Deep subsurface exploration is important for mining, oil and gas industries, as well as in the assessment of geological units for the disposal of chemical or nuclear waste, or the viability of geothermal energy systems. Typically, detailed examinations of subsurface formations or units are performed on cuttings or core materials extracted during drilling campaigns, as well as on geophysical borehole data, which provide detailed information about the petrophysical properties of the rocks. Depending on the volume of rock samples and the analytical program, the laboratory analysis and diagnostics can be very time-consuming. This study investigates the potential of utilizing machine learning, specifically convolutional neural networks (CNN), to assess the lithology and mineral content solely from analysis of drill core images, aiming to support and expedite the subsurface geological exploration. The paper outlines a comprehensive methodology, encompassing data preprocessing, machine learning methods, and transfer learning techniques. The outcome reveals a remarkable 96.7% accuracy in the classification of drill core segments into distinct formation classes. Furthermore, a CNN model was trained for the evaluation of mineral content using a learning data set from multidimensional log analysis data (silicate, total clay, carbonate). When benchmarked against laboratory XRD measurements on samples from the cores, both the advanced multidimensional log analysis model and the neural network approach developed here provide equally good performance. This work demonstrates that deep learning and particularly transfer learning can support extracting petrophysical properties, including mineral content and formation classification, from drill core images, thus offering a road map for enhancing model performance and data set quality in image-based analysis of drill cores.

The analysis of the elemental composition of sediments provides valuable information about input sources, paleoredox conditions, or paleoproductivity. In this study, 1498 rock samples from six wells within the Triassic Montney Formation in Central Alberta, Western Canada, were analyzed using energy-dispersive X-ray fluorescence (ED-XRF). Wavelength dispersive X-ray fluorescence measurements were used to build and validate a Montney-specific calibration to interpret elemental concentrations from the ED-XRF dataset. The calibration accounted for heterogeneities, matrix variations and unique elemental associations that may be present within the Montney Formation. Sediment input sources are interpreted based on cross-plots of proxy elements to be terrigenous, carbonates and biogenic in origins. Analysis of multiple paleoredox proxies indicated the environment of deposition to be mainly oxic. The ratios of potassium (K) to rubidium (Rb) were used as a paleoclimate proxy. Results suggest the predominance of low chemical weathering attributed to a hot and arid paleoclimate during deposition of Montney Formation sediments. Phosphorous concentrations were determined to be strongly associated with Ca, as such it was interpreted to be inorganically sourced, thereby precluding it from being used to assess paleoproductivity. Excess Si interpreted to be biogenically sourced was used in interpreting moderate and high paleoproductivity for the Montney formation. This work demonstrates the versality of matrix-specific calibrated ED-XRF measurements as a valuable, cost efficient, standalone technique in geochemical analysis for characterizing geological formations.

In Switzerland, radioactive waste management requires the safe disposal of nuclear waste within a deep geological argillaceous sequence that includes the Jurassic-age Opalinus Clay. The process of selecting a suitable site for the disposal is ongoing and involves a rigorous site selection process. As part of this process, site-specific physico-chemical data were collected, which include parameters such as cation exchange capacities (CEC) and exchangeable cation occupancies, in addition to other geochemical and mineralogical data. The mineralogy, CEC and exchangeable cation occupancies of rock samples collected from the various lithologies of cores from seven boreholes across the study areas, namely Z & uuml;rich Nordost (ZNO), Nordlich Lagern (NL) and Jura Ost (JO) were investigated. Four different CEC methods were applied, and the results obtained are in good agreement. The general trend in the CEC data follows: Cs -CEC >= (Sigma CATIONS) > Ni -CEC. The higher Cs -CEC values are due to the higher interlayer extraction yields of low hydration cations K+ and NH4+ in illite rich rock samples by the highly selective Cs+. A clear correlation between the 2:1 phyllosilicate content and the CEC values is observed over the entire sequence, so clay minerals primarily govern cation exchange processes and thus the retention of cations. Finally, for the Opalinus Clay of each of the three study areas, porewater chemistries were modelled based on a combination of mineralogical and physico-chemical data. The calculated porewater compositions are in good agreement with the compositions obtained from squeezing and advective displacement experiments in the laboratory.

