Accident modeling of toxic gas-containing flammable gas release and explosion on an offshore platform

doi:10.1016/j.jlp.2020.104118

Journal of Loss Prevention in the Process Industries

Volume 65, May 2020, 104118

https://doi.org/10.1016/j.jlp.2020.104118 Get rights and content

Highlights

•
The chain accident of toxic gas-containing flammable gas release and explosion accidents is firstly concerned.
•
A time-varying leakage rate is adopted considering the impact of the emergency shutdown system and blowdown system.
•
The data-dump technique is adopted so that the dangerous gas leakage and explosion are simulated successively.
•
The risk-based concept and the grid-based concept are adopted to calculate the cumulative adverse effects.

Abstract

Toxic gas-containing flammable gas leak can lead to poisoning accidents as well as explosion accidents once the ignition source appears. Many attempts have been made to evaluate and mitigate the adverse effects of these accidents. All these efforts are instructive and valuable for risk assessment and risk management towards the poisoning effect and explosion effect. However, these analyses assessed the poisoning effect and explosion effect separately, ignoring that these two kinds of hazard effects may happen simultaneously. Accordingly, an integrated methodology is proposed to evaluate the consequences of toxic gas-containing flammable gas leakage and explosion accident, in which a risk-based concept and the grid-based concept are adopted to combine the effects. The approach is applied to a hypothetical accident scenario concerning an H₂S-containing natural gas leakage and explosion accident on an offshore platform. The dispersion behavior and accumulation characteristics of released gas as well as the subsequent vapor cloud explosion (VCE) are modeled by Computational Fluid Dynamics (CFD) code Flame Acceleration Simulator (FLACS). This approach is concise and efficient for practical engineering applications. And it helps to develop safety measures and improve the emergency response plan.

Introduction

Leakage and explosion are major potential chain accidents with high frequency (Ramsay et al., 1994; Kalantarnia et al., 2010; BP, 2010; Ottemöller and Evers, 2010; Yang et al., 2019). Several efforts have been made to predict the consequences of leakage and explosion accidents.

In the field of evaluation of flammable gas dispersion (Gupta and Chan, 2016), conducted a study to model the leakage and dispersion of flammable gas with a time-varying leakage rate. Instantaneous variation of flammable gas cloud was obtained. And the dispersion results obtained by the time-varying leak rate were compared with those obtained using the time-averaged constant leak rate. The study confirms that using the constant leakage rate may be relatively reasonable for systems with slow depressurization rates. But it may lead to inaccurate estimation results when it comes to systems that depressurize at a high rate.

In a study conducted by (Dadashzadeh et al., 2013), FLACS was utilized to model the dispersion of blowout gas and the subsequent vapor cloud explosion (VCE) against the “Deepwater Horizon” accident. Comparing the research results with the investigation results of BP, the authors demonstrated that FLACS was applicable to predict the consequences of dangerous gas leakage and explosion.

In a study conducted by (Savvides et al., 2001), taking the offshore modules as the object, a series of experiments with different configurations were simulated by FLACS. The simulations results were compared with measured data obtained by the JIP Module Experiments. The comparison has confirmed that FLACS can predict the dispersion behavior and accumulation characteristics of released gas with good accuracy.

(Middha et al., 2009) conducted validation effort for FLACS where a series of hydrogen dispersion experiments with different leakage conditions were simulated. Subsonic jets, sonic jets, impinging jets with low and high momentum, and liquid hydrogen releases were considered, respectively. In general, the modeling results are in reasonable agreement with the experimental data.

In a study conducted by (Li et al., 2018a, 2018b), a systematic CFD-based simulation was performed to predict the consequences of a subsea release. Accidental released gas from a subsea gas pipeline releases into seawater and rises to the sea surface. Then a gas pool is generated on the sea surface. The gas pool turns into a leakage source, and the flammable gas disperses into the atmosphere. Besides, a jack-up drilling platform is assumed to be located in the downwind area of the gas pool. The impact of dispersion and deflagration on this drilling platform was discussed. The corresponding risk management measures were also recommended by (Li et al., 2018a, 2018b).

