***Please note venue***
$25.00 Pre-Registration
$35.00 Late Registration
Basin-scale 3D Geologic Modeling of the Permian Basin:
What have we learned about the Delaware and Midland Basins?
Brian Casey
TORA - Tight Oil Resource Assessment Group (University of Texas, Bureau of Economic Geology)
The TORA team (Tight Oil Resource Assessment) at the Bureau of Economic Geology began constructing basin-scale, 3D Geo-cellular models over the Permian Basin in 2015. We modeled the Delaware and Midland sub-basins separately due to their size and computing requirements, and at this time have generated 3 iterations for each sub-basin. The latest Delaware Basin model was released to sponsors in late 2021 and presented at URTeC-2022. The latest Midland Basin model is nearing completion and will be released to sponsors in early 2023.
These are truly basin-scale models. In both sub-basins, we have extended the greater basin regions to their structural boundaries. We have re-interpreted the mappable units within the late Pennsylvanian through upper Leonardian interval using sequence-stratigraphic methods where possible. Faulting that impacts the late Pennsylvanian through lower Leonardian interval has been re-interpreted. A moderately tight grid has been used that is 1500 feet by 1500 feet laterally. In the most productive units, vertical layering is about 2 to 3 feet. The resulting model for Delaware Basin covers 23,630 square miles, has 1874 layers, and contains about 972 million cells. The resulting model for Midland Basin covers 32,540 square miles, contains 1380 layers, and contains about 814 million cells. Within the geo-cellular model for both basins we have interpreted lithofacies, and petrophysical, geomechanical, and hydrocarbon fluid properties.
For the Delaware Basin late Pennsylvanian through upper Leonardian interval, the Wolfcamp and Bone Spring formations have been subdivided into 19 major zones using outcrop, core, and subsurface studies. Uplift and erosion form the western and southern structural margins, with Laramide-related faulting to the west and Late Paleozoic thrust faulting (Marathon Thrust Belt) to the south. The northern structural margin extends to the northern edge of the Northwestern Shelf. The eastern-most structural margin extends over the Central Basin Platform to the faulted, eastern edge of the platform and is contiguous with the western structural boundary of the Midland Basin. Within these structural boundaries, regional dip is toward the basin center.
In the Delaware Basin, depositional limits for each major zone are defined as the basin-slope contact with the time-equivalent shelf edge. For the northern and eastern regions of the basin, these depositional limits occur the structural boundaries. For the western and southern regions of the basin, depositional limits have mostly been obscured or removed by structural events. Over 9800 wells have been used to establish the stratigraphy and depositional limits.
The Delaware Basin fault model uses basement-involved faulting that extends into the Wolfcamp. The mapped faults are recognized through a combination of well penetration and offset, measured active movement, 3D seismic data, and surface mapping. The thickness of mapped zones does change across some of the major faults due to their syn-depositional character, such as along the Grisham Fault zone.
Once the stratigraphic and structural modeling were complete, a lithofacies model was incorporated. The lithofacies model is core and outcrop-based, then extended to subsurface gamma ray and resistivity log data, and further supplemented by petrophysical porosity and mineralogic properties. Facies modeling uses semi-variance correlation (variograms) to characterize the spatial continuity of facies data, and to determine both geometry and facies body orientation. The facies model is then used to influence the data analysis and modeling of porosity and water saturation.
Petrophysical analyses of porosity, water saturation, and total organic carbon are based on the interpretation of over 1100 wells across the Delaware Basin and Central Basin Platform. The petrophysical interpretations from each well use a probabilistic, multi-mineral, simultaneous solver-based methodology. Once the petrophysical data has been upscaled and added to the 3D grid, spatial variability is analyzed per zone and per facies-type using variograms and distributed with Gaussian simulation. In a similar manner, geo-mechanical data is analyzed and modeled.
Hydrocarbon fluid property analysis is complicated due to a range of fluid types from black oil, volatile oil, liquid condensate to dry gas. The data is based on a combination of well production data and PVT analyses, with careful selection of data to minimize results negatively impacted by production depletion. A set of correlations are established to determine formation volume factors for each hydrocarbon type based on GOR, API oil gravity, and pore pressure. The pore pressure gradient is determined from a combination of well production, PVT, and well log data, with pore (or reservoir) pressure then determined by the depth below the ground surface.
A similar process is being used to construct the 3D geo-cellular model and properties for the Midland Basin. This model is mostly complete, with the finalization of hydrocarbon and geo-mechanical properties expected over the next few months.
The learnings from modeling both sub-basins are extensive, so only a few key factors will be noted here. Both basins originally extended into adjacent basin and had multiple regions of provenance. Faulting was mostly active during early Wolfcamp formation development, tapering off by the end of the Wolfcamp, but major faults remained as topographic features. Thus, the sub-basins during Wolfcamp deposition were compartmentalized, with both eustatic and tectonic conditions being significant depositional factors. Major Wolfcamp formation zones tend to prograde toward the basin center from low-angle shelves, but are more aggradational along steeper margins. Compartmentalization is significantly reduced during deposition of the Leonardian units, such as the Bone Spring and Spraberry formations, with eustatic sea-level variations being most significant. However, despite the impact of siliciclastics being prevalent during low-stands and carbonate clastic prevalent during high-stands, the Leonardian interval depositional boundaries are mostly aggradational. Clay content tends to be associated with provenance region, and in turn the clay content plays a significant role in reservoir quality and the reservoir intervals that are under active development.
Further iterations of these 3D geologic models will be produced as we gain additional insight. For example, we have a matrix permeability transform model under construction for the Delaware Basin. We have a formation salinity model for the Delaware Basin that significantly impacts water saturation, and a similar model is planned for the Midland Basin. The history and geo-location of thermal maturation across both basins remains an area that requires further understanding. Depositional boundaries can be improved with additional cross-sectional studies. Facies identification and modeling will continue to evolve, especially as new core accompanied by modern well log data becomes available.
LUNCHEON PROUDLY SPONSORED BY
