Final Report Summary - CMSWIND (Advanced condition monitoring system for the assessment of wind turbines rotating parts)
Executive Summary:
Wind energy is presently the fastest growing renewable energy source in the world. However, the industry still experiences premature turbine component failures, which lead to increased operation and maintenance (O&M) costs and subsequently, increased cost of energy (COE). To make wind power more competitive, it is necessary to reduce turbine downtime and increase reliability. Condition Monitoring helps by reducing the chances of catastrophic failures, enabling cost-effective operation and maintenance practices, and providing inputs to improve turbine operation, control strategy, and component design.
Current wind turbine condition monitoring methodologies can be time-consuming and a costly process and fail to achieve the reliability and operational efficiency required by the industry. For these reasons existing vibration-based Condition Monitoring System (CMS) usually fail to detect defects until they become critical. This project will show the applicability of the CMS enabling the prompt detection of defects. The proposal idea is to use this experience to deploy an existing system that:
• Allows an early detection and identification of any developing defects on several components of the wind turbine’s drive train, thus helping to optimise the maintenance schedule.
• Combine the use of several sensors in order to evaluate the overall operational condition of the turbine’s generator, gearbox, bearings and main shaft.
• Use wireless sensor for rotating components monitoring using high performance powering and energy harvesting technologies.
• Fuse and analyse the data obtained through the different sensors using a single SCADA system.
This project shows the applicability of Condition Monitoring enabling the prompt detection of defects by deploying a system that allows an early detection and identification of any developing defects using Operational Modal Analysis, Motor Current Signature Analysis and Acoustic Emission Techniques applied on the gearbox, generator, and bearings and high speed shaft respectively.
Broadband radio transmission systems were developed to transmit the data from fixed or rotating frames to the ground without losing any signal resolution. The results indicate the feasibility of collecting AE signals from the rotating frame with acceptable level of noise in low to moderate wind speeds. Also other developments were executed in order to verify whether or not the noise level increases appreciably with wind speed and whether such signals can be filtered out.
New procedures, guidelines and training material has also been developed during the CMSWind Project for applying offshore instrumentation and developments and information of use to the whole Wind Energy and NDT Industries.
Project Context and Objectives:
As follows, a description of the project and objectives is presented. The work carried out during the 36 month duration of the project is divided in 9 work packages. The different work package objectives and the achievements of each one is presented.
WP1: System specifications
The objectives of WP1 were: i) To decide on the supervisory hardware that will be used in the system and identify the exact requirements of the end users so that they will be able to take advantage of the CMSWind condition monitoring system fully; ii) To select the most suitable wind turbine for the test installation. For that, a specifications document was produced and submitted to the REA as D1.1. The document comprises the requirements of the condition monitoring system and also describes the critical areas that have been selected for investigation. The document includes some data extracted from the database of vibration and AE data provided by the end-user SKM. SKM has also provided the selected Test Wind Turbine for the consortium, which is the WindMaster300. The coordinator, TWI, led this work package with the support of all the consortium members, especially the rest of the RTDs, the end-user SKM and AEE (Exploitation Manager).
WP2: Motor current signature analysis technique
In this work package all aspects of MCSA methods were thoroughly explored and the procedures that will form the basis of the MCSA technique to be used in the CMSWind System were established. Two deliverables were presented as the tangible work developed within this work package: D2.1 and D2.2 in order to meet the WP2 objectives: i) To determine which of the MCSA methods will be used, apply optimizations and thus form the basis for the MCSA technique for wind turbine generators; ii) To acquire the needed equipment as well as construct and program the interface module with the supervisory system. In D2.1 the specifications of current MCSA methods as well as the obstacles that need to be overcome were explored; an overview of the different kinds of generators for existing wind turbines was also presented, explaining the main differences between them in terms of components and therefore operating principle; the most common faults were also described, together with the existing techniques on current inspection; finally a comprehensive description on the MCSA technique (acquisition, signal processing and fault detection) and the proposed hardware that could be used for CMSWind were presented. After intensive research, the MCSA Technique widely used for off-line inspections has been confirmed as appropriate to be upgraded to an on-line technique that can monitor the generator in a wind turbine and meet the expectations of the CMSWind Project. In D2.2 the acquisition of the needed equipment and construction and programming of the interface module with the supervisory system that will form the MCSA subsystem was described. For that, two lines of work were developed in parallel: on one hand INNORA developed the acquisition cards and its embedded software in order to acquire and save in a folder the current waveform from each of the three phases of the wind turbine generator; on the other hand Brunel University developed the system’s software which will be used for the assessment and remote logging of faults. Such software consists of signal processing (three different signal processing techniques have been applied: FFT, Wavelets and Current Park’s Vector), feature extraction, feature trending and baseline development. When a fault is present the features will fall out of the limits of the baseline and therefore an alarm will be triggered.
WP3: Operational modal analysis technique
Objectives for WP3 are: i) To determine which of the OMA methods plan to be used and form the basis for the OMA technique for wind turbine gearboxes; ii) To acquire the needed equipment as well as construct and program the interface module with the supervisory system. During this work package the OMA methods that form the basis for the OMA technique for wind turbine gearboxes were determined. Also the needed equipment as well as the programme interface module with the supervisory system was constructed. Two deliverables were submitted describing in detail the mentioned work: D3.1 and D3.2. In D3.1 a state-of-the-art in terms of methods for OMA was presented. A comparison of OMA against EMA (Experimental Modal Analysis) was discussed. The approach that will be applied to the wind turbine drive train monitoring was explained, in parallel with the modal parameters to be considered when designing such monitoring system. Finally some preliminary results on laboratory trials for the OMA technique development were drawn where the following information was gathered: position for sensors number and type of sensors, data acquisition requirements, OMA algorithm testing, and evaluation and adaption of the algorithm. The selection of sensors to be used for the OMA module has been proposed. In D3.2 the OMA subsystem is presented were the acquisition cards and its embedded software was developed by INNORA and the software that forms the OMA technique in terms of signal processing, training process (to develop the baseline) and operation process (to carry out the monitoring process) were developed by Fraunhofer. The OMA subsystem was deployed on the test wind turbine in month 12 with the objective of extracting information about: the modes which will be excited when the wind turbine is in operation, optimization of number and position for the sensors, clustering of environmental data for classification, evaluation of signal processing and algorithms for modal parameters, and investigation of the relation between modal parameters and health condition of the wind turbine generator. During month 13 the acquisition hardware developed by INNORA and adapted to the OMA requirements and the acceleration sensors mounted on a plate were validated to form the OMA subsystem.
WP4: Acoustic Emission technique
The objectives of WP4 are: i) To develop an Acoustic Emission technique that can detect defects upon their creation while placed on rotating machine shafts; ii) To construct an AE system that will use an energy harvesting module to give power to the transducers and the data acquisition system and a wireless communication module capable of transmitting the data to a short range receiver. During this WP a suitable AE Module (including acquisition hardware development and suitable algorithms implementation for the detection of faults in the high speed shaft of the wind turbine’s drive train) as well as the system that will use an energy harvesting module to give power to the transducers and the data acquisition system and a wireless communication module capable of transmitting the data to a short range receiver, were developed. The different developments were reported in three deliverables, being these: D4.1 Energy harvesting module, D4.2 Wireless module performance and D4.3 AE final subsystem. In D4.1 the Energy Harvesting system conception is presented, from design to its assembly and deployment. The objective was to supply the Acoustic Emission system during operation when mounted in the rotating equipment. The system includes a battery to last during wind turbine down times, whereas its operation and duration under continuous system operation were investigated. A battery charger was also designed and constructed to charge the system’s battery. The Energy Harvesting System was successfully deployed on the test wind turbine, and the results presented in D4.3. In D4.2 the wireless communication module developed by INNORA with the objective of transmitting the acquired data wirelessly to the module subsystem was presented. Such acquisition hardware was successful tested at both laboratory scale and on the test wind turbine. Finally in D4.3 the final AE subsystem was presented, including acquisition hardware, Energy Harvesting System and AE Technique Software (baseline creation and monitoring process). The initial results obtained from its installation on the test wind turbine have been reported.
WP5: System installation on test wind turbine
Work package 5 consists of installing the three different modules (MCSA, OMA and AE with the supervisory and data logging systems) on the chosen test wind turbine: the WindMaster300 described in Deliverable D1.1. This task is dedicated to ensuring that all installed systems are working as planned. During the second period of the project, such validation on the test wind turbine was achieved successfully. Three tasks form WP5: Task 5.1 SCADA PLC installation with initial sensors, Task 5.2 Installation of MCSA, OMA and AE systems and connection to SCADA, and Task 5.3 Installed systems validation. During these tasks, the RTD partners (TWI, Brunel, Fraunhofer and Innora), with the guide and support from the SME and Associations, installed and validated the hardware and software specifically developed for this application. All the information about the hardware installation procedure and the three subsystems installation and configuration (for the three techniques) with the validation of each individual technique was presented in D5.1.
WP6: Test wind turbine analysis
This work package represents the 2nd phase of the RTD work performed during the project. At this stage, condition monitoring technique development needed raw data to be gathered (Task 6.1) in order to carry out the Final Techniques Assessment (Task 6.2). As soon as the first parts of the supervisory system hardware were installed and communication was established, the measurements begun. During the techniques assessment a correlation between the data gathered and the events occurring in the wind turbine was determined. In this way FHF, Brunel University and TWI were able to see which of the methods used in the techniques provided accurate warnings of the events prior to their occurrence and how the trend of changing measurements can show the current condition of the wind turbine machinery. Installation took place on a Vestas V90 (access granted through TWEA and Borusan). Several choices were considered and negotiation with other wind farm operators and owners (EDP, Iberdrola, Gamesa, etc.) took place through the other Associations in parallel (AEE and AEND), although the V90 in Turkey was the final machine chosen for installation. The data gathered on the V90 was used to understand the final way of manipulating data from the different methods, thus leading to the formation of the final technique guidelines, reported as D6.1. Apart from the general guidelines of usage that will enable the trained personnel to use the new methods to estimate the machinery conditions there will be several occurrences that the new techniques subsystems will record measurements so large that immediate action should be taken. This was part of Task 6.3 Automatic Alerts definition, reported in D6.2 after data analysis from the V90 wind turbine.
WP7: Development of guidelines for application of developed technology
The aim of this WP is to transfer the required expertise to the maintenance teams of the turbine operators and to prepare the training material for the SMEs installing the system. Two tasks form this WP: Task 7.1 Development of training material (reported as D7.1) and Task 7.2 Training program for SMEs. For the Training Material development, clear instruction and service manuals were created. This task met the expectations by producing a 4 instruction and 4 service manuals for the 4 distinct subsystems (MCSA, OMA, AE, SCADA) and 2 half-hour presentations, one on the installation and one on the use of the CMSWind system. This material was updated with every improvement of the system. Regarding Task 7.2 a training session was conducted at TWI facilities, lead and organized by TWI and Brunel University, where the expertise and knowhow described in the developed training material was passed onto the Associations and SMEs for them to transfer this knowledge onto potential operators. All this was supported with photos, videos and a live demonstration of the CMSWind system.
WP8: Result of dissemination and exploitation activities
Work Package 8 objective is to disseminate the technology developed during the previous work packages as widely as possible. For that, the plan was to promote the developed technique and maintain a technology transfer programme that ensures successful implementation and provided a marketing platform for the consortium, as well as to ensure a widespread use and awareness of the developed defect detection system. The tangible deliverables for this work package are a Website (developed by month 5), 2 promotional videos (developed by month 24, and updated at months 30 and then again 36), a Draft Plan for Use of Dissemination of Foreground (submitted by month 15), and the final Plan for Use of Dissemination of Foreground (submitted in month 36). Beside these, further activities have been / will be carried out to contribute to the dissemination and exploitation of the developed work through conferences, events, paper publications, etc.
