Final Report Summary - WINTUR (In situ wireless monitoring of on and offshore wind turbine blades using energy harvesting technology)
Executive summary:
Wind energy is an increasingly important contributor of power within the renewable energy sector. In recent years there have been an increasing number of reports of defective blades contributing towards turbine failure. At present, regular costly inspections are conducted on turbine blades to ensure structural integrity and prevent degradation due to fatigue or impact.
Within larger blade designs, there exists a complex composite structure of glass fibre-reinforced plastic (GFRP) together with other materials. The objective of the WINTUR project was to focus on the development of a structural health monitoring (SHM) system for the purposes of defect detection on turbine blades. Application of an early warning system will extend the blade life expectancy, lower maintenance costs and optimise the efficiency of energy delivery to meet industry and community demands.
The WINTUR project was successful and achieved all objectives. The demonstration on the wind turbine blade confirmed that the SHM development is an effective and reliable inspection device that can perform quick and simple inspection of composite components. The WINTUR project research involved analysis of composite materials when excited by acoustic emission (AE) and long range ultrasonic (LRU) sensors. Signal processing techniques were applied to achieve the signal characteristics most sensitive to material changes within the structure. An analysis of sensor types was conducted with consideration of defect sensitivity, flexibility, conformability. Macrofibre composites (MFCs) provided a good match, detecting both AE and LRU activity. For low-power consumption, energy harvesting (EH) techniques were applied by harnessing the available energy in the surrounding environment. Since EH is an embryonic technology, two harvesting techniques were explored, namely amplified piezoceramic actuators (APAs) were used as well as MFCs. Both systems exploit the vibrations and loading of the blade to which they are attached. Wireless communications technology was developed for the purposes of data transfer from the blade to a control centre with accompanying software for data management and signal visualisation.
(1) Dual-purpose sensor array using AE and LRU to successfully detect defects with good sensitivity. The final trials proved that the WINTUR developed technology is able to detect:
- growing defects up to a distance of 3.8 m,
- blade intrusive-damage down to a thickness of 2 mm,
- impact damage up to a distance of 800 mm.
(2) Signal processing analysis tools able to successfully locate the area of defect with the criteria listed in 1).
(3) Wireless communications for data transfer up to speeds of 2 Mbits per second in 20 seconds maintaining 100% signal integrity.
(4) Two EH systems:
- APAs able to charge the harvester to 5.6 V for a 150 Hz vibrations of 150 mV amplitude with a storage time of 13 s;
- MFCs able to generate 0.5 mWs per deformation cycle at 0.5 Hz - able to power the transmission of 8 datagrams to a receiver node over a distance of 150 m.
(5) A graphical user interface (GUI) providing signal visualisation, a visual status indicator, easy-to-use navigation tools, tabulated display of critical information and a reporting facility.
The WINTUR project has been very successful and achieved all of its objectives. The demonstration on the wind turbine blade confirmed that the SHM development is an effective and reliable inspection device. It is a unique system that can perform quick and simple inspection of composite components.
The public website address for the project is http://www.WINTUR-project.com/ and will be kept as the main platform for any communications related to the WINTUR project beyond the project term.
WINTUR is collaboration between the following research organisations and small and medium-sized enterprises (SMEs) from 6 different European Union (EU) countries: TWI LIMITED, OPTEL sp.Z o.o. SMART MATERIAL GMBH, CEDRAT Technologies, Miyama Composites LTD, Encocam LIMITED, Solent Composite SYSTEMS LTD, Scottish and Southern Energy, Cereteth, Kaunas University of Technology.
Project context and objectives:
The main objective of the project is for the consortium to develop an advanced integrated system for real-time SHM and impending failure detection for wind turbine blades using LRU Technique (LRUT) and AE. The capability of the system will be demonstrated in laboratory experiments at the end of the project.
The project objectives for this period are presented below:
(a) choosing the most appropriate communication system for the data transfer on the blade and data transfer to the wind farm's data centre (work package (WP4));
(b) developing advanced signal processing tools and software (WP5 and WP6);
(c) carry out laboratory and field trials (WP7 & WP8);
(d) produce a plan for use and dissemination of knowledge (WP9).
To accomplish the strategic objectives, the work activities have been organised into a number of discrete WP. Those were divided into a data acquisition and design phase (WPs 1 - 6) and the implementation phase (WPs 7 - 9).
The project has suffered a number of unexpected setbacks during the second half of the project mainly due to partnership changes. As a result it was requested to extend the project end date from month 24 to month 26. The provided extension has given consortium the opportunity to achieve the project objectives for the second period successfully.
Project results:
The first year of the project concentrated mainly on the three first WPs. In the first six months of the project, the majority of the activity focused on WPs 1 and 2. At the 6 month meeting, WP1 was completed and WP2 was in good progress. At this date, the progress and achievements of the project were satisfactory. At the seventh month stage, the project had by then lost two partners: both Nexus Engineering and Ultra Electronics BCF had withdrawn from the project for financial reasons. The investigation into AE techniques was subsequently delayed whilst attempts to find a replacement partner proceeded. WP3 started promptly on month 4 of the project but only with respect to LRU analysis. Periodic Report 1 comprehensively covered WPs 1, 2 and 3.
Subsequent work packages were reported in Periodic Report 2. WP4 on the pulser / receiver unit, EH and wireless system transfer commenced on schedule on month 2 but extended beyond schedule into reporting period two as the scope of the research into EH was extended to investigate two different types of systems. This was executed to ensure satisfactory delivery to the project given the status of the state-of-the-art at time of investigation. WP4 was subsequently delivered in month 12. WP5 on signal enhancement, multi-parametric analysis and flaw location and classification commenced on schedule on month 4 and was reported in Deliverable D5.1 submitted in month 12. Subsequently it was decided that the WP could be further optimised by application of additional signal processing techniques.
WP6 on software development commenced on month 9 and was scheduled to progress to month 14. This work ran slightly over schedule. The European Commission (EC) officer agreed to an extension on the delivery date of the associated report whilst the GUI was optimised to include additional detailed information to assist the system user. The deliverable report was submitted on month 16.
System integration and testing commenced on time according to the revised schedule where sensors were bonded to a full-scale wind turbine blade. This process included analysis of the onset of defects. This concluded the first phase of system tests. The second phase of tests took place one month later when AE sensors, wireless capability and EH techniques contributed to the system. The associated deliverable report was subsequently delayed and is now ready for submission together with the deliverable report for WP8 regarding the demonstration. The demonstration included an optimised integrated system with additional EH contribution.
WP1: Project specifications
Mainly consisted of discussions with consortium partners and research performed by SSE on their fleet. The aims were to establish the industrial requirements for the system to be developed and to establish the component parts of a turbine blade that typically experiences the onset of defects.
At the 3-month meeting, SSE presented the then current monitoring practices (Task 1.1). This involved inspections which are made on an irregular basis and information about blade condition, inferred via a check on the turbine performance curves. Stationary blade inspections are typically performed visually using a tripod mounted telescope and camera. When cracking is observed, practice is to consult the original equipment manufacturer (OEM) either under the terms of a warranty agreement or on a consultancy basis to decide on the appropriate action given the size and location of cracking.
SSE performed an analysis of their fleet. Not all sites were able to contribute given the timescales and data available. The majority of faults reported were cracks in the blade. The WINTUR consortium discussed the inspection possibilities to be able to detect such defects as and when they occur and could be able to identify the crack at an earlier stage in the development cycle.
At project start, SSE's wind turbine fleet was onshore and yearly temperature profiles of a wind farm at Hadyards Hill (from Met Office records) were presented. More extreme temperatures and conditions are experienced in mountainous regions. Using this data for SSE's onshore wind turbine fleet, the operating ambient temperature range requirement for any instrumentation is between - 30 and + 35 degrees of Celsius. However, data on actual blade temperatures in operation was not available within SSE. Also, temperature data for offshore sites was not available, however; analysis by Cereteth revealed that the temperature ranges would be reduced in British coastal waters as sea temperatures have a lower year round variation than those over land. This contribution completed Task 1.2.
