André Biscaya
Research Subjects
Research Interests
André Biscaya has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
João Pedro Santos
Innovation, AI, Data, and Analytics Department, Nadara, Rua João Chagas 53, Piso 5, 1495-072 Algés, Portugal
5 shared publications- Jonathan Shek
Institute for Energy Systems School of Engineering The University of Edinburgh Edinburgh UK
4 shared publications - Max Malyi
Institute for Energy Systems School of Engineering The University of Edinburgh Edinburgh UK
4 shared publications - António Galhardo
Universidade do Porto (Porto, Portugal)
3 shared publications - Catarina Miranda Oliveira
Universidade do Porto (Porto, Portugal)
3 shared publications - Elsa Caetano
CONSTRUCT, Faculty of Engineering, University of Porto, Rua Dr. Roberto Frias s/n, 4200-465 Porto, Portugal
3 shared publications
Latest Publications
- Open Access
- 2 November 2026
- Published by Copernicus GmbH in Wind Energy Science
- Vol. 11 (11) , 4059-4075
Predicting wind turbine aerodynamic loads using laser-scanned blade geometry with validation based on operational and structural data
Aeroelastic wind turbine models are key tools in lifetime extension studies, but their application is often hindered by high uncertainty in aerodynamic load predictions. Manufacturers typically do not disclose blade geometry, which drives aerodynamic loading and is therefore an essential input. Furthermore, model outputs are often only compared with supervisory control and data acquisition (SCADA) data, which contain limited information on structural loading. The present work addresses these challenges by using experimental measurements to develop and validate an aeroelastic model of an onshore multi-megawatt wind turbine. Results from the terrestrial laser scanning of a wind turbine blade are processed with a new methodology based on consensus algorithms to derive spanwise chord, twist, and airfoil distributions, which are used as OpenFAST inputs. Simulation results are compared with structural health monitoring measurements of tower bending moments at two elevations, in addition to SCADA data. Moreover, rotor thrust and torque are estimated from the tower measurements and used as additional comparison metrics. The simulations show good agreement with the experimental data, supporting the proposed methodology. A parametric study assesses how the predicted loads are affected by blade geometry, polars, yaw misalignment, and SCADA wind speed uncertainty. The results highlight the usefulness of laser scanning for improving confidence in aerodynamic model inputs, as well as the value of tower load measurements as a complementary validation source to SCADA data. Therefore, the paper proposes and validates a new strategy to obtain a highly reliable model for load estimation, using a quite unique set of diverse measurements and new data processing approaches.
- 2 November 2026
- Published by Copernicus GmbH in Wind Energy Science
- Vol. 11 (11) , 4059-4075
Open AccessPredicting wind turbine aerodynamic loads using laser-scanned blade geometry with validation based on operational and structural data
Aeroelastic wind turbine models are key tools in lifetime extension studies, but their application is often hindered by high uncertainty in aerodynamic load predictions. Manufacturers typically do not disclose blade geometry, which drives aerodynamic loading and is therefore an essential input. Furthermore, model outputs are often only compared with supervisory control and data acquisition (SCADA) data, which contain limited information on structural loading. The present work addresses these challenges by using experimental measurements to develop and validate an aeroelastic model of an onshore multi-megawatt wind turbine. Results from the terrestrial laser scanning of a wind turbine blade are processed with a new methodology based on consensus algorithms to derive spanwise chord, twist, and airfoil distributions, which are used as OpenFAST inputs. Simulation results are compared with structural health monitoring measurements of tower bending moments at two elevations, in addition to SCADA data. Moreover, rotor thrust and torque are estimated from the tower measurements and used as additional comparison metrics. The simulations show good agreement with the experimental data, supporting the proposed methodology. A parametric study assesses how the predicted loads are affected by blade geometry, polars, yaw misalignment, and SCADA wind speed uncertainty. The results highlight the usefulness of laser scanning for improving confidence in aerodynamic model inputs, as well as the value of tower load measurements as a complementary validation source to SCADA data. Therefore, the paper proposes and validates a new strategy to obtain a highly reliable model for load estimation, using a quite unique set of diverse measurements and new data processing approaches.
