Method for estimating grade of associated cobalt in skarn type iron ore

Abstract

The invention discloses a grade estimation method for associated cobalt in skarn type iron ore. The method comprises the following steps: collecting a sample, analyzing element content, and storing an analysis result into a data set; preprocessing the analysis data; element correlation analysis is carried out, and strong correlation elements are selected; a linear fitting formula between the cobalt grade and the strong correlation element grade is obtained, a residual sum of squares and a decision coefficient between a linear fitting line and the strong correlation element grade data point are calculated, the residual sum of squares is minimized, the closer the decision coefficient is to 1, the optimal linear fitting formula in the strong correlation element grade variable range is selected, and the optimal linear fitting formula is calculated; taking as a cobalt grade estimation formula; and estimating the cobalt grade of the main ore body through a cobalt grade estimation formula. According to the method, accurate estimation of the cobalt element grade in the mining area is realized, and basic data is provided for spatial distribution and grade grading of cobalt elements in the mining area.

Description

Technical Field

[0001] This invention relates to the field of mining technology, specifically to a method for estimating the grade of associated cobalt in skarn-type iron ore. Background Technology

[0002] Cobalt, as a strategically critical mineral, is a mineral resource urgently needed by emerging strategic industries such as new energy. Large quantities of cobalt are found as an associated enrichment in skarn-type iron ore deposits, but the development and utilization of these cobalt resources have not received sufficient attention.

[0003] Generally, for cobalt-rich skarn-type iron ore deposits that have been explored and are currently being mined, only the basic analytical elements TFe and S are measured (mFe is measured in some mines). Cobalt, as an associated element, has only undergone grade analysis in a small number of composite samples, but detailed cobalt grade determination has not been carried out, and it is impossible to determine the richness and poorness distribution and grade classification of cobalt in three-dimensional space. This is detrimental to the development and utilization of cobalt. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention proposes a method for estimating the grade of associated cobalt in skarn-type iron ore. This method can accurately estimate the cobalt grade in the mining area based on existing mine data, providing fundamental data for the spatial distribution and grade classification of cobalt in the mining area.

[0005] To achieve the above objectives, this invention designs a method for estimating the grade of associated cobalt in skarn-type iron ore, which is characterized by including the following steps: S1) Collect samples from the main ore bodies in the mining area, perform elemental analysis on the samples, and store the analysis results in the dataset; S2) Preprocess the analysis data in the dataset to remove outlier data; S3) Perform elemental correlation analysis using the preprocessed analytical data, select other elements that are strongly correlated with cobalt, and obtain the grade data of the strongly correlated elements as variables in the cobalt grade estimation formula in step S4) below. S4) Using the least squares method, the cobalt grade is linearly fitted using the grade data of strongly correlated elements in the sample. The linear fitting formula between the cobalt grade and the grade of strongly correlated elements is obtained. The sum of squared residuals and the coefficient of determination between the linear fitting line and the grade data points of strongly correlated elements are calculated. The optimal linear fitting formula within the range of strongly correlated element grade variables is selected as the cobalt grade estimation formula, with the minimum sum of squared residuals and the coefficient of determination being as close to 1 as possible. The formulas for calculating the sum of squared residuals and the coefficient of determination are as follows: In the formula, SSE represents the sum of squared residuals. x i This indicates the grade data of strongly correlated elements. This represents the cobalt grade data predicted by the linear fitting line. This represents the average value of the grade data for strongly correlated elements. R² represents the coefficient of determination. SST represents the sum of squared deviations; S5) The cobalt grade of the main ore body is estimated using the cobalt grade estimation formula.

[0006] Furthermore, in S1), the sampling locations are at different parts of the main ore bodies in the study mining area, the samples include samples of different grades, and the sampling quantity is no less than 200 pieces.

