Gas pipeline leakages in urban underground utility tunnels pose a severe threat to public safety. Leakages of varying aperture sizes trigger differentiated risks of diffusion and explosion; thus, achieving precise identification of leakage hole size has become a critical issue in safety management. To address the difficulty of traditional methods in effectively separating the acoustic features of different leakage levels within complex utility tunnel environments, this paper proposes a gas pipeline leakage risk level identification method based on a lightweight Spatial–Temporal Graph Neural Network (ST-GNN). First, relying on a real utility tunnel simulation platform, acoustic signals under different pressures and leakage hole size are collected, and time-frequency magnitude features are constructed through Short-Time Fourier Transform (STFT). Furthermore, each acoustic sample is independently converted into a graph with STFT time frames as nodes, where temporal neighborhood edges and K-nearest neighbor edges jointly encode local dynamics and non-local spectral similarities. This transforms unstructured acoustic signals into graph-structured data that embodies spatial–temporal coupling relationships. Building upon this, a lightweight Chebyshev graph convolutional network is designed to progressively extract discriminative features strongly correlated with leakage levels using multi-layer convolution. Experimental results on the actual utility tunnel simulation platform dataset demonstrate that the proposed method achieves excellent performance in a three-level leakage classification task. The t-SNE visualization reveals the effective separation of features, progressing from complete mixing in the input layer to distinct separation in the output layer. Through multiple training statistics and ablation experiments, the impact of dataset size and the number of network layers on the identification performance is analyzed, validating the robustness of the proposed model under limited samples and the effectiveness of its lightweight structure. This provides a feasible solution for the automated and refined identification of gas pipeline leakage levels in underground utility tunnels.

Earthquake-induced soil liquefaction poses a severe threat to the structural safety of underground infrastructure. This paper presents a shaking table model test conducted on sand sites, where two site models—namely, uniform saturated sand and layered dry–saturated sand—are designed to systematically analyze the influences of seismic wave amplitude, seismic wave frequency, and structure burial depth on the development of pore water pressure. The research findings indicate the following: (1) As the input peak ground acceleration (PGA) increases, the output surface peak ground acceleration rises, while the acceleration amplification factor decreases. Under high-intensity earthquakes, sand boiling and water gushing are observed in the uniform saturated sand site, and the overlying unsaturated layer can effectively inhibit the liquefaction development of the underlying saturated sand layer. (2) When the predominant frequency of the input seismic wave is close to the natural vibration frequency of the test site, the surface acceleration response is significantly enhanced, while the resonance effect of low-frequency seismic components is relatively weak. (3) The pore pressure ratio is higher in shallow soil layers, indicating that shallow strata are more prone to liquefaction. Both the pore pressure accumulation rate and the peak pore pressure ratio under uniform saturated sand conditions are higher than those measured in the layered dry–saturated sand site. (4) The sensitivity ranking of the three influencing factors is input PGA > burial depth > seismic wave frequency, where the input PGA is the dominant influencing factor, and the effect of the seismic wave frequency is not statistically significant. The conclusions of this study can provide a reference for seismic response assessment and anti-liquefaction design of shallow-buried structures in liquefiable sites.