Research on Optimization of Forced Ventilation Parameters for Blasting Construction in Large-Section Tunnels Based on CFD

Abstract

Large-section tunnels produce a large amount of dust after drill-and-blast construction. If not removed in a timely manner, the dust will seriously endanger workers’ health. For the purpose of enhancing the working conditions within the tunnel during construction, this investigation employs an integrated methodology that combines computational simulations with on-site measurements. Drawing upon the principles of gas–solid two-phase flow theory, the coupled diffusion law of airflow and dust in large-section tunnels is investigated. A two-factor orthogonal experiment combined with economic analysis is employed to determine the optimal ventilation parameters for the forced ventilation system. The findings indicate that, when the initial ventilation configuration is applied, the airflow field is divided into three stages, and dust diffusion is primarily driven by airflow. The average dust concentration in the 1.6 m breathing zone at 600 s post-blasting is measured to be 36.8 mg/m³. While satisfying the ventilation demand stipulated for the tunnel, the optimal ventilation parameters are determined as an outlet air velocity of 18 m/s and a duct-to-face distance of 40 m. Under these conditions, the dust concentration is reduced to 1.5 mg/m³, representing a 95.9% improvement in dust removal efficiency. Additionally, the hourly electricity cost at 18 m/s is USD 4.39 lower than that at 20 m/s. This study provides valuable insights for optimizing forced ventilation parameters in large-section tunnels, significantly reducing pollutant levels while saving costs.

1. Introduction

Compared with cross-sea bridges, subsea tunnels offer significant inherent advantages in construction cost, operational lifespan, safety, and reliability. In recent years, subsea tunnel projects worldwide have exhibited rapid development, and the drill-and-blast method has become an essential excavation technique in their construction. However, the blasting process generates substantial amounts of dust, prolonging the construction period. When workers are exposed to dust-laden air with high particulate concentrations over long periods, they face a markedly elevated risk of developing pneumoconiosis and other occupationally related respiratory diseases [1,2,3]. Baur et al. [4] conducted a comprehensive analysis on a worker with long-term inorganic dust exposure by means of occupational inquiry, medical imaging examination, histopathological study, and single-particle SEM-EDS detection. It concluded that the patient had mixed-dust pneumoconiosis, with abundant silicon, silicon carbide, and aluminum particles detected in lung tissue, providing evidence for the differential diagnosis of occupational lung diseases.

Scholars globally have conducted extensive research on key aspects such as airflow field characteristics in underground engineering, dust evolution mechanisms, and ventilation scheme optimization [5,6]. Chen et al. explored the impact of varying canyon winds on internal tunnel flow characteristics via numerical simulation and concluded that hydraulic diameter is the primary factor affecting airflow behavior [7]. Meanwhile, studies on dust have also advanced: Guo et al. revealed the spatiotemporal dynamic process of dust diffusion under forced ventilation and determined the best ventilation rate to achieve efficient dust removal [8].

In terms of dust removal technologies, to address dust pollution in long-distance single-heading tunnels, Jiang et al. used numerical modeling to study a ventilation configuration featuring a long forcing duct and a short exhausting duct. They found that the best dust control performance is achieved when the ratio of forced air volume to exhaust air volume is 0.72, the forcing duct outlet is placed 20 m from the tunnel face, and the exhaust duct inlet is located 12 m from the face [9]. Xie et al. [10,11,12], through Fluent-based numerical simulations and multiple field experiments, identified the optimal installation position for forced ventilation duct outlets and corresponding airflow rates. Chen et al. [13] employed an orthogonal experimental design to optimize the ventilation parameters of a forced exhaust air curtain system for dust control. Nie et al. [14] employed CFD simulations within a long-pressure short-exhaust ventilation configuration to identify the ideal positioning of the exhaust duct relative to the working face for achieving peak dust collection performance. To achieve precise control of dust concentration and prevent its spread to surrounding areas, Kokkonen et al. implemented a local exhaust ventilation approach [15]. Liu et al. [16] determined the optimal incident angle (30°) and pressure (8 MPa) of a spray device using CFD simulation and experiments; field application showed a dust reduction efficiency of over 70%, which is 30% higher than the original measure. Liu et al. [17] performed laboratory model tests in conjunction with CFD simulations to identify the most suitable nozzle configuration for a U-shaped dust suppression system installed on a boom-type roadheader. The results identified an optimal configuration consisting of 11 solid-cone nozzles, each 1.6 mm in diameter, operated at an 8 MPa spray pressure. Field measurements indicated that the total dust suppression efficiency reached up to 81.48%, while the efficiency for respirable dust was as high as 81.51%, and efficiencies exceeding 70.41% at other monitoring points. Liu et al. [18] conducted field measurements and numerical simulations, and their results demonstrated that the dust concentration within the lower region was 65.88% greater in comparison with the upper region, with respirable dust of 313.46 mg/m³. The optimal duct-to-face distance is 12 m for the upper partition and 15 m for the lower partition, and the optimal forced airflow is 9.17 m³/s. Yang et al. [19] constructed a full-scale CFD model by combining field measurements, the Poly Hexcore approach, and the sliding mesh technology. Comparing three ventilation systems, their results showed that the combined system incorporating a ventilation duct, a shaft, and an axial fan achieved the highest performance among the three configurations examined. Compared with the duct-only system, wind speed increases by 7.5~30.6%, cooling rate increases by 14.1~17.7%, and CO volume fraction decreases by 26.9~73.9%. Yan et al. [20] synthesized a novel dust suppressant NaAlg-g-OJ through graft copolymerization modification of sodium alginate. Using infrared spectroscopy, thermogravimetric analysis, scanning electron microscopy, spray tests combined with computational fluid dynamics simulations, they demonstrated that the suppressant exhibits good wettability, adhesion, high coagulability, excellent thermal endurance and strong moisture retention, with excellent dust reduction performance.

