大气扩散模型统计性确认评估

使用计算模型预测物理过程之前,必须对该计算模型进行论证,确认评估通过后才能用于物理状态和过程的计算和预测。讨论了基于数据统计属性的确认评估对面向大气扩散物理现象的计算模型之适用性;探究了模型评估工作的基本内容、关键概念和通用原则;给出了确认评估参数的定义、选择原则及示例;分析了评估度量指标的内涵和适用条件;探讨了可接受模型需要满足的指标准则及统计性确认评估指标的选用;最后,指出了基于统计度量的确认评估方法的改进方向。扩散模型评估理论的探讨有助于提高基于模型仿真的大气扩散研究的准确性,也有利于高精度模型和基准试验的设计、开发与遴选。     

HAZDIG: a new software package for assessing the risks of accidental release of toxic chemicals

HAZDIG (HAZardous DIspersion of Gases) is a user-friendly PC- based software for generating scenarios for the emissions and gaseous dispersion of hazardous chemicals. It can simulate accidental as well as normal release but has been specifically developed as a tool for studying accidental release of hazardous chemicals and the consequences. HAZDIG is made-up of five main modules—data, release scenario generation, dispersion, characteristics estimation, and graphics. HAZDIG incorporates the latest models for estimating atmospheric stability and dispersion. The data needed to run the models is easy to obtain and feed—properties of chemicals, operating conditions, ambient temperature, and a few commonly available meteorological parameters. A database containing various proportionality constants and complex empirical data has been built into the system. The graphics module enhances the user friendliness of the software, and enables presentation of the results in an easy-to-understand and visually appealing manner. The output of the software is formatted so that it can be directly used for reporting the results without the need of editing.     

CFD analysis of the dispersion of toxic materials in road tunnels

The present work is aimed at analyzing the evolution of accidental scenarios deriving from the release of toxic materials inside a tunnel. This scenario, compared to the more frequently investigated cases of fire, followed by smoke dispersion, may involve a large variety of common products characterized by widely differing physical properties; nonetheless it has been analysed in the literature less than expected. The present study compares the dispersion of two common toxic chemicals (chlorine and ammonia), in order to derive some preliminary information about the influence of the physical properties and the release rate. A reference road tunnel geometry is assumed, while the release occurs from ground level, at the centre of one lane and in the middle of the tunnel. Two study cases involving a road tanker, transporting the product as liquefied gas under pressure, were considered: a catastrophic release, from a 220 mm hole, emptying the tanker in a few tens seconds (case A), and a continuous release, from a much smaller hole (15 mm), lasting 5 min (case B). For the sake of simplicity, the release is assumed to be in gaseous phase; the dispersion of the toxic is simulated for the 5 min period following the start of the release using a CFD (Computational Fluid Dynamics) analysis, according to an RANS (Reynolds-Averaged Navier–Stokes) approach with the standard k–ε turbulence model, assuming no ventilation conditions. Structured curvilinear grids with hexaedric cells, refined according to the local concentration gradient, are used. For case A scenarios, especially for the whole release duration, dispersion is mainly governed by the “plug-flow” effect caused by the large volume of toxic entering the tunnel in a rather short time; then, the role of diffusivity and gravity becomes more important. Chlorine, heavier than air and with lower diffusivity than ammonia, progressively accumulates towards the floor; the dispersion of ammonia, which is lighter than air, appears more influenced by diffusivity than by gravity, since a limited stratification is observed. These trends are more evident for case B scenarios, where the toxic flow rates are much lower. It is expected the results will give some useful insight into the dispersion phenomenon within highly confined spaces and maybe also provide some suggestion about ventilation systems design and emergency procedures.     

