Modeling, History Matching, Forecasting and Analysis of Shale Reservoirs Performance Using Artificial Intelligence

This article, written by Senior Technology Editor Dennis Denney, contains highlights of paper SPE 121038, "Evaluating Implica tions of Hydraulic Fracturing in Shale- Gas Reservoirs," by J. Daniel Arthur, SPE, Brian Bohm, Bobbi Jo Coughlin, and Mark Layne, SPE, ALL Consulting, prepared for the 2009 SPE Americas E&P Environmental and Safety Conference, San Antonio, Texas, 23-25 March. The paper has not been peer reviewed. Exploration, drilling, and production of shale-gas plays such as the Barnett, Fayetteville, and Haynesville have changed the unconventional-gas industry. Two technologies that enhanced shale-gas development are horizontal drilling and hydraulic fracturing. Hydraulic fracturing of horizontal wells completed in shale reservoirs often requires larger volumes of fracturing fluid than conventional vertical-well stimulations. Rapid development of shale gas across the country has created concerns about the use of infrastructure and about environmental effects. Introduction The Barnett-shale play began in the Fort Worth, Texas, area in the 1980s. Continued improvements in hydraulic-fracturing techniques and technology enhanced development of the Barnett shale. Development uses horizontal drilling and sequenced multistage hydraulic-fracturing technologies. Another source of technological advances in horizontal drilling and hydraulic fracturing is the Bakken shale of the Williston basin in Montana and North Dakota, which produces more oil than gas. Each shale-gas basin has a unique set of exploration criteria and operational challenges. However, the combination of horizontal-well completions and sequenced hydraulic-fracture treatments has been crucial in facilitating the expansion of modern shale-gas development to other shale plays. The low natural permeability of shale has been the limiting factor to production. Shale-Gas Geology Shale gas is natural gas produced from shale formations, and these shales function as both source rock and reservoir for the natural gas. Typically, shale gas is dry gas composed primarily of methane (90% or more). However, some shale-gas formations do produce water. Gas shales are organic-rich formations traditionally viewed as source rocks and seals for stratigraphic gas accumulations. The thin layers, as shown in Fig. 1, that make up shale result in a rock with limited horizontal permeability and minimal vertical permeability. Unfractured shales typically have permeabilities on the order of 0.01 to 0.00001 md, limiting production. The natural matrix permeability of shale sometimes is overcome by natural-fracture networks that developed when overburden pressure was reduced as a result of erosion of overlying rock formations. Other tectonic activity also can cause a change in the pressure, creating additional fractures. Often, fractures are key sources of permeability, facilitating the migration of fluids through fine-grained rocks such as shales.

Shale reservoir has been focused among unconventional resources since the first extraction of free and adsorbed gas from the low-permeable Barnett Shale via horizontal drilling and hydraulic fracturing. In the beginning of production, free gas was rapidly recovered through an artificial fracture system, and then, desorbed gas followed at the final stage due to a decrease of reservoir pressure. This desorbed gas volume commonly occupies 10 to 40% of total gas production in shale gas play although it shows wide variety in cumulative gas volume and production time. The largest gas production in Canada is recovered from either tight sandstone or shale reservoirs. The Montney play in Western Canadian Sedimentary Basin (WCSB) has produced up to 80% of Canadian natural gas production. The desorbed gas production from this play has been reported up to 10% of total produced gas. The distribution and productivity of the desorbed gas have not been fully studied. Therefore, we focus to understand the distribution of the desorbed gas volume of eastern, middle, and northwestern areas in the Montney play. The desorbed gas volume within these areas was estimated from the relationship among canister, illite, and shale volumes in core samples and well logs. The average shale volume fraction in eastern area is 0.38 v / v , the average illite mineral volume fraction is 0.25 v / v , and the average desorbed gas volume refers to 8.52 scf/ton. In middle area, calculated volume represents 0.34 v / v , 0.216 v / v , and 8.15 scf/ton as listed above. The northwestern area also shows 0.65 v / v , 0.4 v / v , and 9.78 scf/ton, respectively. 3D models of each area indicated relatively rich and lack parts of desorbed gas volume. These estimated desorbed gas volume and gas production history were compared in order to understand when and how the desorbed gas would affect to gas production. It shows strong positive relationship, gradually increasing correlation to the later stage (from 24-44 months to 36-44 months) of gas production in the entire areas. This result implies that the estimation of later stage gas productivity is able from the estimated volume of desorbed gas, and also, the total gas production can be forecast in shale gas reservoir. Northwestern area in Montney play preserves relatively abundant desorbed gas volume, which will be dominant after 24 months of production.

