A rapid detection method for the calorific value of coal

Abstract

The present application provides a method for rapid detection of the calorific value of coal, which includes the following steps: S1. Select multiple samples as a fitting set and a validation set respectively, and press the powder of each sample into a tablet; S2. Use an X-ray fluorescence spectrometer to detect each sample in the fitting set to obtain the C, O and S content; and simultaneously determine the true calorific value Q of each sample; S3. Based on the data in step S2, fit a calorific value calculation model: Q=b+K1*C+K2*(O‑S), where b, K1 and K2 are coefficients; S4. Use an X-ray fluorescence spectrometer to detect each sample in the validation set to obtain the C, O and S content, substitute them into the calorific value calculation model, and obtain a predicted calorific value; measure the true calorific value of each sample in the validation set, compare the predicted calorific value with the corresponding true calorific value, and determine that the error range is within a reasonable range, which indicates that the calorific value calculation model is feasible. The present application uses an X-ray fluorescence spectrometer to quickly detect the content of elements related to heat in coal, and further establishes a calorific value calculation model through multiple linear regression analysis, so that the calorific value of coal can be measured quickly and easily.

Description

Technical Field

[0001] The present application relates to a method for quickly detecting the calorific value of coal. Background Art

[0002] Coal is a vital energy source in my country. In 2020, my country's coal production accounted for 51% of global production, and thermal power still accounted for approximately 60% of my country's power generation structure in 2019. Furthermore, since my country's coal production is primarily located in the western and northwestern regions, low-cost seaborne imported coal is an advantageous option for the eastern coastal areas with high energy demand. my country is not only the world's largest coal producer, but also the largest importer and consumer.

[0003] With the increasingly stringent environmental protection requirements in recent years, the country advocates the use of clean coal, which puts forward requirements for the measurement of the calorific value of imported and domestic coal. At the same time, for enterprises that purchase coal and customs departments that import coal, the calorific value of coal is one of the most basic requirements for identifying coal quality. It is also an important indicator for pricing in coal transactions. Therefore, rapid detection of the calorific value of coal is an important demand of the coal industry.

[0004] Conventional coal calorific value determination uses an oxygen bomb calorimeter, a method that requires complex pre-processing and a lengthy testing cycle. Consequently, conventional methods are inadequate for rapid on-site testing and cannot provide a rapid means for guiding coal production, circulation, importation, grading, and regulatory oversight. Traditional Mendeleev models and other rapid coal calorific value estimation methods either lack accuracy or rely heavily on the precise quantification of elements like carbon, hydrogen, oxygen, and sulfur, which are time-consuming and labor-intensive, failing to meet the requirements for rapid testing.

[0005] Therefore, a rapid detection method for the existing coal calorific value is needed. Summary of the Invention

[0006] The present application provides a method for quickly detecting the calorific value of coal, which can quickly measure the calorific value of coal.

[0007] Specifically, the present application is implemented through the following technical solutions: a method for rapid detection of the calorific value of coal, which includes the following steps: S1. Select multiple coal samples as a fitting set, and select multiple coal samples as a verification set, and press the powder of each sample into tablets; S2. Use an X-ray fluorescence spectrometer to detect each sample in the fitting set to obtain the carbon content C, oxygen content O and sulfur content S of each sample in the fitting set; and at the same time use a known measurement method to measure the true calorific value Q of each sample in the fitting set; S3. According to the data in step S2, fit a calorific value calculation model: Q = b + K1*C + K2*(OS), where b, K1 and K2 are coefficients; S4. Use an X-ray fluorescence spectrometer to detect each sample in the verification set to obtain the carbon content C, oxygen content O and sulfur content S of each sample in the verification set, and substitute them into the fitted calorific value calculation model to obtain the predicted calorific value of each sample in the verification set; use a known measurement method to measure the true calorific value of each sample in the verification set; compare the predicted calorific value with the corresponding true calorific value, and determine that the error range is within a reasonable range, which indicates that the calorific value calculation model is feasible.

[0008] According to one embodiment of the present application, the value of coefficient b of the calculation model is between -50 and -48, the value of K1 is between 87 and 89, and the value of K2 is between 66 and 68.