In summary, relatively few studies have focused on ventilation in large-section tunnels (with a cross-sectional area of 80 m² or more). Furthermore, there is a lack of in-depth research on effectively removing pollutants generated by drilling and blasting in such tunnels while simultaneously meeting construction requirements and minimizing costs. This study takes the Qingdao Jiaozhou Bay Second Subsea Tunnel as the engineering background. Using a combination of computational simulations and field observations, and drawing upon the gas–solid two-phase flow framework, the spatiotemporal evolution characteristics of the coupled airflow-dust field under the original ventilation conditions were investigated. A two-factor orthogonal experimental design, considering outlet air velocity and the distance between the ventilation duct and the tunnel face, was adopted to optimize the forced ventilation parameters, with economic factors also taken into account. The combination of ventilation parameters that results in the lowest cost is identified as the optimal solution. The results are anticipated to offer theoretical guidance for managing dust pollution generated during large-section tunnel construction using the drilling and blasting method.

2. Materials and Methods

2.1. Project Overview

The subject of this study is the Qingdao Jiaozhou Bay Second Subsea Tunnel in China. The tunnel is 14.37 km in length and features a horseshoe-shaped cross-section with 14.51 m in bottom width, 8.4 m in height, and covering a cross-sectional area of 102 m². A semi-elliptical excavation invert, measuring 2 m in depth and 12 m in length, is located approximately 85 m from the tunnel face, resulting in a V-shaped longitudinal profile. The tunnel reaches a maximum depth of roughly 110 m beneath sea level. The surrounding rock features favorable integrity and considerable strength. The main surrounding rock types are Grade III and Grade IV, representing 64.6% and 29% of the overall length, respectively. Consequently, the drilling and blasting method serves as the primary excavation approach, with an excavation length of approximately 9.9 km. Figure 1 illustrates the project’s geographic setting together with the actual construction drawings from the site.

2.2. Selection of the Mathematical Model for Airflow and Dust Dispersion

This study employs Ansys Fluent 2022 R1 numerical simulation software to analyze gas–solid coupling behavior. In ventilation systems, air is treated as a continuous, incompressible fluid. It should be noted that the present numerical simulation focuses on the ventilation and dust diffusion processes occurring after blasting, instead of focusing on the immediate blast-induced shock wave generated by the detonation. The shock wave generated by the explosion lasts only a very brief period and exerts virtually no influence on the long-term airflow behavior that follows. Therefore, the assumption of incompressible air remains valid for the purposes of this study. Airflow primarily manifests in the forms of circular orifice jets, separated flows, impinging jets, and swirling flows [21], with the Eulerian and Lagrangian methods serving as the primary approaches for calculating fluid motion. The realizable k-ε model is selected to characterize the turbulent state of the airflow [22,23]. Given that dust particles exist in a discrete state within the ventilation system, they are regarded as a discrete phase, and the Discrete Phase Model based on the Lagrangian coordinate system is adopted for simulation [24,25,26]. Additionally, the simulation process is carried out under the following assumptions:

(1)

The heat exchange between on-site vehicles, equipment and the flow field is neglected.

(2)

The secondary dust generation during vehicle transportation after blasting is not considered.

(3)

The influence of small-scale equipment such as water pipes, wires and ventilation ducts on air flow and dust is ignored.

(4)

The influence of pressure variation between the tunnel interior and the external shaft atmosphere is disregarded, meaning that natural ventilation is excluded from consideration.

