A comprehensive optimization inversion method for logging and mud analysis of altered igneous rock minerals

Abstract

The invention discloses a comprehensive optimization inversion method for logging and logging of altered igneous rock minerals, and the specific steps are as follows: step one: establishing an element logging data standardization method, combining core major element analysis data to evaluate the reliability of standardized element logging data, and laying a data foundation; step two: based on core identification data, considering altered clay minerals and porosity, establishing a multi-mineral volume model of altered igneous rock formations; step three: referring to conventional logging optimization inversion methods, establishing element response equations, and using multivariate regression methods to obtain element characteristic values ​​of different minerals; step four: performing comprehensive inversion of minerals and porosity on actual altered igneous rock formations, and establishing a multi-mineral volume model considering altered clay; the comprehensive optimization inversion obtained by combining the corrected element logging and conventional logging data can invert the content and porosity of multiple mineral components, and at the same time can ensure that the inversion accuracy meets the exploration needs.

Description

Technical Field

[0001] The invention relates to the technical field of altered igneous rock formation mineral and porosity inversion, in particular to an altered igneous rock mineral logging comprehensive optimization inversion method. Background Art

[0002] As one of the most important reservoirs in the world, accurate prediction of the mineral content of altered igneous rocks is of great significance for the exploration of igneous reservoirs. However, compared with common sandstone reservoirs, the complex mineral composition and pore structure of igneous rocks and the ubiquitous alteration phenomenon make it more difficult to identify their lithology and invert minerals and porosity. The common logging optimization mineral analysis method is limited by the conventional logging response equation. The number of mineral types that can be inverted and the accuracy of the content cannot meet the exploration needs of igneous formations, and the prediction results of formation porosity are also poor.

[0003] Mineral analysis is an indispensable part of reservoir exploration. Different mineral contents have a great influence on the reservoir space. Mineral analysis is very important for igneous reservoir exploration. The currently recognized mineral analysis methods are XRD diffraction whole-rock mineral analysis and micro X-ray fluorescence (μXRF); on this basis, some scholars determined the conversion coefficient between the core mineral content and the element content, and converted the element content into the mineral content by relying on the conversion coefficient; later, some scholars used whole-rock chemistry (47 elements) and combined linear multiple regression, artificial neural network, random forest and multivariate adaptive regression spline four regression methods to predict the mineral properties of ores, and found that the multivariate adaptive regression spline method was the best, but these studies were core scale, and it is difficult to apply globally in logging interpretation. At the same time, although the optimization algorithm is a common mineral inversion method in the field of logging, the conventional logging data available for this method is limited, and it is very limited for igneous formations with many mineral components and fast changes. Summary of the invention

[0004] The purpose of the present invention is to provide a comprehensive optimization inversion method for logging and well-logging of altered igneous rock minerals to solve the problem that the common logging optimization mineral analysis method proposed in the above background technology is limited by the conventional logging response equation, and the number of mineral types that can be inverted and the content accuracy cannot meet the exploration needs of igneous rock formations.

[0005] To achieve the above object, the present invention provides the following technical solution: a comprehensive optimization inversion method for logging and well logging of altered igneous rock minerals, the technical steps of which include:

[0006] Step 1: Establish a standardization method for element logging data, and evaluate the reliability of the standardized element logging data in combination with the core major element analysis data to lay the foundation for the data;

[0007] Step 2: Based on the core identification data, a multi-mineral volume model of the altered igneous formation is established taking into account the altered clay minerals and porosity;

[0008] Step 3: Refer to the conventional logging optimization inversion method to establish the element response equation, and use the multivariate regression method to obtain the element characteristic values ​​of different minerals;

[0009] Step 4: Based on conventional logging data and standardized element logging data, comprehensive inversion of mineralogy and porosity of actual altered igneous rock formations is performed.

[0010] Furthermore, the step 1 of establishing a method for standardizing element logging data and evaluating data reliability includes the following steps:

[0011] Step 1 Element logging data standardization method:

[0012] The element logging data include the contents of 17 elements, including sodium (Na), magnesium (Mg), aluminum (Al), silicon (Si), phosphorus (P), sulfur (S), chlorine (Cl), potassium (K), calcium (Ca), barium (Ba), titanium (Ti), vanadium (Vr), manganese (Mn), iron (Fe), nickel (Nr), strontium (Sr) and zirconium (Zr), while the major element analysis of the core includes the relative percentage content data of 10 oxides, including sodium oxide (Na2O), magnesium oxide (MgO), aluminum oxide (Al2O3), silicon dioxide (SiO2), phosphorus oxide (P2O5), potassium oxide (K2O), calcium oxide (CaO), titanium oxide (TiO2), manganese oxide (MnO) and iron oxide (Fe2O3);

