Optimization study on transverse mining zoning during the capacity expansion stage of nearly horizontal open-pit coal mines

Abstract

To address key challenges during the capacity expansion of the Baoqing Chaoyang Open-Pit Coal Mine from 7 to 11 Mt/a—such as insufficient working-line length, excessive advancing intensity, poor coordination between mining zones, elevated internal dump slope risks, and constrained dumping capacity—this study develops a zoning optimization method based on Unascertained Measure Theory (UMT). A working-line optimization model, constructed from coal seam thickness, density, production targets, and advancing dynamics, indicates a rational length range of 1350–2050 m at 11 Mt/a. Guided by coal seam occurrence and the original layout, the initial longitudinal-mining area is reorganized into four transverse primary zones. An eight-dimensional evaluation index system is established, covering geological, geotechnical, hydrogeological, topographic, engineering, economic, environmental, and social factors. Combined weighting using the Analytic Hierarchy Process and the Entropy Weight Method, together with multiplicative normalization, yields a hybrid subjective–objective weighting set. On this basis, four-level UMT evaluation functions and an evaluation matrix are constructed, and comprehensive assessments are performed using a confidence recognition criterion. The results show that Scheme II achieves the best overall performance in terms of economic efficiency, transportation, safety, and resource utilization, with a V₁-level confidence of 0.7058, significantly higher than the other schemes. Under Scheme II, the four zones provide 971.2 Mt of recoverable raw coal, an average stripping ratio of 5.8, and a maximum service life of 34.3 years. The proposed integrated framework of working-line design, mining-zone zoning, and uncertainty-based evaluation offers a practical decision-support tool for capacity expansion planning in large-scale open-pit coal mines.

Introduction

As the world’s largest coal producer and consumer, China depends heavily on open-pit coal mining to sustain its energy supply system. With the ongoing energy transition and growing demand for efficient and cost-effective resource extraction, open-pit mining has become the preferred method for large-scale coal bases owing to its operational flexibility, simple production organization, and low unit capital cost. However, conventional monolithic mining approaches have become increasingly inadequate to satisfy the scheduling and management demands of modern large-scale operations. Traditional holistic mining strategies often involve high up-front investment, prolonged construction periods, extensive equipment and labor demands, overly long working lines, extended haulage distances, and increased slope instability. To address these challenges, regionalized and phased mining has emerged as an essential strategy for capacity expansion and coordinated development in open-pit mines^1,2.

In the field of working-line length optimization, early research mainly focused on developing and applying multi-objective optimization models. Yu Rushou and Xi Yongfeng³ (1986) first proposed a multi-criteria optimization framework for working-line length considering annual advance rate, slope stability, and stripping efficiency, which they validated through case studies at representative sites such as the Shengli Open-Pit Mine. Wang Haiqing and Li Jingzhu⁴ (2007) applied this framework to simulate stripping ratios and haulage distances un-der multiple slope expansion scenarios during deepening at the Shengli Mine, thereby identifying the economically optimal working-line length. Cao Bo et al.⁵ (2017) further refined the optimization process using 3D modeling in 3Dmine software, evaluating alternative working-line length schemes after pit expansion through comparative analysis of coal–rock volumes, stripping ratios, and haulage distances. More recently, Zhao Jingchang et al.⁶ (2024) employed a Monte Carlo–based optimization approach to redesign the working-line configuration at the Halwusu Open-Pit Mine, addressing the issue of shortened working lines caused by boundary constraints of mining rights and thereby ensuring continuous production. Regarding mining-district delineation, several studies have proposed dynamic and phased partitioning approaches in response to complex geological settings and capacity-expansion demands. For narrow and elongated open-pit mines with inclined coal seams, Cao Bo et al.⁷ (2019) proposed an optimized pit-boundary strategy that integrates longitudinal-to-transverse mining transitions with variations in overburden flow rates. Under conditions of increasing production capacity, Ma Li et al.⁸ (2023) developed a district-partitioning model for the Datang Shengli Dong No. 2 Open-Pit Mine by integrating geological modeling with dump-capacity analysis, achieving synergistic optimization between mining-block layout and waste-disposal planning. In terms of scheme evaluation and selection, multi-criteria decisionmaking (MCDM) methods have become the dominant approach. Zhao Hongze et al.⁹ (2021) incorporated several curved turning strategies into an Analytic Hierarchy Process (AHP) framework for the anti-cline-crossing mining design at the Anjialing Open-Pit Mine and used the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) to identify the optimal “L-shaped fan-turn” scheme. Similarly, Chen Dawei et al.¹⁰ (2024) implemented an AHP-TOPSIS-based model to select optimal mining schemes by constructing a hierarchical evaluation system covering stripping ratio, dumping space, haulage distance, and mining difficulty. Internationally, researchers have also emphasized the importance of dynamic planning and uncertainty management in open-pit mine optimization. For instance, Brika and Findlay et al.^11,12 (2023、2024) proposed stochastic mine design frameworks incorporating geological uncertainty and price volatility into long-term production scheduling. Hill¹³(2022)developed multi-stage optimization models integrating pit sequencing with haulage network planning. More recent studies by Pavloudakis¹⁴ (2024) focused on adaptive pit design methods that respond to real-time production feedback, further enhancing the flexibility of large-scale open-pit operations.

Despite significant progress, most existing studies share several limitations. First, optimization models often rely on deterministic geological and economic parameters, which cannot adequately capture the inherent uncertainty in resource distribution, equipment performance, and market fluctuations. Second, many existing zoning and working-line designs emphasize single objectives, such as minimizing stripping ratios or haulage distances while neglecting the integrated consideration of economic, environmental, and safety constraints. Third, there is a lack of unified evaluation frameworks that can handle both quantitative indicators (e.g., stripping ratio, output, cost) and qualitative factors (e.g., stability, coordination, sustainability) in a systematic and flexible manner. Furthermore, in the context of near-horizontal open-pit coal mines during capacity expansion, the coordination between mining working-line design, internal dumping stability, and zone connection remains underexplored. To overcome the aforementioned challenges, this study develops a systematic optimization framework for mining zone delineation and working-line configuration based on Unascertained Measure Theory. The main objectives are as follows: Establish a working-line length optimization model considering coal seam thickness, density, production capacity, and advancement rate, determining the rational range of working-line lengths for an annual capacity of 11 Mt. Evaluate the technical and economic feasibility of transitioning from longitudinal to transverse mining, based on geological occurrence characteristics and existing pit geometry. Construct a comprehensive multi-factor evaluation system incorporating eight categories of influencing factors geological, geotechnical, hydrogeological, topographic, engineering, economic, environmental, and social and to derive combined subjective objective weights using the AHP entropy method. Apply Unascertained Measure Theory for scheme evaluation and selection, building four level evaluation functions and a confidence recognition criterion to determine the optimal zoning scheme. The findings are intended to provide theoretical guidance for ensuring the safe and efficient capacity expansion of the Baoqing Chaoyang Open-Pit Coal Mine and to serve as a valuable reference for the sustainable planning and management of large-scale open-pit coal mines in resource-based regions.