For Task 1.3 Encocam produced a report (WINTUR_Deliverable_D1.1_Cellbond) on the behaviour of composite material and fatigue loading. It was reported that composites fatigue cannot be ignored; a necessary understanding of the mechanisms by which fatigue damage occurs is required. From this, procedures could be prepared to predict the development and accumulation of damage providing information of the life-expectancy of the material. In conjunction, Miyama produced a report (WINTUR_Deliverable_D1.1_Miyama) on the research of typical defects found in composite material used for wind turbine blades. Ultra BCF performed research on lightening. According to standard IEC TR 61400-24, a wind turbine can be divided into different lightning protection zones in order to provide efficient protection of all components on the wind turbine.
TWI and Encocam held a meeting to discuss the specifications (shape, size, material) of the test samples. SCS provided details on the cross section of a large wind turbine blade. The samples were agreed and plans for manufacture initiated. KTU provided a report (WINTUR_Deliverable_D1.1_KTU) on modelling. The report concluded that the most promising method for assessing Guided waves propagating in wind turbine blades were SAFE and finite element (FE) methods. It was also shown that it was necessary to determine the propagating modes, the dispersion curves of phase and group velocities, attenuation, penetration and sensitivity to the defect. Smart Material carried out a research on various different types of MFC transducers that could be used for the WINTUR technology. The selected types are M2807-P1 & M2814- P2.
CEDRAT provided a report on the research carried out on the EH using piezoelements (WINTUR_Deliverable_D1.1_Cedrat).
In summary, the WINTUR system must be able to endure the harsh and remote locations in which wind-farms are located. The monitoring system must remain operational within the 30 and + 35 degree-of-Celsius temperature range. A majority of reports identify cracks on turbine blades. Manufacture of 2 x 1 m GFRP and sandwich samples were created on which to test the WINTUR sensors.
WP2: Theoretical study and modelling
The objective of WP2 was to develop a theoretical model to understand the properties of Guided waves in composite materials and provide a numerical simulation of the novel sensors. This analysis assessed the interaction of guided waves with defects and help improve the probability of defect detection.
Task 2.1 focuses on the properties of guided waves propagating through composite materials. This analysis involved identification of wave modes, wave-speeds and frequency response.
By modelling and through experiments it was demonstrated that using a single contact type transmitter, it is mainly the A0 and S0 guide wave modes in the frequency range 50 - 300 kHz that are excited. In some cases other modes were also excited (but not identified). However, they possess too small amplitudes to be applicable for measurements. The experiments demonstrated that it is best to perform measurements along the blade, because the waves propagating across the blade are almost completely suppressed at the boundary between two types of the composite structures. It was concluded that attenuation measurements showed guided waves propagating further along the length of the blade than those propagating across the blade. For the same propagating distance, the attenuation loss for the S0 mode is 50dB less than the equivalent for the A0 mode revealed the S0 mode. The dispersion curves for the wave modes propagating within the blade are observed in figures included in the 'WINTUR_Deliverable_D2.1' report.
Task 2.2 assesses the performance of two different types of transducer: piezoceramic and MFC. The former preferentially excites out-of-plane modes, e.g. A0 mode, whilst the latter preferentially excites in-plane S0 modes. It had already been identified that S0 modes suffer less attenuation loss than A0 modes. This, together with the knowledge that MFC transducers are light, unobtrusive, flexible and easy to bond, suggested that this is best sensor for defect detection on a turbine blade. Smart Material researched the field pattern of MFC sensors in order that an appropriate array of sensors could be designed for purpose.
In summary, the WINTUR modelling analysis demonstrated that using a single contact type transmitter, it is mainly the fundamental guided wave modes - A0 and S0- that are excited within the frequency range 50 - 300 kHz. Subsequent analysis recorded the following estimation of propagation losses:
- A0 guided wave mode attenuation of approximately 130 dB / m.
- S0 guided wave mode attenuation of approximately 80 dB / m.
Numerical and experimental analysis suggests that maximal distances at which measurements can be performed are between 0.5 m and 1 m using a single contact type transmitter.
WP3: Develop and evaluate transducer and sensors
The objective of WP3 was to develop novel sensors for combined AE and LRU techniques. Transducers were developed that are able to harvest energy from the flexing turbines.
Guided wave analysis on composite materials was reported in Periodic Report 1. This included analysis on attenuation and modal analysis.
The characterisation analysis achieved in reporting period 1 was then used to inform the design of an array of MFCs suitable for the detection of propagating S0 lamb waves - this mode was established as being the mode most sensitive to boundary and material changes. This array successfully detected defects within composite materials.
AE tests were performed to assess the two-dimensional real-time location on the surface of the composite sample. The AE measurements were performed using R15i - AST broadband sensors (250 - 500 kHz).This involved the successful application of sensor attachment, surface coupling and calibration with Hsu Neilson sources (0.3 mm lead break tests) on the surface to be examined to maximise the detection sensitivity to material defects. Attenuation and velocity measurements were taken in order to establish the expected coverage of the AE system. The maximum zonal threshold for a 50 dB signal is 1.5 m - the level of attenuation over that distance was 40 dB. The average velocity of the leading wave component was 3 908 m / s which matched the S0 velocity of the LRU signal. The LRU and AE systems therefore validated one another providing confidence in the system integrity. The AE system software, AEwin, was able to identify the guided wave signals as being separate from the relevant AE data.
TRF techniques were applied to the composite sample after cross-correlation identified the defect location. The received signal was reversed and retransmitted. However this approach was not successful. Other research has used laser interferometers for this purpose, however; this is not a practical solution. The success of TRF is debatable since only signals that have been determined from known defects are time reversed - however, an improved signal-to-noise ratio (SNR) is irrelevant if a defect has already been detected.
Due to withdrawal of Nexus from the project, the AE aspect of the project was not progressed. This resulted in a search for a suitable replacement partner - and when no suitable candidate could be identified, it became a requirement to reallocate this work to TWI. Since this work was not in TWI's original list of tasks, it had an impact on the schedule of completed work and Deliverable Reports.
A fully functioning circuit for synchronised EH of electrical power from ambient mechanical vibrations was developed, in conjunction with CEDRAT's APA 400M-MD device. The circuit has been analysed theoretically and at a hardware level. The result was an optimised EH system sensitive to the response of displacement affecting the material to which the sensor is attached and can harness energy from vibrational fields from a wide range of frequencies. The energy yield is typically dependent on the level of vibration under which the structure is subjected. The electronic circuitry was successfully miniaturised and consumed significantly less energy.
Smart Material developed an EH system using MFCs as energy harvesters. The M8528-P2 MFC was used to power a sensor node with a connected temperature sensor to determine the amount of energy which can be harvested with different types of MFCs at different frequencies.
In summary, the dual-action WINTUR sensors are able to detect:
- AE activity from impact excitation up to 800m from the source,
- LRU signals from a growing defect up to a distance of 3.8 m.
The sensors are lightweight, unobtrusive, conformable and provide good in-plane mechanical coupling an provide reliable early indication of the onset of defects. Two EH systems are researched:
- APAs - on a sub-woofer - able to charge the harvester to 5.6 V for 150 Hz vibrations of 150 mV amplitude with a storage time of 13 s,
- MFCs - on a 1:14 scaled blade model - able to generate 0.5 mWs per deformation cycle at 0.5 Hz - able to power the transmission of 8 datagrams to a receiver node over a distance of 150 m.
WP4: Develop pulser / receiver and communication system
Task 4.2 and Task 4.3 delivered the capability of data transfer from one part of turbine to another using a short-range low-power wireless communications system using Nordic Semiconductors' NRF24L01+ (802.11 protocol), 2.4 GHz range in order to attain the necessary SNR and signal immunity from interference. The communications range was assessed with the successful transfer of data over a distance of 40 m. The wireless PCB was packaged and integrated with the EH circuitry to provide a modular, lightweight and unobtrusive design suitable for implementation within a turbine blade.
Task 4.4 provided the database design for data management. The database stores information concerning user authentication, wind generators, blades, sensor arrays, received LRU measurements, defects, reports, alerts and data meta-tagging. A description of the database tables and stored procedures is detailed further in this section along with a brief description of the capabilities this design attributes to the overall system.
Each table described below has a unique primary key for each entry. That key is the ID column:
- 'Arrays' table stores information concerning blade-installed sensor arrays.
- 'Blades' table provides information on the turbine blades with related turbine ID.
- 'Generator' table provides information on the wind turbine.
- 'Defects' table stores information on detected defects from the automated defect detection and sizing system.
