Summary The reservoir geomechanics simulator (RGMS or RGM simulator), a geomechanics simulator based on the finite element method and parallelized using the Message Passing Interface (MPI), is developed in this work to model the stresses and deformations in subsurface systems. RGMS can be used standalone or coupled with flow and transport models. pTOUGH+HYDRATE (pT+H) V1.5, a parallel MPI-based version of the serial TOUGH+HYDRATE (T+H) V1.5 code that describes mass and heat flow in hydrate-bearing porous media, is also developed. Using the fixed-stress split iterative scheme, RGMS is coupled with the pT+H V1.5 to investigate the geomechanical responses associated with gas production from hydrate accumulations. In the second paper of this series, the parallel performance of the codes is tested on the Texas A&M University Ada Linux cluster using up to 512 processes and on a Mac Pro computer with 12 processes. The investigated problems are: Group 1: Geomechanical problems solved by RGMS in 2D Cartesian and cylindrical domains and a 3D problem, involving 4×106 and 3.375×106 elements, respectively; Group 2: Realistic problems of gas production from hydrates using pT+H V1.5 in 2D and 3D systems with 2.45×105 and 3.6× 106 elements, respectively; Group 3: The 3D problems in Group 2 solved with the coupled RGMS-pT+H V1.5 simulator, fully accounting for geomechanics. Two domain partitioning options are investigated on the Ada Linux cluster and the Mac Pro, and the code parallel performance is monitored. On the Ada Linux cluster using 512 processes, the simulation speedups (a) of RGMS are 218.89, 188.13, and 284.70 in the Group 1 problems, (b) of pT+H V1.5 are 174.25 and 341.67 in the Group 2 cases, and (c) of the coupled simulators are 134.97 and 331.80 in the Group 3 cases. The results produced in this work show the necessity of using full geomechanics simulators in marine hydrate-related studies because of (a) the associated pronounced geomechanical effects on production and displacements and (b) the effectiveness of the parallel simulators developed in this study, which can be the only realistic option in these complex simulations of large multidimensional domains.

Summary Steady-state relative permeability experiments are performed by coinjection of two fluids through core plug samples. The relative permeabilities can be calculated using Darcy’s law from the stabilized pressure drop and saturation of the core if capillary end effects and transient effects are negligible. In most cases, such conditions are difficult to obtain. Recent works have presented ways to extrapolate steady-state pressure drop and average saturation measurements affected by capillary end effects collected at different rates to obtain correct relative permeabilities at correct saturations. Both the considered methods are based on linear extrapolations to determine intercepts. Gupta and Maloney (2016) derived their method intuitively and validated it with numerical and experimental data. Andersen (2021a) derived a method from fundamental assumptions and presented an intercept method in a different form where the saturation and relative permeabilities are found directly and uniquely from straightline intercepts. All system parameters, including saturation functions and injection conditions, appear in the model. In this work, the two methods are compared. It is proven theoretically that Gupta and Maloney’s method is correct in that it produces the correct saturation and pressure drops corrected for capillary end effects. Especially, a constant pressure drop was assumed and here proved to exist, as a result of capillary end effects in addition to the Darcy law pressure drop with no end effects. Their method assumes a well-defined end effect region with length xCEE, but this length can be defined almost arbitrarily. This choice has little impact on average saturation and pressure drop, however. They also assumed that for a defined end effect region, the average saturation was constant and equal to the slope in their saturation plot. It is shown that if the region is defined, the average saturation is indeed constant, but not given by the slope. The correct slope is predicted by the Andersen model. We also comment on theoretical misinterpretations of the Gupta and Maloney method. A few works have correctly calculated that the pressure drop over the end effect region is independent of rate, but not accounted for that its length is rate dependent. We show that the combined pressure drop is equal to a constant plus the Darcy pressure drop over the full core. Examples are presented to illustrate the model behaviors. Literature datasets are investigated showing that (a) apparently rate-dependent CO2-brine relative permeability endpoints can be explained by capillary end effects and (b) the intercept methods can be applied to correct shale relative permeabilities.

