JP7620198B2 - A method for estimating the bulk fracture strength of coke. - Google Patents

Description

The present invention relates to a method for estimating the volumetric fracture strength of coke.

The drum strength of coke is known as an important control index for blast furnace operation, and the drum strength DI ¹⁵⁰ ₁₅ is generally defined as the mass fraction of 15 mm or more after 150 revolutions of a sampled coke in a drum test machine. Fine particles of 15 mm or less are generated by surface or volumetric destruction, and it is known that these destructions are caused by cracks, poor adhesion of non-melted components, flat pores, etc. (see Non-Patent Document 1). It is also known that surface destruction and volumetric destruction correspond to fine particles of 6 mm or less and 15-6 mm, respectively (see Non-Patent Document 2). Therefore, the drum strength DI ¹⁵⁰ ₁₅ of coke is defined as the sum of the surface destruction strength DI ¹⁵⁰ ₆ corresponding to fine particles of 6 mm or less and the volume destruction strength DI ¹⁵⁰ _6-15 corresponding to fine particles in the range of 6-15 mm.

Patent Document 1 describes a method for estimating the amount of fine coke generated by volumetric destruction of coke from the relationship between the resolidification temperature of a coal blend and the amount of fine coke generated having a size of more than 6 mm and not more than 15 mm after a coke strength test.
Patent Document 2 describes a method for blending high- and low-rank coals and estimating the amount of volume-destroyed coke fines based on coal properties measured before carbonization for each blended coal, using the average reflectance of the coal blend, the blending ratio of low-rank coal, the bulk density of the coal blend, and the temperature of a coke oven.

Patent Document 3 describes a method for estimating the volumetric crushing strength DI ¹⁵⁰ _6-15 based on a linear equation that defines the correlation between the volatile matter VM of coal and the volumetric crushing strength DI ¹⁵⁰ _6-15 .
Patent Document 4 describes a method for estimating the volumetric rupture strength DI ¹⁵⁰ _6-15 based on the relationship (α2) between the measured shrinkage rate of a coal blend and the volumetric rupture strength DI ¹⁵⁰ _6-15 of coke obtained from the coal blend.

Japanese Patent Application Publication No. 9-263764 JP 2005-194358 A JP 2007-291262 A JP 2020-094200 A

Peirce, T. J., et al. (1980). "Coke breakage behavior in relation to its structure." Journal of Physics D: Applied Physics 13(6): 953. Arima, T., et al. (1992). "Mechanism of Coke Size Degradation by Mechanical Impact." Tetsu-to-Hagane 78(7): 1101-1108.

However, the estimation methods of Patent Documents 1 to 4 can easily and accurately estimate the volumetric fracture strength DI ¹⁵⁰ _6-15 within the range of conditions (e.g., blending conditions, bulk density, furnace temperature) that are the basis of the estimation, but for coke that does not fall within the range of the conditions, the estimation accuracy decreases, so it is necessary to determine the estimation formula again based on the changed conditions. Since it is necessary to conduct experiments or the like again when reconstructing the estimation formula, there are cases where the volumetric fracture strength DI ¹⁵⁰ _6-15 cannot be estimated quickly.

Therefore, an object of the present invention is to provide a method for estimating the volumetric rupture strength DI ¹⁵⁰ _6-15 with high accuracy and quickly, regardless of the blending conditions of a coal blend.

In order to solve the above problems, the method for producing coke according to the present invention is characterized by having: (1) a first step of performing imaging processing by X-ray CT on coke obtained by carbonizing coal in a test coke oven to obtain information on the crack shape of the coke; a second step of calculating the area of crack surfaces whose crack width is equal to or less than a threshold value by image analysis of the obtained information on the crack shape; a third step of performing the first and second steps for multiple coals having different properties to calculate the area of crack surfaces whose crack width is equal to or less than a threshold value for each coke, and then constructing an estimation formula for estimating the volumetric fracture strength based on the calculated area of each crack surface and the volumetric fracture strength actually measured for each coke; and a fourth step of estimating the volumetric fracture strength from the area of the crack surfaces of the coke obtained by performing the first and second steps using coal for producing coke to be used in a blast furnace and the estimation formula obtained in the third step.

