Search

Increasing the production and utilization of shale gas is of great significance for building a clean and lowcarbon energy system. Sharp decline of gas production has been widely observed in shale gas reservoirs. How to forecast shale gas production is still challenging due to complex fracture networks, dynamic fracture properties, frac hits, complicated multiphase flow, and multi-scale flow as well as data quality and uncertainty. This work develops an integrated framework for evaluating shale gas well production based on data-driven models. Firstly, a comprehensive dominated-factor system has been established, including geological, drilling, fracturing, and production factors. Data processing and visualization are required to ensure data quality and determine final data set. A shale gas production evaluation model is developed to evaluate shale gas production levels. Finally, the random forest algorithm is used to forecast shale gas production. The prediction accuracy of shale gas production level is higher than 95% based on the shale gas reservoirs in China. Forty-one wells are randomly selected to predict cumulative gas production using the optimal regression model. The proposed shale gas production evaluation framework overcomes too many assumptions of analytical or semi-analytical models and avoids huge computation cost and poor generalization for numerical modelling. [ABSTRACT FROM AUTHOR]

Temperature and pressure conditions of deep shale are beyond experiment range, and the amount of adsorbed gas is difficult to determine. To predict the adsorbed gas content of deep shales under formation conditions, isothermal adsorption experiments and model building were conducted on shale samples from Longmaxi Formation in China. A temperature-dependent adsorption model based on the Langmuir equation is proposed, which can be well-fitted by observed isotherms with a high correlation coefficient. Based on the fitted parameters at 303.15 K, the isothermal adsorption curves at 333.15 K, 363.15 K, and 393.15 K are predicted, showing a good agreement with experimental curves available. Compared with previous prediction methods, the biggest advantage of the proposed method is that it can be carried out only based on one-time isothermal adsorption experiment. Based on the predictions, the downward trend of the excess adsorption curves will slow down under high temperature and pressure conditions, and when the pressure reaches a certain level (> 80 MPa), the temperature has little effect on the excess adsorption capacity. While for absolute adsorption, the gas adsorption reaches saturation much slowly at high temperature, it can also reach saturation under formation pressure. Under the burial depth of marine shale, temperature plays a major role in controlling the adsorbed gas, resulting in the decrease of adsorbed gas content in deep shale, and its ratio will further decrease as the depth increases. [ABSTRACT FROM AUTHOR]

An appropriate well spacing plan is critical for the economic development of shale gas reservoirs. The biggest challenge for well spacing optimization is interpreting the subsurface uncertainties associated with hydraulic and natural fractures. Another challenge is the existence of complex natural fractures. This work applied an integrated well spacing optimization workflow in shale gas reservoirs of the Sichuan Basin in China with both hydraulic and natural fractures. The workflow consists of five components: data preparation, reservoir simulation, estimated ultimate recovery (EUR) analysis, economic calculation, and well spacing optimization. Firstly, the multiple realizations of thirteen uncertain parameters of matrix and fractures, including matrix permeability and porosity, three relative permeability parameters, hydraulic fracture height, half-length, width, conductivity, water saturation, and natural fracture number, length, and conductivity, were captured by the assisted history matching (AHM). The fractures were modeled by the embedded discrete fracture model (EDFM) accurately and efficiently. Then, 84 AHM solutions combining with five well spacing scenarios from 517 ft to 1550 ft would generate 420 simulation cases. After reservoir simulation of these 420 cases, we forecasted the long-term gas production for each well spacing scenario. Gas EUR degradation and well interference would imply the critical well spacing. The net present value (NPV) for all scenarios would be calculated and trained by K -nearest neighbors (KNN) proxy to better understand the relationship between well spacing and NPV. In this study, the optimum well spacing was determined as 793 ft, corresponding with a maximum NPV of 18 million USD, with the contribution of hydraulic fractures and natural fractures. [ABSTRACT FROM AUTHOR]

The goal of this study is to develop a new model to simulate gas and water transport in shale nanopores and complex fractures. A new gas diffusivity equation was first derived to consider multiple important physical mechanisms such as gas desorption, gas slippage and diffusion, and non‐Darcy flow. For complex fractures, a state‐of‐the‐art embedded discrete fracture model (EDFM) was implemented. Numerical model is verified against a commercial reservoir simulator for shale gas simulation with multiple planar fractures. After that, a series of simulation studies was performed to investigate the impacts of complex gas transport mechanisms and various fracture geometries on well performance. The critical parameters controlling well performance are identified. The simulation results reveal that modeling of gas production from complex fractures as well as modeling important gas transport mechanisms in shale gas reservoirs is extremely significant. © 2018 American Institute of Chemical Engineers AIChE J, 64: 2251–2264, 2018 [ABSTRACT FROM AUTHOR]

