EOS-Based Compositional Simulation Models for Liquid-Rich Gas Condensate Reservoirs in the Permian Basin

Abstract Across the Delaware Basin in Texas, the reservoir fluid systems and PVT properties vary dramatically. However, with a robust understanding of the fluid systems and behaviors, these dynamic changes can be predicted and leveraged for continuous advancements. The fluids transition from black oil in eastern Loving County to volatile oil in the center of northern Loving County, shifting to near-critical fluid or rich gas condensate on the western part of Reeves County. Understanding the gas/oil ratio (GOR) and condensate-to-gas ratio (CGR) behaviors of near-critical fluid and liquid-rich gas condensate system are crucial to effectively estimate reserves, optimize well spacing, and design efficient completions. Moreover, significant GOR variations with depth are observed from multiple codeveloped Wolfcamp benches. Scientific data and hydraulic fracture models have confirmed significant vertical fracture growth penetrating the high-stress carbonate layers and connecting multiple zones with different producing GORs. Therefore, building a common equation of state (EOS) model (C-EOS) is critical in capturing PVT fluid behavior and vertical composition gradient changes for field-wide co-development scenarios. In this study, the long-term production is forecasted by constructing a C-EOS compositional model to represent various GOR/CGR for codeveloped Wolfcamp benches. The forecast captures long-term CGR behaviors for section-level infill development. This behavior demonstrates early time constant CGR when the system pressure remains above the saturation pressure, followed by a gradual declining profile along with pressure depletion. The C-EOS models are tuned and validated using local PVT data covering a wide range of the GOR and CGR. The composition gradient is applied to represent one generic PVT system, which is initialized with a lighter component downdip (gas condensate with higher C1 mol%), mainly dominated by organic-rich argillaceous siltstone facies, and heavier fluid upward (near-critical fluids) associated with interbedded calcareous silt, mixed carbonate layers, and organic-rich biogenic mudstone beddings. The simulation models are upscaled from the regional subsurface geomodel incorporating facies-controlled petrophysical properties. By integrating microseismic monitoring and scientific learnings on fracture height, hydraulic fracture models are built based on standard completion designs, from which the representative stimulated rock volume (SRV) profiles are extracted. The key rock-fluid properties, SRV/fracture geometry, and well spacing configurations are investigated through the history-matching process and sensitivity analysis. The integrated analytical and numerical modeling workflow captures the generalized GOR/CGR profiles for Wolfcamp formations in Delaware Basin with various PVT regions: volatile oil, near-critical fluid, and rich gas condensate. The field-level EOS-based compositional simulation models help estimate the long-term condensate yield forecast for liquid-rich PVT regions at a given development well spacing. The integration of the dimensionless GOR/CGR, degree of undersaturation, and linear flow pattern/oil originally in place (LFP)/OOIP) ratio offers a way to high-grade a field development strategy by ranking development opportunities based on the PVT fluid system, the degree of undersaturation, and the completion efficiency.