Showing papers in "Journal of Loss Prevention in The Process Industries in 2018"

Development of cement dust suppression technology during shotcrete in mine of China-A review

Abstract: The shotcrete has become a major technique in the support construction of underground mine. Unfortunately, enormous amounts of cement dust may be produced during shotcrete because of deficient mixing of materials and high jet speed. At present, in term of dust suppression, the dust generated from shotcrete is easily ignored when compared with the rock or mine dust in mine exploitation. The paper summarizes the development of cement dust suppression technology during shotcrete. The component and pathological damage of cement dust was analysed. The diffusion of shotcrete dust was uncovered as well as alkaline particle distribution. Efforts to control and reduce the shotcrete dust were done, including developing new shotcrete machines, optimizing process and using new material technology and so on. Finally, challenges of shotcrete proposed in special roadway environment of China’ mine still exist and further efforts need to be made.

An innovative methodology for quickly modeling the spatial-temporal evolution of domino accidents triggered by fire

Abstract: Past accident analyses indicate that fire escalation is responsible for most of the domino effects that happened in the process industries. The evolution of domino accidents triggered by fire is different from domino accidents triggered by other primary scenarios, since the escalation caused by heat radiation is delayed with respect to the start of the fire. In this study, a methodology involving a Domino Evolution Graph (DEG) model and a Minimum Evolution Time (MET) algorithm is proposed to model the spatial-temporal evolution of domino accidents. Synergistic effects and parallel effects of the spatial evolution, as well as superimposed effects of the temporal evolution possibly occurring in complex domino evolution processes, are considered in this study. A case study demonstrates that the methodology is able to not only capture the spatial-temporal dimension but also to overcome the limitation of the “probit model” w.r.t only able to estimate the damage probability of the first level propagation. Besides, different from simulation or Bayesian approaches, our methodology can quickly provide evolution graphs (paths), the evolution time and the corresponding probability given a primary scenario. Therefore our approach can also be applied to domino risk assessment within an industrial park level and provide support for the allocation decision of safety and security resources.

Fuzzy FMECA (failure mode effect and criticality analysis) of LNG storage facility

Abstract: In high hazard industries like nuclear and chemical, a failure could lead to catastrophic events resulting in multiple fatalities, significant economic loss and reputation damage. Failure Mode and Effects Analysis (FMEA) is a safety and reliability analysis tool that systematically identifies the consequences of component failure on systems and determines the significance of each failure mode with regard to the system performance. FMEA uses a Risk Priority Number (RPN) for ranking the failure modes. RPN is commonly calculated as the product of the risk factors occurrence (O), severity (S) and non-detection (D) of the failure modes. When FMEA is used in criticality analysis of failure modes of components, it is also referred to as failure mode effect and criticality analysis (FMECA). Many researchers have pointed out the deficiencies of conventional risk priority number (RPN) used in FMEA. This paper describes a method for developing a fuzzy RPN (FRPN) that may rectify the limitations cited in the literature such as questionable nature of the mathematical formula for calculating RPN and the difficulties in prioritizing the RPN in the case of complex systems due to identical RPN values for different failure modes.The analysis was carried out by using fuzzy linguistic variables for occurrence, severity and non-detection and then using an if-then rule base to interconnect these variables to reach FRPN. The results obtained by using traditional FMECA and the fuzzy FMECA methods are compared. The difficulty in prioritizing the RPN in the case of complex systems such as LNG storage has been addressed in fuzzy FMECA.

A set pair analysis based layer of protection analysis and its application in quantitative risk assessment

Abstract: As an effective risk assessment method, layer of protection analysis (LOPA) is widely used in the evaluation of protection measures, i.e., independent protection layers (IPLs). However, traditional LOPA can only make semi-quantitative assessments for risk. Thus, assessment results with respect to risk will not be accurate or detailed enough, and the evaluation for IPLs may not be scientific or reasonable. By taking advantage of the quantitative analysis of the set pair analysis (SPA), a quantitative LOPA called the set pair analysis-layer of protection analysis (SPA-LOPA) is proposed in this study. The severity of the risk is judged by experts, and expert judgements are reflected by the connection degree (CD) while the corresponding algorithm for the CD is developed. In addition, the diversity degree (DD) and its algorithm are presented to process the CD with respect to the severity, and the assessed severity is measured by the calculated value of the DD. Next, the risk is quantified by the value of the DD and its frequency. Subsequently, the steps of the SPA-LOPA and corresponding assessment flowchart are provided. The SPA-LOPA and semi-quantitative LOPA are utilized to evaluate risks of gas leakage in biomass gasification. It is proved that the SPA-LOPA is more scientific and reasonable in the evaluation for IPLs, based on the comparison results.