Hydraulic fracturing in shale formations containing natural fractures can lead to complex and branched fracture networks which may fail to efficiently access hydrocarbons deep in the formation away from the well. In this study, to minimise the negative effects of natural fractures, the orientation of hydraulically fractured wells in such formations is investigated with the focus on the effects of in-situ differential stresses and toughness of natural fractures. To explore the impact of geomechanical properties on gas production from hydraulically fractured wells, fractures were first up scaled and then gas production from the reservoir was investigated. The results show that low natural fracture toughness and differential in-situ stress both increase the tendency of propagating hydraulic fractures to exploit natural fractures. Under these conditions, it was found that drilling the well in a modified orientation could result in a higher recovery factor compared to the wells drilled parallel to the direction of minimum horizontal stress. This is due to the larger stimulated reservoir volume through a synergy between the hydraulic and natural fractures. However, high differential in-situ stress and fracture toughness both lead to a transition from this tendency, in which hydraulic fractures cross natural fractures rather than exploiting them. As a result, changing the well orientation in these scenarios are less efficient than the industry standard. The results of this study demonstrate a window of geomechanical scenarios in which changing well orientation is beneficial, can be produced as a fast-screening method for engineers to determine the optimised well orientation.

Research on fracture propagation of hydraulic fracturing in a fractured shale reservoir using a novel CDEM-based coupled HM model

Researchers have recently realized that the natural fractures in shale reservoirs are often cemented or sealed with various minerals. However, the influence of cement characteristics of natural fracture on fracturing network propagation is still not well understood. In this work, laboratory-scaled experiments are proposed to prepare model blocks with discrete orthogonal fractures network with different strength of natural fracture, in order to reveal the influence of cemented natural fractures on the interactions between hydraulic fractures and natural fractures. A series of true triaxial hydraulic fracturing experiments were conducted to investigate the mechanism of hydraulic fracture initiation and propagation in model blocks with natural fractures of different cement strength. The results present different responses of interactions between hydraulic and natural fractures, which can be reflected on the pump pressure profiles and block failure morphology. For model blocks with fluctuated pump pressure curves, the communication degree of hydraulic and natural fractures is good, which is confirmed by a proposed new index of “P-SRV.” The most significant finding is that too high and too low strength properties of cemented natural fracture are adverse to generate complex fracturing network. This work can help us better understand how cemented natural fractures affect the fracturing network propagation subsurface and give us reference to develop more accurate hydraulic fracturing models.

The Marcellus Shale play has attracted much attention in recent years. Our full understanding of the complexities of the flow mechanism in matrix, sorption process and flow behavior in complex fracture system (natural and hydraulic) still has a long way to go in this prolific and hydrocarbon rich formation. In this paper, we present and discuss a novel approach to modeling and history matching of hydrocarbon production from a Marcellus shale asset in southwestern Pennsylvania using advanced data mining & pattern recognition technologies. In this new approach instead of imposing our understanding of the flow mechanism, the impact of multi-stage hydraulic fractures, and the production process on the reservoir model, we allow the production history, well log, and hydraulic fracturing data to force their will on our model and determine its behavior. The uniqueness of this technique is that it incorporates the so-called "hard data" directly into the reservoir model, such that the model can be used to optimize the hydraulic fracture process. The "hard data" refers to field measurements during the hydraulic fracturing process such as fluid and proppant type and amount, injection pressure and rate as well as proppant concentration. The study focuses on part of Marcellus shale including 135 wells with multiple pads, different landing targets, well length and reservoir properties. The full-field history matching process was completed successfully. Artificial Intelligence (AI)-based model proved its capability in capturing the production behavior with acceptable accuracy for individual wells and for the entire field.

Development of tight or low permeability reservoirs (e.g. Bakken formation) commences to reveal its potential and significance and expected to become even more essential in the future. The Canadian Bakken has been produced for over 40 years. At present, the combination of horizontal well drilling and the new multi-stage fracturing and completion technologies has been the key to economically unlocking the vast reserves of the Bakken formation. Parametric studies of fracture geometry on well productivity and oil recovery are essential for a successful hydraulic fracturing stimulation treatment. In this paper, we will investigate the best hydraulic fracturing scenario for Bakken formation and the effects of fracture geometries on oil recovery. More specifically, a geological model is firstly built using typical fluid and rock properties in Bakken formation; the horizontal wells with multistage hydraulic fractures will then be modeled using local grid refinement (LGR) and history matched based on the field production data. Various fracture treatment designs are evaluated, which include the investigation of various fracture spacing, lengths, and fracture locations on the horizontal wells. The results of this study show that, employing multistage hydraulic fracturing along horizontal wells can significantly improve oil recovery in Bakken formation. Compared to fracture length, fracture spaces or fracture numbers have greater impacts on oil recovery. More importantly, for the same fractured volume in the reservoir (i.e., the same usage of proppant and fracturing fluid), fractures with longer and smaller fracture numbers lead to a similar oil recovery compared to shorter but larger fracture numbers along the horizontal well laterals.