The governing equations for the airflow field, namely the Navier–Stokes equations, are presented below [27]:

Continuity Equation:

$${\displaystyle \frac{\partial \rho }{\partial t}}+\nabla \cdot (\rho \overrightarrow{v})=0$$

(1)

Momentum Equation:

$${\displaystyle \frac{\partial (\rho \overrightarrow{v})}{\partial t}}+\nabla \cdot (\rho \overrightarrow{v}\overrightarrow{v})=-\nabla p+\nabla \cdot (\overrightarrow{v})+\rho \overrightarrow{g}+\overrightarrow{F}$$

(2)

$$\overrightarrow{\tau }=\mu \left[(\nabla \overrightarrow{v}+\nabla {\overrightarrow{v}}^T)-{\displaystyle \frac{2}{3}}\nabla \cdot \overrightarrow{v}I\right]$$

(3)

where $\rho$ is the gas density; $t$ is time; $\overrightarrow{v}$ is the gas velocity vector; $p$ is the gas pressure; $\stackrel{̿}{\tau }$ is the viscous stress tensor; $\overrightarrow{g}$ is the gravitational acceleration vector; $\overrightarrow{F}$ is the external force vector; $\mu$ is the gas kinetic viscosity; and $I$ is the unit tensor.

The transport equations for turbulent kinetic energy k and its dissipati I have checked and revised the content accordingly. All variables and values in the equations in the original text have been set in italics, as indicated by the blue markings. I have checked and revised the content accordingly. All variables and values in the equations in the original text have been set in italics, as indicated by the blue markings. I have checked and revised the content accordingly. All variables and values in the equations in the original text have been set in italics, as indicated by the blue markings. on rate ε according to the Realizable k-ε model are given below: [27]:

$${\displaystyle \frac{\partial (\rho k)}{\partial t}}+{\displaystyle \frac{\partial (\rho k{u}_j)}{\partial x_j}}={\displaystyle \frac{\partial }{\partial x_j}}\left[\left(\mu +{\displaystyle \frac{{\mu }_t}{{\sigma }_k}}\right){\displaystyle \frac{\partial k}{\partial x_j}}\right]+G_k+G_b-\rho \epsilon -Y_M+S_k$$

(4)

$${\displaystyle \frac{\partial (\rho \epsilon )}{\partial t}}+{\displaystyle \frac{\partial (\rho \epsilon u_j)}{\partial x_j}}={\displaystyle \frac{\partial }{\partial x_j}}\left[\left(\mu +{\displaystyle \frac{{\mu }_t}{{\sigma }_{\epsilon }}}\right){\displaystyle \frac{\partial \epsilon }{\partial x_j}}\right]+\rho C_{1}S\epsilon -\rho C_2{\displaystyle \frac{{\epsilon }^2}{k+\sqrt{\epsilon }}}+C_{1\epsilon }{\displaystyle \frac{\epsilon }{k}}{C}_{3\epsilon }{G}_b+S_{\epsilon }$$

(5)

$${\mu }_t=\rho C_{\mu }{\displaystyle \frac{k^2}{\epsilon }}$$

(6)

$$C_1=\max \left[0.43,{\displaystyle \frac{\eta }{\eta +5}}\right]$$

(7)

$$\eta =S{\displaystyle \frac{k}{\epsilon }},S=\sqrt{2S_{ij}{S}_{ij}}$$

(8)

$$S_{ij}={\displaystyle \frac{1}{2}}\left({\displaystyle \frac{\partial u_i}{\partial x_j}}+{\displaystyle \frac{\partial u_j}{\partial x_i}}\right)$$

(9)

where $x_j$ is the spatial position; $k$ is the turbulent kinetic energy; $\epsilon$ is the turbulent dissipation rate; ${\mu }_t$ is the turbulent viscosity; $G_k$ denotes the turbulent kinetic energy generated by the mean velocity gradient; $G_b$ is the turbulent kinetic energy generated by the buoyancy force; $Y_M$ is the contribution of expansion fluctuations to the total dissipation rate in compressible turbulence; $C_1$, $C_2$ are constants; ${\sigma }_k$ is the turbulent Prandtl number for turbulent kinetic energy; ${\sigma }_{\epsilon }$ is the turbulent Prandtl number for turbulent dissipation rate; $S_k$ as well as $S_{\epsilon }$ denote the source terms; $S_{ij}$ is the strain rate tensor; and $C_{\mu }$ are constants, with $C_{1\epsilon }$ = 1.44, $C_2$ = 1.9, ${\sigma }_k$ = 1.0, and ${\sigma }_{\epsilon }$ = 1.2.

Particle Force balance equation [27]:

$$m_p{\displaystyle \frac{d{\overrightarrow{u}}_p}{dt}}=m_p{\displaystyle \frac{\overrightarrow{u}-{\overrightarrow{u}}_p}{{\tau }_{\tau }}}+m_p{\displaystyle \frac{\overrightarrow{g}({\rho }_p-\rho )}{{\rho }_p}}+\overrightarrow{F}$$

(10)

where $m_p$ is the mass of the particles; $\overrightarrow{u}$ is the fluid velocity; ${\overrightarrow{u}}_p$ is the particle velocity; $\rho$ is fluid density; ${\rho }_p$ is the particle density; ${\tau }_{\tau }$ is the particle relaxation time; and $\overrightarrow{F}$ is the external force acting on particles.