Summary When reservoir fluids are confined by nanoscale pores, pronounced changes in fluid properties and phase behavior will occur. This is particularly significant for the natural gas huff ‘n’ puff (HNP) process as a means of enhanced oil recovery (EOR) technology in unconventional reservoirs. There have been considerable scientific contributions toward exploring the EOR mechanisms, yet almost none considered the effects of nanopore confinement and its proportion on the oil recovery dynamics. To bridge this gap, we developed an approach to calculate fluid phase equilibrium in nanopores by modifying the Rachford-Rice equation and Peng-Robinson equation of state (PR-EOS), completed by considering the shifts of fluid critical properties and oil/gas capillary pressure. Afterward, the effect of nanopore radius (rp) on the phase behavior between the injected natural gas and oil was thoroughly investigated. Compositional simulation was performed using a rigorously calibrated model based on typical properties of a tight reservoir to investigate the production response of natural gas HNP, including the effects of nanopore confinement and its proportion. We demonstrated that the critical pressure and temperature of fluid components decreased with the reduction in rp, especially for heavy constitunts. The saturation pressure, density, and viscosity of the oil in the presence of natural gas all declined linearly with 1/rp in the confined space. The suppression of fluid saturation pressure was indicative of an extended single-phase oil flow period during production. The cumulative oil production was approximately 12% higher if the confinement effect was considered in simulation. Moreover, the average reservoir pressure declined rapidly resulting from this effect, mainly caused by the intensified in-situ gas/oil interaction in nanopores. The results of this paper supplement earlier findings and may advance our understanding of nanopore confinement during natural gas HNP, which are useful for field-scale application of this technique.

Summary Multifrequency complex electric measurements are influenced by the joint effects of mineralogy, spatial distribution of fluids and solids, properties of fluids and minerals, and interfacial polarization mechanisms. The interpretation of multifrequency electrical conductivity and dielectric permittivity dispersion can yield information about the dominant dielectric polarization mechanisms. The dielectric polarization mechanisms detected in sedimentary rocks are linked to grain size, shape, and orientation relative to the externally applied electric field, electrochemical adsorption/desorption complexation reactions, wettability, porosity, and fluid saturations. In this work, we introduce a new workflow for multifrequency interpretation of complex permittivity measurements. The introduced workflow enables simultaneous assessment of dominant wettability, porosity, and hydrocarbon reserves. These properties are estimated by minimizing a cost function using a combination of downhill gradient-descent and an evolutionary algorithm. The cost function quantifies the difference between measured and numerically calculated multifrequency complex dielectric permittivity through a recently introduced rock-physics model. We successfully applied the new interpretation workflow to multifrequency electrical resistivity/permittivity measurements obtained from two water-wet and two hydrocarbon-wet sandstone core samples. Simultaneous estimates of wettability, water saturation, and porosity obtained by applying the new interpretation workflow were in agreement with experimental measurements. Assessment of hydrocarbon reserves, porosity, and dominant wettability solely from the interpretation of multifrequency complex-valued electrical measurements is a unique contribution of the introduced workflow. Additionally, the new workflow honors rock fabric and, thus, it minimizes the necessity of extensive calibration efforts. All the parameters needed as inputs to the new interpretation workflow are associated with physical mechanisms at microscopic- and pore-scale domains or realistic and quantitative pore geometry features of the rock.

Summary A novel technique to predict porosity from mechanical specific energy (MSE), as measured using a downhole drilling dynamics tool, is proposed. The method is based on the fundamental conservation of energy principle and uses rock physics to establish a relationship between the drillability of a rock, its lithological properties, and its porosity. This new technique is particularly useful for formation evaluation of carbonate reef reservoirs drilled using the pressurized mud cap drilling (PMCD) technique. Under these conditions, the near-wellbore formation is often deeply invaded by drilling fluids and cuttings, which make interpretation of conventional logs unreliable and difficult. Since drilling measurements are relatively unaffected by this invasion, porosity determinations from MSE can provide a valuable supplement to conventional logging suites.

