The Chongqing–Kunming (YU–KUN) high-speed railway is one of the longest railway passages on the Beijing–Kunming high-speed railway. It connects east and west China, and once completed, it can facilitate interaction between north and southwest China, and is of great significance to economic growth and national defense construction. The YL tunnel of the YU–KUN high-speed railway is a key project of the whole line, and it is a typical ultra-long deep buried tunnel facing geologic challenges, such as rock bursts, large deformation, water inrush, and toxic or harmful gas leakage, among which rock bursts and large deformation induced by high-stress environments result in the most serious damage. Analyzing and finding solutions to high ground stress in tunnels of the same type is significant to the construction of such tunnels on high-speed railway lines, and even more so in the greater southwest region (Wang et al., 2020).

High ground stress causes two main problems: large deformation of soft rock and rock bursts. Researchers have already investigated the generation mechanism, operating principles, and prediction and prevention measures for such problems (Liu, 2020). The large deformation of soft rock is generally divided into two categories according to the formation mechanism (Aydan et al., 1993): plastic deformation caused by the stress redistribution from tunnel excavation exceeding the strength of the surrounding rock and expansion caused by the chemical reaction between some minerals in the surrounding rock and water. Anagnostou (1993) found that the main factors causing large deformation of soft rock are in situ stress and rock strength. Tan and Wen (1983) attributed the large deformation of the surrounding rock to five aspects: the plastic wedge, expansion of the surrounding rock, rheology, expansion, and deflection. Brox and Hagedorn (1999) point out that stress redistribution caused by tunnel excavation is an important factor that causes large deformation of the surrounding rock in a weak geological environment. He et al. (2007) proposed a failure mechanism of soft rock tunnel deformation according to the rock mass structural plane, force source, and molecular structure of the rock mass itself. Kovačević et al. (2021) used particle swarm optimization to evaluate the long-term deformation of soft rock tunnels on the basis that particle swarm optimization has fast convergence and strong searching capability. Research, to a certain extent, makes the large deformation of soft rock more predictable. Zivaljevic and Tomanovic (2021) carried out long-term creep tests of cyclic loading and gradual unloading on soft rock specimens, making clear the influence of historical loading and unloading on soft rock creep based half on theory and half on empiricism. Zhao et al. (2011) inspected the supporting system of the weak surrounding rock of the Tianpingshan Mountain deep-buried tunnel and analyzed its stress characteristics. Li (2011) analyzed the characteristics of the damaged surrounding rock and support structures in the Maoyushan Mountain Tunnel and proposed that the primary support be closed quickly, and to construct the secondary lining when the time was right, concluding that reserving space for possible deformation and the release of stress in advance are key to tunnel construction in areas where there are chances of soft rock large deformation. Li et al. (2015) analyzed the large deformation of soft rock on the Lanzhou–Chongqing railway line and found a pattern of soft rock deformation in the tunnel, and provided suggestions on the classification standards of deformation of soft rock. Sun et al. (2020) used the Muzhailing highway super-long tunnel to field test different support schemes. These were compared and analyzed, and it was concluded that high prestressed constant resistance and large-deformation anchor cable support measures were more effective in controlling the large deformation of the surrounding rock.

Wang and Huang (1998) experimentally discussed the deformation and failure characteristics of rocks and their rock burst effects under different unloading rates. Xiao and Pan (2005) used rosin to simulate similar materials in experimental research and proposed a rock burst stiffness formula. Wang et al. (1998), Qian (2014), He et al. (2007), Wu et al. (2019), and Peng et al. (2014) carried out research on the prediction of rock bursts through quantitative calculations, numerical simulations, fuzzy mathematics, and acoustic emissions. Ahmad et al. (2021) used J48 and random forest algorithms to predict the risk of rock burst based on multiple engineering examples and made comparisons with other algorithms. They concluded that the random forest algorithm is more accurate than the empirical models (J48 and the artificial network method). Adoko et al. (2013) studied 174 rock bursts and predicted rock bursts with five different models based on the fuzzy inference system and adaptive neuro-fuzzy inference system, and concluded that the results of the adaptive neuro-fuzzy inference system model were more accurate. Gong et al. (2020) compared 20 rock burst proneness criteria with laboratory tests and concluded that using the residual elasticity index as the rock burst proneness criterion is more reasonable and accurate.

