An Innovative Hybrid Structural Retrofit Strategy for Onshore Wind Turbine Repowering

Abstract

This article proposes and validates a hybrid structural reinforcement strategy for onshore wind turbine foundations in repowering projects, enabling the installation of higher-capacity units without demolishing the existing foundation. In a context of increasing demand for renewable energy and infrastructure optimization, the original foundation is reused as the primary element for global stability and serviceability limit state (SLS) requirements, while ultimate limit state (ULS) demands, arising from the replacement of approximately 1.5 MW turbines with 4.1 MW and 6.25 MW units with power ratings representative of various manufacturers’ models in the current market are resisted by a new peripheral reinforced concrete strengthening system. The study considers both shallow (gravity) and piled foundation typologies, which are the most common globally for wind turbines. This solution, applied to a commercially operating wind farm in southern Brazil with actual load data, demonstrated a substantial reduction in concrete volume–up to 80% for shallow foundations and 40% for piled foundations compared to constructing an entirely new foundation. Structural assessment was performed through numerical modeling in SAP2000, employing a shell-beam hybrid model validated against a 3D solid reference, combined with analytical verifications of limit states. Results confirm that the proposed solution ensures global serviceability and adequate ultimate limit state capacity, achieving significant material optimization. This offers a sustainable and efficient alternative for repowering wind turbine foundations, with notable economic and environmental benefits, including the elimination of demolition, transportation, and material disposal costs.

1. Introduction

As the operational lifespan of many wind farms approaches its end, countries will increasingly face decisions regarding whether to repower existing facilities, fully decommission them, install new turbines, or implement combined strategies [1]. According to the Global Wind Energy Council, global installed wind power capacity in 2005 was approximately 59 GW, representing the early stage of utility-scale deployment and defining the cohort of assets that have now exceeded 20 years of operation [2]. Commercial wind turbines were originally designed for an operational lifespan of approximately 20–25 years. Beyond this period, an irreversible performance decline occurs, and their generation capacity becomes significantly lower compared to that of modern turbines [3]. Recent studies emphasize that life extension and repowering are critical for maintaining energy security and economic viability, as observed in European markets [4], where multi-criteria decision analyses highlight the strategic value of reusing existing infrastructure [5].

Partial or complete repowering has emerged as a strategic approach to maximizing the exploitation of high-quality wind sites through the deployment of modern turbine technologies, thereby increasing energy production while reducing the number of installed wind turbines [6]. This transition frequently involves upgrading to higher-capacity turbines, which significantly alters the load profile on the supporting structures [7]. Recent comparative analyses between decommissioning and repowering demonstrate that extending the life of existing wind farms offers substantial economic and environmental benefits, reducing the carbon footprint associated with new construction [8]. Consequently, evaluating the feasibility of fully or partially reusing existing foundations when new turbines are installed at the same locations has become a necessity [9]. This approach can reduce costs and environmental impacts while extending infrastructure service life, aligning with circular-economy principles [10].

The reuse of existing foundations in repowering projects has become a promising strategy for reducing emissions, minimizing construction activities, and limiting waste generation. However, upgrading to modern, higher-capacity turbines significantly increases the aerodynamic loads and overturning moments applied to the base. These massive moments tend to destabilize the structure, making global stability the most critical aspect of the assessment. For shallow onshore foundations, uplift prevention and uplift limitation criteria govern the global stability design [11]. In the case of piled foundations, design is restricted by the need to prevent pile tension under service conditions. While satisfying the normative criteria of “no gapping” and “no pile tension” is a primary challenge that often renders the original geometry inadequate [12], a comprehensive repowering assessment cannot be limited to global stability alone. The foundation must also be rigorously evaluated and appropriately reinforced to withstand the significantly higher internal forces, such as bending moments and shear stresses, imposed by modern turbines under Ultimate Limit State (ULS) conditions. Because these internal forces are directly governed by the soil–structure interaction under extreme eccentric loading, capturing the true base pressure distribution is critical. To accurately predict the distribution of stresses and deformations in these scenarios, advanced modeling techniques, such as the use of spring models for the vertical reaction of the soil, are essential [13]. These models are applicable to both shallow and piled foundations, such as those found in existing wind farms, allowing engineers to verify whether the existing foundation can safely operate for an extended lifespan of 20 years without reinforcement [14].

Current onshore wind turbine foundations typically require very large material quantities, often exceeding 400 m³ of concrete, generally in grades such as C30/37, C35/45, or C40/50, with higher-strength concrete commonly used in the pedestal, and more than 50 tons of reinforcing steel [10]. Standards such as IEC 61400-1 [15], IEC 61400-6 [12], DNVGL-ST-0126 [16], and DNV/Risø [17] provide detailed guidance for new foundations but do not offer specific recommendations for strengthening existing foundations in repowering contexts. Furthermore, the structural design and verification of these concrete elements must comply with rigorous international codes, notably CEN EN 1992-1-1:2004 (Eurocode 2) [18], to ensure structural integrity under the increased ultimate limit state (ULS) and serviceability limit state (SLS) actions.

Despite increasing interest in the topic, important gaps remain regarding the structural behavior of existing foundations subjected to increased loading from higher-capacity turbines. Park-scale studies tend to prioritize energy and economic aspects [19], while classical engineering research primarily addresses the design of new foundations. Notably, Waldron et al. [20] provided one of the few detailed engineering assessments featuring actual structural designs for re-engineered foundations adapted for repowering. Their study demonstrated that traditional retrofitting methods such as expanding the foundation’s footprint or depth require massive increments in concrete and steel volumes. In fact, the material required for these reinforcements often equals or exceeds that of an entirely new foundation, significantly driving up construction costs and environmental impacts, thereby undermining their practical applicability. Recent investigations into strengthening methods have focused specifically on the turbine–foundation interface, such as retrofitting embedded-ring connections with additional anchor bolts [21]. While these methods increase load capacity, they often require complex and highly invasive interventions at the connection level, highlighting the need for simpler and more practical interface adaptations during repowering. Furthermore, optimization strategies for repowering highlight the need for solutions that balance structural reliability with material efficiency [10].

