One major challenge in the development of an unconventional asset is determining optimal perforation cluster spacing during completion of a well. The spacing between perforation clusters influences the geometry of hydraulic fractures, drainage between fractures, drainage radius from the wellbore, production rates, and ultimate recovery of a well. Optimization of cluster spacing is not only concerned with maximizing well productivity but also accounting for the additional cost of various perforation cluster configurations and determining the most economic method to complete a well (Ingram et al. 2014). To add to the challenge, most operators have shifted from drilling single wells to drilling and completing multiple wells simultaneously, often referred to as infill development, where wells are spaced tightly together and compete for production.
The most straight-forward and conclusive method for optimizing cluster spacing typically involves field trials, where a statistically significant number of wellbores is selected to have a new perforation configuration. Production from these wells is then monitored for an extended period of time and compared to a baseline set of wells to determine whether the new completion method was effective. While this method can drive conclusive results, it is both time and capital intensive.
This paper explores a different technique to evaluate cluster spacing that can be used in conjunction with field trials to accelerate the learning curve of an asset. This method uses a single highly instrumented wellbore to build a calibrated fracture model and production-history-matched reservoir model. Sensitivities are then run through the reservoir model on cluster spacing to determine the resulting production of each scenario. Finally, an economic analysis is performed to account for the cost of each case and determine which scenarios are the most cost effective. The particular example in this study evaluates a well in the Eagle Ford shale drilled between two other wellbores spaced 330 ft apart. Both 40- and 20-ft cluster spacing are evaluated to identify which cluster spacing is preferred.