Argillaceous rocks are currently being examined worldwide as potential host rocks for repositories for radio-active waste and spent fuel due to their favourable characteristics such as the self-sealing ability or limited diffusive transport of fluids, solutes or gases. In this study, through-diffusion experiments focused on rock samples with different mineralogical compositions taken from deep boreholes in the Mesozoic rock sequence of northern Switzerland, to determine diffusion parameters of HTO, 36Cl- and 22Na+ perpendicular to the bedding plane. Both effective diffusion coefficients and accessible porosities were calculated from the data obtained. In the Opalinus Clay, the primary selected host rock of the repository, the effective diffusion coefficients (De) and accessible porosity values (epsilon) of the neutral species HTO within and across all study areas are relatively consistent (De = 8.8 +/- 1.9 x 10-12 m2 s- 1; epsilon = 0.12 +/- 0.02), indicating that the material can be considered homogeneous in terms of diffusion properties. Importantly, the chemical composition of the pore water has no effect on the values of the effective diffusion coefficients of HTO. On the other hand, the rocks above and below the Opalinus Clay, up to the next aquifer, consist of tight sedimentary units (carbonates, siliciclastics including argillaceous rocks) with a wider range of diffusion coefficients, indicating a higher level of heterogeneity. The diffusion coefficients of anions, such as 36Cl-, tend to be smaller due to anion exclusion effects, while those of cations, such as 22Na+, tend to be larger due to surface enhanced diffusion. The extent of these effects is largely determined by the composition of the pore water, and - in the case of cations - by the extent of sorption which depends also on the sorption capacity - indicated by the cation exchange capacity -of the solid. As a result, variations in the effective diffusion coefficient of 36Cl- and 22Na+ can be observed firstly between different rock types (i.e. composition, fabric), and secondly between the different study areas due to differences in the chemical composition of the pore waters in contact with the solid phase. Based on the analysis of around 130 samples, it has been observed that the diffusion coefficients of the Mesozoic units in northern Switzerland exhibit a significant correlation with the total clay content of the rocks. These findings pave the way for the establishment of a diffusion database for the relevant elements under consideration in the Swiss radioactive waste programme and the Safety Case.

The Devonian Duvernay Formation in Alberta is a carbonate-siliceous unconventional source rock that produces prolific oil, gas condensate liquids, and gas. In the southern part of the basin and the East Shale Basin, high frequency interbeds of limestone and organic-rich calcareous shales are common at spacing of inches to several feet. Prolific gas liquids are trapped within the shale/mudstone units while the highly cemented and tight carbonate units show a lack of hydrocarbons. Even though most wellbores completed to date have primarily landed in shale units, still, due to geo-steering and other considerations penetration through limestone unit along the lateral cannot be avoided. High contrast in mechanical properties of shale and limestone units along with anisotropy in geomechanical properties of the same rock unit makes optimal hydraulic fracturing breakthrough and placement often challenging. This study reviews examples of high resolution mechanical Duvernay stratigraphy combined with log and core analyses in order to better understand the implications of hydraulic fracture development through such complex reservoirs. Integrated petrophysical well log and core data indicate that there are strong heterogeneities on lithology, and geomechanical properties (hardness, brittleness, and elastic properties) on various scales. The shale units generally have lower Young's modulus and Poisson's ratio, whereas the carbonate units have high Young's modulus and Poisson's ratio. Core data also indicate strong vertical transverse isotropy in the shale units. Combined with occurrence of various types of naturally-occurred sub-vertical fractures and sub-horizontal polished slip faces, weak bedding plane strength, and strong anisotropy in in-situ stresses, and high heterogeneities in lithology and geomechanical properties likely result in complex and poorly constrained and un-optimized hydraulic fractures throughout the Duvernay Formation. The results indicate that high Young's modulus of carbonates causes a narrow fracture width which makes fracture initiation and placement and in particular proppant transport a challenge. Normal and strike slip faulting regimes are observed in some areas of East Basin. The lower anisotropy in horizontal stresses estimated for the shale zones than the carbonate zone may suggest that the complex fractures is more likely to occur in the shale zones than the carbonate zones. In Duvernay minifrac tests, longer fall off data collection (> 2 weeks) is critical for a successful determination of closure pressure and pore pressure estimation. During fracture stimulations, tight cluster spacing, i.e. ≤10m apart, might limit fracture geometry growth on the inner clusters due to higher induced stress. Similarly, due to higher stress shadow effects, tighter stage spacing might affect the fracture growth on the subsequent stages.