(2) The method for estimating the volumetric fracture strength of coke described in (1) above, characterized in that the threshold value is set so that the correlation coefficient between the actual measured value of the volumetric fracture strength and the estimated value using the estimation equation constructed in the third step is 0.85 or more.

According to the present invention, the volumetric fracture strength DI ¹⁵⁰ _6-15 can be estimated quickly and with high accuracy simply by obtaining information about the crack shape of the coke cake using X-ray CT.

FIG. 11 is a transition diagram illustrating a method of image processing. FIG. 13 is a diagram for explaining estimation of the volumetric fracture strength DI ¹⁵⁰ _6-15 using the area of a crack surface with a small crack width. 13 is a graph showing the degree of correlation between the estimation result and the test result of the drum test (Example). 13 is a graph showing the degree of correlation between the estimation result and the test result of the drum test (comparative example);

As described above, the drum strength DI ¹⁵⁰ ₁₅ of coke can be defined as the sum of the surface fracture strength DI ¹⁵⁰ ₆ and the volume fracture strength DI ¹⁵⁰ _{6-15. This embodiment proposes a method for estimating the volume fracture strength DI 150 6-15} ^with _high accuracy and quickly. The method for estimating the volume fracture strength DI ¹⁵⁰ _6-15 of this embodiment will be described in detail below, divided into the following steps S1 to S4.

(Step S1)
The coke cake obtained by carbonizing coal in the test coke oven is imaged, for example, by a three-dimensional medical X-ray CT scanner before being divided into coke blocks (i.e., while placed in a carbonization vessel). By imaging with a three-dimensional medical X-ray CT scanner, information on the crack shape of the coke cake can be obtained three-dimensionally. Instead of the three-dimensional medical X-ray CT scanner, a microfocus X-ray CT scanner can also be used. However, since the three-dimensional medical X-ray CT scanner can obtain the crack shape earlier than the microfocus X-ray CT scanner, it is preferable to use the three-dimensional medical X-ray CT scanner.

It is also possible to image the coke cake using two-dimensional X-ray CT. In this case, by increasing the number of images taken, it is possible to grasp the crack shape of the entire coke cake. It is also possible to grasp the crack shape by observation with a microscope, but this requires work such as filling with resin and polishing, and is time-consuming to process, so it is outside the scope of this invention.

(Step S2)
Next, the information on the crack shape acquired in step S1 is subjected to image analysis to acquire the crack surface, and the area and width of the crack surface are calculated. For the image analysis, known software (for example, Avizo, known as three-dimensional image analysis software) can be used. Specifically, the image analysis can be performed according to the procedure shown in the following steps S2-1 to 2-4. In this embodiment, a three-dimensional medical X-ray CT with a resolution of about 1 mm can be used.

(Regarding step S2-1)
In step S2-1, the captured image is subjected to image processing (for example, binarization processing based on the difference in brightness value) to separate the coke portion from the crack (space) portion. By the binarization processing, for example, the coke portion can be separated into black and the crack portion into white. However, the color-coding method is not limited to this, and other colors can be used. FIG. 1(a) shows a two-dimensional schematic representation of the coke and crack after separation, but in reality, three-dimensional image data is obtained. Note that FIG. 1(b) to FIG. 1(d) cited in the following description are also two-dimensional schematic diagrams, but in reality, three-dimensional image data is obtained.

(Regarding step S2-2)
In step S2-2, a separation process is performed to further separate the coke portion into chunks. Here, as shown in FIG. 1B, the coke block A appears to be one block at first glance, but has a crack x inside, and when an impact is applied, it splits into two starting from the crack x. Therefore, even if a coke block appears to be one block at first glance, if it has a crack x, it is considered to split when it receives an impact, and is determined to be a different block. For the separation process in this step, a Watershed algorithm (see non-patent document: Beucher and Meyer. The morphological approach to segmentation: the watershed transformation. Mathematical morphology in image processing; 34, 433-81 (1993)) that performs processing to identify each particle in an image as a separate entity can be used.