[0009] According to one embodiment of the present application, the method further comprises step S2.1, using the coefficient of determination (R 2 ), squared error (SSE) and root mean square error (RMSE) were used to evaluate the established fitting models.

[0010] According to one embodiment of the present application, in step S2, based on the carbon content C, oxygen content O and sulfur content S of the fitting set samples and the actual calorific value, the correlation coefficients of the C content, the O content O minus the S content S and the calorific value are verified respectively, and it is determined that C is highly positively correlated with Q, and OS is highly negatively correlated with Q.

[0011] According to one embodiment of the present application, the calorific value calculation model is: Q = -48.33 + 88.58*C + 67.08*(OS).

[0012] According to one embodiment of the present application, an oxygen bomb calorimeter is used to measure the true calorific value of each sample in the fitting set and the validation set.

[0013] According to one embodiment of the present application, the X-ray fluorescence spectrometer is a single-wavelength excitation energy dispersive X-ray fluorescence spectrometer, which uses an Ag target or a Cr target and adopts a hyperbolic curved crystal for monochromatic focusing.

[0014] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 The spectrum and element peaks when the sample is detected using an X-ray fluorescence spectrometer.

[0016] Figure 2 This is a flow chart of the method for rapid detection of coal calorific value in this application. DETAILED DESCRIPTION

[0017] Exemplary embodiments are described in detail herein, with examples illustrated in the accompanying drawings. When the following description refers to the drawings, identical numerals in different figures represent identical or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of devices, systems, apparatus, and methods consistent with certain aspects of the present application.

[0018] Due to the wide variety of coal types and complex matrices, as well as the lack of standard sample types, it is impossible and rather complicated to correct for interference between elements and matrix effects using a large number of standard samples. The applicant has found through research that it is more practical to use a basic parameter library and a constructed mathematical model to correct for various effects. This application uses an X-ray fluorescence spectrometer and a mathematical model to quickly detect the calorific value of coal.

[0019] This application uses a single-wavelength excitation energy dispersive X-ray fluorescence spectrometer (HS-XRF) and a fully focusing hyperbolic curved crystal, which can quickly and accurately quantitatively detect carbon, oxygen, and sulfur elements related to the generation and absorption of heat in coal. It can meet the element content accuracy required for fitting the calorific value calculation model, and combined with regression analysis and multivariate fitting, the calorific value can be quickly determined.

[0020] The overall concept of this application's rapid coal calorific value detection method is to establish a calorific value calculation model, use an X-ray fluorescence spectrometer to detect the carbon, oxygen, and sulfur contents of coal with known calorific value, fit the calorific value calculation model to the carbon, oxygen, and sulfur contents, and then verify the feasibility of the calorific value calculation model. During use, the carbon, oxygen, and sulfur content of the coal to be tested is detected using an X-ray fluorescence spectrometer, and the calorific value is calculated by substituting the results into the calorific value calculation model.

[0021] The specific steps to establish a calorific value calculation model are as follows:

[0022] 1. Prepare samples

[0023] Several coal samples were selected as fitting sets for model fitting; several coal samples were selected as validation sets for algorithm verification.

[0024] The number of samples in the fitting set should theoretically be greater than 4, preferably 8, so that the calorific value range of the samples in the fitting set is as large as possible to cover the range of most samples. Similarly, to ensure the rationality of model validation, multiple samples with as wide a calorific value range as possible should be used for validation, preferably 12. The above sample powder is pressed into tablets with a pressure of 20 MPa and the pressure is maintained for 60 seconds. The sample tablets have a diameter of 30 mm and a thickness of about 4 mm.

[0025] The true calorific value of each sample in the fitting set and the validation set can be obtained by a known, relatively accurate but unsuitable for rapid detection measurement method, for example, using an oxygen bomb calorimeter to measure the calorific value of each sample in the fitting set and the validation set.

