Abstract Reliable interpretation and prediction of mechanical stratigraphy is the cornerstone of geomechanical investigations at multiple scales and in different applications. While lithological and petrophysical facies have been used in reservoir characterization to associate with storage capacity and flow characteristics of rocks, mechanical stratigraphy and its role is less known and more complex to identify. Mechanical stratigraphy modeling is the identification and interpretation of geomechanical rock types, which deform distinctly in response to changes in the subsurface system. These changes could occur naturally over geological time due to pore pressure buildup and plate boundary interactions or may be induced through field operations including drilling, hydraulic fracturing, water flood, production, thermal operations and depletion. The different geomechanical rock types will have a combination of lithologies and will often have overlapping petrophysical properties which together lead to variations of stiffness and strength. In this paper, the fundamentals, significance and geomechanical workflows for mechanical stratigraphy modeling are presented and discussed. In addition, evidence of mechanical stratigraphy in the subsurface for unconventional and conventional plays are shown. Applications of mechanical stratigraphy for hydraulic fracture barrier identification, well integrity, wellbore stability and subsurface containment assurance are presented. Introduction All subsurface operations such as drilling, injection and production lead to rock deformation and potentially failure in the reservoir and overburden intervals over the lifespan of a field. While some of these impacts are operationally manageable, in many cases it is critical to predict these changes by properly characterizing the geomechanical properties of the rock during the planning stage. This can help with appropriate design of operations such as injection pressure and mud weight design and mitigate potential geomechanical risks. At the regional scale, mechanical stratigraphy of the rock influences the faulting and folding over geologic time. In compressive settings, the contrast between different geomechanical rock types determines the form and extent of structural features. In extensional settings fracturing is mainly controlled by the mechanical properties of the host formation [1].

Abstract Immediately offsetting the Hydraulic Fracture Test Site II (HFTS II) were two horizontal wells that had been fracture stimulated and produced for several years. Horizontal well image logs recorded in the new HFTS II laterals were used to describe the existing hydraulic fracture geometry for reservoir simulation and to calibrate a 3D hydraulic fracture model. Similarly, in a Rocky Mountain Powder River Basin Turner development a standalone multifractured horizontal well was produced prior to drilling adjacent infill horizontal wells. Horizontal well image logs were recorded in the multiple offsetting laterals to describe the existing hydraulic fracture geometry for hydraulic fracture modeling and reservoir simulation. The image log observations were critical in defining a development plan and well spacing for the Turner formation. When standalone wells are drilled and stimulated by hydraulic fracturing, image logs recorded in the lateral of subsequent offset horizontal wells are a powerful diagnostic tool that can be used to understand hydraulic fracture geometry and to calibrate reservoir simulators. Unlike far-field fiber optic strain-rate measurements or closed-wellbore pressure monitoring, which provide a dynamic observation of hydraulic fracture growth between horizontal wells, image logs provide a static observation of hydraulic fractures that originated from a standalone or "parent" well and crossed an offset wellbore. Fracture swarms observed in an image log can often be extrapolated back to the parent well perforation cluster corresponding to the hydraulic fracture origin. Since hydraulic fractures grow as swarms, an algorithm accounting for swarm width was developed to aid in the correlation of image log hydraulic fracture swarms to perforation clusters in a parent well (or wells) to improve the confidence in the analysis. In HFTS II, 90% of the image log observed fractures had a dip angle greater than 70°, which is often indicative of hydraulic fractures, and 75% of the interpreted fractures were within the expected fracture azimuth window. Fifty-six percent of the image log fractures could be correlated back to the treatment well clusters using a 5-ft swarm window, but the correlation improved to 96% using a 0.5° azimuth window. Hydraulic fractures from both treatment wells impacted the HFTS II infill wells and having two treatment wells increases the difficulty in identifying the originating cluster for each fracture swarm. In the Turner formation, 100% of the fractures had a dip angle greater than 70° and were within the expected fracture azimuth window, which suggests all fractures were hydraulic fractures. Even though a single parent well as a source simplifies the analysis somewhat, only 55% of the image log fractures could be correlated back to discrete clusters. The correlation improved to 98% when the using stage origination as opposed to discrete cluster locations. Lastly, the image logs demonstrated conclusively that hydraulic fracture half-length was greater than 2,640 ft in all treatment stages.

Abstract Drilling and hydraulic fracture stimulation efficiency has been the focus for development of unconventional reservoirs, however, predicting and maintaining lateral well performance remain the most challenging part of these reservoirs' lifecycle. The challenges include, but are not limited to, high water production, high decline rates, casing deformation and frac hits. These challenges can be related to the fact that the formation is highly heterogeneous, reflecting variation in depositional facies, mechanical properties, occurrence of natural fractures, and stratigraphic changes within a specific formation. Surface seismic attributes are used to predict this lateral formation variation; however, it can be only used to a limited extent in understanding and predicting lateral position, assuming wells are steered in a layer cake model. The assumption of lateral homogeneity simply ignores crucial structural, tectonic, and depositional changes that occur ubiquitously on a sub seismic scale, a significant oversight that urges the need for more thorough evaluation. In this paper, we show a novel workflow applied in the Marcellus shale to evaluate lateral geological variations. First, the mechanical properties and minimum horizontal stress are evaluated with data from a slim dipole sonic tool combined with a density log. Next, the geological facies and the natural fractures are identified from oil-based high-resolution borehole image logs. These tools were conveyed through the bit, meeting the typical low-cost and low-risk exposure requirement while maintaining complete well control. Furthermore, the data was used to build a structural model and an entire discrete fracture network (DFN) model based on the image data and constrained by the output of the slim dipole 3D far-field sonic imaging. This enables the operator to investigate the extension of natural fractures up to 40 ft away from the well, a novel approach where modeling of natural fracture length and height is not based only on a statistical or stochastic process fed by assumptions. Instead we consider the continuity of the fractures from the wellbore into the far field.

