Simulation-Based Optimization and Prevention Strategies for Underground Heat Hazards in Menkeqing Coal Mine

Abstract

This study investigates underground heat sources and develops effective strategies for mitigating heat hazards in coal mines, with a focus on the design and optimization of cooling systems. Using the 3107 fully mechanized mining face of Menkeqing Coal Mine as a case study, geological survey data and in situ measurements were combined to evaluate the severity of thermal hazards. Thermodynamic and heat transfer models were applied to quantify heat dissipation from multiple sources. Computational fluid dynamics (CFD) simulations, based on data-driven modeling and geometric reconstruction, tested different equipment layouts and spacing configurations to identify optimal cooling schemes. Field implementation of the designed cooling system confirmed its effectiveness, offering practical guidance for improving heat hazard control and cooling system optimization in deep coal mines.

1. Introduction

As mining depth increases, coal seams are subjected to high geothermal gradients, heat transfer from surrounding rock, and heat generated by large-scale mechanized equipment. These factors intensify underground thermal environments, leading to severe heat hazards that threaten worker safety, reduce production efficiency, and increase ventilation and cooling costs. Previous studies have demonstrated that ventilation improvement alone becomes ineffective when working faces reach persistent secondary heat hazard levels. Artificial refrigeration and optimized cooling system design are therefore recognized as indispensable measures in deep mines.

In recent years, numerical simulation and heat transfer modeling have been widely applied to predict underground thermal behavior and evaluate cooling measures. However, many existing approaches either simplify heat source characterization, neglect coupling effects between multiple heat sources, or lack validation against actual production data. Consequently, there remains a gap between theoretical prediction and practical implementation of effective cooling strategies. To address this gap, this study uses the 3107 fully mechanized mining face of the Menkeqing Coal Mine as a case study. Comprehensive analysis of coal and rock thermal properties, in situ borehole temperature data, and heat release from major underground sources—including surrounding rock, electromechanical equipment, and coal transportation—was performed to quantify the severity and causes of heat hazards. Building on this analysis, computational fluid dynamics (CFD) simulations were employed to optimize cooling equipment spacing and arrangement, with results directly compared with field measurements. Finally, a centralized refrigeration cooling system was designed and implemented, and its performance was evaluated on site.

This integrated approach provides both a theoretical framework for accurately quantifying underground heat sources and a validated engineering solution for cooling system optimization. The findings not only solve pressing heat hazard problems at Menkeqing Coal Mine but also offer transferable methodologies and design principles for other deep coal mines facing similar thermal challenges [1,2,3,4].

The Menkeqing Coal Mine, operated by Zhongtian Hechuang Energy Co., Ltd., is located in the Hujierte mining area of the Dongsheng Coalfield within the Ordos Basin. The regional geothermal field is strongly influenced by basement fault zones, which enhance conductive heat transfer and create a typical high-temperature geological background. The 3107 fully mechanized mining face, which serves as the focus of this study, employs a U-shaped ventilation system. Both the intake and return airways are approximately 2820 m in length, and the face has a strike length of 320 m. The roadway cross-section is rectangular with an average area of 25.53 m². These geological and engineering features establish the spatial framework of the study and provide essential context for analyzing the occurrence and control of heat hazards in this deep coal mine.

2. Evaluation of Heat Hazard Severity at Menkeqing Coal Mine

Geological survey data were obtained from the official production geological reports of Menkeqing Coal Mine (2022–2023), and in situ measurements were continuously carried out during the same period at the 3107 working face. Measurement points included shallow and deep boreholes at multiple roadway sections, ensuring that both vertical and horizontal variations in geothermal gradients were captured. These data cover representative zones of the mine and provide a reliable basis for evaluating the overall heat hazard level.

This section quantifies the severity of underground heat hazards in the Menkeqing Coal Mine. Physical properties of coal and rock samples, combined with survey data and borehole temperature measurements at different depths, were used to calculate geothermal gradients and characterize the local geothermal field. These results provide the basis for hazard classification and temperature prediction. According to the Technical Regulations for the Prevention of Heat Hazards in Mines (State Administration of Work Safety, China, 2011) [5] and related studies, mine heat hazards are commonly classified into primary and secondary categories based on measured rock and airflow temperatures. A working face is considered to be in the primary heat hazard zone when the surrounding rock temperature or airflow temperature lies in the range of approximately 26–31 °C, at which point miners experience increased physiological strain and ventilation becomes less effective. When temperatures further rise to 31–37 °C, the mine is designated as a secondary heat hazard zone, characterized by severe thermal stress, rapid decline in labor efficiency, and urgent need for artificial cooling. Temperatures above 37 °C are typically regarded as serious or extreme heat hazard conditions, under which normal production cannot be sustained without intensive cooling measures. These classifications provide a standardized basis for evaluating underground thermal conditions. In this study, we adopt these definitions to determine the severity of heat hazards at the Menkeqing Coal Mine and to guide the design of appropriate cooling strategies [6,7,8].

