(159al) Calibrating Reservoir Simulation Model Responses using Time-Lapse VSP Data from Farnsworth Field unit, Texas

CO₂-WAG is a popular tertiary hydrocarbon recovery method for exploiting more residual oil after primary and secondary recovery processes by altering the fluid thermodynamic properties, thereby increasing the mobility of the remaining oil. In the event of recovering the residual oil, CO₂ is sequestered in the subsurface reservoir contributing to reduction in greenhouse gas emissions. Time-lapse Vertical Seismic Profile has served as a reservoir monitoring tool. Data obtained from Time-lapse VSP seismic can be used to map the dynamic changes in fluid saturations occurring within a reservoir. The effect of the fluid substitution process within the injection interval is identified by the use of a seismic attribute. In order to monitor the movement of the injected CO₂, 3D vertical seismic profiles are necessary – a baseline and monitor surveys at different injection stages. Time-lapse responses at future timesteps are predicted by integrating data acquired from the measured VSP with the reservoir simulation model. This comparison is made possible by computing seismic attributes from the simulation model and the observed seismic data. A mismatch function is developed either in the amplitude, acoustic impedance, time or depth shift domain. In order to evaluate the feasibility of time-lapse VSP as a reservoir surveillance tool, the Southwest Regional Partnership acquired several time-lapse 3D vertical seismic data from the Farnsworth field located at Ochiltree county, northwest Texas. The movement of the CO₂ was monitored by looking at the differences in the elastic properties inferred from the time-lapse 3D depth shifts.

The objective of this research is to present the calibration of time-lapse depth shifts computed from the Farnsworth unit reservoir simulation model by using corresponding data from the measured time-lapse vertical seismic profile. The current status of the reservoir fluid and rock physics studies are presented in this work. In this study, we present a systematic approach for computing the elastic response of the bulk saturated rock for different fluid saturations. The effective elastic properties of the rock fluid system are evaluated through a combination of the fluid thermophysical properties, rock frame elastic properties and the petrophysical properties of the reservoir rock. These rock physics studies are based on prior analysis of the geophysical logs and core data obtained from science wells in the study area. The results from the rock physics models together with data from the fluid compositional model were used in an elastic forward modeling procedure to create computed time-lapse responses. The 3D depth shifts from the predicted and measured time-lapse seismic data were compared. The spatial differences between the predicted and measured seismic data were minimized by using a Genetic optimization algorithm. Model parameters sensitive to the optimization process were identified and varied within their acceptable tolerance. The relative permeability curves, reservoir permeability and rock elastic parameters were among the sensitive variables.