The El Niño-Southern Oscillation (ENSO) is a driver of global atmosphere-ocean dynamics, but projections of frequency and magnitude in different climate states remain uncertain. Palaeoclimate records offer the potential to improve our understanding of ENSO behaviour but most are fragmentary, suffer low resolution, and/or typically do not cover periods warmer than present day. The Last Interglacial (129-116 kyr BP) was the most recent period during which global temperatures were close to 21st century projections, and potentially provides insights into operation of climate modes of variability in the future. Here we report a continuous, inter-annually resolved record of hydroclimate spanning 220-80 ka from Lynch’s Crater in tropical northeast Australia, a region highly sensitive to ENSO. Our reconstruction is based on a micro-X-ray fluorescence (XRF)-generated elemental profile at 200 µm resolution, combined with loss-on-ignition, magnetic susceptibility, and pollen analysis. We find that during globally warmer periods (including super-interglacial Stage 5e, and 5c), there are significantly larger amplitudes in high-frequency ENSO spectral range (3-8 years), which are absent from the record during the glacial stages MIS6 and MIS4. Our results imply an ENSO dependence on mean climate, with enhanced ENSO variance during interglacials globally warmer than present. These results are consistent with climate model projections for a future slowdown of the Walker circulation and more extreme El Niño events under greenhouse warming.

In the Lower Cretaceous McMurray Formation (Alberta, Canada), many intervals of intensely bioturbated (Bioturbation Index = 5-6) fine-grained sediments are characterized by high gamma-ray (GR) readings. Several methods, including sedimentary facies analysis, thin-section petrography, handheld spectral gamma-ray, portable X-ray fluorescence, X-ray diffraction, inductively coupled plasma-mass spectrometry, microprobe of K-feldspar, energy dispersive spectroscopy, and detrital zircon geochronology by laser ablation-inductively coupled plasma-mass spectrometry, were used to investigate the interval of interest in core samples. The mineralogical analysis shows that these intervals are enriched in heavy mineral grains, and particularly in zircons. The content of radioactive elements is variable. Thorium is commonly elevated up to three times, uranium nil to two times, and potassium content usually remains normal. The studied intervals consist of interbedded, bitumen-saturated cross-bedded and/or ripple cross-laminated sandstone (high-energy deposits) and light-gray bioturbated mudstone (low-energy deposits), commonly addressed as inclined heterolithic strata (IHS). IHS represent tidally influenced, brackish-water, upper point-bar deposits. The zircon grains become concentrated while hydraulic processes interact with bioturbation: the burrowing animals cause significant sediment mixing that allows the lightest sediment particles to go back into the suspension. Additionally, bioturbation increases the surface roughness along the sediment-water interface and, causes more turbulent flow, allowing for quartz and other light grains to be removed by traction and/or saltation, while dispersed heavier zircon grains become trapped and concentrated in open burrows. So far, this study is the first to demonstrate the importance of bioturbation in the enrichment of zircon grains in IHS. The interaction of bioturbation and hydraulic processes explains the apparently counter-intuitive enrichment of heavy minerals in a low-energy depositional setting. This scenario likely applies to numerous intervals characterized by similar GR and/or zirconium spikes across the McMurray Formation. Furthermore, it can be expected that in other sedimentary basins and stratigraphic units, similar studies will demonstrate that the proposed mechanism is universal.