[0026] 2. Detect the content of C, S and O elements in the sample

[0027] Fitting and validation set samples were placed in a single-wavelength excitation energy-dispersive X-ray fluorescence spectrometer (HS-XRF) to analyze the three elements C, S, and O in each sample. For the light elements C and O, the HS-XRF uses an X-ray tube with Ag and Cr targets and a Ge hyperboloid crystal for monochromatic focusing. The monochromatic focused laser energy exceeds the excitation limits of C, S, and O, making it particularly effective for excitation of these two elements. The use of a highly sensitive silicon drift detector enables higher count rates, enabling quantitative analysis of C and O. Once the detection method is established for the HS-XRF, quantitative results for C, O, and S can be directly obtained for the samples tested, along with quantitative results for other elements in the sample.

[0028] Use C respectively ai , O ai 、S ai To represent the C, O, and S element content of the i-th sample in the fitting set, use Q ai The true dry basis high calorific value of the i-th sample in the fitting set; respectively, C bj , O bj ,,S bj To represent the C, O, and S element content of the jth sample in the validation set, use Q bj Indicates the true dry high calorific value of the jth sample in the validation set.

[0029] 3. Determine the variables of the calorific value calculation model

[0030] For the 8 samples in the fitting set, the content of C element is expressed as C a Indicates the content of O element minus the content of S element, S, and is expressed as OS a Indicates the actual heat output by Q a According to the data of 8 samples in the fitting set, C a With Q a The correlation coefficient, and OS a With Q a The correlation coefficient of .

[0031] Correlation coefficient definition:

[0032] X and Y represent two variables used to compare correlations, here C a or OS a With Q a Cov(X, Y) is the covariance of X and Y, Var[X] is the variance of X, and Var[Y] is the variance of Y. The result is C a With Q a The correlation coefficient is 0.9448, which shows that C a With Q a There is a high positive correlation between OS a With Q a The correlation coefficient result is -0.8627, which shows that OS a With Q a There is a high negative correlation between the C a With OS a There is a linear relationship with the heat output.

[0033] 4. Fitting calorific value calculation model

[0034] Using the true heat Q of the fitting set a , carbon content C a , oxygen content O a and sulfur content S a , and the least square method was used for multiple linear regression fitting.

[0035] The calorific value calculation model obtained by multivariate linear regression fitting is:

[0036] Q c =b+K1*C+K2*(OS).

[0037] Among them, Q c is the set of fitted dry basis high calorific values, b, K1, and K2 are the coefficients obtained from the multivariate linear regression fit. The value of b ranges from -50 to -48, the value of K1 ranges from 87 to 89, and the value of K2 ranges from 66 to 68.

[0038] 5. Evaluation of calorific value calculation model

[0039] The established fitting model was then evaluated using the coefficient of determination (R2), sum of square error (SSE), and root mean square error (RMSE). The obtained coefficient of determination (R2) was 0.9873, the sum of square error (SSE) was 0.5843, and the root mean square error (RMSE) was 0.3419, indicating that the calculation results are highly reliable. The specific calculation formula is as follows:

[0040] The coefficient of determination (R2) is defined as:

[0041] Among them, Q bi is the predicted calorific value of the i-th sample, Q ai is the true calorific value of the i-th sample, and n is the number of samples.

[0042] The sum of squared error (SSE) is defined as:

[0043] The root mean square error (RMSE) is defined as:

[0044] The closer R2 is to 1, the better the linearity of the fitting model is; the closer SSE and RMSE are to 0, the smaller the deviation between the predicted results and the actual calorific value is.

[0045] 6. Verify the heat calculation model

[0046] For the 12 samples in the validation set, single wavelength excitation energy dispersive X-ray fluorescence spectrometer was used to detect the quantitative results of C, O and S elements in each sample of the validation set, namely C bj 、S bj , O bj It is understood that this step can be prepared in advance. bj 、S bj , O bj Substitute the calorific value calculation model obtained by multivariate linear regression fitting above to calculate the calorific value and obtain the predicted calorific value.

[0047] Table 1 C, S, O content test data of validation set samples

[0048]

[0049] Comparing the predicted calorific value calculated based on the calorific value calculation model with the actual calorific value, the relative deviation is shown in the following table:

[0050] Table 2 Relative deviation of validation set samples

[0051]

[0052] For the 12 samples in the validation set, the relative deviation between the calculated calorific values (predicted calorific values) and the calorimeter-measured values (true calorific values) was within 0.3% to 6%, a relatively low relative deviation. This indicates good consistency between the calculated results and the calorimeter-measured results, thus demonstrating the accuracy of the calorific value calculation model.

