A geological modeling method for a reservoir with uncertain geological features

Abstract

The present application relates to a geological modeling method for a geological feature uncertain reservoir. The method comprises the following steps: identifying uncertainty factors of the geological feature uncertain reservoir; parameterizing the uncertainty factors; determining possible distribution modes of the uncertainty parameters; respectively performing random sampling on each uncertainty parameter under each distribution mode to obtain a plurality of certain parameter combinations; each certain parameter combination corresponds to a certain geological model; predicting the geological feature uncertain reservoir through the certain geological model to obtain a plurality of prediction results; and statistically analyzing the prediction results to evaluate the prediction probability and prediction risk of each certain geological model. The present application provides a plurality of certain reservoir geological models varying with key geological features such as stratigraphic plane, fault plane and fracture plane, can be fitted through production history, can invert all the modeled uncertainty parameters, can reduce model uncertainty, and is more suitable for practical use.

Description

Technical Field

[0001] This invention belongs to the field of oil and gas extraction technology, and in particular relates to a geological modeling method for reservoirs with uncertain geological characteristics. Background Technology

[0002] With the large-scale development of oil and gas fields both domestically and internationally, conventional oil and gas resources are gradually entering the later stages of development, and their potential for further development is gradually decreasing. Unconventional reservoirs such as tight and fractured reservoirs have become an important battleground for future reserve and production increases. However, most unconventional reservoirs are currently buried at great depths, have complex geological characteristics, and lack clear geological understanding, resulting in significant uncertainties in models.

[0003] Currently, geological modeling research and approaches that consider uncertainty are a hot topic. However, related ideas and methods focus on geostatistical methods that consider uncertainty, primarily on characterizing the uncertainty of lithofacies and property models, while neglecting the uncertainty description of geological features such as reservoir structures, faults / fractures, etc. On the other hand, the geological models established by commercial geological modeling software are mainly corner grid models. This grid system, limited by its unique geometric structure, cannot automatically regenerate the grid according to changes in strata, fault planes, and fracture surfaces. Therefore, there is currently no geological modeling technology that comprehensively considers reservoir uncertainty, nor can it achieve the goal of inverting all modeling uncertainty parameters through production history fitting. Summary of the Invention

[0004] The main objective of this invention is to provide a geological modeling method for reservoirs with uncertain geological characteristics. The technical problem to be solved is how to provide a number of deterministic reservoir geological models that can change with variations in surface planes, fault planes, and fracture surfaces. Furthermore, by fitting production history, all uncertain parameters of the model can be inverted to reduce model uncertainty, thereby making the model more suitable for practical use.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a geological modeling method for reservoirs with uncertain geological characteristics, the method comprising the following steps:

[0006] Identify the uncertainties in reservoirs with uncertain geological characteristics;

[0007] The uncertainty factors are parameterized to obtain the uncertainty parameters;

[0008] Determine the possible distribution patterns of the aforementioned uncertainty parameters;

[0009] For each uncertainty parameter under each distribution mode, random sampling is performed to obtain several deterministic parameter combinations; each deterministic parameter combination corresponds to a deterministic geological model; several deterministic parameter combinations correspond to several deterministic geological models;

[0010] The aforementioned deterministic geological model is used to predict the geological characteristics of the uncertain reservoir, and several prediction results are obtained.

[0011] Statistical analysis was performed on the prediction results to evaluate the prediction probability and prediction risk of each deterministic geological model.

[0012] Preferably, the method further includes back-calculating the predictive consistency of each deterministic geological model through a historical fitting method to reduce the uncertainty of the geological model.

[0013] Preferably, the uncertainty factors include at least one of the following: uncertainty of model geometry, uncertainty of model lithofacies, and uncertainty of model properties.

[0014] Preferably, the uncertainty of the model geometry includes the uncertainty of the layer depth, the uncertainty of the fault location, the uncertainty of the crack location, the uncertainty of the cave location, and the uncertainty of the cave's outer envelope.

[0015] Preferably, the uncertainty of the model properties includes at least one of the following: uncertainty of pore saturation distribution, uncertainty of the conductivity of faults or fractures, and uncertainty of permeability inside the cavern.

