Mohammed Alkatheri, Farzad Hourfar, Ladan Khoshnevisan, Hedia Fgaier, Ali Almansoori, Ali Elkamel
{"title":"A machine learning approach for modelling and optimization of complex systems: Application to condensate stabilizer plants","authors":"Mohammed Alkatheri, Farzad Hourfar, Ladan Khoshnevisan, Hedia Fgaier, Ali Almansoori, Ali Elkamel","doi":"10.1002/cjce.25180","DOIUrl":null,"url":null,"abstract":"<p>Recent advancements in supervised machine learning tools have demonstrated their ability to achieve accurate and efficient prediction results. In this paper, we leverage these tools as alternative approaches to model a specific application in the gas industry, based on operating data. The chosen application is a natural gas condensate stabilization process, in which light end components are removed to reduce condensate vapour pressure, meeting storage and transportation specifications. Here, we develop and evaluate various supervised machine learning models to predict the performance of two industrial condensate stabilizer units. By utilizing large datasets from these units and encompassing comprehensive operating data of input–output variables, we not only demonstrate the capability of these techniques to offer reliable and accurate predictions but also shed light on their potential impacts and implementations. The impacts of applying selected AI and machine learning algorithms are two-fold. First, our research presents an innovative approach to process modelling and optimization in the gas industry, showing the potential for enhanced operational efficiency, profitability, and safety. Second, we propose a data-driven surrogate-based optimization framework, where the generated machine learning models can replace detailed first-principle models, offering a streamlined method to find optimal values for variables to reduce operational energy consumption. Furthermore, we address the validation requirements of our machine learning models, ensuring their robustness and reliability in real-world applications. By incorporating rigorous validation procedures, we guarantee the quality of our predictions and support their practical implementation. In conclusion, our research not only highlights the capabilities of machine learning in gas industry applications but also emphasizes their potential impacts and contributions to operational excellence. So, the presented approach can pave the way for improved performance, efficiency, and profitability in the gas industry.</p>","PeriodicalId":9400,"journal":{"name":"Canadian Journal of Chemical Engineering","volume":"103 3","pages":"1183-1212"},"PeriodicalIF":1.6000,"publicationDate":"2024-01-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Canadian Journal of Chemical Engineering","FirstCategoryId":"5","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/cjce.25180","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"ENGINEERING, CHEMICAL","Score":null,"Total":0}
引用次数: 0
Abstract
Recent advancements in supervised machine learning tools have demonstrated their ability to achieve accurate and efficient prediction results. In this paper, we leverage these tools as alternative approaches to model a specific application in the gas industry, based on operating data. The chosen application is a natural gas condensate stabilization process, in which light end components are removed to reduce condensate vapour pressure, meeting storage and transportation specifications. Here, we develop and evaluate various supervised machine learning models to predict the performance of two industrial condensate stabilizer units. By utilizing large datasets from these units and encompassing comprehensive operating data of input–output variables, we not only demonstrate the capability of these techniques to offer reliable and accurate predictions but also shed light on their potential impacts and implementations. The impacts of applying selected AI and machine learning algorithms are two-fold. First, our research presents an innovative approach to process modelling and optimization in the gas industry, showing the potential for enhanced operational efficiency, profitability, and safety. Second, we propose a data-driven surrogate-based optimization framework, where the generated machine learning models can replace detailed first-principle models, offering a streamlined method to find optimal values for variables to reduce operational energy consumption. Furthermore, we address the validation requirements of our machine learning models, ensuring their robustness and reliability in real-world applications. By incorporating rigorous validation procedures, we guarantee the quality of our predictions and support their practical implementation. In conclusion, our research not only highlights the capabilities of machine learning in gas industry applications but also emphasizes their potential impacts and contributions to operational excellence. So, the presented approach can pave the way for improved performance, efficiency, and profitability in the gas industry.
期刊介绍:
The Canadian Journal of Chemical Engineering (CJChE) publishes original research articles, new theoretical interpretation or experimental findings and critical reviews in the science or industrial practice of chemical and biochemical processes. Preference is given to papers having a clearly indicated scope and applicability in any of the following areas: Fluid mechanics, heat and mass transfer, multiphase flows, separations processes, thermodynamics, process systems engineering, reactors and reaction kinetics, catalysis, interfacial phenomena, electrochemical phenomena, bioengineering, minerals processing and natural products and environmental and energy engineering. Papers that merely describe or present a conventional or routine analysis of existing processes will not be considered.