- 'DefTypes' table will be used to help identify any defects based on library entries.
- 'Reports' table represents reports produced by inspectors. It stores a description of the report, the number of detected or candidate defects, a percentage value illustrating the severity of the results, the date and time of reporting and finally the wind generator the report corresponds to.
- 'Tags' table is used to store information on meta-tags added on visualised measurements.
- 'Users' table stores information on registered users of the system.
- 'ULevels' table contains labels for all user categories when queries are executed.
- 'Permission' table allows mapping of permitted users to wind generators and their sub-components in sequence.
- 'Alerts' table logs information on critical messages for display purposes.
In summary, The WINTUR wireless communication system has the capability to successfully transmit data over a distance of 40m maintaining 100 % signal integrity. The communications link (Nordic NRF24L01+) - the fastest available for low power transmission - is able to transfer data up to speeds of 2 Mbits per second in under 20 seconds. A 'sleep mode' function between data-transfer ensures power consumption is close to zero.
WP5: Develop advanced signal processing and analysis tools and flaw sizing
Signal processing techniques were developed and implemented to extract from the acquired time-series data meaningful parameters that can be easily interpreted to provide meaningful information on the health of the composite blade under inspection.
Composite materials usually results in higher attenuation of ultrasonic waves and as consequence to noisy signals. However, the noise was easily reduced using averaging techniques and thus increased the SNR of the received signals. Wave mode identification was achieved analysing time-of-arrival and dispersion curves.
The cross-correlation function was used and calculates between the measured (monitored) signal and the reference signal. This process proved to provide a distinction between patterns depending upon the material characteristics. However, the reduction of the cross-correlation can be caused also by other factors such as temperature or surface water.
Several techniques were analysed to contribute toward a multi-parametric analysis. These include two-dimensional (2D) fast-Fourier transforms (FFTs) to identify propagating modes, the optimisation technique enabling determination of the attenuation coefficient, and application of B-scan to determine parameters of the propagating guided waves.
In summary, The WINTUR system analysis tools provide an early indication of the onset of blade defects. In addition to signal visualisation, a visual display of the blade status is provided with colour scale system indicating to the user the health of the blade. Processing data from the WINTUR sensors provides good sensitivity providing successful detection:
- of growing defects up to a distance 3.8 m,
- of Blade intrusive-damage down to a thickness of 2 mm,
- of impact excitation up to a distance of 800 mm.
The WINTUR analysis tool is also able to accurately locate the position of defects
WP6: Software development
A GUI was developed (using Microsoft .NET 3.5 Framework), capable of visualising acquired data from sensors on wind generators. The GUI permits the user to connect to the WINTUR database and retrieve data on demand and perform common processing tasks. Once past security protocol and connection is achieved, lists of inspected components are retrieved from the database. The list includes identification of the wind turbines, turbine blades, sensor arrays with the option of the provision of compiled reports using drop-down menus. Graphical visualisation of the acquired AE / LRU signals is also provided. Navigation tools have been developed to permit the user to navigate the signal plot and export the measurements in a common ASCII file format.
The designed database - developed using the MySQL database server and workbench application provided by the MySQL - stores information concerning user authentication, wind generators, blades, sensor arrays, received LRU measurements, defects, reports, alerts and data meta-tagging.
Description of the database tables and stored procedures.
In summary, the WINTUR systems GUI provides easy manipulability of received data providing a visual representation of the inspected blade and its condition. Easy-to-use navigation tools with pull-down menus and a request facility provides the user with access to relevant data. Signal visualisation of the received system data and blade integrity representation is provided to indicate the health of the blade. Time-stamped status reports provide a summary health analysis detailing defect size, defect location, temperature, blade and turbine identification.
WP7: System integration and testing
System integration trials were tried out on the purchased 9.8 m blade purchased by Miyama and residing in Chalkida, Greece. The progress on a system for defect detection in wind-turbine blades was tested with reference to the design, study and implementation to perform non-destructive testing (NDT) testing on composite materials within wind turbine rotor blades. These tests involved
- Successful integration of the dual sensor capability involving AE and LRU. MFCs were successfully connected to the AE system to provide a dual application sensor.
- Testing of an EH system using APAs for integration with a miniaturised, flexible electronics circuit sensitive enough to harness energy from the displacement activity on a sub-woofer over a wide range of frequencies.
- Testing of the low-power wireless communications system using the Nordic Semiconductors' NRF24L01+ (802.11 protocol) for data transfer.
- Navigation using the GUI permitting data visibility and manipulation. Availability of information was only provided upon user authentication. A date-stamped summary report provided a status update on the health of the blade under inspection.
Additional required work to validate included additional testing of the wireless link after a little data corruption at the receiving node over a distance of 40 m. Further validation of the spatial array was also required with further work on EH provision to be completed.
In summary, the integrated WINTUR system provides a reliable early indication of the onset of blade defects. Visual display of the received data and the blade integrity indicates to the user the latest status:
(a) on an area up to 3.8 m on growing defect detection;
(b) on invasive damage down to 2 mm thick;
(c) on an area up to 800 mm on impact damage;
(d) location of defect.
The WINTUR wireless communications system is able to successfully transmit the acquired data at a speed of 2 Mbits per second in 20 seconds.
The WINTUR EH system (APAs) was able to charge the harvester to 5.6 V for a 150 Hz vibrations of 150 mV amplitude with a storage time of 13 s. Lowering the frequency of the 150 mV vibrations to 105 Hz increases the storage time to 70 s.
WP8: Demonstration and large scale laboratory trials
The demo was very successful with satisfaction from each of the SMEs at the level of development achieved. Each of the technology aims were accomplished and it was concluded by the consortium that a viable system for blade monitoring to aid the wind sector attain the desired energy delivery to business and communities whilst reducing operational and maintenance costs could be achieved.
The following system objectives were successfully achieved:
(1) Novel MFC transducers installed on a turbine blade, able to detect the onset of damage that is the occurrence of fibre breakage due to staged development of a hole-defect. Changes in amplitude and changes in boundary condition indicated the presence of defect within the sensor vicinity - a distance of up to 3.8 m for the guided wave sensors facility and 1.5 m for the AE sensor facility. Matrix capture of the spatial array of sensors using signal processing techniques (cross-correlation) successfully achieved defect detection and accurate location on the wind turbine blade. This was the culmination of work from WPs 1, 2, 3 and 5.
(2) EH (developed for the purposes of powering the pulser/receiver unit) successfully harnessed energy using APAs from a simulated ambient mechanical environment (220W sub-woofer) over a wide range of frequencies (35 and 300 Hz). The system did not yield high energy at low vibrational frequencies (8 - 15 Hz). This was expected since this frequency is an order of magnitude lower than the resonance of the APA actuator.
An alternative EH system with mounted MFCs (M8528-P2 MFC) on a scaled blade model (1:14) was explored to assess the yield from a dynamic environment of transient loads and semi- permanent loads.The MFC system was able to yield 0.5mWs per deformation cycle at 0.5 Hz.
(3) The wireless communications system transmitted blade-acquired Guided wave data from a transmitting node to a receiving node located 40 m away. The two sets of data correlated perfectly. The received data was then 'called up' by the data central software programme and visually displayed as time series data. The GUI displayed the blade details and the sensor details and created a time-stamped receipt with user information.
The WINTUR consortium witnessed a successful demonstration of an NDT monitoring system that is fully integrated and applicable to detecting the types of defects typically found in operational turbine blades.
The trials were completed by providing consortium partners with a demonstration on how to use each stage of the inspection system. This was completed through a mixture of a live demo and presentations.
In summary, the complete WINTUR system provides an early defect detection system comprising:
(1) A visual blade display indicating latest blade status over an area of 3.8 m for growing defects; 800mm for impact damage, and sensitivity of fibre-breakage down to 2 mm thick. Location of the defect is also provided.
(2) Wireless communications for data transfer up to speeds of 2 Mbits per second in 20 s.
(3) Two EH systems:
(a) APAs able to charge the harvester to 5.6 V for a 150 Hz vibrations of 150 mV amplitude with a storage time of 13 s;
(b) MFCs able to generate 0.5 mWs per deformation cycle at 0.5 Hz - able to power the transmission of 8 datagrams to a receiver node over a distance of 150 m.
