Proxy modelling for rapid optimisation of miscible CO<SUB align="right">2-WAG injection in heterogeneous clastic reservoirs: a case study from Southern Iraqi oil fields

Abstract

Multiple simulation runs are required to optimise and quantify uncertainty and sensitivity analysis to determine and rank the impact of reservoir and operating parameters on the incremental oil recovery (ΔRF) under miscible CO2-WAG flooding. The central composite design (CCD) was employed to create a statistical proxy model for ΔRF as a function of seven reservoir and operating parameters. Statistical analysis was employed to check the model adequacy. The importance and effect of variables that result to higher predicted ΔRF were investigated. As an analogue of the Nahr Umr reservoir, the reservoir understudy, the proxy model can be roughly applied to predict the ΔRF during the miscible CO2-WAG injection at depending on uncertainty. The reliability of the proxy model has been demonstrated based on many statistical tools. [Received: 29 July 2019; Accepted: 24 November 2019]