KR20200027264A - A method for selecting a sweet spot in shale gas - Google Patents

Abstract

The present invention relates to a method for predicting a sweet spot in shale gas, and more specifically, to a method for predicting a sweet spot in shale gas by constructing an integrated model from a stratigraphic structure model, a rock model, a physical property model, and a natural crack model.

Description

A method for selecting a sweet spot in shale gas}

The present invention relates to a method for predicting a shale gas sweet spot, and more specifically, to a method for predicting a sweet spot of a shale gas by constructing an integrated model from a stratigraphic structure model, a rock model, a physical property model, and a natural crack model. will be.

In general, the shale gas reservoir has a very low permeability, which requires artificial recovery enhancement techniques such as horizontal well drilling and hydraulic fracturing for commercial gas production and requires significant development costs. The economic feasibility of shale gas development can be improved by predicting production variation by region or geological layer caused by total organic carbon content, brittleness, anisotropy, etc. of the shale layer, that is, through sweet spot analysis.

The sweet spot of shale gas is a place having high gas content and a prospect of hydraulic crushing, and refers to an area or location where the maximum yield or potential of production in the play or reservoir is high. Before the Sweet Spot concept was applied to the development of shale blocks, it was mostly a batch development method to develop the entire block, but after discovering that productivity of non-traditional resources is due to the geological, geochemical, and petrochemical properties of the reservoir, shale blocks were developed. There are more and more cases where the concept of sweet spot is applied.

In the sweet spot prediction technology of the shale reservoir, the probabilities of the gas content (reservoir quality) and crushing efficiency (completion quality) suitable for the mine are analyzed using core analysis, borehole and seismic data. Gas content is defined by hydrocarbon properties from a petrophysical point of view indicating the prospect of the reservoir, and includes porosity, gas saturation, permeability, total organic carbon content (TOC), thickness, and thermal maturity. The crushing efficiency is defined as petrochemical properties that mean an environment favorable for the creation, propagation, suppression of hydraulic crushing, and the propant to settle in cracks and maintain the crack open state. Brittleness, closed stress, natural crack, stress field, void pressure, Young's modulus, Poisson's ratio, incompressibility, and stiffness.

The resource development project has high uncertainty in the development of the project due to various factors, such as procurement of developer funds, on-site infrastructure and manpower, and oil and gas prices. Therefore, in order to proceed smoothly, it is required to secure economic feasibility. In general, the economics of a business depends on the productivity of hydrocarbons, and in order to secure the business feasibility of shale gas mine development, exploration technology is needed to predict the region where shale gas can be produced the most.

In the method for deriving the sweet spot of the shale gas reservoir according to the present invention:

Building individual models (S100);

Predicting a sweet spot by constructing an integrated model from the individual models (S200);

Verifying the predicted sweet spot (S300); And

And finally determining a sweet spot of the shale reservoir based on the verification result (S400).

According to the present invention, it is possible to predict a region (sweet spot) in which shale gas can be maximally produced, and has the advantage of being able to perform exploration to secure business feasibility of shale gas mine development.

1 is a schematic flowchart of a method for deriving a sweet spot of a shale gas reservoir according to the present invention.
2 is an illustration of a stratified structural model according to the present invention.
3 and 4 are examples of gas content models according to the present invention.
5 to 7 are examples of the crushing efficiency model according to the present invention.
8 is an example of a physical property model according to the present invention.
9 and 10 are examples of natural crack models according to the present invention.
11 exemplarily shows an integrated model constructed by integrating a natural crack model into a physical property model according to the present invention.
12 is an example of verification of a brittleness model, which is an element of crushing efficiency, among the integrated models according to the present invention.

Hereinafter, a method for predicting sweet spots in a shale reservoir according to an embodiment of the present invention will be described in detail. The accompanying drawings show exemplary forms of the present invention, which are provided to explain the present invention in more detail, and the technical scope of the present invention is not limited thereto.

