Zhuojin Pipe Fitting Co., LtdComparative Study on Horizontal Bearing Characteristics of Steel Pipe Pile Foundations for Xiangshui Offshore Wind Power
Literature Overview
This study by Zhai Endi, Xu Haibin, Guo Shengshan, Jinhua, and Du Xiuli (2019), published in the Acta Energiae Solaris Sinica, presents a comparative analysis of horizontal load capacity prediction methods for steel pipe pile foundations supporting offshore wind turbines at the Xiangshui offshore wind farm in Jiangsu Province, China. Supported by the National Natural Science Foundation of China (Grants 51421005 and 51322813), the research bridges the gap between theoretical prediction methods and actual measured performance, providing essential calibration data for offshore wind foundation design.
Engineering Context and Methodology
The study is based on a 2 m diameter steel pipe pile horizontal loading test at the Xiangshui offshore wind farm. An ANSYS finite element model was developed using design-provided soil parameters. Three analytical methods were compared against measured pile head load-displacement curves, pile deflection profiles, and bending moment distributions:
| Method | Source | Basis |
|---|---|---|
| p-y method | API RP 2A-WSD | API recommended p-y curves |
| m-method | Port Engineering Pile Foundation Code | m-factor approach |
| m-reduced method | Modified m-factor | Reduced m-value approach |
Comparative Results
The comparison revealed the following relative accuracy:
| Method | Prediction Bias | Relative Accuracy |
|---|---|---|
| API p-y method | Conservative (safe side) | Overestimates stiffness, underestimates displacement |
| m-method | Unconservative | Underestimates displacement, overestimates stiffness |
| m-reduced method | Closest to measured values | Best overall agreement |
The API p-y method, while conservative, may lead to uneconomic designs if used directly without calibration. The m-method and m-reduced method both underpredicted displacements but the m-reduced method showed the closest agreement with measured data.
Technical Analysis
Why API p-y is Conservative
The API p-y curves are derived from a broad database of pile tests across various soil types and loading conditions. For the specific soil profile at Xiangshui, the API curves overestimate soil resistance at small to moderate displacements, resulting in predicted pile stiffness that is higher than actual. This conservatism is appropriate for safety but may result in oversized pile sections or unnecessarily deep foundations.
Limitations of m-method Approaches
The m-method relies on a linearized soil modulus concept that does not fully capture the nonlinear soil-pile interaction behavior observed in the field tests. The m-reduced method, which applies a reduction factor to the m-value, partially compensates for this limitation and provides better agreement. However, the appropriate reduction factor may vary with soil type, pile diameter, and loading history, making it difficult to establish universal recommendations.
Engineering Practice Implications
- Design calibration: For offshore wind projects in similar soil conditions to Xiangshui, the m-reduced method with appropriate reduction factors should be considered as the baseline prediction method, supplemented by API p-y for verification.
- Site-specific calibration: The study emphasizes that prediction methods must be calibrated against local soil data. The Xiangshui results may not directly transfer to other locations with different soil profiles.
- Load testing: Horizontal load tests should be conducted on representative piles during construction to validate design assumptions and provide data for future projects.
Key Questions and Reflections
The study acknowledges that the comparison is based on a single project and calls for additional test data and computational analyses to strengthen the conclusions. This is an important caveat: offshore soil conditions vary significantly even within the same coastal region, and the applicability of the m-reduced method to other projects requires site-specific verification. The study also does not address the long-term cyclic loading effects from wind and wave loads, which are critical for offshore wind turbine foundations. Furthermore, the interaction between adjacent piles in a wind farm array was not considered, which could be significant for closely spaced foundations.
Study Insights and Implications
This study provides essential validation data for horizontal load prediction methods in offshore wind foundation design. The finding that the m-reduced method offers the best agreement with measured data is a valuable practical insight, though it should be applied with caution outside the specific site conditions studied. The conservative nature of the API p-y method, while safe, highlights the economic penalty of relying solely on code-based methods without site-specific calibration. For engineers designing offshore wind foundations, this study reinforces the importance of horizontal load testing and the need for method calibration against local soil behavior. The work also underscores the broader challenge in offshore geotechnical engineering: the gap between simplified analytical methods and complex soil-pile interaction behavior, which can only be bridged through systematic testing and data accumulation. As the offshore wind industry expands, projects like this one are essential for building the technical knowledge base that enables safe and economical foundation design.