Tight gas stimulations in the Cooper Basin have been challenged by strike–slip to reverse stress regimes, adversely affecting the hydraulic fracturing treatment. These stress conditions increase borehole breakout and affect log and cement quality, create more tortuous pathways and near-wellbore pressure loss, and reduce fracture containment. These factors result in stimulation of lower permeability, low modulus intervals (e.g. carbonaceous shales and interbedded coals) versus targeted tight gas sands. In the Windorah Trough of the Cooper Basin, several steps have been employed in an ongoing experiment to improve hydraulic fracturing results. First, the wellbore was deviated in the maximum horizontal stress direction and perforations shot 0 to 180° phased to better align the resulting hydraulic fractures. Next, existing drilling and logging-while-drilling data were used to train a machine learning model to improve reservoir characterisation in sections with missing or poor log data. Finally, diagnostic fracture injection tests in non-pay and pay sections were targeted to specifically inform the machine learning model and better constrain permeability and stress profiles. It is envisaged that the improved well and perforation alignment and better targeting of intervals for the fracturing treatment will result in lowered tortuosity, better fracture containment, and higher concentrations of localised proppant, thereby improving conductivity and targeting of desired intervals. The authors report the process and results of their experimentation, and the results relative to the offsetting vertical well where a typical five-stage treatment was employed.
