The selection of candidate wells for refracturing is often very diffi- cult to peform based on the information available at the surface. We propose a systematic methodology to allow a field engineer to eval- uate a well’s potential for refracturing from an analysis of field-pro- duction data and other reservoir data commonly available. The well-selection method was successfully confronted to a case study in the Wattenberg field using data from 300 Codell tight gas wells.

The performance of refracturing treatments has been observed to be highly variable in the Wattenberg field (Colorado), with some wells underperforming while others are restored to initial or even higher production rates. Historically, multiple approaches have been taken to select the best candidate wells, including heu- ristic guidelines, field correlations, and neural networks.

After identifying the physical phenomena that are thought to impact the performance of refracturing operations, five dimen- sionless groups were developed to quantify them. By choosing a dimensionless approach, the goal is to establish refracturing crite- ria that will not limited to one specific field, but may used in dis- tinct oil and gas fields. One potential motivation for refracturing is the stress reorientation occurring around a fractured well, caus- ing the refracture to propagate orthogonally to the initial fracture in underdepleted sections of the reservoir. Numerical simulations of the areal extent of the stress-reversal region as well as tiltmeter measurements confirmed the existence of refracture reorientation in the Codell formation. Guidelines for the selection of refractur- ing candidates were expressed in terms of the potential for stress reorientation, the quality of the initial completion, the initial pro- duction decline, and the reservoir depletion around the well of interest. Two groups of wells showed the most promise for refrac- turing: (a) ineffective initial completions with a small initial pro- duction decline and (b) long initial fractures in underdepleted reservoirs. The dimensionless groups help us identify such wells and provide quantitative criteria for selection of wells that may be good candidates for refracturing.