- Open Access
- 12 October 2026
- Published by Copernicus GmbH in Wind Energy Science
- Vol. 11 (10) , 3841-3862
Sub-minute fatigue monitoring for enhanced lifetime assessment of wind turbines
Accurate fatigue-damage estimation is a critical step for managing and extending the life of wind turbine fleets and the basis on which operators decide whether an ageing asset can be kept in service beyond the end of design life. The most accurate estimation relies on continuous high-frequency strain measurements processed with rainflow cycle counting. However, only a small subset of turbines within a fleet are typically instrumented with strain sensors, and even for these, the data volumes, computational complexity, and cost of continuously processing high-frequency strain data via rainflow counting over long-term operation make this approach impractical to sustain at scale. For this reason, industry practice has long relied on stress data aggregated over 10 min windows already used by supervisory control and data acquisition (SCADA) systems, which are available across the full fleet. This choice is convenient but, when combined with conventional window-based rainflow counting, tends to underestimate fatigue damage generated by cycles whose period exceeds the window length, driven by factors such as slow variations in wind speed and direction, operational transitions, control actions, and seasonal or daily temperature variation, as well as, for offshore turbines, changes in wave parameters. Residual-preserving approaches, most notably low-frequency fatigue dynamics (LFFD), address this limitation by reconstructing cycles that span window boundaries but have so far been applied almost exclusively within the conventional 10 min framework. Technology has recently enabled operators to deploy second-resolution SCADA metrics at scale and at reasonable cost, offering finer-grained statistical descriptors for applications such as load and bending-moment estimation and operational-state classification. This is also an opportunity to revisit fatigue monitoring at a SCADA-aligned timescale that is short enough to limit the truncation of boundary-spanning cycles inherent to conventional window-based counting yet aggregated enough to remain deployable across large fleets. This paper benchmarks the families of methods that make this opportunity actionable against continuous rainflow as a baseline: conventional window-based counting, a modified low-frequency fatigue dynamics (LFFD) formulation that explicitly resolves intra-window and inter-window damage contributions, and two reduced-information extrema-sequence representations – start–peaks–valleys–end (SPVE) and maximum and minimum (max–min). The evaluation uses 6 months of 15 s windows of SCADA and strain measurements acquired at 50 Hz from two instrumented tower sections of an onshore wind turbine and examines the influence of window length, sampling rate, and fatigue-curve formulation on damage computation (Eurocode 3 vs. DNV). Reducing the aggregation window from 10 min to 15 s alone does not improve fatigue estimation, as shorter windows increasingly truncate the low-frequency cycles that dominate accumulated damage. The proposed modified LFFD framework overcomes this limitation by preserving cycle continuity across window boundaries, recovering 99.5 % of the baseline damage while requiring only a fraction of the original data volume. Reduced-information representations further demonstrate that fatigue can be estimated with high accuracy using strongly compressed stress sequences, with start–peaks–valleys–end (SPVE) and maximum and minimum (max–min) retaining 98.8 % and 96.2 % of the baseline damage, respectively. These findings establish a practical pathway towards SCADA-aligned, data-efficient fatigue monitoring, enabling scalable lifetime assessment across large wind turbine fleets while preserving the fidelity of high-resolution structural measurements.