[0013] The principle of the element logging standardization method is: select 10 elements corresponding to the core oxides in the element logging data, namely sodium (Na), magnesium (Mg), aluminum (Al), silicon (Si), phosphorus (P), potassium (K), calcium (Ca), titanium (Ti), manganese (Mn), and iron (Fe) for standardization, and the remaining 7 elements are not involved in the standardization;

[0014] The calculation formula of the relative content of the 10 elements after standardization is as follows:

[0015]

[0016] Where W X is the relative percentage content of X element in mud logging, in %; W LX is the logging data corresponding to the X element; ∑W LX It is the sum of logging data of 10 elements: Na, Mg, Al, Si, P, K, Ca, Ti, Mn and Fe;

[0017] Step 2 Obtain core element analysis data:

[0018] First, based on the molar mass of different elements, the molar mass ratio of non-oxygen elements in different oxides is calculated:

[0019]

[0020] Where W MX is the molar mass percentage of element X; M X is the molar mass of element X, in g / mol; M XO is the molar mass of the oxide corresponding to element X, in g / mol;

[0021] Then find the relative content of different elements:

[0022]

[0023] Where W X1 is the relative percentage content of X element in the core, in %; W XO is the relative content of core oxide corresponding to element X, in %; ∑W XO *W MX It is the sum of the contents of 10 non-oxygen elements, namely Na, Mg, Al, Si, P, K, Ca, Ti, Mn and Fe, in oxides;

[0024] Step 3: Evaluate the reliability of standardized element logging data:

[0025] Formula (1) is used to standardize the logging element data and compare them with the core element data.

[0026] Furthermore, in step 2, based on the core identification data, the process of establishing a multi-mineral volume model of the altered igneous rock formation by considering the altered clay minerals and porosity is as follows:

[0027] Step 1 Analyze the mineral components of altered igneous rocks:

[0028] Taking the diorite and tectonic schist in a depression in the northern part of the South China Sea as an example, according to the identification of core thin sections, it can be known that: in diorite, the primary minerals of the rock are mainly composed of plagioclase, potassium feldspar, common hornblende and biotite, and the accessory minerals are magnetite and zircon, with a content of less than 1%; plagioclase has strong sericite and mud alteration, and the edges are often replaced and interspersed by biotite; tectonic schist presents porphyritic metamorphic structure and residual porphyritic knot structure, and the content of porphyroblastic hornblende is 3%; the matrix is ​​mainly composed of fine-micronized hornblende, biotite, plagioclase, and contains a small amount of accessory mineral magnetite;

[0029] Step 2 Consider the multi-mineral volume model of altered clay:

[0030] The altered igneous strata in the study area have strong heterogeneity, and the mineral composition is difficult to calculate accurately. The mineral components corresponding to different lithologies are obviously different. Table 1 shows the mineral components and corresponding chemical formulas and densities obtained based on the XRD diffraction whole-rock mineral analysis data of different igneous rocks in the study area. There are 21 minerals in total. However, the current optimization algorithm cannot invert 21 minerals at the same time, and it is necessary to establish a suitable multi-mineral volume model.

[0031] Since the element logging data is closely related to the mineral chemical formula, the physical and chemical properties of minerals need to be considered comprehensively when establishing a multi-mineral volume model; a mineral merging scheme is established in combination with the density and chemical formula of the minerals in the study area: quartz, potassium feldspar and plagioclase as the main mineral components are not merged, clay minerals include kaolinite, illite, montmorillonite, chlorite altered clay minerals and lauzenite with physical and chemical properties similar to altered clay, dark minerals include common pyroxene, amphibole and anhydrite with similar density, metal minerals include pyrite, siderite, goethite, hematite, magnesite, anatase, carbonate minerals include calcite, dolomite, aragonite and ankerite; finally, a multi-mineral volume model of altered igneous rock consisting of 8 components: quartz, potassium feldspar, plagioclase, clay minerals, dark minerals, metal minerals, carbonate minerals and pores is obtained.

