Sand production in oil and gas wells is a serious issue for the petroleum industry around the world. The commonly used non-intrusive sand monitoring systems are based on -acoustic emission measurement techniques. This research presents advanced data recognition techniques that can significantly improve the accuracy of sand monitoring. At the first step, factor analysis was used to identify key acoustic features of sand particles. Then, the following machine learning techniques have been applied: support vector machines, logistic regression, random forest method and gradient boosting. For training and testing the recognition system we used the acoustic database obtained in the laboratory of the oilfield service company SONOGRAM LLC (Kazan, Russia). The database consisted of acoustics signals from sand particles impacting on the inside and outside of a pipe wall in various scenarios (dry and wet gas, different flow rates, etc.). It was shown that the use of support vector machines with the Gaussian kernel reduces false positives compared with the algorithm that is based on ultrasound power peaks detection.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.