factors identified for the Chongqing rail traffic system are the terrorist threat, emergency management, and aging infrastructure which need to be investigated as a priority to mitigate risk associated with these infrastructures.     

Safety evaluation of leak in a storage tank using fault tree analysis and risk matrix analysis

The work presented in this paper used a quantitative analysis of relevant risks through the development of fault tree analysis and risk analysis methods to aid real time risk prediction and safety evaluation of leak in a storage tank. Criticality of risk elements and their attributes can be used with real time data to predict potential failures likely to occur. As an example, a risk matrix was used to rank risk of events that could lead to a leak in a storage tank and to make decisions on risks to be allowed based on past statistical data. An intelligent system that recognizes increasing level(s) and draws awareness to the possibility of additional increase before unsafe levels are attained was used to analyse and make critical decisions. After a visual depiction of relationships between hazards and controls had been actualized, dynamic risk modelling was used to quantify the effect controls can potentially have on hazards by applying historical and real-time data into a probabilistic model. The output of a dynamic risk model is near real-time quantitative predictions of risk likelihood. Results from the risk matrix analysis method mixed with RTD and FTA were analyzed, evaluated, and compared.     

Application of fuzzy logic to explosion risk assessment

Safety and health of workers potentially being at risk from explosive atmospheres are regulated by separate regulations (ANSI/AIHA in USA and ATEX in the European Union). The ANSI/AIHA does not require risk assessment whereas it is compulsory for ATEX. There is no standard method to do that assessment. For that purpose we have applied the explosion Layer of Protection Analysis (ExLOPA), which enables semi-quantitative risk assessment for process plants where explosive atmospheres occur. The ExLOPA is based on the original work of CCPS for LOPA taking into account an explosion accident scenario at workplace. That includes typical variables appropriate for workplace explosion like occurrence of the explosive atmosphere, the presence of effective ignition sources, activity of the explosion prevention and mitigation independent protection layers as well as the severity of consequences. All those variables are expressed in the form of qualitative linguistic categories and relations between them are presented using expert based engineering knowledge, expressed in the form of appropriate set of rules. In this way the category of explosion risk may be estimated by the semi-quantitative analysis. However, this simplified method is connected with essential uncertainties providing over or under estimation of the explosion risk and may not provide real output data.In order to overcome this problem and receive more detailed quantitative results, the fuzzy logic system was applied. In the first stage called fuzzification, all linguistic categories of the variables are mapped by fuzzy sets. In the second stage, the number of relation between all variables of analysis are determined by the enumerative combinatorics and the set of the 810 fuzzy rules “IF-THEN” is received. Each rule enables determination of the fuzzy risk level for a particular accident scenario. In the last stage, called defuzzification, the crisp value of final risk is obtained using a centroid method. The final result of the risk presents a contribution of each risk category represented by the fuzzy sets (A, TA, TNA and NA) and is therefore more precise and readable than the traditional approach producing one category of risk only. Fuzzy logic gives a possibility of better insights into hazards and safety phenomena for each explosion risk scenario. It is not possible to receive such conclusions from the traditional ExLOPA calculation results. However it requires the application of computer-aided analyses which may be partially in conflict with a simplicity of ExLOPA.The practical example provides a comparison between the traditional results obtained by ExLOPA and by fuzzy ExLOPA methods.     

Handling and updating uncertain information in bow-tie analysis

Bow-tie analysis is a fairly new concept in risk assessment that can describe the relationships among different risk control parameters, such as causes, hazards and consequences to mitigate the likelihood of occurrence of unwanted events in an industrial system. It also facilitates the performance of quantitative risk analysis for an unwanted event providing a detailed investigation starting from basic causes to final consequences. The credibility of quantitative evaluation of the bow-tie is still a major concern since uncertainty, due to limited or missing data, often restricts the performance of analysis. The utilization of expert knowledge often provides an alternative for such a situation. However, it comes at the cost of possible uncertainties related to incompleteness (partial ignorance), imprecision (subjectivity), and lack of consensus (if multiple expert judgments are used). Further, if the bow-tie analysis is not flexible enough to incorporate new knowledge or evidence, it may undermine the purpose of risk assessment.Fuzzy set and evidence theory are capable of characterizing the uncertainty associated with expert knowledge. To minimize the overall uncertainty, fusing the knowledge of multiple experts and updating prior knowledge with new evidence are equally important in addition to addressing the uncertainties in the knowledge. This paper proposes a methodology to characterize the uncertainties, aggregate knowledge and update prior knowledge or evidence, if new data become available for the bow-tie analysis. A case study comprising a bow-tie for a typical offshore process facility has also been developed to describe the utility of this methodology in an industrial environment.     

