CREATING A RESERVOIR MESH USING A PROLONGED ANISOTROPIC REFINEMENT, ADAPTED TO GEOMETRY, OF A POLYED

专利摘要:
Systems and methods are proposed for creating reservoir meshes using an extended anisotropic refinement, adapted to the geometry, of a polyhedron. In one example, a method includes identifying, based on the specification of the reservoir (204), a set of fractures (206, 208) including fractures (206) for the 2.5 (2.5D) dimension, and other fractures (208). The method also includes generating an intermediate reservoir model (212) including an extrusion mesh that models fractures (206) for 2.5D in a 3D space. In response to the determination that the cells in the mesh must be refined in a direction within the 3D space, the method refines the cells anisotropically in the mesh corresponding to other fractures (208). The method also includes resolving a fracture network within the intermediate reservoir model (212) using the refined cells, and then generating a terrestrial reservoir model using the fracture network.
公开号:FR3028333A1
申请号:FR1559673
申请日:2015-10-12
公开日:2016-05-13
发明作者:Michael Loyd Brewer；Steven Bryan Ward；Gerrick Bivins
申请人:Halliburton Energy Services Inc；
IPC主号:

专利说明:

[0001] TECHNICAL FIELD [0001] The present disclosure generally relates to the modeling of hydrocarbon reservoirs and, more particularly, to techniques for creating reservoir meshes using an extended anisotropic refinement, adapted to the geometry, of a polyhedron. BACKGROUND [0002] In the oil and gas industry, reservoir modeling involves the construction of a computer model of an oil reservoir to improve reserve estimation and decision making for the reservoir. development of the field. For example, geological models can be created to obtain a static description of a reservoir before production. Tank simulation models can be created to simulate and predict fluid flow in a reservoir during its production life. [0003] A problem encountered with reservoir simulation models is the challenge of modeling fractures inside a reservoir, which requires a thorough understanding of the characteristics of the matrix flow, the connectivity of the fracture network and of the matrix-fracture interaction. Fractures may be open cracks or voids within the formation, and may be natural or artificially generated from a wellbore. Precise fracture modeling is important because fracture properties such as spatial distribution, aperture, length, height, conductivity, and connectivity significantly affect the flow of fluid from the reservoir to the well. drilling. [0004] Mesh generation techniques are used in reservoir modeling. Two traditional mesh generation techniques for 3D tank simulation are the structure-based mesh and the extrusion-based mesh. In structured techniques, hexahedra are connected in a 3D logical space i-j-k in which each inner mesh node is adjacent to 8 hexahedra. The extensions of the structured techniques include a local refinement of the mesh in which local regions of an original grid are replaced by finer grids. This can be time-consuming, expensive in computing resources, and prohibitively expensive when dealing with general tank geometries, such as arbitrary 3D fracture surfaces. Due to the 2.5 dimensional inherent nature of existing extrusion techniques, similar limitations apply to these techniques. In addition, there are completely unstructured meshing techniques, such as, for example, tetrahedral and polyhedral meshing schemes. The increased complexity of these techniques often results in a decrease in robustness compared to structured techniques, particularly in the presence of imperfect geometry input (ie, "dirty geometry"). In one or more embodiments of this disclosure, a computer implemented method includes: receiving, at a computing device, a tank specification; identification, based on tank specification, of a set of fractures comprising: fractures allowing 2.5 dimension (2.5D); and other fractures; the generation of an intermediate reservoir model comprising an extrusion mesh that models fractures allowing 2.5D in a three-dimensional space (3D) responding to the determination that the cells in the mesh must be refined in one direction to the inside the 3D space, the refinement of cells anisotropically in the mesh corresponding to other fractures; resolving, by the computing device, a fracture network within the intermediate reservoir model using the refined cells; and the generation of a terrestrial reservoir model using the fracture network. In one or more embodiments of the present disclosure, 2.5D fractures have a geometry that has been discretized in a 2D plane by a collection of line segments, and the generation of the intermediate reservoir model includes , for each line segment associated with each fracture in the 2.5D fractures: the generation of a set of stages at a specified radius from the line segment; generation of closed loops around fracture line segments associated with a 2.5D fracture; and generating the shape elements within the closed loops of the segment of the line, the method also comprising: generating the meshing mesh constrained around the closed loops of the fractures allowing the 2.5D to fill a remaining space in the 2D plane. In one or more embodiments of the present disclosure, the method also comprises: the use of a terrestrial tank model in a reservoir simulation; and displaying the tank simulation on a screen of the computing device. In one or more embodiments of the present disclosure: fractures allowing 2.5D may comprise fractures represented by one or more substantially vertical planes; and the other fractures comprise fractures represented by one or more substantially horizontal planes.
[0002] In one or more embodiments of the present disclosure, the anisotropic refinement of the cells in the mesh comprises: identifying a direction within the 3D space in which the cells are to be refined; and dividing a cell rim, the rim being in the direction within the 3D space. In one or more embodiments of the present disclosure, the flange is divided along a plane that is perpendicular to an axis of a fracture within a fracture network.
[0003] In one or more embodiments of the present disclosure, the anisotropic refinement of the cells comprises: determining a location of a fracture in an intermediate reservoir model; the identification of cells in the region close to the fracture; determining a target size of the edges of cells near the fracture; and determining whether the edges of cells near the fracture in the direction within the 3D space are longer than the size of the target multiplied by a scalar variable, and the cells close to the fracture having edges in the direction inside the 3D space that are longer than the size of the target multiplied by the scalar variable are among the cells that are refined. In one or more embodiments of the present disclosure, the determination of the location of the fracture includes: staggering the edges of cells near the fracture that are longer than the size of the target multiplied by the variable scalar ; and determining whether the stepped flanges intersect the fracture, wherein the near-fracture cells that have stepped flanges that intersect the fracture are cells that are refined.
[0004] In one or more embodiments of the present disclosure, the determination of the location of the fracture also includes: staggering the edges of the cells near the fracture that are longer than the size of the target multiplied by the scalar variable; and for any opposing pair of stepped flanges, determining exactly one of the two stepped flanges intersecting a fracture, wherein the cells close to the fracture that have exactly one step flange crossing the fracture are the cells that are refined.
[0005] In one or more embodiments of the present disclosure, the determination of the location of the fracture also includes: determining whether the edges of cells near the fracture that are longer than the size of the target multiplied by the scalar variable have two or more hanging nodes, wherein the cells close to the fracture having two or more hanging nodes are refined.
[0006] In one or more embodiments of the present disclosure, the anisotropic refinement of the cells in the mesh also includes, for any opposite pair of flanges of near-fracture cells that are longer than the size of the target multiplied by the scalar variable, the determination if exactly one of the two edges comprise two or more suspended nodes, and in which the cells close to the fracture which comprise two or more suspended nodes are refined. In one or more embodiments of the present disclosure, the anisotropic refinement of the cells in the mesh also comprises: staggering of the edges of the cells close to the fracture which are longer than the size of the target multiplied by the variable scalar ; and determining whether cells near the fracture having stepped flanges intersect a fracture while no near fracture-like cells cross a fracture, wherein the near-fracture cells that have stepped flanges that intersect the fracture are refined . In one or more embodiments of the present disclosure, the anisotropic refinement of cells in the mesh further includes: determining a location of a fracture in an intermediate reservoir model; the identification of cells in the region close to the fracture; the identification of a direction within the 3D space in which the cells close to the fracture must be refined; and determining whether the cells near the fracture having a ridge in the identified direction also have not more than one hanging knot, wherein the cells close to the fractures that do not have more than one hanging knot are refined. In one or more embodiments of the present disclosure, the anisotropic refinement of the cells in the mesh further includes: determining a location of a fracture in an intermediate reservoir model; the identification of cells in the region close to the fracture; and for any opposite pair of cell flanges proximate to the fracture, determining if exactly one of the two opposite flanges has two or more suspended nodes, wherein the cells close to the fracture that have the two or more suspended nodes are refined. In one or more embodiments of the present disclosure, a computer readable storage medium has executable instructions stored on said storage medium, which, when executed by a computing device, enables the computing device to perform operations, the instructions comprising: instructions for receiving a tank specification; instructions for identifying, based on tank specification, a set of fractures comprising: 2.5 dimensional (2.5D) size fractures; and other fractures; instructions for generating an intermediate reservoir model comprising an extrusion mesh that models fractures allowing 2.5D in a three-dimensional (3D) space; responding to the determination that the cells in the mesh must be refined in a direction within the 3D space, instructions for the anisotropic refinement of the cells in the mesh corresponding to the other fractures; instructions for resolving a fracture network within the intermediate reservoir model using the refined cells; and instructions for generating an Earth tank model using the fracture network. In one or more embodiments of the present disclosure, the instructions for anisotropic refinement of the cells in the mesh include: instructions for identifying a direction within the 3D space in which the cells must be refined; and instructions for dividing a cell rim, the rim being in the direction within the 3D space.
[0007] In one or more embodiments of the present disclosure: fractures allowing 2.5D may comprise fractures represented by one or more substantially vertical planes; and the other fractures comprise fractures represented by one or more substantially horizontal planes. In one or more embodiments of the present disclosure, a system comprises: a processor; a display device; and a memory having instructions stored thereon which, when executed by a processor, enables the processor to perform operations including: receiving a tank specification; identification, based on tank specification, of a set of fractures comprising: fractures allowing 2.5 dimension (2.5D); and other fractures; the generation of an intermediate reservoir model comprising an extrusion mesh that models fractures allowing 2.5D in a three-dimensional space (3D) responding to the determination that the cells in the mesh must be refined in one direction to the inside the 3D space, the refinement of cells anisotropically in the mesh corresponding to other fractures; resolving a fracture network within the intermediate reservoir model using the refined cells; and the generation of a terrestrial reservoir model using the fracture network. In one or more embodiments of the present disclosure: fractures allowing 2.5D include fractures represented by one or more vertical planes; and other fractures including fractures represented by one or more substantially horizontal planes.
[0008] In one or more embodiments of the present disclosure, the system also comprises: displaying, on the display device, the terrestrial tank model.
[0009] Brief Description of the Drawings [0005] FIG. 1 is a flowchart of an extended anisotropic refinement system, adapted to the geometry, allowing the creation of reservoir meshes according to some illustrative examples of the present disclosure. [0006] FIG. 2 is a schematic diagram illustrating a method for creating reservoir meshes using an extended, geometry-adapted anisotropic refinement of a polyhedron, according to some illustrative examples of the present disclosure. [0007] FIG. Figure 3 illustrates examples of refinement configuration for a single hexahedron cell within a U-V-W 3D space, according to some illustrative examples of the present disclosure. Figures 4 and 5 illustrate representative triangular prism cells showing a logical coordinate system in 2D U-W and a number scheme for the edges of the prisms, according to some illustrative examples of the present disclosure. [0009] FIG. Figure 6 illustrates refinement patterns for a single prism cell using a single U-direction paradigm within the 2D U-W logic coordinate system, according to some illustrative examples of the present disclosure. [0010] FIG. 7 illustrates 4-way refinement patterns utilizing a U-directional paradigm within the 2D U-W logic coordinated system, according to some examples of the present disclosure. [0011] FIG. Figure 8 illustrates an interactive user interface for displaying a model containing hexahedron and triangular prisms, according to some aspects of the present disclosure. [0012] FIG. 9 illustrates an interactive user interface for displaying refined fractures in a model using a single U-direction paradigm, according to some illustrative examples of the present disclosure. [0013] FIG. 10 illustrates a representative non-triangular prism (eg, a general prismatic element) showing the logical coordinate system in 2D, according to certain features of the present disclosure. [0014] FIG. 11 illustrates a possible refinement of a general prismatic element by subdividing the element into hexahedra, according to some illustrative examples of the present disclosure. [0015] FIG. 12 illustrates a representative cell showing the local coordinate system in 3D and a number scheme, according to some illustrative aspects of the present disclosure. Figures 13 and 14 illustrate the application of the scaling functions to a cell and its edges, respectively, according to certain illustrative features of the present disclosure. Figures 15 and 16 illustrate two different refinements generated using the illustrative methods of the present disclosure. Figures 17 and 18 are examples of generated meshing images using the illustrative aspects of the present disclosure. Figures 19 and 20 are complete terrestrial models generated using some of the exemplary methods of the present disclosure. Figures 21 and 22 are flow diagrams illustrating an extended anisotropic cell refinement method, adapted to geometry, according to some illustrative examples of the present disclosure. [0021] FIG. Figure 23 illustrates a 3D fracture image modeled using certain illustrative examples of the present disclosure. [0022] FIG. 24 illustrates an example of a set of discretized 3D fractures intersecting a 2D plane, according to some illustrative aspects of the present disclosure. [0023] FIG. 25 is a flowchart of a method for modeling 3D fractures, according to some illustrative examples of the present disclosure. [0024] FIG. 26 is