[0016] Preferably, the parameterization of the uncertainty in the layer depth includes the following steps:

[0017] Establish a deterministic level model for level location;

[0018] In the aforementioned deterministic level model, a set of reference points p are arranged. t , where t represents a natural number;

[0019] A set of reference points p are arranged at each well point. m , where m represents a natural number and represents the actual data of the layer at the well point;

[0020] Reference point p t and p m The data is merged into a single data source, using its depth offset dz. t and dz m As observational data, the depth offset of the layer depth is calculated using Kriging interpolation and directly applied to the layer depth attribute.

[0021] Preferably, the parameterization of the uncertainty in the fault location includes the following steps:

[0022] A deterministic level model of the fault location is established; points located on the level of the deterministic level model are represented by three-dimensional coordinates of the x-axis, y-axis, and z-axis;

[0023] The deterministic level model is translated by dx, dy, and dz in the x, y, and z axes, respectively, to obtain new fault locations.

[0024] Preferably, the parameterization method for the uncertainty of the crack location is as follows:

[0025] The geometry of the crack is described by a combination of parameters;

[0026] Mesh modeling is performed, and the vertex coordinates of the cracks are calculated using the aforementioned parameter combination.

[0027] Preferably, the uncertainty of the location of the cave is described by the coordinates (x0, y0, z0) of the geometric center point of the cave; wherein the geometric center point is the centroid or the center of gravity.

[0028] Preferably, the parameterization method for the uncertainty of the cave's outer envelope includes the following steps:

[0029] An ellipsoid is used for description, and the lengths of the three axes of the ellipsoid are represented according to the parameters of the cave surface.

[0030] Perform mesh modeling, calculate the coordinates of the ellipsoid based on the ellipsoid model, and then move the center of the ellipsoid to the coordinate point (x0, y0, z0).

[0031] Preferably, the method for determining the possible distribution of the uncertainty parameter is as follows:

[0032] Obtain basic information for the corresponding block;

[0033] The variation range and distribution pattern of each uncertainty parameter are given; wherein, the variation range of the uncertainty parameter is its upper limit and lower limit; the distribution pattern includes normal distribution, uniform distribution or triangular distribution.

[0034] Preferably, the random sampling includes equally spaced sampling, Monte Carlo pure random sampling, Latin hypercube sampling, or orthogonal array sampling.

[0035] Preferably, obtaining a deterministic geological model through the deterministic parameter combination includes the following steps:

[0036] Based on the uncertainties in the depth of the strata, the location of the fault, the location of the fracture, the location of the karst cave, and the outer envelope of the karst cave, the geometric coordinates of the corresponding strata, fault plane, fracture surface, and karst cave surface are calculated, resulting in several deterministic geometric feature models.

[0037] In the aforementioned deterministic geometric feature model, the definite location of the fracture is calculated; using discrete fracture modeling technology, a peri-well infiltration zone is delineated around the fracture, and unstructured mesh generation is performed to obtain a mesh model;

[0038] In the aforementioned mesh model, interpolation is performed on the model's lithofacies and model properties.

[0039] The technical effects and advantages of this invention are as follows:

[0040] 1. The geological modeling method proposed in this invention is a set of geological modeling methods based on stochastic theory, which realizes the establishment of a geological model that comprehensively considers various uncertainties. During the modeling process, geological features such as stratigraphic structure, fault planes, fracture surfaces, lithofacies, and porosity / permeability can all change. Simultaneously, each change in geological feature is uniquely determined by several random parameters. Therefore, the entire geological model can be generated using a set of random parameters, and the uncertainties of the entire model can be further described. Using this method, multiple deterministic concrete implementations can be established, which can be used for reserve calculations or numerical simulations of oil and gas reservoirs, etc.

[0041] 2. The geological modeling method proposed in this invention can, on the one hand, understand the sensitivity of each uncertainty parameter by performing data statistics on the calculation results, such as reserves, production, and recovery rate; on the other hand, it can fit and invert the deterministic geological model obtained by this set of random parameters through historical production data, thereby narrowing the possible distribution range of each uncertainty parameter and achieving the purpose of reducing model uncertainty.

[0042] 3. The geological modeling method proposed in this invention can provide the distribution range of oilfield reserves and development indicators in the early stages of oilfield development when drilling data is scarce and uncertainties are high. This is of great significance for oilfield development investment decisions and scientific management.

