The Pliocene B4 sandstone is an important gas-bearing reservoir in the Oilbird field, Columbus Basin, southeast Trinidad. The B4 sandstone is dominated by fine-grained, massive, and parallel-laminated sandstones interbedded with thinly laminated siltstone and very fine-grained sandstones deposited in a shelfal setting. A detailed sedimentologic study of the B4 reservoir was performed by integrating 60 ft (18.3 m) of core, 3770 ft (1149 m) of borehole image data, open-hole logs, mud logs, and biostratigraphic data. A total of 12 sedimentary facies were described and interpreted on the basis of sedimentation processes using a genetically oriented facies analysis approach. Six facies associations were identified based on the gamma ray (GR) pattern and the vertical facies association. Core data show mostly massive sandstones with very little to no bioturbation. These massive sandstones commonly alternate with intervals having diffuse lamination, which appear as a recurrent feature within the massive bodies. The transitional recurrence of massive and parallel-laminated sandstones indicates velocity fluctuating and sustained turbulent flows. Sedimentologic evidences suggest that the origin of the B4 sandstone could be related to the paleo-Orinoco River-related product of turbidity (hyperpycnal) outflows that extended into the Columbus Basin. The associated occurrence of plant debris favors the interpretation of a direct fluvial supply by rivers in flood stage (hyperpycnal systems).

The lateral correlation of facies associations throughout seven wells allowed the identification of six depositional units, named from base to top, A, B, C, D, E, and F. Facies maps for the B4 reservoir have been developed and were constrained by paleocurrent data extracted from image data. The B4 reservoir sand shows a progradational pattern reflecting the infill of a fault-controlled depocenter. Paleoflow data indicate axial transport roughly parallel to the main fault system, indicating that faults were active and controlled the accommodation space. The proposed hyperpycnal depositional model will introduce substantial changes for the prediction of the geometry and position of sandstone accumulations. The hyperpycnal model predicts the occurrence of sand accumulations in the lower parts of the paleolandscape, whereas the higher parts of the basin (margins) are characterized by fine-grained sediments. Thus, this new depositional model could represent a drastic shift in the prospectivity guide for new exploration plays.