2.3. Construction of the Physical Model of Large-Section Tunnel

Based on site investigation data and construction drawings, a spatial three-dimensional geometric model of a tunnel section was developed at a 1:1 scale using Rhino 8.0. As shown in Figure 2, the model section is horseshoe-shaped, with dimensions of 180 m × 14.51 m × 8.4 m, and a cross-sectional area of about 102 m². Compared with conventional road tunnels, this tunnel is classified as a large-section construction tunnel. After tunnel blasting, forced ventilation was implemented, with the air duct outlet positioned 50 m from the tunnel face and a diameter of 1.8 m. The physical model also included other simplified equipment.

2.4. Mesh Division and Mesh Independence Test

The reliability of numerical simulations is highly dependent on mesh quality, necessitating a mesh independence verification. In this work, three grid schemes with different densities were adopted, including a coarse (612,956 elements), medium (1,379,056 elements), and fine (2,657,835 elements) were generated using the Meshing module. As depicted in Figure 3, the air velocity distributions at the tunnel breathing zone height (Z = 1.5 m, X = 7.5 m, 0 ≤ Y ≤ 180 m) were compared across the three schemes. The results demonstrate strong agreement between the medium and fine mesh simulations, whereas the coarse mesh exhibits notable discrepancies. This confirms that the medium mesh scheme satisfies the grid independence requirement. Comprehensively weighing calculation accuracy against resource consumption, the medium mesh scheme was adopted for all subsequent simulations and analyses in this study.

The model employs an unstructured mesh with tetrahedral elements. For the medium mesh scheme, a total of 13,790,56 elements were generated. The quality of the elements varied between a peak value of 1.0 and a minimum of 0.24, yielding a mean quality of 0.825. Notably, 99.51% of the elements exhibited a quality greater than 0.35. Since the grid has no negative value, the quality is good and meets the requirements of numerical simulation. Figure 4 shows the percentage of grid quality after meshing.

2.5. Dust Particle Size Characterization Experiments and Boundary Condition Configuration

Before performing the numerical simulation, the size distribution of the post-blasting dust must be measured through experimental means. During field sampling, a CCZ-20(A) dust sampler (Zhejiang Hengda Instrumentation Co., Ltd., Hangzhou, Zhejiang, China) was used to manually collect dust at the site, and an LS-909 laser particle size analyzer (Zhuhai Omec Instruments Co., Ltd., Zhuhai, Guangdong, China) was applied to determine the particle size distribution of the collected dust. Figure 5 illustrates the particle size distribution of the dust sample.

As illustrated in the preceding figure, the relationship between dust particle dimensions and their corresponding volumetric proportions follows the Rosin-Rammler distribution. The dust particle diameters vary between 0.87 μm and 163 μm, exhibiting a median value of 18.5 μm and a characteristic size of 22.1 μm. The volume fraction exhibits an increasing trend within the range of 0.87 μm to 20.32 μm, reaching a peak at 20.32 μm, followed by a continuous decrease thereafter.

Dust generated after tunnel blasting originates from the tunnel face, which is therefore designated as a dust injection source. The airflow enters the computational fluid domain from the outlet wind of the air duct, which is set as the velocity inlet. At the tunnel exit, the pressure exit condition is applied, and the no-slip condition and standard wall function are used at the wall to solve the turbulence near the wall. According to the field measurements by Zhang et al. [28], for each kilogram of explosive used, 0.0542 kg of dust is generated. In this research, the actual amount of explosive used per tunnel blast is approximately 305 kg, yielding a calculated total dust mass of 16.531 kg. To prevent simulation instability caused by an extremely short dust generation duration, the start and end times of dust injection are set to 0 s and 3.5 s, respectively, corresponding to a mass flow rate of 4.72 kg/s. The specific boundary condition settings are presented in

Table 1. Parameter settings for numerical simulation of tunnel drill-and-blast ventilation.

Project	Name	Parameter	Project	Name	Parameter
General	Gravity	−9.81 m/s2	Dust information	Minimum Diameter	8.7 × 10−7 m
	Solver	Pressure base		Maximum Diameter	1.63 × 10−4 m
Sticky model	K-epsilon	Realizable		Median Diameter	1.85 × 10−5 m
	Wall function	Standard wall		Total flow rate	4.72 kg/s
Boundary condition	Duct inlet	Velocity-inlet		Resistance Law	Spherical
	Tunnel Exit	Pressure-outlet		Particle size distribution	Rosin-Rammler
	Inlet velocity	10 m/s	Solution methods	Pressure velocity coupling	PISO
	Turbulence intensity	2.77%		Turbulent kinetic energy	Second order upwind
	Shear condition	Non-slip wall		Turbulent dissipation rate	Second order upwind
	DPM	Reflect
		Trap
		Escape

.