Abstract Which reservoir properties are most significant in optimizing hydraulic fracture propagation? In this work, we performed statistical analyses of hydraulic fractures along a slant core drilled in the Hydraulic Fracture Test Site (HFTS1) in the Midland Basin in West Texas. We combined these data with a high-resolution 3D reservoir model constructed to find the critical factors that control fracture initiation and propagation. The data collected from the HFTS1 has enabled us to build an accurate geomodel and perform statistical analysis of the associated attributes. The 3D reservoir model incorporates geological, geophysical, geomechanical, and production data from the HFTS1, following Dommisse et al. (2018)'s approach. The centerpiece of this model is 600 feet of core taken from the Wolfcamp A and B. This core passes by two horizontal producers on the 11-well pad and contains natural and hydraulic fractures that have been analyzed. Previous works have detailed the nature of the fractures recovered in the core, identifying 375 hydraulic fractures, 309 calcite-sealed natural fractures, and 8 faults (Gale et al., 2018). We studied the likelihood for fractures to be initiated or reactivated, compared to the reservoir properties and completion processes for neighboring wells. Analysis tools include Bluback Investigator and frequentist and Bayesian statistical inference libraries. The results of this work show hydraulic fracture initiation was dependent on the distance to the well and well productivity at the HFTS1 pad is highly varied. We have found that hydraulic fractures were less likely in highly naturally fractured rock. Of the three sets of natural fractures, the oldest were never reactivated. The youngest were most likely to be reactivated due to their orientation being closest to the modern-day SHmax. We also found that, controlling for other factors, natural and hydraulic fractures of were less common in rock that was more carbonate rich. Other variables such as mechanical stratigraphy and upward fracture growth play a significant role. Fractures have grown at least 70 feet upwards (and 15 feet horizontally) from the nearest hydraulically fractured well to the slant core. HFTS1 has the data to change the way the industry understands the fracturing process. Much of the data collected at the pad has now been released. This study places the unique HFTS1 slant core in a three-dimensional geomodel. We used this to place faults in the geomodel and to extract petrophysical information from the model at each fracture. This work brings together geologic, geophysical, and petrophysical data and statistics to improve our knowledge of the subsurface. These results are applicable for the Midland basin and for tight oil and geothermal projects around the world.

Abstract Objectives: Previous work in the Midland Basin Hydraulic Fracture Test Site project (HFTS1) has shown the value of core recovered from a stimulated volume in providing information on the geometry and extent of hydraulic fractures. This project (HFTS2) is in the Wolfcamp Formation of the Delaware Basin. A slant core through the stimulated volume was acquired above and below a stimulated well and the results of the fracture description are presented here. Methods: Nine, 3-inch-diameter whole cores from the slant well were examined for fractures prior to slabbing, together with a CT scan of the core. We used criteria developed during the HFTS1 project to identify hydraulic, natural, drilling-induced and core-handling fractures. Material collected from a few fracture faces was cleaned, sieved, and examined under a microscope to look for proppant. Results: A total of 1261 fractures, including 500 hydraulic fractures, were described in 948 feet of core. Hydraulic fractures are not evenly distributed through the cores; they tend to occur in clusters. Many are remarkably planar, smooth, and featureless, while others have twist hackles, steps, plumose, and kinks. Natural, subvertical fractures, mostly less than 1 mm in width, sealed with blocky calcite, are oriented NE-SW (Set 1) and WNW-ESE (Set 2). Two Set 1 fractures have drusy cement on parted faces and were possibly open prior to drilling. A few subvertical sealed fractures with different orientations are also present. Bed-parallel natural fractures, most having fibrous cement (beef), and some showing top-to-north bed-parallel shear are present. Faults with normal/oblique movement were found in one part of the core. Drilling induced and core-handling breaks were identified in all cores, and were also noted to have reactivated some natural fractures. Significance: The results from this work will allow hydraulic fracture distribution relative to the stimulated well to be assessed. The number of fractures is generally greater than the number of perforations, but these can be compared for different stages and perforation clusters. Surface features provide information about the direction of hydraulic fracture propagation and about segmentation and bifurcation that might increase the number of fracture strands. Reactivation of natural fractures, including bedding-parallel fractures, occurs although the mechanism of parting is mostly unknown. Interdisciplinary Components: The wider HFTS2 project will make use of this fracture description to improve hydraulic fracture modeling (geomechanics and engineering), to help explain pressure depletion and fiber optics observations (reservoir depletion and frac geometry), to help verify indirect fracture diagnostic techniques such as microseismic monitoring (geophysics), and to help calibrate horizontal well image logs.