In this context, there is a clear need for practical retrofitting strategies that enable repowering with higher-capacity turbines while avoiding the material, cost, and environmental penalties associated with full foundation replacement. No studies have been identified that: (i) explicitly propose a strengthening strategy that fully preserves the existing foundation as a stabilizing mass to meet the “no tension” and “no gapping” criteria; (ii) introduce a supplementary reinforced-concrete system capable of resisting ULS actions associated with repowering; and (iii) demonstrate, through real-case applications, that the additional material volume can be lower than that of an equivalent new foundation, thereby avoiding demolition of the original structure.

This study proposes and evaluates a hybrid reinforcement strategy in which the existing foundation is retained to ensure global stability and serviceability, and a new peripheral reinforced concrete system is designed to resist all ULS actions. The goal is to demonstrate the benefits of continued generation through repowering compared to decommissioning, verifying whether the existing foundations can safely operate for an extended lifespan or if they require the proposed hybrid reinforcement. Demonstrating, on a real wind farm in southern Brazil, that this configuration safely accommodates repowering with less additional material than an equivalent new foundation highlights its technical relevance, economic benefits, and sustainability potential.

2. Materials and Methods

The methodology adopted in this study provides a comprehensive framework for evaluating the structural and geotechnical feasibility of repowering onshore wind turbines. Based on data from an existing wind farm in southern Brazil, the assessment considers two prevalent foundation typologies: a shallow gravity foundation and a piled foundation. Although this research is based on a specific case study, both foundations feature a circular geometry which, alongside the shallow and piled typologies, represents most onshore wind turbine installations worldwide [22]. This inherent characteristic ensures that the proposed solutions and findings are highly representative of global repowering challenges. The primary objective is to determine whether these existing structures can safely operate for an extended lifespan of 20 years under the significantly increased loads of two specific modern turbines with capacities of 4.1 MW and 6.25 MW.

As illustrated in the flowchart below, the process begins with a comprehensive verification of the original foundation against Ultimate Limit State (ULS), Serviceability Limit State (SLS), and global stability criteria. The structural analyses were performed using SAP2000 (v.26, Computers and Structures, Inc. (CSI), Berkeley, CA, USA). Complementary structural verifications and the geotechnical assessment were conducted using analytical formulations. If the existing structure fails to meet these normative requirements, particularly the “no gapping” and “no pile tension” conditions, the proposed hybrid reinforcement strategy is applied. This strategy retains the existing concrete block to ensure global stability while introducing a new peripheral reinforced concrete system to resist the increased ULS actions. Ultimately, the methodology culminates in a comparative assessment between repowering and decommissioning, highlighting the material and environmental benefits of the proposed solution. Figure 1 presents the complete flowchart of this methodological approach.

2.1. Materials of the Existing Foundation and Reinforcement

The original foundation is in adequate service condition, with no degradations affecting its integrity. The original foundation utilizes A500 steel and C30/37 class concrete, whereas the reinforcement system (main shaft, radial beams, and transition section) uses A500 rebar and C40/50 class concrete (

Table 1. Material properties.

Material Properties
Concrete	C30/37	C40/50
Characteristic compressive strength fck, (MPa)	30	40
Approx. cube strength fck cube, (MPa)	37	47
Mean compressive strength fcm, (MPa)	38	48
Young’s modulus Ecm, (GPa)	33	36
Mean tensile strength fctm, (MPa)	2.9	3.51
Characteristic tensile strength fctk 0.05, (MPa)	2.03	2.46
Unit weight, (kN/m3)	25	25
Poisson’s ratio ν	0.2	0.2
Rebar
Characteristic yield strength fyk, (MPa)	500	-
Young’s modulus Es, (GPa)	210	-
Poisson’s ratio ν	0.3	-
Unit weight, ${\gamma }_b$, (kN/m3)	77	-

).

Site characterization was determined from a comprehensive campaign across the wind farm, which comprises 62 existing foundations (58 shallow and 4 piled). For this study, the two most unfavorable profiles (one for shallow and one for piled foundations) were selected, integrating standard penetration tests within the soil layers and rotary coring in the underlying rock. This ensures that the geotechnical parameters used represent the most critical conditions within the park, making the proposed solution applicable to all 62 units. At the location of the existing shallow foundation, the profile reveals a heterogeneous sequence of sandy silts and partially weathered residual soils overlying competent bedrock encountered at approximately 2–3 m depth. Additionally, plate load tests were performed at the foundation settlement depth for all 58 shallow units [23]. The geotechnical parameters adopted in this analysis are directly based on these field test results, providing an adequate representation of the site conditions. This condition is typical of onshore wind turbine sites situated on elevated and exposed terrain, where shallow rock surfaces are frequent due to intense weathering. At the location considered for the deep (piled) foundation alternative, similar surficial materials were identified; however, the competent rock material is encountered only at greater depths, between approximately 11 and 13 m, thus justifying the use of piles for a new foundation solution. The parameters adopted in this study, therefore, represent the minimum geotechnical information required for bearing capacity, subgrade reaction, and global stability verifications associated with the direct and indirect foundation configurations. These geotechnical parameters are incorporated into numerical and analytical checks only to the extent required to verify the criteria of ‘no gapping’ and ‘no pile tension’ as prescribed in IEC 61400-6 [12].

2.2. Geometry of Existing and Reinforced Foundations

2.2.1. Existing Foundations

The two existing foundation types evaluated have a circular truncated cone shape, 15 m in diameter, 2 m shaft height, and 1.3 m edge thickness, totaling 298 m³ of concrete and 30 t of steel, using materials as previously specified. The tower is connected by a metallic flange. The shallow (gravity) foundation is equipped with 16 active anchors, each with 400 kN capacity, while the piled foundation consists of 32 piles, 300 mm in diameter and 12 m deep. Figure 2, show the construction stages of the analyzed existing wind turbine foundation.