- 12 October 2026
- Published by Copernicus GmbH in Wind Energy Science
- Vol. 11 (10) , 3841-3862
Open AccessSub-minute fatigue monitoring for enhanced lifetime assessment of wind turbines
Accurate fatigue-damage estimation is a critical step for managing and extending the life of wind turbine fleets and the basis on which operators decide whether an ageing asset can be kept in service beyond the end of design life. The most accurate estimation relies on continuous high-frequency strain measurements processed with rainflow cycle counting. However, only a small subset of turbines within a fleet are typically instrumented with strain sensors, and even for these, the data volumes, computational complexity, and cost of continuously processing high-frequency strain data via rainflow counting over long-term operation make this approach impractical to sustain at scale. For this reason, industry practice has long relied on stress data aggregated over 10 min windows already used by supervisory control and data acquisition (SCADA) systems, which are available across the full fleet. This choice is convenient but, when combined with conventional window-based rainflow counting, tends to underestimate fatigue damage generated by cycles whose period exceeds the window length, driven by factors such as slow variations in wind speed and direction, operational transitions, control actions, and seasonal or daily temperature variation, as well as, for offshore turbines, changes in wave parameters. Residual-preserving approaches, most notably low-frequency fatigue dynamics (LFFD), address this limitation by reconstructing cycles that span window boundaries but have so far been applied almost exclusively within the conventional 10 min framework. Technology has recently enabled operators to deploy second-resolution SCADA metrics at scale and at reasonable cost, offering finer-grained statistical descriptors for applications such as load and bending-moment estimation and operational-state classification. This is also an opportunity to revisit fatigue monitoring at a SCADA-aligned timescale that is short enough to limit the truncation of boundary-spanning cycles inherent to conventional window-based counting yet aggregated enough to remain deployable across large fleets. This paper benchmarks the families of methods that make this opportunity actionable against continuous rainflow as a baseline: conventional window-based counting, a modified low-frequency fatigue dynamics (LFFD) formulation that explicitly resolves intra-window and inter-window damage contributions, and two reduced-information extrema-sequence representations – start–peaks–valleys–end (SPVE) and maximum and minimum (max–min). The evaluation uses 6 months of 15 s windows of SCADA and strain measurements acquired at 50 Hz from two instrumented tower sections of an onshore wind turbine and examines the influence of window length, sampling rate, and fatigue-curve formulation on damage computation (Eurocode 3 vs. DNV). Reducing the aggregation window from 10 min to 15 s alone does not improve fatigue estimation, as shorter windows increasingly truncate the low-frequency cycles that dominate accumulated damage. The proposed modified LFFD framework overcomes this limitation by preserving cycle continuity across window boundaries, recovering 99.5 % of the baseline damage while requiring only a fraction of the original data volume. Reduced-information representations further demonstrate that fatigue can be estimated with high accuracy using strongly compressed stress sequences, with start–peaks–valleys–end (SPVE) and maximum and minimum (max–min) retaining 98.8 % and 96.2 % of the baseline damage, respectively. These findings establish a practical pathway towards SCADA-aligned, data-efficient fatigue monitoring, enabling scalable lifetime assessment across large wind turbine fleets while preserving the fidelity of high-resolution structural measurements.
- Open Access
- 28 July 2026
- Published by Copernicus GmbH
- p. 1-28
Comment on wes-2026-108
Accurate fatigue damage estimation is a critical step for managing and extending the life of wind turbine fleets, and the basis on which operators decide whether an ageing asset can be kept in service beyond the end of design life. The most accurate estimation relies on continuous high-frequency strain measurements processed with rainflow cycle counting. However, the data volumes, computational complexity, and associated costs of doing this over long-term periods and across large fleets make it impractical at fleet scale. For this reason, industry practice has long relied on stress data aggregated over 10-minute windows already used by Supervisory Control and Data Acquisition (SCADA) systems. This choice is convenient but tends to underestimate fatigue damage generated by cycles whose period exceeds the windows, driven by slow variations in wind speed, operational transitions and control actions. Technology has recently enabled operators to deploy second‑resolution SCADA metrics at scale and at reasonable cost. This is an opportunity to revisit fatigue monitoring at a SCADA‑aligned timescale that is short enough to retain the cycles a 10‑minute window discards, yet aggregated enough to remain deployable across large fleets. This paper benchmarks the families of methods that make this opportunity actionable against continuous rainflow as a baseline: conventional window‑based counting, a modified Low‑Frequency Fatigue Dynamics (LFFD) formulation that explicitly resolves intra‑window and inter‑window damage contributions, and two reduced‑information extrema‑sequence representations — Start-Peaks-Valleys-End (SPVE) and Maximum and Minimum (Max–Min). The evaluation uses six months of 15‑second windows of SCADA and strain measurements acquired at 50 Hz from two instrumented tower sections of an onshore wind turbine and examines the influence of window length, sampling rate, and fatigue‑curve formulation on damage computation (Eurocode 3 vs DNV). Reducing the aggregation window from 10-minute to 15-second alone does not improve fatigue estimation, as shorter windows increasingly truncate the low-frequency cycles that dominate accumulated damage. The proposed modified LFFD framework overcomes this limitation by preserving cycle continuity across window boundaries, recovering 99.5 % of the baseline damage while requiring only a fraction of the original data volume. Reduced-information representations further demonstrate that fatigue can be estimated with high accuracy using strongly compressed stress sequences, with Start-Peaks-Valleys-End (SPVE) and Maximum and Minimum (Max–Min) retaining 98.8 % and 96.2 % of the baseline damage, respectively. These findings establish a practical pathway towards SCADA-aligned, data-efficient fatigue monitoring, enabling scalable lifetime assessment across large wind turbine fleets while preserving the fidelity of high-resolution structural measurements.