Abstract Individual well performance in the Marcellus Shale of northeastern Pennsylvania varies markedly, even in areas where the lithology, fluid composition, and completion design are consistent. A primary reason for this is the natural fracture system, which influences hydraulic fracture growth, dynamic fluid flow, reservoir pressure and stress behavior. Chief Oil and Gas (Chief) contracted Schlumberger to conduct an integrated study using an innovative modeling approach to quantify the impact of these natural fractures on field development. Working together, the team created an approach that consisted of constructing and coupling three models: a 3D geomechanical model, an unconventional fracture model (UFM), and a 3D dynamic dual-porosity model. The geomechanical model is composed of a discrete fracture network (DFN) containing both regional (J1 and J2 sets) and tectonic fractures. These are interpreted from seismic attributes (anisotropy azimuth, seismic velocity anisotropy) and ant tracking. The UFM model simulates the growth of hydraulic fractures and their interaction with natural fractures in the DFN. Portions of the natural fracture network are assumed to be open tectonic fractures, and their flow properties adjusted (porosity and permeability) to match well performance. Adjustments are also made to account for production-related perturbations in dynamic stress magnitude and azimuth, which impact later wells. These modifications to the fracture network are critical for history matching the dual-porosity model. The production history match showed that hydraulic fractures and open tectonic natural fractures are key production drivers in the study area, and that the spatial variability of the natural fracture network exerts more influence on well performance than initially thought. The connection between the hydraulic fracture network and portions of the open tectonic natural fracture system enhances parent well access to larger drainage areas. This controls the strongly variable well production observed in the study area. Subsequent stress perturbation resulting from parent well depletion is detrimental to the completion efficiency of the child wells, even with better frac designs with higher proppant loading. The modeling work also shows that the gas-in-place is consistent with volumetric and rate transient analysis (RTA) estimates. The coupling of the three models reasonably approximated changing reservoir conditions and created a nexus of domain expertise including geology, geophysics, geomechanics, stimulation, completions engineering and reservoir engineering. This enabled an understanding of the complex reservoir behavior of the naturally-fractured Marcellus Shale and generation of an optimized fit-for-purpose development plan. Chief was already implementing changes in spacing and putting more distance between offset PDP (Proved Developed Producing) wells and this study affirmed that revised development plan.

Abstract Hydraulic Fracturing Test Site-2 (HFTS-2) is a field-based research experiment performed in the Wolfcamp Formation of the Permian (Delaware) Basin. This paper focuses on integration, advanced geological characterization, and 3D subsurface modeling of the comprehensive HFTS-2 dataset. The study showcases a multidisciplinary reservoir characterization approach that incorporates geology, petrophysics, geochemistry, geomechanics, microseismic, and subsurface engineering analysis. Subsurface characterization of organic-rich mudstone formations requires understanding complex hydraulic fracture network growth in relation to inherent lithology, geomechanical properties, and interaction with pre-existing natural fractures. This paper presents a characterization workflow incorporating pre- and post-stimulation subsurface data, unique to the HFTS-2 dataset. The study integrated: (1) rock properties from logs, cores, and thin sections; (2) natural and hydraulic fracture descriptions from cores and image logs; (3) local and regional stresses; (4) geomechanics; (5) microseismic; (6) fiber optic (FO) and bottomhole pressure gauge (BHPG) response; and (7) produced fluids analysis. During a stimulation treatment, creation of the stimulated rock volume (SRV) is influenced by several subsurface factors. Key contributing factors include structural context, stress conditions, lithology, facies architecture, pre-existing natural fractures, and geomechanical properties. The HFTS-2 subsurface data integration indicates that the SRV is comprised of a complex juxtaposition of hydraulic fracture swarms, as evidenced by image logs analysis, core description, and microseismic monitoring. The HFTS-2 microseismic event density was used to generate 3D heat maps that serve as a representative SRV footprint, corroborated by secondary datasets. These maps were further integrated with petrophysical and geomechanical characteristics, as well as responses from FO and BHPG, to estimate the lateral and vertical dimensions of the effective fractures. The geological characterization for the HFTS-2 dataset combined with 3D modeling for petrophysical and geomechanical properties provides a strong foundation for subsurface simulation and optimization studies. The workflow improved our understanding of HFTS-2 hydraulic fracture propagation and characteristics in relation to offset pressure depletion and interaction with pre-existing natural fractures. Analysis showed that fracture geometry varies by stage and by well, and a complex fracture network is generated with varying fracture density. The multidisciplinary workflow presented herein for integration and characterization serves as a foundation to evaluate completion efficiency and estimate areal and vertical stimulation and depletion extent for the project. The workflow and learnings can also be transferred to other unconventional plays.