[0043] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0044] Figure 1 A schematic diagram of a geological modeling method for reservoirs with uncertain geological characteristics;

[0045] Figure 2 A schematic diagram of the stratigraphy and well locations of the study block estimated before uncertainty modeling;

[0046] Figure 3 A schematic diagram illustrating the arrangement of reference points for a depth-uncertain plane and the alteration of the depth of the entire plane through interpolation;

[0047] Figure 4 This is a schematic diagram of the ground stratum model obtained after adjusting the depth.

[0048] Figure 5 This is a schematic diagram of the large cracks contained in the model;

[0049] Figure 6 This is a schematic diagram illustrating multiple implementations of various large cracks;

[0050] Figure 7 It is a grid-refined area delineated based on the distribution of large cracks;

[0051] Figure 8 It is a mesh model generated based on a set of defined geometric parameters;

[0052] Figure 9 It is a permeability field generated by interpolating attribute uncertainty parameters based on a grid model. Detailed Implementation

[0053] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0054] To further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, the following detailed description, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a geological modeling method for reservoirs with uncertain geological characteristics proposed according to the present invention.

[0055] In the geological modeling of reservoirs with uncertain geological characteristics, it is necessary to fully consider the uncertainties in geological understanding, especially the uncertainties in key geological features such as bedding planes, fault planes, and fracture surfaces. The uncertainties in oil and gas reservoir geological modeling come from multiple aspects, including uncertainties in geological structure, lithofacies distribution, porosity / permeability / saturation properties, and key geological features such as fractures / cavities.

[0056] Geological structure determines the spatial location of a geological model and has a significant impact on lithofacies distribution and porosity-permeability-saturation distribution. Therefore, these are all important factors that need to be considered when modeling uncertain reservoir geology.

[0057] Besides geological structures, lithofacies modeling and property modeling also involve geostatistical methods, each containing a large number of random parameters. Geological features such as fractures and caves are also major controlling factors for formation fluid flow, but due to limitations in geological exploration methods, the location and scale of these key features are subject to considerable uncertainty.

[0058] In summary, to comprehensively describe the uncertainties and risks inherent in reservoir geological characteristics, the geological modeling process needs to consider parameter uncertainties from various sources, forming a geological modeling process under given uncertain parameters. This modeling method can be directly integrated with the historical data fitting process, thereby using historical data fitting of production dynamics to infer uncertain parameters. Ultimately, a relatively accurate geological model is obtained.

[0059] Therefore, this invention proposes a geological modeling method for reservoirs with uncertain geological characteristics, specifically as follows: Figure 1 As shown, the method includes the following steps:

[0060] Identify the uncertainties in reservoirs with uncertain geological characteristics;

[0061] The aforementioned uncertainty factors are parameterized to obtain uncertainty parameters;

[0062] Determine the possible distribution patterns of the aforementioned uncertainty parameters;

[0063] For each uncertainty parameter under each distribution mode, random sampling is performed to obtain several deterministic parameter combinations; each deterministic parameter combination corresponds to a deterministic geological model; several deterministic parameter combinations correspond to several deterministic geological models;

[0064] The aforementioned deterministic geological model is used to predict the geological characteristics of the uncertain reservoir, and several prediction results are obtained.

[0065] Statistical analysis was performed on the prediction results to evaluate the prediction probability and prediction risk of each deterministic geological model.

[0066] Using the aforementioned geological modeling method for reservoirs with uncertain geological characteristics, countless possible deterministic reservoir models can be obtained by randomly sampling and assigning values ​​to parameters. These numerous deterministic reservoir models are then fitted using historical data to verify the conformity of the proposed deterministic reservoir model.

[0067] The uncertainty of geological models can be reduced by back-calculating the predictive consistency of each deterministic geological model through historical fitting methods.

[0068] The historical fitting method can be either a manual historical fitting method or an automated historical fitting method, such as the ensemble Kalman filter method. The back-calculation method is used to verify the conformity between the geological model prediction results determined by the set of random parameters and the actual production data, thereby reducing the uncertainty of the geological model.

[0069] Preferably, the uncertainty factors include at least one of the following: uncertainty of model geometry, uncertainty of model lithofacies, and uncertainty of model properties.

[0070] Preferably, the uncertainty of the model geometry includes the uncertainty of the layer depth, the uncertainty of the fault location, the uncertainty of the crack location, the uncertainty of the cave location, and the uncertainty of the cave's outer envelope.

[0071] Since the distance between two adjacent layers is the formation thickness, the uncertainty of the layer depth can also be called the uncertainty of the formation thickness. These two concepts are equivalent.