2.2.2. Reinforced Concrete Strengthening System

The proposed retrofit comprises a main shaft, 3.5 m tall and 6 m in diameter, connected to 8 radial beams spaced at 45° intervals. For shallow foundations, the beams rest on a circular footing; for piled foundations, on 8 pile caps. This layout enlarges the foundation lever arm and concentrates ULS demands in the new reinforcement. Figure 3 presents the geometry of the shallow and deep foundations.

Figure 4 shows plan views of the piled (left) and shallow (right) foundations. Both share identical shaft, beam, and transition geometries, differing only in their support structures: pile caps versus a continuous footing.

2.2.3. Monolithic Action

Interfaces between new and existing concrete are common in structural systems cast in multiple stages, in monolithic precast assemblies, and in repaired components where concrete is placed at different ages [24]. The connection between the new shaft and the existing foundation shaft was assumed to act as a fully bonded interface, an assumption widely accepted in strengthening interventions and supported by classic experimental research [25,26], as well as more recent studies on old–new concrete interfaces [27,28]. To ensure adequate force transfer, the interface surface was roughened by light scabbling, and drilled holes were used to install anchored bars bonded with epoxy adhesive, thereby providing the required mechanical continuity between the existing and newly cast concrete.

2.3. Turbine Loads and Load Combinations

Onshore wind turbines are continuously subjected to repeated overturning forces throughout their service life, resulting from wind loads acting from any direction (360°). Because these wind loads occur randomly in both space and time, they are classified as a variable action and are therefore most appropriately represented using statistical descriptions [29].

Loads are based on actual data provided by the turbine manufacturers: Goldwind GW 77/1500 (China), existing foundation, WEG AGW147/4100 (Brazil), and Mingyang MySE 172/6250 (China). Each turbine provides axial forces, horizontal forces, overturning, and torsional moments, under both service (serviceability limit state [SLS]) and ultimate (ultimate limit state [ULS]) conditions. As per IEC 61400-1 [15], extreme loads are factored and combined with aerodynamic, operational, and transient actions. These factored values are applied directly to the numerical model, respecting the manufacturer defined load points.

Table 2. Characteristic loads applied to the studied wind turbines.

Load, Set (Turbine)	Limit State	Safety Factor	Loads
			Axial Forces (kN)	Horizontal Forces (kN)	Torsional Moment (kN.m)	Overturning (kN.m)
I-GW77	SLS	1.00	2972	302	949	24,175
II-GW77	ULS	1.35	2918	629	681	59,010
III-AGW147	SLS	1.00	6930	721	3080	76,200
IV-AGW147	ULS	1.35	9410	1146	7047	120,015
V-MySE172	SLS	1.00	6008	966	747	84,411
VI-MySE172	ULS	1.35	6139	1819	14,873	173,544

presents the loads for both foundations along with the corresponding safety factors.

2.4. Design Codes and Verification

Structural and geotechnical verifications for the original and reinforced foundations were based on IEC 61400-1, IEC 61400-6, DNVGL-ST-0126, and Eurocode 2, alongside classic wind turbine foundation formulations [12,15,16,18]. Specifically, IEC 61400-6 served as the primary guideline for global stability and Serviceability Limit State (SLS) criteria, enforcing ‘no tension’ conditions for piled foundations and ‘no gapping’ for shallow units. DNVGL-ST-0126 provided supplementary limits for soil–structure interaction and deformation. For the reinforced concrete retrofit, Eurocode 2 was applied to size the reinforcement and verify critical sections under the Ultimate Limit State (ULS), as well as to evaluate cracking and fatigue under the Serviceability Limit State (SLS). Consequently, the existing foundation is utilized primarily for its mass, acting in conjunction with the hybrid reinforcement to satisfy the ‘no gapping’ and ‘no pile tension’ criteria. Meanwhile, the proposed new structural system is designed to independently resist all Ultimate Limit State (ULS) loads, undergoing comprehensive SLS and ULS evaluations. Only the standards and expressions directly applied in these analyses are detailed in this section.

For shallow gravity foundations, the concrete block is commonly modeled as a rigid body resting on a circular base with diameter D, radius R = D/2, and area A = πR². When subjected to an axial load and an overturning moment about its centroid, the soil contact pressure is assumed to follow a linear distribution, consistent with classical rigid body mechanics. The maximum contact pressure is expressed by Equation (1):

$${\sigma }_{max}={\displaystyle \frac{N}{A}}\pm {\displaystyle \frac{M_R}{\left(\frac{\pi \ast R^3}{4}\right)}} \le {\displaystyle \frac{q_{ult}}{\gamma F}}$$

(1)

where ${\sigma }_{max}$ is the maximum soil pressure, N is the applied vertical load, $M_R$ is the overturning resultant moment, $q_{ult}$ is the ultimate bearing resistance of the soil, and $\gamma F$ is the geotechnical partial safety factor.

The bearing capacity adopted herein was determined from plate load tests performed at the foundation bearing level. This standardized procedure, following guidelines such as [23], consists of applying incremental static loads to a rigid plate under slow loading and unloading cycles, recording the corresponding deflections and settlements. The interpretation of these results and their extrapolation to the in situ behavior of the circular wind turbine foundation were based on established methodologies from the geotechnical literature [30,31]. From the ultimate bearing capacity obtained in the test, an appropriate safety factor is applied to determine the design bearing capacity. These values are then utilized for the verification of soil–structure interface stresses and to obtain fundamental parameters, such as the vertical soil reaction.

