Introduction

Accelerated urbanization and expanding infrastructure development have elevated the strategic importance of mountain tunnel engineering across critical domains including transportation networks, water resource allocation, and energy transmission systems. However, tunnel excavation within densely populated areas with existing structures presents significant challenges for conventional blasting techniques. Persistent issues such as structural vibration impacts and compounded operational risks necessitate effective solutions.Novel fracturing technology based on the phase transition principle of liquid CO₂ offers an innovative approach for rock excavation adjacent to existing buildings, leveraging its controllable energy release mechanism. This technique exhibits primary advantages including substantially reduced vibration, effective noise control, enhanced operational safety, and improved excavation precision1,2.

CO₂ phase change fracturing technology achieves effective fragmentation while significantly mitigating environmental disturbances through its directed pressure wave propagation mechanism3. Current theoretical advancements in tunnel blasting predominantly focus on energy transfer models, parameter matching algorithms, and quantitative effect evaluation systems. In mountain highway tunnel excavation, gradient optimization of fracturing parameters has enabled the compression of the blast disturbance radius to within 15 m, significantly reducing strata displacement risks4.Regarding efficient coalbed methane (CBM) extraction, field applications of liquid CO₂ phase change fracturing technology demonstrate its capacity to substantially enhance gas drainage efficiency. Furthermore, its inherent spark-free nature ensures heightened operational safety during implementation5,6,7. CO₂ phase change fracturing technology is also widely utilized for ice-breaking emergency operations, where energy released during the phase change rapidly clears ice-blocked sea or river channels8,9,10. Abi et al.11 demonstrated that the peak pressure generated during CO2 phase change fracturing is approximately 1/3.36 of that produced by conventional explosive blasting, with peak particle velocity (PPV) ranging between 1/74 and 1/78 of equivalent explosive charges. Sun12 recorded a maximum PPV of 0.376 cm/s during field applications, confirming the technique’s suitability for cutting and rock-breaking operations and providing a valuable reference for engineering practices such as rock tunnel development and urban underground excavation. Through numerical simulations, Wang et al.13,14 systematically optimized the construction parameters for CO₂ phase change fracturing. Asheghi et al.15,16 employed ultrasonic testing to assess crack initiation and propagation in tunnel rock masses, complemented by 3D finite element modeling using ABAQUS 2022 software for comprehensive simulation analysis.

Within urban renewal projects, this technology provides low-disturbance solutions for the safe demolition of existing structures, effectively balancing engineering efficiency with environmental protection17. Practical applications on operational railway lines confirm the technique’s ability to reduce blast-induced vibration amplitudes affecting track facilities by up to 60%, concurrently enabling effective noise level management18. For subgrade modification, it demonstrates a level of precision unattainable with conventional blasting. By establishing an energy gradient release model, it successfully mitigates chain reactions triggered by abrupt stress changes in surrounding rock within complex strata, reducing blast positioning errors to within ± 15 cm19. A case study from a shaft construction project on Shenzhen Metro Line 12 illustrates its application in densely built areas. Utilizing precision shallow drilling (depth < 15 m) and spacing gradient optimization to ensure balanced energy propagation, the project team successfully maintained peak ground vibration velocities below 2.5 cm/s20. Addressing the confined workspaces of deep urban metro excavations, the technology facilitates green construction systems via vibration suppression, confining underground pipeline displacement within a safe threshold of 3 mm21. The technology exhibits remarkable versatility, extending beyond conventional rock fragmentation to demonstrate distinctive advantages under challenging geological conditions.

As mentioned earlier, Existing research primarily focuses on qualitative descriptions of the application scenarios of CO2 fracturing blasting. However, there is a notable lack of in-depth investigation into the calculation, control, and safety verification of fracturing energy under complex environmental conditions. Due to the absence of a reliable verification model, the blasting design process often relies heavily on empirical trial-and-error, which is neither efficient nor reliable for ensuring safety. This study investigates the phase change pressurization mechanism and operational characteristics of CO2 fracturing blasting. By employing unmanned aerial vehicle (UAV) for safety inspections, the fracturing energy of a 280 MPa MZL-51 CO2 fracturer was calculated. A comprehensive safety verification methodology for fracturing energy is proposed. Subsequent vibration monitoring of structures near the blast source confirmed that the PPV at all measured points remained within safe limits. The findings of this research provide a novel quantitative calculation method for CO2 phase change fracturing blasting based on the protection of target structures.

