Qualitative and quantitative determination of oil base mud using non-negative matrix factorizationQualitative and quantitative determination of oil base mud using non-negative matrix factorization Πλήρης Δημοσίευση σε Συνέδριο Conference Full Paper 2018-06-272016enThe aim of this study was to develop an off-line, fast, efficient and reliable analytical methodology, capable to determine, both qualitatively and quantitatively, the content of a common oil base mud filtrate in contaminated formation fluids samples. We demonstrate the performance of a methodology based on Gel-Permeation Chromatographic (GPC) analytical signals, combined with a modified factorization algorithm to estimate the oil fractions and an optimization procedure to calculate their concentrations within the mixtures. The proposed scheme is innovatively applied for the determination of the mud filtrate in contaminated oil samples and self-adapted to data without any former information. The results of this work indicate the efficiency of the algorithmic framework to reveal the content of a complex petroleum mixture analyzing the chromatograms of a limited number of samples. The presented technique has a great potential in upstream oil industry, enabling the un-mixing of analytical signals from mixtures of oil samples contaminated with specific petroleum fractions. http://creativecommons.org/licenses/by/4.0/362-367IEEE International Conference on Imaging Systems and TechniquesIST 2016 - 2016 IEEE International Conference on Imaging Systems and Techniques, Proceedings Livanos Georgios Λιβανος Γεωργιος Pasadakis Nikos Πασαδακης Νικος Zervakis Michail Ζερβακης Μιχαηλ Giakos George C. Institute of Electrical and Electronics Engineers Chromatography Mud filtrates Non-negative matrix factorization Signal decomposition