2.1. Thermal Properties of Coal and Rock

A PDR-300 thermal conductivity tester was employed to measure the thermal physical properties of coal and rock samples, which are essential input parameters for calculating the surrounding rock temperature field and heat dissipation at the mining face. Following the requirements of thermal property testing and relevant sampling standards, freshly exfoliated coal and newly collapsed roof rock were collected from the 3107 working face of the Menkeqing Coal Mine. The samples were sealed on site and transported to the laboratory for preparation and testing. The results showed that the coal samples had an average thermal diffusivity of 1.1682 m²/s, an average thermal conductivity of 1.7970 W/(m·°C), and an average volumetric heat capacity of 1.55 × 10⁶ J/(m³·°C). For the rock samples, the average thermal diffusivity was 1.1094 m²/s, the average thermal conductivity was 3.0723 W/(m·°C), and the average volumetric heat capacity was 2.99 × 10⁶ J/(m³·°C). These parameters provide a robust basis for numerical simulation and thermal analysis of the mining environment.

2.2. Borehole Temperature Measurement Methods

Prior to use in thermodynamic calculations and CFD modeling, raw measurement data were subjected to preprocessing, including the removal of abnormal values caused by equipment noise, interpolation of missing points, and calibration of thermocouples using the ice point method. Statistical regression and least-squares fitting were applied to smooth fluctuations and ensure consistency across datasets.

2.2.1. Preparation of Temperature Measurement Equipment

The heat transfer characteristics of surrounding rock and coal seam surfaces at the mining face were measured using a WRNT-01 thermocouple. A digital multimeter with an accuracy of 200 μV was used to acquire temperature data. Prior to measurement, the relationship between the thermoelectric voltage $V$ and the temperature $t$ was calibrated. The calibration function is expressed as follows:

$$t=mV+t_0,$$

(1)

In the equation, $t$ is the temperature at the hot junction of the thermocouple (°C), $m$ is the reciprocal of the thermoelectric coefficien (°C/mV), $V$ is the thermoelectric voltage between the hot and cold junctions (mV), and $t_0$ is the temperature at the cold junction of the thermocouple (°C).

Calibration in this study was carried out using the ice point method. The results were classified into two groups according to borehole depth: shallow boreholes (1.5 m) and deep boreholes (20 m). For the shallow borehole group, the fitted calibration equations produced values of the parameter $m$ ranging from 24.326 to 27.666 and values of $t_0$ ranging from 42.752 to 49.276, with a minimum correlation coefficient of 0.9900. For the deep borehole group, the parameter $m$ ranged from 23.877 to 26.530, and $t_0$ ranged from 43.753 to 53.809, also with a minimum correlation coefficient of 0.9900. The results confirm that the calibration equations for both groups achieved the required accuracy for measuring surrounding rock temperatures in the mining environment [9,10,11].

2.2.2. Surrounding Rock Temperature Measurement via Shallow and Deep Boreholes

Two methods were employed to measure surrounding rock temperature: shallow borehole and deep borehole measurements. At the 3107 working face, the measurement location was set at 800 m along the main haulage roadway. From this point, three boreholes were drilled at 20 m intervals, using the same specifications for both shallow and deep drilling. Shallow boreholes were drilled to a depth of 1.5 m, while deep boreholes extended to 20 m. Each borehole was measured three times, and the average of the three readings was recorded as the final value. The results showed average temperatures of 31.8 °C for shallow boreholes (1.5 m depth) and 32.6 °C for deep boreholes (20 m depth). These measurements provide essential input for analyzing the geothermal gradient and heat transfer characteristics of rock strata in the mining area.

2.3. Results of Surrounding Rock Temperature Distribution

Temperature data from multiple boreholes were analyzed to investigate variations in geothermal gradients at different depths. The geothermal gradient was calculated using the following equation:

$$T_{G100}=100\times {\displaystyle \frac{T_2-T_1}{H_2-H_1}},$$

(2)

In the equation, $T_{G100}$ is the geothermal gradient (°C/100 m), $T_1$ is the temperature of the isothermal zone (°C), $H_1$ is the depth of the isothermal zone (m), $T_2$ is the temperature of the surrounding rock at the measurement borehole (°C), and $H_2$ is the burial depth of the measurement borehole (m).