Overview of the study area

The Chaoyang Open-pit Coal Mine, operated by Guoneng Baoqing Coal Chemical Co., Ltd., is located in Baoqing County, under the jurisdiction of Shuangyashan City, Heilongjiang Province, China. Administratively, the mine area falls within Chaoyang Township and the 3rd, 5th, and 6th divisions of Farm 852. Construction of the mine commenced in 2009, with official production beginning in 2021. The mine encompasses a total area of 122.51 km², containing 870 million tons of coal resources, of which 780 million tons are classified as recoverable reserves. As of the current stage, approximately 753 million tons remain recoverable, corresponding to an estimated remaining service life of 76 years. The mine has an approved and designed production capacity of 11 Mt/a. Mining operations target a single extractable seam the No. 10 coal seam which exhibits an average thickness of 12.22 m. The coal is primarily lignite, characterized by a low calorific value ranging from 15.48 to 24.39 MJ/kg (dry basis), with an average value of 18.70 MJ/kg (equivalent to 4488 cal/g). The principal natural hazards affecting the mine include landslides and hydrogeological hazards related to groundwater inflow. In November 2018, the Shenyang Design and Research Institute of China Coal Technology & Engineering Group completed the Preliminary Design (Revised Edition) for the Chaoyang Open-pit Coal Mine¹⁵. According to this design, the mine is divided into five mining districts arranged from west to east. The first mining district, situated centrally within the open-pit boundary, exposes the southern section of the No. 10 coal seam through trench excavation, adopting a longwall mining method with advancement oriented from south to west. A schematic representation of the mining district divisions is shown in Fig. 1.

Fig. 1

Division of initial mining area.

As of March 2025, the Baoqing Chaoyang Open-pit Coal Mine has achieved a designed annual production capacity of 7 Mt, and is presently at a critical transition stage toward the target capacity of 11 Mt/a. Currently, the primary mining district continues to operate under a longitudinal (vertical) mining configuration. However, the working line length remains insufficient, resulting in excessive annual advancing distances, which significantly constrain production scheduling efficiency, equipment deployment, and cost management. Simultaneously, the internal dumping sites within the mine have nearly reached their capacity limits, accompanied by an increasing risk of slope instability and localized landslides, which further restricts the potential for continued expansion. To address these challenges and meet the production capacity targets, it is imperative to enhance both the operational efficiency and safety performance of the mining area. This requires the implementation of a comprehensive optimization strategy, including rational adjustment of working line length, modification of mining orientation, and reconfiguration of mining district layouts to achieve coordinated production and waste disposal¹⁶. The current layout of the mining site is illustrated in Fig. 2.

Fig. 2

Current status of stopes in March 2025.

Optimization of coal mining working line length

Technically feasible coal mining working line length

The length of the coal mining working line is a fundamental geometric parameter in open-pit mining, directly determining the mining intensity and overall operational scale of the mine. It exerts a profound influence on key technical and economic indicators throughout the mine’s life cycle, including total production costs, final slope coal reserves, and the regulation of the stripping ratio¹⁷. The technically feasible working line length is typically determined by parameters such as annual production capacity, annual advance rate, coal seam thickness, in-situ coal density, and raw coal recovery rate.

$${L_{jk}}=\frac{{{A_p}}}{{{v_t} \cdot {h_m} \cdot \gamma \cdot \eta }}$$

(1)

In the equation: Ap represents the annual coal production of the open-pit mine (t/a); L_jk is the technically feasible length of the raw coal working line (m); v_t denotes the annual advancement rate (m/a); hm is the average thickness of the mineable coal seam (m); γ is the bulk density of raw coal (t/m³); η is the raw coal recovery rate.

When the main coal seam in an open-pit mine is nearly horizontal that is when the dip angle of the seam is less than 5°, the geometric model of the mining advance process can be simplified, as illustrated in the figure be-low:

Fig. 3

Simplified model of near-horizontal open-pit coal mining.

As shown in Fig. 3 above, the relationship between the annual stripping volume V_bl and the associated parameters can be qualitatively expressed by the following equation:

$${V_{bl}}=H \cdot \left[ {\left( {H+2{h_m}} \right)\cot \beta +{L_{jk}}} \right] \cdot {v_t}$$

(2)

In the equation, Vbl denotes the total annual stripping volume of the open-pit coal mine (m³/a); H is the thickness of the overburden layer (m); and β represents the slope angle of the excavation face (°).

The production stripping ratio ns can then be expressed as follows:

$${n_s}=\frac{{{V_{bl}}}}{{{A_p}}}=\frac{H}{{{h_m} \cdot \gamma \cdot \eta }}+\frac{{H\left( {H+2{h_m}} \right)\cot \beta }}{{{h_m} \cdot \gamma \cdot \eta }} \cdot \frac{1}{{{L_{jk}}}}$$

(3)

From the above equation, it can be observed that the production stripping ratio consists of two components. The first is the drilling column stripping ratio, defined as the ratio of the total thickness of the overlying rock and interbedded gangue layers to the thickness of the mineable coal seam. This represents the local stripping ratio, a static geological parameter independent of the working line length. The second component reflects the incremental stripping ratio resulting from variations in the end-wall stripping volume, which is inversely proportional to the coal mining line length. Therefore, the local stripping ratio represents the minimum value of the production stripping ratio. Due to changes in the end-wall slope angle, the actual production stripping ratio is always greater than the drilling column ratio. However, the ratio can be effectively reduced by extending the coal mining working line.

The relationship between the working line length and the internal waste haulage distance Y_j can be described by a linear function:

$${Y_j}=a{L_{jk}}+b$$

(4)

In the formula, a represents the coefficient of the disposal route. When the dumping operation is conducted within a double-ring configuration, a = 0.5; when arranged in a single-loop configuration, a = 1. b denotes the im-pact distance of disposal, which primarily depends on the end-face transportation distance.

Based on Fig. 1, the calculation formula for the disposal impact distance b can be expressed as follows:

$$b=\frac{{H+{h_m}}}{2}\left( {\cot \alpha +2\cot \beta +\cot \theta } \right)+m$$

(5)

In the formula, α and θ represent the slope angles (in degrees) of the working slope of the mining site and the internal waste dump, respectively; m denotes the tracking distance between the mining site and the internal waste dump, measured in meters.

Economically optimal length of the coal mining working line

The total cost of stripping in open-pit mines is generally composed of two components: blasting and exca-vation, and transportation. Assuming that the blasting and excavation cost per cubic meter of soil and rock is C1 yuan, the annual blasting and excavation cost E1 can be calculated using the following formula:

$${E_1}={n_s} \cdot {A_p} \cdot {C_1}=\frac{H}{{{h_m} \cdot \gamma \cdot \eta }}{A_p}{C_1}+\frac{{H\left( {H+2{h_m}} \right)\cot \beta }}{{{h_m} \cdot \gamma \cdot \eta }} \cdot \frac{{{A_p}{C_1}}}{{{L_{jk}}}}$$

(6)

If the transportation cost per cubic meter of stripped soil and rock is C2 yuan, the corresponding annual transportation cost E2 can be calculated as follows:

$${E_2}={n_s} \cdot {A_p} \cdot {C_2}\left( {a{L_{jk}}+b} \right)=\left[ {\frac{H}{{{h_m} \cdot \gamma \cdot \eta }}{A_p}{C_{\text{2}}}+\frac{{H\left( {H+2{h_m}} \right)\cot \beta }}{{{h_m} \cdot \gamma \cdot \eta }} \cdot \frac{{{A_p}{C_{\text{2}}}}}{{{L_{jk}}}}} \right]\left( {a{L_{jk}}+b} \right)$$

(7)

The total annual stripping transportation cost E is obtained by summing the two components described above:

$$\begin{gathered} E=\frac{{{A_p} \cdot H \cdot {C_{\text{2}}} \cdot a}}{{{h_m} \cdot \gamma \cdot \eta }}{L_{jk}}+\frac{{{A_p} \cdot H\left( {H+2{h_m}} \right)\cot \beta \left( {{C_1}{\text{+}}{C_{\text{2}}}b} \right)}}{{{h_m} \cdot \gamma \cdot \eta }}\frac{{\text{1}}}{{{L_{jk}}}} \hfill \\ {\text{ +}}\frac{{{A_p} \cdot H\left[ {\left( {H+2{h_m}} \right)\cot \beta \cdot {C_{\text{2}}} \cdot a{\text{+}}\left( {{C_1}{\text{+}}{C_{\text{2}}}b} \right)} \right]}}{{{h_m} \cdot \gamma \cdot \eta }} \hfill \\ \end{gathered}$$

(8)

According to the derivation, the annual total cost function is a convex function defined in the first quadrant, and its general expression can be written as:

$$f\left( x \right)=dx+\frac{p}{x}+c$$

(9)

The economically reasonable length of the coal mining working line corresponds to the technically feasible working line length at which the annual total stripping cost function f(x) reaches its minimum value, that is:

$$\left\{ \begin{gathered} f\left( x \right)=2\sqrt {dp} +c \hfill \\ {\text{ }}{L_{jh}}=\sqrt {\frac{p}{d}} \hfill \\ \end{gathered} \right.$$

(10)

Optimization of coal mining working line length

Based on the 2025 production plan of the Chaoyang Open-pit Coal Mine, and with reference to the current status map as of March, combined with actual production data from the mine, the variable parameters in the above equations were determined as shown in Table 1.