风险分析的质量评价研究

风险分析的质量直接影响风险分析技术的应用 ,随着风险分析技术及其应用范围的扩大 ,建立评价风险分析的内容、结果及其方法和标准是十分必要的。为此 ,讨论了风险分析的质量概念 ;提出了一种评价风险分析质量的方法 ,该方法基于对风险分析过程的评价 ,利用检查表来发现风险分析的缺陷 ,同时研究了该方法的有效性和可靠性 ;最后 ,对方法的局限性以及提高风险分析质量的途径等其他相关问题进行了讨论。研究结果表明 :评价风险分析质量的方法能揭示风险分析中大多数的重大缺陷     

基于Apriori算法的易燃易爆危险化学品储运火灾爆炸事故关键致险因素的挖掘

导致易燃易爆危险化学品储运火灾爆炸事故的因素繁多、关系复杂,挖掘关键致险因素是减少管理成本、提高防控效率的关键。研究了200例事故等级为较大事故以上的易燃易爆危险化学品储运火灾爆炸事故的原因,采用事故树分析法建立易燃易爆危险化学品储运火灾爆炸事故树,并运用频度统计法遴选出致险因素;在此基础上,建立基于Apriori算法的关联规则模型进行数据挖掘,共得到14个关键致险因素。通过对关键致险因素与易燃易爆危险化学品储运火灾爆炸事故之间关联规则的分析表明,关键致险因素与事故之间存在强关联规则,单一关键致险因素或其组合的存在必将导致事故的发生,为实现危险化学品储运精准化安全管理提供参考。     

基于主逻辑图的安全风险建模研究

概率风险评估已成为研究复杂系统安全风险较为成熟的方法 ,其风险模型的建立是基于故障树/事件树的 ,风险分析具有众多的静态特性。然而 ,大的复杂系统往往存在诸多动态因素 ,在风险研究需要考虑这些动态因素的情况下 ,基于传统故障树 /事件树的模型则难以提供支持。笔者对此进行了分析 ,提出了基于主逻辑图的安全风险建模 ;主要介绍了主逻辑图的概念及其建模方法 ,并对模型进行了分析 ;在此基础上 ,基于主逻辑图分析事故场景 ,对场景风险的量化评估进行了简要介绍 ,并结合某核反应堆例子进行了分析。     

Analyzing system safety and risks under uncertainty using a bow-tie diagram: An innovative approach

A bow-tie diagram combines a fault tree and an event tree to represent the risk control parameters on a common platform for mitigating an accident. Quantitative analysis of a bow-tie is still a major challenge since it follows the traditional assumptions of fault and event tree analyses. The assumptions consider the crisp probabilities and “independent” relationships for the input events. The crisp probabilities for the input events are often missing or hard to come by, which introduces data uncertainty. The assumption of “independence” introduces model uncertainty. Elicitation of expert's knowledge for the missing data may provide an alternative; however, such knowledge incorporates uncertainties and may undermine the credibility of risk analysis.This paper attempts to accommodate the expert's knowledge to overcome missing data and incorporate fuzzy set and evidence theory to assess the uncertainties. Further, dependency coefficient-based fuzzy and evidence theory approaches have been developed to address the model uncertainty for bow-tie analysis. In addition, a method of sensitivity analysis is proposed to predict the most contributing input events in the bow-tie analysis. To demonstrate the utility of the approaches in industrial application, a bow-tie diagram of the BP Texas City accident is developed and analyzed.     