A CFD based explosion risk analysis methodology using time varying release rates in dispersion simulations

A full probabilistic Explosion Risk Analysis (ERA) is commonly used to establish overpressure exceedance curves for offshore facilities. This involves modelling a large number of gas dispersion and explosion scenarios. Capturing the time dependant build up and decay of a flammable gas cloud size along with its shape and location are important parameters that can govern the results of an ERA. Dispersion simulations using Computational Fluid Dynamics (CFD) are generally carried out in detailed ERA studies to obtain these pieces of information. However, these dispersion simulations are typically modelled with constant release rates leading to steady state results. The basic assumption used here is that the flammable gas cloud build up rate from these constant release rate dispersion simulations would mimic the actual transient cloud build up rate from a time varying release rate. This assumption does not correctly capture the physical phenomena of transient gas releases and their subsequent dispersion and may lead to very conservative results. This in turn results in potential over design of facilities with implications on time, materials and cost of a project.In the current work, an ERA methodology is proposed that uses time varying release rates as an input in the CFD dispersion simulations to obtain the fully transient flammable gas cloud build-up and decay, while ensuring the total time required to perform the ERA study is also reduced. It was found that the proposed ERA methodology leads to improved accuracy in dispersion results, steeper overpressure exceedance curves and a significant reduction in the Design Accidental Load (DAL) values whilst still maintaining some conservatism and also reducing the total time required to perform an ERA study.     

大气扩散计算流体力学模型综合评估方法研究

模型评估方法研究是模型研发工作的重要组成部分。基于科学性验证与统计性确认提出了大气扩散CFD模型的综合确认评估方法,通过示例的方式展示了方法的使用效果。综合确认评估方法可以提高模型筛选的速度,降低统计性评估对确认性试验数据的需求量,从总体上提高模型确认的效率。评估理论的研究有助于提高基于模型的大气扩散研究的准确性,也有利于高精度模型和试验的设计、开发与遴选。     

Machine learning based quantitative consequence prediction models for toxic dispersion casualty

Incidental release of toxic chemicals can pose extreme danger to life in the vicinity. Therefore, it is crucial for emergency responders, plant operators, and safety professionals to have a fast and accurate prediction to evaluate possible toxic dispersion life-threatening consequences. In this work, a toxic chemical dispersion casualty database that contains 450 leak scenarios of 18 toxic chemicals is constructed to develop a machine learning based quantitative property-consequence relationship (QPCR) model to estimate the affected area caused by toxic chemical release within a certain death rate. The results show that the developed QPCR model can predict the toxic dispersion casualty range with root mean square error of maximum distance, minimum distance, and maximum width less than 0.2, 0.4, and 0.3, which indicates that the constructed model has satisfying accuracy in predicting toxic dispersion ranges under different lethal consequences. The model can be further expanded to accommodate more toxic chemicals and leaking scenarios.     

Methodology for global sensitivity analysis of consequence models

A methodology is presented for global sensitivity analysis of consequence models used in process safety applications. It involves running a consequence model around a hundred times and using the results to construct a statistical emulator, which is essentially a sophisticated curve fit to the data. The emulator is then used to undertake the sensitivity analysis and identify which input parameters (e.g. operating temperature and pressure, wind speed) have a significant effect on the chosen output (e.g. vapour cloud size). Performing the sensitivity analysis using the emulator rather than the consequence model itself leads to significant savings in computing time.To demonstrate the methodology, a global sensitivity analysis is performed on the Phast consequence model for discharge and dispersion. The scenarios studied consist of above-ground, horizontal, steady-state discharges of dense-phase carbon dioxide (CO₂), with orifices ranging in diameter from ½ to 2 inch and the liquid CO₂ stagnation conditions maintained at between 100 and 150 bar. These scenarios are relevant in scale to leaks from large diameter above-ground pipes or vessels.Seven model input parameters are varied: the vessel temperature and pressure, orifice size, wind speed, humidity, ground surface roughness and height of the release. The input parameters that have a dominant effect on the dispersion distance of the CO₂ cloud are identified, both in terms of their direct effect on the dispersion distance and their indirect effect, through interactions with other varying input parameters.The analysis, including the Phast simulations, runs on a standard office laptop computer in less than 30 min. Tests are performed to confirm that a hundred Phast runs are sufficient to produce an emulator with an acceptable degree of accuracy. Increasing the number of Phast runs is shown to have no effect on the conclusions of the sensitivity analysis.The study demonstrates that Bayesian analysis of model sensitivity can be conducted rapidly and easily on consequence models such as Phast. There is the potential for this to become a routine part of consequence modelling.     