3. Simulation Results and Analysis Under the Original Ventilation Conditions

3.1. Airflow Field Distribution Characteristics

The tunnel construction employs a single-duct forced ventilation system. Under the driving force of the air current, the dust produced by blasting is passively carried back toward the tunnel outlet. To more clearly illustrate the evolution characteristics, the upper limits of air velocity are assigned as 6 m/s and 2 m/s, while the lower limit is fixed at 0 m/s; these values are indicated by red and blue in the colored columns. Figure 6 presents the streamline diagram of the air flow trajectory. Figure 7 presents the breathing zone and cross-section air velocity clouds at different distances from the palm face.

The airflow leaves the outlet of the ventilation duct at a speed of 10 m/s, and by the time it arrives at the tunnel face, its speed has dropped to 1.44 m/s. Upon impingement with the tunnel walls, the airflow reverses direction, with the velocity dropping to 0.66 m/s, forming a recirculation zone on the opposite side. Ventilation dead zones appear in localized areas on both sides of the tunnel face. Due to entrainment effects, a vortex zone forms between the jet and the recirculated flow, where the air speed is reduced compared to the adjacent areas. In the transition zone, the recirculated flow is divided by the excavation trolley, generating a fan-shaped recirculation pattern. Two small irregular vortex structures develop within 59–67 m and 67–74 m from the tunnel face. In the stable zone, the airflow is uniformly distributed; although the secondary lining trolley causes slight obstruction, the air current finally reaches a steady state with a velocity ranging from 0.20 to 0.23 m/s. Because of the tunnel’s large cross-sectional area, the mean air velocity is reduced when compared to tunnels with smaller cross-sections under identical forced airflow conditions, resulting in reduced pollutant removal efficiency.

3.2. Spatiotemporal Diffusion Characteristics of the Dust Field

In the simulation of the airflow field, dust particles were added as a discrete phase to enable a coupled analysis of gas–solid two-phase flow behavior. On account of computational limitations, a simulation duration of 600 s was selected to maintain a balance between accuracy and efficiency. Within this duration, the dust migration behavior under the original ventilation conditions was investigated. For clear representation of dust concentration levels, the upper limit of 300 mg/m³ is indicated in pink and the lower limit of 0 mg/m³ in blue within the color bars, respectively. Figure 8 illustrates the temporal evolution of dust dispersion throughout the tunnel, while Figure 9 presents the dust distribution cloud maps in the breathing zone (at a height of 1.6 m measured from the tunnel floor) at different time points.

As shown in Figure 8 and Figure 9:

(1)

A substantial amount of dust is generated immediately after blasting. Coarse particles settle rapidly by gravity, whereas fine particles remain suspended and couple with the airflow, forming a dust-laden flow. Driven by the airflow, high-concentration dust is primarily distributed on the side opposite the ventilation duct, with dust accumulation occurring in the vortex zone due to entrainment effects.

(2)

At t = 100 s, the dust has propagated as far as 75.79 m from the source. By t = 400 s, the average dust concentration close to the tunnel face drops from 445.13 mg/m³ to 49.50 mg/m³, while dense dust clouds remain present near the excavation trolley and air duct. At t = 440 s, the dust reaches the tunnel exit, and its further diffusion is hindered by the secondary lining trolley. At t = 600 s, the average dust concentration in the 1.6 m breathing zone is 36.8 mg/m³, which exceeds the safety limit of 2 mg/m³ stipulated in the Technical Specification for Construction of Highway Tunnels [29].

Figure 10 illustrates the variation in dust concentration within the breathing zone over time (1.6 m above the tunnel floor) at four time points after blasting. Overall, the dust concentration exhibits a pattern of “concentration, diffusion, and attenuation.” As time progresses, the peak concentration decreases from 963 mg/m³ to 177 mg/m³, and the corresponding peak location shifts progressively from 80 m to 140 m downstream. This indicates a clear pattern in which dust gradually moves toward the downstream section of the tunnel as time advances, and its concentration continuously decreases over time. Owing to the vortex zone, dust persistently builds up in the vicinity of the tunnel heading face, resulting in concentrations higher than those observed in the excavation trolley area.

The simulation results indicate that under the original ventilation conditions (outlet velocity: 10 m/s; duct-to-face distance: 50 m), the dust concentration in the tunnel remained high 600 s after blasting and still exceeded the permissible limit for construction. Therefore, to improve dust removal efficiency, it is essential to explore the optimal parameter combination of forced ventilation systems for large-section tunnels.

4. Optimization Study of Forced Ventilation Parameters for Large-Section Tunnels

The simulation results indicate that under the original ventilation conditions, blasting-induced pollutants in large-section tunnels exhibit high concentrations, while the baseline air velocity remains relatively low, resulting in dust concentrations that fail to meet construction requirements. Therefore, in this section, a two-factor orthogonal experimental design is adopted to investigate the effects of outlet air velocity and duct-to-face distance on dust removal efficiency, while satisfying the required airflow rate for the tunnel. Furthermore, once the construction requirements are met, economic factors are also considered to select the most cost-effective ventilation parameter combination.