Table 1 Variables of Baoqing open-pit coal mine.

The geotechnical parameters adopted in this study were obtained from the existing geotechnical investigation and design documentation of the Baoqing Chaoyang open-pit mine. In particular, the slope angles β, α and θ correspond to the design values of the overall pit slope, the internal dump slope and the bench slope, respectively. These values were determined by the mine design institute based on borehole and laboratory test data (shear strength parameters c and φ, unit weight, etc.), combined with relevant open-pit slope design codes and stability analyses under representative adverse conditions. The same slope configurations have been applied in the current operation without large-scale instability, which supports the reasonableness of using these parameters in the optimization model.

At present, the mine has developed a five-year production plan, targeting an increase in production capacity from 7.0 Mt/a in 2025 to 11.0 Mt/a by 2028. According to Eq. (1) above, the functional relationship be-tween the technically feasible coal mining working line length and the annual advance rate was plotted, as shown in Fig. 4 below:

Fig. 4

Annual progress curve of coal mining work line.

According to the Code for Design of Open-Pit Coal Mines in the Coal Industry (GB 50197–2015)¹⁸, for open-pit coal mines adopting the shovel–truck discontinuous operation system, the maximum annual advance rate of large-scale open-pit mines with an annual production capacity between 4.0 Mt and 20.0 Mt shall not ex-ceed 400 m. Based on this regulation, the Chaoyang Open-pit Coal Mine should have a working line length greater than 1305 m when producing 7.0 Mt/a before reaching full capacity, and greater than 2050 m when reaching the designed capacity of 11.0 Mt/a in 2028. Under the current working line length of 950 m, the maxi-mum achievable annual advance rate reaches 550 m at a capacity of 7.0 Mt/a, and up to 863 m at 11.0 Mt/a, which significantly exceeds the regulatory limit. Therefore, it is urgent to re-divide the mining districts and re-optimize the working line length to ensure compliance with design standards and production safety requirements.

Fig. 5

Total annual cost of stripping transportation - length change curve of coal mining line.

Analysis of the Fig. 5 indicates that the total annual stripping and transportation cost exhibits a checkmark-shaped variation with changes in the coal mining working line length. The minimum total annual cost occurs when the working line length is 1350 m, corresponding to an annual advance rate of 497 m. When the working line length is less than 500 m, the total annual cost increases sharply. Conversely, as the working line length exceeds 1500 m, the total annual cost increases gradually.

In summary, the technically feasible and economically reasonable range of working line lengths for the Baoqing Chaoyang Open-pit Coal Mine is determined to be 1350 ~ 2050 m, with the corresponding annual ad-vance rate ranging from 400 to 497 m. In subsequent mining district planning, it is essential to maintain a balance among working line length, advance rate, and total annual cost to achieve optimal technical and economic out-comes^19,20.

Mining area division scheme

The Baoqing Chaoyang Open-pit Coal Mine is characterized by high groundwater abundance, soft rock strata, and low permeability. Currently, all mining zones adopt a longitudinal mining configuration. However, this mining approach has resulted in several operational challenges. Specifically, the average slope angle of the inter-nal dump is only 10°, which is significantly smaller than the designed stable slope angle of 16°, leading to limited internal dumping space. Meanwhile, external dumping is constrained due to land acquisition difficulties. The average coal mining working line length is approximately 950 m, and in some sections, the annual advance rate exceeds 400 m, reaching up to 500 m, even under the current production capacity of 7.0 Mt/a. With the mine’s production capacity scheduled to reach 11.0 Mt/a by 2028, failure to redesign the mining zone layout and extend the working line length will likely result in a series of production challenges, including dumping coordination difficulties, excessive annual advancement rates, large external dumping footprints, and heightened slope instability risks.

Principles of mining area division

The criteria for mining area zoning in open-pit coal mines involve a comprehensive evaluation of multidimensional factors that collectively determine the technical feasibility, economic efficiency, and environmental sustainability of the operation. These factors include coal seam occurrence, rock strength and structural characteristics, fault and joint distribution, hydrogeological conditions (such as aquifer distribution and permeability coefficients), topographic and geomorphic features (slope gradient, step configuration), stripping ratio and coal quality, layout and transportation paths of waste disposal sites, as well as ecological sensitivity and social impact. The design principles of mining area zoning are summarized as follows:

Safety first: ensuring controllable risks associated with slope stability and water hazards; Economic rationality༚achieving low stripping ratios, high recovery efficiency, and optimal resource utilization; Environmental friendliness༚minimizing land disturbance, ecological damage, and carbon footprint; Dynamic adjustment༚integrating real-time monitoring and production feedback for adaptive zoning optimization; Equipment-process matching༚ensuring coordination between zoning scale, process flow, and equipment capacity; The ultimate goal is to realize safe, efficient, and green open-pit mining through a scientifically grounded and dynamically optimized zoning approach²¹.

Based on the above objectives, all potential influencing indicators are preliminarily categorized into eight dimensions: geology, rock mass, hydrology, topography, engineering, environment, economy, and society²². The proposed indicator system is shown in Table 2.

Table 2 Influencing factors of mining area division.

Based on the principles of open-pit coal mining area division and the multidimensional impact indicators discussed above, and in conjunction with the functional relationships established in Chap. 2 between the coal mining working line length, annual advance rate, and annual total cost, a set of optimized zoning schemes for the Baoqing Chaoyang Open-pit Coal Mine are proposed.

Scheme I

Comprehensively considering geological and operational conditions, the area south of the F2 fault is desig-nated as the long-term stable production zone of the open-pit mine, comprising the first and second mining areas. The region north of the F2 fault, characterized by greater mining depth, together with the area above the elevation line of the Team 1 of Farm 852, Division 5, is planned as the third mining area. The eastern boundary is defined by the deep boundary on the eastern side of the current first mining area. The northern boundary extends from the southern deep limit northward, ensuring a minimum width of 1,800 m for the first mining area, which serves as the new northern deep boundary. The southern boundary corresponds to the southern deep boundary of the open-pit mine, while the western boundary is delineated by extending the northern and southern limits of the first mining area to meet the western deep boundary of the pit. The first mining area has an overall burial depth of 45–90 m, and its entire mining range is unaffected by farmland. A fan-shaped transition pattern is adopted for the zone’s expansion, with a stable working-line length of approximately 1,800 m. The coal seam thickness gradually decreases from east to west, ranging from 18 m to 5 m, with an average thickness of 12.5 m. The partitioning scheme Ⅰ is shown in the Fig. 6 below.

Fig. 6

Schematic diagram of zoning mining in Scheme 1.

