The fundamental concept and procedures for the basic and detailed design of offshore wind turbine substructures are illustrated in Figure 3 [8]. During the basic design phase, design factors and criteria are established, and an appropriate substructure method [9] is selected based on the soil characteristics of the seabed installation conditions. In the subsequent detailed design stage, detailed structural components are analyzed under environmental and operational conditions to ensure they meet the necessary structural safety requirements. The design of the pin piles used to secure the foundation is also scrutinized, taking into account soil effectiveness.

Figure 3: 
Design flow chart and check items

Before proceeding to the detailed design, reaction force data from operational conditions are utilized to assess the substructure's structural integrity. Prior to finalizing the manufacturing drawings, analysis results for structural members are incorporated, and if any members fail to meet safety requirements, a process of redesign and reinforcement is undertaken.

Following the basic design procedure (as depicted in Figure 3), the specific wind farm and installation seabed site in the South-western Sea of Korea are selected. The loads and moments acting on the primary structural members are computed based on the substructure's basic design. Additionally, a design methodology considering the characteristics of the detailed soil-bearing capacity is developed using information specific to the chosen seabed.

To begin, the self-weight component affecting the jacketed offshore wind turbine structure is divided into the upper structure's weight (nacelle, blade, tower) and the lower structure's weight (see Figure 2). Environmental loads are designed by assessing the structure's safety, taking into account dynamic load components associated with a 50-year return period, including wind, waves, wind-induced generator vibration, and earthquakes.

Wave loads are determined using the Morrison formula, which incorporates a stream function describing particle trajectories in steady wave flows within a sea area. The wave load impacting a structure primarily depends on wave height, which exhibits period-dependent effects. However, establishing a direct relationship between wave height and wave direction proves challenging due to location-specific variations. Consequently, this study employs a wave height of 10.3 m, a wave period of 13.2 s, and a wind speed of 50 m per second for the 50-year return period design wave, calculated through long-term wave analysis. Moreover, meteorological conditions during typhoons are considered.

Earthquake loads are omitted from consideration due to the pin pile's connection to the rock layer, in line with the characteristics of the jacket-type substructure.

Furthermore, the wind load affecting the upper structure induces changes in sea level amplitude as wind speed increases, leading to heightened turbulence intensity. The design accounts for the effect of sea-surface roughness [10], factoring in changes in wind speed.

The structural shape chosen for the substructure is an inclined jacket-type, minimizing displacement and stress at the substructure and connection points during wind turbine installation, and offering optimal dynamic response to loads. This substructure, featuring four pin piles, provides excellent stability and a wide fixation area compared to vertical alternatives.

These advantages facilitate a design that avoids resonance with the wind turbine, while minimizing the surface area exposed to environmental forces such as currents, waves, and wind.

Detailed dimensions and structural members based on the substructure FE model are outlined in Figure 4 and Table 1. Structural simulations were conducted using the SACS computer structural analysis program [11]. All structural steel employed in the construction is high-tensile steel (EH36) with a yield stress of 355 MPa. In the leg and brace connection joints (Cans), localized thickness increases are applied due to in-plane stress concentration phenomena, with modeling results depicted in Figure 4 using member outer diameter information.

Figure 4: 
FE model and diameter of leg and brace members

Loading conditions [12] are categorized into force and moment components, as outlined in Table 2 and Table 3, respectively.

To assess the seabed's bearing capacity at the designated wind farm location in the South-western Sea of Korea, we assumed that the ground characteristics consisted of clay layers extending approximately 17 m from the seabed, transitioning to rocks beyond that depth.

In particular, we examined the limit state conditions for the clay layer, utilizing clay layer stiffness data from the North Sea and Gulf of Mexico (GoM), along with penetration test data from an offshore wind farm in the Southwestern Sea of Korea [13]-[14]. An FE model was applied to simulate the length of the pin pile penetrating the substructure, as depicted in Figure 5. Typically, a pin pile is installed in the soil depth direction, comprising vertical stiffness (K_z), horizontal stiffness (K_x), and rotational stiffness (K_θ), as illustrated in Figure 6. The detailed soil stiffness values applied are presented in Table 4. Notably, it was observed that the coefficient of soil stiffness in the North Sea exceeded that in the Southwest Sea for the same clay layer by more than tenfold. This discrepancy can be attributed to the distinctive soil characteristics found in the target seabed.