Abstract Numerical simulation results of multi-stage hydraulic fracturing in a naturally fractured reservoir with explicit representation of a realistic number and geometry of discrete natural fractures is presented. The parametric study considers different strengths and hydrological properties of DFN. The fracture paths obtained at the end of the four zipper stages stimulation are a combination of pre-existing fracture segments as well as new fractures generated in the intact rock. The stimulation responses in the DFN simulation cases are overall quite similar but consistent, with a larger fractured footprint for lower fracture friction and lower fracture cohesion. Comparison of results with and without a DFN show that both the extension of the fractured zone in the direction perpendicular to the minimum principal stress and the fracture evolution pressure are larger in the case without a DFN, as expected. Introduction Although interaction of a hydraulic fracture with a single discontinuity has been studied extensively (experimentally, analytically and numerically), the authors are not aware of a comprehensive study of multi-stage hydraulic fracturing in naturally fractured reservoirs with explicit representation of a realistic number and geometry of natural fractures in a reservoir. In this work, a physics-based numerical tool is used to investigate the effect of a discrete fracture network (DFN) on reservoir stimulation considering different mechanical and hydrological properties of the DFN. A number of numerical techniques have been developed to simulate hydraulic fracturing based on different methods such as Boundary Element Method (BEM), Finite Element Method (FEM), Extended Finite Element Method (XFEM), Finite Difference Method (FDM), Displacement Discontinuity Method (DDM), Discrete Element Method (DEM) and some hybrid methods. The formulation of some of the methods and example applications are published by Rougier et al. (2011, 2012); Zhao et al. (2015); and Lisjak et al. (2015). An overview of numerical models for modeling of coupled hydro-mechanical process in naturally fractured rock masses is provided by Lei et al. (2017). The numerical method used to conduct analyses presented in this paper is based on DEM implementation using the lattice approach.

Abstract The Diyab unconventional play is the first unconventional play to emerge in the Middle East. Reservoir characterization assessments have been made and several horizontal exploration wells have been drilled to date proving productivity of the play in the United Arab Emirates (UAE). In this paper, we apply a modeling workflow that integrates geological, geomechanical, and pre-existing completions information to optimize the future well landing points and completion designs to potentially enhance stimulated rock volume (SRV) and production. Known subsurface static properties from core-calibrated petrophysical logs and 1D geomechanical models were extrapolated across the appraisal area in a 3D static reservoir model with the addition of a discrete fracture network (DFN) derived from numerous image logs. Afterwards, completion data from three previously existing multi-stage hydraulically fracked horizontal wells was integrated into a sector model. Simulation of hydraulic fracturing have been performed and the sector model was calibrated to the real hydraulic fracturing data. This has allowed for a better understanding of the fracture propagation and stimulated rock volume in the reservoir at the various landing points. The production of the SRV has been simulated and calibrated to the field data. Lastly, sensitivities were run to design an improved well landing point and completion strategy for future horizontal wells. Applying a fully integrated geology, geomechanics, completion and production workflow has allowed us to observe the role that geology, natural fractures, rock mechanical properties and stress regimes play in the fracture propagation for the unconventional Diyab play. The model simulations show how the hydraulic fractures propagate and how they interact with pre-existing natural fractures. Sensitivity runs on the calibrated model will allow for an improved well landing point and completion strategy of future horizontal wells. This multi-disciplined approach has delineated the most crucial impacts to the hydraulic fracture stimulation job in the Diyab unconventional play onshore UAE and how they can be possibly mitigated while designing the hydraulic fracture job, thus improving SRV and ultimately production performance.

Abstract In unconventional reservoirs, the existence of natural fractures is evident from the analysis of data gathered with techniques like well logs, core studies, and micro-seismic interpretation. In hydraulic stimulation of these reservoirs, propagating hydraulic fracture (HF) generally intersects natural fractures (NF) complicating the geometry of the stimulated volume and the estimation of proppant flow and transport. The effectiveness of reservoir stimulation by hydraulic fracturing depends on flow area and proppant pack permeability; hence, proppant transport and deposition is an essential component of hydraulic fracturing stimulation. In this paper, we present a modeling study of the problem using an advanced simulator "GeoFrac-3D" that can consider irregular fracture geometries and non-orthogonal intersection between the HF and NF, thereby capturing realistic flow and proppant transport pathways and deposition sites. The proppant transport and deposition within the fractures are modeled by treating the mixture of fluid and proppant particles as a slurry. The "GeoFrac-3D" is based on the combination the displacement discontinuity method for the rock deformation and the finite element method for the fracture fluid flow and proppant transport simulation, and linear elastic fracture mechanics approach for the fracture propagation. The deformation of the NF's is implemented using a linear joint model. The proppant transport and deposition within the fractures are modeled by treating the mixture of fluid and proppant particles as slurry. Proppant settling velocities and thus proppant distribution is affected by fluid velocity, micro-proppant size, fluid rheology, fracture aperture, hydraulic and natural fracture interaction and near wellbore tortuosity. The results demonstrate that the micro-proppants (i.e., proppant particle diameter equal to or less than 100 mesh size) being smaller size particles have strong potential for the effective uniform proppant placement into the complex fractured unconventional reservoirs; hence, to increase their conductivity for the oil and gas in-flow. The numerical results demonstrate the proppant transport and deposition in the HF-NF networks strongly depend on the reservoir in-situ stress state and orientations of the natural fractures.

