DATA-DRIVEN ANALYSES OF LOW SALINITY WATERFLOODING IN CARBONATES

Abstract

Maximizing crude oil recovery is a main objective of the oil and gas industry. Oil recovery by natural production in carbonates is usually lower than 30%. Thus, Enhanced Oil Recovery (EOR) methods are used to increase the oil production in carbonate reservoirs. Low salinity water (LSW) injection is a promising EOR technique, which have been studied by many researchers for potential improvement of oil recovery. LSW flooding in carbonates has been widely evaluated by coreflooding tests in prior studies. A closer look in the literature on LSW in carbonates indicates a number of gaps and shortcomings. It is difficult to understand the exact relationship between different controlling parameters and the LSW effect in carbonates. The active mechanisms involved in oil recovery improvement are still uncertain, and more analyses are required. To predict the LSW performance and study the mechanisms of oil displacement, data collected from available experimental studies on LSW injection in carbonates were analyzed using data analysis approaches. In this thesis, I collected data from 26 secondary and 117 tertiary coreflooding tests. Machine learning (ML) and statistical approaches were utilized to analyze the extracted main parameters. We used a linear regression model to study the linear relationship between single parameters and incremental recovery factor (RF). Correlations between rock, oil, brine properties and tertiary RF were negligible and weak. Subsequently, we analyzed the effect of brine and oil/brine parameters (oil acidity, alteration in salinity and active ions concentration) on LSW performance using multivariable linear regression. Relatively stronger linear correlation was found for a combination of oil/brine parameters and RF. We also studied the nonlinear relationship between parameters by applying ML nonlinear models, such as Artificial Neural Networks (ANN), Support Vector Machine (SVM) and Decision Tree (DT). These models showed better data fitting results compared to linear regression. Strong and very strong relationships between properties and RF were achieved by ML models. Among the used ML models, DT provided the best correlation for oil/brine parameters, as ANN and SVM overfitted the testing data. Finally, different mechanisms involved in the LSW effect were analyzed based on the changes in the effluent PDIs concentration, interfacial tension, pH, zeta potential, pressure drop. Wettability alteration by LSW was commonly observed in coreflooding tests.