Abstract Monitoring the deformation of arch dams is critical for assessing structural health and operational safety. This study provides an in-depth analysis of the deformation characteristics of extra-high arch dams during their operational phase, using a 300-meter-high dam as a primary case study. By analyzing spatiotemporal deformation patterns and establishing a multiple regression statistical model—accounting for hydrostatic pressure, temperature fluctuations, and aging effects—the research reveals that while the dam initially shifts downstream during reservoir impoundment, it eventually transitions into a stable elastic state. The decreasing rate of deformation over time indicates structural convergence and a healthy operational status.
Introduction
Extra-high arch dams represent some of the most complex engineering challenges globally. Due to their immense scale, these structures are subjected to significant and overlapping loads, including hydrostatic pressure from the reservoir, seasonal thermal cycles, and long-term material "creep" or aging of both the concrete and the underlying bedrock. Distinguishing between normal environmental responses and potential structural warnings is vital for dam safety. This paper establishes a "spatiotemporal law" for these deformations, providing a baseline for normal behavior in ultra-high concrete structures during their first decade of operation.
Monitoring Network and Methodology
The study utilizes a comprehensive monitoring network at the Xiluodu Hydropower Plant in China. The system includes 29 normal perpendicular lines and 8 inverted vertical lines, allowing for precise mapping of relative movements across various dam sections and elevations.
To isolate the drivers of movement, researchers employed a Hydrostatic-Season-Time (HST) statistical regression model. This model decomposes total observed deformation ($\delta$) into three components:
- Hydrostatic ($\delta_H$): Movement driven by water level changes, modeled via a fourth-order polynomial.
- Seasonal ($\delta_S$): Cyclic movement caused by temperature variations, modeled using multi-period harmonic functions.
- Time/Aging ($\delta_T$): Irreversible movement from concrete creep and foundation adjustment, modeled using linear and logarithmic functions.
The dataset spans from May 2013 to December 2023, covering ten reservoir storage cycles and nine subsidence cycles.
Key Results and Spatiotemporal Trends
The analysis confirms that the dam is currently in a "good elastic working state." A high degree of model accuracy was achieved, with correlation coefficients ($R$) exceeding 0.99 for typical measurement points and standard deviations under 1mm.
Key observations include:
- Sectional Activity: The arch crown (the center of the curve) exhibits the largest radial displacement. Section 15, located near the center, recorded the highest annual variation, reaching a maximum of 38.84mm.
- Convergence of Aging Deformation: Crucially, the "aging" component of deformation—the movement not attributed to water or weather—is decreasing annually. This convergence indicates that the structure is reaching a new equilibrium and is not prone to accelerating, unstable movement.
- Directional Shifts: While the dam initially deformed downstream during early water storage, the aging component has since trended upstream, suggesting the valley and dam foundation are stabilizing.
Conclusion
The first decade of monitoring at the Xiluodu dam demonstrates that the structure has successfully transitioned from its initial adjustment phase into a predictable, elastic operational state. While the current trends are positive, the study highlights the need for continued long-term monitoring. Future work could integrate more granular geological data to better understand how specific rock mass behaviors influence the aging component of the HST model, ensuring safety throughout the dam's 50- to 100-year lifespan.
Summary updated