However, we suggest that there are still some shortcomings in existing theories that need to be further studied. Previous research focused on the occurrence mechanism of two types of geological disasters, but rarely considered the coupling effect with other disasters in the complex geological environment, which often increases the risk of disasters and deviates greatly from the prediction results in practical engineering. Therefore, combined with existing research, this study evaluated, analyzed, and predicted the high ground stress problem of the YL tunnel, put forward construction suggestions and prevention measures, and conducted a qualitative analysis of the amplification effect of various disasters in the aforementioned environment. Based on global research, this study investigated the difficulties resulting from high ground stress in the YL Tunnel and put forward suggestions for preventing rock bursts and other geohazards encountered during construction.

With social development and economic growth, ultra-long deep tunnels such as the YL Tunnel are being increasingly needed, but this also means that there are an increasing number of engineering geological problems (Zhu et al., 2019). The main problems caused by the high ground stress in the YL Tunnel included the large deformation of soft rock and rock bursts. The current and commonly used method to predict hazards caused by high ground stress is to measure the initial ground stress, and interpret the results comprehensively in accordance with hydraulic fracturing, the stress-relief method, and the stress-recovery method.

However, an increasing number of engineering accidents in construction show that obvious discrepancies exist between predictions and actual measurements. To uncover the reasons behind this, this study conducted an analysis from two aspects. First, when making predictions, disturbances in the surrounding rock were not considered. For example, in complex structural areas, the surrounding rock itself stores considerable elastic potential owing to the multistage geological tectonic stress field. When the tunnel is excavated at this point, the surrounding rock disturbance causes the deformation energy to be released to the empty surface, resulting in engineering hazards. Especially in soft rock areas where rocks are weak and cracks develop, any water adsorption causes expansion and therefore determines its poor anti-disturbance ability. The rock in this area is extremely sensitive to construction disturbances, such as blasting and excavation, and is prone to engineering problems beyond prediction. Second, the amplification effects caused by the coupling of geohazards were not considered because the engineering geological conditions of the tunnel site were not determined by a single condition, and the problems generated under multiple geological conditions were often more complicated. Such tunnels not only face unusually complex structural, high ground stress, and, active fracture issues, but also more complex engineering hazards caused by the coupling of multiple engineering geological issues.

The abovementioned geological conditions also make it difficult to make risk predictions in the tunnel, resulting in obvious discrepancies between predictions based on measurements as well as experiments and actual construction. The coupling effect between different hazards often has an amplification effect, leading to more serious geological engineering problems. As for the YL Tunnel, the possible amplified hazards may include coupling among large deformations, groundwater, and harmful gases caused by high ground stress.

The coupling amplification effect of different hazards is not discussed in this study but will be the focus of future research. This study only discusses similar problems that this type of tunnel may encounter.

Engineering geological problems in the YL Tunnel of the Chongqing–Kunming high-speed railway include karst, rock burst, large deformation of soft rock, bedding bias, high ground temperature, toxic and harmful gases, and rock piles. Among these, the tunnel damage caused by high ground stress is particularly prominent.

Based on field surveys, spot and indoor tests, engineering geology profiles, hydrogeological data, and ground stress data, the high-risk sections of the tunnel prone to large deformations of soft rock and rock bursts were predicted accordingly. The following areas were most prone to rock burst: D1K338 + 945–D1K339 + 025, D1K339 + 075–D1K339 + 220. The prevention and countermeasures were presented in this study. We present relevant prevention measures and construction suggestions. The tunnel body passes through the vertical seepage zone to the seasonal variation zone, with medium to strong karst development and high water inrush risks. In addition, there is a risk of gas diffusion when the tunnel body passes through coal-bearing strata.

The coupling of multiple undesirable geological processes often has an amplifying effect and is typically uncertain, irregular, and unpredictable. Obvious discrepancies exist between predictions and actual measurements in similar sections of tunnel construction. When constructing ultra-long and deep tunnels, such as the YL Tunnel, special attention should be paid to the coupling of multiple undesirable geological effects. To prevent hazards from occurring, a detailed analysis should be conducted based on geological data, and any risks should be predicted and monitored in advance.