CN115407407A - Three-dimensional geological model construction method for carbonate rock ancient karst cave and filling thereof - Google Patents

Abstract

The invention discloses a three-dimensional geological model construction method for carbonate rock ancient karst caves and filling thereof, which comprises the following steps: establishing a karst cave compound model according to seismic interpretation data of an area to be analyzed; acquiring a single karst cave geometric parameter based on field outcrop data, and establishing a single karst cave internal filling distribution model by respectively using the single karst cave geometric parameter and the karst cave complex model as input conditions and constraints; obtaining a training model representing the filling state in the single karst cave based on field outcrop data and well drilling and logging interpretation data, and performing multipoint geological statistics on the filling distribution model in the single karst cave based on the training model to obtain a corresponding simulation model; and performing seismic forward modeling on the simulation model to obtain a seismic data volume for calculation, and generating an optimal geological model for representing the structure in the karst cave by combining actual seismic interpretation data. The method improves the three-dimensional representation precision and accuracy of the underground ancient karst cave reservoir.

Description

Three-dimensional geological model construction method for carbonate rock ancient karst cave and filling thereof

Technical Field

The invention relates to the technical field of oil and gas reservoir characterization, in particular to a three-dimensional geological model construction method for carbonate rock ancient karst caves and filling thereof.

Background

The carbonate rock fracture-cave type oil-gas reservoir is the most main oil-gas reservoir type of Ordovician series in a Tarim basin, is one of the most main formations for increasing storage and increasing production at present in China, and is typical such as a medium petrochemical Tahe oil field, a northward oil field, a medium petroleum Ha Laha pond, an England-force oil field and the like. Years of exploration and development practice show that the large ancient karst cave is the most main reservoir body of the fracture-cavity type oil reservoir, almost occupies 80% of geological reserves, and the establishment of an accurate underground ancient karst cave distribution model has important guiding significance for evaluating the geological reserves and efficiently developing.

The existing three-dimensional geological modeling related research on the ancient karst cave reservoir body can be summarized into the following categories: (1) And (3) quantitatively analyzing the large karst cave based on karst outcrop and establishing an outcrop model. (2) And (3) exploring and researching a multivariate and multiscale fracture-cavity type oil reservoir three-dimensional modeling method. (3) And (5) researching the filling characteristics in the large karst cave and a logging interpretation technology.

In the process of implementing the present invention, the inventor finds that at least the following problems exist in the prior art:

(1) The scale relation of the representation of the earthquake, geological data and the underground ancient karst cave needs to be further determined, and the geological meaning represented by the earthquake attribute needs to be determined in the process of developing earthquake karst cave prediction;

(2) The existing ancient karst cave modeling method mainly focuses on classification and scale division modeling and fusion, namely a deterministic modeling method is mainly adopted for a large karst cave, a sequential indication or multipoint statistics is mainly adopted for random simulation for a karst cave (small karst cave), and then fusion is carried out according to a primary-secondary relation, so that the method has practicability from the perspective of production practice, and has two problems to be improved from the perspective of method principle: firstly, a large karst cave model established by adopting an attribute truncation method based on earthquake prediction is not a true single karst cave, and the geological meaning of the large karst cave model needs to be further defined and the internal structure needs to be refined by a geological modeling method; secondly, different reservoirs 'violate' the principle of phase proportion conservation in statistical modeling in the process of fusion according to primary and secondary relations, and need to be further improved from the optimization of a modeling method;

(3) The existing research on filling characteristics inside the karst cave mostly focuses on outcrop and well logging explanation, and the established filling structure mode is mainly two-dimensional and lacks of expansion and application to a three-dimensional model. In addition, a corresponding exploration is lacked in the aspect of a filling structure three-dimensional modeling method in the underground ancient karst cave reservoir body.

Therefore, the prior art needs to provide a three-dimensional quantitative geologic model construction scheme for an ancient karst cave of carbonate rock and an internal filling structure thereof, so as to solve one or more of the technical problems.

Disclosure of Invention

In order to solve the technical problems, the invention provides a three-dimensional geological model construction method for carbonate rock ancient karst caves and filling thereof, which comprises the following steps: establishing a karst cave compound model according to the seismic interpretation data of the area to be analyzed; obtaining a single karst cave geometric parameter based on field outcrop data, and establishing a single karst cave internal filling distribution model by respectively using the single karst cave geometric parameter and the karst cave compound model as input conditions and constraints; obtaining a training model representing the filling state in the single karst cave based on the field outcrop data and the well drilling and logging data, and carrying out multipoint geological statistics on the filling distribution model in the single karst cave based on the training model to obtain a corresponding simulation model; and performing seismic forward modeling on the simulation model to obtain a seismic data volume for calculation, and generating an optimal geological model for representing the structure in the karst cave by combining actual seismic interpretation data.

Preferably, in the process of establishing the karst cave compound model, the method comprises the following steps: extracting a seismic texture attribute data volume from the seismic interpretation data, and converting the seismic texture attribute data volume into a seismic probability volume representing the contour of the karst cave complex; and taking the actually measured development thickness of the karst cave complex drilled as a threshold value, referring to field outcrop data, and taking the earthquake probability body as a constraint to establish a distribution model of the karst cave complex.

Preferably, in the process of establishing the filling distribution model inside the single cavern, the method comprises the following steps: according to the field outcrop data, carrying out field measurement on a plurality of single karst caves in the karst cave complex, and counting the height, width, extension length and section morphological characteristics of each single karst cave; and taking the karst cave complex model as a frame constraint, taking the geometric dimension statistical data of the single karst cave as an input condition, and combining logging interpretation data to perform simulation calculation on each single karst cave in the karst cave complex model to obtain a filling distribution model in the single karst cave.

Preferably, in the process of generating the training model, the method comprises: and obtaining a karst cave filling longitudinal sequence based on well drilling interpretation data, and modeling the karst cave filling longitudinal sequence to obtain the training model of the internal longitudinal filling sequence of the single karst cave.

Preferably, the training model is scanned according to the actually measured well data, and a search tree is established; and establishing a plurality of random lines, and searching the filling distribution model in the single karst cave based on a search tree aiming at each random line so as to generate the corresponding simulation model.

Preferably, the plurality of simulation models are respectively compared with actual seismic interpretation data in a visual-based structural mode, and the error between each simulation model and the actual seismic interpretation data is calculated, so that the optimal geological model is optimized according to the error.

