Automated diagnosis of malfunctions based on object orientated programming

The diagnostic event analyser (DEA) is a system for automated diagnosis of malfunction in continuous processes. DEA interprets real-time sequences of abnormal events such as measurements out of normal ranges and violation of quantitative algebraic constraints. DEA can be used to supplement an existing alarm system, providing a higher-level interpretation of patterns of abnormal events for the process operator. To use DEA, the user must supply only standard ‘form’ inputs of the root causes and relationships between abnormal events. DEA is implemented using object- oriented programming, a technique which allows general rules and procedures to be applied to process-specific objects. This architecture facilities treatment of multiple malfunctions and robustness to variation in event sequences.     

Measuring and analyzing adaptive capacity at management levels of resilient systems

Resilience engineering (RE) is capable of handling disruptive events and controlling their consequences in process industries such as petrochemical plants. This study aims at analyzing the level of adaptive capacity and identifying effective factors on developing adaptive capacity in the organizational structure of process industries. The data of this study were obtained through direct observation and a structured questionnaire in a petrochemical plant. Managers at all levels participated in the survey. Data envelopment analysis (DEA), which is a mathematical approach, was used to compute and analyze the role of the factors contributing to adaptive capacity. The results indicated that reporting safety issues played a central role in enhancing adaptive capacity at all management levels. Both middle management and low-level management emphasized the importance of management commitment, whereas top-level management considered flexibility as a vital factor in managing disruptions and reducing accidents. The findings of this study could be useful for managers and other decision-makers to improve safety in process industries.     

Fault-dynamic modelling of a phthalic anhydride reactor

A new methodology, fault-dynamic modelling, has been developed for analysis of potentially hazardous situations in the process industries. Traditional fault-tree analysis is used to determine the combinations of component failure that can lead to a particular process upset condition. Realistic dynamic modelling is then used to calculate the time available for corrective action once the upset has started. The method is applied to a phthalic anhydride reactor. The results of the analysis identify three process upsets that can lead to catastrophic failure in 2–5 min if left uncorrected. Other process upsets lead to safe conditions.     

Focus on mission success: Process safety for the Atychiphobist

Modern process plants are complex engineering systems. While thorough reviews of system safeguards are performed, catastrophic events continue to occur, often unfolding in unforeseen ways. Success in process safety demands safe processes, and understanding rare, high consequence events is central to the traditional process safety approach. This philosophy is common to all high-hazard industries, offering the potential for sharing approaches, experience, and lessons learned. The problem, however, is that people (and organizations and entire industries) who fear failure (atychiphobia) sometimes obsess about failure so much that they miss opportunities to succeed.This paper examines selected risk management practices in the power generation and aerospace industries and how those practices have led to improved performance. Risk informed decision making (RIDM) has had widespread application in the nuclear and aerospace industries, and is undergoing enhancements to become a key framework for risk management. Additionally, rather than focusing on avoidance of loss, there are emerging approaches supporting achievement of success. This approach provides a more direct link of risk to business and operational objectives, but does challenge conventional risk approaches founded in a loss prevention-centric view. The paper reflects upon risk informed decision making and success modeling, and suggests how these methods may be applied in the field of process safety. Specific examples are drawn from the defense in depth approach from the nuclear power industry and mission success concepts developed for NASA.     

Performance evaluation of integrated resilience engineering factors by data envelopment analysis: The case of a petrochemical plant

Petrochemical units are potentially prone to incidents that have catastrophic consequences such as explosion, leakage of toxic materials, and the stoppage of the production process. Resilience engineering (RE) is a new method that can control incidents and limit their consequences. It includes top-level commitment, reporting culture, learning, awareness, preparedness, and flexibility. However, this study introduces a new concept of RE (referred to as integrated RE or IRE) which includes the above factors in addition to self-organization, teamwork, redundancy and fault-tolerant. This study evaluates performance of IRE in a petrochemical plant through considering the obtained data from questionnaires and data envelopment analysis (DEA) approach. Moreover, the performance of RE and the new IRE are compared and discussed. The results show that although there is a strong direct correlation between the DEA results in two frameworks, the mean scores of efficiency in IRE is slightly higher than RE. This is the first study that introduces an integrated approach for RE. In addition, this study is amongst the first ones that examine the behavior of resilience engineering by DEA. Moreover, the superiority of IRE is shown through robust statistical analysis.     