In addition, irrespective of the reference numerals, the same or corresponding components will be given the same reference numerals, and redundant description thereof will be omitted, and the size and shape of each illustrated member may be exaggerated or reduced for convenience of description. have.

According to the present invention, the method for predicting sweet spots in the shale reservoir is largely based on an individual model building step (S100), an integrated model building from individual models to predict the sweet spot (S200), and a step of verifying the predicted sweet spot (S300). ), And finally determining the sweet spot of the shale reservoir based on the verification result (S400).

The individual model building step (S100) includes a stratified structure model building step (S110), a rock image model building step (S120), a physical property model building step (S130), and a natural crack model building step (S140).

First, a stratified structure model construction step (S110) is performed. The stratified structure model construction step (S110) is a step for distinguishing how the stratified structure model was deposited in the past and how its perceptual characteristics change. In other words, it is a step to identify the boundary of the strata through periodic division. To do this, there are two major ways to acquire data in the field. One is a borehole analysis method that analyzes the results obtained by drilling a borehole wall by drilling through a borehole as a direct exploration method. The other is an indirect sensing method, which is an elastic wave sensing method that transmits a signal from the surface to the basement and receives and measures a wave reflected from the layer.

In summary, the times are classified based on the data obtained by boreholes at a specific location, and since the data obtained by boreholes contain only information about a specific location, the data obtained through seismic exploration are connected to the structural structure of the strata. Model the structure of the strata by grasping the frame.

At this time, since the physical properties are not the same across the entire strata, for example, over 100 km ² , it is possible to enter the physical properties for each lattice by performing lattice on the strata obtained in the stratigraphic structure model.

In the stratified structure model construction step (S110), for example, a stratified structural model can be built using Petrel (Schlumberger Co.), which is geological modeling software, and in FIG. 2, the stratified structure is constructed in the stratified structure model construction step (S110). Illustratively. “Top”, “badheart”,… “Top_Upper_montney”, “top_lower_montney”,… The parts shown as “Base” represent each strata separately. In addition, as described above, grids are generated for each of these layers, and property information is input to the grids. Examples of the input property information include brittleness (BI), closed stress (CSS), density (P-impedence), and S-impedence (S-impedence) information.

Next, a phase model building step (S120) is performed. For the stratigraphic structural model obtained in step S110, spatial rock information is predicted using seismic waves and borehole rock information. Spatial rock information includes, for example, high content more siliceous, siliceous, argillaceous, carbonaceous, and mixed.

Various geological properties can be inferred from seismic data. In this way, the information obtained from the seismic data and the borehole data are combined and modeled to make the rock distribution visible in three dimensions. To this end, it is necessary to calculate the impedance, elasticity information (Young's modulus, Poisson's ratio, incompressibility, stiffness, etc.) calculated from the borehole physics layer, rock information, and three-dimensional impedance and elasticity information derived from seismic inversion. The analysis procedure shows the rock information in the crossplot of the borehole impedance and elastic information, analyzes the distribution of each rock, and selects the probability density function of each rock. Then, the 3D rock distribution can be obtained by linking the impedance and elastic information derived from the seismic inversion to the probability distribution of each rock calculated in the borehole.

Next, a physical property model construction step (S130) is performed. The physical property model construction step (S130) largely includes a step of building a gas content (Reservoir Quality: RQ) model (S131) and a step of building a completion quality (CQ) model (S132). The step of constructing the gas content model (S131) and the step of constructing the crushing efficiency model (S132) may be performed either of them, or may be performed simultaneously.

First, the step of building a physical property model (S130) includes a step of building a gas content model (S131). In the step of constructing the gas content model (S131), the four main factors related to gas content, total organic carbon (TOC), gas saturation rate, porosity, and water permeability, are divided into the main reservoirs as follows: After building the model, we integrate the four elements. At this time, the properties of the seismic wave used for RQ factor prediction include amplitude, frequency, frequency filter, phase, geological density, MuRho, Young's modulus, crack density, Vp / Vs Ratio, and porosity (used when predicting permeability).

