COMPUTATIONAL TIME REDUCTION OF COMPOSITIONAL RESERVOIR SIMULATION MODEL WITH WAG INJECTION AND GAS RECYCLE SCHEME THROUGH NUMERICAL TUNING OF SUBMODELS

Abstract

This work shows a procedure to build fast and reliable numerical models with WAG-CO2-rich injection scheme. This novel and practical approach to numerical tuning high-complexity reservoir models can save days or even months of work. Improving step 2 of the 12-step reservoir characterization and modeling methodology proposed by Schiozer et al. (2015) leads to an optimization of the numerical control of the model based on the critical compositional numerical parameters and performance diagnostics. We show the results of a probabilistic risk analysis application. For the complex case scenario presented, results show that applying the proposed technique can save roughly 80% of the total time spent to perform a risk study. Furthermore, we found that time saving tends to increase as the number of simulations increases. This work improvement comes from making a methodology that includes both compositional and black-oil numerical solver parameters in every step of the numerical tuning optimization, rendering a broader and more robust method.