4.1. Calculation of Required Air Volume for Large-Section Tunnels

Significant differences exist in the airflow requirements corresponding to different construction operation types and working conditions. In general, the maximum required air volume among all considered working conditions can satisfy the ventilation demand of the tunnel. Without considering duct leakage, the required air volume for the tunnel should be determined as the maximum of the following: the air volume required for personnel, the minimum air volume requirement inside the tunnel, and the air volume needed to dilute emissions from internal combustion engines [30].

(1) Air volume required for construction workers:

$$Q_1=k\cdot q\cdot n$$

(11)

where $Q_1$ is the required air volume for construction workers’ respiration; $k$ is the air volume reserve coefficient; q is the air supply rate per person at the working face; $n$ is the maximum number of personnel working simultaneously in the tunnel. ($k$ = 1.2, $q$ = 3 m³/(person·min), $n$ = 80)

According to the above formula, $Q_1$ = 4.8 m³/s.

(2) Air volume required for the allowable minimum ventilation velocity:

$$Q_2=60\cdot A\cdot v$$

(12)

where $Q_2$ is the allowable minimum ventilation velocity of the required air volume; $A$ is the tunnel excavation section area; $v$ is the hole’s allowable minimum ventilation velocity. ($A$ = 102.8 m², $v$ = 0.15 m/s)

According to the above formula, $Q_2$ = 15.42 m³/s.

(3) Total power of internal combustion machinery and equipment:

$$Q_3=q_1{\displaystyle \sum N_i}{T}_i$$

(13)

where $T_i$ is the utilization coefficient of diesel engine equipment operating simultaneously (0.65 for excavators, loaders, and slag transport vehicles, and 0.5 for tank trucks); $q_1$ is the air volume demanded by per unit power of the internal combustion engine ($q_1$ = 0.075 m³/s); $N_i$ is the power rating of a single internal combustion engine (183 kW for slag transport vehicles and 230 kW for tank trucks. The tunnel is considered to be equipped with 3 slag transport vehicles and 1 tank truck at the tunnel face.

According to the above formula, $Q_3$ = 35.38 m³/s.

(4) Air volume required for diluting blasting fumes:

$$Q_4={\displaystyle \frac{7.8}{t}}\sqrt[3]{G{(A{L}_0)}^2}$$

(14)

$$L_0=15+{\displaystyle \frac{G}{5}}$$

(15)

where $t$ is the ventilation time after blasting, taken as 15 min; $G$ is the quantity of explosive used for blasting, taken as 305 kg; $A$ is the tunnel excavation cross-sectional area (m²); $L_0$ is the diffusion length of blasting fumes, taken as 76 m.

According to the above formula, $Q_4$ = 22.95 m³/s.

Now the above calculation results take the maximum Q = max (Q₁, Q₂, Q₃, Q₄) as the control design air volume. The final required air volume is determined to be 35.38 m³/s, which corresponds to an air velocity of about 13.91 m/s at the duct outlet. This indicates that the original air outlet velocity fails to meet the required airflow rate for the tunnel.

4.2. Optimization of Ventilation Parameters Based on Orthogonal Experiments

In this study, factor A represents the air outlet velocity, while factor B denotes the distance separating the air duct from the tunnel face. Based on the required air volume formula, the air outlet velocity is calculated to be 13.92 m/s. Considering the strength limitation of the air duct, the fan cannot operate at its maximum power. Therefore, the air outlet velocity is set at 14, 16, 18 and 20 m/s for a total of 4 levels, and the spacing from the air duct outlet to the working face is assigned four distinct levels, namely 35 m, 40 m, 45 m, and 50 m.

Table 2. Table of factor levels.

Factor Level	A Air Velocity (m/s)	B Distance from the Tunnel Face (m)
1	14	35
2	16	40
3	18	45
4	20	50

presents the selected factors along with their corresponding levels used in this investigation.

To improve efficiency and ensure result reliability while minimizing the quantity of experimental runs, the orthogonal testing approach was adopted for the two-factor design with four levels each. The four levels of factor B were combined into two: 35 m and 40 m as level 1, and 45 m and 50 m as level 2. Based on the factor levels, the mixed-level orthogonal array L₈ (4 × 2) was selected, where “8” denotes the number of experiments, “4” represents one four-level column, and “2” represents one two-level column.

Table 3. Orthogonal experiment schedule.

Factor Experiment Number	A	B
1	A1	B1
2	A1	B4
3	A2	B2
4	A2	B3
5	A3	B2
6	A3	B4
7	A4	B1
8	A4	B3

provides a summary of the chosen factors and their corresponding levels, with each row corresponding to an experimental scheme. Take Experiment 1 as an example, where the air outlet velocity is set to 14 m/s and the duct is positioned 35 m away from the tunnel face. To ensure experimental balance, the four levels of factor B appear with equal frequency. The complete orthogonal experimental design is presented in Table 3.

