Growing concerns over global climate change is also increasing the interest in developing technologies to reduce the concentration of carbon dioxide (CO2) in the atmosphere. Geological Carbon Sequestration which injects CO2 into subsurface formations such as deep saline aquifers, depleted hydrocarbon reservoirs are some of the viable options to reduce CO2 emissions into the atmosphere. Deep saline aquifers are a particularly good choice due to their wide geographical distribution and proximity to emission sources that provide easy accessibility and storability of CO2. Saline aquifers, however, are data-poor systems that requires a thorough understanding of the impact of all factors and their uncertainties on long-term CO2 storage for risk assessment. This study considers a wide range of geologic and dynamic model uncertainties, including structural uncertainties, petrophysical heterogeneities, dynamic flow parameters and geochemical reactions to define the most critical parameters for different CO2 trapping mechanisms. Experimental design over stochastic modelling and sampling is used to minimize the computational cost of the study. A commercially available, compositional reservoir simulator with reactive transport modeling capability has been used in this study to account for the impact of potential geochemical reactions. Assumed reactive transport model considers the water solubility, ionization, and mineral trapping mechanisms of CO2 in saline reservoirs for a range of components of dissolved salts in the brine as well as the pH of the brine. Chemical reactions that may occur with the rock minerals (Kaolinite, Anorthite and Calcite) when adding the CO2 are prescribed as input into the simulator. Simulations also include effects of hysteresis and diffusion processes. Study shows that the most important parameter for all trapping mechanisms is the permeability since it controls the injection capacity. Beyond that, different combination of parameters with different range of uncertainties alter the ranking of factors influencing dissolution, ionization, hysteresis, mineralization and the movement of CO2 within the aquifer. There is no unique set of parameters that maximizes all storage mechanisms. There is a significant overlap and change in amount the amount of CO2 stored via different mechanisms depending on the parameter combinations. Study results also provide insights into how one can prioritize data gathering needs depending on the objectives, data uncertainties and their sensitivities for an aquifer site under consideration.

Fracture diagnostic data for shale wells show that the fracture system after hydraulic fracturing is quite complex. Accurate and efficient simulation modeling of complex hydraulic and natural fracture networks is critical for evaluation of well performance and stimulation effectiveness in unconventional oil and gas reservoirs. The traditional method based on local grid refinement (LGR) has limitations in handling 3D hydraulic and natural fracture geometry. In addition, its computational efficiency is low, especially for dealing with a large number of hydraulic fractures and multiple horizontal wells. In this study, we develop a new workflow which, for the first time, utilizes embedded discrete fracture model (EDFM) method coupled with a parallel reservoir simulator (PRS) to simulate all types of 3D hydraulic and natural fractures. EDFM can easily avoid re-gridding matrix cells containing hydraulic and natural fractures. More realistic 3D fracture geometry from either fracture propagation simulation or user definition and geological model with corner point can be accurately honored. The input fractures can be smoothly embedded into the model grid through EDFM processing We validated the parallel reservoir simulator with EDFM by comparing the simulation results using dual porosity dual permeability (DPDP) model. After benchmarking, we applied this new workflow to simulate three synthetic field cases. The simulation results are also compared with a commercial reservoir simulator (CRS) with the EDFM method. Well performance for the real case with and without natural fractures can be efficiently simulated. The new EDFM workflow enables to model 3D hydraulic and natural fractures with any strike and dip angels efficiently and accurately in the parallel reservoir simulator. Modifications of fracture properties can be easily done. This new workflow enables a much faster and more robust fracture modeling process, which is highly effective for the fracture model calibration and development optimization in unconventional oil and gas reservoirs.

This study presents a new workflow by integrating a robust 3D hydraulic fracture propagation model in conjunction with reservoir simulation through the embedded discrete fracture model (EDFM). Specifically, the hydraulic fracture model is applied to simulate more realistic fracture geometry considering 3D geomodel with rescue format and 3D natural fractures. Multiple wells with varying well spacing are considered. After fracture simulation, the predicted hydraulic fractures, activated natural fractures, and non-activated natural fractures are easily transferred to the reservoir simulator. The EDFM method can accurately and efficiently deal with the 3D complex fractures. It can avoid the complex gridding process and low computational efficiency of the traditional local grid refinement method. Based on long-term performance of different scenarios with varying well spacing, the optimal well spacing is determined. The new workflow was applied to optimize well spacing in a real shale gas case with 3D geomodel and natural fractures. The geomodel with rescue format, 3D natural fracture, 3D non-uniform distribution of stress and rock properties are all honored in the fracture model. Two horizontal wells with varying well spacing of 200 m, 300 m, and 400 m with actual completion data were considered and simulated. Subsequently, the same reservoir and fracture model was used to perform production simulation for 20 years. The distribution of heterogenous properties in matrix was considered. By comparing the estimated ultimate recovery of different well spacing scenarios, the optimal well spacing was identified. This work, for the first time, performs well spacing optimization by coupling fracture model and reservoir model considering more realistic geomodel and fracture system. This new workflow significantly improves the accuracy and efficiency of well spacing optimization process in unconventional oil and gas reservoirs compared to the traditional workflow presented in the literature.