[0007] Furthermore, in S1), the steps for analyzing the elemental content of the sample include: first, crushing and sieving the sample, and then performing X-ray fluorescence spectroscopy and inductively coupled plasma mass spectrometry analysis to obtain a dataset of the elemental grades of TFe, mFe, S, and Co.

[0008] Furthermore, in S2), the data preprocessing method is as follows: S21) Perform descriptive statistics on the collected sample test results, then plot a histogram to observe the overall data characteristics; S22) Identify outlier data, interpret the outlier data, and selectively remove outlier data.

[0009] Furthermore, in S3), the method for selecting strongly correlated elements is as follows: S31) Perform correlation analysis on the elements in the dataset and calculate the correlation coefficient between two elements. The correlation coefficient is obtained by the following formula: r In the formula, r represents the correlation coefficient, with values ​​ranging from -1 to 1. The closer r is to 1, the stronger the correlation. Negative values ​​represent negative correlation, and positive values ​​represent positive correlation. y i This indicates the cobalt grade data in the sample. x i This indicates the grade data of other elements; S32) The other elements corresponding to the calculated maximum correlation coefficient are taken as the strongly correlated elements.

[0010] Further, in S4), the linear fitting formula is: =a+bx i +cx i2 and / or =a+bx i .

[0011] In the formula, x represents the cobalt grade data predicted by the linear fitting line. i This represents the grade data of strongly correlated elements in the linear fitting line, where a, b, and c represent constants.

[0012] Furthermore, in S5), the steps for estimating the cobalt grade of the main ore body using the cobalt grade estimation formula are as follows: S51) Construct three-dimensional solid models and block models of the main ore bodies in the mining area based on all exploration projects in the mining area; S52) Obtain the grade data of strongly correlated elements in each block of the block model, calculate the grade value of cobalt in each block using the cobalt grade estimation formula, and then perform cobalt grade classification, spatial distribution and reserve estimation in the ore body based on the calculated grade.

[0013] The advantages of this invention are: 1. This invention is based on existing exploration data and basic analytical elements. It collects samples from a small number of representative locations, determines the content of cobalt and the original basic analytical elements, and selects elements that are strongly correlated with cobalt by using correlation coefficient calculation method. The grade data of these strongly correlated elements are used as estimation variables. The optimal fitting curve method is used to estimate the grade of cobalt, which is largely missing in the mining area. Based on this, the cobalt resource quantity of the mining area can be estimated and the spatial distribution law of cobalt can be studied. This avoids the problems of missing and insufficient data, fills the gap in cobalt grade estimation technology for this type of mine, and avoids the need for mines to collect mineralized samples and determine cobalt content in all exploration projects.

[0014] Based on the optimal linear fitting formula, this invention uses the original basic analysis element results in the three-dimensional block model to estimate the cobalt element of all ore blocks in three-dimensional space. The application of this method can provide basic data for the development and utilization of cobalt resources in mines, save a lot of time and money in mine development, and also provide a relatively accurate cobalt grade estimate for areas in old mines where sampling is not possible. It has a certain generalization ability and can supplement cobalt grade data for many old mines and mines that have been explored. This invention provides a method for estimating the grade of associated cobalt in skarn-type iron ore. By using a correlation coefficient calculation method to select elements that are strongly correlated with cobalt, and using the grade data of these strongly correlated elements as estimation variables, the method of optimal fitting curve is used to estimate the grade of cobalt, which is largely missing in the mining area. This achieves an accurate estimation of the cobalt grade in the mining area and provides basic data for the spatial distribution and grade classification of cobalt in the mining area. Attached Figure Description

[0015] Figure 1 This is a flowchart of the present invention; Figure 2 This is a sample distribution diagram of a certain iron ore mine at -517m level in the example; Figure 3 This is a sample distribution diagram of a certain iron ore mine at -535m level in the example; Figure 4 This is the histogram used in the embodiment; Figure 5 a~5d are graphs of different fitting equations and line graphs with the true values ​​in the examples; Figure 6 a~6d are the piecewise linear fitting graphs and line graphs with the true values ​​in the examples. Detailed Implementation

[0016] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] In the description of this invention, it should be understood that the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention.