CFD modeling of large-scale flammable cloud dispersion using FLACS

Abstract: The number of accidents in oil and gas refineries/storage terminals are increasing worldwide. Such events are disastrous to both human beings and infrastructure. It is therefore necessary to utilize the best methods to study worst-case scenarios associated with a process and/or plant. Computational Fluid Dynamics (CFD) models are appropriate to perform 3D modeling of major events with all necessary details. The present work reports the 3D CFD modeling of large-scale flammable cloud dispersion in a real configuration. The most widely approved CFD code for dispersion and explosion simulation FLACS (Flame Acceleration Simulator) is used to simulate the gases released with different flow rates in storage terminals. It was assumed that leak began near the pipes supplying fuel to the storage tank. The flow rate, surrounding condition and release duration were varied to study their influence on overall vapor cloud size i.e. diameter, height and explosive strength. Depending on the extent of LFL and UFL (Lower and Upper Flammability Limit) total flammable volumes of the clouds were predicted. It was found that such detailed modeling helped to understand the dispersion behavior much better than the phenomenological models. The strategic decisions on gas detectors layout can also be made for loss prevention and control. The simulation of worst-case scenario provided guidelines for pre- and post-incident mitigation measures.

Investigation of the acoustic emission characteristics during deformation and failure of gas-bearing coal-rock combined bodies

Abstract: With increasing mining depth, coal-gas compound dynamic disasters have become an important factor restricting mining safety. In the present study, conventional triaxial compression (CTC) tests were conducted on the gas-bearing coal, gas bearing coal-mudstone combination and gas bearing coal-sandstone combination using the RLW-500G triaxial experimental system. The gas bearing coal-sandstone combined samples were subjected to unloading tests, including unloading confining pressure (UCP) under constant axial tests and UCP-reloading axial stress (UCP-RAS) tests. In addition, the acoustic emission (AE) signals and permeabilities were measured simultaneously during the mechanical process. The experimental results indicate that the deformation of the coal-rock body is stronger under unloading conditions than in the CTC tests. Moreover, the damage of the coal-rock combination body is more severe in the UCP-RAS tests than in the UCP tests. Under lower confining pressure, the AE cumulative counts and the energy are higher for the gas-bearing coal-rock body than the gas bearing coal body. As the confining pressure increases, the AE cumulative counts and the energy are lower for the gas-bearing coal-rock body than for the gas bearing coal body. The AE cumulative counts and the energy of the three specimens under different stress paths decrease with the increase in the confining pressure or with the decrease in the gas pressure. The AE cumulative counts and energy of the gas-bearing coal-rock body are highest for the UCP-RAS test, followed by the UCP test and the CTC test. This study provides some references for understanding the mechanisms of coal-gas compound dynamic disasters and the basis for an early warning system.

Kick control reliability analysis of managed pressure drilling operation

Abstract: Offshore drilling involves complex operations and equipment; thus, faces many operational challenges, including well control. Managed pressure drilling has been proved to resolve most of these challenges; however, this technology, for the most part, is still in its infancy. This paper explores the safety and reliability assessment of a managed pressure drilling operation by investigating the kick control operation of constant bottomhole pressure technique of managed pressure drilling. In addition, this study seeks to understand the components interactions in an MPD system and their modes of failure. Failure scenarios are first built using a Fault Tree model and then analyzed using a Bayesian Network model. The reliability assessments of kick control operation are performed in two ways: a basic-components approach and a system-barrier elements approach. The analysis identifies communication related components, including network device damage as the most safety-critical component due to their high-level of influence while flowline and pump line blockages/rupture are ranked second-most critical but with limited-level of influence. However, the system-barrier element approach ranks the managed pressure drilling control system as the most safety-critical equipment. Further detailed analysis confirms that the monitoring equipment are the most safety-critical components of the managed pressure drilling control system with Coriolis flow meter and Rig pump exhibiting the most critical monitoring equipment. Additionally, the managed pressure drilling system's components show high degree of dependencies on one another and exhibit non-sequential modes of failure during kick control operation.