WP9: Management and coordination
Work Package 9 refers to Project Management and Coordination, which has been an on-going task throughout the project duration. This work package also includes Project Management of Technical and Administrative activities and all Project Reports (Progress Reports, Technical Reports, WP Reports, and Final Project Report). As part of this Work Package the Project Coordinator planned, organized and monitored the project for administrative, legal and contractual matters, quality and standards representation and implementation. Allocated budgets & cost statements were collated and reviewed prior to submission to the EC. The Project Manager planned, organized and reviewed at 6 month intervals the technology, product and knowledge management work packages with the Work Package Leaders. Also the Project Manager dealt with the day to day inter-participant or consortium level issues.
Project Results:
WP1: System Specifications
Partners involved: TWI, AEND, Co-Services, INESCO, SKM, DGZfP, UBRUN, FHF, Innora, AEE, TWEA.
System Specifications and selected test wind turbine
1. Overview
This work package decided on the supervisory hardware used in the system, as well as take into account the end users requirements in order to select the most appropriate test installation system (wind turbine).
2. System Requirements
The system has the following capabilities:
• Operate Acoustic Emission (AE), Motor Current Signature Analysis (MCSA) and Vibration sensors (OMA).
• Operate parametric sensors such as wind speed, proximity probes (RPM) and power measurements.
• Collect the data following a specific protocol.
• Save the raw data and process it.
• Generate a signature (Baseline) that will be used as characteristic of the Wind Turbine machinery.
• Continuously monitor the Wind Turbine through the execution of the same measurements.
• Generate alarms on criteria already implemented.
The system works following the next steps:
• The system records the wind speed and power output at regular intervals of 30 minute and stores these values.
• Parametric measures such as wind speed/power/RPM will be used as a trigger to start collecting data from AE sensors and accelerometers.
• For a specific data set, the current wind speed/power is compared against a table representing the bins and the number of data per bins collected (ex. 5m/s = 0, 5.1m/s = 1 etc…). The bins are divided in increments. For example of 0.1m/s from a minimum of 3m/s to a maximum of 30 m/s for wind speed measurements.
• The mean wind speed (and mean Power output and RPM) will be the average wind speed collected in the first 5 seconds (Figure 1).
The following description is based on the wind speed acquisition:
• If this speed matches the bin requirement, the program proceeds in collecting the rest of the data. If the bin requirement is not satisfied, the values thus far collected are cleared.
• If the mean wind speed is between 3m/s and 30m/s the raw data is collected and stored in binary in order to be accessed and analysed on a later date.
• The operating parameters are tested every minute - until a required wind speed is attained at which point it samples the AE sensors and accelerometers and stores its values. The MCSA deal with as a parametric input such as wind speed, power and RPM.
• The MCSA, RPM, wind speed and power output measurement is occuring during the first 5 seconds and not overlapping with the A/D data acquisition (vibration and AE), which is always limited to 1 second of AE data and 5 seconds of vibration date (Figure 2).
• Acoustic emission data is sent through a band-pass filter (10kHz - 250kHz).
• Vibration data is sent through a low-pass filter (10kHz)
• The files are binned according to the power output of the turbine (5kW bins) and the different analysis values are averaged within the bin (Figure 2).
3. Operation process
When the system is installed in a test rig or in a real wind turbine it operates at three levels: pre-acquisition, acquisition and warning. At the pre-acquisition stage, the user has to decide the number of sensors which are necessary to monitor the machine under study and attach them.
Once this has been done, the acquisition is started. The user has to run the software setting up previously some parameters (sampling rate, number of samples, number of files per bin, …) . The system starts acquiring data. A baseline is generated as soon as all the bins have been filled. This baseline is characteristic of each machine and lets establish whether there is a fault or even if it is been initiated.
In the last stage, the software follows an algorithm which is different for each of the three techniques to assess the health of the machine studied.
4. Test Wind Turbine data
It was necessary to integrate the drivetrain provided by SKM in an operation process that allowed the investigation of fault signatures in order to characterize a failure deviation from baseline. For this reason TWI and BIC investigated potential mechanical connection of the hub to a variable speed/rotor torque.
The specifications given for the wind turbine test rig were provided by SKM:
• 300KW rated power output
• 50RPM nominal speed for the low speed shaft
All equipment is aquired by TWI in order to build a laboratory test facility. The objective was to (i) characterize the fault detection by seeding defects in the machinery and (ii) build facility for training and certification.
WP2: Motor Current Signature Analysis Technique
Partners involved: TWI, AEND, Co-Services, INESCO, SKM, DGZfP, UBRUN, FHF, Innora, AEE, TWEA.
Motor Current Signature Analysis
• Overview
This workpackage serves as a comprehensive description of the methods that form the basis of the MCSA technique that has been used in the CMSWind System further to WP1 on System Specifications.
The specifications of current MCSA methods have been defined as well as the obstacles that need to be overcome. An overview of the different kinds of generators for existing wind turbines was presented, explaining the main differences between them in terms of components and therefore operating principle. The most common faults were also described, together with the existing techniques on current inspection. Finally a comprehensive description on the MCSA technique (acquisition, signal processing and fault detection) and the proposed hardware that could be used for CMSWind are presented.
• Summary of the work done
Intensive research effort has been focused on the technique of monitoring and diagnosis of electrical machines, to name some: time domain analysis of the electromagnetic torque and flux phasor, temperature measurement, infrared recognition, radio frequency (RF), emission monitoring, detection by space vector angular fluctuation (SVAF), noise and vibration, acoustic noise measurements, harmonic analysis of motor torque and speed, artificial intelligence and neural network based techniques and MCSA.
Of all the above mentioned techniques, MCSA has been pointed out as the most suitable one for induction motors monitoring, since it is non-intrusive, it is not affected by the type of load or other asymmetries and can target all relevant faults in the induction machine. MCSA is a condition monitoring technique that is now widely used to diagnose problems such as broken rotor bars, abnormal levels of air gap eccentricity, shorted turns in low voltage stator windings, and certain mechanical problems. It has therefore been confirmed as appropriate to monitor the generator in a wind turbine and meet the expectations of the CMSWind Project .
A lot of research work on MCSA for induction machines has been carried out with the aim of understanding the state-of-the-art in this field. Hereafter, the techniques to be employed specifically for the generator as part of the CMSWind Project are further investigated and developed in the following work packages.
MCSA Final Subsystem
• Overview
The objectives were acquiring the needed equipment and constructing and programming the interface module with the supervisory system that will form the MCSA subsystem, as per the Project DoW.
Brunel University developed the system’s software which is used for the assessment and remote logging of faults on the SCADA unit and their evaluation by expert engineers acting as the system’s supervisors. In the event of a critical fault signal being received by the SCADA unit from the wind farm server, the system alerts the supervisor who re-assesses the automated data analysis carried out by the intelligent unit on the wind turbine for that particular fault. The supervising engineer then confirms to the maintenance crew the remedial actions to be taken to resolve the problem and update the maintenance schedule accordingly.
The target specifications for the MCSA subsystem are summarized in this WP according to WP1 Specifications Document.
The hardware that use for the MCSA subsystem is presented, where the acquisition system consisting in Acquisition Card and Main Board, was developed by consortium member INNORA.
The different MCSA methods found in literature and described in detail in the previous heading that assess the generator current have been programmed in LabVIEW as sub.vi(s). The higher level monitoring software that creates the baseline and raise alarms in the event of a fault has been encoded with an architecture based on state machine programming, also with LabVIEW.
• Summary of the work done
Modern measurement techniques in combination with advanced computerized data acquisition and processing show new ways in the field of induction machine monitoring by the use of spectral analysis of operational process parameters. In the present WP, the process of data acquisition, signal processing and fault detection for the diagnosis of an induction machine was presented, along with the hardware required for signal acquisition and data processing.
The MCSA methods to be used which form the basis for the MCSA approach for wind turbine generator monitoring were determined and programed in LabVIEW, being these: Fast Fourier Transform, Wavelets Analysis (based on Multiresolution Techniques), and Current Park’s Vector. Fast Fourier Transform can identify faults related to Broken Rotor Bars, Shorted Turns, Air Gap Eccentricity, Gearboxes, and Bearings. Wavelet Analysis and Current Park’s Vector can support the FFT Technique by detecting some of the mentioned faults at no-load conditions, which is the main limitation for the FFT Technique.
The next stage, as per the DoW, was testing and validation of the MCSA subsystem. As
described in WP1 and previously in WP2 the MCSA subsystem has to be validated both at laboratory scale (in a test rig from SpectraQuest), and tested in the WindMaster300 provided by consortiummember SKM. The MCSA subsystem has to be optimized for an enhanced performance as part of an integrated solution for the complete CMSWind Condition Monitoring System.
WP3: Operational Modal Analysis Technique
Partners involve: FHF, Innora.
Operation Modal Analysis
The operational modal analysis (OMA) does not rely on specific excitation. The method rather assumes a white noise spectrum as input. The loads which are applied directly to the structural system cannot be measured and are not known. Thus it is not possible to separate the dynamic influence of the environment and the structure itself.
Both are recorded in the sensor network to record the structure’s response. However this is actually a great advantage of the method, since modal parameters are extracted for structures under actual operational conditions. Results of previous publications have shown that modal parameters for the static and operational conditions may differ greatly.
To identify a system description and derive the needed modal parameters, multiple approaches are available for OMA. The methods can be classified in the frequency domain and time domain variants.
• Frequency-Domain Decomposition (FDD)
• Enhanced Frequency-Domain Decomposition (EFDD)
• Curve-Fit Freq. Domain Decomposition (CFDD)
• Stochastic Subspace Identification (SSI)
After identifying spectral peaks using one of the above mentioned methods, one has to divide the spectral peaks into frequencies coming from noise and periodic distortion (both unwanted) or structural frequencies (wanted). For this kurtosis analysis will be applied. The efficiency of the method depends on the quality and amount of the recorded signals as well as the statistical properties of the distortions and the modal frequencies.
Operation Modal Analysis Final subsystem
This is a summary of the work performed within tasks T3.2 and T3.3 which has the objectives of acquiring the needed equipment and constructing and programming the interface module with the supervisory system that now form the OMA subsystem, as per the Project DoW.
The target specifications for the OMA subsystem were summarized according to WP1 Specifications.
The hardware that will be used for the OMA subsystem is presented, the acquisition system used in the Acquisition Card and Main Board was developed by consortium member INNORA. The acquisition software use for the technique (a suite of 5 programs designed to set up the acquisition systems and receive the data) was also developed by INNORA.
The INNORA acquisition system was successfully tested for application of acceleration signal acquisition for modal analysis. The next thing to do is the realisation of triggering by environmental events and the integration of software tools in a LabVIEW environment for averaging of signals and modal parameter extraction.
WP4: Acoustic Emission Technique
Partners involve: TWI, Innora.
Energy Harvesting Module
One of the objectives of this WP is to present the development of an Energy harvesting module with potential applications on rotating equipment such as the low speed or high speed shaft.
In the CMSWind project an Acoustic Emission (AE) system, mounted directly on the main rotating shaft of the wind turbine, is used to detect defects.
As the system (sensor and node) is attached to a rotating shaft, it is not possible to power it from the main by means of a cable connection. In view of this, the design of an Energy Harvesting System is necessary.
When the shaft is rotating, the AE system is powered and operates continuously. However, during a wind turbine down time the shaft is not rotating and a battery is required to power the system. Moreover, a battery charger is needed to ensure the storage of the energy harvested.
The former components allow securing enough energy from the shaft to power the AE system. A schematic of the entire system described here is shown in Figure 5. Figure 6 and Figure 7 show the Energy Harvesting system designed.
