RU2530324C2 - Method for determining position of marker depth coordinates when constructing geological model of deposit - Google Patents

Abstract

FIELD: physics, computer engineering.

SUBSTANCE: invention relates to a method, a device and a computer-readable data medium used for building a geological model of an oil or other mineral deposit, particularly for determining correlation coefficients for a set of well-logging curves and finding the position of marker depths, for which the correlation coefficient value is maximum. The method enables, for a marker already having marks at a certain so-called reference group of wells, to calculate said marks for a well from another group. For each well W, on which the marker depth value is being searched for, wells of a reference group lying at a given distance from well W are selected, and from among them a well is selected with the highest correlation coefficient, wherein the point at which said maximum is reached is denoted as the desired mark of the marker. Verification tests are used to search for wells in which the correlation function is less than the maximum correlation coefficient, and the correlation quality coefficient is greater than the maximum correlation coefficient. The found well is added to the reference group of wells.

EFFECT: improved accuracy of calculation of parameters used to build a geological model of the location of oil or other deposits.

8 cl, 9 dwg

Description

The invention relates to a method, device and computer-readable storage medium for constructing a geological model of an oil or other field. In particular, the invention relates to a method, apparatus, and computer-readable storage medium for determining correlation coefficients for a complex of GIS curves and for finding marker depth positions for which the correlation coefficient value is maximum.

State of the art

From the prior art, the source EA 200600036 A1, E21B 7/04, 12/30/2008, which describes a package of programs stored in memory at an automated workstation or in another computing computer system, designed to build a model of a single well, is known.

The well-known source US 2007/0276604 A1, G01V 1/00, 11.29.2007, which discloses a method of visualization and organization of these oil and gas fields. The method discloses processing of logging diagrams of wells, which consists in using raster images of logging diagrams, which are converted to digital format for subsequent application of a marker on them.

Source US 2010/0004864 A1, G01V 9/00, January 7, 2010 discloses a method for correlating well logs, which consists in automatically correlating data from a plurality of well logs describing information in different locations.

All of the above sources, to one degree or another, reveal the general principles of modeling the geological model of a mineral deposit, however, none of the above sources provides and does not imply the generation of a high-precision geological model based on correlation coefficients between large groups of wells relative to the positions of depth markers, which turn allows you to increase the accuracy of determining the location of mineral deposits.

Disclosure of invention

The task to which the claimed invention is directed is to build a geological model with precise determination of the positions of oil or other fields.

The technical result is to increase the accuracy of the calculation of parameters used in constructing a geological model of the location of oil or other fields.

The task of correlation of complexes of well logs suggests the presence of a group of wells for which well logs have been studied by methods sufficient for stratigraphic and lithological analysis. Without loss of generality, it is believed that as a result of these studies, a set of well logs was obtained for each well of the group, one for each method. To build a geological model of a deposit, an important step is to trace the boundaries of stratigraphic complexes or lithological properties. These boundaries can be identified along the wells and extended to the study area by interpolation. Such boundaries, called markers, have deep marks on the wells, which are located at the places of the combined features of the well logs.

In one embodiment of the invention, the method includes determining correlation coefficients for a set of well logs on pairs of wells located no further than a predetermined distance from each other, and finding the positions of the depths of the marker for which this coefficient is maximum. Marker depths are searched for wells that are not included in the support group for which these depths are predefined. The method also consists in repeated repetition of the described search with the inclusion of newly found wells in the support group at each iteration. The search is filtered by a set of tests that increase the accuracy of calculations. The method may involve trend markers, the use of which increases its effectiveness.

The method allows for a marker that already has marks on a certain, called reference, group of wells, to be calculated for wells from another group. For each well W, on which the depth of the marker is searched, the wells of the reference group are selected that are separated from the well W at a given distance, and among them the well with the highest value of the correlation coefficient is selected. Then the point at the well W, at which this maximum is reached, is assigned as the desired marker mark. Assuming further the result of the described algorithm by iterating a generalized algorithm, and adding to the support group after each iteration, the wells of its solution, the process is repeated until the algorithm of the method gives an empty result, i.e. It will not report that all wells for which marker marks have been detected are analyzed and identified in a given area. Such “looping” of the main algorithm allows us to obtain a solution for a large number of wells. To increase the reliability of the calculation, a number of tests have been added to the method that do not allow setting marker marks in “bad” places. These tests are as follows: correlation threshold, correlation quality, transitivity level, restrictions on the values of functions in the well.