Abstract For optimizing and enhancing hydrocarbon recovery from unconventional plays, the technological race is currently focused on development and production of state-of-the-art surfactants that reduce interfacial tension to mitigate obstructive capillary forces and thus increase the relative permeability to hydrocarbon (kro). This study provides insight into the pore-scale evaluation of the latest flowback enhancer technologies currently applied in the Permian Basin, Texas, USA. A multidisciplinary approach, including concepts of nanotechnology, was used to assess fluid-fluid and rock-fluid interactions occurring at the nanopore scale and their implications on enhancing oil recovery. A heterogeneous dual-porosity dual-permeability microfluidic chip was designed and developed with pore geometries representing shale formations. This micro-chip simulated Wolfcamp shale with two distinct regions: (i) a high-permeability fracture zone (20 µm pore size) interconnected to (ii) a low-permeability nano-network zone (100 nm size). The fluorescent microscopy technique was applied to visualize and quantify the performance of different flowback enhancers during injection and flowback processes. This study highlights results from the nanofluidic analysis performed on Wolfcamp Formation rock specimens using a microreservoir-on-a-chip, which showed the benefits of the multi-functionalized surfactant (MFS) in terms of enhancing oil/condensate production. Test results obtained at a simulated reservoir temperature of 113°F (45°C) and a testing pressure of 2,176 psi (15 MPa) showed a significant improvement in relative permeability to hydrocarbon (kro) in the nanomodel when MFS was added to the stimulation fluids at loadings as low as 0.05% v/v. The results were compared against other premium flowback enhancers. Measurements using a high-resolution spinning drop tensiometer showed a 40-fold reduction in interfacial tension when the stimulation fluid containing MFS was tested against Wolfcamp crude at 113°F (45°C). Also, MFS outperformed other premium surfactants in Amott spontaneous imbibition analysis when tested with Wolfcamp core samples. This work used a nanofluidic model that appropriately reflected the inherent nanoconfinement of shale/tight formation to resolve the flowback process in hydraulic fracturing, and it is the first of its kind to visualize the mechanism behind this process at nanoscale. This platform also demonstrated a cost-effective alternative to coreflood testing for evaluating the effect of chemical additives on the flowback process. Conventional lab and field data were correlated with the nanofluidic analysis.

The fast pace of drilling and completion of unconventional reservoirs in North America is challenging engineers, geoscientists and petrophysicists who have to make prompt and reasonable plans for drilling and completion strategies. One of the main issues is understanding the physical rock properties, such as clay and organic content, and mechanical properties or stress behaviour, in these often highly anisotropic reservoirs. In many cases, these parameters are selected or determined based only on well log data, and if available, a few data points from core derived from close or far offset wells. This practice continues despite the fact that, during drilling, valuable cuttings material is available; in addition, government archives store valuable cores or cuttings material from older, offset wells.

High resolution chemostratigraphy of the Montney Formation across Alberta identifies three unique and distinct chemostratigraphic units, referred to as CS1-3, with several subunits (i.e. CS1a, lb, 3a, 3b). The lowermost two units (CS1-2) are mostly monotonous, finely laminated dolomitic "siltstones" with low geochemical (e.g. element, ratio, factor) variability except for CS1b, where both sedimentary and geochemical variability is present. The overlying units (i.e. CS3a-3b) contains dolomitic siltstones with sandy beds that show abundant sedimentary variability, pervasive bioturbation and abundant geochemical variability. This study focuses on the lowermost units (CS la, lb, 2) of the Montney Formation in Alberta, deposited during the Triassic Induan Period and proposes a new depositional model that largely contradicts the current depositional models of a basinal turbidite system. Evidence for a new model are derived from geochemical, petrographic and SEM data. Gcochemical data (i.e. Si/Ii, As/Zr ratios, etc) suggest that the lower chemostratigraphic units (CS1-2) had minimal terrestrial influx and geostatistical analysis infers three distinct sources during the time of deposition, i.e. 1) marine benthic carbonates; 2) planktonic siliceous ooze; and 3) terrestrial material. Petrographic/SEM analyses of CS1-2 corroborate the geochemical findings revealing the dominance of authigenic minerals of quartz, dolomite, feldspars and pyrite, with few detrital grains of quartz, mica, feldspars and clay. Hence, the depositional model for the Lower Triassic Montney Formation in Alberta proposes a back-arc, semi-restricted shallow carbonate shelf with prolific benthic/planktonic shell productivity due to limited species diversity following the Permian-Triassic extinction event and limited terrestrial influx (varies locally). High benthic biogenic productivity suggests oxic bottom water conditions and possibly elevated salinity (to limit species diversity) that led to excess NaCl in the pore waters after deposition. Trace elements and total organic carbon (TOC) contents support oxic to sub-oxic sea water conditions, contradicting previous hypotheses of bottom water anoxia. Mineralogical, geochemical and organic carbon data also suggest limited influx of terrestrial and volcanic material. Since the majority of the CS1-2 sediments examined in Alberta display finely laminated dolomitic siltstones, diagenetic processes that generate dolomitic quartz-feldspar laminae from initial bedded shell debris are elaborated upon. A new proposal for the diagenetic pathway is based on findings from isolated carbonate concretions that demonstrate how in-situ porous biogenic debris (metastable aragonite, high Mg-calcite, biogenic silica and rare detrital and organic matter) can transform diagenetically during burial to "pseudo-laminated dolomitic siltstones".

