基于贝叶斯网络的危化品爆炸事故电力系统风险评估模型*

为提高危化品爆炸事故电力应急预警的准确性,建立基于贝叶斯网络的危化品爆炸事故电力系统风险评估模型。基于危化品爆炸事故电力应急典型情景分析,建立综合考虑突发事件、承灾载体和应急管理等风险因素的贝叶斯网络结构。应用概率刻画风险因素信息的不确定性及其相互影响,定量分析事件后果。结合一般条件和典型情景等的应用实例,分析评价方法和风险因素对风险等级的影响。结果表明:该模型能够在危化品爆炸事故发生后,评价电力应急预警等级;能够在危化品爆炸事故发生前,分析典型情景的风险和风险因素的影响,为应急准备提供支持;“最大概率法”较“概率加权求和法”得出的事件等级可能较低。     

Dynamic risk assessment using failure assessment and Bayesian theory

To ensure the safety of a process system, engineers use different methods to identify the potential hazards that may cause severe consequences. One of the most popular methods used is quantitative risk assessment (QRA) which quantifies the risk associated with a particular process activity. One of QRA's major disadvantages is its inability to update risk during the life of a process. As the process operates, abnormal events will result in incidents and near misses. These events are often called accident precursors. A conventional QRA process is unable to use the accident precursor information to revise the risk profile. To overcome this, a methodology has been proposed based on the work of Meel and Seider (2006). Similar to Meel and Seider (2006) work, this methodology uses Bayesian theory to update the likelihood of the event occurrence and also failure probability of the safety system. In this paper the proposed methodology is outlined and its application is demonstrated using a simple case study. First, potential accident scenarios are identified and represented in terms of an event tree, next, using the event tree and available failure data end-state probabilities are estimated. Subsequently, using the available accident precursor data, safety system failure likelihood and event tree end-state probabilities are revised. The methodology has been simulated using deterministic (point value) as well as probabilistic approach. This Methodology is applied to a case study demonstrating a storage tank containing highly hazardous chemicals. The comparison between conventional QRA and the results from dynamic failure assessment approach shows the significant deviation in system failure frequency throughout the life time of the process unit.     

危险化学品事故后果计算过程探讨

基于事故后果的风险评估方法是国内外进行安全评价和土地使用安全规划的基础方法之一。本文基于危险化学品种类和危险装置类型，对可能发生的事故情景和相应事故后果计算模型的选择方法进行了阐述，针对易燃、易爆、有毒危险化学品的事故后果计算过程进行了系统分析。最后对本文提出的事故后果分析方法进行了实例应用。本文的研究成果可为安全评价或土地使用安全规划等工作的开展提供必要的参考。     

Research on three-dimensional visualization system of Natech events triggered domino accidents in oil-gas depots

Natech events and domino accidents happen worldwide and usually lead to severe consequences, especially in hazardous areas such as Oil-Gas depots. With the continuous development of the Oil-Gas industry and information technology, it is essential to realize the three-dimensional management and monitoring of hazardous substances. To evaluate the consequences caused by Natech events triggered domino accidents (accident chain), a matrix loop method was proposed to calculate the probability and evolution path of the accident chain. The actual layout and the real-time data of the Oil-Gas depot are input into the evaluation method in matrices, thus making the results reliable and updated. Besides, a B/S architecture system is developed to present the evaluation consequences of the proposed method. The three-dimensional visualization effects of natural disasters, technological accidents and the dynamic propagation process of the accident chain are also realized in the system to enhance the user's experience. A system application regarding lightning-triggered domino accidents was carried out to demonstrate the feasibility and rapidity of the proposed evaluation method and perform the system's operation process and visualization effects. The application results show that the system can provide effective decision-making assistance to safety management workers before the accident and guidance for emergency rescue operations during the accident.     