Summary The Reservoir GeoMechanics Simulator (RGMS or RGM simulator), a geomechanics simulator based on the finite element method (FEM) and parallelized using the Message Passing Interface (MPI), is developed in this work to model the stresses and deformations in subsurface systems. RGMS can be used standalone or coupled with flow and transport models. pTOUGH+ HYDRATE (pT+H) V1.5, a parallel MPI-based version of the serial TOUGH+HYDRATE (T+H) V1.5 code that describes mass and heat flow in hydrate-bearing porous media, is also developed. Using the fixed-stress split iterative scheme, RGMS is coupled with the pT+H V1.5 codes to investigate the geomechanical responses associated with gas production from hydrate accumulations. In the first paper of this series, we discuss the governing equations underlying physics and their mathematical representation in the modeling of the geomechanics, methane hydrate, and coupled problems as well as the numerical methods and the parallelization processes (involving a domain decomposition method based on the MPI approach) used for the parallel simulators. Two 2D problems (in Cartesian and radial-cylindrical coordinates) and a 3D Cartesian coordinate problem are created to validate the FEM and the parallelization method in RGMS. The displacements and the maximum principal effective stresses obtained from the RGMS solution of these three problems are compared to those from the commercial software Ansys Mechanical and are shown to practically coincide. The parallelization of pT+H V1.5 is validated by comparing its results to those from the serial T+H V1.5 code in a study that involved (a) fluid production from a large-scale 2D cylindrical system describing a real-life oceanic hydrate deposit and (b) a simplified geomechanical model based on hydrate-dependent pore compressibility. The coupling method is validated by comparing the numerical results to the analytical solutions of the Terzaghi and the McNamee-Gibson problems. The parallelization validation of the coupled simulator is achieved by comparing the results obtained for different numbers of processes in the solution of the problems used for the pT+H V1.5 parallelization validation with the full geomechanical model. The results clearly demonstrate the validity and reliability of the parallel codes (a) RGMS, (b) pT+H V1.5, and the (c) coupled pT+H V1.5 and RGM simulators, which can be used to solve the large-scale physics of complex problems.

Summary Reservoir rock wettability plays an important role in waterflooding especially in fractured carbonate reservoirs because oil recovery tends to be inefficient from the mixed-wet or oil-wet rock matrix. Improved oil recovery has been observed by adjusting the concentrations of potential-determining ions (PDIs) to alter the wettability of carbonate rocks. Our previous study showed that the oil recovery from carbonate reservoirs by waterflooding can be enhanced by the addition of glycine, the simplest amino acid. The interaction of glycine anion and oil-wet carbonate surfaces was confirmed in contact angle measurements and yielded the incremental oil recovery in imbibition experiments. This paper presents a surface complexation model (SCM) for the interaction between glycine and oil-wet carbonate surfaces that considers the impact of temperature, pH, salinity, and the concentrations of PDIs. The proposed SCM was tuned based on the zeta potential (ZP) data reported for synthetic calcite in glycine solutions. The tuned model predicts that the strong affinity of glycine for the oil-wet carbonate surface causes the carboxylic acids to desorb from the surface. The concentration of adsorbed carboxylic acids was in qualitative agreement with the water-wetting state of carbonate rocks inferred from the reported contact angle and spontaneous imbibition experiments. Moreover, the model indicates that glycine’s performance as a wettability alteration agent improves significantly at high temperatures. It also suggests that enhanced oil recovery (EOR) by formation brine (FB) with 1 to 3 wt% glycine at high temperatures should be similar to the injection of low-salinity seawater (LSW). The proposed SCM supports glycine’s potential to change the wettability of oil-wet carbonates at high salinity, high hardness, and high temperature.

Summary Two of the most critical properties for multiphase flow in a reservoir are relative permeability (Kr) and capillary pressure (Pc). To determine these parameters, careful interpretation of coreflooding and centrifuge experiments is necessary. In this work, a machine learning (ML) technique was incorporated to assist in the determination of these parameters quickly and synchronously for steady-state drainage coreflooding experiments. A state-of-the-art framework was developed in which a large database of Kr and Pc curves was generated based on existing mathematical models. This database was used to perform thousands of coreflood simulation runs representing oil-water drainage steady-state experiments. The results obtained from the corefloods including pressure drop and water saturation profile, along with other conventional core analysis data, were fed as features into the ML model. The entire data set was split into 70% for training, 15% for validation, and the remaining 15% for the blind testing of the model. The 70% of the data set for training teaches the model to capture fluid flow behavior inside the core, and then 15% of the data set was used to validate the trained model and to optimize the hyperparameters of the ML algorithm. The remaining 15% of the data set was used for testing the model and assessing the model performance scores. In addition, K-fold split technique was used to split the 15% testing data set to provide an unbiased estimate of the final model performance. The trained/tested model was thereby used to estimate Kr and Pc curves based on available experimental results. The values of the coefficient of determination (R2) were used to assess the accuracy and efficiency of the developed model. The respective crossplots indicate that the model is capable of making accurate predictions with an error percentage of less than 2% on history matching experimental data. This implies that the artificial-intelligence- (AI-) based model is capable of determining Kr and Pc curves. The present work could be an alternative approach to existing methods for interpreting Kr and Pc curves. In addition, the ML model can be adapted to produce results that include multiple options for Kr and Pc curves from which the best solution can be determined using engineering judgment. This is unlike solutions from some of the existing commercial codes, which usually provide only a single solution. The model currently focuses on the prediction of Kr and Pc curves for drainage steady-state experiments; however, the work can be extended to capture the imbibition cycle as well.

