Accelerating Clean Coal Gasifier Designs with Hybrid MPI/OpenMP High Performance Computing

Abstract

In this study, we present the impact of high performance computing in the development of advanced environmentally-clean coal-based power generation technologies using MFIX, an open source computational fluid dynamics code. MFIX is a legacy code with over two decades of development towards state-of-the-art models for simulating coal gasification, which was recognized for its uniqueness with a 2007 R&D 100 award. The computational algorithm involves iterative solution of twentytwo nonlinearly coupled conservation equations at each time step. We present how open source scientific software could be used to achieve high-fidelity simulations to address industrial scale problems by reducing the computational time, using high performance computing platforms so that the results can be incorporated in the design cycle. In order to take advantage of the modern high performance computing platforms, a number of improvements were implemented. Preliminary results show that hybrid mode operation can yield substantial improvement in time-to-solution when utilizing thousands of multi-core processors. We expect our experiences and lessons learnt to date would be valuable for other groups using legacy codes.