Analysis and Improvement of Oversize Goaf Backfill Engineering Based on Fuzzy Theory

Abstract

For the mine of large goaf, there are many factors that affect the filling quality. In order to improve the filling quality, it is necessary to identify the factors that have a significant impact. In this paper, the fuzzy mathematical method was used to analyze the five main factors (PC32.5cement sand ratio, slurry concentration, number of slurry lowering, dehydration and drainage, and tailing particle size grading) affecting the filling quality. The priority of each influence factor was calculated, and the priority set was established. It is found that the slurry launching point and particle size grading of the tailings have an obvious effect on filling quality. After increasing the slurry launching point and optimizing the particle size grading, the strength and uniformity of the filling body are improved.

1. Introduction

As a relatively popular research direction in recent years, much research about backfill methods has provided beneficial help to solve practical problems [1,2,3]. A methodology for selecting a mineral deposit development technology and a rational backfill composition were developed. The structure of the methodology includes fuzzy models and algorithms that provide the processing of large amounts of information [4]. An algorithm for choosing the optimal development system for safe mining is developed. The authors proposed a zonal division of the mass to determine the initial structure of a multi-agent system [5]. In order to increase the strength characteristics of the filling material, fullerene starlene is used as a nanomodified additive [6]. The rheological behavior is important for transportation in the pipe [7]. In the daily mining process of metal deposits, in order to improve production efficiency and reduce production costs, enterprises generally choose the Vertical Crater Retreat mining method (VCR method) according to the condition that the stability of the ore body and surrounding rock allow [8]. However, when a large quantity of ore body is continuously mined from the ore bed, significant quantities of goaves will be produced. In order to prevent the occurrence of such secondary man-made geological disasters, the goaves must be backfilled to effectively protect the surface and ensure safe and continuous large-scale mining [9,10,11]. The filling quality of the mined-out area has an important influence on the second-step pillar mining. If the homogeneity or strength of the backfill is poor, the blasting vibration in the second step of the mining process may cause the stability problem of the backfill. A large area collapse of the backfill will cause a surge in the ore loss rate and dilution rate, thus greatly increasing the cost of mining and beneficiation. After fully considering the actual situation of backfill, the quality of backfill in the study of two-step mining technology was analyzed [12]. In order to control the filling quality from the source, the main influencing factors in the design of the filling system were regulated, which plays a helpful role in the application of filling [13,14].

The operation effect of the filling system is usually a comprehensive reflection of the interaction of multiple factors, including both the performance of the filling material and the operation stability of the filling system. Therefore, it is necessary to analyze the main influencing factors and take appropriate technical measures to continuously improve the filling quality on the original basis [15,16,17]. For example, in Wancheng Lead-zinc Mine, expanding production scale means the capacity of the original filling system can no longer meet the demand for goaf filling. Therefore, the optimization design of sand bin circulation, slurry preparation, transportation efficiency and other aspects has been carried out, and the demand for safe mining has been met after the improvement of filling capacity [18]. In the ultra-deep mining process of the Hongtuoshan Copper and Zinc Mine, the conveying range of filling slurry has been extended by laying an underground vertical sand silo and pumping system, and the failure rate of transportation has been reduced while the filling quality has been significantly improved [19]. In order to improve the backfill material of mechanical properties, some research about filling aggregate and PC32.5 cement materials was conducted. The effect of the mineral composition of filling materials on the micro and macro performance of backfill was analyzed. Under reasonable conditions of production, these studies promote the improvement of the quality of the filling material [20,21,22]. A fast and reliable detection method of elemental analysis was used to investigate the effect of sulfur ion migration on the strength and quality of backfill [23]. The use of alkali excitation can promote the hydrolysis of blast furnace slag, generating sufficiently stable hydrolytic products to effectively consolidate high-sulfur tailings [24]. In terms of improving the performance of coarse aggregate filling materials, researchers conducted performance tests on cementitious materials and crushed stone composite materials and obtained an optimal plan for the grinding time of cementitious materials through experimental analysis [25]. In terms of investigating the properties of some filling materials, relevant research adopts neural network principles to highlight the adjustment of main influencing factors. By improving material combinations and some engineering conditions, reasonable solutions are provided for obtaining excellent mechanical properties of filling materials [26,27].