Brittleness and plasticity indices in hydrocarbon reservoirs are calculated to understand how rocks behave under stress, and for assessing the fracturing performance of clay-rich shale reservoirs and assessing borehole stability. Evaluating shale plasticity/brittleness requires careful analysis of clay mineral composition in target shales and the development of fracking strategies for optimal shale stimulation. Here we report on the mineralogical variability of two Permian lacustrine shale units, the Roseneath and Murteree shales in the Cooper Basin, Australia, that are considered to have potential as unconventional hydrocarbon producers. The study involved a combination of X-ray diffraction, scanning electron microscopy and petrophysical modelling of the Roseneath and Murteree shales in order to obtain a better understanding of the compositions and microfabrics of these two units. This is part of a larger investigation of the shale gas potential of these two units in the Cooper Basin, and the results presented here may ultimately lead to improved reservoir stimulation techniques in both units. Core data has been integrated with wireline logging data to better identify brittle and plastic zones within the Roseneath and Murteree shales. Mineralogical analysis shows that both units are composed mainly of detrital quartz and clay/mica minerals with siderite cement. The clay mineral composition is dominated by illite/mica, and kaolinite in both units. However, based on the relative mineralogical differences between the two units, the Murteree Shale has more favourable brittle properties than the Roseneath Shale, and is considered to be more amenable to hydraulic fracturing for gas exploitation. However, the Roseneath Shale also has potential for gas stimulation, especially in intervals where siderite cement is prevalent.

The Devonian Duvernay Formation in Alberta, characterized as a carbonate-siliceous source rock, is ramping up to be one of the largest and most prolific shale oil plays in Canada. In the southern part of the Duvernay Shale Basin (i.e. East Shale Basin), tight limestone beds are interbedded with laminated organic-rich calcareous shales, which show an organic maturity ranging mostly from early oil- to condensate-window. This new light oil shale play is still in the initial stages of development and the nature of these deposits requires hydraulic fracturing to increase stimulated rock volume. General completion programs involve ≥50 clustered plug ‘n’ perf stages with slickwater treatments in excess of 40,000 m3 with ~4000 tonnes of proppant per well. The large water volume treatments will inevitably interact directly with the rock surface in the stimulated area and cause both oil-water and rock-water interactions. Post-hydraulic fracturing water retention is especially pronounced in light oil shale plays. The oil-wet nature of the Duvernay, along with calcareous and siliceous shale lithologies, adds to the complexity of water retention and perceived water-blockage. In addition, because of operational delays such as road bans and pipeline constraints, some wells may be shut-in after the fracturing treatment for weeks and even months, which will affect rock-oil-water behavior (i.e. production). The extent of water displacing into the matrix of the rocks of the Duvernay Formation in the East Shale Basin, as measured by load fluid recovery, varies significantly and appears to heavily rely on the choice of surfactant. Although the use of surfactants is generally accepted for this play, detailed understanding of the rock-fluid interaction mechanisms is still incomplete. This paper investigated the response of Duvernay Shale rocks from the East Shale Basin to various types of surfactants and analyzed production and fluid flowback data. Amott Cell analyses, which test for spontaneous oil displacement using various stimulation fluid types, demonstrated that in the East Shale Basin, nano-sized surfactants including multi-functional surfactants (MFS) and microemulsions significantly outperformed common surfactant chemistry when tested with mixed wettability shale core samples. The results provide an estimate as to extent of water migration into the matrix of the Duvernay as a result of the choice of surfactant. Our analysis is made possible from publicly available cores, laboratory analysis and high quality well production data from the Alberta Energy Regulator.

