Nazim Hasan, Shadma Tasneem, Othman Hakami, Waleed M. Alamier, Marjan Goodarzi
{"title":"Development of machine learning models for predicting thermophysical properties of VR/VGO nanofluids applicable in enhanced oil recovery","authors":"Nazim Hasan, Shadma Tasneem, Othman Hakami, Waleed M. Alamier, Marjan Goodarzi","doi":"10.1007/s10973-024-13857-3","DOIUrl":null,"url":null,"abstract":"<div><p>Previous studies indicate that nanotechnology can significantly enhance the effectiveness of enhanced oil recovery (EOR) techniques, particularly hot fluid injection. The efficiency of the injection process can be significantly influenced by the nanofluid's thermophysical properties (TPPs). However, laboratory examinations of TPPs on the performance of nanofluid-based EOR techniques are time-consuming and expensive, and theoretical models can also be inaccurate. To address this challenge, machine learning (ML) models can efficiently predict nanofluid TPPs and their impact on oil recovery performance. This study uses six ML algorithms: the K-Nearest Neighbors (KNN), the Theil-Sen regressor, the decision tree (DT), the lightGBM, the Bayesian ridge, and the LASSO. Using the VR-fluid thermal conductivity and temperature created the formula and the model for thermal conductivity of the VR-VGO nanofluid with 2 mass% CuAeg, and also by the VR-VGO nanofluid with 2 mass% CuAeg thermal conductivity and temperature created the formula and model for VR-fluid thermal conductivity. The maximum R-squared (<i>R</i><sup>2</sup>) for the VR-VGO nanofluid with 2 mass% CuAeg for the formula and the model attained by the Theil-Sen regressor and the KNN, which respectively are 0.997 and 0.999 in linear formation, and the best <i>R</i><sup>2</sup> for the VR-fluid for the formula and the model reached by the BR and the KNN respectively are 0.98 using the polynomial formation, and 0.995 using the linear form. The VR-fluid viscosity and temperature calculate the VR-VGO nanofluid with 2 mass% CuAeg viscosity and the VR-fluid viscosity are calculated using the VR-VGO nanofluid with 2 mass% CuAeg viscosity and temperature. The best <i>R</i><sup>2</sup> for the VR-VGO nanofluid with 2 mass% CuAeg reached by the Theil-Sen regressor for the formula, the KNN for the model, and the <i>R</i><sup>2</sup> score for the VR-VGO nanofluid with 2 mass% CuAeg viscosity respectively are 0.999 and 0.999 in linear form. The best <i>R</i><sup>2</sup> value for the VR-fluid for the formula by the BR is 0.999, and for the model by the KNN is 0.999 in linear form.</p></div>","PeriodicalId":678,"journal":{"name":"Journal of Thermal Analysis and Calorimetry","volume":"150 1","pages":"691 - 705"},"PeriodicalIF":3.0000,"publicationDate":"2024-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Thermal Analysis and Calorimetry","FirstCategoryId":"5","ListUrlMain":"https://link.springer.com/article/10.1007/s10973-024-13857-3","RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"CHEMISTRY, ANALYTICAL","Score":null,"Total":0}
引用次数: 0
Abstract
Previous studies indicate that nanotechnology can significantly enhance the effectiveness of enhanced oil recovery (EOR) techniques, particularly hot fluid injection. The efficiency of the injection process can be significantly influenced by the nanofluid's thermophysical properties (TPPs). However, laboratory examinations of TPPs on the performance of nanofluid-based EOR techniques are time-consuming and expensive, and theoretical models can also be inaccurate. To address this challenge, machine learning (ML) models can efficiently predict nanofluid TPPs and their impact on oil recovery performance. This study uses six ML algorithms: the K-Nearest Neighbors (KNN), the Theil-Sen regressor, the decision tree (DT), the lightGBM, the Bayesian ridge, and the LASSO. Using the VR-fluid thermal conductivity and temperature created the formula and the model for thermal conductivity of the VR-VGO nanofluid with 2 mass% CuAeg, and also by the VR-VGO nanofluid with 2 mass% CuAeg thermal conductivity and temperature created the formula and model for VR-fluid thermal conductivity. The maximum R-squared (R2) for the VR-VGO nanofluid with 2 mass% CuAeg for the formula and the model attained by the Theil-Sen regressor and the KNN, which respectively are 0.997 and 0.999 in linear formation, and the best R2 for the VR-fluid for the formula and the model reached by the BR and the KNN respectively are 0.98 using the polynomial formation, and 0.995 using the linear form. The VR-fluid viscosity and temperature calculate the VR-VGO nanofluid with 2 mass% CuAeg viscosity and the VR-fluid viscosity are calculated using the VR-VGO nanofluid with 2 mass% CuAeg viscosity and temperature. The best R2 for the VR-VGO nanofluid with 2 mass% CuAeg reached by the Theil-Sen regressor for the formula, the KNN for the model, and the R2 score for the VR-VGO nanofluid with 2 mass% CuAeg viscosity respectively are 0.999 and 0.999 in linear form. The best R2 value for the VR-fluid for the formula by the BR is 0.999, and for the model by the KNN is 0.999 in linear form.
期刊介绍:
Journal of Thermal Analysis and Calorimetry is a fully peer reviewed journal publishing high quality papers covering all aspects of thermal analysis, calorimetry, and experimental thermodynamics. The journal publishes regular and special issues in twelve issues every year. The following types of papers are published: Original Research Papers, Short Communications, Reviews, Modern Instruments, Events and Book reviews.
The subjects covered are: thermogravimetry, derivative thermogravimetry, differential thermal analysis, thermodilatometry, differential scanning calorimetry of all types, non-scanning calorimetry of all types, thermometry, evolved gas analysis, thermomechanical analysis, emanation thermal analysis, thermal conductivity, multiple techniques, and miscellaneous thermal methods (including the combination of the thermal method with various instrumental techniques), theory and instrumentation for thermal analysis and calorimetry.