For shallow foundations under the Serviceability Limit State (SLS), the ‘no gapping’ condition requires full compression across the soil–structure interface, which is satisfied when the load eccentricity is less than one-quarter of the foundation radius (Equation (2)). Under the Ultimate Limit State (ULS), this full compression constraint is relaxed; instead, it is required that at least 50% of the foundation area remains in contact with the soil, applying the same eccentricity-based formulation (Equation (3)):

$${\displaystyle \frac{M}{N}}=e\le 0.25\ast R$$

(2)

$${\displaystyle \frac{M}{N}}=e\le 0.25\ast \pi \ast R$$

(3)

Overturning stability is evaluated according to IEC 61400-6 [12] by comparing the design stabilizing moment with the design overturning moment. The destabilizing moment includes wind actions and all horizontal and torsional effects, applying unfavorable partial safety factors. Conversely, the stabilizing moment is derived from the total gravity load (including the foundation, turbine, and backfill contributions) multiplied by the foundation radius, using favorable resistance factors. The foundation is considered stable when the ratio between the stabilizing and overturning moments exceeds unity under ULS conditions (Equation (4)):

$${\gamma }_{RD}={\displaystyle \frac{M_{d,stabilizing}}{M_{d,overturning}}}$$

(4)

where $M_{d,stabilizing}$ is the design stabilizing moment and $M_{d,overturning}$ is the design overturning moment and the criterion is satisfied when ${\gamma }_{RD}$ > 1.

Sliding stability is assessed by comparing the design horizontal load at the soil–foundation interface with the design sliding resistance, in accordance with IEC 61400-6 [12]. The design shear stress is determined from the horizontal action using unfavorable partial factors, while the sliding resistance is calculated based on the soil shear capacity incorporating favorable resistance factors. Stability under ULS is ensured when the design shear demand does not exceed the available resistance (Equation (5)):

$${\tau }_{Ed}={\displaystyle \frac{H_d}{A}}<{\tau }_{Rd}$$

(5)

where $H_d$ is the design horizontal load acting at the interface, A is the effective contact area around the resultant vertical load, ${\tau }_{Ed}$ is the design shear stress, ${\tau }_{Rd}$ is the design sliding resistance.

For piled foundations, the pile group is generally idealized as a rigid section formed by n piles uniformly arranged at a constant radius around the foundation center. When the system is subjected to a vertical load $N$ and a resultant overturning moment $M_R$, and assuming that plane sections remain plane, the axial load in the i-th pile located at coordinate $z_i$ along the bending direction can be expressed by Equation (6):

$$N_i={\displaystyle \frac{N}{n}}\mp {\displaystyle \frac{M_R\ast z_i}{{\sum }_{j}z2}}$$

(6)

In this formulation, $N_i$ represents the axial load acting on pile i, $N$ is the total applied vertical load, $n$ is the number of piles in the group, $M_R$ is the resultant bending moment acting on the foundation, $z_i$ is the position of pile I measured along the moment axis, and ${\sum }_{j}z2$ corresponds to the second geometric moment of the pile layout, representing the contribution of all pile positions to resisting the applied moment. The “no tension” condition for the pile group, as per IEC 61400-6, is checked as follows (Equation (7)):

$$N_{mín}\ge 0$$

(7)

For pile foundations, the total axial resistance is the sum of the base resistance and the shaft resistance mobilized along the embedded length. The characteristic base resistance and the characteristic shaft resistance are expressed by (Equations (8) and (9)):

$$R_{b,k}=q_{b,k}\ast A_b$$

(8)

$$R_{s,k}=\mathsf{\Sigma } (q_{s,k}\ast A_i)$$

(9)

where $R_{b,k}$ is the characteristic base resistance, $q_{b,k}$ is the characteristic unit base resistance acting at the pile tip, and $A_b$ is the area of the pile base. $R_{s,k}$ denotes the characteristic shaft resistance, Σ represents the contribution of all soil layers intersected by the pile, $q_{s,k}$ is the characteristic unit shaft resistance in soil layer i, and $A_i$ is the lateral surface area of the pile in contact with that layer.

Design of the structural reinforcement elements (i.e., the main shaft, radial beams, and transition section) was performed for ULS considering global actions derived from the numerical model. All sections were assumed to behave linearly elastically up to the resistance limit. The radial beams are treated as reinforced concrete beams in simple flexure, evaluated under ULS per Eurocode 2. The objective is to ensure sufficient flexural (Equation (10)) and shear capacity, with geometry predetermined (Equation (11)):

$$A_s={\displaystyle \frac{M_{Ed}}{(z\ast f_{yd})}}$$

(10)

$$V_{Rd,c}=\left[C_{Rd,c}\ast k\ast {\left(100\ast \rho l\ast f_{ck}\right)}^{{\displaystyle \frac{1}{3}}}+k_1\ast {\sigma }_{cp}\right]\ast b_w\ast d$$

(11)

where $A_s$ denotes the required tensile steel area, $M_{Ed}$ is the design bending moment, $z$ is the internal lever arm, and $f_{yd}$ is the design yield strength of the reinforcement. The shear resistance expression includes the design concrete shear capacity $V_{Rd,c}$, the shear coefficient $C_{Rd,c}$, the size-effect factor $k$, the longitudinal reinforcement ratio $\rho l$, and the concrete compressive strength $f_{ck}$. The term $k_1$ accounts for the contribution of the mean compressive stress ${\sigma }_{cp}$ arising from the axial load, while $b_w$ and $d$ are the effective web width and effective depth of the section, respectively.

The transition section is checked at ULS as a compressed and bent member, following Eurocode 2 [18] standards, based on the N-M interaction diagram representing simultaneous compressive and flexural resistance. The normal force and overturning moment from the numerical model are checked at selection section to ensure the representative point falls within the resistance envelope. The geometry and longitudinal/transverse reinforcement are sized to ensure sufficient flexural, compressive, and global stability (Equation (12)):

$${\displaystyle \frac{N_{Ed}}{N_{Rd\left(x\right)}}}+{\displaystyle \frac{M_{Ed}}{M_{Rd\left(x\right)}}}\le 1$$

(12)

where $N_{Ed}$ is the design axial force acting on the section; $N_{Rd}$ is the design axial resistance; $M_{Ed}$ is the design bending moment; $M_{Rd\left(x\right)}$ is the design flexural resistance.