Through the analysis of the filling case of many mines, it is found that the main factors influencing the strength of the filling body are continuous and stable aggregate and binding material, the admixture, water and the preparation of the filling paste. Therefore, according to the field conditions, some mining needs to actively seek potential filling materials. Due to the great differences in the performance of filling materials in different regions, it is necessary to conduct corresponding analysis and research for specific mines to guide on-site production. For example, Chen conducted rheological and mechanical properties research on cemented paste backfill with the addition of anionic polyacrylamide, and research on heavy metal leaching characteristics provided a reliable basis for ecological and environmental protection treatment of solid waste [28]. As a large-scale underground mining mine in the Huaibei Plain area, the current production capacity of the Caolou Iron mine in this study has reached three million t/a. The surrounding area of the mine is densely populated and covered with fertile fields, which require high-quality backfill to protect the surface from collapse. Therefore, it is extremely urgent to study its daily and future backfill conditions. In order to improve mining efficiency and reduce the cost, the Caolou Iron Mine mainly uses the VCR method to mine ore bodies [29,30]. After sampling by geological drilling, it was found that the backfill slurry of the Caolou Iron Mine had been seriously segregated and stratified, the partial backfill body showed the character of loose sand, and the RQD value was low. This will bring great safety risks to pillar mining [31]. Therefore, in order to realize the safe mining of the second-step pillar, it is necessary to take appropriate measures to improve the quality of the filling to provide reliable engineering conditions. In the two-step mining process, the safety and stability of the ore body and rock layers can significantly improve the utilization rate of mineral resources, create economic benefits for enterprises, and at the same time, bury harmful solid waste deep underground, thus achieving the goal of eliminating the harm of heavy metals and harmful substances.

As a large-scale underground mining iron mine, Caolou Iron Mine produces a goaf of about 1 million m³ every year. With the deepening of mining, stress concentration has occurred in the local rock stratum, which is mainly manifested in the local fracture of the roof of the drill chamber, the cracking and falling of reinforced shotcrete layers, and the cracks of different degrees in the walls of individual houses on the surface. It is important to fill the goaf to maintain the stability of the rock stratum and surface. Based on the geological drilling investigation in the early stage, the quality of the backfill core was measured and analyzed. It is found that the overall quality of the backfill is poor. Various engineering influencing factors are diversity and complexity. Identification of some uncertain factors is the key to solving engineering problems. In order to fully grasp the causes of the low filling quality of mine, it is necessary to combine related similar studies and adopt corresponding technical and theoretical means to quantify the influencing factor index so as to provide solutions to practical engineering problems [32,33,34].

In some engineering and related issues, research is often necessary to analyze the impact of multiple factors on the implementation effect, among which the discrimination of non-quantitative factors is more complex. In this process, while identifying key factors, the analysis process is transformed into quantification, and finally, the quantitative conclusions guide subsequent work [35,36]. In complex traffic engineering problems, researchers use fuzzy mathematics and neural network methods to collect the main factors that affect traffic engineering, analyze the variation patterns of variables, and provide important technical support for rail transit safety [37,38]. In the mining of granite quarries, researchers introduced the fuzzy VIKOR method to analyze environmental impact factors, effectively avoiding the impact of subjective consciousness on evaluators. The application of this method has played a demonstration role in promoting similar problems [39]. In terms of engineering site selection, researchers analyzed 16 alternative solutions based on GIS and fuzzy Einstein ordinal priority methods and determined a reasonable engineering site after sorting the calculation results. This research result also provides a reference for similar problems [40]. Usually, only a single method is used to analyze the filling quality while ignoring the comprehensive impact of multiple factors. In this study, the importance of five main factors, including cement sand ratio, slurry concentration, number of slurry lowering, dehydration and drainage, and tailing particle size grading, was evaluated by fuzzy mathematical methods. Corresponding quantitative indicators were given. Evaluation sets were developed after determining the elements of the evaluation factor set and their weights. Finally, the influence degree of various factors on the filling quality was ordered on the basis of the priorities themselves. These factors would then be improved in an order according to their priorities.

