What We Learned from a Study of Re-fracturing in Barnett Shale: An Investigation of Completion/Fracturing, and Production of Re-fractured Wells

The success of the Barnett shale helped to revitalize the oil and gas industry in the United States, and sparked a wave of shale development programs. At the same time, it has caused practitioners to rethink how, or even if, we can apply certain theories and technologies developed and implemented in conventional reservoirs to unconventional reservoirs. One fundamental difference in unconventional reservoirs is several orders of magnitude lower rock matrix permeability, largely requiring stimulation for economic development. Re-completion/fracturing in shale reservoirs is a means to increase production by accessing reserves beyond the volume achieved from initial completions. Re-fracturing may also be applied to wells with inefficient initial completions, less-thanoptimum designs or executions, and in depleted wells in which it is possible to treat previously un-stimulated rock volume. We performed this study to discover how production is impacted by recompletion and re-fracturing. By studying specific treatment and production data sets in the Barnett, we searched for new insights about how shale plays should be developed. In this work, we extracted about 200 re-fractured or re-completed wells with treatment information and production data in the Barnett Shale play from a service company’s database and a public well database. In this analysis, we took into account treatment data in both initial and refracturing, and evaluated their impact on long-term recovery. Further, we considered other available data in this investigation. In order to identify hidden, potentially predictive information in the data set, exploratory data analysis, univariate and multivariate/ cluster analysis techniques were applied using statistical analysis software such as SAS, combined with Geographic Information System (GIS) software. The paper is organized mainly in two parts. The first part is a data mining study on production and treatment data to discover certain treatment parameters controlling production for both initial completion and refracturing; in the second part, we interpret those observation and findings from perspective of mechanism.