Journal of Loss Prevention in The Process Industries最新文献

Methanol, as an important chemical raw material and clean energy source, easily forms explosive droplet clouds during its production, processing, and usage. Therefore, understanding the evolution and propagation mechanisms of methanol spray explosion flames is crucial for its widespread use and the design of safety measures. This paper studies the flame propagation behavior of methanol spray explosions using a 16.2-L visualized enclosed vessel. The results indicate that the flame structure of methanol spray explosions is similar to that of premixed gas explosions but significantly differs from traditional fossil fuels. It demonstrates homogeneous combustion characteristics primarily governed by the kinetics-controlled regime. As the methanol spray concentration increases, both the maximum flame propagation speed and the maximum explosion pressure initially increase and then decrease, reaching their peak at a concentration of 224.68 g/m³, with values of 9.96 m/s and 0.72 MPa, respectively. The heat losses during combustion exhibit a trend opposite to that of explosion pressure. Numerical simulation results indicate that the concentrations of key radicals such as O, H, and OH vary significantly between lean and rich combustion states. The increase in H radical concentration enhances the elementary reactions that suppress flame temperature and promotes chain-terminating reactions.

Extinguished lithium-ion battery fires can experience temperature rebound and re-ignition, necessitating proper fire extinguishing and cooling measures for battery safety. This study introduces a novel approach that integrates C₆F₁₂O and water mist as a strategy to prevent the spread of thermal runaway in lithium-ion batteries, evaluating the fire-suppression and cooling ability under various intermittent spray modes. The findings demonstrate that the strategy has a stronger suppression effect compared to continuous water spray and could prevent the spread of thermal runaway. C₆F₁₂O can extinguish fires in just 1 s. During the cooling phase, the cooling capacity generally decreases with increasing cycle period and duty cycle. Among different cycle periods, C₆F₁₂O combined with intermittent water mist spray (DC = 0.5, Pt = 2 s) exhibited the most effective cooling, significantly reducing peak temperature and providing the highest heat suppression. In terms of duty cycles, C₆F₁₂O combined with intermittent water mist spray (DC = 0.1, Pt = 20 s) achieved the longest cooling duration, lowering the temperature of Cell #3 to below 50 °C. The combination of C₆F₁₂O and intermittent water mist spray rapidly extinguishes flames, maximizes the cooling impact of water mist, and ultimately hinders the propagation of thermal runaway.

Large-scale plants, with their concentrated facilities, are at risk of domino effects in the event of escalating accidents. Specifically, the fire-induced domino effects can have considerable consequences. The prevalent probit model is inherently limited to capturing the synergistic effects of multiple fires, thereby constraining the assessment of the overall consequences subject to domino effects. Therefore, this paper introduces a novel numerical method that considers the cumulative effect of all escalation vectors, employing the concept of ‘residual thermal dose’. Monte Carlo simulations were used to model the uncertainties related to effective mitigation measures. Subsequently, we developed two formulas to assess overall human vulnerability to domino effects, providing a visualisation of potential consequences through vulnerability maps. We validated the merits and applicability of our methodology through a case study of a 50-tank oil storage plant. The proposed method can facilitate vulnerability assessment of domino accidents within large-scale plants and identification of critical hazard installations, thereby supporting risk assessment and security management.

In this work, a novel Physics-of-Failure (PoF) model-based framework is proposed for accounting for the aging of safety barriers in the risk assessment of industrial facilities. A PoF model of aging is integrated into the barriers fragility models to provide fragility surfaces as functions of both the hazard magnitude and the barrier age. Such modelling is integrated in a Dynamic Bayesian Network (DBN) for the risk assessment. The proposed framework is applied to a case study of a chemical facility exposed to Natural hazard-induced Technological (NaTech) risk due to earthquakes. The results demonstrate the feasibility and effectiveness of the framework.

This paper outlines the challenges and opportunities involved in developing a safe and sustainable hydrogen infrastructure. The growing global energy demand and environmental impacts of fossil fuels have sparked interest in alternative energy sources. Hydrogen, as an environmentally friendly and sustainable energy carrier, offers a promising solution. However, the widespread adoption of hydrogen technologies faces significant safety and data reliability challenges. This paper reviews existing literature on hydrogen safety, encompassing hydrogen leak diffusion, fire and explosion, hydrogen deflagration to detonation transition (DDT), risk assessments, and mitigation techniques associated with different hydrogen facilities. Multiple approaches, including probabilistic risk analysis, computational fluid dynamics (CFD), experimental measurements, and machine learning algorithms (MLAs), to ensure hydrogen safety are also explored. Existing hydrogen-related accidents are also extensively analysed. Despite the progress in hydrogen safety research, challenges and limitations still exist. These include a lack of reliable data, limited AI applications due to data availability issues, the need for safe and economic hydrogen storage, and the importance of providing personnel with adequate safety awareness and knowledge. Moreover, the article identifies future research opportunities in investigating auto-ignition mechanisms, collecting more experimental data, integrating AI and CFD to investigate hydrogen dispersion behaviour, exploring the sensor's technology, developing inherently safer designs, and studying the integrated impacts of evolving accident scenarios. In conclusion, the paper emphasises the importance of addressing safety challenges to establish a secure and dependable hydrogen infrastructure. It highlights the need for further research to enhance safety protocols, establish robust standards, and support the long-term sustainability goals of the hydrogen industry. The insights provided in this study can contribute to identifying research areas, improving safety measures, and developing future hydrogen infrastructure.