an illustration of the generation of a computerized mesh around a segment of a fracture line, according to certain illustrative features of the present disclosure. [0025] FIG. 27 is an illustration of the generation of a computerized mesh around segments of a cross-fracture line, according to some illustrative examples of the present disclosure. [0026] FIG. 28 illustrates an example of computerized meshes around a complex network of segments of a fracture line, according to some illustrative examples of the present disclosure. [0027] FIG. 29 illustrates an unstructured grid generated around complex geometries that include cross-fracture line segments, according to some illustrative examples of the present disclosure. [0028] FIG. 30 is a diagram of an example of a computer system in which certain embodiments of the present disclosure may be implemented. DETAILED DESCRIPTION [0029] Certain aspects and characteristics relate to the generation of hybrid computerized meshes around complex and discrete fractures in order to simulate reservoirs. Illustrative examples and related methodologies of the present disclosure are described above such that they can be used in a system that applies extended, geometry-adapted anisotropic refinement of a polyhedron (AGAR) to create a mesh. of tank. In one embodiment, an extended AGAR algorithm is used in association with 2.5 D meshers, such as a stadimetric mesher. Examples of stadimetric meshers are described with respect to FIGS. 24-28 In one embodiment, an AGAR algorithm uses a model of the Earth composed entirely of hexahedra as the basic structure in which it solves the fracture networks that are modeled. The extension of the AGAR algorithm allows the systems and methods described here to use more general topologies in an input-based model. The AGAR extension provides a workflow that resolves fracture networks in two phases: a 2.5D mesh phase; and a refinement phase in 3D. This combination can maintain the advantages of computational speed and control orientation of a 2.5D mesher when desired, while also invoking an AGAR algorithm to solve general 3D fracture networks. An extended AGAR algorithm is used so that the embodiments are not limited to the use of a structured mesh. In particular, some embodiments initially use a 2.5D stadimetric mesher that provides an unstructured mesher, and then the use of an extended, enhanced version of the AGAR algorithm for refining 3D fractures. Some examples use AGAR techniques to scale edges and cells. These techniques can be applied directly to additional types of elements, such as, for example, triangular prisms and other general prismatic elements. According to some embodiments, a set of rules is defined to refine the more general elements and to ensure that the refinement of the cell allows a conformal mesh topology. In one embodiment, an AGAR extension still allows anisotropic refinement for the majority of elements in most models while eliminating the restriction that the elements may be hexahedrons. A feature of an example of a combined workflow (see, eg, the workflow shown in Figure 2), is that the user can get the 2.5D mesh relatively quickly and efficiently for the job. counting cells whenever fractures allow it with no specific need to know which fractures allow or not such a solution. Often, the most conductive fractures are represented using a 2.5 D geometry, and in these cases, the examples of techniques presented here can produce a highly refined mesh in these regions, while still resolving all other fractures, although with a lower resolution. Some embodiments can maintain the advantages of the 2.5D mesh, where possible, while also allowing the use of completely general fractures as input. These examples can improve non-extended AGAR algorithms by eliminating the need for elements to be hexahedra. Some examples can also improve the techniques that only depend on a 2.5D stadia mesh by eliminating the need that the fractures used allow a 2.5D mesh. While freeing users from these restrictions, the disclosed AGAR extension and its association with a 2.5D stadian mesh free users from the need to understand, and to be significantly involved in, the mesh process. Illustrative methods described herein can modify a structured grid of a model of the Earth to then create a quantitative mesh that 1) can be used in a reservoir simulation, 2) respects the properties and geometry of the model of the Original Earth and 3) solves additional natural hydraulic fracturing systems. The various examples described herein may use an extended AGAR algorithm to insert fracture networks into an existing Earth model, which exists as a structured grid having physical property values associated therewith. The fracture network may be in the form of a surface mesh (manifold or not) in a 3D space. The structured grid of the Earth model is then anisotropically refined to resolve fractures and allow proper cell rank classification at the near-fracture region. As a rule, the cells can be examined to identify those that need refinement. These cells can be refined in the appropriate directions (thus, anisotropic). When no cell needs refinement, the process can be terminated. The refinement, as used here, describes the division of cells into 3D space which, ultimately, results in the formation of a higher mesh resolution Earth pattern. A fracture network can then be solved within the Earth model using refined cells. The resolution of a fracture network can include the generation of a model of the Earth and the use of a high resolution mesh, in which fractures and areas close to fractures in the model have a higher mesh resolution. higher than the other areas of the model. Since the mesh can be used for a numerical simulation, in general, smaller cells (after refinement) can give more accurate solutions compared to larger cells. Since reservoir simulators generate approximations of mathematical formulas, the smaller the cells, the closer the approximation mimics mathematical formulas. Since the methods apply an extended anisotropic approach when handling the cells, the counting of data and its associated computing needs can be greatly reduced. Therefore, the method can be faster and more efficient than conventional methods that apply brute force analysis, requiring more resource and data intensive calculations. In addition, because of the relationship of the model of the Earth generated compared to the original Earth model, the interpolation of the characteristics is minimized and the model of the Earth generated also respects the values of the original characteristics. Thus, certain embodiments of the present disclosure allow modeling with higher mesh resolution while reducing computational computing requirements. [0036] FIG. 1 is a flowchart of a reservoir simulation and modeling system 100. The system 100 includes a processor 102, a computer-readable non-transitory storage 104, a transmitter / receiver / network communication module 105, input / output devices optional output (1/0) 106 and an optional display device 108 (eg, a display for a user interface), all interconnected through a bus system 109. Software instructions that can be executed by a processor 102 to implement the software instructions stored inside the AGAR engine 110 may be stored in a storage 104 or on any non-transitory computer readable storage medium. Similarly, the instructions executable by a processor 102 for implementing the 2.5D stagger mesh module 116 may be stored in a storage 104. As illustrated in FIG. 1, the stadimensional mesh module 2.5D 116 can implement a 2.5D mesh algorithm or be embodied as a stadimetric mesh 2.5D. Although not explicitly illustrated in FIG. 1, the system 100 may be connected to one or more public or private networks through one or more appropriate network connections. As shown with reference to the computer system 2900 shown in FIG. 29, the software instructions which include the extended AGAR engine 110 and the 2.5D stagger mesh module 116 may also be downloaded to the storage 104 from a CD-ROM or other suitable storage medium through wired connections or not. In addition, the techniques disclosed herein can be realized with a variety of configurations of computer systems, including portable devices, multiprocessor systems, consumer electronics components based on multiprocessors or programmable, mini computers, servers, central computers, etc. Any number of computer systems and computer networks are acceptable for use with the present disclosure. The techniques disclosed herein can be performed in distributed computing environments in which tasks are performed by remote processing devices that are connected through communication networks. For example, the techniques disclosed herein may be realized by a cluster of computing devices operating in a group on a server farm.
[0010] In a distributed computing environment, the program modules can be located both on a local or remote computer storage medium, including memory storage devices. The present disclosure may therefore be implemented in connection with various hardware, software or a combination thereof in a computer system or other processing system. In certain embodiments, the extended AGAR engine 110 comprises a fracture system module 112 and a ground modeling module 114. The fracture system module 112 provides the geometric definition of the fracture network, with Petrophysical and mechanical property distributions for fractures. Examples of fracture system modules can be implemented using, for example, the Fracpro® or Gohfer® platforms. Other platforms and software systems may, however, be used to implement the fracture system module 112. [0039] The extended AGAR engine 110 also includes a model for modeling the earth 114. The modeling model of the Earth 114 integrates with the fracture data contained in the module of the fracture system 112 to allow stratigraphic visualization under the surface of the fractures and the reservoir. Visualization may include, for example, geo-scientific interpretation, oil system modeling, geochemical analysis, stratigraphic mesh, facies, net cell volume, and petrophysical property modeling. In addition, each earth modeling module 114 models well paths, in addition to a cut across the facies and porosity data. Earth modeling platforms include platforms such as, eg, DecisionSpace®, which is commercially available from Halliburton Energy Services Inc., Houston, Texas. A variety of other earth modeling platforms, however, may also be used with or with the present disclosure. The extended AGAR engine 110 may also include processing circuitry for robust data recovery and integration of historical or real-time reservoir-related data that covers all aspects of well planning, processes and construction and completion such as drilling, concreting, cable logging, testing and well simulation. In addition, such data may include drilling data, well trajectories, rock petrophysical property data, rock mechanical property data, surface data, fault, data from surrounding wells, data derived from geostatistics, etc. The database (not shown) that stores this information may be in the module of the fracture system 112 or at a remote location. A database platform is, for example, the INSITE® software, commercially available from Halliburton Energy Services Inc., Housto, Texas. A variety of other database platforms, software platforms, and associated systems can be used to retrieve, store, and integrate well-related data as described herein. In addition, an extended AGAR engine 110 may also include multidomain workflow automation capabilities, which can connect a variety of desired technical applications. Thus, the output from one application, or one module, can become the input for another, thereby providing the ability to analyze how various changes affect well placement and / or fracture design. Tradespeople who are aware of this disclosure realize that there is a variety of workflow platforms that can be used for this purpose. As is generally described above, the methods and embodiments of the present disclosure describe algorithms for the automatic refinement of the mesh to solve characteristics in a reservoir for applications which include, without limitation, the modeling and reservoir simulation.
[0011] An important aspect of the present disclosure is the ability to selectively apply anisotropic refinement to reduce the degrees of freedom for subsequent analyzes using the generated Earth model. Sample Workflow [0043] FIG. 2 illustrates an example of a workflow 200 for the creation of reservoir meshes using an extended anisotropic refinement, adapted to the geometry, of a polyhedron. Input 202 to workflow 200 includes a reservoir model (eg, technical data of reservoir 204), and two sets of fractures 206 and 208. An example input model for input 202 is given in FIG. 8. In one embodiment, the workflow 200 begins upon receiving the technical data of a reservoir provided by a user 204. At block 214, an extended AGAR algorithm is used in association with a stadimetric meshing in step. 2.5D 210. The technical data of tank 204 is the input required for the 2.5D 210 stadian mesh. As illustrated, the 2.5D 210 stadian mesh can be an extrusion mesh. As an additional component of the input 202 in the workflow 200, a user provided fracture network can be received. This fracture network can be represented as geometry and fracture properties. In the example of the embodiment of FIG. 2, the fracture set is divided into 2 groups: those to be solved in the 2.5D mesh (eg, fractures allowing the 2.5D 206); and those that need to be solved only in the final 3D mesh (eg, other fractures 208). The fracture characteristics allowing the 2.5D 206 are thus called because they allow a 2.5D mesh. As illustrated, the fracture characteristics for 2.5D 206 may include fractures represented by vertical planes. Fracture characteristics allowing 2.5D 206 may also include fractures represented by substantially vertical planes with respect to the horizon. Other fractures 208 may include fractures represented by horizontal planes. The other fractures 208 may also include fractures represented by substantially horizontal planes (eg, almost horizontal planes). Fracture sets 206 and 208 are not necessarily exclusive. That is, some subsets of fractures can be found in both sets of fractures 206 and 208. In workflow 200, however, all input fractures are categorized as at least 1 one of the two sets of fractures 206, 208. In the workflow 200, the 2.5D mesh is first created by using a stadimetric mesher in 2.5D 210 in order to solve the fractures allowing the 2 5D 206. The workflow 200 may use the 2.5D 210 stadian mesh to produce an 2.5D extruded mesh, which is illustrated as the intermediate reservoir model 212 in FIG. 2. The 2.5D 210 stadian mesher can generate the intermediate reservoir model 212 so that it includes an extrusion mesh that models the 2.5D 206 fractures in a 3D space. The extrusion mesh models a certain volume of the reservoir (eg the reservoir corresponding to the technical data of reservoir 204), including fractures allowing 2.5D 206. Examples of 2.5D stadimetric meshes are described in detail with reference to FIGs. 24-28 next. Next, the workflow 200 applies the extended AGAR algorithm at block 214 to solve the other fractures 208. As illustrated, the intermediate reservoir model 212 and the result of the resolution of the other fractures 208 are associated to Filled reservoir model 216 may be included in output 218 of workflow 200. In one example, output 218 may be presented graphically in a user interface, such as the example of interface given in FIG. 9. In an alternative example, if only the fractures allowing 2.5D 206 are included in an entry 202, the block 214 can be skipped, and the intermediate reservoir model 212 is used as the full reservoir model 216. Examples of Refinement techniques [0046] In general, the illustrative refinement algorithms described herein refine the cell edges up to "n" times and divide no rim (and, thus, divide the cell) that would produce a shorter rim than the `targetSize. ' As defined herein, the largetSize 'represents the desired mesh size, or flange length, for resolving the width of the fracture. In some embodiments, the value of the targetSize is provided to the system based on a desired level of precision and the desired time to solution. In order to generate the Earth models, the processes can be determined if a cell is to be refined and in which direction these cells are to be