Decision analysis on fire safety design based on evaluating building fire risk to life

This study presents a framework of decision analysis on fire safety design alternatives based on evaluating building fire risk to life. A probabilistic risk assessment method for occupant life is presented with consideration of some uncertainties of evacuation process and fire development at first. For occupant evacuation time assessment, occupant pre-movement time is characterized by normal distribution. For onset time to untenable conditions assessment, its uncertainty is considered as probability distribution according to the range of design fires. Based on event tree technique, probable fire scenarios are analyzed with consideration of the effect of fire protection systems, employees extinguishing, etc. It is difficult to make a precise assessment on probability and consequence of every fire scenario, but their lower bound and upper bound can be achieved based on statistical data. Therefore, Supersoft Decision Theory [Malmnäs, P.E., 1995. Methods of Evaluation in Supersoft Decision Theory. Department of Philosophy, Stockholm University, 365 Stockholm; Johansson, H., Malmnäs, P.E., 2000. Application of supersoft decision theory in fire risk assessment. Journal of Fire Protection Engineering 14, 55–84] that allows decision maker to utilize vague statement is utilized to integrate with risk assessment to make decision on different fire safety design alternatives. To express how to make decision on different fire safety design alternatives based on evaluating building fire risk to life, two hypothetical fire safety design alternatives for a commercial building are presented.     

Consequence risk analysis using operating procedure event trees and dynamic simulation

Faults due to human errors cost the petrochemical industry billions of dollars every year and can have adverse environmental consequences. Unquantified human error probabilities exist during process state transitions performed each day by process operators using standard operating procedures. Managing the risks associated with operating procedures is an essential part of managing the overall safety risk. Additional operator training and safety education cannot eliminate all such faults due to human errors; therefore, we propose an operating procedure event tree (OPET) like analysis with branches and events specifically designed to perform risk analysis on operating procedures. The OPET method adapts event trees to analyze the risk due to human error while performing operating procedures. We consider human error scenarios during the procedure and determine the likely consequences by applying dynamic simulation. The modified event tree provides an estimate of the error frequencies.Operating procedure steps were developed, and potential operator faults were determined for two typical equipment switching procedures found in chemical plant operations. Then, dynamic simulation using Aspen HYSYS software was applied to determine the overpressure related consequences of each fault. Finally, the error frequencies resulting from those scenarios were analyzed using operating procedure event trees. We found that a typical ethylene plant gas header would overpressure with 0.6% frequency per manual dryer switch. Since dryer switches occur from every few days up to once per shift, these results suggest that dryer switching should be automated to ensure safe and environmentally friendly operation. Process dryer switching performed manually by operators opening and closing gate valves can be automated with control valves and a distributed control system. A sample distillation column was found to overpressure with 0.85% frequency per manual reflux pump switch.     

Cost-based fire risk assessment in natural gas industry by means of fuzzy FTA and ETA

Fire is the most prevalent accident in natural gas facilities. In order to assess the risk of fire in a gas processing plant, a fault tree analysis (FTA) and event tree analysis (ETA) has been developed in this paper. By utilizing FTA and ETA, the paths leading to an outcome event would be visually demonstrated. The framework was applied to a case study of processing plant in South Pars gas complex. All major underlying causes of fire accident in a gas processing facility determined through a process hazard analysis (PHA). Fuzzy logic has been employed to derive likelihood of basic events in FTA from uncertain opinion of experts. The outcome events in event tree has been simulated by computer model to evaluate their severity. In the proposed methodology the calculated risk has the unit of cost per year which allows the decision makers to discern the benefit of their investment in safety measures and risk mitigation.     

Joint applicability test of software for laboratory assessment and risk analysis

Most of the available risk management methods are not directly applicable to academic research laboratories. One solution to systematically perform risk analyses in this environment is the Laboratory Assessment and Risk Analysis (LARA) method. This method was developed to allow untrained personnel to identify of possible risks and rank them according to their importance. The purpose of this study was to find out, if this method can be used as a holistic risk management technique in different environments, and which are the differences when comparing the results to other, well established risk analysis techniques. The risk analyses were performed at two European universities and for various procedures. The results show, that the LARA procedure is more easily performable than the other methods and gives comparably adequate results. Being applicable by non-experts, this holistic risk analysis method for research laboratories can help to reduce the accident rate in the academic environment.     

Safety study of an LNG regasification plant using an FMECA and HAZOP integrated methodology

A safety analysis was performed to determine possible accidental events in the storage system used in the liquefied natural gas regasification plant using the integrated application of failure modes, effects and criticality analysis (FMECA) and hazard and operability analysis (HAZOP) methodologies. The goal of the FMECA technique is the estimation of component failure modes and their major effects, whereas HAZOP is a structured and systematic technique that provides an identification of the hazards and the operability problems using logical sequences of cause-deviation-consequence of process parameters. The proposed FMECA and HAZOP integrated analysis (FHIA) has been designed as a tool for the development of specific criteria for reliability and risk data organisation and to gain more recommendations than those typically provided by the application of a single methodology. This approach has been applied to the risk analysis of the LNG storage systems under construction in Porto Empedocle, Italy. The results showed that FHIA is a useful technique to better and more consistently identify the potential sources of human errors, causal factors in faults, multiple or common cause failures and correlation of cause-consequence of hazards during the various steps of the process.     

