Microsoft Word - 6laurent.docx CHEMICAL ENGINEERING TRANSACTIONS VOL. 36, 2014 A publication of The Italian Association of Chemical Engineering www.aidic.it/cet Guest Editors: Valerio Cozzani, Eddy de Rademaeker Copyright © 2014, AIDIC Servizi S.r.l., ISBN 978-88-95608-27-3; ISSN 2283-9216 Effect of Large Obstacles on High Momentum Jets Dispersion Marco Pontiggiaa, Valentina Busini*b, Martina Ronzonib , Giovanni Uguccionia, Renato Rotab aD'Appolonia S.p.A. - via Martiri di Cefalonia 2, 20097 – San Donato Milanese (MI) - Italy bPolitecnico di Milano - Dip. di Chimica Materiali e Ingegneria Chimica "G. Natta" - P.zza Leonardo da Vinci 32 - 20133 Milano valentina.busini@polimi.it Dispersion of toxic and flammable materials from Chemical industries represents a major issue in Risk Analysis; presently, integral models are generally used to assess dispersion consequences, due to the low CPU and time requirements connected to the use of these tools. Nevertheless, they are mainly developed and tuned for releases in open field (open spaces without relevant obstacles), and therefore they cannot properly account for the geometrical features of the dispersion domain. Computational Fluid Dynamic, on the other hand, allows a full 3D analysis, thus accounting for all the obstacles influence on the flow field, but it involves large computational requirements. In case of gas discharge directed towards nearby large obstacles, an impinged jet is expected: if the jet hits a nearby obstacle, the gas velocity suddenly drops, minimizing the inertial dispersion phase, thus reducing the relevant air entrainment and generally increasing the damages distances. Impinged release models are included in some commercial integral models for consequences assessment even if a clear method to decide when to use them is often missing. The aim of this work is to provide a comparison between the two approaches (CFDs vs. integral tools) in predicting damage thresholds for both impinged and non impinged jets. A realistic case-study of industrial interest was set-up and the fine tuning of all the involved models and parameters (turbulence modeling, geometry description, mesh independence, etc.) was finalized. 1. Introduction Dispersions of toxic and flammable materials from Chemical industries represent a major issue in Risk Analysis, since they usually reach very large damage distances thus potentially involving a great number of people both inside and outside the plant; presently, integral models (simplified, uni-dimensional models) are generally used to assess dispersion consequences, such as DEGADIS, SLAB, ALOHA (BernatikLibisova, 2004) and UDM (Pandya et al., 2012), due to the low CPU and time requirements connected to the use of these tools. Integral models are lumped-parameter models, usually pseudo one- dimensional, which account for some physical phenomena using semi-empirical relationships whose parameters are tuned on field test data (Hanna, 1994). Thus, they are mainly developed and tuned for releases in open field (open spaces without relevant obstacles), and therefore they cannot properly account for the geometrical features of the dispersion domain (Brook et al., 2003). Significant obstacles produce eddies, wakes, stagnation and recirculation points that can enhance or reduce mixing with fresh air, thus strongly influencing damage distances (Calhoun et al., 2000). Computational Fluid Dynamic, on the other hand, consists of the numerical solution of the Navier-Stokes transport equations over a computational domain spatially discretized through the definition of a calculation grid. This approach allows for a full 3D analysis, thus accounting for all the obstacles influence on the flow field as discussed in previous works (Busini et al., 2012, Busini et al., 2011, BusiniRota, 2014, Derudi et al., 2014, Pontiggia et al., 2012, Pontiggia et al., 2010, Pontiggia et al., 2009, Pontiggia et al., 2011), but it involves large DOI: 10.3303/CET1436088 Please cite this article as: Pontiggia M., Busini V., Ronzoni M., Uguccioni G., Rota R., 2014, Effect of large obstacles on high momentum jets dispersion, Chemical Engineering Transactions, 36, 523-528 DOI: 10.3303/CET1436088 523 computational requirements (Tauseef et al., 2011, Steffens et al., 2013, AiMak, 2013, TominagaStathopoulos, 2013). In case of gas discharge directed towards nearby large obstacles, an impinged jet is expected: high