(4) A GUI providing signal visualisation, a visual status indicator, easy-to-use navigation tools, tabulated display of critical information and a reporting facility.
WP9: Exploitation and dissemination
The work performed during the first period informed the dissemination and exploitation activities in the second half.
Marketing material was created for the purposes of dissemination at conferences and exhibitions.
Technical papers were presented at the following conferences:
- BINDT, 2011, United Kingdom (UK)
- Engineering Structural Integrity Assessment 2011, UK
- BINDT, 2010, UK.
The following conferences were attended by WINTUR partners where dissemination of the project took place:
- ISPA - International Symposium on Piezocomposite Applictaions, 2011, Germany
- IDTechEx- EH & Storage Europe, 2011, Germany
- Wind Turbine 1st technology Forum, 2011, Italy
- EWEA 2011, Belgium
- Sensors Expo & Conference, 2010, United States (US)
- SPIE Smart Structures / NDE, 2010, US
- Smart 2009, Portugal
Further details on these events are provided in Section 4.2 (A1) and in Deliverable D9.2 which includes information on conference papers, flyers and feedback via the project website.
The main dissemination will continue with the WINTUR project website beyond the project term. The established partner benefits include:
(a) ownership and licensing status of each exploitable product with particular detail;
(b) foreground IP from the successful exploitation of the Background IP;
(c) benefits of exploitable IP and successful implementation of technical integration for the reliable detection and prediction of defects within composite blade components;
(d) an approach for exploitation of the integrated WINTUR technology beyond the project term and an established path permitting SMEs to serve new markets;
(e) established routes for further dissemination of the project achievements through workshops, publications, conferences and the WINTUR website.
The main results of the project and therefore the key selling points of the WINTUR technology were identified by the consortium as below:
- Reliable blade inspection - reducing operational and maintenance costs by minimising downtime due to advanced defects causing critical component failure. Harnessing of environmental energy to ensure low sensor power consumption and avoid routine battery replacement.
- Dual-purpose sensors - proven integration of AE and LRU technologies able to detect and locate flaws in turbine and extending component life-expectancy by detecting flaws at an incipient stage of development. Reduced onsite maintenance: reducing the frequency of labour-intensive inspection providing efficient maintenance service for operators. Alternatively, regular tests can be run safely from a central base.
Data transfer without human intervention - remote inspection (via wireless transmission) to provide processed blade integrity data for high level decision making. Reduced insurance premiums - early detection limits damage to component parts thereby extending their life-expectancy and reducing insurance premium costs. The overall exploitation and dissemination planning in WINTUR has been geared towards maximising the economic benefit to the SMEs as quickly as possible following the end of the project. Marketing flyers in different languages to promote WINTUR technology have also been created for wider audience dissemination.
Dissemination was further provided via the project website where regular project updates were reported. The consortium member sites contained a link to the project website to attract more hits. A workshop held in Greece was also used to establish wider recognition for the work conducted within the wider renewable energy industry.
Potential impact:
The WINTUR project has been very successful. The field trials carried out on the purchased wind turbine blade confirmed that the SHM development is an effective and reliable inspection device that has the potential to be of considerable interest for the wind energy industry. It is a unique system that can perform quick and simple inspection of composite components.
The consortium had identified the need for such a system for utilisation in what is an expanding wind energy market. The development of the project has been widely available on the WINTUR website (see http://www.wintur-project.com/ online) launched by TWI in the opening weeks of the project in order to facilitate project dissemination. A Google search of the word 'WINTUR' provides details of the project as the first item. The website was maintained by the project coordinator with the facility for updates by consortium partners. It remains the principal internet platform where information, contacts, news and developments about the WINTUR project system were accessible to the public for the duration for the project and will remain so beyond project completion. In addition, all partners were asked to add a link of the WINTUR project website onto their respective websites. Whilst conducted by all partners, Optel together with Miyama have provided the momentum behind dissemination activities.
The investment cost for such a system is considered to be high in the current status. This is mainly due to the electronic instrumentation that can be easily replaced by a more simple and affordable instrument. At the final stage of the WINTUR project, the price of the prototype was discussed but not defined. It was agreed that a drive to lower the costs should be taken seriously without the threat to system capability and integrity.
In terms of benefits, this project will benefit the SMEs and reduce long-term costs for utility companies. There is currently no standard condition monitoring technique available that can perform routine autonomous inspection providing details of the blade integrity. The technique presently used is visual inspection - a relatively slow, expensive technique of providing the user with the blade information sought. At present, the industry deploys hands-on-action as deemed necessary - usually long before a critical failure occurs. The aim is to confine repairs to straight-forward up-tower repairs to prevent costly down-tower serious repairs that may require extended shutdowns.
The benefits of fitting an SHM system such as WINTUR will become apparent to the industry when a way to calculate the point at which a condition monitoring system becomes cost-effective from the perspective of the owner-operator. There are a variety of reasons for the delay in standard implementation of a condition monitoring system. These include the complicated nature of wind turbines, the remoteness of their installations and the random nature of the industry's early years. Indeed, the argument regarding the remoteness of installations cuts both ways and can be converted into a case for a cost-saving monitoring system. The viewpoint that offshore is ideally suited to SHM is somewhat comprehensive throughout the industry. This is where a monitoring system like that developed by the WINTUR consortium has a part to play. An additional strength of the WINTUR project is the high-level statistical matching (via cross-correlation) and 'duality' (AE combined with LRU) to avoid misrepresentation and reduce the risk of overabundant supply of information.
Contact has been established with Vestas one of the world's largest turbine manufacturers who expressed considerable interest in the techniques utilised in WINTUR project. This communication is ongoing and will establish a route for feedback regarding OEM's concerns and requirements.
There is further evidence that a project such as WINTUR will find a market-place in the medium-term: statistics suggest that large numbers of wind turbines will come out of warranty within the next couple of years. Costs will then automatically transfer to the owners (i.e. utility companies). When the costs of repairs become apparent to owners, mass implementation of condition monitoring will take place. Thus, the SMEs within the WINTUR consortium are well-placed to exploit this cycle of maturity set to occur within the wind industry.
A workshop held simultaneously with the project demonstration attracted the interest of Creative Systems Engineering (CSE) - energy consultant in Greece. This company was impressed with the technology developed and believes that the WINTUR project does have a part to play when the roll-out of condition monitoring systems gathers momentum.
The WINTUR consortium recognised and agreed the importance of understanding the strength, weakness of the developed technologies and to identify the opportunities and threats for the end results as a realisation process. The strengths, weaknesses, opportunities, and threat (SWOT) analysis carried out in the first reporting period was the first step of strategic planning. It helped all partners to be aware of the different aspects of the market and guide the project to focus on key issues.
Threats for the system are not overly critical and depend upon the way the information resulting from the WINTUR project will be disseminated and exploited.
The weaknesses presented can be solved with additional resources and time to enhance the system. The price of the WINTUR system has to be addressed to make it a viable option in the industry sector.
The SWOT analysis shows that the system opens the door to many other opportunities. With further improvements, additional features can be brought to the WINTUR system and this will contribute to increased viability.
After lengthy discussion with the WINTUR partners, it was concluded that there are two ways in which the WINTUR project can be brought to the market:
(1) As a factory-installed system controlled by the OEM: However, one 'threat' of this approach is that it permits the OEM to control the data and hide flawed equipment. This can be averted by buyers through a process of negotiation at the contractual stage.
(2) Retro-fitting the system onto ageing turbines: Whilst this may not be possible whilst the turbine is under warranty, when the owner takes full responsibility for turbine operation, there is considerable scope for mass implementation.
A number of dissemination activities were performed during the course of the project to expose the project results and to make the developments visible to the industry. Conferences and seminars related to wind energy technology were attended by the project partners and presentations were given. However, since information explaining how to use the WINTUR technology is commercially sensitive (because the WINTUR consortium is planning to exploit the results beyond the project), it was decided that this information will not be published in any format (report, guideline or wall chart) other than in the deliverable reports only visible to project partners.
Flyers were the main marketing tools for the project and were handed out at conferences by partners. Dissemination activities after completion of the project were considered critical for the successful implementation and commercialisation of the product.
