Coking coal preparation method for improving selectability of extremely-difficult-to-separate coal through mixed washing

Abstract

The invention discloses a coking coal preparation method for improving selectability of extremely-difficult-to-separate coal through mixed washing, and belongs to the technical field of coal washing and processing. The method comprises the following steps: firstly, carrying out a floating and sinking test on target extremely-difficult-to-separate coal and generating a continuous density-yield distribution function; then, by taking a density distribution complementation degree index as a compatibility criterion, quantitatively screening compatible coal types which are complementary with the target coal in density defect in the separation key density interval from the candidate coal types; determining an optimal ratio for maximizing the target ash content yield under the coking performance constraints that the clean coal caking index is not less than 80 and the gelatinous layer thickness is not less than 16.5 mm; and feeding the mixed coal into a dense medium cyclone for separation, and synchronously optimizing the stability of the suspension and a slime water system. According to the method, the overall density distribution curve of the mixed raw coal is remarkably improved, the yield of the clean coal is increased by 12-18%, finally qualified coking clean coal is produced, and conversion of the extremely difficult separation coal from low-value power coal to high-value coking coal is achieved.

Description

Technical Field

[0001] This invention relates to the field of coal washing and processing technology, specifically to a method for preparing coking coal that improves the washability of extremely difficult coals through blending and washing. Background Technology

[0002] Coking coal is the basic raw material for metallurgical coke production, and the strength and reactivity of coke are directly determined by the quality of the coal fed into the furnace. The national standard GB / T 16417-2011, "Methods for Evaluating the Washability of Coal," uses a separation density of ±0.1 g / cm³. 3 The content of near-heavy matter within the specified range (δ±0.1) is used as a criterion for coal washability. When δ±0.1 exceeds 40%, it is judged as extremely difficult to wash. Among China's proven coking coal resources, a considerable proportion in major producing areas such as Shanxi, Shaanxi, and Anhui are extremely difficult to wash.

[0003] The core characteristic of extremely difficult-to-prepare coal is its excessively high content of intermediate-density materials. Described by the density-yield distribution function, this is expressed as 1.3 g / cm³. 3 Up to 1.5 g / cm 3 The cumulative yield increases sharply within the density range, and the float-sink curve is almost vertical in this range. This means that at any single separation density, the density difference between the materials on both sides of the separation interface is extremely small, making it difficult for heavy medium cyclones to achieve effective separation even under optimal operating conditions. When coal washing plants use 10.5% as the ash content index for clean coal, the yield of clean coal washed alone is often as low as below 30%, which is economically unsustainable. If the ash content is relaxed to increase the yield, the caking index G and the plastic layer thickness Y of the clean coal will not meet the standards, making it unsuitable as a coking feedstock.

[0004] Currently, there are two main processing strategies for extremely difficult-to-separate coals. The first is to improve the coal preparation process by adding fine separation equipment such as spiral separators, TBS (Total Particulate Filter) separators, or flotation columns to perform secondary or tertiary refining of the coarse-separated products. This method suffers from drawbacks such as high equipment investment, high operating energy consumption, and sensitivity to feed conditions. Furthermore, it essentially attempts to improve separation accuracy under the same material conditions without fundamentally changing the density composition of the feed, thus offering limited improvement for extremely difficult-to-separate coals. The second approach is to directly downgrade the extremely difficult-to-separate coal to thermal coal for sale, circumventing the separation problem but resulting in significant resource value loss. Under current market conditions, the price difference between coking coal and thermal coal is typically between 300 and 800 yuan per ton. For a coal preparation plant with an annual processing capacity of 3 million tons, downgrading and selling the coal results in an annual economic loss of hundreds of millions of yuan.

[0005] In the coking industry, coal blending technology has been maturely applied in the coking process. The principle is that different types of clean coal differ in their petrographic composition, volatile matter, and caking properties; by rationally blending them, the overall coking performance of the coal entering the furnace can be optimized. However, current coal blending operations all occur after coal preparation and before entering the coke oven, implicitly assuming that each type of coal participating in the blend has already been washed and processed into qualified clean coal. For extremely difficult-to-prepare coals, separate washing cannot produce qualified clean coal, thus preventing them from entering the subsequent coal blending and coking process. This constitutes a break in the technological path.

[0006] It is worth noting that there is often complementarity between the density-yield distribution functions of different coal types: the density distribution function of extremely difficult-to-wash coals is concentrated in the intermediate density range (where the defects lie), while the density distribution function of some easily washable coals is concentrated in the low density range (where the advantages lie). If the two are mixed in an appropriate proportion before washing, the synthetic density distribution curve of the mixed coal will show the characteristics of a decrease in the proportion of intermediate density grades and an increase in the proportion of low density grades, thereby improving the washability of the feed from the source. However, there is currently no systematic method to quantitatively evaluate the degree of complementarity between the two coals in the density distribution dimension, and there is also a lack of technical means to determine the optimal blend ratio under the condition of simultaneously meeting coking performance constraints. Existing float-sink test standards and coal preparation process specifications are all designed for single coal types and do not cover the blending and washing scenario.

[0007] Therefore, there is an urgent need for a blending and washing method that can be implemented in the coal preparation process. This method should include: quantitative diagnostic methods for density defects in the target extremely difficult coal to be prepared, quantitative evaluation criteria for the compatibility of candidate coal types, a blending optimization mechanism under coking performance constraints, and process synergistic control in conjunction with heavy medium cyclone separation, so as to transform extremely difficult coal to be prepared into qualified coking coal and fill the gap in this technology. Summary of the Invention

[0008] Technical Objective: Addressing the problem that existing extremely difficult-to-wash coals suffer from unfavorable density composition due to excessively high content of intermediate-density materials, resulting in high ash content, low yield, and substandard coking performance under individual washing conditions, thus hindering their effective utilization as coking raw materials, and that existing fine separation technologies fail to fundamentally change the feed density composition, have limited effectiveness, and require significant investment, this invention discloses a method for preparing coking coal that improves the washability of extremely difficult-to-wash coals through blending and washing. The core idea is to move the blending operation forward to the coal preparation stage: by quantitatively analyzing the density composition defects of the target extremely difficult-to-wash coal, using density distribution complementarity index as a quantitative criterion, select compatible coal types that complement the target coal's defects in the key separation density range. Under the constraint of nonlinear coking performance, determine the optimal blending ratio, and use a conventional heavy medium cyclone separator to separate the blended coal, ultimately obtaining qualified coking clean coal.

