Timothy R. Carr

Organic-rich Marcellus Shale lithofacies modeling and distribution pattern analysis in the Appalachian Basin

The Marcellus Shale is considered to be the largest unconventional shale-gas resource in the United States. Two critical factors for unconventional shale reservoirs are the response of a unit to hydraulic fracture stimulation and gas content. The fracture attributes reflect the geomechanical properties of the rocks, which are partly related to rock mineralogy. The natural gas content of a shale reservoir rock is strongly linked to organic matter content, measured by total organic carbon (TOC). A mudstone lithofacies is a vertically and laterally continuous zone with similar mineral composition, rock geomechanical properties, and TOC content. Core, log, and seismic data were used to build a three-dimensional (3-D) mudrock lithofacies model from core to wells and, finally, to regional scale. An artificial neural network was used for lithofacies prediction. Eight petrophysical parameters derived from conventional logs were determined as critical inputs. Advanced logs, such as pulsed neutron spectroscopy, with log-determined mineral composition and TOC data were used to improve and confirm the quantitative relationship between conventional logs and lithofacies. Sequential indicator simulation performed well for 3-D modeling of Marcellus Shale lithofacies. The interplay of dilution by terrigenous detritus, organic matter productivity, and organic matter preservation and decomposition affected the distribution of Marcellus Shale lithofacies distribution, which may be attributed to water depth and the distance to shoreline. The trend of normalized average gas production rate from horizontal wells supported our approach to modeling Marcellus Shale lithofacies. The proposed 3-D modeling approach may be helpful for optimizing the design of horizontal well trajectories and hydraulic fracture stimulation strategies.

Effects of soil‐moisture content on shallow‐seismic data

Repeated shallow-seismic experiments were conducted at a site on days with different near-surface moisture conditions in unconsolidated material. Experimental field parameters remained constant to ensure comparability of results. Variations in the seismic data are attributed to the changes in soil-moisture content of the unconsolidated material. Higher amplitudes of reflections and refractions were obtained under wetter near-surface conditions. An increase in amplitude of 21 dB in the 100-300 Hz frequency range was observed when the moisture content increased from 18% to 36% in the upper 0.15 m (0.5 ft) of the subsurface. In the time-domain records, highly saturated soil conditions caused large-amplitude ringy wavelets that interfered with and degraded the appearance of some of the reflection information in the raw field data. This may indicate that an intermediate near-surface moisture content is most conducive to the recording of high-quality shallow-seismic reflection data at this site. This study illustrates the drastic changes that can occur in shallow-seismic data due to variations in near-surface moisture conditions. These conditions may need to be considered to optimize the acquisition timing and parameters prior to collection of data.

Marcellus Shale Lithofacies Prediction by Multiclass Neural Network Classification in the Appalachian Basin

Marcellus Shale is a rapidly emerging shale-gas play in the Appalachian basin. An important component for successful shale-gas reservoir characterization is to determine lithofacies that are amenable to hydraulic fracture stimulation and contain significant organic-matter and gas concentration. Instead of using petrographic information and sedimentary structures, Marcellus Shale lithofacies are defined based on mineral composition and organic-matter richness using core and advanced pulsed neutron spectroscopy (PNS) logs, and developed artificial neural network (ANN) models to predict shale lithofacies with conventional logs across the Appalachian basin. As a multiclass classification problem, we employed decomposition technology of one-versus-the-rest in a single ANN and pairwise comparison method in a modular approach. The single ANN classifier is more suitable when the available sample number in the training dataset is small, while the modular ANN classifier performs better for larger datasets. The effectiveness of six widely used learning algorithms in training ANN (four gradient-based methods and two intelligent algorithms) is compared with results indicating that scaled conjugate gradient algorithms performs best for both single ANN and modular ANN classifiers. In place of using principal component analysis and stepwise discriminant analysis to determine inputs, eight variables based on typical approaches to petrophysical analysis of the conventional logs in unconventional reservoirs are derived. In order to reduce misclassification between widely different lithofacies (for example organic siliceous shale and gray mudstone), the error efficiency matrix (ERRE) is introduced to ANN during training and classification stage. The predicted shale lithofacies provides an opportunity to build a three-dimensional shale lithofacies model in sedimentary basins using an abundance of conventional wireline logs. Combined with reservoir pressure, maturity and natural fracture system, the three-dimensional shale lithofacies model is helpful for designing strategies for horizontal drilling and hydraulic fracture stimulation.

