Low Resistivity Pay intervals are often present along carbonate reservoirs sections and bring a lot of problems in fluid determination and saturation assessment. An integrated methodology has been developed from a well documented Libyan field case to tackle these problems. The resistivity logs run in this reservoir formation, drilled with water based mud, lead after interpretation to predict values of water saturation higher than 70%, despite low BSW observed after perforation. The feeling was that actual water saturation was much lower, from core observations showing that the problematic interval was very heterogeneous, with macro porous oil bearing patches embedded in a micro porous water bearing background. This particular texture could have generated an electrical bypass of hydrocarbon filled connected porosity, leading to a biased estimation of oil volume while interpreting resistivity with conventional Archie's m=n=2 parameters. The workflow designed to improve such formations characterisation follows 3 steps:
Acquisition and processing of Computed Tomography scanner images to visualise reservoir heterogeneities in 3D along the cored section, completed by minipermeameter acquisitions to characterise them in terms of reservoir quality.
Building of a core scale geological model of the cored interval and of its wellbore section, constrained by a CT scan derived facies cube and petrophysical properties based on core measurements (conventional plugs, core minipermeameter, SCAL data, including Interfacial Tension and capillary Pressure measurements). A reference average saturation data was obtained from this core scale geomodel.
Forward modelling of the resistivity response of a 2D section derived from the core geomodel, using a research software simulating the invasion process, to be compared with field acquired resistivity curves.
This approach allowed to verify that the bias in saturation evaluation was, as expected, directly explained by the complex 3D heterogeneity of the reservoir, not detected by conventional logs, and not accurately represented by the conventional values of Archie's parameters. Average water saturation value of 51%, derived from core geomodel was found more consistant with production data, through fractional flow analysis, than the 77% one obtained from resistivity logs interpretation, considering conventional Archie's parameters. These results can be used to correct the log saturation of equivalent facies intervals in case resistivity tools only are available, providing image log or geological knowledge lead to their identification.