First, as an individual element, the step of constructing a total organic carbon (TOC) model (S131-1) is as follows. Since the shale gas mine has the same source and reservoir rocks as the traditional oil and gas field, the total organic carbon content has a high correlation with the gas content of the reservoir, and core data is used to determine the amount of organic hydrocarbons in the reservoir. Should be analyzed. There are three main methods for the prediction of total organic carbon content (TOC). In the first method, the total organic carbon content is calculated using the correlation between the core total organic carbon content measured in the core sample experiment and the resistivity sample layer. In the second method, the total organic carbon content is calculated using the correlation between the total organic carbon content measured in the core sample experiment and the neutron-density sample layer difference. In the third way, Passey et al . (1990)? Log ?, which converts the difference between the acoustic and non-resistive samples proposed in terms of total organic carbon content? As a method, the total organic carbon content of the borehole is calculated using the resistivity log, sonic log, and level of maturity (LOM) parameters. This can be verified with the total organic carbon content measured at the core, and the prediction error of the total organic carbon content can be minimized by applying the artificial neural network technique that performs RQ seismic properties and learning by spatial visualization.

Next, as an individual element, the step of constructing a gas saturation model (S131-2) is as follows. After calculating the gas saturation rate for each borehole using the Waxman-smits equation (Waxman and Smits, 1968) and verifying it with the core data, an artificial neural network technique that performs learning and RQ seismic properties that minimize the prediction error of the gas saturation rate By applying, the gas saturation rate is visualized spatially. Since the gas saturation rate is inversely related to the water saturation rate, regions with low water saturation rates show high gas saturation rates and gas content.

Next, as an individual element, the step of constructing a porosity model (S131-3) is as follows. After verifying the porosity calculated from the bulk density with the porosity measured at the core, the prediction of the porosity is minimized and spatially visualized through learning of RQ seismic properties and multi-property linear regression technique. Since a place having a high porosity can provide a space for containing a lot of gas, a place having a high porosity appears to have a high gas content.

Next, as an individual element, the step of constructing a permeability model (S131-4) is as follows. The relationship between the porosity and permeability of the core was derived, and the permeability was predicted by applying it to another borehole without a core. The spatial visualization of the permeability is based on RQ seismic properties and the artificial neural network technique that performs learning based on the permeability calculated at each borehole and minimizes the predicted permeability. The permeability derived from modeling tended to be very similar to the porosity. This phenomenon is similar to the linear relationship between porosity and permeability derived from the core experiment, i.e., the characteristics of good permeability in areas with good porosity.

Next, the individual elements as described above for analysis to select a place with high gas productivity for each of the four individual elements (that is, total organic carbon (TOC, Total Organic Carbon), gas saturation rate, porosity and permeability) After constructing, in order to determine which individual element has the highest correlation with gas productivity, the step of constructing a gas content model by integrating the four elements obtained in steps S131-1 to S131-4 (S131-5) ). First, the four elements are normalized for each element.

First of all, for comparison with the results of weighting individual elements, the gas content model can be constructed by assigning the same 25% weight to 4 elements (see FIG. 3).

Alternatively, after constructing an integrated model with different weights, different weights may be assigned and analyzed for each of the four factors in order to determine whether there is consistency with production data. For example, unlike traditional gas fields, shale gas can be assigned a weight of 40% to the TOC and 20% to the other 3 elements because it exists in the reservoir after gas is generated from the source rock. Yes (see Figure 4).