Characteristics and precursor information of electromagnetic signals of mining-induced coal and gas outburst

Abstract: Coal and gas outburst disasters can cause serious casualties and property losses, which can be effectively avoided by means of early warning. In this paper, based on the theoretical analysis of the electromagnetic radiation (EMR) response law of mining-induced coal and gas outburst, this paper studied the variation law and precursory characteristics of EMR signals before and after the outburst. The results show that: (1) Coal and gas outburst is induced under the disturbance of mining and promoted by gas pressure and highly concentrated mining stress. (2) The EMR signal is related to changes in stress state, fracturing activity and gas flow state in the process of mining, and the increase of the outburst danger will change characteristics of EMR signal in time and frequency domains. (3) In the time domain, the EMR intensity is positively correlated with the outburst danger of coal seam. The variable coefficient of EMR intensity is greater than 0.15 during the dynamic appearance. The peak value of variable coefficient of EMR intensity is positively correlated to the local instability of coal. (4) In the frequency domain, the amplitude of EMR grows, and the signal frequency moves toward a high level. The dominant frequency rises from 6.6 kHz to 17.4 kHz, and the EMR signal components become more complicated. In practical application, the evolution process of coal and gas outburst danger can be monitored in real time through a comprehensive use of EMR intensity value and its variable coefficient, waveform and frequency distribution characteristics, so that the occurrence of coal and gas outburst disasters can be eliminated.

Developing a dynamic model for risk analysis under uncertainty: Case of third-party damage on subsea pipelines

Abstract: Third-party damage is an important factor leading to subsea pipelines failure, and risk analysis an efficient approach to mitigate and control such events. However, available crisp probabilities for input events are usually limited, missing or unknown, which introduces data uncertainty. Furthermore, conventional risk analysis methods are known to have a static structure, which introduces model uncertainty. This paper presents a dynamic model for risk analysis under uncertainty and illustrates it by a case of third-party damage on subsea pipelines. Proposed model makes use of fuzzy set theory and evidence theory to handle data uncertainty, and utilizes Bayesian network (BN) to address model uncertainty. Primary accident scenario is developed by the FT-ESD approach, and it is transformed into BN to circumvent model uncertainty by relaxing the limitations of conventional methods. Expert elicitation is integrated into fuzzy set theory and evidence theory to obtain the crisp probabilities of input events in BN. Based on the model, a robust probability reasoning is conducted, through which the most probable factors contributing to the occurrence of unexpected consequence are identified. As new observations become available, potential accident probabilities are updated over time to produce a dynamic risk profile. The case study demonstrates the applicability and effectiveness of the model, which indicates that it is an alternative approach for risk analysis in the process industries under uncertainty.

Risk identification of third-party damage on oil and gas pipelines through the Bayesian network

Abstract: This paper aims to identify the risks influencing oil and gas (O&G) pipeline safety caused by third-party damage (TPD). After comprehensive literature study, we found that the traditional risk identification of TPD suffers from defining binary states of risk only and ignores the risk factors after pipeline failure. To overcome this problem, we investigated incident reports to identify previously unrecognized additional factors. This work also developed a graphic model by using Bayesian theory to cope with the multistate risks arising from third parties and to present the incident evolution process explicitly. Furthermore, this paper included a leakage case study conducted to verify the logicality of this model. The results of case study inspire that the proposed methodology can be used in a dual assurance approach for risk mitigation, namely learning from previous incidents and continuously capturing new risk information for risk prevention.