The design and construction of the Inertia Energy Harvesting System was completed to supply the Acoustic Emission system during operation when mounted in rotating equipment. The system includes a battery to last during wind turbine down times, whereas its operation and duration under continuous system operation were investigated. The different parts of the system were identified, designed and constructed. A battery charger was also designed and constructed to charge the system’s battery.
Wireless Module performance
Another objective of this WP was to develope the wireless transmitter/receiver module necessary to implement the Acoustic Emission technique on wind turbine machinery rotating parts. Specifically, the objective of this report is to present the selection process of the wireless protocols, the microcontroller to wireless module connection method, and the assessment methodology for the validation of the developed solution.
It should be noted here that, in CMSWind project, we developed a system that digitize the sensor data with a high accuracy A/D converter and transmit it wirelessly to a receiver placed nearby in the wind turbine nacelle. AE sensor, A/D converter and wireless module are powered by an Energy Harvesting module and the whole system will be integrated in a package as miniature as possible. This whole effort is done so that the sensor can be placed on the rotating shaft of the wind turbine. Figure 7 depicts the proposed connection architecture of the wireless sensors.
CMSWind project required Acoustic Emission sensors installed on rotating parts of Wind turbines to locate cracks on the materials. The rotating nature of these parts makes wired sensors unusable for the task. Moreover use of contact rings is expensive, difficult to install and would have to be tailor made to every wind turbine model. Wireless data transmission, energy harvesting and small analog to digital converters were proposed as a viable alternative solution. In this WP we have configured a wireless transmission scheme that is capable of transferring the volume of data required and has sufficiently low power consumption. The proposed solution fulfills all the system requirements as described in WP2 and was validated through two tests.
Acoustic Emission Integrated
The AE sub-system was designed, integrated and tested. To do these tasks TWI developed the hardware and software which is used for the assessment and remote logging of AE data.
The hardware that is used for the AE sub-system consists on an Acquisition Card and Main Board was developed by consortium member INNORA.
The different AE techniques that assesses the machinery were programmed in LabVIEWTM as sub.vi(s). The higher level monitoring software creates the baseline and raise alarms in the event of a fault. The software was encoded with an architecture based on state machine programming, also with LabVIEWTM.
An integration process was held between TWI and INNORA in order to incorporate the wireless data acquisition hardware.
Various tests were run in the laboratory, in order to initialise and set up the system. Once everything was working in the laboratory the system was installed in the test wind turbine. Data was gathered for the full speed operation range and then processed in order to look for the AE signals of interest.
Some mounting methods were tested in order to assess the sensor response to the AE events depending on the way it was attached. Finally, a magnetic coupling was chosen as it gave the best results and was the most reliable mounting process.
The Energy Harvesting System was mounted and tested both statically and dynamically. Due to the centrifugal force and the inertia of the system at high speed, the static part of the system was attached magnetically to the generator.
Using this attachment method, the rotating movement is fully controlled and the EHS is protected from inertial damage. The designed piece to support those efforts is cross shape and has four magnets to ensure a good grip for the bearing. The plastic part has to mesh with two gears located in the rotating part. These gears were joined to the motors and produce its movement. As working as generators, the electricity produced by the motors is used to supply the whole electronic system and sensors.
All the parts that constitute the acquisition system have a place in the rotating plate. The locations were developed using little plastic marks to attach their boxes. Apart from the sensors, which were located in two holes near to shaft, acquisition cards and battery charger circuits are placed in different compartments.
A correlation between AE parameters and the shaft rotation speed was carried out for the analysis of the data gathered.
WP5: System Installation on test wind turbine
Partners involve: TWI, AEND, CO-Services, INESCO, SKM, DGZfP, UBRUN, FHF, Innora, AEE, TWEA.
Validation of the system in the test wind turbine
Here it is summarized the work performed within tasks T5.1 T5.2 and T5.3 where the Acoustic Emission, Motor Current Signature Analysis (MCSA) and the Operational Modal Analysis (OMA) subsystems developed in the previous Work Packages, were installed on a WINDMASTER 300 wind turbine.
During these tasks RTD partners, TWI, Brunel, Fraunhofer and Innora installed, and validated the hardware and software that was developed for the project.
The three techniques used for the monitoring of the wind turbine rotating parts were validated in the WINDMASTER 300:
• MCSA Validation Results
Some of the results obtained from the gathered data of the WINDMATER300 nacelle are presented in Figure 8. Ten files were acquired per bin in order to generate the average and standard deviation of the crest factor feature to define the different baselines of each interval.
Different baselines that allow health monitoring for the generator of the WINDMASTER300 nacelle were developed. Using this baseline generation approach, the generator monitoring process can be performed by comparing each new set of data acquired to the original baseline created during the initial stage.
• AE Validation Results
Some of the results obtained from the gathered data of the WINDMATER300 Nacelle are presented in Figure 9 and Figure 10. Ten files have been acquired per bin in order to generate the average and standard deviation of the AERMS signal.
Two resonant AE sensors were installed on the low and high speed shafts of the test wind turbine for feature extraction and baseline development. The extracted feature was the AERMS signal, gathered at 2,5 M Samples/s during 0,8 seconds, where ± standard deviation of the AERMS define the operational limits for the different parts under study (Figure 9 and Figure 10).
The AE subsystem was installed in the test wind turbine. The data acquired was used to develop a healthy system signature including the AERMS parameter. As expected, there was a clear correlation between the rotational speed and the AERMS mean and standard deviation for each of the speed bins. This denoted that the baseline represents a pattern for AE monitoring that increases alarm definition accuracy.
Further research and data processing was carried out during WP6 where the final techniques developed are presented and the alarm levels definition studied.
• Vibration Validation Results
First the identification of the structural modal frequencies was carried out by FDD on the non–rotating drive train, based on impact excitation at the low speed shaft. The estimated first singular values were evaluated in order to build the reference for all further rotation experiments. Then for each ‘Bin’ the parameter ‘First singular value’ were determined as a reference signature representing this specific ‘Bin’. If the state changes, this can be detected for example by the correlation coefficient between data extracted during operation and reference data. For further parameter extraction frequency ranges of interests were specified as shown in Figure 11. Within this frequency ranges signal parameter were calculated like peak frequency and power of the frequency band.
During rotation, the signals were strongly influenced by the frequencies of the shaft rotation, the gear engagements and the bearings. Figure 12 shows the spectrum of time signal at Bin 2: (frequency of the high speed shaft: 8.3 Hz). The structural modes were very weak and some are not excited, for example in the range below 20 Hz.
In Figure 12, the first singular values of all Bins of the y-axis were collected in a surface plot. The frequency range around 80 Hz and 110 Hz (blue marked region in12). The modes below 20 Hz were excited only in the BIN’S with the higher rpm (6-10). The mode around 80 Hz and 110 Hz were better excited in the Bins one to six.
Referring to the frequency ranges estimated in non-rotating experiments, the parameters for the monitoring database were generated.
During the experiments, a database containing the reference singular values for each Bin was established for monitoring application for the TWI WINDMASTER. Global parameters based on singular value decomposition of all vibration data as well as parameters for each sensor within the identified frequency range were extracted and can be used for further investigations.
Hardware and were tested successfully. Because the parameters were specific for the used drive train, the reference data set was adapted to the specific device. So not all algorithms, integrated in the LabViewTM Software, runed automatically. A learning process must be performed by the user to define the frequency ranges of interest for the further calculation of parameters and to save them as reference. Description of the developed LabViewTM Software is contained in WP7.
WP6: Test Wind Turbine Analysis
Partners involved: Test Wind Turbine Analysis
The objective of this work package was the installation of the advanced Condition Monitoring System which has been developed as part of the CMSWind FP7 Project on a real wind turbine. Different approaches have been considered in order to complete the work package successfully.
It is presented first of all, the CMSWind system integration in terms of hardware and the embedded software that governs the system.
The CMSWind consortium received great interest from the Wind Industry operators in different countries: EDF in UK, Iberdrola in Spain, Borusan in Turkey, CRES in Greece, etc., for the system to be installed on the most relevant wind turbines available in the market (Vestas, Gamesa, etc.).
The option which was implemented was installation on a Vestas V90 operated by Borusan. The results of the CMSWind System validation in a real wind turbine are presented in this report.
Final Techniques development
The final CMSWind system installed on an operating wind turbine. Results from ongoing data gathering process and data analysis for the proposed techniques Motor Current Signature Analysis (MCSA), Acoustic Emission (AE) and Operational Modal Analysis (OMA) have been presented in D6.1. The algorithms proposed in the deliverables D2.2 D3.2 and D4.3 for the three techniques before were optimised and final algorithms were presented.
The location selected to perform the trials was the Bandirma (Turkey) Wind Energy Power Plant where Borusan have wind turbines. The installation was carried out by BIC and TWI in collaboration with Borusan and Vestas. The wind turbine selected to validate the CMSWind System was a Vestas V90 – 3MW.
Parametric outputs such as power output and wind speed were provided by Borusan weekly as TWI/BIC were not allowed to acquire that information using the CMSWind System. As an example, Figure 13 shows the variations in the power output and the speed of the rotor during in May 2015. The power generated by the wind turbine changes between 0 and 3MW while the speed at the input of the gearbox changes between 0 and 16rpm. That means that the maximum speed the generator spins at is 1672rpm.
The system was installed in a Vestas V90-3MW. The wind turbine selected was number 14 as there was not any other Condition Monitoring System installed that could interfere with the performance of the system developed within the project. Figure 14 shows how the CMSWind System was lifted using a crane.
Four magnets (Figure 15a) were placed on the Condition Monitoring System case to achieve the correct attachment between the wall of the Nacelle and the integrated system. Figure 15b and Figure 15 c depict how the DAQ system was installed at the front left hand side of the nacelle (just on top of the gate).
The data gathered by the system is sent via Ethernet cable to a laptop where the data is stored. A USB 3G Dongle was used to have remote access to the CMSWind System. The software which gives access to the computer is ‘Go To My PC’. This software is a powerful tool for flexible work which allows the user to perform daily checks on whether or not the software is running and that the data is being acquired correctly. It also enables the user to transfer files between the laptop in Turkey and the one where the post processing is going to be carried out.
Sensors installation
For the OMA application at the Vestas wind turbine three accelerometers were attached at the gearbox as shown in Figure 17:
• OMA 1: First Planetary stage
• OMA 2: Second planetary stage
• OMA 3: Helical stage
The probe selected to monitor the condition of the 3-phase generator (MCSA) was the GMC-IProsys ACP 3005/24 Current Probe & Clamp. The probe is suitable for the 1000V cabling on the Vestas V90 generator. Further specifications are summarized in D6.1. Each phase is split in four cables to reduce the amount of current travelling through the wires. Therefore, the probe was installed in one of them (Figure 18).
Figure 19 depicts the location selected to place the Acoustic Emission sensors. The shaft which connects the gearbox with the generator was covered so there was not easy access to the bearings on both sides of the shaft. The transducers were attached as close as possible to the bearings. Acoustic Emission sensor 1 was attached using grease and a magnetic holder to the gearbox case close to the output bearing (Figure 19a). Acoustic Emission 2 was attached following the same procedure close to the main bearing located at the generator input (Figure 19b). Signals coming from each sensor are gathered by two different cards. Signals gathered by AE1 are picked up by the card 00x20 whereas the ones coming from AE2 are picked up by the card 00x13.
The final version of the CMSWind Software for post-processing is presented as follows, with enhanced functionalities and improved characteristics, as compared to its inception described in the first stages of the work (D2.2 D3.2 and D4.3).