[0072] Model lithofacies refer to rocks or rock assemblages formed in a specific sedimentary environment and are the main components of sedimentary facies. Lithofacies and sedimentary facies are subordinate to each other, not the same.

[0073] Preferably, the uncertainty of the model properties includes at least one of the following: uncertainty of pore saturation distribution, uncertainty of the conductivity of faults or fractures, and uncertainty of permeability inside the cavern.

[0074] After identifying the aforementioned uncertainties, the next step is to parameterize these uncertainties. For the parameterization of uncertainties related to lithofacies and properties, since this type of uncertainty is relatively simple to parameterize, it mainly includes random parameters from geostatistical methods used in lithofacies or property modeling based on geological grid models, parameters describing fault conductivity, and parameters related to fracture / cavity porosity and permeability, etc., which can be parameterized using existing techniques in the field. For the parameterization of uncertainties related to model geometry, the technical solution of this invention proposes the following parameterization method, detailed below:

[0075] The parameterization of the uncertainty in the layer depth includes the following steps:

[0076] A. Establish a deterministic level model for level location;

[0077] B. In the aforementioned deterministic level model, a set of reference points p are arranged. t , t = 1, 2, 3, ..., where t represents a natural number, and its depth offset is dz. t , t=1,2,3…, are called uncertainty parameters;

[0078] C. Arrange a set of reference points p at each well point. m m = 1, 2, 3…, where m represents a natural number, and dz represents the actual data of the layer at the well point, its depth offset. m m = 1, 2, 3... are all equal to 0;

[0079] D. The above reference point p t and p m The data is merged into a single data source, using its depth offset dz. t and dz m As observational data, the depth offset of the layer depth is calculated using Kriging interpolation and directly applied to the layer depth attribute.

[0080] The original stratigraphic location still needs to be established using geological modeling software such as Petrel or Gocad. However, the deterministic stratigraphic models created by these software cannot reflect the uncertainty of stratigraphic depth. This is because while the depths at well points are accurate in these deterministic models, in areas without well control constraints, the depths are only estimates that conform to geological statistical laws and contain significant uncertainties.

[0081] To quantitatively describe this uncertainty, this invention proposes a method for layer depth parameterization, which allows for the placement of a set of reference points p in areas without well control constraints or where uncertainty is considered high. t (t = 1, 2, 3, ..., representing the reference point number), and set the depth offset dz at the reference point. t (t = 1, 2, 3, ...) are uncertain parameters. Then, given all dz... t After (t=1,2,3,…), the depth of the entire formation can be readjusted using the following method: First, assume that there is also a reference point p at each well point. m (m = 1, 2, 3, ..., representing the number of wells with hard data for the bedding planes), and considering that the bedding plane depth at the well point is accurate (because hard data correction is used), the depth offset dz at that point is... m (m=1,2,3,…) are all equal to 0. Then, the two sets of reference points (p) are set to 0. t and p m Merge them, and then use their depth offset (dz) t and dz m Using the observation data, the Kriging method is used to interpolate and calculate the depth offset across the entire field, which is then directly applied to the depth attributes of the entire layer. This achieves the goal of depth adjustment.

[0082] The parameterization of the uncertainty in the fault location includes the following steps:

[0083] A. Establish a deterministic level model of the fault location; points located on the level of the deterministic level model are represented by three-dimensional coordinates of the x-axis, y-axis, and z-axis;

[0084] B. Translate the aforementioned deterministic level model by dx, dy, and dz in the x-axis, y-axis, and z-axis directions, respectively, to obtain new fault locations.

[0085] The initial location and morphology of the fault still need to be established using geological modeling software such as Petrel or Gocad. However, the deterministic fault geometry model established by such software cannot reflect the uncertainty of the fault location and shape, so it still needs to be parameterized.

[0086] This invention proposes using translational offsets to describe the uncertainty of fault location. Specifically, by adding translational variables (dx, dy, dz) to the fault plane formed during modeling, a new fault location is generated. It should be noted that fault movement may alter their intersection patterns. For example, previously intersecting faults may no longer intersect, previously non-intersecting faults may intersect, and the intersection line of previously intersecting faults may change. Generally, changes in intersection patterns do not affect subsequent modeling processes, but in some extreme cases, they may increase the computational difficulty of modeling. Therefore, in situations that might disrupt the modeling process, intersecting faults can be moved as a whole.