Summary Prediction of well production from unconventional reservoirs is often a complex problem with an incomplete understanding of physics and a considerable amount of data. The most effective way for dealing with it is to use the gray-box approach that combines the strengths of physics-based models and machine learning (ML) used for dealing with certain components of the prediction where physical understanding is poor or difficult. However, the development of methodologies for the incorporation of physics into ML is still in its infancy, not only in the oil and gas industry, but also in other scientific and engineering communities, including the physics community. To set the stage for further advancing the use of combining physics-based models with ML for predicting well production, in this paper we present a brief review of the current developments in this area in the industry, including ML representation of numerical simulation results, determination of parameters for decline curve analysis (DCA) models with ML, physics-informed ML (PIML) that provides an efficient and gridless method for solving differential equations and for discovering governing equations from observations, and physics-constrained ML (PCML) that directly embeds a physics-based model into a neural network. The advantages and potential limitations of the methods are discussed. The future research directions in this area include, but are not limited to, further developing and refining methodologies, including algorithm development, to directly embed physics-based models into ML; exploring the usefulness of PIML for reservoir simulations; and adapting the new developments of how the physics and ML are incorporated in other communities to the well-production prediction. Finally, the methodologies we discuss in the paper can be generally applied to conventional reservoirs as well, although the focus here is on unconventional reservoirs.

Summary Benchmarking the recovery factor and production performance of a given reservoir against applicable analogs is a key step in field development optimization and a prerequisite in understanding the necessary actions required to improve hydrocarbon recovery. Existing benchmarking methods are principally structured to solve specific problems in individual situations and, consequently, are difficult to apply widely and consistently. This study presents an alternative empirical analog benchmarking workflow that is based upon systematic analysis of more than 1,600 reservoirs from around the world. This workflow is designed for effective, practical, and repeatable application of analog analysis to all reservoir types, development scenarios, and production challenges. It comprises five key steps: (1) definition of problems and objectives; (2) parameterization of the target reservoir; (3) quantification of resource potential; (4) assessment of production performance; and (5) identification of best practices and lessons learned. Problems of differing nature and for different objectives require different sets of analogs. This workflow advocates starting with a broad set of parameters to find a wide range of analogs for quantification of resource potential, followed by a narrowly defined set of parameters to find relevant analogs for assessment of production performance. During subsequent analysis of the chosen analogs, the focus is on isolating specific critical issues and identifying a smaller number of applicable analogs that more closely match the target reservoir with the aim to document both best practices and lessons learned. This workflow aims to inform decisions by identifying the best-in-class performers and examining in detail what differentiates them. It has been successfully applied to improve hydrocarbon recovery for carbonate, clastic, and basement reservoirs globally. The case studies provided herein demonstrate that this workflow has real-world utility in the identification of upside recovery potential and specific actions that can be taken to optimize production and recovery.

Summary Accurate production forecasting is an essential task and accompanies the entire process of reservoir development. With the limitation of prediction principles and processes, the traditional approaches are difficult to make rapid predictions. With the development of artificial intelligence, the data-driven model provides an alternative approach for production forecasting. To fully take the impact of interwell interference on production into account, this paper proposes a deep learning-based hybrid model (GCN-LSTM), where graph convolutional network (GCN) is used to capture complicated spatial patterns between each well, and long short-term memory (LSTM) neural network is adopted to extract intricate temporal correlations from historical production data. To implement the proposed model more efficiently, two data preprocessing procedures are performed: Outliers in the data set are removed by using a box plot visualization, and measurement noise is reduced by a wavelet transform. The robustness and applicability of the proposed model are evaluated in two scenarios of different data types with the root mean square error (RMSE), the mean absolute error (MAE), and the mean absolute percentage error (MAPE). The results show that the proposed model can effectively capture spatial and temporal correlations to make a rapid and accurate oil production forecast.