[0009] Technical solution: To achieve the above technical objectives, the present invention adopts the following technical solution:

[0010] A method for preparing coking coal with improved washability for extremely difficult coal preparation includes the following steps:

[0011] Step S1: Target Coal Density Composition Analysis: Conduct float-sink tests on the target extremely difficult-to-benefit coal according to at least six standard density grades to obtain yield and ash content data for each density grade. Generate continuous density-yield distribution functions and density-ash content curves through interpolation fitting, and calculate the theoretical clean coal yield per unit coal at the target ash content. The six standard density grades include 1.3 g / cm³. 3 1.4g / cm 3 1.5g / cm 3 1.6g / cm 3 1.7g / cm 3 and 1.8g / cm 3 The interpolation fitting method uses cubic spline interpolation, with a density resolution of no less than 0.01 g / cm³. 3 This is to ensure the accuracy of subsequent synthesis calculations.

[0012] Step S2, Coal Compatibility Screening: Perform float-sink tests on multiple candidate coal types and obtain their respective density-yield distribution functions. Calculate the density distribution complementarity index between each candidate coal type and the target extremely difficult-to-separate coal. The physical meaning of this index is: a quantitative characterization of the two coals in the critical separation density range (1.3 g / cm³). 3 Up to 1.6 g / cm 3 The degree of complementarity of density distribution defects within the coal type is considered. Candidate coal types whose density distribution complementarity index meets a preset threshold (e.g., not less than 0.70) are selected as blending coal types.

[0013] The specific calculation method for the density distribution complementarity index is as follows: Define an ideal selectivity distribution function. This function satisfies the low density level (below 1.4 g / cm³) within the critical density range for sorting. 3 Yield percentage higher than 45%, intermediate density grade (1.4 g / cm³) 3 Up to 1.6 g / cm 3 The condition that the yield percentage is less than 25% is considered. This ideal distribution represents the density composition characteristics of easily washable coal. The deviation function between the density-yield distribution function of the target coal and the ideal distribution is calculated. And the deviation function of candidate coal type k Density distribution complementarity index The calculation formula is:

[0014]

[0015] The integration interval is 1.3 g / cm³. 3 Up to 1.6 g / cm3 . The value of is between 0 and 1. The closer the value is to 1, the stronger the complementarity of the density distribution defects of the two coals: when the deviation functions of the two coals are relatively large in different density sub-intervals (i.e., the defect of one coal is exactly in the dominant interval of the other coal), the product integral value in the numerator is smaller. If the two values ​​approach 1, then... Approaching 0.

[0016] Based on density complementarity screening, surface property compatibility evaluation can be further conducted: the contact angle and Zeta potential of the target coal and each candidate coal type are measured. Coal types with an absolute difference in contact angle not exceeding 15° and an absolute difference in Zeta potential not exceeding 10 mV are considered surface property compatible. Only candidate coal types that simultaneously meet the density complementarity threshold and surface property compatibility conditions are determined as the final blending coal types. The significance of surface property compatibility evaluation is that if the surface hydrophobicity and charge characteristics of the two coals differ too much, their wetting behavior in the heavy medium suspension after mixing will be inconsistent, which may affect the interfacial mass transfer between the medium and coal particles and reduce the separation efficiency.

[0017] Step S3, Blending Optimization Step: The density-yield distribution function of the target coal and the density-yield distribution function of the blended coals are weighted and superimposed according to different blending ratios ω to generate the theoretical washability curve of the blended coal. Where ω is the mass percentage of the target coal. The optimal mix ratio is determined under the following constraints: (a) The ash content of the clean coal does not exceed the target ash content. (b) The caking index (G) of the clean coal is not less than 80; (c) The thickness (Y) of the plastic layer is not less than 16.5 mm; (d) The mass percentage of the target coal in the blended coal is not less than 50%; (e) The density of the blended coal within ±0.1 g / cm³ during sorting is within ±0.1 g / cm³. 3 The content of near-heavy substances within the range δ±0.1 does not exceed 40%.

[0018] It is important to note that the caking index G and the plastic layer thickness Y do not follow a linear weighted average law. This invention employs a nonlinear prediction model (such as support vector machine regression, XGBoost, or BP neural network) to predict the G and Y values ​​of the blended coking coal. This nonlinear prediction model uses the dry ash-free volatile matter content of each coal type as the basis for prediction. The model is trained using vitrinite and inertinite content as input features, and G and Y values ​​as output targets under different laboratory-measured ratios. On the validation set, the prediction accuracy of the model should reach a mean absolute error of no more than 3 units for the G value and no more than 0.5 mm for the Y value.

[0019] The ratio optimization can be carried out using a multi-objective optimization method, with the dual objectives of clean coal yield and economic value. A systematic search is conducted within the above-mentioned constraint space to obtain the Pareto optimal ratio set, from which the scheme that takes into account both yield and economy is selected as the optimal ratio.

[0020] Step S4, Mixing and Washing Separation: The target coal is mixed with the blended coal types according to the optimal ratio. An independent coal feeder is used to quantitatively feed each coal type. An electronic belt scale provides real-time feedback on the actual feed amount and performs closed-loop correction of ratio deviations. After thorough mixing in a homogenization bin, the mixed raw coal enters the preparation workshop for crushing and screening. The crushed mixed raw coal is then fed into a heavy medium cyclone separator for density separation. The separation density is set according to the separation density value corresponding to the target ash content on the theoretical washability curve of the mixed coal.

[0021] During the sorting process, the stability of the heavy media suspension is controlled: the density and viscosity of the feed suspension are monitored in real time, and if the density deviation exceeds ±0.02 g / cm³ of the set sorting density, the system will take action. 3 When the viscosity exceeds 25 mPa·s, adjust the amount of magnetite powder added or the flow rate of the circulating medium to correct the deviation; when the viscosity exceeds 25 mPa·s, increase the processing capacity of the medium purification circuit to reduce the content of non-magnetic fine mud. Perform online ash content testing on the clean coal product and finely adjust the sorting density in real time based on the test results.