2. Backfill Engineering and System Operation

2.1. Overview of Mine

The Caolou Iron Mine was originally designed to have an annual yield of 1.5 million tons. In recent years, its production capacity has been expanded to three million tons per year and up to 3.3 million tons. The primary production and sales businesses of the mine include iron ore mining, mineral processing, and the production of iron powder. Because the mine is covered by 170 m of Quaternary strata, there are 1–4 flow sand layers between the two layers of strata, and the surface area is occupied by dense villages, ponds, and farmland. There are high risks of surface subsidence and sand conduction damage in the mining environment. In order to prevent the occurrence of such secondary man-made geological disasters, the goaves must be backfilled to effectively protect the surface and ensure safe and continuous large-scale mining. The ore loss ratio and the dilution ratio need to be reduced to achieve the comprehensive benefits of the mining enterprise.

2.2. Construction of Backfill Preparation System

To effectively use the backfill mining method in the Caolou Iron Mine, multiple backfill approaches were evaluated. Considering the differences in the backfill slurry preparation and transportation processes, the principle of maximizing the utilization of the solid wastes produced by the mining process and the tailings of the dressing plant were selected as the aggregate for the backfill. Based on an earlier laboratory test and a semi-industrial test, the backfill system in the Caolou Iron Mine was designed and constructed in 2005–2006. It would generate more than goaf of 1 × 10⁶ m³ during the production of three million tons of ore. After project defense and on-site investigation, the plant and processing facilities of the backfill station were constructed in the form of an overall steel structure to shorten the construction period and start operation as soon as possible. With an increase in the mining capacity, the backfill system was expanded and optimized in 2013–2015 to meet the demand for a backfill slurry preparation capacity of 1 × 10⁶ m³. The construction process of the backfill system is shown in Figure 1. The constructed backfill system is shown in Figure 2.

2.3. Operation of Backfill System

After the plan design, geological engineering survey, process construction drawing design, equipment and pipeline installation and system debugging were completed, the backfill system was smoothly put into daily operation. The tailings slurry was settled circularly in four vertical tailings silos with a volume of 1100 m³. When a predetermined concentration was reached, the 1.8–2.5 m³ of clear water were reserved on the sand surface to adjust the tailing concentration. The air of high pressure was provided by a screw air compressor to perform the pneumatic mixing. The concentration of the tailings slurry was approximately 59–65%. The tailings slurry was uniform and viscous. Under the strong mixing, the tailings slurry continuously flowed up from the bottom of the tailings silo, thus forming dynamic convective circulation from the surface of the material to the bottom of the silo. The positions of the tailings particles with different particle sizes can be dynamically exchanged in this stage. Water, tailings (with a SiO₂ content of 68 to 72% in tailings), and PC32.5 (11~20%) cement were added to the mixer in turn. They are premixed with a dual horizontal shaft mixer with a rotational speed of 40 to 65 rpm/min and then strongly stirred with a high-speed activation mixer with a rotational speed of 300 to 600 rpm/min. The mixing and homogenization time is about 30 to 45 s. When the outflow of filler slurry does not contain agglomerated cement particles, it is considered that various materials have been mixed evenly. After the preparation of the slurry using the compressed air was finished, the tailings were discharged from the vertical tailings silo. A high concentration of 70–72.5% was achieved for the backfill slurry. A high-speed activation mixer and a double-shaft mixer were used to stir the slurry to make it homogeneous and ensure that it easily flowed. Driven by its own gravity, the slurry was continuously transported to the goaf through a long-distance filling pipeline.

