Statistical prediction of waterflooding performance by K-means clustering and empirical modeling

Statistical prediction is often required in reservoir simulation to quantify production uncertainty or assess potential risks. Most existing quantification procedures aim decompose the input random field independent variables, and may suffer from curse of dimensionality if correlation scale small compared domain size. In this work, we develop test a new approach, K-means clustering assisted empirical modeling, for efficiently estimating waterflooding performance multiple geological realizations. This method performs single-phase flow simulations large number realizations, uses select only few representatives, on which two-phase are implemented. The models then adopted describe relation between solutions using these representatives. Finally, all realizations can be predicted readily. applied both 2D 3D synthetic shown perform well P10, P50 P90 rates, as probability distributions illustrated by cumulative density functions. It able capture ensemble statistics Monte Carlo results with computational cost significantly reduced.