4.3. Visualization Analysis of Orthogonal Experiment Results

The computational simulations were carried out in accordance with the mesh arrangement detailed in Section 2.4 and the orthogonal experimental design specified in Section 4.2. By comparing the spatiotemporal evolution characteristics of the coupled airflow-dust field and the average dust concentration in the tunnel at 600 s after blasting across the eight experimental cases, the experimental groups that met the construction dust concentration standards were identified.

Figure 11 illustrates the airflow field distribution across various experimental schemes. The airflow field distribution (including the vortex zone, transition zone and stable zone) of the 8 orthogonal experiments is almost the same, but there are certain differences in the distribution range of each group. With a reduction in the separation distance between the air duct and the tunnel face, the range of the vortex zone shrinks, the stable zone expands accordingly, and the morphology of irregular vortices inside the transition zone also changes. Under the condition that the spacing between the air duct and the tunnel face is held unchanged, a higher air outlet velocity leads to a greater return air velocity. Similarly, a higher air outlet velocity leads to an increased airflow speed within the stable zone.

Figure 12 reveals the dust distribution corresponding to various experimental scenarios. According to the orthogonal test findings, the mean dust concentration within the tunnel exhibits a notable reduction as the air outlet velocity rises and the distance from the air duct outlet to the tunnel face shortens. In Experiments 1–4, the dust concentration in the rear section of the tunnel remained relatively high, primarily concentrated near the secondary lining trolley. In Experiments 5, 7, and 8, the average dust concentration fell below the safety limit of 2 mg/m³. Although Experiment 7 exhibited the best dust removal performance, with a concentration of 0.8 mg/m^3, further increasing the outlet velocity—while satisfying the required airflow rate and safety concentration (below 2 mg/m³)—would lead to higher fan power consumption and, consequently, unnecessary energy and construction costs.

4.4. Range Analysis of Orthogonal Experimental Results

The preceding analysis indicates that ventilation parameters greatly affect dust removal performance, and it is crucial to further investigate the influence of each ventilation parameter on dust suppression. Accordingly, a range analysis is performed in this section on the basis of the orthogonal test results. The numerical simulation results (i.e., the experimental indicators) of each experiment are shown in

Table 4. Orthogonal experiment results.

Factor Experiment	A	B	Evaluation Index
			Average Dust Concentration (mg/m3)
1	1	1	12.5
2	1	4	14.7
3	2	2	5.7
4	2	3	6.9
5	3	2	1.5
6	3	4	2.7
7	4	1	0.8
8	4	3	1.2

.

In this paper, the span analysis method is abbreviated as the “R-method”. The span value is defined as the difference obtained by subtracting the minimum average from the maximum average for each factor level [31]. The range values enable the identification of both the most favorable level for each factor and the relative importance of that factor with respect to the experimental outcomes. A higher R-value signifies that the corresponding factor exerts a stronger effect on the measured indicator.

Table 5. Experimental index range.

Parameter	Factors of Index
	A	B
K1	13.60	6.65
K2	6.30	3.60
K3	2.10	4.05
K4	1.00	8.70
R	12.6	5.10

presents the average values of the experimental index (i.e., dust concentration).

In the table: $K_i={\displaystyle \frac{y_{i1}+y_{i2}}{2}},R=\mathrm{Max}{K}_i-\mathrm{Min}{K}_i$, where i = 1, 2, 3, 4, j = 1,2, and $y_{ij}$ denotes the experimental result of the $j$-th experiment at the $i$-th level.

Factor A exerts a stronger effect on the variation in the experimental index compared to Factor B. Therefore, the significance order of ventilation parameters on dust removal efficiency under forced ventilation is as follows: outlet air velocity > separation distance from the air duct to the tunnel face. Figure 13 presents the factor index diagram for the experimental index. The horizontal axis denotes the factor levels, and the vertical axis signifies the mean value of the experimental index.

(1)

When the outlet air velocity (Factor A) rises from 14 m/s to 20 m/s, the average dust concentration in the tunnel exhibits a declining tendency. At 20 m/s, the dust removal efficiency is optimal. However, the marginal benefit of further increasing the air velocity diminishes, leading to a gradual decline in the incremental effectiveness of dust removal.

(2)

When the spacing between the air duct and the tunnel face (Factor B) is raised from 35 m to 40 m, the average dust concentration exhibits a decreasing trend. When the distance increases from 40 m to 50 m, the average dust concentration shows an increasing trend. Therefore, the dust removal efficiency is optimal when the duct-to-face distance is 40 m.

4.5. Economic Analysis of Construction

By adjusting the fan power, different airflow conditions can be achieved, albeit at the expense of higher construction costs. To minimize costs while meeting tunnel construction requirements, this section investigates the hourly electricity cost of the fan under various outlet air velocities. The tunnel is ventilated by a TV-18 axial flow fan with a rated motor power of 2 × 315 kW, and the industrial electricity price is assumed to be 0.11 USD/kw·h. The correlation between the airflow rate and the fan power consumption is obtained from the characteristic performance curve of the fan.