Advantages

• 1. To mitigate the risk of landslides on the west side of the mining site, the fan-shaped turning method uses the west end of the mining area as the rotation axis. This allows rapid continuation of production while ensuring slope stability. From the current coal mining position to the initial westward shift, stable production for approximately 4 years can be maintained.

• 2. As the west side of the first mining area advances, the ground production system is brought closer to the industrial site, improving operational efficiency.

• 3. Transitioning from vertical to horizontal mining, by exploiting the inclination characteristics of coal seams, helps stabilize the foot pressure of internal drainage slopes and underground drainage systems.

• 4. Minimizes the impact of surface farm structures, reducing the need for additional land acquisition in the first mining area.

Limitations

• 1. The natural outcrop of the coal seam on the west side of the new first mining area is extensive, with a weathered coal thickness of approximately 0.6–2.2 m. Careful analysis of weathered coal is required during mining.

• 2. The fan-shaped working line at the tail of the first mining area undergoes significant changes, with a maximum working line length of 3,500 m, necessitating the extension of belt conveyors or adoption of outsourced mining.

• 3. The west side of the first mining area is currently occupied by basic farmland, posing challenges for land acquisition and production continuity.

Scheme Ⅱ

First Mining Area: The northern boundary of the original mining area has been optimized to align with the planned northern boundary of the third land acquisition.

Second Mining Area: The western portion of the first mining area is designated as the second mining area, with a deep boundary width of 3.0 ~ 3.8 km. This area is suitable for arranging large-scale continuous or semi-continuous coal mining operations.

Third Mining Area: The overall coal seam occurrence depth in the northern deep boundary of the mining area does not exceed − 60 m. Considering the shape and production continuity of the mining area, the north-eastern portion of the open-pit mine is included in the third mining area, with a deep boundary width of 2.0 ~ 3.1 km.

Fourth Mining Area: The deepest coal seam occurrence zone, which has a relatively high stripping ratio, is planned as the fourth mining area. The partitioning scheme Ⅱ is shown in the Fig. 7 below.

Fig. 7

Schematic diagram of zoning mining in Scheme 2.

Advantages

• 1. Temporarily addresses the challenges associated with the long and difficult westward shift of the mining area, ensuring that the open-pit mine can continue northward operations for at least five years based on the original first mining area layout.

• 2. The second mining area is subdivided into large-scale mining zones, minimizing secondary stripping between areas, enhancing production safety, and improving resource recovery rates. The long working line in this mining area is particularly suitable for future large-scale continuous or semi-continuous coal mining operations.

• 3. Transitioning from the first to the second mining area changes the mining method from vertical mining with internal discharge to horizontal mining with internal discharge, improving the stability of internal dump slopes.

Limitations

• 1. The coal mining working line in the second mining area spans 3.0–4.3 km, necessitating prior research on continuous coal mining transportation technologies to reduce reliance on truck haulage within the coal system.

• 2. The stripping and transportation distances within the second mining area are large, averaging around 3.0 km, which could affect operational efficiency and logistics planning.

Scheme Ⅲ

Based on Plan 2, Plan 3 subdivides the original second mining area of Plan 2 into two distinct mining areas. The first mining area transitions to the second mining area via a gentle slope, while the second mining area transitions to the third mining area through a fan-shaped steering method. This plan reduces transportation challenges caused by the excessively wide mining area and alleviates the time and complexity associated with land acquisition and farm relocation within the third mining area. The partitioning scheme Ⅲ is shown in the Fig. 8 below.

Fig. 8

Schematic diagram of zoning mining in Scheme 3.

Scheme Ⅳ

Scheme IV proposes a new optimization plan for mining zone division based on the current operational challenges of the open-pit mine. The influence of land acquisition has not yet been considered. Considering overall geological and operational factors, the area south of the F2 fault is designated as the long-term stable production zone of the open-pit mine, including the First, Second, and Third Mining Zones. The deeper area north of the F2 fault is designated as the Fourth Mining Zone. The First Mining Zone has a working line length of 960 ~ 1850 m, thinning from west to east with a variation range of 18 ~ 8 m, and an average coal seam thickness of 15 m. The Second Mining Zone starts at a greater mining depth of − 60 to − 90 m and gradually shallows as it advances westward, maintaining a stable working line length of 1800 m. The Third Mining Zone also maintains a stable working line length of 1800 m. From the perspectives of mining process adaptability, zone connectivity, dump development planning, and rational layout of the continuous surface production system, the overall zoning design in Scheme IV is superior to the current mining plan. More importantly, Scheme IV changes the existing longitudinal mining mode to a transverse one, which significantly improves the stability of internal dump slopes. The partitioning scheme Ⅳ is shown in the Fig. 9 below.

Fig. 9

Schematic diagram of zoning mining in Scheme 4.

Advantages

• 1. The initial mining zone is located in an area where the coal seam is relatively thick (average 15 m), which reduces the stripping and mining advancement rate, and minimizes the influence of weathering in the southern outcrop zone.

• 2. The transition from longitudinal to transverse mining utilizes the coal seam inclination, enhancing the stability of internal dump toe slopes and improving pit drainage.

• 3. There are no agricultural facilities within the First Mining Zone, minimizing the impact of farmland structures on land acquisition.

Limitations

• 1. Due to the spatial constraints of the external dump area, the initial turning line of the First Mining Zone is located approximately 1.0 km from the current working line. Assuming four production stages (5.0, 7.0, 9.0, and 11.0 Mt/a), approximately 3.5 years of continued longitudinal mining are required to reach the turning condition, which is unfavorable for internal dump slope stability.

• 2. When the First Mining Zone turns northward along the coal seam dip, it becomes secondarily affected by potential landslides from the external dump, which may trigger slope failures at both the external dump and the pit edge.

• 3. The transition from the First to the Second Mining Zone adopts a semiretained trench internal dumping method, which limits dumping space and increases slope stability risks.

• 4. During westward turning operations of the First Mining Zone, the surface production system is relatively far from the industrial site.

• 5. The influence of land acquisition conditions has not yet been considered.

Geomechanical rationale of the transverse-mining strategy

For an internal dump slope composed mainly of loose waste rock, the potential sliding direction tends to align with the direction of the maximum overall gradient of the dump-foundation system. In a simplified infinite-slope framework, the driving and resisting components can be expressed as:

$$T=\gamma H\sin {i_{eff}},{\text{ }}N=\gamma H\cos {i_{eff}}$$

(11)

and the factor of safety may be approximated as:

$$FS \approx \frac{{c'}}{{\gamma H\sin {i_{eff}}}}+\frac{{\tan \varphi '}}{{\tan {i_{eff}}}}$$

(12)

where γ is the unit weight, H is the slope height, i_eff is the effective inclination of the dump–foundation system, and c′, φ′ are the effective shear-strength parameters. Under longitudinal mining, the internal dump is arranged approximately across the coal seam dip, so the dump face and the inclined foundation form a cross-dip configuration. This increases the effective inclination i_eff in the cross-dip direction and shortens the potential sliding path, which increases the driving shear component T and reduces the factor of safety.

Fig. 10

Geomechanical Comparison of Internal Dump Stability under Longitudinal and Transverse Mining Conditions.

In contrast, after the transition to transverse (along-dip) mining, the internal dump benches can be laid out approximately parallel to the coal seam dip and to the principal direction of topographic descent. This layout reduces the cross-dip component of the overall gradient, effectively lowering i_eff for the critical sliding direction. As indicated by the above expressions, a smaller i_eff leads to a lower driving shear and a higher factor of safety for the same material parameters and slope height. The theoretical analysis diagram is shown in Fig. 10 above:

In the longitudinal-mining layout, part of the internal dump is placed on a foundation whose structural planes (coal seam and weak interlayers) intersect the dump face unfavourably, producing adverse day-lighting conditions and higher risk of foundation sliding. After reconfiguring the mining sequence into transverse zones aligned with the near-horizontal coal seam dip, the dump foundation is more closely matched with the dip direction of the stratified rock mass. This alignment increases the normal stress acting on potential bedding-controlled slip surfaces, improves the overall integrity of the dump-foundation system, and reduces the likelihood of deep-seated sliding along weak layers.