- 28 July 2026
- Published by Copernicus GmbH
- p. 1-28
Open AccessComment on wes-2026-108
Accurate fatigue damage estimation is a critical step for managing and extending the life of wind turbine fleets, and the basis on which operators decide whether an ageing asset can be kept in service beyond the end of design life. The most accurate estimation relies on continuous high-frequency strain measurements processed with rainflow cycle counting. However, the data volumes, computational complexity, and associated costs of doing this over long-term periods and across large fleets make it impractical at fleet scale. For this reason, industry practice has long relied on stress data aggregated over 10-minute windows already used by Supervisory Control and Data Acquisition (SCADA) systems. This choice is convenient but tends to underestimate fatigue damage generated by cycles whose period exceeds the windows, driven by slow variations in wind speed, operational transitions and control actions. Technology has recently enabled operators to deploy second‑resolution SCADA metrics at scale and at reasonable cost. This is an opportunity to revisit fatigue monitoring at a SCADA‑aligned timescale that is short enough to retain the cycles a 10‑minute window discards, yet aggregated enough to remain deployable across large fleets. This paper benchmarks the families of methods that make this opportunity actionable against continuous rainflow as a baseline: conventional window‑based counting, a modified Low‑Frequency Fatigue Dynamics (LFFD) formulation that explicitly resolves intra‑window and inter‑window damage contributions, and two reduced‑information extrema‑sequence representations — Start-Peaks-Valleys-End (SPVE) and Maximum and Minimum (Max–Min). The evaluation uses six months of 15‑second windows of SCADA and strain measurements acquired at 50 Hz from two instrumented tower sections of an onshore wind turbine and examines the influence of window length, sampling rate, and fatigue‑curve formulation on damage computation (Eurocode 3 vs DNV). Reducing the aggregation window from 10-minute to 15-second alone does not improve fatigue estimation, as shorter windows increasingly truncate the low-frequency cycles that dominate accumulated damage. The proposed modified LFFD framework overcomes this limitation by preserving cycle continuity across window boundaries, recovering 99.5 % of the baseline damage while requiring only a fraction of the original data volume. Reduced-information representations further demonstrate that fatigue can be estimated with high accuracy using strongly compressed stress sequences, with Start-Peaks-Valleys-End (SPVE) and Maximum and Minimum (Max–Min) retaining 98.8 % and 96.2 % of the baseline damage, respectively. These findings establish a practical pathway towards SCADA-aligned, data-efficient fatigue monitoring, enabling scalable lifetime assessment across large wind turbine fleets while preserving the fidelity of high-resolution structural measurements.
- Open Access
- 6 July 2026
- Published by Copernicus GmbH in EGUsphere
- Vol. 2026, 1-28
Sub-minute Fatigue Monitoring for Enhanced Lifetime Assessment of Wind Turbines
Accurate fatigue damage estimation is a critical step for managing and extending the life of wind turbine fleets, and the basis on which operators decide whether an ageing asset can be kept in service beyond the end of design life. The most accurate estimation relies on continuous high-frequency strain measurements processed with rainflow cycle counting. However, the data volumes, computational complexity, and associated costs of doing this over long-term periods and across large fleets make it impractical at fleet scale. For this reason, industry practice has long relied on stress data aggregated over 10-minute windows already used by Supervisory Control and Data Acquisition (SCADA) systems. This choice is convenient but tends to underestimate fatigue damage generated by cycles whose period exceeds the windows, driven by slow variations in wind speed, operational transitions and control actions. Technology has recently enabled operators to deploy second‑resolution SCADA metrics at scale and at reasonable cost. This is an opportunity to revisit fatigue monitoring at a SCADA‑aligned timescale that is short enough to retain the cycles a 10‑minute window discards, yet aggregated enough to remain deployable across large fleets. This paper benchmarks the families of methods that make this opportunity actionable against continuous rainflow as a baseline: conventional window‑based counting, a modified Low‑Frequency Fatigue Dynamics (LFFD) formulation that explicitly resolves intra‑window and inter‑window damage contributions, and two reduced‑information extrema‑sequence representations — Start-Peaks-Valleys-End (SPVE) and Maximum and Minimum (Max–Min). The evaluation uses six months of 15‑second windows of SCADA and strain measurements acquired at 50 Hz from two instrumented tower sections of an onshore wind turbine and examines the influence of window length, sampling rate, and fatigue‑curve formulation on damage computation (Eurocode 3 vs DNV). Reducing the aggregation window from 10-minute to 15-second alone does not improve fatigue estimation, as shorter windows increasingly truncate the low-frequency cycles that dominate accumulated damage. The proposed modified LFFD framework overcomes this limitation by preserving cycle continuity across window boundaries, recovering 99.5 % of the baseline damage while requiring only a fraction of the original data volume. Reduced-information representations further demonstrate that fatigue can be estimated with high accuracy using strongly compressed stress sequences, with Start-Peaks-Valleys-End (SPVE) and Maximum and Minimum (Max–Min) retaining 98.8 % and 96.2 % of the baseline damage, respectively. These findings establish a practical pathway towards SCADA-aligned, data-efficient fatigue monitoring, enabling scalable lifetime assessment across large wind turbine fleets while preserving the fidelity of high-resolution structural measurements.