Summary A common approach to forecast production from unconventional reservoirs is to extrapolate single-phase flow solutions. This approach ignores the effects of multiphase flow, which exist once the reservoir pressure falls below the bubble/dewpoint. This work introduces a new two-phase (oil and gas) flow solution suitable to extrapolating oil and gas production using scaling principles. In addition, this study compares the application of the two-phase and the single-phase solutions to estimates of production from tight-oil wells in the Wolfcamp Formation of west Texas. First, we combine the oil and the gas flow equations into a single two-phase flow equation. Second, we introduce a two-phase pseudopressure to help linearize the pressure diffusivity equation. Third, we cast the two-phase diffusion equation into a dimensionless form using inspectional analysis. The output of the model is a predicted dimensionless flow rate that can be easily scaled using two parameters: a hydrocarbon pore volume and a characteristic time. This study validates the solution against results of a commercial simulator. We also compare the results of both the two-phase and the single-phase solutions to forecast wells. The results of this research are the following: First, we show that single-phase flow solutions will consistently underestimate the oil ultimate recovery factors (URFs) for solution gas drives. The degree of underestimation will depend on the reservoir and flowing conditions as well as the fluid properties. Second, this work presents a sensitivity analysis of the pressure/volume/temperature (PVT) properties, which shows that lighter oils (more volatile) will yield larger recovery factors for the same drawdown conditions. Third, we compare the estimated ultimate recovery (EUR) predictions for two-phase and single-phase solutions under boundary-dominated flow (BDF) conditions. The results show that single-phase flow solutions will underestimate the ultimate cumulative oil production of wells because they do not account for liberation of dissolved gas and its subsequent expansion (pressure support) as the reservoir pressure falls below the bubblepoint. Finally, the application of the two-phase model provides a better fit when compared with the single-phase solution. The present model requires very little computation time to forecast production because it only uses two fitting parameters. It provides more realistic estimates of URFs and EURs, when compared with single-phase flow solutions, because it considers the expansion of the oil and gas phases for saturated flow. Finally, the solution is flexible and can be applied to forecast both tight-oil and gas condensate wells.

Summary Overpressure is a common feature among productive unconventional shale reservoirs, such as the Bone Spring (BSPG) and Wolfcamp (WFMP) Formations of the Delaware Basin (DB) of west Texas and southeastern New Mexico, and is thought to be a strong driver of well productivity. Compared with conventional reservoirs and shales in normal pressured conditions, the effects of overpressure on the mechanical properties of shales is not well understood. Here we present an analysis of overpressure in clay-bearing siliciclastic facies of the BSPG and WFMP Formations of the DB and implications for mechanical properties of the reservoir. Estimation of the effects of overpressure on mechanical properties of unconventional shale reservoirs is determined through use of the sonic overpressure indicator (SOPI). The method requires log model results that accurately characterize variations in lithology and porosity for the formations of interest. The SOPI (ΔT/ΔTN)2, where ΔT is the measured compressional sonic transit time, and ΔTN is the forward-modeled result for normally pressured conditions, can be used with elastic moduli and their interrelationships to compare estimates of mechanical properties including Poisson’s ratio ν, the Biot or effective stress coefficient α, and Young’s modulus E, in normal and overpressured conditions. Results presented here are broadly applicable to overpressured unconventional reservoirs that contain significant clay volume (>0.1 v/v) and exhibit low porosity (<0.08 v/v), comparable to that of siliciclastic-rich facies of the WFMP Formation. To account for increased VP/VS ratio, we regard overpressurization of shaly facies as an irreversible thermodynamic process that transforms a normally pressured siliciclastic system. At stress below the yield point, which is taken as the limit of normal pressure, the system responds elastically to stress; beyond this point, during overpressurization, the system responds as an elastic/plastic medium with strain hardening. We regard elastic moduli as descriptive of mechanical energy stored in this system. This perspective enables Poisson’s ratio for the overpressured system νOP to be computed from an estimate of the normally pressured system νN using (ΔT/ΔTN)2. Overpressure also results in a limited increase of the Biot or effective stress coefficient α. Moreover, recognition that overpressure results in a decrease of Young’s modulus, that is, EOP/EN < 1, provides a means of estimating the amount of strain energy stored by the formation due to overpressurization. We believe that when exposed to lower pressures by wellbore construction, this strain energy stored in overpressured unconventional reservoirs drives creep, which affects interpretations made using geomechanical models. We have developed and tested computational models based on biaxial or plane strain for vertical wells and uniaxial strain for horizontal wells that describe how creep likely affects estimation of minimum horizontal stress Shmin and pore pressure from instantaneous shut-in-pressure (ISIP) measurements. Thus, for overpressured unconventional reservoirs, ISIP determinations differ from tectonic Shmin by an amount related to ν and EOP/EN.