[0053] At the same time, the classic t-test method was used to compare the two analysis methods. The data summary results are shown in Table 2 below:

[0054] Table 3 T-test data of the difference between the calculated and measured calorific value of the same sample

[0055]

[0056] in

[0057]

[0058]

[0059] Q ak is the calorific value of all 20 samples measured by known accurate method, Q bk The calorific value was measured for all 20 samples using this method.

[0060] The results show that: according to T 统计量 <T 95％，n-1 It can be seen that there is no significant difference between the results obtained by the above calorific value calculation model and the measurement results of the calorimeter, which proves that the calorific value calculation model is more accurate and can be used to quickly test the calorific value of actual samples.

[0061] This application establishes a calorific value calculation model based on the carbon, oxygen, and sulfur contents of coal. During testing, simply illuminate the sample using an X-ray fluorescence spectrometer. The system then feeds the measured carbon, oxygen, and sulfur contents into the calorific value calculation model, allowing for a simple and quick calculation of the sample's calorific value.

[0062] The single-wavelength excitation energy dispersive X-ray fluorescence spectrometer used in this application has high sensitivity and low detection limit. It can quantitatively analyze the content of light elements such as carbon and oxygen. Combined with multivariate linear regression fitting, it can fill the gap in rapid detection methods on the market, and provide a new technical route for the safe processing and smooth circulation of coal. It is of great significance to ensure the rapid circulation of coal in the market and the rapid detection of imported coal.

[0063] The above is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

Claims

1. A method for rapid detection of coal calorific value, characterized in that: It includes the following steps: S1. Select multiple coal samples as the fitting set and multiple coal samples as the validation set, and compress the powder of each sample into tablets; S2. Use X-ray fluorescence spectrometry to measure each sample in the fitting set to obtain the carbon content C, oxygen content O, and sulfur content S of each sample in the fitting set; and simultaneously measure the true calorific value Q of each sample in the fitting set using a known determination method; S3. Based on the data in step S2, a multivariate linear regression fit is performed using the least squares method to fit the calorific value calculation model: Q = b + K1*C + K2*(OS), where b, K1, and K2 are the coefficients obtained from the multivariate linear regression fit; S4. Use an X-ray fluorescence spectrometer to measure each sample in the validation set to obtain the carbon content C, oxygen content O, and sulfur content S of each sample in the validation set. Substitute these values into the fitted calorific value calculation model to obtain the predicted calorific value of each sample in the validation set; and use a known determination method to measure the actual calorific value of each sample in the validation set. Compare the predicted calorific value with the corresponding actual calorific value and make sure that the error range is within a reasonable range, which means that the calorific value calculation model is feasible; The X-ray fluorescence spectrometer is a single-wavelength excitation energy dispersive X-ray fluorescence spectrometer, which uses an Ag target or a Cr target and adopts a hyperbolic curved crystal for monochromatic focusing.

2. The rapid detection method according to claim 1, characterized in that The values of coefficient b of the calculation model are between -50 and -48, the values of K1 are between 87 and 89, and the values of K2 are between 66 and 68.

3. The method for rapid detection of calorific value of coal according to claim 1, characterized in that: Also included is step S2.1, using the coefficient of determination (R 2 ), squared error (SSE) and root mean square error (RMSE) were used to evaluate the established fitting models.

4. The method for rapid detection of coal calorific value according to claim 1, characterized in that: In step S2, based on the carbon content C, oxygen content O, and sulfur content S of the fitting set samples and the actual calorific value, the correlation coefficients of the C content, the O content O minus the S content S, and the calorific value are verified respectively, and it is determined that C is highly positively correlated with Q, and OS is highly negatively correlated with Q.

5. The method for rapid detection of coal calorific value according to claim 1, characterized in that: The calorific value calculation model is: Q = -48.33 + 88.58*C + 67.08*(OS).

6. The method for rapid detection of coal calorific value according to claim 1, characterized in that: The true calorific value of each sample in the fitting set and the validation set was measured using an oxygen bomb calorimeter.

Images