Validation of discharge and atmospheric dispersion for unpressurised and pressurised carbon dioxide releases

This paper discusses the validation of discharge and subsequent atmospheric dispersion for both unpressurised and pressurised carbon dioxide releases using the consequence modelling package Phast.The paper first summarises the validation of the Phast dispersion model (UDM) for unpressurised releases. This includes heavy gas dispersion from either a ground-level line source (McQuaid wind-tunnel experiments) or an area source (Kit-Fox field experiments). For the McQuaid experiments minor modifications of the UDM were made to support line sources. For the Kit Fox experiments steady-state and 20-s finite-duration releases were simulated for both neutral and stable conditions. Most accurate predictions of the concentrations for finite duration releases were obtained using the UDM Finite Duration Correction method.Using experiments funded by BP and Shell and made available via DNV's CO2PIPETRANS JIP, the paper secondly summarises the validation of the Phast discharge and dispersion models for pressurised CO₂ releases. This modelling accounted for the possible presence of the solid CO₂ phase following expansion to atmospheric pressure. These experiments included both high-pressure steady-state and time-varying cold releases (liquid storage) and high-pressure time-varying supercritical hot releases. Both the flow rate and the concentrations were found to be predicted accurately.The above validation was carried out with no fitting whatsoever of the Phast extended discharge and dispersion models.     

Modelling of discharge and atmospheric dispersion for carbon dioxide releases

This paper discusses the modelling of the discharge and subsequent atmospheric dispersion for carbon dioxide releases using extensions of models in the consequence modelling package Phast. Phast examines the progress of a potential incident from the initial release to the far-field dispersion including the modelling of rainout and subsequent vaporisation. The original Phast discharge and dispersion models allow the released chemical to occur only in the vapour and liquid phases. As part of the current work these models have been extended to also allow for the occurrence of liquid to solid transition or vapour to solid transition. This applies both for the post-expansion state in the discharge model, as well as for the thermodynamic calculations by the dispersion model. Solid property calculations have been added where necessary. The above extensions are generally valid for fluid releases including CO₂. Using the extended dispersion formulation, a sensitivity study has been carried out for mixing of solid CO₂ with air, and it is demonstrated that solid effects may significantly affect the predicted concentrations.     

Modelling of time-varying dispersion for elevated pressurised releases without rainout

Many commonly used atmospheric dispersion models are limited to continuous or instantaneous releases only, and cannot accurately simulate time-varying releases. The current paper discusses a new enhanced dispersion formulation accounting for time-varying effects resulting from a pressure drop in a vessel or pipe, and presuming no rainout. This new formulation is implemented in the Unified Dispersion Model (UDM), and is planned to be included in a future version of Phast.First existing methods are summarised for modelling finite-duration and time-varying releases, and limitations of these methods are identified.Secondly the new mathematical model is summarised. The new formulation presumes a number of ‘observers’ to be released at successive times from the point of discharge. The UDM carries out pseudo steady-state calculations for each observer, where the release data correspond to the time at which the observer is released. Subsequently the model applies a correction to the observer concentrations to ensure mass conservation when observers move with different velocities. Finally effects of along-wind diffusion (due to ambient turbulence) are included by means of Gaussian integration over the downwind distance. This results in reduced concentrations while the cloud travels in the downwind direction.The benefits of the new UDM methodology are illustrated for the case of a H₂S toxic release from a long pipeline representative of some extremely sour fields in the Middle East that are now being developed. Using corrected observer concentrations and along-wind diffusion significantly reduces toxic effect distances when compared to the current Phast 7.1 approach.     