[0087] The parameterization method for the uncertainty of the crack location is as follows:

[0088] A. Describe the geometric morphology of the crack through parameter combinations;

[0089] The parameter combination is set as follows: the crack is defined by the crack's starting point, side length, dip angle, and azimuth angle; wherein, a two-dimensional crack is represented by the crack's starting point coordinates (x0, y0), length l, and dip angle θ; a three-dimensional crack is represented by the crack surface's vertex coordinates (x0', y0', z0'), side lengths l1 and l2, dip angle θ', and azimuth angle ψ.

[0090] B. Mesh modeling; wherein, during mesh modeling, the vertex coordinates of the crack are calculated using the aforementioned parameter combination.

[0091] To perform unstructured mesh generation, the geometric information of cracks is generally represented using vertex coordinates. For two-dimensional problems, this is represented by the starting coordinates (x0, y0) and the ending coordinates (x1, y1); for three-dimensional problems, it is represented by the coordinates of the four vertices of the crack surface (x0, y0, x1, y1). i y i zi The parameters are represented by i = 0, 1, 2, 3. While this parameterization method is simple and intuitive, it fails to reflect key characteristics such as the crack's scale and dip angle. Therefore, the technical solution of this invention proposes defining the crack by its starting point, side length, dip angle, and azimuth angle, and then performing uncertainty parameterization based on this.

[0092] The uncertainty of the location of the cave is described by the coordinates (x0, y0, z0) of the geometric center point of the cave; wherein the geometric center point is the centroid or centroid.

[0093] The parameterization method for the uncertainty of the outer envelope of the cave is as follows:

[0094] The shape of the outer envelope of the cave is described by an ellipsoid, where the parameters of the cave surface are a, b, and c, which represent the lengths of the three axes of the ellipsoid, respectively.

[0095] When performing mesh modeling, firstly, based on the ellipsoid equation x... 2 / a 2 +y 2 / b 2 +z 2 / c 2 =1 Calculate the coordinates of the ellipsoid; then move the center of the ellipsoid to the coordinate point (x0, y0, z0).

[0096] The shape of the karst cave is irregular, but due to the limitations of current exploration technology, it is difficult to accurately describe its shape. The technical solution of this invention proposes using an ellipsoid to describe the shape of the karst cave; of course, other suitable shapes can also be used to describe it based on the actual shape of the karst cave. When using an ellipsoid to describe the shape of the karst cave, the parameterization method for the uncertainty of its outer envelope is as described above.

[0097] Preferably, the method for determining the possible distribution of the uncertainty parameter is as follows:

[0098] A. Obtain the basic information for the corresponding block;

[0099] B. Based on experience in oil and gas reservoir engineering, the variation range and distribution pattern of each uncertainty parameter are given; wherein, the variation range of the uncertainty parameter is its upper limit and lower limit; the distribution pattern includes normal distribution, uniform distribution or triangular distribution.

[0100] Preferably, the random sampling includes equidistant sampling, Monte Carlo pure random sampling, Latin hypercube sampling, or orthogonal array sampling.

[0101] It should be noted that for some variables with a large range of values, such as crack permeability, the possible units of value are from 1×10⁻⁶. 3 mD to 1×10 6 mD has a large range, and its logarithmic values ​​are generally sampled. By randomly sampling each uncertain parameter, a set of deterministic parameter combinations can be obtained, and this set of parameter combinations corresponds to a deterministic geological model.

[0102] Preferably, obtaining a deterministic geological model from the deterministic parameter combination includes the following steps:

[0103] A. Establishment of deterministic geometric feature models: Based on the uncertainties of the surface depth, fault location, fracture location, cave location, and cave envelope, the geometric coordinates of the corresponding ground surface, fault surface, fracture surface, and cave surface are calculated to obtain several deterministic geometric feature models.

[0104] B. Unstructured mesh generation: In the deterministic geometric feature model, the location of the fractures is calculated; using discrete fracture modeling technology, a perimeter infiltration zone is delineated around the fractures, and unstructured mesh generation is performed to obtain a mesh model;

[0105] C. Attribute interpolation within the mesh: In the mesh model, interpolate the model lithofacies and model attributes within it.