In the second embodiment of the invention, the method includes calculating the functionals, which are the sum of the correlation coefficients for the set of well logs at the marker points, which are the initial approximation, on pairs of wells located no further than a given distance from each other, and finding their maxima using adapted to the problem of the gradient descent algorithm.

This method allows for markers selected as an initial solution to calculate marker depths at each well that provide the best overall correlation. For each marker included in the set, a functional is determined, which is the sum of the correlation coefficients of the complex of well logging methods for pairs of wells located no further than a given distance from each other. For this functional, partial derivatives are calculated and the vector thus obtained is smoothed out and used to find a larger value of the functional on a certain segment along this vector. If a larger value is not found, then the last position of the marker marks is considered the solution to the problem, and if found, then the solution point is smoothed out and the process is repeated again. At each iteration of the algorithm, marker depths are sorted.

In a third embodiment, the present invention provides an apparatus for determining a marker depth mark on wells W when constructing a geological model of a field. The device may be, but not limited to: a supercomputer, a personal computer, a laptop computer, a tablet computer, a PDA, a smartphone, and the like. The device necessarily contains one or more processors that are designed to execute computer instructions or codes stored in the device’s memory in order to ensure the execution of the methods of the first or second embodiments of the present invention, a computer-readable storage medium (memory) and input / output modules (I / O ) I / O modules are, but are not limited to, typical and prior art device controls: a mouse, keyboard, joystick, touchpad, trackball, electronic pen, stylus, touch screen, and the like. Also, I / O modules are, but are not limited to, typical and known from the prior art means of displaying information: monitor, projector, printer, plotter and the like. By way of example, and not limitation, a computer-readable storage medium may include random access memory (RAM); read-only memory device (ROM); Electrically Erasable Programmable Read-Only Memory (EEPROM); flash memory or other memory technologies; CDROM, digital versatile disc (DVD) or other optical or holographic storage media; magnetic cassettes, magnetic tape, magnetic disk storage device or other magnetic storage devices, wave carriers or other storage medium that can be used to encode the required information and which can be accessed using the device described above.

In a fourth embodiment, the present invention provides a computer-readable storage medium comprising program code that causes a processor and / or processors to perform actions according to the method described in the first or second embodiments of the invention and which, accordingly, are not further described. By way of example, and not limitation, a computer-readable storage medium may include random access memory (RAM); read-only memory device (ROM); Electrically Erasable Programmable Read-Only Memory (EEPROM); flash memory or other memory technologies; CDROM, digital versatile disc (DVD) or other optical or holographic storage media; magnetic cassettes, magnetic tape, magnetic disk storage device or other magnetic memory devices, wave carriers or other storage medium that can be used to encode the required information, and can be accessed by the device described in the third embodiment of the invention, and which is accordingly not further described.

Brief Description of the Drawings

Figure 1 illustrates a General scheme of the method according to the first embodiment of the invention.

Figure 2 illustrates the curves of the GIS method for wells W and W _0, respectively, and the correlation function of the GIS method.

Figure 3 illustrates the step of determining the distance between groups of wells.

Figure 4 illustrates the step of determining the depth z _max corresponding to the maximum correlation value.

Figure 5 illustrates the step of adding to the reference group of wells, the depth of which was determined at the stage of determining the depth z _max .

6 illustrates a general flowchart of a method according to a second embodiment of the invention.

7 illustrates the step of compiling a gradient vector for the current solution point.

Fig. 8 illustrates the step of determining, in the direction of the gradient vector, a functional value greater than the previous one.

Figure 9 illustrates the step of smoothing the solution found.

Detailed Description of Drawings

The method for solving the problem, described below in relation to the first embodiment of the invention, consists in finding the depths of the marker by calculating the correlation coefficient of the complex of well logs for pairs of wells, for one of which this problem has already been solved. Namely, let for some subset of wells, which we call reference, the depths of the marker have already been set. Then, choosing one of the reference wells W ₀ with marker depth z ₀ , for any well W that does not belong to the reference group, we calculate the correlation function C (z), whose values are the correlation coefficients of the set of well logs on well W at point z, and on well W ₀ at the point z ₀ . Denote by z _{max the} maximum of this function. If wells of the support group located in its predetermined neighborhood were found for well W, then the depth z _max corresponding to the reference well with the highest value of the maximum of the function C (z) is considered to be the possible depth of the marker on well W. The correlation function takes values from -1 to 1 and for one GIS method, F is defined as follows:

, $$C(z)=\frac{{\displaystyle \int F_W(T(z+x))F_{W_0}({\mathrm{<figure-callout\; id="0"\; label="z"\; filenames="00000014.png"\; state="\{\{state\}\}">z</figure-callout>}}_0+x)dx}}{\sqrt{{\displaystyle \int F_W^2(T(z+x))dx}}\sqrt{{\displaystyle \int {\mathrm{<figure-callout\; id="0"\; label="F\; W"\; filenames="00000014.png"\; state="\{\{state\}\}">F\; </figure-callout>}}_{W_{}}^{{}_0}({\mathrm{<figure-callout\; id="0"\; label="z"\; filenames="00000014.png"\; state="\{\{state\}\}">z</figure-callout>}}_0+x)dx}}}$$

,

- кривые метода F на скважинах W и W₀ соответственно. Функция T имеет вид T(z)=az+b, причем а отлично от 1, a b от 0 только в том случае, когда заданы один или два трендовых маркера, соответственно. В этих случаях а и b вычисляются как решение системы очевидных линейных уравнений. Для комплекса методов $$\left\{F_i\right\}$$

,i=0,…,n, функция корреляции определяется следующим образом:where f _w and $$F_{W_0}$$

- curves of method F on wells W and W _0, respectively. The function T has the form T (z) = az + b, and a is different from 1, ab from 0 only if one or two trend markers are specified, respectively. In these cases, a and b are calculated as a solution to the system of obvious linear equations. For a set of methods $$\left\{F_i\right\}$$

, i = 0, ..., n, the correlation function is defined as follows:

, $$C(z)=\frac{{\displaystyle \sum _{i=0}^n{w}_i{C}_i(z)}}{{\displaystyle \sum _{i=0}^n{w}_i}}$$

,

where C _i (z) is the correlation function for the method F _i , aw _i are weighting coefficients (Figure 2).

In accordance with FIG. 1, the steps of the method will be described in detail below.

At the first stage of the method, there are wells located in a given neighborhood of the support group of wells, that is, located no further than a predetermined distance R from any well of the support group. The distance between the wells can be determined in several ways: as the distance between the points of the well at a fixed depth, as the distance between the points of a given marker or as the distance between the points of intersection of the wells with a given surface (Figure 3).

The second step of the method consists in calculating the correlation functions and choosing for each well W found at the first stage, the corresponding well from the reference group W ₀ with the maximum correlation coefficient and the corresponding point z _max at the well W.

In the third stage, each well W selected in the second stage is tested for a series of tests and, if this test is successful, the well selection is considered approved, and the depth z _max corresponding to the maximum correlation value is assigned to the marker depth in the well W (Figure 4).

In the fourth stage of the method, wells with the selection approved in the third stage are added to the support group and the transition again to the first stage occurs (Figure 5).

This process is repeated until a single well is found in the first, second, or third stage. The tests included in the verification at the third stage are:

1) the excess of the correlation function at the point z _max specified threshold value. Large threshold correlation values provide greater accuracy of the algorithm, but reduce the number of wells found. For example, if a threshold value of 0.9 is specified, and the maximum of the correlation function at z _max is 0.88, then the well is considered not to have passed the test.

2) excess of a given threshold value by the correlation quality factor at a point. The correlation quality coefficient is defined as the coefficient of deviation of the correlation function closest in value to the local maximum of its correlation function from its largest local maximum at z _max . Large values of this coefficient mean that the found maximum of the correlation coefficient is significantly larger than other local maxima of the correlation function. For example, if the maximum value of the correlation function is 0.9 and the closest maximum is 0.89, then the value of the correlation quality coefficient will be (0.9-0.89) /0.9, which is approximately 0.01. Then, with a threshold value of the quality factor equal to, for example, 0.5, the well will be considered not to have passed the test.

3) exceeding a given threshold value by the degree of transitivity. The degree of transitivity is defined as the number of previous iterations of the algorithm for which the maximum correlation function of the found reference well with the well of the current iteration passes test 1. This test increases the degree of reliability of the method. For example, if the threshold for the values of the correlation function from test 1 is set to 0.9, and well A passes this test when correlating with well B from the reference group, and well B, in turn, was validated by correlation with well C at the previous iteration, then the value the transitivity degree threshold equal to 2 requires obtaining a maximum of the correlation function not lower than 0.9 for well A and C as a reference value for A. The transitivity degree threshold value is assumed to be equal to the number of previous iterations of the algorithm, if this number is less than the specified thresholds e value.

