Wellbore Integrity Challenges in Saltwater Disposal: Case Study on the Fort Worth Basin

Abstract

Abstract This paper examines wellbore integrity in saltwater disposal (SWD) operations via modeling the stress-distribution evolutions within a well's casing-cement sheath-rock formation (C/CS/RF) system. Injection from SWD wells is thought to be responsible for elevated seismicity in regions of high oil-and-gas activity, such as the Fort Worth Basin in northeastern Texas. This modeling approach is useful for practitioners to take the appropriate preventative measures in the form of hazard-mitigation guidelines. Induced (or "triggered") seismicity phenomena depend on the fluid-injection parameters, as well as the reservoir's petrophysical and geomechanical properties, including the presence of naturally-occurring pre-existing faults (PEFs) and their frictional behavior. The increase of the pore pressure within the FWB region, following water injection from SWD wells in 12 counties was obtained in the literature and used as inputs assessing the integrity of the C/CS/RF system against a total of ten modes of mechanical degradation. These degradation modes are induced by collapse/burst and compressive/tensile stress loads within the casing layer, inner or outer debonding, radial cracking, shear cracking, and disking within the CS layer, and tensile fracture initiation in the adjacent-RF. The potential impacts of this wellbore-integrity compromise in SWD operations are discussed, as some of the aforementioned mechanical degradation modes are likely to adversely impact wellbore integrity a-priori to the occurrence of induced seismicity. Tendencies towards radial cracking and disking are identified to occur a-priori to any induced seismicity occurrences. This mechanical degradation means can facilitate for saltwater leakages towards neighboring zones in the subsurface and influence injectivity negatively. Such saltwater leakages can have a stabilizing effect that may suppress induced seismicity. Three-dimensional (3D) hydromechanical modeling from past studies attempted to draw spatial and temporal correlations between seismic-activity observations and areas of pore-pressure elevation. Several localized regions of predicted-pore-pressure buildup did not indicate strong correlations with the recorded earthquake occurrences.