[0032] Furthermore, in step 3, the process of establishing the element response equation by referring to the conventional well logging optimization inversion method and obtaining the element characteristic values ​​of different minerals by using the multivariate regression method is as follows:

[0033] Step 1 Establish the response equation of the element:

[0034] Natural gamma (GR), acoustic transit time (AC), compensated neutron (φ N ), compensation density (ρ b ) The four logging curve values ​​and the standardized element logging data are taken as the actual values. The volume content of pore fluid (φ) and the volume content of each mineral component in different volume models are calculated by the optimization algorithm. The corresponding response equation is:

[0035] Well logging response equation:

[0036] Element response equation: W E =∑W Ei *W i (i=1,2,3...N) (5)

[0037] In the formula, i represents the i-th component; φ Ni represents the apparent neutron porosity value of the i-th component; ρ i represents the density value of the i-th component; AC i Represents the acoustic time difference of the i-th component; GRi represents the natural gamma value of the i-th component; W E W is the weight percentage content of each element after standardization; Ei is the element characteristic value of the i-th component; W i is the weight percentage content of the i-th component;

[0038] Because the sum of the pore volume and the volume content of each mineral in the volume model is 1, an additional equilibrium equation is added:

[0039]

[0040] Thus, equations (4) to (6) form a set of equations (7), based on which the optimal solution is obtained using two optimization algorithms; therefore, the optimization comprehensive inversion problem is transformed into the problem of selecting an appropriate algorithm to obtain the optimal solution of the following set of equations:

[0041]

[0042] Where T is the actual data set, R is the logging and element logging response parameters of the i-th component, V i is the volume of the i-th component;

[0043] The optimization method is based on the volume model. Therefore, the weight percentage data of element logging needs to be converted between weight percentage and volume percentage using the density of different minerals when participating in the optimization solution. Formula (8) to Formula (10) are the conversion formulas for converting mineral weight content into volume content:

[0044]

[0045] Where V mineral The volume of a mineral, W mineral is the weight percentage of a mineral; ρ mineral is the density of a certain mineral; is the percentage of minerals in the skeleton; is the volume percentage of a mineral in the entire rock; is the rock porosity; the volume content of minerals in the rock can be converted through formulas (8) to (10);

[0046] Step2 Multiple regression calculation of element eigenvalues:

[0047] For elemental characteristic values, different minerals have different chemical compositions and their elemental compositions vary greatly. Therefore, it is necessary to determine the elemental composition of mineral components in different models based on the multi-mineral volume model of igneous rocks in the study area, and then use multiple linear regression to obtain elemental characteristic values. Multiple linear regression analysis is a mathematical analysis method based on the correlation between independent variables and dependent variables. In this paper, the element logging data of different minerals are used as independent variables, and the corresponding weight contents of multiple minerals are used as dependent variables to perform multiple linear regression analysis. Assuming that the dependent variable y is related to the i independent variables x i The general formula of the multiple linear regression model is:

[0048] y=β1x1+β2x2+β3x3……+β n x n +ε (11)

[0049] In the formula, y is the dependent variable, representing the element content; β1, β2, ... β n are regression coefficients; x1, x2, ... x n is the independent variable, representing the content of different minerals; ε is the residual, which obeys the normal distribution N(0, σ 2 );

[0050] It is usually necessary to normalize the dependent variables and predictor variables, but in order to better meet the actual physical meanings between different elements and minerals, the relevant data are not normalized when performing multivariate fitting;

[0051] Next, the regression equation between elements and minerals is established according to different mineral components; the mineral components of igneous rocks in the study area are complex. According to previous statistical results, the common minerals in igneous rocks in this area are quartz, potassium feldspar, plagioclase, amorphous, lauzenite, clay minerals, amphibole, pyroxene, anhydrite, pyrite, siderite, goethite, hematite, magnesite, anatase, aragonite, calcite, dolomite, and ankerite;

[0052] Taking the element characteristic parameter value of a certain element (N) contained in different minerals as an example, we can get:

[0053] W N =∑β X W X (12)

[0054] Where W N is the value of element N in the element logging data after standardization; W X Represents the weight percentage content of the Xth mineral component in the rock mineral model; β X is the regression coefficient of the Xth mineral component, which is also the target we want to obtain, that is, the characteristic value of the element response equation;

[0055] When a mineral component does not contain a certain element, its corresponding element characteristic value is considered to be 0. Then the element characteristic values ​​of different minerals can be regressed according to formula (12). At the same time, since the element logging measurement is based on the rock skeleton, the element characteristic values ​​of the pore fluid are all 0.

