Topic 1 – The application of big data and advanced statistical techniques
Value from free-text maintenance records: converting wind farm work orders into quantifiable, actionable information using text mining
Presenting author: Erik Salo, University of Strathclyde
Co-authors: Dr David McMillan
Prof Richard Connor
The aim of this project is to demonstrate how data and text mining techniques can help wind farm operators to extract unique, quantifiable, site- and asset-specific maintenance information from historic work orders. Understanding how maintenance efforts have been distributed in the past can help develop a more evidence-based maintenance strategy for the future in terms of labour intensity, budgeting and logistics of spare parts. However, work order records – where significant information is entered by a human in the form of free text – can present a particularly complex data source for analysis.
Our approach introduces a novel combination of machine learning techniques supported by a database of domain vocabulary and expert judgement. Significant focus is on term recognition, aided by spelling error correction and semantic matching of synonyms and abbreviations. Task descriptions can thereby be classified by meaning, not just the words present. In the first instance this creates a frequency distribution of all the different tasks carried out. Categorical data can then be extracted about maintenance of different functional locations and subsystems, as well as the occurrence of different failure modes.
Data from major onshore wind farms in Scotland was used to test our approach against undertaking a similar analysis manually. Potential savings were identified on the order of weeks of effort, or £ 9k in labour cost per wind farm, in addition to the benefits of an improved maintenance strategy.
The remaining challenges mainly lie in increasing accuracy and reducing operator input. These are being addressed by our continued research, but also provide opportunities for collaboration and standardisation across the wind energy industry to maximise the value of data.
Analysis and benchmark-oriented evaluation of energy yields with focus on the performance assessment of wind turbines
Presenting author: Philip Görg, Fraunhofer IEE
The contribution deals with the topic of analyzing low energetic yields as well as the identification of their different causes and the derivation of possible optimization approaches. For this purpose, a benchmark concept is created which describes the different steps and includes the information required to build a holistic framework which can be used for the performance assessment of a wind turbine (WT).
The contribution compromises the implementation of the suggested concept as well as the deduction of different key indicators which can be used to assess multiple energetic loss causes in terms of their magnitudes. Different key indicators are found which allow to quantitatively assessing the performance, the nacelle misalignment, and the wind measurement system of a WT, respectively. The methodology for deriving these indicators is then applied to the operational data of the Fraunhofer IWES WInD-Pool to determine a critical threshold for each considered loss cause to identify potentials of optimization.
The results show that the majority of all analyzed turbines have a “non-critical performance” in at least 98.50% of their operational time. Moreover, an approximately reverse square correlation between the mean generation ratio and the operational lifetime of all analyzed turbines can be seen. While the turbines performance increases significantly throughout the first year of operation from an initially low level, a continuous decline of the generation ratio can be seen after some operational years..
The usage of status codes from the turbines operating system to identify critical performance events is only partially effective. Most of all data points found to be “critical” cannot be linked to a specific status code (approx. 14%).
Applying the proposed key indicators to identify yaw misalignments of turbine nacelles suggests a threshold of 2.94% (time-based ratio). Furthermore, onshore WT show higher values than offshore WT.
The wind measuring system of all considered WT shows a significantly higher deviation between the measured value and a reference value (wind farm mean) in standstill than in operational times. Additionally, all considered turbine types show a higher tendency to underestimate the wind speed compared to the underlying references during standstill times (median: -0.59m/s).
Using Operational SCADA Data to Optimise Assets
Presenting author: Charles Plumley, Wood
Co-authors: David Robb
While performance is often measured by just looking at budgeted outputs, availability and occasionally power curves, a more in depth analysis of the SCADA data can reveal trends and opportunities for optimisation that might otherwise be passed by. Wood’s optimiser data analysis approach involves:
- Looking at individual event codes to identify common issues across a farm, portfolio or turbine type
- Monitoring performance in a comparative manner to identify issues at an individual turbine or farm
- Exploring a wide variety of SCADA signals to identify anomalies and suboptimal designs
- And predicting the potential for life extension and uprating through comparing site conditions with the wind turbine class
By delving more deeply into the available SCADA data, conclusions can be drawn to highlight where best to focus effort to improve performance. For example, following the data analysis, endemic problems such as pitch or electrical failures can be pursued with the OEM, targeted forestry felling can be justified, controller and hardware improvements can be applied, unnecessary curtailments or operational modes can be removed, and uprating or life extension considered, but only where appropriate.
These techniques have benefited over 1 GW of capacity so far and the lessons learnt and experiences gained are shared here to benefit the wider community.
Visualisation and automation data to decisions
Presenting author: Mark Spring, Lloyd’s Register
Co-authors: Philip Knaute, Matthew Zhang
Through working with wind farm owners it has become clear that improvements in performance are hampered by limited access to comprehensive turbine operational and contextual information. New software tools developed have enabled both an increase in revenue from electricity generated and a reduction in operational costs, due for instance to on-site interventions. Technical hurdles in the path include pitch angle errors, yaw misalignments, power curtailment, mechanical and electrical losses and production lost during inefficiently scheduled turbine downtime.
Implementations of the approaches described in this paper have demonstrated that seeing a problem clearly has been the first step towards addressing it, diagnosing the causes, mitigating future recurrences and predicting the risk of experiencing each particular mode of failure. We have succeeded in automating the display of key relationships and combinations of data. By using domain knowledge from design engineers, technicians, business managers and specialists with relevant experience from outside the wind sector, we have trained the data science tools so as to automate the process of diagnosis and simulation of future operation. Future risks of failure have been estimated by understanding modes of operation, accumulation of damage, the physics of failure, interactions between environment and turbines, incorporating the influence of maintenance teams, parts suppliers. This understanding has been incorporated into practices and software.
The process started with a formal FMECA. Automation of the subsequent steps of data classification, fault diagnosis and failure prediction have relied on novel combinations of proven data science algorithms, face-to-face workshops, user-interface mock-up exercises and condition monitoring systems.
For a system to be resilient it needs to exhibit the following characteristics: plasticity, adaptivity, redundancy, autonomy and self organisation. This project has demonstrated progress along the path towards this goal for wind power plants.
Catching the context: cross-fleet benchmarking enabled by context-sensitive fingerprinting of SCADA data
Presenting author: Tom Tourwe, Sirris
Co-authors: Elena Tsiporkova, Alessandro Murgia, Mathias Verbeke
Wind turbine data, whether it is CMS or SCADA data, is typically analysed using advanced data mining and machine learning algorithms that exploit the time series. Time-sensitive features extracted from hourly, daily, weekly or monthly periods do not reflect explicitly the actual context in which the turbine is operating however. As a consequence, the resulting models are turbine-specific and are difficult to generalise to the whole fleet or portfolio. In this presentation, we will present an alternative approach that is based on characterising the operational context in which the turbines are operating. We will explain how this context can be represented, how each turbine’s performance within the different contexts can be modeled, how this allows to identify groups of turbines that behave similarly, and how ultimately these context-sensitive fingerprints can be exploited to derive context-aware performance and cross-fleet benchmarking.