momentum jets in open field are characterized by high velocities relative to the ambient air, thus involving a significant air entrainment; if the jet hits a nearby obstacle, the gas velocity suddenly drops, reducing the air entrainment and generally increasing the damages distances. Impinged release models are included in some commercial integral models for consequences assessment; since the use of the impinged vs. the open field model can produce large differences in damage distance, a reliable criterion to select the most appropriate model based on jet characteristics and release geometry is required. The aim of this work was to work out a comparison between the two approaches (CFDs vs. integral models) in predicting damage thresholds for both impinged and non-impinged jets with a realistic case- study of industrial interest. Results of the two approaches were compared in order to obtain: 1) a cross check of the CFD results in open field (thus validating the effectiveness of the CFD models in predicting the effect of atmospheric turbulence in open field, where the integral model are largely validated); 2) a validation of the CFD capacity in describing the interaction between high momentum jets and geometrical obstacles in the near field, verifying that the initial loss of momentum of the impinged jet would resolve in lower initial fresh air entrainment and therefore longed damage distances; 3) a comparison between CFD and integral methods predictions of impinged jet damage distances, to verify the over-conservative approach of integral methods; 4) a solid and time effective simulation approach to be applied in a massive number of CFD runs to build up a criterion to evaluate the best available model for impinged jet dispersions. 2. Materials and Methods In this work, CFD was used to perform the simulations coupled with the AsSM (Pontiggia et al., 2009) for the description of an atmospheric stability class consistent with Monin-Obhukov similarity theory profiles across the integration domain. Thus fully developed vertical profiles of velocity, temperature, turbulence intensity, and dissipation rate were used as boundary conditions at the wind inlet boundary. Standard boundary conditions were used for all the other boundaries (as reported in Table 1). The commercial package Fluent 12 (ANSYS Inc., 2009) was used for all the computations. 3. Results and Discussion The case-study treated is part of a regasification plant in which an accidental release of methane gas was hypothesized. The jet is coming out of an Open Rack Vaporizer (ORV) and disperses in atmosphere leading to a steady-state release. In the vaporizer the methane is stocked in gas phase at the absolute pressure of 65 bar and at a temperature of 4,5 °C. The hole diameter is 0,0254 m (1 inch) and positioned at the centre of the pipe external surface. Firstly the open field dispersion of natural gas was modelled for a neutral stability class and 5 m/s wind speed at 10 m above the ground with the suite package PHAST in order to define the dimension of the expanded diameter, the final velocity of the jet (i.e., the velocity of the jet in correspondence of the expanded diameter) and the mass flow rate. The release is sonic and the gas calculated velocity after the atmospheric expansion is about 377 m/s, the expanded diameter is 0.7037 m and the mass flow rate is 5.54 kg/s and the gas temperature is -65.72°C. The CFD simulations where performed considering the estimated condition of the release, the density of the gas was modelled as an ideal gas at constant pressure (thus providing only the dependency of density upon temperature). The geometry of the case-study comprehend the ORV (16mx8mx8m) and the pipeline (10 m long with a diameter of 0.4 m and 1 m far from the ground) as sketched in Figure 1. The simulated domain was 300mx50mx50m, thus wide enough to ensure the independencies of the simulations results from the chosen domain; it was meshed with a triangular grid for the faces and a tetrahedral one for the volumes, paying attention at the density of the cells, which must be higher in correspondence of the critical spots (such as the hole and the obstacle); to mesh the open field case, about 6.5.105 cells were used. To verify the independence of the results from the used mesh, the simulations were carried on also with a mesh of about 140.105 cells; the results of the two different configurations were comparable. 