System safety assessment using safety entropy

Safety assessment has a primary role in hazardous operations. Most studies on safety assessment focus on risk and accident modeling, in which safety is absent. These top-down methods are highly dependent on the occurred accidents to establish accidental scenarios, which may make the assessment approach lagging behind the evolving modern systems. Moreover, this “special to general” logic is scientifically suspect in safety assessment. There is a call for the development of safety assessment methods in the presence of system safety to complement risk-focused safety analysis. These methods should provide a framework based on a bottom-up approach to examine system safety from the operational perspective. This paper has attempted to provide a potential solution. In particular, a novel concept of safety entropy is proposed to integrate with The Functional Resonance Analysis Method (FRAM), which is used to form the qualitative understanding of a system. A formula consisted of safety entropy, functional conformability, and system complexity has been established to determine the spontaneity of the safety state-changing process. The proposed method is applied to the safety assessment of a propane feed-control system. The results show the applicability of the method. Nevertheless, the model still needs to be further improved to fulfill better support for safety-related decision problems.     

ExSys-LOPA for the chemical process industry

The chemical process industries are characterized by the use, processing, and storage of large amounts of dangerous chemical substances and/or energy. Among different missions of chemical plants there are two very important ones, which: 1. provide a safe work environment, 2. fully protect the environment. These important missions can be achieved only by design of adequate safeguards for identified process hazards. Layer of Protection Analysis (LOPA) can successfully answer this question. This technique is a simplified process of quantitative risk assessment, using the order of magnitude categories for initiating cause frequency, consequence severity, and the likelihood of failure of independent protection layers to analyze and assess the risk of particular accident scenarios. LOPA requires application of qualitative hazard evaluation methods to identify accident scenarios, including initiating causes and appropriate safeguards. This can be well fulfilled, e.g., by HAZOP Studies or What-If Analysis. However, those techniques require extensive experience, efforts by teams of experts as well as significant time commitments, especially for complex chemical process units. In order to simplify that process, this paper presents another strategy that is a combination of an expert system for accident scenario identification with subsequent application of LOPA. The concept is called ExSys-LOPA, which employs, prepared in advance, values from engineering databases for identification of loss events specific to the selected target process and subsequently a accident scenario barrier model developed as an input for LOPA. Such consistent rules for the identification of accident scenarios to be analyzed can facilitate and expedite the analysis and thereby incorporate many more scenarios and analyze those for adequacy of the safeguards. An associated computer program is under development. The proposed technique supports and extends the Layer of Protection Analysis application, especially for safety assurance assessment of risk-based determination for the process industries. A case study concerning HF alkylation plant illustrates the proposed method.     

Bi-directional connectivity diagram for accident propagation analysis considering the interactions between multiple-process units

An incident may propagate to an accident with different severity dependent on its propagation scenarios. Since the accident propagation is a two-way process, the current research is focusing on the one-way analysis. This paper aims to analyze the combined effect of multi-units sources and their interactions during the accident propagation. The bi-directional connectivity diagram (BDCD) is applied to visualize the interactions between multiple process units as hazardous sources. The deployed safety barriers interrupt the connection between the hazardous sources and thus minimize the influence of one BDCD node on another. Through which, the accident propagation is reduced. The proposed method can be suitable to the general accidents, and it is applied to a case study of the LNG terminal station to assess the potential consequences of explosion caused by the leakage, in which the cost of the safety barrier is also considered. The BDCD approach is found more effective than traditional single-hazardous source methods for analyzing the accident propagation of multi-units sources in the chemical plant and achieving intrinsic safety.     

Using risk tolerance criteria to determine safety integrity levels for safety instrumented functions

Standards and industry guidelines for Safety Instrumented Systems (SISs) describe the use of hazard and risk analysis to determine the risk reduction required, or Safety Integrity Levels (SILs), of Safety Instrumented Functions (SIFs) with reference to hazardous events and risk tolerance criteria for them. However, significant problems are encountered when putting this approach into practice. There is ambiguity in the meaning of the term hazardous event. Notably, even though it is a key concept in the process-sector-specific SIS standard, IEC 61511/ISA 84, it is not defined in the standard. Consequently, risk tolerance criteria for hazardous events are ill-defined and, therefore, they are not the most appropriate criteria to use. Most current approaches to SIL determination use them and therefore they are flawed fundamentally.An informed decision on the tolerability of risk for a facility cannot be made by determining only the tolerability of risk for individual hazardous events. Rather, the tolerability of the cumulative risk from all hazard scenarios and their hazardous events for a facility must be determined. Such facility risk tolerance criteria are the type used by regulators. This issue applies to all per event risk tolerance criteria. Furthermore, determining the tolerability of risk for a facility based only on the risks of single events, be they hazard scenarios or hazardous events, and comparing them to risk tolerance criteria for the events is not meaningful because there is no consideration of how many such events can actually occur and, therefore, no measure of the total risk. The risks from events should be summed for a facility and compared with overall facility risk tolerance criteria.This paper describes and illustrates SIL determination using a risk model implemented within the framework of Layers of Protection Analysis (LOPA) that overcomes these problems. The approach allows the allocation of risk across companies, facilities, processes, process units, process modes, etc. to be managed easily.     