Summary Geoscientific and engineering experiments in petrophysics, rock physics, and rock mechanics depend on multiple, costly, and sometimes rare samples used to characterize the properties of natural rocks. Testing these samples helps in modeling various hydrocarbon recovery and stimulation scenarios, as well as understanding the fluid-rock interactions in the subsurface under various pressure and temperature conditions. Over the last decade, 3D printing has matured to become a more commonly available tool to enable repeatable experiments with controllable materials and pore system geometries to investigate petrophysical, geomechanical, and geophysical properties of porous rocks. This review introduces the development, characteristics, and capabilities of 3D printing technology that are specifically used in research. Applications in the realm of petrophysics highlight the issues of replicating the pore network geometry and subsurface physics, aiming at understanding fluid flow in porous media problems. Using 3D-printed models in rock mechanics experiments focuses on generating comparable geomechanical properties and reproducing fractures, joint surfaces, and other rock structures, whereas in rock physics, geophysical forward modeling is highlighted to take advantage of 3D printing technology. By summarizing the recent advances in 3D printing as applied to petrophysics, rock physics, and rock mechanics, this review paper presents the current state of the art and the challenges in scale, cost, time, and materials, as well as the directions for advancing this frontier discipline to answer various fundamental questions regarding porous media research using 3D printing technology.

Summary When the oil price is volatile, maximizing steam allocation and noncondensable gas (NCG) is essential to ensuring a profit but reducing risk. Minimizing risk entails moving the distribution of lower tail returns closer to the expected return. Thus, there is a risk-reward tradeoff during optimization. Real-time risk-return optimization with first-principle models is computationally demanding. Sibaweihi et al. (2019) presented a real-time steam-assisted gravity drainage (SAGD) recovery optimization with varying steam availability workflow. The workflow cannot handle uncertainty, and the data-driven model may forecast out of the physical range of the model output parameters. As a result, data-driven process modeling incorporating physical or operational constraints and an optimization problem formulation that references a decision-makers' metrics to a benchmark is crucial. This study proposes data-driven input-output normalization to incorporate operating constraints based on their physical range. The workflow includes model training updating based on the concept of forgetting factor to adapt the data-driven model to the current state of the reservoir. A robust optimization (RO) problem scheme in which economic risk is mitigated by formulating the objective as a tradeoff of expected returns and risk is managed in real time. A modified Modigliani’s risk-adjusted performance has been implemented to minimize the possibility of selecting the wrong optimal risk-return tradeoff of nonsymmetric return realizations in this work. In this work, the risk is quantified through variance, minimum, semivariance (down side risk), and conditional-value-at-risk of the returns realizations because of oil price volatility. Application of the proposed workflow on a synthetic reservoir with steam NCG co-injection showed the data-driven calibrated model forecast performance shows a reasonable agreement with the synthetic reservoir throughout the optimization period. In addition, the optimization study with the proposed workflow also showed a net present value (NPV) increase of approximately 25–77% and a decrease in the cumulative steam-oil-ratio (cSOR) from 4.5 to 6.7% compared with the continuous steam injection base case. The reduction in cSOR indicates a lower steam requirement. An increase in methane sequestered demonstrates workflow ability to reduce greenhouse gas emissions while improving SAGD NCG co-injection key performance indicators.

Summary The use of fiber optics in reservoir surveillance is bringing valuable insights into fracture geometry and fracture-hit identification, stage communication, and perforation cluster fluid distribution in many hydraulic fracturing processes. However, given the complexity associated with field data, its interpretation is a major challenge faced by engineers and geoscientists. In this work, we propose to generate distributed strain sensing (DSS)/distributed acoustic sensing (DAS) synthetic data of a crosswell fiber deployment that incorporates the physics governing hydraulic fracturing treatments. Our forward modeling can be used to add value to the interpretation task. The forward modeling is based on analytical and numerical solutions. The analytical solution is developed integrating two models: 2D fracture (e.g., Khristianovic-Geertsma-de Klerk known as KGD) and Sneddon’s induced stress. DSS is estimated using the plane strain approach that combines calculated stresses and rock properties (e.g., Young’s modulus and Poisson’s ratio). On the other hand, the numerical solution is implemented using the displacement discontinuity method (DDM), a type of boundary element method, with net pressure and/or shear stress as the boundary condition. In this case, the fiber gauge length concept is incorporated deriving displacement (i.e., DDM output) in space to obtain DSS values. In both methods, DAS is estimated by the differentiation of DSS in time. The analytical technique considers a single fracture opening and is used in a sensitivity analysis to evaluate the impact that rock/fluid parameters can promote on strain time histories. Moreover, advanced cases including multiple fractures failing in tensile or shear mode are simulated applying the numerical technique. Results indicate that our models are able to capture typical characteristics present in field data: heart-shaped pattern from a fracture approaching the fiber, stress shadow, and fracture hits. In particular, the numerical methodology captures relevant phenomenon associated with hydraulic and natural fractures interaction, which is often interpreted purely in terms of opening fractures. We can anticipate that the developed forward modeling, when embedded in a classification or regression artificial intelligence framework, will be an important tool adding substantial insights related to field fracture systems that ultimately can lead to production 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.