CMSWind Final Modules
The software was developed in such a way to facilitate the adoption of the technology by potential users. As a result, the software code was developed in LabVIEW, a graphical programming software development environment used by engineers and scientists to develop sophisticated test, measurement and control applications. LabVIEW and its graphical dataflow programming language is characterised by ease of use and provides a better way of solving problems than traditional lower level alternatives. The practical benefit of the graphical approach is that it puts more focus on data and the operations being performed on that data and abstracts much of the administrative complexity of computer programming such as memory allocation and language syntax. As a result, new programmers typically report shorter learning curves compared to other programming languages because they can relate LabVIEW code to flow charts and other familiar visual representations of processes.
The software developed within the project has the capability to generate a specific baseline during the first weeks of installation: Signature Stage. After the baseline was created, the new data acquired by the system was compared to the baseline in order to find out the status of the machine from that moment onwards.
Data is stored in bins according to the speed of the high speed shaft or the wind turbine power output (as the user prefers). The system calculates the mean of the files stored in each bin and then it establishes the maximum and minimum limits which delimit the normal operation area of the machine being monitored. It is right after that when the monitoring stage begins.
As long as the new data acquired falls within the specified limits, the machine is said to be under normal operation conditions which means that the machinery inside the wind turbine is healthy. Otherwise the alarms will be triggered indicating that the system is out of normal operation conditions, showing the presence on potential defects in the part of the drive-train under monitoring.
Results
From the Acoustic Emission analysis, the following conclusions were extracted:
- The Acoustic Emission module of the CMSWind System was validated in the present study using real data gathered from an operating 3MW wind turbine in Turkey (Bandirma Wind Energy Power Plant).
- Given a specific working condition, the software output can be regarded as the baseline features of measured AE signals. Any significant deviation of practically measured AE and vibration signal features from the baseline can be explained to be due to some kind of abnormality inside turbine components and/or system. Consequently, this information can be effectively used to achieve the objective of condition monitoring and fault diagnosis.
- The RMS and Peak Value of the Acoustic Emission measurements from the turbine drive train were used for the establishment of the baseline model.
- The analysis of the real data collected from the wind turbine was processed for three different numbers of files per bin.
- The baseline generation process is sensitive to changes in the number of files used. The higher the number of files is the more accurate the limits are, reducing the number of false alarms.
From the Operational Modal Analysis could be concluded for the application:
- The vibrational data are reliable for Operational modal analysis if he exceed a minimum level
- The generator rpm I the most important parameter for binning. The optimal value is 30- 60 rpm.
- The harmonic excitation due rotation causes the most dominant frequency component for the wind turbine in good condition
- As monitoring parameters has to be used the frequency of peaks in the first singular and values ,their damping and the vibration signal power of all sensors at the identified peak frequency (modal frequency or harmonic excitation).
- The parameter are extracted within frequency ranges, which has to be defined by analysing data from reference measurements of the healthy turbine and from analysing data in the case of damages . The parameters for each bin and frequency range are compared with their reference level and limits.
Report on suggested automatic alarm limits
The basic principle of the proposed automatic alert system is a reference method. There were extracted baselines from the set of reference data. The statistical properties of these parameters as mean deviation and extreme values were regarded for the process of alarm definition. The parameters were supervised and warnings were created if the defined limits for the parameter are exceeded. The limits for the monitoring parameters were established from the data gathering process of the wind turbine during operation in good condition. The monitoring parameters from the three techniques AE, MCSA and OMA are automatically collected in bins, defined by wind turbine operational data and are monitored within this bins. If damages occur and can be identified, the data base for reference data could be actualised and will be stated more precisely.
• OMA Strategy
Base of the automatic alert system is the reference data set. From this data set the baselines and limits for the monitoring parameter: frequency, damping and vibration signal power were extracted. For the OMA algorithm the range of rpm for one bin is important and should be 60 min-1. The first reference data set was collected after installing the system on the healthy wind turbine. Following the proposal in 3.1. the limits for the parameters were set to
Limit (parameter) = Mean (Parameter) 2*standard deviation (parameter)
If too much false alarms are generated during the data acquisition period from the healthy wind turbine, the limit definition has to be adapted by including the data which produced false alarms.
During the monitoring process the actual data is first sorted into the referred bin directory after analysing the wind turbine generator rpm. The specified number of data is collected within the bins for averaging (20 measurements). If the bin is complete the OMA parameters are calculated and averaged as described in deliverable D6.1. After comparing the parameter with the baseline limits, a warning is created if the parameters are out of limits. If a specified number of warnings is exceeded, it must be proved if there a damage at the wind turbine is occurred. Nevertheless the algorithm is universal and can be applied to other types of wind turbine. The binning process for the reference data as well as for monitoring is described in deliverable 6.1.
• AE Strategy
The purpose of data analysis is to provide representative AE level across a range of wind speeds or power output for improving the reliability of wind turbines and thereby help define appropriate AE limits.
If the limits are too far from the target value, small deviations from the target value may go undetected but if the limits are too close to the target value there will be a large number of false alarms (that is there will be a signal for action when the process mean is on target and no action is necessary).
It has been convenient to set the warning limits so that, if the mean is on target, 95% of sample means will lie within them. Since we are dealing with sample means it is reasonable to assume they are normally distributed. Choosing 1.96 times the standard deviation only 5% of the values will be outside the limits if the features under study are in the same condition as when the signature was calculated. In general, the following warning limits will be set:
Where μ is the target value and σ is the standard deviation. The standard deviation will be an estimate of the true value and 95% is somewhat arbitrary figure. For these reasons and for simplicity the limits are often set at:
In this specific case, the target value is not known as there are not standards available for establishing Acoustic Emission limits in rotating machinery. When this happens, a large sample should be taken when the process is running satisfactory and the sample mean used as a target value.
In order to inform machine operators or engineers of important structural damages in the wind turbine machinery which is being monitored, an alert strategy has been accomplished. Figure 3 shows the flow which is followed by the software developed in LabVIEW.
Every time an Acoustic Emission (AE) is gathered by the data acquisition system, the software needs to allocate that piece of data in a specific bin as the speed varies constantly. The bin is selected according to the power or the wind speed provided by the SCADA system. Firstly, the software double checks whether or not the bin is full of data. If it is not full, that piece of data is stored within that bin. If the number of files is higher than the maximum number of files per bin allowed, it means that the software has enough data to build up the baseline. That measurement will be used for monitoring purposes.
This AE value can fall inside or outside the limits. It cannot be considered that there is a defect just because there is an outlier. It could just be a false alarm. For this reason, a criteria to establish the alarm limit has to be established. Due to the fact that a new set of data is collected every 30 minutes, the alarm will be activated whether the average of the last three files gathered for that speed falls outside the limits.
The selection of the number of files is completed taking into account that the time interval between measurements is 30 minutes and the power value changes very fast. Therefore, the minimum temporal distance between the last three files stored for a specific bin is 90 minutes. This distance will be usually higher as not many files are stored in the same bin consecutively. If the average of these three files falls outside the limits the software will switch on the alarms to let the operator know about the problem.
• MCSA Strategy
The limits created in the signature mode for the selected features (please see deliverables D2.1 and D2.2) are set as follows. Once a statistically significant sample for one bin has been acquired, all the parameters are averaged and the standard deviation is calculated. The limit selection will affect the sensibility of the system. The lower the alarm limit is selected the smaller size of defects can be detected. More false alarms can be generated though. The equation used to calculate the limits is set as follows:
Where σ is the standard deviation.
The limit selection has been investigated statistically. Considering the distribution of the features extracted from the signal as normal distribution (Gaussian function) only 68.27% of the values lie within one standard deviation of the mean. Therefore, even if the bearings are in good condition 32.72% of the values will be outside the limits. Choosing double the standard deviation, only 4.55% of the values will be outside the limits if the features under study are in the same condition as when the signature was calculated. Thus, the possibility of generating false alarms is reduced significantly.
In order to evaluate real trends and not only one value that can fall off the established limits, the following strategy is implemented.
Every time a new value is acquired, the software checks whether we are at the baseline stage or at the monitoring stage. If all the values required for the baseline stage had been previously gathered (i.e. the array for a particular bin is full) then we go to the monitoring stage, where each new value acquired is compared against the baseline.
At this point a new array is defined: the ‘OUTSIDE’ array, which is an empty array of a specific size (the number of values, ie the size of the array, can be modified by the user). Every time a new value falls outside the limits established by the baseline it becomes stored in the ‘OUTSIDE’ array. For the ‘OUTSIDE’ array to be populated, the values falling outside the limits must be consecutive, ie if the new value falls within the limits established by the baseline the ‘OUTSIDE’ array is emptied.
Once a number of successive atypical events is gathered, hence the ‘OUTSIDE’ array is completed, the array is analysed. If the trend of the ‘OUTSIDE’ array is crescent (n1
WP7: Development of guidelines for application of developed technology
Partners involved: TWI, AEND, Co-Services, INESCO, SKM, DGZfP, UBRUN, FHF, Innora, AEE, TWEA.
Training Material
Wind energy is presently the fastest growing renewable energy source in the world. However, the industry still experiences premature turbine component failures, which lead to increased operation and maintenance (O&M) costs and subsequently, increased cost of energy (COE). To make wind power more competitive, it is necessary to reduce turbine downtime and increase reliability. Condition monitoring helps by reducing the chances of catastrophic failures, enabling cost-effective operation and maintenance practices, and providing inputs to improve turbine operation, control strategy, and component design.
This WP provides training information of wind turbine drivetrain condition monitoring based on the use of CMSWind system.
Since the gearbox and generator have shown to have the longest downtime and are the most costly subsystems to maintain throughout a turbine’s 20 years of design life, they have been chosen as the main targeted subsystems to be monitored. Installation of CMSWind system and application of Wind turbine drivetrain condition monitoring practices will be addressed in detail along this report.
The custom hardware developed specifically for CMSWind project as well as the development and adoption of the three novel monitoring techniques are explained throughout this training material.
The training material is divided in the four following sections:
- CMSWind hardware architecture and configuration.
- MCSA module description and installation.
- OMA module description and installation.
- AE subsystem description and installation.
The current training manual presents the use and installation of the different CMSWind systems.
The installation process has been developed and explained taking into consideration the test wind turbine installed at TWI for testing purposes during CMSWind project. Due to this fact, additional considerations comprising both health and safety and operational conditions should be taken into account if the system is to be installed in a different wind turbine. However, and thanks to the architecture of the CMSWind system, the applicability of the different techniques and use of the hardware is scalable to different sizes and configurations of wind energy systems.
The training material was developed with a focus on the potential user which does not necessarily need to know technical details about the use of advanced condition monitoring techniques and systems in order to understand the information provided by the CMSWind system. This was fully ensured through the evaluation of professionals not related with the wind energy maintenance or diagnostic techniques sectors in order to ensure that it is fully understandable by the potential user.
WP8: Result of Dissemination and explotation activities
Partners involved: TWI, AEND, Co-Services, INESCO, SKM, DGZfP, UBRUN, FHF, Innora, AEE, TWEA.
Website
As part of the CMSWind Project, Brunel University hosts a website on behalf of the consortium with the domain name www.cmswind.eu.
The website consists of the following tabs:
• Home: Title and summary of the CMSWind Project. The flyers specifically developed for the CMSWind Project can also be found in the home page.
They are in three different languages, for dissemination purposes.
• Project Background: a deeper overview of the project is presented in this section, including more details on project objectives, the proposed solution and work packages.
• Consortium Members: a brief description of each of the CMSWind consortium partners is included in this tab. The logo of each company has also been incorporated.
• News & Events: A very important part of the CMSWind Project is the dissemination activities and events. In this tab, all the conferences and events where information on CMSWind Project is published and disseminated by the consortium members are presented.
• Contact us: contact details for Project Coordinator, Dr. Slim Soua, TWI Limited.
• Private Area: secure member area to upload deliverables, meeting minutes and other documents for the consortium partners.
