A new method to forecast reservoir performance during immiscible and miscible gas-flooding via transfer functions approach

Abstract

<jats:title>Abstract</jats:title><jats:p>In order to improve oil recovery during gas flooding, it is crucial to afford an accurate estimation of future performance. So far, different approaches have been developed to present forecasting. The most common reservoir simulator, grid-based, which has the highest accuracy, suffers from some weaknesses; time-consuming computation and also need for large quantity of data. Sometimes, a quick overview of reservoir performance is adequate or all required data are not accessible. Therefore, in this study a fast simulator is introduced to provide a quick overview with the minimum amount of data.</jats:p><jats:p>In this study, the development of a method based on transfer functions (TF) is presented to model immiscible and miscible gas flooding. TF is a mathematical representation in Laplace domain which demonstrates the relation between the input and output signals of a system. In the proposed method, reservoirs are modeled with a combination of TFs. The order and arrangement of the TFs are chosen based on the physical conditions of the reservoir which are ascertained by checking several cases. Injection and production mass rates act as input and output signals respectively. TF parameters are calculated using history matching. Also, a fractional flow model is introduced and coupled with the TF system to obtain oil production rate as the final output.</jats:p><jats:p>In an attempt to validate the approach, it is required to compare the results with the grid-based method. Six different synthetic cases are constructed and used to validate the developed approach. The results state a good agreement with those obtained from the numerical simulators.</jats:p><jats:p>This approach is a quick way to predict gas flooding performance with minimum amount of data, production and injection rates data are the only requirements. It can be a new window for the future of fast simulators. It provides estimations with plausible certainty. On the other hand, the analytical solution of method enables its utilization in finding optimum rates for gas injection in a short period of time. The method also presents some key parameters such as well connectivity. It is fair to state that the use of the model is limited to situations when a rapid estimation is looked for and/or adequate data is not accessible.</jats:p>