In modeling of fractured ultratight gas-condensate reservoirs it is particularly important to represent liquid dropout, and its impact on fluid flow behavior, accurately. This, in turn, allows for design and optimization of reservoir production strategies, to mitigate potential production loss below the dewpoint pressure. Unfortunately, the traditional transfer functions used to describe the mass transfer between matrix/fracture segments in dual-porosity models (DPM’s) do not explicitly represent the complex physics of the inherent multiphase problem, which may complicate the application of existing simulators. Therefore, the objective of this work is to develop a two-phase transfer function that improves the modeling of gas-condensate systems via a DP representation. Commercial simulators typically offer a simple matrix/fracture mass transfer model to perform calculations in DP systems: The default shape factors are based on a pseudo-steady state (PSS) approximation, and for multiphase problems, the transfer function of Kazemi et al. (1976) is commonly used. Without sub-gridding, the traditional transfer functions cannot represent the extended transient behavior observed in ultratight fractured reservoirs. Furthermore, in compositional modeling of gas-condensate reservoirs, the averaging of matrix block properties masks the compositional changes near the matrix/fracture interface, that is critical to prediction of production behavior below the dewpoint. To overcome the limitations of the current models, we introduce a new transfer function that includes 1) a time-dependent shape factor and 2) a modified representation of the two-phase flow. We utilize the correction factor introduced by Zhang et al. (2022) allowing the shape factors to vary with time. For two-phase problems, instead of evaluating the mass transfer at an average pressure and composition, we propose a novel approach that evaluates the transfer function at conditions that reflect/approximate the pressure and composition/saturation gradients within the matrix block. By approximating the fluid state near the matrix/fracture interface, we represent the liquid buildup that dictates the fluid mobility more accurately. The open-source MATLAB Reservoir Simulation Toolbox (MRST) was used as a platform for the work presented in this paper: The existing single-porosity (SP) compositional model of MRST was first extended and validated (with the analytical solution for the single-phase pressure diffusion equation) to allow for DP modeling and simulation of condensate systems. We present calculation results for a single-block DPM, representative of a fractured ultratight gas-condensate reservoir, to demonstrate the limitations of the traditional mass transfer modeling approach. We then introduce the new transfer function and compare all calculation results with fine-grid SP reference models. In this work, we consider two condensate fluid descriptions: 1) a 4-component analog rich gas, and 2) a realistic 24-component fluid description from a gas-condensate reservoir. We demonstrate that the new transfer function represents the physical mechanisms at play more accurately and provides for a substantial improvement over the traditional formulation in terms of the amounts and compositions of the produced phases. The proposed method is motivated by the physics of the problem and requires no sub-gridding. The enhanced simulation of gas-condensate systems enables improved decision making in reservoir management and supports optimization of field development plans. In addition, the improved calculation of the surface streams, obtained from the new method, provides for a more accurate design of field surface facilities.

Embedded discrete fracture models (EDFM's) are commonly used to study fluid flow in unconventional reservoirs. The EDFM, however, requires extensive pre-processing of non-neighboring connections. More importantly, it becomes further demanding when grid refinement is required to accurately model mass transfer between the fractures and the tight matrix rock, primarily to resolve transients in the matrix. This renders the EDFM an expensive option to model large-scale fractured reservoirs where the number of fracture segments can be substantial. In this paper, an upscaling procedure is proposed to construct a dual-porosity model (DPM) from a discrete fracture characterization. The implicit representation of the fractures provides for an improved simulation efficiency. While the DPM's are often perceived as simple sugar-cube representations of complex fracture networks, the upscaling technique presented here, demonstrates the capability of DPM's in providing accurate and efficient solutions for a broad range of complex fractured systems. The potential of the DPM is unlocked via application of a flexible matrix/fracture transfer function, that is similar in form to the generalized Vermeulen transfer function (gVer) introduced by Zhang et al. (2022). Unlike the traditional class of transfer functions, first introduced by Warren and Root (1963), the gVer transfer function incorporates a time-dependent correction factor to resolve the transient within the matrix without the need for sub-gridding. The use of an advanced DPM becomes key to our proposed upscaling workflow which also includes the evaluation of effective rock properties and a unique set of matrix/fracture interaction parameters. We first examine several test cases of ultra-tight fractured systems and establish the need for advanced modeling techniques to accurately capture the matrix transient, as observed from single-porosity reference models (SPM's). In the context of EDFM, this can be accomplished by refining the mesh for the matrix grid at an increased computational cost, while the use of gVer transfer function to describe the mass transfer in the DPM is demonstrated to accomplish the same at a marginal computational cost. We then apply the approach to solve more complex problems including models with fracture networks, exhibiting arbitrary fracture orientations and geometries, in a heterogeneous nano-darcy rock. Calculations for such examples are performed using EDFM, and an equivalent upscaled DPM. We furthermore demonstrate the versatility of our approach by refining the DPM in regions of higher spatial variation to capture the details of the fractured rock as dictated by the original EDFM representation. The proposed upscaling procedure overcomes the commonly assumed limitation of the classical DP approach and allows for modeling of unconventional reservoirs without losing the realism of a discrete fracture characterization. It is demonstrated that the proposed technique reproduces the correct response with significantly fewer grid-blocks and hence enables reduction of computational cost of advanced optimization studies in unconventional reservoirs.