2.1 Underground caverns and geological conditions in this hydropower plant

The water division and power generation system in the Baihetan project include water intakes, pressure tunnels, powerhouse caverns, main transform tunnels, tailrace surge chambers, and tailrace tunnels. A part of the 3D geometry model of this project at the left bank is shown in Figure 1. The powerhouse cavern has the dimensions of 438.00 m (Long) × 34.00 m (Wide) × 88.70 m (Hide). The cavern axis orientation is N20° E.

               

The geological condition in the powerhouse cavern zone is complicated. The strata at the left bank are mainly P2β23, P2β24, and P2β31 basalt and breccia lava. The hard basalt is mainly composed of cryptocrystalline, porphyritic, and amygdaloidal basalt. The surrounding rock masses of the powerhouse are classified as grade III by the rock mass rating system (RMR) system. The burial depth of the powerhouse is 950–1050 m in the horizontal direction and 260–330 m in the vertical direction. The initial maximum principal stress is 19–23 MPa, and the measured maximum horizontal principal stress is 33.39 MPa. The angle between the initial maximum principal stress and the main axis of the powerhouse is 60°–70°.

2.2 Mechanical parameters of the interlayer shear zone

Interlayer shear zone C2 intersects the underground openings at the left bank (Figure 1). It is well developed between P2β24 and P2β31 basalt, with an average thickness of about 20 cm (Figure 2). It is composed of tuff, but also contains clay with rock fragments. Hence, its strength is low and can be easily softened by water. In addition, some small-scale faults and large fractures are distributed around the underground powerhouse at the right bank, such as f717, f718, f719, f720, f723, and LS3152. Three sets of joints are distributed around the underground powerhouse: (1) a strike of N30°–70° W, a SW tendency, and a dip angle of 65°–90°; (2) a strike of N50°–70° E, a SE tendency, and a dip angle of 50°–60°; and (3) a strike of N20°–50° E, a SE tendency and a dip angle of 10°–35°. The joint lengths are generally 2–5 m with spacing larger than 50 cm and 20–50 cm in local areas. The fracture planes are closed, straight, and rough.

               

Strength parameters of the weak interlayer shear zone are critical to the stability analysis of the surrounding rocks. Five in situ direct shear tests were conducted and the results are shown in Figure 3. In order to avoid the size effect of the in situ experiments, numerical direct shear experiments were also conducted using 3DEC. The Coulomb-slip model in 3DEC was used to model opening, compressing, and shearing behaviors of structural planes (Itasca, 2010). The unloading relaxation of the interlayer shear zones is closely related to the thickness of the interlayer shear zone and the mechanical properties of its soft filling materials. In this study, both normal stiffness and shear stiffness of this interlayer shear zone were derived by the equivalent deformational response. Specifically, the normal stiffness Kn is 0.2 GPa/m and shear stiffness is 0.074 GPa/m.

               

Figure 4 shows the comparison between numerical simulations and field experiments. As can be seen, they agree very well with each other. The parameters used in numerical simulations are listed in Tables 1 and 2. The strength parameters of the interlayer shear zone on a large scale were determined as follows: the cohesion was 0.09 MPa and the friction angle was 15.64°.

               

3.1 Effect of the interlayer shear zone on sidewall deformation

When the initial maximum principal stress is horizontal, the deformation of the sidewall upstream is larger than that downstream. This is due to the height difference of vertical sidewall excavation. The numerical modeling conducted in 2011 during the basic design period shows that the interlayer shear zone obviously changed the deformation mode of the high sidewall (Figure 5). The interlayer shear zone C2 strikes the upstream. For the sidewall at the downstream side, C2 undergoes tensile shear deformation during excavation. This can cause large sidewall sliding along C2 toward the free face of the hanging wall. When C2 cuts the sidewall at a lower position, the deformation of the hanging wall at the downstream side reaches 125 mm (Figure 5a). In contrast, for the sidewall at the upstream side, C2 undergoes compressive shearing deformation during excavation and its shearing movement is quite small. When the layered zones are revealed after excavation, toppling deformation of the sidewall occurs at the footwall of C2 (Figure 5b). Thus, the large deformation of surrounding rocks at the upstream sidewall mainly exists in the footwall of the shallow part of C2. When C2 cuts the sidewall at a higher place, local deformation reaches up to 70–90 mm. Thus, this C2 causes a large difference in the discontinuous deformation at the upstream and the downstream sidewalls of this underground powerhouse cavern. It exerts a larger impact on the deformation difference of the downstream sidewall.