Failure probability analysis for emergency disconnect of deepwater drilling riser using Bayesian network

Abstract: Drilling risers are the crucial connection of subsea wellhead and floating drilling vessel. Emergency disconnect (ED) is the most important protective measure to secure the risers and wellhead under extreme conditions. This paper proposes a methodology for failure probability analysis of ED operations using Bayesian network (BN). The risk factors associated with ED operations and the potential consequences of ED failure were investigated. A systematic ED failure and consequence model was established through fault tree and event sequence diagram (FT-ESD) analyses and, then the FT-ESD model was mapped into BN. Critical root causes of ED failure were inferred by probability updating, and the most probable accident evolution paths as well as the most probable consequence evolution paths of ED failure were figured out. Moreover, the probability adaptation was performed at regular intervals to estimate the probabilities of ED failure, and the occurrence probabilities of consequences caused by ED failure. The practical application of the developed model was demonstrated through a case study. The results showed that the probability variations of ED failure and corresponding consequences depended on the states of critical basic events (BEs). Eventually, some active measures in drilling riser system design, drilling operation, ED test, and operation were proposed for mitigating the probability of ED failure.

Exploring inherent process safety indicators and approaches for their estimation: A systematic review

Abstract: The index-based approach is one of the most popular ways to measure the inherent safety degree of a chemical route or process during the early design stages. One of the main shortcomings of current indices is the limited set of aspects which are considered and are influencing inherent safety. In addition, the minimal knowledge of process designers regarding inherent safety hazards can exacerbate this problem. In this study, we identify the inherent safety indicators (within the period 1990–2017) used to measure the inherent safety degree of a process, and describe existing approaches to estimate these indicators. Bibliographic sites, including the Web of Science, ScienceDirect, Springer, ACS publications and Online Library, were searched based on various search strategies. A total of 62 resources were selected, and 35 indicators were found that were classified into six categories: (i) the 'chemical and physical properties of a chemical substance (11 indicators); (ii) the 'process conditions' (5 indicators); (iii) the 'equipment' (5 indicators); (iv) the 'reaction/decomposition properties' (3 indicators); (v) the 'activities and operations characteristics' (4 indicators); and (vi) the 'consequences' (7 indicators). We also found six estimation approaches, including the relative rating, an advanced mathematical approach (statistical, numerical descriptive and fuzziness), the risk-based, graphical, equational (or formula) based approach and the hybrid approach. This study can provide a quick guide for non-experienced researchers being enthusiast to work on inherent safety measurements using an index-based approach.

Process safety education: A literature review

Abstract: In this article, an extensive literature review has been carried out about process safety education. We drafted a process safety model able to systematize the literature review and investigated scientific papers as well as professional articles and so-called grey literature. The presence of a common background emerged, although possibilities for optimization of university curricula are possible, as well as harmonization within universities in different countries and between universities and industry. More collaboration in the field of process safety education is recommended, thereby also involving government agencies and/or control authorities and inspection bodies. In the light of the prevention of major accidents in the chemical industry, the process safety education topic deserves to receive more attention from all parties involved, that is, academia, industry and authorities.

Application of polymer nanocomposites in the flame retardancy study

Abstract: Application of polymer products is almost universal; however, the flammability of these hydrocarbon-based materials needs attention. Flame retardant additives have been studied for reducing the flame spread in a fire involving polymers. Polymer nanocomposites is a relatively new class of flame retardant material that has shown improved thermal stability and flammability properties and therefore, have received attention from the scientific community. This paper provides an overview of different types of flame retardant additives and their mechanisms, especially polymer nanocomposites. The synthesis and characterization techniques for analyzing the morphology, thermal and flammability properties of polymer nanocomposites have also been discussed. The overall objective is to provide a summary of the application of polymer nanocomposites in the field of flame retardancy and associated techniques to study these materials.

BLEVE fireball modeling using Fire Dynamics Simulator (FDS) in an Algerian gas industry

Abstract: BLEVE is one of major accidents observed in gas industry causing severe damage to people and environment. Its effects are manifested in three ways: shock wave propagation, fireball radiation and fragments projection. To assess these effects, risk decision-makers often use Quantitative Risk Analysis (QRA). In most cases, QRA data are obtained from empirical correlations. However, these correlations are not very satisfactory because they generally overestimate BLEVE effects and do not take into account geometry effects. In order to overcome the limitations of these empirical approaches, CFD modeling appears as a powerful tool able to provide more accurate data to better realize QRA. In this paper, the objective is to develop a CFD methodology in order to predict BLEVE thermal effects. Numerical simulations are carried out using the CFD code FDS. A sensitivity analysis of numerical models is performed in order to choose the right parameters allowing to model the fireball dynamics. The models retained are based on a single-step combustion using EDC model coupled with a LES turbulence model. Predictions show good agreement in comparison with results issued from three large-scale experiments. Furthermore, a case study on a propane accumulator in an Algerian gas processing unit is carried out.