[0056] While regressing to obtain the characteristic values ​​of different mineral elements, the regression correlation coefficient can also be calculated. The higher the correlation coefficient, the better the relationship between the element and the mineral and the more effective the characteristic is. The lower the correlation coefficient, the poorer the relationship between the mineral and the element, and the characteristic value is a redundant characteristic. When performing optimization calculations, the more response equations there are, the slower the calculation speed. In order to improve the optimization calculation speed and avoid redundant characteristics that reduce the calculation accuracy, elements with higher correlation with minerals can be selected to participate in the calculation according to the size of the correlation coefficient.

[0057] Furthermore, the process of performing comprehensive inversion of minerals and porosity of the actual altered igneous rock formation based on conventional logging data and standardized element logging data in step 4 is as follows:

[0058] Step1L-M comprehensive optimization inversion and result analysis:

[0059] Taking the diorite and tectonic schist strata of Well X-1 in a depression in the northern South China Sea as an example, based on conventional logging and standardized element logging data, the LM optimization method was used to perform a comprehensive inversion of multiple minerals and porosity; the input data included four conventional logging curves of natural gamma, acoustic time difference, compensated neutron, and compensated density, and five standardized element logging data of magnesium, aluminum, silicon, calcium, and iron, which fully met the inversion requirements of simultaneously obtaining eight components in the volume model, and the inversion results were very good.

[0060] Beneficial effects:

[0061] The present invention combines core element data to create a correction method for element logging data, allowing the application of unstable element logging data; establishes a multi-mineral volume model of altered clay minerals, effectively coping with the complex mineral components of altered igneous rock formations; establishes element response equations for minerals based on the corrected element logging data, breaking through the limitation of the limited number of minerals in conventional logging optimization inversion; uses a multivariate linear regression method to calculate the element characteristic values ​​of different minerals and combines them with conventional logging to perform LM optimization inversion, solving the problem that the element characteristic values ​​of mineral components are difficult to determine; and the comprehensive optimization inversion obtained by combining the corrected element logging with conventional logging data can not only simultaneously invert the content and porosity of multiple mineral components, but also ensure that the inversion accuracy meets exploration needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the drawings of the embodiment are briefly introduced below.

[0063] The drawings described below are only related to some embodiments of the present invention, but are not intended to limit the present invention.

[0064] In the attached picture:

[0065] Figure 1 It is a comparison diagram between the standardized logging element data and the core data of an embodiment of the present invention.

[0066] Figure 2 This is an identification diagram of an igneous rock core thin section according to an embodiment of the present invention.

[0067] Figure 3 It is a volume model diagram of altered igneous rock multi-minerals according to an embodiment of the present invention.

[0068] Figure 4 This is a diagram of the optimized comprehensive inversion result of multiple minerals and pores of altered igneous rocks according to an embodiment of the present invention.

[0069] Figure 5 is a flow chart of an embodiment of the present invention. DETAILED DESCRIPTION

[0070] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0071] The present invention provides the following technical solution: a comprehensive optimization inversion method for logging and well logging of altered igneous rock minerals, the steps of the technology comprising:

[0072] Step 1: Establish a standardization method for element logging data, and evaluate the reliability of the standardized element logging data in combination with the core major element analysis data to lay the foundation for the data;

[0073] Step 2: Based on the core identification data, a multi-mineral volume model of the altered igneous formation is established taking into account the altered clay minerals and porosity;

[0074] Step 3: Refer to the conventional logging optimization inversion method to establish the element response equation, and use the multivariate regression method to obtain the element characteristic values ​​of different minerals;

[0075] Step 4: Based on conventional logging data and standardized element logging data, comprehensive inversion of mineralogy and porosity of actual altered igneous rock formations is performed.

[0076] Among them, the establishment of the element logging data standardization method and the evaluation of data reliability in step 1 include the following steps:

[0077] Step 1 Element logging data standardization method:

[0078] The element logging data includes the contents of 17 elements, including sodium (Na), magnesium (Mg), aluminum (Al), silicon (Si), phosphorus (P), sulfur (S), chlorine (Cl), potassium (K), calcium (Ca), barium (Ba), titanium (Ti), vanadium (Vr), manganese (Mn), iron (Fe), nickel (Nr), strontium (Sr) and zirconium (Zr), while the main element analysis of the core is the relative percentage content data of 10 oxides, including sodium oxide (Na2O), magnesium oxide (MgO), aluminum oxide (Al2O3), silicon dioxide (SiO2), phosphorus oxide (P2O5), potassium oxide (K2O), calcium oxide (CaO), titanium oxide (TiO2), manganese oxide (MnO) and iron oxide (Fe2O3). It can be seen that the dimensions of different methods are different, and the difference in element types also affects the accuracy of element logging data. Therefore, before using element logging data for analysis, we need to first determine the correction method of element logging data, and determine the accuracy of logging elements. Only then can the relevant research carried out on this basis be reliable.