Second, the step of building a physical property model (S130) includes a step of building a crushing efficiency model (S132). First, in order to develop a shale gas mine and produce gas, hydraulic crushing must be performed to expand the gas flow path of the reservoir, unlike traditional gas field mines (Kim et al., 2014). In order to grasp, it is necessary to find a section that satisfies the conditions that the propagation range of crushing is wide when water pressure is applied to the reservoir and the flow path secured after crushing can be maintained for a long time. To this end, in the step of constructing a CQ model (S132), three factors among the characteristics of rock mechanics associated with the hydraulic crushing of the reservoir, such as brittleness, natural cracking zone density, and pore pressure, are used for the corresponding photosphere analysis. 3D modeling is performed using multi-attributive linear regression and stochastic artificial neural network techniques based on core, physical layer, and seismic properties. The three factors related to hydraulic fracturing are classified into major reservoirs as follows, and the model is first constructed individually in the same way as the RQ (Reservoir Quality), and then the CQ integrated model is constructed. Seismic properties used for predicting CQ factors include amplitude, frequency, frequency filter, differential operator, phase, geological density, Poisson's ratio, P-impedance, crack density, and permeability (used when predicting void pressure). .

First, as an individual element, a step (S132-1) of building a brittleness index model is as follows. Brittleness modeling is performed by deriving the Poisson's ratio and Young's modulus using p-sonic and transverse (s-sonic) acoustic samples and combining the two factors. Spatial visualization was achieved by applying a multi-attributive linear regression technique that minimizes the predictability of brittleness by performing learning with borehole brittleness and CQ seismic properties. As the result of core analysis, the shale gas mine is mainly composed of mudstone, quartz, and carbonate, so it is necessary to select a place with a lot of quartz and carbonate rocks. After that, select a place with few components of green clay (Smectite) that absorb moisture and block the gas flow path.

Next, as an individual element, the step (S132-2) of constructing a natural fracture intensity model is as follows. Natural Fracture helps to increase gas flow by expanding the gas flow path during hydraulic fracturing, so it is necessary to select a place where many natural cracks are distributed (Taylor et al., 2013). In addition, the distribution of multiple angles as the crack density is organically connected rather than distributed in a single direction helps to improve the hydraulic crushing efficiency (Sayers and Le Calvez, 2010). In the step of constructing the natural cracking zone density model, the degree of anisotropy due to the speed difference in each direction angle can be calculated as the density of the natural crack.

Next, as an individual element, the step (S132-3) of constructing a pore pressure model is as follows. Another important factor in hydraulic fracturing is the pore pressure. The higher the pore pressure, the stronger the energy that the gas can move to the ground after hydraulic crushing, so the higher the pore pressure, the higher the gas productivity. The step of constructing the pore pressure model is to calculate the pore pressure of the borehole by using the pore pressure prediction method using the p-sonic proposed by Eaton (1972, 1975), to determine the pore pressure spatially. Can be visualized. As a method for predicting the void pressure, specifically, first, density data not measured at the top of the reservoir are connected through an extrapolation technique. Next, hydrostatic pressure and overburden pressure are calculated. Next, the baseline used for calculating the supply pressure is determined using the acoustic sample data. Next, check whether the calculated void pressure fluctuates within the range of the hydrostatic pressure and the load pressure. Next, a process of comparing the predicted pore pressure with borehole DFIT pressure data may be additionally performed.

Next, after constructing the individual elements as described above for the analysis for selecting a place where the hydraulic crushing efficiency is high for each of the three individual elements (ie brittleness, natural crack density, pore pressure), any individual element In order to determine whether the correlation with the hydraulic fracturing is high, a step (S131-4) of building a CQ model by integrating the three elements obtained in steps S132-1 to S132-3 is performed. First, three elements are normalized for each element. The normalization of each element is based on the maximum value (average), which accounts for a ratio of 1% or more in the histogram, and the range of values is minimum 0 and maximum 1. The reason for this process is that it is easy to quantify the weight comparison of each property and the selection of sweet spots after building the integrated model.

After normalizing the three elements, as in the case of RQ analysis, the crushing efficiency model can be constructed by assigning the same 33% weight to the three elements without assigning weights to individual elements (see FIG. 5). ).