[0106] This invention employs a discrete fracture model as the mesh system for modeling. Unlike traditional corner mesh models, the discrete fracture mesh model is an unstructured mesh, allowing for flexible control of mesh density at different locations. This enables mesh refinement near key geological features, resulting in a more accurate simulation of flow processes. The geological modeling method in this invention uses an unstructured mesh model, where the mesh can be of various types, such as tetrahedrons, triangular prisms, PEBI, and EDFM. The peri-well refinement region is used for subsequent mesh refinement during modeling; that is, within this region, a finer and denser mesh can be used for subdivision, thereby accurately simulating peri-well flow processes.

[0107] The geological modeling methods described above describe the stochasticity of a geological model and establish a single deterministic implementation, thereby creating a deterministic geological model. This deterministic geological model is automatically generated by geological simulation software. Using these methods, multiple deterministic geological models can be established, enabling the calculation of reserves, numerical simulation of oil and gas reservoirs, and history fitting. This work has two significant implications: first, by statistically analyzing the calculation results (reserves, production, recovery rate, etc.) and generating cross-validation plots or tornado diagrams, the sensitivity of various uncertainty parameters can be studied; second, through history fitting, the possible distribution range of each uncertainty parameter can be narrowed, thereby reducing model uncertainty.

[0108] The following detailed description uses specific examples: Unlike conventional modeling methods, the geological modeling method described in this invention is characterized by the fact that all parameters are uncertain during the modeling process, and values ​​need to be randomly selected according to the distribution pattern of each parameter, thereby establishing a specific and definite geological model.

[0109] This embodiment uses a simple but general model for illustration, which is a small fault-block reservoir, as shown in the attached figure. Figure 2 As shown, it contains two smaller layers, and from top to bottom, there are three ground layers, denoted as I, II, and III. The model contains several large cracks. A specific embodiment of the present invention includes the following steps:

[0110] Step 1: Summarize the uncertainty variables of the statistical model. The uncertainty parameters in this embodiment include:

[0111] 1) In the construction parameters, the depth of layer II is uncertain, and the location of faults in the model is uncertain;

[0112] 2) The random parameters in the geostatistical methods used for attribute interpolation in the model have uncertainty.

[0113] Step 2: Based on the uncertainty range of the stratum depth, generate an example of the ground stratum:

[0114] As attached Figure 3 As shown, in this embodiment, the depth variation of the intermediate formation II is uncertain, but at two well points (as shown in the attached diagram)... Figure 3 The depths at the top left and bottom right corners are fixed and constant. However, in areas without well points, the depth is random. We perform geological modeling according to the technical solution of this invention, in the attached... Figure 3Two reference points are placed at the upper right and lower left corners, respectively, and the depth variation of these reference points is assumed to be a random variable. For a specific implementation, once these two reference points have determined values, the depths at other locations across the entire plane can be calculated using Kriging interpolation, as shown in the attached figures. Figure 3 and attached Figure 4 As shown.

[0115] Step 3: Example of generating a crack model: The model includes several large cracks, as shown in the attached image. Figure 5 As shown. However, the location of the crack is uncertain. Let's take any one of the cracks as an example, for instance, the one attached... Figure 5 Taking the light gray crack below as an example, assuming that its length, orientation, and center point are all uncertain, a series of specific implementations can be generated according to the methods in the implementation steps, as shown in the appendix. Figure 6 As shown, each of the fissures corresponds to a deterministic geological model.

[0116] Step 4: Delineate the Refinement Area: In this embodiment, since a triangular prism mesh (a type of unstructured mesh) will be used for mesh modeling later, this model allows for local mesh refinement. Considering that a denser mesh is needed near the crack to characterize the flow, we delineate the area near the crack and refine it during mesh modeling, as shown in the attached diagram. Figure 7 As shown.

[0117] Step 5: Unstructured Mesh Modeling: The above steps have assigned specific values ​​to all geometric parameters. Based on this, a mesh model can be built. (See attached image) Figure 8 As shown. (Attached) Figure 8 Different shades of color are used to represent the upper and lower strata.

[0118] Step 6: Based on the mesh model, perform geological attribute interpolation: Extract the center point of each triangular prism mesh, and then use geostatistical methods such as Kriging to interpolate attributes such as porosity, permeability, and saturation. This process is a standard geological modeling technique and will not be elaborated further. The permeability distribution after attribute modeling is shown in the appendix. Figure 9 As shown.