A comprehensive model for simulating gas transport in shale formation with complex fracture geometry is still lacking in the petroleum industry. First, the current models are challenging to efficiently handle the complex nonplanar fractures with varying fracture width and fracture permeability along fracture length caused by very-complicated gridding issues, a high computational cost, and complexities in development of computational codes. Second, the conventional gas-diffusivity equation needs to be revised to include all important gas-transport mechanisms such as gas slippage, gas diffusion, and gas desorption. Hence, the goal of this study was to develop a new model to fill this gap. We present an efficient semianalytical model by dividing fractures into several fracture segments to represent the complex nonplanar- fracture geometry and present a revised diffusivity equation to consider all the important gas-transport mechanisms for shale-gas reservoirs. We verified this model against a numerical reservoir simulator and an analytical solution for a single planar fracture. Transient-flow behavior between planar fractures and nonplanar fractures with and without considering the natural fractures was compared. Furthermore, a well with actual production data from the Marcellus Shale was analyzed. First, we use a complex hydraulic-fracture-propagation model to generate nonplanar-fracture geometry. Then, we apply the fracture geometry in the semianalytical model to evaluate well performance. Contributions of each mechanism to gas recovery are examined. By use of history matching and production forecasting, the difference of cumulative gas production between the nonplanar fractures and the converted planar fractures with equal fracture length is approximately 12% at 30 years of production. In addition, the contribution of gas diffusion, gas slippage, and gas desorption to gas recovery at 30 years of production compared with that without considering them is approximately 11, 16, and 25%, respectively. In total, the contribution of these three important mechanisms is approximately 56%. Hence, modeling of gas production from the nonplanar fractures and modeling the important gas-transport mechanisms in shale-gas reservoirs are significant. [ABSTRACT FROM AUTHOR]

The total of the gas in shale gas reservoirs is sourced from a combination of free, adsorbed and dissolved/diffused gas. The mechanisms of production of free and adsorbed gas have been studied by numerous researchers. In contrast, the evolution of the dissolved gas and its contribution to the total gas production is not well understood. In this study we model the effect of pore micro-structure in organic matter (OM) on the rate of gas production in shale reservoirs. In this regard, first, we solve the gas-in-solid diffusion equation over a general doubly connected spatial domain with external Neumann and internal Dirichlet boundary conditions. The obtained solution is applied systematically to multi-pore porous OM domains and then the effect of pore morphology on the rate of gas production is studied. Our model results show that pore geometry has a slight effect on the gas diffusion process, while total organic carbon, and OM porosity, pore size distribution and specific surface area, are dominant parameters. An abundance of very small pores in OM tremendously increases the diffuse gas contribution in the total gas reserve and production. [ABSTRACT FROM AUTHOR]

In this study, the non-intrusive EDFM (embedded discrete fracture model) method was presented to investigate the impact of different choke management strategies on well performance. Through the EDFM method, accurate simulation can be conducted to efficiently evaluate the fracture complexities. First, by implementing this powerful technology, a horizontal well with multi-stage hydraulic and natural fractures was set up, where the permeability can be distributed sequentially in each hydraulic fracture segment. Then various pressure drawdown scenarios from conservative to aggressive strategy were designed. The different levels of fracture closure can be properly modeled in each state. Additionally, pressure distribution for the matrix and fractures was depicted to provide straightforward insights for the choke management under two extreme strategies. Subsequently, a series of sensitivity studies were presented to evaluate the impacts of various factors on shale gas production, including fracture permeability modulus, fracture closure, and natural fractures network. The simulation results show that choke management can be simulated effectively by applying EDFM. After considering the fracture closure behavior and complex fracture networks, the conservative drawdown strategy can be addressed as the optimal strategy for the EUR, as it improves the cumulative gas production by maintaining the hydraulic fracture open through a steady pressure decline. The remained proppants enhance the fracture conductivity, thereby expanding its drainage influence towards larger zones of the reservoir. The influence of natural fractures, including the fracture length, fracture number, and fracture conductivity, are also studied. All these three variables play a significant impact on well performance. Consequently, the model becomes a valuable stencil to design fracture closure and complex fracture networks, which can be applied to optimize the choke management design for unconventional reservoirs. • A new model of choke management optimization for shale gas reservoirs was developed. • Hydraulic fractures with complex fracture networks were efficiently modeled by EDFM. • Fracture closure under different pressure drawdown strategies were evaluated. • Multiple uncertain natural fracture parameters were analyzed. [ABSTRACT FROM AUTHOR]