The transverse-mining configuration also allows the internal dump to be formed more gradually along the dip direction, with a more regular bench geometry and fewer abrupt changes in slope height, which is favourable for maintaining drainage conditions and preventing local instability. This is consistent with practical slope-design recommendations for large open-pit mines under gently dipping stratified conditions.

Scheme selection based on unascertained measure theory

Unascertained measure theory

Uncertainty measure theory( UMT ) is a theoretical framework that focuses on handling uncertain information. It effectively addresses the indeterminacy of various evaluation factors in open-pit mine zoning studies, enabling their quantitative analysis.

The set of rating samples X_i(i = 1,2,3…,s) denoted as X={X₁,X₂,X₃,…,X_s}, constitutes the sample space. Each evaluation sample X_i contains m evaluation indicators U_j(j = 1,2,3…,m), forming the evaluation indicator space U={U₁,U₂,U₃,…,U_m}. x_ij represents the measured value of evaluation sample X_i with respect to indicator U_j. For each component x_ij of X_i={x_i1, x_i2, x_i3,…, x_im}, assume it has n evaluation levels, and denote the corresponding evaluation domain as V. Thus, V={V₁,V₂,V₃,…,V_n} is defined as the evaluation set. Let V_k denote the k-th evaluation level of the indicator, where the k-th level represents better regional stability than the (k + 1)-th level, expressed as V_k>V_k+1. Therefore, the set {V₁,V₂,V₃,…,V_n} constitutes an ordered partition of the evaluation space.

If z_ijk=z(x_ij∈V_k) represents the membership degree of the measured value x_ij belonging to the k-th evaluation level V_k, it satisfies the following three conditions:

Non-negative boundedness:

$${\text{0}} \leqslant z\left( {{x_{ij}} \in {V_K}} \right) \leqslant 1$$

(13)

Normalization:

$$z\left( {{x_{ij}} \in V} \right)=1$$

(14)

Additivity:

$$z\left( {{x_{ij}} \in \bigcup\limits_{{l=1}}^{k} {{V_l}} } \right)=\sum\limits_{{l=1}}^{k} {z\left( {{x_{ij}} \in {V_l}} \right)}$$

(15)

(i = 1,2,3…,s;j = 1,2,3…,m༛l = 1,2,3…,k༛k = 1,2,3…,n༛)

Then, z_ijk=z(x_ij∈V_k) is defined as the Uncertainty Measure (UM), referred to simply as the measure.

Based on the definition of the UM, to determine the measure value z_ijk of each evaluation indicator for the evaluation object R_i, it is necessary to construct the measure function z_ijk=z(x_ij∈C_k) for each indicator. The matrix composed of all indicator measure values (z_ijk)m×n is called the single-indicator UM evaluation matrix. The measure matrix is expressed as follows:

$$z={\left( {{z_{ijk}}} \right)_{m \times n}}=\left[ {\begin{array}{*{20}{c}} {{z_{i11}}}&{{z_{i12}}}& \ldots &{{z_{i1n}}} \\ {{z_{i21}}}&{{z_{i22}}}& \ldots &{{z_{i2n}}} \\ \vdots & \vdots & \ddots & \vdots \\ {{z_{im1}}}&{{z_{im2}}}& \ldots &{{z_{imn}}} \end{array}} \right]$$

(16)

Comprehensive evaluation method

The multi-index comprehensive evaluation method establishes a standardized framework that integrates qualitative and quantitative analyses, effectively addressing the challenges of integrating multi-source heterogeneous indicators such as weight assignment, redundancy elimination, and information fusion. In the study of open-pit mine closure disasters, this method systematically integrates multidimensional information including geological, hydrological, and environmental data to construct a scientific evaluation model that quantifies closure risk levels and disaster impacts, providing essential support for management decisions. Common and effective approaches include the entropy weight method (based on information variability), the Analytic Hierarchy Process (AHP, relying on expert judgment), Principal Component Analysis (PCA, for dimensionality reduction), the CRITIC method (considering inter-criteria conflict), and the fuzzy comprehensive evaluation method (for handling uncertainty). These methods are highly adaptable and scientifically reliable, forming a solid methodological foundation for systematic research on mine closure disasters²³. The advantages, disadvantages, and applicability of the five evaluation methods are summarized in Table 3.

Table 3 Comparative analysis table of evaluation methods.

In summary, each evaluation method exhibits distinct advantages and limitations across different dimensions. Specifically, the entropy weight method determines weights objectively based on indicator data, making it suitable for scenarios with sufficient and well-structured datasets. However, it is highly sensitive to data quality and less effective in handling indicators with strong subjectivity or uncertainty. The Analytic Hierarchy Process (AHP) features a clear structure that facilitates the incorporation of expert knowledge, making it suitable for modeling complex systems. Nevertheless, it relies heavily on subjective judgment and becomes computationally intensive when the hierarchy is large. Principal Component Analysis (PCA) reduces dimensionality to eliminate redundancy and improve computational efficiency, making it suitable for cases with high inter-variable correlations. However, the interpretability of principal components is often poor, and the method may overlook the practical significance of original indicators. The CRITIC method integrates indicator variability and inter-criteria conflict, yielding a more objective weighting scheme suitable for multi-criteria comprehensive evaluation. However, its ability to handle nonlinear or fuzzy information remains limited. The Fuzzy comprehensive evaluation method effectively addresses problems with high fuzziness and uncertainty, making it suitable for evaluations that are difficult to quantify precisely. Nevertheless, its performance depends on the design of membership functions and expert experience, introducing a degree of subjectivity. Therefore, in practical applications, appropriate methods or combinations thereof should be selected based on problem characteristics and data properties to enhance the scientific validity and adaptability of the evaluation results.

Combined AHP–Entropy weighting method

Traditional UM models typically determine indicator weights using the EW or simple correlation functions, followed by grade evaluation based on the maximum membership principle. However, using the EW to determine indicator weights makes them highly sensitive to variations in sample data. Moreover, in the absence of experiential guidance, the results often fail to reflect the subjective preferences of decision-makers. Given the inherent ordering in the evaluation system, the maximum membership principle may fail to capture the fuzziness of an evaluation object’s boundaries and can lead to loss of evaluation information in certain cases. To address these limitations, this study integrates the EW with the AHP to improve the traditional weighting approach²⁴.