refined. In some embodiments, there may be two types of rules for determining whether a cell is to be refined: 1) the gradation rules (the rules that provide slow transitions on the element side) and 2) the crossover rules ( rules that ensure that fractures are adequately represented). In one example to summarize the process, consider the U directions (a direction in a 3D space). It should be noted that U is a direction in a topological sense and not in a Euclidean sense. The "U direction" for each cell is independent of the U direction for a neighboring cell. In addition, in one cell, an 'Il' rim may point in a slightly different direction than another 'Il' rim. As will be described in more detail below, considering the U direction, the extended AGAR engine 110 will scan all the cells in one model and refine almost fractured cells in the U direction if all the U flanges are longer than C x targetSize and at least one of the following rules is satisfied: 1) a stepped U-edge crosses a fracture; 2) for any pair of opposite flanges in the U direction, exactly one of the two stepped flanges intersects a fracture; 3) a rim U has two or more nodes 'suspended'; 4) for any pair of opposite flanges in the U direction, exactly one of the two stepped flanges have two or more hanging knots; or 5) the non-staged cell crosses a fracture but no staggered rim of the cell crosses a fracture. Referring to the C x targetSize, for a scalar variable specific to an implementation or provided by a user, C, the system 100 will not divide any rim (hence, cells) that is shorter than C x targetSize. In this illustrative method, C is chosen as - / 2, but may be other variables such as, for example, 4/3 or any finite value. This process is repeated in each direction in the 3D space. As used herein, the term "suspended node" is generally used to describe a node created during the refinement of a neighboring cell, which is not necessary to maintain the underlying geometry of the cell of interest. For example, the underlying geometry may be a hexahedron (as in the examples of FIGs 3, 8 and 12-13), a prism (as demonstrated by the examples of FIGs 4-8 and 10-11). , or other types of extruded elements. Consider an H cell, when a neighboring cell is refined, an extra node is added to the hexahedron H-cell. This node is "suspended" with respect to cell H. In still other examples, the requirement of edge length may be omitted for rules 3 and 4 if strict adherence to the hanging node directive is adopted. [0049] FIG. 3 illustrates a set 300 of possible refinement patterns 306, 312 and 316 for a single hexahedron cell 304 within a 3D space U-V-W 302. In particular, FIG. 3 illustrates how, for a hexahedron, some embodiments may utilize a local UVW logical space 302, and each cell is analyzed and refined into a certain subset of these three dimensions, U, V, and W as demonstrated by the steps 308, 310 and 314 of FIG. Once the refinement directions have been determined for a cell, an extended AGAR engine 110 performs the refinement (i.e., divide the edges) for that cell. FIG. 3 illustrates eight possible refinement configurations for a single cell within a UVW 302 3D space. The extended AGAR engine 110, via the algorithm, can carry out an iterative refinement by first determining whether the cells close to the fracture must be refined in the U, V or W dimensions / directions (steps, 308, 310, and 314 shown in FIG 3), and then, if so, the division of the flanges in the directions illustrated for creating refinements of cells 306, 312 and / or 316. The eight possible cell refinement patterns 304, 306, 312 and 314 range from refinement without refinement direction (hexahedron 304) to refinement in all 3 refinement directions. (the most right refinement 316). Since these cells will be those that will be positioned in the region close to the fracture, the mesh resolution of the images thus obtained will be greatly improved. [0051] FIG. 4 illustrates a representative triangular prism cell 404 and a 2D U-W logic coordinate system 402 and a number scheme for the edges of the prism 404. Some embodiments include two alternatives for extending the aforementioned refinement concept to the prisms. One is the "single U direction" option shown in FIG. 4. As illustrated in FIG. 4, the single U direction encompasses all the triangular face flanges of the prism 404. According to this definition of directions, the refinement, when it occurs on the triangular faces of the prism 404, is always isotropic with respect to the face. In the example of FIG. 4, using the single U direction option 400, each of the flanges 1 to 6 is considered to be in the U direction. The refinement in U will refine all the edges and the refinement can be done in any of the subassemblies. of {U, W}. Another option is called here the "multiple U direction" option. This is illustrated in FIG. 5. FIG. 5 illustrates a triangular prism cell 504 and a U-W 502 logic coordinate system 502 and a number scheme for the rims of the prism 504. Using the "multiple U direction" option 500 of FIG. 5, the refinement would be performed in the U-direction pairs. Thus, the refinement could be performed in a subset of directions 1U1U2, U1U3, U2U3, W} 502 shown in FIG. 5. As a general rule, in the rules applied by some embodiments, not all 16 permutations in the 4 directions (eg, W, U1, U2, and U3) are allowed during the step of refinement. In the example of FIG. 5, the multiple directions U are considered in pairs. For example, refinement in U1U2 will refine flanges 1, 2, 4, 5 of prism 504. [0052] FIG. Figure 6 illustrates an example of the U 600 single-directional paradigm. Following the general notation and the extended AGAR algorithm presented above, and the use of the canonical numbering of the triangular prism in FIG. 6, label the edges of the cell, Ci, as teid) for J = 1 ... 9. Consider now an abstract Boolean function on the edges, Test ((), eg the cell can be refined in: - [6. ,, 1, ei, 2, ei.3, and. Er, U if any flange from the satisfied set Test (teid)) (see step 606 of FIG.6), and - .W if any flange from the set e 1.7 'satisfies Test (teJ) (see step 610 of FIG 6) The aforementioned condition is called the "main condition" for gradation, and we will now define "opposing conditions." For quadrilaterals in prism 604, the rule is the same for hexahedra (eg, as shown above with reference to FIG.3) If, say, two opposing flanges on the face, exactly one of these flanges satisfies Test (tei /}), then the direction which divides these 2 edges will also be refined.For example, the edges tei, 4, 11 are opposite edges on a quadrilateral face.If the edge tei, 41 sa Satisfy Test (tei, i}) but not the teial edge, then this condition will be satisfied, and the direction W would be added to the refinement set. In the same way, if the flange satisfies satisfied Test (te1, /}) but not the flange [e.] The direction W would be added to the refinement game. As for the triangular face, if exactly a satisfied Test (tei, /}), then U is added to the refinement game. For example, if the flange tei, 4) is satisfied with Test (id) but not the flanges i, 5 'i, 6), the direction W would be added to the refinement set. With respect to the crossing rules of the extended AGAR algorithm, we define Test (teid) as true if and only if you have been refined at least twice. With respect to the AGAR algorithm gradation rules, we define Test (teid) as true if the step flange intersects a fracture of interest. FIG. 6 illustrates the possible combinations of refinements 608 and 612 using the single U-direction paradigm 600. [0055] FIG. 7 illustrates how one can, in an alternative of the single U direction presented above with reference to FIG. 6, select a multiple-directional paradigm U 700. In particular, FIG. 7 illustrates refinements 708, 712, 716, 720 and 722 in the 4 logical directions 702 for a U700 multi-directional paradigm of a 704 prism. Using the U 700 multi-directional paradigm, the cell will be refined in - U1U2 and U1U3 if any rim from the set tei, 1 satisfies Test (teEd}) (to steps 706 and 710 of FIG 7), - te, r satisfies U1U2 and U2U3 if any rim of the set E, 2 Test (ti, ./)) (see steps 706 and 714 of FIG 7), U1U3 and U2U3 if any edge of the set te ei, 3 satisfies 1 ei, satisfies Test ( tei, /}) see Test (td-) (see steps 710 and 714 of FIG 7), and -te W if any flange of the assembly of step 718 of FIG. 7). [0056] In addition to these main conditions, we have opposite conditions. For quadrilaterals, the two U paradigms are very similar. If, taking two opposite flanges 20 on the face, exactly one of these flanges satisfies Test (teid), then the directions that divide the two flanges will also be refined. If the flange te [, 7) satisfies Test (te 1) but not the flange tet, 81, the direction W would be added to the refinement game. In one example, for the rule of the triangular face, if exactly one rim satisfies Test (t JI), then U1U2 and U1U3 can be added to the refinement game. In this paradigm, we do not add any opposite conditions for the triangular faces. However, alternative rules can be determined and tested in the paradigm framework beyond the paradigms 600 and 700 illustrated in FIGs. 6 and 7, respectively. Figures 8 and 9 are examples of user interfaces. In particular, FIG. 8 illustrates a user interface 800 displaying an input template containing both hexahedral 802 and triangular 804 prisms 804. FIG. 9 illustrates a user interface displaying a refined fracture set 902 in the model of FIG. 8 using the unique U-direction paradigm. As illustrated in FIG. 9, the anisotropic advantages are greatly reduced for prisms, using the single U-direction paradigm, but only in the horizontal directions. Most of the anisotropic behavior is maintained in the vertical direction. This is because the vertical direction is not necessarily refined. FIG. 10 illustrates a representative non-triangular prism 1004 (e.g., a general prismatic element) showing the logical coordinate system in 2D 1002. FIG. 11 illustrates a possible refinement of a general prismatic element 1104 by subdividing the element into hexahedrons. As described above, an extended AGAR algorithm may be used to implement methods that extend the hexahedron-based AGAR scheme to also include triangular prisms, such as those illustrated in FIGS. 4-7. For long quadrilateral and triangular meshes, this combination is sufficient. Assuming a coherent extrusion topology, the paradigms presented above with reference to FIGs. 4-7 will result in the formation of a valid mesh even if some of the elements are degenerate. Figures 10 and 11 illustrate how some embodiments can directly extend the single U-direction paradigm to non-triangular prisms using a half-point subdivision algorithm. For non-triangular prisms, such as the non-triangular prism 1004 of FIG. 10, the anisotropy can be maintained in the U-W 1002 space. As can be seen in the refinement 1104 of FIG. 11, using this refinement algorithm, the non-hexahedral prismatic elements 1004 are subdivided into hexahedrons. Since hexahedra have 3 logical dimensions, compared to the two illustrated in FIGS. 10 and 11 for non-hexahedral prismatic elements, the opportunity for anisotropic refinement is increased after the refinement of the original element. To further generalize, we can apply a mid-point subdivision algorithm for all convex polyhedron cells, even non-extruded cells. It should be noted that this will modify the refinement algorithm for the triangular prisms that are presented above with reference to FIGs. 4-7. However, for non-extruded cells, we use a single-dimensional U-space, ruling out the anisotropic refinements for these general cells. In cases where the majority of cells are extruded (hexahedron cells or other prism cells), anisotropy for these cells will still provide a significant advantage over completely isotropic refinement. In addition, the anisotropic refinement on a regular cell surrounding an irregular cell does not violate the compliance requirement on the mesh. The polygonal face of the neighboring irregular cell is simply divided. [0060] Now that a brief description of an exemplary refinement process has been given, a more detailed discussion of examples of techniques for creating reservoir mesh will now be presented. In order to generate the Earth's reservoir models, an example of a process first determines which cells need to be refined and in which direction these cells should be refined. To accomplish this in this example, an extended AGAR engine 110 first begins with a structured, uniform and axially aligned hexahedron grid. However, in other examples, it is not necessary to start with a uniform, axially aligned grid. Nevertheless, when data are input to the refinement algorithm, an extended AGAR engine 110 requires: 1) M = Cd, the initial set of mesh cells to be refined, 2) F = {FI}, the facet set triangular fractures, 3) co, the desired "width" of the cells to adequately resolve the fractures, and 4) a, the size of the desired mesh "along the" fractures. Because the initial grid, M, is uniform, there is a characteristic grid size, TT. Thus, the extended AGAR engine 110 calculates two numbers: niso = logi Eq (1), and T 2 Eq (2), where ntot = g1- -T 2 rani = ntot-niso. which will define the number of refinement levels achieved by the extended AGAR engine 110. The extended AGAR engine 110 will perform Iso n.sode isotropic refinement to obtain IX, close to the fractures. The AGAR engine extended 110 mesh size, will perform nn cal year / year isotropic refinement to obtain a mesh of "width", ntot 2, close to fractures. It should be noted that we have a = -1a and [0062] Therefore, in an illustrative method, the general algorithm applied to an extended AGAR engine 110 may be expressed as: refine (n, doIsotropic, targetSize) co = 25 for iteration = 1 ..., -n-min (n, iteration) sizeLimit = targetSize * z for C in M // loop above each cell in the mesh for this iteration directions = determineDirections (C, F, iteration, sizeLimit, n, doIsotropic) refineInGivenDirections (directions, C) if no change has been made during this iteration output routine [0063] In this routine 1, the "Refine" algorithm is called with an integer to indicate the number of refinement levels to execute and a Boolean to indicate whether it is isotropic refinement or not. It should be noted that in this example the routine is called twice. The extended AGAR engine 110 calls refining (niso, true, â) followed by nam, easiness, Essentially, to execute n refinement levels, the extended AGAR engine 110 loops over the cells until no another change is necessary. Since the extended AGAR engine 110, via the algorithm, prohibits the division of any edge that is shorter than C x targetSize during an iteration, in the end the algorithm will end. One of the most complex parts of the algorithm is encapsulated in the method "determinateDirections", given in routine 2 below which returns a subset of local direction in 3D {U, V, W} depending on the directions which must be refined. It should be noted that "sizeLimit", used to define the smallest edge during an iteration, is large for early iterations. By the iteration, and for all subsequent iterations, sizeLimit is equal to targetSize. Routine 2 can be expressed as: determineDirections (C, F, iteration, sizeLimit, n, doIsotropic) directions = 0 // Add any refinement direction needed because of fracture / cell crossover // This is n is allowed that during the n first iterations if the iteration <= n // in the crossing stage directions = getDirectionsFromIntersections (C, F) // add any direction of refinement necessary because of the gradation // this will only be necessary only for the iteration> 1 If the iteration> 1 directions = directions U getDirectionsFromGradation (C) C) // if we want isotropic and we find any direction, perform all directions if & treotrop & (directions 0) directions = ALL // remove all forbidden directions, regardless of what we found directions = directions - getDisallowedDirections (C, F, sizeLimit) return directions [0064] In this illustrative Routine 2, `determineDirections' mainly uses getDirectionsFromGradation and getDirectionsFRomIntersections subroutines. As demonstrated by `determineDirections' from Routine 2, it is only during the first n iterations that the extended AGAR engine 110 refines the cells close to the fracture as a function of the fracture interaction, which will be described in more detail below. During a first iteration after the extended AGAR engine 110 refines the cells based on the 'hanging node rule'. That is, if a rim of the cell has been split more than once (due to a neighbor being refined), the cell can be refined. If the extended AGAR engine 110 achieves "isotropic" refinement, then, whenever it needs to refine a cell in any direction, it refines the cell in all directions. All directions identified for refinement must pass a final test before being accepted. By refining in a given direction, 4 edges can be divided. If any of these 4 edges is shorter than C x targetSize, then the direction is rejected by the extended AGAR engine 110. Since the initial mesh may comprise a hexahedron, each cell may be represented by a canonical numbering system described below with reference to FIG. 12. FIG. 12 illustrates a representative cell 1204 showing the local 3D coordinate system 1202 and the numbering scheme 1200. In FIG. 12, the U-V-W directions are defined by the mapping of the nodes on the numbering system (note that it is drawn as a left-handed system). In the numbering scheme 1200, the numbers representing the edges are encircled, while the numbers representing the nodes are not. The mapping of U-V-W 1202 on the x-y-z system may differ from neighboring cells. In this example, take the flanges {0,2,4,6} as the flanges U, the flanges {1,3,5,7} as the flanges V, and the flanges {8,9,10,11} as the flanges W. [0067] Using the canonical scanning scheme in FIG. 12, the edges of the cell, Ci, are labeled as {ev} for j = 0 ... 