We have developed a new pore-scale method to quantify petrophysical properties of hydrocarbon (HC)-bearing shale. Recent studies indicate that slip flow, Knudsen diffusion, Langmuir desorption, and diffusion in kerogen contribute to the unconventional production properties of shale-gas formations. Conventional petrophysical interpretation methods do not account for the aforementioned phenomena and are often inconclusive when estimating petrophysical properties in shale formations. We constructed a pore-scale representation of the lower Eagle Ford Shale based on focused-ion-beam-scanning-electron-microscope (FIB-SEM) images. Permeability is calculated via previously developed finite-difference methods for the cases with and without slip flow and Knudsen diffusion. The method also calculates streamlines to describe sample pore connectivity. Weighted throat-size distributions are defined based on streamlines to represent the most resistive paths for fluid flow in the FIB-SEM image. Subsequently, permeability is estimated from the dominant throat size in the weighted throat-size distribution. We used a new fluid percolation model for HC-bearing shale that expands HC from kerogen surfaces and water from grain and clay surfaces into the pore space to vary fluid saturation. Isolated pores are randomly distributed within kerogen to increase kerogen maturity in the model. Electrical resistivity is calculated with a finite-difference solution of Kirchhoff's voltage law applied at the pore scale. A parallel conductor model was used based on Archie's equation for water conductivity in pores and a parallel conductive path for the Stern-diffuse layer. Calculations were compared with Waxman-Smits' and Archie's predictions of macroscopic electric conductivity. Using practical modeling parameters, the parallel conductor model yields the most accurate prediction of pore-scale sample conductivity for various cases of water saturation and conductivity. [ABSTRACT FROM AUTHOR]

Highlights: [•] History matching is performed based on two field production data from Barnett Shale and Marcellus Shale. [•] Gas desorption and geomechanics effects are first evaluated systematically for Barnett Shale and Marcellus Shale. [•] Comparison of gas desorption effect from different shale formations is present. [•] Critical parameters controlling gas desorption process are obtained using Design of Experiment. [Copyright &y& Elsevier]

In order to account for big uncertainties such as well interferences, hydraulic and natural fractures' properties and matrix properties in shale gas reservoirs, it is paramount to develop a robust and efficient approach for well spacing optimization. In this study, a novel well spacing optimization workflow is proposed and applied to a real shale gas reservoir with two-phase flow, incorporating the systematic analysis of uncertainty reservoir and fracture parameters. One hundred combinations of these uncertainties, considering their interactions, were gathered from assisted history matching solutions, which were calibrated by the actual field production history from the well in the Sichuan Basin. These combinations were used as direct input to the well spacing optimization workflow, and five "wells per section" spacing scenarios were considered, with spacing ranging from 157 m (517 ft) to 472 m (1550 ft). An embedded discrete fracture model was used to efficiently model both hydraulic fractures and complex natural fractures non-intrusively, along with a commercial compositional reservoir simulator. Economic analysis after production simulation was then carried out, by collecting cumulative gas and water production after 20 years. The net present value (NPV) distributions of the different well spacing scenarios were calculated and presented as box-plots with a NPV ranging from 15 to 35 million dollars. It was found that the well spacing that maximizes the project NPV for this study is 236 m (775 ft), with the project NPV ranging from 15 to 35 million dollars and a 50th percentile (P50) value of 25.9 million dollars. In addition, spacings of 189 m (620 ft) and 315 m (1033 ft) can also produce substantial project profits, but are relatively less satisfactory than the 236 m (775 ft) case when comparing the P25, P50 and P75 values. The results obtained from this study provide key insights into the field pilot design of well spacing in shale gas reservoirs with complex natural fractures. [ABSTRACT FROM AUTHOR]