Factors influencing and a statistical method for describing dust explosion parameters: A review

Abstract: The maximum and minimum values of dust explosion parameters are indeterminate because the possibility of explosion has not been quantified under conditions that approach the explosion limit. In this review, the influencing factors, laws, and mechanisms of dust explosion characteristics are described by comparing the differences in the methods for measuring the parameters in the ASTM, EN, GB, and ISO standards. The variation law of explosion parameters changes when crossing the limit values of the characteristic diameters and dust concentration. The similarities of certain influence mechanisms are also introduced. In addition, a new method is introduced to quantify explosion probability. The statistical nature of ignition is revealed by discussing the variability of the data overlap region of the “explosion” and the “no explosion” near the MEC and MIE measurements. The influence of sample size and positive factors on the statistical phenomena of ignitions were analyzed. The variability of the results increases when the conditions are not conducive to explosions. The values with 50% probability of explosion are independent of the number of trials and are less affected by the regression models. Reliable statistics can be obtained by sequential analysis methods in a small number of tests. The present work is a more accurate guide for the risk assessment of dust explosions, and introduces a potential statistical method to more accurately analyze the actual explosion characteristics.

STAMP-based causal analysis of China-Donghuang oil transportation pipeline leakage and explosion accident

Abstract: The methods used by researchers and accident investigators to analyze or investigate accidents are critical for understanding the underlying causes and for proposing improvement measures. As a systematic analysis model, Systems-Theoretic Accident Model and Processes (STAMP) is used in the aviation, maritime, and railway transportation industries, among others. To the best of the author's knowledge, no systematic analysis approach has been applied in the long-distance pipeline transportation industry. However, the pipeline transportation system is increasingly being considered a complex socio-technical system that requires the exploration of accident causes from a systematic viewpoint. The aims of this paper are to identify the causal factors of the China-Donghuang oil transportation pipeline leakage and explosion accident through a systematic method and illustrate the appropriateness of applying STAMP to the analysis of incidents in the long-distance pipeline transportation industry. To achieve this, a systematic analysis based on STAMP is conducted on the China-Donghuang oil transportation pipeline leakage and explosion accident. The analysis results expand the causal analysis beyond immediate failures, to causes from a systematic perspective, and illustrate the utility of applying the STAMP model to the pipeline transportation domain.

The impact of nature on chemical industrial facilities: Dealing with challenges for creating resilient chemical industrial parks

Abstract: In this paper, a conceptual framework is developed to improve NaTech safety in the chemical industry. The concept is called EPIC, indicating that emphasis should be put on Education, learning and training, Proactive risk minimization and safety innovation, Intensified informed inspection and analysis, and Cooperation and transparency. Concrete initiatives addressing NaTech for every domain of the EPIC conceptual framework are given. The innovativeness of the EPIC framework resides in the potential of the simultaneous application of initiatives within the four areas of improvement (E, P, I, and C) to make chemical clusters much more resilient with respect to nature-related disasters. As such, the proposed EPIC framework may lead to a much-needed revolution of NaTech safety in the chemical industry.

Estimating probability of success of escape, evacuation, and rescue (EER) on the offshore platform by integrating Bayesian Network and Fuzzy AHP

Abstract: Reliable Escape, Evacuation, and Rescue (EER) could have averted or reduced the catastrophic consequences of marine disasters. This paper presents a model to estimate the probability of successful EER on the offshore platforms. The proposed model consists of two parts. The first part uses fault tree method to analyze factors that affect the success of EER qualitatively, and these influencing factors are treated as nodes of Bayesian network (BN) in the next phase. A quantitative analysis model is constructed using BN in the second part. The BN is mapped from the fault tree model in the first part. Since one of the most important steps in BN analysis is to determine the CPT between nodes, fuzzy analytical hierarchy process (fuzzy AHP) and decomposition method are applied to estimate the CPTs of BN. Probabilities of successful EER are calculated by using BN, and the most influencing factors for the success of EER are determined based on sensitivity analysis (SA). In order to demonstrate the proposed method, a case study is made and results show that the proposed model is capable of finding out potential critical risk factors in EER and carrying out quantitative analysis of EER.