In the storage of the concentrated tailings slurry and the production of the slurry using compressed air, the cohesive force and the friction force between the tailings particles were difficult for the pneumatic stirring force to overcome. When the concentration of the tailings in the silo was high, the slurry gradually changed from saturated to unsaturated. The tailings particles that lost water gathered. When some of the 15 nozzles at the bottom of the sand storage were blocked by the extremely fine tailings particles, the tailings accumulation problem in the conical bottom of the silo became serious. Due to this major problem, only 67.5% (in volume) of the tailings could be discharged, and 32.5% of the tailings were blocked at the bottom of the storage silo. Only high-pressure water could wash out the deposited tailings. After the tailings slurry was diluted, it was discharged through a discharge pipe. This process was the cause of the large fluctuations in the concentration and flow rate, which affected the backfill quality of the goaf. In the processes of pressurized slurry production and material mixing, the displayed values of the flow meter and the concentration meter and measured value were quite anastomotic, providing the principle to set the mass of the cemented backfill material to match the sand discharge flow rate and the concentration. The process of compressed air pulping and filling slurry preparation was observed, which was used to visually analyze the problems in the filling slurry preparation process. Figure 3 is the pneumatic mixing tailings slurry in the tailings silo. Figure 4 is the uniform backfill slurry produced after adding cemented materials.

3. Fuzzy Analysis of Backfill Quality

3.1. Backfill Conditions of Goaf

When the two-step pillar mining method is used, it is difficult to determine the physical and mechanical properties of the backfill body in the goaf. To grasp the actual situation of the backfill body and provide reliable technical parameters for pillar mining, a geological drilling sampling survey of the backfill body in the goaf was carried out. According to the drilling investigation and the strength tests in the filling field, under the existing conditions of #17R, it is difficult to ensure the stability of the backfill body during the pillar mining process in the local area [41]. The strength distribution of backfill body #17R was extremely non-uniform. The core of the backfill body extracted from the drilling location was loose, and sometimes the core could not even be extracted. The setting and hardening conditions of the backfill body at the bottom of the mine and at the top near the upper plate area were satisfactory. The middle of the backfill body and its top near the lower plate area was loose and had an extremely low strength. According to the stability analysis of the backfill body, it was difficult for the local backfill body to meet the independence and stability at a certain height. When the exposed height was large, the risk of instability and collapse of the backfill body was very high. Therefore, appropriate technical measures should be taken in the process of pillar mining to ensure that the backfill body meets the mining requirements.

Based on the geological investigation, drilling tests and stability analysis of the backfill body at #17R, there was an excessively low concentration of backfill slurry, water accumulation (water of 5000 m³ was discharged using a submersible pump), failure of drainage due to blocking by the filter cloth, and excessive air pressure fluctuation of the air compressor. The reasons for these problems are that the backfill system was being operated for the first time, so the management and operation personnel were not proficient. Moreover, there were defects in the detection mechanisms related to the concentration and flow rate of the backfill slurry, the vertical sand bin level, the strength of the backfill body, and the addition of gelling material. The operation personnel did not have a sufficient understanding of the purpose and significance of the goaf backfill. As a result, the above factors resulted in the backfill body having uneven strength and poor quality. Because single-point feeding and a low concentration were used, the backfill slurry experienced partial segregation when it flowed along the long axis of the mining site. In the area relatively enriched in coarse particles, there were insufficient cementing materials, resulting in a loose and wet filling body with extremely low strength. Therefore, the quality of the goaf backfill urgently needs to be improved in the future.

3.2. Fuzzy Analysis of the Backfill Quality

The backfill process is a complex engineering system. Similar to many other studies [42,43,44,45], understanding and managing the different factors also affects the backfill system. Generally, filling material is similar to cementitious materials, and their strength is greatly affected by the amount of cement and the water–cement ratio. For concrete with different particle sizes, their strength also varies greatly. Therefore, when analyzing the mechanical properties of filling materials with less binder material, some analysis methods of ordinary concrete can still be used for reference. In order to determine the importance of the factors affecting the quality of the backfill body at #17R, such as the lime–sand proportion control A₁, slurry concentration control A₂, slurry launching point A₃, coalmine dehydration drainage A₄, and particle size grading of the tailings A₅, eight experts in the filling field were invited to evaluate these factors using fuzzy math methods.

