Identification and characterization of thermally induced fractures using modern analytical techniques

Thermally Induced Fractures (TIFs) are common in waterflooded fields where the injection of (relatively) cold water leads to a cooling front that reduces the formation's in-situ stresses. TIF initiation and growth can thus occur when the previously measured, fracture propagation pressure is not exceeded. The presence of TIFs severely distorts the flood front, decreasing sweep efficiency. This paper applies several published, waterflood analysis methods to TIF characterization in synthetic and real case studies. These methods include well injection history data using modified Hall analysis, injectivity index and pressure vs. rate analysis, etc. to identify the onset of TIF, its propagation and its properties. The Capacitance Resistance Model (CRM) is chosen to quantify the change in inter-well communication due to TIF, the likely TIF direction and its impact on the flood front. Application of these models is illustrated and verified using reservoir geomechanical/fluid flow simulations. In addition, a field case study is used to demonstrate the value of integrating analytical and semianalytical models when developing an improved understanding of TIF evolution in waterflood operations Results from synthetic examples using the proposed analytical models confirm that the models correctly identify the onset, monitor the evolution of TIF, and provided valuable insights into the overall well injection performance. The CRM model quantified the connectivity between well pairs, identified TIF initiation and its subsequent impact on well connectivity and flood efficiency. The same techniques are used to analyse data from a real-field case history. TIF initiation and propagation events were clearly identified, even when their existence was much less certain with the other techniques. A correlation between the injection water temperature and the TIF growth dynamics was observed while the well production data and FMI images supported our analysis. The novelty of the presented approach is in efficiently integrating the recent, analytical and semianalytical models to identify onset, propagation, characteristics, and impact of TIF from the generally available, well injection production history data. The practical workflow employed here will help engineers detect and monitor TIFs; as well as evaluate the metrics describing waterflood performance, namely flood efficiency, inter-well communication and pressure maintenance.