In the field of toxic gas dispersion evaluation (Lovreglio et al., 2016), proposed an integrated approach to predict the consequences of toxic gas dispersion. The approach was applied to a hypothetical scenario where the dynamic assessment results were compared with the results obtained by the static approach which ignoring the evacuation movement, and the proposed approach proved to improve the accuracy of assessment results significantly.

(Lyu et al., 2018) pointed out the importance of performing the off-site consequence analysis for all chemical dealing companies. Moreover, the effect of safety measure regarding mitigation barriers was confirmed by two representative accident scenarios.

The consequences of flammable gas or toxic gas release, dispersion and consequent explosion accidents were extensively studied by Centre for Offshore Engineering and Safety Technology of China University of Petroleum (East China) (Deng et al., 2012; Liu et al., 2015; Zhu et al., 2010; Shi et al., 2019). Different research methods, such as experiments and numerical simulations, were used. A series of accident scenarios based on different parameters or different installations were studied (Li et al., 2019; Shi et al., 2018a, Shi et al., 2018b; Wei et al., 2014; Zhang and Chen, 2010; Zhu and Chen, 2010).

However, assuming that the flammable gas containing toxic gas, and exposed to ignition, the poisoning effect and explosion effect will occur continuously. There are comprehensive studies on consequence prediction about flammable gas containing toxic gas leakage accidents. E.g., a blowout accident involving hydrogen sulfide was simulated by (Yang et al., 2017). Poisoning risk area and explosion risk area are identified separately, i.e., the poisoning effect and explosion effect are considered respectively. Also, there are some attempts focusing on the domino effect (Khakzad et al., 2013, 2018; Ni et al., 2016), predicting the cumulative adverse effects due to accident escalation. However, the theory of domino effect is of inadequacy to the current work, since the domino effect always refers to the accident starting from one unit and spreading to different units with great randomness and strong uncertainty. The current study focuses on the flammable gas containing toxic gas leakage and explosion accident, in which only a facility is involved and the cumulative adverse effects are inevitable.

The above studies did not consider the dual consequences against flammable gas containing toxic gas release and explosion accident, i.e., the poisoning effect and the explosion effect were considered respectively. A combination of these two kinds of effects is imperative because the poisoning effect and explosion effect are inevitable during flammable gas containing toxic gas leakage and explosion accident. Accordingly, comprehensive consideration of the explosion effect and the poisoning effect is emphasized, and a systematic approach is proposed against similar accidents. Then the proposed approach is presented through a hypothetical scenario concerning H₂S-containing natural gas leakage and explosion chain accidents on an offshore platform. The main innovation of the current study is that the explosion effect and the poisoning effect are considered comprehensively, while the previous studies take these two hazard effects into account separately. The dispersion of the released gas and the subsequent VCE are predicted by FLACS. FLACS is a leading 3D CFD software which has been widely used in process industry (Huser and Kvernvold, 2000; Qiao and Zhang, 2010; Li et al., 2014; Huang et al., 2017; Shi et al., 2018a, Shi et al., 2018b), and the availability and accuracy of the tool have been validated against numerous experiments with different scales and different scenarios (Savvides et al., 2001; Middha et al., 2009; Hansen et al., 2010; Bleyer et al., 2012).

Section snippets

Review of past accidents

In 2007, a natural gas containing hydrogen sulfide leakage accident occurred in the Kab-121 platform in the Gulf of Mexico (Yang et al., 2019). Then a fire accident followed due to the contact between an unknown source of ignition and the vapor cloud. The consequences caused by fire and hydrogen sulfide were 21 deaths.

In January 2018, a tanker collision accident between bulk freighter CF Crystal and tanker Sanchi happened in the East China Sea (Li et al., 2018a, 2018b), which caused a cargo

Integrated methodology

Fig. 1 introduces the proposed methodology for toxic gas-containing flammable gas leakage and explosion accident consequence modeling. The detailed procedure is described below.