Here, if the width of the crack x is extremely small, the coke lump will not break even if it is subjected to an impact. Therefore, in order to consider only the cracks that affect the breakup when it is subjected to an impact, cracks with a predetermined width or less that do not contribute to the breakup when it is subjected to an impact are filled by image processing. This predetermined width can be obtained by an experiment or the like. For example, a coke that has actually been subjected to a rotational impact is observed by X-ray CT or a microscope to identify multiple coke cracks remaining in the coke lump. Then, the crack with the largest crack width is identified from among these cracks, and the width of this identified largest crack can be set as the predetermined width. Here, as an example, the width that does not contribute to the breakup when it is subjected to a rotational impact can be set to 0.1 mm. The rotational impact can be applied, for example, by a rotating drum, and the volumetric fracture strength DI ¹⁵⁰ _6-15 can be evaluated by setting the rotation speed to 150 revolutions.

(Step S2-3)
The expansion process is uniformly performed for each separated coke lump, and image processing is performed to fill in the cracks, thereby extracting the boundaries between the coke lumps separated in step S2-2 (see FIG. 1(c)).

(Step S2-4)
The boundary extracted in step S2-3 is compared with the crack location in step S2-1, and a surface that is the boundary and located at the crack location is extracted as a crack surface CF, and a crack width P is calculated (see FIG. 1(d)). Here, the crack surface CF is a surface that extends along the crack toward the paper surface side of FIG. 1(d), and the crack width P is a distance between the coke lumps facing each other across the crack surface CF.
As shown in Fig. 2(a), when each crack surface CF is defined as crack surfaces CF1 to CF4, in this embodiment, the crack surfaces CF are divided into two sections according to the size of the crack width P. Fig. 2(b) is a diagram after the division, in which the crack surfaces CF1 to CF3 with a relatively large crack width P are shown by solid lines, and the crack surface CF4 with a relatively small crack width P is shown by a dashed line.

In this manner, in this embodiment, the crack surface CF is divided into two sections according to the size of the crack width P, and as described below, the volumetric fracture strength DI ¹⁵⁰ _6-15 is estimated based on the area of the crack surface CF with a relatively small crack width P. The reason for focusing on the area of the crack surface CF with a relatively small crack width P is as follows.

It is known that among the cracks formed in the coke, cracks with a relatively large crack width are related to the grain size of the coke, and cracks with a relatively small crack width are related to the destruction of the coke. Therefore, it is presumed that only cracks with a small crack width should be considered for the volume destruction of the coke. The inventors presumed that cracks with a crack width of a threshold value (X mm) or less affect the volume destruction from the prior art (Sakimoto et al. ISIJ International; 56(11), 1948-55 (2016)). Here, the threshold value (X mm) depends on the resolution of the X-ray CT, but since the example described later was 2.5 mm, it is presumed to be approximately several mm. Specifically, the threshold value (X mm) is set to a value that satisfies the estimation accuracy required in the linear equation described later. The required estimation accuracy is not particularly limited, but is preferably 0.85 or more when expressed in terms of the correlation coefficient (R ² ) described later. In other words, it is desirable to set the threshold value (X mm) so that the correlation coefficient between the estimated value by the linear expression and the actual measured value is 0.85 or more.

As described above, the resolution of the three-dimensional medical X-ray CT used in this embodiment is about 1 mm. However, if a three-dimensional medical X-ray CT with higher resolution is available, the crack widths below a threshold value (X mm) may be further divided into multiple sections, and a weighting process may be performed taking into account the influence of the crack width of each section, thereby estimating the volumetric fracture strength DI ¹⁵⁰ _6-15 .

Three-dimensional crack widths can be evaluated using the evaluation method shown in the non-patent literature (Hildebrand and Ruegsegger. A new method for the model-independent assessment of thickness in three-dimensional images. Journal of Microscopy; 185(1), 67-75 (1997)). Specifically, a method can be used in which the largest sphere that touches the boundary between the target three-dimensional region and an outside region is placed inside the target three-dimensional region, and the width of the target object can be measured from the diameter of the placed sphere. In the case of two dimensions, two-dimensional crack widths can be evaluated by placing a circle instead of a sphere in three dimensions, but in the case of two dimensions, the width may change even at the same location depending on the direction of the cross section, so three-dimensional evaluation is preferable.