[0018] like Figure 1 As shown, the present invention discloses a method for estimating the grade of associated cobalt in skarn-type iron ore, comprising the following steps: S1) Collect samples from the main ore bodies in the mining area, perform elemental analysis on the samples, and store the analysis results in the dataset.

[0019] Preferably, the sampling locations are different parts of the main ore bodies in the study mining area, the samples include samples of different grades, and the number of samples is not less than 200.

[0020] In this embodiment, as Figure 2 and 3 As shown, the sampling area comprises the main ore bodies III, V, and VI of a certain iron ore mine. Sampling locations are primarily in the -517m and -535m sections of ore bodies III and V, as well as core samples from Zk392 and Zk452 (mainly from ore body VI). The samples are dominated by massive and disseminated magnetite, interspersed with magnetite-mineralized skarn, granite, and garnet-bearing skarn.

[0021] Figure 2In the diagram, the numbers starting with "B" indicate the sample number, the red line represents the ore body boundary, the blue line represents the boundary between skarn and other geological formations, and the green line represents mining engineering. Ⅲ and Ⅴ indicate the ore body number, SK indicates skarn, and δμ indicates diorite porphyry.

[0022] Figure 3 In the diagram, the numbers starting with "B" represent the sample number, the red line represents the ore body boundary, and the black line represents the mining project; III and V represent the ore body number.

[0023] Specifically, the steps for analyzing the elemental content of the sample include: first, crushing and sieving the sample, and then performing X-ray fluorescence spectroscopy and inductively coupled plasma mass spectrometry analysis to obtain a dataset of the elemental grades of TFe, mFe, S, and Co.

[0024] S2) Preprocess the analysis data in the dataset to remove outlier data.

[0025] Specifically, the data preprocessing method is as follows: S21) Perform descriptive statistics on the collected sample test results, then plot a histogram to observe the overall data characteristics; S22) Identify outlier data, interpret the outlier data, and selectively remove outlier data.

[0026] In this embodiment, descriptive statistical analysis was performed on the data, as shown in Table 1. The grades of TFe, S, and Co were determined for all test samples, with the mFe grade measured for 97 samples. Statistical analysis of the test results showed that the average TFe grade was 35.37%, with a minimum of 5.45% and a maximum of 59.24%; the average mFe grade was 26.11%, with a minimum of 1.35% and a maximum of 55.28%; the average S grade was 4.27%, with a minimum of 0.5% and a maximum of 24.63%; and the average Co grade was 0.023%, with a minimum of 0.005% and a maximum of 0.43%.

[0027] Table 1. Descriptive statistics of TFe-mFe-S-Co Observe TFe-mFe-S-Co using histograms, such as Figure 3 As shown, the TFe and mFe grade data are close to a normal distribution, while the S and Co grade data show skewed distribution characteristics and are non-normally distributed data.

[0028] The abnormal data processing in this embodiment includes: First, the objective of this invention is to predict the Co element content in the ore within the iron ore mining area, therefore, the non-ore portion of the data with TFe (wt%) < 20 is removed. Second, statistics show that although there are high Co grade data (Co = 0.43%) in this sampling data, most data are below 0.10%, and only two sets of data have Co grade greater than 0.20%. Therefore, the data with a content greater than 0.20% are treated as extremely high value anomalies and removed.

[0029] S3) Perform elemental correlation analysis using the preprocessed analytical data, select other elements that are strongly correlated with cobalt, and obtain the grade data of the strongly correlated elements as variables in the cobalt grade estimation formula in step S4) below.