STAMP-based analysis of deepwater well control safety

Abstract: The complexity of a deepwater well control system makes defining appropriate safety requirements with traditional safety analysis methods difficult. Hence, there is a need for a complex systems approach for better understanding the development and prevention of accidents during deepwater drilling. Differing from traditional methods based on reliability theory, we use the system-theoretic accident model and process (STAMP) and system-theoretic process analysis (STPA) methods to establish a hierarchical control and feedback loop model of a well control system. In view of the characteristics of complexity and dynamism, the safety analysis is regarded as a system control and feedback problem in this work. Using this model, the systematic hazards and safety-related constraints are firstly defined, the safety control structure (SCS) is then established, the inappropriate control actions (ICAs) are identified in the next step, and the key factors that contribute to ICAs are finally determined. Guided by the STAMP/STPA, we construct the process of well control to simulate the consequences of ICAs during deepwater drilling by using the dynamic multiphase simulation software, OLGA. The simulation takes the lack of control actions and the provided control actions of a late shut-in as examples. The STAMP/STPA method proves to be an effective solution for evaluating the safety of deepwater well control from the perspectives of control and constraint. By modeling the process of a well kick, shut-in, and well killing, it was quantitatively indicated that the rational control actions within a certain time period can prevent accidents from occurring and escalating, thereby ensuring the safety of the system. The differences between traditional methods and STAMP/STPA are compared in this paper, and the limitations that need to be solved in the future are pointed out.

Influence of particle size polydispersity on coal dust explosibility

Abstract: The effect of particle size polydispersity (σD) on coal dust explosibility was studied using a 20-L spherical explosion test apparatus. Four kinds of blends with a fixed median diameter (D50) of 95 μm but varying σD of 1.18, 1.70, 1.96 and 2.40, were prepared by mixing original coal samples with relative low σD. Experimental results showed that the values of maximum pressure rise (Pex) and maximum rate of pressure rise ((dP/dt)ex·V1/3) of coal blends with the same D50 of 95 μm varied largely with different σD, and bigger σD produced larger Pex and (dP/dt)ex·V1/3 values. The Sauter mean diameter (D3,2) presented the best correlation between particle size and the explosion parameters for the blended coal samples. Risk assessment evaluation of coal dust should be reported in terms of D3,2 and σD to avoid serious underestimation.

Explosion hazards of aluminum finishing operations

Abstract: Metal dust deflagrations have become increasingly common in recent years. They are also more devastating than deflagrations involving organic materials, owing to metals' higher heat of combustion, rate of pressure rise, explosion pressure and flame temperature. Aluminum finishing operations offer a particularly significant hazard from the very small and reactive aluminum particles generated, and thus require high attention to details of operation and explosion safety management. This paper presents available statistics on metal dust explosions and studies the specific explosion hazards of aluminum finishing operations. The analysis of seven case studies shows that the proper design, monitoring and maintenance of dust collection systems are particularly important. Furthermore, the isolation of deflagrations occurring in dust collection systems, as well as good housekeeping practices in buildings, are critical safeguards to avoid the occurrence of catastrophic secondary explosions.