[0022] Simultaneous optimization of the coal slurry water system: Mineral composition analysis of the coal slurry water generated during washing is performed. Based on the main clay mineral types, the appropriate flocculant type and dosage are selected to reduce the turbidity of the coal slurry water to below the preset standard. The solid content in the circulating water is monitored; when it exceeds the preset concentration limit, the concentration overflow discharge and clean water replenishment are initiated.

[0023] The present invention also provides a coking coal preparation system for improving the washability of extremely difficult coal preparation, including a coal quality analysis module, a proportioning optimization module, a coal feeding module, a heavy media separation module, and a quality monitoring module.

[0024] The coal quality analysis module is configured to collect and process float-sink test data for the target extremely difficult coal to be made and multiple candidate coal types, generate the density-yield distribution function of each coal type, calculate the density distribution complementarity index between the target extremely difficult coal to be made and each candidate coal type, and screen out the compatible coal types that meet the preset threshold.

[0025] The proportioning optimization module is connected to the coal quality analysis module and is configured to receive the density-yield distribution function and coking performance parameters of each coal type, generate the theoretical sorting curve of the mixed coal by weighted superposition, solve the optimal proportion to maximize the clean coal yield under coking performance constraints and ash content constraints, and output the optimal proportion scheme and the corresponding sorting density setting value.

[0026] The coal blending and feeding module is signal-connected to the proportioning optimization module and configured to control the feeding amount of the target extremely difficult coal and the blended coal according to the optimal proportioning scheme, and to transport the mixed coal to the inlet of the heavy medium cyclone.

[0027] The heavy medium separation module includes a heavy medium cyclone separator and a medium circulation system, configured to perform density separation on the mixed coal according to the separation density set value to obtain coking coal;

[0028] A quality monitoring module is installed at the clean coal output end of the heavy medium separation module. It is configured to detect the ash content of the clean coal in real time and feed the detection results back to the ratio optimization module and the heavy medium separation module for adjusting the ratio or separation density.

[0029] Beneficial Effects: The coking coal preparation method for improving the washability of extremely difficult coal preparation provided by the present invention has the following beneficial effects:

[0030] First, this invention introduces for the first time a quantitative criterion of density distribution complementarity, transforming the selection of blended coal types from traditional qualitative and empirical judgments to quantitative evaluation based on float-sink test data. This index, starting from the perspective of the complementarity of density distribution defects between two types of coal in the key density range of separation, provides an evaluation standard with clear physical meaning, a definite calculation method, and repeatability. This is an innovative criterion that does not exist in existing coal preparation and blending technologies.

[0031] Secondly, this invention moves the coal blending operation forward to the coal preparation stage, breaking through the inherent path of existing technologies that involve separate washing and preparation before blending for coking. Existing technologies assume that each type of coal participating in blending must first be washed and prepared into qualified clean coal, while this invention breaks this assumption by directly blending and mixing the raw coal before washing. This innovative approach makes it possible, for the first time, for extremely difficult-to-prepare coals to be utilized as coking raw materials.

[0032] Third, this invention introduces a nonlinear prediction model to handle the nonlinear mixed effect of the bonding index G and the plastic layer thickness Y, solving the problem of large prediction errors (over 16%) in the traditional weighted average method. The nonlinear model accurately predicts the coking performance of the blended coal, ensuring the reliability of the proportioning optimization results in terms of coking performance.

[0033] Fourth, this invention can effectively separate extremely difficult-to-separate coals using a conventional heavy medium cyclone separator, eliminating the need for additional fine separation equipment, resulting in low equipment investment and operating costs. Industrial trials show that the yield of clean coal at the target ash content can be increased by 12 to 18 percentage points, the clean coal caking index G value is not less than 80 and the plastic layer thickness Y value is not less than 16.5 mm, meeting the quality requirements of coking coal. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0035] Figure 1 This is a schematic diagram of the overall process of the method of the present invention. Figure 1 The sequence and data flow of the four steps from target coal density composition analysis (S1), coal type screening (S2), blending optimization (S3) to mixed washing and separation (S4) are shown.

[0036] Figure 2 This is a schematic diagram illustrating the calculation principle of the density distribution complementarity index. Figure 2 The density distribution function of the target coal is displayed with density on the horizontal axis and yield distribution function values ​​on the vertical axis. Density distribution function of a candidate coal type and ideal choice distribution function The relationship between them, and the first deviation function Second deviation function The meaning of .

[0037] Figure 3 This is a schematic diagram comparing the selectivity curves of target coal washed alone and after blending. Figure 3 Using the horizontal axis as the ash content of clean coal and the vertical axis as the yield of clean coal, the theoretical washability curves of the target coal when washed alone and the theoretical washability curves after being washed with the optimal ratio are shown. The yield of the curve after washing with the optimal ratio is significantly higher than that of the curve when washed alone at the same ash content.

[0038] Figure 4 This is a schematic diagram of the module composition and data flow of the system of the present invention. Figure 4 The layout of the coal quality analysis module, the proportioning optimization module, the coal feeding module, the heavy media separation module, and the quality monitoring module, as well as the data flow and control signal flow between the modules, are shown. The quality monitoring module provides feedback signals to the proportioning optimization module and the heavy media separation module. Detailed Implementation

[0039] The present invention will now be described more clearly and completely by way of a preferred embodiment in conjunction with the accompanying drawings, but this does not limit the invention to the scope of the described embodiment.

[0040] Example 1

[0041] See Figure 1 This embodiment uses extremely difficult-to-process prime coking coal (hereinafter referred to as target coal A) from a mining area in Shanxi Province as the object to illustrate the complete implementation process of the method of the present invention. The method of the present invention includes four steps in sequence: target coal density composition analysis (S1), coal type screening (S2), blending ratio optimization (S3), and mixed washing and separation (S4).

[0042] Step S1, Target Coal Density Composition Analysis:

[0043] A representative sample of 200 kg of target coal A was taken and subjected to a float-sink test according to GB / T 478-2008 "Coal Float-Sink Test Method". The six selected standard density grades were 1.3 g / cm³. 3 1.4g / cm 3 1.5g / cm 3 1.6g / cm 3 1.7g / cm 3 and 1.8g / cm 3 The results of the buoyancy and sinking test are as follows:

[0044] Density level below 1.3 g / cm³ 3 Yield 18.6%, ash content 4.2%.

