An Integrated Approach to Reduce Water Production in the San Andres Formation

摘要

San Andres formation is a dolomitized carbonate and an oil-and gas-bearing member of the Upper Permian in the Permian Basin. It has been produced with vertical wells that were either waterflooded or CO2 injected for primary, secondary, and tertiary recovery. Some of the early waterfloods date back to the 1960s. Historically, the San Andres formation is associated with high water cuts. This paper proposes an integrated solution from planning to completion of a horizontal well, with an aim to control water production. The workflow involves the integration of various technologies such as crosswell seismic, petrophysical logs, geomechanical logs, fracture modeling, and real-time microseismic monitoring. Stratigraphic surfaces were created using the crosswell seismic data tied in with well tops from offset vertical wells. The 3D model was populated with high-tier petrophysical and geomechanical properties from the pilot well. The zones with high water saturations were identified in the subsurface, which gave an insight into the source of the produced water. Fully 3D planar fracture modeling was performed using the log data from the vertical offset well at the landing point of the horizontal well. Sensitivities were performed on various fluid systems, job volumes, and pump rates, and a playbook was created for real-time operations as part of a contingency plan. Lateral measurements included petrophysical and geomechanical data, which were used to place perforations, within a stage, in similar type of rock. Depending on the real-time microseismic events, the stimulation design was changed on the fly. Stages with events that were growing out of the zone were modified first, and a final pump schedule was established after the first five stages. The production results indicated a 20% decrease in water cut, which is a notable improvement compared to the historical production data in the San Andres. The workflow proved that the water production can be significantly reduced by applying a methodology that includes integration of data from multiple domains, thereby improving the economics of a well.