Developing a new alternative risk assessment framework in the work sites by including a stochastic and a deterministic process: A case study for the Greek Public Electric Power Provider

An individual method cannot achieve the optimum risk-assessment result in the worksites, and future perspectives should focus on the parallel application of a deterministic approach with a stochastic approach. In particular, the risk analysis and assessment techniques of the deterministic (DET) approach are classified into three main categories: (a) the qualitative, (b) the quantitative, and (c) the hybrid techniques (qualitative-quantitative, semi-quantitative). Furthermore, the stochastic (STO) approach includes the classic statistical approach (CSA) and the accident forecasting modeling (AFM). The objective of this paper is triple: (a) the presentation and classification of the main risk analysis and risk assessment methods and techniques of the deterministic approach and the stochastic approach as well, (b) the development and presentation of a new alternative risk assessment framework (called as STODET) including a stochastic and a deterministic process, and (c) the application of STODET in the Greek Public Power Corporation (PPC) by using occupational accidents that have been recorded, during the 17-year period of 1993-2009. In particular, the STODET application proves that required actions (or suppressive measures) are essential and must be taken in a medium-term period (1 working year) for abolishing the hazard sources.     

聚丙烯车间12m3釜聚合岗危险分析及安全措施

聚合岗位是石油化工厂聚丙烯车间重要且非常危险的岗位,一旦发生意外事故,将会导致人员的重大伤亡和巨大的经济损失。笔者依据聚丙烯车间生产现状,详细地介绍了12m3 釜聚合岗位工艺流程,用“系统安全”的理论和方法对其主要物质、生产工艺、操作过程的危险性进行了辨识,且绘制了主要危险点分布图;通过危险源辨识可知,聚合釜具有的超温、超压特点是该岗位众多危险源中最为严重的潜在危险,应用“事故树法”对聚合釜超温、超压爆炸事故进行了危险分析,找出其爆炸潜在的危险因素有2 4种,该事故树的最小割集共有6 6个,表明聚合釜爆炸可能性是很大的。依据分析结果,针对聚合釜爆炸可能性最大的危险因素,提出了安全对策与措施,以避免或减少爆炸事故的发生     

Incorporating the 2007 recommendations of the International Commission on Radiation Protection into the J-value analysis of nuclear safety systems

The newly released findings by the International Commission on Radiation Protection (ICRP) led to a review of the lifetime risk coefficients for fatal cancer used in J-value analysis of nuclear safety systems. The change in life expectancy a safety system brings about by averting a radiation exposure needs to be estimated in order to calculate the safety system's J-value, and this is done following the ICRP's practice of using risk coefficients that are uniform across both genders and all ages in the defined population group (either workers or the general population). The ICRP predicted uniformly lower radiation risks in 2007 than in 1990 on a like-for-like basis, but it was found that the ICRP's new risk coefficients needed to be multiplied by a compensating factor specific to each population when used in calculating the radiation-induced change in life expectancy. Incorporating the new compensating factor leads to a decrease in the J-value calculated of about 5% for workers and 15% for the general population compared with earlier, reported results. These adjustments are not large compared with the uncertainties associated with radiation harm and the economics of installing a safety system, but will strengthen slightly the case for spending on a nuclear safety measure.     

Handling data uncertainties in event tree analysis

Event tree analysis (ETA) is an established risk analysis technique to assess likelihood (in a probabilistic context) of an accident. The objective data available to estimate the likelihood is often missing (or sparse), and even if available, is subject to incompleteness (partial ignorance) and imprecision (vagueness). Without addressing incompleteness and imprecision in the available data, ETA and subsequent risk analysis give a false impression of precision and correctness that undermines the overall credibility of the process. This paper explores two approaches to address data uncertainties, namely, fuzzy sets and evidence theory, and compares the results with Monte Carlo simulations. A fuzzy-based approach is used for handling imprecision and subjectivity, whereas evidence theory is used for handling inconsistent, incomplete and conflicting data. Application of these approaches in ETA is demonstrated using the example of an LPG release near a processing facility.