               

3.2 Deformation of the interlayer shear zone

The discontinuous deformation of the sidewall varies with the cutting height of the powerhouse cavern sidewall by the interlayer shear zone. Figure 6 shows the underground structure of the powerhouse cavern and eight conduits (named as unit 1 to unit 8) (Meng, Fan, et al., 2016). The interlayer shear zone strikes the upstream side, as shown in Figures 5b and 6. Shearing deformation of the interlayer shear zone is also different at the upstream and downstream sides and the shearing deformation of C2 is larger at the downstream sidewall than that at the upstream sidewall. Along the axis direction of the underground powerhouse cavern, the shearing deformation of C2 is larger at the south unit than at the north unit. The largest shearing deformation appears at the sidewall of units #2, #3, and #4 and the value is 60–70 mm. Furthermore, the shearing deformation in the interlayer shear zone is rather small and almost disappears as the distance from the powerhouse cavern exceeds the cavern width of 33.4 m.

               

3.3 Rock mechanical problems caused by the interlayer shear zone

The high sidewall of the underground powerhouse cavern undergoes strong stress relaxation. Being very soft in nature, the interlayer shear zone C2 becomes the moving boundary of the surrounding rocks under high stress conditions. If the rock masses above and below the interlayer shear zone are relatively intact, an overall shearing movement may occur at the hanging wall and footwall with overall stability. However, for rather fractured rock masses, the stress relaxation of the interlayer shear zone and the secondary structural plane may cause local stability in the rock masses.

The cavern axis orientation is N20° E. As shown by geological investigations, the angle between the major NWW joint set and the powerhouse axis is very large (Figure 1). Thus, large rock blocks will not be formed. However, the NE joint set may combine with C2 and the NWW joint set to form potential large deformation and unstable blocks. It should be noted that even though this study does not consider the creep effect, the creep of the interlayer shear zone containing clay may affect the long-term stability of the underground powerhouse cavern. In fact, the creep effect would be a good topic for future study.

4.1 Comparison of tunnel-replacement schemes

A tunnel-replacement scheme was proposed to replace C2 with concrete. This control measure used is to reduce the shearing deformation of C2 and thus prevent severe unloading relaxation of the hanging wall and footwall of rock masses. The tunnels should not only control the shearing deformation of the interlayer shear zone but also ensure their safety. Under high in situ stresses, the key issue is the proper distance from the cavern sidewall to the replacement tunnels. If the replacement tunnels are too far away from the excavation face, shearing deformation can not be controlled effectively. If the replacement tunnels are too close to the excavation surface, the replacement tunnels may yield or even be cut off by the large shear stress. Thus, three replacement schemes were proposed during the detailed design stage in 2012. Concrete was used to fill in the tunnels in C2. The details of the plan and the corresponding effects are presented in Figure 7 and described as follows.

               

Scheme 1 involves deep tunnel replacement. This scheme involves replacement of the interlayer shear zone 13–19 m away from the sidewall. As shown in Figure 7a, the deep shear deformation can be cut off effectively and systematically through this scheme. However, the shallow deformation is still large and the local shear deformation during excavation is 55–65 mm.

Scheme 2 involves serial tunnel replacement. This scheme involves replacement of the interlayer shear zone with serials of concrete tunnels perpendicular to the excavation surface. As shown in Figure 7b, this scheme can effectively control the shallow shear deformation along C2. However, the shear deformation between tunnels is still large and the local shear deformation during excavation is 50–60 mm.

