Determination of in situ hydrocarbon contents in shale oil plays. Part 1: Is routine Rock–Eval analysis reliable for quantifying the hydrocarbon contents of preserved shale cores?

An accurate determination of in situ hydrocarbon (oil) content is critical for the assessment of shale oil plays in terms of their total resources in place and recovery potential. Although correction of light hydrocarbon loss to the Rock–Eval S1 parameters on conventional core and cuttings samples has been recently attempted using preserved (e.g., sealed or pressure–preserved) cores, hydrocarbon evaporation loss during sample preparation and delay time before the actual commencement of Rock–Eval analysis has not received sufficient attention. In this study, a combination of petrophysical and geochemical techniques including Dean–Stark extraction, hydrogen nuclear magnetic resonance (1H-NMR) T1-T2 mapping, thermal desorption–gas chromatography (TD–GC) analysis, and Rock–Eval programmed pyrolysis were used in parallel on a set of 30 preserved shale cores to determine their total hydrocarbon contents. The preserved shale cores were analyzed in large pieces (e.g., 3–5 cm in size) by the former two petrophysical methods while using fine powder (e.g., <60 mesh) by the latter two thermal analytical methods. The total free hydrocarbon contents (equivalent of Rock–Eval S1) determined on the studied lacustrine shale samples with a Ro ranging from 1.00% to 1.51% followed the order of Dean–Stark (average 10.27 mg/g) ≈ NMR (10.34) > TD–GC (6.65) > Rock–Eval (3.93). The core crushing and grinding process required for both TD-GC and Rock–Eval pyrolysis analyses was found to cause a loss of approximately 35% to the in situ hydrocarbons. In addition, Rock–Eval analysis delay time (of 5 min) could lead to another 25% reduction to the total free hydrocarbons. In combination, sample preparation and analysis delay resulted in up to 60% reduction to the Rock–Eval S1 peaks for the preserved shale cores, and a 30% loss to the liquid (C6+) hydrocarbons in particular. To the best of our knowledge, this is the first reported observation that in situ light hydrocarbon contents can be severely underestimated by Rock–Eval and TD–GC analyses using powdered forms of preserved core samples. The extents of light hydrocarbon loss in conventional cores have been underestimated using the Rock–Eval S1 difference between preserved shale and its long–duration exposed replicates. Hence, when quantifying the hydrocarbon contents of the preserved shale samples, techniques utilizing larger rock pieces are highly recommended.