A transportation network assessment tool for hazardous material cargo routing: Weighing exposure health risks,proximity to vulnerable areas,delay costs and trucking expenses

The inherent risks associated with accidental releases of hazardous materials during transport have drawn attention and concerns in the recent decades. The aim of this study is to propose a tool for evaluation and comparison of the transportation networks which can be used to assess the routing options between origins and destinations of the cargos for their suitability for transporting hazardous material cargos by tanker trucks and to identify routes which provide lower accidental release risks, lower public exposure risks, and offer economical benefits. Each route segment of transportation networks were evaluated using specific criteria which included health risk and cost of delay in case of an accidental release of materials, trucking cost and proximity to vulnerable areas. Since, the health impact of hazardous materials differ depending on the characteristics of the material being transported as well as release quantities and atmospheric conditions; this paper aimed in providing a tool that can be used to estimate the impact radius (for health risks) after accidental release of hazardous materials by taking into account different atmospheric conditions based on the meteorological data and solar elevation angle. The Gaussian air dispersion model paired with ArcGIS using Python programming were employed to estimate the health risk impact zones by considering the meteorological data, and accordingly to analyze road segments for cost impacts (delay and trucking costs), and the proximity to vulnerable areas. The route assessment tool was demonstrated with a case study. The results of this study can efficiently aid decision makers for transportation of hazardous materials.     

Evaluation of hazard range for the natural gas jet released from a high-pressure pipeline: A computational parametric study

In the present study, the hazard range of the natural gas (NG) jet released from a high-pressure pipeline was investigated. A one-dimensional integral model was combined with a release model to calculate the length and width (i.e., size), and the shape of NG jet release. The physical parameters affecting the jet release of NG were categorized into three types: source release, environmental and time parameters. The effects of each type of parameters on the gas jet release rate, size and shape were evaluated systematically. The results show that all of these parameters have important influence on the hazard range of NG jet release. The source release parameters, including the pipeline length, the operation pressure of the pipeline, the release hole diameter and the pipe diameter, dominate the gas release rate through a hole and therefore the length and width of gas jet release. The gas jet release rate and size are found to be highly correlative with these parameters in terms of power curve regression analysis. The environmental parameters including the atmospheric stability, the ambient wind speed and the source height, have no influence on the gas jet release rate but have influence on the hazard range of gas jet by the turbulent mixing and dilution of NG with air. The time parameters including the concentration averaged time and the valve closing time which are related to the unsteady state jet release of NG, also show the influence on the hazard range of gas jet release. The results show that the decreasing valve closing time and increasing gas concentration averaged time are in favor of reducing the length and width of gas jet release. In addition, these computational parametric studies indicate that the parameters of source release and time have no significant influence on the shape of gas jet release (i.e., jet length/width ratio, LWR) which can maintain the values between 7 and 8. However, the environmental parameters have influence on the shape of gas jet release. These comprehensive investigations provide useful database of evaluating the hazard range for NG jet released from a hole on a high-pressure pipeline and also provide the foundation of decision-making for further fire and/or explosion evaluation and people evacuation.     

HGSYSTEM: a review, critique, and comparison with other models

HGSYSTEM is a package of computer models for modeling release and atmospheric dispersion of hazardous substances, which has capabilities that may not be found together elsewhere (e.g., hydrogen fluoride/air/water thermodynamics, aerosol release, release and dispersion of mixtures, multi-component evaporation, downward vertical jets, plume lift-off, deposition, and street-canyon). There are, however, some shortcomings in its spill models and in the transition from a spill to a dispersion model. The model's strong and weak points and limited comparisons with ALOHA and DEGADIS are discussed herein.     

海底管道泄漏油气扩散规律数值仿真与对比分析

为研究海底原油与天然气单相泄漏扩散规律的差异性,合理制定应急响应策略,减小事故损失,针对海底管道失效所致的原油与天然气泄漏问题,基于计算流体动力学CFD方法,建立海底油气管道泄漏事故后果预测与评估模型,对特定事故场景下的海底原油与天然气泄漏扩散过程进行模拟与分析,从泄漏扩散过程、工况因素影响、泄漏后果及应对策略4个方面对比原油与天然气的泄漏扩散特性。结果表明:相同工况下,海底原油与天然气在泄漏速率、扩散时间、扩散形态及水平最大扩散距离方面存在显著差别;与天然气相比,原油泄漏扩散行为对工况因素具有更高的敏感性;原油泄漏会引发严重的环境灾害,天然气泄漏则会影响海上结构物的稳定性及引发火灾爆炸事故,据此需合理制定具有针对性的应对策略。     