524 Figure As jet locate accord Simul open distan report We ca case, open metho e 1: domain to t-impinging ob ed 6 m far fro ding to typical ations results field conditio nce reached b ted. This allow an see that in thus validatin field. Howeve odologies emb Table 1 Bo W Win Top Latera G G op global view bstacle, a cylin m the hole, a l layout of real are sketched ns, the LFL f by the LFL wit ws for an easy n open field c ng the effectiv er, the agreem bedded in inte : Boundary c oundary Wind Inlet nd Outlet boundary al boundary Ground Gas Inlet Walls w nder 2 m high as sketched in l regasification d in Figure 2, footprints in p th and without comparison b conditions the veness of the ment worsens f gral models to onditions used velocit velocit velocit wall, Velocity wall@30 and with a di n Figure 1. Ob n plant. Figure 3 and presence of t t the obstacle between Fluen CFD results e CFD in pred for the imping o overestimate d in all the sim Type ty inlet, velocit pressure outl ty inlet, velocit ty inlet, velocit roughness = inlet, 377 m/s 0 K, roughnes iameter of 1.5 bstacle dimen Figure 4 in w the obstacle (i.e., in case nt and PHAST are in very g dicting the eff ed jet, therefo e the hazardo mulations ty profile et ty profile ty profile 0.05 m , -65.72 °C ss = 0.05 m 5 m was added sions and loc which the the L and the parit of impingeme T results. good agreeme fect of atmosp ore confirming us distance. d in the doma cation were se LFL footprints ty plot of dow ent and free je ent with the P pheric turbule the tendency ain and elected under wnwind et) are PHAST ence in y of the 525 Figure Figure e 2: LFL footp e 3: LFL footp rint of methan rint of methan ne jet under op ne jet under im pen field condi mpinging condi ition predicted ition predicted d by the CFD d by the CFD 526 4. Co This w disper studie jet hit signifi entrai Integr never avoid produ In ord The a 1 2 3 4 Figure 4 Pa onclusions work is focuse rsions are oft es, and their p ts a large obs cantly limited nment in the i ral methods rtheless, releva extreme over ce jet-impinge der to analyze aim of the work ) To validat turbulence accomplis conditions releases, obstacles. of CFD in 2) To investig able to pre obstacle in recirculatio CFDs in c air entrain distances 3) To compa in this wo distances geometric and the m impingeme 4) To establ simulation arity plot of do ed on the mod ten the most otential outco stacle in the d, thus reduc nitial high mo are generally ant results ca r-conservative ement and wh this problem k was fourfold te CFD appro e modeling) hed by means : since integ relevant perf Results (exp simulating hig gate the phys edict effects o nfluence in te on point. On t correlating the ning in the hi in terms of dis re impinged g ork has confir for impinged arrangement most suitable ent). ish a robust with CFD cod wnwind distan deling of impin critical event mes can beco near field (w cing the relat mentum dispe y provided w n be over-con e results, it is i ich obstacles , CFD tools w : oach (source for gas dis s of a compar gral model pa formances ar ressed in term gh momentum sical phenome of geometrica erms of turbule the other han initial velocity igh momentu stance to haza gas distances med the gen jet. This feat (i.e. distance e model to b and time/res des: such an nce reached b nged jet dispe ts in terms of ome even wor here the gas tive velocity ersion phase. with specific nservative; to o mportant to d are too small were applied to term represe persion from ison among C arameters are re well estab ms of distance jet gas dispe ena involved l features on ence increasi nd, performed y loss due to m gas disper ardous gas co forecast by C eral tendency ure raises the e between jet be used in c sources effec approach will by the LFL with ersion in the e f consequenc rse due to the momentum i with the atm models for j obtain credible istinguish in a or too distant o a realistic ca entation, geom m high mome CFD and integ e tuned on e blished in abs e to flammabili rsion in open f in the jet imp the flow field ng due to the simulations, nearby obstac rsion and the oncentration. FDs and integ y of integral m e necessity fo and obstacle consequences ctive approach be adopted, h and without environment: t es distances impingement is still high), mospheric air, jet impingem e consequenc a proper way w to trigger this ase of flamma metry discreti entum