SummaryUnderstanding the mechanism of multicomponent mass transport in the nanopores of unconventional reservoirs, such as Eagle Ford, Niobrara, Woodford, and Bakken, is of great interest because it influences long-term economic development of such reservoirs. Thus, we began to examine the phase behavior and flow characteristics of multicomponent flow in primary production in nanoporous reservoirs. Besides primary recovery, our long-term objectives included enhanced oil production from such reservoirs. The first step was to evaluate the phase behavior in nanopores on the basis of pore-size distribution. This was motivated because the physical properties of hydrocarbon components are affected by wall proximity in nanopores as a result of van der Waals molecular interactions with the pore walls. For instance, critical pressure and temperature of hydrocarbon components shift to lower values as the nanopore walls become closer. In our research, we applied this kind of critical property shift to the hydrocarbon components of two Eagle Ford fluid samples. Then, we used the shifted phase characteristics in dual-porosity compositional modeling to determine the pore-to-pore flow characteristics, and, eventually, the flow behavior of hydrocarbons to the wells. In the simulation, we assigned three levels of phase behavior in the matrix and fracture pore spaces. In addition, the flow hierarchy included flow from matrix (nano-, meso-, and macropores) to macrofractures, from macrofractures to a hydraulic fracture (HF), and through the HF to the production well.From the simulation study, we determined why hydrocarbon fluids flow so effectively in ultralow-permeability shale reservoirs. The simulation also gave credence to the intuitive notion that favorable phase behavior (phase split) in the nanopores is one of the major reasons for production of commercial quantities of light oil and gas from shale reservoirs. It was determined that the implementation of confined-pore and midconfined-pore phase behavior lowers the bubblepoint pressure, and this, in turn, leads to a slightly higher oil recovery and lesser gas recovery. Also it was determined that the implementation of midconfined-pore and confined-pore phase-behavior shift reduces the retrograde liquid-condensation region, which in turn, leads to lower liquid yield while maintaining the same gas-production quantity. Finally, the important reason that we are able to produce shale reservoirs economically is “rubblizing” the reservoir matrix near HFs, which creates favorable permeability pathways to improve reservoir drainage. This is why multistage hydraulic fracturing is so critical for successful development of shale reservoirs.

Compared to conventional reservoirs, shale reservoirs are characterize by the extraordinary low porosity and permeability, which contribute to the large-scale application of the technology of hydraulic fracturing in the development. The flow characteristics of shale oil is quite complicated in the hydraulic and natural fractures and the flow characteristics of shale oil in fractures is vital significant to oilfield development. In this paper, a model is developed to simulate the flow characteristics in fractured shale reservoirs, which fully coupling the geomechanics and fluid flow. Specifically, an embedded discrete fracture model is adopted to characterize the hydraulic fractures and the flow between natural and hydraulic fractures. Correspondingly, the Multiple Interacting Continua model is adopted to characterize the flow between matrix and natural fractures. The influence on production caused by the stress sensitivity of two kinds of fractures is considered. Then, the sequential iterative solution method is adopted to solve the model which coupled the geomechanics and fluid flow, and the finite volume method and the finite element method are applied to solve the fluid-governing equations and geomechanics-governing equations, respectively.

