Development of numerical simulator for predicting non-equilibrium steam-water flow behavior and its application to estimating steam-water relative permeability

Geothermal power generation is one of the most noteworthy renewable energies because it can steadily generate electricity for 24 hours a day and for 365 days a year. Geothermal reservoirs generation, however, have a problem of steam shortage caused by excessive steam production and a large volume of recharge water injection is necessary. In order to develop the geothermal energy effectively, a simulation study is useful, which can predict the fluid flow behavior in a geothermal reservoir. The relative permeability, which is generally estimated based on experimental data, is a key data for the rigorous simulation. The steam-water relative permeability, however, is not straightforward to measure due to the complex phase change. In this research, therefore, it was attempted to develop a program to estimate the steam-water relative permeability numerically using a geothermal simulator as an engine of this program. First, in this study, we developed a two-component, two-phase simulator to predict the flow behavior of geothermal fluid considering the non-equilibrium state. We confirmed that the simulator worked accurately by comparing the basic (equilibrium) calculation results with those by STARS developed by Computer Modelling Group Ltd. in Canada. Then, using the simulator thus developed, the sensitivity of the geothermal reservoir performances, to the kinetic constants for mass transfer from water phase to steam phase and vice versa, was investigated. Finally, we developed a program to estimate the steam-water relative permeability based on the results of steam flooding experiment. This program optimizes the relative permeability so as to reproduce the core flooding experimental results using the simulator developed in the above, by means of the Iterative Latin Hypercube Sampling which is one of the non-gradient optimization methods. It was confirmed that this program could successfully optimize/estimate the steamwater relative permeability based on the hypothetical core flooding experimental data obtained numerically.