Preferably, performing petrophysical analysis on the simulation model to establish a corresponding initial model; and converting the initial model into the seismic data volume for calculation by a seismic convolution or wave equation forward modeling method.

Preferably, the thickness data of the karst cave complex encountered by each well is obtained based on the single well interpretation result information; establishing a preset grid model, and coarsening the thickness data into the preset grid model to form discrete karst cave complex data and a corresponding probability percentage curve aiming at the current karst cave complex; converting the discrete karst cave complex data into pseudo-Gaussian data; fitting the variation function by using the scale data of the karst cave complex obtained by outcrop measurement; based on the pseudo-Gaussian data, a three-dimensional continuous variable Gaussian distribution field is obtained by using a preset Gaussian theoretical model and a fitted variation function, taking the seismic probability body as a constraint and adopting a sequential Gaussian simulation method; and performing Gaussian field truncation processing on the three-dimensional continuous variable Gaussian distribution field by taking the probability percentage curve as a threshold value curve to obtain the karst cave complex distribution model containing the discrete variable value information of each grid point.

Preferably, predicting the central point position of a single karst cave from the karst cave complex model; carrying out consistency matching on the karst cave complex model and the karst cave interpretation data of the drilled well, and entering the next step after matching; generating an empirical cumulative probability distribution function about the size characteristics of different positions in the current single karst cave according to the geometric dimension statistical data and the drilling and logging information, extracting the height of the current single karst cave from the empirical cumulative probability distribution function, and then assigning the height to the corresponding position of the karst cave complex model; calculating the actual width of the karst cave through the geometric dimension statistical data, and carrying out width assignment on the karst cave compound model; and setting a target function and a threshold value according to the height and the actual width of the single karst cave and the statistical data of the geometric dimension, and performing simulation calculation on the karst cave compound model subjected to height and width assignment to form the filling distribution model in the single karst cave.

Preferably, the measured well data are correspondingly marked at corresponding positions of the filling distribution model in the single karst cave; and scanning the training model according to a preset data template to obtain the search tree.

Compared with the prior art, one or more embodiments in the above scheme can have the following advantages or beneficial effects:

the invention discloses a three-dimensional geological model construction method for carbonate rock ancient karst caves and filling thereof. The method considers the scale of the karst cave scale and the precision of the representation data, and proposes to establish a three-dimensional model of the karst cave complex by adopting a truncated Gaussian method under seismic inversion constraint; under the restriction of the karst cave complex, further based on karst cave field outcrop measurement data, a single karst cave three-dimensional model is established by adopting a target-based indicative point process method; in order to further improve the representation precision of the karst cave reservoir stratum, based on the knowledge of the karst cave filling structure, a multi-point geostatistics method is adopted to establish a karst cave filling structure model, and a model with the highest actual goodness of fit is optimized through an earthquake forward modeling method to realize condition modeling. According to the comprehensive modeling method, three-dimensional modeling method research of a karst cave complex, a single karst cave and a filling structure inside the karst cave is carried out from large scale to small scale through a three-step walking strategy, a set of hierarchical constraint ancient karst cave and filling structure multi-level modeling method is formed, three-dimensional representation precision and accuracy of an underground ancient karst cave reservoir are improved, earthquake, geology, field outcrop, well logging and well drilling data are utilized to the maximum extent, and modeling method optimization of different representation scales is realized. The comprehensive karst cave modeling technology provided by the invention has strong applicability to the fine description of the tower river fracture-cave type reservoir bodies.

Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

Drawings

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:

fig. 1 is a step diagram of a three-dimensional geological model construction method for carbonate rock palsy and filling thereof according to an embodiment of the present application.

Fig. 2 is a specific flow diagram of a three-dimensional geological model construction method for carbonate rock palsy and filling thereof according to an embodiment of the present application.

Fig. 3 is a schematic cross-sectional view of a seismic probability volume in the three-dimensional geological model construction method for carbonate rock palsy and filling thereof according to the embodiment of the present application.

Fig. 4 is a schematic diagram of a measurement principle of a single karst cave geometric parameter in a three-dimensional geological model construction method for an ancient carbonate karst cave and filling thereof according to an embodiment of the present application.

Fig. 5 is a schematic diagram illustrating an effect of a karst cave complex distribution model in the three-dimensional geological model construction method for an ancient karst cave of carbonate rock and filling thereof according to the embodiment of the present application.

Fig. 6 is a schematic cross-sectional effect diagram of a single karst cave internal filling distribution model in the three-dimensional geological model construction method for the carbonate rock ancient karst cave and filling thereof according to the embodiment of the present application.

Fig. 7 is a schematic diagram illustrating the obtaining principle of a training model in the method for constructing the three-dimensional geological model for the carbonate rock palsy and filling thereof according to the embodiment of the present application.

Fig. 8 is an exemplary diagram of a cross section of a simulation model in the three-dimensional geological model construction method for the carbonate rock palsy and the filling thereof according to the embodiment of the present application.

Detailed Description

The following detailed description of the embodiments of the present invention will be provided with reference to the drawings and examples, so that how to apply the technical means to solve the technical problems and achieve the technical effects can be fully understood and implemented. It should be noted that, as long as there is no conflict, the embodiments and the features of the embodiments of the present invention may be combined with each other, and the technical solutions formed are within the scope of the present invention.

Additionally, the steps illustrated in the flow charts of the figures may be performed in a computer system such as a set of computer-executable instructions. Also, while a logical order is shown in the flow diagrams, in some cases, the steps shown or described may be performed in an order different than here.

