Search

The unpredicted discovery of downdip oil in Permian–Carboniferous Unayzah sandstone reservoirs along the southern flank of the predominantly gas condensate Ghazal field in Saudi Arabia highlighted the need for a thorough assessment of controls on hydrocarbon properties, with the aim to improve prediction and reduce uncertainty. Primary controls on hydrocarbon property variations (condensate-to-gas ratio) here appear to be compartmentalization, multiple charging, presumably from a Lower Silurian Qusaiba kitchen to the south and east within the late-oil to wet-gas generation window, and dysmigration to the north and west. Gas on production represents mostly unmixed thermogenic accumulations of variable maturities, with no sign of biodegradation or thermal cracking. Light oil (1.06–1.21% calculated vitrinite reflectance [Rc]) in the southern part of the field represent late-mature accumulations that apparently escaped flushing by subsequent gas condensate charges generated near the wet-gas generation window (1.41–1.55% Rc) and currently occupying the field to the north. Constant maturity variations for both gasoline range and heavier components in each sample testify to a lack of mixing between the hydrocarbon phases. Petroleum inclusions similarly comprise either light oil or gas condensate that coexist with significantly overlapping homogenization temperatures, suggesting considerable overlap in their trapping history (135–35 Ma). Less mature petroleum has not been identified in the field, possibly because of a lack of trap closure or migration focus prior to the inferred Late Jurassic filling or because of displacement to updip structures, thereby high grading the probability of undiscovered oil west and northwest of Ghazal. If thermal generation instead of retrograde condensation or phase separation controls this petroleum system, then the potential for deeper oil accumulations is limited.

3D seismic acoustic impedance inversion was performed for a Permo-Carboniferous sandstone reservoir interval based on limited well control. The impedance volume was then depth converted and used to target development well locations. The drilling campaign yielded mixed results; some areas of the field showed a high correlation between impedance computed from the wells and that derived from seismic, and other areas showed a low correlation. This difference in correlation presented a challenge when using the impedance model to condition the porosity model. Changes to the impedance volume would be needed in areas of low confidence in the seismic data. To address this problem, the depth-converted impedance volume was rebuilt using a 3D intensity function. This technique assigns appropriate weights to the seismic data in different areas of the field, based on confidence in the seismic data quality. Using this technique produced a more realistic porosity model, when conditioned by the revised impedance model. The subsequent construction of an object-based facies model also used intensity functions to locate seed positions of the sand dune objects, which were then tied to specific impedance ranges in the updated impedance model. Paleosol and sheet-sand facies, unrestricted to specific porosity ranges, were constructed separately using probabilities of occurrence from each well. The separate models were then merged together to produce the final facies model.