Optimizations in an numerical method code for two-phase fluids flow in porous media using the SDumont supercomputer
Palavras-chave:
Porous media, Numerical methods, Message Passing Interface (MPI), High-performance computing, Decomposition domain strategy, Supercomputing for enabling large-scale advanced simulationsResumo
In petroleum reservoir simulations, the level of detail incorporated into the geologic model typi-
cally exceeds the capabilities of traditional flow simulators. In this sense, such simulations demand new
high-performance computing techniques to deal with a large amount of data allocation and the high
computational cost to compute the behavior of the fluids in the porous media. This paper presents opti-
mizations performed on a code that implements an explicit numerical scheme to approximate the solution
of the governing differential equation for water saturation in a two-phase flow problem with heterogeneous
permeability and porosity fields. The experiments were performed on the SDumont Supercomputer using
2nd Generation Intel® Xeon® Scalable Processors (formerly Cascade Lake architecture). We employed
the domain decomposition method to split the 3-D reservoir onto n sub-domains that are solved using each
Message Passing Interface (MPI) process. We analyzed the performance of the domain decomposition
strategies and identified communication bottlenecks as the mesh size and the number of computational
nodes increase. The results show that the optimizations implemented in the numerical code remarkably
reduce the execution time of the simulations.