Abstract:
The seismic safety of pumped storage power stations situated in mountainous terrain necessitates thorough investigation of topographic effects on ground motion parameters. This study conducts a comprehensive numerical analysis of seismic mountain topography effects at both upper and lower reservoir sites under two distinct engineering scenarios: the natural terrain condition and the post-excavation expanded reservoir condition. A three-dimensional numerical model was developed using FLAC3D finite difference software based on engineering design documentation, with model dimensions of 3 955.53 m×4 118.84 m×2 274.14 m. The model incorporated 20 187 gridpoints and 89 158 zones, with mesh refinement to 1 m×1 m at critical topographic features including monitoring points a1 and p1.
The geological formation was characterized as Class Ⅱ−Ⅲ granite bedrock with specified mechanical parameters: elastic modulus of 60 GPa, Poisson’s ratio of 0.25, density of 2750 kg/m3, and P-wave velocity of 5500 m/s. Seismic excitation was simulated through application of pulse time-history inputs simultaneously in three directions at the model base, with effective frequency band of 0−10 Hz. Rayleigh damping was implemented as the damping model for the simulation.
Monitoring points were strategically distributed throughout the model, including a1 through a6 in the lower reservoir basin, ad1 through ad7 at the lower dam foundation, p1 through p6 in the upper reservoir basin, pd1 through pd6 at the upper dam foundation, and additional points b through k along the water conveyance tunnel alignment. Comparative analysis encompassed root mean square acceleration, 90% energy duration, and transfer function evaluation through Fourier spectral ratio analysis.
The analytical results demonstrate that under both natural and excavation conditions, vertical direction root mean square acceleration values exceed horizontal direction values at locations with steep topographic slopes. This phenomenon is attributed to the reflection and transmission of P-waves at free surfaces and subsequent generation of SV-waves. Following reservoir excavation, substantial increases in root mean square acceleration were observed at dam foundation locations, with approximate doubling of acceleration values compared to natural conditions. Concurrently, 90% energy duration at dam foundations increased significantly from approximately 0.99 seconds under natural conditions to approximately 4.33 seconds following excavation.
Spectral analysis revealed critical frequency domain modifications, with the predominant frequency band shifting from 10−20 Hz under natural conditions to below 5 Hz following excavation. Fourier spectral ratio analysis confirmed pronounced amplification of frequency components below 5 Hz in post-excavation conditions, accompanied by reduced amplification in the 15−30 Hz frequency range. This redistribution of seismic energy toward lower frequencies increases resonance potential with long-period dam structures.
The investigation establishes that reservoir excavation and expansion significantly alter local topographic amplification characteristics through two primary mechanisms: loss of horizontal bedrock constraint at dam foundation locations and creation of wave-focusing geometric configurations at reservoir boundaries. These modifications collectively generate enhanced seismic demand characterized by amplified acceleration amplitudes, prolonged strong motion duration, and preferential enrichment of low-frequency seismic energy content.
These findings demonstrate limitations in conventional seismic design approaches that utilize ground motion parameters derived exclusively from pre-construction topographic conditions. The study concludes that comprehensive seismic safety evaluation for pumped storage power stations in mountainous terrain must incorporate explicit consideration of construction-induced topographic modifications. Engineering practice should advance toward implementation of analytical frameworks that independently assess seismic response under both natural and post-excavation conditions, with separate determination of ground motion parameters for upper and lower reservoir areas. Particular emphasis should be placed on vertical ground motion components and systematic evaluation of low-frequency resonance hazards in structural design methodologies for critical infrastructure in complex topographic environments.