No abstract

The Colombian Caribbean region has become an important exploratory target, and recent discoveries confirm its potential as a gas province to overcome the expected near-future gas deficit. A petrophysical and dynamic characterization workflow was implemented for this challenging deepwater play, where the depositional environment is the result of turbidity current processes. The reservoirs consist mostly of thin to very thin sand layers, corresponding mainly to the Ta, Tb, and Tc divisions of the Bouma sequence as observed in the cored intervals. Bouma divisions Td and Te are related with the lowest rock quality and represent the nonreservoir intervals. The greatest challenge in the characterization of this particular reservoir is the vertical resolution, given the very low thickness of the layers, which becomes very difficult to detect using standard resolution logs. Thus, tomography images, resultant CT-scan curves, and their integration with routine and special core analyses were used to reveal the true nature of this complex reservoir. The proposed methodology focuses on the integration of routine and special core analysis for the petrophysical and dynamic characterization of the reservoir, where the pore-throat-radius distribution from high-pressure mercury injection becomes the basis of the differentiation between what is considered reservoir and what is not. Pore-throat radius estimated from high-pressure mercury injection (R35) correlates extremely well with textural features and clay content in the rock; therefore, this parameter (R35) was used to define the different classes for rock typing. The approach taken was to develop a multilinear regression model of R35 as a function of very high-resolution tomography outputs in the cored zones and then see how it may be extrapolated to the uncored zones using available high-resolution logs. Special petrophysical analyses, such as NMR low field, porous-plate capillary pressure, electrical properties, and relative permeability curves (steady state), showed correlation with the defined rock types and, in turn, allowed for a determination of the gas accumulation potential of the area. Finally, rock and fluid (dry-gas) properties have been used to build a single-well radial model to design initial well tests (DST) and predict production performance from each interval (selective tests). The simulation model represents the lateral and vertical heterogeneity related to the geological environment (turbidites). The final results have defined the flow and shut-in times during tests to optimize the budget.