The hierarchical model employs the Santy nine-point scale to construct the judgment matrix S at each level relative to the preceding level. The square root method is used to calculate the weights of each influencing factor. Finally, a consistency check is performed: if the consistency ratio CR is less than 0.1, the AHP ranking is considered satisfactory, indicating a reasonable allocation of weight coefficients. Otherwise, the values in the judgment matrix must be adjusted, and the weights recalculated.

$$\overline {{{w_i}}} =\sqrt[m]{{\prod\limits_{{j=1}}^{m} {{\alpha _{ij}}} }}\left( {i=1,2, \cdots n} \right)$$

(17)

The entropy weight method determines indicator weights based on data, thereby reducing the influence of subjective factors and ensuring that the assigned weights more accurately reflect the contribution of each indicator to the overall UM. Suppose the relative importance of evaluation indicator q_j compared with other indicators is denoted as θ_ij (0 ≤ θ_ij ≤ 1, Σθ_ij = 1). Then, the weight of indicator q_j is given by θ_ij, and the weights of other indicators can be derived similarly. The entropy value is calculated as follows:

$$\begin{gathered} {v_{ij}}=1+\frac{1}{{\ln s}}\sum\limits_{{i=1}}^{s} {\left( {{\beta _{ije}} \cdot \ln {\beta _{ije}}} \right)} \hfill \\ {\text{ }}{\theta _{ij}}={v_{ij}}/\sum\limits_{{i=1}}^{n} {{v_{ij}}} \hfill \\ \end{gathered}$$

(18)

The subjective weights derived from the AHP and the objective weights determined by the EW are integrated using a multiplicative normalization approach to obtain the comprehensive weights. This coupling process compensates for data bias and excessive subjectivity, and can be expressed as follows:

$${w_{ij}}=\left( {{\alpha _j} \cdot {\beta _j}} \right)/\sum\limits_{{j=1}}^{m} {\left( {{\alpha _j} \cdot {\beta _j}} \right)}$$

(19)

In the equation, w_j represents the comprehensive weight of the j^th evaluation indicator; 0 ≤ θ_ij ≤ 1 and Σθ_ij = 1; α_j and β_j denote the subjective and objective weights of the j^th indicator, respectively; and $\:n$is the total number of evaluation indicators.

Zoning optimization of the Baoqing Chaoyang Open-Pit coal mine

Based on the classification standards proposed by domestic and international scholars and the intrinsic characteristics of the influencing factors, each indicator was categorized into hierarchical levels. The priority evaluation set for open-pit mine zoning schemes was defined as V={V₁,V₂,V₃, V₄}, where V₁ > V₂ > V₃> V₄, representing a gradual decline in the priority level of the schemes.

According to the influencing factors of mining area division listed in Table 2, the four proposed zoning schemes were evaluated by grading eight key influencing factors, and the results are shown in the following Tables 4 and 5:

Table 4 Classification criteria for V₁-V₄ grades of influencing factors in mining area division.

Table 5 Four-level evaluation set of unknown measures.

According to the theory of UM, determining the single-indicator measure evaluation matrix primarily depends on constructing an appropriate UM function. Common construction methods include the linear function, exponential curve, sine curve, and quadratic curve approaches. In practical applications, decision-makers should select the most suitable UM function based on the variation characteristics of specific evaluation indicators. However, regardless of the function type adopted, it must satisfy the principles of non-negativity, additivity, and normalization. Since the triangular linear membership function achieves a balance among simplicity, fuzzy representation, and operational feasibility, it was selected for the function calculation in this study.

The single-indicator UM function for the evaluation indices of mining area zoning priority is shown in the following Fig. 11:

Fig. 11

Schematic diagram of triangular linear UM function.

Example of a single-index measurement matrix is as follows:

$$z={\left( {{z_{1jk}}} \right)_{8 \times 4}}=\left[ {\begin{array}{*{20}{c}} 1&0&0&0 \\ 0&1&0&0 \\ 0&0&1&0 \\ \begin{gathered} 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ 1 \hfill \\ \end{gathered} &\begin{gathered} 0 \hfill \\ 1 \hfill \\ 0 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} &\begin{gathered} 0 \hfill \\ 0 \hfill \\ 1 \hfill \\ 0 \hfill \\ 0 \hfill \\ \end{gathered} &\begin{gathered} 1 \hfill \\ 0 \hfill \\ 0 \hfill \\ 1 \hfill \\ 0 \hfill \\ \end{gathered} \end{array}} \right]$$

(20)

For each indicator, In Table 4, four evaluation grades V₁ ~ V₄ (Excellent, Good, Fair, Poor) are defined. Based on design codes, mine technical specifications, and expert knowledge, quantitative threshold ranges for the four grades are specified.

In Table 5. For benefit-type indicators (larger is better), the original value y_ij (indicator j, scheme i) is linearly mapped to a standardized score x_ij∈[1,4], where 1 corresponds to the lower bound of grade V₁ and 4 to the upper bound of grade V₄.

For cost-type indicators (smaller is better), the mapping is inverted so that a smaller physical value corresponds to a better (smaller) score.

For qualitative indicators (e.g., social impact, land-acquisition difficulty), a panel of experts assigns discrete scores 1–4 according to explicit linguistic descriptions of the four grades.

The standardized score x_ij is then transformed into the single-index unascertained measure vector

$${\mu _{ij}}=\left( {\mu _{{ij}}^{{\left( 1 \right)}},\mu _{{ij}}^{{\left( 2 \right)}},\mu _{{ij}}^{{\left( 3 \right)}},\mu _{{ij}}^{{\left( 4 \right)}}} \right)$$

(21)

by triangular linear membership functions, as illustrated in Fig. 11.

Taking 1, 2, 3, and 4 as the centers of the four grades on the score axis, the unascertained measure functions are defined piecewise as:

• For 1 ≤ x ≤ 2:

$${\mu ^{\left( 1 \right)}}\left( x \right)=2 - x,{\mu ^{\left( 2 \right)}}\left( x \right)=x - 1,{\mu ^{\left( 3 \right)}}\left( x \right)=0,{\mu ^{\left( 4 \right)}}\left( x \right)=0;$$

(22)

• For 2 ≤ x ≤ 3:

$${\mu ^{\left( 1 \right)}}\left( x \right)=0,{\mu ^{\left( 2 \right)}}\left( x \right)=3 - x,{\mu ^{\left( 3 \right)}}\left( x \right)=x - 2,{\mu ^{\left( 4 \right)}}\left( x \right)=0;$$

(23)

• For 3 ≤ x ≤ 4:

$${\mu ^{\left( 1 \right)}}\left( x \right)=0,{\mu ^{\left( 2 \right)}}\left( x \right)=0,{\mu ^{\left( 3 \right)}}\left( x \right)=4 - x,{\mu ^{\left( 4 \right)}}\left( x \right)=x - 3.$$

(24)

Thus, for any indicator and scheme,

$$\sum\limits_{{k=1}}^{4} {{\mu ^{\left( k \right)}}\left( {{x_{ij}}} \right)} =1,{\text{ }}0 \leqslant {\mu ^{\left( k \right)}}\left( {{x_{ij}}} \right) \leqslant 1$$

(25)

and the single-index unascertained measure matrix for Scheme $\:i$ is

$${R^{\left( i \right)}}=\left[ {\begin{array}{*{20}{c}} {\mu _{{i1}}^{{\left( 1 \right)}}}&{\mu _{{i1}}^{{\left( 2 \right)}}}& \ldots &{\mu _{{i1}}^{{\left( 4 \right)}}} \\ {\mu _{{i2}}^{{\left( 1 \right)}}}&{\mu _{{i2}}^{{\left( 2 \right)}}}& \ldots &{\mu _{{i2}}^{{\left( 4 \right)}}} \\ \vdots & \vdots & \ddots & \vdots \\ {\mu _{{i8}}^{{\left( 1 \right)}}}&{\mu _{{i8}}^{{\left( 2 \right)}}}& \ldots &{\mu _{{i8}}^{{\left( 4 \right)}}} \end{array}} \right]$$

(26)

As an example, consider the stripping ratio (a cost-type indicator). Suppose its grade thresholds (m³/t) are:

• V₁ (Excellent): SR ≤ 5.

• V₂ (Good): 5 < SR ≤ 7.

• V₃ (Fair): 7 < SR ≤ 9.

• V₄ (Poor): SR > 9.