Over the past two decades, numerous service companies and universities have offered chemostratigraphic analysis for correlations and mechanical profiling. The advantage of XRF analysis is that data collection can be relatively fast and does not require elaborate and time-consuming preparation or extensive data modeling such as XRD mineralogy. Energy-dispersive XRF systems (ED-XRF) allow elemental measurements from a few ppm to 100% concentrations. Hence, core scanning and detailed chemostratigraphic analysis have become a common tool in the world of unconventional shale exploration and assessments. In exploration and research studies, XRF chemostratand mineral model-profiling combined with downhole lithologs appear to present an attractive package potentially providing information such as mineral phase, brittleness and/or other mechanical property changes along horizontal or vertical wells. The information is used to better identify zones with different rock properties in order to optimize well stimulations and hence production.

This study investigates petrophysical characteristics of lacustrine Permian Murteree and Roseneath shales in relation to reservoir evaluation of the most prospective gas shale plays in the Cooper Basin, Australia. Both shales were investigated for gas volumes by employing unconventional petrophysical techniques through a combination of source rock parameters acquired by geochemical analysis, and integrating the extracted parameters into log interpretation and core studies. Modeling mineralogical composition using wireline logs require the selection of a proper mineral model. In this study, the mineral model was built in the Interactive Petrophysics (IP's) Mineral Solver module by integrating all regional sedimentological, petrographic, SEM (Scanning electronic microscope), pulse decay and X-ray diffraction data (XRD) from core and chip cutting samples. This study developed a mineral grouping framework to assist in the selection of a proper model to easily solve complex shale gas reservoirs for gas volumes. Furthermore, the permeability of both shales depends on in-situ confining stress and permeability of these cores and can be calculated through decay rate of a pressure pulse applied to experimental data. Subsequent to the integrated study as explained above, it is concluded on the basis of extruded parameters (shale porosity, permeability, volume of kerogen, volume of brittle minerals and water saturation) that Murteree formation exhibits better potential than Roseneath formation in and around Nappameri, Patchawarra and Tenappera troughs, while poor potential is exhibited in the Allunga trough. The only location where Roseneath exhibits better potential is in Encounter-01 well.

The Permo-Triassic Cooper Basin is one of the largest intracratonic basins in Australia, covering approximately 130,000km2 in South Australia and Queensland. The basin is one of Australia's major onshore hydrocarbon province and most prospective region for both conventional and unconventional hydrocarbon explorations. Organic petrography and thermal maturity of two Permian lacustrine shale units in the Cooper Basin, the Murteree and Roseneath shales, were investigated on 21 wells with the objective of evaluating the gas generating potential of these units. Vitrinite reflectance values for the Murteree and Roseneath shales range between 1.17% and 2.00%. Macerals show systematic changes in properties relative to maturity rank. A range of maceral compositions, dominated by vitrinite group macerals, are present in both units, which vary between rich and very rich in organic content. Rock-Eval data suggest fair to very good kerogen quality (of kerogen types II, III, and IV ranging from immature to mature) and imply a mostly gas-prone generation potential in the shales.

To estimate the resources of Permian Roseneath and Murteree gas shales in the Cooper Basin, Australia, geochemical analysis, log interpretation and core analysis techniques were combined to conduct mineralogical modelling and define petrophysical parameters of both formations. With the sedimentologic, petrographic, SEM and XRD data derived from analysis of cores and cuttings, a mineralogical model was built for target formations. Moreover, based on the results of conventional core analysis, wireline logging, SEM analysis, XRD analysis, and geochemical and petrographic analysis, a petrophysical model was established for key wells. Then, these models were used to analyse the mineral composition and petrophysical properties of Roseneath and Murteree gas shales. The results show that both Roseneath and Murteree gas shales are composed of clay, quartz, carbonate and kerogen, as well as a small quantity of auxiliary minerals (e.g. feldspar and siderite). According to porosity, permeability, TOC, water saturation, mineral composition and other parameters, it is concluded that Murteree shale has higher potential than Roseneath shale within the basin, except for areas in and around Well Encounter 1 where Roseneath exhibit excellent potential.