- 6 July 2026
- Published by Copernicus GmbH in EGUsphere
- Vol. 2026, 1-28
Open AccessSub-minute Fatigue Monitoring for Enhanced Lifetime Assessment of Wind Turbines
Accurate fatigue damage estimation is a critical step for managing and extending the life of wind turbine fleets, and the basis on which operators decide whether an ageing asset can be kept in service beyond the end of design life. The most accurate estimation relies on continuous high-frequency strain measurements processed with rainflow cycle counting. However, the data volumes, computational complexity, and associated costs of doing this over long-term periods and across large fleets make it impractical at fleet scale. For this reason, industry practice has long relied on stress data aggregated over 10-minute windows already used by Supervisory Control and Data Acquisition (SCADA) systems. This choice is convenient but tends to underestimate fatigue damage generated by cycles whose period exceeds the windows, driven by slow variations in wind speed, operational transitions and control actions. Technology has recently enabled operators to deploy second‑resolution SCADA metrics at scale and at reasonable cost. This is an opportunity to revisit fatigue monitoring at a SCADA‑aligned timescale that is short enough to retain the cycles a 10‑minute window discards, yet aggregated enough to remain deployable across large fleets. This paper benchmarks the families of methods that make this opportunity actionable against continuous rainflow as a baseline: conventional window‑based counting, a modified Low‑Frequency Fatigue Dynamics (LFFD) formulation that explicitly resolves intra‑window and inter‑window damage contributions, and two reduced‑information extrema‑sequence representations — Start-Peaks-Valleys-End (SPVE) and Maximum and Minimum (Max–Min). The evaluation uses six months of 15‑second windows of SCADA and strain measurements acquired at 50 Hz from two instrumented tower sections of an onshore wind turbine and examines the influence of window length, sampling rate, and fatigue‑curve formulation on damage computation (Eurocode 3 vs DNV). Reducing the aggregation window from 10-minute to 15-second alone does not improve fatigue estimation, as shorter windows increasingly truncate the low-frequency cycles that dominate accumulated damage. The proposed modified LFFD framework overcomes this limitation by preserving cycle continuity across window boundaries, recovering 99.5 % of the baseline damage while requiring only a fraction of the original data volume. Reduced-information representations further demonstrate that fatigue can be estimated with high accuracy using strongly compressed stress sequences, with Start-Peaks-Valleys-End (SPVE) and Maximum and Minimum (Max–Min) retaining 98.8 % and 96.2 % of the baseline damage, respectively. These findings establish a practical pathway towards SCADA-aligned, data-efficient fatigue monitoring, enabling scalable lifetime assessment across large wind turbine fleets while preserving the fidelity of high-resolution structural measurements.