Engineering applications of CO₂ fracturing technology

Technical principle and characteristics

The fundamental mechanism of CO₂ fracturing technology is derived from the energy release effect associated with the phase change of a liquid substance22. When liquid CO₂ undergoes rapid vaporization within a controlled environment, its volumetric expansion generates ultra-high pressure exerted upon the rock mass structure, The principle of CO2 phase change explosion is shown in Fig. 1. During operation, liquid CO₂, injected into pre-drilled boreholes, experiences rapid heating, triggering its phase change. This process instantaneously generates a pressure field on the order of hundreds of megapascals (MPa). This sudden surge in fluid pressure creates a stress gradient within the rock strata, promoting directional fracturing of the rock along predetermined failure planes23. Compared to conventional blasting methods, this technology enables precise management of energy output through controlled regulation of the gas release rate. This approach effectively maintains the requisite fragmentation intensity while concurrently significantly attenuating the impact of vibration waves and shock energy on the surrounding environment24,25.

Fig. 1
Fig. 1
Full size image

CO2 phase change explosion principle.

Construction environment analysis

This study focuses on the Yanmenzi Reservoir project in Tongzi County. The originally designed excavation scheme for the water intake and drainage tunnel employed conventional explosive blasting(Fig. 2). However, approximately 38 residential dwellings with earth-timber structures are situated directly above the tunnel alignment (Fig. 3), with the closest structure located merely 30 m from the excavation zone. Critically, the majority of these dwellings are structurally compromised. Proceeding with the initially planned explosive blasting posed a significant risk of triggering collective incidents, including work stoppages and petitions by residents. Following comprehensive on-site investigation and technical evaluation by all project stakeholders, the adoption of CO₂ fracturing technology for tunnel excavation was determined to be the viable alternative. This method is currently under implementation for the tunnel construction.

The tunnel portals traverse moderately to highly weathered rock masses, primarily comprising silty mudstone and siltstone, characterized by well-developed steeply dipping fissures, intense surface weathering, poor rock mass integrity, and highly unstable surrounding rock, classified as Class V according to relevant standards. The tunnel body section consists of silty mudstone and siltstone interbedded with feldspathic quartz sandstone, exhibiting sparsely developed fractures. Given the stringent low-vibration requirements, the implementation of CO₂ fracturing technology enabled effective vibration control. Peak PPV at building monitoring points were consistently maintained below the target threshold of 1.0 cm/s (for vibration impact assessment, refer to Table 1). This performance represents a 50% reduction compared to the conventional explosive blasting benchmark of 2.0 cm/s, thereby achieving the safety objective for the earth-timber structures.

Table 1 Medvedev seismic intensity scale.
Full size table
Fig. 2
Fig. 2
Full size image

Tunnel excavation opening.

Fig. 3
Fig. 3
Full size image

Distribution of houses around the tunnel.

Technical implementation procedures

In mountain tunnel construction adjacent to existing structures, the application of CO₂ fracturing technology necessitates meticulous planning tailored to specific site conditions to ensure simultaneous achievement of engineering safety and operational efficiency. Geological exploration data constitute the fundamental basis for blast design, mandating the concurrent determination of borehole layout patterns and initiation parameters.

Optimization of borehole layout parameters

To ensure borehole quality, the hole spacing was optimized based on in-situ rock properties, rock mass fracturing degree, and stability classification of the surrounding rock (Table 2). Survey engineers precisely deployed borehole positions using total stations according to design drawings, achieving deviations within ± 30 mm for spacing and ± 50 mm for depth. Cut holes were concentrated in the central zone of the excavation profile, with 6 to 9 holes utilized for concentrated expansion fracturing (black dots in Fig. 4). Perimeter holes were detonated in three successive layers from the interior outward to progressively expand the contour and adjust the tunnel profile (white dots in Fig. 4), with the actual field configuration shown in Fig. 5. Borehole layout design requires precise calculation of depth and row spacing parameters integrated with rock mass characteristics to prevent blasting failures or safety risks arising from parametric deviations26,27.

Table 2 Optimized pore parameters.
Full size table
Fig. 4
Fig. 4
Full size image

Design drawing of tunnel blasting section.

Fig. 5
Fig. 5
Full size image

Site construction drawings.

Construction process

Firing circuit network connection represents a critical operational stage. When initiating a large number of fracturing holes simultaneously, zonal circuit partitioning is essential to minimize total network resistance. During partitioning, resistance balance across parallel branches must be ensured to guarantee uniform current delivery to each fracturing device.For wired fracturing device verification, set the ohmmeter to the 200Ω range and measure resistance between the initiation terminal and tube casing. A reading of approximately 2Ωconfirms functional integrity; devices deviating from this specification must be rejected.