Serviceability limit state verifications for the main shaft, radial beams, and transition section were conducted. However, due to their geometry and the load levels obtained from the numerical model, no SLS check was critical for final sizing. The following checks are presented for code completeness. Crack width is limited per Eurocode 2 to ensure durability and prevent reinforcement corrosion (Equation (13)):

$$w_k=S_{r,max}\ast \left({ϵ}_{sm}-{ϵ}_{cm}\right)\le 0.30 \mathrm{m}\mathrm{m}$$

(13)

where $w_k$ is the characteristic crack width in the serviceability limit state, $S_{r,max}$ is the maximum crack spacing determined by the steel-concrete bond-mechanics model, ${ϵ}_{sm}$ is the mean strain in the tensile reinforcement after crack formation, and ${ϵ}_{cm}$ is the mean concrete strain between cracks, accounting for the combined effects of shrinkage, creep, and stress transfer through bond. Concrete fatigue was assessed by typical EC2/classical relations (Equation (14)):

$${\displaystyle \frac{{\sigma }_{c,max}}{f_{cd,fat}}}\le 0.5+0.45\ast {\displaystyle \frac{{\sigma }_{c,min}}{f_{cd,fat}}}\le 0.9 for f_{ck}\le 50 \mathrm{M}\mathrm{P}\mathrm{a}$$

(14)

where ${\sigma }_{c,max}$ is the maximum compressive stress in the concrete under the considered fatigue load combination, ${\sigma }_{c,min}$ is the corresponding minimum compressive stress in the same loading cycle. The variable $f_{cd,fat}$ corresponds to the design compressive strength of concrete for fatigue verification, incorporating the appropriate material partial factor and the fatigue reduction coefficient applied to the characteristic compressive strength $f_{ck}$. The upper limit of 0.9 is applicable for concretes with $f_{ck}\le 50 \mathrm{M}\mathrm{P}\mathrm{a}$, as prescribed by the fatigue criteria for concrete in the governing standards. Fatigue in reinforcement is checked against the stress range (Equation (15)):

$${\Delta }_{\sigma s}\le {\Delta }_{\sigma Rsk}=70 \mathrm{M}\mathrm{P}\mathrm{a}$$

(15)

where ${\Delta }_{\sigma s}$ is the variation in steel stress induced by the cyclic load range considered in the fatigue verification and ${\Delta }_{\sigma Rsk}$ is the characteristic fatigue stress range capacity of the reinforcing steel. For conventional reinforcing steel, the standard limit for the characteristic fatigue stress range is ${\Delta }_{\sigma Rsk}=70 \mathrm{M}\mathrm{P}\mathrm{a}$, which establishes the upper admissible difference between the maximum and minimum steel stresses within each load cycle.

2.5. Numerical Modeling Strategy

The reinforcement analysis was carried out using the finite element method with a hybrid modeling strategy that combines shell (membrane) and frame (beam) elements. This approach provides a balance between geometric fidelity and computational efficiency, as full solid modeling would lead to prohibitive numerical cost. Shell elements were used to model the laminar surfaces of the existing foundation (footing or block) and the new peripheral supporting component (continuous footing or pile caps). Based on the Reissner-Mindlin formulation, these shell finite elements provide six degrees of freedom per node and capture the coupled bending and in-plane membrane behavior that governs foundation response under large bending moments. Specifically, thick shell elements were adopted to accurately represent the physical characteristics of the footing. This formulation allows for the explicit definition of the foundation’s thickness and the concrete reinforcement cover parameters, ensuring a highly realistic representation of the structural section [32]. The discretization level of the finite element mesh is directly related to the accuracy of the numerical solution: higher mesh refinement improves result precision but increases computational demand [33]. Figure 5 presents the hybrid models adopted for both foundation configurations.

Simultaneously, frame elements are utilized to represent the eight radial beams and transition sections. These elements are ideal for components dominated by their longitudinal dimension, as they directly supply internal forces (bending moments and shear). Within the hybrid model, the rigid connection between the shell elements (representing the block) and the frame elements (representing the radial beams) ensures the correct force transfer between the main shaft and the peripheral supports. This approach enables a faithful representation of the complex 3D reinforcement geometry while drastically reducing computation time compared to full solid modeling.

To validate this approach, a secondary model utilizing solid (volumetric) elements typically eight-node hexahedra was developed. While these solid elements yield the most accurate representation of the structural geometry and behavior, they are computationally expensive and were therefore used primarily to calibrate and verify the efficiency of the hybrid shell-and-frame model. Unlike shell elements, which represent the behavior through a median plane, the solid element models the entire spatial configuration. Its degrees of freedom are exclusively translational (three per node), without rotations. While the implemented model considers concrete as a homogeneous material (direction-independent properties), the solid element formulation allows modulus variations by direction, making it suitable for studying anisotropic materials such as composites.

3. Results

This section presents the validation of the numerical model, followed by the comprehensive structural and geotechnical assessment of the reinforced foundations under the upgraded turbine loads. While the original foundations were validated for their initial 20-year design life without apparent pathologies, this study focuses exclusively on the performance of the proposed structural reinforcement (for both shallow and piled typologies) under the more demanding loads.

To illustrate the structural response of the reinforced system under these critical loading scenarios, Figure 6 details the stress distribution across the concrete elements. The figure highlights the stress concentrations within the new radial beams and the central pedestal, particularly at the interface where the anchor cage Section (TFS) is embedded, reflecting the primary load transfer path.

3.1. Hybrid Numerical Model Validation

Results showed that the hybrid model reproduces the structural behavior of the reinforced foundation with high accuracy. Maximum differences in vertical and horizontal displacements are within acceptable limits for structural engineering analysis. The following comparison illustrates the critical interface between the main shaft and the most heavily loaded radial beam: the S11 stress value was 3.08 MPa (compression) and −1.5 MPa (tension), with the hybrid beam element at 2355.15 kN·m. Linearizing the critical section, converting S11 stresses into overturning moments, demonstrated that the moment from this analysis matches the beam (frame) element value in the hybrid model extremely closely, with variation under 0.5%, (solid 2364.58 kN.m vs. 2355.15 kN.m), confirming the hybrid model’s fidelity in representing flexural behavior and supports its further verification.

Table 3. Comparison of equivalent moments (solid).

Section	W (m3)	σtop (MPa)	σbot (MPa)	σb (MPa)	M (kN.m)
Critical	1.0417	3.08	−1.5	2.27	2364.58

presents the linearization of the stresses obtained in the solid model for the worst-case scenario at the junction between the shaft and the most heavily loaded beam, where the bending moments at this location were determined.

