While useful, screening-level analyses often don’t provide the high-fidelity modeling required for an in-depth examination of complex hazard scenarios. Analyzing a problem with Computational Fluid Dynamics (CFD) can better predict the consequences of postulated scenarios. In comparison to a screening level analysis, CFD analysis results typically lead to more effective and more cost-efficient mitigation solutions. The experts at BakerRisk leverage their comprehensive experience in research, development, and the application of CFD tools to better simulate event scenarios and provide high-quality solutions.
BWTI©
BakerRisk developed BWTI© (Blast Wave Target Interaction) to simulate the generation and propagation of blast and shock waves and their interaction with structures. BWTI© has been extensively used for designing test enclosures to protect against high-pressure pneumatic failures and high explosives for industrial and government clients. It excels in replicating and modeling blast waves resulting from VCEs, bursting pressure vessels, and high-explosive detonations. Its primary application lies in engineering-level analyses, particularly within scenarios where the nuanced aspects of blast wave shielding and focusing come into play.
Flacs
The FLACS (Flame Acceleration Simulator) code is an industry-standard CFD tool that our team at BakerRisk utilizes in many projects to model flammable gas and evaporating liquid releases and dispersions, toxic dispersion distances, blast loads due to vapor cloud explosions (VCE), and exhaust heat plumes for both on-shore and off-shore facilities. FLACS is commonly used to analyze typical refining and processing facilities, chemical facilities, and offshore production. In recent years, it has been extended to model hazards associated with all phases of renewable hydrogen (production, storage, transportation, and consumption).
Unique applications of this software include evaluating the efficacy of mitigation strategies, such as vented deflagrations, gas detector layout and response, dispersion barriers, and DDT prevention through modifying congestion configuration.
Fds
FDS (Fire Dynamics Simulator) was developed by the National Institute of Standards and Technology within the U.S. Department of Commerce. It’s a sophisticated CFD model designed to simulate heat and smoke transport from fires with a versatile toolset that allows for the computation of various other non-thermal-related low-speed fluid flow simulations.
For example, FDS has been used to effectively model fire spread through buildings, liquid pool fires, mist dispersion, and buoyant mixing calculations to accurately determine hazards at industrial and private facilities. In combination with the Pathfinder program, it is able to determine personnel evacuation efficiency during hazardous fire events. A primary component of the FDS tool is Smokeview, a companion program capable of producing high-quality images and animations, making it an indispensable tool for assessing hazards.
Validation
BakerRisk has tested and validated many of the above computer models and technologies within our very own test site. As a result, you can trust our fluid dynamics experts because we don’t just perform modeling: we ensure that computational models work properly before leveraging them in real-world scenarios. We’ve also published major studies and papers relating to this field.
Our Specialists
Related Resources
The downhole tool industry commonly conducts tests that require pressurization of a vessel. The most common failure mode is through the launching of end caps, plugs, and fittings as opposed to a catastrophic rupture of the vessel body. This type of vessel failure during high-pressure gas testing can produce significant threats to nearby personnel in the form of high-energy projectiles and blast waves that must be blocked or dissipated before reaching personnel. Adequately designing a structure to contain this energy depends on how well a worst case scenario event can be modeled.
Although jet fires may pose a major hazard in the chemical processing industry, they are typically modeled with semi-empirical models that were developed based on experimental results of natural gas. In most semi-empirical models, it is typically assumed that all liquid is instantly vaporized upon discharge and combusted fully in the jet. In reality, depending on discharge fuel momentum, elevation, temperature, and liquid fuel properties, a significant percentage of liquid fuel may fall to the ground and burn as a pool rather than contribute to the jet fire. As a result, such a model may predict overly conservative results when used to model liquid spray jet fires.
BakerRisk will present eight topics at the AIChE 2020 Spring Meeting and GCPS conference.
Fixed water spray systems may be designed to mitigate flammable vapors. They work primarily by entraining air into the spray cone and mixing the air and water droplets with the flammable vapor, thereby diluting the vapor. The current industry standards - including NFPA 15: Standard for water spray fixed systems for fire protection, while providing information on vapor mitigation using fixed water spray (section 7.5) - do not provide the design criteria on how much water and pressure is required to effectively mitigate a specified minor vapor leak. API 2030 does not address the use of fixed water spray for vapor mitigation. In addition, in the event of an ignition, the amount of water and pressure required to control the fire is not defined. This has resulted in the inconsistent application of NFPA 15 for vapor mitigation by engineers and consultants. The goal of this research project is to evaluate the effectiveness of fixed water sprays to suppress the development of flammable vapor clouds and control the fire in case of ignition.