The first step is geometric modeling. In this step, the geometric modeling is conducted by the pre-processor of FLACS. All structural components can be converted into primitives, such as boxes, cylinders, ellipsoids, generally truncated cones, and complex polyhedrons. However, in general, only the basic primitives,

Turbulence model

The gas dispersion process satisfies the continuity equation, the laws of mass conservation, momentum conservation, and energy conservation. The dispersion process also satisfies mixture fraction and fuel mass fraction transport equation considering the multiple compositions in released gas. All these conversation equations can be represented in general as (GexCon, 2015): $\frac{\partial}{\partial t} (ρ φ) + \frac{\partial}{\partial x_{j}} (ρ u_{j} φ) = \frac{\partial}{\partial x_{j}} (ρ Γ_{φ} \frac{\partial φ}{\partial x_{j}}) + S_{φ}$ where ϕ represents the general variable, including variables such as mass, momentum,

Case study

The proposed methodology is applied to a hypothetical H₂S-containing natural gas leakage and explosion accident on an offshore platform. The offshore platform consists of a process module and an accommodation module. There are three decks in the process module, i.e., the main deck, the middle deck, and the lower deck. Process equipment is mainly concentrated on the middle deck and the lower deck, including power device, compressor, fuel gas device, acid gas, dehydration & mercury removal

Poisoning assessment

The monitored data for the concentration of H₂S of each monitoring point is extracted, and the concentration of H₂S at every moment is of concern, i.e., the H₂S concentration-time curve at each coordinate is expected rather than the maximum concentration. Then the probit variables of toxic gas can be calculated according to Eq. (2). Converting the probit variables into probabilities of fatality (Eq. (3)), then the risk index of the poisoning effect based on the probit model can be estimated

Conclusions

Toxic gas-containing flammable gas leakage is very dangerous, and catastrophic accidents may be induced if the gas is exposed to an ignition source. Both the poisoning effect and explosion effect do indeed exist and cannot be neglected in the chain accident. Moreover, these two hazard effects have nothing with the domino effect.

An integrated approach is proposed to fill the gap that the poisoning effect and explosion effect are considered separately in the current efforts. To evaluate the

CRediT authorship contribution statement

Dongdong Yang: Writing - review & editing, Writing - original draft, Investigation, Formal analysis, Methodology, Conceptualization. Guoming Chen: Supervision, Project administration, Funding acquisition. Ziliang Dai: Validation, Data curation.

Acknowledgments

The authors gratefully acknowledge the financial support provided by the National Key R&D Program of China (No: 2017YFC0804501).

References (45)

A. Bleyer et al.
Comparison between FLACS explosion simulations and experiments conducted in a PWR Steam Generator casemate scale down with hydrogen gradients
Nucl. Eng. Des.
(2012)
M. Dadashzadeh et al.
Explosion modeling and analysis of BP Deepwater Horizon accident
Saf. Sci.
(2013)
S. Gupta et al.
A CFD based explosion risk analysis methodology using time varying release rates in dispersion simulations
J. Loss Prev. Process. Ind.
(2016)
O. Hansen et al.
Validation of FLACS against experimental data sets from the model evaluation database for LNG vapor dispersion
J. Loss Prev. Process. Ind.
(2010)
O. Hansen et al.
Equivalent cloud methods used for explosion risk and consequence studies
J. Loss Prev. Process. Ind.
(2013)
Y. Huang et al.
Multi-level explosion risk analysis (MLERA) for accidental gas explosion events in super-large FLNG facilities
J. Loss Prev. Process. Ind.
(2017)
A. Huser et al.
Explosion risk analysis - development of a general method for gas dispersion analyses on offshore platforms
Parall. Comput. Fluid Dyn.
(2001)
M. Kalantarnia et al.
Modelling of BP Texas City refinery accident using dynamic risk assessment approach
Process Saf. Environ. Protect.
(2010)
N. Khakzad et al.
How to address model uncertainty in the escalation of domino effects?
J. Loss Prev. Process. Ind.
(2018)
J. Li et al.
New correlation for vapor cloud explosion overpressure calculation at congested configurations
J. Loss Prev. Process. Ind.
(2014)

J. Li et al.

Gas dispersion risk analysis of safety gap effect on the innovating FLNG vessel with a cylindrical platform

J. Loss Prev. Process. Ind.

(2016)

J. Li et al.

Optimal blast wall layout design to mitigate gas dispersion and explosion on a cylindrical FLNG platform

J. Loss Prev. Process. Ind.

(2017)

X. Li et al.

Gas dispersion and deflagration above sea from subsea release and its impact on offshore platform

Ocean Eng.

