Influence of Property Modeling Practices on Estimated Hydrocarbon Pore Volume

Summary The emergence in geoscience of 3D geocellular modeling has raised questions about the correspondence of hydrocarbon volumetric deliverables with those derived from 2D zonal modeling. These differences of approach are compounded because the contemporary 3D methods use net-reservoir volumetric algorithms whereas the more established 2D methods have traditionally used net-pay volumetric algorithms. Both methods are initialized using 1D “along-wellbore” datasets. Key parameters for comparison are initial hydrocarbon pore volume (IHPV) for 2D and 3D modeling and equivalent hydrocarbon column (EHC) for 1D along-wellbore modeling. The focus is twofold. The first objective has been to generate a functional petrophysical model as a basis for volumetric comparisons. The reservoir notionally comprises an oil accumulation within a water-wet, lithologically-clean sandstone that is partially saturated with high-salinity brine. The sandstone comprises five rock types (RTs), each of which has a defined set of interpretive algorithms. The second objective has been to compare static volumetric estimates of EHC and/or IHPV for a range of case models. This objective was approached using three workflows. Initially, 1D along-wellbore screening studies used different case models representing various stratigraphic sequences. These allowed a preliminary assessment of results arising from the use of net-reservoir and net-pay volumetric algorithms without the influence of mapping practices. The findings were corroborated by field studies. Second, 2D zonal modeling led to values of IHPV based on both net-reservoir and net-pay algorithmic protocols. Third, 3D geocellular modeling also led to values of IHPV based on both protocols. These data allowed equitable comparisons of 2D zonal deliverables with those from 3D geocellular models because a consistent inter-well interpolation methodology was used for all 2D and 3D cases. The analysis incorporated the influence of stratigraphic sequences of the five RTs with their different petrophysical characteristics. Comparisons of 2D and 3D models showed that IHPV values delivered by established 2D zonal models with net-pay algorithmic protocols are mostly lower than those values delivered by contemporary 3D geocellular models with net-reservoir protocols by approximately 4% on average, but the differences are highly variable. These outcomes, which have implications for reserves estimation, are strongly governed by the stratigraphic distribution of the RTs. They re-emphasize that each case must be investigated separately and thoroughly.