According to long-term observation data and geological survey records, the depth of the isothermal zone ranges from 80 to 120 m, with temperatures between 10 and 14.9 °C. The calculated geothermal gradients are summarized in

Table 1. Geothermal gradients calculated from borehole temperature measurements at different depths in the Menkeqing Coal Mine.

Borehole ID	Temperature Transition Zone (°C)		Isothermal Zone (°C)		Temperature Increasing Zone (°C)			Geothermal Gradient (°C/100 m)
	Temperature at 20 m (°C)	Temperature at 80 m (°C)	Temperature at 80 m (°C)	Temperature at 120 m (°C)	Temperature at 120 m (°C)	Depth (m)	Geothermal Temperature (°C)
HS11	10.4	10.7	10.7	11.3	11.3	860	33.6	3.01
MS19	9.1	10.3	10.3	11.0	11.0	900	33.1	2.83
MS26	8.6	11.2	11.2	11.7	11.7	880	32.0	2.67
MS32	8.4	11.2	11.2	12.0	12.0	885	33.2	2.77
MS33	8.0	10.0	10.0	10.7	10.7	900	31.0	2.60
MS34	13.0	14.4	14.4	14.9	14.9	880	34.6	2.59
Average value	9.58	11.3	11.3	11.9	11.9		32.9	2.75

. The values for individual boreholes ranged from 2.59 to 3.01 °C/100 m, with an average of 2.75 °C/100 m across the study area.

Measured data indicate that the geothermal temperature at the current mining level generally exceeds 30 °C. Using the least-squares method, a linear equation was derived to describe the relationship between original rock temperature and burial depth as follows:

$$t=0.0375h+6.634.$$

(3)

Similarly, the linear relationship between geothermal gradient and burial depth can be expressed as

$$td=0.0016h+1.819.$$

(4)

Based on geothermal data from the mine’s geological report, the onset depths of primary and secondary heat hazards at each borehole and the geothermal temperature at the maximum 3-1 seam depth were calculated from the geothermal gradient. The results are presented in

Table 2. Calculated initial burial depths of primary and secondary heat hazard zones and geothermal temperatures at the maximum burial depth of the 3-1 coal seam in the Menkeqing Coal Mine.

Borehole ID	Measured Borehole Depth (m)	Measured Bottom Temperature (°C)	Corrected Bottom Temperature (°C)	Onset Depth of Primary Heat Hazard (m)	Onset Depth of Secondary Heat Hazard (m)	Geothermal Temperature at Maximum 3-1 Seam Depth (°C)
B03	900	43.1	43.1	599	748.3	35.06
B04	900	38.8	39.4	664.7	832.8	31.66
B05	900	42.6	43.3	596.3	744.4	34.5
HS11	860	33.6	34.1	759.4	954.2	28.67
MS19	900	33.1	33.6	809.4	1018.5	27.36
MS26	880	32	32.5	826.4	1040.7	26.96
MS32	885	33.2	33.7	792.8	997.6	27.78
MS33	900	31	31.5	880.8	1110.7	25.78
MS34	880	34.6	35.2	746.2	937.3	28.95
Average value				741.7	931.6	29.6

.

As shown in Table 2, the highest geothermal temperature at the maximum burial depth of the 3-1 coal seam is 35.06 °C, the lowest is 25.78 °C, and the average is 29.6 °C. According to the classification of heat hazard zones, only boreholes B03, B04, and B05 suggest that the 3-1 coal seam has entered the primary heat hazard zone (31–37 °C).

Using temperature data from deep and shallow boreholes together with formulas (2) and (3), the deep geothermal temperature of the mine was predicted. The results indicate that areas above an elevation of 486 m fall within the primary heat hazard zone, while areas below 486 m belong to the secondary heat hazard zone, classified as a high-temperature anomaly area.

2.4. Prediction and Measurement of Air Temperature at the Mining Face

To predict the thermal environment at the working face, we derived a mathematical model from the heat exchange differential equation of airflow in underground roadways. This equation accounts for convective heat transfer between the airflow and surrounding rock, as well as heat dissipation from coal and gangue during transportation. The governing equation was solved using the finite difference method, which enables estimation of airflow temperature distribution along the roadway. To calibrate the model parameters and validate prediction accuracy, statistical regression analysis (least-squares fitting) was applied to compare model outputs with measured temperature data from multiple boreholes and working faces. This approach allowed adjustment of empirical coefficients in the heat transfer terms to minimize prediction error. The final model equations (Equations (5)–(7)) provide quantitative relationships between airflow temperature, roadway length, and heat source intensity and were verified against field measurements at the 3107 working face, showing deviations within ±1.2 °C. This combined mathematical–statistical framework ensures that the prediction method can be generalized and applied to other mining faces with similar geological and thermal conditions.