• A 3D in-situ comparison method for non-destructive identification of minerals and pore structure in shale. • Quantitative evaluation of CO 2 -water-shale reaction process through 3D digital core. • The depth of reaction, types of reactive minerals, and structural failure characteristics are quantitatively analyzed through non-destructive methods. With the continuous improvement of carbon emission requirements and the promotion of the strategic position of shale gas energy, CO 2 flooding shale gas technology has become one of the key technologies to reduce carbon emission and improve the recovery efficiency of shale gas wells. After carbon dioxide enters the shale reservoir, a series of physical and chemical reactions will occur under the combined influence of water and temperature, resulting in a decrease in the mechanical strength of the shale and a change in reservoir properties such as porosity. The cores always are destroyed by the previous mineral identification method, which is hard to achieve the analysis without hurt for reaction depth, location and the three-dimensional change of pore structures of the same sample. Therefore, this manuscript establishes a set of shale mineral identification method based on the dual-energy CT theory which is verified by using sandstones and reaction experiments. The three-dimensional digital cores and mineral distribution of six shale samples are gotten by this method. The samples are carried out carbon dioxide-water-rock reaction experiments, and the influences of different water, temperature and reaction time on reaction are clarified, the maximum reaction depth of the reaction between CO 2 and shale with water under 30 ℃ is 50.9 μm, the depth with water under 80 ℃ is 152 μm, and the reaction depth of different size samples with water under 30 ℃ for 60 days are 220 μm and 195 μm respectively. The results show that carbon dioxide can hardly react with shale without water, and with the increase of temperature, the rate of CO 2 -water-shale reaction increases. Larger samples have a larger surface area and more minerals participate in heterogeneous surface reactions, resulting in more minerals being ultimately dissolved. After the reaction, the porosity of the shale surface is significantly improved, and its permeability will be greatly enhanced. [ABSTRACT FROM AUTHOR]

• A novel procedure was developed for solving viscoelastic engineering problems based on fractional calculus. • Fractional order is in direct proportion to the total content of clay and kerogen in shale samples. • Simulation results based on fractional model in long-term and short-term scales are quite different. • The delayed proppant production was explained by our closure stress relaxation model. With the development of geothermal energy and unconventional reservoirs, shales have become the popular target formation. This kind of organic-rich or clay-rich geomaterial has significant viscoelastic behavior which may lead to severe accidents, due to the creep rupture and the relaxation of pre-stress. In order to understand the viscoelastic behavior and simulate it quickly and accurately, we have developed a novel procedure for solving viscoelastic problems based on fractional calculus. Fractional viscoelastic constitutive equations were built by introducing a springpot into the classical component models. Fractional and classical constitutive equations were compared by the fitting results of triaxial experimental data of shale samples. Fitting efficiency of fractional constitutive equation is enhanced if compared to classical equations. Besides, we also found that fractional order is directly correlated to the amount of total volume content of clay and kerogen. This implies that the physical meaning of fractional model is clearer and its extrapolation is more accurate. Fracture width creep model for unpropped fractures and closure stress relaxation model for well propped fractures were developed and analyzed. Their simulation results in terms of long-term and short-term scales differ tremendously and hence should be studied individually. The influence of lateral compliances of real shale samples on fracture width creep or closure stress relaxation is not significant and may be ignored in long-term simulations. Simulation results of closure stress relaxation could explain the abnormal proppant production, which occurs several days later than the start of natural flow-back. [ABSTRACT FROM AUTHOR]

• A fractional model for simulating long-term fracture conductivity decay was developed. • The damage rate of cumulative production induced by shale creep is unignorable. • The formation viscous modulus determines the strength of shale viscoelastic mechanical behaviors. • Cumulative production damage rate caused by shale creep is obvious when formation viscous modulus is less than 30GPa. Due to the high-content of clay and kerogen, shale will always show viscoelastic mechanical behaviors during the development of shale gas, which will lead to the increase of proppant embedment and do harm to gas recovery from shale reservoirs. In the previous studies, viscoelastic mechanical behaviors of reservoir rock had not been coupled with productivity model of gas well and the long-term fracture conductivity was overestimated. In order to figure out the creep damage to long-term fracture conductivity and predict the estimated ultimate recovery of shale gas wells accurately, a novel time-varying fracture conductivity model based on fractional calculus was developed for simulating the long-term decay of fracture conductivity. Proppant deformation, proppant embedment, and shale creep described by fractional Kelvin model were taken into consideration in the time-varying fracture conductivity model and it was coupled with embedded discrete fracture model (EDFM) to simulate the well performance. Simulation results show that: Daily gas production curve simulated by the EDFM with creep damage under field conditions of a shale gas well in Sichuan Province, China was aligned with its production data. The damage rate of cumulative production considering shale creep is more than twice that only caused by proppant deformation and proppant embedment. And the formation viscous modulus determines the strength of the shale viscoelastic mechanical behaviors. According to formation viscous modulus, the shale reservoirs can be divided into three types: Type I with strong creep damage (E 2 v < 5GPa), Type Ⅱ with moderate creep damage (5GPa < E 2 v < 30GPa), and Type III with few creep damage (E 2 v > 30GPa). [ABSTRACT FROM AUTHOR]

