Optimizing Waterflood Management in a Giant UAE Carbonate Oil Field Using Simulation-Based Streamlines

Abstract

AbstractStreamlines have been acknowledged as a powerful tool for modeling and optimizing waterfloods in oil reservoirs. Streamlines provide visualization of reservoir fluid flow and a unique way to conceptualize and quantify injector-producer well connectivity. Well allocation factors (WAFs) between injectors and producers as well as injection efficiency can be calculated and used to optimize water injection and reduce injection water cycling while focusing on maximizing oil recovery. The objective of this study was to develop a methodology, tool, and workflow for optimizing pattern waterflood management and to demonstrate application in a giant offshore carbonate reservoir. This methodology utilizes history-matched simulation models to generate streamlines from the pressure field and fluxes computed by a finite-difference reservoir simulator, EMpower. A finite-difference simulator is more rigorous than streamline simulators in terms of its ability to model complex fluid flow, and therefore, provides a more realistic solution. The streamlines thus obtained take into account the detailed geology, well locations, phase behavior, and flow behavior of the history-matched models. This methodology utilizes the concept of pair "injection efficiency" (IE) and pair "voidage replacement ratio" (VRR) which was necessary for the subject giant oil field and is equally applicable to other fields.The optimization process is started by generating streamlines and associated WAFs at the current time using the simulation model. The injection efficiency is calculated and injector-producer well pairs are identified with high and low IE. New producer rates are then calculated by increasing producer rates in high IE well pairs and decreasing rates in low IE well pairs based on a weight factor. The injector rates are then calculated by applying a target VRR to each individual well pair. This results in redistribution of injection and production rates in a more balanced/optimal manner. The process is repeated at regular intervals (e.g., yearly or quarterly) while simulation is run in the prediction mode. Results show higher total oil production rates or lower water production and water injection rates. The WOR versus cumulative oil production clearly demonstrates that the overall injection efficiency increases as the optimization process is progressed.