$$t_2=t_1·\exp \left(-{\displaystyle \frac{N}{1+E^{\phi }}}\right)+[1-\mathrm{e}\mathrm{x}\mathrm{p}(-{\displaystyle \frac{N}{1+E^{\phi }}})](t_s+{\displaystyle \frac{F}{N}}),$$

(5)

$$N=(K_{l}UL+2.4\times {10}^{-3}{C}_K{M}_K{L}^{0.8})/M_B{C}_B,$$

(6)

$$F=(\sum Q_M-8.4\times {10}^{-3}{C}_K{M}_K{L}^{0.8})/M_B{C}_B.$$

(7)

Based on the airflow temperature prediction model, the temperature at the 3107 working face was estimated to range between 29.54 °C and 31.46 °C. This prediction offers a reference for subsequent simulation optimization of the 3107 face, as well as for heat hazard prevention and forecasting in the 4-1 coal seam and other areas.

Field measurements of airflow temperatures were carried out at several mining faces. At the 3107 mining face, the airflow temperature was 29.2 °C, while the return airway heading face recorded a range of 28–30 °C. The maximum temperature at the bottom yard of the 3-1 coal 11th panel reached 32 °C, classifying it as a secondary heat hazard zone. Similarly, the southern wing of the same panel recorded a peak of 30.6 °C, also indicating a secondary heat hazard zone. Considering both original rock and airflow temperature classifications, the mine as a whole is categorized as a secondary heat hazard mine. Given the persistent secondary heat hazard conditions at the southern wing and bottom yard of the 3-1 coal 11th panel, strengthening ventilation and increasing airflow velocity in the 11-3107 working face are no longer effective. Therefore, artificial refrigeration measures must be implemented to reduce heat hazards.

3. Identification of Major Heat Sources in the Working Face

To enhance cooling efficiency beyond the current ventilation and cooling measures, this study optimizes the cooling system layout through thermal modeling and energy simulation. Before calculating heat release from each section of the 3107 working face, the effect of surface temperature on the underground thermal environment was analyzed. This step is essential to reduce the influence of external factors and ensure accurate heat source energy calculations.

3.1. Influence of Surface Temperature on Underground Conditions

To evaluate the seasonal influence of surface climate on underground conditions and to accurately assess annual variations in heat hazards, seasonal air temperatures from both the surface and underground sites were collected and analyzed. The analysis focused on the impact of surface climate on the temperature at the 3107 coal mining face. Surface air temperatures at the Menkeqing Coal Mine exhibited strong seasonal cycles with large fluctuations, while the 3107 working face showed minimal annual variation, with a fluctuation of only 0.7 °C. This trend is illustrated in Figure 1a.

Surface meteorological parameters and underground climatic conditions were continuously monitored at multiple locations in the mining area to assess the influence of surface climate on air temperature variations at the 3107 working face. To avoid interference from production activities, measurements were taken between 8:00 a.m. and 11:00 a.m. Air temperatures were recorded at the auxiliary shaft entrance and at several points within the working face, including the upper and lower corners, over a three-hour period. The results are shown in Figure 1b. Considering both seasonal and daily fluctuations, it can be reasonably concluded that surface temperature has little effect on the working face temperature.

3.2. Heat Sources and Their Quantification

The thermal environment of the mining face is shaped by multiple underground heat sources. The fundamental principle of heat exchange between these sources and the airflow forms the theoretical basis for analyzing this environment. To achieve accurate estimates of heat dissipation, the heat release from each individual source must be examined [12,13,14,15].

3.2.1. Heat Generation from Air Compression

As air descends from the surface to the shaft bottom, it is compressed by gravitational force, leading to increased density and reduced volume. According to the ideal gas law, the temperature of the compressed air rises accordingly. The calculation formula for heat generated by air compression is as follows:

$$Q_Z=G·A·(Z_1-Z_2)·E.$$

(8)

In the equation, $Q_Z$ is the heat generated by air compression (kW); $G$ is the mass flow rate of the airflow (kg/s); $A$ is the conversion factor of air thermal work, 9.81 × 10⁻³ (kJ/(kg·m)); $Z_1$ is the elevation of the airflow inlet (m); $Z_2$ is the elevation of the lowest point of the airflow at the shaft bottom (m); and $E$ is the coefficient representing the fraction of compression heat absorbed by the airflow, with $E\le 1$.

Due to the convective heat exchange between the airflow and the shaft wall, part of the compression heat is absorbed by the shaft wall when the airflow temperature is higher than that of the wall. However, since a coupled heat balance calculation method involving both air compression heating and convective heat exchange with the shaft wall is adopted, the coefficient $E$ is taken as 1. The compression heat generated by the descending airflow reaching the shaft bottom is calculated to be 323.8 kW [16,17].

