Wind energy, a sustainable and environmentally friendly energy source, offers an innovative path for global efforts in addressing energy shortages. The installed capacity of offshore wind turbines has shown consistent annual growth. These turbines are typically anchored to the seabed, often involving marine clay layers. Foundation costs, comprising approximately 25%–35% of total costs (Bhattacharya, 2014), necessitate various marine clay parameters for the geotechnical design of diverse foundation solutions. Under complex marine environmental loads, such as wind, waves, storms, and earthquakes, marine clay demonstrates notable stiffness degradation characteristics (Yang et al., 2018; Pan et al., 2021; Xiao et al., 2023) and can even lose strength entirely. Thus, understanding the cyclic resistance of marine clay is crucial for foundation design. However, its precise determination and rational design pose significant challenges in engineering construction (Lunne et al., 2006; Berre et al., 2022; Gao et al., 2024).

The predominant method to determine marine clay’s cyclic resistance involves laboratory cyclic tests, including cyclic triaxial tests, cyclic torsional shear tests, and cyclic direct simple shear tests. These tests demand high-quality soil samples and skilled testers. Nevertheless, soil sample disturbance during drilling and transport can compromise the accuracy of these tests in representing marine clay’s in-situ cyclic resistance. In addition, the requirement for a large volume of soil samples substantially increases the cost. Consequently, the engineering community urgently seeks effective methods to ascertain marine clay’s cyclic resistance with fewer tests.

Comprehensive in-situ and laboratory tests are indispensable in acquiring site geotechnical properties and soil parameters. Cone penetrometer (CPTu) tests are preferred for in-situ testing due to their high accuracy, convenience, and speed (Cai et al., 2012; Cai et al., 2016; Duan et al., 2017; Meng and Pei, 2023). CPTu data comprehensively represent soil strength and deformation capacity under static and dynamic loading. Over the past decades, CPTu-based undrained shear strength evaluation methods for soils have been extensively developed (Sandven, 1990; Eslami, 1997). Subsequently, researchers began developing cyclic strength evaluation methods based on in-situ and laboratory tests (Olsen, 1994; Robertson and Wride, 1998; Robertson, 2009). Juang et al. (2008) developed a deterministic CPTu-based cyclic resistance evaluation method applicable to various soil types. Juang et al. (2012) and Ku and Juang (2012) refined this model. Notably, these models were formulated using CPTu data for terrestrial soils. However, marine clays, influenced by factors like high salt content, low-temperature seawater environments, unique cementitious materials, and complex hydrodynamics, develop flocculated structures distinct from terrestrial soils. Thus, terrestrial clay cyclic resistance models are not directly transferable to marine clays. Limited research focuses on evaluating marine clay cyclic resistance. He et al. (2021) and Wang et al. (2022) explored the cyclic behavior of marine soils using CPTu tests, but they did not develop predictive models. Therefore, creating and refining a cyclic resistance evaluation method for marine clay based on in-situ and laboratory tests is essential.

This study conducts a case study involving CPTu tests at the Shaba offshore wind farm in southern China to establish a cost-effective method to evaluate marine clay cyclic resistance. Soil stratigraphy is delineated, highlighting the soil’s multilayered and complex sedimentation. In addition, consolidated undrained cyclic triaxial tests on marine clay are conducted to assess cyclic resistance, a vital parameter for foundation design. The well-established CPTu-based cyclic resistance evaluation methods in existing literature are not fully applicable to marine clays. Ultimately, this research develops a CPTu-based cyclic resistance evaluation method for marine clay at the Shaba offshore wind farm. The study’s findings will offer assistance and guidance for constructing offshore wind farms in China.

The Yangjiang Shaba Offshore Wind Farm is situated in the southern sea area of Shaba Town, Yangxi County, Yangjiang City, Guangdong Province, as illustrated in Figure 1 . This wind farm lies approximately 20 kilometers offshore. The area’s sea is expansive, lacks islands, and features relatively flat terrain. Water depths vary from 23 to 27 meters, with an average tidal range of around 2.73 meters. Summer months bring significant typhoon impacts, leading to maximum wave heights near 8 meters.

The site is positioned at the Rudong River’s mouth, where the sedimentary environment is notably complex, shaped by the combined influences of river flow and ocean waves. Sediments primarily comprise marine, alluvial marine, and residual deposits, characterized by a swift sedimentation rate. Soil stratification will be elaborated upon based on CPTu and borehole sampling results. In addition, the sea area at this location is spacious, devoid of surrounding islands. The seafloor topography is predominantly gentle, showing a trend of higher elevation in the northwest and lower in the southeast. No potential submarine geological hazards, such as underwater landslides, have been identified.