jet; t ral model resu experimental sence of sig ity limits) show field. pingement sce , are fully cap e formation of highlighted a cles impingem erefore provid gral model: the model in over or improved c , dimension o s calculation h for high ve as a future wo the obstacle toxic and flam in risk asses phenomenon the axial velo and, therefo ment represen ces assessme which obstacle s phenomenon able gas dispe zation, atmos his validation ults under ope data of open nificant geom wn a good cap enario: CFDs, pable to acco f eddies, wake lso the capab ment with the ding higher da e analysis des restimating da correlations be of the obstacle (impinged vs elocity imping ork, in order to mmable ssment n: if the ocity is ore air ntation: ent and es can n. ersion. spheric n was en field n field metrical pability being unt for es and bility of limited amage scribed amage etween e, etc.) s. non ged jet o work 527 out a massive number of simulation to investigate an empiric relation between scenario description and most suitable modeling approach. References Ai Z. T., Mak C. M., 2013, CFD simulation of flow and dispersion around an isolated building: effect of inhomogeneous ABL and near-wall treatment, Atmospheric Environment, 77, 568. Ansys Inc. 2009. ANSYS Fluent 12 User's guide. Lebanon, NH, USA. Bernatik A., Libisova M., 2004, Loss prevention in heavy industry: risk assessment of large gasholders, Journal of Loss Prevention in the Process Industries, 17, 271-278. Brook D. R., Felton N. V., Clem C. M., Strickland D. C. H., Griffiths I. H., Kingdon R. D., 2003, Validation of the Urban Dispersion Model (UDM), International Journal of Environment and Pollution, 20, 11- 21. Busini V., Lino M., Rota R., 2012, Influence of Large Obstacles and Mitigation Barriers on Heavy Gas Cloud Dispersion: a Liquefied Natural Gas Case-Study, Industrial & Engineering Chemistry Research, 51, 7643-7650. Busini V., Pontiggia M., Derudi M., Landucci G., Cozzani V., Rota R., 2011, Safety of LPG Rail Transportation, Chemical Engineering Transactions, 24, 1321-1326. Busini V., Rota R., 2014, Influence of the shape of mitigation barriers on heavy gas dispersion, Journal of Loss Prevention in the Process Industries, 29, 13-21. Calhoun R., Chan S., Leone R. L., Shinn J., Stevens D. Year. Flow patterns around a complex building. In: 11th Joint Conference on the Applications of Air Pollution Meteorology with the A&WMA, 2000 AMS, 45 Beacon Street, Boston 02108. 47–52. Derudi M., Bovolenta D., Busini V., Rota R., 2014, Heavy gas dispersion in presence of large obstacles: selection of modeling tools, Industrial & Engineering Chemistry Research Hanna S. R., 1994, Hazardous Gas-Model Evaluations - Is an Equitable Comparison Possible, Journal of Loss Prevention in the Process Industries, 7, 133-138. Pandya N., Gabas N., Marsden E., 2012, Sensitivity analysis of Phast's atmospheric dispersion model for three toxic materials (nitric oxide, ammonia, chlorine), Journal of Loss Prevention in the Process Industries, 25, 20-32. Pontiggia M., Busini V., Gattuso M., Uguccioni G., Rota R., 2012, Consequences assessment of an accidental toxic gas release through a cfd tool: Effect of the terrain and major obstacles, Chemical Engineering Transactions, 26, 537-542. Pontiggia M., Derudi M., Alba M., Scaioni M., Rota R., 2010, Hazardous gas releases in urban areas: Assessment of consequences through CFD modelling, Journal of Hazardous Materials, 176, 589- 596. Pontiggia M., Derudi M., Busini V., Rota R., 2009, Hazardous gas dispersion: A CFD model accounting for atmospheric stability classes, Journal of Hazardous Materials, 171, 739-747. Pontiggia M., Landucci G., Busini V., Derudi M., Alba M., Scaioni M., Bonvicini S., Cozzani V., Rota R., 2011, CFD model simulation of LPG dispersion in urban areas, Atmospheric Environment, 45, 3913-3923. Steffens J. T., Heist D. K., Perry S. G., Zhang K. M., 2013, Modeling the effects of a solid barrier on pollutant dispersion under various atmospheric stability conditions, Atmospheric Environment, 69, 76-85. Tauseef S. M., Rashtchian D., Abbasi S. A., 2011, CFD-based simulation of dense gas dispersion in presence of obstacles, Journal of Loss Prevention in the Process Industries, 24, 371-376. Tominaga Y., Stathopoulos T., 2013, CFD simulation of near-field pollutant dispersion in the urban environment: a review of current modeling techniques, Atmospheric Environment, 79. 528