The 3D solid model is presented below, Figure 7, illustrating the stresses that develop at the interface between the tower shaft and the perimeter beam. These results provided the stress values used for comparison with the hybrid model shown in Figure 8.

In the hybrid model, Figure 8 shows that the perimeter beam is represented using a combined shell-and-frame discretization. In this approach, the shell elements capture the flexural and membrane behavior of the concrete section, while the frame elements model the global stiffness and load-transfer mechanisms along the beam axis. This modeling strategy enables a consistent representation of the bending and shear fields while achieving greater computational efficiency compared to a full 3D solid formulation. Among all the perimeter beams, the most heavily loaded one was selected for detailed stress evaluation. For this beam, the internal forces and stress resultants from the shell-frame interaction were extracted at the critical cross-section near the shaft–beam interface, enabling a direct comparison with the stresses computed in the 3D solid model.

Moreover, the hybrid model proved vastly more computationally efficient, reducing processing time by approximately 92% compared to the solid model. This reduction was directly observed from the processing time reports generated by SAP2000 during the final processing run of the models, where the shell-beam model completed calculations in 28 s, whereas the full solid model required 350 s. This efficiency renders it suitable for parametric and sensitivity studies without sacrificing reliability.

3.2. Structural and Geotechnical Assessment

A sensitivity analysis was conducted to evaluate the hybrid structural solution under the 4.1 and 6.25 MW turbine loads, using the previously validated numerical model. The following results summarize the behavior of the foundation soil, piles, and reinforcement elements under higher-than-original design loads.

3.2.1. Soil Stress vs. Bearing Capacity

Contact stress at the foundation-soil interface was compared to the bearing capacity derived from the plate load test performed at the founding depth [30,31]. For all load cases, maximum stress remained below the allowable value, validating the performance of the shallow foundation at ULS and confirming that the reinforcement does not induce critical pressure increases.

Table 4. Plate load test results for soil parameters.

Parameter	Value
Plate diameter, D (m)	0.80
Applied load, Q (kN)	402
Measured settlement, δ (m)	0.0025
Plate area, A (m2)	0.5026
Applied contact pressure, σapplied (kN/m2)	799.75
Vertical subgrade reaction modulus kv (kN/m3)	319,901
Horizontal subgrade reaction modulus kh (kN/m3)	86,185
Allowable bearing pressure, qadm (kN/m2)	400
Soil elastic modulus (ν = 0.3), Es (MPa)	141

lists the input parameters and results of the static plate load test used to determine the soil bearing capacity. The test reached a bearing pressure of 800 kN/m² with a settlement of 2.5 mm. Applying a global safety factor of 2 yields an allowable bearing pressure of 400 kN/m², adopted as the design value (q_adm = 400 kN/m²).

Table 5. Site-applied soil stresses from wind turbine loads.

LS Type	σmax (Soil)	σmin (Soil)	σmax (Soil)	σmin (Soil)
	WEG	WEG	Mingyang	Mingyang
Ultimate limit state	240.88	−49.94	330.42	−113.28
Serviceability limit state	177.07	0.88	196.90	0.26

summarizes the maximum soil contact stresses obtained for both turbine foundations. For the Mingyang turbine, the foundation develops a maximum compressive stress of 330.42 kN/m², while the WEG turbine reaches 240.88 kN/m². Both values remain below the allowable bearing pressure of 400 kN/m², confirming that the ultimate limit state is satisfied with adequate geotechnical safety margins. Under serviceability conditions, both foundations remain entirely in compression, indicating stable load transfer and proper performance of the soil-foundation system.

3.2.2. “No Gapping” Verification in the Shallow Foundation

For the shallow foundation under serviceability conditions,

Table 6. Results of the geotechnical stability assessment.

LS Type	Overturning	Sliding	Eccentricity		Overturning	Sliding	Eccentricity
	FS	FS	Limit (m)	(m)	FS	FS	Limit (m)	(m)
	WEG				Mingyang
Ultimate limit state	2.63	8.84	8.25	4.00	1.96	6.54	8.25	5.36
Serviceability limit state	4.04	14.88	2.63	2.60	3.99	12.27	2.63	2.63

indicates eccentricities of 2.60 m for the WEG turbine and 2.63 m for the Mingyang turbine, both at or below the allowable limit of 2.63 m, thereby ensuring full soil compression with no tensile zones. Under ultimate limit state loading, the eccentricities increase to 4.00 m and 5.36 m, respectively, remaining well below the 8.00 m limit and maintaining more than 50% of the contact area in compression. Stability checks for overturning and sliding also resulted in safety factors exceeding the minimum values required by current design standards.

3.2.3. Pile Tension Verification

For the deep foundation, the axial response of the pile cap was assessed under both loading scenarios. Under serviceability conditions, the pile cap remains fully in compression, with no tensile demand transferred to the piles. At the ultimate limit state, the pile axial capacity is governed by the structural resistance of 1413.17 kN, while the geotechnical resistance is 1661.67. The most heavily loaded pile in the Mingyang foundation reaches 1378.20 kN, thereby satisfying the axial capacity criterion with an adequate safety margin (

Table 7. Pile loads for new wind turbines.

LS Type	Compression (kN)	Tension (kN)	Compression (kN)	Tension (kN)
	WEG		Mingyang
Ultimate limit state block	6607	−737	6891	−2672
Serviceability limit state block	5176	-	5326	-
Ultimate limit state pile	1321.4	−147.4	1378.2	−534.4

).

