A novel approach to high-pressure gas releases simulations
Introduction
The field of consequences assessment and damage quantification are crucial aspects in the industrial risk assessment, especially for those plants which are considered risk-relevant like Oil & Gas (O&G) or nuclear ones. In this plants, many activities are performed and involve hazardous substances at high pressure inside equipment and pipes (Paté-Cornell 1993; Necci et al., 2019) e.g. the most recent design of fusion installations foresees the possibility to use high-pressure gaseous coolants (Zappatore et al., 2020). Moreover, consequences are often worsened by the congested environment that additionally challenges the management of the accidental scenarios (Uggenti et al., 2017).
In the framework of safety assessment of offshore installations, the European Union (EU) Offshore Safety Directive 2013/30/EU affirms that Quantitative Risk Analysis (QRA) tools are the most appropriate for damage estimation of major hazards, in particular fire and explosions.
Nowadays, the major part of the consequences analysis for industrial applications is performed through semi-empirical models such as turbulent free-jet models (Chen and Rodi 1980), gas dispersion models (Davidson 1989) and jet fire models (TNO 2005; Zamejc 2014). These tools, based on massive measurements and few theoretical principles, neglect the geometry and approximate the physical phenomena. On the one hand these assumptions guarantee the semi-empirical models’ simple and rapid implementation, essential for the simulations of a huge number of accidental scenarios, typical aspect of a QRA. On the other hand, the previous hypotheses lower the accuracy of the results, leading to consequences overestimation and, accordingly, to overprotected systems, overloaded bearing structures and waste of materials and money especially if highly congested plants are considered. Typical examples are represented by nuclear power plants, since, for safety reasons, the major part of the equipment is in the containment building and O&G offshore platforms, in which all the equipment must be arranged in a small volume due to space issues.
The alternative to the empirical methods is represented by more accurate models, for example based on the Computational Fluid Dynamics (CFD). They guarantee more realistic results, taking into account the complex geometries and phenomena overlooked by the semi-empirical models. As a drawback, the computational effort and the implementation complexities limit the use of these tools during the preliminary phases of the design and become incompatible with the project schedule (Impalà et al., 2017). Usually, then, only few scenarios are CFD analyzed, and the results are available when the key design choices are already made, reducing the CFD role to a final verification.
To overcome this drawback, this work presents a new approach, called Source Box Accident Model (SBAM) (Carpignano et al., 2017), targeted to be a compromise between the semi-empirical and pure CFD models and aiming at decreasing the calculation costs maintaining a good results accuracy.
The approach is focused on toxic and/or flammable accidental gas releases from pipelines or tanks in complex geometries. Additionally, this tool can be used also for asphyxiating gases (e.g. CO2). Due to the frequent presence of high-pressure components in industrial plants, accidental releases from 10 bar or more are considered. This phenomenon is complex being characterized by a multi-physics and a multiscale nature: the jet is supersonic and compressible near the release point and becomes a subsonic and incompressible dispersion in the remaining portion of the domain, the largest one.
From a literature survey, resulted that several papers treated the CFD simulation of gaseous releases and dispersions. Some papers are focused only on supersonic free-jets (Birkby and Page 2001; Fairweather and Ranson 2006), and do not account for neither the jet-objects interaction nor the gas dispersion. On the contrary, in (Lin et al., 2021; Li et al., 2019; Wang et al., 2020; Dong et al., 2017) only the gas dispersion due to a low-inertia release is modeled. More complex high-pressure releases in big environments are treated in (Venetsanos et al., 2008; Choi et al., 2013; Liu et al. 2014, 2015; Deng et al., 2018); however the initial supersonic release is never described by CFD tools, but it is substituted by semi-empirical approaches or other simplified models based on conservation laws. For example, two of the most used CFD software for QRA, FLACS (Gexcon 2019) and KFX (DNV GL 2019), use the Birch model (Birch et al., 1984) to evaluate the initial jet-expansion, and a simplified formulation of the flow field governing equations (Porosity Distributed Resistance CFD approach). In (Wilkening and Baraldi 2007) both release and dispersion are simulated in a single CFD simulation, nonetheless the resulting computational cost is prohibitive and a High Performance Computing is employed. SBAM is developed to handle both the supersonic release and dispersion via CFD in a detailed and computationally sustainable way.
