Doctoral Dissertations


"Wettability alteration with surfactants has been widely applied to improve oil recovery in carbonate reservoirs. An effective surfactant huff-puff application design requires comprehensive guidelines about where, how, and when this method could be applied. Besides, traditional methods to design an effective application and optimize surfactant performance are dependent on extensive experiments and are time-consuming. This study is to assist in the design of surfactant huff-puff via statistical methods and machine learning techniques.

In this study, a dataset including 402 effective surfactant imbibition tests is established by collecting information from nearly 50 publications. The dataset provides a foundation for further data analysis. Descriptive statistical analysis methods are used to establish comprehensive application guidelines for surfactant huff-puff treatments. A Random Forest (RF) model was built to predict surfactant performance. Based on RF model, Shapley additive explanations approach was applied to obtain new insights to promote the understanding of wettability alteration with surfactants. Also, a new procedure integrating RF model with Powell’s method was built to optimize surfactant performance. Besides, principal component analysis and hierarchical clustering algorithm was applied to uncover the hidden pattens embedded in global surfactant huff-puff treatments. The analog reasoning was utilized to provide valuable experiences for designing surfactant treatments and predicting performance for new candidate projects. Finally, an adaptive, offline, and friendly desktop application was developed to facilitate data analysis and support decision making"--Abstract, p. iv


Wei, Mingzhen

Committee Member(s)

Bai, Baojun
Flori, Ralph E.
Dunn-Norman, Shari
Siau, Keng, 1964-


Geosciences and Geological and Petroleum Engineering

Degree Name

Ph. D. in Petroleum Engineering


Missouri University of Science and Technology

Publication Date

Summer 2023


xix, 249 pages

Note about bibliography

Includes_bibliographical_references_(pages 247-248)


© 2023 Ya Yao, All Rights Reserved

Document Type

Dissertation - Open Access

File Type




Thesis Number

T 12304

Electronic OCLC #