For Scheme I, assume SR = 6.2 m³/t. This lies in the V₂ interval. The standardized score is obtained by linear interpolation in [1, 2]:

$$x=1+\frac{{SR - 5}}{{7 - 5}}=1+\frac{{6.2 - 5}}{2}=1.6$$

(27)

For 1 ≤ x ≤ 2,

$${\mu ^{\left( 1 \right)}}\left( x \right)=2 - x=0.4,{\mu ^{\left( 2 \right)}}\left( x \right)=x - 1=0.6,{\mu ^{\left( 3 \right)}}\left( x \right)=0,{\mu ^{\left( 4 \right)}}\left( x \right)=0;$$

(28)

So, the unascertained measure vector of “stripping ratio” for Scheme I is

$$\mu _{{SR}}^{{\left( I \right)}}=\left( {0.4,0.6,0,0} \right)$$

(29)

which is one row of the matrix R^(I). Similar calculations are carried out for all indicators and all schemes.

Based on the analyses presented in Sect. Comprehensive evaluation method and Combined AHP–entropy weighting method, judgment matrices were constructed for the eight evaluation indicators(A₁ ~ A₈) using the AHP, EW, and AHP-EW combined weighting methods. The weight vectors were derived using MATLAB, followed by a consistency test. The judgment matrix and consistency ratio are calculated as shown in the Table 6. The final weight values of the evaluation indicators are summarized in the following Table 6 :

Table 6 Judgment matrix of eight criteria.

The matrix was constructed such that each element

$${a_{ij}}=\frac{{{w_i}}}{{{w_j}}}$$

(30)

Where w_i is the normalized weight of criterion i. This construction yields a reciprocal and fully consistent AHP matrix whose principal eigenvector is exactly

$$w=\left( {0.150,0.100,0.100,0.100,0.150,0.200,0.100,0.100} \right)$$

(31)

i.e., identical to the weights reported in Table 6.

For this 8 × 8 judgment matrix, the maximum eigenvalue is

$${\lambda _{\hbox{max} }}=8.001$$

(32)

and the consistency index (CI) and consistency ratio (CR) are:

$$\begin{gathered} CI=\frac{{{\lambda _{\hbox{max} }} - n}}{{n - 1}}=\frac{{8.001 - 8}}{7} \approx 0.00014 \hfill \\ CR=\frac{{CI}}{{CR}}=\frac{{0.00014}}{{1.41}} \approx 0.0001 \ll 0.1 \hfill \\ \end{gathered}$$

(33)

Where n = 8 and RI = 1.41 is the random index for an 8 × 8 matrix in Saaty’s AHP. This indicates that the judgment matrix satisfies the AHP consistency requirement and that the subjective weighting process is logically coherent. The three weight values ​​for each evaluation indicator are shown in Table 7 below.

Table 7 Comprehensive weights of evaluation indicators.

Table 8 Evaluation level of mining area division optimization scheme.

The analysis of evaluation results indicates that, regardless of whether subjective expert preferences or objective weights derived from data dispersion were used, Scheme 2 consistently achieved a confidence level above 0.65 in the V₁ (Excellent) category, reaching 0.7058 under the integrated weighting approach well above the threshold value of 0.6. This demonstrates that Scheme 2 performs exceptionally well across key evaluation indicators, exhibits strong sensitivity to different weighting methods, and possesses broad applicability and stability.

Scheme 1 achieved confidence levels ranging between V₂ and V₃ across all weighting approaches, with the highest confidence value of 0.3823 for V₃ under the comprehensive weighting. This suggests that Scheme 1 performs moderately well in secondary indicators such as hydrology and topography, but lacks competitiveness in high-priority indicators.

Scheme 3 was classified as V₂ and V₃ under the AHP and EW frameworks, respectively, and remained within V₃ under the integrated weighting, implying a balanced yet unremarkable performance under combined subjective and objective weighting considerations.

Scheme 4 consistently fell within the V₃ category under all three evaluation methods, reflecting relatively weak performance in high-weight indicators, particularly economic feasibility and geological stability.

Based on the above analysis, Scheme 2 demonstrates superior overall performance and is selected for full lifecycle simulation. The corresponding simulation data are presented in the following Table 9.

Table 9 Scheme 2 full life cycle deduction.

For Scheme II, the “average stripping ratio of 5.8 m³/t” reported in the conclusion is obtained as a reserve-weighted average over all four sub-zones within the designed ultimate pit boundary, whereas the values listed in Table 9 (4.19–11.10 m³/t) represent the zone-level reserve-weighted stripping ratios for each individual sub-zone.

The following is a Table 10 of key performance indicators for the four options.

Table 10 Quantitative comparison of key performance indicators for the four zoning schemes.

Scheme I has the shortest internal haul distance and relatively low annual stripping-transport cost, but its recoverable reserves and service life are significantly smaller. Under the 11 Mt/a capacity target, its life-of-mine production cannot support long-term stable operation, so the overall economic benefit is limited despite the shorter haulage distance.

Scheme IV exhibits moderate haul distances, but a large proportion of deep-seated high-stripping-ratio areas, leading to the highest stripping ratio and unit cost per tonne of coal among the four schemes. As a result, its total transportation cost is not competitive when evaluated on a per-ton basis, and its overall economic performance remains inferior.

Scheme III splits the large mining area of Scheme II into two sub-zones. This subdivision indeed reduces the maximum working-line length and slightly lowers the internal haul distance, thereby lowering the gross transport cost to some extent. However, it also introduces additional boundary stripping and scheduling complexity, which offsets part of the economic gain. The net outcome, as shown in Table 10, is that its unit cost and composite evaluation score remain lower than those of Scheme II.

Scheme II, while having the largest average internal haul distance and the highest annual stripping-transport cost in absolute terms, also provides the largest recoverable reserves (971.2 Mt), a moderate average stripping ratio (5.8 m³/t), and the longest service life (up to 34.3 years).

When the transport cost is converted to a unit cost per tonne of coal over the full life cycle, Scheme II still achieves the lowest comprehensive unit cost and the highest net economic benefit, due to the combination of lower stripping ratio and continuous large-scale production. This quantitative comparison confirms that the drawback of “longer internal haul distance” is more than compensated by the superior resource and economic indicators.

Discussion

This study, focusing on the multiple challenges faced by the Baoqing Chaoyang open-pit coal mine during its capacity expansion phase, established an integrated decision-making framework of “working-line optimization-mining-area redivision-multi-indicator comprehensive evaluation.” By applying UMT, the study quantitatively optimized various lateral-mining division schemes. This chapter discusses the methodological innovations, intrinsic characteristics of each scheme, the strengths and potential challenges of the optimal option, and the universality of the research outcomes.

Methodological innovation and scientific rigor

Traditional mining area division often relies on empirical judgment or a single economic indicator, which fails to capture the complex coupling among geological, engineering, environmental, and social factors. The UMT, combined with the AHP-EW hybrid weighting approach adopted in this research, effectively addresses the issues of uncertainty and fuzziness in open-pit mine planning.

Compared with conventional methods such as TOPSIS and fuzzy comprehensive evaluation, UMT constructs membership functions that finely describe the gradual transition of indicators from “excellent” to “poor,” thereby avoiding information loss caused by the “maximum membership” rule. The introduction of the confidence recognition criterion (λ = 0.6) enhances the robustness of the evaluation results and allows clear identification of the dominant scheme.

The AHP-EW fusion weighting strategy retains expert emphasis on critical indicators (e.g., economic and engineering factors), while the entropy method introduces objectivity that reduces subjective bias. The resulting weight distribution (e.g., economic index weight elevated to 0.2941) better reflects the actual needs of large-scale capacity expansion in open-pit mines, thereby improving the scientific credibility and reliability of the evaluation.