Deep neural network and random forest classifier for source tracking of chemical leaks using fence monitoring data

Abstract: Chemical plant leak accidents are classified as one of the major industrial accidents that can spread secondary and tertiary major disasters. It is very important to keep track and diagnose the source location(s) and notify the plant manager and emergency responders promptly to alleviate secondary and tertiary damages, improving the effectiveness of emergency responses. In this study, we propose an emergency response system that can cope with leak accidents of a chemical plant by monitoring sensor data and track down the suspected leak source using machine learning: Deep-learning and Random Forest classifiers. It is also difficult to get enough chemical leak accident scenario data or perform actual leak experiments on real plants due to high risk and cost factors. Consequently, Computational Fluid Dynamics (CFD) simulations are used to derive fence monitoring data for chemical leak accident scenarios. These data are to train the machine learning models to predict leak source locations. Six time-series Deep Neural Network (DNN) structures and three Random Forest (RF) structures are trained using CFD dispersion simulation results for 640 leak accident scenarios of a real chemical plant, divided as training and test datasets. As a result, on DNN model using 25 hidden layers and on RF model using 100 decision trees, 75.43% and 86.33% prediction accuracy are achieved, respectively, classifying the most probable leak source out of 40 potential leak source locations. Analyzing the predicted leak source locations that are wrongly classified, those predicted leak sources are also quite adjacent to the actual leak location and hardly called as misclassifications. Considering the superb performance of DNN and RF classifiers for chemical leak tracking, the proposed method would be very useful for chemical emergency management and is highly recommended for real-time diagnosis of the chemical leak sources.

How to address model uncertainty in the escalation of domino effects

Abstract: Modeling potential domino scenarios in process plants includes the prediction of the most probable sequence of events and the calculation of respective probabilities, so-called escalation probabilities, so that appropriate prevention and mitigation safety measures can be devised. Domino effect modeling, however, is very challenging mainly due to uncertainties involved in estimation of escalation probabilities (parameter uncertainty) and prediction of the sequence of events during a domino effect (model uncertainty). In the present study, a methodology based on dynamic Bayesian network is developed for identification of the most likely sequence of events in domino scenarios while accounting for model uncertainty. Verifying the accuracy of the methodology based on a comparison with previous studies, the methodology is applied to model single-primary-event and multiple-primary-event domino scenarios in process plants.

Risk analysis of subsea blowout preventer by mapping GO models into Bayesian networks

Abstract: Bayesian network (BN) is commonly used in probabilistic risk quantification due to its powerful capacity in uncertain knowledge representation and uncertainty reasoning. For the formalization of BN models, this paper presents a novel approach on constructing a BN from GO model. The equivalent BNs of the seventeen basic operators in GO methodology are developed. Therefore, the existing GO model can be mapped into an equivalent BN on basis of these developed BNs of the operators. Subsea blowout preventer (BOP) system plays an important role in providing safety during the subsea drilling activities. A case of closing the subsea BOP in the presence of pump failures is used to illustrate the mapping process. First, its GO model is presented according to the flowchart of the case. Then, BN is obtained based on the presented GO model. The developed BN relaxes the limitations of GO model and is capable of probability updating and probability adapting. Sensitivity analysis is performed to find the key influencing factor. The three-axiom-based analysis method is used to validate the developed BN.

Cause analysis and safety evaluation of aluminum powder explosion on the basis of catastrophe theory

Abstract: A safety state catastrophe model of aluminum product plant was established in this study by analyzing the causes of aluminum powder explosions. This model, which was developed from the perspectives of humans and objects, was created to explore the cause of aluminum powder explosions and establish a safety evaluation model. Weight is influenced by subjective factors in the safety evaluation of plants. Thus, a safety evaluation method based on catastrophe theory was proposed to analyze aluminum powder explosion. Fuzzy analysis method was combined with the hierarchical decomposition of the evaluation object. Normalization formula was used to compute the overall value of catastrophe subordinate function to realize the evaluation and dynamic judgment of the safety level of plants. Management system, operator, facilities and equipment, and the environment are the main factors that affect the safety level of aluminum product plants. The total value of safety catastrophe subordinate function of an aluminum product factory is determined by the minimum value of catastrophe subordinate function values at criterion layers. The four factors must be comprehensively considered. The evaluation method involves simple calculation, requires no index weighing, and realizes the dynamic and safe evaluation of aluminum product factories using multiple assessment indices.