[0045] Density grade 1.3 g / cm³ 3 Up to 1.4 g / cm 3 Yield 22.3%, ash content 9.8%.

[0046] Density grade 1.4 g / cm³ 3 Up to 1.5g / cm 3 Yield 24.1%, ash content 16.5%.

[0047] Density grade 1.5 g / cm³ 3 Up to 1.6 g / cm 3 Yield 13.7%, ash content 23.2%.

[0048] Density grade 1.6 g / cm³ 3 Up to 1.7 g / cm 3 Yield: 8.2%, ash content: 31.6%.

[0049] Density grade: 1.7 g / cm³ 3 Up to 1.8 g / cm 3 Yield: 5.4%, ash content: 40.8%.

[0050] Density level higher than 1.8 g / cm³ 3 Yield: 7.7%, ash content: 68.5%.

[0051] The discrete data above are fitted to a continuous density-yield distribution function using cubic spline interpolation. Density-ash content relationship curve, with a density resolution of 0.01 g / cm³. 3 Calculations based on the sorting curve show that the theoretical clean coal yield of target coal A at a target ash content of 10.5% is 32.8%, and the sorting density is 1.45 g / cm³.3 ±0.1g / cm 3 The near-heavy matter content (δ±0.1) within this range is 47.2%, classifying it as an extremely difficult coal to prepare. The yield of this clean coal is economically infeasible, and the resulting clean coal has a caking index (G value) of only 72 and a plastic layer thickness (Y value) of only 14.8 mm, both failing to meet the quality requirements for coking coal (G≥80, Y≥16.5 mm).

[0052] The petrographic analysis results of target coal A are as follows: vitrinite content 67.3%, inertinite content 24.8%, and chrysrinite content 7.9%; dry ash-free volatile matter... It is 23.6%.

[0053] Step S2, Coal type screening:

[0054] Four candidate coal types (coal types B, C, D, and E) were selected from the surrounding mining areas, and float-sink tests were conducted according to the same six standard density grades as the target coal A to obtain their respective density-yield distribution functions.

[0055] The key property parameters of each candidate coal type are as follows:

[0056] Coal type B: Density grade less than 1.3 g / cm³ 3 The yield was 41.5%, and the density was 1.3 g / cm³. 3 Up to 1.6 g / cm 3 The total yield was 32.1%, the adhesion index (G value) was 88, and the adhesive layer (Y value) was 18.2 mm. 21.8%, vitrinite content 73.2%, inertinite content 19.5%.

[0057] Coal type C: Density grade less than 1.3 g / cm³ 3 The yield was 28.7%, and the density was 1.3 g / cm³. 3 Up to 1.6 g / cm 3 The total yield was 43.5%, the adhesion index (G value) was 90, and the adhesive layer (Y value) was 19.0 mm. 24.1%, vitrinite content 76.8%, inertinite content 16.3%.

[0058] Coal type D: Density grade less than 1.3 g / cm³ 3 The yield was 35.2%, and the density was 1.3 g / cm³. 3 Up to 1.6 g / cm 3 The total yield was 38.6%, the adhesion index (G value) was 75, and the adhesive layer (Y value) was 15.0 mm. 25.3%, vitrinite content 61.5%, inertinite content 30.2%.

[0059] Coal type E: Density grade less than 1.3 g / cm³3 The yield was 25.3%, and the density was 1.3 g / cm³. 3 Up to 1.6 g / cm 3 The total yield was 48.2%, the adhesion index (G value) was 85, and the adhesive layer (Y value) was 17.5 mm. 22.0%, vitrinite content 70.5%, inertinite content 22.1%.

[0060] See Figure 2 First, we define the ideal choice distribution function. The critical density range for sorting is 1.3 g / cm³. 3 Up to 1.6 g / cm 3 Internally, the ideal distribution satisfies: low density (below 1.4 g / cm³). 3 Yield percentage higher than 45%, intermediate density grade (1.4 to 1.6 g / cm³) 3 The yield share is less than 25%. This ideal distribution represents the density composition characteristics of easily sifted coal.

[0061] Then calculate the deviation function between the density distribution function of target coal A and the ideal distribution (corresponding to...). Figure 2 (b) the gray-filled area), the deviation function between each candidate coal type and the ideal distribution (corresponding to) Figure 2 (b) the area filled with diagonal lines, and calculate the density distribution complementarity index according to the formula described in the invention content section of the specification. Figure 2 The thick solid line curve in (b) represents the deviation product function. The smaller the integral value of the function over the integration interval, the stronger the complementarity of the density defects between the two coal types. The higher the value, the better. The results of the complementarity index calculation between each candidate coal type and target coal A are as follows:

[0062] Coal type B: .

[0063] Coal type C: .

[0064] Coal type D: .

[0065] Coal type E: .

[0066] The threshold for the density distribution complementarity index is set at 0.70. Coal type B ( ) and coal type D ( The threshold condition is met.

[0067] Further surface property compatibility evaluation was conducted. The contact angle and Zeta potential of each coal type were measured as follows:

[0068] Target coal A: contact angle 72°, Zeta potential -28 mV.

[0069] Coal type B: Contact angle 68°, Zeta potential -32mV. Absolute value of contact angle difference 4°, absolute value of Zeta potential difference 4mV. Both meet the compatibility conditions (contact angle difference not exceeding 15°, Zeta potential difference not exceeding 10mV).

[0070] Coal type D: Contact angle 55°, Zeta potential -15mV. The absolute value of the contact angle difference is 17°, which exceeds the 15° threshold and does not meet the compatibility requirements.

[0071] Therefore, coal type B was ultimately selected as the blending coal type.

[0072] Step S3, Optimization of proportions:

[0073] The mixing density curves of target coal A and coal type B were synthesized and coking performance was predicted according to different blending ratios. The calculations were performed stepwise from 50% to 90% (in 5% increments) based on the mass percentage ω of target coal A. The synthesis formula is as follows: .