Research on the prediction model of hazardous chemical road transportation accidents

In recent years, hazardous chemicals road transport accidents have occurred frequently, causing huge casualties and property losses, and accident risk assessment has become the focus of researchers' research. To predict the risk probability value of hazardous chemical road transport accidents, first, we compiled data on road transportation accidents of hazardous chemicals in China in the past five years. And the nine nodes in the Bayesian network (BN) structure were defined in combination with relevant classification standards. The optimal Bayesian network structure for hazardous chemical road transport accidents was determined based on the K2 algorithm and the causalities between the nodes. Second, the node conditional probabilities were derived by parameter learning of the model using Netica, and the validity of the model was verified using the 5-fold cross-validation method. Last, the Bayesian network model of hazardous chemical road transport accidents is used to analyze accident examples, and the accident chain of “rear-end-leakage” is predicted, and the accident is most likely to be disposed of within 3–9 h. The study shows that the derived accident prediction model for hazardous chemical road transportation can reason reasonably about the evolution of accident scenarios and determine the probability values of accident risks under different parameter conditions.     

基于危险的事故场景及其形式化描述

事故场景的鉴别是复杂装备系统安全风险分析的基础,其完备性直接影响风险评估的有效性,因为任何事故场景的遗漏即意味着设计上可能存在潜在的事故隐患。提出从鉴别系统危险出发。建立危险与其原因(关联引发事件)、事件与系统要素以及系统各要素之间的映射关系,通过对这些关系的运算来获取事故场景的新方法并给出其形式化表述,以期在理论上确保事故场景鉴别的完备性。     

Identification of accident scenarios caused by internal factors using HAZOP to assess an organic hydride hydrogen refueling station involving methylcyclohexane

Organic hydride hydrogen refueling stations have been remarked as stations that can employ a practicable method based on the organic chemical hydride system involving methylcyclohexane (MCH) for the transport of hydrogen. This station has advantages in that the storage and transportation of MCH does not require a large amount of energy compared to compressed and liquefied hydrogen, and the system can use existing infrastructure. This type of station involves some hazardous materials, and thus, scenario identifications and risk assessments have been performed by researchers. However, the sample of studies available have employed a conceptual design model, and they did not identify concrete scenarios triggered by internal factors. Therefore, the purpose of this study is to identify accidental scenarios caused by internal factors that can affect an organic hydride hydrogen refueling station. In this study, we used Hazard and Operability study (HAZOP) and examined safety measures for the scenarios. As a result of the HAZOP, 105 accidental scenarios were identified and classified into the two following groups; (i) the scenarios assumed that the substances were ignited after they were released to the atmosphere, and (ii) the scenarios assumed that the substances were ignited in the process before they were released. Significant scenarios in group (i) were MCH or toluene pool fires, hydrogen jet fires, vapor gas explosions, or flash fires. The significant scenarios classified in (ii) were newly identified in this study. The scenarios include the explosion of the explosive mixture formed by the gaseous phase of toluene and oxygen from the vent line connected to the tank due to the static electric charge in the tank. For each scenario, safety measures to prevent the progression of the accident scenario were examined with reference to the current laws and regulations in Japan.     