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.     

Continuous performance assessment and improvement of integrated HSE and maintenance systems by multivariate analysis in gas transmission units

In this study, an integrated “Health, Safety, and Environment (HSE)” and maintenance systems are presented. Multivariate analysis is used for continuous performance assessment and improvement of these systems. The two subjects of “Maintenance System” and “HSE” have been individually investigated several times in different studies. However, few studies have been done to integrate these two systems and provide an integrated system for their implementation. This study evaluates current maintenance and HSE systems of a Gas Transmission Unit by Data Envelopment Analysis (DEA) and Principal Component Analysis (PCA). Moreover, decision making units (DMUs) are examined and ranked. Employees are considered as DMUs. This is done through investigating and measuring their efficiencies and identifying the inefficient and less-efficient units. Since an increase in the number of inputs does not lead to an increase in the number of outputs with the same scale, an output-oriented DEA with a Variable Returns to Scale (VRS) is used. The Fuzzy DEA (FDEA) is also used in this research to decrease uncertainty existing in qualitative indicators and human error. Finally, suggestions are given to improve those DMUs. The managers and employees of the gas transmission unit constitute the statistical population of the study. To achieve the objectives of this study, standard questionnaires with respect to HSE and Maintenance system are completed by operators. The proposed approach would help policy makers and top managers of Gas Transmission Company to have a more comprehensive and thorough understanding the working conditions with respect to the maintenance and HSE features.     

Case study analysis of the financial impact of catastrophic safety events

Several catastrophic events in the process industry have caused extensive damage to life and property and have forever changed the process safety landscape. Despite the commitment from the industry, incidents still occur with similar root causes, impacting major companies in various ways. This work examines the financial impact of such incidents on the stability and future of affected companies in different industries. Several devastating incidents were analyzed to evaluate the direct impact in the short- and long-term and to discuss factors influencing the ability of affected companies to withstand the consequence of catastrophic events.     

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.     

Resilience assessment of chemical industrial areas during Natech-related cascading multi-hazards

In chemical industrial areas, technological accidents triggered by natural events (Natech events) may escalate. Complex cascading multi-hazard scenarios with high uncertainties may be caused. Resilience is an essential property of a system to withstand and recover from disruptive events. The present study focuses on the change of the resilience level due to (possible) interactions between cascading hazards, chemical installations and safety barriers during the dynamic evolution of fire escalations triggered by a natural hazard (certain cascading multi-hazard scenarios). A quantitative resilience assessment method is developed to this end. The state transition of a system facing accidents in the context of resilience is explored. Moreover, the uncertainties accompanying an accident evolution are quantified using a Dynamic Bayesian Network, allowing a detailed analysis of the system performance in different time steps. System resilience is measured as a time-dependent function with respect to the change of system performance. The applicability of the proposed methodology is demonstrated by a case study, and the effects of different configurations of safety barriers on improving resilience are discussed. The results are valuable to support disaster prevention within chemical industrial areas.     

Modeling and optimizing efficiency gap between managers and operators in integrated resilient systems: The case of a petrochemical plant

The reliability issue in complex industrial systems such as oil, gas, petrochemical companies, nuclear and aviation industries has been of great importance. Resilience engineering (RE) is a new attitude toward the improvement of safety and reliability in the stated systems. One of the challenges a resilient system might face is the gap between work as imagined by managers and work as actually done by operators. This study will introduce a new framework named integrated resilience engineering (IRE) as a result of developing the concept of RE. The data used in this research have been obtained by means of questionnaire from a petrochemical company. Thereafter, the efficiency of operators and managers are calculated in RE and IRE frameworks through data envelopment analysis (DEA) approach. Then, the gaps between managers and operators are analyzed in two frameworks. The results are indicative of a significant growth in the number of efficient operators and managers in IRE framework compared to RE framework. Besides, the efficiency mean of managers and operators in IRE framework has experienced the growth of 1.8% and 5% compared to RE framework, respectively. The efficiency gap between managers and operators in IRE framework has also enjoyed the improvement of 88% compared to RE framework. Generally, it can be said that the suggested items of this research has led to the betterment of managers and operators’ efficiency and of the efficiency gap between them. Therefore, these items can improve the resilience and safety of complex systems. The results of Spearman test show that there is a strong direct correlation between the DEA results in two frameworks. This is the first study that examines the efficiency gap between operators and managers based on the RE principles and by means of DEA approach.     