Evaluation of multi-phase atmospheric dispersion models for application to Carbon Capture and Storage

A dispersion model validation study is presented for atmospheric releases of dense-phase carbon dioxide (CO₂). Predictions from an integral model and two different Computational Fluid Dynamics (CFD) models are compared to data from field-scale experiments conducted by INERIS, as part of the EU-funded CO₂PipeHaz project.The experiments studied consist of a 2 m³ vessel fitted with a short pipe, from which CO₂ was discharged into the atmosphere through either a 6 mm or 25 mm diameter orifice. Comparisons are made to measured temperatures and concentrations in the multi-phase CO₂ jets.The integral dispersion model tested is DNV Phast and the two CFD models are ANSYS-CFX and a research and development version of FLACS, both of which adopt a Lagrangian particle-tracking approach to simulate the sublimating solid CO₂ particles in the jet. Source conditions for the CFD models are taken from a sophisticated near-field CFD model developed by the University of Leeds that simulates the multi-phase, compressible flow in the expansion region of the CO₂ jet, close to the orifice.Overall, the predicted concentrations from the various models are found to be in reasonable agreement with the measurements, but generally in poorer agreement than has been reported previously for similar dispersion models in other dense-phase CO₂ release experiments. The ANSYS-CFX model is shown to be sensitive to the way in which the source conditions are prescribed, while FLACS shows some sensitivity to the solid CO₂ particle size. Difficulties in interpreting the results from one of the tests, which featured some time-varying phenomena, are also discussed.The study provides useful insight into the coupling of near- and far-field dispersion models, and the strengths and weaknesses of different modelling approaches. These findings contribute to the assessment of potential hazards presented by Carbon Capture and Storage (CCS) infrastructure.     

Risk assessment methodology for high-pressure CO2 pipelines incorporating topography

This paper presents a risk assessment methodology for high-pressure CO₂ pipelines developed at the Health and Safety Laboratory as part of the EU FP7 project CO2Pipehaz.Traditionally, consequence modelling of dense gas releases from pipelines at major hazard impact levels is performed using integral models with limited or no consideration being given to weather bias or topographical features of the surrounding terrain. Whilst dispersion modelling of CO₂ releases from pipelines using three-dimensional CFD models may provide higher levels of confidence in the predicted behaviour of the cloud, the use of such models is resource-intensive and usually impracticable. An alternative is to use more computationally efficient shallow layer or Lagrangian dispersion models that are able to account for the effects of topography whilst generating results within a reasonably short time frame.In the present work, the proposed risk assessment methodology for CO₂ pipelines is demonstrated using a shallow-layer dispersion model to generate contours from a sequence of release points along the pipeline. The simulations use realistic terrain taken from UK topographical data. Individual and societal risk levels in the vicinity of the pipeline are calculated using the Health and Safety Laboratory's risk assessment tool QuickRisk.Currently, the source term for a CO₂ release is not well understood because of its complex thermodynamic properties and its tendency to form solid particles under specific pressure and temperature conditions. This is a key knowledge gap and any subsequent dispersion modelling, particularly when including topography, may be affected by the accuracy of the source term.     

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.     

Sub-cooled and flashing liquid jets and droplet dispersion I. Overview and model implementation/validation

Many accidents involve two-phase releases of hazardous chemicals into the atmosphere. This paper describes the results of a third phase of a Joint Industry Project (JIP) on liquid jets and two-phase droplet dispersion. The aim of the project is to increase the understanding of the behaviour of sub-cooled non-flashing and superheated flashing liquid jets, and to improve the prediction of initial droplet size, droplet dispersion and rainout.Phase III of the JIP first included scaled experiments for materials with a range of volatilities (water, cyclohexane, butane, propane and gasoline). These experiments were carried out by Cardiff University including measurements of flow rate and initial droplet size across the full relevant range of superheats. See the companion paper II for further details of these experiments and the derivation of new refined correlations for droplet size distribution and Sauter Mean Diameter. Furthermore large-scale butane experiments were carried out by INERIS (France) to ensure that for more realistic scenarios the derived droplet size correlations are accurate.Model validation and model improvements were carried out by DNV Software, including validation of release rate and initial droplet size against the above scaled and large-scale experiments. New correlations for droplet size distribution and Sauter Mean Diameter (SMD) were implemented into the Phast discharge model. These were compared against a range of other droplet size and rainout correlations published in the literature, in conjunction with validation against an extensive set of experiments. It was shown that the new droplet size correlation agrees better against experimental data than the existing Phast correlation. To further improve the rainout prediction, the Phast dispersion model (UDM) was also extended to allow simultaneous modelling of a range of droplet sizes and distributed rainout (rather than rainout at one point).     