11. The return set from getDirectionsFromGradation (C1) includes: 1) U if any edge of the set {e , 0, e1,2, e1,4, e1,6} has been refined at least 2 times, 2) V if any edge of the set {éi, Co, e, 5, has been refined at least 2 times, and 3) W if any edge of the set {e, 8, e1,9, e1,10, e1,11} has been refined at least twice. This condition is called the "main condition" for gradation. In addition, if we take two opposite flanges on one face, exactly one of these flanges has been refined 2 times, then the direction that divides these 2 flanges will also be refined by the extended AGAR engine 110. For example, the flanges {e, 8, Co} are opposite edges on one side. If the flange {e, 8} has been refined twice but the flange {e, 9} has been refined once, then this condition would be satisfied, and the direction U would be added to the refinement set. This test is called the "opposite condition" for gradation. Note that for all tests in getDirectionsFromGradation, extended AGAR engine 110 considers only the number of times the edge has been split (ie, refined) before the current iteration. In this way, the extended AGAR engine 110 decreases its dependence on the order in which the cells can be visualized in the algorithm.
[0012] Cell-Facet Crossing Test [0068] One aspect of the refinement algorithm used in some embodiments is the definition of a test for determining whether a given cell, Ci, crosses a given fracture facet, Fk. See FIGs. 4, 5 and 12, for example, the numbering schemes used by the extended AGAR motor 110 for the nodes and the edges of the cell. In order to continue describing this illustrative method, the extended AGAR engine 110 may define the scaling transform, Sfl (), which scales the distance between the center of an object (the node) and each node by the scalar value, fi . In this illustrative algorithm, fi is defined as 2, but the determination of the most appropriate value of fi is a design choice and may be modified, as will be understood by one skilled in the art who is aware of this disclosure. Note that the extended AGAR engine 110 can apply this transform to the cell, producing a staggered cell, or at the edges, giving a set of stepped edges. These two applications give different results, as shown in FIGs. 13 and 14, which illustrates the scaled function, Sfl (), applied to a cell, and the scaling function, Sfl (), applied to the edges of the cell, respectively. Still with reference to FIGs. 13 and 14, the Boolen function Tfl (C'F) is true if and only if there is a fracture Fk EF so that the crossing of the staggered cell (Se (C3 n Fk 0) and the fracture n ' is not empty For the edges, the function is slightly altered so that it will be false each time the edge is in the plane of the facet, so for the edges, Tfl (e, J, F) if and only if there is a fracture Fk EF so that both the crossing of the step flange and the fracture are not empty the normal of Fk is not perpendicular (or transversal in other examples) to the flange (R.S'e (e,) n Fk # 0) & (normal (F k) e1> E) Essentially, the function is true when the scaled entity intersects the facets, except in the case where the flange is located in the plane of the facet, this leads to the division of the flange along a plane that is transverse (eg perpendicular) to an axis of the fracture within the fracture network [0070] Now, an illustrative algorithm used in Routine 2 that defines getDirectionsFromIntersections (C'F) is specified. Note that, as in the case of the grading tests above, the crossover tests include similar "main conditions" and "opposite conditions". This routine 3 can be expressed as, e.g. : Set <Directions> getDirectionsFromIntersections (C'F) directionSet = // do nothing if no facet crosses the scaled cell If Tfl (C'F) == false return directionSet // the "main" condition // If a step flange intersects the fracture set, we carry out a refinement perpendicular to it T j3 (e F) IT F) IT 13 (e 0, F) IT directionSet = directionSet UU (e 0, F) 1T (e, F) IT i3 (ei, s, F) IT i, 7 directionSet = directionSet UV for a certain c small and specified).  if (T (e48, F) I Tp, (eo, F) I Tp, (e00, F) IT / 3 (e directionSet = directionSet UW // the "opposite" condition // If the "opposite" edges have different, perform a refinement of the division direction 'T (e8, F) # T f3 (F) T e (e 00, F) # T f3 (e F) if T 13 (e F) T 15, (e F) IT 13 (ei, 5, F) # T 13 (e, 7, F) directionSet = directionSet UF) #T e (e F) IT / 3 (e F) #T e (e 00, F) si T 13 (e F) # T # (e 42, F) IT i3 (e F) # T 13 (e 46, F) directionSet = directionSet UV, a (e 43, F) #T e (e 47, F ) IT (e F) # T) 5, (e F) if T (e 40, F) # T / 3 (in IT / 3 (ei, 2, F) # T (e 46, n directionSet = directionSet UW // If the game is empty but a facet crosses the non-scaled cell, refine all If (directionSet, 0 & T0 (C, F)) directionSet = UUVUW [0071] Note that the algorithm "Refine" can be invoked or invoked by an extended AGAR engine 110 with an integer to indicate the number of refinement levels to be executed and a Bouléen to indicate whether it is an isotropic refinement or not.  Figures 15 and 16 illustrate the 1500 images of the extended AGAR meshes generated with illustrative features of the present disclosure.  In some process examples, the "opposite conditions" for determining refinement directions can be used to restrict anisotropic refinement in a localized region (e.g. , close to the fracture).  Eg. take the FIG.  15, which demonstrates an AGAR refinement close to the end of a plane fracture 1500, and FIG.  16, which shows the same model of FIG.  15 using an AGAR extension, the "opposite conditions" being disabled.  In the examples of FIGs.  15 and 16, the fracture 1500 is illustrated to be surrounded by various cells 1502 in the region close to the fracture.  In the absence of the "opposite conditions" for selecting the direction of refinement as shown in FIG.  16, the end of fracture 1500 would not have been resolved and the anisotropic refinement would have continued through the top of image 1500.  Figures 17 and 18 are additional images of the generated AGAR meshes using illustrative aspects of the present disclosure.  FIG.  Figure 17 illustrates one end of fracture 1500 with = 2, thus giving a mesh containing 1500 cells.  FIG.  Figure 18 illustrates one end of fracture 1500 with = 4, thereby providing a mesh with 2260 cells.  Figures 19 and 20 show a reservoir model filled with both hydraulic and natural fracture meshes using the illustrative methods described herein.  In FIG.  19, a slice through an example of mesh generated with the refinement algorithm is illustrated, including hydraulic fracture 1900 and natural fractures 1902.  In the example of FIG.  19, the mesh size characteristic of the original mesh was 250 feet and the target sizes were w = 3.125 feet, a = 250 feet.  C. -to-d. for this mesh, none of the levels use isotropic refinement, which gives flani = 6.  As illustrated in FIG.  19, the resolution of the mesh allowed by the cells close to the fracture is higher in the zones surrounding the 1900 and 1902 fractures in comparison with the zones that do not surround the 1900 or 1902 fractures.  In FIG.  20, the model of FIG.  19 is illustrated with some overlapping fractures 2000, 2002.  Examples of Methods [0075] In light of the foregoing detailed description, an illustrative method will now be described with reference to the flowcharts of FIGs.  21 and 22.  The methods 2100 and 2200 illustrated in FIGS.  21 and 22, respectively, are described with reference to the exemplary features of FIG.  1.  However, the methods 2100 and 2200 are not limited to these exemplary features.  As illustrated in FIG.  21, the method 2100 begins at block 2102.  In one example, block 2102 includes an initialization system 100 to generate a terrestrial tank model.  During initialization, the extended AGAR engine 110 downloads data from the module of the fracture system 112 and the Earth modeling module 114.  Such data may include, e.g. , a geocellular grid and fracture data.  At block 2104, extended AGAR engine 110 analyzes all cells in a 3D Earth model to determine whether cells in a mesh must be anisotropically refined (e.g. , refined in any direction in 3D space).  If an extended AGAR engine 110 determines that the cell being analyzed does not require refinement, control is transferred to block 2106.  At 2106, the extended AGAR engine 110 then determines whether there are other cells to be analyzed.  If it is determined that there are no other cells to be analyzed, control is transferred to block 2116 and process 2100 ends.  Otherwise, if it is determined that there is another cell remaining to be analyzed, the command is retransmitted to block 2104 where this cell will now be analyzed to determine if refinement might be necessary.  If, at block 2104, the extended AGAR engine 110 determines that this cell requires refinement, the algorithm goes back to the block 2108 in which this cell is refined.  As previously described, the gradation and crossover rules can be used to determine whether a cell needs to be refined.  Once an extended AGAR engine 110 refines the cell in the directions determined to be necessary, the algorithm proceeds to block 2109 where a determination is made to see if there are other cells to be analyzed.  If it is determined that there are other cells to be analyzed, the command is retransmitted to block 2106.  Otherwise, if it is determined that there are no other cells to be analyzed, the command is retransmitted to block 2110.  At block 2110, the extended AGAR engine 110 is used to resolve the fracture within the Earth model using the refined cells.  The model thus obtained will have a higher level of mesh resolution in these areas surrounding the fractures (e.g. , areas close to the fracture and fractures) and a lower mesh resolution in these areas close to the fracture.  FIG.  22 is a flowchart of an illustrative refinement process 2200 used in block 2104 shown above with reference to FIG.  21.  To begin this analysis, it should be noted that the location of the fractures within the fracture system is known from the data from the module of the fracture system 112.  Using these data, as well as with the previously described gradation and crossover rules, the extended AGAR engine 110 can locate these cells in regions close to the fracture of the model that requires refinement.  For cells that require refinement, all rims (and thus cells) will be divided in two in each refining direction.  The newly created cells will then be analyzed again in the next iteration, ultimately increasing the mesh resolution.  The high resolution mesh thus obtained provides more accurate numerical results in the regions close to the fracture and the regions of the fracture.  The C x targetSize (where C = - / 2, eg. ) is determined as previously described above in Section A.  Nevertheless, at block 2104 (i), the extended AGAR engine 110 determines whether all the edges in the U-V-W directions for the cell (each direction is analyzed separately) are longer than C x targetSize.  If these flanges are longer than C x targetSize, no refinement is required, and the command is passed to block 2106.  If, however, all the edges are longer than C x targetSize, then the 5 rules can be applied.  At this point, at blocks 2104 (ii) - (vi), the extended AGAR engine 110 determines if one or more of the following are satisfied: ii) a stepped directional edge crosses a fracture; iii) for any pair of opposing flanges in the 3D direction, exactly one of the two stepped flanges intersects a fracture; iv) a directional ledge has two or more "hanging" nodes; v) for any pair of opposing flanges in the 3D direction, exactly one of the two flanges has two or more hanging knots; and vi) the non-staged cell crosses a fracture but no step rim of the cell crosses a fracture.  At block 2104 (h), with reference to FIG.  13, the extended AGAR engine 110 determines whether the scaled version of cell 1300 (c. -to-d. , 1300S) have a rim in the analyzed direction that crosses a fracture.  At block 2104 (iii), extended AGAR engine 110 (for example) may analyze a pair of flanges in a given direction (e.g. , the flanges 5 and 1 in FIG.  12) to determine if only one of the stepped flanges intersects the fracture.  At block 2104 (iv), the AGAR engine will determine whether a rim in the analyzed direction has 2 or more suspended nodes.  In FIG.  22, element 2209 is given as an illustration of this principle.  Element 2209 is described below.  As illustrated in FIG.  22, the element 2209 comprises two adjacent cell faces 2210 and 2220 (shown in 2D for simplicity) having two nodes 2212 and 2214 positioned between them.  In the example of element 2209, the first refinement would have divided cell 2210 into two using node 2212.  The second refinement would have split one of the new cells into 2 using the 2214 node.  Therefore, the nodes 2212 and 2214 are "suspended" with respect to the cell 2220 because, even though the cell 2210 has been refined (in a previous iteration), the cell 2220 has not been refined.  Since the cell 2220 has at least two hanging nodes, the extended AGAR engine 110 will also refine it in later iterations.  At block 2104 (v) of the 2200 process, the extended AGAR engine 110 analyzes pairs of opposite flanges in the analyzed direction to determine if only one of the flanges has 2 or more hanging nodes.  At block 2104 (vi), the extended AGAR engine 110 determines whether a non-staggered cell crosses a fracture but that no staggered rim of the cell crosses a fracture.  If the determination is "yes" to any one of the rules described above with reference to block 2104 (i-v), the method passes the command to block 2108 where this cell will be refined.  Otherwise, if the determination is "no" to any of the 5 rules, the command is retransmitted to block 2106.  This process will continue iteratively until each cell in the model has been analyzed.  Once the analysis is completed, the extended AGAR engine 110 will generate a model of the Earth using the refined cells.  The generated Earth model can be used in a variety of applications, such as in reservoir simulation.  Then, the reservoir simulation can be used to design and complete a wellbore completion and to consistently implement downhole operations.  Examples of Stadimetric Meshing Techniques [0084] FIG.  23 illustrates a 2300 image of 3D fractures that can be modeled with the disclosed features.  As can be seen in Figure 2300, the layers of the Earth's formation include fractures within the formation.  These fractures can be described as open cracks or voids within the formation, and they can be either natural or artificially generated from a wellbore.  Understanding and modeling the characteristics of these fractures is important as fractures allow and influence the flow of fluids from the reservoir to the wellbore.  Images such as image 2300 can be obtained or generated using image catalogs.  Image catalogs can use a rotary transducer to measure acoustic impedance across the entire borehole wall to identify the presence and directions of rock fractures, as well as to understand the direction of the tilt of the rock. stratigraphy.  [0085] FIG.  24 illustrates a broad perspective of an example of a 3D fracture set 2410 that intersects a 2400 2D plane at line segments 2420.  As can be seen in FIG.  24, a closed loop stadium game can be generated around each of the line segments 2420 and a constrained mesh fills the remaining space of the 2D 2400 plane.  The closed loop stage set can be generated using the method 2500 described below with reference to FIG.  25.  FIG.  25 is a flowchart illustrating a method 2500 for 3D fracture modeling using a 2.5D stadimetric mesher.  The method 2500 begins at block 2502 by receiving a set of fractures with a geometry that has been discretized in a 2D plane by a collection of line segments.  In addition, the method 2500 can begin by performing the discretization of a set of fractures in a 2D plane by a collection of line segments.  The line segment collection can represent the intersection between the 2D plane and the 3D geological fractures.  As shown in block 2503, each fracture can be represented by a collection of straight line segments to approximate a curvature of the fracture.  According to an example, at block 2504, for each segment of the fracture line in the 2D plane, the method 2500 realizes block 2506 to generate a set of stages with a specified radius from the segment of the respective fracture line.  Then, at block 2508, method 2500 then generates closed loops around all line segments of a fracture.  In some embodiments, block 2508 includes a method for generating closed loops around line segments of the fracture.  This may include computing a cross on all sides of the stages for each specified radius for each line segment of the fracture, as demonstrated by block 2509 in FIG.  25, and the rejection of the contained segments for each line segment in each fracture line segment that is completely contained by the stage of other line segments in the segment of the fracture line, as shown in block 2511.  After completion of block 2508 (and optional blocks 2509 and 2511 in some embodiments), method 2500 proceeds to block 2510.  At block 2510, shape elements may be generated within the closed loops of the straight line segment.  Eg. in one embodiment, the method 2500 generates parametric segments along a length and radius of each straight line segment, as shown in block 2513.  The method 2500 then forms quadrilaterals where possible within the structured region at block 2515, and forms polygons within the remaining regions of closed loops at block 2517. .  After generating the shape elements, the method 2500 can generate a constrained mesh around the closed loops of the set of fracture line segments, filling the remainder of the 2D plane at block 2512).  In one embodiment, a Delaunay triangulation algorithm can be used to generate the constrained mesh around the closed loops of the set of segments of the fracture line.  Thus, the 2D plane now consists entirely of cell elements of the segments of the fracture line and the constrained mesh.  At this point, the method 2500 can extrude each of the cells in the 2D plane into a third dimension for creating one or more 3D cell layers.  Cells within a closed loop of a segment of the fracture line may represent a 3D fracture, while cells within the mesh may represent rocky layers encompassing these fractures.  Thus, the method 2500 can assign reservoir properties such as, without limitation, porosity and permeability, to each of the cells in 3D in order to model the fluid flow of the reservoir, as shown by block 2516.  Finally, the method 2500 can capture the 3D cellular model in a simulation program, such as, without limitation the Nexus® reservoir simulation software, for performing a numerical simulation and for evaluating the flow of fluid, such as demonstrates block 2518).  FIG.  Fig. 26 gives a closer view of method 2500 illustrating an example for generating a computer mesh around a single fracture line segment in accordance with the disclosed embodiments.  Starting with diagram 2602, a set of stages is generated around a 2600 line segment.  As shown in Diagram 2602, each stage in the game of stages is composed of two linear sides connected by two arcs in order to completely surround the straight line segment.  The distance from each side to the straight line segment is a constant radius.  In some embodiments, the radius distance may be a user changeable variable value.  In the diagram 2604, parametric segments along a length and radius of each straight line segment are generated in accordance with block 2513 of method 2500.  Quadrilateral elements may then be formed where possible within the structured region as referenced by block 2513 of method 2500.  Diagram 2608 illustrates the constrained mesh generated around the closed loops of the 2600 segment line.  [0094] FIG.  Figure 27 provides another close-up view of process 2500 illustrating an example of the generation of computer meshes around cross-fracture line segments.  Eg. , Diagram 2702 illustrates a stadium game generated around 3 fracture line segments.  The result of the chart 2702 requires that the method calculate a crossing of all sides of the stages for each specified radius for each of the cross-fracture line segment as referenced in block 2509 and reject segments contained for each line segment. fractures which are completely contained by stages of other fracture line segments as referenced in block 2511.  Diagram 2704 illustrates the results of the generation of the shape elements within the closed loops of the segments of the fracture line as referenced in block 2510 of FIG.  25.  As can be seen, the parametric segments along a length and radius of each fracture line segment are generated in accordance with block 2513 of FIG.  25.  Quadrilateral elements may then be formed where possible within the structured region as referenced by block 2515 shown in diagram 2706.  In addition, polygons may be formed within the remaining regions of the closed loops of the segments of the fracture line as indicated at block 2517 of FIG.  25.  Diagram 2708 illustrates a constrained mesh generated around closed loops of segments of the fracture line as referenced at block 2512 of method 2500.  [0096] FIG.  Figure 28 illustrates an example of computer meshes generated around a complex network of fracture line segments.  In particular, the diagram 2802 indicates a set of fractures with a geometry that has been discretized in a 2D plane by a collection of line segments.  Diagram 2804 illustrates the results of a set of stages that is generated around each of the segments of the fracture line.  Diagram 2806 illustrates an exploded view of the segments of the fracture line resulting from the execution of the remaining process 2500 described above with reference to FIG.  25.  As illustrated in FIG.  28, the disclosed algorithm can quickly generate unstructured grids using elements structured around complex geometries.  The cells in the 2D plane shown in FIG.  28 can then be extruded in a third dimension to form 3D cell layers.  3D cells can be assigned reservoir properties to allow numerical simulation.  As another example, FIG.  29 presents a 2900 illustration of complex geometry involving a plurality of cross-fracture line segments in which some embodiments can quickly generate a 2D grid cell that can be extruded into 3D elements to perform numerical simulations.  Examples of Computer System Implementation [0099] Although exemplary embodiments have been described in terms of devices, systems, services, and methods, it is contemplated that certain features described herein may be implemented in software applications. on microprocessors, such as the microprocessor chip embedded in computing devices, such as the computer system 3000 shown in FIG.  30.  In various embodiments, one or more of the functions of the various components may be implemented in software that controls a computing device, such as the computer system 3000, which is described below with reference to FIG.  30.  [0100] In order to implement the various features and functions described above, some or all of the elements of the systems (e.g. , the system 100 of FIG.  1), workflows and processes (eg. methods 200, 2100 and 2500 of FIG.  2, 21, and 25) can be implemented using elements of the computer system of FIG.  30.  More particularly, FIG.  Figure 30 illustrates an example of a computer system 3000 for implementing the techniques in accordance with the present disclosure.  The aspects of the present disclosure illustrated in FIGs.  1-28, or any one or more parts and one or more functions thereof, may be implemented using hardware, software modules, firmware, tangible computer readable medium having logic or instructions stored thereon. or a combination thereof and can be implemented in one or more computer systems or other processing systems.  [0102] FIG.  30 illustrates an example of a computer system 3000 in which embodiments of the present disclosure, or parts thereof, can be implemented as instructions or computer readable code.  Eg. , certain functionalities realized by the system 100 illustrated in FIG.  1, can be implemented in computer system 3000 using hardware, software, firmware, computer-readable non-transitory medium having instructions stored thereon, or a combination thereof and can be implemented in one or more computer systems or other processing systems.  Hardware, software, or any combination thereof may embody certain modules and components used to implement the system 100 and workflow 200 illustrated in FIGs.  1 and 2 presented above.  Similarly, hardware, software, or any combination thereof may embody certain modules and components used to implement the steps in the flowcharts shown in FIGs.  21, 22 and 25 presented above.  If a programmable logic is used, such logic can run on a commercially available processing platform or on a specialized device.  One skilled in the art may appreciate that embodiments of the subject of the disclosure may be practiced with various configurations of computer systems, including multi-core multiprocessor systems, mini computers, mainframes, linked or grouped computers with distributed functions, as well as ubiquitous or miniature computers that can be integrated into virtually any device.  For ex. at least one processor and memory device may be used to implement the embodiments described above.  A processor may be a single processor, a plurality of processors, or combinations thereof.  Processors may include one or more processor "cores".  Some embodiments are described in terms of this exemplary computer system 3000.  After reading this description, it will be obvious to a specialist in the relevant field how to implement the disclosed examples using other computer systems and / or other computer architectures.  Although the operations may be described as a sequential method, some of the operations may, in fact, be performed in parallel, simultaneously and / or in a distributed environment, and with the program code stored locally or remotely for a access by single processor or multiple processor devices.  In addition, in some embodiments the order of operations may be rearranged without departing from the spirit of the disclosed object.  The processor 3004 may be a specialized processor or a multi-purpose processor.  As will be understood by those skilled in the relevant field, the processor 3004 may also be a single processor system in a multicomputer multiprocessor system, such a system operating alone or in a cluster of computing devices or in a cluster or farm of servers.  The processor 3004 is connected to a communication infrastructure 3006, e.g. , a bus, a message queue, a network or a multichannel message transfer scheme.  In one embodiment, the bus system 109 of the system 100 described above with reference to FIG.  1 may be embodied as a communication infrastructure, 3006 shown in FIG.  30.  According to one embodiment, the processor 102 of the system 100 described above with reference to FIG.  1 may be embodied as the processor 3004 shown in FIG.  30.  The computer system 3000 also includes a main memory 3008, for example, a RAM memory and may also include a secondary memory 3010.  The secondary memory 3010 may comprise, e.g. , a hard disk 3012, and a removable storage disk 3014.  The removable storage disk 3014 may include a floppy disk, a magnetic tape disk, an optical disk, a flash memory, and so on.  In a non-limiting embodiment, the storage 104 of the system 100 of FIG.  1 may be embodied as the main memory 3008 shown in FIG.  30.  The removable storage disk 3014 reads and / or writes to a removable storage unit 3018 in a well-known manner.  The removable storage unit 3018 may include a floppy disk, a magnetic tape disk, an optical disk, and the like. which is read by the removable storage disk 3014 which can also record on it.  As will be understood by those skilled in the relevant field, the removable storage unit 3018 includes a computer-readable non-transitory storage medium having stored therein computer software and / or data.  In alternative implementations, a secondary memory 3010 may include other similar means to allow the computer program or other instructions to be downloaded to the computer system 3000.  Such means may include, e.g. , a removable storage unit 3022 and an interface 3020.  Examples of such means may include a software module and an interface module (such as that found in video game machines), a removable memory chip (such as an EPROM or EEPROM) and the associated jack, and other removable storage units 3022 and interfaces 3020 that allow the transfer of software and data from the removable storage unit 3022 to the computer system 3000.  The computer system 3000 may comprise a communication interface 3024.  The communication interface 3024 allows the transfer of software and data between a computer system 3000 and external devices.  The communication interface 3024 may include a modem, a network interface (such as an Ethernet card), a communication port, a location and a PCMCIA card, or the like.  The software and data transferred through the communication interface 3024 may be in the form of signals, which may be electronic, electromagnetic, optical, or other signals capable of being received by the communication interface 3024.  These signals may be provided to the communication interface 3024 through a communication channel 3026.  The communication path 3026 carries signals and may be implemented with a wire or cable, optical fiber, telephone line, cellular telephone link or RF link or other communication channels.  As used herein, the terms "computer-readable medium" and "computer-readable non-transitory media" are generally used to describe media such as memories, such as main memory 3008 and a memory. secondary 3010, which may be semiconductor memories (e.g. , DRAM, etc. ).  A computer-readable medium and a computer-readable non-transitory medium may also describe a removable storage unit 3018, a removable storage unit 3022, and a hard disk installed in the hard disk 3012.  The signals carried through the communication path 3026 may also embody the logic described herein.  These computer program products may also be a means of providing software to the computer system 3000.  The computer programs (also called computer control logic) can be stored in the main memory 3008 and / or in the secondary memory 3010.  Computer programs can also be received through the 3024 communication interface.  Such computer programs, when executed, enable the computer system 3000 to implement the present examples as described herein.  In particular, the computer programs, when executed, enable the processor 3004 to implement the methods of the present disclosure, such as the steps in the workflow illustrated in FIG.  2 and the methods illustrated by the flow charts of FIGS.  21, 22 and 25, presented above.  Therefore, such computer programs represent the commands of the computer system 3000.  When features are implemented using software, it can be stored in a computer program and downloaded to the computer system 3000 using a removable storage disk 3014, an interface 3020 and a hard disk 3012, or a communication interface 3024.  In some embodiments, one or more of the display devices 108 of a system 100 and displays used to display the user interfaces illustrated in FIGS.  8 and 9 could be computer monitors 3030 shown in FIG.  30.  The computer monitor 3030 of the computer system 3000 may be implemented as a touch device (c. -to-d. , a touch screen).  