Summary Reservoir cooling by water or wastewater injection can significantly change the stress in the target injection zones and the bounding layers. Out-of-zone fracture growth is substantially affected by these poro-thermo-elastic stress changes occurring in heterogeneous rock layers. No previous study has systematically investigated the influence of both heat conduction and convection and the associated stress alteration and fracture height growth during long-term water injection in multiple layers. Without understanding this coupled effect, it is not possible to predict the injection-induced fracture geometry or the conditions under which these fractures will breach the bounding shale layers. This paper presents a model and results that clearly show that accounting for thermal conduction between the injection sand and the bounding shale is crucial in predicting fracture containment during water injection. We developed a fully coupled compositional reservoir/fracturing simulator that solves poro-thermo-elastic equations. We used it to simulate 3D fracture propagation induced by cold water injection and, at the same time, calculate changes in the stress field induced by thermo-poro-elastic effects in heterogeneous reservoir layers. The stress in the bounding layer is shown to change significantly as a result of thermal conduction from the layer below. We first validate our model with existing analytical solutions and then present synthetic cases and simulate a field case. Simulation results show that fracture height growth can be underpredicted if heat conduction between the target injection sands and bounding shales is ignored. We identify the effects of fluid properties, rock properties, and injection temperature on thermal-conduction-induced stress changes and fracture containment for the first time.

Summary Production forecasting is usually performed by applying a single model from a classical statistical standpoint (point estimation). This approach neglects: (a) model uncertainty and (b) quantification of uncertainty of the model’s estimates. This work evaluates the predictive accuracy of rate-time models to forecast production from tight-oil wells using Bayesian methods. We apply Bayesian leave-one-out (LOO) and leave-future-out (LFO) cross-validation (CV) using an accuracy metric that evaluates the uncertainty of the models’ estimates: the expected log predictive density (elpd). We illustrate the application of the procedure to tight-oil wells of west Texas. This work assesses the predictive accuracy of rate-time models to forecast production of tight-oil wells. We use two empirical models, the Arps hyperbolic and logistic growth models, and two physics-based models: scaled slightly compressible single-phase and scaled two-phase (oil and gas) solutions of the diffusivity equation. First, we perform Bayesian inference to generate probabilistic production forecasts for each model using a Bayesian workflow in which we assess the convergence of the Markov chain Monte Carlo (MCMC) algorithm, calibrate, and evaluate the robustness of the models’ inferences. Second, we evaluate the predictive accuracy of the models using the elpd accuracy metric. This metric provides a measure of out-of-sample predictive performance. We apply two different CV techniques: LOO and LFO. The results of this study are the following. First, we evaluate the predictive performance of the models using the elpd accuracy metric, which accounts for the uncertainty of the models’ estimates assessing distributions instead of point estimates. Second, we perform fast CV calculations using an important sampling technique to evaluate and compare the results of the application of two CV techniques: leave-one-out cross-validation (LOO-CV) and leave-future-out cross-validation (LFO-CV). While the goal of LOO-CV is to evaluate the models’ ability to accurately resemble the structure of the production data, LFO-CV aims to assess the models’ capacity to predict future-time production (honoring the time-dependent structure of the data). Despite the difference in their prediction goals, both methods yield similar results on the set of tight-oil wells under study. The logistic growth model yields the best predictive performance for most of the wells in the data set, followed by the two-phase physics-based flow model. This work shows the application of new tools to evaluate the predictive accuracy of models used to forecast production of tight-oil wells using: (a) an accuracy metric that accounts for the uncertainty of the models’ estimates and (b) fast computation of two CV techniques, LOO-CV and LFO-CV. To our knowledge, the proposed approach is novel and suitable to evaluate and eventually select the rate-time model(s) with the best predictive accuracy of models to forecast hydrocarbon production.