(2018)

X. Li et al.

Analysis of underwater gas release and dispersion behavior to assess subsea safety risk

J. Hazard Mater.

(2019)

R. Lovreglio et al.

A dynamic approach for the impact of a toxic gas dispersion hazard considering human behaviour and dispersion modelling

J. Hazard Mater.

(2016)

B. Lyu et al.

Additional data on damage reduction strategies against chemical accidents by using a mitigation barrier in Korean chemical risk management

Data Brief

(2018)

P. Middha et al.

Validation of CFD-model for hydrogen dispersion

J. Loss Prev. Process. Ind.

(2009)

Z. Ni et al.

Relative risk model for assessing domino effect in chemical process industry

Saf. Sci.

(2016)

A. Qiao et al.

Advanced CFD modeling on vapor dispersion and vapor cloud explosion

J. Loss Prev. Process. Ind.

(2010)

C. Ramsay et al.

Quantitative risk assessment applied to offshore process installations. Challenges after the piper alpha disaster

J. Loss Prev. Process. Ind.

(1994)

J. Shi et al.

Vented gas explosion overpressure prediction of obstructed cubic chamber by Bayesian Regularization Artificial Neuron Network–Bauwens model

J. Loss Prev. Process. Ind.

(2018)

J. Shi et al.

Robust data-driven model to study dispersion of vapor cloud in offshore facility

Ocean Eng.

(2018)

Cited by (23)