The impacts of the fluid type and stress shadow on proppant transport during hydraulic fracturing in a Marcellus shale horizontal well with multiple hydraulic fracture stages were investigated. The available data from a Marcellus Shale horizontal well were utilized to determine the hydraulic fracture properties and to develop a reservoir model. The reliability of the model was confirmed by the close agreement between the predicted production by the model and the well production data. The impact of fluid type and stage spacing on proppant transport during hydraulic fracturing and gas production from Marcellus shale horizontal well were then investigated by employing the reservoir models. Slickwater, compared to crosslinked gels and high viscosity friction reducers (HVFRs), is more efficient in horizontal Marcellus shale wells that are fractured in multiple stages. The application of the HVFRs and crosslinked gels resulted in fracture height growth out of the pay zone. While the use of slickwater directed all stimulation energy towards the pay zone and resulted in improved hydraulic fracture properties in the Marcellus shale and lead to high gas recovery. INTRODUCTION The advancements in horizontal drilling and hydraulic fracturing technologies have made shale production a reality for the past decade. With the technical and economical limitations posed by hydraulic fracturing operations in unconventional reservoirs, it is imperative to create an optimized fracturing design to achieve economic production. A number of treatment design considerations including the type of fluid, proppant type and size, and the additives must be evaluated to achieve successful well stimulation results. In general, proppant transport is dependent on proppant size, density, and type; fluid rheology; temperature changes; pumping rate, among others. The fracturing fluid should be compatible with the formation and promote the increase in the stimulated volume. In order to transport the proppants deep into the fracture, the fluid should be capable of keeping the proppants suspended. With an inherent viscosity, fluid should be able to develop the necessary fracture width to accept proppants. Also, it should have low friction pressure loss and sufficient stability to retain its viscosity throughout the treatment and finally, it should be easy to remove after the treatment. The type of fluid used therefore has pronounced impact on the transport of proppants and the properties of the created hydraulic fracture and consequently the hydrocarbon production. Proppant selection is a contributing factor in the dimensions of a propagating fracture. The most prevalent hydraulic fracturing treatment fluids in shale formations is slickwater with no significant amount of thickening agent. The application of this low viscosity fluid system contradicts conventional fracturing theory. However, slickwater is the preferable fracturing fluid in shale because of its cost-effectiveness and low potential for damaging the shale. In addition, slickwater has the potential to prevent high fracture height growth leading to better fracture containment within the pay zone.

Production data analysis of shale gas using fractal model and fuzzy theory: Evaluating fracturing heterogeneity

Accurate modeling of hydrocarbon production is a necessary, yet challenging step for economic exploitation of unconventional resources. One of the main challenges is to model flow to a horizontal well from a complex network of hydraulic and natural fractures. Many unconventional reservoirs comprise well-developed natural fracture networks with multiple orientations and complex hydraulic fracture patterns based on microseismic data. Conventional dual porosity and dual permeability models are not adequate for modeling these complex networks of natural and hydraulic fractures. Also, it is neither practical nor advantageous to model a large number of pre-existing fractures with a discrete fracture model. Therefore, an appropriate approach to model production from low-permeability reservoirs is to perform discrete fracture modeling for hydraulic fractures and employ a dual continuum approach for numerous natural fractures. We have developed a coupled dual continuum and discrete fracture method to simulate production from unconventional oil and gas reservoirs. Large-scale hydraulic fractures (macro-fractures) are modeled explicitly using a discrete fracture model, called EDFM, and numerous small-scale natural fractures (micro-fractures) are modeled using a dual continuum approach. The hybrid model includes three domains: matrix, discrete-fracture, and continuum-fracture domains. A systematic approach is devised to calculate transport parameters between all three domains. Moreover, EDFM allows for not only transverse and longitudinal hydraulic fractures but also macro-fractures of any orientation. Thus, the coupled model provides an effective and reliable environment to improve stimulation designs and completion strategies. We present several examples in this study to show the applicability, robustness, and performance of the hybrid method for the simulation of unconventional oil and gas reservoirs. We examine multi-stage hydraulic fractures with multiple configurations in the presence of numerous pre-existing fractures. Simulations show a noticeable contribution from natural fracture networks on total production. Furthermore, for the tight oil reservoir examined in this study, the stimulation scheme with longer hydraulic fractures improves cumulative oil production compared to the scheme with larger number of shorter hydraulic fractures. We also examine production from a tight gas reservoir wherein hydraulic fractures partially penetrate the formation height. Simulations indicate that inefficient fracture treatment can result in significant loss of production.

Numerical investigation on the propagation behavior of hydraulic fractures in shale reservoir based on the DIP technique

