| dc.description.abstract | Hot waterflooding is a well-known method of thermal enhanced oil recovery. Prediction of the reservoir behavior is generally obtained through numerical simulation. The finite volume simulator is the most common tool to simulate the thermal reservoir. However, the computational cost of modeling large-scale, fine-grid, heterogeneous reservoirs is high due to the nonlinearity and complexity of the non-isothermal problem. The streamline-based simulator complements the finite volume simulator in generating faster results by decoupling the 3D flow equation into several 1D problems along the streamlines. Implementation of the streamline simulator in the thermal model has been introduced by several researchers. As of now, the heat conduction part is mostly solved using the operator splitting technique at the end of a global time step. This technique can introduce the operator splitting error and may not be a unique solution.
This work proposes a new approach to handle heat conduction in a streamline simulation by solving it simultaneously with the pressure equation along the gridblocks. A 2D, quarter five-spot well pattern with one injector and one producer was tested on the new streamline simulator, and the results are in agreement with the commercial finite volume simulator. Compared to the operator splitting method, the new approach is slightly slower due to a larger set of linear equation systems. However, the better accuracy compensates for the longer computational time, and it is still much faster than the commercial thermal simulator. | en |
