To economically and efficiently develop unconventional resource plays, the industry has been spending tremendous resources to optimize completion and well spacing by piloting – a trial-and-error approach. However, the approach tends to take long time and cost significant amount of money. As the complex fracturing modeling technology advances, we question: "Can we use the latest complex fracturing modeling and reservoir simulation technologies to optimize completion and well spacing?", so that the industry can significantly save piloting time and money, and quickly find the optimal well spacing and corresponding optimal completion.
A recent case study in Permian Basin has answered the question well. For a Wolfcamp well completed with crosslinked gel and wide cluster spacing in 2012, we first built a 3-D geological and geomechanical model, and a full wellbore fracturing propagation model, and then calibrated it with multi-stage fracturing pumping history; the resulting complicated fracture network model was then converted into an unstructured grid-based reservoir simulation model, which was then calibrated with the well production history. During the process, discrete natural fracture network (DFN) and stress anisotropy were systematically evaluated to study their impact on fracture growth. Microseismic and tracer log data were used to validate the hydraulic fracturing modeling results. To test if the calibrated geomechanical and reservoir models can be used to optimize well completion design, we then ran the fracturing model with the latest completion design (tighter cluster spacing, slick-water, and more fluid and proppant) and forecasted the well performance. We found out that the resulting well performance is very similar to the performance of those wells with similar completion designs in the same area.
After establishing the confidence on the capacity of those models, we then further studied the impact of different completion designs on fracture dimensions and well performance. We examined the distributions of fracture length along the wellbore resulted from different cluster spacings, fracturing fluid types and volume, and proppant amount. We found out (1) the hydraulic fracture length and network complexity mainly depend on DFN and stress anisotropy, and fracturing fluid viscosity; and (2) the fracture length of those fractures initiated from different perforation clusters along wellbore is in a log-normal distribution depending on completion designs, which provides crucial insights to well interference and furthermore on well spacing. Therefore, we can reasonably model complicated fracture propagation and corresponding well performance with the latest modeling technologies, and then optimize well spacing, which should help operators save significant time and money on well completion and spacing piloting projects, and thus speed up field development decision.
The paper demonstrates our novel workflow as an effective way to optimize completion design and well spacing by integrating advanced multi-stage fracture modeling with reservoir simulation in unconventional resource plays.