3.2.4. Ultimate Limit State Checks for Reinforced Concrete Elements

The structural performance of the reinforcement elements—namely the central pedestal, radial beams, and transition sections—was evaluated at the Ultimate Limit State (ULS) using the internal forces extracted from the numerical model. Regarding the central pedestal, the contact stresses at the tower–foundation interface strictly comply with the IEC 61400-6 [12] limits. Specifically, the stresses remain below 0.60 fck for characteristic load combinations and 0.45 fck for permanent loads, operating well within the maximum resistance thresholds. Furthermore, the radial beams were evaluated as reinforced concrete members subjected to flexure and shear, in accordance with Eurocode 2. The analysis confirms that these elements maintain significant design margins, with the selected reinforcement ratios exceeding the minimum code requirements to ensure adequate safety. Finally, Figure 9 highlights the two most heavily loaded radial beams, which effectively form a load couple to resist the resultant overturning moment.

Table 8. Bending moments and shear forces results for new wind turbines.

Type	Turbine	Beam	Bending Moments (kN.m)	Shear Forces (kN)
Shallow	WEG	01	+19,001.94	+3862.00
Shallow	WEG	05	−4312.57	−826.80
Deep	WEG	01	+19,107.00	+3865.00
Deep	WEG	05	−4305.00	−829.00
Shallow	MY	01	+25,074.00	+4950.65
Shallow	MY	05	−10,502.00	−1939.00
Deep	MY	01	+25,204.00	+4955.00
Deep	MY	05	−10,520.00	−1944.00

summarizes the bending moments and shear forces obtained for the most highly loaded peripheral beams, 01 and 05, in both shallow and deep foundation configurations for the WEG and Mingyang turbines. The results highlight the distinct structural demand associated with each turbine model and foundation type, with beam 1 and 5 consistently exhibiting the governing internal forces.

Due to the high loads imposed by large onshore wind turbines, the connection between the central pedestal and the peripheral beams experiences significant stress concentrations. Consequently, the beam height cannot be substantially reduced. This decision is primarily driven by geometric detailing requirements and the need to manage stress concentrations effectively at the pedestal connection, rather than solely by structural capacity. The resulting large structural depth provides both high bending stiffness and sufficient cross-sectional area for steel reinforcement, resulting in very high flexural (MRd) and shear (VRd) resistances. As a result, all peripheral beams exhibit low utilization ratios (μ = MEd/MRd), well below 1.0, indicating considerable reserve capacity. In practice, these beams could safely support loads approximately double the current structural demand (

Table 9. Bending moments and shear forces reserve capacity.

Type/Turbine	MEd (kN.m)	μM = MEd/MRd	VEd (kN)	μV = VEd/VRd
Shallow WEG	19,001.9	0.233	3862	0.386
Deep WEG	19,107.0	0.234	3865	0.387
Shallow MY	25,074.0	0.307	4950.7	0.495
Deep MY	25,204.0	0.309	4955	0.496

).

The pedestal columns, due to their large cross-sectional dimensions required to accommodate the high forces transmitted by the peripheral beams, also exhibit substantial load-carrying capacity, even with reinforcement ratios close to the minimum specified by EC2. The transition section was verified using axial force-bending moment (N–M) interaction diagrams, and all analyzed load combinations remained well within the resistance envelope.

3.2.5. Serviceability Limit State Checks for Concrete Elements

As established in the previous section, the SLS verification according to Eurocode 2 includes checks for crack control, concrete fatigue, and reinforcement fatigue. In this case study, only the radial beams exhibited significant service-load effects and are therefore the focus of these assessments. The calculated crack widths remained below the recommended limit of Wk = 0.30 mm, ensuring adequate crack control. With respect to fatigue, concrete stress variations were within permissible ranges, indicating no tendency toward cumulative damage. Similarly, the reinforcement stress amplitudes remained below 70 MPa, confirming sufficient fatigue safety under cyclic loading.

4. Discussion

4.1. Volumetric and Economic Implications

A comparative assessment was conducted between the hybrid reinforcement strategy and a conventional new foundation to quantify the actual material savings, economic implications, and construction-related impacts. This comparison did not involve ULS verifications, or the evaluation of internal force demands in the new foundation. Instead, the analysis focused exclusively on volumetric efficiency and the structural adequacy associated with global stability requirements. Particular emphasis was placed on the no-gapping criterion, which governs the design of large-diameter shallow foundations, and on the no-pile-tension condition for pile-supported foundations.

For comparison consistency, a 24 m diameter was adopted for the new foundation, typical for 4–6 MW turbines in the literature and the wind market. Stability verification for quasi-permanent load showed that this shallow foundation would reach the allowable eccentricity of about 3.0 m (≈0.25·R), satisfying the no gapping criterion. To match the reinforcement solution, the new foundation also employs the same level of vertical perimeter anchorage: eight tie rods of 900 kN each (P = 7200 kN).

Table 10. Concrete volume for new turbines without reinforcement (shallow foundation).

Wind Turbine	N (kN)	P (kN)	Wf (kN)	Σ (kN)	M (kN m)	e (m)	New Foundation (m3)
III-AGW147	6930	7200	12,125	26,255	78,724	3.00	485
V-MySE172	6008	7200	16,100	29,308	87,792	3.00	644

presents the global loads considered and calculated volumes for both the new foundation and the hybrid reinforcement. The adopted geometries yield, for the WEG AGW147 (4.1 MW) and Mingyang MySE172 (6.25 MW) turbines, new foundation volumes of 485 and 644 m³, respectively, while the hybrid reinforcement (suitable for both turbines) uses just 357.31 m³. Beyond reducing the concrete volume and corresponding structural weight, this also means proportional reductions in steel, formwork, and excavation operations.

For the 6.25 MW turbine, the new foundation’s volume is about 80% above the hybrid reinforcements; for the 4.1 MW turbine, the difference is 36%. This disparity arises primarily from the requirement to satisfy the no gapping criterion for large-diameter shallow foundations, necessitating substantial mass to limit admissible eccentricity under service loads. Because the hybrid solution reuses the existing foundation for global stability, the new reinforcement need not duplicate this volume but is sized only for ULS action resistance.

In addition to volumetric reduction, the hybrid reinforcement eliminates the costs and impacts associated with demolition of the existing foundation, which can constitute 10–25% of the new foundation’s total cost and extends turbine downtime. Reduced excavation and earthmoving also significantly decrease emissions from material transport and heavy equipment use.