(Step S2-5)
The area of the crack surface whose crack width is equal to or less than the threshold value (X mm) extracted in step S2-4 is normalized by the volume of the entire analysis region, which is the sum of the coke portion and the crack portion, to calculate S _{-X mm} .

(Step S3)
The above-mentioned steps S1 and S2 are carried out for each of a plurality of coals having different properties, and S _-Xmm is calculated for each coke. From the viewpoint of constructing an estimation equation, it is preferable that the volumetric fracture strength DI ¹⁵⁰ _6-15 of each coke is more dispersed. Therefore, it is preferable that the blending conditions (coal brand, blending ratio, bulk density, presence or absence of coke powder, etc.) and heating conditions (heating rate, ultimate temperature, heating time) of the blended coals used in the production of each coke are different from each other. In addition, the volumetric fracture strength DI ¹⁵⁰ _6-15 is obtained in advance by carrying out a drum test for each coke.
By the above-mentioned processing, a set of data consisting of S _-Xmm and the actual measured value of the volumetric fracture strength DI ¹⁵⁰ _6-15 is obtained the same number of times as the above-mentioned steps S1 and S2 are performed using a plurality of coals having different properties. By fitting these data to a linear function based on the least squares method, the following linear equation (1) is constructed as an estimation equation.

(Step S4)
The volumetric fracture strength DI 150 6-15 is estimated by substituting S _-Xmm of the coke (coke obtained by carbonizing coal for producing coke to be used in a blast furnace in a test coke oven) newly measured by carrying out steps S1 to ^S3 into the linear _equation (1).

According to the method of the present embodiment, the volumetric fracture strength DI ¹⁵⁰ _6-15 can be estimated simply by acquiring information on the cracks in the coke cake and calculating the area of the crack surface with a crack width equal to or less than a threshold value (X mm). Even if conditions such as the blending conditions of the coal or the presence or absence of coke powder are changed, there is no need to reconstruct the estimation formula (1), so the volumetric fracture strength DI ¹⁵⁰ _6-15 can be estimated by a quick and simple method. Furthermore, after constructing the estimation formula (1), a drum test for determining the volumetric fracture strength DI ¹⁵⁰ _6-15 is no longer necessary, so the effort and time required for the drum test can be reduced.

Next, the present invention will be specifically described with reference to examples. Coal blends having the properties shown in Table 1 were carbonized in a test coke oven to produce cokes C1 to C25. In Table 1, ΣTD (%) is the weighted average value of the total expansion coefficient TD of each single coal contained in the coal blend, and ΣVM (mass%) is the weighted average value of the volatile matter of each single coal contained in the coal blend. The size of the carbonization vessel of the test coke oven was set to oven width (W): 400 mm, oven length (L): 600 mm, and oven height (H): 420 mm. The charged bulk density (BD) of the coal was set to 850 (kg/m ³ ) or 900 (kg/m ³ ) on a dry basis. The final temperature of the test coke oven was set to 950°C, 1050°C, or 1150°C, and the carbonization time was set to 18.5 (hours) in each case. The reason why the average particle diameters of the coke fines C16 and C18 are expressed by one digit below, and the average particle diameters of the other coke fines are expressed by two digits below, is because the accuracy of the sieve meshes used to calculate the average particle diameter is different. The addition rate of the coke fines (mass%) is expressed as an additional number.

As a process corresponding to step S1 described above, after carbonization, each coke was placed in a carbonization vessel and a three-dimensional image was obtained using a three-dimensional medical X-ray CT. After the imaging, a drum test was performed to measure the volumetric fracture strength DI ¹⁵⁰ _6-15 .

The imaging conditions for the X-ray CT were a tube voltage of 120 kV, a tube current of 350 mA, and a slice pitch of 0.5 mm. The image size was 512 x 512 pixels. The image resolution was 1.132 mm/pixel. For the separation process (corresponding to step S2), the segmentation module provided by Avizoo was used.