A snapshot of the website is presented in Figure 21.
Promotional videos
As part of the CMSWind Project, TWI developed two promotional videos in collaboration with the CMSWind project consortium.
These two videos were developed in parallel by the RTD’s and TWI as coordinator formatted and merged the different clips into a video.
Additionally, the promotional videos were published in the CMSWind Website for public access and dissemination of project results.
The different tasks carried out in order to develop the videos are described as follows:
• Recording and development of clips explaining the system installation.
• Recording and development of clips explaining system configuration.
• Recording and development of clips explaining system operation.
• Provision of clips and pictures related to wind energy sector from project partners.
• Integration and formatting of videos into a common style.
• Development and synchronization of speech text to be embedded in the video.
These videos complement other dissemination materials produced by the CMSWind consortium as well as banners and posters and are being used for project dissemination.
Draft Plan for Use of Dissemination of Foreground
A Plan for Use and Dissemination of Foreground (PUDF) was carried out as set out in the Description of Work for the CMSWind project. The interim PUDF has been reviewed and amended during the course of the project as directed by the Project Steering Committee (PSC) and implemented by the Exploitation Manager, Mr Kilian Rosique from the Asociación Empresarial Eólica, who has been responsible for the planning and coordination of all activities related to adequate and timely dissemination of the project results within the consortium and externally within EU and global NDT and Wind energy community ensuring an appropriate level of disclosure for each action.
The PUDF includes a plan to attract investment for the CMSWind technology beyond the duration of the project.
Final Plan for Use of Dissemination of Foreground
The consortium has looked into the full spectrum of exploitation opportunities of the project results and not just the product development. Therefore, it is intended that exploitable results in the project will arise in many forms. Apart from the technologies developed within the project which will form the basis for commercial products and protected through patents and IPR agreements, exploitation opportunities include: Input to standardization activities, know-how into further EU-sponsored projects, know how into national and industry-sponsored research projects, development of new services based on the prototypes, methods and tools developed by the consortium and finally, the creation of start-up businesses to commercialize the results.
The CMSWind consortium has recognized from the early stages of this initiative the features that will comprise the proposed systems and the industrial, financial, societal and environmental benefits of its successful implementation as a whole.
The role of the industrial partners in the exploitation plan of the project results will be extremely important as they have agreed to proactively pursue the aforementioned exploitation plans. The industrial partners of the consortium will take a leading role in the exploitation of the results since these partners have the greatest commercial experience and play a vital part for the results to reach the appropriate markets.
Furthermore, there are no existing anticipated business agreements which may impose limitations on the subsequent exploitation or information or inventions generated as a result of the project.
The Final version of the PUDF is divided into two sections:
➢ A public one related to results that will be disseminated and the corresponding dissemination activities (specifying the target audience and the applied communication strategy, and presented in a verifiable way to ensure that the EC can keep track of them).
➢ A confidential one describing exploitable results and related planned activities. This section should include:
• A verifiable list of all intellectual property rights that have been applied for or registered (eg a European patent has been applied for).
• A list of all results that may have commercial or industrial applications (eg software, inventions, prototypes, compiled information and data, etc).
• An outline of the owner of each particular element of foreground, whether it is a single participant or several of them (in a situation of joint ownership).
• An explanation of how the foreground has been or is going to be used, in either further research or commercial exploitation activities, including elements such as the following:
Purpose, main features and benefits of each technology or product, derived from the research results: innovative aspects in comparison intended audience.
With technologies and products already available, needs for further R&D activity and implied risks, collaboration needs for exploitation (technology transfer activities).
Customer detection: identification of the potential customers and the factors that affect their purchasing decisions.
Features of the target market: size, growth rate, share that the technology/product could reach, driving factors likely to change the market, legal, technical and commercial barriers, other technologies likely to emerge in the near future.
How the participant (or other entity) entitled to the technology exploitation is positioned (or should be positioned) in the market, competing businesses/applications/technologies.
The PUDF is aimed at the following audiences and respectively at the fulfilment of the following objectives:
➢ European Commission: to communicate the consortium´s strategy and report on dissemination activities.
➢ Consortium partners: to inform about participants ´rights and obligations, as well as notify to other participants partners´ intentions in order to enable them to exercise their objection right in case their legitimate interest could be impaired.
Potential Impact:
CMSWind will allow several environmental benefits to be drawn by supporting the growth of the wind energy industry. It will assist in the minimisation of failures and associated costs while it will considerably improve the reliability of wind turbines. Moreover, this project will contribute to the optimisation of the efficiency of preventive maintenance thus extending the operational lifetime of assets and reducing the need for new components which contribute to the depletion of valuable natural resources.
The rise in greenhouse gas emissions from energy is unsustainable. By 2030, global greenhouse emissions could more than double due to rising use of fossil fuels, notably in developing countries. At the same time most climate experts suggest that carbon dioxide emissions need to be halved if the worst impacts of climate change are to be avoided. The IPPC states that CO2 emissions will need to reach their peak by 2015 at the latest and immediately start declining for the world to stay below a two degree Celsius increase in average temperature. New nuclear fusion and carbon capture will not be available within that timeframe. Therefore, the need for a long-lasting solution that is environmentally benign, economically sound and can be put quickly and efficiently into place is more urgent than ever. Wind energy fills all these criteria and it is one of the most important tools that Europe possesses in decarbonising power generation while maintaining the economic growth and prosperity of the region. Some member states already acquire a significant proportion of their electricity production from wind energy, eg Denmark has 21%, Spain 11% and Germany 7%.
Wind energy, apart from the fact that it does not emit any carbon dioxide emissions, does not deplete natural resources in the way that fossil fuels do, nor does it cause environmental damage through resource extraction, transportation, or waste management. It even operates without water consumption and can be deployed in sites where there is shortage of this resource, a critical factor for the decades to come particularly for the Mediterranean countries of Europe. The EU’s total installed wind energy capacity in 2007 avoided the emission of 91 million tonnes of CO2; the equivalent of taking 46 million cars off the road and equal to 26% of the EU’s Kyoto obligation. An overall production of 180 GW through wind energy by 2020 would achieve a saving of 328 million tonnes of CO2 which is the equivalent to taking 165 million cars off the road, accounting for three quarters of the EU’s overall car fleet.
The impact for Europe as a whole from the CMSWind project is summarised next:
• To support the REA’s 2007 Renewable Energy Road Map in building a more sustainable future.
• To support the Directive 2001/77/EC of the European Parliament and the Council of 27th September 2001, on the promotion of electricity from renewable energy sources in the internal electricity market.
• To support Decision No 646/2000/EC of the European Parliament and of the Council of 28th February 2000, adopts a multiannual programme for the promotion of renewable energy sources in the Community (Altener1998 to 2002), by encouraging private and public investment in the production and use of energy from renewable sources.
• To support the strategic initiatives of many member states (Germany, the UK, Denmark, Spain, etc.) in substantially increasing the market share of wind energy and promoting the decarbonisation of their economies.
• To support the binding measures of the Kyoto Protocol and the decisions made during the 2007 Bali Convention for the protection of the environment and the substantial reduction in the emission of global warming gases.
The CMSWind consortium intends to contribute profoundly in the improvement of reliability within the wind power generation industry by delivering the technology required in order to substantially reduce unexpected wind turbine failures and unnecessary costs that result from them. By increasing the reliability of the European wind turbine fleet the wind farm operators will be able to improve their maintenance strategies, minimise operating costs leading to a reduction of the cost per MWh produced, improve their efficiency, increase public confidence in renewable energy sources. Moreover it will allow the use of the financial resources recovered as a direct result of the reduction of operating costs for the improvement of existing infrastructure and the construction of new wind farms throughout Europe. The long-term implication is that it will be possible to plan the economic future of the region on the basis of known and predictable cost of electricity, derived from an indigenous energy source free of all the security, economic and environmental disadvantages associated with oil and gas. The improvement in the efficiency of maintenance procedures and operational reliability of wind turbines will also play a significant role on the way towards the decarbonisation of the European economy and achieving energy independence from oil and gas imports.
List of Websites:
www.cmswind.eu
Wind energy is presently the fastest growing renewable energy source in the world. However, the industry still experiences premature turbine component failures, which lead to increased operation and maintenance (O&M) costs and subsequently, increased cost of energy (COE). To make wind power more competitive, it is necessary to reduce turbine downtime and increase reliability. Condition Monitoring helps by reducing the chances of catastrophic failures, enabling cost-effective operation and maintenance practices, and providing inputs to improve turbine operation, control strategy, and component design.
Current wind turbine condition monitoring methodologies can be time-consuming and a costly process and fail to achieve the reliability and operational efficiency required by the industry. For these reasons existing vibration-based Condition Monitoring System (CMS) usually fail to detect defects until they become critical. This project will show the applicability of the CMS enabling the prompt detection of defects. The proposal idea is to use this experience to deploy an existing system that:
• Allows an early detection and identification of any developing defects on several components of the wind turbine’s drive train, thus helping to optimise the maintenance schedule.
• Combine the use of several sensors in order to evaluate the overall operational condition of the turbine’s generator, gearbox, bearings and main shaft.
• Use wireless sensor for rotating components monitoring using high performance powering and energy harvesting technologies.
• Fuse and analyse the data obtained through the different sensors using a single SCADA system.
This project shows the applicability of Condition Monitoring enabling the prompt detection of defects by deploying a system that allows an early detection and identification of any developing defects using Operational Modal Analysis, Motor Current Signature Analysis and Acoustic Emission Techniques applied on the gearbox, generator, and bearings and high speed shaft respectively.
Broadband radio transmission systems were developed to transmit the data from fixed or rotating frames to the ground without losing any signal resolution. The results indicate the feasibility of collecting AE signals from the rotating frame with acceptable level of noise in low to moderate wind speeds. Also other developments were executed in order to verify whether or not the noise level increases appreciably with wind speed and whether such signals can be filtered out.
New procedures, guidelines and training material has also been developed during the CMSWind Project for applying offshore instrumentation and developments and information of use to the whole Wind Energy and NDT Industries.
Project Context and Objectives:
As follows, a description of the project and objectives is presented. The work carried out during the 36 month duration of the project is divided in 9 work packages. The different work package objectives and the achievements of each one is presented.
WP1: System specifications
The objectives of WP1 were: i) To decide on the supervisory hardware that will be used in the system and identify the exact requirements of the end users so that they will be able to take advantage of the CMSWind condition monitoring system fully; ii) To select the most suitable wind turbine for the test installation. For that, a specifications document was produced and submitted to the REA as D1.1. The document comprises the requirements of the condition monitoring system and also describes the critical areas that have been selected for investigation. The document includes some data extracted from the database of vibration and AE data provided by the end-user SKM. SKM has also provided the selected Test Wind Turbine for the consortium, which is the WindMaster300. The coordinator, TWI, led this work package with the support of all the consortium members, especially the rest of the RTDs, the end-user SKM and AEE (Exploitation Manager).