[0119] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A geological modeling method for reservoirs with uncertain geological characteristics, characterized in that, The method includes the following steps: Identify the uncertainty factors of reservoirs with uncertain geological characteristics. The uncertainty factors include the uncertainty of model geometry, the uncertainty of model lithofacies, and the uncertainty of model properties. The uncertainty of model geometry includes the uncertainty of bedding depth, the uncertainty of fault location, the uncertainty of fracture location, the uncertainty of cave location, and the uncertainty of the cave envelope surface. The uncertainty factors are parameterized to obtain the uncertainty parameters; Determine the possible distribution patterns of the aforementioned uncertainty parameters; For each uncertainty parameter under each distribution mode, random sampling is performed to obtain several deterministic parameter combinations; each deterministic parameter combination corresponds to a deterministic geological model; several deterministic parameter combinations correspond to several deterministic geological models; obtaining a deterministic geological model from the deterministic parameter combinations includes the following steps: Based on the uncertainties in the depth of the strata, the location of the fault, the location of the fracture, the location of the karst cave, and the outer envelope of the karst cave, the geometric coordinates of the corresponding strata, fault plane, fracture surface, and karst cave surface are calculated, resulting in several deterministic geometric feature models. In the aforementioned deterministic geometric feature model, the definite location of the fracture is calculated; using discrete fracture modeling technology, a peri-well infiltration zone is delineated around the fracture, and unstructured mesh generation is performed to obtain a mesh model; In the aforementioned mesh model, interpolation is performed on the model lithofacies and model properties within it; The aforementioned deterministic geological model is used to predict the geological characteristics of the uncertain reservoir, and several prediction results are obtained. Statistical analysis was performed on the prediction results to evaluate the prediction probability and prediction risk of each deterministic geological model.

2. The geological modeling method according to claim 1, characterized in that, The method also includes back-calculating the predictive consistency of each deterministic geological model through historical fitting methods to reduce the uncertainty of the geological model.

3. The geological modeling method according to claim 1, characterized in that, The uncertainties in the model properties include at least one of the following: the uncertainty of pore saturation distribution, the uncertainty of the conductivity of faults or fractures, and the uncertainty of permeability inside the cavern.

4. The geological modeling method according to claim 1, characterized in that, The parameterization of the uncertainty in the layer depth includes the following steps: Establish a deterministic level model for level location; In the aforementioned deterministic level model, a set of reference points are arranged. , where t represents a natural number; Set up a set of reference points at each well point , where m represents a natural number and represents the actual data of the layer at the well point; Reference point and The data is merged into a single data source, using its depth offset. and As observational data, the depth offset of the layer depth is calculated using Kriging interpolation and directly applied to the layer depth attribute.

5. The geological modeling method according to claim 1, characterized in that, The parameterization of the uncertainty in the fault location includes the following steps: A deterministic level model of the fault location is established; points located on the level of the deterministic level model are represented by three-dimensional coordinates of the x-axis, y-axis, and z-axis; The deterministic level model is translated by dx, dy, and dz in the x, y, and z axes, respectively, to obtain new fault locations.

6. The geological modeling method according to claim 1, characterized in that, The parameterization method for the uncertainty of the crack location is as follows: The geometry of the crack is described by a combination of parameters; Mesh modeling is performed, and the vertex coordinates of the cracks are calculated using the aforementioned parameter combination.

7. The geological modeling method according to claim 1, characterized in that, The uncertainty in the location of the cave is determined by the coordinates of its geometric center point. The description is as follows; wherein the geometric center point is the centroid or the center of gravity.

8. The geological modeling method according to claim 1, characterized in that, The parameterization method for the uncertainty of the outer envelope of the cave includes the following steps: An ellipsoid is used for description, and the lengths of the three axes of the ellipsoid are represented according to the parameters of the cave surface. Perform mesh modeling, calculate the coordinates of the ellipsoid based on the ellipsoid model, and then move the center of the ellipsoid to the coordinate system. point.

9. The geological modeling method according to claim 1, characterized in that, The method for determining the possible distribution of the uncertainty parameter is as follows: Obtain basic information for the corresponding block; The variation range and distribution pattern of the uncertainty parameter are given; wherein, the variation range of the uncertainty parameter is its upper limit and lower limit; the distribution pattern includes normal distribution, uniform distribution or triangular distribution.

10. The geological modeling method according to claim 1, characterized in that, The random sampling includes equidistant sampling, Monte Carlo pure random sampling, Latin hypercube sampling, or orthogonal array sampling.

Images