Scheme 3 involves the combination of deep tunnel replacement, serial tunnel replacement, and concrete plug. Actually, this scheme is the combination of Scheme 1 and Scheme 2. As shown in Figure 7c, this scheme can reinforce both deep and shallow surrounding rocks at the same time. The local shear deformation of C2 during excavation can be significantly reduced to 20–30 mm. Therefore, Scheme 3 boasts of obvious advantages in deformation control.

4.2 Prediction of deformation of the sidewall and the interlayer shear zone in the construction stage

During the construction period, most of the underground excavation and support systems were determined. Although Scheme 3 was the most effective in controlling the shear deformation of the interlayer shear zone, it was difficult to ensure the safety of the concrete plug. In view of these considerations, the final control measure was a modified Scheme 3 that cancelled the concrete plug and reduced the spacing of serial tunnel replacement. The updated numerical simulations on the tunnel replacement of the interlayer shear zone were conducted in 2014. Both the surrounding rock deformation and the shear deformation of the interlayer shear zone C2 under high in situ stresses were predicted. The prediction results are presented below.

As shown in Figure 8, unloading deformation of the hanging wall of the interlayer shear zone around units #1–#5 is still large at the downstream side. The maximum displacement of the sidewall reaches up to 100–110 mm (Figure 8b). These places are in need of key reinforcement. Meanwhile, the largest shear deformation of the interlayer shear zone appears at around units #2–#4 at the downstream side of the powerhouse and could be reduced by 10–20 mm. The final largest shear deformation of C2 is 50–55 mm after tunnel replacement (Figure 9). However, shear displacement of the interlayer shear zone at the upstream side does not change much after tunnel replacement. This is inferred from the shear deformation of the interlayer shear zone around unit #3 (Figure 8b). Thus, replacement tunnels are not used in the interlayer shear zone at the upstream side.

               

               

5.1 Monitoring data on the deformation of surrounding rocks

The powerhouse cavern at the left bank is 438 m long, 34 m wide, and 88.7 m high. The excavation started from 2014 and lasted over 4 years up to 2018. The cavern was excavated in 10 stages, as shown in Figure 10. This excavation mode is called layered excavation. Each layer of excavation was conducted carefully and supported in a timely manner by strict analysis and double checking to prevent severe collapse. Figure 11 shows the photo of the powerhouse cavern taken at the site. The hanging wall sliding along C2 toward the free face, which is caused by tensile shear deformation of the shear zone, can be obviously seen on the sidewall at the downstream side (Figure 11a). This is a result of excavation and stress relief.

               

               

Monitoring instruments were installed in the construction stage. The extensometer Mzc0+042-1 was installed at the downstream sidewall of the powerhouse at the left bank, as shown in Figure 8b. The deformation of surrounding rock during and after the cavern excavation was monitored using an extensometer. Figure 12 presents the displacement at different distances of 0, 1.5, 4.5, 10.0, and 20.0 m away from the excavation face monitored from March 20, 2017 to March 15, 2020. The displacement of the sidewall was 75 mm on June 28, 2017. It can be seen that the deformation of the surrounding rocks increased quite rapidly during the first 3 months after the main powerhouse cavern was excavated from EL.575 m to EL.568 m. This was due to the strong stress relaxation after excavation stage VII. The displacement still increased rapidly after 3 months. It was 89.25 mm on the sidewall on August 19, 2017. The displacement of surrounding rock remained almost stable after July 2018. The final displacement of surrounding rocks was 105 mm at the excavation face and 25 mm at a distance of 20 m away from the excavation face. The displacement of surrounding rock 20 m away from the excavation face was smaller than 25 mm during the whole monitoring period. Both the predicted deformation of 110 mm in Figure 8 and the monitoring data from 2017 to 2020 (Figure 12) share similar change patterns. These site monitoring data verified the numerical simulation results of the surrounding rock deformations predicted earlier in 2012 (see Section 4.2).