The carbonate rock fracture-cave type oil-gas reservoir is the most main oil-gas reservoir type of Ordovician series in a Tarim basin, is one of the most main formations for increasing storage and increasing production at present in China, and is typical such as a medium petrochemical Tahe oil field, a northward oil field, a medium petroleum Ha Laha pond, an England-force oil field and the like. Years of exploration and development practice show that the large ancient karst cave is the most main reservoir body of the fracture-cavity type oil reservoir, almost occupies 80% of geological reserves, and the establishment of an accurate underground ancient karst cave distribution model has important guiding significance for evaluating the geological reserves and efficiently developing. With regard to the three-dimensional geological modeling related research of the ancient karst cave reservoir body, a great deal of work is carried out by scholars, and the overall research can be summarized into the following categories:

(1) And (3) quantitatively analyzing the large karst cave based on karst outcrop and establishing an outcrop model. First, taking the outcrop of the oridoite large-scale ancient karst cave in northwest edge of the Tarim basin as an example, the method integrates actual exploration and Google Earth image analysis, divides the large-scale ancient karst cave into 3 main types according to morphological structure characteristics, and divides the large-scale ancient karst cave into a single karst cave and a karst cave group according to the existing form, and provides the concept of the karst cave group. And moreover, the distribution rule and control factors of each type of karst cave are analyzed, the quantitative scale of each type of karst cave is counted, and an ancient karst cave development concept geological model integrating the characteristics and the control factors of the large ancient karst cave is established. And secondly, by utilizing the technical characteristics of three-dimensional laser scanning, a three-dimensional digital outcrop characterization technology based on laser scanning is provided, a three-dimensional laser scanner is utilized to acquire a three-dimensional model of a karst cave of modern karst to observe and analyze the karst phenomenon, the data acquisition, data processing and three-dimensional data model establishment of the Hunan Longwang cave are realized, and the acquired and processed model has high precision and good integrity and systematicness. And thirdly, selecting a Cyckel section of a Tarim basin as a key research object, and discussing development characteristics and evolution modes of the paleo-karst reservoir. The scheme indicates that the karst reservoir of the reef bank body of one room is small in karst pores and wide in karst cave distribution; the eagle mountain group mainly comprises marlite and sand-dust limestone, and is mainly a big-hole large-gap type with formed gaps as main control factors; the research area also found that hydrothermal action formed deep caverns, but both in small quantities and on a small scale. And fourthly, aiming at the problem that the heterogeneity of the carbonate reservoir is difficult to accurately predict by the conventional exploration and development technology, acquiring a field high-precision digital three-dimensional geological outcrop by using an outcrop digital modeling technology, establishing a three-dimensional geological model by using velocity data, and effectively improving the reservoir prediction precision by carrying out seismic forward simulation and seismic data comparative analysis on the three-dimensional velocity model.

(2) And (3) exploring and researching a multivariate and multiscale fracture-cavity type oil reservoir three-dimensional modeling method. Firstly, aiming at the characteristics of the carbonate fractured-vuggy reservoir such as large size, irregular shape, discontinuous distribution and the like, a three-dimensional geological modeling method of the carbonate fractured-vuggy reservoir is provided by taking a Tahe carbonate fractured-vuggy reservoir as a prototype, namely, a single type reservoir model is established firstly, then a homothetic condition assignment algorithm is adopted to fuse the single type reservoir model into a multi-scale discrete fractured-vuggy reservoir three-dimensional geological model, and the distribution characteristics of the fractured-vuggy reservoir in the three-dimensional space are quantitatively depicted. Secondly, aiming at the characteristic that the wave impedance attribute can accurately identify the large karst cave but cannot identify the small karst cave, the three-dimensional geological modeling method of the karst reservoir bodies of different types and different scales is researched by combining the karst development mode. Firstly, truncating different karst zones by adopting different wave impedance attribute values, and establishing a large karst cave model according to a karst mode man-machine interaction correction method; the method comprises the steps of taking large karst cave models of different karst zones as training images, adopting a multipoint geostatistics algorithm to establish a small karst cave model, enabling the established model to be matched with hard data, and reflecting the spatial distribution characteristics of karst cave reservoirs.

(3) And (5) researching the filling characteristics in the large karst cave and a logging interpretation technology. Firstly, taking a typical slotted hole unit of a tower river oil field as an example, calibrating conventional logging information by using a rock core and imaging logging information, preferably selecting cave sensitive parameters and crack sensitive parameters, establishing a normalized weighted identification function of the small slotted hole of the tower river oil field, identifying various cave filling material types by using a rendezvous chart, wherein the identification success rate is over 80%; meanwhile, more than 70% of the space of the karst fissure cavern is filled with substances such as underground river sediment sand mud, cave collapse gravel and the like, wherein the collapse gravel accounts for about 30%, and the karst fissure has better storage performance and oil-containing property when the collapse gravel has larger particles. Secondly, taking the karst outcrop in the Corkshel-Bachu area of the North of the Tower as an example, the filling and filling types of the karst cave are analyzed, 2 or 3 filling combinations can appear in the same karst cave to form a composite filling, and the cause mode of the karst cave is discussed, so that the filling identification and filling process research of the fracture-cave type reservoir of the oilfield of the Tower river is guided.

To sum up, the current state of the art in this field is:

(1) The scale relation of the seismic and geological data and the representation of the underground ancient karst cave needs to be further defined. A large number of single karst cave scale parameters are counted in the existing outcrop research work, and the height range of the outcrop section in the karst cave structure main parameter range is 1.5-6.5 meters and is far lower than the underground actual seismic resolution (30 meters) of the Tahe oil field. Therefore, the 'large karst cave' obtained based on the seismic data carving is often the concept of 'karst cave complex', that is, the reflection of a single string of beads on the seismic section is equivalent to the superposition of multiple layers of karst caves on the outcrop, and it is very difficult to obtain a single karst cave through the seismic data prediction. Therefore, the geological meaning represented by the seismic attributes needs to be clarified in the process of developing seismic karst cave predictions.

(2) The existing ancient karst cave modeling method mainly focuses on classification and scale modeling and fusion, namely a deterministic modeling method is mainly adopted for a large karst cave, sequential indication or multipoint statistics is mainly adopted for a karst cave (small karst cave) to carry out random simulation, and then fusion is carried out according to primary and secondary relations. The method has its utility from the perspective of production practice, while there are two problems to be improved from the perspective of method principles: firstly, a large karst cave model established by adopting an attribute truncation method based on earthquake prediction is not a true single karst cave, and the geological meaning of the large karst cave model needs to be further defined and the internal structure needs to be refined by a geological modeling method; secondly, different reservoirs 'violate' the principle of phase proportion conservation in statistical modeling in the process of fusion according to primary and secondary relations, and need to be further improved from the optimization of a modeling method.

(3) The existing research on filling characteristics inside the karst cave mostly focuses on outcrop and well logging explanation, and the established filling structure mode is mainly two-dimensional and lacks of expansion and application to a three-dimensional model. In addition, corresponding exploration is lacked in the aspect of a filling structure three-dimensional modeling method in the underground ancient karst cave reservoir body.