For the deep foundation, for comparison, a new 20 m diameter foundation (smaller than the shallow one), 19 m between piles, and 36 piles, was considered. The “no pile tension” condition, for quasi-permanent load, was checked by assigning concrete volume (weight) until all piles remained in compression, as detailed in

Table 11. Concrete volume for new turbines without reinforcement (deep foundation).

Wind Turbine	N (kN)	P (kN)	Wf (kN)	Σ (kN)	M (kN m)	New Foundation (m3)
III-AGW147	6930	-	9650	16,580	78,724	386
V-MySE172	6008	-	12,475	18,483	87,792	499

. The table shows the loads and concrete volumes for the WEG AGW147 (4.1 MW) and Mingyang MySE172 (6.25 MW) turbines: new foundation volumes of 386 and 499 m³, respectively, versus just 357.31 m³ for the hybrid reinforcement, applicable to both turbines. As previously stated, this reduction yields proportional decreases in steel, formwork, and excavation operations.

For the 6.25 MW turbine, the volume of the new foundation is approximately 8% greater than that of the hybrid reinforcement; for the 4.1 MW turbine, it is 40% greater. This increase is primarily due to the need to satisfy the “no pile tension” criterion for large-diameter deep foundations, which requires substantial mass to prevent pile tension under service conditions. In contrast, the existing foundation is reused.

Table 12. Concrete volume comparison: hybrid reinforcement vs. the two new wind turbine foundations (shallow and pilled).

Concrete Volume Comparison	m3	%
Reinforcement (original foundation)	357	0
New foundation AGW147 Piled	386	8
New foundation MySE172 Piled	499	40
New foundation AGW147 Shallow	485	36
New foundation MySE172 Shallow	644	80

lists the concrete volumes associated with all foundation alternatives considered in this study: the reinforced original foundation, the shallow foundations for the WEG and Mingyang turbines, and the corresponding deep foundation solutions supported by piles. In every case, reinforcing the existing foundation results in the lowest concrete volume and, consequently, the lowest steel demand, indicating superior material efficiency compared to any new foundation alternative. This comparison accounts only for structural and volumetric quantities. Additional advantages of the reinforcement solution would be even more significant when considering ancillary operations required by a full replacement, such as demolition of the existing block, transportation logistics, and disposal at licensed environmental facilities.

4.2. Comparative Analysis with the Literature

To contextualize the proposed hybrid strategy within the current state of the art, it is important to highlight the scarcity of literature on wind turbine foundation reinforcement that presents detailed structural design, allowing for direct comparisons of material volumes or reinforcement tactics. Despite this limitation, a comparative analysis was conducted against the few documented traditional retrofitting methods, focusing on global stability, load capacity, material optimization, and constructability.

Compared to traditional base expansion methods (e.g., Waldron et al. [20]), which successfully increase both ULS and SLS capacities but require massive concrete additions to increase the footprint and stabilizing mass, the proposed hybrid strategy proves to be a vastly superior solution. While base expansion often demands material volumes equal to or greater than a new foundation and involves extensive excavation, the hybrid method is highly optimized, reducing concrete volume by up to 80%.

On the other hand, localized interface retrofits (e.g., Wang et al. [21]) focus primarily on the local connection rather than global bending moments, failing to address the global stability demands (such as overturning moments and “no gapping” criteria) imposed by modern 4.1 to 6.25 MW turbines. Furthermore, while such methods can be highly invasive and complex at the critical pedestal area, the proposed hybrid solution offers significant constructive advantages. In our strategy, the new embedded ring is easily positioned within the newly cast peripheral shaft, avoiding complex interventions in the existing congested reinforcement zones. This functional separation not only ensures high operational reliability and constructive ease but also reinforces the economic and environmental viability of life extension over full decommissioning.

5. Conclusions

This study presented and validated a hybrid reinforcement concept for onshore wind turbine foundations in repowering scenarios. Based on the structural and geotechnical assessments, the following main conclusions are drawn: (i) As a primary scientific result, the proposed strategy successfully introduces a functional separation in foundation retrofitting, where the existing foundation ensures global stability and serviceability (SLS, including no-gapping and no-tension criteria), while a new peripheral system resists all ultimate limit state (ULS) actions. (ii) Furthermore, regarding scientific validation, the use of a hybrid shell–beam finite element model demonstrated high accuracy in representing the complex load transfer mechanisms and the rigid composite action at the old–new concrete interface, offering a computationally efficient alternative to full 3D solid modeling. (iii) In terms of applied results, the application to a real wind farm with 4.1 MW and 6.25 MW turbine loads confirmed that the hybrid solution maintains structural safety while significantly reducing material consumption, achieving concrete volume reductions of up to 80% for shallow foundations and 40% for piled foundations compared to constructing entirely new structures. (iv) Finally, as a broader applied outcome, the strategy eliminates the costs, environmental impacts, and downtime associated with the demolition, transportation, and disposal of the existing foundation, proving to be a highly sustainable and practically feasible alternative for extending the life of wind energy assets.

To the best of the authors’ knowledge, no equivalent studies in the available literature have evaluated the structural reinforcement of in-service wind turbine foundations specifically designed to accommodate the new, higher loads imposed by more powerful turbines during repowering. This absence reinforces the innovative character of the proposed method. While this study is supported by a comprehensive geotechnical campaign, the soil characterization is predominantly site-specific, and the structural assessment remains limited to a single existing block typology and one specific radial beam reinforcement geometry. Future research should assess the proposed reinforcement strategy under different soil conditions and alternative radial beam configurations, alongside probabilistic approaches for material and degradation uncertainties. Furthermore, leveraging advanced computer vision techniques—such as drone imagery combined with robust, domain-agnostic AI models—presents a promising avenue for long-term structural health monitoring. Architectures like DeepLab [34] and EfficientNet [35], while widely recognized for their precision in medical and general image segmentation, excel at pixel-level anomaly detection and can be readily adapted for automated crack and defect classification in concrete surfaces. Overall, the hybrid strategy represents a promising and adaptable solution for repowering, supporting a safer and more sustainable life extension of wind energy assets.