In addition, it can be analyzed by assigning different weights to each of the three elements. For example, the map obtained from the three individual element models shows microroseismic, taking into account the increase and decrease in the value of each element and the degree of coincidence with the distribution / direction of the development (see FIG. 6), the brittleness is 50% by weight. , Natural crack density can be assigned a weight of 30%, and the pore pressure can be assigned a weight of 20% (Fig. 7). Since the micro-earthquake event is a result generated according to the characteristics of the crushing efficiency (CQ) factors, it is possible to grasp the influence characteristics of the CQ factors. In the region where the brittleness (BI) has a high value, many microscopic events occur, and in the region having a low value, the number of microscopic earthquake events is low. Micro Earthquake events show high correlation with brittleness, moderate correlation with natural crack density, and low correlation with void pressure (PP). Therefore, as described above, the CQ model may be constructed by assigning a 50% weight to the brittleness, a 30% weight to the natural crack density, and a 20% weight to the pore pressure.

For example, referring to FIGS. 5 and 7, when weight is given, layer A is better than layer B, and when weight is not applied, layer B is better than layer A. On the other hand, the goodness of the C layer is the same.

The step (S133) of building a physical property model by integrating a gas content model and a crushing efficiency model is performed. A physical property model is constructed to select a place where the gas content is high and the hydraulic crushing efficiency is good at the same time (see FIG. 8). Specifically, as described above, the gas content model includes the total organic carbon content, the porosity, the gas saturation rate, and the water permeability, and is integrated after giving the same weight to each element. In addition, the CQ model includes brittleness, natural crack density, supply pressure, and assigns and weights different weights to each element based on the micro earthquake event data.

For example, based on the analysis results of RQ and CQ individual elements when constructing a physical property model, four factors that are not weighted to the RQ model related to gas content (TOC 25% + gas content 25% + porosity 25% + The permeability rate of 25%) and the weighted crushing efficiency CQ model use three weighted factors (50% brittleness + 30% natural crack stand density + 20% void pressure).

Next, the gas content model and the crushing efficiency model, which include the gas content including four physical properties and the three physical properties, are integrated (added) by applying a weight (50%) of the same ratio.

Alternatively, the gas content model and the crushing efficiency model may be combined (added) by assigning different weights. For example, it is possible to assign a weight of 60% to the gas content model and a weight of 40% to the crushing efficiency model. Alternatively, a weight of 40% may be assigned to the gas content model, and a weight of 60% may be assigned to the crushing efficiency model. This weighting method is easy to grasp the effect of gas content and crushing efficiency model on productivity. That is, it can be determined which factor among the two factors (gas content, crushing efficiency) is more affected by the productivity of shale gas. For example, the kiwigana mine in the Hornriver Basin, Canada, has been found to have more impact on crushing efficiency than gas content.

Next, a step (S140) of building a natural crack model is performed. First, in order to build a natural crack model, there must be borehole data and seismic data. Only then can the 3D distribution be predicted. If the seismic data is largely classified, the underground structure itself appears convex, and it can be said that a crack appears to some extent with such curvature. As the direction angle (azimuth angle) data, seismic waves when the surface of the earth is viewed from 360 degrees as a whole can be obtained. When the signal is propagated (reflected) for each acquired direction, there is a difference in amplitude seen in the acoustic waves such as 45 degrees and 145 degrees for each direction angle, which is the difference in amplitude due to cracking. In addition, there is a difference in speed, but if there are many cracks, if the direction in which the signal propagates and the direction of the cracks are orthogonal, the speed becomes slow. Conversely, if the speed and crack direction are the same, the speed is much faster. The difference between the slow speed and the fast speed can be determined whether the cracks are distributed more or less. In addition to these seismic data, borehole data are also used to build a natural model. Analysis of the direction of the crack by looking at the data from the borehole's inlet (ie, the image of the borehole being scanned by scanning the wall) do. Cracks are plotted by entering the direction, size (horizontal and vertical) of the crack, height and width, the viscosity of the fluid as the fluid flows, and the thickness of the flow, and correspond to each grid of the stratified structural model. Accordingly, for each grid, information on cracks is displayed. That is, a natural crack model is obtained.