Identifying organic-rich Marcellus Shale lithofacies by support vector machine classifier in the Appalachian basin

Unconventional shale reservoirs as the result of extremely low matrix permeability, higher potential gas productivity requires not only sufficient gas-in-place, but also a high concentration of brittle minerals (silica and/or carbonate) that is amenable to hydraulic fracturing. Shale lithofacies is primarily defined by mineral composition and organic matter richness, and its representation as a 3-D model has advantages in recognizing productive zones of shale-gas reservoirs, designing horizontal wells and stimulation strategy, and aiding in understanding depositional process of organic-rich shale. A challenging and key step is to effectively recognize shale lithofacies from well conventional logs, where the relationship is very complex and nonlinear. In the recognition of shale lithofacies, the application of support vector machine (SVM), which underlies statistical learning theory and structural risk minimization principle, is superior to the traditional empirical risk minimization principle employed by artificial neural network (ANN). We propose SVM classifier combined with learning algorithms, such as grid searching, genetic algorithm and particle swarm optimization, and various kernel functions the approach to identify Marcellus Shale lithofacies. Compared with ANN classifiers, the experimental results of SVM classifiers showed higher cross-validation accuracy, better stability and less computational time cost. The SVM classifier with radius basis function as kernel worked best as it is trained by particle swarm optimization. The lithofacies predicted using the SVM classifier are used to build a 3-D Marcellus Shale lithofacies model, which assists in identifying higher productive zones, especially with thermal maturity and natural fractures.

The application of improved NeuroEvolution of Augmenting Topologies neural network in Marcellus Shale lithofacies prediction

The organic-rich Marcellus Shale was deposited in a foreland basin during Middle Devonian. In terms of mineral composition and organic matter richness, we define seven mudrock lithofacies: three organic-rich lithofacies and four organic-poor lithofacies. The 3D lithofacies model is very helpful to determine geologic and engineering sweet spots, and consequently useful for designing horizontal well trajectories and stimulation strategies. The NeuroEvolution of Augmenting Topologies (NEAT) is relatively new idea in the design of neural networks, and shed light on classification (i.e., Marcellus Shale lithofacies prediction). We have successfully enhanced the capability and efficiency of NEAT in three aspects. First, we introduced two new attributes of node gene, the node location and recurrent connection (RCC), to increase the calculation efficiency. Second, we evolved the population size from an initial small value to big, instead of using the constant value, which saves time and computer memory, especially for complex learning tasks. Third, in multiclass pattern recognition problems, we combined feature selection of input variables and modular neural network to automatically select input variables and optimize network topology for each binary classifier. These improvements were tested and verified by true if an odd number of its arguments are true and false otherwise (XOR) experiments, and were powerful for classification.

Comparison of ground-penetrating radar response and rock properties in a sandstone-dominated incised valley-fill deposit

Alex Martinez*, Exxon Exploration Company, Houston, TX (formerly Kansas Geological Survey, Lawrence, KS); Alan P. Byrnes, Kansas Geological Survey, Lawrence, KS; D. Scott Beaty, Chevron USA, Midland, TX formerly Kansas Geological Survey, Lawrence, KS); Timothy R. Carr, Kansas Geological Survey, Lawrence, KS; and James M. Stiles, Department of Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS

Integrated core-log petrofacies analysis in the construction of a reservoir geomodel: A case study of a mature Mississippian carbonate reservoir using limited data

Small independent operators produce most of the Mississippian carbonate fields in the United States mid-continent, where a lack of integrated characterization studies precludes maximization of hydrocarbon recovery. This study uses integrative techniques to leverage extant data in an Osagian and Meramecian (Mississippian) cherty carbonate reservoir in Kansas. Available data include petrophysical logs of varying vintages, limited number of cores, and production histories from each well. A consistent set of assumptions were used to extract well-level porosity and initial saturations, from logs of different types and vintages, to build a geomodel. Lacking regularly recorded well shut-in pressures, an iterative technique, based on material balance formulations, was used to estimate average reservoir-pressure decline that matched available drillstem test data and validated log-analysis assumptions.Core plugs representing the principal reservoir petrofacies provide critical inputs for characterization and simulation studies. However, assigning plugs among multiple reservoir petrofacies is difficult in complex (carbonate) reservoirs. In a bottom-up approach, raw capillary pressure (Pc) data were plotted on the Super-Pickett plot, and log- and core-derived saturation-height distributions were reconciled to group plugs by facies, to identify core plugs representative of the principal reservoir facies, and to discriminate facies in the logged interval. Pc data from representative core plugs were used for effective pay evaluation to estimate water cut from completions, in infill and producing wells, and guide-selective perforations for economic exploitation of mature fields.The results from this study were used to drill 22 infill wells. Techniques demonstrated here can be applied in other fields and reservoirs.

