A Novel Design to Reduce Uncertainty in Remaining Oil Saturation Estimation Through Integrated Laboratory Core Measurements: A Case Study from Kuwait's Heterogeneous Oolitic Carbonate Reservoir

This study aims to reduce uncertainty in remaining oil saturation (ROS) estimation in Kuwait's heterogeneous Minagish Oolite carbonate reservoir. A progressive experimental workflow across three batches integrates Liquid Trapper coring, Dean-Stark, Core flooding, Centrifuge and NMR (T1/T2) techniques. Objectives include improving material balance, distinguishing hydrocarbon from water signals, and validating ROS and Sor using multiple methods. Deuterium tracer and doped brine enhance saturation accuracy. Results guide EOR efficiency and recovery factor estimation. A progressive stepwise laboratory workflow was designed to quantify ROS using cores acquired via Liquid Trapper coring. Learnings from every batch experiments were used for improvements in later phases. The workflow incorporated routine core analysis, Dean- Stark extraction, core flooding, centrifuge testing, and NMR including native-state and doped-brine measurements. Deuterium tracers assessed mud invasion, while doped brine enhanced hydrocarbon-water signal separation. Batch-1 showed fluid loss during core retrieval. In response, batches 2 and 3 integrated improved NMR protocols and centrifuge methods. These refinements enhanced material balance, reduced uncertainty in ROS estimation, and enabled reliable saturation measurements in swept and unswept zones. A Batchwise experimental workflow adopted to evaluate ROS in cores obtained by Liquid Trapper coring helped in optimising core sample selection process, enhancing cost economics and time efficiency in laboratory experiments. Batch-1 results indicated unreliable ROS estimates due to hydrocarbon losses possibly during core retrieval and handling. RCA and Dean-Stark analyses yielded material imbalance (So+Sw < 0.7), though trends aligned with log data. To address this, Batch-2 incorporated NMR (native, two-phase, single-phase) and core flooding, improving data quality and achieving better material balance (So+Sw > 0.95). However, some plugs still failed to yield oil during flooding, especially in partially swept zones. Batch-3 further enhanced the workflow by integrating centrifuge and doped-brine NMR methods. This allowed clearer hydrocarbon-water phase separation, confirmed by distinct NMR T2 signals. Doped-brine NMR improved hydrocarbon quantification and validated helium and saturation porosity results. Overall, the multi-method approach, including NMR and deuterium tracers, significantly reduced uncertainty in ROS estimation, revealed losses likely occurred during coring/lifting, and improved the confidence in oil saturation measurements—especially in complex, partially swept zones. The study confirms that a combined use of core flooding, Dean-Stark, NMR, and centrifuge methods is essential for accurate Sor/ROS evaluation in such reservoirs. This study advances the state of knowledge by demonstrating a robust, multi-technique workflow for accurately measuring Remaining Oil Saturation (ROS) in complex carbonate reservoirs. By combining core flooding, NMR (including doped-brine), centrifuge and Dean- Stark methods, it addresses fluid loss challenges during coring and improves saturation estimates. This enhances the reliability of EOR project evaluations, guides better reservoir management, and sets a benchmark for ROS analysis in heterogeneous, partially swept reservoirs.