Therefore, in order to solve the technical problems, the application provides a three-dimensional geological model construction method for the carbonate rock ancient karst cave and the filling structure thereof. Starting from development characteristics and characterization data of the ancient karst cave, firstly, establishing a large-scale karst cave compound model based on seismic data according to a hierarchy constraint principle; secondly, establishing a single karst cave distribution three-dimensional model by adopting a multipoint geostatistics method on the basis of the karst cave compound model; and then, under the constraint of a single karst cave model, establishing a filling structure model in the karst cave by adopting a multipoint geostatistics method. Based on logging, earthquake and field outcrop data, the invention develops three-dimensional modeling method research of karst cave complex, single karst cave and filling structure in the karst cave from large scale to small scale, forms a set of ancient karst cave and filling structure multilevel modeling method under hierarchical constraint, further improves three-dimensional representation precision and accuracy of underground ancient karst cave reservoir, and serves for efficient development of the oil and gas reservoirs.

Fig. 1 is a step diagram of a three-dimensional geological model construction method for carbonate rock palsy and filling thereof according to an embodiment of the present application. Fig. 2 is a specific flow diagram of a three-dimensional geological model construction method for carbonate rock palsy and filling thereof according to an embodiment of the present application. A three-dimensional geological model construction method (hereinafter referred to as "three-dimensional geological model construction method") for carbonate rock palsy and filling thereof according to an embodiment of the present invention will be described below with reference to fig. 1 and 2.

Step S110, a karst cave complex model is established according to the seismic interpretation data of the area to be analyzed. After obtaining the data of rock core, well logging, earthquake, field outcrop and the like aiming at the current region to be analyzed, firstly determining the relation among the karst cave complex, the single karst cave and the filling structure in the karst cave in the current region, and judging the matching among various types of measured data, thereby further utilizing the step S110 to establish a karst cave complex three-dimensional model with larger scale by adopting a truncated Gaussian simulation method based on the earthquake prediction result of the current region to be analyzed. For example: based on typical karst outcrop research, a karst cave complex is usually composed of a plurality of (usually 3-8) single karst caves, according to the scale measurement result of the single karst caves, the width range is about 1.9-5.5 m, the height range is 1.5-6.5 m, therefore, the scale of the width and the height of the karst cave complex is about 15-45 m, and the karst cave complex has better matching performance with the identification precision of deep seismic data of a Tahe river, and therefore, the relation among the karst cave complex, the single karst caves and the filling structure in the karst caves in the current region is determined.

Further, (step S111, not shown) a seismic texture attribute data volume is extracted from the seismic interpretation data of the current region to be analyzed, and the seismic texture attribute data volume is converted into a seismic probability volume representing the contour of the karst cave complex.

The seismic attribute is a main technical means for predicting the underground ancient karst cave distribution, the seismic main frequency of a deep layer (> 6000 m) of a Tahe is about 25HZ, the longest seismic reflection of a single string bead is about 8ms, the reduced height is about 30m, and the scale of the karst cave complex is basically met, so that the prediction of the karst cave complex based on the earthquake has higher reliability.

As the cavern type reservoir body is characterized by a string of beads on the Seismic reflection, and most Seismic attributes have certain response, the Seismic Texture attribute (Seismic Texture) is introduced to predict the distribution probability of the cavern. The seismic texture attribute is mainly combined and strengthened with information similar to a spatial waveform structure in a waveform clustering mode, is sensitive to abnormal geologic body reflection, is more suitable for detection of karst cave reservoirs, and can indicate the possibility of karst cave development in a detection probability (value range 0-1) mode.

Therefore, in step S111, an earthquake texture attribute data volume is extracted after denoising through earthquake interpretation data, then the earthquake texture attribute data volume is converted into a gridding model, and an attribute value of each grid in the earthquake texture attribute data volume is converted into a corresponding probability value through a cluster analysis method to represent the probability that the current grid develops into the karst cave, so as to form an earthquake probability volume representing the contour of the karst cave complex, namely, a karst cave development earthquake probability volume, and the probability volume is used as an earthquake probability volume for constraining modeling of the karst cave complex. Fig. 3 is a schematic cross-sectional view of a seismic probability volume in the three-dimensional geological model construction method for carbonate rock palsy and filling thereof according to the embodiment of the present application. As shown in fig. 3, in the gridded seismic texture attribute data volume, each grid corresponds to a corresponding texture attribute value, the strength of the value represents the probability of karst cave development, and a higher texture attribute value indicates a higher probability of karst cave development.

Referring to fig. 2, after the karst cave development earthquake probability body is generated, (step S112, not shown), the actually measured karst cave complex development thickness after drilling is used as a threshold value, the field outcrop data and the logging and logging data are referred to, the current karst cave development earthquake probability body is used as a constraint, and a truncated gaussian simulation method is adopted to establish a distribution model of the karst cave complex body.

The truncated Gaussian simulation method is mainly used for discrete sedimentary facies (lithofacies) simulation and is characterized in that a simulation result can reflect the spatial change of a phase sequence. The simulation process is to cut off the three-dimensional continuous variable through a series of threshold values and threshold rules to establish the three-dimensional distribution of the type variable. In the embodiment of the invention, the obvious contrast relation between the karst cave development and the surrounding rock background is considered, so that a truncated Gaussian simulation method is introduced, and a three-dimensional model of the karst cave complex is established under the constraint of the karst cave development earthquake probability body.

Specifically, in the first step, the thickness data of the karst cave complex encountered by each well is obtained based on the single well interpretation result information obtained by the logging and logging data. And (4) counting the thickness corresponding to drilling and encountering karst cave complexes at different positions of each well in the current region to be analyzed by utilizing a statistical method based on the single-well interpretation result. And secondly, establishing a preset grid model, and coarsening the counted thickness data into the current preset grid model to form discrete karst cave complex data and a corresponding probability percentage curve aiming at the current karst cave complex. Establishing a three-dimensional grid model (a preset grid model) based on Petrel/RMS/Gocad software, reasonably dividing the sizes of a plane grid and a longitudinal grid, assigning interpretation data (thickness data required by a karst cave complex drilled at different positions of each well) of a single-well karst cave complex to corresponding positions in the established preset grid model, counting discrete karst cave complex data formed by the thickness data of each grid in the grid model after the thickness assignment on one hand, and counting the probability that each grid in a full grid model can be developed into the karst cave complex at the same time, thereby forming a karst cave complex probability percentage curve.