[0074] The nonlinear prediction model employs Support Vector Machine Regression (SVR), with a Radial Basis Function (RBF) kernel, a penalty parameter C=100, and a kernel parameter γ=0.1. The training data comes from laboratory-prepared target coal A and coal type B in 10 different blending tests in small coking ovens (each group was repeated 3 times and the average was taken, for a total of 30 data sets). The weighted average, the weighted average of vitrinite content, and the weighted average of inertinite content were used as input features. The model's prediction accuracy in 5-fold cross-validation was: G-value mean absolute error 2.1 units (coefficient of determination R0). 2 =0.927), the mean absolute error of the Y value is 0.38 mm (R 2 =0.912).

[0075] The calculation and prediction results of key indicators under each ratio are as follows:

[0076] ω=90% (A accounts for 90%, B accounts for 10%): Theoretical clean coal yield is 35.1%, clean coal ash content is 10.4%, δ±0.1 is 44.3%, predicted G value is 74, and predicted Y value is 15.3 mm. The G value constraint and δ±0.1 constraint are not satisfied.

[0077] ω=85% (A accounts for 85%, B accounts for 15%): Theoretical clean coal yield is 38.2%, clean coal ash content is 10.4%, δ±0.1 is 42.1%, predicted G value is 77, and predicted Y value is 15.8 mm. The G value constraint and δ±0.1 constraint are not satisfied.

[0078] ω=80% (A accounts for 80%, B accounts for 20%): Theoretical clean coal yield 41.5%, clean coal ash content 10.3%, δ±0.1 is 39.6%, predicted G value 79, predicted Y value 16.2 mm. The G value constraint is not met.

[0079] ω=75% (A accounts for 75%, B accounts for 25%): theoretical clean coal yield 45.1%, clean coal ash content 10.3%, δ±0.1 is 36.8%, predicted G value 82, predicted Y value 16.8 mm. All constraints are satisfied.

[0080] ω=70% (A accounts for 70%, B accounts for 30%): theoretical clean coal yield 48.6%, clean coal ash content 10.2%, δ±0.1 is 33.5%, predicted G value 83, predicted Y value 17.1 mm. All constraints are satisfied.

[0081] ω=65% (A accounts for 65%, B accounts for 35%): theoretical clean coal yield 51.2%, clean coal ash content 10.1%, δ±0.1 is 30.2%, predicted G value 84, predicted Y value 17.3 mm. All constraints are satisfied.

[0082] ω=60% (A accounts for 60%, B accounts for 40%): theoretical clean coal yield 53.5%, clean coal ash content 10.0%, δ±0.1 is 27.8%, predicted G value 85, predicted Y value 17.5 mm. All constraints are satisfied.

[0083] ω=55% (A accounts for 55%, B accounts for 45%): theoretical clean coal yield 55.7%, clean coal ash content 9.9%, δ±0.1 is 25.1%, predicted G value 86, predicted Y value 17.8 mm. All constraints are satisfied.

[0084] ω=50% (A accounts for 50%, B accounts for 50%): Theoretical clean coal yield is 57.3%, clean coal ash content is 9.8%, δ±0.1 is 22.6%, predicted G value is 87, and predicted Y value is 18.0 mm. All constraints are satisfied.

[0085] From a yield perspective, the smaller ω (the higher the proportion of blended coal type B), the higher the yield of refined coal. However, from an economic perspective, the procurement cost and supply constraints of coal type B need to be considered. The procurement price of target coal A is set at 350 yuan per ton (pithead price), the procurement price of coal type B at 520 yuan per ton, the selling price of coking coal at 1200 yuan per ton, and the selling price of thermal coal at 650 yuan per ton. The net benefit per ton of blended raw coal is used as the economic objective function:

[0086]

[0087] Given a processing cost of 50 yuan per ton, calculate the net benefit of each proportion that satisfies all constraints:

[0088] ω=75%: Net benefit = 0.451×1200 - [0.75×350+0.25×520] - 50 = 541.2 -392.5 - 50 = 98.7 yuan / ton.

[0089] ω=70%: Net benefit = 0.486×1200 - [0.70×350+0.30×520] - 50 = 583.2 -401.0 - 50 = 132.2 yuan / ton.

[0090] ω=65%: Net benefit = 0.512×1200 - [0.65×350+0.35×520] - 50 = 614.4 -409.5 - 50 = 154.9 yuan / ton.

[0091] ω=60%: Net benefit = 0.535×1200 - [0.60×350+0.40×520] - 50 = 642.0 -418.0 - 50 = 174.0 yuan / ton.

[0092] ω=55%: Net benefit = 0.557×1200 - [0.55×350+0.45×520] - 50 = 668.4 -426.5 - 50 = 191.9 yuan / ton.

[0093] ω=50%: Net benefit = 0.573×1200 - [0.50×350+0.50×520] - 50 = 687.6 -435.0 - 50 = 202.6 yuan / ton.

[0094] Pareto analysis shows that net benefit increases monotonically with the increase in the proportion of coal type B in the blend. However, considering the limited market supply capacity of coal type B (annual procurement volume of approximately 1 million tons) and the risk of supply stability, the maximum sustainable blending ratio of coal type B is approximately 30% under the condition of processing 3 million tons per year. Therefore, ω=70% (target coal A accounts for 70%, coal type B accounts for 30%) is selected as the optimal blending scheme. Figure 3 As shown, under this scheme, the yield of clean coal increases by 15.8 percentage points compared to washing target coal A alone (from 32.8% to 48.6%). Figure 3 The difference in the vertical axis of the two selectivity curves at the target ash content of 10.5% is the yield increase. All coking performance indicators meet the requirements. The G value of the washed coal increased from 72 to 83 (not less than 80), and the Y value increased from 14.8mm to 17.1mm (not less than 16.5mm), turning from unqualified to qualified, resulting in an annual net benefit increase of approximately 397 million yuan.

[0095] Step S4, Mixing and Washing Sorting:

[0096] Industrial trials were conducted at the coal preparation plant. Coal type B was transported to the plant's independent storage silo via a dedicated transport channel. It was then fed onto the main conveyor belt at a set ratio (70:30) using independent quantitative feeders. The quantitative feeder for target coal A was set to 700 t / h, and the quantitative feeder for coal type B was set to 300 t / h. Two electronic belt scales were installed on their respective feed belts to collect real-time data on the actual feed rate and feed it back to the proportioning controller. When the actual proportioning deviation exceeded ±2%, the proportioning controller automatically adjusted the speed of the corresponding feeder to correct it. After mixing, the raw coal was processed in a homogenization silo (effective volume 500 m³). 3 After thorough mixing (with a residence time of approximately 30 minutes), the mixture is transferred to the preparation workshop.