4) the belonging of the values of a given function in the well at the point of maximum correlation to a given range of values. This test allows you to discard knowingly bad places based on the values of any function containing this information. As such a function, the method can use, for example, the coherence function or the deviation function of the wavelet transform extrema with increasing period.

,i=0,…,n глубинные отметки на скважинах W_i. Определим функционал на n-мерном пространстве следующим образом:The method for solving the problem, described below in relation to the second embodiment of the invention, consists in finding the depths of the markers by calculating the maxima of the functionals characterizing the degree of similarity of the complexes of well logs at the points of the markers in the wells. Namely, let $$\left\{z_i\right\}$$

, i = 0, ..., n depth marks on wells W _i . We define a functional on an n-dimensional space as follows:

, $$C({\mathrm{<figure-callout\; id="0"\; label="z"\; filenames="00000014.png"\; state="\{\{state\}\}">z</figure-callout>}}_0,...z_n)={\displaystyle \sum _{i=0}^n{\displaystyle \sum _{j=i+\mathrm{one}}^{n}B(i,j)\cdot C(z_i,z_j)}}$$

,

Here, C _k (z _i , z _j ) denotes the correlation coefficient for the k- _th well logging method at points z _i and z _j at wells i and j, respectively. This coefficient takes values from -1 to 1 and is calculated by the following formula:

$$C_k(z_i,{\text{z}}_{\text{j}})=\frac{{\displaystyle \int F_{k,i}(z_i+x)F_{k,j}(z_j+x)dx}}{\sqrt{{\displaystyle \int F_{k,i}^2(z_i+x)dx}}\sqrt{{\displaystyle \int F_{k,j}^2(z_j+x)dx}}}$$

The maxima of these functionals are found using the gradient descent method (in this case, ascent), which consists in calculating the gradient vector, the coordinates of which are the partial derivatives of the functional, and then searching for the largest values along this vector. To neutralize the typical problems of applying the gradient descent method, a special technique of “shaking” the intermediate solutions of the algorithm is implemented, which reduces to smoothing the current solution point and the current gradient vector. Such smoothing is performed with a predetermined coefficient decreasing at each iteration, and practically does not occur at the last iterations of the algorithm. Smoothing is performed, for example, by the moving average method in the window, and the smoothing factor in this case is the radius of the window. Since when changing the current solution point, its depth at some wells can deviate greatly from the depths at other wells and may be closer in depth to other set markers, the depth of the set markers at each well is sorted. As a result of this sorting, the depths corresponding to one set marker can be assigned to another. This simple technique reduces the scatter of marker depths across different wells.

.At the first stage, the value of the functional C is calculated at the initial point of the solution $$\left\{z_i\right\}$$

.

(Фиг.7).At the second stage of the method, partial derivatives of the functional are calculated and a gradient vector is compiled for the current solution point $$\left\{z_i\right\}$$

(Fig.7).

At the third stage, the gradient vector is smoothed, that is, each component of the vector in the well W is replaced by the average value of the components in the wells that are in the vicinity of the well W of a given radius R.

и заданной длины в направлении вектора градиента с заданным шагом ищется значение функционала, большее предыдущего. Если такое значение не найдено, то алгоритм прекращает работу и текущее решение $$\left\{z_i\right\}$$

принимается за окончательное (Фиг.8).In the fourth stage for a segment with a starting point $$\left\{z_i\right\}$$

and a given length in the direction of the gradient vector with a given step, the value of the functional is searched, greater than the previous one. If such a value is not found, then the algorithm stops working and the current solution $$\left\{z_i\right\}$$

taken as final (Fig. 8).

At the fifth stage, the solution found is refined by searching for a larger value within the step for which a maximum is found, with a smaller step.

полагается точка, в которой достигнут максимум функционала и производится сглаживание этой точки тем же способом, что вектора градиента на третьем этапе, а радиус сглаживания R уменьшается на заданную величину (Фиг.9).In the sixth step, the current decision $$\left\{z_i\right\}$$

it relies on the point at which the maximum of the functional is reached and this point is smoothed in the same way as the gradient vector in the third stage, and the smoothing radius R is reduced by a predetermined value (Fig. 9).

At the seventh stage, the marks of the set markers are sorted by depth and the transition to the second stage of the algorithm is performed.

