Inconsistency between Fracture and Anisotropic Systems in Predicting Horizontal-sink Productivity.

Abstract

Modeling a reservoir containing unidirectional natural fractures by an equivalent anisotropic system is one of the most common methods in the realm of fractured reservoir engineering. The universal validity of such a macroscopic treatment, however, has not been quantitatively confirmed. This study primarily examined whether an equivalent anisotropic system is consistent with a fracture system through horizontal-sink productivity problems.Numerical experiments for a horizontal sink completed parallel or perpendicular to fractures with 1000 fracture realizations were performed to compare the productivity predictions based on naive fracture systems and equivalent anisotropic systems. The results showed the anisotropic approach yields erroneous productivities higher than the true values with mean relative errors of 11% and 13% for parallel and perpendicular completions, respectively. The conversion of fracture systems to equivalent anisotropic systems is not always appropriate for flow in the vicinity of a horizontal sink and may mislead the productivity predictions.A pseudo-skin factor was introduced to correct the overestimation of the anisotropic systems. Mean pseudoskin factors of 0.366 and 0.258 were required for parallel and perpendicular completions, respectively. Versatility of the correction technique requires a simple means to evaluate the pseudo-skin factor, thus a set of correlation equations with minimal information (horizontal sink length, total fracture length, and geometric mean of fracture lengths) was established. The inconsistency between fracture and anisotropic systems could be corrected by the systematic procedure proposed in this study.