The evaluation factor set was defined as U = {u₁, u₂, u₃, u₄, u₅}, where u₁ is the technical level, u₂ is the operating difficulty, u₃ is the self-control level, u₄ is the human factor, and u₅ is the awareness level. The weighting of the evaluation factor set was W⁰ = {0.30, 0.15, 0.10, 0.20, 0.25}. The importance of the influencing factors on the backfill quality was classified into five levels using an evaluation set V = {v₁, v₂, v₃, v₄, v₅}, where v₁ is extremely important (0.75 point), v₂ is very important (0.65 point), v₃ is important (0.45 point), v₄ is average (0.25 point), and v₅ is unimportant (0.15 point).

According to the evaluation scale, each evaluation factor for influencing factor A₁ was evaluated. It is shown that two experts believed u₁ was extremely important for A₁, three experts believed it was very important, two experts believed it was important, and one expert believed it was average. Therefore, the membership degrees of the evaluation scales were calculated as follows:

$$\begin{gathered} r_{11}=\frac{2}{8}=0.250, r_{12}=\frac{3}{8}=0.375, r_{13}=\frac{2}{8}=0.250, r_{14}=\frac{1}{8}=0.125, r_{15}=\frac{0}{8}=0.000, \\ {r}_{21}=\frac{2}{8}=0.250, r_{22}=\frac{2}{8}=0.250, r_{23}=\frac{4}{8}=0.500, r_{24}=\frac{0}{8}=0.000, r_{25}=\frac{0}{8}=0.000, \\ {r}_{31}=\frac{2}{8}=0.250, r_{32}=\frac{3}{8}=0.375, r_{33}=\frac{3}{8}=0.375, r_{34}=\frac{0}{8}=0.000, r_{35}=\frac{0}{8}=0.000, \\ {r}_{41}=\frac{2}{8}=0.250, r_{42}=\frac{1}{8}=0.125, r_{43}=\frac{3}{8}=0.375, r_{44}=\frac{2}{8}=0.250, r_{45}=\frac{0}{8}=0.000, \\ {r}_{51}=\frac{4}{8}=0.500, r_{52}=\frac{3}{8}=0.375, r_{53}=\frac{1}{8}=0.125, r_{54}=\frac{0}{8}=0.000, r_{55}=\frac{0}{8}=0.000. \end{gathered}$$

Based on the information in

Table 1. Factor weighting and evaluation scale.

Evaluation Factor		u1 Technical Level	u2 Operation Difficulty	u3 Self-Control Level	u4 Human Factor	u5 Depth of Understanding
Weighting W0		0.3	0.15	0.1	0.2	0.25
Evaluation scale of ratio control	0.75	2	2	2	2	4
	0.65	3	2	3	1	3
	0.45	2	4	3	3	1
	0.25	1	0	0	2	0
	0.15	0	0	0	0	0
Evaluation scale of concentration control	0.75	2	1	4	3	4
	0.65	3	5	2	3	1
	0.45	2	2	2	0	2
	0.25	1	0	0	2	1
	0.15	0	0	0	0	0
Evaluation scale of the drop point	0.75	4	2	4	2	5
	0.65	3	5	2	6	2
	0.45	1	1	1	0	0
	0.25	0	0	1	0	1
	0.15	0	0	0	0	0
Evaluation scale of dehydration and drainage	0.75	3	2	5	4	3
	0.65	3	3	2	1	4
	0.45	2	3	0	3	1
	0.25	0	0	1	0	0
	0.15	0	0	0	0	0
Evaluation scale of particle size grading	0.75	5	5	4	7	6
	0.65	2	3	4	1	1
	0.45	1	0	0	0	1
	0.25	0	0	0	0	0
	0.15	0	0	0	0	0

, the membership matrix of A₁ was obtained.

$$R_1=\left[\begin{array}{ccccc}0.25& 0.375& 0.25& 0.125& 0\\ 0.25& 0.25& 0.5& 0& 0\\ 0.25& 0.375& 0.375& 0& 0\\ 0.25& 0.125& 0.375& 0.25& 0\\ 0.5& 0.375& 0.125& 0& 0\end{array}\right]$$