Figure 5: 
Substructure model with pin pile

Figure 6: 
Soil application around the pin pile

To account for the corrosion impact of seawater on jacket structural members, we categorized them based on the magnitude of wave load fluctuations [15]. As depicted in Figure 7, we assumed that the inner and outer surfaces of the pipes would equally degrade by 0.1 mm per year from the seabed to a depth of 34.2 m, resulting in a total thickness loss of 2.0 mm. From the water surface to the seabed at a depth of 9.4 m, the inner surface of the pipe was assumed to lose 0.2 mm, while the outer surface would lose 0.4 mm per year. Additionally, we considered the area above the water surface, unaffected by seawater, which would experience a reduced corrosion rate of 2 mm over a span of 20 years.

Figure 7: 
Application of corrosion based on DNV requirement

The maximum wave height was determined to be 19.3 m, based on the 50-year return period wave height for the wind turbine and substructure's design wave conditions. Consequently, in the case of the substructure, we accounted for a 50 mm increase in thickness due to the accumulation of marine sediments below sea level. Additionally, for a conservative structural strength design, we configured the directions of local and environmental loads generated during generator operation to be the same.

Table 5 provides details of five load conditions derived from the return period of the ULS design. It was possible to assess all five combinations and formulate a design that met the criteria mentioned above. Based on the limit states during the load combinations, three ULSs are categorized as presented in Table 6 (a) and (b) correspond to conditions primarily influenced by environmental loads, while Table 6 (c) represents an abnormal wind load situation condition.

Figure 8 illustrates the outcomes of applying current and wave loads to the analysis model at an incident angle of 90°. In the case of wave loads, details such as wave theory, height, and period are provided, with the incident angle's direction indicated by an arrow. Furthermore, the current load was considered, taking into account the velocity gradient from the seabed to the surface, with a maximum speed of 2.2 m/s explained.

Figure 8: 
Current load wave load at the heading angle of 90 degrees

We employed the SACS code, a commercial offshore finite element analysis tool, to model and analyze the structural strength. Figure 9 illustrates the jacket's lower structure, including the four pillars used in our study. The modeling, depicting the structure connecting the wind turbine installed on the upper part and the lower structure, is presented on the left side of Figure 9. The right side of the figure shows the state without a transition piece. In this study, we utilized the reaction force generated during wind turbine operation for load combinations, evaluating the lower structure's strength using the modeling displayed on the left side of Figure 9.

Figure 9: 
Analysis model of the substructure (left: with transition piece, right: without transition piece)

The ends of the four pillars in the substructure incorporate pin piles that penetrate the interior. These pillars are comprised of 232 joints and 291 members, with 1,844 member sections used. We classified single loads into 93 types and derived 193 types of load combinations. The structural members' material properties were considered to be high-tensile steel and mild steel, with detailed information provided in Table 7.

In this study, we applied EUROCODE [16] as the criterion for structural safety evaluation. While EUROCODE is not exclusive to offshore wind turbine substructure design, it enjoys widespread adoption in evaluating the safety of jacket-type structures due to its high reliability.

Among the ultimate limit state design conditions (ULS-(a)), we analyzed the highest unity check (UC) results, specifically 0.995 for the main column member and 0.610 for the brace. These results satisfy the structural strength criteria, with values less than the acceptance criterion of 1.0. The member size and stress result information are summarized in Table 8.

Figure 10 presents a comparison of the maximum combined stress under the pin-pile condition (simply supported condition) and the assumption of fixed ends for the substructure. Similar stress ratios were observed when applying ground-fixing effects of the pin pile and arbitrary boundary conditions. This can be attributed to the substantial confinement effect resulting from the stiffness of the surrounding soil, given the sufficiently deep penetration length of the pin pile.

Figure 10: 
Combined unity ratio according to end fixity

To assess the substructures' structural strength based on boundary conditions, Figure 10 and Table 9 compare the analysis results when fixed conditions were applied, incorporating the stiffness of three clay layers. The maximum U.C. of the vertical column yielded similar results to those when the pin pile was applied. The highest values were observed in the North Sea. Without the pin pile, the evaluation was 1% smaller, based on the maximum U.C.

Figure 11: 
Results of maximum combined UC according to soil stiffness

Considering these results, we acknowledge the challenge of obtaining geological data during the initial design and the considerable experience and time required for nonlinear soil analysis. Consequently, in the initial weight review, we determined that substituting the effect of the pin pile with a fixed condition, based on the findings of this study, is sufficient to establish the steel's order.