TY - CONF
T1 - Development of program coupling flow simulator with geomechanics simulator using adaptive mesh refinement model and analytical solution for simple one-dimensional rock deformation
AU - Sawata, Masaru
AU - Iwata, Shigeki
AU - Tenma, Norio
AU - Kurihara, Masanori
N1 - Funding Information:
This work was supported by funding from the Research Consortium for Methane Hydrate Resources in Japan (MH21 Research Consortium) planned by the Ministry of Economy, Trade, and Industry (METI), Japan.
Publisher Copyright:
© 2018 Society of Petrophysicists and Well-Log Analysts. All Rights Reserved.
PY - 2018
Y1 - 2018
N2 - In general, conventional oil and gas reservoirs are not deformed significantly. However, some unconventional reservoirs including those of methane hydrate and heavy oil/bitumen are soft and deformable. Therefore, it is important to introduce the geomechanical effects such as the changes in porosity/permeability associated with the deformation of a reservoir for accurately predicting production performances for these unconventional reservoirs. Recently, the reservoir fluid flow behavior has been simulated in conjunction with geomechanics simulation, using various coupling methods such as fully implicit method, explicit method, iterative method, etc. All of these methods have both advantages and disadvantages. For example, the fully implicit method, which solves flow and geomechanical behavior simultaneously, is accurate, but requires a huge computer time. In addition, it is difficult to utilize existing flow simulators and geomechanics simulators in this method. Meanwhile, the explicit method does not require much computational time, but is not accurate since the geomechanical behavior is predicted without being reflected enough in flow simulation. The iterative method is accurate because geomechanical information and flow information is transferred to each other until the flow performances become consistent with geomechanical behavior. This method, however, requires considerable computer time. Therefore, we conducted this research seeking for a new method coupling fluid flow with geomechanics that enables the accurate prediction in a reasonably short computer time. In this research, after developing the 3-D (dimensional), 3-phase black oil type simulator and the elastic type geomechanics simulator, the programs combining these two simulators by various methods were developed. To shorten the computational time, the geomechanics simulator developed in this research has the function of computing stress-strain field using Adaptive Mesh Refinement (AMR) model. In addition, the program that roughly calculates the formation deformation applying the analytical solution for simple 1-D deformation was coded. Furthermore, the flow simulator solving one-dimensional deformation equation implicitly with flow equations was constructed. Case studies using these simulators/programs revealed that: i) the solutions given by the explicit method were unreliable, ii) the application of the analytical solution of 1-D deformation was very fast, but might not be accurate in the complex 3-D case, iii) the iterative method could provide accurate solutions, but was slow, and iv) the best method was to couple the flow simulator with the geomechanics simulator equipped with AMR model, which could reduce the computational time to less than a half of the conventional iterative methods without reducing the accuracy of prediction.
AB - In general, conventional oil and gas reservoirs are not deformed significantly. However, some unconventional reservoirs including those of methane hydrate and heavy oil/bitumen are soft and deformable. Therefore, it is important to introduce the geomechanical effects such as the changes in porosity/permeability associated with the deformation of a reservoir for accurately predicting production performances for these unconventional reservoirs. Recently, the reservoir fluid flow behavior has been simulated in conjunction with geomechanics simulation, using various coupling methods such as fully implicit method, explicit method, iterative method, etc. All of these methods have both advantages and disadvantages. For example, the fully implicit method, which solves flow and geomechanical behavior simultaneously, is accurate, but requires a huge computer time. In addition, it is difficult to utilize existing flow simulators and geomechanics simulators in this method. Meanwhile, the explicit method does not require much computational time, but is not accurate since the geomechanical behavior is predicted without being reflected enough in flow simulation. The iterative method is accurate because geomechanical information and flow information is transferred to each other until the flow performances become consistent with geomechanical behavior. This method, however, requires considerable computer time. Therefore, we conducted this research seeking for a new method coupling fluid flow with geomechanics that enables the accurate prediction in a reasonably short computer time. In this research, after developing the 3-D (dimensional), 3-phase black oil type simulator and the elastic type geomechanics simulator, the programs combining these two simulators by various methods were developed. To shorten the computational time, the geomechanics simulator developed in this research has the function of computing stress-strain field using Adaptive Mesh Refinement (AMR) model. In addition, the program that roughly calculates the formation deformation applying the analytical solution for simple 1-D deformation was coded. Furthermore, the flow simulator solving one-dimensional deformation equation implicitly with flow equations was constructed. Case studies using these simulators/programs revealed that: i) the solutions given by the explicit method were unreliable, ii) the application of the analytical solution of 1-D deformation was very fast, but might not be accurate in the complex 3-D case, iii) the iterative method could provide accurate solutions, but was slow, and iv) the best method was to couple the flow simulator with the geomechanics simulator equipped with AMR model, which could reduce the computational time to less than a half of the conventional iterative methods without reducing the accuracy of prediction.
UR - http://www.scopus.com/inward/record.url?scp=85059416821&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85059416821&partnerID=8YFLogxK
M3 - Paper
AN - SCOPUS:85059416821
T2 - 24th Formation Evaluation Symposium of Japan
Y2 - 11 October 2018 through 12 October 2018
ER -