Abstract Stable isotope composition of gas is widely used in hydrocarbon exploration to determine the composition and thermal maturity of source rocks. Many isotope classification systems used for gas to source rock correlation and thermal maturity determination are primarily based on empirical observations made in conventional reservoirs and the kinetic isotope effects observed during pyrolysis experiments performed on source rocks. However, such relationships may not be readily applicable to onshore unconventional reservoirs due to the strong molecular and isotope fractionation that occur during extensive gas expulsion associated with basin uplift and depressurization. Degassing studies of freshly recovered core samples can provide useful insight into the behaviour of gas molecules in unconventional reservoirs during basin uplift. The analyses of Australian coal and marine shale samples demonstrate that during desorption both molecular and isotopic compositions of gas change at variable rates. Gas initially desorbed from the samples is mostly CH 4 , whereas later desorbed gas becomes increasingly enriched in C 2 H 6 and higher hydrocarbons. Hydrocarbon molecules also fractionate according to their isotopic composition, where the early released gas is enriched in 12 C causing the remaining gas in the reservoir to be enriched in the heavier 13 C isotope. During the release of gas from the Bowen Basin coals the C isotope ratio of CH 4 ( δ 13 C 1 ) changes by up to 21‰ (VPDB), whereas that for C 2 H 6 ( δ 13 C 2 ) and C 3 H 8 ( δ 13 C 3 ) changes by <6‰. Similar changes in the isotope composition can be seen during the release of gas from marine source rocks of the Beetaloo Sub-basin. In a fully gas-mature middle Velkerri shale sample, δ 13 C 1 changes by up to 28‰ and δ 13 C 2 by up to 3‰ with no appreciable change occurring in δ 13 C 3 . The extent of molecular fractionation during gas flow through carbonaceous rocks is primarily related to the adsorption–desorption properties of organic matter and diffusivity through the overall rock matrix. Using the current dataset, the magnitude of the contributions exerted by the desorption and diffusion processes cannot be readily distinguished. However, both Bowen Basin coals and Beetaloo Sub-basin shale show similar fractionation effects during gas flow, where the heavier alkane molecules, including those containing more 13 C, desorb and move slowly compared with the lighter components, in particular CH 4 . Different rates of isotope fractionation between hydrocarbon molecules during gas flow cause the shape of compound-specific-isotope (CSI) curve to change with time. Early released gas is characterized by a normal CSI trend where the short-chain hydrocarbons are isotopically lighter compared with the longer-chain hydrocarbons. Because CH 4 and C 2 H 6 molecules enriched in 12 C desorb and diffuse more readily than the heavier hydrocarbons (including those enriched 13 C), the gas remaining in the coal and shale samples after extensive desorption shows a reversed CSI trend where CH 4 and C 2 H 6 are isotopically heavier compared with the longer chain hydrocarbons. Reversed isotope trends may also develop over geological time, particularly where the source rock is fully gas-mature and has expelled hydrocarbons due to prolonged degassing. As seen in the Beetaloo Sub-basin, the CSI trend in the dry-gas-mature Velkerri shale is reversed, possibly due to the loss of a large proportion of originally generated CH 4 during post-Cambrian basin uplift.

ABSTRACT Scanning electron microscopy (SEM) has revolutionized our understanding of shale petroleum systems through microstructural characterization of dispersed organic matter (OM). However, as a result of the low atomic weight of carbon, all OM appears black in SEM (BSE [backscattered electron] image) regardless of differences in thermal maturity or OM type (kerogen types or solid bitumen). Traditional petrographic identification of OM uses optical microscopy, where reflectance (%R o ), form, relief, and fluorescence can be used to discern OM types and thermal maturation stage. Unfortunately, most SEM studies of shale OM do not employ correlative optical techniques, leading to misidentifications or to the conclusion that all OM (i.e., kerogen and solid bitumen) is the same. To improve the accuracy of SEM identifications of dispersed OM in shale, correlative light and electron microscopy (CLEM) was used during this study to create optical and SEM images of OM in the same fields of view (500× magnification) under white light, blue light, secondary electron (SE), and BSE conditions. Samples ( n = 8) of varying thermal maturities and typical of the North American shale petroleum systems were used, including the Green River Mahogany Zone, Bakken Formation, Ohio Shale, Eagle Ford Formation, Barnett Formation, Haynesville Formation, and Woodford Shale. The CLEM image sets demonstrate the importance of correlative microscopy by showing how easily OM can be misidentified when viewed by SEM alone. Without CLEM techniques, petrographic data from SEM such as observations of organic nanoporosity may be misinterpreted, resulting in false or ambiguous results and impairing an improved understanding of organic diagenesis and catagenesis.

