In the Lance formation of the Jonah field, a neural network was utilized to determine best completion practices. In addition to other findings, the study indicated that reducing gel concentrations, reducing the use of 100-mesh sand, and reducing pad volume by 20% improved production. Although the study was not conclusive, these may indicate that increased fracture conductivity was beneficial.
The Lance formation in the Jonah field is an overpressured, tight-gas sand that requires hydraulic fracturing for economic production. Because of large gross intervals containing several individual sands, limited entry has been the typical fracturing technique. Wells can have more than 30 individual sands that are completed with multiple fracture treatments; however, production log data indicate that only 58% of the perforated sands contribute to production. Production optimization is dependent on improving the percentage of completed pay contributing to production.
A detailed field study was conducted to determine pay identification and best practices for completions. In the study, 44 wells in the Stud Horse Butte area were analyzed. The study comprised detailed log analysis, fracture treatment data, production data, reservoir analysis, and completion practices. Some of the processes used were log normalization, traditional statistical analysis, and reservoir modeling. Log and treatment data were also analyzed with an artificial neural network (ANN).
Author(s): Halliburton Energy Services, M.J. Eberhard, M.J. Mullen, C.A. Seal, Inc.; B.P. Ault, Ultra Petroleum
Paper Number: SPE 59790