Contact details:
Kenneth Burnham
kenneth.burnham@twi.co.uk
Project Coordinator
TWI
Granta Park,
Cambridge, CB21 6Al
Tel: +44-012-23899334
List of websites: http://www.wintur-project.com
Wind energy is an increasingly important contributor of power within the renewable energy sector. In recent years there have been an increasing number of reports of defective blades contributing towards turbine failure. At present, regular costly inspections are conducted on turbine blades to ensure structural integrity and prevent degradation due to fatigue or impact.
Within larger blade designs, there exists a complex composite structure of glass fibre-reinforced plastic (GFRP) together with other materials. The objective of the WINTUR project was to focus on the development of a structural health monitoring (SHM) system for the purposes of defect detection on turbine blades. Application of an early warning system will extend the blade life expectancy, lower maintenance costs and optimise the efficiency of energy delivery to meet industry and community demands.
The WINTUR project was successful and achieved all objectives. The demonstration on the wind turbine blade confirmed that the SHM development is an effective and reliable inspection device that can perform quick and simple inspection of composite components. The WINTUR project research involved analysis of composite materials when excited by acoustic emission (AE) and long range ultrasonic (LRU) sensors. Signal processing techniques were applied to achieve the signal characteristics most sensitive to material changes within the structure. An analysis of sensor types was conducted with consideration of defect sensitivity, flexibility, conformability. Macrofibre composites (MFCs) provided a good match, detecting both AE and LRU activity. For low-power consumption, energy harvesting (EH) techniques were applied by harnessing the available energy in the surrounding environment. Since EH is an embryonic technology, two harvesting techniques were explored, namely amplified piezoceramic actuators (APAs) were used as well as MFCs. Both systems exploit the vibrations and loading of the blade to which they are attached. Wireless communications technology was developed for the purposes of data transfer from the blade to a control centre with accompanying software for data management and signal visualisation.
(1) Dual-purpose sensor array using AE and LRU to successfully detect defects with good sensitivity. The final trials proved that the WINTUR developed technology is able to detect:
- growing defects up to a distance of 3.8 m,
- blade intrusive-damage down to a thickness of 2 mm,
- impact damage up to a distance of 800 mm.
(2) Signal processing analysis tools able to successfully locate the area of defect with the criteria listed in 1).
(3) Wireless communications for data transfer up to speeds of 2 Mbits per second in 20 seconds maintaining 100% signal integrity.
(4) Two EH systems:
- APAs able to charge the harvester to 5.6 V for a 150 Hz vibrations of 150 mV amplitude with a storage time of 13 s;
- MFCs able to generate 0.5 mWs per deformation cycle at 0.5 Hz - able to power the transmission of 8 datagrams to a receiver node over a distance of 150 m.
(5) A graphical user interface (GUI) providing signal visualisation, a visual status indicator, easy-to-use navigation tools, tabulated display of critical information and a reporting facility.
The WINTUR project has been very successful and achieved all of its objectives. The demonstration on the wind turbine blade confirmed that the SHM development is an effective and reliable inspection device. It is a unique system that can perform quick and simple inspection of composite components.
The public website address for the project is http://www.WINTUR-project.com/ and will be kept as the main platform for any communications related to the WINTUR project beyond the project term.
WINTUR is collaboration between the following research organisations and small and medium-sized enterprises (SMEs) from 6 different European Union (EU) countries: TWI LIMITED, OPTEL sp.Z o.o. SMART MATERIAL GMBH, CEDRAT Technologies, Miyama Composites LTD, Encocam LIMITED, Solent Composite SYSTEMS LTD, Scottish and Southern Energy, Cereteth, Kaunas University of Technology.
Project context and objectives:
The main objective of the project is for the consortium to develop an advanced integrated system for real-time SHM and impending failure detection for wind turbine blades using LRU Technique (LRUT) and AE. The capability of the system will be demonstrated in laboratory experiments at the end of the project.
The project objectives for this period are presented below:
(a) choosing the most appropriate communication system for the data transfer on the blade and data transfer to the wind farm's data centre (work package (WP4));
(b) developing advanced signal processing tools and software (WP5 and WP6);
(c) carry out laboratory and field trials (WP7 & WP8);
(d) produce a plan for use and dissemination of knowledge (WP9).
To accomplish the strategic objectives, the work activities have been organised into a number of discrete WP. Those were divided into a data acquisition and design phase (WPs 1 - 6) and the implementation phase (WPs 7 - 9).
The project has suffered a number of unexpected setbacks during the second half of the project mainly due to partnership changes. As a result it was requested to extend the project end date from month 24 to month 26. The provided extension has given consortium the opportunity to achieve the project objectives for the second period successfully.
Project results:
The first year of the project concentrated mainly on the three first WPs. In the first six months of the project, the majority of the activity focused on WPs 1 and 2. At the 6 month meeting, WP1 was completed and WP2 was in good progress. At this date, the progress and achievements of the project were satisfactory. At the seventh month stage, the project had by then lost two partners: both Nexus Engineering and Ultra Electronics BCF had withdrawn from the project for financial reasons. The investigation into AE techniques was subsequently delayed whilst attempts to find a replacement partner proceeded. WP3 started promptly on month 4 of the project but only with respect to LRU analysis. Periodic Report 1 comprehensively covered WPs 1, 2 and 3.
Subsequent work packages were reported in Periodic Report 2. WP4 on the pulser / receiver unit, EH and wireless system transfer commenced on schedule on month 2 but extended beyond schedule into reporting period two as the scope of the research into EH was extended to investigate two different types of systems. This was executed to ensure satisfactory delivery to the project given the status of the state-of-the-art at time of investigation. WP4 was subsequently delivered in month 12. WP5 on signal enhancement, multi-parametric analysis and flaw location and classification commenced on schedule on month 4 and was reported in Deliverable D5.1 submitted in month 12. Subsequently it was decided that the WP could be further optimised by application of additional signal processing techniques.
WP6 on software development commenced on month 9 and was scheduled to progress to month 14. This work ran slightly over schedule. The European Commission (EC) officer agreed to an extension on the delivery date of the associated report whilst the GUI was optimised to include additional detailed information to assist the system user. The deliverable report was submitted on month 16.
System integration and testing commenced on time according to the revised schedule where sensors were bonded to a full-scale wind turbine blade. This process included analysis of the onset of defects. This concluded the first phase of system tests. The second phase of tests took place one month later when AE sensors, wireless capability and EH techniques contributed to the system. The associated deliverable report was subsequently delayed and is now ready for submission together with the deliverable report for WP8 regarding the demonstration. The demonstration included an optimised integrated system with additional EH contribution.
WP1: Project specifications
Mainly consisted of discussions with consortium partners and research performed by SSE on their fleet. The aims were to establish the industrial requirements for the system to be developed and to establish the component parts of a turbine blade that typically experiences the onset of defects.
At the 3-month meeting, SSE presented the then current monitoring practices (Task 1.1). This involved inspections which are made on an irregular basis and information about blade condition, inferred via a check on the turbine performance curves. Stationary blade inspections are typically performed visually using a tripod mounted telescope and camera. When cracking is observed, practice is to consult the original equipment manufacturer (OEM) either under the terms of a warranty agreement or on a consultancy basis to decide on the appropriate action given the size and location of cracking.
SSE performed an analysis of their fleet. Not all sites were able to contribute given the timescales and data available. The majority of faults reported were cracks in the blade. The WINTUR consortium discussed the inspection possibilities to be able to detect such defects as and when they occur and could be able to identify the crack at an earlier stage in the development cycle.
At project start, SSE's wind turbine fleet was onshore and yearly temperature profiles of a wind farm at Hadyards Hill (from Met Office records) were presented. More extreme temperatures and conditions are experienced in mountainous regions. Using this data for SSE's onshore wind turbine fleet, the operating ambient temperature range requirement for any instrumentation is between - 30 and + 35 degrees of Celsius. However, data on actual blade temperatures in operation was not available within SSE. Also, temperature data for offshore sites was not available, however; analysis by Cereteth revealed that the temperature ranges would be reduced in British coastal waters as sea temperatures have a lower year round variation than those over land. This contribution completed Task 1.2.