ABSTRACT Organic matter (OM) in petroleum source rocks is a mixture of organic macerals that follow their own specific evolutionary pathways during thermal maturation. Understanding the transformation of each maceral into oil and gas with increasing thermal maturity is critical for both source rock evaluation and unconventional shale oil/gas reservoir characterization. In this study, organic petrology was used to document the reflectance, abundance, color, and fluorescence properties of primary organic macerals and solid bitumen (SB) in 14 Upper Devonian New Albany Shale samples (kerogen type II sequence) from early mature (vitrinite reflectance [VR o ] of 0.55%) to post-mature (VR o 1.42%). Micro-Fourier transform infrared (micro-FTIR) spectroscopy analyses were conducted on these samples to derive information on the evolution of the chemical structure of organic macerals and SB with increasing thermal maturity. Primary OM (amorphous organic matter, alginite, vitrinite, and inertinite) and secondary organic matter (SB) were identified in early mature samples. Amorphous organic matter (AOM) was the dominant organic component in early mature samples and was observed up to the maturity equivalent to VR o 0.79% but could not be identified at VR o 0.80%. An organic network composed of AOM and SB was observed from VR o 0.55 to 0.79%, which, together with the decrease in AOM content being accompanied by an increase in SB content, suggests that with the onset of petroleum generation, SB gradually replaced the original AOM. Alginite, represented by Tasmanites cysts, started to transform to pre-oil bitumen at a maturity corresponding to VR o 0.80%. It shows weak orange-yellow fluorescence at this maturity, a change from strong greenish-yellow fluorescence in early mature samples. Alginite could not be identified at VR o 0.89%, and generated bitumen remained in place or migrated over short distances. Petrographic observations and micro-FTIR study of alginite indicate that substantial hydrocarbon generation from alginite does not start until alginite is completely transformed to pre-oil bitumen. In contrast to AOM and alginite, vitrinite and inertinite derived from terrestrial woody materials occur as dispersed particles and do not change significantly during thermal maturation. A linear relationship between vitrinite and SB reflectance exists for the studied samples. The reflectance of vitrinite is higher than that of SB until VR o 0.99%, and at higher maturities, SB reflectance exceeds vitrinite reflectance. The inclusion of pre-oil SB converted from alginite in reflectance measurements could result in a lower average SB reflectance and, therefore, caution should be applied when using SB reflectance as an indicator of thermal maturity.

Abstract An integrated thermal history study of the Beaufort-Mackenzie basin of northern Canada is underway using multiparameter data gathered as part of a twelve year petroleum systems research project. New and legacy percent vitrinite reflectance (%Ro) data for approximately 81 wells have been compiled and standardized in order to make maps and cross sections showing thermal maturity trends and to provide paleotemperature constraints for thermal models. Data were quality-assessed by comparing measured %Ro values with other temperature-sensitive indicators (Rock-Eval pyrograms, liptinite fluorescence, and degree of apatite fission track (AFT) thermal annealing) and inconsistent legacy %Ro data were reinterpreted in conjunction with new sample analyses. Extensive organic matter recycling is the major issue affecting data quality and interpretation whereas other factors such as sample caving and oil staining (%Ro suppression) are important but less significant. Multikinetic AFT thermochronology data were obtained for 60 (mainly core) samples from 25 key wells with standardized %Ro data to provide constraints on the timing and rates of burial and exhumation across the study area. Also, thermal maturity and shale compaction trends were used to estimate net erosion magnitudes related to multiple phases of Tertiary deformation and exhumation. Measured %Ro is highest in exhumed strata along the southern basin margin and lowest in thick Cenozoic strata offshore (north) on the Beaufort shelf. Thermally immature strata persist to depths of > 4 km and to temperatures > 100°C on the outer Beaufort shelf due to rapid deposition of the thick (> 2.5 km) Plio-Pleistocene Iperk Sequence. Shale compaction trends suggest that 0.5-2 km of postrift Tertiary strata have been eroded from southern and southeastern onshore areas. %Ro-based erosion estimates are more variable and can give much higher values (by a factor of 3 or 4) than the compaction-based estimates, particularly in the Tertiary fold belt of the western and southern areas of the basin. There is a discontinuous increase in maturity across the unconformity that separates Jurassic synrift and Permian prerift strata along the southern basin margin. In the area south of Richards Island, maturity trends suggest that up to 4 km of Permo-Triassic strata may have been eroded prior to the deposition of Jurassic sediments. In the Anderson Plain to the east, erosion has been more extensive and Devonian rocks lie near the surface. Thermal modeling of AFT data from a Devonian sample in the Kugaluk N-02 well (AFT age: 216 Ma) suggests that exhumation was well underway in the Triassic, possibly in association with the onset of rifting. Exhumation may have continued until the Early Cretaceous, followed by reburial under 1-2 km of Cretaceous-Cenozoic sediments that were removed subsequently by erosion. The %Ro data provide important maximum paleotemperature constraints for integrated thermal history models based on multikinetic, AFT thermochronological data collected for the synrift and postrift successions.