3.2.2. Heat Release from Surrounding Rock

In mine roadways, the surrounding rock serves as a heat source due to its inherent thermal properties. Heat conduction within the rock is unsteady; even if the surface wall temperature remains constant, heat transfer from the rock interior to the surface changes over time. Similarly, heat exchange between the surrounding rock and the airflow is unsteady, and the heat released by the rock is given by

$$Q_w=K_{\tau }·(t_y-t_k)·U·L.$$

(9)

In the equation, $t_y$ is the original rock temperature (°C); $t_k$ is the average air temperature in the roadway (°C); $L$ is the length of the roadway (m); $U$ is the perimeter of the roadway (m); and $K_{\tau }$ is the unsteady heat transfer coefficient (kW/(m²·°C)), generally taken as 0.0073 kW/(m²·°C) [18,19,20].

Heat released by surrounding rock is the dominant source of elevated underground temperatures. Based on analysis and field investigations and excluding return-air roadway dissipation, the main heat release from surrounding rock at the 3107 working face is 785 kW.

3.2.3. Heat Release from Coal and Gangue Transportation

Detailed calculations of heat dissipation must account for coal and gangue transported in the mine as they significantly influence underground heat and moisture transfer. According to heat transfer theory, the heat released from coal and gangue can be expressed as

$$Q_y=m_k·C_y{·∆t}_k.$$

(10)

In the equation, $Q_y$ represents the heat released by coal and gangue during transportation (kW), $m_k$ is the mass flow rate of coal and gangue during transportation (kg/s), $C_y$ denotes the average specific heat capacity of coal and gangue (kJ/(kg·°C)), and ${∆t}_k$ is the temperature difference representing the cooling of coal and gangue within the considered roadway section (°C).

Analysis and measurements indicate that heat exchange between coal or gangue and the airflow during transportation at the 3107 working face amounts to 22.78 kW.

3.2.4. Heat Dissipation from Electromechanical Equipment

At the 3107 working face, a mobile train carrying electromechanical devices—including a mobile transformer and spray pump water tank—was positioned 600 m from the intake corner in the main haulage roadway. Using an infrared thermal imaging camera, temperature and emissivity measurements were taken at three points on the equipment surface: upstream, midsection, and downstream relative to the airflow direction. Temperature distributions of electromechanical equipment surfaces at different measurement points are shown in Figure 2. To evaluate the heat dissipation characteristics of electromechanical equipment, surface temperatures were measured at multiple points along the equipment body. Each measurement location was assigned a number in sequential order, and the average values were plotted on a line chart. Here, the x-axis label “Number” corresponds to these designated measurement points. A line chart was selected because it more clearly illustrates both the overall temperature range and the variation trend across equipment surfaces, compared with bar or scatter plots.

During both production and non-production periods, the upper surfaces of electromechanical and power distribution equipment exceeded 32 °C, with most devices at about 34 °C, some reaching 40 °C, and a maximum of 59.8 °C. Lower surfaces were all above 31 °C, typically near 34 °C, with a few exceeding 40 °C. Surrounding equipment temperatures also exceeded 31 °C, averaging around 34 °C, with some as high as 60.5 °C. Because these surface temperatures are higher than the airflow temperature, heat transfer occurs from the equipment to the airflow, consistent with heat conduction theory. Heat dissipation from electromechanical equipment is calculated as

$$Q_j=N_m·K_m.$$

(11)

In the equation, $N_m$ is the rated power of the equipment (kW), and $K_m$ is the comprehensive coefficient, generally taken as 0.2.

Based on analysis and on-site investigations and excluding low-power equipment below 20 kW, the total rated power is 950 kW under normal operation and 1832.7 kW when all equipment is fully utilized.

3.2.5. Oxidation of Coal and Gangue

Oxidation heat released from coal or gangue contributes to temperature increases at the working face. This heat can be calculated as follows:

$$Q_h=q_0·F·W^{0.8}.$$

(12)

In the equation, $q_0$ is the equivalent oxidation heat release coefficient (kW/m²), representing the heat released per unit area of roadway surface per unit time due to oxidation when the roadway airflow velocity is 1 m/s; $W$ is the average airflow velocity in the roadway (m/s); and $F$ is the oxidation surface area (m²).

At the 3107 working face, analysis of the underground environment identified two primary oxidation heat sources: 141.8 kW from rock walls and 122.7 kW from coal on the conveyor belt. Together, these account for a total oxidation heat release of 264.5 kW.