In other words, when constructing a natural crack model, the characteristics of the density, direction, etc. of natural cracks existing underground are measured through OBMI or FMI samples. Through this, it is possible to construct a natural crack model using the acquired image layer information.

On the other hand, natural cracks are usually made only in a small number of boreholes due to measurement cost, utilization, etc., and it is not easy to assume the existence of natural cracks in the area where the cracks are made. For this reason, seismic properties associated with natural cracks can be extracted from seismic data and used for spatial modeling. Seismic properties include, for example, curvature, coherence, discontinuity, and azimuth (VVAZ / AVAZ). Among the properties of the seismic wave, in the case of constructing a natural crack distribution model predicted using a seismic direction azimuth velocity difference, the parameters for the natural crack model construction are the shape of the crack, the length of the crack, There are mean dip, mean dip azimuth, crack distribution concentration, crack thickness, and crack permeability. The shape of the crack is assumed to have four faces and half the length of the width and length. The length of the crack is the length detected by the seismic properties utilized as spatial guiding (eg, minimum 40 ft, maximum 2000 ft). The average inclination and the average inclination direction angle may use a Fisher model, for example, 35.5 °, 315 ° / 45 ° obtained from an OBMI specimen, and the distribution concentration of cracks may be about 70 °. It can be assumed that the thickness of the crack follows a log-normal function form with an average of 0.000075ft and a standard deviation of 0.0000015ft in the range of 0 to 0.005ft, for example. The crack permeability can be regarded as following the cubic-law that the crack thickness and permeability are correlated. 9 and 10 show examples of the natural crack model accordingly. FIG. 9 exemplarily shows the results of a natural crack model including A, B, and C strata constructed using natural crack densities in the 315 ° and 45 ° directions predicted through the crack modeling parameters and VVAZ inversion. The model results constructed by setting the crack modeling parameters in FIG. 10 and the curvature seismic properties are exemplarily shown. In the constructed model, it can be seen that the A layer has the largest number of natural cracks, and the B and C layers gradually decrease the number of natural cracks.

Next, a step (S200) of predicting sweet spots by building an integrated model from individual models is performed. First, the values of the gas content model and the crushing efficiency model derived in the previous step of building a physical property model (S130) are normalized values as described above. Each element is based on the maximum value (average) that accounts for a ratio of 1% or more in the histogram, and the range of values is a minimum of 0 and a value normalized to a maximum of 1. This is because it is easy to quantify for comparing weights and selecting sweet spots for each property after building an integrated model.

The step of constructing an integrated model from individual models to predict the sweet spot (S200) includes the step of constructing an integrated model by integrating a natural crack model into the physical property model (S210). The integrated model is constructed by combining the 3D physical property model obtained in step S130 of building the physical property model and the 3D natural crack model obtained in natural crack model S140. 11 exemplarily shows an integrated model constructed by integrating a natural crack model into a physical property model.

Next, a step (S220) of predicting a sweet spot from the integrated model is performed. Specifically, in the step of constructing the integrated model (S210), the region having the highest value is selected based on the result of quantified physical properties, and the promising shale gas reservoir is calculated by comparing the average values of the models for each reservoir and calculating the ranking and regionally within the mine area. The promising area is selected using the average map for each reservoir layer. At this time, the previously built natural crack model is included as a natural crack density factor, which is partially used for sweet spot evaluation, and is applied as an essential factor to provide a flow path in the reservoir simulation and hydraulic fracture simulation. It is used to estimate gas production and to estimate the ultimate ultimate recovery (EUR).