Research on gas diffusion and explosion characteristics in a ship engine room
2024, International Journal of Hydrogen Energy
This study focuses on potential gas leakage scenarios during the natural gas transportation process in ship engine room. A three-dimensional model of the ship engine room is constructed to investigate the diffusion patterns of leaked gas and the resulting explosion and disaster risks. The results reveal that natural gas exhibits the highest concentration along the jet direction, with the concentration remaining above 5% by 360 s after the leakage. When an explosion intensity of 5 is chosen, the calculated results from TNO multi-energy method show an average relative error of less than 10% compared to the numerical simulations beyond 20 m. In the rear of obstacles, the peak overpressure decreases by 60%, and the time to reach the peak overpressure extends to 1.8 times the original duration. The range of slight building damage caused by explosion overpressure is the largest, averaging about twice the range of minor injuries to personnel.
A hybrid association analysis framework of accident reports for offshore drilling platforms
2023, Journal of Loss Prevention in the Process Industries
As the key work unit of offshore oil and gas exploration and exploitation activities, an offshore drilling platform (ODP) integrates high-density equipment and facilities in a relatively narrow area, making it prone to frequent serious accidents under the influence of environmental factors. The mining association rules of the hidden dangers from the accumulated historical accident reports can help find the potential causes of the risks and proactively reduce the risk of accidents and damages. However, due to the professionalism of the corpus linguistics of the offshore oil and gas industry and the complex correlations between multiple named entities, mining valuable information from massive transaction datasets has increasingly become a challenge for academia and industry. To this end, a hybrid association analysis framework for ODPs is proposed in this paper, which is divided into four stages: data preprocessing, information extraction, association rule mining and making a rectification plan. In the key information extraction stage, we propose a three-layer deep learning network based on the BERT-IDCNN-BiLSTM-CRF (BIBC) model, which makes effective use of the semantic information of the context to extract the trituples of the hidden dangers (i.e., equipment, location and phenomenon). Based on the extracted entities, an effective multientity association rule mining algorithm based on hash technology and an inverted itemset (HTApriori) as a variant of Apriori for ODP databases is developed to understand the many-to-one, one-to-many and many-to-many risky causes between the entities. A series of real-world experimental results show that the proposed framework offers more reliable and effective performances of information extraction and association rule mining than the comparison methods. By analyzing the association results, the hidden dangers and their association rules, which occur with some small probability but are critical in ODPs, can be found. This work is helpful in significantly improving the rationality of the rectification plans. In addition, the application results on the CoNLL2003 dataset further prove the generalization performance of the proposed method.
Numerical simulation on natural gas explosion and prevention measures design under water–gas compartment in utility tunnel
2022, Tunnelling and Underground Space Technology
Currently, in utility tunnels in China, the gas pipelines were installed under an isolated compartment, which meant the cost of installation and maintenance is high, and the novel design of the water–gas compartment was proposed to lay water pipeline and gas pipeline. The possible explosive accidents were obtained by simulation software FLACS, including shock wave and flame damage. The results show that the shape of vapor cloud influences the explosive intensity more than fuel mass, and the maximum overpressure shows at the fire-proof partition rather than the explosive center. The shock wave will hurt people and the flame will influence the adjacent compartment. Pressure-relief panels and fine water mist were examined to compare the protective effects, and the pressure-relief panels are more effective. The results are helpful in evaluating future water–gas compartment design and related explosive prevention measures, also instructive in escaping for personnel in utility tunnels.
Numerical modelling of wind-influenced above sea gas dispersion and explosion risk analysis due to subsea gas release on multileveled offshore platform
2022, Applied Ocean Research
Citation Excerpt :
In fact, studies have utilised FLACS to perform large-scale dispersion modelling in offshore platforms (Shi et al., 2018; Dasgotra et al., 2018). Furthermore, taking offshore platforms as the object of study, studies performed combined gas dispersion and subsequent explosion analysis using FLACS (Qiao & Zhang, 2010; Yang et al., 2020). Due to increased industry demand and operations, natural gas underwater pipeline leakages frequently occur posing risk to the above sea environment and the people offshore (Jackson et al., 2014).
Above-sea gas dispersion due to subsea release could lead to explosion on nearby offshore facility. Less congested single-deck offshore platform has been considered in past for numerical modelling of above-sea gas dispersion and deflagration. Offshore platforms are designed compactly and comprise of multiple decks due to space restrictions. The approaching flammable gas may disperse differently at each deck, and high overpressures could develop in event of explosion in confined decks causing severe damages. Therefore, investigating effect of above-sea gas release and dispersion and resulting vapour cloud explosion on equipment damage and human fatality in complex multi-deck platform is worth investigating. This paper presents above-sea gas dispersion and explosion modelling on realistic multileveled fixed offshore platform under varying wind speeds using FLACS. Simplified steps are provided to model gas release from large area sea surface gas pool along with atmospheric dispersion and subsequent explosion in FLACS. It was found that generally, lower decks had near-stoichiometry concentrations thus being most susceptible to explosion consequences. High wind speed increased flammable cloud accumulation, where 50-60% of decks were filled at 3m/s wind speed whereas 100% filled under 7m/s. It is suggested to have adequate ventilation on lower platform decks specially under high wind circumstances to avoid possible explosion hazard due to cloud accumulation. Congested and confined regions with low flammable concentration produced more overpressure than the areas with same concentration but no or low congestion and confinement. Stoichiometric concentration together with congestion produced incredibly high localised overpressure reaching 3.55-4 barg. The probability of equipment damage and human fatality was reaching 1 for explosions at base and deck 2 under 7 m/s wind speed. Since roof was open, the probability of equipment damage and human fatality was least.
