Splitting of multiphase flow from a single flowline into a dual riser
Prof. Dr. ir. Ruud Henkes – Shell/TU Delft
A new CFD solver for high accuracy multiphase flow simulations in Oil & Gas
Dr. Andrew Jackson – Engys UK
Avoiding surge pressure failures in pipelines due to waterhammer and enclosed air pockets
Dr. ir. Richard Fawcett – DRG
Shockwave interaction on Liquid slugs – Validating a 1D numerical model with large scale experiments
Ir. Michiel Tukker – Deltares
Solving vibration problems in a two-phase flowline
Ir. Luuk Hennen – DRG
A possible new development concept for Floating LNG may include a single flowline along the sea floor that splits into dual or more flexible risers. Since the gas co-produces condensate and water, design rules are needed for the splitting of the phases at the riser base manifold. Ideally the phase volume ratio of the multiphase flow should be fully equal over the two risers and remain the same as in the flow line. The hypothesis is that maldistribution can occur if the gas flow rate in the risers is so low that it gives churn flow or hydrodynamic slug flow in the risers, whereas an equal split is expected if the gas flow rate is sufficiently high to produce annular flow. To study this we have carried out lab experiments and CFD simulations.
The flow facility at the Shell Technology Centre in Amsterdam transports air and water through a 2”, 100 m long flowline, splitting into dual 15 m high risers, having a diameter of 2” or 1.25”. The pressure is atmospheric at the riser top. For the splitting configuration, we tested both a non-symmetric lay-out (so called Branching Tee) and a symmetric lay-out (so-called Impacting Tee). We also created the splitting curve by systematically changing the opening of the chokes at the top of the risers. At low gas flows, for both splitting configurations, a non-symmetric flow split was found, with flip-flopping and hysteresis in the risers. For example, a stagnant liquid flow could develop in one riser, and churn flow in the other riser, with a sudden swap of flow behaviour between the two risers. This maldistribution gradually disappeared as the gas flow rate was increased.
CFD simulations were carried out with the Fluent package using a Volume of Fluid approach for the multiphase flow through the symmetric splitter. As in the experiments, the CFD simulations also give the preference of all flow to be produced though a single riser, with a stagnant liquid column in the other riser. However, in the CFD prediction this misdistribution is found at a lower gas production than in the experiments.
Packed beds and multiphase flows are two common features of industrial processes such as furnaces  and reactor vessels , and yet the combination of the two is sparsely covered by the literature on numerical modelling. Further applications include wave action on harbour breakwaters  and multi-fluid flows in rock beds .
In this presentation, we describe our efforts to improve upon the multiphase porous modelling capabilities in the OpenFOAM® toolset. These developments comprise improvements to the handling of free surfaces on non-orthogonal meshes, as well as a numerical method for handling arbitrary heterogeneous porosity fields, possibly including discontinuities, without introducing instabilities or spurious oscillations. This allows for a general porosity distribution to be specified as just another field.
For the two-fluid (free-surface) modelling the volume of fluid (VoF) method is used with the HiRAC  interface capturing algorithm. Flow through the porous medium is modelled using the volume averaging scheme of Whittaker  and Ergun’s standard model of porous drag . Accepted values from literature are used to determine the empirical drag coefficients .
Consistent inter-cell interpolations are derived which allow for discontinuities in porosity and density (across the liquid-gas interface) to co-exist without introducing numerical oscillations or instabilities. The application of the derived numerical scheme is demonstrated for realistic 3D test problems involving wave impact on harbour structures and metal tapping from a smelter.
 M. Kadkhodabeigi: Modeling of Tapping Processes in Submerged Arc Furnaces, Ph.D. thesis, Norwegian University of Science and Technology, Trondheim (2011)
 H. Jacobsen: Chemical Reactor Modeling, Springer (2008)
 M. del Jesus, J.L. Lara, I.J. Losada: Three-dimensional interaction of waves and porous coastal structures Part I: Numerical model formulation, Coastal Engineering, Vol. 64 pp. 57–72 (2012)
 P. Bastian: Numerical Computation of Mulitphase Flows in Porous Media, Habilitationsschrft, Christian-Albrechts-Universität Kiel (1999)
 J.A. Heyns, A.G. Malan, T. M. Harms, O.F. Oxtoby: Blended surface capturing formulation for modelling free-surface flow using the volume-of-fluid approach, International Journal for Numerical Methods in Fluids, Vol. 71, pp. 788–804 (2013)
 S. Whitaker: Diffusion and Dispersion in Porous Media, American Institute of Chemical Engineers Journal, Vol. 13, pp. 420–427 (1967)
 S. Ergun: Fluid flow through packed columns, Chemical Engineering Progress, Vol. 48, pp. 89–94 (1952)
 B. Jensen, N.G. Jacobsen, E.D. Christensen: Investigations on the porous media equations and resistance coefficients for coastal structures, Coastal Engineering, Vol. 84, pp. 56–72 (2014)
This presentation discusses the validation of the 1D pipeline simulation software WANDA with experiments. The Wanda simulation software is used to perform multiple studies for gas wells and transport pipelines to assess the functioning of High Integrity Pressure Protection Systems (HIPPS) and the protection against excessive high pressures. The experiments have been performed in the large-scale multiphase flow research facility, the Alpha Loop, a 320 m long pipeline system with a diameter of 200 mm.
The project is investigates the behaviour of a slug at rest, which is accelerated due to expanding gas in a long pipeline system. Measurements of the pressures, temperatures and slug position are recorded during the experiment. The movement of the slug has also been recorded using a high-speed camera to get a better understanding of the slug.
The experiment setup has been modelled in WANDA and the model was calibrated using measurements where there is no slug, but only gas flows through the pipeline. The simulation results are then compared with the measured slug velocities, pressures and temperatures from the experiment. The results show good correlation between the calibrated model and the measurements. However, observed 3D phenomena in the experiment due to the large scale are not included in the current 1D model of Wanda, causing the 1D model to predict higher slug velocities than were recorded in the experiment.
Multiphase flow in piping systems can pose a serious hazard to system integrity. Slugs that form in piping, due to a two-phase flow that is fed to the system, can impose big shock loads on piping components such as elbows, tees and valves. This can be a one-time event or cyclic by nature. Such phenomenon can therefore result in dynamic behaviour of the piping itself such as vibrations or shifting of the entire line.
In contrast to pipe vibrations due to connecting rotating equipment, multiphase flow often results in vibrations with a low frequency and high amplitude. Those vibrations can be easily observed by the naked eye. Such vibrations are regularly experienced in furnaces, water feed lines or wellhead lines and can result in mechanical fatigue failure, immediate pipe failure or broken supports. Additionally, the presence of entrapped gas in the system may result in large pressures, significantly exceeding the system design pressures.
Different approaches for solving vibration issues due to two-phase flow will be presented from risk assessment to stress calculations and associated mitigation methods. Often accelerometer measurements are taken to substantiate dynamic calculations. Challenges for such projects include balancing between required system flexibility and sufficiently supporting to absorb slug loads and the assessment of the actual slugging conditions.