               

5.2 Data monitoring on shear deformation of the interlayer shear zone

In order to monitor shear deformation of the interlayer shear zone at the downstream side, a drilling hole with an inclinometer was set 6 meters away from the excavation face in busbar tunnel #3 in February 2017 (Figure 13a). The inclinometer INzmd3-0+023-1 captured successive and significant shear deformation in the process of layered excavation (Figure 13b). One month later, the monitored shear deformation was 13.5 mm on February 19, 2017. The largest shear deformation monitored was 52.91 mm on June 10, 2017. It took just 4 months for the shear zone to reach such a large shear deformation. Both the deformation pattern and the magnitude of deformation agree well with the numerical simulation results presented in Section 4.2 (Figure 9). Accordingly, it can be concluded that the deformation of the interlayer shear zone C2 and the underground powerhouse cavern during the construction stage can be predicted well using the numerical simulation procedure.

               

6.1 Effects of replacement tunnel location and replacement tunnel cracks

The location of replacement tunnels is a key parameter for controlling the shear deformation of the interlayer shear zone and the extent of rock mass stress relaxation. After many trials, replacement tunnels with a size of 6 m × 6 m were set 13 m away from the sidewall of the powerhouse cavern in the construction. This ensured that the replacement tunnels would not be completely cut even at places where large shearing deformation occurred.

The mechanical mechanism of deep replacement tunnels involves passing the shear loading along C2 from the hanging wall to the footwall of rock masses. As shown in Figure 14a, the shear stress inside the replacement tunnels increases during excavation of the cavern and the interlayer shear zone. After the interlayer shear zone is exposed, unloading relaxation of rock masses is induced at the footwall. Thus, the deformation difference between the hanging wall and the footwall becomes relatively small. The inclinometer readings show that shearing deformation decreases from 52.91 to 39.00 mm (Figure 13b). Therefore, shear loading of the replacement tunnels decreases in the meantime (Figure 14a). The shearing deformation along the interlayer shear zone is controlled and thus prompt anchoring is possible. However, replacement tunnels might still yield in this situation. Discontinuous cracks might occur in the concrete as well.

               

During layered excavation, cracks were actually observed in replacement tunnels in field investigations since they carried large loading of the surrounding rock masses and could yield at some places, as shown in Figure 14b. The distribution of unconnected cracks was consistent with that of shearing deformation of the interlayer shear zone and loading transport of replacement tunnels obtained in the numerical analysis. When the powerhouse cavern was further excavated to a lower altitude, the deformation of rock masses of the footwall would increase, and thus the shear deformation of the replacement tunnels would decrease.

6.2 Evaluation of replacement tunnel stability

The stability of replacement tunnels was evaluated using a strength reduction method (Liang et al., 2019). Specifically, the influence of fracture cracking and long-term stability of the replacement tunnels is determined by identifying the shearing deformation development of C2. Accordingly, it can be checked whether the replacement tunnels would undergo even more severe cracking or not. Numerical results show that the largest deformation still appears at the downstream sidewall from generator #2 to #4 during the strength reduction of tunnel concrete. As shown in Figure 15, when the strength reduction factor (SRF) is 3.0, the shearing deformation increment is smaller than 4 mm in the interlayer shear zone and smaller than 5 mm in the surrounding rocks; axial force increment of the bolt is smaller than 50 kN; and plastic zone and tensile stress zone increment is smaller than 1%.

               

The replacement tunnel can play a very important role in transferring shearing loading from the hanging wall to the footwall. It would not go further cracking under such conditions. Even if it did, and even if the mechanical parameters of replacement tunnels decrease considerably, it would not cause a major change in sidewall deformation or the plastic deformation zone. It would not affect the anchoring system either. Therefore, the overall stability of the sidewall of the powerhouse cavern can be guaranteed.