|dc.description.abstract||Downhole temperature measured by distributed temperature sensors shows distinct response during multi-stage fracture treatments. A forward model is needed to interpret the measured dynamic temperature data during injection and shut-in of a well in complex flow systems to quantitatively diagnose fracturing treatments and characterize created fractures.
In this study, a forward thermal model predicting temperature distribution along a wellbore is established considering formation, wellbore, and fracture heat transfer. The fracture model can predict fracture propagation, fluid distribution, and fracture temperature based on mass and energy conservation equations. Temperature distribution in the reservoir can also be obtained by coupling a reservoir model with fracture model and wellbore model. For multi-stage fracture treatments, a sequential simulation method is applied by introducing real time control. Using the algorithm from single-stage treatment, a work flow for multi-stage fracture simulation is created by performing a single-stage stimulation, shutting in the stage, and moving along the wellbore to the next stage. Warm-up of the entrained fracturing fluid during shut-in periods is simulated by
removing the fluid injection term and implementing different boundary conditions.
Due to the large temperature difference between the injection fluid and the surrounding formation, simulated results show temperature signal change occurs at fracture locations during the injection period. Warm-back behavior is also obvious at fracture locations after shut-in the well. The effect of injection flowrate, fluid distribution, fluid properties, and reservoir characteristics on temperature behavior are investigated. At first initiation of the fracture, injection flowrate plays an important role on fracture half-length and leak-off front. Heat conduction is the dominant mechanism governing temperature response during shut-in. For a shale formation, the time to reach thermal equilibrium is on the order of weeks. Sensitivity of observed temperature to fluid distribution, and reservoir parameters in the simulation allow for fracture diagnosis using distributed temperature data during stimulation operations.||en