Streamline simulation is a complementary technology that is gaining popularity in the industry by providing fast desktop simulation of large reservoir models or multiple realizations. Researchers at SINTEF have worked with this technology for more than 15 years and offer several interesting and challenging student projects:

• High-resolution methods in streamline simulation - A key to high efficiency of streamline simulators is that fluid transport is computed by solving one-dimensional transport problems along streamlines rather than on a 3D grid in physical space. Our inhouse simulators have so far used either unconditionally stable front-tracking methods or simple upwind methods for the one-solution operators. In this project, you will explore different high-resolution solvers (WENO, central-difference, etc) and possibly investigate efficient implementations using programmable graphics cards.
• Tracing on unstructured polyhedral grids - Currently, our inhouse solvers have efficient methods for tracing streamlines on hexahedral, tetrahedral, and corner-point grids. In this project, your task will be to develop efficient tracing methods for PEBI, prismatic, and fully ustructured grids.
• GPU acceleration - use CUDA or OpenCL on programmable GPUs to accelerate the solution of 1-D transport problems along streamlines and gravity lines and/or pvt and flash calculations. This project is in collaboration with our research group on heterogeneous computing.

We offer supervision at NTNU and at the universities of Bergen and Oslo. For all the projects listed above, you must work in close collaboration with the GeoScale team and we expect you to do quite a bit of programming.

Contact: Knut-Andreas Lie