A pressure vessel burst (PVB) is an explosion scenario commonly encountered at chemical processing and petroleum refining facilities. Existing methodologies are available to predict the blast loads resulting from a spherical or cylindrical PVB source, with the PVB source either at grade or at an elevation. In the case of an elevated PVB source, the resulting blast wave will reflect from the ground at an angle. This ground level reflection will result in the formation of a Mach stem at certain angles between the incident blast wave and ground, with the required angles dependent on the blast wave overpressure. The triple point associated with the Mach stem moves upwards as the Mach stem progresses forwards, which can create a region of high blast pressure. This paper focuses on the investigation of a methodology that can be used to determine the high-pressure region generated by the Mach stem, along with the associated blast pressure, as a function of the PVB source elevation and incident blast pressure.
Shock wave attenuation in a straight tunnel (or pipe) can be evaluated using existing methodologies. Shock attenuation is enhanced when there are right-angle turns along the length of the tunnel over which the shock is transmitted. A repeated set of such turns is generally defined as a blast trap. Little guidance is available in the open literature regarding the blast attenuation enhancement due to a right-angle turn or a blast trap in a tunnel. This paper presents guidance for shock wave attenuation as a function of the number of right-angle turns and blast wave parameters (i.e., peak pressure and duration). Characteristic parameters are utilized in order to define shock wave properties and tunnel dimensions. The shock attenuation due to up to four consecutive right-angle turns is evaluated. The purpose of this work is to provide a database of the shock attenuation within a tunnel due to multiple right-angle turns for use in designing tunnel structural components and evaluating the response of such components to postulated transmitted shock loads.
Drag loads due to the gas flows generated by vapor cloud explosions (VCEs) can damage piping and vessels, particularly near or in the flammable cloud. There is little guidance available to the piping and vessel designer regarding the magnitude of such drag loads, beyond evaluating drag loads on a case-by-case basis [e.g., using computational fluid dynamics (CFD)]. This paper presents a newly developed set of drag loads vs standoff distance for a range of flame speeds and vapor cloud sizes. Nine flame speeds were evaluated, up to and including supersonic flame speeds. Cloud sizes are characterized using energy scaled standoff distances (i.e., Sachs scaling). The purpose of this work is to provide a database of drag loads that can be used to evaluate the structural response of piping, vessels, equipment, and supporting structures. The database presentation is similar to that used for existing VCE blast load charts, such as those used in the Baker-Strehlow-Tang (BST) and TNO MEM (Multi-Energy Model) methods.
This paper presents a comparison between the results from a test program carried out to characterize the blast load environment within BakerRisk’s Deflagration Load Generator (DLG) test rig, and predictions made using the FLACS computational fluid dynamics (CFD) code. The test data was also compared to internal peak pressure predictions made using the National Fire Protection Association’s Standard on Explosion Protection by Deflagration Venting (NFPA 68) [1]. The purpose of these tests was to provide data for comparison with standard methods used to predict internal blast loads in a vented deflagration. The tests also provided a characterization of the internal DLG blast load environment for equipment qualification testing.
Blast walls are frequently considered as a potential mitigation option to reduce the applied blast loading on a building or structure in cases where unacceptably high levels of blast damage are predicted. There are three general explosion types of interest with respect to blast loading: High Explosive (HE), Pressure Vessel Burst (PVB), and Vapor Cloud Explosion (VCE). The blast waves resulting from these explosion types can differ significantly in terms of blast wave shape and duration. The effectiveness of a blast wall depends on these blast wave parameters (shape and duration), as well as the blast wall parameters (e.g., height, width and standoff distance from the protected structure). The effectiveness of a blast wall in terms of mitigating the blast loading on a protected structure depends on the combination of the blast wave and blast wall parameters. However, little guidance is available on the effectiveness of blast walls as a mitigation option for non-HE explosion sources. The purpose of this paper is to characterize the effect of blast wave parameters on the effectiveness of a blast wall and to provide guidance on how to determine whether a blast wall is an effective and practical blast damage mitigation option for a given blast loading.
High-pressure pipeline ruptures are a credible explosion hazard at many industrial facilities. The blast field generated by a pipe rupture is highly directional. However, there have been few evaluations of the directional blast loads produced by pipe ruptures. This paper addresses the blast loads generated by a typical “fish mouth” type pipe rupture. The effects of five key parameters on the resulting directional blast field were examined: rupture opening speed, final rupture opening area, pipe diameter, initial gas pressure, and initial gas temperature. The resultant blast loads were compared to those based on existing blast curves for Pressure Vessel Bursts (PVB), the most common of which is based on an assumption of a spherical vessel geometry and instantaneous failure of the entire pressure vessel boundary. The effective gas volume (i.e., number of pipe diameters) required to achieve reasonable agreement between the blast load based on existing PVB blast curves and that resulting from a high-pressure pipeline fish mouth rupture for a specified direction was determined.
Elongated congested volumes are common at chemical processing and petroleum refining facilities due to the arrangement of processing units, but there have been relatively few evaluations reported for the blast loads produced by elongated vapor cloud explosions (VCEs). The accidental VCE that occurred at the Buncefield, UK facility in 2005 involved an elongated congested volume formed by the trees and undergrowth along a portion of the site boundary. Some of near-field damage indicators present at the Buncefield site could not be reasonably explained using existing standard VCE blast load prediction techniques that are based on an assumption that the congested volume filled with flammable gas cloud is hemispherical and located at grade level.
Elongated congestion patterns are common at chemical processing and petroleum refining facilities due to the arrangement of processing units. The accidental vapor cloud explosion (VCE) which occurred at the Buncefield, UK facility involved an elongated congested volume formed by the trees and undergrowth along the site boundary. Although elongated congested volumes are common, there have been few evaluations reported for the blast loads produced by elongated VCEs. Standard VCE blast load prediction techniques do not directly consider the impact of this congested volume geometry versus a more compact geometry.
Enclosed chemical processes, laboratory facilities, boilers and reformer furnaces are typical examples of facilities and equipment where an internal VCE may be postulated to occur. The applied blast load history on the enclosure surfaces is required in order to assess the response of the enclosure to the postulated VCE. The internal blast pressure history and associated applied blast loads depend on a number of factors, such as: (1) the maximum flame speed attained in a flammable cloud, (2) the ratio of the cloud volume to the total enclosure volume, (3) the cloud location within the enclosure, and (4) the environment temperature. The purpose of work described in this paper was to investigate the dependence of internal blast pressure history and applied blast loads on the aforementioned factors. It was found that the applied loads on the enclosure surfaces can be roughly classified into two regimes: quasi-static and dynamic, depending on the combination of these factors. It can be important to identify the appropriate blast loading regime in order to properly analyze the structural response of the enclosure.
The purpose of this paper is to characterize elongated vapour cloud explosions (VCEs) and to identify the resultant difference in blast wave shape relative to those predicted by well-known VCE blast prediction methodologies (e.g., BST, TNO, etc.), the blast curves for which are normally based on an assumption that the flammable gas cloud is hemispherical and located at grade level. BakerRisk’s Blast Wave Target Interaction (BWTI™) computational fluid dynamics (CFD) code was used to investigate the VCE blast loads resulting from elongated flammable gas clouds. The BWTI™ predictions were first validated against the recent Buncefield JIP test data. A blast wave diagram was developed to illustrate typical blast waves generated during the initial flame acceleration process and the subsequent flame propagation. It was found that the aspect ratio of the elongated cloud/congestion volume (i.e., ratio of length to width to height) has a significant impact on the resultant blast wave shape. The blast wave shape in elongated VCEs is was found to be controlled by the negative phase, which is generated by the primary blast wave during the initial acceleration phase in which an effective spherical (or cylindrical) flame front still exists. This primary negative phase plays a key role in the follow-on flame acceleration process, which either attains a steady flame speed or experiences a deflagration-to-detonation transition (DDT).
The purpose of the research was to improve prediction of response of buildings to blast waves by including the negative phase and considering clearing of both positive and negative phases. Commonly used structural design practices, which trace their origins to military design manuals, often ignore the negative phase as well as positive phase clearing. For high explosive threats, this approach is conservative in most circumstances. However, negative phase clearing had not previously been studied for blast waves, and the implications for structural response had not been evaluated. This paper presents results of modeling negative phase blast clearing behavior for a typical blast wave and discusses the differences from positive phase clearing. The implications of including positive and negative phase clearing in building blast damage analysis are also investigated through single-degree-of-freedom (SDOF) analyses.