3.2.6. Heat Dissipation from Personnel and Lighting

The heat released by underground workers primarily depends on the intensity and duration of their work. The heat dissipation from personnel can be calculated using the following formula:

$$Q_r=n·q.$$

(13)

In the equation, $Q_r$ is the heat dissipation from personnel (kW); $n$ is the number of personnel (persons); and $q$ is the average heat dissipation per person (kW), taken as 275 W considering moderate physical labor.

Within the mining face, heat contributions from lighting and hot water trenches are negligible compared with total heat dissipation and can be disregarded. Because incoming surface air is cooler than the working face air, it does not add heat to the underground environment. Field investigations indicated that up to 16 workers (the maximum allowed) operate simultaneously at the 3107 working face, contributing a total personnel heat dissipation of 4.4 kW.

Using calculation methods for various heat sources, the total heat dissipation at the 3107 working face was estimated at 2880.5 kW. The dominant contributors are surrounding rock and electromechanical equipment, which together account for 71.4% of the total, identifying them as the primary causes of elevated temperatures at the working face.

4. Simulation-Based Optimization of Cooling System

Building on the heat dissipation analysis of underground heat sources and the expected cooling effects, additional simulations were performed to examine the impact of radiator spacing, equipment train arrangement, and radiator placement on cooling efficiency. To establish the numerical model, the physical system of the cooling setup was first reconstructed using SolidWorks 2023 (Dassault Systèmes, Vélizy-Villacoublay, France), based on the actual geometry of the 3107 intake roadway, the equipment train, and the cooling units. The geometry was then imported into ANSYS Fluent 2023 (Ansys, Inc., Canonsburg, PA, USA) for meshing and thermal–fluid simulation. A hybrid grid system with approximately 2.1 million cells was generated, using hexahedral elements in the roadway and tetrahedral refinement around equipment and cooling units to capture complex flow fields. Grid independence tests confirmed that further mesh refinement resulted in less than 2% variation in predicted airflow temperature. Boundary conditions were defined as follows: a velocity inlet with measured intake air parameters (30 °C, 2.8 m³/s), a pressure outlet at the return airway, and constant wall temperatures for roadway and equipment surfaces based on field measurements. Heat source intensities from surrounding rock, electromechanical equipment, coal transport, and oxidation were assigned according to the calculations in Section 3. The airflow was modeled as an incompressible fluid under steady-state turbulent flow. The standard k–ε turbulence model with standard wall functions was employed, and the Boussinesq approximation was applied for buoyancy effects. The governing equations were solved using the SIMPLE algorithm with second-order discretization. Convergence was assumed when residuals fell below 10⁻⁴ for continuity, momentum, and energy equations. The simulation results were validated by comparing predicted airflow temperatures at key monitoring points with field measurements from the 3107 working face. The deviation between simulation and measured data was within ±1.5 °C, confirming the accuracy and reliability of the model. The goal was to identify the optimal cooling configuration. Predictive modeling combined with field measurement comparisons was used to assess the cooling requirements of the 3107 working face. The preliminary design aimed to lower the air temperature from 29.2–30.6 °C (after passing through the equipment train) to 19.2–25.6 °C and further to 15.6–17.8 °C after passing through all cooling equipment along the 3107 working face. For simulation purposes, the inlet air temperature was set at 30 °C, the temperature after the cooling equipment at 17 °C, and the temperature after the equipment train at 22 °C. Comprehensive simulations were conducted under these conditions to analyze the temperature field and airflow distribution, incorporating all identified heat dissipation sources.

4.1. Model Establishment and Assumptions

To enable numerical simulation, the physical system of the cooling setup was simplified. Based on the actual layout and operating conditions, a model was constructed in SolidWorks (Figure 3). The abstraction considered both the structure and working principles of the cooling units while minimizing complexity. For simplification, the internal condensation process within the cooling equipment was excluded, and the simulation focused on the heat extraction and cooling effects of the units. Since no active cooling occurs between the working face and the return airway, the modeling was restricted to the intake roadway where the cooling devices were installed. A physical model of the intake roadway was then developed in SpaceClaim using the actual dimensions of the 3107 fully mechanized coal mining face. The simplified model included the roadway, equipment train, and radiator arrangement, forming the basis for thermal-fluid simulations in subsequent analyses.

4.2. Initial Condition Assumptions

During the numerical simulation of the thermal environment in the coal mining face, the following simplifications are assumed: The roadway surrounding rock in the mining face is homogeneous and isotropic. The underground airflow is treated as an incompressible fluid, and the heat dissipation caused by viscous work of the airflow is neglected. The airflow is considered to be in steady-state turbulent flow, satisfying the Boussinesq approximation, where density variations are only considered when calculating buoyancy forces. Moisture exchange between the surrounding rock and airflow, as well as condensation caused by mixing of hot and cold airflows, is neglected. The standard wall function method is used to treat the near-wall region. The wall temperatures of the intake airway, mining face, and return airway, as well as equipment-related temperatures, are assumed to be constant [21,22,23,24].