And thirdly, converting the discrete karst cave complex data obtained in the second step into pseudo-Gaussian data. And fourthly, fitting the variation function by using the scale data of the karst cave complex obtained by outcrop measurement. In the fourth step, firstly, the following karst cave complex scale data are obtained by using field outcrop measurement data of the current region to be analyzed, and the variation function required by truncated Gaussian simulation calculation is fitted by using the karst cave complex scale data. And fifthly, establishing a three-dimensional continuous variable Gaussian distribution field by using the karst cave development earthquake probability body as constraint and adopting a sequential Gaussian simulation method based on the pseudo Gaussian data obtained in the third step and by using a preset Gaussian theoretical model and the fitted variation function. And finally, taking the formed karst cave complex probability percentage curve as a threshold value curve, and performing Gaussian field truncation processing on the established three-dimensional continuous variable Gaussian distribution field to obtain a karst cave complex distribution model. Fig. 5 is a schematic diagram illustrating an effect of a karst cave complex distribution model in the three-dimensional geological model construction method for an ancient karst cave of carbonate rock and filling thereof according to the embodiment of the present application. FIG. 5 (a) is a two-dimensional section of a distribution model of a cavern complex; fig. 5 (b) is a three-dimensional (stereo) display diagram of the karst cave complex distribution model.

Thus, the embodiment of the present invention establishes a three-dimensional model of the karst cave complex with a larger scale by using a truncated gaussian simulation method based on the seismic prediction result, see fig. 5, and then proceeds to step S120.

As shown in fig. 1, a single karst cave geometric parameter (single karst cave geometric dimension data) is obtained based on the field outcrop data of the region to be analyzed, and the single karst cave geometric parameter and the karst cave complex model constructed in step S110 are respectively used as an input condition and a constraint to establish a single karst cave internal filling distribution model. The karst cave complex is composed of a plurality of single karst caves of the same type, the built karst cave complex model is used as constraint according to the hierarchical constraint principle, the single karst cave geometric parameters obtained through outcrop description are used as input conditions, and a target-based indicative point process simulation method is adopted to build a filling distribution model in the single karst caves.

Specifically, (step S121, not shown) a plurality of single caverns inside the cavern complex are measured in the field according to the field outcrop data of the current region to be analyzed, and the geometric dimension (statistical) data of each single cavern is counted. The geometric (statistical) data includes, but is not limited to, height, width, extension length and cross-sectional morphology of a single cavity.

Fig. 4 is a schematic diagram of a measurement principle of a single karst cave geometric parameter in a three-dimensional geological model construction method for an ancient carbonate karst cave and filling thereof according to an embodiment of the present application. Referring to fig. 4, based on field outcrop measurement data for the current area to be analyzed and previous research knowledge, a plurality of single caverns inside the current cavern complex are measured in the field, and single cavern geometric dimension data including data of height, width, possible extension length, and cross-sectional morphological features (including but not limited to oval, bean pod, etc.) of the single cavern are obtained, see table 1. Table 1 shows the geometric (statistical) data of each single cavern contained in the current cavern complex obtained by field outcrop measurement.

TABLE 1

The variation range of the height (1.6-7.7 m), the variation range of the width (1.5-5.1 m) and the variation range of the extension length (12-18 m) of a single karst cave are respectively summarized from the statistical results (table 1), so that the ancient karst cave scale parameters obtained currently are further used as important input parameters of the following single karst cave simulation.

Referring to fig. 2, after obtaining the geometric size data and the ancient cave scale parameters of each single cave in the current karst cave complex, (step S122, not shown), the above-mentioned established karst cave complex distribution model is used as a frame constraint, and the geometric size statistical data and the ancient cave scale parameters (karst cave complex scale data) of the single cave are used as input conditions, and the logging interpretation data is combined to perform simulation calculation on each single cave in the current karst cave complex distribution model, so as to obtain a single cave internal filling distribution model for the current karst cave complex.

In step S122, first, in step S, the central point position of a single karst cave is predicted from the karst cave complex distribution model established in step S110. Particularly, in a karst cave complex distribution model, a random sampling method is adopted to predict a corresponding central point for each single karst cave. And secondly, carrying out consistency matching on the karst cave complex distribution model and the karst cave interpretation data of the drilled well in the current region to be analyzed, and entering the next step after matching. In the second step, the central point position of each single karst cave is used as a positioning basis to judge whether the currently generated karst cave complex distribution model conflicts with the karst cave interpretation data of each drilled well position obtained by logging and seismic interpretation data (namely whether the currently simulated karst cave thickness is consistent with the drilled well thickness is mainly judged), if so, the input karst cave scale parameter range is continuously adjusted, the simulation is continued until the karst cave scale parameter range does not conflict, and if not, the next step is carried out.

And step three, generating an empirical cumulative probability distribution function related to the size characteristics of different positions in the current single karst cave according to the obtained geometric size statistical data of the single karst cave, the ancient karst cave scale parameters and the drilling and logging information, extracting the height of the current single karst cave from the empirical cumulative probability distribution function, and assigning the height values to corresponding positions of a karst cave complex distribution model for completing geological conflict judgment. In the third step, size characteristic data of different positions (depth position and longitudinal position) in each single karst cave in the current karst cave complex are generated according to well drilling and logging data, ancient karst cave scale parameters and single karst cave geometric dimension statistical data obtained through statistics from field outcrop measurement data, so that a corresponding experience cumulative probability distribution function is formed, then the height of the single karst cave is randomly extracted from the current experience cumulative probability distribution function, and the height value is assigned to the corresponding grid position of the karst cave complex distribution grid model for completing geological conflict judgment.

And fourthly, calculating the actual width of the karst cave according to the obtained single karst cave geometric dimension statistical data, and assigning the width of the karst cave complex model. In the fourth step, the mapping relationship between the height and the width of the single karst cave, which is obtained according to the geometric dimension statistical data of the single karst cave, is used to calculate the actual width of each single karst cave based on the randomly extracted height data of each single karst cave obtained in the third step, and the actual width data is assigned to the corresponding grid position of the karst cave complex distribution grid model with the completed height amplitude.

And finally, setting a target function and a threshold value required for constructing a single karst cave model according to the obtained height and actual width data of the single karst cave and the statistical data of the geometric dimension of the single karst cave, and performing simulation calculation on the karst cave compound model subjected to height and width assignment to form a filling distribution model in the single karst cave. In the fifth step, an objective function (the objective function refers to a spatial position function where a single karst cave may be placed) and a threshold (the threshold includes a threshold of the number of single karst caves and/or a threshold of volume ratios between different single karst caves) required for constructing a single karst cave model need to be set, and simulation calculation is performed on the karst cave composite model with the assigned height and width until the number and/or volume ratio of the single karst caves reaches a given threshold, so as to finish the simulation calculation, thereby obtaining a three-dimensional simulation model of the distribution of the single karst caves.