[0097] The crushed and screened mixed raw coal (with an upper limit of 50mm) is fed into a Φ710mm three-product heavy medium hydrocyclone for separation. Based on the theoretical washability curve of the mixed coal, the separation density of the first-stage hydrocyclone is set at 1.48 g / cm³. 3 The separation density of the two-stage hydrocyclone was set to 1.80 g / cm³. 3 The heavy medium suspension was prepared from magnetite powder, with the particle size (D95) controlled below 75 μm. The density of the suspension in the first-stage hydrocyclone was maintained at 1.48 ± 0.02 g / cm³ using an automatic density control system. 3 Within the specified range, the feed pressure is controlled at 9D (D is the diameter of the hydrocyclone, approximately 120 kPa).

[0098] Stability control of heavy medium suspension: A Coriolis mass flow meter is installed on the hydrocyclone inlet pipe to monitor the suspension density in real time. Simultaneously, an online viscometer (rotary type, range 0 to 100 mPa·s) is installed in the qualified medium tank. If the density deviation exceeds ±0.02 g / cm³, the suspension will be monitored. 3 When the viscosity exceeds 25 mPa·s, the density controller adjusts the frequency of the inverter of the qualified medium pump; when the viscosity exceeds 25 mPa·s, the system automatically increases the speed of the magnetic separator feed pump and the water spray volume of the desliming screen to accelerate the discharge of fine mud.

[0099] The ash content of the refined coal is monitored in real time using a dual-energy X-ray online ash content analyzer installed on the refined coal conveyor belt, with a standard deviation of ±0.15%. When the ash content of the refined coal exceeds 10.5% for 5 consecutive minutes, the system issues an early warning to the operator and automatically reduces the density of the first-stage sorting section by 0.01 g / cm³. 3 When the ash content is below 9.5% for 5 consecutive minutes, it indicates that the sorting density can be increased to improve the yield.

[0100] Coal slurry water system synergistic optimization: X-ray diffraction mineral composition analysis was performed on samples of the coal slurry water generated from the mixed washing process. The clay minerals in the slurry of target coal A were mainly kaolinite and illite, while those in the slurry of coal type B were mainly kaolinite. After mixing, the clay minerals in the slurry remained predominantly kaolinite. Based on this, anionic polyacrylamide (molecular weight 12 million) was selected as the flocculant, with a dosage of 30 mg / L. The turbidity of the treated coal slurry water was reduced to below 200 NTU. The solids content of the circulating water was controlled below 10 g / L through periodic sampling and testing; when this upper limit was exceeded, the thickener overflowed and clean water was replenished.

[0101] Industrial test results:

[0102] The actual quality indicators of coking coal obtained after 72 hours of continuous industrial testing are as follows:

[0103] The ash content of the clean coal is 10.4% (meeting the target requirement of not exceeding 10.5%).

[0104] The yield of clean coal was 47.8% (an increase of 15.0 percentage points compared to the 32.8% yield of target coal A washed separately).

[0105] The adhesion index G value is 81 (meeting the requirement of not less than 80).

[0106] The thickness of the adhesive layer (Y value) is 16.8 mm (meeting the requirement of not less than 16.5 mm).

[0107] Dry ash-free volatile matter It is 22.5%.

[0108] The content of near-heavy substances δ±0.1 is 34.1% (improved from extremely difficult to select to relatively difficult to select).

[0109] Heavy medium cyclone separator separation efficiency index: possible deviation Ep is 0.035 g / cm³ 3 (Within the normal range), the actual value of the sorting density ρ50 is 1.483 g / cm³. 3 (Compared to the set value of 1.48g / cm) 3 The deviation is within 0.01 g / cm 3 within).

[0110] Economic Benefit Analysis: The output of coking coal per ton of raw coal increased from 328 kg when washed alone to 478 kg. Based on a price of 1200 yuan per ton of coking coal and 650 yuan per ton of thermal coal, the product value per ton of raw coal increased from 393.6 yuan to 573.6 yuan, an increase of approximately 45.7%. After deducting the premium for blended coal purchases, the net increase in benefit per ton of mixed raw coal is approximately 132 yuan. For a coal preparation plant with an annual processing capacity of 3 million tons, the annual increase in economic benefit is approximately 396 million yuan.

[0111] Example 2

[0112] This embodiment uses extremely difficult-to-process coking coal (hereinafter referred to as target coal F) from a mining area in Shaanxi Province as the object to verify the applicability of the method of the present invention under different coal types and mining area conditions.

[0113] The float-sink test results for target coal F show that its density is below 1.3 g / cm³. 3 The yield was 15.3%, and the density was 1.3 g / cm³. 3 Up to 1.6 g / cm 3 The total yield was 65.2%, and the near-heavy matter content (δ±0.1) reached 51.6%, classifying it as extremely difficult to prepare coal. The theoretical clean coal yield per coal at a target ash content of 10.0% was only 28.5%, with a caking index (G value) of 76 and a plastic layer (Y value) of 15.2 mm, all failing to meet coking requirements.

[0114] Candidate coal types (G, H, and I) were selected from three adjacent mines in the same region, and float-sink tests and density distribution complementarity index calculations were performed on each:

[0115] Coal type G: The threshold is not met.

[0116] Coal type H: The thresholds are met. Low-density yield is 39.8%, G value is 92, and Y value is 19.5 mm. Surface property evaluation: absolute contact angle difference is 8°, and absolute Zeta potential difference is 6mV, meeting compatibility requirements.

[0117] Coal type I: The threshold is not met.

[0118] Coal type H was determined to be a compatible coal type.

[0119] After optimization, the optimal blending ratio was determined to be 65% target coal F and 35% coal H. The theoretical clean coal yield under this ratio is 46.3% (an increase of 17.8 percentage points compared to single coal), the clean coal ash content is 9.8%, the δ±0.1 is 34.7% (improved from extremely difficult to beneficiate to moderately beneficiate), the predicted G value is 85, and the Y value is 17.6 mm.