Claims (8)

1. A method for determining the position of the depth coordinate of the well marker W when constructing a geological model of a mineral deposit, comprising the steps of:
1) determine the support group of wells;
2) determine the wells W located in a given neighborhood of the support group of wells, while the neighborhood does not exceed the value of the distance R from any of the wells of the support group;
3) determine the correlation function C (z) and select, for each well W found at the previous stage, the corresponding well from the reference group with the maximum value of the correlation coefficient and the corresponding point Z _max at the well W;
4) check the well W using validation tests and, if this test is successful, then the well selection is considered approved, and the depth corresponding to the maximum correlation value Z _max is assigned to the depth of the marker on the well W, and at least one validation test is a validation test where a check is made for the maximum exceeding the correlation function C (z) at the point Z _{max of the} specified threshold value, and the maximum value of the correlation function during the execution of the test test carbonize up to tenths of a larger value, and if the value of the correlation maximum during rounding is equal to or greater than the specified threshold value, then the tested well is considered to have failed the test, and at least one additional verification test is an additional verification test, which is carried out checking for excess of the correlation quality factor by a predetermined threshold value, and the value of the correlation quality coefficient is defined as the open coefficient -toxic nearest local maximum value of correlation function from its highest point in the local max Z _max, and if the correlation value of quality factor is less than a predetermined threshold, the tested well is considered not passed the test;
5) add the found well W to the support group of wells; and
6) iteratively repeat all the steps of the method until no wells W are identified in steps 1) or 2) or 3) of the method. 2. The method according to claim 1, characterized in that the additional checking test checks for exceeding a predetermined threshold value by the degree of transitivity, which is defined as the number of previous iterations of the algorithm for which the maximum correlation function of the found reference well with the current iteration well does not exceed the maximum of the function correlations C (z) at the point Z _max . 3. The method according to claim 2, characterized in that an additional verification test establishes a predetermined interval of values and checks for the presence of wells that are not included in the specified interval for subsequent screening. 4. A device for determining the position of the coordinates of the depths of marker wells W when constructing a geological model of a mineral deposit, containing:
one or multiple processors;
input / output modules (I / O);
A computer-readable storage medium (memory) that contains program code that, when executed, causes the processor or processors to perform the steps in which:
1) determine the support group of wells;
2) determine the wells W located in a given neighborhood of the support group of wells, while the neighborhood does not exceed the value of the distance R from any of the wells of the support group;
3) determine the correlation function C (z) and select, for each well W found at the previous stage, the corresponding well from the reference group with the maximum value of the correlation coefficient and the corresponding point Z _max at the well W;
4) check the well W using validation tests and, if this test is successful, then the well selection is considered approved, and the depth corresponding to the maximum correlation value Z _max is assigned to the depth of the marker on the well W, and at least one validation test is a validation test where a check is made for the maximum exceeding the correlation function C (z) at the point Z _{max of the} specified threshold value, and the maximum value of the correlation function during the execution of the test test carbonize up to tenths of a larger value, and if the value of the correlation maximum during rounding is equal to or greater than the specified threshold value, then the tested well is considered to have failed the test, and at least one additional verification test is an additional verification test, which is carried out checking for excess of the correlation quality factor by a predetermined threshold value, and the value of the correlation quality coefficient is defined as the open coefficient -toxic nearest local maximum value of correlation function from its highest point in the local max Z _max, and if the correlation value of quality factor is less than a predetermined threshold, the tested well is considered not passed the test;
5) add the found well W to the support group of wells; and
6) iteratively repeat all the steps of the method until no wells W are identified in steps 1) or 2) or 3) of the method. 5. The device according to claim 4, characterized in that the additional checking test checks for exceeding a predetermined threshold value by the degree of transitivity, which is defined as the number of previous iterations of the algorithm for which the maximum correlation function of the found reference well with the current iteration well does not exceed the maximum of the function correlations C (z) at the point Z _max . 6. The device according to claim 5, characterized in that an additional verification test sets the specified interval of values and checks for the presence of wells that are not included in the specified interval for subsequent screening. 7. The device according to claim 4, characterized in that the additional checking test sets the specified interval of values and checks for the presence of wells that are not included in the specified interval for subsequent screening. 8. A computer-readable storage medium suitable for use in a device according to any one of claims 4 to 7, which contains program code, which upon execution causes the processor or processors to perform actions of the methods according to any one of claims 1 to 3.

Images