The comprehensive evaluation vector S₁ was calculated.

$$S_1=W^0{R}_1=(0.3125 0.30625 0.29375 0.0875 0).$$

Then, the priority N₁ of A₁ was given by

$$N_1=S_1^0{V}^{\mathrm{T}}=0.587500$$

The membership matrices R₂–R₅ for A₂–A₅ and the corresponding priority levels N₂–N₅ were calculated using the same method, based on the membership degree rᵢ₁ of each evaluation scale as follows:

$$\begin{gathered} R_2=\left[\begin{array}{ccccc}0.25& 0.375& 0.25& 0.125& 0\\ 0.125& 0.625& 0.25& 0& 0\\ 0.5& 0.25& 0.25& 0& 0\\ 0.375& 0.375& 0& 0.25& 0\\ 0.5& 0.125& 0.25& 0.125& 0\end{array}\right] \\ {R}_3=\left[\begin{array}{ccccc}0.5& 0.375& 0.125& 0& 0\\ 0.25& 0.625& 0.125& 0& 0\\ 0.5& 0.25& 0.125& 0.125& 0\\ 0.25& 0.75& 0& 0& 0\\ 0.625& 0.25& 0& 0.125& 0\end{array}\right] \\ {R}_4=\left[\begin{array}{ccccc}0.375& 0.375& 0.25& 0& 0\\ 0.25& 0.375& 0.375& 0& 0\\ 0.625& 0.25& 0& 0.125& 0\\ 0.5& 0.125& 0& 0& 0\\ 0.375& 0.5& 0.125& 0& 0\end{array}\right] \\ {R}_5=\left[\begin{array}{ccccc}0.625& 0.25& 0.125& 0& \begin{array}{c}\end{array}0\\ 0.625& 0.375& 0& 0& \begin{array}{c}\end{array}0\\ 0.5& 0.5& 0& 0& \begin{array}{c}\end{array}0\\ 0.875& 0.125& 0& 0& \begin{array}{c}\end{array}0\\ 0.75& 0.125& 0.125& 0& \begin{array}{c}\end{array}0\end{array}\right]. \\ {S}_2=W^0{R}_2=(0.34375 0.3375 0.2625 0.11875 0) \\ {N}_2=S_2^0{V}^T=0.625 \\ {S}_3=W^0{R}_3=(0.44375 0.50625 0.06875 0.04375 0) \\ {N}_3=S_3^0{V}^T=0.70375 \\ {S}_4=W^0{R}_4=(0.40625 0.36875 0.1625 0.0875 0) \\ {N}_4=S_4^0{V}^T=0.639375 \\ {S}_5=W^0{R}_5=(0.69375 0.2125 0.06875 0 0) \\ {N}_5=S_5^0{V}^T=0.689375 \end{gathered}$$

Therefore, the importance of these five factors can be ranked as follows to provide useful decision information for technical improvement.

$$N_3>N_5>N_4>N_2>N_1$$

The importance ranking of the factors from high to low is as follows: slurry launching point A₃, particle size grading of the tailings A₅, coalmine dehydration drainage A₄, slurry concentration control A₂, and lime-sand proportion control A₁.

3.3. Measures to Improve Backfill Quality

The quality of the goaf backfill is affected by multiple factors. As a complex engineering system, its operation effectiveness is affected by the material supply, the preparation method in the surface backfill workshop, and the underground goaf environment. Necessary measures should be taken to solve the practical problems revealed in the fuzzy analysis of the backfill engineering system.

(1) In the backfill process at #17R, it was found that a single feeding point may cause slurry segregation. However, because blasting was implemented on both sides of the mine and the construction was near the end, the feeding pipe could not be installed on two sides of the upper drilling cavern. This issue resulted in the entire mine only having one slurry feeding location. Therefore, it is necessary to use two or three scattered feeding points in the continuous backfill of other goaves.