WP2: Motor current signature analysis technique
In this work package all aspects of MCSA methods were thoroughly explored and the procedures that will form the basis of the MCSA technique to be used in the CMSWind System were established. Two deliverables were presented as the tangible work developed within this work package: D2.1 and D2.2 in order to meet the WP2 objectives: i) To determine which of the MCSA methods will be used, apply optimizations and thus form the basis for the MCSA technique for wind turbine generators; ii) To acquire the needed equipment as well as construct and program the interface module with the supervisory system. In D2.1 the specifications of current MCSA methods as well as the obstacles that need to be overcome were explored; an overview of the different kinds of generators for existing wind turbines was also presented, explaining the main differences between them in terms of components and therefore operating principle; the most common faults were also described, together with the existing techniques on current inspection; finally a comprehensive description on the MCSA technique (acquisition, signal processing and fault detection) and the proposed hardware that could be used for CMSWind were presented. After intensive research, the MCSA Technique widely used for off-line inspections has been confirmed as appropriate to be upgraded to an on-line technique that can monitor the generator in a wind turbine and meet the expectations of the CMSWind Project. In D2.2 the acquisition of the needed equipment and construction and programming of the interface module with the supervisory system that will form the MCSA subsystem was described. For that, two lines of work were developed in parallel: on one hand INNORA developed the acquisition cards and its embedded software in order to acquire and save in a folder the current waveform from each of the three phases of the wind turbine generator; on the other hand Brunel University developed the system’s software which will be used for the assessment and remote logging of faults. Such software consists of signal processing (three different signal processing techniques have been applied: FFT, Wavelets and Current Park’s Vector), feature extraction, feature trending and baseline development. When a fault is present the features will fall out of the limits of the baseline and therefore an alarm will be triggered.
WP3: Operational modal analysis technique
Objectives for WP3 are: i) To determine which of the OMA methods plan to be used and form the basis for the OMA technique for wind turbine gearboxes; ii) To acquire the needed equipment as well as construct and program the interface module with the supervisory system. During this work package the OMA methods that form the basis for the OMA technique for wind turbine gearboxes were determined. Also the needed equipment as well as the programme interface module with the supervisory system was constructed. Two deliverables were submitted describing in detail the mentioned work: D3.1 and D3.2. In D3.1 a state-of-the-art in terms of methods for OMA was presented. A comparison of OMA against EMA (Experimental Modal Analysis) was discussed. The approach that will be applied to the wind turbine drive train monitoring was explained, in parallel with the modal parameters to be considered when designing such monitoring system. Finally some preliminary results on laboratory trials for the OMA technique development were drawn where the following information was gathered: position for sensors number and type of sensors, data acquisition requirements, OMA algorithm testing, and evaluation and adaption of the algorithm. The selection of sensors to be used for the OMA module has been proposed. In D3.2 the OMA subsystem is presented were the acquisition cards and its embedded software was developed by INNORA and the software that forms the OMA technique in terms of signal processing, training process (to develop the baseline) and operation process (to carry out the monitoring process) were developed by Fraunhofer. The OMA subsystem was deployed on the test wind turbine in month 12 with the objective of extracting information about: the modes which will be excited when the wind turbine is in operation, optimization of number and position for the sensors, clustering of environmental data for classification, evaluation of signal processing and algorithms for modal parameters, and investigation of the relation between modal parameters and health condition of the wind turbine generator. During month 13 the acquisition hardware developed by INNORA and adapted to the OMA requirements and the acceleration sensors mounted on a plate were validated to form the OMA subsystem.
WP4: Acoustic Emission technique
The objectives of WP4 are: i) To develop an Acoustic Emission technique that can detect defects upon their creation while placed on rotating machine shafts; ii) To construct an AE system that will use an energy harvesting module to give power to the transducers and the data acquisition system and a wireless communication module capable of transmitting the data to a short range receiver. During this WP a suitable AE Module (including acquisition hardware development and suitable algorithms implementation for the detection of faults in the high speed shaft of the wind turbine’s drive train) as well as the system that will use an energy harvesting module to give power to the transducers and the data acquisition system and a wireless communication module capable of transmitting the data to a short range receiver, were developed. The different developments were reported in three deliverables, being these: D4.1 Energy harvesting module, D4.2 Wireless module performance and D4.3 AE final subsystem. In D4.1 the Energy Harvesting system conception is presented, from design to its assembly and deployment. The objective was to supply the Acoustic Emission system during operation when mounted in the rotating equipment. The system includes a battery to last during wind turbine down times, whereas its operation and duration under continuous system operation were investigated. A battery charger was also designed and constructed to charge the system’s battery. The Energy Harvesting System was successfully deployed on the test wind turbine, and the results presented in D4.3. In D4.2 the wireless communication module developed by INNORA with the objective of transmitting the acquired data wirelessly to the module subsystem was presented. Such acquisition hardware was successful tested at both laboratory scale and on the test wind turbine. Finally in D4.3 the final AE subsystem was presented, including acquisition hardware, Energy Harvesting System and AE Technique Software (baseline creation and monitoring process). The initial results obtained from its installation on the test wind turbine have been reported.
WP5: System installation on test wind turbine
Work package 5 consists of installing the three different modules (MCSA, OMA and AE with the supervisory and data logging systems) on the chosen test wind turbine: the WindMaster300 described in Deliverable D1.1. This task is dedicated to ensuring that all installed systems are working as planned. During the second period of the project, such validation on the test wind turbine was achieved successfully. Three tasks form WP5: Task 5.1 SCADA PLC installation with initial sensors, Task 5.2 Installation of MCSA, OMA and AE systems and connection to SCADA, and Task 5.3 Installed systems validation. During these tasks, the RTD partners (TWI, Brunel, Fraunhofer and Innora), with the guide and support from the SME and Associations, installed and validated the hardware and software specifically developed for this application. All the information about the hardware installation procedure and the three subsystems installation and configuration (for the three techniques) with the validation of each individual technique was presented in D5.1.
WP6: Test wind turbine analysis
This work package represents the 2nd phase of the RTD work performed during the project. At this stage, condition monitoring technique development needed raw data to be gathered (Task 6.1) in order to carry out the Final Techniques Assessment (Task 6.2). As soon as the first parts of the supervisory system hardware were installed and communication was established, the measurements begun. During the techniques assessment a correlation between the data gathered and the events occurring in the wind turbine was determined. In this way FHF, Brunel University and TWI were able to see which of the methods used in the techniques provided accurate warnings of the events prior to their occurrence and how the trend of changing measurements can show the current condition of the wind turbine machinery. Installation took place on a Vestas V90 (access granted through TWEA and Borusan). Several choices were considered and negotiation with other wind farm operators and owners (EDP, Iberdrola, Gamesa, etc.) took place through the other Associations in parallel (AEE and AEND), although the V90 in Turkey was the final machine chosen for installation. The data gathered on the V90 was used to understand the final way of manipulating data from the different methods, thus leading to the formation of the final technique guidelines, reported as D6.1. Apart from the general guidelines of usage that will enable the trained personnel to use the new methods to estimate the machinery conditions there will be several occurrences that the new techniques subsystems will record measurements so large that immediate action should be taken. This was part of Task 6.3 Automatic Alerts definition, reported in D6.2 after data analysis from the V90 wind turbine.
WP7: Development of guidelines for application of developed technology
The aim of this WP is to transfer the required expertise to the maintenance teams of the turbine operators and to prepare the training material for the SMEs installing the system. Two tasks form this WP: Task 7.1 Development of training material (reported as D7.1) and Task 7.2 Training program for SMEs. For the Training Material development, clear instruction and service manuals were created. This task met the expectations by producing a 4 instruction and 4 service manuals for the 4 distinct subsystems (MCSA, OMA, AE, SCADA) and 2 half-hour presentations, one on the installation and one on the use of the CMSWind system. This material was updated with every improvement of the system. Regarding Task 7.2 a training session was conducted at TWI facilities, lead and organized by TWI and Brunel University, where the expertise and knowhow described in the developed training material was passed onto the Associations and SMEs for them to transfer this knowledge onto potential operators. All this was supported with photos, videos and a live demonstration of the CMSWind system.
WP8: Result of dissemination and exploitation activities
Work Package 8 objective is to disseminate the technology developed during the previous work packages as widely as possible. For that, the plan was to promote the developed technique and maintain a technology transfer programme that ensures successful implementation and provided a marketing platform for the consortium, as well as to ensure a widespread use and awareness of the developed defect detection system. The tangible deliverables for this work package are a Website (developed by month 5), 2 promotional videos (developed by month 24, and updated at months 30 and then again 36), a Draft Plan for Use of Dissemination of Foreground (submitted by month 15), and the final Plan for Use of Dissemination of Foreground (submitted in month 36). Beside these, further activities have been / will be carried out to contribute to the dissemination and exploitation of the developed work through conferences, events, paper publications, etc.
WP9: Management and coordination
Work Package 9 refers to Project Management and Coordination, which has been an on-going task throughout the project duration. This work package also includes Project Management of Technical and Administrative activities and all Project Reports (Progress Reports, Technical Reports, WP Reports, and Final Project Report). As part of this Work Package the Project Coordinator planned, organized and monitored the project for administrative, legal and contractual matters, quality and standards representation and implementation. Allocated budgets & cost statements were collated and reviewed prior to submission to the EC. The Project Manager planned, organized and reviewed at 6 month intervals the technology, product and knowledge management work packages with the Work Package Leaders. Also the Project Manager dealt with the day to day inter-participant or consortium level issues.
Project Results:
WP1: System Specifications
Partners involved: TWI, AEND, Co-Services, INESCO, SKM, DGZfP, UBRUN, FHF, Innora, AEE, TWEA.
System Specifications and selected test wind turbine
1. Overview
This work package decided on the supervisory hardware used in the system, as well as take into account the end users requirements in order to select the most appropriate test installation system (wind turbine).
2. System Requirements
The system has the following capabilities:
• Operate Acoustic Emission (AE), Motor Current Signature Analysis (MCSA) and Vibration sensors (OMA).
• Operate parametric sensors such as wind speed, proximity probes (RPM) and power measurements.
• Collect the data following a specific protocol.
• Save the raw data and process it.
• Generate a signature (Baseline) that will be used as characteristic of the Wind Turbine machinery.
• Continuously monitor the Wind Turbine through the execution of the same measurements.
• Generate alarms on criteria already implemented.
The system works following the next steps:
• The system records the wind speed and power output at regular intervals of 30 minute and stores these values.
• Parametric measures such as wind speed/power/RPM will be used as a trigger to start collecting data from AE sensors and accelerometers.
• For a specific data set, the current wind speed/power is compared against a table representing the bins and the number of data per bins collected (ex. 5m/s = 0, 5.1m/s = 1 etc…). The bins are divided in increments. For example of 0.1m/s from a minimum of 3m/s to a maximum of 30 m/s for wind speed measurements.
• The mean wind speed (and mean Power output and RPM) will be the average wind speed collected in the first 5 seconds (Figure 1).
The following description is based on the wind speed acquisition:
• If this speed matches the bin requirement, the program proceeds in collecting the rest of the data. If the bin requirement is not satisfied, the values thus far collected are cleared.
• If the mean wind speed is between 3m/s and 30m/s the raw data is collected and stored in binary in order to be accessed and analysed on a later date.
• The operating parameters are tested every minute - until a required wind speed is attained at which point it samples the AE sensors and accelerometers and stores its values. The MCSA deal with as a parametric input such as wind speed, power and RPM.
• The MCSA, RPM, wind speed and power output measurement is occuring during the first 5 seconds and not overlapping with the A/D data acquisition (vibration and AE), which is always limited to 1 second of AE data and 5 seconds of vibration date (Figure 2).
• Acoustic emission data is sent through a band-pass filter (10kHz - 250kHz).
• Vibration data is sent through a low-pass filter (10kHz)
• The files are binned according to the power output of the turbine (5kW bins) and the different analysis values are averaged within the bin (Figure 2).
3. Operation process
When the system is installed in a test rig or in a real wind turbine it operates at three levels: pre-acquisition, acquisition and warning. At the pre-acquisition stage, the user has to decide the number of sensors which are necessary to monitor the machine under study and attach them.
Once this has been done, the acquisition is started. The user has to run the software setting up previously some parameters (sampling rate, number of samples, number of files per bin, …) . The system starts acquiring data. A baseline is generated as soon as all the bins have been filled. This baseline is characteristic of each machine and lets establish whether there is a fault or even if it is been initiated.
