Integration of Improved Asymmetric Frac Design Using Strain Derived From Geomechanical Modeling in Reservoir Simulation

Abstract

Capital investment and field development plans are based on evaluation of well performance. Numerical simulation models are widely used in that matter to derive estimated ultimate recovery (EUR) and depletion patterns of the stimulated reservoir volume (SRV). The main challenge when building a simulation model of hydraulically fractured wells is to realistically represent the heterogeneous conductivity distribution of the propped volume and its interaction with natural fractures. Field data such as microseismic and chemical tracers show evidence of horizontal and vertical asymmetric geometry of the hydraulic fractures. However, in practice they are modeled using symmetric bi-wing geometry due to over-simplified fracture models, where homogeneous stress field, rock properties and pore pressure are assumed. Simulation results obtained from such over simplistic models would lead to miscalculation of depletion patterns causing deficient spacing plans and performance forecasting. Previously, successful attempts were made to overcome the issue by calculating a volumetric enhanced permeability in two sub regions: in the near vicinity of the well and the SRV region, both extracted from strain derived as a result of the dynamic geomechanical simulation of the interaction between hydraulic and natural fractures. This approach improved the understanding of pressure depletion patterns while accounting for geomechanical heterogeneity between stages and wells in a pad; however, the transition from the dynamic geomechanical simulation of the frac complexity directly to the reservoir flow simulation left out a major tool commonly used in the design of the hydraulic fractures: frac design software and its resulting practical recommendations. To incorporate the frac design in the workflow, this paper illustrates how the frac complexity represented earlier as a volume can be discretized into multiple mathematical hydraulic fracture planes. This planar hydraulic fracture representation is a realistic mathematical approximation of the volumetric frac complexity that has been captured and validated with microseismic data in the dynamic geomechanical simulation. To keep the realism of asymmetric stimulation derived in the dynamic geomechanical simulation, the proposed workflow uses an asymmetric pseudo-3D hydraulic fracture model that accounts for the variation in height considering asymmetric half lengths due to lateral stress gradients in a heterogeneous reservoir. Such a frac design model results in a realistic field validated asymmetrical fracture geometry and conductivity for each activated fracture, calibrated to the proposed fracture treatment, or post-frac data. In this paper, the previous volumetric approach of modeling frac complexity and exporting it directly to reservoir simulation is briefly reviewed and the new workflow that represents the volumetric frac complexity with multiple asymmetric fracture planes is presented. Two scenarios are considered: 1) a coarse approximation of the frac complexity volume with a single fracture per stage and 2) a fine detailed representation of the stimulated volume using multimathematical hydraulic fracture planes per stage solution. The reservoir simulation results indicate that the fine multi fracs per stage representation is able to match the early pressure which contrasts with the coarse single frac per stage approximation unable to represents the complex physics occurring when the early flow is dominated by the high permeability created around the hydraulic fractures. The ease with which the early pressure was matched indicates that the fine multi frac stages approximation of the frac complexity volume is a viable solution to mathematically represent the complex realities of the hydraulic fracturing and its consequences on the asymmetric depletion. In addition of having a validated approach that can 1) predict microseismicity during the dynamic geomechanical simulation, 2) honor the frac treatment data during stimulation and 3) match the early pressure and beyond during production, we must emphasize another major benefit of this workflow is that it uses existing reservoir simulators, requiring from the user no additional reservoir simulation budget expenditures or training.