[0030] Specifically, the method for selecting strongly correlated elements is as follows: S31) Perform correlation analysis on the elements in the dataset and calculate the correlation coefficient between two elements. The correlation coefficient is obtained by the following formula: r In the formula, r represents the correlation coefficient, with values ​​ranging from -1 to 1. The closer r is to 1, the stronger the correlation. Negative values ​​represent negative correlation, and positive values ​​represent positive correlation. y i This indicates the cobalt grade data in the sample. x i This indicates the grade data of other elements.

[0031] S32) The other elements corresponding to the calculated maximum correlation coefficient are taken as the strongly correlated elements.

[0032] In this embodiment, the correlation coefficients between some elements calculated using the above correlation coefficient formula are shown in Table 2.

[0033] Table 2 Correlation Coefficients of Some Elements Table 2 shows that Co is highly correlated with S and (TFe-mFe), but not highly correlated with TFe and mFe. Further analysis shows that Co is not significantly correlated with TFe and mFe, but is positively correlated with (TFe-mFe), which represents iron in pyrite after magnetite removal.

[0034] Although Co and Fe have very similar ionic radii (0.0745 nm Vs Fe²⁺ 0.078 nm) and electronegativity, they can undergo isomorphous substitution. In skarn-type iron ores, Co mainly substitutes Fe²⁺ into sulfide lattices such as pyrite in an isomorphous manner, while it rarely enters Fe²⁺ in magnetite.

[0035] Statistical data show a moderate positive correlation between S and Co (r=0.65), which is highly consistent with the mineral occurrence pattern of Co in this deposit. X-ray diffraction and electron probe microanalysis confirmed that more than 85% of the cobalt in the study area replaced Fe²⁺ in pyrite (FeS₂) in an isomorphous manner, forming (Fe,Co)S₂ solid solution. Since pyrite mainly forms in the sulfide stage (mineralization temperature 200-300℃), while oxides such as magnetite (Fe₃O₄) have crystallized out in large quantities in the aforementioned mineralization stage, the S content and Co enrichment in the sulfide phase show a significant synchronous increasing trend.

[0036] Since Co mainly enters sulfides such as pyrite, statistical data show that Co has a relatively obvious correlation with S and (TFe-mFe). However, considering that most of the original exploration samples in the mine did not have magnetite grade (mFe) data measured, but S grade data is complete as a basic analytical sample.

[0037] Therefore, in this embodiment, element S is selected as a strongly correlated element, and the grade data of element S is used to estimate the grade data of element Co.

[0038] S4) Using the least squares method, the cobalt grade is linearly fitted using the grade data of strongly correlated elements in the sample. The linear fitting formula between the cobalt grade and the grade of strongly correlated elements is obtained. The sum of squared residuals and the coefficient of determination between the linear fitting line and the grade data points of strongly correlated elements are calculated. The optimal linear fitting formula within the range of strongly correlated element grade variables is selected as the cobalt grade estimation formula, with the sum of squared residuals being minimized and the coefficient of determination being as close to 1 as possible.

[0039] The formulas for calculating the sum of squared residuals and the coefficient of determination are as follows: In the formula, SSE represents the sum of squared residuals. x i This indicates the grade data of strongly correlated elements. This represents the cobalt grade data predicted by the linear fitting line. This represents the average value of the grade data for strongly correlated elements. R² represents the coefficient of determination. SST stands for Sum of Squared Deviations.

[0040] Specifically, the linear fitting formula is as follows: =a+bx i +cx i 2 and / or =a+bx i .

[0041] In the formula, x represents the cobalt grade data predicted by the linear fitting line. i This represents the grade data of strongly correlated elements in the linear fitting line, where a, b, and c represent constants.

[0042] In this embodiment, the S-Co fitting formula is used to estimate the Co grade. Specifically, two methods, S-Co univariate linear equation and multivariate linear equation, are selected to fit and predict Co, resulting in formula ① and formula ②.

[0043] The fitted equation is as follows: Co = 0.00493 + 0.00387×S ① Co = 0.01162 + 0.000817×S + 0.000186×S 2 ② like Figure 5 As shown in Figures a through 5d, 5a and 5b are univariate linear fitting equations, while 5c and 5d are multivariate linear fitting equations.