A risk assessment methodology for high pressure CO2 pipelines using integral consequence modelling

This paper presents a risk assessment methodology for high pressure CO₂ pipelines developed at the Health and Safety Laboratory (HSL) as part of the EU FP7 CO2Pipehaz project. Until recently, risk assessment of dense phase and supercritical CO₂ pipelines has been problematic because of the lack of suitable source term and integral consequence models that handle the complex behaviour of CO₂ appropriately. The risk assessment presented uses Phast, a commercially available source term and dispersion model that has been recently updated to handle the effects of solid CO₂. A test case pipeline was input to Phast and dispersion footprints to different levels of harm (dangerous toxic load and probit values) were obtained for a set of pipeline specific scenarios. HSL's risk assessment tool QuickRisk was then used to calculate the individual and societal risk surrounding the pipeline. Knowledge gaps that were encountered such as: harm criteria, failure rates and release scenarios were identified and are discussed.     

Validation of CFD-model for hydrogen dispersion

To be able to perform proper consequence modelling as a part of a risk assessment, it is essential to be able to model the physical processes well. Simplified tools for dispersion and explosion predictions are generally not very useful. CFD tools have the potential to model the relevant physics and predict well, but without proper user guidelines based on extensive validation work, very mixed prediction capability can be expected. In this article, recent dispersion validation effort for the CFD tool FLACS–HYDROGEN is presented. A range of different experiments is simulated, including low-momentum releases in a garage, subsonic jets in a garage with stratification effects and subsequent slow diffusion, low momentum and subsonic horizontal jets influenced by buoyancy, and free jets from high-pressure vessels. LH₂ releases are also considered. Some of the simulations are performed as blind predictions.     

Modelling three-phase releases of carbon dioxide from high-pressure pipelines

This paper describes the development and experimental validation of a three-phase flow model for predicting the transient outflow following the failure of pressurised CO₂ pipelines and vessels. The choked flow parameters at the rupture plane, spanning the dense-phase and saturated conditions to below the triple point, are modelled by maximisation of the mass flowrate with respect to pressure and solids mass fraction at the triple point. The pertinent solid/vapour/liquid phase equilibrium data are predicted using an extended Peng–Robinson equation of state.The proposed outflow model is successfully validated against experimental data obtained from high-pressure CO₂ releases performed as part of the FP7 CO2PipeHaz project (www.co2pipehaz.eu).The formation of solid phase CO₂ at the triple point is marked by a stabilisation in pressure as confirmed by both theory and experimental observation. For a fixed diameter hypothetical pipeline at 100 bar and 20 °C, the flow model is used to determine the impact of the pipeline length on the time taken to commence solid CO₂ discharge following its rupture.     

Two-phase jet releases,droplet dispersion and rainout I. Overview and model validation

Many accidents involve two-phase releases of hazardous chemicals into the atmosphere. This paper describes the results of the fourth phase of a Joint Industry Project (JIP) on liquid jets and two-phase droplet dispersion. The objective of Phase IV of the JIP was to generate experimental rainout data for non-flashing experiments, and to develop recommendations for the best methodology to predict rainout [total rainout mass and its spatial distribution (‘distributed’ rainout)].Phase IV of the JIP first included rainout experiments by the UK Health and Safety Laboratory (HSL) for sub-cooled releases of water and xylene with a range of orifice sizes and stagnation pressures. See the companion paper II for further details. Secondly model validation was carried out by DNV Software for these experiments using different correlations for the initial droplet size (Sauter Mean Diameter, SMD), i.e. the CCPS SMD correlation and the Phase III JIP SMD correlation. The validation includes flow rates, droplet size, distributed rainout and cloud temperature drop. Subsequently validation was considered for a wider range of experiments from the literature (sub-cooled and superheated releases) for both SMD and total rainout. Adopted rainout methods comprised both methods including explicit modelling of the droplets (using an extended version of Phast dispersion model UDM), as well as more simple methods based on rainout correlations without droplet modelling. Recommendations are made for the most accurate droplet size and rainout modelling. A modified CCPS UDM droplet size correlation has been shown to agree best against experimental rainout data.     