- Open Access
- 22 April 2026
- Published by Copernicus GmbH in EGUsphere
- Vol. 2026, 1-30
Predicting wind turbine aerodynamic loads using laser scanned blade geometry with validation based on operational and structural data
Aeroelastic wind turbine models are key tools in lifetime extension studies, but their application is often hindered by high uncertainty in aerodynamic load predictions. Manufacturers typically do not disclose blade geometry, which drives aerodynamic loading and is therefore an essential input. Furthermore, model outputs are often only compared with Supervisory Control and Data Acquisition (SCADA) data, which contains limited information on structural loading. The present work addresses these challenges by using experimental measurements to develop and validate an aeroelastic model of a utility-scale wind turbine. Results from the terrestrial laser scanning of a wind turbine blade are processed with a new methodology based on consensus algorithms to derive spanwise chord, twist, and airfoil distributions, which are used as OpenFAST inputs. Simulation results are compared with structural health monitoring measurements of tower bending moments at two elevations, in addition to SCADA data. Moreover, rotor thrust and torque are estimated from the tower measurements and used as additional comparison metrics. The simulations show good agreement with the experimental data, supporting the proposed methodology. A parametric study investigates how several uncertainties in the model inputs affect the predicted aerodynamic loads. The results highlight the usefulness of laser scanning for improving confidence in aerodynamic model inputs, as well as the value of tower load measurements as a complementary validation source to SCADA data. Therefore, the paper proposes and validates a new strategy to obtain a highly reliable model for load estimation, using a quite unique set of diverse measurements and new data processing approaches.
- 22 April 2026
- Published by Copernicus GmbH in EGUsphere
- Vol. 2026, 1-30
Open AccessPredicting wind turbine aerodynamic loads using laser scanned blade geometry with validation based on operational and structural data
Aeroelastic wind turbine models are key tools in lifetime extension studies, but their application is often hindered by high uncertainty in aerodynamic load predictions. Manufacturers typically do not disclose blade geometry, which drives aerodynamic loading and is therefore an essential input. Furthermore, model outputs are often only compared with Supervisory Control and Data Acquisition (SCADA) data, which contains limited information on structural loading. The present work addresses these challenges by using experimental measurements to develop and validate an aeroelastic model of a utility-scale wind turbine. Results from the terrestrial laser scanning of a wind turbine blade are processed with a new methodology based on consensus algorithms to derive spanwise chord, twist, and airfoil distributions, which are used as OpenFAST inputs. Simulation results are compared with structural health monitoring measurements of tower bending moments at two elevations, in addition to SCADA data. Moreover, rotor thrust and torque are estimated from the tower measurements and used as additional comparison metrics. The simulations show good agreement with the experimental data, supporting the proposed methodology. A parametric study investigates how several uncertainties in the model inputs affect the predicted aerodynamic loads. The results highlight the usefulness of laser scanning for improving confidence in aerodynamic model inputs, as well as the value of tower load measurements as a complementary validation source to SCADA data. Therefore, the paper proposes and validates a new strategy to obtain a highly reliable model for load estimation, using a quite unique set of diverse measurements and new data processing approaches.
- Open Access
- 1 January 2026
- Published by Institution of Engineering and Technology (IET) in IET Renewable Power Generation
- Vol. 20 (1) , e70327
Exploratory Semantic Reliability Analysis of Wind Turbine Maintenance Logs Using Large Language Models
A wealth of operational intelligence is locked within the unstructured free‐text of wind turbine maintenance logs, a resource largely inaccessible to traditional quantitative reliability analysis. While machine learning has been applied to this data, existing approaches typically stop at classification, categorising text into predefined labels. This paper addresses the gap in leveraging modern large language models (LLMs) for more complex reasoning tasks. We introduce an exploratory framework that uses LLMs to move beyond classification and perform deep semantic analysis. We apply this framework to a large industrial dataset to execute four analytical workflows: failure mode identification, causal chain inference, comparative site analysis and data quality auditing. The results demonstrate that LLMs can function as powerful “reliability co‐pilots,” moving beyond labelling to synthesise textual information and generate actionable, expert‐level hypotheses. This work contributes a novel and reproducible methodology for using LLMs as a reasoning tool, offering a new pathway to enhance operational intelligence in the wind energy sector by unlocking insights previously obscured in unstructured data.