Pressure vessel burst (PVB) is an explosion scenario commonly encountered at chemical processing facilities. PVBs pose both blast and fragmentation hazards. Blast prediction methods specific to PVBs were first developed in the 1970s and revised blast curves were published in 1995. The published blast curves were developed for spherical vessel bursts. However, most pressure vessels are cylindrical rather than spherical. The blast wave originating from a cylindrical PVB is not spherical (i.e., as with a spherical vessel). Rather, the blast to the sides of a cylindrical vessel is stronger than on the ends, creating nonspherical pressure contours, particularly near the vessel. The cylindrical vessel directional blast effect has recently been investigated by the authors, resulting in a correlation to account for the directional effects. However, it was assumed in the prior work that the vessel was at ground level. This paper extends the prior work to elevated PVBs. Both elevated spherical and cylindrical PVBs are examined to provide new correlations for blast overpressure and impulse for a range of vessel geometries and burst conditions.
A key component of explosion hazard evaluations is the determination of standoffs to given blast overpressure values. Many such evaluations use a simplified methodology which assumes that the blast wave propagates from the explosion source to the target location without interacting with intervening buildings or structures (i.e., without blast wave shielding). This is obviously a perfectly acceptable approach for a screening study, but blast wave shielding effects can be significant in certain circumstances (e.g., within a building group). A methodology was proposed by the UK Health & Safety Laboratory (HSL) in 2001 to account for the blast shielding due to buildings/structures between the explosion source and target location. The HSL methodology is based on the blast waves generated by high explosives (HE). This paper extends the blast shielding evaluation to blast waves generated from pressure vessel bursts (PVB) and vapor cloud explosions (VCE). The influences of blast wave shape parameters (overpressure, duration and rise time) on blast wave shielding are examined. The results indicate that the degree of blast shielding is strongly dependent on the source of the blast wave (i.e., on the blast wave shape parameters) and that the shielding factors obtained with HE blast waves are not always directly applicable for PVB and VCE blast waves.
The impact of thermal hazards on process buildings is an important component of site hazard evaluations. API RP-752 recommends that process facilities analyze thermal hazards and assess their impact to onsite buildings and their occupants. Thermal loads resulting from fires in process units and equipment can have a significant impact on buildings, especially if the building is close to the fire source. Some buildings may be designed for blast and toxic protection, which allows the buildings to be located near process units and equipment, but possibly exposed to thermal hazards from a potential fire.
Computational fluid dynamic (CFD) analysis of flammable gas dispersion can provide a realistic evaluation of the resulting flammable gas cloud. CFD dispersion analysis can provide a more realistic assessment of the resulting gas cloud in heavily congested and/or confined environments relative to simplified dispersion analysis methods. In contrast to simplified methods, the gas cloud predicted using CFD analysis in congested/confined environments is typically complex, both in terms of shape and concentration gradients. The blast loads predicted using CFD vapor cloud explosion (VCE) analysis for the resulting flammable gas cloud are dependent upon the representation chosen for the gas cloud. Directly utilizing the predicted dispersed gas cloud, with the corresponding complexities in shape and concentration gradients, requires an extended series of analyses, since both ignition location within the cloud and the cloud location can have a strong influence on the predicted blast load. A common approach is to replace the dispersed gas cloud with a simplified representation for the purposes of performing a CFD VCE blast load assessment, such as a cubic shaped cloud with a uniform near-stoichiometric fuel concentration. This approach is very attractive since it greatly simplifies the blast load assessment. However, the relationship between the blast loads predicted using a simplified cloud representation versus those predicted for the actual dispersed cloud is not well defined.
Jet fire is a major hazard in the chemical processing industry. During a fire incident, one or more obstruction(s) may be present between the flames and the target of interest, which can affect the flame shape and position, and partially block the thermal radiation. As a result, the thermal radiation intensity and resulting level of damage experienced by the target of interest may be significantly affected by the obstruction(s). Thus, Computational Fluid Dynamic (CFD) modeling of the fire is required, in order to accurately capture these effects. Like any other modeling, CFD modeling requires validation against experimental data before it can be used as a defensible, predictive tool for hazards/risk assessment and fire protection design.
Obtaining pressure and impulse data from explosions where fragments are present can be challenging. One way of protecting airblast gauges from damage due to fragment debris is to place a small but physically robust obstruction directly between the charge and the target. Typically, a vertical metal pole (steel pipe, usually) is positioned directly in-line between the charge and the gauge. The presence of this pipe, however, reduces the amount of blast energy that reaches the target and affects the measured pressure record. A widely implemented rule of thumb suggests that placing the pole at a standoff of 10 pole diameters from the gauge causes a negligible reduction in the airblast impulse that reaches the target.