Next, a step (S300) of verifying the predicted sweet spot is performed. The step of verifying the predicted sweet spot (S300) includes comparing the predicted yield and the actual production amount by simulation with respect to the predicted sweet spot (S310). In order to determine the suitability of the integrated model, it is necessary to check whether it is consistent with production data for each region (pad) and reservoir section. For example, for the predicted sweet spot, GEM, Eclipse, etc. can be used to predict the production of 90 days, for example, and analyze the productivity and consistency of the integrated model by using production data from the initial 90 days. You can. In addition, the step of verifying the predicted sweet spot (S300) includes a step of verifying the brittleness model, which is an element of crushing efficiency in the integrated model, through X-ray fluorescence analysis (XRF) data (S320). That is, in addition to verifying the integrated model through productivity, the integrating model's completion quality (CQ) model can be verified through the obtained cancer piece-based X-ray fluorescence analysis (XRF) data. In other words, verification of sweet spots is possible through comparison of actual and predicted production over a period of time, and core-based X-ray fluorescence analysis (XRF) data obtained from boreholes has a single brittleness, which is a crushing efficiency model of the integrated model. It is to verify the model. That is, X-ray fluorescence analysis is performed using the obtained core specimen of the shale gas reservoir after drilling on the derived sweet spot, and the measured brittleness value is compared with the brittleness model value. That is, the results obtained from drilling results (core specimens) in the sweet spot area are compared with the model. The physical properties (porosity, TOC, brittleness, etc.) measured through experiments on the core specimen are more accurate and reliable than the physical properties (porosity, TOC, brittleness, etc.) calculated from the physical specimen.

12 is an example of verification of the brittleness model, which is an element of crushing efficiency among the integrated models according to the present invention, and the Young's modulus obtained by performing X-ray fluorescence analysis (XRF) based on the obtained cancer specimens of the crushing efficiency (CQ) model among the integrated models. This is an example of converting the Poisson's ratio to brittleness and comparing it with the brittleness of the crushing efficiency model.

Next, a step (S400) of finally determining the sweet spot of the shale reservoir based on the verification result. In step S300, if the difference between the production during the predetermined period and the production during the actual initial predetermined period predicted by the simulation at the corresponding sweet spot is within an effective range, the corresponding sweet spot is finally determined. In this step, the crack shape (shape), crack length (length), mean dip, mean dip azimuth, crack distribution concentration, and crack thickness (corresponding to the natural cracking characteristics) Since aperture and crack permeability directly affect the flowability of shale gas, productivity is evaluated by considering artificial cracks generated by hydraulic fracturing at the same time. Adjustment of crack shape, crack length, mean dip, mean dip azimuth, crack distribution concentration, crack thickness, and crack permeability values Through this, the difference between the predicted simulation output and the actual output is corrected.

It will be appreciated that the technical configuration of the present invention described above may be implemented in other specific forms by those skilled in the art without changing the technical spirit or essential features of the present invention. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. In addition, the scope of the present invention is shown by the claims below, rather than the above detailed description. In addition, all changes or modifications derived from the meaning and scope of the claims and equivalent concepts should be construed as being included in the scope of the present invention.

S100: steps to build individual models
S200: Step to predict sweet spot by building an integrated model from the individual models
S300: verifying the predicted sweet spot
S400: Final determination of the sweet spot of the shale reservoir based on the verification result

Claims (21)

