Enhanced oil recovery using smart water combined with sodium dodecyl sulfate and cetyltrimethylammonium bromide surfactants: A data-driven artificial neural network framework
Engineering Applications of Artificial Intelligence, cilt.181, 2026 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 181
- Basım Tarihi: 2026
- Doi Numarası: 10.1016/j.engappai.2026.115441
- Dergi Adı: Engineering Applications of Artificial Intelligence
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Applied Science & Technology Source, Compendex, INSPEC, Academic Search Ultimate (EBSCO), Engineering Source (EBSCO)
- Anahtar Kelimeler: Artificial neural network, Contact angle, Enhanced oil recovery, Surfactant flooding, Wettability alteration
- İstanbul Gelişim Üniversitesi Adresli: Evet
Özet
A data-driven artificial neural network framework is developed to quantitatively characterize the nonlinear interfacial and displacement behavior of mixed sodium dodecyl sulfate and cetyltrimethylammonium bromide surfactant systems within the investigated concentration domain. The concentrations of sodium dodecyl sulfate and cetyltrimethylammonium bromide are treated as independent compositional variables, while contact angle and oil recovery are treated as dependent responses representing wettability alteration and macroscopic displacement efficiency, respectively. Model generalization capability is examined using a ten-fold cross-validation strategy. Following optimization, the overall mean square error reached 0.3140 for contact angle and 0.1567 for oil recovery, while training, validation, and testing errors remained within the same numerical order of magnitude. Regression analysis demonstrated strong statistical concordance, with correlation coefficients exceeding 0.995 for contact angle and remaining above 0.98 for oil recovery across all subsets. Independent test evaluation revealed low relative prediction deviations within the investigated dataset, ranging from 0.0030 to 0.6875 percent for contact angle and from 0.0958 to 0.8667 percent for oil recovery. The combined stability of cross-validation results, regression consistency, bounded predictive errors, and structured sensitivity trends indicated that the proposed framework provided a statistically robust representation of concentration-dependent surfactant concentration-response relationships associated with wettability and enhanced oil recovery performance.