[0120] In the industrial trial, the same coal blending, heavy media separation, and coal slurry water treatment processes as in Example 1 were used. The actual clean coal ash content was 9.8%, the yield was 45.1%, the G value was 83, and the Y value was 17.2 mm. This verified the universality of the method of the present invention for different extremely difficult coal types and different mining area conditions.

[0121] System Implementation Examples

[0122] See Figure 4 The present invention also provides a coking coal preparation system for improving the washability of extremely difficult coal preparation, comprising the following five modules:

[0123] The coal quality analysis module includes a buoyancy and sedimentation data digitization unit and a complementarity calculation unit. The buoyancy and sedimentation data digitization unit receives yield and ash dispersion data for each coal type at various standard density levels, and converts them into continuous density-yield distribution functions and density-ash relationship curves through cubic spline interpolation. The complementarity calculation unit receives the density-yield distribution functions for each coal type and calculates the complementarity index between the target coal and each candidate coal type according to the formula. Value, and according to The output results are sorted in descending order of values. This module also integrates an interface for inputting surface property data (contact angle, zeta potential), supporting joint screening of density complementarity and surface compatibility.

[0124] The blending optimization module and the coal quality analysis module are connected via a data bus, receiving density-yield distribution functions, coal petrographic data, and coal price information for each coal type. Internally, the module includes: a weighted superposition calculation engine for synthesizing theoretical sorting curves for blended coal at different blending ratios; a built-in nonlinear prediction model (SVR) for predicting the G and Y values ​​of the blended clean coal; and a multi-objective optimization solver for solving the Pareto optimal blending set under dual objectives of yield and economic value, within all constraints (ash content, G value, Y value, lower limit of proportion, upper limit of δ±0.1). This module outputs the optimal blending scheme and the corresponding sorting density setpoint.

[0125] The coal blending and feeding module is connected to the proportioning optimization module via an industrial Ethernet signal. This module includes separate quantitative feeders (speed adjustment accuracy ±1%) for the target coal and the blended coal types, an electronic belt scale (accuracy class 0.5), and a homogenization silo (volume 500 m³). 3 The quantitative coal feeder adjusts the feeding rate of each type of coal according to the mass ratio signal output by the proportioning optimization module, and the electronic belt scale feeds back the actual feeding amount to the proportioning controller in real time for closed-loop correction.

[0126] The heavy medium separation module includes a heavy medium hydrocyclone, a magnetite powder media preparation system, a media circulation and purification loop, and an automatic density control system. The separation density setpoint is issued by the proportioning optimization module. The automatic density control system monitors the suspension density in real time using a Coriolis density meter; if the deviation exceeds ±0.02 g / cm³, the system will detect the change. 3 The frequency of the frequency converter for the qualified medium pump is automatically adjusted in real time. The magnetic separator in the medium purification circuit is responsible for recovering magnetite powder, and the amount of water used for washing the desliming screen is adjusted according to viscosity feedback.

[0127] The quality monitoring module is located at the output end of the clean coal conveyor belt and includes an online ash content analyzer and an online rapid caking property evaluation unit. The online ash content analyzer can use dual-energy X-ray transmission (ash content detection standard deviation ±0.15%) or transient gamma neutron activation analysis (ash content detection standard deviation ±0.15%, and can also determine total sulfur content). The online rapid caking property evaluation unit uses near-infrared spectroscopy to analyze the petrographic characteristics of the clean coal and, combined with a preset caking property mapping model, quickly estimates the G and Y values. The detection results from the quality monitoring module are transmitted to the blending optimization module and the heavy medium separation module via feedback channels, forming a two-way feedback closed loop: when the ash content is too high, the separation density can be fine-tuned or the proportion of blended coal can be increased; when the caking property index is close to the lower limit, the proportion of blended coal can be appropriately increased.

[0128] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for preparing coking coal with improved washability for extremely difficult coal preparation, characterized in that, Specifically, the following steps are included: S1. Conduct float and sink tests on the target extremely difficult coal according to at least six standard density grades to obtain the yield and ash content data of each density grade. Generate a continuous density-yield distribution function and density-ash content relationship curve by interpolation fitting, and calculate the theoretical clean coal yield of a single coal under the target ash content. S2. Perform float-sink tests on multiple candidate coal types and obtain their respective density-yield distribution functions. Calculate the density distribution complementarity index between each candidate coal type and the target extremely difficult coal to be prepared. The density distribution complementarity index is used to quantitatively characterize the degree of complementarity between the target coal and the candidate coal types in the key density range of separation. Select candidate coal types whose density distribution complementarity index meets a preset threshold as blending coal types. S3. The density-yield distribution function of the target extremely difficult coal to be made and the density-yield distribution function of the blended coal are weighted and superimposed according to different ratios to generate multiple sets of theoretical coal beneficiation curves; under the constraints of coking performance (coal caking index not less than 80 and plastic layer thickness not less than 16.5 mm) and the mass constraint (coal ash content not exceeding the target ash content), the optimal ratio that maximizes the theoretical yield of the coal at the target ash content is determined. S4. Mix the target extremely difficult coal with the blended coal according to the optimal ratio, and send the mixed raw coal into a heavy medium cyclone separator for density separation. The separation density is set according to the separation density value corresponding to the target ash content on the theoretical coal washability curve to obtain coking coal that meets the target ash content requirements.

2. The method for preparing coking coal with improved washability for extremely difficult coal preparation according to claim 1, characterized in that, In step S2, the critical density range for sorting is 1.3 g / cm³. 3 Up to 1.6 g / cm 3 The density distribution complementarity index is calculated as follows: The first deviation function between the density-yield distribution function of the target extremely difficult coal to be prepared and the preset ideal selectivity distribution function, and the second deviation function between the density-yield distribution function of the candidate coal and the ideal selectivity distribution function are obtained respectively. The product of the first deviation function and the second deviation function is integrated over the critical density interval of sorting to obtain the deviation co-integral value, and the square of the larger of the first deviation function and the second deviation function is integrated to obtain the deviation baseline integral value. The density distribution complementarity index is obtained by normalizing the ratio of the deviation synergy integral value to the deviation benchmark integral value. The ideal selectivity distribution function is a standard distribution function in which the proportion of low-density grade yield is higher than a preset first proportion threshold and the proportion of intermediate-density grade yield is lower than a preset second proportion threshold within the critical density range of sorting.

