Full text loading...
-
Deterministic Smart Tools to Predict Recovery Factor Performance of Saline Water Injection in Carbonated Reservoirs
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 82nd EAGE Annual Conference & Exhibition, Oct 2021, Volume 2021, p.1 - 5
Abstract
Enhanced oil recovery (EOR) processes have been investigated in recent years due to scarcity of petroleum sources and weak performance of conventional waterflooding. One of the most practical method in EOR is low salinity water injection. While the experimental studies are costly and time-consuming, it is required to have a reliable and a cost-effective tool to predict the recovery performance of saline waterflooding, especially in carbonated reservoirs. In this paper, two deterministic tools including artificial neural network (ANN) and multigene genetic programming (MGGP) are developed for prediction of the recovery factor (RF) of saline waterflooding in carbonated reservoirs. For this purpose, 145 experimental data points of coreflooding tests were extracted from the literature. As well as, permeability, porosity, temperature, injection rate, total dissolved salinity (TDS), oil viscosity at experimental condition, and initial water saturation were selected as the input parameters. Besides, utilizing MGGP model to perform parametric sensitivity analysis. Results indicated that ANN has a great performance for predicting the RF. Coefficient of determination of ANN for training, testing and total data are 0.996, 0.9305, and 0.9804, respectively. Sensitivity analysis outcomes illustrate that permeability, porosity and TDS are the most effective parameters on the RF performance in carbonated reservoirs.