4.3. Numerical Simulation and Optimization of the Cooling System

4.3.1. CFD Simulation of Cooling Equipment Spacing

Airflow and heat transfer in the fully mechanized coal mining face are governed by fundamental heat transfer principles. In mine roadways, airflow is typically turbulent, so CFD simulations require an appropriate turbulence model. Here, the standard k-ε model was used. To maximize cooling efficiency, the layout strategy placed cooling units upstream of the heat-dissipating equipment train. With the positions of the first cooling unit and the equipment train fixed, simulations were conducted by adjusting the spacing between cooling units to determine the optimal configuration.

A baseline simulation was first performed with 50 m spacing. The resulting temperature contours (Figure 4b) closely matched field observations, confirming the accuracy of the model. Using this as a reference, additional simulations were carried out with spacings of 30 m, 70 m, and 90 m. Cooling effects were then evaluated by comparing temperature cloud maps and profiles (Figure 4), and the effectiveness of each configuration was analyzed.

Comparison of temperature distributions in Figure 4 shows that simulated heat dissipation and related parameters align well with field conditions, confirming that the model and simulation settings are reasonable. Cooling effects at different intervals revealed that a 30 m spacing produced weak cooling at the roadway end, especially beyond the equipment train. In contrast, spacings of 70 m and 90 m significantly improved cooling performance at the roadway end. For detailed analysis, a straight-line path along the intake roadway was used to record pointwise temperatures, as shown in Figure 5a. The results identified 90 m as the optimal spacing, with cooling units placed 50 m upstream of the equipment train.

4.3.2. Simulation of Cooling Equipment Arrangement Schemes

To further refine the layout, the relative position of the equipment train and cooling units was adjusted. Specifically, the cooling performance was evaluated when the equipment train was moved to a position between cooling units. In the model, the spacing between cooling units was fixed at 90 m, while the distance between the equipment train and the nearest cooling unit was set at 50 m. Four layout schemes were considered, with zero, one, two, or three cooling units placed downstream of the equipment train. The scheme with no downstream units corresponded to the previously identified optimal 90 m spacing configuration. CFD simulations were conducted for each scenario, and the results are presented in Figure 6.

Comparison of the temperature distributions in Figure 6 shows that cooling effectiveness at the roadway end decreased when the equipment train was repositioned between cooling units. The original optimized scheme—with 90 m spacing and the equipment train located at the end—still provided the best performance. For detailed evaluation, a straight line along the intake airway was selected for pointwise temperature analysis, and the results are shown in Figure 5b. These results confirm that the optimal configuration remains a 90 m spacing between cooling units with the equipment train positioned at the end.

5. Design and Implementation of Cooling Measures

5.1. Calculation of Cooling Load

Refrigeration cooling was designed for the 3107 fully mechanized mining face. Considering that heat exchange occurs before air reaches the face, even after being cooled to 22 °C, the practical cooling target was set to reduce the ambient temperature at the face to 28 °C. Cooling requirements were estimated using average air parameters for the hottest month. Following the measurement points defined in the Mine Cooling Technical Specifications, a backward segmented thermal calculation was applied. The cooling load was calculated using the enthalpy difference method [25,26,27,28]:

$$Q=G·\rho ·\left(h_1{-h}_2\right).$$

(14)

In the equation, $G$ is the volumetric flow rate (m³/s); $\rho$ is the air density, taken as the density after cooling, in kg/m³; $h_1$ is the air enthalpy before cooling (kJ/kg); and $h_2$ is the air enthalpy after cooling (kJ/kg).

According to the Technical Regulations for the Prevention of Heat Hazards in Mines, the ambient temperature at the working face should be reduced from 32 °C to 28 °C, with relative humidity lowered from 95% to 85%. With an air volume of 2800 m³/min, the cooling load was calculated as 1427 kW, requiring a refrigeration capacity of at least 1712 kW.

5.2. Configuration of the Cooling System

The cooling system comprised three subsystems: a chilled water system, a cooling water system, and a spray water system. In the chilled water circuit, the main refrigeration unit produced water at a minimum of 3 °C. The chilled water was pumped through insulated pipes to six cooling cabinets installed in the intake airway, where it exchanged heat with the airflow, reducing both temperature and humidity in stages. After heat exchange, the chilled water temperature rose to about 14.4 °C before returning to the evaporator of the refrigeration unit. In the cooling and spray water circuits, heat absorbed by the refrigerant was transferred to the cooling water, which was circulated to a closed cooling tower in the return airway. There, heat was dissipated through spray evaporation and exchange with return airflow, lowering the water temperature to 3–5 °C before recirculation. The system flow is shown in Figure 7a.

