Water Saturation Profile Analysis and Modeling: Case Study in a Tight Carbonate Reservoir

Abstract

Abstract The complexity in carbonate reservoir evaluation stems from their intricate pore topology and mineralogical classifications. Classical approach to determine fluid distributions is through resistivity logs, however, this is marred by drawbacks as the resultant water saturation (Sw) can sometimes exhibit a profile that has limited to no usage as it is modeled with porosity and lithological volumes only. Significant uncertainties in Sw calculations from logs in heterogeneous tight carbonate rocks combined with the difficulties in measuring reliable pressure gradients, render the interpretation of transition zones and free water levels challenging. This case study describes the integration of limited core capillary pressure (Pc) measurements, formation pressures, conventional and Nuclear Magnetic Resonance (NMR) logs to define a saturation height model that explains the Sw profiles of a tight gas carbonate reservoir. The log derived Sw and its associated uncertainty have been computed using input logs and electrical parameters from cores. The NMR log is at the center of the saturation modeling. It integrates the pore size distribution into a synthetic Pc, establishes the irreducible water saturation by adjusting a variable T2 cut-off, and defines a permeability transform adjusted to core data and adapted for wells with conventional logs data only. The pressure gradients analysis, core and synthetic Pc and the petrophysical properties determined from logs have ultimately been combined to define a Saturation Height Function (SHF) for all the wells. The Sw profiles and the differences in free fluid level positions between the wells cannot be explained by structural or stratigraphic compartments. The assumption of undisturbed initial saturation conditions has been questioned, and further supported by the initial pressure data that could involve earlier production from nearby field. The porosity permeability and Swi estimation from logs are coherent with core data including routine and special core analysis. A key observation was the mismatch of transition zone from core Pc data relative to the Sw profile and Free Water Level (FWL) from gradient analysis; therefore, Sw profile is believed to be more reliable. Uncertainty in Sw calculations exists at high Sw, however, it seems that residual gas exists below the "current" FWL. Moreover, GWC in some wells cannot be explained with the existing reservoir compartmentalization. This study reveals how the analysis and modeling of the Sw profile can change the understanding of reservoir compartments and field history and have significant impact on its development. A major concern highlighted is regarding those fields having wells at original saturation conditions as if there is a common deeper FWL and variable swept areas leading to different contacts. Finally, relating poro-perm rock typing to lithofacies association and depositional environment can be advanced through core descriptions and thin sections.