Fig. 6 is a schematic cross-sectional effect diagram of a single karst cave internal filling distribution model in the three-dimensional geological model construction method for the carbonate rock ancient karst cave and filling thereof according to the embodiment of the present application.

Thus, the embodiment of the invention generates the central point space distribution of the objects according to the distribution rule of the geometric objects in the space based on the internal dimension characteristics and scale characteristics of the single karst cave and the ancient karst cave scale data obtained by field outcrop measurement, well drilling and well logging information, and then marks the physical properties (such as geometric shape, size, direction and the like) on each discrete point. Therefore, each discrete grid point in the current karst cave complex space region is subjected to value marking, single karst cave modeling inside the karst cave complex is fully carried out, and each discrete point in the modeling process is equivalent to each single karst cave. Therefore, the invention further establishes a single karst cave three-dimensional model based on the karst cave field outcrop measurement data under the restriction of the karst cave complex, referring to fig. 5, and then the step S130 is proceeded to.

Step S130 is based on the field outcrop data and the well drilling and well logging interpretation data of the current region to be analyzed, a training model for representing the filling state in the single karst cave is obtained, and then according to the current training model, multipoint geology statistics is carried out on the single karst cave internal filling distribution model constructed in the step S120, and a corresponding simulation model is obtained. The storage performance of a karst cave reservoir stratum is determined by filling characteristics inside the karst cave, and the karst cave reservoir stratum is mainly divided into sand-shale filling, collapse gravel filling, calcite cemented filling and unfilled filling according to the filling material properties. The sandstone pores, the holes among the cobbles and the unfilled cave part form a main storage space of the karst cave storage body, and the proportion and the distribution of the filling structures of all the parts are more critical for evaluating the properties of the ancient karst cave storage. As the multipoint geostatistics has stronger advantages for representing the contact relation between different lithologies, the embodiment of the invention introduces the multipoint geostatistics method to simulate the filling structure characteristics in the karst cave, and the main contents comprise two parts of training image acquisition and multipoint geostatistics simulation.

First, (step S131, not shown) a karst cave filling longitudinal sequence is obtained based on the well drilling interpretation data, and the karst cave filling longitudinal sequence is modeled to obtain a training model of the internal longitudinal filling sequence of the single karst cave. Based on the observation and recognition of well drilling and outcrop, the karst cave filling has certain regularity, the bottom is mostly collapsed gravel filling, the upper part is filled with calcite formed under the cementing action, the upper part is filled with sand shale formed under the hydrodynamic action, and the upper part is mostly filled with residual unfilled karst cave space. In step S131, a longitudinal sequence of the current karst cave complex filling is obtained from the well drilling and logging interpretation data and recorded as a filling longitudinal sequence; then, modeling is carried out on the current karst cave filling longitudinal sequence by adopting a deterministic modeling method, and a (ideal) model of the longitudinal filling sequence in the single karst cave is manufactured, so that a corresponding training model is formed, and each lithologic contact surface shows certain spatial variation as much as possible on the premise of ensuring that the filling sequence is not changed, and reference is made to fig. 7. Fig. 7 is a schematic diagram illustrating the obtaining principle of a training model in the method for constructing the three-dimensional geological model for the carbonate rock palsy and filling thereof according to the embodiment of the present application. FIG. 7 (a) is a three-dimensional distribution diagram of a single cave filling sequence model; FIG. 7 (b) is a grid diagram of a single cave filling sequence model; fig. 7 (c) and 7 (d) are two-dimensional cross sections of a single cave-filling sequence grid pattern in different directions, respectively. Because the drilling and outcrop measurement can only actually observe one-dimensional and two-dimensional karst cave profiles and the three-dimensional karst cave filling structure is difficult to detect, the embodiment of the invention uses the constructed training model as the three-dimensional training image of the single karst cave filling structure to carry out further structural simulation on the internal filling structure of each single karst cave.

After the construction of the three-dimensional training model is completed, (step S132, not shown), a simulation calculation based on statistics is performed on the filling distribution model inside the single karst cave according to the current training model. First, (step S1321, not shown) a training model is scanned based on measured well data to create a search tree. Specifically, firstly, obtaining actually measured well data from well logging and well drilling interpretation data, and correspondingly marking the well data at corresponding positions of the filling distribution model in the single karst cave (namely, correspondingly marking the actually measured well data on the nearest grid node in the filling distribution model in the single karst cave); and then, scanning the established three-dimensional training model according to a preset three-dimensional data template, and establishing a corresponding search tree so as to obtain a training mode library. It should be noted that the search tree is a template for searching the filling structure of the single karst cave internal filling distribution model, the template covers the geological features of the internal filling structure of each single karst cave in the karst cave complex to be analyzed, which are obtained based on actual measurement, and the richer the template is, the stronger the capability of training the geological features of the model on line is.

Then, after the search tree building is completed, (step S1322, not shown) a plurality of (different) random lines are further built, and search tree-based search is performed on the filling distribution model inside the single karst cave for each random line, so that a corresponding simulation model is generated for each random line to reproduce the actual geological features (i.e., filling structures) inside each single karst cave by using the plurality of simulation models.

Since the search flow for each random line is the same, the embodiment of the present invention is described by taking the search flow of one random line as an example. Specifically, (1) determining a random path corresponding to a single karst cave internal filling distribution model which currently completes the data annotation of an actually measured well (wherein the random path can access each point to be simulated in the path), retrieving the point to be simulated through a search tree at a first point to be simulated in the random path of the current grid model, and solving a conditional probability distribution function corresponding to the current point to be simulated; (2) extracting a value from the conditional probability distribution of the current point to be simulated (wherein the current value extracted by, for example, a Monte Carlo random sampling method represents the possibility of a certain filling lithology, the value range is 0-1, and the larger the value is, the higher the possibility of the certain lithology is shown) as a random simulation value of the current point to be simulated, thereby adding the simulation value into a condition data set as condition data extracted by a simulation value of the next point to be simulated; (3) and (3) judging whether the current point to be simulated is the end point of the current random line or not according to the current random path, if not, carrying out random simulation on the next node to be simulated, repeating the steps (1), (2) and (3), and if so, ending the simulation aiming at the current random line so as to generate a corresponding random simulation implementation, namely a simulation model. Therefore, after the search simulation of all random lines is completed, a plurality of simulation models are obtained. Wherein, the simulation model is quantitatively marked with filling structure data of each single karst cave in the current karst cave complex.