[0044] Formula ① has a sample size of 154, a residual sum of squares of 0.02683, an R² of 0.54035, and an adjusted R² of 0.53732. The fitted curve for Co element and the comparison curve between the actual and fitted values ​​are shown below. Figure 5 a、5b.

[0045] Formula ② has a sample size of 154, a residual sum of squares of 0.01600, an R² of 0.66275, and an adjusted R² of 0.65885. The fitted curve for Co element and the comparison curve between the true and fitted values ​​are shown below. Figure 5 c, 5d.

[0046] Formulas ① and ② both show good fitting results, but formula ② is significantly better.

[0047] To more accurately predict Co, formula ② is further optimized by segmenting S to obtain formulas ③ and ④: When S = 0 ~ 3, Co = 0.00334 + 0.00532 × S ③ When S > 3, Co = 0.02315 - 0.00253 × S + 0.000321304 × S 2 ④ like Figure 6 As shown in Figures a to 6d, 6a and 6b are the fitting curves when S = 0 to 3, and 6c and 6d are the fitting curves when S > 3.

[0048] Plot a line graph of predicted and actual values ​​for formulas ③ and ④. Figure 6 (b, 6d) The predicted values ​​and the actual values ​​in the figure are highly fitted, showing that the formula has good prediction results for Co grade in different intervals.

[0049] In this embodiment, the relevant parameter table of the above fitting equation is shown in Table 3.

[0050] Table 3. Parameters related to the fitted equation As shown in Table 3, Formula ② has the highest coefficient of determination R², which is 0.65885 after adjustment. The sum of squared residuals is also smaller than that of Formula ①. However, in the low sulfur range (0~3%), the predicted value of Co is 2-3 times larger than the actual value, resulting in a large error. Formula ③, on the other hand, re-estimates and fits Co when S is between 0 and 3, which makes up for the shortcomings of Formula ②. At the same time, in order to ensure the continuity between the segment points, the polynomial fitting of S-Co is re-performed in the part where S>3.

[0051] Based on the above analysis results, this embodiment determines to use formulas ③ and ④ to estimate the two S grade values ​​in segments. Taking all factors into consideration, formulas ③ and ④ for the S-Co segmentation are selected: When S = 0 ~ 3, Co = 0.00334 + 0.00532 × S ③ When S > 3, Co = 0.02315 - 0.00253 × S + 0.000321304 × S 2 ④ S5) Estimate the cobalt grade of the main ore body using the cobalt grade estimation formula. Specifically, the steps for estimating the cobalt grade of the main ore body using the cobalt grade estimation formula are as follows: S51) Construct three-dimensional solid models and block models of the main ore bodies in the mining area based on all exploration projects in the mining area; S52) Obtain the grade data of strongly correlated elements in each block of the block model, calculate the grade value of cobalt in each block using the cobalt grade estimation formula, and then perform cobalt grade classification, spatial distribution and reserve estimation in the ore body based on the calculated grade.

[0052] In this embodiment, the cobalt grade classifications of the main ore bodies No. III, No. V, and No. VI are shown in Table 4 below.

[0053] Table 4. Cobalt Grade Grading Table for Major Mining Areas This invention provides a method for estimating the grade of associated cobalt in skarn-type iron ore. By using a correlation coefficient calculation method to select elements that are strongly correlated with cobalt, and using the grade data of these strongly correlated elements as estimation variables, the method of optimal fitting curve is used to estimate the grade of cobalt, which is largely missing in the mining area. This achieves an accurate estimation of the cobalt grade in the mining area and provides basic data for the spatial distribution and grade classification of cobalt in the mining area.