基于事件提取和改进MMHC的航空旅客运输事故征候贝叶斯网络建模*

为表征航空旅客运输事故征候演化机理,提出事故征候贝叶斯网络的建模方法。基于事故征候中致因事件、结果事件及分类标准的定义,以7 265起事故征候案例为样本,利用事件提取算法,识别事故征候叙述文本中的致因事件,利用改进的最大最小爬山算法实现网络建模;依据事件提取的测试集验证与结构学习的交叉验证,检验建模算法的准确性与有效性;基于证据敏感性指标,识别关键致因事件。结果表明:航空旅客运输事故征候贝叶斯网络模型包含94个节点和247条有向弧。空降冲突、严重设备故障、机组成员疾病及火灾烟雾是模型中高风险关联的致因事件,在安全监管过程中消除或减弱关键致因事件的发生能有效控制系统风险。     

Process Safety Management (PSM) for managing contractors in process plant

Accidents involving contractors continue to occur with regular frequency. By using the standard set within the PSM 29 CFR 1910.119(h) regulations, it has been identified that certain aspects of the way contractors do work are not up to the stipulated regulatory requirements especially regarding matters like not providing mandatory personal protective equipment (PPE) to the workers, not discussing hazards related to handling procedures with the workforce and improper control of non-routine activities during changes in shifts. The PSM 29 CFR 1910.119(h) regulations promulgated in 1992 provides standards that covered processes to obtain and evaluate data regarding contractors' health and safety programs as well as the contractors' performance evaluation. Many of the accidents involving contractors are direct result of poor training of contractors and/or poor control of the contracted work. Even though most organizations have their own contractor management systems, there are issues in meeting the requirements of PSM. The PSM standard only state “what to do” not “how to do it”. This is known as self-regulatory policy which depends on the industries understanding to interpret the standard that also contribute to this problem. This paper presents a structured and easy technique to plan and implement a practical and comprehensive contractors' management system in process industries that will comply with OSHA CFR 1910.119. A model has been developed based on this technique and its application has been tested in a pilot plant for compliance to PSM regulation. The model is beneficial to the process industries as any deficiencies in the PSM contractors' management program will be highlighted by the model which will then easily correct the identified deficiency so as to minimize and prevent catastrophic accidents.     

Identification of managerial shaping factors in a petrochemical plant by resilience engineering and data envelopment analysis

Resilience engineering (RE) is a novel approach that is capable of controlling and limiting incidents and accidents. This study identifies managerial shaping factors in a petrochemical plant by RE and data envelopment analysis (DEA). To do this, a standard questionnaire containing resilient factors is completed by managers of a petrochemical plant. Then, the best DEA model is selected based on average efficiency and statistical test. In addition, sensitivity analysis are performed to identify the most important shaping factors. Reporting culture, management commitment, and preparedness are identified as the most important factors in this paper, respectively. Finally, the proposed model is validated and verified through statistical experiment. The proposed approach would help managers to have a comprehensive understanding of the plant with respect to the RE. To the best of our knowledge this is the first paper that examines resilient shaping factors in a petrochemical plant with respect to management and organization by DEA.     

Dynamic failure assessment of an ammonia storage unit: A case study

Chemical Process Industries usually contain a diverse inventory of hazardous chemicals and complex systems required to perform process operations such as storage, separation, reaction, compression etc. The complex interactions between the equipment make them vulnerable to catastrophic accidents. Risk and failure assessment provide engineers with an intuitive tool for decision making in the operation of such plants. Abnormal events and near-miss situations occur regularly during the operation of a system. Accident Sequence Precursors (ASP) can be used to demonstrate the real-time operating condition of a plant. Dynamic Failure Assessment (DFA) methodology is based on Bayesian statistical methods incorporates ASP data to revise the generic failure probabilities of the systems during its operational lifetime.In this paper, DFA methodology is applied on an ammonia storage unit in a specialized chemical industry. Ammonia is stored in cold storage tanks as liquefied gas at atmospheric pressure. These tanks are susceptible to failures due to various abnormal conditions arising due process failures.Tank failures due to three such abnormal conditions are considered. Variation of the failure probability of the safety systems is demonstrated. The authors use ASP data collected from plant specific sources and safety expert judgement. The failure probabilities of some safety systems concerned show considerable deviation from the generic values. The method helps to locate the components which have undergone more degradation over the period and hence must be paid attention to. In addition, a Bayesian predictive model has been used to predict the number of abnormal events in the next time interval. The user-friendly and intuitive nature of the tool makes it appropriate for application in safety assessment reports in process industries.     