Deep Neural Network-based Optimization Framework for Safety Evacuation Route during Toxic Gas Leak Incidents
2022, Reliability Engineering and System Safety
Citation Excerpt :
Based on the outstanding accuracy of CFD modeling, numerous studies have attempted to utilize CFD results for toxic gas dispersion for various practical purposes in industrial facilities. Yang et al. evaluated the consequences of both toxicity and flammability on an offshore platform based on the output of a CFD dispersion simulation [19]. The mitigation effects of a physical barrier [20], a water curtain [21], and strategic decisions regarding the gas detector layout [22] can be assessed using three-dimensional CFD modeling of toxic and flammable gas dispersion.
Evacuation planning is important for reducing casualties in toxic gas leak incidents. However, most evacuation plans are too qualitative to be applied to unexpected practical situations. Here, we suggest an evacuation route proposal system based on a quantitative risk evaluation that provides the safest route for individual evacuees by predicting dynamic gas dispersion with a high accuracy and short calculation time. Detailed evacuation scenarios, including weather conditions, leak intensity, and evacuee information, were considered. The proposed system evaluates the quantitative risk in the affected area using a deep neural network surrogate model to determine optimal evacuation routes by integer programming. The surrogate model was trained using data from computational fluid dynamics simulations. A variational autoencoder was applied to extract the geometric features of the affected area. The predicted risk was combined with linearized integer programming to determine the optimal path in a predefined road network. A leak scenario of an ammonia gas pipeline in a petrochemical complex was used for the case study. The results show that the developed model offers the safest route within a few seconds with minimum risk. The developed model was applied to a sensitivity analysis to determine variable influences and safe shelter locations.
Quantitative risk assessment of gas leakage and explosion accidents and its security measures in open kitchens
2021, Engineering Failure Analysis
Citation Excerpt :
Dan et al. [22] assessed the risk of fire and explosion in the LNG floating production, storage and offloading by quantitative risk analysis, including frequency and consequence analyses. Yang et al. [23] developed an integrated methodology to evaluate the toxic gas-containing flammable gas leakage and explosion consequences using CFD code FLACS. A risk-based concept and the grid-based concept were adopted to combine theses effects.
In this study, a quantitative risk assessment methodology is developed for gas leakage and explosion accidents in open kitchens. Meanwhile, the corresponding security measures are investigated. The physical and computational model of a typical open kitchen and the corresponding traditional kitchen is firstly developed, using CFD code FLACS, respectively. Based on the similarity principle, a small scale low-pressure natural gas leakage and explosion experiment platform like a kitchen was constructed to validate the accuracy of CFD model. Subsequently, gas leakage and dispersion behavior is performed to analyze the gas concentration distribution. Meanwhile, a time-of-exposure model is introduced to calculate the overall ignition probability during the whole gas leakage and dispersion process. And then, the explosion simulation with a real inhomogeneous flammable gas cloud is carried out to analyze the overpressure distribution. In addition, the maximum weighted average overpressure is estimated accurately. According to the overall ignition probability and the maximum weighted average overpressure, the integrated risk caused by gas leakage and explosion accidents is assessed. Furthermore, the active and passive security measures are investigated to reduce the integrated risk in open kitchens. This methodology and its security measures are applied to a typical open kitchen, located in Harbin, China. The results show that the simulation results reproduced by the CFD model are in good agreement with the experimental results. The explosion accident probability in open kitchens is much higher than that in traditional kitchens, whereas the explosion accident consequences in open kitchens are less than that in traditional kitchens. The integrated risk in open kitchens is greater than that in traditional kitchens. These achievements provide an effective theoretical guidance and technical support for safety gas supply design of open kitchens.

View all citing articles on Scopus

View full text

Accident modeling of toxic gas-containing flammable gas release and explosion on an offshore platform

Highlights

Abstract

Introduction

Section snippets

Review of past accidents

Integrated methodology

Turbulence model

Case study

Poisoning assessment

Conclusions

CRediT authorship contribution statement

Acknowledgments

Nucl. Eng. Des.

Saf. Sci.

J. Loss Prev. Process. Ind.

J. Loss Prev. Process. Ind.

J. Loss Prev. Process. Ind.

J. Loss Prev. Process. Ind.

Parall. Comput. Fluid Dyn.

Process Saf. Environ. Protect.

J. Loss Prev. Process. Ind.

J. Loss Prev. Process. Ind.

J. Loss Prev. Process. Ind.

J. Loss Prev. Process. Ind.

Ocean Eng.

J. Hazard Mater.

J. Hazard Mater.

Data Brief

J. Loss Prev. Process. Ind.

Saf. Sci.

J. Loss Prev. Process. Ind.

J. Loss Prev. Process. Ind.

J. Loss Prev. Process. Ind.

Ocean Eng.