Accident consequence calculation of ammonia dispersion in factory area

With the widespread use of ammonia in the process industry, more and more accidents were caused by ammonia leakage and dispersion. The dispersion of ammonia is determined by its physical properties, release source conditions and atmospheric environment. Full-scale numerical simulation based on CFD theory was carried out to study the dispersion law of ammonia in a food factory. It was found that ammonia concentrated on the symmetric plane and showed an upward movement near the source. Moreover, the effect of pressure on the dispersion of ammonia was explored showing that the concentration of ammonia near the source increased with the increase of pressure, while the dispersion of ammonia far from the source is mainly influenced by wind field. Last but not the least, the dangerous area completely covers the obstacle region according to the harmful concentration, but the lethal concentration range and explosion range both only existed near the release source. Correspondingly, the concentration of ammonia in the region far from the symmetric plane can be regarded as a safe area. When the accident happens, one should stay away from the release source and evacuate towards the sides in a timely manner. We hope that this work can provide an effective method in predicting the impact of ammonia dispersion and can arouse concerns over the public safety.     

Analysis of meteorological parameters for dense gas dispersion using mesoscale models

The current study focuses on characterizing the atmospheric details required for dense gas dispersion analysis resulting from release of cryogenic liquids like LNG. The study investigates the effectiveness of coupling the prognostic MM5 mesoscale model with the CALMET diagnostic model for producing meteorological conditions that is characteristic of dry and arid regions like Qatar, with non-neutral boundary conditions. MM5-CALMET wind fields and temperature data are compared with the meteorological field observations from the Doha International Airport (DIA) on a monthly basis, daily basis and hourly basis to study the effect of different averaging periods. The monthly averages replicate the annual patterns of meteorological parameters very well. However difference in observation and model are observed for wind speed and wind direction variable. The daily averages obtained from the model are in good agreement with the observation for wind speed and temperature. For hourly averaging, the model is found capable of mimicking the temperature of a given location, but not wind speed and direction. The prediction of wind direction parameter using MM5-CALMET is moderate for any averaging period. The sub-optimal performance of wind direction variable is attributed to grid resolution of vertical and horizontal layers of MM5-CALMET model. Additionally a case study is performed to illustrate the effect of variation of meteorological parameters on the lower flammability limits (LFL) resulting from flammable dense gas release of LNG. The case study demonstrates the issues that arise in a risk analysis study when “wrong” meteorological data could be used. The overall study indicates that utilizing the coarsest prognostic meteorological model output in a diagnostic model provides an effective option for generating meteorological inputs for dispersion studies.     

Modeling the two-phase cloud evolution from instantaneous flashing release using CFD

Computational Fluid Dynamics (CFD) approach has been successfully applied to simulate the small-scale instantaneous flashing release experiment by Pettitt. A model for dispersion of the release event is provided based on relevant theories and existing experimental data. An application of the CFD method to the dispersion simulation is illustrated. Furthermore, a new methodology based on discrete phase model for setting computational initial conditions is provided. An initial expansion and subsequent turbulence dispersion can be characteristically identified from both volume and temperature variation of the cloud obtained by the simulation. The possible mechanism for these phenomena has also been discussed and analyzed. The study deepens the understanding of the physical process of this event and provides one more reliable tool for relevant safety systems.     

A multivariable model for estimation of vapor cloud explosion occurrence possibility based on a Fuzzy logic approach for flammable materials

In this paper, a new method based on Fuzzy theory is presented to estimate the occurrence possibility of vapor cloud explosion (VCE) of flammable materials. This new method helps the analyst to overcome some uncertainties associated with estimating VCE possibility with the Event Tree (ET) technique. In this multi-variable model, the physical properties of the released material and the characteristics of the surrounding environment are used as the parameters specifying the occurrence possibility of intermediate events leading to a VCE. Factors such as area classification, degree of congestion of a plant and release rate are notably affecting the output results. Moreover, the proposed method benefits from experts' opinions in the estimation of the VCE possibility. A refrigeration cycle is used as the case study and the probability of VCE occurrence is determined for different scenarios. In this study, sensitivity analysis is performed on the model parameters to assess their effect on the final values of the VCE possibility. Furthermore, the results are compared with the results obtained using other existing models.     