That is, in the process corresponding to step S2-1, the coke portion and the crack (space) were separated by binarization. In the process corresponding to step S2-2, the coke portion was further separated into chunks using the Watershed algorithm. In this case, a parameter (marker extent) that determines the strength of the separation process was set to 1, which indicates strong separation, and coke chunks having cracks with a width of more than 0.1 mm were separated. In the process corresponding to step S2-3, an expansion process was performed on each separated coke chunk, and image processing was performed to fill the cracks. In the process corresponding to step S2-4, the boundary portion extracted in step S2-3 was compared with the crack portion in step S2-1, and the surface that is the boundary portion and located at the crack portion was extracted as a crack surface, and the crack width was calculated.

As a process corresponding to step S2-5, a crack width of 2.5 mm was set as a threshold, and for cracks with a crack width of 2.5 mm or less, the area of the crack surface was standardized by the volume of the entire analysis area, which was the sum of the coke portion and the crack portion (however, the area of the void between the coke and the distillation vessel was excluded from the standardized volume), to calculate S _{-2.5 mm} .
The threshold value of the crack width may change depending on the resolution of the three-dimensional medical X-ray CT scanner or the value of the correlation coefficient to be obtained. Therefore, the threshold value of the crack width in the present invention is not limited to 2.5 mm.

The above-mentioned process was carried out for each coke (corresponding to step S3). In this embodiment, for comparison, the area of the crack surface (regardless of the crack width) was also calculated by standardizing the volume of the entire analysis region obtained by adding up the coke portion and the crack portion (however, the region of the _void between the coke and the carbonization vessel was excluded from the volume to be standardized).

The volumetric fracture strength DI ¹⁵⁰ _6-15 , which is the result of the drum test, and S _-2.5mm calculated by the above-mentioned method were fitted to a linear function based on the least squares method, and the following formula (2) was calculated as an estimated formula (corresponding to an example). Also, the volumetric fracture strength DI ¹⁵⁰ _6-15 , which is the result of the drum test, and S _all , which was calculated by the above-mentioned method, were fitted to a linear function based on the least squares method, and the following formula (3) was calculated as an estimated formula (corresponding to a comparative example).

Figure 3 is a graph showing the degree of correlation between the estimation results based on formula (2) and the test results of the drum test, with the horizontal axis showing the estimated values and the vertical axis showing the test results of the drum test (actual values). Figure 4 is a graph showing the degree of correlation between the estimation results based on formula (3) and the test results of the drum test, with the horizontal axis showing the estimated values and the vertical axis showing the test results of the drum test (actual values).

3 and 4, the correlation coefficient ^R2 of the estimation method of this embodiment is 0.860, and the correlation coefficient of the estimation method of the comparative example is 0.246. Therefore, it was found that the volumetric fracture strength DI ¹⁵⁰ _6-15 can be accurately estimated by determining the area of the crack surface with a crack width of 2.5 mm or less for the coke cake carbonized in the test coke oven and substituting the determined crack surface area into formula (2). When the correlation coefficient ^R2 is less than 0.850 and an estimation accuracy of 0.850 or more is required for the correlation coefficient ^R2 , the threshold value of the crack width may be changed and set so as to satisfy the estimation accuracy.

In addition, for the coke cake obtained in the test coke oven, the conventional drum test required approximately 1.5 hours per level, but in this embodiment, this was shortened to approximately 10 minutes per level.

Claims (1)

A first step of acquiring information about the shape of cracks in a coke by performing imaging processing using an X-ray CT on the coke obtained by carbonizing coal in a test coke oven;
A second step of calculating an area of a crack surface having a crack width equal to or smaller than a threshold value by performing image analysis on the acquired information on the crack shape;
a third step of calculating an area of a crack surface of each coke having a crack width equal to or smaller than a threshold value by performing the first step and the second step for a plurality of coals having different properties, and then constructing an estimation equation for estimating the volumetric fracture strength based on the calculated area of each crack surface and the volumetric fracture strength actually measured for each coke;
A fourth step of estimating a volumetric fracture strength from the crack surface area of the coke obtained by performing the first step and the second step using coal for producing coke to be used in a blast furnace and the estimation formula obtained in the third step;
having
A method for estimating the volumetric fracture strength of coke, characterized in that the threshold value is set so that the correlation coefficient between the actual measured value of the volumetric fracture strength and the estimated value using the estimation equation constructed in the third step is 0.85 or more .