Risk analysis of deepwater drilling operations using Bayesian network

Deepwater drilling is one of the high-risk operations in the oil and gas sector due to large uncertainties and extreme operating conditions. In the last few decades Managed Pressure Drilling Operations (MPD) and Underbalanced Drilling (UBD) have become increasingly used as alternatives to conventional drilling operations such as Overbalanced Drilling (OVD) technology. These newer techniques provide several advantages however the blowout risk during these operations is still not fully understood. Blowout is regarded as one of the most catastrophic events in offshore drilling operations; therefore implementation and maintenance of safety measures is essential to maintain risk below the acceptance criteria. This study is aimed at applying the Bayesian Network (BN) to conduct a dynamic safety analysis of deepwater MPD and UBD operations. It investigates different risk factors associated with MPD and UBD technologies, which could lead to a blowout accident. Blowout accident scenarios are investigated and the BNs are developed for MPD and UBD technologies in order to predict the probability of blowout occurrence. The main objective of this paper is to understand MPD and UBD technologies, to identify hazardous events during MPD and UBD operations, to perform failure analysis (modelling) of blowout events and to evaluate plus compare risk. Importance factor analysis in drilling operations is performed to assess contribution of each root cause to the potential accident; the results show that UBD has a higher occurrence probability of kick and blowout compared to MPD technology. The Rotating Control Devices (RCD) failure in MPD technology and increase in flow-through annulus in UBD technology are the most critical situations for kick and blowout.     

Operational risk assessment: A case of the Bhopal disaster

Accidental releases of hazardous chemicals from process facilities can cause catastrophic consequences. The Bhopal disaster resulting from a combination of inherently unsafe designs and poorly managed operations is a well-known case. Effective risk modeling approaches that provide early warnings are helpful to prevent and control such rare but catastrophic events. Probability estimation of these events is a constant challenge due to the scarcity of directly relevant data. Therefore, precursor-based methods that adopt the Bayesian theorem to update prior judgments on event probabilities using empirical data have been proposed. The updated probabilities are then integrated with consequences of varying severity to produce the risk profile.This paper proposes an operational risk assessment framework, in which a precursor-based Bayesian network approach is used for probability estimation, and loss functions are applied for consequence assessment. The estimated risk profile can be updated continuously given real-time operational data. As process facilities operate, this method integrates a failure-updating mechanism with potential consequences to generate a real-time operational risk profile. The real time risk profile is valuable in activating accident prevention and control strategies. The approach is applied to the Bhopal accident to demonstrate its applicability and effectiveness.     

A development in energy flow/barrier analysis

In this paper, the occupational accidents and their effects on people are modeled. The basis of energy flow/barrier analysis is used to define an accident as the impact of a hazardous agent on a target, as a result of failure of control and protective barriers. This definition is enhanced to include serial and parallel barriers and to distinguish energy barriers from target barriers. The barrier attributes are defined and used to create a quantitative scenario-building model. The probability and severity of various accidents are estimated, by studying barrier reliability and efficiency. This approach is used to develop and simulate accident scenarios and to calculate their consequences. This model can be used in complex systems for analyzing the risk and estimating the importance of barriers.     

Reliability assessment of safety instrumented systems subject to different demand modes

Safety instrumented systems (SISs) are commonly used in the process industry, to respond to hazardous events. In line with the important standard IEC 61508, SISs are generally classified into two types: low-demand systems and high-demand systems. This article explores this classification by studying the SIS reliability for varying demand rates, demand durations, and test intervals. The approach is based on Markov models and is exemplified by two simple system configurations. The SIS reliability is quantified by the probability of failure on demand (PFD) and the frequency of entering a hazardous state that will lead to an accident if the situation is not controlled by additional barriers. The article concludes that very low-demand systems are similar and may be treated as a group. The same applies to very high-demand system. Between these group, there is a rather long interval where the demand rate is neither high-demand nor low-demand. These medium-demand systems need a specific treatment. The article shows that the frequency of entering into a hazardous state increases with the demand rate for low-demand systems, while it is nearly independent of both the demand rate and the demand duration for high-demand systems. The PFD is an adequate measure for the SIS reliability for low-demand systems, but may be confusing and difficult to interpret for high-demand systems.     