Considering cone resistance as an indicator of the failure strength of soils in situ (Yu, 2006) and sleeve friction f _s as a measure of soil strength post-failure, Robertson and Wride (1998) proposed a complex cyclic resistance evaluation method. This approach considers in-situ vertical stress, soil behavior type index I _c, and modified cone resistance q _t1N. Juang et al. (2008) simplified the parameters and developed a method using the soil behavior type index I _c,BJ and modified cone resistance q _t1N. Compared to Robertson and Wride’s model, Juang et al.’s approach accounts for the influence of excess pore water pressure, offering insights into soil consolidation and permeability properties (Chai et al., 2011). In contrast, Olsen’s method (Olsen, 1994) does not consider soil type and excess pore water pressure. These three models, based on CPTu tests on terrestrial soils, are summarized in Table 5 . Figure 8 presents the field cyclic resistance ratio (CRR _field) calculated using these three models for four borehole locations. It reveals generally consistent trends in results calculated by each model with depth, albeit with notable differences in numerical values. At borehole locations CPTu 1, CPTu 3, and CPTu 4, Olsen’s model yields the highest CRR _field, followed by the Robertson model, with Juang’s model providing the lowest values. In contrast, at borehole locations CPTu 2, Olsen’s model results in the lowest CRRfield, Juang’s model is intermediate, and the Robertson model calculates the highest values.

The laboratory test conditions are a simplified representation of field conditions. Differences often arise when applying results from laboratory cyclic triaxial tests to field situations. Seed (1979) proposed a conversion factor, C _r, to modify CRR _lab, yielding CRR _field, as demonstrated in the subsequent Equation (3):

where 0.9 is the correction factor for converting the laboratory cyclic resistance ratio under unidirectional loading to the cyclic resistance ratio under multiple-direction loading conditions in the field, C _r is taken as 0.7.

Comparisons of CRR _field, as calculated by various models at respective depths, with CRR _field-lab determined by the test results are depicted in Figure 9 . This figure reveals that Robertson’s and Juang’s approaches exhibit comparable effectiveness, with most errors remaining under 30% relative to CRR _field-lab. However, Olsen’s method, which does not account for soil type and excess pore water pressure, shows the least accuracy, with errors surpassing 60%.

Accordingly, while CRR _field models derived from CPTu data show some applicability, the overall errors are significant. In addition, these models lack a clear functional relationship, potentially limiting their utility in practical engineering projects. As detailed in Section 3, the geotechnical properties of soil layers vary considerably at different depths. Incorporating soil characteristics indicated by q _t, u ₂, and fs, and amalgamating laboratory test outcomes with CPTu data, a discernible functional relationship emerges among the effective cone tip resistance q _E (= q _t − u ₂), f _s, and CRR _field-lab. Thus, an empirical model to predict CRRfield-lab for marine clay was formulated, using q _E, f _s, and H as independent variables, with CRR _field-lab as the dependent variable, as delineated in Equation 4. Notably, H accounts for the in-situ stress of soils. Through multiple regression analysis, a new CRR _field-lab evaluation method based on depth-corrected CPTu index [EXP(q _E/f _s)^0.3/H] was developed, as Equation (4) shown, and its application to assess CRR _field-lab for marine clay in the Yangjiang Wind Farm offshore area has demonstrated a robust fit, as illustrated in Figure 10 .

Noted that the CPTu-based CRR _field-lab prediction model proposed in this study can effectively predict the cyclic resistance of marine clay, which makes up for the difficulty of sampling and high testing costs in offshore engineering. Also, compared with the above three well-developed prediction models (Olsen, 1994; Robertson and Wride, 1998; Juang et al., 2008), the CRR _field-lab prediction model contains just only fewer basial physical parameters, i.e., if the depth H and CPTu data of marine clay are determined, then the CRR _field-lab can be evaluated quickly and efficiently, which provides a significant advantage in the evaluation of liquefaction triggering of marine soils in practice.

This study presents the site investigation and cyclic resistance of marine clay, utilizing CPTu tests and advanced laboratory tests, taking the Shaba wind farm in southern China as a case study. An evaluation method for the cyclic resistance of marine clay, grounded in CPTu data, is introduced. The key conclusions are as follows:

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

QW: Conceptualization, Data curation, Project administration, Resources, Validation, Visualization, Writing – original draft, Writing – review & editing. EZ: Data curation, Investigation, Methodology, Validation, Writing – original draft, Writing – review & editing. XX: Data curation, Investigation, Methodology, Validation, Visualization, Writing – review & editing. YL: Conceptualization, Methodology, Resources, Supervision, Validation, Visualization, Writing – review & editing. GC: Conceptualization, Data curation, Funding acquisition, Supervision, Validation, Visualization, Writing – review & editing.