The computer monitor 3030 can connect to the communication infrastructure through a display interface 3002 to display the received electronic content.  Eg. , the 3030 computer monitor can be used to display refined input models and fractures.  But also, eg , the computer monitor 3030 can be used to display the user interfaces 800 and 900 shown in FIGs.  8 and 9.  Certain embodiments may be directed also to a computer program comprising software stored on any medium that can be used by a computer.  Such software, when run on one or more data processing devices, causes the operation of one or more data processing devices as described herein.  Some embodiments use a readable or computer usable medium.  Examples of computer usable media include, but are not limited to, main storage devices (e.g. , any type of random access memory), secondary storage devices (e.g. , hard disks, floppy disks, CD ROMs, ZIP disks, tapes, magnetic storage devices and optical storage devices, MEMS, nanotechnological storage devices, etc. ), and communication media (e.g. , wired and non-wired communication networks, local area networks, wide area networks, and the intranet, etc. ).  Although the subject of the present disclosure has been described in detail with respect to the specific embodiments thereof, it will be appreciated that the specialists in the field, when they have acquired an understanding of the foregoing description, can easily bring alterations to, and variations of, and equivalents of such embodiments.  Therefore, it should be understood that the present disclosure has been presented for the purpose of example rather than limitation, and does not preclude the inclusion of such modifications, variations and / or additions to the subject matter of this disclosure. as it would be easily obvious to a tradesman.  Indeed, the methods and systems described herein can be embodied in a variety of other forms; furthermore, various omissions, substitutions and changes in the form of the methods and systems described herein may be made without departing from the spirit of the present disclosure.  The attached claims and their equivalents are intended to cover such forms or modifications as are within the scope and spirit of this disclosure.  [0116] Except in the case of specific mention to the contrary, it is appreciated that throughout this specification discussions using terms such as "in process", "calculate", "calculation", "determination" and "identification" and Similar terms describe actions or methods of a computing device, such as one or more computers or similar electronic computing device (s), that manipulate or transform data represented as electronic or magnetic quantities to the computer. internal memories, registers or other information storage devices, transformation devices or display devices of the computer platform.  The system (s) presented here are not limited to any architecture or hardware configuration.  A computing device may include any component arrangement that provides a result dependent on one or more inputs.  Suitable computing devices include versatile microprocessor-based computer systems that access stored software that programs or configures the computer system of a versatile computing apparatus to a specialized computing apparatus implementing one or more of the embodiments of the present invention. the subject of this disclosure.  Any type of programming, script or any other type of language or language combination may be used to implement the teachings contained herein in software that is to be used in the programming or configuration of a computing device.  Examples of the methods described herein can be realized in the operation of such computer devices. The order of the blocks presented in the above-mentioned examples can be varied, e.g. , the blocks can be rearranged, combined and / or divided into sub-blocks.  Some blocks or processes can be run in parallel.  The conditional language used here, such as among others "may", "could", "could", "may", "e.g.  And similar words, except where specifically stated otherwise, or otherwise understood in the context used, are generally intended to convey that some examples include, while other examples do not include, certain characteristics, elements and / or steps.  Thus, such conditional language is not generally intended to imply that features, elements and / or steps are in any way necessary for one or more examples or that one or more of the examples necessarily include logic. for decision-making, with or without an author entry or prompt, whether its features, elements, and / or steps are included or not, or that they should be performed in a given example.  The terms "comprising", "including", "having" and similar terms are synonymous and are used inclusively, in an open manner and do not exclude elements, features, actions, additional operations, etc.  But also, the term "or" is used in its inclusive sense (and not in its exclusive meaning) so that when used, eg. To link a list of elements, the word "or" means one, some or all of the elements in the list.  In addition, the use of "based on" is intended to be open and inclusive, in that a method, step, calculation or other action "based on" one or more of the described conditions or values may, in practice , be based on conditions or values additional to that described.  The various features and methods described above may be used independently of one another, or may be associated in a variety of ways.  All possible combinations and subcombinations are intended to be within the scope of this disclosure.  In addition, some processes or process blocks may be omitted in some implementations.  The methods and processes described herein are also not limited to any given sequence, and blocks or states related thereto may be made in other sequences that are appropriate.  Eg. the blocks and states described may be in an order other than that specifically disclosed, or multiple blocks or states may be combined into a single block or state.  Examples of blocks or states can be made in series, in parallel, or in some other way.  Blocks or states may be added or removed from the disclosed examples.  In the same way, the examples of systems and components described herein may be configured differently as described.  Eg. elements may be added to, or removed from, or rearranged in comparison to the disclosed examples.  The foregoing description of the embodiments, including the illustrated embodiments, has been presented for illustrative purposes only and it is not intended that the description be exhaustive or be limited to the precise forms disclosed.  Many modifications, adaptations and uses of it will be apparent to a specialist in the field.  The embodiments and methods described herein also relate to one or more of the following paragraphs.  As used above, any reference to a series of examples should be understood as a reference to each of these examples disjunctively (e.g. , "Examples 1-4" should be understood as "Examples 1, 2, 3 or 4").  [0124] Example 1 is a computer-implemented method that includes receiving, at a computing device, a tank specification.  The method identifies, according to the reservoir specification, a set of fractures comprising: Fractures (2.5D) allowing 2.5D; and other fractures.  The method also includes generating an intermediate reservoir model comprising an extrusion mesh that models fractures allowing 2.5D in a 3D space.  In response to the determination that the cells in the mesh must be refined in a direction within the 3D space, the process refines the cells anisotropically in the mesh corresponding to other fractures.  The method also includes solving, by the computing device, a fracture network within the intermediate reservoir model using the refined cells; and generating a model of Earth reservoir using the fracture network.  Example 2 is the method of Example 1, in which the 2.5D fractures have a geometry that has been discretized in a 2D plane by a collection of line segments, and in which the generation Intermediate reservoir model includes, for each line segment associated with each fracture in fractures allowing the 2.5D: the generation of a set of stages with a specified radius from the line segment; the generation of closed loops around all line segments associated with a 2.5D fracture; and generating the shape elements within the closed loops of the segment line.  In Example 2, the method also includes the generation of grid mesh constrained around closed loops of fractures allowing the 2.5D to fill a space remaining in the 2D plane.  Example 3 is the method of Examples 1 or 2 in which the method also comprises the use of the earth tank model in a reservoir simulation, and then the display of the reservoir simulation on a screen of the computer device.  Example 4 is the method of Examples 1-3, wherein: 2.5D fractures comprise fractures represented by one or more substantially vertical planes; and the other fractures comprise fractures represented by one or more substantially horizontal planes.  Example 5 is the method of Examples 1-4, wherein the anisotropic refinement of the cells in the mesh comprises identifying a direction within the 3D space in which the cells are to be refined. , and the division of a rim of the cells, the rim being in the direction inside the 3D space.  Example 6 is the method of Example 5, wherein the flange is divided along a plane that is perpendicular to an axis of a fracture within a fracture network.  Example 7 is the method of Examples 1-6, wherein the anisotropic refinement of cells comprises: determining a location of a fracture within an intermediate reservoir model; the identification of cells in the region close to the fracture; determining a target size of the edges of cells near the fracture; and determining whether the edges of the cells near the fracture in the direction within the 3D space are longer than the size of the target multiplied by a scalar variable.  The cells near the fracture have flanges in the direction inside the 3D space that are longer than the size of the target multiplied by the scalar variable are among the cells that are refined.  Example 8 is the method of Example 7, wherein the determination of a location of the fracture comprises staggering the edges of cells near the fracture which are longer than the size of the multiplied target. by the scalar variable, and the determination if the staggered edges intersect the fracture.  The cells close to the fracture that have stepped flanges that cross the fracture are cells that are refined.  Example 9 is the method of Example 7 or 8, wherein the determination of a location of the fracture also includes staggering of the edges of cells near the fracture that are longer than the size of the fracture. the target multiplied by the scalar variable, and for any opposing pair of stepped edges, determining whether the stepped flanges intersect the fracture.  The cells close to the fracture that have exactly one ridge that crosses the fracture are the cells that are refined.  Example 10 is the method of Example 7-9, in which the determination of the location of the fracture also includes determining whether the edges of cells near the fracture that are longer than the size of the fracture. the target multiplied by the scalar variable, comprise two or more suspended nodes, and in which the cells close to the fracture which comprise two or more suspended nodes are refined.  Example 11 is the method of Example 7-10, wherein the anisotropic refinement of the cells in the mesh also comprises, for any opposite pair of edges of the cells near the fracture which are longer than the size of the target multiplied by the scalar variable, the determination if exactly one of the two edges comprise two or more suspended nodes, and in which the cells close to the fracture which comprise two or more suspended nodes are refined.  Example 12 is the method of Example 11, wherein the anisotropic refinement of the cells in the mesh comprises staggering the edges of cells near the fracture which are longer than a size of the target multiplied by the scalar variable, and the determination if the near-fracture cells with stepped flanks intersect a fracture while no near-fracture cells cross a fracture.  The cells close to the fracture that have staggered edges that cross the fracture are refined.  Example 13 is the method of Examples 7-11, wherein the anisotropic refinement of the cells in the mesh also comprises: determining a location of a fracture within an intermediate reservoir model ; identification of cells in a region close to the fracture; the identification of a direction within the 3D space in which the cells close to the fracture must be refined; and determining whether cells near the fracture having a ridge in the identified direction also have no more than one suspended node.  The cells close to the fractures that do not have more than one hanging knot are refined.  Example 14 is the method of Examples 7-11, wherein the anisotropic refinement of the cells also comprises: determining a location of a fracture within an intermediate reservoir model; the identification of cells in the region close to the fracture; and for any opposing pair of cell flanges near the fracture, determining if exactly one of the two opposite flanges has two or more suspended nodes.  The cells close to the fracture that have two or more hanging nodes are refined.  [0138] Example 15 is a computer readable storage medium that has executable instructions stored thereon.  If the instructions are executed by a computing device, the instructions allow the computing devices to perform the operations.  The instructions include: instructions for receiving a tank specification.  The instructions also include instructions for identifying, based on the tank specification, a set of fractures including 2,5D fractures (2.5D) and other fractures.  The instructions also include instructions for generating an intermediate reservoir model comprising an extrusion mesh that models fractures allowing 2.5D in a 3D space.  In response to the determination that the cells in the mesh must be refined in a direction within the 3D space, the instructions include instructions for refining the cells anisotropically in the mesh corresponding to the other fractures.  The instructions also include instructions for resolving a fracture network within the intermediate reservoir model using the refined cells, and instructions for generating a terrestrial reservoir model using the fracture network.  Example 16 is the storage medium of Example 15, in which the instructions for the anisotropic refinement of the cells in the mesh include instructions for identifying a direction within the 3D space in which cells must be refined, and instructions for dividing a rim of a cell, the rim being in the direction within the 3D space.  Example 17 is the storage medium of Examples 15 or 16, where fractures allowing 2.5D comprise fractures represented by one or more substantially vertical planes, and where the other fractures comprise fractures represented by one or more substantially horizontal plans.  [0141] Example 18 is a system that includes a processor, a display, and a memory having instructions stored thereon, which, when executed by a processor, causes the processor to execute operations.  Operations include receiving a tank specification and identifying, based on the tank specification, a set of fractures including 2.5D fractures allowing 2.5D, and other fractures.  The operations also include generating an intermediate reservoir model comprising an extrusion mesh that models fractures allowing 2.5D in a 3D space.  In response to the determination that the cells in the mesh must be refined in a direction within the 3D space, the operations include refining the cells anisotropically in the mesh corresponding to the other fractures.  The operations also include instructions for resolving a fracture network within the intermediate reservoir model using the refined cells, and then generating an Earth reservoir model using the fracture network.  Example 19 is the system of Example 18, where the 2.5D fractures comprise fractures represented by one or more vertical planes, and the other fractures comprise fractures represented by one or more horizontal planes.  Example 20 is the system of Examples 18 or 19, wherein the operations also include the display of the Earth tank model on the display device. 30