Schaben field, Kansas: Improving performance in a Mississippian shallow-shelf carbonate

Schaben field (Kansas), located along the northeastern shelf of the Hugoton embayment, produces from Mississippian carbonates in erosional highs immediately beneath a regional unconformity. Production comes from depths of around 4400 ft (1342 m) in partially dolomitized shelf deposits. A detailed reservoir characterization/simulation study, recently performed as part of a Department of Energy Reservoir Class Oil Field Demonstration Project, has led to important revision in explanations for observed patterns of production. Cores recovered from three new data wells identify three main facies: spicule-rich wackestone-packstone, echinoderm wackestone/packstone/grainstone, and dolomitic mudstone-wackestone. Reservoir quality is highest in spicule-rich wackestone/packstones but is subject to a very high degree of vertical heterogeneity due to facies interbedding, silicification, and variable natural fracturing. The oil reservoir is underlain by an active aquifer, which helps maintain reservoir pressure but supports significant water production. Reservoir simulation, using public-domain, PC-based software, suggests that infill drilling is an efficient approach to enhanced recovery. Recent drilling directed by simulation results has shown considerable success in improving field production rates. Results from the Schaben field demonstration project are likely to have wide application for independent oil and exploration companies in western Kansas.

Seismic Attributes Application for the Distributed Acoustic Sensing Data for the Marcellus Shale: New Insights to Cross-Stage Flow Communication

Recently, oil and gas companies started to invest in fiber optic technology to remotely monitor subsurface response to stimulation. Distributed Acoustic Sensing (DAS) and Distributed Temperature Sensing (DTS) record vibration and temperature around the fiber, respectively. In this research, we introduce new seismic attributes calculated from the DAS data that could suggest cross-stage fluid communication during hydraulic fracturing. The DAS data covers the entire 28 stimulated stages of the lateral MIP-3H well close to Morgantown, WV. We calculated the energy attribute for the DAS data of the studied stages. Subsequently, a Hilbert transform is applied to the DAS data to evaluate the instantaneous frequency of each trace in the DAS. In addition, we applied a fast Fourier transform to each trace for all the SEGY files to calculate the dominant frequency with a 30 second temporal window. The dominant frequency is compared to the DTS data and energy attribute for the stages in the horizontal MIP-3H well. The DTS analysis shows that stimulation of the stages 10 causes a temperature rise in the previous stage 9; in contrast, stage 18 stimulation does not affect stage 17 temperature. We suggest that the common low frequency zone identified in instantaneous frequency and dominant frequency attributes between stages 10 and 9 is related to presence of fluid and gas that transferred cross-stage during hydraulic fracturing. The fluid and results in the frequency damping of the vibrations around the fiber. We show that the frequency attribute reveals increases detail about the stimulation than conventional signal energy attribute of the DAS data.

Coupled laboratory and field investigations resolve microbial interactions that underpin persistence in hydraulically fractured shales

Significance Microorganisms persisting in hydraulically fractured shales must maintain osmotic balance in hypersaline fluids, gain energy in the absence of electron acceptors, and acquire carbon and nitrogen to synthesize cell building blocks. We provide evidence that that cofermentation of amino acids (Stickland reaction) meets all of these organismal needs, thus functioning as a keystone metabolism in enriched and natural microbial communities from hydraulically fractured shales. This amino acid-based metabolic network can be rationally designed to optimize biogenic methane yields and minimize undesirable chemistries in this engineered ecosystem. Our proposed ecological framework extends to the human gut and other protein-rich ecosystems, where the role of Stickland fermentations and their derived syntrophies play unrecognized roles in carbon and nitrogen turnover. Hydraulic fracturing is one of the industrial processes behind the surging natural gas output in the United States. This technology inadvertently creates an engineered microbial ecosystem thousands of meters below Earth’s surface. Here, we used laboratory reactors to perform manipulations of persisting shale microbial communities that are currently not feasible in field scenarios. Metaproteomic and metabolite findings from the laboratory were then corroborated using regression-based modeling performed on metagenomic and metabolite data from more than 40 produced fluids from five hydraulically fractured shale wells. Collectively, our findings show that Halanaerobium, Geotoga, and Methanohalophilus strain abundances predict a significant fraction of nitrogen and carbon metabolites in the field. Our laboratory findings also exposed cryptic predatory, cooperative, and competitive interactions that impact microorganisms across fractured shales. Scaling these results from the laboratory to the field identified mechanisms underpinning biogeochemical reactions, yielding knowledge that can be harnessed to potentially increase energy yields and inform management practices in hydraulically fractured shales.