In the last stage, the software follows an algorithm which is different for each of the three techniques to assess the health of the machine studied.
4. Test Wind Turbine data
It was necessary to integrate the drivetrain provided by SKM in an operation process that allowed the investigation of fault signatures in order to characterize a failure deviation from baseline. For this reason TWI and BIC investigated potential mechanical connection of the hub to a variable speed/rotor torque.
The specifications given for the wind turbine test rig were provided by SKM:
• 300KW rated power output
• 50RPM nominal speed for the low speed shaft
All equipment is aquired by TWI in order to build a laboratory test facility. The objective was to (i) characterize the fault detection by seeding defects in the machinery and (ii) build facility for training and certification.
WP2: Motor Current Signature Analysis Technique
Partners involved: TWI, AEND, Co-Services, INESCO, SKM, DGZfP, UBRUN, FHF, Innora, AEE, TWEA.
Motor Current Signature Analysis
• Overview
This workpackage serves as a comprehensive description of the methods that form the basis of the MCSA technique that has been used in the CMSWind System further to WP1 on System Specifications.
The specifications of current MCSA methods have been defined as well as the obstacles that need to be overcome. An overview of the different kinds of generators for existing wind turbines was presented, explaining the main differences between them in terms of components and therefore operating principle. The most common faults were also described, together with the existing techniques on current inspection. Finally a comprehensive description on the MCSA technique (acquisition, signal processing and fault detection) and the proposed hardware that could be used for CMSWind are presented.
• Summary of the work done
Intensive research effort has been focused on the technique of monitoring and diagnosis of electrical machines, to name some: time domain analysis of the electromagnetic torque and flux phasor, temperature measurement, infrared recognition, radio frequency (RF), emission monitoring, detection by space vector angular fluctuation (SVAF), noise and vibration, acoustic noise measurements, harmonic analysis of motor torque and speed, artificial intelligence and neural network based techniques and MCSA.
Of all the above mentioned techniques, MCSA has been pointed out as the most suitable one for induction motors monitoring, since it is non-intrusive, it is not affected by the type of load or other asymmetries and can target all relevant faults in the induction machine. MCSA is a condition monitoring technique that is now widely used to diagnose problems such as broken rotor bars, abnormal levels of air gap eccentricity, shorted turns in low voltage stator windings, and certain mechanical problems. It has therefore been confirmed as appropriate to monitor the generator in a wind turbine and meet the expectations of the CMSWind Project .
A lot of research work on MCSA for induction machines has been carried out with the aim of understanding the state-of-the-art in this field. Hereafter, the techniques to be employed specifically for the generator as part of the CMSWind Project are further investigated and developed in the following work packages.
MCSA Final Subsystem
• Overview
The objectives were acquiring the needed equipment and constructing and programming the interface module with the supervisory system that will form the MCSA subsystem, as per the Project DoW.
Brunel University developed the system’s software which is used for the assessment and remote logging of faults on the SCADA unit and their evaluation by expert engineers acting as the system’s supervisors. In the event of a critical fault signal being received by the SCADA unit from the wind farm server, the system alerts the supervisor who re-assesses the automated data analysis carried out by the intelligent unit on the wind turbine for that particular fault. The supervising engineer then confirms to the maintenance crew the remedial actions to be taken to resolve the problem and update the maintenance schedule accordingly.
The target specifications for the MCSA subsystem are summarized in this WP according to WP1 Specifications Document.
The hardware that use for the MCSA subsystem is presented, where the acquisition system consisting in Acquisition Card and Main Board, was developed by consortium member INNORA.
The different MCSA methods found in literature and described in detail in the previous heading that assess the generator current have been programmed in LabVIEW as sub.vi(s). The higher level monitoring software that creates the baseline and raise alarms in the event of a fault has been encoded with an architecture based on state machine programming, also with LabVIEW.
• Summary of the work done
Modern measurement techniques in combination with advanced computerized data acquisition and processing show new ways in the field of induction machine monitoring by the use of spectral analysis of operational process parameters. In the present WP, the process of data acquisition, signal processing and fault detection for the diagnosis of an induction machine was presented, along with the hardware required for signal acquisition and data processing.
The MCSA methods to be used which form the basis for the MCSA approach for wind turbine generator monitoring were determined and programed in LabVIEW, being these: Fast Fourier Transform, Wavelets Analysis (based on Multiresolution Techniques), and Current Park’s Vector. Fast Fourier Transform can identify faults related to Broken Rotor Bars, Shorted Turns, Air Gap Eccentricity, Gearboxes, and Bearings. Wavelet Analysis and Current Park’s Vector can support the FFT Technique by detecting some of the mentioned faults at no-load conditions, which is the main limitation for the FFT Technique.
The next stage, as per the DoW, was testing and validation of the MCSA subsystem. As
described in WP1 and previously in WP2 the MCSA subsystem has to be validated both at laboratory scale (in a test rig from SpectraQuest), and tested in the WindMaster300 provided by consortiummember SKM. The MCSA subsystem has to be optimized for an enhanced performance as part of an integrated solution for the complete CMSWind Condition Monitoring System.
WP3: Operational Modal Analysis Technique
Partners involve: FHF, Innora.
Operation Modal Analysis
The operational modal analysis (OMA) does not rely on specific excitation. The method rather assumes a white noise spectrum as input. The loads which are applied directly to the structural system cannot be measured and are not known. Thus it is not possible to separate the dynamic influence of the environment and the structure itself.
Both are recorded in the sensor network to record the structure’s response. However this is actually a great advantage of the method, since modal parameters are extracted for structures under actual operational conditions. Results of previous publications have shown that modal parameters for the static and operational conditions may differ greatly.
To identify a system description and derive the needed modal parameters, multiple approaches are available for OMA. The methods can be classified in the frequency domain and time domain variants.
• Frequency-Domain Decomposition (FDD)
• Enhanced Frequency-Domain Decomposition (EFDD)
• Curve-Fit Freq. Domain Decomposition (CFDD)
• Stochastic Subspace Identification (SSI)
After identifying spectral peaks using one of the above mentioned methods, one has to divide the spectral peaks into frequencies coming from noise and periodic distortion (both unwanted) or structural frequencies (wanted). For this kurtosis analysis will be applied. The efficiency of the method depends on the quality and amount of the recorded signals as well as the statistical properties of the distortions and the modal frequencies.
Operation Modal Analysis Final subsystem
This is a summary of the work performed within tasks T3.2 and T3.3 which has the objectives of acquiring the needed equipment and constructing and programming the interface module with the supervisory system that now form the OMA subsystem, as per the Project DoW.
The target specifications for the OMA subsystem were summarized according to WP1 Specifications.
The hardware that will be used for the OMA subsystem is presented, the acquisition system used in the Acquisition Card and Main Board was developed by consortium member INNORA. The acquisition software use for the technique (a suite of 5 programs designed to set up the acquisition systems and receive the data) was also developed by INNORA.
The INNORA acquisition system was successfully tested for application of acceleration signal acquisition for modal analysis. The next thing to do is the realisation of triggering by environmental events and the integration of software tools in a LabVIEW environment for averaging of signals and modal parameter extraction.
WP4: Acoustic Emission Technique
Partners involve: TWI, Innora.
Energy Harvesting Module
One of the objectives of this WP is to present the development of an Energy harvesting module with potential applications on rotating equipment such as the low speed or high speed shaft.
In the CMSWind project an Acoustic Emission (AE) system, mounted directly on the main rotating shaft of the wind turbine, is used to detect defects.
As the system (sensor and node) is attached to a rotating shaft, it is not possible to power it from the main by means of a cable connection. In view of this, the design of an Energy Harvesting System is necessary.
When the shaft is rotating, the AE system is powered and operates continuously. However, during a wind turbine down time the shaft is not rotating and a battery is required to power the system. Moreover, a battery charger is needed to ensure the storage of the energy harvested.
The former components allow securing enough energy from the shaft to power the AE system. A schematic of the entire system described here is shown in Figure 5. Figure 6 and Figure 7 show the Energy Harvesting system designed.
The design and construction of the Inertia Energy Harvesting System was completed to supply the Acoustic Emission system during operation when mounted in rotating equipment. The system includes a battery to last during wind turbine down times, whereas its operation and duration under continuous system operation were investigated. The different parts of the system were identified, designed and constructed. A battery charger was also designed and constructed to charge the system’s battery.
Wireless Module performance
Another objective of this WP was to develope the wireless transmitter/receiver module necessary to implement the Acoustic Emission technique on wind turbine machinery rotating parts. Specifically, the objective of this report is to present the selection process of the wireless protocols, the microcontroller to wireless module connection method, and the assessment methodology for the validation of the developed solution.
It should be noted here that, in CMSWind project, we developed a system that digitize the sensor data with a high accuracy A/D converter and transmit it wirelessly to a receiver placed nearby in the wind turbine nacelle. AE sensor, A/D converter and wireless module are powered by an Energy Harvesting module and the whole system will be integrated in a package as miniature as possible. This whole effort is done so that the sensor can be placed on the rotating shaft of the wind turbine. Figure 7 depicts the proposed connection architecture of the wireless sensors.
CMSWind project required Acoustic Emission sensors installed on rotating parts of Wind turbines to locate cracks on the materials. The rotating nature of these parts makes wired sensors unusable for the task. Moreover use of contact rings is expensive, difficult to install and would have to be tailor made to every wind turbine model. Wireless data transmission, energy harvesting and small analog to digital converters were proposed as a viable alternative solution. In this WP we have configured a wireless transmission scheme that is capable of transferring the volume of data required and has sufficiently low power consumption. The proposed solution fulfills all the system requirements as described in WP2 and was validated through two tests.
Acoustic Emission Integrated
The AE sub-system was designed, integrated and tested. To do these tasks TWI developed the hardware and software which is used for the assessment and remote logging of AE data.
The hardware that is used for the AE sub-system consists on an Acquisition Card and Main Board was developed by consortium member INNORA.
The different AE techniques that assesses the machinery were programmed in LabVIEWTM as sub.vi(s). The higher level monitoring software creates the baseline and raise alarms in the event of a fault. The software was encoded with an architecture based on state machine programming, also with LabVIEWTM.
An integration process was held between TWI and INNORA in order to incorporate the wireless data acquisition hardware.
Various tests were run in the laboratory, in order to initialise and set up the system. Once everything was working in the laboratory the system was installed in the test wind turbine. Data was gathered for the full speed operation range and then processed in order to look for the AE signals of interest.
Some mounting methods were tested in order to assess the sensor response to the AE events depending on the way it was attached. Finally, a magnetic coupling was chosen as it gave the best results and was the most reliable mounting process.
The Energy Harvesting System was mounted and tested both statically and dynamically. Due to the centrifugal force and the inertia of the system at high speed, the static part of the system was attached magnetically to the generator.
Using this attachment method, the rotating movement is fully controlled and the EHS is protected from inertial damage. The designed piece to support those efforts is cross shape and has four magnets to ensure a good grip for the bearing. The plastic part has to mesh with two gears located in the rotating part. These gears were joined to the motors and produce its movement. As working as generators, the electricity produced by the motors is used to supply the whole electronic system and sensors.
All the parts that constitute the acquisition system have a place in the rotating plate. The locations were developed using little plastic marks to attach their boxes. Apart from the sensors, which were located in two holes near to shaft, acquisition cards and battery charger circuits are placed in different compartments.
A correlation between AE parameters and the shaft rotation speed was carried out for the analysis of the data gathered.
WP5: System Installation on test wind turbine
Partners involve: TWI, AEND, CO-Services, INESCO, SKM, DGZfP, UBRUN, FHF, Innora, AEE, TWEA.
