Fast estimation of supercritical CO2 thermal conductivity by a supervised learning machine: Implications for EOR

A. Rostami, M. Arabloo, E. Joonaki*, S. Ghanaatian, A. Hassanpour Youzband

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Selection of CO2 as an injection gas for enhanced oil recovery (EOR), is more desired as it could be employed for the aims of CO2 sequestration resulting in the greenhouse gas reduction as well. Thermal conductivity of supercritical CO2 is one of the most important thermophysical properties in designing CO2 injection into the reservoir. In detail, thermal conductivity of supercritical CO2 can highly affect the transport and fluid properties of supercritical CO2 like I FT, density, viscosity, solubility in crude oil and oil swelling factor through temperature variations in both transportation lines and reservoir systems. In this study, a dominant soft computation model called, least square support vector machine optimized by coupled simulating annealing (LSSVM-CSA), was developed and validated as a powerful tool for accurate estimation of supercritical CO2 thermal conductivity. This model is based on a function of temperature and pressure. In continuum, by the several statistical parameters and visual tools the superiority of the suggested model was verified in comparison with other commonly applied correlations in the literature. As a result of this study, the proposed tool can be of great significance for engineers in designing any CO2 process for enhanced oil recovery and CO2 sequestration.

Original languageEnglish
Title of host publication79th EAGE Conference and Exhibition 2017
PublisherEuropean Association of Geoscientists and Engineers, EAGE
ISBN (Electronic)9789462822177
Publication statusPublished - 1 Jan 2017
Event79th EAGE Conference and Exhibition 2017: Energy, Technology, Sustainability - Time to Open a New Chapter - Paris, France
Duration: 12 Jun 201715 Jun 2017

Conference

Conference79th EAGE Conference and Exhibition 2017: Energy, Technology, Sustainability - Time to Open a New Chapter
Country/TerritoryFrance
CityParis
Period12/06/1715/06/17

Fingerprint

Dive into the research topics of 'Fast estimation of supercritical CO2 thermal conductivity by a supervised learning machine: Implications for EOR'. Together they form a unique fingerprint.

Cite this