Small unmanned aerial vehicles (UAVs) are increasingly utilized in a variety of environments, including those containing flammable substances, where the risk of ignition is a significant concern. In this study, we investigate the risk of sparks generated by UAV propeller collisions with personal protective gear to ignite these gases. We utilized a high-speed infrared camera to measure the temperatures of sparks produced during such collisions. The experiments involved collisions between various types of protective gear and rotating propellers, during which spark generation was observed upon impact. The high-speed infrared camera recorded the spark temperatures at the moment of impact. The sparks produced by these collision reached temperatures ranging from 375 [°C] and 900 [°C], exceeding the auto-ignition temperatures of hydrogen, diesel oil, and gasoline vapors. Notably, even protective gear that did not produce visible sparks reached temperatures up to 375 [°C], which surpasses the auto-ignition temperatures of some flammable substances, such as gasoline and kerosene. Based on the experimental results, we discuss the potential risks of igniting hydrogen, diesel oil, gasoline, and kerosene due to propeller collisions with protective equipment.

The computational explosion venting analyzer (EVA), a zero-dimensional (0D) model based on the conservation of mass and energy, is being developed to model centrally- and rear-ignited explosions. For this purpose, a series of explosion venting experiments in a cylinder with vent areas of $132.7$, $86.6$, $67.9$ cm² (corresponding to the vent area ratios of $K=0.47,0.31,0.24$, respectively) is performed. Two equivalence ratios of $\varphi =0.8$ and $1$ are considered to represent the fuel-lean and stoichiometric methane-air mixtures, respectively. The dynamics of explosions is studied through the observation of flame propagation and pressure measurements. In rear ignition experiments, laminar, so-called “finger flame” propagation is observed, while in the case of center ignition, a flame initially expands spherically and then is pulled by the vent, acquiring a half-elliptical and half-spherical shape. The peak pressures obtained from rear ignition exceed their counterparts in the center ignition experiments. The EVA is compared with the experimental matrix. No turbulence is implemented in stoichiometric simulations, and slight turbulence has been accounted for in the lean mixture simulations. It is found that the large vent, generally, imposes more disturbances on the flame shape and the fuel-lean mixtures are more prone to the diffusional-thermal instabilities. It is shown that such a simple numerical tool, as the EVA is, can estimate a complicated problem such as pressure evolution resulted from a vented gas explosion.

To investigate the fire danger of the valve hall, a 3D numerical model of the DC converter substation's valve hall is created by using the fire dynamics simulator (FDS) program and the fire burning process of the valve hall is simulated. In particular, the procedure of simulating the fire burning in the valve hall under various fire source positions is accomplished by varying certain parameters. The smoke spreading process, ambient temperature field, and heat release rate curve under various fire source placements are compared based on the results of the simulation calculations. The findings demonstrate that under various fire source locations, the valve hall's combustion process varies as well. This study compares the time for smoke to fill the entire valve hall, the maximum temperature within the valve hall, and the maximum fire source heat release efficiency in different fire scenes. The results indicate that in fire scenes where the fire source is closer to the middle position, the time for smoke to spread throughout the valve hall is shorter, and the fire source heat release rate is higher. Conversely, in fire scenes where the fire source is closer to the edge, the maximum temperature within the valve hall is higher. The numerical simulation of the valve hall in this DC converter station assessed the hazards under different fire scenes, aiming to maximize the safety of the valve hall. It provides reliable guidance for maximizing the safety of the valve hall and facilitating firefighting and rescue efforts, thus protecting personal safety and minimizing property damage.

In the present study, Computational Fluid Dynamics (CFD) simulations were carried out on a separator subjected to three different fire scenarios: pool fire, jet fire and fireballs. The empirical models applicable to such plants are not developed and conventional radiation models leads to an overestimation of the heat flux and safety distances. Therefore an alternative tool like CFD along with empirical models is used to model the potential scenarios mentioned above in a model oil and gas separator plant. The developed model were validated by comparing literature data on the flame length, maximum flame temperature and its location above pan surface in case of pool fire and comparing them with literature data. It is found that in case of flame length the deviation from literature data was about 10% and less than 5% for the maximum flame temperature. For jet fire the average surface temperature was 561 K while this value is 423 K in case of pool fire and 317 K for fireballs. When it comes to the radiative flux, an average value of 29 kW/m${}^2$ is found for jet fires while this value was only 14 kW/m${}^2$ for pool fire and 23 kW/m${}^2$ for fireballs. The thermal loads on the separator vessel due to above fire scenarios were also calculated. The obtained value of thermal stresses were around 285 MPa and 600 MPa for pool and jet fires respectively. Thus it can be concluded that from safety point of view jet fires are the most critical one. The results obtained form simulations were also compared with that of data from empirical models.

Conventional safety analyses in complex systems like air separation units (ASUs) often attributed accidents to linear, deterministic causes, such as operator error. However, acknowledging the intricate interdependence of process components necessitates a shift towards recognizing the complexity of incident causation. This study proposes a novel model that integrates Function Resonance Analysis Method (FRAM) and fuzzy logic analysis to address this growing need. The model facilitates the identification of emerging risks and assesses the impact of influential factors within a mixed qualitative and quantitative framework. The FRAM method is initially employed to identify emerging risks within the ASU. Subsequently, fuzzy multi-criteria decision-making methods are utilized to establish the relationships and weightage of influential factors. Data collection encompasses semi-structured interviews, direct observation, process workflow analysis, and the involvement of a panel of engineers and operators from the investigated ASU. Utilizing FMV software for FRAM analysis, functions associated with air compression, distribution, and storage exhibit high resonance. This signifies substantial variability and a heightened potential for incidents or deviations in these functions and higher-level tasks. Furthermore, Fuzzy TOPSIS analysis reveals that education and experience emerge as the most impactful factors governing newly emerging risk. This model demonstrates significant merit for risk assessment and incident investigation. Its non-linear and dynamic nature empowers the proactive identification and examination of processes, incidents, and emerging risks before deviations or accidents occur.