Carbonates pose extreme challenges in petrophysical studies that lead to higher uncertainty levels for static and dynamic modeling of reservoirs. Consequently, economic viability studies for the field suffer from low levels of confidence. The difficulties in mapping "controls of fluid flow" exponentially increase when/if basic data availability is hindered by limited or failed wireline logs. We had taken two wells in a carbonate structure in which the control well had wealth of wireline, core, formation and well test data while the second well lacked vital wireline data and other information due to operational difficulties. The objectives of this study were to generate all petrophysical data from the control well and to develop models for propagating them to the original well and beyond.
In carbonates, no simple/apparent linear-relationship exists between permeability and porosity. Therefore, we used multivariate multiple regression models to evaluate statistical dependency of wireline logs to Flow-Zone-Indicator1 (FZI). FZI amalgamates permeability, porosity, tortuosity, shape factor in a single term lessening the un-correlating nature of permeability. Explicit-dependency of FZI to sonic-transit time and density log was revealed and used after calibration of NMR permeability to formation-tester data for the computation of FZI.
The subject carbonate was minimally fractured and had minimal number of vugs (a well-behaving carbonate). Although, lesser challenge was experienced in modeling, possible remedies for extreme heterogeneities were also formulated and presented. To incorporate the effects of "vug connectivity2 "on permeability, the modified form2 of Timur-Coates permeability model was fed with the profile of connectivity factor (p) from sonic-log4. Then, the permeability profile was re-computed for statistically tying transmissivities of matrix and connected-vugs and making the power-law model averages analogues to average permeability from WFT (wireline formation tester) data for the given intervals.
A new formulation was derived for determining the irreducible wetting phase saturation for the cases with no NMR log data in which resistivity-based water saturation (Swtt) can be in check with for more realistic prediction/determination of "dry" or "wet" production. Simplification and re-arrangement of the permeability model equation produced bulk-volume-irreducible (BVI) profile leading to Swirr-modeling.
Having the Sw, Swirr, permeability, porosity, and Corey-Burdine type relative permeability correlations for gas-brine carbonate systems made it possible to compute well productivity indices for gas and brine for various scenarios of skin factor, drawdown, drainage radius and well placements. The outcome of the model-driven petrophysical data set and reservoir engineering treatment of it for productivity estimations were in agreement with the actual productivity of well X1-2 verifying the reasonable accuracy of the models.