In practice, three refrigeration units operated in series in the chilled water circuit to provide stepwise cooling. Low-temperature chilled water was delivered to the six combined cabinets, arranged according to the optimized simulation scheme. The cabinets’ air systems were connected in series along the coal transportation roadway. The equipment train was located downstream of the cooling setup, with the six cabinets providing staged cooling and dehumidification of intake air, as illustrated in Figure 7b.

5.3. Field Application and Cooling Effects

The underground centralized cooling system with return-air heat exhaust was applied for the first time at the fully mechanized face of Menkeqing Coal Mine. After being put into operation at the 3107 working face, the system ran continuously for two weeks, achieving outlet air temperatures as low as 17.4 °C and humidity reduced to 80%, closely matching simulation predictions. The system operated steadily, providing effective cooling and dehumidification. On-site test data collected during stable operation showed consistent and stable temperature and humidity trends. Figure 8a,b compare temperature and humidity before and after implementation of the return-air cooling system at a representative time during the operation period.

Performance evaluation of the refrigeration system showed that the 3107 working face temperature decreased from 29.2–30.6 °C to 19.2–25.6 °C after commissioning, a reduction of 5–10 °C, or more than 16.3%. The greatest drop occurred 50 m from the equipment conveyor, where temperatures fell from about 29.2 °C to 15.6–17.8 °C, a maximum reduction of 46.6%. After applying the return-air cooling system, face temperatures fully met the “Coal Mine Safety Regulations” limit of 26 °C. Humidity decreased from about 98% to 76–80%, a reduction of 18–22% (exceeding 18.4%), with the largest drop (from 98% to 74%) also occurring 50 m from the conveyor. The actual cooling performance closely matched simulation results, demonstrating effective cooling and dehumidification that largely solved the heat hazard problem at the 3107 working face.

6. Discussion

This study builds upon and extends existing research on deep mine heat hazard prevention. Previous work has often emphasized either ventilation improvements [1,12] or the application of refrigeration systems [2,3], but few studies have systematically quantified the contributions of different heat sources and validated optimized cooling layouts under real production conditions. By combining comprehensive heat source characterization, CFD-based cooling optimization, and field verification, our work provides a more holistic framework that bridges theoretical modeling and engineering application.

The findings also have broader implications. First, the identification of surrounding rock and electromechanical equipment as the dominant heat sources (accounting for 71.4% of total dissipation) highlights priority targets for energy-efficient cooling design. Second, the demonstrated agreement between CFD predictions and field data strengthens confidence in simulation-based optimization as a practical design tool for underground cooling systems. Finally, the successful application of a centralized refrigeration system at Menkeqing illustrates a transferable methodology that can inform design and hazard prevention in other deep coal mines worldwide.

Future work should focus on refining dynamic heat source models, integrating moisture and condensation processes into thermal simulations, and exploring hybrid cooling strategies that couple refrigeration with renewable energy or heat recovery. Such advances would further improve efficiency and sustainability in deep mine climate management.

The methodology developed in this study—comprising heat source quantification, CFD-based cooling optimization, and field verification—can be applied to other underground mines beyond the 3107 working face of Menkeqing Coal Mine. However, several limitations should be noted. First, the present analysis is based on coal seam conditions, whereas metal mines may exhibit different dominant heat sources, such as geothermal gradients or radioactivity-related heat, requiring model parameter recalibration. Second, the study focuses on a specific depth, seam thickness, and geological context, which may limit the direct transferability of results. Third, the optimization strategy mainly considers refrigeration-based cooling and does not fully integrate renewable or hybrid energy-saving methods. Future work will aim to extend the framework to different mine types, including metal mines, and to a range of seam depths and thicknesses. Moreover, dynamic heat source modeling, incorporation of multi-energy coupling strategies, and multi-mine comparative validation will be pursued to enhance the generalizability and practical applicability of the approach.

7. Conclusions

This paper investigated the severity, causes, and mitigation of heat hazards at the 3107 working face of the Menkeqing Coal Mine through a combined approach of field measurement, heat source quantification, CFD simulation, and engineering implementation. The results show that surrounding rock and electromechanical equipment are the dominant contributors to heat accumulation and that optimized cooling system layouts can substantially improve thermal conditions. The implementation of a centralized refrigeration system reduced face temperatures by more than 16% and lowered humidity by over 18%, achieving compliance with mine safety regulations.

Beyond the case study, this work demonstrates the value of integrating detailed thermal characterization with simulation-based design and field validation. The framework established here provides both a reference model for hazard evaluation and a practical methodology for cooling system optimization in other deep coal mines facing similar challenges.