Abstract Distributed Strain Sensing (DSS) and Distributed Acoustic Sensing (DAS) are two emerging technologies that provide important insights into many challenging issues related to hydraulic fracturing design and operation. In particular, we highlight fracture geometry and fracture-hit inference, stage communication and perforation cluster fluid distribution. However, given the complexity associated with multiple sources of information that constitute realistic field data, its interpretation is a major challenge faced by engineers and geoscientists. In this paper, we propose a numerical modeling of DSS/DAS that is capable of incorporating the interaction phenomenon existing between hydraulic and natural fractures. The model is used in a sensitivity study varying orientation, distribution, and length of natural fractures, with the objective to assist and optimize field data interpretation in cases where such type of fracture is present. The numerical model incorporates rock and fluid mechanics associated with the hydraulic fracturing process and computes strain fields in the medium using Displacement Discontinuity Method (DDM). DDM is a type of Boundary Element Method (BEM) that applies net pressure for hydraulic fractures and shear stress for natural fractures to specify boundary conditions. In this case, the fiber gauge length concept is incorporated deriving displacement (i.e. DDM output) in space to obtain DSS values. DAS results are estimated by the differentiation of computed DSS in time. The impact of different configurations of natural fractures on strain and strain rate is investigated by a sensitivity study in which three different parameters describing natural fractures are varied: (1) orientation angle ranging from ±20° to ±60°, (2) fracture density examining 5 to 20 natural fractures per cluster, and (3) length varying from 2.8 ft to 29.6 ft. This study considers multiple perforation clusters and a cross-well fiber deployment with monitor and operation wells separated by a distance of 100 ft. Results indicate that we can capture typical characteristics present in field data: heart-shaped pattern from a fracture approaching the fiber, stress shadow and fracture hits. The incorporation of shear stress into the simulation generates DSS/DAS time histories that entail complexities beyond what is observed in cases where tensile stress is the unique failure mechanism considered. Intrinsic characteristics of natural fractures can be identified in waterfall plots depending on their orientation, distribution and length. Furthermore, the variation of such components has the potential to impact stress shadowing effect and consequently the generation of throughgoing fracture systems from different perforation clusters. The model framework captures a wide range of relevant phenomenon and provides a solid foundation for generating accurate and rich synthetic data representing multiple distinct scenarios leading to interpretation optimization. Also, the development of specific packages (commercial or otherwise) that explicitly model both DSS and DAS, incorporating the impact of fracture opening and slippage on strain and strain rate, is still in its infancy. This paper is novel in this regard and opens up new avenues of research and applications of synthetic DAS/DSS in hydraulic fracturing processes.

Abstract A physics-assisted deep learning model is presented to facilitate transfer learning in unconventional reservoirs by integrating the complementary strengths of physics-based and data-driven predictive models. The developed model uses a deep learning architecture to map formation properties to their corresponding production responses using a low-dimensional feature space representation. Transfer learning is accomplished by first training the network weights using production data from a mature shale play and combining the learned weights with limited data from a new unconventional field to generate a predictive model. The simulated data provides approximate production predictions for input parameters of the target field, for which the source data may not provide a good prediction. The resulting model has a superior performance to simulation-based and data-driven predictions alone. The results indicate that (1) physics-based simulated data can facilitate production predictions when out-of-range (unseen) input parameters have to extrapolate from data, and (2) transferring the weights learned from the source field to the target field can add valuable information to enhance the prediction performance of the target field. Introduction There has been a significant increase in the development of unconventional reservoirs in recent decades, particularly tight oil reservoirs in the United States. These tight oil formations often have extremely low permeability and are not economically exploitable using traditional drilling and completion techniques used for conventional high permeability reservoirs. Fortunately, due to the advancement in key technologies, in particular hydraulic fracturing and horizontal drilling, these low permeability shale formations can be stimulated to induce the production of hydrocarbons (King, 2010). However, the traditional flow equations that are well documented and studied for conventional wells have not been applied to tight oil reservoirs. Traditionally, the development of conventional reservoirs relies heavily on numerical reservoir simulators (Altman et al., 2020, Aziz et al., 1979). These simulators provide a reliable prediction that assists in decision-making for resource development. They are built using well-studied and trusted physics-based analytical expressions that are solved during the simulation to provide prediction responses. Since these physics-based simulation models are based on analytical equations, they have the advantage of being able to produce a prediction response for any possible range of input variables, and this allows them to easily extrapolate to any range of data. Additionally, simulation models can be run for any combination of input variables and enable the collection of large amounts of simulated data for all possible scenarios. However, the equations used in simulation models are not able to provide dependable predictions for unconventional tight oil resources since the complex physical relation of flow from tight formations and fractures along with the fracture generation is poorly understood. Additionally, there are various components in the field that are not always accounted for when building simulation models (Fung et al., 2016). These limitations make simulation models unreliable to be used to develop unconventional reservoirs.