The ignition of a flammable gas mixture contained within a piping system can lead to damage or failure of the piping or system components. Flame propagation and acceleration within piping systems have been extensively studied. It has been well documented that, given sufficient flame propagation distance and/or the presence of turbulence generating features, flame acceleration within a pipe can lead to a deflagration-to-detonation transition (DDT). The high overpressures associated with a DDT can increase the potential for deformation or failure of the piping system relative to the loads associated with either a fast deflagration or steady-state detonation. This paper presents the results of numerical evaluations to predict the pressure distributions within a pipe run due to a DDT. The blast overpressure associated with a DDT was found to depend on a number of parameters, including: the rate of flame acceleration prior to the DDT, the length of piping occupied by the flammable mixture, the initial gas pressure and the flammable mixture concentration distribution along the pipe. This paper also provides a comparison of the blast loads associated with a steady-state detonation relative to those due to a DDT.
Pressure vessel burst (PVB) is a class of explosion for which there are hazards at virtually all chemical processing facilities. PVBs present both airblast and fragmentation hazards. Blast prediction methods specific to PVBs were first developed in the 1970s and revised blast curves were published in 1995. The published blast curves were developed for spherical vessel bursts, whereas most pressure vessels in use in industry are cylindrical. Blast effects around a bursting cylindrical vessel are not uniform as with a spherical vessel. The blast to the side of a cylindrical vessel is stronger than off the ends, creating non-circular pressure contours. The directional effects diminish with distance as the expanding shock wave approaches a spherical shape. A correlation was developed in the 1970s to account for directional effects using high explosive test data, the best available resource at the time. Like all test programs, pressure transducers extended to limited distances from the explosive charge, yet the data are often extrapolated to a far greater distance. This paper presents the results of recent work on directional effects specific to bursting cylindrical pressure vessels and provides new correlations for blast overpressure and impulse for a range of vessel geometries and burst conditions. The results can be used to predict the airblast hazards from cylindrical PVBs over the range of standoff distances for which directional effects exist.
Modular metal buildings are widely used in petrochemical facilities for equipment centers as they are relatively inexpensive and can be constructed and finished out much more quickly than conventional buildings in a turn-key operation. These buildings typically consist of self-framing stiffened metal panels for the roof and walls, and are supported on top of a steel beam frame at floor level. Single-Degree-of-Freedom (SDOF) analysis is often performed to design and predict relevant component responses. Finite Element Analysis (FEA), however, can more precisely evaluate and predict both component and global responses. This paper details an FEA study of the structural response of modular metal buildings to typical blast loads generated from a vapor cloud explosion. The roof and walls are analyzed both individually and as a combined structure to determine geometrical requirements to sustain the design blast loads, and connections are also evaluated. The FEA results are compared to those attained from SDOF models and limited previously published experimental data. Additional design considerations, such as recommendations for appropriate design criteria for equipment shelters, are also provided.
Deflagration to detonation transitions (DDT) were previously observed in a set of large-scale unconfined vapor cloud explosion (VCE) tests involving ethylene-air mixtures in a moderately congested environment. The test rig congested region was 48 feet long by 12 wide by 6 feet high, with a regular array of vertical circular tubes providing the congestion (pitch-to-diameter ratio of 4.5, area and volume blockage ratios of 22% and 4.1%, respectively). The test rig was configured without any confinement (i.e., no wall or roof sections). The fuel-air mixture completely filled the congested region in these tests, but did not extend beyond it. DDTs were observed at ethylene concentrations ranging from 5.9% (lean) to 9.3% (rich). A comparison has now been made between a DDT criteria established for the FLACS code and this test data. The FLACS DDT criteria utilize the maximum dimensionless pressure gradient and were established by benchmarking primarily against hydrogen-air tests in confined geometries (i.e., pipes and channels). The comparison with the existing ethylene-air VCE DDT data indicates the critical value for the maximum dimensionless pressure gradient could differ from that developed based on the confined hydrogen-air tests.
In this paper, the interaction between a blast wave and an array of cylindrical objects (pipe or vessel) has been numerically investigated to evaluate blast load distribution around the objects. It was found that the interaction leads to a blast focusing within the gap between two adjacent pipes, and that the focused overpressure may exceed a fully reflected overpressure. This paper examines the effects of a combination of the arrangement of pipes (diameter, gap distance) and incident blast wave characteristic parameters (positive phase duration) on the focusing phenomena and the resultant blast load distribution.
This paper presents example applications of a computational fluid dynamics (CFD) computer code to the evaluation of blast wave propagation and interaction problems relevant to the chemical processing and petroleum refining industries. The specific phenomena illustrated in these example applications are: clearing of a blast wave over a wall, enhancement of blast loads behind a door blast shield, and the development of an enhanced negative phase on an enclosure wall.
Contact: +1 (210) 824-5960