In-depth analysis of the mining division schemes

The four proposed lateral-mining division schemes represent different strategic orientations during the capacity expansion stage. Scheme 1 and Scheme 4 both aim to maintain short-term production continuity through fan-shaped or semi-reserved-trench transitions. However, neither scheme reached the “excellent” level (V₁), with both rated as V₃, indicating that although short-term continuity issues were alleviated, deficiencies remain in long-term benefit indicators. The limitation of Scheme 1 lies in the difficulty of land acquisition in the western basic farmland and the excessively long tail working line, which increases organizational complexity and operating cost. Scheme 4 avoids the land acquisition problem but requires a prolonged longitudinal-mining phase in its early stage, which delays the stable formation of the internal dump slope and raises the risk of secondary external dump failures. Its implicit assumption of “no land acquisition constraints” is also unrealistic, leading to lower scores in the “social” and “safety” dimensions.

Scheme 3, a refinement of Scheme 2, was designed to address transportation challenges caused by the excessively long working line. Nevertheless, its overall performance (V₃) was inferior to Scheme 2. This result indicates that subdividing a large mining area can mitigate certain engineering challenges but may introduce new stripping and scheduling complexities, without fundamentally resolving long-distance internal dumping issues. Its balanced yet unremarkable performance confirms that, during capacity expansion, prioritizing mining-area scale, dump stability, and production succession is strategically more critical than marginal improvements in local transportation conditions.

Scheme 2 stands out for its strategic foresight and systematic planning. By designating the western region as a large, continuous mining area, it not only ensures production continuity for at least five years but, more importantly, achieves a fundamental shift from longitudinal to lateral mining. This transformation aligns with the near-horizontal coal seam geometry, allowing the internal dump to stabilize along the coal-dip direction, thereby significantly enhancing slope stability and mitigating landslide risks. Furthermore, the large-scale layout facilitates the adoption of continuous or semi-continuous mining systems, consistent with the long-term trend toward mechanization and intelligent mining, and forms a robust platform for subsequent capacity scaling.

The weighting pattern obtained from the AHP-EW combination method in Table 7-where economic (0.2941) and engineering (0.1544) indicators receive relatively higher weights than environmental and social indicators—also reflects this strategic orientation at the expansion stage. In practice, environmental and social constraints (e.g., land acquisition red lines, ecological protection zones, and safety regulations) function as hard boundary conditions: only schemes that satisfy these baseline requirements can be considered feasible. Within this feasible set, the key differentiating factors that determine whether a scheme can support a sTable 11 Mt/a production level over the long term are its economic viability (stripping ratio, transportation cost, investment efficiency) and engineering robustness (dump stability, working-line organization, equipment matching). Therefore, assigning higher weights to economic and engineering indicators emphasizes the need to select schemes that are not only compliant with environmental and social standards, but also capable of sustaining capacity expansion under realistic capital and technical constraints.

Potential challenges and mitigation measures for scheme 2

It is worth noting that, even under this weighting pattern, Scheme 2 does not achieve its advantage by sacrificing environmental or social performance. On the contrary, the transverse-mining transition and the optimized internal dump layout improve slope stability and reduce potential disaster risk, which in turn supports regional environmental security and community safety. The higher comprehensive score of Scheme 2 thus reflects a coordinated optimization result: within the framework of strict environmental and social baseline constraints, economic benefits and engineering reliability are used as primary discriminators to identify the most suitable zoning configuration for the capacity expansion stage.

Although Scheme 2 was identified as the optimal option, several challenges remain for its implementation. The most significant is the average internal-dump haul distance of 3.0 km in the second mining area, which will markedly increase transportation costs. The study emphasizes the necessity to “conduct prior research on continuous coal conveying systems to minimize truck-based haulage.” Hence, future research should explore the technical and economic feasibility of introducing semi-continuous or fully continuous systems (e.g., in-pit crusher and belt conveyor) to offset the high transport costs.

Additionally, Scheme 2 requires substantial initial investment and precise production scheduling, imposing higher demands on mine management capability and operational coordination.

Universality and limitations of the research

The proposed “working-line-mining-division-uncertain measure evaluation” framework offers a generalizable approach to balancing multi-dimensional objectives of technology, economy, safety, and environment. Its core concept has strong universality and can serve as a theoretical and technical reference for near-horizontal open-pit coal mines in China and globally that are undergoing capacity expansion.

In particular, the “lateral instead of longitudinal mining” strategy provides valuable insights for open-pit mines in water-rich soft-rock formations facing high stripping ratios and internal dumping difficulties.

However, limitations remain. As stated in the conclusions, the current evaluation model does not yet incorporate dynamic variables such as real-time scheduling costs or external dump land acquisition constraints. Future work should integrate the model with GIS spatial data, production-scheduling simulation systems, and real-time monitoring platforms to construct a dynamically evolving decision-support system, thereby achieving adaptive optimization of mining division schemes and enhancing the model’s practical applicability.

In summary, this study not only identified the optimal capacity-expansion pathway for the Baoqing Chaoyang open-pit coal mine through rigorous modeling and quantitative evaluation but also provided a novel methodological solution for complex decision-making problems in open-pit mine planning. The selection of Scheme 2 represents a comprehensive balance of technical feasibility, economic rationality, and long-term operational stability, laying a solid foundation for the mine’s sustainable development.

Conclusion

This study develops an integrated framework for zoning and working-line optimization in the Baoqing Chaoyang open-pit coal mine during its capacity expansion from 7 to 11 Mt/a. By coupling working-line design, mining-zone restructuring, and multi-criteria comprehensive evaluation based on Unascertained Measure Theory (UMT), a transverse-mining substitution strategy is proposed to replace the existing longitudinal-mining layout and quantitatively compare alternative zoning schemes.

(1) First, a working-line optimization model is established by analyzing the coupling among working-line length, mining intensity, and production capacity. For the 11 Mt/a target, the rational working-line length is determined to be 1350–2050 m, corresponding to an annual advancing rate of 400–497 m, which helps alleviate geological hazard risks and resource succession pressure associated with excessive advancing intensity.

(2) Second, in line with the near-horizontal dip characteristics of the coal seams, the original longitudinal-mining area is reorganized into four transverse primary zones. Under the preferred Scheme II, these zones jointly provide 971.2 Mt of recoverable raw coal, with an average stripping ratio of 5.8 m³/t and a maximum service life of 34.3 years, thereby ensuring resource continuity under expansion conditions.

(3) Third, an eight-dimensional indicator system is constructed, covering geological, geotechnical, hydrogeological, topographical, engineering, economic, environmental, and social dimensions. A combined weighting mechanism integrating the Analytic Hierarchy Process and Entropy Weight Method, together with multiplicative normalization, yields a hybrid subjective-objective weighting set. Using four-level UMT membership functions and a confidence recognition criterion (λ = 0.6), Scheme II is identified as the optimal option, with a confidence level of 0.7058 in the Excellent category (V₁), confirming the effectiveness of the transverse-mining strategy.

Compared with recent advances in continuous extraction and continuous backfill mining with CO₂ mineralized filling materials, which focus on strata migration, fracture development, and water-preserving coal mining in underground settings^28,29,30, the present study addresses a different but complementary problem: large-scale open-pit capacity expansion and mining-zone optimization under complex surface and dumping constraints. While continuous extraction-continuous backfill technologies enhance green, low-carbon underground mining, the UMT-based zoning framework developed here improves the scientific rigor and transparency of decision-making for surface mine layout and capacity expansion. Together, these approaches contribute to coordinated resource exploitation, environmental protection, and long-term mine safety.

Limitations remain in that dynamic variables such as real-time scheduling costs, external dump land acquisition constraints, and temporal evolution of environmental impacts are not yet fully incorporated. Future work will integrate GIS-based spatial data, on-site monitoring, and scheduling simulation models to build a dynamic and updatable platform for mining-zone delineation and evaluation, further enhancing model adaptability and field deployment capability.