Abstract Enhanced Geothermal Systems (EGS) have gained great attention since they promise to delivers geographically disperse, carbon-free energy with minimal environmental impact. Natural fracture networks have a major impact on EGS heat extraction. The objective of our model is to evaluate the impact of natural fracture networks on EGS producing temperature profiles. In this work, we apply the thermal embedded discrete fracture model (thermal EDFM) to handle thermal modeling of complex fracture networks. The thermal non-neighbor connections enable us to efficiently simulate temperature behavior within a massive complex fracture network. Our EGS model provides a high-resolution temperature evaluation since we consider the true flow and temperature behavior in each fracture without any upscaling. In this work, we compare EGS with different fracture network connectivities to analyze heat extraction performance. Introduction In recent years, many models have been developed for geothermal applications. TOUGH2 is a general-purpose simulator for multiphase fluid flow and heat transfer in porous media. AD-GPRS (Automatic Differentiation General Purpose Research Simulator) is widely applied to simulate steam heat transfer in the EGS using an automatic difference solution scheme. However, most EGS models cannot handle the realistic features including complex hydraulic and natural fracture networks, massive fracture numbers, and corner points due to the limited model capability and model calculation efficiency. In recent years, a novel model called the embedded discrete fracture model (EDFM) is developed to efficiently model the flow between the matrix and complex fracture network (Lee et al 2001; Li et al. 2008; Hajibeygi et al. 2011; Xu et al. 2017). Moinfar et al. (2014) extended the EDFM method to 3D simulation. Cavalante Filho et al. (2015) adapted the EDFM method to the corner-point grid. Matei et al. (2017) developed a projection-based EDFM method (pEDFM). This method automatically adjusts the matrix transmissibility field in accordance with the conductivity of the neighboring fracture networks, which extends the EDFM method applications to an impermeable flow barrier. Chai (2018) developed a compartmental EDFM (cEDFM) method by splitting the matrix blocks when intersecting fractures. This method largely improved accuracy when simulating flow across instead of along the fracture. Xu et al. (2019) extended the EDFM application to the corner point. Sun et al. (2020) applied the EDFM method to thermal modeling of distributed temperature sensing process. Sun et al. (2021) applied the thermal EDFM method to model the enhanced geothermal system with horizontal wells.

The Middle Devonian Marcellus shale play has emerged as a major world-class hydrocarbon accumulation and represents one of the largest and most prolific shale plays in the world. According to many outcrop studies in the region, natural fractures are well developed in the Marcellus Shale. However, evaluating fractures in the subsurface is often a significant challenge due to a lack of sufficient data. Therefore, in the Marcellus Shale Energy and Environment Laboratory (MSEEL) consortium project, significant efforts have been made to acquire high-quality image logs in the Marcellus laterals. The project provided tremendous opportunities to characterize the natural fractures and sub-seismic faults and to evaluate their impact on well stimulation. In this study, about 70,000 ft of acquired high-resolution logging while drilling (LWD) acoustic images from five long laterals located in Monongalia County, West Virginia, were processed and interpreted. In addition, the study used high-quality micro-resistivity images from a pilot well, allowing the evaluation of natural fractures in the entire Marcellus vertical sequence. Based on the available acoustic images, the natural fractures were classified into three basic categories: high-amplitude fractures, low-amplitude fractures, and faults. Further, larger open fractures can also be determined when a low-amplitude fracture is evident on caliper images. The fractures in the Marcellus usually have a medium to high angle dip; however, multiple fracture sets in terms of strike orientation were clearly observed in all the laterals. The fracture set with a strike at NE-SW (or 60-240 deg) seems to be the predominant one in all the wells. A few other sets, including those with N-S, NWW-SEE, and E-W strikes, were also observed. Several sub-seismic faults, with mostly a low dip angle and a NE-SW strike, have also been seen in two of the laterals. The fracture density is variable across all the laterals, ranging from very low (or none) to very high (up to 5 fractures per ft). The average fracture density for all the laterals is about 1 fracture per 10 ft. In the vertical sequence, the natural fracture development showed a clear preference for shale or shaly facies over carbonate-rich or thin limestone layers. The interpreted fracture and fault data were used as input data for the stimulation design with the purpose of better understanding the fractures' impact on well stimulation. Production data from the laterals were also used to evaluate the natural fractures’ influence on well performance. The quality image database and the consistent interpretation results for the entire project enabled a systematic approach to characterizing fractures and, more importantly, to evaluating the impact of fractures on well stimulation and production.

The Upper Triassic Xujiahe formation is very tight formation located in highly tectonically stressed zones in the Western Sichuan Basin of China. The second and fourth members (abbreviated as T3X2 and T3X4, respectively) of the Xujiahe formation are gas-bearing tight sandstones. Bedding planes and natural fractures with low deviations are developed in the sandstones. This tectonically active in-situ stress state and the low angle natural fractures make the hydraulic fracture to propagate either horizontally or vertically but with small fracture height, which are unfavorable for enhancing gas production. Multidisciplinary data from the field and experiments were used to determine the in-situ stress state. Minifrac test data were analyzed to determine the minimum horizontal stress. The FMI logging and stress polygon method were used to constrain the maximum horizontal stresses. The differential strain analysis was also performed to measure the principal stresses. A series of laboratory tests were conducted to determine the influences of in-situ stresses and natural fractures on hydraulic fracture propagations. The cohesions and internal friction angles of the bedding cracks were obtained by direct shearing tests using the downhole cores of T3X2. To imitate preexisting low angle fractures in the laboratory tests, man-made fractures were cut and then cemented in the outcrop rock samples, and the true tri-axial fracturing experiments were conducted in these samples to analyze the effects of in-situ stress state on the hydraulic fracture propagation in the tight sandstone. The following are the results obtained in this paper: (1) The T3X2 reservoirs are in the strike-slip faulting stress state, which is beneficial to generate vertical hydraulic fractures; while the T3X4 reservoirs are close to the reverse faulting stress state where the hydraulic fracture height might be lower than that in the T3X2 reservoirs; (2) The true tri-axial fracturing experiments under the conditions of the in-situ stress states of the T3X2 and T3X4 reservoirs show that the hydraulic fracture geometry is mainly composed of the vertical fractures accompanying with few horizontal fractures. When the ratio of the overburden stress to the minimum horizontal stress is less than 1.2, hydraulic fracture opens or slides along the low angle natural fracture and may form a limited fracture height as a result; (3) The hydraulic fracturing target should be selected at the intervals that have the lowest horizontal stress, and the intervals having low angle natural fractures should be avoided. The pad fluid with high viscosity guar fluid is recommended to use to initiate the main vertical hydraulic fracture. This study provides an integrated method and a guide for evaluating and designing the stimulation of the Xujiahe tight gas reservoirs in the Western Sichuan Basin. This method may be applicable to similar reservoirs.