Simulation study on near-field structure and flow characteristics of high-pressure CO2 released from the pipeline

The accidental release of high-pressure carbon dioxide (CO₂) can cause serious damages to both humans and pipeline equipment. Therefore, it is of great significance to have a deeper understanding about the release characteristics of high-pressure CO₂ for improving the safety level of Carbon Capture and Storage (CCS) technologies. Both industrial-scale and laboratory-scale studies have been carried out to predict the release behaviors. In recent years, computational fluid dynamics (CFD) simulation has become a crucial method to study the instantaneous changes and microscopic details of the fluid behaviors. In this paper, the simulation method was employed to study the near-field structure and flow characteristics of high-pressure CO₂ released from pipelines. The Peng-Robinson Equation of State (EOS) was used to compute the thermodynamic properties of high-pressure CO₂, and SST k-ω model was applied to simulate the structure and physical parameters of the under-expanded jet. In addition, the multi-phase mixture model was introduced to study the phase transition. The non-equilibrium liquid/vapor transition is modeled by introducing ‘source terms’ for mass transfer and latent heat. Compared to the experimental results, the simulation results showed good agreement. Furthermore, the influences of operating conditions, including different stagnation pressure, stagnation temperature, and nozzle size, were analyzed.     

A protocol for the evaluation of LNG vapour dispersion models

The international transport, storage and utilisation of LNG is growing rapidly. Whilst the LNG industry has an excellent safety record, the possibility of an accidental release cannot be discounted. Internationally-accepted standards, such as the 59A Standard of the US National Fire Protection Association (NFPA), provide direction on the assessment of LNG spill hazards and hazard range criteria which must be met. Modelling of the atmospheric dispersion of LNG vapour from accidental spills is one of the critical steps in such hazard analyses. This paper describes a comprehensive evaluation protocol devised for the 59A Standard, specifically for the assessment of LNG vapour dispersion models. The evaluation protocol is based on methodologies developed in previous European Union studies, which have been extended, significantly adapted and tailored to the specific requirements of the evaluation of models for the dispersion of LNG vapour. The protocol comprises scientific evaluation of the numerical and physical basis of models for the dispersion of LNG vapour, model verification, and validation; resulting in a comprehensive model evaluation report which includes qualitative and quantitative criteria for model acceptance. A supporting suite of validation data, and guidance on the use of this data, has also been produced. The NFPA 59A (2009) standard states that LNG vapour dispersion models are acceptable for use if they have been evaluated in accordance with this protocol.     

Are dispersion models suitable for simulating small gaseous chlorine releases?

Dispersion models are mostly validated on the basis of historical dispersion experiments. The latter imply large quantities of hazardous products (flammable or toxic gases), and are dedicated to study the dispersion of the resulting clouds on great distances from the source to reach a better knowledge of the different phases of gas dispersion (slumping, creeping, passive dispersion…).However, dispersion models have hardly been validated on small releases and therefore require more validation on small plumes of dangerous gases. Indeed, what is their reliability in case of accidents involving small amounts (e.g., chlorine leakages at swimming pools’ installations), and for small distances downwind the gas source? This information is of prime interest in so far as small releases are more likely to occur than larger ones.This paper reports on chlorine small-scale dispersion experiments and deals with the comparison between experimental data of ground level concentrations in the plume and predicted concentrations obtained from several dispersion models.     

Modelling accidental releases of dangerous gases into the lower troposphere from mobile sources

The article reports the results of different methods of modelling releases and dispersion of dangerous gases or vapours in cases of major accidents from road and rail transportation in urban zones. Transport accidents of dangerous substances are increasingly frequent and can cause serious injuries in densely inhabited areas or pollution of the environment. For quantitative risk assessment and mitigation planning, consequence modelling is necessary.

