A method to determine seismic fragility of buried pipelines

Zhenyi Kong, Zhongjie Lu, Changsheng Xu, Li ChunOpenalex Unknown 2025 · DOI · OpenAlex

Abstract

Underground pipelines are vital to a nation’s economy, environment, and public well-being, and often regarded as societal lifelines. However, they are vulnerable to seismic actions, necessitating fragility analysis as part of lifecycle management. This paper proposes a probabilistic method to assess pipeline seismic fragility, considering uncertainties in earthquake occurrence and intensity. Ground motions are generated using a random function-spectrum model, and nonlinear dynamic analyses simulate pipeline responses under varying peak ground accelerations. By repeating these steps, the pipeline’s dynamic response statistics and probability distribution function are obtained. Given an acceptable response limit, a limit state function can be constructed, and failure probability can be determined using structural reliability methods. For different acceptable limits, a fragility curve can be plotted for different peak ground accelerations. A worked example is presented to demonstrate the method’s accuracy and efficiency. This analysis provides a scientific basis for pipeline design, operation, and risk management.

Introduction: The Vulnerability of Lifeline Infrastructure

Buried pipelines serve as the critical "lifelines" of modern society, facilitating the transport of essential resources such as water, gas, and electricity. Despite their importance, protecting these assets against seismic events remains a significant engineering challenge. The vulnerability of a pipeline is dictated by a complex interaction between the structural properties of the pipe material and the dynamic behavior of the surrounding soil.

Traditional risk assessment methods often struggle with two main issues: the scarcity of site-specific historical seismic records and the reliance on "stationary" ground motion assumptions. Stationary models treat earthquakes as consistent signals, failing to capture the chaotic, time-evolving nature of real seismic events. Furthermore, many studies focus on isolated failure modes, often overlooking the critical interplay between compressive buckling and tensile rupture. This research introduces a realistic, multi-mode probabilistic framework to address these gaps.

Methodology

The proposed analytical framework consists of three integrated stages:

  1. Non-Stationary Excitation Modeling: To overcome the limitations of historical data, the study employs the Random Function-Spectral Representation Method (RFSRM). This approach generates synthetic ground motions that realistically modulate in both frequency and intensity over time, aligned with the Clough-Penzien power spectrum model.
  2. High-Fidelity Structural Simulation: Numerical models were developed in ABAQUS utilizing a Winkler foundation beam model. As shown in Figure 1, the pipeline is represented by beam elements, while the soil-pipe interaction is modeled through axial nonlinear spring elements. A tri-linear constitutive model for L415 steel was used to accurately account for plastic deformation.
  3. Fragility Curve Construction: By conducting numerous simulations across a range of Peak Ground Accelerations (PGA), the researchers calculated the exceedance probabilities for three damage states: Minor, Moderate, and Major. These results were then fitted to a lognormal cumulative distribution function using maximum likelihood estimation.
Figure 1: Schematic of the Finite Element (FE) model for buried pipelines, illustrating the pipe-soil interaction via nonlinear springs.

Key Insights and Results

The study yielded several critical findings regarding pipeline behavior under seismic stress:

  • Buckling vs. Tensile Sensitivity: A primary insight is the significant divergence in failure modes. The analysis reveals that buckling—manifesting as local wrinkling or beam-type warping—occurs at much lower seismic intensities compared to tensile rupture.
  • Validation and Efficiency: The generated fragility curves demonstrated high consistency with traditional Monte Carlo Simulations (MCS), validating the method's accuracy while maintaining computational efficiency.
  • Dominant Failure Mode: For low-to-moderate seismic intensities, the probability of buckling failure is markedly higher than that of tensile failure. Figure 2 illustrates the fragility curves for buckling across different damage thresholds.
  • Design Implications: These results suggest that engineering efforts should prioritize enhancing axial compression resistance (buckling) in seismic zones, as this typically represents the "weakest link" in the pipeline's structural integrity.
Figure 2: Fragility curves for buckling failure across minor, moderate, and major damage states.

Conclusion and Future Directions

While the proposed framework offers a robust tool for risk management, certain limitations remain. The use of the Winkler foundation model simplifies soil into independent springs, which may not capture complex 3D effects like soil-flow or liquefaction. Additionally, the results are specific to L415 steel and particular soil profiles; different materials or saturated conditions may produce different fragility signatures.

Future research will aim to expand this framework to include multi-hazard scenarios, such as the combined impact of long-term corrosion and seismic loading over a pipeline's multi-decade lifespan. This will provide a more comprehensive scientific basis for the design, operation, and resilient management of these vital infrastructure systems.

Summary updated

Abstract

Underground pipelines are vital to a nation’s economy, environment, and public well-being, and often regarded as societal lifelines. However, they are vulnerable to seismic actions, necessitating fragility analysis as part of lifecycle management. This paper proposes a probabilistic method to assess pipeline seismic fragility, considering uncertainties in earthquake occurrence and intensity. Ground motions are generated using a random function-spectrum model, and nonlinear dynamic analyses simulate pipeline responses under varying peak ground accelerations. By repeating these steps, the pipeline’s dynamic response statistics and probability distribution function are obtained. Given an acceptable response limit, a limit state function can be constructed, and failure probability can be determined using structural reliability methods. For different acceptable limits, a fragility curve can be plotted for different peak ground accelerations. A worked example is presented to demonstrate the method’s accuracy and efficiency. This analysis provides a scientific basis for pipeline design, operation, and risk management.
BibTeX
@inproceedings{kong2025method,
  author = {Zhenyi Kong and Zhongjie LU and Changsheng Xu and Li Chun},
  title = {{A method to determine seismic fragility of buried pipelines}},
  booktitle = {Openalex Unknown},
  year = {2025},
  doi = {10.1201/9781003595120-197},
}
Figure