An innovative simulation methodology was developed that incorporates fine scale features of complex reservoir geology, the details of hydraulic and natural fracture networks, and compositional modeling to account for the injected propane-oil interaction. The simulation models were used as an operational tool to make key design and operational decisions in the Bakken Propane Injection EOR field test and were critical to the success of the test. Multiple reservoir model realizations were built using a fine simulation grid, dual porosity/dual permeability option, multi-component compositional fluid description, and robust representations of the hydraulic and natural fractures. Geo-mechanical simulation was applied for hydraulic fracture representation in the individual stages of horizontal and vertical wells. A software tool was developed for the hydraulic fracture representation in the structured simulation grid matching geometry, porosity, and conductivity of the geo-mechanical simulation. A discrete fracture network (DFN) was used to represent the natural fractures. Each model was history matched with all the production and pressure data obtained during the pre-propane injection period. The models were updated continually with the field test surveillance data and were used to forecast production phase oil and gas rates and propane concentrations in the produced gas. The model predictions guided field operational strategies optimizing propane injection volumes and soaking duration. The effective optimization methodology was applied for optimizing operational strategies. The simulation predictions of incremental oil, water, and gas production rates and the propane concentration in the produced gas were corroborated by field data, indicating the robustness of the simulation methodology. Novel methodologies were used for representing hydraulic fractures in the structured grid simulation model and for optimizing field enhanced oil recovery (EOR) operational strategies. These features enhanced the robustness and predictive capability of the reservoir simulation workflow. Validation of the modeling workflow with field EOR test data has given us the confidence to apply it for design and operation of upcoming field pilots.

Distributed temperature sensing (DTS) is an enabling technology for fracture diagnosis and multiphase flow measurement in unconventional areas. Fracture characterization and flow profiling are crucial to evaluate the performance of hydraulic fractures. The objective of our DTS data analysis approach is to provide a high-resolution quantitative diagnosis of hydraulic and natural fractures with the full-physics model, which will benefit the fracturing operation design and decision-making process in the unconventional reservoir. In this work, we developed a comprehensive numerical forward model for DTS data analysis. Our model includes reservoir and wellbore models. Also, the flow and thermal models are fully coupled. A thermal embedded discrete fracture model (TEDFM) is developed to handle the thermal modeling of complex fracture networks. Subsequently, we developed ensemble smoother with multiple data assimilation (ESMDA) as the inverse model to match field data and characterize fractures. ESMDA can handle large volume of data and model parameter sets with acceptable computational time. The DTS analysis with our model provides a high-resolution solution since the fracture diagnosis and flow profiling are performed for each fracture. The hydraulic and natural fracture properties and geometry such as fracture half-length, height, and fracture conductivity are evaluated. With this analysis, we obtain a deeper understanding of the effectiveness of the field hydraulic fracturing operation. Although numerous simulators are developed for DTS data analysis, relatively few existing models can handle the full-physics such as complex fracture geometry and multiphase flow. Our inverse model provides an improved DTS data match result. Our model is more rigorous than the prior models to simulate and match the field DTS data. Introduction The key technology for shale gas reservoir is hydraulic fracturing: pressurized fracturing fluid containing water, sand, and other proppants is injected into a wellbore to create cracks in the formation. The high contact area of the fracture surface enables high oil and gas production. However, in the field applications, usually, 20 percent of fractures contribute 80 percent of production. The reason behind this is a lack of understanding of fracture geometry and fracture property. Due to the pre-existing natural fractures, the complex fracture network may be generated during the hydraulic fracturing operations. (Maxwell et al. 2002; Fisher et al. 2004; Cipolla et al. 2010; Cipolla et al 2011)