In the method of deriving the sweet spot of the shale gas reservoir:
Building individual models (S100);
Predicting a sweet spot by constructing an integrated model from the individual models (S200);
Verifying the predicted sweet spot (S300); And
And finally determining a sweet spot of the shale reservoir based on the verification result (S400). The method of claim 1,
Step of building the individual model (S100):
Building a stratified structural model (S110);
Building a rock image model (S120);
Building a physical property model (S130); And
A method of deriving a sweet spot, comprising the step of building a natural crack model (S140). According to claim 2,
The step of predicting the sweet spot by constructing an integrated model from the individual models (S200):
Building an integrated model by integrating the natural crack model into the physical property model based on the stratigraphic structure model and the rock image model (S210); And
And predicting a sweet spot from the integrated model (S220). According to claim 2,
In the step of constructing the stratigraphic structural model (S110), a borehole analysis and a signal that analyzes the results obtained by drilling a borehole wall by drilling through a borehole and sends a signal to the basement to grasp the structural frame of the stratum in the layer to identify the strata The method of deriving sweet spots is to model and lattice the structure. The method of claim 3,
The step of constructing the rock image model (S120) is to predict spatial rock information by using seismic waves and borehole rock information with respect to the stratigraphic structural model, and the spatial rock information is high siliceous, A method for deriving sweet spots comprising low content siliceous, argillaceous, mixed and carbonaceous. According to claim 2,
The step of building the physical property model (S130) is:
Constructing a gas content model by normalizing the total organic carbon content (TOC, total saturation rate, porosity and permeability), and then integrating them (S131);
Normalizing the brittleness, natural crack density, and pore pressure values, and then integrating them to construct a crushing efficiency model (S132); And
And integrating the gas content model and the crushing efficiency model. The method of claim 6,
A method for deriving sweet spots in which total weight of organic carbon (TOC), gas saturation rate, porosity, and water permeability are each given the same weight. The method of claim 6,
A method for deriving sweet spots in which a total weight is given to each of total organic carbon (TOC), gas saturation rate, porosity and permeability. The method of claim 6,
A method for deriving sweet spots, wherein the same weight is given to each of the brittleness, the natural crack density, and the pore pressure. The method of claim 6,
A method for deriving sweet spots wherein different weights are given to each of the brittleness, the natural crack density, and the pore pressure. The method of claim 8,
A method for deriving sweet spots in which total organic carbon (TOC), gas saturation, porosity, and permeability are weighted by 40%, 20%, 20%, and 20%, respectively. The method of claim 8,
A method for deriving sweet spots in which total organic carbon (TOC), gas saturation, porosity, and permeability are weighted by 0%, 40%, 60%, and 0%, respectively. The method of claim 10,
A method for deriving sweet spots in which the brittleness, natural crack density, and pore pressure are weighted by 50%, 30%, and 20%, respectively. The method of claim 6,
The method of deriving a sweet spot in the construction of the physical property model, the same weight is given to each of the gas content model and the crushing efficiency model. The method of claim 6,
A method for deriving sweet spots in which a weight value is assigned to each of the gas content model and the crushing efficiency model when constructing a physical property model. The method of claim 15,
When constructing a physical property model, a sweet spot derivation method in which a gas content model and a crushing efficiency model are weighted by 40% and 60%, respectively. The method of claim 15,
When constructing a physical property model, a sweet spot derivation method in which the gas content model and the crushing efficiency model are weighted by 60% and 40%, respectively. The method of claim 3,
The step of constructing the natural crack model (S140) is a method for deriving a sweet spot from which a natural crack model is built from a difference in amplitude of an elastic wave and a difference in velocity of an elastic wave in the direction angle data of the elastic wave. The method of claim 17,
The step of constructing the natural crack model (S140) includes crack shape, crack length, mean dip, mean dip azimuth, crack distribution concentration, and crack thickness. (aperture), and a method for deriving a sweet spot, wherein a natural crack model is constructed with parameters of crack permeability. The method of claim 1,
The step of verifying the predicted sweet spot (S300) is performed by comparing the production amount for a predetermined period and an actual predetermined period for the predicted sweet spot, which is an element of crushing efficiency in the integrated model. A method for deriving sweet spots, which verifies brittleness models through X-ray fluorescence analysis (XRF) data. The method of claim 20,
The predetermined period is 90 days, sweet spot derivation method.

Images