[0079] The principle of the element logging standardization method is: select 10 elements corresponding to the core oxides in the element logging data, namely sodium (Na), magnesium (Mg), aluminum (Al), silicon (Si), phosphorus (P), potassium (K), calcium (Ca), titanium (Ti), manganese (Mn), and iron (Fe) for standardization, and do not participate in the standardization of the remaining 7 elements, so as to reduce the error caused by different types of elements;

[0080] The calculation formula of the relative content of the 10 elements after standardization is as follows:

[0081]

[0082] Where W X is the relative percentage content of X element in mud logging, in %; W LX is the logging data corresponding to the X element; ΣW LX It is the sum of logging data of 10 elements: Na, Mg, Al, Si, P, K, Ca, Ti, Mn and Fe;

[0083] Step 2 Obtain core element analysis data:

[0084] First, based on the molar mass of different elements, the molar mass ratio of non-oxygen elements in different oxides is calculated:

[0085]

[0086] Where W MX is the molar mass percentage of element X; M X is the molar mass of element X, in g / mol; MXO is the molar mass of the oxide corresponding to element X, in g / mol;

[0087] Then find the relative content of different elements:

[0088]

[0089] Where W X1 is the relative percentage content of X element in the core, in %; W XO is the relative content of core oxide corresponding to element X, in %; ∑W XO *W MX It is the sum of the contents of 10 non-oxygen elements, namely Na, Mg, Al, Si, P, K, Ca, Ti, Mn and Fe, in oxides;

[0090] Step 3: Evaluate the reliability of standardized element logging data:

[0091] Formula (1) is used to standardize the logging element data and compare it with the core element data. Figure 1 The following is a comparative analysis chart: the first track is the depth track; the second to eleventh tracks are the comparative analysis of the standardized element logging data and core elements of Si, Ca, Na, Mg, Al, K, Fe, Ti, Mn, and P elements, respectively. The red bar curve is the weight content of the core elements, and the black dots are the standardized element logging data. It can be seen from the figure that the W_X and W_X1 of the Na, Mg, Al, Si, P, K, Ca, Ti, Mn, and Fe elements are basically consistent, proving that the standardized element logging data are reliable and the logging element data can be standardized according to formula (1).

[0092] The process of establishing a multi-mineral volume model of altered igneous rock formations in step 2 based on core identification data and considering altered clay minerals and porosity is as follows:

[0093] Step 1 Analyze the mineral components of altered igneous rocks:

[0094] Taking the diorite and tectonic schist in a depression in the northern South China Sea as an example, according to the identification of core thin sections, it can be known that diorite, such as Figure 2 As shown in (a), the primary minerals of the rock are mainly composed of plagioclase, potassium feldspar, common hornblende, and biotite. The accessory minerals are magnetite and zircon, with a content of less than 1%. Plagioclase has strong sericite and mud alteration, and its edges are often replaced and interspersed by biotite. Tectonic schists, such as Figure 2 (b) shows a porphyritic metamorphic structure and a residual porphyritic knot structure. The content of porphyroblastic hornblende is 3%. The matrix is ​​mainly composed of fine-grained hornblende, biotite, plagioclase, and contains a small amount of accessory mineral magnetite. It can be seen that the mineral composition of altered igneous rocks is very complex, and the alteration is very common.

[0095] Step 2 Consider the multi-mineral volume model of altered clay:

[0096] The altered igneous strata in the study area have strong heterogeneity, and the mineral composition is difficult to calculate accurately. The mineral components corresponding to different lithologies are obviously different. Table 1 shows the mineral components and corresponding chemical formulas and densities obtained based on the XRD diffraction whole-rock mineral analysis data of different igneous rocks in the study area. There are 21 minerals in total. However, the current optimization algorithm cannot invert 21 minerals at the same time, and it is necessary to establish a suitable multi-mineral volume model.

[0097] Since the element logging data is closely related to the mineral chemical formula, the physical and chemical properties of minerals need to be considered comprehensively when establishing a multi-mineral volume model; a mineral merging scheme is established in combination with the density and chemical formula of the minerals in the study area: quartz, potassium feldspar and plagioclase as the main mineral components are not merged, clay minerals include kaolinite, illite, montmorillonite, chlorite altered clay minerals and lauzenite with physical and chemical properties similar to altered clay, dark minerals include common pyroxene, amphibole and anhydrite with similar density, metal minerals include pyrite, siderite, goethite, hematite, magnesite, anatase, carbonate minerals include calcite, dolomite, aragonite and ankerite; finally, an altered igneous rock multi-mineral volume model consisting of quartz, potassium feldspar, plagioclase, clay minerals, dark minerals, metal minerals, carbonate minerals and pores is obtained, as shown in Figure 3 shown.