The existence of natural fractures in tight reservoirs causes great uncertainty in the infill-well completion. However, it is difficult to quantify the effects of natural fractures on stress evolution and frac hits due to the stochastic manner of the natural fracture system. In this work, we investigate the impact of two types of lateral complexity, i.e., parallel fracture complexity and transverse fracture complexity, on the stress redistribution and propose suggestions to mitigate frac hits during interwell fracturing. An in-house 3D coupled geomechanics and compositional simulator is used to model the pressure and stress distribution, followed by a displacement discontinuity method hydraulic fracture model to simulate the infill-well fracture propagation. Numerical results show that the parallel natural fracture serves as an attractor of stress reorientation, whereas the transverse natural fracture acts as a disperser of the orientation change. The existence of the parallel fracture complexity induces significant local stress heterogeneity around the natural fracture tip; thus, parallel rather than staggered perforation of the infill-well is favored to reduce the risk of frac hits. Since the transverse fracture complexity leads to more uniform stress reorientation between parent-well and infill-well, the perforation location is not as important as that in the parallel complexity case. By uncovering the fundamental mechanisms of lateral fracture complexity on frac hits, this work provides some insights into the interwell fracturing of tight reservoirs and will facilitate the on-site hydrocarbon production. Introduction Field observations have shown the complex growth of hydraulic fractures in both lateral and vertical directions (Cipolla et al., 2008; Soliman et al. 2010; Wilson, 2015). The fracture network complexity is generally induced by the interaction between hydraulic fractures and pre-existing natural fractures, fissures, or faults (Cruz et al., 2017; Wang et al., 2020). One possible situation is that the hydraulic fractures propagate in parallel with microfractures (Nguyen and Fleming, 2012; Sierra, 2016), as shown in Figure 1a . Most small parallel branches are connected to the horizontal wellbore but do not communicate with the hydraulic fractures. Another common practice is the interconnected fracture system (Cipolla et al., 2008; Warpinski et al., 2009) created by hydraulic fractures crosscutting natural fractures, as described in Figure 1b . Moreover, the natural fractures in tight formations are found highly clustered in most cases (Gale et al., 2014; Forand et al., 2017). Based on seismic interpretation and image logs, the complex natural fractures can be grouped into several dominant sets before we run reservoir simulations.

Water flowback behaviors of unconventional gas reservoirs have been studied for gaining useful insights pertinent to fracture geometry characterization after fracturing stimulation. Field observations indicate that flowback profiles, such as fluid flow rates, flowback time, fracturing fluid recovery, may vary significantly among different wells. Stimulated fracture geometry has been widely postulated to be a dominant factor influencing flowback characteristics. This study attempts to identify the flowback signatures under various fracture geometries and configurations and to investigate effects of stimulated fracture geometry on flowback characteristics. In this study, an in-house three-dimensional coupled fluid flow and geomechanics simulator is used to simulate the flowback behaviors in unconventional gas reservoirs. The developed model incorporates essentially all the relevant physical mechanisms controlling flowback characteristics: capillarity, fracture closure, stress-dependent reservoir properties, and gravity segregation. Various complex fracture geometries are efficiently modeled through the Embedded Discrete Fracture Model (EDFM). Geomechanical deformations of the hydraulic and natural fractures are quantified based on widely-adopted fracture constitutive models. The flowback signatures under various fracture geometries are characterized by plotting different diagnostic plots. Fluid phase and pressure distributions are presented to help explain the mechanisms controlling observed characteristics. Two prominent regimes are observed for the whole flowback process: two-phase transient flow in fractures (early flowback period, EFP) and transient linear flow with considerable gas influx from matrix to fractures (late flowback period, LFP). Semi-log plots of water rate vs. cumulative water production (RVP) show near straight-line trends in EFP, and the slope of this straight-line increases as the total fracture volume increases. The curve moves left downward as the fracture interface increases under the same fracture volume. Complex fracture geometry leads to lower water recovery due to more water trapping in fractures. The existence of gas-filled natural fractures increases both the gas rate and ultimate water recovery, characterized by high gas water ratio (GWR) and a large slope in RVP plot during EFP. This study examines the salient signatures in flowback profiles for a variety of fracture geometries. The outcomes have offered new insights for the effect of fracture geometry on water flowback, facilitating the reduction in uncertainty of the stimulated fracture characterization using flowback data.

The Diyab Reservoir is an unconventional prospect in Abu Dhabi where previous exploration efforts revealed thin low permeability targets, unsuccessful stimulation, and poor well placement. This early field experience coupled with limited knowledge about local in situ stress and the impact of faults and subseismic fractures to affect hydrofracs led to a concentrated field data acquisition program together with a fully integrated 3D reservoir simulation approach to assess Producibility, Fracability and Geohazards in the Diyab. The fit-for-purpose field data acquisition program, 3D hydraulic fracture simulation, and reservoir flow modeling helped establish a baseline understanding of the relationship between 3D reservoir characterization and proppant transport. Seismic data was interpreted to understand structural components and natural fracture characterization. Elastic inversions were performed on this data using petrophysical models to populate high-resolution 3D geological and geomechanical models calibrated against log derived rock properties and in situ tests. A detailed analysis of core identified bitumen layers and thinly laminated mudstones that have the ability to undermine completions by causing horizontal hydraulic fracture growth and undesirable proppant migration. Physics-driven frac simulation of this fully integrated geomodel was performed to determine design completion and fracking strategies for the target reservoirs. Post frac analysis for three Diyab wells reveals that the ISIP varied significantly among the stages and the largest post frac pressure drops occurred at stages intersected by small-scale faults or natural fracture zones, particularly when they are well oriented for shear slip. Results of high-resolution 3D simulation of frac stages on earlier wells showed proppant distribution closely followed reservoir property distributions of low stress and similar Young’s Modulus values. Wells were landed in specific target reservoirs, however, the simulation demonstrated that proppant from some stages was inadvertently placed in overlying reservoirs. The natural fractures play a significant role in stage efficiency indicating the need to utilize non-geometric completion design. Simulating the role of natural fractures to create reservoir access indicated significant differences in propped height and length within naturally fractured stages. Dual permeability pre-frac reservoir modeling based on frac simulation results predicted cumulative gas rates for the new wells.