Table 6. Hourly electricity costs under different air velocity conditions.

Air Velocity (m/s)	Air Volume (m3/s)	Power (kW)	Hourly Electricity Costs (USD)
14	35.61	84.5	9.32
16	40.70	120.9	13.34
18	45.78	161.6	17.83
20	50.87	201.4	22.22

presents the statistical results of hourly electricity costs under different air velocity operating conditions.

According to the outcomes of the numerical simulations, the dust concentrations in Experiments 5, 7, and 8 all fell below the safety limit. Experiment 7 achieved the best dust removal performance; however, the hourly electricity cost of the fan when the outlet velocity is 20 m/s was USD 4.39 higher than that at 18 m/s. Considering economic factors, Scheme 5—which satisfies the tunnel construction requirements at a lower cost—was selected as the optimal ventilation parameter combination for forced ventilation in large-section tunnels. This scheme corresponds to an outlet air velocity of 18 m/s and a duct-to-face distance of 40 m, which represents a 95.9% enhancement in dust removal performance relative to the initial ventilation setup.

4.6. Field Measurements and Error Analysis

For the purpose of verifying the reliability of the computational simulations, field tests were conducted using the optimized ventilation parameters. Dust concentrations at each measuring point were measured 200 s after blasting and compared with the corresponding simulation results. The measurement instruments included a CCZ-20(A) dust sampler (Zhejiang Hengda Instrumentation Co., Ltd., Hangzhou, Zhejiang, China) and a TSI-9535-A handheld anemometer (TSI Incorporated, Shoreview, MN, USA). As illustrated in Figure 14, the measuring points were selected based on an (X, Y, Z) coordinate system. Along the Y-direction (longitudinal), eight measurement cross-sections were set up at locations 20 m, 30 m, 40 m, 60 m, 80 m, 100 m, 120 m, and 140 m away from the tunnel face. The Z-coordinate was fixed at 1.6 m to represent the breathing zone. The X-direction (transverse) distance from the tunnel wall (opposite the air duct side) to the measuring points was set to 7.5 m. At each measurement location, the recorded value represents the arithmetic mean obtained from three replicate readings. The comparison between the numerical simulation outcomes and the field measurement data is shown in Figure 15 and Figure 16, where the calculated relative errors are also provided.

At positions A and B, the relative errors for air velocity fall within 2.5% and 13.2%, respectively, while those for dust concentration are within 2.4% and 11.6%. Overall, all errors fall within acceptable limits, thereby validating the accuracy of the numerical simulation results.

5. Conclusions

This study adopts a combined approach of numerical simulation and field monitoring to examine the spatiotemporal distribution and dispersion features of blast-induced airflow and dust within a large cross-section tunnel under the original forced ventilation scheme. On this basis, a two-factor orthogonal experimental method, incorporating economic factors, was adopted to determine the optimal ventilation parameter combination. The primary findings can be summarized as follows:

(1)

The airflow field can be categorized into three distinct zones: the disordered zone, transition zone, and stable zone. The core air velocity within the vortex zone remains comparatively low. As a result of entrainment, a portion of the recirculated flow is drawn into the high-speed jet, forming vortex structures, while the remaining recirculated flow propagates toward the tunnel rear in a fan-shaped pattern. Because of the tunnel’s substantial cross-sectional area, the mean air velocity remains low, dropping to roughly 0.2 m/s before stabilizing. The transport of dust within the tunnel is primarily governed by advective airflow.

(2)

Dust transport within the tunnel is primarily governed by airflow advection. Under the original ventilation conditions, coarse particles settle rapidly due to gravity, whereas fine particles are conveyed by the airflow. The dust concentration in the vortex zone is significantly higher than that in the lateral regions. At t = 440 s, dust has diffused to the tunnel exit. By t = 600 s, the average dust concentration in the breathing zone (1.6 m above ground) reaches 36.8 mg/m³, exceeding the construction safety limit.

(3)

A two-factor orthogonal experimental design was adopted to identify the ventilation parameter combination that minimizes construction costs while satisfying the tunnel air demand and reducing dust concentration to within the safety limit. The results indicate that the significance of the key parameters affecting dust removal efficiency is ranked as follows: outlet air velocity > duct-to-face distance. Based on the experimental results and economic considerations, the optimal ventilation parameters for the forced ventilation system were determined as an outlet velocity of 18 m/s and a duct-to-face distance of 40 m. Under this configuration, the average dust concentration at 600 s after blasting is 1.5 mg/m³, corresponding to a 95.9% improvement in dust removal efficiency compared to the original conditions. Furthermore, the hourly electricity cost at 18 m/s is USD 4.39 lower than that at 20 m/s, thereby enhancing worker occupational safety while reducing project costs. Both numerical simulations and field measurements confirm the reliability and practical applicability of the proposed method.