Fig. 8 is an exemplary diagram of a cross section of a simulation model in the three-dimensional geological model construction method for carbonate rock ancient cavern and filling thereof according to the embodiment of the application. As can be seen from fig. 8, the embodiment of the present invention obtains the internal filling geological types of each single cavern of the simulation model and the distribution characteristics of each type of filling types in the single cavern. Thus, the embodiment of the invention further improves the representation precision of the karst cave reservoir stratum by using the step S130, and proposes to form a karst cave filling structure model by using a multi-point geostatistics method based on the knowledge of the karst cave filling structure, thereby entering the step S140.

Step S140 performs seismic forward modeling on the plurality of simulation models obtained in step S130 to obtain corresponding seismic data volumes for calculation, and generates an optimal geological model representing the karst cave internal structure by combining with actual seismic interpretation data, so as to represent the contour of the karst cave complex, the single karst cave distribution, and the filling structure characteristics inside the single karst cave in the current region to be analyzed by using the optimal geological model.

Specifically, in step S140, the plurality of simulation models obtained in step S130 are respectively compared with the actual seismic interpretation data, and an optimal simulation model is preferably selected from the plurality of simulation models, so that the optimal simulation model is used as a model with the highest goodness of fit to represent an optimal large-scale karst cave internal filling structure model for the current karst cave complex, and is used as an optimal three-dimensional geological model for the current carbonate rock paleo-cavern (to-be-analyzed area) and the internal filling structure thereof.

Further, step S140 performs a visual-based structural comparison between the simulation models obtained in step S130 and the actual seismic interpretation data, and calculates an error between each simulation model and the seismic interpretation data, so as to preferably select an optimal geological model according to the error data corresponding to each simulation model. And taking the simulation model corresponding to the minimum error data as the optimal geological model.

In order to improve the comparison accuracy, before comparing the plurality of simulation models with the actual seismic interpretation data, in step S140, the embodiment of the present invention performs a corresponding seismic reflector conversion process on each simulation model, so as to perform a visual-based structural comparison between each seismic reflector and the actual seismic interpretation data. Specifically, performing petrophysical analysis on each simulation model, and establishing a corresponding initial model for each simulation model; then, each initial model is converted into a corresponding seismic data volume (seismic reflector) for calculation through a seismic convolution forward simulation method or a wave equation forward simulation method; and finally, performing visual-based structural comparison on each seismic data volume for calculation and actual seismic interpretation data respectively, and calculating the error between each seismic data volume for calculation and the seismic interpretation data respectively, so as to preferably select the optimal geological model according to the error data corresponding to each seismic data volume for calculation.

Therefore, after a large-scale karst cave internal filling structure model is established, the calculated seismic reflection data volume is compared with actual seismic data through convolution forward modeling or wave equation forward modeling, and the model with the highest goodness of fit is preferably selected as a final simulation result.

Taking the tahe oil field T unit as an example, the flow of establishing the three-dimensional geological model for the underground ancient cavern of the tahe oil field T unit by developing the above three-dimensional geological model construction method will be specifically described.

The ancient karst cave reservoir of carbonate rock, which is typical for development of Ordovician building groups and eagle mountain building groups in Tahe oil fields, is particularly affected by non-integration of the top of the building group, large karsts are developed along the vicinity of the interface of the building group, and through years of development and practice, a large number of well drilling prove the existence and internal structural characteristics of the karsts. By utilizing the three-dimensional geological model construction method, the data such as rock cores, well logging, earthquakes, field outcrops and the like are integrated, the relation among the karst cave complex, the single karst cave and the filling structure in the karst cave is firstly determined, and then the three-dimensional geological modeling research is developed by adopting step-by-step constrained train of thought and stratification based on the data precision.

Firstly, establishing a three-dimensional model of the karst cave complex. Based on the analysis of seismic data, the seismic structure (Texture) attribute reflecting the contour of the karst cave complex is extracted, the strength of the attribute value represents the karst cave development probability, and the higher the value is, the higher the possibility of developing the karst cave is (figure 3). Based on a truncation Gaussian method, a drilled actual measurement karst cave development proportion is used as a truncation threshold value, an outcrop measurement assignment variation function is referred, and a karst cave complex three-dimensional model (figure 5) is established under the attribute cooperative constraint of a seismic structure (Texture).

And secondly, establishing a single karst cave model. The single karst cave internal filling distribution model of the T unit is built by taking the built karst cave compound model as frame constraint, taking the ancient karst cave scale parameters of field outcrop statistics as input conditions and adopting a target-based indicative point process simulation method (figure 6).

And step three, establishing a filling structure model in the karst cave. Based on outcrop description and well drilling interpretation, firstly establishing an ideal structure model filled in a single karst cave by a deterministic modeling method, and taking the ideal structure model as a three-dimensional training image (figure 7); the training image is scanned through the multi-point data sample plate, a search tree is constructed, and a plurality of simulation realizations are generated by adopting a Snesim multi-point geostatistical modeling algorithm.

And fourthly, all the geological modeling methods adopted in the three steps are random modeling categories, the realization of a plurality of groups of geological models can be completed through parameter adjustment, and the preliminary selection of the models is carried out through comparison with the drilling data and the geological knowledge.

And fifthly, for realizing a large number of in-hole structure models established by multipoint geostatistics, a group of models with the best goodness of fit is preferably selected as a final simulation result (figure 8) of the T unit by a method of earthquake forward modeling and actual earthquake data comparison, and the model has better rule consistency with the existing geological model from the modeling result.

According to the steps, all links of three-dimensional fine modeling of the ancient karst cave are completed, and a set of complete technical process and method are formed.