How can we improve process hazard identification? What can accident investigation methods contribute and what other recent developments? A brief historical survey and a sketch of how to advance

Abstract: Risk assessment is essential for various purposes such as facility siting, safeguarding, and licensing. Hazard identification (HAZID), which suffers greatly from incompleteness, is still the weakest link in risk assessment. Of course, this recognition is not new and many efforts have been spent to improve the situation, of which some have been rather successful. To find out what can go wrong, creative divergent thinking is required. Hazard identification should result in scenario definition. In that respect, applying the present tools as HAZOP and FMEA there is still a great emphasis on the material and equipment aspects. In contrast, underlying management and leadership failure in its many forms reflecting in organizational and human failure, due to complexity, attracts much less attention. Unlike in HAZID, in accident investigation the occurrence of an event with nasty consequences is no doubt a fact, so there must be one or more causes and the traces will lead to them. Over the years, methods for accident and incident investigation have gone through a significant evolution. From the early-on simplistic domino stone model and the human operator always at fault, via models of latent failure due to failing management involvement and via extensive root cause analysis (RCA) to a system approach. Hence, in accident investigation, management failure appearing in the many possible forms of human and organizational factors, obtained already 30 years ago with the RCA technique much attention, while it nowadays culminates in the socio-technical system approach. So, the question arises whether for improved HAZID we can learn from the accident investigation experience. In addition, safer design and advances from static risk assessment towards more accurate predictive operational dynamic risk assessment and management, will also be enabled by possibilities offered by big data and analytics. Digitization, automation and simulation, hence computerization, will be of great help in improving the identification of hazards and tracing the corresponding scenarios. The paper reviews the developmental history of both accident investigation and hazard identification methodology; incidentally it will identify commonality and differences. On the basis of the comparison and of recent advances in computerization, the paper will investigate to what extent beneficial modifications and additions can be made to obtain a higher degree of completeness in HAZID.

In search of causes behind offshore incidents: Fire in offshore oil and gas facilities

Abstract: In order to eliminate fire incidents from occurring onboard offshore oil and gas facilities, it is crucial to have a better understanding of the causes behind them. Such understanding can be achieved through identification of the underlying causes that led to the previous incidents. Current paper focuses on analyzing investigation reports of 137 fire incidents reported to Bureau of Safety and Environmental Enforcement (BSEE) by oil and gas facilities located in the outer continental shelf of the US. The analysis digs as far as possible into the investigation reports to provide a statistical representation of the technical, operational, human and organizational factors that contributed to these incidents and to identify the lagging causes and the leading measures that needs to be tackled in order to prevent future disasters. Although the investigation reports indicated equipment failure and human error as the most common direct causes, further analysis showed that job safety analysis, procedure and maintenance related issues were the top contributors to such incidents.

Petri-net based evaluation of emergency response actions for preventing domino effects triggered by fire

Abstract: In industrial chemistry, many flammable materials are handled and/or stored in various facilities. It is possible that major fires occur at these facilities possibly leading to domino effects due to the failure of neighboring facilities caused by thermal radiation. It will take a certain time that thermal radiation causes nearby facilities to fail and that a domino effect occurs. The time needed for escalation to take place allows emergency actions as a response to the primary fire to prevent the propagation of the fire. In the present study, a Timed Colored Hybrid Petri-net (TCHPN) based methodology is introduced to evaluate different emergency response actions based on their efficiency in preventing or delaying the propagation of domino effects. A TCHPN model of emergency response to flammable liquid tank fire is established, and a time based analysis of emergency response actions for preventing domino effects is performed. Based on the simulation analysis, the probability of domino effects is calculated and response actions are compared.

A framework and method for the assessment of inherent safety to enhance sustainability in conceptual chemical process design

Abstract: Inherent safety indices have been suggested to choose between alternative process and chemical reaction routes in conceptual chemical process design. The indices are relative ranking methods that add up parameters without considering the difference in the magnitude of the hazard, complexity of the procedure, or expert opinion. We propose an improved framework based on fuzzy logic using chemical properties, process data, and chemical accident databases. The proposed methodology is applied to the methyl methacrylate (MMA) process as a case study. The results are compared with existing methods and experts' rankings by using three risk-rules, which are related to the experts' opinions and the tendency of decision makers. The risk-standard rule showed same results to that of the expert's scoring, while the ranking results are slightly different based on risk-easy and risk-hard rules. This methodology can facilitate the ranking of alternatives for decision making in the preliminary design stage.