During implementation, establish standardized safety operation procedures, clarify the technical indicators and operation standards for each process, and ensure the stability of the fracturing device operation. Implement a regional responsibility system, clarify job responsibilities, including personnel division, hole allocation, and time arrangement. Operators need to have a deep understanding of key operations such as blasthole positioning, depth, angle, detonator segmentation, and network connection to prevent mutual interference between operations. The specific operation steps are as follows (Fig. 6):

Fig. 6
Fig. 6
Full size image

Schematic diagram of CO₂ fracturing technology.

UAV deployment for environmental reconnaissance

Risk identification is defined as the systematic process of recognizing potential safety hazards that may lead to casualties, environmental damage, economic losses, or project delays28. In this project, we implemented UAV-based aerial reconnaissance for risk identification (Figs. 7 and 8), significantly enhancing the safety and reliability of CO₂ fracturing technology during mountain tunnel construction adjacent to existing structures.

Fig. 7
Fig. 7
Full size image

UAV takes off in an open environment.

Fig. 8
Fig. 8
Full size image

Safety inspection.

The UAV operations strictly adhered to predefined flight parameters: altitude maintained at 30–40 m above ground level (adjustable for surface obstacles such as power lines and trees) to ensure optimal ground identification; speed controlled at 3–5 m/s to achieve high-resolution image capture and stable video transmission; and flight path following a “lawnmower” pattern covering a 100 m radius from the blast source to ensure 100% coverage of both the blast influence zone and personnel evacuation routes. The corresponding risk checklist is provided in Table 3.

Table 3 Risk register.
Full size table

All personnel implemented remote initiation procedures while strictly maintaining the minimum standoff distance of 100 m. Prior to CO₂fracturing operations, all personnel were positioned behind designated safety barriers at this distance, conducting remote monitoring and initiation commands through the UAV’s first-person view (FPV) video transmission system. The decision-making workflow for initiation authorization is presented in the accompanying flowchart(Figure 9).

Fig. 9
Fig. 9
Full size image

Decision flowchart.

CO₂ fracturing performance verification

CO₂ fracturing achieves high-efficiency rock fragmentation with minimal vibration and shockwaves, exhibiting controlled ejection distances under 5 m. This represents a 90% reduction compared to traditional explosive blasting exceeding 50 m, significantly mitigating construction hazards. Post-fracturing debris primarily comprises uniformly graded fragments (Fig. 10), with 80% of particles within the 10ཞ50cm size range. This optimized gradation eliminates secondary crushing requirements and facilitates muck removal. Post-fragmentation, material is excavated for transport to designated recycling sites, discharged via coordinated scraper conveyor operations (Fig. 11), and processed into engineered-grade aggregates at crushing plants.

Fig. 10
Fig. 10
Full size image

Crushed stone after explosion.

Fig. 11
Fig. 11
Full size image

Material discharge of the crusher.

Verification and analytical methodology for CO₂ fracturing

CO₂ fracturing energy verification methodology

Risk mitigation and quality control constitute critical elements in the implementation of CO₂ fracturing technology during construction. The energy released during the fracturing process (pressure vessel rupture) primarily depends on the gas pressure, vessel volume, and the phase-state properties of the medium within the vessel. This study employed MZL-51 liquid CO₂ reservoirs with a volume of 1.5 L. According to existing experimental results, the SD390 bursting discs have a rupture pressure of 280 MPa(high-pressure CO2 gas). Previous studies29,30 have established the energy calculation formula (1) for liquid CO₂ fracturing, providing the theoretical basis for TNT equivalence conversion.

$$\:{\:E}_{g}=\frac{PV}{K-1}\left[\text{1-}{\left(\frac{\text{0.1013}}{\text{P}}\right)}^{\frac{\text{K}\text{-}\text{1}}{\text{K}}}\right]\times\:{10}^{3}$$
(1)

where Eg is the energy generated by CO₂ fracturing, kJ. P is the absolute gas pressure within the vessel, MPa. V is the vessel volume, m³. K is the adiabatic index of the gas, and the adiabatic index of CO₂ is 1.295.

$$\:{{W}}_{{TNT}}{=}\frac{{{E}}_{{g}}}{{{Q}}_{{TNT}}}$$
(2)

where WTNT is the equivalent TNT mass for CO₂ phase change fracturing device, kJ. QTNT is the explosive energy of 1 kg TNT, and the 1 kg TNT is 4250 kJ/kg.