The invention discloses a construction method of a three-dimensional geological model for carbonate rock ancient karst caves and filling of the carbonate rock ancient karst caves. The method considers the scale of the karst cave scale and the precision of the representation data, and proposes to establish a three-dimensional model of the karst cave complex by adopting a truncated Gaussian method under seismic inversion constraint; under the restriction of the karst cave complex, further based on karst cave field outcrop measurement data, a single karst cave three-dimensional model is established by adopting a target-based indicative point process method; in order to further improve the representation precision of the karst cave reservoir stratum, based on the knowledge of the karst cave filling structure, a multi-point geostatistics method is adopted to establish a karst cave filling structure model, and a model with the highest actual goodness of fit is optimized through an earthquake forward modeling method to realize condition modeling. According to the comprehensive modeling method, three-dimensional modeling method research of a karst cave complex, a single karst cave and a filling structure inside the karst cave is carried out from large scale to small scale through a three-step walking strategy, a set of hierarchical constraint ancient karst cave and filling structure multi-level modeling method is formed, three-dimensional representation precision and accuracy of an underground ancient karst cave reservoir are improved, earthquake, geology, field outcrop, well logging and well drilling data are utilized to the maximum extent, and modeling method optimization of different representation scales is realized. The comprehensive karst cave modeling technology provided by the invention has strong applicability to the fine description of the tower river fracture-cave type reservoir bodies.

The above description is only a preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

It is to be understood that the disclosed embodiments of the invention are not limited to the particular structures, process steps, or materials disclosed herein but are extended to equivalents thereof as would be understood by those ordinarily skilled in the relevant arts. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to be limiting.

Reference in the specification to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the invention. Thus, the appearances of the phrase "one embodiment" or "an embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment.

Although the embodiments of the present invention have been described above, the above description is only for the convenience of understanding the present invention, and is not intended to limit the present invention. It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims (10)

1. A three-dimensional geological model construction method for carbonate rock ancient karst caves and filling thereof comprises the following steps:

establishing a karst cave compound model according to the seismic interpretation data of the area to be analyzed;

obtaining a single karst cave geometric parameter based on field outcrop data, and establishing a single karst cave internal filling distribution model by respectively using the single karst cave geometric parameter and the karst cave compound model as input conditions and constraints;

obtaining a training model representing the filling state in the single karst cave based on the field outcrop data and the well drilling and logging data, and carrying out multipoint geological statistics on the filling distribution model in the single karst cave based on the training model to obtain a corresponding simulation model;

and carrying out earthquake forward modeling on the simulation model to obtain a seismic data volume for calculation, and generating an optimal geological model for representing the structure in the karst cave by combining actual earthquake interpretation data.

2. The method for constructing the three-dimensional geological model according to the claim 1, which comprises the following steps in the process of constructing the karst cave compound model:

extracting a seismic texture attribute data volume from the seismic interpretation data, and converting the seismic texture attribute data volume into a seismic probability volume representing the contour of the karst cave complex;

and taking the actually measured development thickness of the karst cave complex drilled as a threshold value, referring to field outcrop data, and taking the earthquake probability body as a constraint to establish a distribution model of the karst cave complex.

3. The method for constructing a three-dimensional geological model according to claim 1 or 2, wherein in the process of constructing the filling distribution model in the single karst cave, the method comprises the following steps:

according to the field outcrop data, carrying out field measurement on a plurality of single karst caves in the karst cave complex, and counting the height, width, extension length and section morphological characteristics of each single karst cave;

and taking the karst cave complex model as a frame constraint, taking the geometric dimension statistical data of the single karst cave as an input condition, and combining logging interpretation data to perform simulation calculation on each single karst cave in the karst cave complex model to obtain a filling distribution model in the single karst cave.

4. The method for constructing a three-dimensional geological model according to any of the claims from 1 to 3, characterized in that in the process of generating said training model, it comprises:

and obtaining a karst cave filling longitudinal sequence based on well drilling interpretation data, and modeling the karst cave filling longitudinal sequence to obtain the training model of the internal longitudinal filling sequence of the single karst cave.

5. The three-dimensional geological model construction method according to claim 1 or 4,

scanning the training model according to the actually measured well data, and establishing a search tree;

and establishing a plurality of random lines, and searching the filling distribution model in the single karst cave based on a search tree aiming at each random line so as to generate the corresponding simulation model.

6. The method of constructing a three-dimensional geological model according to claim 5,

and carrying out visual-based structural comparison on the plurality of simulation models and actual seismic interpretation data respectively, and calculating the error between each simulation model and the actual seismic interpretation data respectively, so as to preferably select the optimal geological model according to the error.

7. The method of constructing a three-dimensional geological model according to claim 5 or 6,

performing rock physical analysis on the simulation model, and establishing a corresponding initial model;

and converting the initial model into the seismic data volume for calculation by a seismic convolution or wave equation forward modeling method.

8. The method of constructing a three-dimensional geological model according to claim 2,

acquiring thickness data of karst cave complex encountered by drilling of each well based on single well interpretation result information;

establishing a preset grid model, and coarsening the thickness data into the preset grid model to form discrete karst cave complex data and a corresponding probability percentage curve aiming at the current karst cave complex;

converting the discrete karst cave complex data into pseudo-Gaussian data;

fitting the variation function by using the scale data of the karst cave complex obtained by outcrop measurement;

based on the pseudo-Gaussian data, a three-dimensional continuous variable Gaussian distribution field is obtained by using a preset Gaussian theoretical model and a fitted variation function, taking the seismic probability body as a constraint and adopting a sequential Gaussian simulation method;

and performing Gaussian field truncation processing on the three-dimensional continuous variable Gaussian distribution field by taking the probability percentage curve as a threshold value curve to obtain the karst cave complex distribution model containing the discrete variable value information of each grid point.

9. The method of constructing a three-dimensional geological model according to claim 3,

predicting the central point position of a single karst cave from the karst cave compound model;

carrying out consistency matching on the karst cave complex model and karst cave interpretation data of a drilled well, and entering the next step after matching;

generating an empirical cumulative probability distribution function about the size characteristics of different positions in the current single karst cave according to the geometric dimension statistical data and the drilling and logging information, extracting the height of the current single karst cave from the empirical cumulative probability distribution function, and then assigning the height to the corresponding position of the karst cave complex model;

calculating the actual width of the karst cave through the geometric dimension statistical data, and carrying out width assignment on the karst cave compound model;

and setting a target function and a threshold value according to the height and the actual width of the single karst cave and the statistical data of the geometric dimension, and performing simulation calculation on the karst cave compound model subjected to height and width assignment to form the filling distribution model in the single karst cave.

10. The method of constructing a three-dimensional geological model according to claim 5,

correspondingly marking the measured well data at the corresponding position of the filling distribution model in the single karst cave;

and scanning the training model according to a preset data template to obtain the search tree.

Images