[0054] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A method for estimating the grade of associated cobalt in skarn-type iron ore, characterized in that, Includes the following steps: S1) Collect samples from the main ore bodies in the mining area, perform elemental analysis on the samples, and store the analysis results in the dataset; S2) Preprocess the analysis data in the dataset to remove outlier data; S3) Perform elemental correlation analysis using the preprocessed analytical data, select other elements that are strongly correlated with cobalt, and obtain the grade data of the strongly correlated elements as variables in the cobalt grade estimation formula in step S4) below. S4) Using the least squares method, the cobalt grade is linearly fitted using the grade data of strongly correlated elements in the sample. The linear fitting formula between the cobalt grade and the grade of strongly correlated elements is obtained. The sum of squared residuals and the coefficient of determination between the linear fitting line and the grade data points of strongly correlated elements are calculated. The optimal linear fitting formula within the range of strongly correlated element grade variables is selected as the cobalt grade estimation formula, minimizing the squared residuals and making the coefficient of determination as close to 1 as possible. The formulas for calculating the sum of squared residuals and the coefficient of determination are as follows: In the formula, SSE represents the sum of squared residuals. x i This indicates the grade data of strongly correlated elements. This represents the cobalt grade data predicted by the linear fitting line. This represents the average value of the grade data for strongly correlated elements. R² represents the coefficient of determination. SST represents the sum of squared deviations; S5) The cobalt grade of the main ore body is estimated using the cobalt grade estimation formula.

2. The method for estimating the grade of associated cobalt in skarn-type iron ore according to claim 1, characterized in that: In S1), the sampling locations are different parts of the main ore bodies in the study mining area, and the samples include samples of different grades, with a sampling quantity of no less than 200 pieces.

3. The method for estimating the grade of associated cobalt in skarn-type iron ore according to claim 2, characterized in that: In S1), the steps for analyzing the elemental content of the sample include: first, crushing and sieving the sample, and then performing X-ray fluorescence spectroscopy and inductively coupled plasma mass spectrometry analysis to obtain a dataset of the elemental grades of TFe, mFe, S, and Co.

4. The method for estimating the grade of associated cobalt in skarn-type iron ore according to claim 1, characterized in that, In S2), the data preprocessing method is as follows: S21) Perform descriptive statistics on the collected sample test results, then plot a histogram to observe the overall data characteristics; S22) Identify outlier data, interpret the outlier data, and selectively remove outlier data.

5. The method for estimating the grade of associated cobalt in skarn-type iron ore according to claim 4, characterized in that, In S3), the method for selecting strongly correlated elements is as follows: S31) Perform correlation analysis on the elements in the dataset and calculate the correlation coefficient between two elements. The correlation coefficient is obtained by the following formula: r In the formula, r represents the correlation coefficient, with values ​​ranging from -1 to 1. The closer r is to 1, the stronger the correlation. Negative values ​​represent negative correlation, and positive values ​​represent positive correlation. y i This indicates the cobalt grade data in the sample. x i This indicates the grade data of other elements; S32) The other elements corresponding to the calculated maximum correlation coefficient are taken as the strongly correlated elements.

6. The method for estimating the grade of associated cobalt in skarn-type iron ore according to claim 1, characterized in that: In S4), the linear fitting formula is: =a+bx i +cx i 2 and / or =a+bx i ; In the formula, x represents the cobalt grade data predicted by the linear fitting line. i This represents the grade data of strongly correlated elements in the linear fitting line, where a, b, and c represent constants.

7. The method for estimating the grade of associated cobalt in skarn-type iron ore according to claim 6, characterized in that, In S5), the steps for estimating the cobalt grade of the main ore body using the cobalt grade estimation formula are as follows: S51) Construct three-dimensional solid models and block models of the main ore bodies in the mining area based on all exploration projects in the mining area; S52) Obtain the grade data of strongly correlated elements in each block of the block model, calculate the grade value of cobalt in each block using the cobalt grade estimation formula, and then conduct further research on the grade classification, spatial distribution and reserve estimation of cobalt in the ore body based on the calculated grade.

Images