CFD for two phase flow in a splitter used for offshore oil and gas production

The student will carry out the work at the office of Dynaflow in Zoetermeer.


To produce the remaining hydrocarbon gas reserves, more advanced technologies are needed. This is because such reserves are at off-shore locations, more remotely and in deep waters. One of such new technologies is Floating LNG (FLNG). Here, the gas plant is formed by a floating vessel, positioned close to the remote location. Due to the strong currents the vessel needs to be able to rotate through a so-called turret, which is where multiphase subsea flexible risers are connected to the vessel. A multiphase mixture of gas, oil and water is transported through the flexible risers, which bridges the full water depth. Due to the weight of the risers, its diameter is limited. Therefore multiple risers are needed to bring the production from the single subsea pipeline to the vessel. As a consequence the multiphase flow needs to be spitted from the pipeline into multiple risers.

Recently two Master students of the TU Delft have carried out laboratory experiments to investigate how a non-symmetric split of the gas/liquid flow between the multiple risers can occur, and how this can be mitigated. A non-symmetric distribution is undesired as it gives strong pressure fluctuations and an irregular outflow of gas and liquid at the riser top. It has been investigated how using control through actuated valves at the top of the risers can improve the symmetry. Thereto, laboratory experiments with air and water were carried out in a 100 m long pipeline (with 5 cm diameter), connected to dual 15 m high risers. These experiments were carried out at the Shell Research Centre in Amsterdam.

Current engineering design tools used in the oil and gas industry are unable to properly predict the split of the two-phase flow at the splitter location. This is because these models are all one-dimensional, whereas the three-dimensional shape of the splitter geometry has a large effect on how the precise flow is established. Therefore a three-dimensional model is needed using Computational Fluid Dynamics.

Tasks in Master End Project

• Give overview of the literature on computational models for two-phase flow in splitters.
• Build computational models for the various splitter configurations as recently used in the experiments at Shell. Thereto, the OpenFOAM tool will be used, which is an open-source computational framework for solving the Navier-Stokes equations with turbulence models.
• Compare model predictions with available experimental data.
• Define simple correlations (if possible) that can be used in the one-dimensional design tools used in the oil and gas industry.
• Write and defend Master Thesis.

Figure: Sketch of flow patterns as found in the lab experiments with a non-symmetric split from a single horizontal flowline to dual vertical risers. (green = liquid, white=gas).

Supervisors: Dr. ir. F.M. Bos (Dynaflow Research Group BV Zoetermeer) and Prof. dr. ir. R.A.W.M. Henkes


Involved People:

Facilities used: