This paper describes a massively parallel simulation method and its application for modeling multiphase flow and multicomponent transport in porous and fractured reservoirs. The parallel-computing method has been implemented into the TOUGH2 code and its numerical performance is tested on a Cray T3E-900 and IBM SP. The efficiency and robustness of the parallel-computing algorithm are demonstrated by completing two simulations with more than one million gridblocks, using site-specific data obtained from a site-characterization study. The first application involves the development of a three-dimensional numerical model for flow in the unsaturated zone of Yucca Mountain, Nevada. The second application is the study of tracer/radionuclide transport through fracture-matrix rocks for the same site. The parallel-computing technique enhances modeling capabilities by achieving several-orders-of-magnitude speedup for large-scale and high resolution modeling studies. The resulting modeling results provide many new insights into flow and transport processes that could not be obtained from simulations using the single-CPU simulator.
机构:
New Mexico State Univ, Plant & Environm Sci Dept, Las Cruces, NM 88003 USANew Mexico State Univ, Plant & Environm Sci Dept, Las Cruces, NM 88003 USA
Gonzalez-Delgado, Amir M.
Shukla, Manoj K.
论文数: 0引用数: 0
h-index: 0
机构:
New Mexico State Univ, Plant & Environm Sci Dept, Las Cruces, NM 88003 USANew Mexico State Univ, Plant & Environm Sci Dept, Las Cruces, NM 88003 USA