3. The method for preparing coking coal with improved washability for extremely difficult coal preparation according to claim 1, characterized in that, Step S2 also includes a step of evaluating the surface property compatibility of candidate coal types: The contact angle and Zeta potential between the target extremely difficult coal and various candidate coal types were determined; Calculate the absolute values ​​of the contact angle difference and the absolute values ​​of the Zeta potential difference between the target extremely difficult coal preparation and each candidate coal type; Candidate coal types with an absolute value of contact angle difference not exceeding 15° and an absolute value of Zeta potential difference not exceeding 10mV are classified as surface property compatible coal types. Only candidate coal types that simultaneously meet the density distribution complementarity index threshold and surface property compatibility conditions are determined as the final coal types to be blended.

4. The method for preparing coking coal with improved washability for extremely difficult coal preparation according to claim 1, characterized in that, Step S3 further includes: The minimum requirement is that the proportion of extremely difficult-to-select coal in the mixed coal should not be less than 50%. With the mixed coal having a separation density of ±0.1 g / cm³ 3 The requirement that the content of near-heavy substances within the specified range not exceed 40% is an optional constraint. A multi-objective optimization model is established with the theoretical yield and economic value of refined coal as dual objectives. By systematically searching within the proportion space that satisfies the lower limit constraint, selectivity constraint, coking performance constraint, and quality constraint, the Pareto optimal proportion set is obtained. From the Pareto optimal proportion set, the proportion scheme that balances yield and economy is selected as the optimal proportion.

5. The method for preparing coking coal with improved washability for extremely difficult coal preparation according to claim 1, characterized in that, In step S3, the coking performance constraints of the coking coal caking index G and the plastic layer thickness Y are obtained through a nonlinear prediction model. The nonlinear prediction model takes the volatile matter, vitrinite content and inertinite content of each type of coal as input variables, and the caking index G and plastic layer thickness Y of the blended coking coal as output variables. It is trained and verified using measured coal blending coking test data.

6. The method for preparing coking coal with improved washability for extremely difficult coal preparation according to claim 1, characterized in that, Step S4 also includes a heavy media suspension stability control step: Real-time monitoring of the density and viscosity of the feed suspension to the heavy medium cyclone separator; When the deviation between the suspension density and the set sorting density exceeds ±0.02 g / cm³ 3 At the same time, adjust the amount of magnetite powder added or the flow rate of the circulating medium to correct the density deviation; When the viscosity of the suspension exceeds 25 mPa·s, increase the processing capacity of the media purification circuit to reduce the content of non-magnetic fine mud in the suspension. The ash content of the sorted clean coal products is tested online, and the sorting density is finely adjusted in real time based on the test results.

7. The method for preparing coking coal with improved washability for extremely difficult coal preparation according to claim 1, characterized in that, Step S4 also includes a collaborative optimization step for the coal slurry water system: Mineral composition analysis was performed on the coal slurry water generated from the mixed washing process, and the main clay mineral types in the coal slurry water were determined based on the analysis results. Select the appropriate type of flocculant based on the main clay mineral type, and control the amount of flocculant added to reduce the turbidity of the coal slurry water to below the preset turbidity standard. Monitor the solid content in the circulating water. When the solid content exceeds the preset upper limit of concentration, initiate the concentration overflow discharge and clean water replenishment operation.

8. The method for preparing coking coal with improved washability for extremely difficult coal preparation according to claim 1, characterized in that, In step S1, at least six standard density grades include 1.3 g / cm³. 3 1.4g / cm 3 1.5g / cm 3 1.6g / cm 3 1.7g / cm 3 and 1.8g / cm 3 The interpolation fitting employs a cubic spline interpolation method with a density resolution of not less than 0.01 g / cm³. 3 .

9. A coking coal preparation system for improving the washability of extremely difficult coal preparation, characterized in that, A method for preparing coking coal with improved washability for extremely difficult coal preparation as described in any one of claims 1-8, comprising: The coal quality analysis module is configured to collect and process float-sink test data for the target extremely difficult coal to be made and multiple candidate coal types, generate the density-yield distribution function of each coal type, calculate the density distribution complementarity index between the target extremely difficult coal to be made and each candidate coal type, and screen out the compatible coal types that meet the preset threshold. The proportioning optimization module is connected to the coal quality analysis module and is configured to receive the density-yield distribution function and coking performance parameters of each coal type, generate the theoretical sorting curve of the mixed coal by weighted superposition, solve the optimal proportion to maximize the clean coal yield under coking performance constraints and ash content constraints, and output the optimal proportion scheme and the corresponding sorting density setting value. The coal blending and feeding module is signal-connected to the proportioning optimization module and configured to control the feeding amount of the target extremely difficult coal and the blended coal according to the optimal proportioning scheme, and to transport the mixed coal to the inlet of the heavy medium cyclone. The heavy medium separation module includes a heavy medium cyclone separator and a medium circulation system, configured to perform density separation on the mixed coal according to the separation density set value to obtain coking coal; A quality monitoring module is installed at the clean coal output end of the heavy medium separation module. It is configured to detect the ash content of the clean coal in real time and feed the detection results back to the ratio optimization module and the heavy medium separation module for adjusting the ratio or separation density.

10. A coking coal preparation system for improving the washability of extremely difficult coal preparation according to claim 9, characterized in that, The coal quality analysis module includes a buoyancy data digitization unit and a complementarity calculation unit. The buoyancy data digitization unit is configured to convert discrete yield and ash content data of each standard density level into a continuous density-yield distribution function through cubic spline interpolation. The complementarity calculation unit is configured to calculate the density distribution complementarity index based on the density-yield distribution function and output the ranking results. The quality monitoring module includes an online ash content analyzer, which employs dual-energy X-ray transmission or transient gamma neutron activation analysis. The proportioning optimization module has a built-in nonlinear prediction model. The nonlinear prediction model uses the volatile matter, vitrinite content and inertinite content of each coal type as input variables to predict the caking index G value and the plastic layer thickness Y value of the mixed clean coal.

Images