A New Methodology for Data Assimilation Using Capacitance-Resistance Model Combined with Fractional Flow and Aquifer Modeling

Summary Data assimilation for uncertainty reduction (DAUR) using reservoir simulators demands high computational time and resources. Depending on the reservoir model, the process can take days or weeks. Therefore, it is desirable to speed up the process to increase its efficiency, keeping the quality of the result. Our aim in this paper is to present a new methodology for data assimilation (DA) using the capacitance-resistance model (CRM), integrated with fractional flow (FF) models, to reduce the computational time and resources in the process. The methodology brings novel contributions, such as (a) applying the CRM in a probabilistic manner for DA, (b) developing a new FF approach, and (c) proposing a new approach for aquifer modeling under uncertainty. The methodology was successfully applied in a real field case. To validate the CRM results, we compared the DA process using the CRM with the DA results using the reservoir simulator. This comparison showed a very good agreement between the results. Notably, the CRM approach was up to 279 times faster than the process using the reservoir simulator. We also validated the CRM results using several kinds of multidisciplinary geoengineering data from the same studied case, including (1) historical injection, (2) hydraulic communication analysis based on static well pressure, (3) water salinity measurements analysis, and (4) 4D seismic analysis. These analyses showed the consistency of the CRM results in terms of physical representativeness. After an extensive validation process, we can state that the CRM approach, combined with the novel FF proposed in this work, has great potential to be applied in DA, reservoir management, and production strategy optimization, thus contributing to the acceleration of the decision-making process.