权利要求:
Claims (17)
[0001]
REVENDICATIONS1. A computer-implemented method, characterized by comprising: receiving, at a computing device (3000), a tank specification (204); identifying, based on the reservoir specification (204), a set of fractures (206, 208) comprising: fractures (206) for 2.5D; and other fractures (208); generating an intermediate reservoir model (212) comprising an extrusion mesh that models fractures (206) allowing 2.5D in a 3D space; responding to the determination that the cells in the mesh must be refined in a direction within the 3D space, refining the cells anisotropically in the mesh corresponding to the other fractures (208); resolving, by the computing device, a fracture network within the intermediate reservoir model (212) using the refined cells; and the generation of a terrestrial reservoir model using the fracture network.
[0002]
The computer-implemented method of claim 1, wherein the 2.5D-enabling fractures (206) have a geometry that has been discretized in a 2D plane by a collection of line segments, and wherein the generation of the model intermediate reservoir assembly (212) includes, for each line segment associated with each fracture in the fractures (206) for 2.5D: generating a set of stages at a specified radius from the line segment; generating closed loops around fracture line segments associated with a fracture (206) for 2.5D; and generating the form members within the closed loops of the segment of the line, the method also comprising: generating the grid mesh constrained around the closed loops of the fractures allowing the 2.5D to fill a space remaining in the 2D plane.
[0003]
The computer-implemented method of claim 1 or 2, further comprising: using a terrestrial tank model in a reservoir simulation; and displaying the reservoir simulation on a screen (108, 3030) of the computing device (3000).
[0004]
The computer-implemented method of any one of claims 1 to 3, wherein: the fractures (206) enabling the 2.5D may comprise fractures represented by one or more substantially vertical planes; and the other fractures comprise fractures represented by one or more substantially horizontal planes.
[0005]
The computer-implemented method of any one of claims 1 to 4, wherein the anisotropic refinement of the cells in the mesh comprises: identifying a direction within the 3D space in which the cells must to be refined; and dividing a cell rim, the rim being in the direction within the 3D space.
[0006]
The computer implemented method of claim 5, wherein the flange is divided along a plane that is perpendicular to an axis of a fracture within a fracture network.
[0007]
The computer-implemented method of any one of claims 1 to 6, wherein the anisotropic refinement of the cells comprises: determining a location of a fracture (206) in an intermediate reservoir model (212); the identification of cells in the region close to the fracture; determining a target size of the edges of cells near the fracture; and determining whether the edges of the cells near the fracture (206) in the direction within the 3D space are longer than the size of the target multiplied by a scalar variable, in which the cells close to the fracture having edges in the direction inside the 3D space that are longer than the size of the target multiplied by the scalar variable are among the cells that are refined.
[0008]
The computer-implemented method of claim 7, wherein determining the location of the fracture (206) comprises: staggering the edges of cells near the fracture that are longer than the size of the target multiplied by the scalar variable; and determining whether the stepped flanges intersect the fracture (206), wherein the near-fracture cells that have stepped flanges that intersect the fracture (206) are cells that are refined.
[0009]
The computer-implemented method of claim 7, wherein determining the location of the fracture (206) also includes: staggering the edges of cells near the fracture that are longer than the size of the multiplied target by the scalar variable; and for any opposing pair of stepped flanges, determining if exactly one of the two stepped flanges intersects a fracture (206), in which the cells close to the fracture (206) that have exactly one step flange crossing the fracture are the cells that are refined.
[0010]
The computer-implemented method of claim 7, wherein determining the location of the fracture (206) also includes: determining whether the edges of cells near the fracture (206) that are longer than the size of the fracture the target multiplied by the scalar variable comprises two or more suspended nodes, wherein the cells close to the fracture (206) having two or more hanging nodes are refined.
[0011]
The computer-implemented method of claim 7, wherein the anisotropic refinement of the cells in the mesh also includes, for any opposite pair of flanges of near-fracture cells (206) that are longer than the size of the target multiplied by the scalar variable, determining if exactly one of the two edges comprises two or more suspended nodes, and wherein the cells close to the fracture (206) having two or more suspended nodes are refined.
[0012]
The computer-implemented method of claim 11, wherein the anisotropic refinement of the cells in the mesh further comprises: staggering the edges of the cells near the fracture (206) that are longer than the target size multiplied by the scalar variable; and determining whether the near-fracture cells having stepped flanges intersect a fracture (206) while no near-fracture-like cells cross a fracture, wherein the near-fracture-like cells (206) that have step flanges who cross the fracture are refined.
[0013]
The computer-implemented method of claim 7, wherein the anisotropic refinement of cells in the mesh further comprises: determining a location of a fracture (206) in an intermediate reservoir model (212); identifying the cells in the region close to the fracture (206); identifying a direction within the 3D space in which cells near the fracture (206) are to be refined; and determining whether the cells near the fracture (206) having a ridge in the identified direction also have not more than one hanging knot, wherein the near fracture cells (206) that do not have more than one hanging knot are refined.
[0014]
The computer-implemented method of claim 7, wherein the anisotropic refinement of the cells in the mesh further comprises: determining a location of a fracture (206) in an intermediate reservoir model (212); identifying the cells in the region close to the fracture (206); and for any opposite pairs of cell flanges proximate to the fracture (206), determining if exactly one of the two opposite flanges has two or more suspended nodes, wherein the near-fracture cells (206) having the two or more hanging nodes are refined.
[0015]
A computer readable storage medium having executable instructions stored on said storage medium, which, when executed by a computing device (3000), enables the computing device (3000) to perform operations, characterized in that the instructions include instructions for carrying out a method according to any one of claims 1 to 14.
[0016]
16. System (100) characterized in that it comprises: a processor (102); a display device (108); and a memory having instructions stored thereon which, when executed by the processor (102), enables the processor to perform operations including operations for carrying out a method according to one any of claims 1 to 14.
[0017]
The system (100) of claim 16, further comprising: displaying, on the display (108), the Earth tank model.