Worst-case identification of gas dispersion for gas detector mapping using dispersion modeling

To quickly and accurately quantify the material release in process units, gas detectors may be placed according to the results of gas dispersion modeling. DNV's PHAST software is one of the most useful and reliable tools for material dispersion modeling. In this software, fluid dispersion is modeled based on the process conditions, the weather conditions and the specifications of the material release point. However, varying weather conditions throughout the year and the exact determination of the release point on the plot plan and the release elevation are problematic; these issues cause the results to be non-exact and non-integrated. Choosing the most appropriate conditions is challenging. In this paper, a scheme was provided to select the most appropriate conditions for gas dispersion modeling. This scheme approaches modeling based on the worst-case scenario (the situation in which the dispersed gas reaches the detector later in comparison to the other cases). Therefore, different weather conditions, release elevations and release points on the plot plan were modeled for an absorber tower of the Gonbadli Dehydration Unit of the Khangiran Refinery. The worst case of each release condition was then chosen. Finally, gas detectors were placed using the gas dispersion modeling results based on the worst-case scenario.     

A preliminary explore to the forced ventilation on the toxic gas release/dispersion and the hazard mitigation within a petrochemical plant

More than thirty-five years ago, the Bhopal disaster shook the whole world and investigators found out that many people survived just because they turned on the fans in their bedrooms. It was postulated that the forced ventilation played an important role in diluting the toxic gas and saved these people. In order to provide evidence to solve this old mystery, this research employed FLACS software to assess the hazardous degree of a toxic gas (hydrogen sulfide) leakage within a petrochemical process. Series of gas dispersion simulations were performed to actualize the hazardous characteristics and the corresponding risks of the release accident. The study shows that the hazardous level and the hazard range can be greatly influenced when parameters, such as the gas leakage circumstances (atmospheric conditions and wind speed) and the mitigation measures (direction of fans and their speed) are altered.By using explosion-proof fans in different positions and ventilation directions, combined with the natural wind in a certain direction, this research attempts to detect the best combination from various mitigation designs and to compare the influence of fan directions on hazard mitigation. It is also the first time of its kind to simulate the effect of forced ventilation on hazard mitigation within a process plant. The results show that the hazardous level of a toxic release can be effectively alleviated, when the direction of the mechanical ventilation is against the natural wind direction. With the help of the CFD simulation and the quantitative risk analysis technique, different loss prevention strategies can be tested via this method in order to establish a safer working environment.     

Estimation of safety distances in the vicinity of fuel gas pipelines

In this paper, safety distances around pipelines transmitting liquefied petroleum gas and pressurized natural gas are determined considering the possible outcomes of an accidental event associated with fuel gas release from pressurized transmission systems. Possible outcomes of an accidental fuel gas release were determined by performing the Event Tree Analysis approach. Safety distances were computed for two pipeline transmission systems of pressurized natural gas and liquefied petroleum gas existing in Greece using real data given by Greek Refineries and the Greek Public Gas Enterprise. The software packages chetah and breeze were used for thermochemical mixture properties estimation and quantitative consequence assessment, respectively. Safety distance determination was performed considering jet fire and gas dispersion to the lower flammable limit as the worst-case scenarios corresponding to immediate and delayed cloud ignition. The results showed that the jet fire scenario should be considered as the limiter for safety distances determination in the vicinity of natural and petroleum gas pipelines. Based on this conclusion, the obtained results were further treated to yield functional diagrams for prompt safety distance estimation. In addition, qualitative conclusions were made regarding the effect of atmospheric conditions on possible events. Thus, wind velocity was found to dominate during a jet fire event suppressing the thermal radiation effect, whereas gas dispersion was found to be affected mainly by solar radiation that favors the faster dissolution of fuel gas below the lower flammable limit.     

Calculating concentration contour co-ordinates using integral plume models

A mathematical model is proposed which allows the prediction of the contour, for a given concentration, of a plume caused by a release into the atmosphere of substances either heavier or lighter than air. The application of the model to different cases has given results close to those obtained from other models. Its introduction into the computer code of any dispersion model provides a tool useful in risk analysis and environmental assessment.