Formula (1) yields an energy release of 1193.37 kJ from the MZL-51 reservoir. Using Eq. (2), this corresponds to approximately 281 g TNT equivalent. With reference to the aforementioned vibration velocity limit and actual site conditions, a safety threshold of V ≤ 1.0 cm/s was established for the timber-structure house at 30 m (requiring no structural damage). The geological survey report for the tunnel alignment indicates that the surrounding rock primarily comprises competent sandy mudstone and siltstone interbedded with feldspathic quartz sandstone with uniaxial compressive strength ranging between 80 and 90 MPa, classifying as hard rock. For such lithology, the site coefficient Kv typically ranges 50–150 and the attenuation coefficient α generally varies 1.3–1.5. Following recommendations from the “Blasting Safety Regulations” and incorporating findings from comparable studies31,32,33, we adopted Kv=100 and α = 1.4. The maximum permissible TNT equivalence for CO₂ fracturing was subsequently back-calculated using Sadovsky’s Eq. (3):

$$Q_{MAX} = R^{3} \times \left(\frac{v}{K_v}\right)^{3/\alpha} = 30^3 \times \left(\frac{1}{1000}\right)^{3/1.4} \approx 1.4kg$$
(3)
$$\:{{W}}_{{TNT}}{=281g}{{\leq Q}}_{{MAX}}{=1400g}$$
(4)

where v is the peak particle velocity, cm/s. Kv is the site coefficient, and the site coefficient of hard rock is 100. R is the blast center distance, m. α is the attenuation coefficient, and the attenuation coefficient of hard rock is 1.4.

Calculations confirm the CO₂ fracturing energy equivalence at 281 g TNT. Formula (4) mandates less than 1400 g TNT for adjacent structural safety, with CO₂ fracturing constituting only 20% of this permissible maximum. This minimal impact satisfies operational requirements at the construction site. The methodology enables pre-blast determination of both single-stage CO₂ charge yield and maximum allowable equivalence through known source-to-receptor distances and structure criticality assessments, establishing a validated safety computation framework. This precision energy-regulated system concurrently ensures construction safety, environmental compatibility, and process optimization, delivering an innovative computational paradigm for CO₂ fracturing engineering.

Sensitivity analysis

This section presents a sensitivity analysis to evaluate the influence of various parameters on the conversion of CO₂ fracturing energy to TNT equivalence34. The primary sources of uncertainty in the calculation were identified, including fluctuations in reservoir burst pressure, manufacturing tolerances in volume, and the selection of TNT explosive energy. The impact of variations in each parameter on the TNT equivalence was systematically analyzed (Table 4).

Table 4 Parameter sensitivity Analysis.
Full size table

The sensitivity coefficients (percentage change in TNT equivalence resulting from a 1% variation in each parameter) are summarized in Table 4. The sensitivity analysis reveals that pressure and volume exhibit the most substantial influence on the results, classifying them as high-sensitivity parameters. TNT explosive energy represents a moderate-sensitivity parameter, as employing different standard values (4250 kJ/kg versus 4500 kJ/kg) introduces approximately 5.4% variation in the calculated equivalence. The adiabatic index of the gas demonstrates the least influence, qualifying as a low-sensitivity parameter; this coefficient remains relatively stable during the CO₂ phase change process. Table 5 further analyzes the practical ranges of parameter fluctuations encountered in engineering applications and their underlying causes.

Table 5 Parameter fluctuation range.
Full size table

Estimation of total uncertainty \(\:{\text{U}}_{\text{y}}\:\):

$$U_{y} = \sqrt{{(1.00 \times 5)^2\: + \:(1.00 \times 2)^2\:+\:(0.47\times 1)^2\: + \:(1.00\times 3)^2}} =6.2\%$$

Consequently, pressure and volume, being the most influential parameters, require precise control and measurement. It is recommended to maintain pressure control accuracy within ± 2% and volume calibration accuracy within ± 1%. Engineering practice should explicitly specify the adopted TNT explosive energy value, while the actual adiabatic index used should be documented and reported. The total calculated uncertainty is approximately 6.2%. Given that the TNT equivalence of the CO₂ fracturing charge constitutes merely 20% of the maximum permissible limit, this level of uncertainty is acceptable for engineering applications. This confirms that the calculated result of 281 g remains reliable within reasonable parameter fluctuation ranges.