海底管道泄漏风险演化复杂网络分析

为了控制海底管道泄漏连锁风险,基于复杂网络,提出针对管道系统泄漏演化的半定量风险演化评价方法,将复杂的事故风险发展过程转化为简洁的网络分析计算。首先,构建包含30个风险节点与54条连接边的海底管道泄漏演化复杂网络模型;其次,采用无权有向网络中的节点出入度和聚类系数进行风险分析,确定影响管道泄漏的关键节点,提出断链控制方案;最后,将演化模型转化为带权的有向网络,采用Dijkstra算法计算各初始事件导致泄漏事故的最短路径。结果表明:海底管道系统泄漏网络的聚类系数为0.13,网络聚集程度偏低而演化性较强;各初始事件的最短路径均不超过10,表现出明显的小世界网络特征,初始事件的风险经少数几步传递即可导致泄漏事故的发生。海底管道泄漏风险演化规律的研究可为抑制初始事件、控制传递事件和减轻后果事件提供理论依据,对预防海底管道泄漏事故发生、保障管道持续安全运行具有现实意义。     

A bayesian network-based safety assessment method for solid propellant granule-casting molding process

The safety of the solid propellant molding process is vital for the stable production of high-quality propellants. Failure events caused by abnormal parameters in the molding process may have catastrophic consequences. In this paper, a Bayesian network (BN) model is proposed to assess the safety of the solid propellant granule-casting molding process. Fault tree analysis (FTA) is developed to construct a causal link between process variables and process failures. Subsequently, expert experience and fuzzy set theory (FST) are used to obtain failure probabilities of the basic events (BEs). Based on the mapping rules, FTA provides BN with reliable prior knowledge and a network structure with interpretability. Finally, when new evidence is obtained, the probability is updated with the diagnostic reasoning capability of BN. The results of the sensitivity analysis and diagnostic inference were combined to identify key parameters in the granule-casting molding process, including curing temperature, vacuum degree, extrusion, calendering roll distance, length setting value, holding time, and polish time. The results of this paper can provide effective supporting information for managers to conduct process safety analysis.     

基于安全结构理论的系统外部扰动演化规律研究*

为解决目前安全结构理论体系中缺少对系统外部扰动演化规律研究的问题,基于工程学外部扰动理论与安全结构理论,研究系统外部扰动对安全事件发生、发展及演化的作用规律。给出活动环境系统与外部因素的定义,参考星形拓扑网络建立外部因素拓扑结构;通过分析外部因素状态随时间的演化形式,阐明活动因素状态受外部扰动的响应机制,揭示系统外部扰动对安全事件演化的影响及作用规律。研究结果表明:外部扰动导致系统内活动因素状态的动态响应是安全事件形态演变的内在动力,外部因素状态随时间变化呈3类波动形式,活动因素状态的扰动响应变化规律受外部因素临界状态、极限斜率等特殊阈值的限制。通过控制活动环境系统外部因素的状态维持在阈值范围内来实现调控目的,可避免外部扰动对活动环境系统安全产生影响。     

Markov reliability model research of monitoring process in digital main control room of nuclear power plant

Monitoring process is an important part in a high safety digital main control room of nuclear power plant (NPP), it is the source extracted information and found abnormal information in time. As the human factors events arisen from monitoring process recently take place more and more frequent, the authors propose a reliability Markov model to effectively decrease these abnormal events. The model mainly analyzes next monitoring object probability in terms of current information and plant state. The authors divide digital human–machine interface into two parts that are referred as logical homogeneous Markov and logical heterogeneous Markov. For the former, a series of methods of probability evaluation are proposed, such as, Markov transition probability with condition, probability distributed function with human factors, system state and alarm; for the latter, the authors propose the calculation of probability of correlation degree between last time and next time and probability calculation methods with multi-father nodes. The methods can effectively estimate the transition probability from a monitoring component to next monitoring component at time t, can effectively analyze which information is more important in next monitoring process and effectively find more useful information in time t + 1, so that the human factors events in monitoring process can greatly be decreased.