Abstract A manual history matching is a time-consuming process and too risky for an investment decision because it cannot capture the uncertainty of subsurface information and production forecast. Therefore, Assisted History Matching (AHM) was adopted to obtain multiple reservoir realizations for the probabilistic production forecast. Markov Chain Monte Carlo (MCMC) is one of the AHM in the Bayesian statistics that have benefits of quantifying uncertainty without bias, not being trapped in any local minima. However, the AHM in shale reservoirs has not been widely studied due to the unconventional shale's complexities such as fracture modeling. The continuum approach cannot accurately model the connection between hydraulic fractures and matrix blocks in shale reservoirs. While, Discrete Fracture Modeling (DFM) by using local grid refinement can overcome the issue of the continuum approach, there is still a limitation of computational efficiency. Therefore, Embedded Discrete Fracture Modeling (EDFM) has been developed to combine the benefits of dual-continuum model and DFM. While DFM is an intrusive technique to model fractures, EDFM can account the mass transfer between matrix cells and fractures without being intrusive to the model. Besides, EDFM provides more efficient computational time than DFM. In this study, we proposed a practical and efficient iterative AHM workflow by integrating the benefits of using multiple objective functions, multiple proxy-based MCMC algorithm, EDFM and commercial reservoir simulator. Then, we demonstrated an application of the proposed workflow to a real shale gas well by matching both gas and water rates. The field application shows that the proposed AHM workflow efficiently finds the history matching solutions of 64 out of 300 simulation runs (21%) with a total computational time of one day. Furthermore, the workflow captures the uncertainty of reservoir and fracture parameters in terms of posterior distribution. Also, we performed the probabilistic production forecast and obtained P10-50-90 estimated ultimate recovery of gas production. Highlights • An assisted history matching workflow for shale gas reservoir is developed. • Reservoir simulator, EDFM, and multiple proxy-based MCMC sampling algorithm are coupled. • The workflow was applied to history match a real field case of shale gas reservoir. • Posterior distributions for each reservoir and effective fracture properties are determined. [ABSTRACT FROM AUTHOR]

History matching is an important process for uncertainty quantification of reservoir performance. Hundreds of reservoir simulation cases, or more, are typically required to sample the posterior probability density function of the uncertain parameters. The process is extremely challenging for shale reservoirs due to large uncertainties including reservoir heterogeneity, wide range of permeability, gas desorption, and fracture properties. Direct application of stimulated rock volume or dual continuum model fails to capture fracture connectivity and fracture conductivity that dominate the fluid transport in most shale developments. Embedding discrete fractures in reservoir simulation model is thus a preferable method to attain more realistic reservoir behavior. However, the option of using local grid refinement to embed the fractures is computationally expensive as well as intrusive to a reservoir model. Even more challenging is generating a huge set of fracture-embedded reservoir cases during the history matching process. Nevertheless, recent developments in a methodology called Embedded Discrete Fracture Model (EDFM) have successfully overcome the computational complexity to include discrete fractures in reservoir simulations. We demonstrate the implementation of the assisted history matching workflow coupled with EDFM to history match a real field case of shale gas condensate reservoir in Duvernay Shale, Canada. In the workflow, the EDFM preprocessor is fully coupled with a commercial reservoir simulator and proxy-based Markov chain Monte Carlo (MCMC) algorithm. Moreover, to further reduce the computational cost, the single-fractured submodel is used to represent the full reservoir model for this case study. Using the proposed assisted history matching workflow and the appropriate submodel enables efficient and successful uncertainty quantification of this shale play. [ABSTRACT FROM AUTHOR]

