To effectively drain hydrocarbon from unconventional plays, operators have been optimizing their drilling spacing unit (DSU) well spacing for many years. This paper presents a continuous improvement of Bakken well spacing using trial pads with smaller spacing and multidisciplinary data gathering to understand the most effective spacing between wells.
The operator's standard well spacing between Middle Bakken (MB) wells in the East Nesson (EN) area [Alger Field] ranges from 500 to 700 ft. To understand the optimum spacing between wells in this area, the operator trialed well spacing of 500 ft between like formations. The spacing between Middle Bakken and Three Forks wells was 250 ft. Data gathered during the spacing trial included microseismic, microseismic depletion delineation (MDD), radioactive (RA) tracers, chemical tracers, image logs, pressure measurements during completion/flowback/early-time production, and diagnostic fracture injection tests (DFITs). The data was used to calibrate advanced hydraulic fracture models and guide next-generation reservoir simulation history matching to characterize multiwell production behavior.
The MDD in the parent well provided data to "map" drainage patterns, showing that drainage was limited to the MB formation. However, microseismic showed that hydraulic fracture height extended from the Three Forks second bench (TF2) up through the Three Forks first bench (TF1), Middle Bakken, and into the overlying Lodgepole (LP)—connecting the entire Bakken petroleum system. The microseismic also showed asymmetric fracture growth toward the parent well in the DSU, but the asymmetry diminished as completions progressed away from the parent well. In addition to the MB-TF connectivity indicated from the microseismic, RA tracers pumped from a TF2 well were detected in a Middle Bakken well. The implied transport of proppant from a lower TF completion to the MB increases the likelihood of production communication between formations. Chemical tracers (oil and water) pumped during hydraulic fracturing operations were found from one end of the pad to another (over 2,500 ft) regardless of formation; another confirmation of hydraulic communication between Middle Bakken and Three Forks wells.
The hydraulic fracture model was calibrated using microseismic data and used for subsequent reservoir simulation history matching. The workflow consisted of modeling the parent well hydraulic fractures and history matching production, performing geomechanical modeling to determine the effects of parent well depletion on 3D stress state for hydraulic fracture modeling of the infill wells, and production history matching of the entire pad. The modeling showed significant fracture-to-fracture communication and well-to-well interference. The operator quickly used these learnings to optimize well spacing across the area to maximize DSU value.