- 1 January 2026
- Published by Institution of Engineering and Technology (IET) in IET Renewable Power Generation
- Vol. 20 (1) , e70327
Open AccessExploratory Semantic Reliability Analysis of Wind Turbine Maintenance Logs Using Large Language Models
A wealth of operational intelligence is locked within the unstructured free‐text of wind turbine maintenance logs, a resource largely inaccessible to traditional quantitative reliability analysis. While machine learning has been applied to this data, existing approaches typically stop at classification, categorising text into predefined labels. This paper addresses the gap in leveraging modern large language models (LLMs) for more complex reasoning tasks. We introduce an exploratory framework that uses LLMs to move beyond classification and perform deep semantic analysis. We apply this framework to a large industrial dataset to execute four analytical workflows: failure mode identification, causal chain inference, comparative site analysis and data quality auditing. The results demonstrate that LLMs can function as powerful “reliability co‐pilots,” moving beyond labelling to synthesise textual information and generate actionable, expert‐level hypotheses. This work contributes a novel and reproducible methodology for using LLMs as a reasoning tool, offering a new pathway to enhance operational intelligence in the wind energy sector by unlocking insights previously obscured in unstructured data.
- Open Access
- 1 January 2026
- Published by Institution of Engineering and Technology (IET) in IET Renewable Power Generation
- Vol. 20 (1) , e70212
A Comparative Benchmark of Large Language Models for Labelling Wind Turbine Maintenance Logs
Effective operation and maintenance (O&M) is critical to reducing the levelised cost of energy (LCOE) from wind power, yet the unstructured, free‐text nature of turbine maintenance logs presents a significant barrier to automated analysis. We address this by presenting a novel, reproducible framework for benchmarking large language models (LLMs) on classifying these complex industrial records, which is publicly available as an open‐source tool. We demonstrate the framework's utility by systematically evaluating a diverse suite of state‐of‐the‐art proprietary and open‐source LLMs, providing a practical case study for the foundational assessment of their trade‐offs in reliability, operational efficiency and model calibration. Benchmarked against a qualitatively selected high‐performing reasoning model, our study demonstrates how the tool quantifies a clear performance hierarchy to help analysts identify top models exhibiting high alignment and trustworthy confidence scores. Furthermore, we show that classification performance depends heavily on semantic ambiguity, with all models displaying higher consensus on objective component identification than on interpretive maintenance actions. Given that no model achieves perfect accuracy and calibration varies dramatically, we conclude the most responsible application is a human‐in‐the‐loop system where LLMs act as assistants to accelerate and standardise data labelling, enhancing O&M data quality and downstream reliability analysis.
- 1 January 2026
- Published by Institution of Engineering and Technology (IET) in IET Renewable Power Generation
- Vol. 20 (1) , e70212
Open AccessA Comparative Benchmark of Large Language Models for Labelling Wind Turbine Maintenance Logs
Effective operation and maintenance (O&M) is critical to reducing the levelised cost of energy (LCOE) from wind power, yet the unstructured, free‐text nature of turbine maintenance logs presents a significant barrier to automated analysis. We address this by presenting a novel, reproducible framework for benchmarking large language models (LLMs) on classifying these complex industrial records, which is publicly available as an open‐source tool. We demonstrate the framework's utility by systematically evaluating a diverse suite of state‐of‐the‐art proprietary and open‐source LLMs, providing a practical case study for the foundational assessment of their trade‐offs in reliability, operational efficiency and model calibration. Benchmarked against a qualitatively selected high‐performing reasoning model, our study demonstrates how the tool quantifies a clear performance hierarchy to help analysts identify top models exhibiting high alignment and trustworthy confidence scores. Furthermore, we show that classification performance depends heavily on semantic ambiguity, with all models displaying higher consensus on objective component identification than on interpretive maintenance actions. Given that no model achieves perfect accuracy and calibration varies dramatically, we conclude the most responsible application is a human‐in‐the‐loop system where LLMs act as assistants to accelerate and standardise data labelling, enhancing O&M data quality and downstream reliability analysis.