Summary Rayleigh frequency-shift-based distributed strain sensing (RFS-based DSS) is a fiber-optic-based diagnostic technique, which can measure the strain change along the fiber. The spatial resolution of RFS-based DSS can be as low as 0.2 m, and the measuring sensitivity is less than 1 με. Jin et al. (2021) presented a set of DSS data from the Hydraulic Fracture Test Site 2 project to demonstrate its potential to characterize near-wellbore fracture properties and to evaluate perforation efficiency during production and shut-in periods. Extensional strain changes are observed at locations around perforations during a shut-in period. At each perforation cluster, the observed responses of strain changes are significantly different. However, the driving mechanisms for the various observations are not clear, which hinders accurate interpretations of DSS data for near-wellbore fracture characterization. In this study, we applied a coupled flow and geomechanics model to simulate the observed DSS signals under various fractured reservoir conditions. The objective is to improve understanding of the DSS measurements and characterize near-wellbore fracture geometry. We used our in-house coupled flow and geomechanics simulator, which is developed by a combined finite-volume and finite-element method, to simulate strain responses within and near a fracture during shut-in and reopen periods. Local grid refinement was adopted around fractures and the wellbore, so that the simulated strain data can accurately represent the DSS measurements. The plane-strain condition is assumed. Numerical models with various fracture geometries and properties were constructed with representative parameters and in-situ conditions of the Permian Basin. The simulated well was shut-in for 4 days after producing 240 days, and reopened again for 1 day, following the actual field operation as shown in Jin et al. (2021). The characters of the strain changes along the fiber were analyzed and related to near-wellbore fracture properties. A novel diagnostic plot of relative strain change vs. wellbore pressure was presented to infer near-wellbore fracture characteristics. The impacts of permeability and size of the near-wellbore-stimulated region, fracture length, and near-perforation damage zone on strain responses were investigated through sensitivity analysis. The strain responses simulated by our model capture the observed signatures of field DSS measurements. During the shut-in period, clear positive strain changes are observed around the perforation locations, forming a “hump” signature. The shape of the “hump” region and peak value of each “hump” are dependent on the size and permeability of the near-wellbore fractured zone. Once the well is reopened, the strain changes decrease as the pressure drops. However, in one cycle of shut-in and reopen, the strain-pressure diagnostic plot shows path dependency. The discrepancy between the shut-in and reopen periods is highly influenced by the properties of near-wellbore fractured zones. The differences in the strain-pressure diagnostic plots can help to identify the conductive fractures. This study provides better understandings of the DSS measurements and their relations to the near-wellbore fracture properties, which is of practical importance for near-wellbore fracture characterization and completion/stimulation optimization.

Summary Shale wells are often shut in after hydraulic fracturing is completed. Shut-in often lasts for an extended period in the perceived hope of improving the ultimate oil recovery. However, current literature does not show a strong consensus on whether shut-in will improve ultimate oil recovery. Because of the delayed production, evaluating the benefits of shut-in in improving the ultimate oil recovery is crucial. Otherwise, shut-in would merely delay the production and harm the economic performance. In this paper, we use a numerical flow-geomechanical modeling approach to investigate the effect of imbibition on shut-in potentials to improve the ultimate oil recovery. We propose that imbibition is one of the strongly confounding variables that causes mixed conclusions in the related literature. The investigation methodology involves probabilistic forecasting of three reservoir realization models validated based on the same field production data. Each of the models represents different primary recovery-driving mechanisms, such as imbibition-dominant and compaction-dominant recovery. A parametric study is conducted to explore and identify the specific reservoir conditions in which shut-in tends to improve shale oil recovery. Ten reservoir parameters that affect the imbibition strength are studied under different shut-in durations. Comparison among the three models quantitatively demonstrates that shut-in tends to improve both the ultimate oil recovery and net present value (NPV) only if the shale reservoir demonstrates imbibition-dominant recovery. A correlation among ultimate oil recovery, flowback efficiency, and NPV also shows that there is no strong relationship between flowback efficiency and ultimate oil recovery. This study is one of the first to emphasize the importance of quantifying the imbibition strength and its contribution in helping recover the shale oil for optimum flowback and shale well shut-in design after hydraulic fracturing.