Increase in energy demand has played a significant role in the persistent exploitation and exploration of unconventional oil and gas resources. Shale gas reservoirs are one of the major unconventional resources. Advancements in horizontal drilling and hydraulic fracturing techniques have been the key to achieve economic rates of production from these shale gas reservoirs. In addition to their ultra-low permeability, shale gas reservoirs are characterized by their complex gas transport mechanisms and complex natural and induced (hydraulic) fracture geometries. Production from shale gas reservoirs is predominantly composed of two-phase flow of gas and water. However, proper modeling of the two-phase behavior as well as incorporating the complex fracture geometries have been a challenge within the industry. Due to the limitation of local grid refinement approach, hydraulic fractures are assumed to be planar (orthogonal), which is an unrealistic assumption. Although more flexible approaches are available, such as the use of unstructured grids, they require significantly high computational resources. In this paper, we introduce an embedded discrete fracture model (EDFM) to explicitly model complex fracture geometries. The EDFM efficient approach is capable of explicitly modeling complex fracture geometries without increasing the computational demand. Utilizing EDFM alongside a commercial simulator, we constructed a 3D reservoir model to investigate the effect of complex fracture geometries on the two-phase flow of a shale gas well. In our investigation, varying degrees of hydraulic fracture complexity with 1-set and 2-set of natural fractures were tested. The simulation results confirm the importance of properly modeling fracture complexity, highlighting that it plays an integral part in the estimation of gas and water recoveries. In addition, the simulation results hint to the pronounced effect of fracture interference as fracture complexity increases. Finally, variable fracture conductivities and initial water saturation values were analyzed to further assess their effect on the two-phase production behavior of the shale gas well. This study examines the effect of non-orthogonal complex fracture geometry on the two-phase flow of shale gas wells. The work can provide a significant insight toward understanding the extent to which fracture complexity can affect the performance of shale gas wells. [ABSTRACT FROM AUTHOR]

• EDFM was used to model hydraulic fractures. • Neural-network Markov chain Monte Carlo algorithm were used for history matching. • Three shale gas wells were automatically history matched using assisted history matching workflow. • Effective hydraulic fracture and matrix properties were determined. We investigated the difference in production performances of three shale gas wells with different completions and fracturing designs. We used assisted history matching technique to perform production match automatically and obtain multiple history matching solutions. The history matching algorithm used in this study is neural networks Markov chain Monte Carlo. In addition, we utilized embedded discrete fracture model technique to model hydraulic fractures instead of local grid refinement method to speed up the simulation runs. The history matching parameters consist of fracture half-length, fracture height, fracture conductivity, fracture width and fracture water saturation for hydraulic fractures properties and include matrix porosity, matrix permeability and relative permeability curves for matrix properties. We found that fracture width, fracture half-length, fracture conductivity and matrix permeability were significantly different in three shale gas wells while fracture height, fracture water saturation, porosity and relative permeability curves were less different. The information from this study will be used further to improve completion and fracturing design. [ABSTRACT FROM AUTHOR]

• A new assisted history matching workflow for shale gas reservoirs was developed. • K-nearest neighbors and neural networks as proxy model was compared. • Hydraulic fracture properties were identified in a probabilistic manner. • Hydraulic fractures were modeled using EDFM. We performed assisted history matching (AHM) in a real shale gas well using uncertain parameters such as fracture geometry, fracture conductivity, matrix permeability, matrix and fracture water saturation, and relative permeability curves. We also investigated the performance of two proxy models including K-nearest neighbors (KNN) and neural networks (NN) to be used in multiple-proxy-based Markov chain Monte Carlo (MCMC) algorithm. We emphasized the performance of both proxy models by comparing the number of history matching solution found and elapsed time. While, KNN required less elapsed time by half than NN, we found that NN performed better in terms of accuracy and predictability than KNN. In other words, NN required a smaller number of simulations by half than KNN in order to obtain the same number of history matching solutions. Therefore, it depends on what is more important to each problem either number of simulations or elapsed time. For history matching result, both proxy models in the multiple-proxy-based MCMC algorithm have similar results of posterior distribution of uncertain parameters. This confirms the robustness of the proposed history matching algorithm. The benefits of this study are that we can characterize fracture geometry and reservoir properties in a probabilistic manner. These multiple realizations can be further used for a probabilistic production forecast, future fracturing design, and well spacing optimization and planning. This AHM workflow can be applied to any hydraulic-fractured wells with historical production data. [ABSTRACT FROM AUTHOR]