For Task 1.3 Encocam produced a report (WINTUR_Deliverable_D1.1_Cellbond) on the behaviour of composite material and fatigue loading. It was reported that composites fatigue cannot be ignored; a necessary understanding of the mechanisms by which fatigue damage occurs is required. From this, procedures could be prepared to predict the development and accumulation of damage providing information of the life-expectancy of the material. In conjunction, Miyama produced a report (WINTUR_Deliverable_D1.1_Miyama) on the research of typical defects found in composite material used for wind turbine blades. Ultra BCF performed research on lightening. According to standard IEC TR 61400-24, a wind turbine can be divided into different lightning protection zones in order to provide efficient protection of all components on the wind turbine.
TWI and Encocam held a meeting to discuss the specifications (shape, size, material) of the test samples. SCS provided details on the cross section of a large wind turbine blade. The samples were agreed and plans for manufacture initiated. KTU provided a report (WINTUR_Deliverable_D1.1_KTU) on modelling. The report concluded that the most promising method for assessing Guided waves propagating in wind turbine blades were SAFE and finite element (FE) methods. It was also shown that it was necessary to determine the propagating modes, the dispersion curves of phase and group velocities, attenuation, penetration and sensitivity to the defect. Smart Material carried out a research on various different types of MFC transducers that could be used for the WINTUR technology. The selected types are M2807-P1 & M2814- P2.
CEDRAT provided a report on the research carried out on the EH using piezoelements (WINTUR_Deliverable_D1.1_Cedrat).
In summary, the WINTUR system must be able to endure the harsh and remote locations in which wind-farms are located. The monitoring system must remain operational within the 30 and + 35 degree-of-Celsius temperature range. A majority of reports identify cracks on turbine blades. Manufacture of 2 x 1 m GFRP and sandwich samples were created on which to test the WINTUR sensors.
WP2: Theoretical study and modelling
The objective of WP2 was to develop a theoretical model to understand the properties of Guided waves in composite materials and provide a numerical simulation of the novel sensors. This analysis assessed the interaction of guided waves with defects and help improve the probability of defect detection.
Task 2.1 focuses on the properties of guided waves propagating through composite materials. This analysis involved identification of wave modes, wave-speeds and frequency response.
By modelling and through experiments it was demonstrated that using a single contact type transmitter, it is mainly the A0 and S0 guide wave modes in the frequency range 50 - 300 kHz that are excited. In some cases other modes were also excited (but not identified). However, they possess too small amplitudes to be applicable for measurements. The experiments demonstrated that it is best to perform measurements along the blade, because the waves propagating across the blade are almost completely suppressed at the boundary between two types of the composite structures. It was concluded that attenuation measurements showed guided waves propagating further along the length of the blade than those propagating across the blade. For the same propagating distance, the attenuation loss for the S0 mode is 50dB less than the equivalent for the A0 mode revealed the S0 mode. The dispersion curves for the wave modes propagating within the blade are observed in figures included in the 'WINTUR_Deliverable_D2.1' report.
Task 2.2 assesses the performance of two different types of transducer: piezoceramic and MFC. The former preferentially excites out-of-plane modes, e.g. A0 mode, whilst the latter preferentially excites in-plane S0 modes. It had already been identified that S0 modes suffer less attenuation loss than A0 modes. This, together with the knowledge that MFC transducers are light, unobtrusive, flexible and easy to bond, suggested that this is best sensor for defect detection on a turbine blade. Smart Material researched the field pattern of MFC sensors in order that an appropriate array of sensors could be designed for purpose.
In summary, the WINTUR modelling analysis demonstrated that using a single contact type transmitter, it is mainly the fundamental guided wave modes - A0 and S0- that are excited within the frequency range 50 - 300 kHz. Subsequent analysis recorded the following estimation of propagation losses:
- A0 guided wave mode attenuation of approximately 130 dB / m.
- S0 guided wave mode attenuation of approximately 80 dB / m.
Numerical and experimental analysis suggests that maximal distances at which measurements can be performed are between 0.5 m and 1 m using a single contact type transmitter.
WP3: Develop and evaluate transducer and sensors
The objective of WP3 was to develop novel sensors for combined AE and LRU techniques. Transducers were developed that are able to harvest energy from the flexing turbines.
Guided wave analysis on composite materials was reported in Periodic Report 1. This included analysis on attenuation and modal analysis.
The characterisation analysis achieved in reporting period 1 was then used to inform the design of an array of MFCs suitable for the detection of propagating S0 lamb waves - this mode was established as being the mode most sensitive to boundary and material changes. This array successfully detected defects within composite materials.
AE tests were performed to assess the two-dimensional real-time location on the surface of the composite sample. The AE measurements were performed using R15i - AST broadband sensors (250 - 500 kHz).This involved the successful application of sensor attachment, surface coupling and calibration with Hsu Neilson sources (0.3 mm lead break tests) on the surface to be examined to maximise the detection sensitivity to material defects. Attenuation and velocity measurements were taken in order to establish the expected coverage of the AE system. The maximum zonal threshold for a 50 dB signal is 1.5 m - the level of attenuation over that distance was 40 dB. The average velocity of the leading wave component was 3 908 m / s which matched the S0 velocity of the LRU signal. The LRU and AE systems therefore validated one another providing confidence in the system integrity. The AE system software, AEwin, was able to identify the guided wave signals as being separate from the relevant AE data.
TRF techniques were applied to the composite sample after cross-correlation identified the defect location. The received signal was reversed and retransmitted. However this approach was not successful. Other research has used laser interferometers for this purpose, however; this is not a practical solution. The success of TRF is debatable since only signals that have been determined from known defects are time reversed - however, an improved signal-to-noise ratio (SNR) is irrelevant if a defect has already been detected.
Due to withdrawal of Nexus from the project, the AE aspect of the project was not progressed. This resulted in a search for a suitable replacement partner - and when no suitable candidate could be identified, it became a requirement to reallocate this work to TWI. Since this work was not in TWI's original list of tasks, it had an impact on the schedule of completed work and Deliverable Reports.
A fully functioning circuit for synchronised EH of electrical power from ambient mechanical vibrations was developed, in conjunction with CEDRAT's APA 400M-MD device. The circuit has been analysed theoretically and at a hardware level. The result was an optimised EH system sensitive to the response of displacement affecting the material to which the sensor is attached and can harness energy from vibrational fields from a wide range of frequencies. The energy yield is typically dependent on the level of vibration under which the structure is subjected. The electronic circuitry was successfully miniaturised and consumed significantly less energy.
Smart Material developed an EH system using MFCs as energy harvesters. The M8528-P2 MFC was used to power a sensor node with a connected temperature sensor to determine the amount of energy which can be harvested with different types of MFCs at different frequencies.
In summary, the dual-action WINTUR sensors are able to detect:
- AE activity from impact excitation up to 800m from the source,
- LRU signals from a growing defect up to a distance of 3.8 m.
The sensors are lightweight, unobtrusive, conformable and provide good in-plane mechanical coupling an provide reliable early indication of the onset of defects. Two EH systems are researched:
- APAs - on a sub-woofer - able to charge the harvester to 5.6 V for 150 Hz vibrations of 150 mV amplitude with a storage time of 13 s,
- MFCs - on a 1:14 scaled blade model - able to generate 0.5 mWs per deformation cycle at 0.5 Hz - able to power the transmission of 8 datagrams to a receiver node over a distance of 150 m.
WP4: Develop pulser / receiver and communication system
Task 4.2 and Task 4.3 delivered the capability of data transfer from one part of turbine to another using a short-range low-power wireless communications system using Nordic Semiconductors' NRF24L01+ (802.11 protocol), 2.4 GHz range in order to attain the necessary SNR and signal immunity from interference. The communications range was assessed with the successful transfer of data over a distance of 40 m. The wireless PCB was packaged and integrated with the EH circuitry to provide a modular, lightweight and unobtrusive design suitable for implementation within a turbine blade.
Task 4.4 provided the database design for data management. The database stores information concerning user authentication, wind generators, blades, sensor arrays, received LRU measurements, defects, reports, alerts and data meta-tagging. A description of the database tables and stored procedures is detailed further in this section along with a brief description of the capabilities this design attributes to the overall system.
Each table described below has a unique primary key for each entry. That key is the ID column:
- 'Arrays' table stores information concerning blade-installed sensor arrays.
- 'Blades' table provides information on the turbine blades with related turbine ID.
- 'Generator' table provides information on the wind turbine.
- 'Defects' table stores information on detected defects from the automated defect detection and sizing system.
- 'DefTypes' table will be used to help identify any defects based on library entries.
- 'Reports' table represents reports produced by inspectors. It stores a description of the report, the number of detected or candidate defects, a percentage value illustrating the severity of the results, the date and time of reporting and finally the wind generator the report corresponds to.
- 'Tags' table is used to store information on meta-tags added on visualised measurements.
- 'Users' table stores information on registered users of the system.
- 'ULevels' table contains labels for all user categories when queries are executed.
- 'Permission' table allows mapping of permitted users to wind generators and their sub-components in sequence.
- 'Alerts' table logs information on critical messages for display purposes.
In summary, The WINTUR wireless communication system has the capability to successfully transmit data over a distance of 40m maintaining 100 % signal integrity. The communications link (Nordic NRF24L01+) - the fastest available for low power transmission - is able to transfer data up to speeds of 2 Mbits per second in under 20 seconds. A 'sleep mode' function between data-transfer ensures power consumption is close to zero.
WP5: Develop advanced signal processing and analysis tools and flaw sizing
Signal processing techniques were developed and implemented to extract from the acquired time-series data meaningful parameters that can be easily interpreted to provide meaningful information on the health of the composite blade under inspection.
Composite materials usually results in higher attenuation of ultrasonic waves and as consequence to noisy signals. However, the noise was easily reduced using averaging techniques and thus increased the SNR of the received signals. Wave mode identification was achieved analysing time-of-arrival and dispersion curves.
The cross-correlation function was used and calculates between the measured (monitored) signal and the reference signal. This process proved to provide a distinction between patterns depending upon the material characteristics. However, the reduction of the cross-correlation can be caused also by other factors such as temperature or surface water.
Several techniques were analysed to contribute toward a multi-parametric analysis. These include two-dimensional (2D) fast-Fourier transforms (FFTs) to identify propagating modes, the optimisation technique enabling determination of the attenuation coefficient, and application of B-scan to determine parameters of the propagating guided waves.
In summary, The WINTUR system analysis tools provide an early indication of the onset of blade defects. In addition to signal visualisation, a visual display of the blade status is provided with colour scale system indicating to the user the health of the blade. Processing data from the WINTUR sensors provides good sensitivity providing successful detection:
- of growing defects up to a distance 3.8 m,
- of Blade intrusive-damage down to a thickness of 2 mm,
- of impact excitation up to a distance of 800 mm.
The WINTUR analysis tool is also able to accurately locate the position of defects
WP6: Software development
A GUI was developed (using Microsoft .NET 3.5 Framework), capable of visualising acquired data from sensors on wind generators. The GUI permits the user to connect to the WINTUR database and retrieve data on demand and perform common processing tasks. Once past security protocol and connection is achieved, lists of inspected components are retrieved from the database. The list includes identification of the wind turbines, turbine blades, sensor arrays with the option of the provision of compiled reports using drop-down menus. Graphical visualisation of the acquired AE / LRU signals is also provided. Navigation tools have been developed to permit the user to navigate the signal plot and export the measurements in a common ASCII file format.
The designed database - developed using the MySQL database server and workbench application provided by the MySQL - stores information concerning user authentication, wind generators, blades, sensor arrays, received LRU measurements, defects, reports, alerts and data meta-tagging.
Description of the database tables and stored procedures.
In summary, the WINTUR systems GUI provides easy manipulability of received data providing a visual representation of the inspected blade and its condition. Easy-to-use navigation tools with pull-down menus and a request facility provides the user with access to relevant data. Signal visualisation of the received system data and blade integrity representation is provided to indicate the health of the blade. Time-stamped status reports provide a summary health analysis detailing defect size, defect location, temperature, blade and turbine identification.
WP7: System integration and testing
System integration trials were tried out on the purchased 9.8 m blade purchased by Miyama and residing in Chalkida, Greece. The progress on a system for defect detection in wind-turbine blades was tested with reference to the design, study and implementation to perform non-destructive testing (NDT) testing on composite materials within wind turbine rotor blades. These tests involved
- Successful integration of the dual sensor capability involving AE and LRU. MFCs were successfully connected to the AE system to provide a dual application sensor.
- Testing of an EH system using APAs for integration with a miniaturised, flexible electronics circuit sensitive enough to harness energy from the displacement activity on a sub-woofer over a wide range of frequencies.
- Testing of the low-power wireless communications system using the Nordic Semiconductors' NRF24L01+ (802.11 protocol) for data transfer.
- Navigation using the GUI permitting data visibility and manipulation. Availability of information was only provided upon user authentication. A date-stamped summary report provided a status update on the health of the blade under inspection.
Additional required work to validate included additional testing of the wireless link after a little data corruption at the receiving node over a distance of 40 m. Further validation of the spatial array was also required with further work on EH provision to be completed.
In summary, the integrated WINTUR system provides a reliable early indication of the onset of blade defects. Visual display of the received data and the blade integrity indicates to the user the latest status:
(a) on an area up to 3.8 m on growing defect detection;
(b) on invasive damage down to 2 mm thick;
(c) on an area up to 800 mm on impact damage;
(d) location of defect.
The WINTUR wireless communications system is able to successfully transmit the acquired data at a speed of 2 Mbits per second in 20 seconds.
The WINTUR EH system (APAs) was able to charge the harvester to 5.6 V for a 150 Hz vibrations of 150 mV amplitude with a storage time of 13 s. Lowering the frequency of the 150 mV vibrations to 105 Hz increases the storage time to 70 s.
WP8: Demonstration and large scale laboratory trials
The demo was very successful with satisfaction from each of the SMEs at the level of development achieved. Each of the technology aims were accomplished and it was concluded by the consortium that a viable system for blade monitoring to aid the wind sector attain the desired energy delivery to business and communities whilst reducing operational and maintenance costs could be achieved.
The following system objectives were successfully achieved:
(1) Novel MFC transducers installed on a turbine blade, able to detect the onset of damage that is the occurrence of fibre breakage due to staged development of a hole-defect. Changes in amplitude and changes in boundary condition indicated the presence of defect within the sensor vicinity - a distance of up to 3.8 m for the guided wave sensors facility and 1.5 m for the AE sensor facility. Matrix capture of the spatial array of sensors using signal processing techniques (cross-correlation) successfully achieved defect detection and accurate location on the wind turbine blade. This was the culmination of work from WPs 1, 2, 3 and 5.
(2) EH (developed for the purposes of powering the pulser/receiver unit) successfully harnessed energy using APAs from a simulated ambient mechanical environment (220W sub-woofer) over a wide range of frequencies (35 and 300 Hz). The system did not yield high energy at low vibrational frequencies (8 - 15 Hz). This was expected since this frequency is an order of magnitude lower than the resonance of the APA actuator.
An alternative EH system with mounted MFCs (M8528-P2 MFC) on a scaled blade model (1:14) was explored to assess the yield from a dynamic environment of transient loads and semi- permanent loads.The MFC system was able to yield 0.5mWs per deformation cycle at 0.5 Hz.
(3) The wireless communications system transmitted blade-acquired Guided wave data from a transmitting node to a receiving node located 40 m away. The two sets of data correlated perfectly. The received data was then 'called up' by the data central software programme and visually displayed as time series data. The GUI displayed the blade details and the sensor details and created a time-stamped receipt with user information.
The WINTUR consortium witnessed a successful demonstration of an NDT monitoring system that is fully integrated and applicable to detecting the types of defects typically found in operational turbine blades.
The trials were completed by providing consortium partners with a demonstration on how to use each stage of the inspection system. This was completed through a mixture of a live demo and presentations.
In summary, the complete WINTUR system provides an early defect detection system comprising:
(1) A visual blade display indicating latest blade status over an area of 3.8 m for growing defects; 800mm for impact damage, and sensitivity of fibre-breakage down to 2 mm thick. Location of the defect is also provided.
(2) Wireless communications for data transfer up to speeds of 2 Mbits per second in 20 s.
(3) Two EH systems:
(a) APAs able to charge the harvester to 5.6 V for a 150 Hz vibrations of 150 mV amplitude with a storage time of 13 s;
(b) MFCs able to generate 0.5 mWs per deformation cycle at 0.5 Hz - able to power the transmission of 8 datagrams to a receiver node over a distance of 150 m.
(4) A GUI providing signal visualisation, a visual status indicator, easy-to-use navigation tools, tabulated display of critical information and a reporting facility.
WP9: Exploitation and dissemination
The work performed during the first period informed the dissemination and exploitation activities in the second half.
Marketing material was created for the purposes of dissemination at conferences and exhibitions.
Technical papers were presented at the following conferences:
- BINDT, 2011, United Kingdom (UK)
- Engineering Structural Integrity Assessment 2011, UK
- BINDT, 2010, UK.
The following conferences were attended by WINTUR partners where dissemination of the project took place:
- ISPA - International Symposium on Piezocomposite Applictaions, 2011, Germany
- IDTechEx- EH & Storage Europe, 2011, Germany
- Wind Turbine 1st technology Forum, 2011, Italy
- EWEA 2011, Belgium
- Sensors Expo & Conference, 2010, United States (US)
- SPIE Smart Structures / NDE, 2010, US
- Smart 2009, Portugal
Further details on these events are provided in Section 4.2 (A1) and in Deliverable D9.2 which includes information on conference papers, flyers and feedback via the project website.
The main dissemination will continue with the WINTUR project website beyond the project term. The established partner benefits include:
(a) ownership and licensing status of each exploitable product with particular detail;
(b) foreground IP from the successful exploitation of the Background IP;
(c) benefits of exploitable IP and successful implementation of technical integration for the reliable detection and prediction of defects within composite blade components;
(d) an approach for exploitation of the integrated WINTUR technology beyond the project term and an established path permitting SMEs to serve new markets;
(e) established routes for further dissemination of the project achievements through workshops, publications, conferences and the WINTUR website.
The main results of the project and therefore the key selling points of the WINTUR technology were identified by the consortium as below:
- Reliable blade inspection - reducing operational and maintenance costs by minimising downtime due to advanced defects causing critical component failure. Harnessing of environmental energy to ensure low sensor power consumption and avoid routine battery replacement.
- Dual-purpose sensors - proven integration of AE and LRU technologies able to detect and locate flaws in turbine and extending component life-expectancy by detecting flaws at an incipient stage of development. Reduced onsite maintenance: reducing the frequency of labour-intensive inspection providing efficient maintenance service for operators. Alternatively, regular tests can be run safely from a central base.
Data transfer without human intervention - remote inspection (via wireless transmission) to provide processed blade integrity data for high level decision making. Reduced insurance premiums - early detection limits damage to component parts thereby extending their life-expectancy and reducing insurance premium costs. The overall exploitation and dissemination planning in WINTUR has been geared towards maximising the economic benefit to the SMEs as quickly as possible following the end of the project. Marketing flyers in different languages to promote WINTUR technology have also been created for wider audience dissemination.
Dissemination was further provided via the project website where regular project updates were reported. The consortium member sites contained a link to the project website to attract more hits. A workshop held in Greece was also used to establish wider recognition for the work conducted within the wider renewable energy industry.
Potential impact:
The WINTUR project has been very successful. The field trials carried out on the purchased wind turbine blade confirmed that the SHM development is an effective and reliable inspection device that has the potential to be of considerable interest for the wind energy industry. It is a unique system that can perform quick and simple inspection of composite components.
The consortium had identified the need for such a system for utilisation in what is an expanding wind energy market. The development of the project has been widely available on the WINTUR website (see http://www.wintur-project.com/ online) launched by TWI in the opening weeks of the project in order to facilitate project dissemination. A Google search of the word 'WINTUR' provides details of the project as the first item. The website was maintained by the project coordinator with the facility for updates by consortium partners. It remains the principal internet platform where information, contacts, news and developments about the WINTUR project system were accessible to the public for the duration for the project and will remain so beyond project completion. In addition, all partners were asked to add a link of the WINTUR project website onto their respective websites. Whilst conducted by all partners, Optel together with Miyama have provided the momentum behind dissemination activities.
The investment cost for such a system is considered to be high in the current status. This is mainly due to the electronic instrumentation that can be easily replaced by a more simple and affordable instrument. At the final stage of the WINTUR project, the price of the prototype was discussed but not defined. It was agreed that a drive to lower the costs should be taken seriously without the threat to system capability and integrity.
In terms of benefits, this project will benefit the SMEs and reduce long-term costs for utility companies. There is currently no standard condition monitoring technique available that can perform routine autonomous inspection providing details of the blade integrity. The technique presently used is visual inspection - a relatively slow, expensive technique of providing the user with the blade information sought. At present, the industry deploys hands-on-action as deemed necessary - usually long before a critical failure occurs. The aim is to confine repairs to straight-forward up-tower repairs to prevent costly down-tower serious repairs that may require extended shutdowns.
The benefits of fitting an SHM system such as WINTUR will become apparent to the industry when a way to calculate the point at which a condition monitoring system becomes cost-effective from the perspective of the owner-operator. There are a variety of reasons for the delay in standard implementation of a condition monitoring system. These include the complicated nature of wind turbines, the remoteness of their installations and the random nature of the industry's early years. Indeed, the argument regarding the remoteness of installations cuts both ways and can be converted into a case for a cost-saving monitoring system. The viewpoint that offshore is ideally suited to SHM is somewhat comprehensive throughout the industry. This is where a monitoring system like that developed by the WINTUR consortium has a part to play. An additional strength of the WINTUR project is the high-level statistical matching (via cross-correlation) and 'duality' (AE combined with LRU) to avoid misrepresentation and reduce the risk of overabundant supply of information.
Contact has been established with Vestas one of the world's largest turbine manufacturers who expressed considerable interest in the techniques utilised in WINTUR project. This communication is ongoing and will establish a route for feedback regarding OEM's concerns and requirements.
There is further evidence that a project such as WINTUR will find a market-place in the medium-term: statistics suggest that large numbers of wind turbines will come out of warranty within the next couple of years. Costs will then automatically transfer to the owners (i.e. utility companies). When the costs of repairs become apparent to owners, mass implementation of condition monitoring will take place. Thus, the SMEs within the WINTUR consortium are well-placed to exploit this cycle of maturity set to occur within the wind industry.
A workshop held simultaneously with the project demonstration attracted the interest of Creative Systems Engineering (CSE) - energy consultant in Greece. This company was impressed with the technology developed and believes that the WINTUR project does have a part to play when the roll-out of condition monitoring systems gathers momentum.
The WINTUR consortium recognised and agreed the importance of understanding the strength, weakness of the developed technologies and to identify the opportunities and threats for the end results as a realisation process. The strengths, weaknesses, opportunities, and threat (SWOT) analysis carried out in the first reporting period was the first step of strategic planning. It helped all partners to be aware of the different aspects of the market and guide the project to focus on key issues.
Threats for the system are not overly critical and depend upon the way the information resulting from the WINTUR project will be disseminated and exploited.
The weaknesses presented can be solved with additional resources and time to enhance the system. The price of the WINTUR system has to be addressed to make it a viable option in the industry sector.
The SWOT analysis shows that the system opens the door to many other opportunities. With further improvements, additional features can be brought to the WINTUR system and this will contribute to increased viability.
After lengthy discussion with the WINTUR partners, it was concluded that there are two ways in which the WINTUR project can be brought to the market:
(1) As a factory-installed system controlled by the OEM: However, one 'threat' of this approach is that it permits the OEM to control the data and hide flawed equipment. This can be averted by buyers through a process of negotiation at the contractual stage.
(2) Retro-fitting the system onto ageing turbines: Whilst this may not be possible whilst the turbine is under warranty, when the owner takes full responsibility for turbine operation, there is considerable scope for mass implementation.
A number of dissemination activities were performed during the course of the project to expose the project results and to make the developments visible to the industry. Conferences and seminars related to wind energy technology were attended by the project partners and presentations were given. However, since information explaining how to use the WINTUR technology is commercially sensitive (because the WINTUR consortium is planning to exploit the results beyond the project), it was decided that this information will not be published in any format (report, guideline or wall chart) other than in the deliverable reports only visible to project partners.
Flyers were the main marketing tools for the project and were handed out at conferences by partners. Dissemination activities after completion of the project were considered critical for the successful implementation and commercialisation of the product.
Contact details:
Kenneth Burnham
kenneth.burnham@twi.co.uk
Project Coordinator
TWI
Granta Park,
Cambridge, CB21 6Al
Tel: +44-012-23899334
List of websites: http://www.wintur-project.com