- Open Access
Impact of Strain Gauge Preprocessing Methods on Load Measurements and Fatigue Estimation in Wind Turbine Towers
Electrical strain gauges are essential for monitoring wind turbine tower loads and fatigue, but accurate load measurements from these sensors require calibration over time to correct the zero-drift found in long-term measured signals. Calibration is often performed using nacelle rotation events for cable untwisting, where the tower mechanical load is known; however, non-uniform solar heating during these events can introduce thermal stresses that are misinterpreted as drift, causing systematic errors. This study evaluates six preprocessing methods for correcting zero-drift and thermal stresses in strain gauges, using measurements from two tower cross-sections—one with temperature sensors and one without. Performance is quantified using the scatter of the 10 min mean bending moments in the fore–aft and side-to-side directions and the cumulative fatigue damage over the monitoring periods. Results show that modelling the thermal stresses using a linear regression model with temperature measurements as inputs yields the most physically consistent load curves. If temperature measurements are unavailable, the effects of thermal stresses can be partly mitigated by restricting calibration to nighttime events or using solar-position variables in a regression model (instead of temperatures). As expected, the choice of preprocessing method significantly impacts load curves, but its influence on fatigue damage estimates is limited.
Open AccessImpact of Strain Gauge Preprocessing Methods on Load Measurements and Fatigue Estimation in Wind Turbine Towers
Electrical strain gauges are essential for monitoring wind turbine tower loads and fatigue, but accurate load measurements from these sensors require calibration over time to correct the zero-drift found in long-term measured signals. Calibration is often performed using nacelle rotation events for cable untwisting, where the tower mechanical load is known; however, non-uniform solar heating during these events can introduce thermal stresses that are misinterpreted as drift, causing systematic errors. This study evaluates six preprocessing methods for correcting zero-drift and thermal stresses in strain gauges, using measurements from two tower cross-sections—one with temperature sensors and one without. Performance is quantified using the scatter of the 10 min mean bending moments in the fore–aft and side-to-side directions and the cumulative fatigue damage over the monitoring periods. Results show that modelling the thermal stresses using a linear regression model with temperature measurements as inputs yields the most physically consistent load curves. If temperature measurements are unavailable, the effects of thermal stresses can be partly mitigated by restricting calibration to nighttime events or using solar-position variables in a regression model (instead of temperatures). As expected, the choice of preprocessing method significantly impacts load curves, but its influence on fatigue damage estimates is limited.
- 1 October 2025
- Published by Springer Nature
- p. 1009-1015
Evaluation of Wind Turbine Foundation Repair with Accelerometer Measurements
- 1 October 2025
- Published by Springer Nature
- p. 1009-1015
Evaluation of Wind Turbine Foundation Repair with Accelerometer Measurements
- Open Access
Exploratory Semantic Reliability Analysis of Wind Turbine Maintenance Logs using Large Language Models
A wealth of operational intelligence is locked within the unstructured free-text of wind turbine maintenance logs, a resource largely inaccessible to traditional quantitative reliability analysis. While machine learning has been applied to this data, existing approaches typically stop at classification, categorising text into predefined labels. This paper addresses the gap in leveraging modern large language models (LLMs) for more complex reasoning tasks. We introduce an exploratory framework that uses LLMs to move beyond classification and perform deep semantic analysis. We apply this framework to a large industrial dataset to execute four analytical workflows: failure mode identification, causal chain inference, comparative site analysis, and data quality auditing. The results demonstrate that LLMs can function as powerful "reliability co-pilots," moving beyond labelling to synthesise textual information and generate actionable, expert-level hypotheses. This work contributes a novel and reproducible methodology for using LLMs as a reasoning tool, offering a new pathway to enhance operational intelligence in the wind energy sector by unlocking insights previously obscured in unstructured data.
Open AccessExploratory Semantic Reliability Analysis of Wind Turbine Maintenance Logs using Large Language Models
A wealth of operational intelligence is locked within the unstructured free-text of wind turbine maintenance logs, a resource largely inaccessible to traditional quantitative reliability analysis. While machine learning has been applied to this data, existing approaches typically stop at classification, categorising text into predefined labels. This paper addresses the gap in leveraging modern large language models (LLMs) for more complex reasoning tasks. We introduce an exploratory framework that uses LLMs to move beyond classification and perform deep semantic analysis. We apply this framework to a large industrial dataset to execute four analytical workflows: failure mode identification, causal chain inference, comparative site analysis, and data quality auditing. The results demonstrate that LLMs can function as powerful "reliability co-pilots," moving beyond labelling to synthesise textual information and generate actionable, expert-level hypotheses. This work contributes a novel and reproducible methodology for using LLMs as a reasoning tool, offering a new pathway to enhance operational intelligence in the wind energy sector by unlocking insights previously obscured in unstructured data.