S. Goodarzi, A. Settari, M. Zoback, David William Keith
{"title":"A coupled geomechanical reservoir simulation analysis of carbon dioxide storage in a saline aquifer in the Ohio River Valley","authors":"S. Goodarzi, A. Settari, M. Zoback, David William Keith","doi":"10.1306/EG.04061111002","DOIUrl":null,"url":null,"abstract":"With almost 200 coal-burning power plants in the region, the Ohio River Valley is an important region to evaluate potential formations for carbon dioxide (CO2) storage. In this study, we consider whether injection-induced stress changes affect the viability of the Rose Run Sandstone, considered as a potential effective storage unit. Our study uses a coupled geomechanical and reservoir simulator that couples fluid flow to induced stress and strain in all the significant stratigraphic units from the surface to the crystalline basement. The pressure and stress variations were modeled during CO2 injection, focusing on injection from a single well. The model uses a constant pressure condition on the boundary of the system. Both reservoir and surface deformation were simulated, and the possibility of reaching shear failure in the reservoir was tested. Carbon dioxide injection in the Rose Run Sandstone aquifer is not likely to cause any significant surface deformation. To consider the potential of increasing injectivity, simulation of a static fracture with a half-length of 300 m (984.3 ft) was considered. As the modeling shows that, with constant injection rate, the fracture can propagate beyond the propped length, a dynamic fracture propagation was also studied. This was achieved by allowing the fracture to grow as a function of a propagation criteria based on effective stress. Because of the favorable stress state of the Rose Run Sandstone, the propagation is primarily in the lateral direction, and no upward fracture propagation through the cap rock has been observed in the model. Finally, we demonstrate that dynamic fracture propagation significantly increases the possible injection rates, and its modeling is useful for determining optimal injection rates.","PeriodicalId":11706,"journal":{"name":"Environmental Geosciences","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2011-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1306/EG.04061111002","citationCount":"16","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Environmental Geosciences","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1306/EG.04061111002","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"Earth and Planetary Sciences","Score":null,"Total":0}
引用次数: 16
Abstract
With almost 200 coal-burning power plants in the region, the Ohio River Valley is an important region to evaluate potential formations for carbon dioxide (CO2) storage. In this study, we consider whether injection-induced stress changes affect the viability of the Rose Run Sandstone, considered as a potential effective storage unit. Our study uses a coupled geomechanical and reservoir simulator that couples fluid flow to induced stress and strain in all the significant stratigraphic units from the surface to the crystalline basement. The pressure and stress variations were modeled during CO2 injection, focusing on injection from a single well. The model uses a constant pressure condition on the boundary of the system. Both reservoir and surface deformation were simulated, and the possibility of reaching shear failure in the reservoir was tested. Carbon dioxide injection in the Rose Run Sandstone aquifer is not likely to cause any significant surface deformation. To consider the potential of increasing injectivity, simulation of a static fracture with a half-length of 300 m (984.3 ft) was considered. As the modeling shows that, with constant injection rate, the fracture can propagate beyond the propped length, a dynamic fracture propagation was also studied. This was achieved by allowing the fracture to grow as a function of a propagation criteria based on effective stress. Because of the favorable stress state of the Rose Run Sandstone, the propagation is primarily in the lateral direction, and no upward fracture propagation through the cap rock has been observed in the model. Finally, we demonstrate that dynamic fracture propagation significantly increases the possible injection rates, and its modeling is useful for determining optimal injection rates.