The modelling of dangerous substance dispersion by standard methods does not fully represent the behaviour of toxic or flammable clouds in obstructed areas such as street canyons. Therefore the predictions from common software packages as ALOHA, EFFECTS, TerEx should be augmented with computational fluid dynamics (CFD) models or physical modelling in aerodynamic tunnels, and further studies are planned to do this.

The goal of this article is to present the results of the first approach of modelling using these standard methods and to demonstrate the importance of the next development stage in the area of transport accident modelling of releases and dispersions of dangerous substances in urban zones in cases of major accident or terrorist attacks.     

Validation of FLACS against experimental data sets from the model evaluation database for LNG vapor dispersion

The siting of facilities handling liquefied natural gas (LNG), whether for liquefaction, storage or regasification purposes, requires the hazards from potential releases to be evaluated. One of the consequences of an LNG release is the creation of a flammable vapor cloud, that may be pushed beyond the facility boundaries by the wind and thus present a hazard to the public. Therefore, numerical models are required to determine the footprint that may be covered by a flammable vapor cloud as a result of an LNG release. Several new models have been used in recent years for this type of simulations. This prompted the development of the “Model evaluation protocol for LNG vapor dispersion models” (MEP): a procedure aimed at evaluating quantitatively the ability of a model to accurately predict the dispersion of an LNG vapor cloud.This paper summarizes the MEP requirements and presents the results obtained from the application of the MEP to a computational fluid dynamics (CFD) model – FLACS. The entire set of 33 experiments included in the model validation database were simulated using FLACS. The simulation results are reported and compared with the experimental data. A set of statistical performance measures are calculated based on the FLACS simulation results and compared with the acceptability criteria established in the MEP. The results of the evaluation demonstrate that FLACS can be considered a suitable model to accurately simulate the dispersion of vapor from an LNG release.     

Self-ignition and explosion during discharge of high-pressure hydrogen

The phenomenon of self-ignition and explosion during discharge of high-pressure hydrogen was investigated. To clarify the ignition conditions of high-pressure hydrogen jets, rapid discharge of the high-pressure hydrogen was examined experimentally. A diaphragm was used to allow rapid discharge of the high-pressure hydrogen. The burst pressure was varied from 4 to 30 MPa. The downstream geometry of the diaphragm was a flange and extension pipes, with the pipe length varying from 3 to 300 mm. The diameter of the nozzle was 5 or 10 mm. When short pipes were used, the hydrogen jet did not ignite. However, the hydrogen jet showed an increasing tendency to ignite in the pipe as the length of the pipe became longer. At higher burst pressures, a diffusion jet flame was formed from the pipe. The blast wave from the fireball formed on self-ignition of the hydrogen jet resulted in an extremely rapid pressure rise.     

An innovative and comprehensive approach for the consequence analysis of liquid hydrogen vessel explosions

Hydrogen is one of the most suitable solutions to replace hydrocarbons in the future. Hydrogen consumption is expected to grow in the next years. Hydrogen liquefaction is one of the processes that allows for increase of hydrogen density and it is suggested when a large amount of substance must be stored or transported. Despite being a clean fuel, its chemical and physical properties often arise concerns about the safety of the hydrogen technologies. A potentially critical scenario for the liquid hydrogen (LH₂) tanks is the catastrophic rupture causing a consequent boiling liquid expanding vapour explosion (BLEVE), with consequent overpressure, fragments projection and eventually a fireball. In this work, all the BLEVE consequence typologies are evaluated through theoretical and analytical models. These models are validated with the experimental results provided by the BMW care manufacturer safety tests conducted during the 1990's. After the validation, the most suitable methods are selected to perform a blind prediction study of the forthcoming LH₂ BLEVE experiments of the Safe Hydrogen fuel handling and Use for Efficient Implementation (SH₂IFT) project. The models drawbacks together with the uncertainties and the knowledge gap in LH₂ physical explosions are highlighted. Finally, future works on the modelling activity of the LH₂ BLEVE are suggested.