Optimizing Oil Production by Novel Technology Integration – Well Flow Modeling

Oil is used in a wide range of applications worldwide, and it is a primary resource for energy systems. In energy systems, oil is used as a combustible fuel. In 2020 oil is expected to be responsible for 32% of the share of the global energy system. The global demand for hydrocarbons is expected to increase towards 2030, before declining due to a worldwide energy transformation.

The present thesis uses three-dimensional computational fluid dynamics (CFD) to study subsurface fluid flow in oil wells. The thesis numerically investigates the impact of different completions on the near-well reservoir, the impact of a well on the reservoir by coupling CFD with reservoir simulators, and erosion modeling of completion components.

More specifically, the thesis investigates different aspects of the well flow to optimize the important design process. The design process is important, as the wells are generally expected to be operational for 20 - 30 years with only minor interventions, as major interventions are extremely costly.

In this work, we develop an algorithm that couples a CFD code and a reservoir simulator. The two-way coupling scheme is weak and it is based on fixed-point iterations, to find a solution to the system of non-linear equations. The fixed-point iteration scheme solves the individual domain separately while keeping the others fixed.

The coupling algorithms are implemented in Matlab and CFD is controlled by developed implementations of algorithms in C. Test cases of simple reservoirs show increased accuracy by using the coupling framework compared to simple well models that are commonly used in both industry and research community.

Furthermore, to accommodate the need for the coupling in the real reservoir, the framework is able to convert the near-well region of a real three-dimensional reservoir into a full CFD model. The conversion is handled by Matlab and uses implementations in java and C to build a functional model. The conversion is tested on a horizontal well producing in the North Sea.

The well is cemented and perforated. The thesis also study another horizontal well production in the North Sea. The study investigates the inflow profile and the productivity index for different formation damage scenarios, using a reduced model. A hybrid Navier-Stokes/Darcy-Forchheimer model is used in the CFD to capture the flow in both the reservoir formation and in the well.

Depending on the formation damage, the inflow profile and the productivity index for the well are affected. It is shown, that critical formation damage can influence the productivity of the well without the proper well design that reaches beyond the damage. Fluid flow in several components can reach velocities that might cause erosion by migrated sand particles.

Erosion is investigated in an inflow control device (ICD) by injecting Lagrangian particles within an Eulerian phase. Different designs of the ICD are investigated to reduce the erosion in the component. Furthermore, the ICD was investigated both experimentally and numerically in a flow loop to validate the CFD model of the component.