In section 2 the problem statement is introduced by describing the main features of the physics involved in the phenomena treated by SBAM, with particular attention to highly under-expanded jets; two different flow regimes are identified: the supersonic release and the low velocity dispersion.
In section 3, the rationale behind SBAM is explained. The concept of Source Box (SB) is introduced, i.e. a small domain in which the release phase is simulated. Then, it is explained how the dispersion is simulated taking as input the results of the SB. Hence the coupling of these two simulations is described.
In section 4, the two analyses carried out on SBAM are described: a sensitivity analysis on the coupling parameters and a numerical benchmark.
In section 5, a typical case study, which can be treated by SBAM is introduced: an accidental high-pressure gas release in an offshore platform. The numerical methods employed in the SBAM simulations are discussed.
In section 6, the results are discussed with reference to the typical SBAM output: damage areas, damage metrics as flammable volume and mass.
In section 7, final considerations and conclusion on the model are summarized.
Section snippets
Physics overview: problem statement
This paper considers an accidental rupture on a piece of equipment (pipe, tank, etc.) through which the contained substance, a high-pressure (>10 bar) hazardous gas, exits in the surrounding environment at ambient pressure. These releases have similar fluid-dynamic and geometrical characteristics, and they are defined as highly under-expanded jets as the necessary condition shown in Eq. (1) between the release pressure (prel) and the ambient pressure (pa) is satisfied (Franquet et al., 2015):
Methods: SBAM
The release is simulated in a small domain, the SB, considering the compressibility effects with enough accuracy to capture the under-expanded jet features. The compressible inertia driven flow is treated using suitable models and refined meshes, and the simulation is performed in steady-state as for highly under-expanded jets the transient evolution of the jet is fast enough to be neglected; in fact, as demonstrated by (Tang et al., 2017), the jet stabilization occurs in ~40 μs.
The results of
Tests on SBAM parameters and performances
A sensitivity analysis on the coupling parameters is performed: alternatives with respect to the coupling method explained in section 3.2 are investigated. The second analysis concerns a numerical benchmark of SBAM in comparison to a standard CFD simulation.
Case study and numerical setup
This work considers an offshore platform for natural gas extraction located in the Adriatic Sea: all environmental conditions refer to this site. The analysis is carried out considering only the platform production deck. It can be assumed separated from the adjacent ones since plated grounds are commonly used in gas extraction platforms to contain the spreading of potential gaseous leakages. In Fig. 6 the CAD and the dimensions of the simplified domain are depicted; all the components are
Source-box results
The results shown in Fig. 9, are obtained by mirroring the solution with respect to the symmetry planes. In Fig. 9 (left) a volume rendering of the flow inside the SB shows the jet velocity spatial distribution: the high inertia of the jet is concentrated in the near-field while towards the boundaries the gas flows at subsonic velocity, confirming the correctness of the assumptions on the SB dimensions (Eq. (8)). In Fig. 9 (right) a velocity contour plot at the transversal SB plane shows that
Conclusions
In this paper, a novel CFD approach called SBAM is extensively described and discussed. It aims at simulating via CFD high-pressure (>10 bar) gas releases in big congested environments, guaranteeing a good computational cost-accuracy compromise.
Firstly, two different flow regimes are defined, the release and the dispersion. SBAM and its main features are presented: the release is simulated in a small domain (the SB) accounting for the supersonic compressible flow, the dispersion is simulated in
Author contribution statement
Alberto Moscatello: Methodology, Software, Formal analysis, Investigation, Data curation, Writing-Original Draft, Writing-Review & Editing, Visualization. Anna Chiara Uggenti: Conceptualization, Methodology, Investigation, Writing-Review & Editing. Raffaella Gerboni: Conceptualization, Investigation, Writing-Review & Editing, Supervision. Andrea Carpignano: Conceptualization, Resources, Writing-Review & Editing, Supervision, Project administration, Funding acquisition.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgment
The research presented in this paper has been sponsored by the Italian Ministry of Economic Development's Directorate General for Safety - National Mining Office for Hydrocarbons and Georesources.
References (43)
- et al.
A CFD simulation of hydrogen dispersion for the hydrogen leakage from a fuel cell vehicle in an underground parking garage
Int. J. Hydrogen Energy
(2013) Simultaneous trajectory and dilution predictions from a simple integral plume model
Atmos. Environ.
(1989)- et al.
A method for simulating the release of natural gas from the rupture of high-pressure pipelines in any terrain
J. Hazard Mater.
(2018) - et al.
Simulation of heavy gas dispersion in a large indoor space using CFD model
J. Loss Prev. Process. Ind.
(2017) - et al.
Free underexpanded jets in a quiescent medium: a review
Prog. Aero. Sci.
(2015) - et al.
Simulation and assessment of gas dispersion above Sea from a subsea release: a CFD-based approach
Int. J. Naval Arch. Ocean Eng.
(2019) - et al.
CFD simulations on high-buoyancy gas dispersion in the wake of an isolated cubic building using steady RANS model and LES
Build. Environ.
(2021) - et al.
Source strength and dispersion of CO2 releases from high-pressure pipelines: CFD model using real gas equation of state
Appl. Energy
(2014) - et al.
Study of the consequences of CO2 released from high-pressure pipelines
Atmos. Environ.
(2015) - et al.
Lessons learned from offshore Oil and gas incidents in the arctic and other ice-prone seas
Ocean. Eng.
(2019)
Numerical investigation of a high pressure hydrogen jet of 82 MPa with adaptive mesh refinement: the starting transient evolution and Mach disk stabilization
Int. J. Hydrogen Energy
CFD modelling of hydrogen release, dispersion and combustion for automotive scenarios
J. Loss Prev. Process. Ind.
A dynamic approach for evaluating the consequences of toxic gas dispersion in the chemical plants using CFD and evacuation modelling
J. Loss Prev. Process. Ind.
CFD modelling of accidental hydrogen release from pipelines
Int. J. Hydrogen Energy
API standard 521 new alternative method to evaluate fire relief for pressure relief device sizing and depressuring system design
J. Loss Prev. Process. Ind.
Mach disk and Riemann wave in underexpanded jet flows
Fluent 18.2 User's Guide
The structure and concentration decay of high pressure jets of natural gas
Combust. Sci. Technol.
Numerical predictions of turbulent underexpanded sonic jets using a pressure-based methodology
Proc. IME G J. Aero. Eng.
Modelling of a supersonic accidental release in Oil&Gas offshore: characterisation of a source box
Geoingegner. Ambient. Miner.
Vertical Turbulent Buoyant Jets
Cited by (8)
A novel approach combining bootstrapped non-intrusive reduced order models and unscented transform for the robust and efficient CFD analysis of accidental gas releases in congested plants
2023, Journal of Loss Prevention in the Process IndustriesCFD gas release model performance evaluation through wind tunnel experiments
2022, Journal of Loss Prevention in the Process IndustriesCitation Excerpt :The SB results, specifically the gas concentration profiles, and the velocity profiles calculated on its faces, are used as boundary conditions for the dispersion simulation, where the propagation of the cloud within the platform structure is modelled. Details on SBAM are discussed in (Moscatello et al., 2021). It is important to specify that the accidental release is assumed continuous in time and both simulations are carried out in steady state, since the final purpose is the evaluation of risk related quantities like flammable volume/mass in the worst condition that is the one reached at the end of the transient when the gas reach its maximum volume.
Change of location class in gas pipelines from a regulatory perspective
2024, Process Safety Progress