类似技术:

---

公开号 | 公开日 | 专利标题

---

FR3028333A1|2016-05-13|CREATING A RESERVOIR MESH USING A PROLONGED ANISOTROPIC REFINEMENT, ADAPTED TO GEOMETRY, OF A POLYEDRE

---

Pellerin et al.2014|Automatic surface remeshing of 3D structural models at specified resolution: A method based on Voronoi diagrams

---

FR3041803A1|2017-03-31|

---

FR2987903A1|2013-09-13|GEOLOGICAL FAILURE STRUCTURES CONTAINING NONCONFORMITIES.

---

WO2014126959A2|2014-08-21|Systems, methods, and computer-readable media for modeling complex wellbores in field-scale reservoir simulation

---

EP2924232A2|2015-09-30|Method for constructing an optimised network for reservoir simulation in a subterranean formation

---

US10324228B2|2019-06-18|Gridless simulation of a fluvio-deltaic environment

---

US9880321B2|2018-01-30|Defining non-linear petrofacies for a reservoir simulation model

---

EP2873989A1|2015-05-20|Method for building a grid representing the distribution of a physical property of a subsurface formation by conditional multi-point statistical simulation

---

GB2529957B|2019-11-13|2.75D meshing algorithm

---

AU2013393306B2|2017-08-10|2.5D stadia meshing

---

FR3048300A1|2017-09-01|3D GEOCELLULAR HYBRID REPRESENTATION OF SELECTED NATURAL FRACTURE NETWORK SUBASSEMBLIES

---

US20160104320A1|2016-04-14|3d stadia algorithm for discrete network meshing

---

FR3055723A1|2018-03-09|POINT-VECTOR-BASED MODELING OF OIL TANK PROPERTIES FOR TANK-FREE SIMULATION MODEL

---

FR3027134A1|2016-04-15|USE OF A REPRESENTATIVE ELEMENTARY VOLUME FOR DETERMINING A SUB-SET VOLUME IN A GROUND MODEL OF A ZONE OF INTEREST

---

EP3017430B1|2021-04-07|Lofting algorithm for discrete network meshing

---

US10310116B2|2019-06-04|Method for determining a stacked model describing architectural elements

---

US10605055B2|2020-03-31|Integrated hydrocarbon fluid distribution modeling

---

EP3058398B1|2021-04-28|Method and device for the controlled determination of channels

---

EP3058397A1|2016-08-24|Method and device for determining groups of geological facies

---

FR2983615A1|2013-06-07|Method for refining simulation model used to simulate behavior of environment with sedimentary basin, involves producing chart based on portion of scalar field values, and applying chart to variable distributed in space in area

同族专利:

---

公开号 | 公开日

---

CA2963928C|2019-06-25|

---

WO2016076847A1|2016-05-19|

---

GB2545608B|2020-06-17|

---

GB2545608A|2017-06-21|

---

US20170299770A1|2017-10-19|

---

AR102183A1|2017-02-08|

---

US10564317B2|2020-02-18|

---

GB201705594D0|2017-05-24|

---

CA2963928A1|2016-05-19|

引用文献:

---

公开号 | 申请日 | 公开日 | 申请人 | 专利标题

---

US5740342A|1995-04-05|1998-04-14|Western Atlas International, Inc.|Method for generating a three-dimensional, locally-unstructured hybrid grid for sloping faults|

---

EP1107187A1|1999-12-10|2001-06-13|Institut Francais Du Petrole|Method for generating a grid on a heterogene formation traversed by one or more geometric discontinuities with the aim of realising simulations|

---

FR2801710B1|1999-11-29|2002-05-03|Inst Francais Du Petrole|METHOD FOR GENERATING A HYBRID MESH FOR MODELING A HETEROGENEOUS FORMATION CROSSED BY ONE OR MORE WELLS|

---

US7369973B2|2000-06-29|2008-05-06|Object Reservoir, Inc.|Method and system for representing reservoir systems|

---

AU7585201A|2000-06-29|2002-01-14|Object Reservoir, Inc.|Feature modeling in a finite element model|

---

US7248259B2|2001-12-12|2007-07-24|Technoguide As|Three dimensional geological model construction|

---

FR2870621B1|2004-05-21|2006-10-27|Inst Francais Du Petrole|METHOD FOR GENERATING A THREE-DIMENSIONALLY THREADED HYBRID MESH OF A HETEROGENEOUS FORMATION CROSSED BY ONE OR MORE GEOMETRIC DISCONTINUITIES FOR THE PURPOSE OF MAKING SIMULATIONS|

---

US7809537B2|2004-10-15|2010-10-05|Saudi Arabian Oil Company|Generalized well management in parallel reservoir simulation|

---

FR2891383B1|2005-09-26|2008-07-11|Inst Francais Du Petrole|METHOD FOR SIMULATING FLUID FLOWS WITHIN A DISCRETE MEDIA BY A HYBRID MESH|

---

US7925482B2|2006-10-13|2011-04-12|Object Reservoir, Inc.|Method and system for modeling and predicting hydraulic fracture performance in hydrocarbon reservoirs|

---

US8190414B2|2008-03-26|2012-05-29|Exxonmobil Upstream Research Company|Modeling of hydrocarbon reservoirs containing subsurface features|

---

WO2010033710A2|2008-09-19|2010-03-25|Chevron U.S.A. Inc.|Computer-implemented systems and methods for use in modeling a geomechanical reservoir system|

---

WO2010065774A2|2008-12-03|2010-06-10|Chevron U.S.A. Inc.|System and method for predicting fluid flow characteristics within fractured subsurface reservoirs|

---

MX2011005663A|2008-12-03|2011-06-16|Chevron Usa Inc|System and method of grid generation for discrete fracture modeling.|

---

FR2948215B1|2009-07-16|2011-06-24|Inst Francais Du Petrole|METHOD FOR GENERATING A HEXA-DOMINANT MESHING OF A FAILLED UNDERGROUND MEDIUM|

---

US8494827B2|2009-09-25|2013-07-23|Exxonmobil Upstream Research Company|Method of predicting natural fractures and damage in a subsurface region|

---

WO2011043862A1|2009-10-07|2011-04-14|Exxonmobil Upstream Research Company|Discretized physics-based models and simulations of subterranean regions, and methods for creating and using the same|

---

EP2499567A4|2009-11-12|2017-09-06|Exxonmobil Upstream Research Company|Method and apparatus for reservoir modeling and simulation|

---

US8727017B2|2010-04-22|2014-05-20|Exxonmobil Upstream Research Company|System and method for obtaining data on an unstructured grid|

---

US8731873B2|2010-04-26|2014-05-20|Exxonmobil Upstream Research Company|System and method for providing data corresponding to physical objects|

---

US8463586B2|2010-06-22|2013-06-11|Saudi Arabian Oil Company|Machine, program product, and computer-implemented method to simulate reservoirs as 2.5D unstructured grids|

---

US9058445B2|2010-07-29|2015-06-16|Exxonmobil Upstream Research Company|Method and system for reservoir modeling|

---

EP2649271A2|2010-12-10|2013-10-16|Baker Hughes Incorporated|A method to improve reservoir simulation and recovery from fractured reservoirs|

---

CN102110183B|2010-12-30|2011-09-28|中国石油化工股份有限公司胜利油田分公司地质科学研究院|Numerical simulation method for reflecting fluid channeling of fluid along great rifts of reservoir|

---

US9051825B2|2011-01-26|2015-06-09|Schlumberger Technology Corporation|Visualizing fluid flow in subsurface reservoirs|

---

US9194222B2|2011-04-19|2015-11-24|Halliburton Energy Services, Inc.|System and method for improved propped fracture geometry for high permeability reservoirs|

---

WO2013039606A1|2011-09-15|2013-03-21|Exxonmobil Upstream Research Company|Optimized matrix and vector operations in instruction limited algorithms that perform eos calculations|

---

FR2987138A1|2012-02-22|2013-08-23|Schlumberger Technology Corp|TITLE NOT RENTED.|

---

EP2820451A4|2012-04-10|2016-05-18|Halliburton Energy Services Inc|Method and apparatus for generating seismic to map subterranean fractures|

---

NO339744B1|2012-08-10|2017-01-30|Schlumberger Technology Bv|Hybrid, local non-reconciliation method for multiphase flow simulations in heterogeneous cracked media|

---

US20140136171A1|2012-11-13|2014-05-15|Chevron U.S.A. Inc.|Unstructured Grids For Modeling Reservoirs|

---

US20140236559A1|2013-02-18|2014-08-21|Saudi Arabian Oil Company|Systems, methods, and computer-readable media for modeling complex wellbores in field-scale reservoir simulation|

---

CN105474272A|2013-07-02|2016-04-06|兰德马克绘图国际公司|2.75D meshing algorithm|

---

EP3017429A4|2013-07-02|2017-05-10|Landmark Graphics Corporation|2.5d stadia meshing|

---

EP3167387A4|2014-07-11|2018-02-28|Landmark Graphics Corporation|Anisotropic geometry-adaptive refinement for reservoir mesh creation|

---

US10571601B2|2014-10-09|2020-02-25|Schlumberger Technology Corporation|Geologic fault representation for flow simulation|

---

WO2017082862A1|2015-11-09|2017-05-18|Halliburton Energy Services, Inc.|Fracture network fluid flow simulation with junction area modeling|EP3017429A4|2013-07-02|2017-05-10|Landmark Graphics Corporation|2.5d stadia meshing|

---

GB2556598B|2015-09-24|2021-07-21|Landmark Graphics Corp|Simulating fractured reservoirs using multiple meshes|

---

US20170184760A1|2015-12-23|2017-06-29|Paradigm Sciences Ltd.|Device, system and method for a structure and stratigraphy preserving transformation of a geological model|

---

US10255720B1|2016-10-13|2019-04-09|Bentley Systems, Incorporated|Hybrid mesh from 2.5D and 3D point data|

---

CN106991244B|2017-04-13|2019-06-28|河海大学|A method of Fracture Networks connectivity and seepage calculation based on graph theory|

---

CN108287371A|2018-01-31|2018-07-17|中南大学|Background grid Adaptive meshing method in dc resistivity element-free menthod|

---

US11156744B2|2019-01-10|2021-10-26|Emerson Paradigm Holding Llc|Imaging a subsurface geological model at a past intermediate restoration time|

---

US10520644B1|2019-01-10|2019-12-31|Emerson Paradigm Holding Llc|Imaging a subsurface geological model at a past intermediate restoration time|

法律状态:
2016-10-21| PLFP| Fee payment|Year of fee payment: 2 |

---

2017-10-26| PLFP| Fee payment|Year of fee payment: 3 |

---

2018-03-02| PLSC| Search report ready|Effective date: 20180302 |

---

2018-09-28| PLFP| Fee payment|Year of fee payment: 4 |

---

2019-10-30| PLFP| Fee payment|Year of fee payment: 5 |

---

2021-07-09| ST| Notification of lapse|Effective date: 20210605 |

优先权:

---

申请号 | 申请日 | 专利标题

---

PCT/US2014/065177|WO2016076847A1|2014-11-12|2014-11-12|Reservoir mesh creation using extended anisotropic, geometry-adaptive refinement of polyhedra|

---