[0098] Among them, in step 3, the process of establishing the element response equation by referring to the conventional logging optimization inversion method and obtaining the element characteristic values ​​of different minerals by using the multivariate regression method is as follows:

[0099] Step 1 Establish the response equation of the element:

[0100] Natural gamma (GR), acoustic transit time (AC), compensated neutron (φ N ), compensation density (ρ b ) The four logging curve values ​​and the standardized element logging data are taken as the actual values. The volume content of pore fluid (φ) and the volume content of each mineral component in different volume models (quartz, potassium feldspar, plagioclase, clay minerals, dark minerals, metal minerals, carbonate minerals) are calculated by the optimization algorithm. The corresponding response equation is:

[0101] Well logging response equation:

[0102] Element response equation: W E =∑W Ei *W i (i=1,2,3...N) (5)

[0103] In the formula, i represents the i-th component; φ Ni represents the apparent neutron porosity value of the i-th component; ρ i represents the density value of the i-th component; AC i Represents the acoustic time difference of the i-th component; GR i represents the natural gamma value of the i-th component; W E W is the weight percentage content of each element after standardization; Ei is the element characteristic value of the i-th component; W i is the weight percentage content of the i-th component;

[0104] Because the sum of the pore volume and the volume content of each mineral in the volume model is 1, an additional equilibrium equation is added:

[0105]

[0106] Thus, equations (4) to (6) form a set of equations (7), based on which the optimal solution is obtained using two optimization algorithms; therefore, the optimization comprehensive inversion problem is transformed into the problem of selecting an appropriate algorithm to obtain the optimal solution of the following set of equations:

[0107]

[0108] Where T is the actual data set, R is the logging and element logging response parameters of the i-th component, V i is the volume of the i-th component;

[0109] The optimization method is based on the volume model. Therefore, the weight percentage data of element logging needs to be converted between weight percentage and volume percentage using the density of different minerals when participating in the optimization solution. Formula (8) to Formula (10) are the conversion formulas for converting mineral weight content into volume content:

[0110]

[0111] Where V mineral The volume of a mineral, W mineral is the weight percentage of a mineral; ρ mineral is the density of a certain mineral; is the percentage of minerals in the skeleton; is the volume percentage of a mineral in the entire rock; is the rock porosity; the volume content of minerals in the rock can be converted through formulas (8) to (10);

[0112] Step2 Multiple regression calculation of element eigenvalues:

[0113] For elemental characteristic values, different minerals have different chemical compositions and their elemental compositions vary greatly. Therefore, it is necessary to determine the elemental composition of mineral components in different models based on the multi-mineral volume model of igneous rocks in the study area, and then use multiple linear regression to obtain elemental characteristic values. Multiple linear regression analysis is a mathematical analysis method based on the correlation between independent variables and dependent variables. In this paper, the element logging data of different minerals are used as independent variables, and the corresponding weight contents of multiple minerals are used as dependent variables to perform multiple linear regression analysis. Assuming that the dependent variable y is related to the i independent variables x i The general formula of the multiple linear regression model is:

[0114] y=β1x1+β2x2+β3x3……+β n x n +ε (11)

[0115] In the formula, y is the dependent variable, representing the element content; β1, β2, ... β n are regression coefficients; x1, x2, ... x n is the independent variable, representing the content of different minerals; ε is the residual, which obeys the normal distribution N(0, σ 2 );

[0116] It is usually necessary to normalize the dependent variables and predictor variables, but in order to better meet the actual physical meanings between different elements and minerals, the relevant data are not normalized when performing multivariate fitting;

[0117] Next, the regression equation between elements and minerals is established according to different mineral components; the mineral components of igneous rocks in the study area are complex. According to previous statistical results, the common minerals of igneous rocks in this area are quartz, potassium feldspar, plagioclase, amorphous, lauzenite, clay minerals, amphibole, pyroxene, anhydrite, pyrite, siderite, goethite, hematite, magnesite, anatase, aragonite, calcite, dolomite, and ankerite. Among them, clay minerals include chlorite, kaolinite, illite and montmorillonite;

[0118] Taking the element characteristic parameter value of a certain element (N) contained in different minerals as an example, we can get:

[0119] W N =∑β X W X (12)

[0120] Where W N is the value of element N in the element logging data after standardization; W X Represents the weight percentage content of the Xth mineral component in the rock mineral model (including quartz, potassium feldspar, plagioclase, clay minerals, dark minerals, metal minerals and carbonate minerals); β Xis the regression coefficient of the Xth mineral component, which is also the target we want to obtain, that is, the characteristic value of the element response equation;

[0121] When a mineral component does not contain a certain element, its corresponding element characteristic value is considered to be 0. Then the element characteristic values ​​of different minerals can be regressed according to formula (12). At the same time, since the element logging measurement is based on the rock skeleton, the element characteristic values ​​of the pore fluid are all 0.

[0122] While regressing to obtain the characteristic values ​​of different mineral elements, the regression correlation coefficient can also be calculated. The higher the correlation coefficient, the better the relationship between the element and the mineral and the more effective the characteristic is. The lower the correlation coefficient, the poorer the relationship between the mineral and the element, and the characteristic value is a redundant characteristic. When performing optimization calculations, the more response equations there are, the slower the calculation speed. In order to improve the optimization calculation speed and avoid redundant characteristics that reduce the calculation accuracy, elements with higher correlation with minerals can be selected to participate in the calculation according to the size of the correlation coefficient.

[0123] Among them, the process of comprehensive inversion of minerals and porosity of actual altered igneous rock formations based on conventional logging data and standardized element logging data in step 4 is as follows:

[0124] Step1L-M comprehensive optimization inversion and result analysis:

[0125] Taking the diorite and tectonic schist formations of Well X-1 in a depression in the northern South China Sea as an example, based on conventional logging and standardized element logging data, the LM optimization method is used to perform a comprehensive inversion of multiple minerals and porosity; the input data include four conventional logging curves of natural gamma, acoustic time difference, compensated neutron, and compensated density, and five standardized element logging data of magnesium, aluminum, silicon, calcium, and iron, which fully meet the inversion requirements of obtaining eight components (pores and seven minerals) in the volume model at the same time, and the inversion results are very good;

[0126] like Figure 4 As shown in the figure, based on conventional logging and standardized element logging data, the mineral contents of Well X-1 obtained by LM optimization comprehensive inversion are basically consistent with the XRD data, and the porosity is also consistent with the core porosity, indicating that the multi-mineral and porosity comprehensive inversion method can not only solve the shortcomings of conventional logging optimization, but also ensure the accuracy of the calculation results, indicating that the comprehensive inversion method of logging optimization is more universal and effective.

[0127] The specific flow chart of this method is as follows Figure 5As shown in the figure: first, the element logging data is corrected to obtain standardized element logging data; a reasonable multi-mineral volume model is established based on the mineral analysis results of core thin section identification; then the element response equation is established with reference to the conventional logging optimization response equation, and combined with the standardized element logging data, the element characteristic parameters of different mineral components are obtained by multivariate linear regression; finally, the standardized element logging and conventional logging data are combined for comprehensive optimization inversion to obtain the mineral content and porosity inversion results.

