Speed Up of Lattice Gas Automata Simulation of Polymer Flow Using a Cluster System

The performance of polymer flooding may be estimated from microscopic behaviour of polymer particles in the pores of porous media. Lattice gas automata method is a modelling method that has been used by researchers to study flow behaviour on a particle scale. It had been also used to simulate polymer flow in porous media for studying microscopic interactions between rock and polymer. However, the computation time was too lengthy for even a 5 cm long porous media sample. On this study, parallel computing using standalone computers and a cluster system was investigated in an effort to decrease the computation time of modelling polymer flooding using lattice gas automata. From the results, speedup due to parallel processing was greater than 3.75 times for four processors and less. Differences in the estimations of displacement efficiency and saturations between sequential and parallel programming were less then three percent. It was concluded that parallel programming was successfully used to speed up computations in polymer flooding without causing significant variations in the results