Summary This paper presents new guidelines for determining net pay thickness in low-permeability, multilayered gas wells. These criteria were developed from a sensitivity study performed with an analytical solution for compmlex multilayered reservoirs. The purpose of this study was to determine whether many layers now considered to contribute to net pay actually have transmissibilities too low for the layer to be productive, causing performance projections from current single-layer descriptive models to be too optimistic. Introduction Devonian shales are multilayered, frequently naturally fractured reservoirs with a large variance in permeability between layers. Many Devonian shale wells have had hydraulic fracture treatments to increase productivity; however, the fractures may exist only in some of the many layers of the reservoir. According to conventional post-fracture pressure-buildup test analysis, stimulation of the Devonian shale wells has fallen short of design expectations. This paper presents results from a study using an analytical solution for multilayered wells that suggest that conventional analysis of tests from these wells is inadequate because of the simplicity of the reservoir model used to describe them. We developed a new analytical solution, based on Laplace transform methods, similar to Spath et al.'s1 recent solution. This solution accurately models the complexity of Devonian shale wells. Devonian shale completions frequently consist of at least five layers commingled at the wellbore. Some layers are naturally fractured, interspersed with homogeneous-acting silt layers. Some or all of the layers may contain a hydraulic fracture, and some layers may be of limited areal extent. Gas production from these wells is at a constant bottomhole pressure (BHP). The analytical solution we have programmed can model all these conditions. We studied two-layer systems systematically to find a minimum transmissibility for a layer to be considered part of the net pay thickness. This study considered homogeneous or naturally fractured layers, with and without a hydraulic fracture in the layer. The results from this two-layer study were extended and confirmed for multilayered cases. Unless layers are tested individually, conventional pressure-buildup analysis results in a total transmissibility, kh, and an equivalent fracture half-length for the entire net pay zone. When these results are used in performance projections with a single-layer model, forecasts have consistently proved to be optimistic. We used the analytical solution to compare the impact of a fracture treatment that affects only one layer in a five-layer system with a fracture of the same length in the total system. The input parameters for the analytical simulation for the five-layer case were obtained from geophysical logs, production logs, and well tests in typical eastern Devonian shale wells. We illustrate the power of the model through a field case described with individual-layer test results. Here, we first describe the model used to simulate the complex Devonian shale formation. We verified this model using examples from the literature2-6 and a commercial analytical simulator; detailed results are given in Ref. 3. We then present our results from both the two-layer and the multi-layer studies. Finally, we state conclusions drawn from the results of the research. Model Description The analytical solution can model n layers. Each layer can be either naturally fractured, with pseudo-steady-state or transient interporosity flow, or homogeneous-acting. Individual layers may have hydraulic fractures of various lengths, with either infinite conductivity or uniform flux. The layers can be infinite-acting or can act as closed systems with various drainage areas. Each layer can have a different permeability, thickness, porosity, wellbore radius, skin, fracture half-length, or natural fracture characteristics. The multilayered system is commingled at the wellbore, with no crossflow between layers. The well may be produced at constant BHP or at constant total flow rate. The solutions were developed for slightly compressible liquids, but the use of normalized pseudopressure in the solution allowed us to consider gas production. Isothermal conditions, negligible gravitational effects, production of a single-phase fluid, laminar flow behavior (gas velocities low enough that non-Darcy flow effects are negligible), permeability independent of pressure, and a viscosity/compressibility, µct, product that is constant with pressure were assumed. The Appendix presents a derivation of the analytical solution. This reservoir model will have many applications in reservoir description beyond those discussed.

Summary Advancements in horizontal-well drilling and multistage hydraulic fracturing have enabled economically viable gas production from tight formations. Reservoir-simulation models play an important role in the production forecasting and field-development planning. To enhance their predictive capabilities and to capture the uncertainties in model parameters, one should calibrate stochastic reservoir models to both geologic and flow observations. In this paper, a novel approach to characterization and history matching of hydrocarbon production from a hydraulic-fractured shale is presented. This new methodology includes generating multiple discrete-fracture-network (DFN) models, upscaling the models for numerical multiphase-flow simulation, and updating the DFN-model parameters with dynamic-flow responses. First, measurements from hydraulic-fracture treatment, petrophysical interpretation, and in-situ stress data are used to estimate the initial probability distribution of hydraulic-fracture and induced-microfracture parameters, and multiple initial DFN models are generated. Next, the DFN models are upscaled into an equivalent continuum dual-porosity model with analytical techniques. The upscaled models are subjected to the flow simulation, and their production performances are compared with the actual responses. Finally, an assisted-history-matching algorithm is implemented to assess the uncertainties of the DFN-model parameters. Hydraulic-fracture parameters including half-length and transmissivity are updated, and the length, transmissivity, intensity, and spatial distribution of the induced fractures are also estimated. The proposed methodology is applied to facilitate characterization of fracture parameters of a multifractured shale-gas well in the Horn River basin. Fracture parameters and stimulated reservoir volume (SRV) derived from the updated DFN models are in agreement with estimates from microseismic interpretation and rate-transient analysis. The key advantage of this integrated assisted-history-matching approach is that uncertainties in fracture parameters are represented by the multiple equally probable DFN models and their upscaled flow-simulation models, which honor the hard data and match the dynamic production history. This work highlights the significance of uncertainties in SRV and hydraulic-fracture parameters. It also provides insight into the value of microseismic data when integrated into a rigorous production-history-matching work flow.