[0128] Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to replace some of the technical features therein. Any modification, replacement, and improvement made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A comprehensive optimization inversion method for logging and logging of altered igneous rock minerals, characterized by: The following steps are involved: Step 1: Establish a standardization method for element logging data, and evaluate the reliability of the standardized element logging data in combination with the core major element analysis data; Step 2: Based on the core identification data, a multi-mineral volume model of the altered igneous formation is established taking into account the altered clay minerals and porosity; Step 3: Refer to the conventional logging optimization inversion method to establish the element response equation, and use the multivariate regression method to obtain the element characteristic values ​​of different minerals; Step 4: Based on conventional logging data and standardized element logging data, conduct comprehensive inversion of minerals and porosity of the actual altered igneous rock formation; The step 1 of establishing the element logging data standardization method and evaluating the data reliability comprises the following steps: Step 1 Element logging data standardization method: The element logging data include the contents of 17 elements, including sodium (Na), magnesium (Mg), aluminum (Al), silicon (Si), phosphorus (P), sulfur (S), chlorine (Cl), potassium (K), calcium (Ca), barium (Ba), titanium (Ti), vanadium (Vr), manganese (Mn), iron (Fe), nickel (Nr), strontium (Sr) and zirconium (Zr), while the major element analysis of the core includes the relative percentage content data of 10 oxides, including sodium oxide (Na2O), magnesium oxide (MgO), aluminum oxide (Al2O3), silicon dioxide (SiO2), phosphorus oxide (P2O5), potassium oxide (K2O), calcium oxide (CaO), titanium oxide (TiO2), manganese oxide (MnO) and iron oxide (Fe2O3); The principle of the element logging standardization method is: select 10 elements corresponding to the core oxides in the element logging data, namely sodium (Na), magnesium (Mg), aluminum (Al), silicon (Si), phosphorus (P), potassium (K), calcium (Ca), titanium (Ti), manganese (Mn), and iron (Fe) for standardization, and the remaining 7 elements are not involved in the standardization; The calculation formula of the relative content of the 10 elements after standardization is as follows: Where W X is the relative percentage content of X element in mud logging, in %; W LX is the logging data corresponding to the X element; ∑W LX It is the sum of logging data of 10 elements: Na, Mg, Al, Si, P, K, Ca, Ti, Mn, and Fe. Step 2 Obtain core element analysis data: First, based on the molar mass of different elements, the molar mass ratio of non-oxygen elements in different oxides is calculated: Where W MX is the molar mass percentage of element X; M X is the molar mass of element X, in g / mol; M XO is the molar mass of the oxide corresponding to element X, in g / mol; Then find the relative content of different elements: Where W X1 is the relative percentage content of X element in the core, in %; W XO is the relative content of core oxide corresponding to element X, in %; ∑W XO *W MX It is the sum of the contents of 10 non-oxygen elements, namely Na, Mg, Al, Si, P, K, Ca, Ti, Mn and Fe, in oxides; Step 3: Evaluate the reliability of standardized element logging data: Formula (1) was used to standardize the logging element data and compare them with the core element data; In step 3, the process of establishing the element response equation by referring to the conventional well logging optimization inversion method and obtaining the element characteristic values ​​of different minerals by using the multivariate regression method is as follows: Step 1 Establish the response equation of the element: Natural gamma (GR), acoustic transit time (AC), compensated neutron Compensation density (ρ b ) The four logging curve values ​​and the standardized element logging data are used as actual values ​​to calculate the volume content of pore fluid through the optimization algorithm. As well as the volume content of each mineral component in different volume models, the corresponding response equation is: Well logging response equation: Element response equation: W E =∑W Ei *W i (i=1,2,3...N) (5) In the formula, i represents the i-th component; represents the apparent neutron porosity value of the i-th component; ρ i represents the density value of the i-th component; AC i Represents the acoustic time difference of the i-th component; GR i represents the natural gamma value of the i-th component; W E W is the weight percentage content of each element after standardization; Ei is the element characteristic value of the i-th component; W i is the weight percentage content of the i-th component; Because the sum of the pore volume and the volume content of each mineral in the volume model is 1, an additional equilibrium equation is added: Thus, equations (4) to (6) form a set of equations (7), based on which the optimal solution is obtained using two optimization algorithms; therefore, the optimization comprehensive inversion problem is transformed into the problem of selecting an appropriate algorithm to obtain the optimal solution of the following set of equations: Where T is the actual data set, R is the logging and element logging response parameters of the i-th component, V i is the volume of the i-th component; The optimization method is based on the volume model. Therefore, the weight percentage data of element logging needs to be converted between weight percentage and volume percentage using the density of different minerals when participating in the optimization solution. Formula (8) to Formula (10) are the conversion formulas for converting mineral weight content into volume content: Where V mineral The volume of a mineral, W mineral is the weight percentage of a mineral; ρ mineral is the density of a certain mineral; is the percentage of minerals in the skeleton; is the volume percentage of a mineral in the entire rock; is the rock porosity; the volume content of minerals in the rock can be converted through formulas (8) to (10); Step2 Multiple regression calculation of element eigenvalues: Multiple linear regression analysis was performed with the element logging data of different minerals as independent variables and the corresponding weight contents of various minerals as dependent variables. i The general formula of the multiple linear regression model is: y=β1x1+β2x2+β3x3……+β n x n +e (11) In the formula, y is the dependent variable, representing the element content; β1, β2, ... β n are regression coefficients; x1, x2, ... x n is the independent variable, representing the content of different minerals; ε is the residual, which obeys the normal distribution N(0,σ 2 ); Establish regression equations between elements and minerals according to different mineral components; By the following formula: IN N =∑β X IN X (12) Where W X Represents the weight percentage content of the Xth mineral component in the rock mineral model; β X is the regression coefficient of the Xth mineral component; W N is the value of element N in the standardized element logging data; It can be obtained that the element characteristic parameter values ​​of a certain element (N) contained in different minerals.

Images