Upscaling in the context of reservoir modeling is the process of averaging the reservoir properties to obtain coarse dynamic simulation models from high resolution static geologic models. The objective of upscaling is to preserve the maximum amount of heterogeneity with the minimum possible cost of simulation. However, the existing flow based upscaling methods rely on Steady State (SS) concepts of flow which cannot preserve the dynamic connectivity from the fine to coarse scale. This can result in overly homogenized simulation models which systematically yield optimistic results. The objective of this work is to develop an upscaling technique which preserves the dynamic connectivity from the geologic models to simulation models. Application of pressure transient concepts enables us to distinguish between well-connected pay and weakly connected payin the region of upscaling.This allows us to quantify the sub-volume of the resource that is not effectively producible. In terms of monetary value, the approach allows fundamental control on well planning. Over simplified simulation models systematically overestimate sweep, leading to a lower well count in the planning stage. By preserving the dynamic connectivity, the proposed approach helps in preparing a more accurate economic model for field development through proper well planning.Also, capturing the fine scale heterogeneity is important for any forward prediction of complex flow patterns in the reservoir.
The concepts of diffusive time of flight and drainage volume were utilized to accurately capture the local pressure transients. Also, a new volume averaging approach was developed to obtain the effective coarse transmissibility that accounts for non-uniform fluxes and pressure gradients while averaging.The proposed method was validated on SPE10 synthetic model and an onshore tight gas reservoir. The comparison of performance prediction against fine scale numerical simulation and Steady State incompressible upscaling demonstrate the accuracy of the proposed approach.