Quantitative impact assessment of Blast-Induced vibrations

China’s current safety criterion prioritizes PPV at critical protection targets per Blasting Safety Regulations35. This project specifically monitors two parameters at protected structures: foundation-level PPV and dominant vibration frequency. Pre-construction permissible charge calculations for surrounding buildings validated the proposed CO₂ fracturing energy methodology through vibration monitoring results, mandating PPV below 1.0 cm/s for structures adjacent to tunnel excavations.

Fig. 12
Fig. 12
Full size image

PPV at different monitoring points.

Vibration monitoring was conducted during the tunnel excavation process using a TC-4850 vibrometer with three-channel parallel acquisition and DC accuracy error < 0.5%. The nearest structure was located 30 m from the blast source, establishing this point as the monitoring origin with subsequent points at 5-meter intervals. Detailed measurement results are provided in Appendix A, with the PPV representing the maximum value among the three directional components.Measurement results identified the vertical (Z) component as the dominant direction of vibration effects. The recorded PPV values at different distances are presented in Fig. 12, with subsequent attenuation analysis performed on the dataset. All sensors were calibrated by the Chengdu Metrology and Verification Institute within their valid certification period to ensure measurement accuracy and reliability.

Monitoring data demonstrated progressively decreasing PPV with increasing distance from the blast source. At the closest monitoring location (30 m), the maximum recorded velocity in the Z-direction was Vmax = 0.18 cm/s, significantly below the safety threshold of V = 1.0 cm/s. Directional vibration spectra are presented in Fig. 13. All field-measured PPV values remained within safe limits, inducing negligible structural impact and imperceptible vibration levels. Post-excavation inspections confirmed absence of structural damage or cracking.Comparative analysis with conventional explosives reveals substantially superior safety performance of CO₂ fracturing. Literature35 reports recorded PPV of 2.61 cm/s at 30 m from explosive blasting, representing at least five times greater vibration intensity than CO₂ fracturing under comparable conditions. These findings validate that the proposed CO₂ fracturing energy methodology satisfies target safety requirements and demonstrates significantly enhanced safety for rock excavation near protected structures.

Fig. 13
Fig. 13
Full size image

Field-measured PPV at 30 m distance from fracturing source.

Conclusions

Based on the engineering case study of the Yanmenzi Reservoir mountain tunnel construction in Tongzi County utilizing CO₂fracturing technology, this study investigated the operational principles, construction methodologies, UAV implementation processes, and TNT equivalence conversion of CO₂ phase change fracturing. A comprehensive safety verification methodology for CO₂ fracturing energy was proposed and comparatively analyzed with field vibration monitoring data, yielding the following conclusions:

  1. (1)

    CO₂phase change fracturing technology demonstrated significant effectiveness in mountain tunnel construction adjacent to existing structures, providing an solution to challenges associated with conventional blasting. Engineering implementation requires adjustment of fracturing parameters according to geological conditions, integrated with UAV surveillance and vibration monitoring to establish a multi-dimensional safety protection system.

  1. (2)

    (TNT equivalence conversion for MZL-51 CO₂fracturing devices at 280 MPa yields 281 g TNT equivalent. The maximum permissible TNT equivalence for CO₂ fracturing energy was determined as 1400 g through back-calculation using Sadovsky’s equation.

  1. (3)

    Under CO₂phase change fracturing, structural vibration velocities exhibited exponential attenuation with increasing distance from the blast source, with the maximum recorded velocity measuring merely 0.18 cm/s.

  1. (4)

    Project validation confirmed that the proposed CO₂ energy calculation methodology successfully restricted vibration velocities below the 1.0 cm/s control threshold, with all field-measured PPV remaining within safe parameters. This enables safety control of structural impact on adjacent timber-framed buildings.

The significance of this research is twofold. From a theoretical perspective, it advances the understanding of CO2 fracturing by moving from qualitative description to quantitative analysis, establishing a vital link between theoretical energy calculations and practical engineering controls. From a practical standpoint, the successful application of our proposed framework in the Yanmenzi Reservoir tunnel project in Tongzi County demonstrates its efficacy. Our method ensured that the PPV affecting a nearby wooden structure house was meticulously controlled to a mere 0.18 cm/s, far below the safety limit of 1.0 cm/s. This provides a reliable, replicable, and innovative solution for safe construction in spatially constrained and environmentally sensitive areas.

However, certain limitations must be acknowledged: different lithologies (such as extremely hard rock or weak fractured zones) may affect energy transfer efficiency and rock-breaking mechanisms. Furthermore, the practical application encounters “tube ejection” phenomena, presenting safety concerns that require resolution through equipment refinement in future research.