Abstract
Improved production forecasting allows completion engineers to compare cause and effect on alternative completion strategies and reservoir managers to optimize spacing and determine reserves. Paramount in production forecasting is recognition of the transition from linear flow into the fracture face to boundary dominated flow caused by the distance of investigation from adjacent fracture facies along the horizontal wellbore to meet. If all completions were the same and permeability constant, this transition point would be identical. Or if permeability and distance between fracture stimulation treatments were known, the time to this transition could be calculated. Without the constant reservoir/completion assumption and/or known reservoir/completion information an alternative approach can be employed using a stochastic technique exploiting offset well data where this transition can be determined. This study looks at horizontal oil wells that have been producing, approximately three to seven years, in the Bakken Shale, in McKenzie and Williams Counties, North Dakota. The technique utilized in this study is based on a finite conductivity model match of reciprocal rate versus cumulative production, this study uses monthly production data from public data records.
Inherent in this method model is the simultaneous fit of the boundary dominated Arps' b exponent during boundary dominated flow and the transition point. There will be an optimum fit of the boundary dominated flow production rate yielding an instantaneous rate and instantaneous decline that match the rate and decline at the end of linear flow. This linked Arps' method ensures a model continuum while honoring simulation models and field observations for finite conductivity fractures.
Once the transition time (end of linear flow/beginning of boundary dominated flow) has been determined a statistical approach was used to determine the probability that transition time will occur at various times using approximately 70 horizontal wells.
The latter in the well's life this transition occurs the higher the reserves. So for a well still performing in a linear flow regime, a probabilistic approach can be applied to ascertain proved, probable, and possible reserves.