HazMat transportation safety assessment: Analysis of a “Viareggio-like” incident in the Netherlands

Relevant safety issues are associated with hazardous materials transportation, especially when transport routes cross populated areas. On March 6th, 2015, a passenger train collided with the last rail car of a freight train in Tilburg, the Netherlands. The last car contained 50 t of liquefied 1,3-butadiene. As a result of the collision, the last car showed deformation; a small leakage occurred but fortunately with no relevant consequences. However, extremely severe consequences could have happened, such as in the rail accident that occurred in Viareggio, Italy in 2009. In this work, the case of Tilburg was firstly outlined and explored by qualitative methods, in order to identify possible realistic final scenarios that could have happened. Second, the potential consequences of the identified scenarios were estimated through conventional integral model for physical effects evaluation. Comparison with the Viareggio case was also shown in order to support the discussion of the results obtained. Finally, lessons learned after the incident, policy making considerations, and indications for the risk mitigation of hazardous materials transportation are given.     

Dynamic safety analysis of process systems by mapping bow-tie into Bayesian network

Among the various techniques used for safety analysis of process systems, bow-tie (BT) analysis is becoming a popular technique as it represents an accident scenario from causes to effects. However, the BT application in the dynamic safety analysis is limited due to the static nature of its components, i.e. fault tree and event tree. It is therefore difficult in BT to take accident precursors into account to update the probability of events and the consequent risk. Also, BT is unable to represent conditional dependency. Event dependency is common among primary events and safety barriers. The current paper illustrates how Bayesian network (BN) helps to overcome these limitations. It has also been shown that BN can be used in dynamic safety analysis of a wide range of accident scenarios due to its flexible structure. This paper also introduces the application of probability adapting in dynamic safety analysis rather than probability updating. A case study from the U.S. Chemical Safety Board has been used to illustrate the application of both BT and BN techniques, with a comparison of the results from each technique.     

Predicting major accidents in the process industry based on the barrier status at scenario level: A practical approach

OCI Nitrogen wants to gain knowledge of (leading) indicators regarding the process safety performance of their ammonia production process. This paper answers the question whether indicators can be derived from the barrier system status to provide information about the development and likelihood of the major accident processes in the ammonia production process.The accident processes are visualized as scenarios in bowties. This research focuses on the status of the preventive barriers on the left-hand side of the bowtie. Both the quality – expressed in reliability/availability and effectiveness – and the activation of the barrier system give an indication of the development of the accident scenarios and the likelihood of the central event. This likelihood is calculated as a loss of risk reduction compared to the original design. The calculation results in an indicator called “preventive barrier indicator”, which should initiate further action. Based on an example, it is demonstrated which actions should be taken and what their urgency is.     

基于 MIMAH 的工业事故场景辨识

介绍了事故场景概念,并根据欧盟ARAMIS项目框架下提出的MIMAH（辨识重要事故危险方法）,即从危险设备的角度来辨识与设备相关的关键事件,并利用事故树（ FTA ）、事件树（ ETA）,建立一个以关键事件为中心的蝴蝶结结构图来描述事故场景。通过运用这种方法,能够对事故场景的辨识更加具有系统性、针对性。最后,以液氨储罐装置作为示例进行说明。     

Surface installations intended for Carbon Capture and Sequestration: Atypical accident scenarios and their identification

With the advent of Carbon Capture and Sequestration (CCS) technology the extent of CO₂ handling is set to increase dramatically. However, lack of substantial operational experience in such a novel process can lead to significant difficulties in identifying the associated hazards. This field may be characterized by atypical accident scenarios, i.e. scenarios not captured by common HAZard IDentification (HAZID) techniques because of omissions, errors or lack of knowledge. Recent atypical events evidence that consequences may exceed by far those of worst-case reference scenarios. Identification of atypical scenarios related to CCS is a challenge, considering also the public concern that this technology raises. This study focuses on new and emerging technologies of carbon capture and transport. A HAZID analysis was carried out by means of two different approaches (‘top-down’ and ‘DyPASI’). This allowed not only for a double check of results, but also for the comparative assessment of the methodologies and of their applicability. A general overview of the accident scenarios related to these technologies was given. No absolute showstoppers were found. Rather, a number of potential hazards were identified which will require the adoption of safe design principles to eliminate, prevent, control or mitigate them. Some possible safety barriers required for implementation were identified as a starting point in this process.