Integration of Capacitance Resistance Model with Reservoir Simulation

Abstract

Abstract Capacitance resistance models (CRM) provide an analytical framework to characterize and optimize water flooding fields. There are significant challenges in adapting these methods to actual field data with many complexities. In this research, we present a novel approach to integrate CRM models with reservoir simulation to improve reservoir characterization as well as identify new injection strategies to increase oil recovery. The results of this integration are presented for a specific field with challenging geological uncertainties. In particular, the historical production and injection data taken from water flooding field has been analyzed and separated as training and prediction part. First, inter-well connectivity (gain) values are estimated based on training part of the data which are further verified with streamline simulation results. A comparison of allocation factors between wells from the reservoir simulation model and the inter-well connectivity is presented to showcase consistency. Next, the prediction capability of the model has been verified with the testing section of data. Finally, for the next future 3 years, oil production rates obtained from CRM is further verified with results from reservoir simulation. The large number of wells and the uncertainties associated with a producing field makes CRM modeling a challenge especially for predictions. Accordingly, initial screening and analysis of field data is an essential part for successful CRM application. After examining field data with 5 injectors and 21 producers which is subject of this study, it has been observed that only 6 wells are positively correlated with injectors which has been used as an input for CRM. Results further indicate that, with correct application of CRM, for the selected wells future oil recovery can be predicted with less than 2% differences compared to simulation results. This shows the capability and practicality of CRM. The model is next used to obtain new injection schedule to increase oil recovery. The new injection rates obtained from the model predict an increase in oil recovery by 5-10%. The continuous recording of production and injection data presents a new opportunity to integrate analytical approaches with traditional reservoir simulation approach for effective reservoir management. The presented model that includes verification of CRM results (with streamlines as well as reservoir simulation) helps benchmark these techniques that are simple and effective. The presented approach is particularly applicable for mature fields where historical production and injection data is available so that data-driven models (CRM) can be integrated with traditional reservoir simulation models.