(2) The particle size grade and the chemical composition of the tailings are critical to the transportation performance of the backfill slurry. For the tailings of the Caolou Iron Mine, coarse particles were abundant: the diameter of particles exceeding 20 μm was approximately 77.89%. Because the SiO₂ content was 71.06%, the hardness of the tailings particles was high, and the wear on the equipment was greater than the wear produced by the fine tailings particles in other mines. Therefore, the local loss of very fine tailings should be reduced as much as possible when the tailings enter the tailings silo. A flocculation sedimentation study was carried out in the Caolou Iron Mine and was applied in the production. These efforts improved the grading properties of the tailings particles.

(3) A suspension corrugated filter tube was used to achieve drainage. A hole was drilled at the top of the rock cavern on the side opposite the feeding location. A rebar bolt of ϕ28 mm was installed. The top of the filter tube was firmly fixed to the anchor rod. The bottom part of the filter tube passed through the V-shaped trench and a reserved hole in the retaining wall. Usually, there are two filter tubes. Because the tailings were coarse and settled fast, the water precipitated quickly. Through field observations and practice, 2–4 corrugated filter pipes must be added in appropriate positions in order to discharge the water from the slurry as quickly as possible.

(4) The slurry concentration control is crucial to ensuring the quality of the backfill. According to a computer record and a manual verification test, when the concentration was 68–72%, the slurry flowed smoothly, and the corresponding slump was 27.5–24.5 cm. When the concentration of the backfill slurry was high, the bleeding of the slurry decreased. Therefore, the concentration of the backfill should be increased as much as possible to ensure acceptable fluidity when another goaf is backfilled in the future. Moreover, the water used to clean the pipes before and after the backfill should not be discharged into the backfill goaf to avoid the slurry being diluted, which would negatively affect the quality of the backfill.

(5) The lime–sand proportion control is a key factor affecting the cost of backfill. The above improvement provided a technical basis for controlling the amount of gelled material. After optimization of the aggregate grading and the improvement of the concentration, the lime–sand proportion control needs to be optimized. Because the backfill capacity of the mine was close to 1 × 10⁶ m³ per year and the daily backfill capacity was greater than 3000 m³, the PC32.5cement consumption was huge. Therefore, a cementitious material was developed to represent PC32.5cement to decrease the comprehensive cost. A cementitious material production line was developed in the mine, which will play a positive role in controlling the cost of backfill.

4. Conclusions

In the investigation of filling quality at the early stage of second-step pillar mining in Caolou Iron Mine, the fuzzy mathematical method was used to analyze the five main factors affecting the filling quality, e.g., the PC32.5cement sand ratio A₁, slurry concentration A₂, number of slurry lowering A₃, dehydration and drainage A₄, and tailing particle size grading A₅. The evaluation sets consisted of U₁, U₂, U₃, U₄, and U₅ elements, respectively, representing technical level, operation difficulty, degree of self-control, human factors, and height of understanding. On the basis of calculating evaluation scale r_i1, when calculating the membership degree, the membership matrix R_n(n = 1–5) relating to evaluation sets A_n(n = 1–5) was obtained the according to the factor weights and evaluation scale data. The priority N_n of each influence factor, A_n, was then calculated, and the priority set was established.

According to the importance of the influence factors of the corresponding priority index, for practical engineering problems, fuzzy mathematics methods are effective for the analysis and quality improvement of very large goaf filling. The top priority indexes were N₃ = 0.70375 and N₅ = 0.689375, which indicates that the most urgent problems are to increase the number of slurry feeding points and improve the tailing particle size gradation. The minimum priority index N₁ = 0.5875 was obtained.

When analyzing engineering problems using fuzzy mathematics theory, in order to better guide actual production with research conclusions, it is necessary to further select the influencing factors after preliminary screening. The effects of different influencing factors may vary with working conditions, so specific analysis is required for specific problems. The engineering case of goaf filling in this study is complex and variable. The method of converting qualitative indicators to quantitative indicators makes it easy to judge and adjust the influencing factors, thereby enabling the ingenious application of fuzzy mathematics in the process of improving mining efficiency and providing a good reference for expanding the application of mathematical knowledge in mining engineering and other aspects.