The stability of the compressed air energy storage (CAES) cavern is influenced by ground stresses during the construction phase and by cyclic high internal pressure (≥10 MPa) during the operational phase, resulting in different mechanical responses of the surrounding rock during the unloading and loading phases. This study investigates the stress paths of the CAES cavern from the perspectives of stress evolution and damage modes throughout both construction and operational phases. The findings indicate that during the excavation phase, active support technology significantly reduces deformation by adjusting the principal stress field of the surrounding rock, particularly in low-strength rock formations. Additionally, the load transfer effect decreases the burden on initial supports, such as steel arches. In the operational phase, the high internal pressure causes the sequence of principal rock stresses to shift, with radial stress increasing to the maximum principal stress and circumferential stress transitioning to the minimum principal stress. When exposed to hydrostatic pressure, characterized by low local stress and high rock strength, the surrounding rock at the tunnel wall may experience tensile failure during the high-pressure gas storage stage; otherwise, shear failure will occur at the tunnel wall. When the lateral pressure coefficient () is less than 1, shear failure occurs in the surrounding rock at the side walls. In cases where the lateral pressure coefficient is particularly low, tensile failure may still happen at the arch crown. Conversely, when exceeds 1, shear failure is observed in the surrounding rock at the arch. If the lateral pressure coefficient is notably high, tensile failure may still occur at the side walls. As pressure increases, the area of tensile fractures may subsequently undergo shear failure. Further analysis reveals that conventional grouting and radial bolt systems significantly enhance stability only during the excavation phase. During operation, radial bolts not only lose their reinforcing capabilities but also exacerbate stress concentration in the surrounding rock under compression. In contrast, enhanced circumferential restraint effectively improves the stress state of the surrounding rock. For surrounding rock of good quality and high strength, grouting has limited efficacy in filling fractures. Conversely, grouting measures have been shown to increase the shear strength of formations characterized by low strength. Based on these insights, this study proposes an innovative solution for the stability control of CAES caverns subjected to complex stress paths by forming an anchor network system using crossed diagonal anchors and optimizing the inclination angle of the anchors to achieve a full-cycle tensile state.
As a unique geological hazard in a high-geo-stress environment, rockburst happens with strong suddenness, randomness, and destructiveness, but the mechanism of its occurrence in a deep-buried tunnel in a high-geo-stress environment needs further study. Based on the analyses of the stress field of the rockburst section of Ping’an Tunnel, which is over-long and deep-buried, the occurrence mechanism of rockburst is figured out. Furtherly, the intensity and location of rockburst are predicted by using the rockburst criterion. Results show that there exists large compressive stress at the side wall of the tunnel, which is the main cause of rockburst. Under a high-geo-stress condition, due to the existence of unfavorable factors like structural planes, high intensity rockburst is likely to happen in the middle of the tunnel face after the sudden release of original rock stress. Arranging stress-releasing holes in the rockburst section can effectively reduce the possibility of rockburst by releasing the original rock stress in advance. The research results can deepen the understanding of the mechanism of rockburst in a high-geo-stress environment and provide scientific basis for the prevention and control of rockburst in similar engineering projects.
This paper proposes a stepwise expert-teaching reinforcement learning framework for intelligent frequency control in hydro–thermal–wind–solar–compressed air energy storage (CAES) integrated systems under high renewable energy penetration. The proposed method addresses the frequency stability challenge in low-inertia, high-volatility power systems, particularly in Southwest China, where large-scale renewable-energy-based energy bases are rapidly emerging. A load frequency control (LFC) model is constructed to serve as the training and validation environment, reflecting the dynamic characteristics of the hybrid system. The stepwise expert-teaching PPO (SETP) framework introduces a stepwise training mechanism in which expert knowledge is embedded to guide the policy learning process and training parameters are dynamically adjusted based on observed performance. Comparative simulations under multiple disturbance scenarios are conducted on benchmark systems. Results show that the proposed method outperforms standard proximal policy optimization (PPO) and traditional PI control in both transient response and coordination performance.