Power Curve Analysis using high-frequency SCADA Data
Presenting author: Gianmarco Pizza, Nispera AG
In the present study, the power curve of a modern multi-megawatt wind turbine is analysed using high-frequency (1 second) SCADA data. Different transients are investigated and described in detail, like e.g. turbine startup, turbine yawing, turbine response to wind speed ramps, and comparison with commonly used 10-minutes averaged data is provided.
Preliminary results have been presented recently at Windenergietage
Topic 2 – Predicting and enhancing turbine performance
A review of data uncertainty of floating lidars enbling use in analysis of operational wind farms
Presenting author: Breanne Gellatly, Director of European Operations, Axys Technologies
Floating LiDAR Systems (FLS) as a wind resource assessment tool has seen increasing market acceptance in recent years with multiple commercial deployments and new applications such as power performance testing, as an onsite weather station to feed into operational decision-making tools such as weather forecasting and wind farm fatigue forecasting. While systems have demonstrated performance to the accuracy standards laid out in various industry documents, the offshore wind industry is now requiring a greater understanding of the levels of uncertainty in floating LiDAR measurements. Greater coherence across this topic will benefit the industry by providing banks and other finance providers with a more accurate basis for the assessment of FLS derived P50s and P90s, leading to better rate pricing of risk and supporting a reduction in the cost of offshore wind finance through to operational phases.
RETEX on 4 Years of Wind Turbine Performance Testing and Optimisation using Independent Measurement System Based on 2D Nacelle Lidar.
Presenting author: Guillaume Coubard-Millet, WPO
Co-authors: Alban JEHU, Antoine ROBBES, Philippe QUINET
Asset management of wind farms depends heavily on the SCADA data the wind farm produces. The main concern of operators today is to maximize the technical availability of the turbines. For this purpose, the LOGs of the turbines are the main source of information, when the wind speed is used as an indicator of wind presence more than an accurate measurement. Nonetheless, with the lowering cost of energy, turbine efficiency and their optimization become a more and more prevalent topic.
In order to target turbines that requires performance improvement, the classic approach is to compute power curves based on the available data (SCADA data) and using methodologies inspired by standards such as IEC61400-12-2. Turbine wind sensors being placed behind the rotor plan, their measurement are highly disturbed by the blades washing. Therefor, to reflect the wind flow conditions as if the turbine was not present, nacelle anemometers measurement are corrected by a function called Nacelle Transfer Function. This correction can induce potential bias, leading to potentially wrong performance diagnostic (correction on the Nacelle anemometer) or under-performance (wind vane signals correction).
To address these issues, WPO pre analyzed on behalf of its clients 126MW of asset based on SCADA power curves and innovative data analysis methods, allowing to target specific turbine for an independent performance assessment and optimization campaign.
For this purpose, WPO developed a complete solution. The first part of this solution is an innovative SCADA data pre analysis methodology that permits to select under performing turbines for an independent measurement campaign. The second part of the solution is installed in the turbine, and consists in an independent logger that gathers signals from power meters, a meteo station, a GPS Compass and a 2D Nacelle Lidar measuring at 2.5D for an independent wind speed measurement. The system measures and collect all information required to perform an IEC-like power curve. The use of nacelle Lidar allowed the characterization of the Yaw misalignment, the Nacelle Transfer Function, the turbulence intensity level and the turbine power curve – power coefficient following the IEC61400-12-1 guidelines.
3.Main body of abstract
In this study, WPO presents a RETEX on 8 campaigns that were carried throughout Europe. The targeted turbines were purposefully chosen for being erected maximum 5 years prior the measurement campaign, in order to exclude the wearing of the blades and the electromechanical chain as source of under-performance. Results are ordered in 4 categories by order of frequency.
Category 1: Slight to no NTF deviation, Yaw misalignment issue and under performing turbine: The turbine under performance is related to the yaw misalignment. The turbine improvement after yaw misalignment correction was about +1.8% in average.
Category 2: Large NTF deviation, no Yaw misalignment and under performing turbine: By sending a biased wind speed information to the wind turbine controler, the pitch and RPM are not set optimally, leading to a sub optimal operation of the turbine.
Category 3: Slight to no NTF deviation, no Yaw misalignment and under performing turbine: The source of underperformance is not identified. Indirectly, the shape of the independent power curve can indicate a turbulence / wind shear induced underperformance. If the underperformance is equal on the whole wind speed range, a pitch angle analysis is advised.
Category 4: NTF deviation, no Yaw misalignment and False detection of under-performance: It is believed that a NTF deviation between 3% to 7% do not induce under performance. Nonetheless, the NTF deviation can biased the understanding of the performance of their turbine performance.
Thanks to innovative pre analysis, WPO was able to reduce the uncertainty around the veracity of the under performance diagnostic based on SCADA data, and limit the false detection. This uncertainty reduction is critical for operators, and optimize the operator’s investment.
The WPO measurement system was able to perform IEC like power curves and to characterize under performance causes with measurement campaigns 15days long (yaw misalignment characterization) to 3 months long (full analysis). The time efficiency was allowed by the use of a nacelle Lidar and efficient data analysis monitoring.
Our track record shows that the yaw misalignment issue represents the main source of under performance.
Perform IEC like power curve assessment with 2D Wind Iris Nacelle Lidar.
Statistics on the cause of turbine under performances.
On the importance of an efficient wind turbine performance pre analysis prior deciding on an independent measurement campaign.
Wind turbine performance under the influence of wind characteristics: a case study
Presenting author: Minh-Thang DO, Meteodyn
Co-authors: Julien BERTHAUT-GERENTES
The influence of air density to the performance of wind turbine has been known for a long time and has been introduced by several manufacturers as different power curve for each level of air density. In 2013, IEC proposed a correction of the air density in the standard of the industry IEC 61400-12-2.
The influence of other wind characteristics, such as: turbulence intensity, shear, veer or inflow angle… has also been investigated in several researches. But how to apply this theoretical knowledge to better estimate the wind turbine output power is still an open question. The main obstacle is the difficulty to obtain these variables in practice.
Our approach proposes to estimate these wind characteristics from the measured wind direction and a look-up table obtained from a pre-processed CFD simulation (MeteodynWT). The influence of wind shear and wind inflow angle to the performance of wind turbine is then detected in this case study. A correction surface based on a polynomial model is applied to the SCADA power curve and a reduction of about 15% of the standard deviation is obtained at the most dispersed zone on the power curve.
This polynomial correction is then translated to several powers curves corresponding to several levels of wind shear and inflow angle. This simplified discrete model conforms to the different power curves for each level of air density proposed by manufacturers.
Reducing the cost of energy through OPEX cost optimisation
Presenting author: Selena Farris, Natural Power Consultants Ltd.
While wind farm owners will have a range of operational strategies, they will all ultimately be driven by a ROI over a set number of years, short-term for some, and long-term for others. An O&M approach designed to optimise the ROI for the owner’s priorities by leveraging benchmarked data and a total asset management approach is utilised. Costs are fed in for each activity and the ROI over the required timeframe is assessed to provide evidence for any OPEX costs.
A three step process for reducing the cost of energy through OPEX cost optimisation is undertaken. First, the asset owner/operator is given a clear understanding of the performance of their assets through analysis of SCADA and other available data. Secondly, a comprehensive review of the management of the wind farm site is undertaken. Finally, and most importantly, the intelligence gained from the performance analysis is fed into a performance cost model which produces O&M plans that consider the ROI for each task recommended. From this, an optimised O&M strategy is defined. Case studies are presented to demonstrate results.
Performance monitoring of wind turbines using Spinner Anemometers
Presenting author: Sowjanya Subramaniam Iyer, ROMO Wind
Co-authors: Nick Gerardus Cornelis Janssen
Eduardo Gil Maron
The rapid growth of wind turbines and wind farms has created the importance of monitoring the performance of wind turbines. Throughout these years, there have been several conventional ways to do this. One of the more recent approaches is using spinner anemometers which is being established at ROMO Wind using iSpin. As a step towards performance monitoring, at ROMO Wind we have developed the concept of iSpin Guardian. Using iSpin Guardian, we are able to compare the measured power curve for a specific wind turbine type to other turbines of the same type, independent of location and climatic conditions. The performance of all turbines can be compared to the iSpin Guardian power curve i.e. the average performance of the turbine type. In this manner, the under-performing turbines can be identified and later verify improvement in terms of increased AEP.
In order to demonstrate the ability of iSpin to measure and compare wind turbine performance and be a robust source for performance monitoring of wind turbines, the Performance Transparency Project (PTP) was initiated. In this project, 60 iSpin systems have been installed in 6 different locations for two turbine types with varying terrain complexities. Using iSpin measurements from these different sites, this project aims at proving that iSpin wind speeds are stable and turbine independent which implies that measuring wind speeds at spinner is more valuable thereby promoting the use of spinner anemometers for performance monitoring of wind turbines.
The first results of the PTP will be showcased in this work. For the first two sites for V112 3MW turbines, iSpin guardian showcases AEP of all the twenty turbines within 1.6% of each other. These are results from two sites: one offshore in Belgium and other simple terrain onshore site in Croatia. The results from the third site for V112 is expected soon. The results for 30 E-82 2.3 MW turbines is expected to be in place within the next three months. From the results for all the 6 sites, we are able to highlight the effectiveness of performance monitoring using spinner anemometers.
Case study: Comparison of two Nacelle Mounted LiDAR technologies for a wind turbine yaw misalignment analysis.
Presenting author: Alix Pradel, UL
Co-authors: Alix Pradel, Julien Dalmas, Christophe Lepaysan
UL DEWI, Lyon, France, +33 4 27 18 10 25, [email protected]
VALEMO, Bagles, France, +33 5 57 96 96 54, [email protected]
EPSILINE, Toulouse, France, +33 5 32 10 83 50, [email protected]
1. General summary
Yaw misalignment is a frequent cause of turbine underperformance during the operation of a wind turbine as it can cause production losses and undesired loads. The detection and measurement of such a misalignment is usually performed using Nacelle Mounted LiDARs.
Indeed, a LiDAR mounted on the top of the nacelle and accurately aligned with the rotor is used to measure the wind speed and the relative wind direction in front of the rotor before the wind is perturbed by the rotating blades. The measured wind parameters can therefore be considered more representative of the actual wind that faces the turbine than the usual wind sensors installed on the nacelle. By analysing the LiDAR relative wind direction, yaw misalignment angle can be calculated.
Two different LiDAR technologies were installed on the top of a wind turbine in France and measured the wind during the same one month period. An Avent Wind Iris LiDAR, which already has a long track record on many turbine models, is compared with a more recent LiDAR product, the EPSILINE YawAdvisor. The goal was to check the wind turbine misalignment as well as the behaviour of both devices.
At the end of the campaign, it has been possible to check that the wind turbine was not correctly aligned with the wind. The measurements of both LiDARs were also consistent in term of static yaw misalignment and dynamic yaw misalignment.
The Avent Wind Iris LiDAR (two beams) measures the horizontal wind speed and direction from 80m to 400m upwind of the turbine.
The EPSILINE YawAdvisor enables wind direction measurement immediately in front of the wind turbine (at 10 m ahead of the LiDAR) and provides information regarding the angle between the actual wind flow and the nacelle centerline.
During approximately one month, both Nacelle Mounted LiDARs were measuring simultaneously on a GE2.5 xl wind turbine part of the French wind farm Laucourt, owned by VALOREM and operated by VALEMO. UL DEWI and VALEMO performed the calculation of the Yaw Misalignment with the Avent WindIris LiDAR and the EPSILINE YawAdvisor LiDAR data. Both LiDARs were accurately aligned with the rotor thanks to a specific procedure using a laser device that was aligned with the main shaft.
At the end of the campaign it has been possible to check that the final yaw misalignment angle results for both technologies were quite similar with an angle of 7.1Â° for the Wind Iris and 7.3Â° for the Yaw Advisor.
The YawAdvisor wind speed measurement is affected by the rotor proximity (induction zone) due to the distance of measurement (at 10 m ahead of the LiDAR), it is therefore not possible to use this wind speed for performance analysis and power curve measurements. However, for comparison purpose, the wind speeds measurements of both LiDARs were compared. The 10 minute wind speed correlation coefficient was calculated to be around 0.98 between the Wind Iris first range gate (80 m) and the YawAdvisor. The average wind speed measured by the YawAdvisor was 21% below the wind speed measured by the Wind Iris at a distance of 2.4 diameters due to the induction zone.
The evolution of the yaw misalignment and its standard deviation with the wind speed were also determined for both technologies and showed similar trends: in this case the yaw misalignment was not depending on the wind speed while the standard deviation of the yaw misalignment was decreasing for higher wind speeds as the turbine controller is more reactive and the wind direction possibly more stable.
Two different LiDAR technologies were installed on the top of the nacelle during the same measurement period, the data gathered by both devices are consistent and it has been concluded that the studied wind turbine was misaligned.
These results are encouraging as they show that the relative wind direction measured 10 m ahead of the wind turbine with the YawAdvisor is comparable to the relative direction measured at longer distances by the Wind Iris. Therefore, this would mean that the wind direction measured by the YawAdvisor is suitable for yaw misalignment detection.
5. Learning Objectives
By measuring at longer distances, the Wind Iris can also be used to assess the operational power curve and quantify the gain of a yaw realignment correction. The YawAdvisor can only be used to calculate the yaw misalignment angle as the wind speed is affected by the rotor proximity.
The two LiDAR technologies could be seen as complementary in the wind industry. In order to check the yaw alignment on many turbines in a wind farm, the YawAdvisor could be a quick and simple solution while the Wind Iris could be installed in some sample turbines for an additional analysis of the operational power curve.
The results are based on one case study, they should be reproduced to get stronger conclusions about the consistency of the YawAdvisor technology for yaw misalignment angle calculation.
Topic 3 – Post-construction yield analysis
Calculation of energetic loss of operating wind farms based on SCADA Data
Presenting author: Till Schorer, UL
Co-authors: Volker Barth
The German Renewables Act 2017 (EEG 2017) requires to assess for all turbines erected under that act the actually fed-in energy as well as the energy that could have been potentially produced after 5, 10 and 15 years of operation. While some criteria are defined in the renewables act itself, a detailed procedure how to calculate energetic losses to fulfil these new legal requirements was not described so far.
The presentation gives a brief introduction about the Technical Guideline 10 (TR10) which is scheduled to be published in early 2018 by FGW e.V. This guideline describes how energetic losses have to be calculated from 10 minutes SCADA data in connection with status codes from the turbine logfiles. The presentation gives a detailed description how status codes shall be summarized according to the guideline and which 10 minutes data have to be used to create a representative power curve, which is ultimately used to calculate the energetic losses. Periods without data available have to be filled by using data from nearby turbines or wind data from Reanalysis models based on correlation matrix. We will also show how the energetic availability will be derived in the end. With regard to the amount of data in a 5 years SCADA time series an automated processing approach will be necessary and presented.
Although the defined process in the TR10 is directly related to the German Renewables Act, it is the first guideline describing a standard how to calculate energetic losses based on SCADA data, which can be easily adopted to each wind turbine in operation everywhere in the world to define an independent and standardized assessment of the energetic loss in operation creating the base for further performance assessment. Therefore, the Technical Guideline 10 has as well a global relevance.
Requirements and Achievements in Post Construction Energy Yield Assessment
Presenting author: Martin Strack, Deutsche WindGuard Consulting GmbH
The post construction energy yield assessment is an important measure to provide the most accurate view on the long term expectancy of the project revenue. The presenter has performed such assessments for several GW of installed capacity, within the scope of project evaluation, vending or refinancing processes. The highest requirements for the work, but often also the highest motivation because of uncertainty reduction, can be observed for offshore projects. So the proposed talk will highlight but not limit to experiences gained in offshore projects.
The basis of any post construction energy yield assessment is a qualified operational data assessment. The talk will describe experiences gained and requirements derived for assessment and filtering of performance issues. The utilisation of the nacelle anemometer, which is a popular and helpful instrument for this purpose, is discussed on basis of findings from more than 2 GW offshore operational data evaluation and several measurement campaigns performed . A solution in terms of an applicable and robust procedure, which ensures the usability of the nacelle anemometer, is explained .
The wind turbine interactions or exposure to external wake effects can be non-representative and thus significantly disturb an operational data evaluation. Based on examples from offshore wind farms, it will be shown how the availability status of the turbines can influence the park efficiency and thus bias any evaluation which bases on the mean wake losses or wake losses calculated (statically) for the operational period. It is shown how this can be resolved by performing a dynamical wake calculation, considering the actual operational and availability status of each turbine in time series resolution. This is also a solution to accurately describe the temporal variation of external wake losses (and other exposure factors), which can be an important aspect also for onshore turbines.
The long-term assessment of operational data is usually the largest single uncertainty component. Different approaches to model the wind farm power variation, and possible pitfalls when doing this, will be shown on basis of findings from offshore and onshore examples. Regarding the correct determination of the mean wind conditions, which still can be a challenge, the author has recently conducted an extensive study for a leading wind energy lender in Germany. This work comprises the evaluation of 235 years of operational wind farm data (onshore). The main findings from this, including experiences derived for different (international) long-term data sources, will be summarized.
Based on anonymized real examples, it will be shown how the upside in uncertainty or P90 level can be, when a qualified uncertainty assessment is performed which combines the results based on different data sources, considering the dependency of different evaluation steps.
The proposed talk will give an overview of the main aspects of post construction energy yield assessment, in terms of requirements but also feasible solutions, based on sound experience and real-world examples. This may provide valuable knowledge or suggestions for the audience involved in these tasks.
 Martin Strack: Quantification of grid related energy yield losses: Experience gained in German offshore projects. Contribution to WindEurope Summit 2016, Hamburg, Germany.
 Martin Strack et.al.: Big data approach of wind resource and operational data analysis in cold climate. Presentation held at WindEurope Summit 2016, Hamburg, Germany.
 Martin Strack: Quantification of grid related energy yield losses: Experience gained in German offshore projects. Presentation held at German Wind Energy Conference (DEWEK) 2017, Bremen, Germany.
 Martin Strack et.al.: Big data approach of wind resource and operational data analysis. Presentation held at German Wind Energy Conference (DEWEK) 2017, Bremen, Germany.
Long-term Operational Energy Assessments
Presenting author: Claudia Puyals, AWS Truepower, a UL Company
Co-authors: Jesica Pion, Brian Kramak and Stephen Lightfoote
Before the construction of a wind farm, pre-construction long-term energy production estimates are necessary to anticipate generation and estimated revenue, but generally, these pre-construction assessments are calculated using only wind data collected prior to the construction of the project. After the wind farm has been in operation for a minimum recommended time period of one year, it is possible to derive more accurate results and reduce uncertainties. Operational yield assessments are performed using actual production data of the plant and for this reason, they require fewer assumptions and it is possible to produce results with lower uncertainty “typically two thirds to one half of the pre-construction energy estimates”.
In this presentation, the authors explain how operational energy estimates are performed using self-developed programming tools, including future blade degradation losses as well as future availability estimates based on the past performance of the plant.
In case of owning a portfolio of wind projects, this allows owners to balance local resource fluctuations and reduce the volatility of the revenue stream. By evaluating long-term energy production of a portfolio as a whole, the result is that the error distribution for the combined output of the projects is often narrower than the summed error distributions of the individual projects. This may translate into lower risk to the net income for the portfolio owner.
During the presentation, we will describe the proposed methodology and analyze the results obtained in a number of projects in Europe.
Topic 4 – Asset reporting: availability and other key metrics
An estimation of key metrics from scada data – a third-party’s view
Presenting author: Minh-Thang DO, Research Engineer, Meteodyn
Co-authors: Julien Berthaut-Gerentes, Meteodyn
SCADA-based monitoring is a cost-effective approach to reduce O&M budget in wind energy. However, this approach requires a clean and complete status-code signal, which is usually reconstructed from an error log. This procedure is complicated and sometime is not feasible due to the missing or the incompleteness of the error log.
For this reason, we introduce a detailed and dedicated classification algorithm, which reconstructs a valuable status code, including partial stops and curtailments. It requires only the nacelle wind speed and the output power from the SCADA system. When available, other SCADA parameters, such as the pitch control, are automatically added to enhance the precision of the classification algorithm. The operational power curve is then estimated from the dataset of normal status. The detection method has a high robustness, which can give a good estimation of the power curve, even in the case that the normal status takes less than 50% of the total SCADA data. From the detected status code and the estimated power curve, key metrics of each machine during a given period are identified: total production, losses, time based availability and energy based availability.
In a second time, the power performance of each machine is calculated on a monthly base. This power performance index, which is insensitive to wind exposure variations, is able to measure simply and rapidly the effect of a maintenance (e.g. repair of a mechanical part), a settings change (e.g. recalibration of an anemometer), and to detect a functional anomaly.
3rd Classfication of Windcube achieved
Presenting author: Paul Mazoyer, Leosphere
Co-authors: Florian Rebeyrat
Windcube fully complies with the requirements of the latest edition of the IEC 61400-12-1 for power measurement testing (PPT) as it has been subject to the remaining tests. Implementing a PPT with a wind sensor requires comprehensive information of the sensitivity of the accuracy to environmental parameters. This sensitivity is tested through a process called the classification which is fully documented in the standard. The test is to be done against a mast. This test has to last several months, to be undertaken on two different testing sites and for two different units. The classification may seem cumbersome but it is crucial. Indeed, it allows for a comprehensive assessment of the uncertainty of an RSD. In 2017, Deutsche WindGuard tested a Windcube unit on two different sites. This adds to the previous classification of another Windcube unit at Rysum test site in 2012. The test is then completed for the Windcube.
Leosphere aims to present the classification process done by Windcube , along with the actual results. Especially, the study confirms that the only impacting parameters are the turbulence intensity and the shear, as observed in 2012. The final accuracy class of the Windcube will be presented.
The objective of the presentation is to provide a clear overview on the use of RSD for PPT along with on the uncertainties of PPT with RSD following the IEC standard.
Asset Reporting Targeting Wind Farm Optimisation
Presenting author: Charles Plumley, Wood
Co-authors: David Robb
Asset reporting typically focusses on month-to-month operation, with a general performance overview looking at current revenue compared to the predicted (the P50), availability indicators, and sometimes power curves. Whilst this kind of analysis highlights underperforming assets, it fails to highlight what can be done to improve performance and the root causes of any underperformance.
Wood has developed a method that specifically focusses on improving the performance of all wind farm assets, known as an Optimiser Phase 1. This analysis uses a wide range of data sources and metrics to identify the root causes of underperformance and quantify potential solutions. This then acts as a springboard to target the most cost effective solutions to improve asset performance.
In particular the analysis utilises:
Reports and discussions with the operators and onsite personnel identify baselines, trends and any known issues.
Event logs are used to determine availability and the lost energy due to any faults or curtailments.
SCADA data is analysed to compare performance across the farm, check anemometry, wind turbine operation and the potential for life extension.
Topographic maps and knowledge of land management plans allow for analysis of the effects of targeted tree felling and other topographic impacts.
Financial modelling to determine the cost benefit and impact on Net Present Value (NPV) for the implementation of identified performance improvements.
A wide range of metrics are therefore used to quantify not only any unnecessary losses, but also the added value that applying different solutions will bring to the assets. The methods, metrics and outcomes from Wood’s experience of carrying out these assessments will be detailed in this paper.
WEBS: a framework for aligning with international standards for asset reporting
Presenting author: Craig Stout, Wind Energy Benchmarking Services
Co-authors: Conaill Soraghan (WEBS Engineering Specialist)
Jeff Bryan (WEBS Commercial Manager)
Reporting on key performance indicators (KPIs) is a crucial business process but fleet-wide and industry-wide comparison of these metrics can be problematic due to inconsistencies in methodologies. Addressing this issue is a key aim of the Wind Energy Benchmarking Service (WEBS). WEBS provides independent anonymised benchmarking data for the wind industry and currently aggregates operational data from wind farms across 6 countries.
This presentation will provide an overview of the WEBS system and demonstrate how it has created a framework that is facilitating the standardisation of wind farm performance reporting.
Participation in WEBS provides owners/operators with unique insight into successes or performance gaps and the ability to develop performance improvement plans. Of particular interest for this presentation, WEBS provides a framework to align with international standards which has been adopted from best practice in the offshore wind industry.
WEBS benchmarks involve ranking performance of KPIs against that of industry peers and leading performers. The WEBS system benchmarks more than 120 performance, availability, reliability and maintenance metrics across a range of dimensions, such as geographical region, turbine type and age. The WEBS KPIs can be categorised as:
- Production (yield) based availability [IEC 61400 26-2] : Wind farm availability based on the actual production at site as a proportion of the potential production over the period.
- Time Weighted Run Time Availability [IEC 61400 26-1] – Measurement of the proportion of time the full wind farm system was available for generation; capturing the availability when the turbines were generating or could have generated if the environmental conditions were right.
- Lost Energy Production – Production energy lost due to specified events; such as scheduled maintenance, forced outages and requested shutdowns.
Major System Repairs
- Details of major system repairs at site, characterised by major components with significant downtime. Lost energy production and downtime due to specified number of major component failures is provided.
Sub-System Forced Outages
- When an immediate action to disable generating function of the turbine(s) is required as unforeseen damage, faults, failures or alarms are detected.
- Application of the Reference Designation System for Power Plants [RDS-PP] taxonomy for components.
- These metrics provide additional context to the benchmarks; including metrics such as SCADA data availability, Hub-height wind speed and Number of non-access days due to weather.
One of the unique aspects of WEBS is that the architecture of the system ensures participants can benefit from industry wide benchmarks while their individual KPIs remain anonymous. Commercial sensitivity is a priority at WEBS and is why all data is protected by several layers of cloud-based data protection and security measures. This presentation will provide a high-level design of the system architecture that preserves anonymity.
Building on the success of a similar system called SPARTA which is being used by the offshore wind industry, WEBS is a relatively new benchmarking platform designed exclusively for operational onshore wind farms. The lessons learned that have been incorporated in the WEBS metric definitions and system architecture will be shared in this presentation.
Peer-based OPEX benchmarking as a tool for identifying areas of improvement and understand the industry baseline
Presenting author: Grzegorz Skarzynski, Pexapark LTD
To share with workshop participants with the OPEX Benchmarking Initiative experience, as a practical tool to enhance asset effectiveness
To present view on asset effectiveness from commercial side
– OPEX benchmark as a continuous activity – recurring every year (always open for new peers) allows to see improvements from past efforts and whether market moved allowing for new optimization
– methodology developed by Pexapark for OPEX Benchmarking Initiative
– selected 7 main categories, 20 subcategories and 7 filters (to make those categories comparable)
– OPEX benchmarking gives peers chance to quantify performance, through:
-comparison OPEX across categories and
– adjustments for filters (size, technology,
– compare cost ratios on realized production/capacity
– OPEX benchmarking as a tool for identifying opportunities:
– identify the market baseline and possible gap to
– seize improvement opportunity based on trends
– connect with peers to share insights and knowledge
Lessons learned from the Initiative
– pitfalls in cost allocations
– one -offs considerations for making results
– continuous actions – increased peer group
– minimal peer group to assure results relevance
– selection for categories and sub – categories
Topic 5 – Repowering, extending life or decommissioning?
New approach for the Condition Monitoring System within Life Extension Strategies
Presenting author: Enrique Camacho Cuesta, Ingeteam Power Technology – Service
Co-authors: Luis Moreno
The energy output from wind turbines has increased dramatically over the past thirty years from 50 kW to 8 MW, while 10-20 MW turbines are currently in the design stage. The increasing numbers of multi-MW turbines being installed offshore require an accurate and cost-effective condition monitoring system (CMS) to be put in place in order to achieve high availability, reliability and maintain profitability of the wind farm.
The ROI (return of investment) for CMS including installation, maintenance and licensing is between 3 and 5 years depending on the wind turbine power. Nowadays these figures are only interesting, from of investment point of view, for turbines bigger than 1 : 1.5 MW of power.
Most turbine manufacturers build their machines with 20-year operational life but in practice around 90% of turbines can potentially carry on working for up to a decade beyond their original operational life.
Assuming a turbine lifespan of 20 years, more than 86 GW of wind capacity is expected to be decommissioned in Europe progressively from 2016. Life extension programs of turbines offer European operators a competitive alternative to repowering in markets where regulated payments have dropped in recent years. These turbines are generally powers lower than 1 MW and it was an unreachable market for installing CMS so far.
Life-extension programs involve the installation of new or additional condition monitoring equipment so it brings the opening of a new scenario for the CMS:
- Need to install new CMS
- Need to adapt the actual CMS with more capabilities for the hardware as the adaptability and configurability.
- Need to have one unique software to integrate the different CMS technologies that they are already installed in the wind turbines
In this paper we will discuss the needs of this new market for the CMS including the installation expectations.
Predicting RUL (Remaining Useful Life) of structural components by means of IAM (Indpendant Aerolastic Models) and how to reduce its uncertainties.
Presenting author: Santiago Lopez, UL
Co-authors: Jose Javier Ripa
Predicting RUL (Remaining Useful Life) of structural components by means of IAM (Indpendant Aerolastic Models) and how to reduce its uncertainties.
LTE analysis is based on a comparison between the design conditions and the actual wind farm conditions. The analysis considers the wind conditions, the operation conditions and the maintenance activities carried out on the wind turbines. The comparison is performed using a life time model based on an Independent Aero-elastic model or IAM, which is a function of the dynamic and aerodynamics of the specific turbine model. The wind model and the O&M conditions provides time series of loads and other variables as rotor speed, power in several points of the wind turbine, deflections, and more. A probabilistic approach is used and the results of the analysis show a different life estimation per component as the material properties and the way the loads affect differ from one component to other. The results are present by means of Gaussian distributions, which have more or less width depending on the uncertainty. The more precise is the analysis, the lower economical or safety risks will be faced during the extended operation. Finally, life time expectancies are related to different inspection programs connected to a specific risk base scenario. Among the analysis, the failure rate is key factor to understand the life extension. The turbines are designed and certified for ten to the power of minus 4 per year. The certification bodies check that this limit is not reached before the year 20. If the site or operating conditions are more benign, this failure rate will be reached later. Uncertainties plays a key role in this parameter, as these can be reduce by reducing the uncertainty of the material by means of selected inspections. Owners and/or operators can define the maximum acceptable FR to continue operating assets in safe conditions.
Uncertainties are key to value the risks of the Life Extension Plan. The analytical model is supported in the study of many external parameters, with their respective uncertainties and sensitivities. Some are large and require specific measures to reduce them: Proper Data Treatment, Wind Modelling and Time Series Generation, Aerolastic model adjustments with accredited tests and Inspections plan.
How Predictive Analytics Can Better Inform Decisions Related to Remaining Useful Life of Turbines
Presenting author: Alex Byrne, Uptake
Making decisions about the operating life of a wind farm is a complicated undertaking: most do not keep exact or extensive records of the entire multi-decade operating history which is needed for analysis; predicting the future is challenging; and owners typically do not receive comprehensive design documentation from wind turbine manufacturers. These factors combine to produce very high uncertainty around the estimated remaining useful life (RUL) of a wind turbine; one can hardly make a decision about retrofitting, repowering, decommissioning or continuing to run business as usual if the predicted life of the asset is, for example, 30 years with an uncertainty of 15 years.
Even in situations where uncertainty is high, RUL calculations can still be helpful for down-selecting options for life extension or its alternatives and for generally understanding the severity of environmental loading at a specific site. However, the mean calculated RUL should not be used to inform a detailed operating strategy, the risks of being wrong are too great. Instead, monitoring, inspections, and predictive analytics should and can be used to support safe continued operations past a turbine’s design life.
This presentation will describe the methods of calculating remaining useful life of wind turbine components, with a particular focus on realistic reporting of uncertainty involved with those calculations. Secondly, a complementary approach to life extension will be discussed: the role of monitoring, inspections and predictive analytics can be a powerful way to get the most years of operation out of an asset while still maintaining a high level of safety, in fact the authors would argue that it is the only way.
Uptake is a leading independent and comprehensive predictive analytics software company that creates a world that always works. Uptake provides solutions for industry giants, such as Berkshire Hathaway Energy, and offers a SaaS product suite for companies of all sizes. Based in Chicago, Uptake works with customers to set new standards for productive, secure, safe and reliable operations.
ABOUT THE SPEAKER
Alex Byrne is Uptake’s Senior Innovation Manager for wind and brings more than a decade of energy engineering experience. Prior to Uptake, she worked for DNV GL as a Principal Engineer supporting stakeholders in evaluation of repowering, life extension, site suitability and performance upgrades. She has a Master’s degree in Mechanical Engineering from the University of California, Berkeley.
Topic 6 – Innovative techniques for enhanced performance
How to boost wind farm profitability with smart automated operations
Presenting author: Alejandro Cabrera, Green Eagle Solutions SL
Co-authors: Alejandro Cabrera
The levelized cost of renewable energy continues to decrease each year, making wind and solar energy even more competitive than conventional sources.
The energy market requires constant improvement and optimization of O&M activities to preserve the long term viability of investments. However, as prices decrease, so do the margins of the entire supply chain. How low can the cost of an O&M service be while still preserving the viability and quality of the O&M service? Will the demand for low cost services negatively impact the lifespan of wind farms?
Our research has focused on finding the most cost effective O&M strategy to minimize downtime and maximize wind energy yield.
With our experience in SCADAs, Green Eagle Solutions wanted to measure the impact of software retrofit and compare it to other solutions such as turbine retrofit or repowering. The solutions needed to be technologically independent with a low cost of installation.
By automating the decision making in remote operations, it is possible to customize the operations protocol up to single turbine level, taking into account not only errors or alarms coming from the turbine, but also meteorological data, energy pool price, or any other input that the operator would like to introduce in the operations protocol.
With this strategy, O&M can provide an immediate but also smarter response, minimizing production losses due to resettable errors, increasing availability, and improving production.
In this study, we analyze the results of implementing this new strategy in five different wind farms equipped with GAMESA, MADE and NORDEX technologies. We highlight a business case of the implementation of our solution, CompactSCADA® Virtual Operator, in a Spanish wind farm, with a total installed capacity of 49MW. The results show an estimated yearly savings of over a 50.000 and improve turbines operations after implementing the new strategy.
Smart automated operations will become an essential tool to enhance wind farms’ performance, availability, and profitability worldwide.
An innovative way of processing iSpin measurements leads to a factor 2 reduction in system cost.
Presenting author: Nick Janssen, ROMO Wind
iSpin measurement technology has reached a new level of cost-competitiveness; Making it available for every single wind turbine in a fleet. This was concluded by ROMO Wind after a factor 2 reduction in cost of an iSpin measurement system was achieved by successfully shifting from a multi-sensor setup to a single-sensor measurement setup.
Over the past years, iSpin technology has settled itself into the market and is generally accepted as a valid wind speed measurement device. Its value lies in the fact that wind conditions are measured before they cross the rotor plane, rather than behind the rotor plane. This means that iSpin measurements provide a low uncertainty and can reliably be used to optimize turbine performance.
A traditional iSpin system consists of a controller, and three ultrasonic wind speed sensors on the spinner of a wind turbine. If the iSpin system could be reduced from a multi-sensor system to a single sensor, the cost of an iSpin system can be reduced by a factor 2. This can be achieved, because each sensor travels along the same path and therefore collects the same information during a rotation. Storing historical information makes it possible to make up for the missing sensors.
Results from a single-sensor measurement setup were compared to a multi-sensor setup. It was concluded that measurements are equally accurate when using a single sensor. This means that it is safe to reduce the amount of iSpin sensors on the spinner from three to one. A downside of a single-sensor measurement setup is however that historical data is needed from up to a minute back in time. More pros an cons of this innovative method are elaborated and discussed.
A more efficient method for static yaw misalignment detection and correction
Presenting author: Nicolas Quievy, ENGIE
Co-authors: Bruno Declercq
Wind turbines are designed to face the wind. When the nacelle is not aligned with the wind direction, the wind turbine produces less power than expected. This static yaw misalignment, i.e. the non-zero mean error of yaw alignment, is usually measured by means of a nacelle-LiDAR campaign. This method is time consuming, expensive and can only be applied to one turbine at a time. Therefore a more efficient measurement technique is aimed. A method using a ground-based LiDAR (GBL), measuring the absolute wind direction, and nacelle mounted Professional Heading and Positioning Global Navigation Satellite System (GNSS) compass on several turbines, measuring the absolute nacelle direction, was developed. This method was tested and validated on a wind farm located in flat terrain. The yaw error before and after correction was indeed compared to simultaneous nacelle-LiDAR measurements. Results were consistent with nacelle-LiDAR measurements although a slightly higher uncertainty was observed. However this has a minimal impact on annual energy production after correction, demonstrating the potential of this new method to bring significant cost reduction and duration of yaw misalignment campaign by correcting several turbines simultaneously.
Case Study: Optimising the performance of a small UK wind farm through wind turbine power modes
Presenting author: Matthew Zhang, Lloyd’s Register
Co-authors: David Pullinger,
Qadir Singapore Wala,
Nathan Hill and
Whole wind farm optimisation is a hot-topic within the wind industry. Current research areas are investigating many different routes for improving wind farm performance when the system is considered as a whole instead of the traditional wind turbine optimisation approach previously followed. This study will focus on the application of power mode optimisation to a small onshore wind farm using the already available building blocks within the wind turbine controller.
The advantage of this approach is that implementation is straightforward and there is no requirement for significant changes of the control system, or any risk of increased loading due to alternative strategies such as wake bending. By assessing the performance in isolation, the benefits of the power mode variations studied can be thoroughly understood, before more complex techniques are further incorporated within a wind farm optimisation.
The work presented will focus on a case study project (name and details kept confidential) and proof of concept for a 9 wind turbine project located in the UK. Through an analysis of the operational SCADA data, the tuning of the site wind flow and wake model and an understanding of the power modes already specified within the wind turbine control system several different configurations are analysed.
The different configurations represent different objectives that a wind farm owner or operator may have for their project including maximum power output and various load scenarios for the turbines. The results of this study will quantifiably demonstrate the impacts of the different optimisation strategies on both the project energy yield but also the loading conditions that the turbines experience.
Results of Concurrent Power Performance Measurements Using an IEC mast, a Profiling Lidar, and a 4-beam Nacelle Lidar
Presenting author: Claudia Puyals, Senior Analyst, AWS Truepower, a UL Company
Co-authors: Matthew V. Filippelli, AWS Truepower, a UL Company
Paul Mazoyer – Project Manager / Wind Energy – Leosphere
Linda Sloka – Senior Engineer – AWS Tuepower, LLC
AWS Truepower, Leosphere and a confidential wind farm owner have developed and deployed a multi-sensor measurement program to accelerate the commercial application of lidar in wind turbine power performance measurement campaigns. Power performance measurement plays an important role in wind project financing, warranties and performance assessment. The significant cost, timing, and logistical requirements of conducting these analyses according to the current IEC standards have fostered the development of alternate tools and methods. These methods are expected to significantly expand and accelerate power performance measurements, but have not yet been broadly adopted for commercial applications. The technical efficacy of nascent techniques and sensors, including the IEC 61400-12-1 ed. 2 procedures and Leosphere’s suite of profiling and nacelle-mounted lidar systems, have been demonstrated in a variety of environments; however, the public library of direct comparisons with current standards and uncertainties : upon which commercial acceptance is partly based : is still relatively limited.
This measurement campaign is uniquely scoped to characterize a single turbine’s power curve using concurrent observations from multiple sensor platforms and analysis procedures. A new multi-megawatt turbine, installed at a simple terrain site in the Midwest USA, has undergone power performance measurements according to the current IEC 61400-12-1 standards with a hub-height mast. The measurement site is also equipped with a Windcube profiling lidar to evaluate turbine performance in line with IEC 61400-12-1, ed. 2. The assessed turbine will also be additionally instrumented with a commercial 4-beam Wind Iris nacelle-mounted lidar capable of measuring hub height speeds at various distances upwind, along with vertical shear and veer across the rotor swept area.
The specific analyses will incrementally include comparison of power curve results and uncertainties. Examination of the nacelle-mounted lidars will include characterization of expected uncertainties in power curve measurement. It will also present assessment of key filtering criteria including, shear, TI, and veer, in accordance with the recommendations of the Power Curve Working Group. The resulting data sets are expected to provide multiple, comparable case studies with the new measurement methods benchmarked against current standard procedures.
WindGainHub: OEM independent strategy
Presenting author: Carlo Durante, eTa Blades
Best practices by WindGainHub, a pioneering initiative supported by a network of Independent Technology and Service Providers, to enable Wind-Farmers improving the Technical Control of their Assets and unlock their potential, maximizing IRRs and lowering LCOEs.
WindGainHub tailors methodologies and solutions according to Asset / Site Specific Conditions (Wind, Operations, Turbine, Legislation, Permitting, Compliance, etc.) on each Wind-Farm with the aim to maximise Wind-Farmers’ IRRs through the optimal combination of Life-Extension, Power Production Upgrades, Long-Term Tailored Maintenance, Damage Mitigation Strategies.
WindGainHub is conceived to be OEM independent providing unique equipment and tools which are compatible, dedicated and tailored for each asset, covered by Certification, at the same level of technical skills and safety as OEMs and proven to deliver higher and better results than standard packs available in the market.
Topic 7 – Offshore operations: lowering the costs
Improving offshore wind farm performance through novel access strategies
Presenting author: Fernando Sevilla Montoya, DNV GL
Due to the challenges of accessing offshore wind turbines, an optimal O&M strategy is critical in achieving a high level of wind farm performance. One key area of the O&M strategy is the crew transportation and turbine access solution for the project.
In order to increase wind farm accessibility, improve project performance and reduce operational costs, crew transportation and turbine access solutions are scrutinised during the operational phase of the project. The choice of an access strategy is highly site specific, depending predominantly on the distance to the site, number of turbines and site conditions. Current access strategies, such as the deployment of standard catamaran Crew Transfer Vessels, can achieve accessibilities at offshore projects in the North Sea in the range of 50 to 70%.
As technology has evolved, new vessel technology and methods of access have emerged and are now available for operational projects, enabling potential benefits to accessibility and hence project performance, that were not previously attainable. Based on detailed time-domain and Monte Carlo simulations, DNV GL has scrutinised the benefits to project availability and O&M costs that novel access strategies, such as recently-trialled vessels (Surface Effect Ship), helicopter support and use of Service Operations Vessels (SOV), could represent.
For this study, DNV GL has performed computer simulations that recreate all relevant real operational activities required for three operational offshore wind projects and estimated their current availability and operational expenditure. DNV GL has then changed the O&M strategy via implementation of novel vessel and access technologies, enabling detailed analysis of the benefits that these technologies could unlock in offshore wind farm availability and operational expenditure.
The results of this work show that a significant increase in wind turbine availability in the range of 1-2% can be achieved through changes to current O&M strategies.
Active maintenance optimisation of a large offshore wind farm based on mesoscale day-ahead power forecasts
Presenting author: Ken Tay, Lloyd’s Register
Co-authors: Qadir Singapore Wala, Rebecca Sykes
The scheduling of planned maintenance for offshore wind farms often does not take into account the value of the energy lost whilst maintaining individual wind turbines. Typical strategies involve the use of a passive scheduling strategy : this work demonstrates that the passive approach is inefficient based on real offshore wind farm data and quantifies the benefit of active optimisation strategies.
We developed a methodology to actively schedule maintenance activities based on a day-ahead power forecast. In the study, we use a physics-based mesoscale model to simulate a month-long case study of a mid-latitude offshore wind farm in the summer. We compare the following maintenance scheduling approaches: ACTIVE, ADVERSE and REALISTIC.
In all maintenance strategies, a list of turbines that require scheduled maintenance are generated at the start of the month. On each day, 4 selected turbines are shut-down and maintained for 8 daylight hours. A total of 120 unique turbines will be serviced over the month.
Based on the daily power forecasts, the ACTIVE case selects the 4 lowest yielding turbines while ADVERSE selects the 4 highest yielding turbines. Comparing ACTIVE and ADVERSE cases, there is a 37% reduction in downtime power loss. This reduction in power loss represents a theoretical maximum that can be achieved (worst case vs. best case).
The REALISTIC case is based on actual recorded SCADA data where a daily list of turbines shut-
down for scheduled maintenance is generated. Using the simulation data, the power forecast of these shut-down turbines is compared against maintaining the same number of turbines on that day using the ACTIVE strategy. In this case study, we found a 32% reduction in downtime power loss : which directly relates to project revenue.
Topic 8 – Operating wind farms in hybrid mode
Market opportunities for hybrid windfarm and battery plants – a critical review
Presenting author: Henrik Sundgård Pedersen, EMD International
Co-authors: Anders Andersen
The price drop in electrical batteries offer new opportunities for economic optimizing the operation in the electricity markets of wind farms equipped with a large batteries, but requires a more complex daily operation of such hybrid plants.
This presentation examines through market simulations which added value a hybrid plant may have in 4 opportunities, compared to only operating the windfarm alone:
- The hybrid plant improves sale in Intra-day and Day-ahead markets
- The hybrid plant may reduce punishment for imbalances created in whole sale markets
- The hybrid plant may improve active participation in balancing markets
- The hybrid plant may improve sale when there is restricted grid capacity
It is in the presentation shown, that the earnings depend on the specific location of the hybrid plant, considering the price volatility in the electricity markets and how the whole sale and balancing markets are organized.
A discussion will be made of these 4 opportunities compared to the opportunities for the hybrid plant delivering other ancillary services.