Validation of the system in the test wind turbine
Here it is summarized the work performed within tasks T5.1 T5.2 and T5.3 where the Acoustic Emission, Motor Current Signature Analysis (MCSA) and the Operational Modal Analysis (OMA) subsystems developed in the previous Work Packages, were installed on a WINDMASTER 300 wind turbine.
During these tasks RTD partners, TWI, Brunel, Fraunhofer and Innora installed, and validated the hardware and software that was developed for the project.
The three techniques used for the monitoring of the wind turbine rotating parts were validated in the WINDMASTER 300:
• MCSA Validation Results
Some of the results obtained from the gathered data of the WINDMATER300 nacelle are presented in Figure 8. Ten files were acquired per bin in order to generate the average and standard deviation of the crest factor feature to define the different baselines of each interval.
Different baselines that allow health monitoring for the generator of the WINDMASTER300 nacelle were developed. Using this baseline generation approach, the generator monitoring process can be performed by comparing each new set of data acquired to the original baseline created during the initial stage.
• AE Validation Results
Some of the results obtained from the gathered data of the WINDMATER300 Nacelle are presented in Figure 9 and Figure 10. Ten files have been acquired per bin in order to generate the average and standard deviation of the AERMS signal.
Two resonant AE sensors were installed on the low and high speed shafts of the test wind turbine for feature extraction and baseline development. The extracted feature was the AERMS signal, gathered at 2,5 M Samples/s during 0,8 seconds, where ± standard deviation of the AERMS define the operational limits for the different parts under study (Figure 9 and Figure 10).
The AE subsystem was installed in the test wind turbine. The data acquired was used to develop a healthy system signature including the AERMS parameter. As expected, there was a clear correlation between the rotational speed and the AERMS mean and standard deviation for each of the speed bins. This denoted that the baseline represents a pattern for AE monitoring that increases alarm definition accuracy.
Further research and data processing was carried out during WP6 where the final techniques developed are presented and the alarm levels definition studied.
• Vibration Validation Results
First the identification of the structural modal frequencies was carried out by FDD on the non–rotating drive train, based on impact excitation at the low speed shaft. The estimated first singular values were evaluated in order to build the reference for all further rotation experiments. Then for each ‘Bin’ the parameter ‘First singular value’ were determined as a reference signature representing this specific ‘Bin’. If the state changes, this can be detected for example by the correlation coefficient between data extracted during operation and reference data. For further parameter extraction frequency ranges of interests were specified as shown in Figure 11. Within this frequency ranges signal parameter were calculated like peak frequency and power of the frequency band.
During rotation, the signals were strongly influenced by the frequencies of the shaft rotation, the gear engagements and the bearings. Figure 12 shows the spectrum of time signal at Bin 2: (frequency of the high speed shaft: 8.3 Hz). The structural modes were very weak and some are not excited, for example in the range below 20 Hz.
In Figure 12, the first singular values of all Bins of the y-axis were collected in a surface plot. The frequency range around 80 Hz and 110 Hz (blue marked region in12). The modes below 20 Hz were excited only in the BIN’S with the higher rpm (6-10). The mode around 80 Hz and 110 Hz were better excited in the Bins one to six.
Referring to the frequency ranges estimated in non-rotating experiments, the parameters for the monitoring database were generated.
During the experiments, a database containing the reference singular values for each Bin was established for monitoring application for the TWI WINDMASTER. Global parameters based on singular value decomposition of all vibration data as well as parameters for each sensor within the identified frequency range were extracted and can be used for further investigations.
Hardware and were tested successfully. Because the parameters were specific for the used drive train, the reference data set was adapted to the specific device. So not all algorithms, integrated in the LabViewTM Software, runed automatically. A learning process must be performed by the user to define the frequency ranges of interest for the further calculation of parameters and to save them as reference. Description of the developed LabViewTM Software is contained in WP7.
WP6: Test Wind Turbine Analysis
Partners involved: Test Wind Turbine Analysis
The objective of this work package was the installation of the advanced Condition Monitoring System which has been developed as part of the CMSWind FP7 Project on a real wind turbine. Different approaches have been considered in order to complete the work package successfully.
It is presented first of all, the CMSWind system integration in terms of hardware and the embedded software that governs the system.
The CMSWind consortium received great interest from the Wind Industry operators in different countries: EDF in UK, Iberdrola in Spain, Borusan in Turkey, CRES in Greece, etc., for the system to be installed on the most relevant wind turbines available in the market (Vestas, Gamesa, etc.).
The option which was implemented was installation on a Vestas V90 operated by Borusan. The results of the CMSWind System validation in a real wind turbine are presented in this report.
Final Techniques development
The final CMSWind system installed on an operating wind turbine. Results from ongoing data gathering process and data analysis for the proposed techniques Motor Current Signature Analysis (MCSA), Acoustic Emission (AE) and Operational Modal Analysis (OMA) have been presented in D6.1. The algorithms proposed in the deliverables D2.2 D3.2 and D4.3 for the three techniques before were optimised and final algorithms were presented.
The location selected to perform the trials was the Bandirma (Turkey) Wind Energy Power Plant where Borusan have wind turbines. The installation was carried out by BIC and TWI in collaboration with Borusan and Vestas. The wind turbine selected to validate the CMSWind System was a Vestas V90 – 3MW.
Parametric outputs such as power output and wind speed were provided by Borusan weekly as TWI/BIC were not allowed to acquire that information using the CMSWind System. As an example, Figure 13 shows the variations in the power output and the speed of the rotor during in May 2015. The power generated by the wind turbine changes between 0 and 3MW while the speed at the input of the gearbox changes between 0 and 16rpm. That means that the maximum speed the generator spins at is 1672rpm.
The system was installed in a Vestas V90-3MW. The wind turbine selected was number 14 as there was not any other Condition Monitoring System installed that could interfere with the performance of the system developed within the project. Figure 14 shows how the CMSWind System was lifted using a crane.
Four magnets (Figure 15a) were placed on the Condition Monitoring System case to achieve the correct attachment between the wall of the Nacelle and the integrated system. Figure 15b and Figure 15 c depict how the DAQ system was installed at the front left hand side of the nacelle (just on top of the gate).
The data gathered by the system is sent via Ethernet cable to a laptop where the data is stored. A USB 3G Dongle was used to have remote access to the CMSWind System. The software which gives access to the computer is ‘Go To My PC’. This software is a powerful tool for flexible work which allows the user to perform daily checks on whether or not the software is running and that the data is being acquired correctly. It also enables the user to transfer files between the laptop in Turkey and the one where the post processing is going to be carried out.
Sensors installation
For the OMA application at the Vestas wind turbine three accelerometers were attached at the gearbox as shown in Figure 17:
• OMA 1: First Planetary stage
• OMA 2: Second planetary stage
• OMA 3: Helical stage
The probe selected to monitor the condition of the 3-phase generator (MCSA) was the GMC-IProsys ACP 3005/24 Current Probe & Clamp. The probe is suitable for the 1000V cabling on the Vestas V90 generator. Further specifications are summarized in D6.1. Each phase is split in four cables to reduce the amount of current travelling through the wires. Therefore, the probe was installed in one of them (Figure 18).
Figure 19 depicts the location selected to place the Acoustic Emission sensors. The shaft which connects the gearbox with the generator was covered so there was not easy access to the bearings on both sides of the shaft. The transducers were attached as close as possible to the bearings. Acoustic Emission sensor 1 was attached using grease and a magnetic holder to the gearbox case close to the output bearing (Figure 19a). Acoustic Emission 2 was attached following the same procedure close to the main bearing located at the generator input (Figure 19b). Signals coming from each sensor are gathered by two different cards. Signals gathered by AE1 are picked up by the card 00x20 whereas the ones coming from AE2 are picked up by the card 00x13.
The final version of the CMSWind Software for post-processing is presented as follows, with enhanced functionalities and improved characteristics, as compared to its inception described in the first stages of the work (D2.2 D3.2 and D4.3).
CMSWind Final Modules
The software was developed in such a way to facilitate the adoption of the technology by potential users. As a result, the software code was developed in LabVIEW, a graphical programming software development environment used by engineers and scientists to develop sophisticated test, measurement and control applications. LabVIEW and its graphical dataflow programming language is characterised by ease of use and provides a better way of solving problems than traditional lower level alternatives. The practical benefit of the graphical approach is that it puts more focus on data and the operations being performed on that data and abstracts much of the administrative complexity of computer programming such as memory allocation and language syntax. As a result, new programmers typically report shorter learning curves compared to other programming languages because they can relate LabVIEW code to flow charts and other familiar visual representations of processes.
The software developed within the project has the capability to generate a specific baseline during the first weeks of installation: Signature Stage. After the baseline was created, the new data acquired by the system was compared to the baseline in order to find out the status of the machine from that moment onwards.
Data is stored in bins according to the speed of the high speed shaft or the wind turbine power output (as the user prefers). The system calculates the mean of the files stored in each bin and then it establishes the maximum and minimum limits which delimit the normal operation area of the machine being monitored. It is right after that when the monitoring stage begins.
As long as the new data acquired falls within the specified limits, the machine is said to be under normal operation conditions which means that the machinery inside the wind turbine is healthy. Otherwise the alarms will be triggered indicating that the system is out of normal operation conditions, showing the presence on potential defects in the part of the drive-train under monitoring.
Results
From the Acoustic Emission analysis, the following conclusions were extracted:
- The Acoustic Emission module of the CMSWind System was validated in the present study using real data gathered from an operating 3MW wind turbine in Turkey (Bandirma Wind Energy Power Plant).
- Given a specific working condition, the software output can be regarded as the baseline features of measured AE signals. Any significant deviation of practically measured AE and vibration signal features from the baseline can be explained to be due to some kind of abnormality inside turbine components and/or system. Consequently, this information can be effectively used to achieve the objective of condition monitoring and fault diagnosis.
- The RMS and Peak Value of the Acoustic Emission measurements from the turbine drive train were used for the establishment of the baseline model.
- The analysis of the real data collected from the wind turbine was processed for three different numbers of files per bin.
- The baseline generation process is sensitive to changes in the number of files used. The higher the number of files is the more accurate the limits are, reducing the number of false alarms.
From the Operational Modal Analysis could be concluded for the application:
- The vibrational data are reliable for Operational modal analysis if he exceed a minimum level
- The generator rpm I the most important parameter for binning. The optimal value is 30- 60 rpm.
- The harmonic excitation due rotation causes the most dominant frequency component for the wind turbine in good condition
- As monitoring parameters has to be used the frequency of peaks in the first singular and values ,their damping and the vibration signal power of all sensors at the identified peak frequency (modal frequency or harmonic excitation).
- The parameter are extracted within frequency ranges, which has to be defined by analysing data from reference measurements of the healthy turbine and from analysing data in the case of damages . The parameters for each bin and frequency range are compared with their reference level and limits.
Report on suggested automatic alarm limits
The basic principle of the proposed automatic alert system is a reference method. There were extracted baselines from the set of reference data. The statistical properties of these parameters as mean deviation and extreme values were regarded for the process of alarm definition. The parameters were supervised and warnings were created if the defined limits for the parameter are exceeded. The limits for the monitoring parameters were established from the data gathering process of the wind turbine during operation in good condition. The monitoring parameters from the three techniques AE, MCSA and OMA are automatically collected in bins, defined by wind turbine operational data and are monitored within this bins. If damages occur and can be identified, the data base for reference data could be actualised and will be stated more precisely.
• OMA Strategy
Base of the automatic alert system is the reference data set. From this data set the baselines and limits for the monitoring parameter: frequency, damping and vibration signal power were extracted. For the OMA algorithm the range of rpm for one bin is important and should be 60 min-1. The first reference data set was collected after installing the system on the healthy wind turbine. Following the proposal in 3.1. the limits for the parameters were set to
Limit (parameter) = Mean (Parameter) 2*standard deviation (parameter)
If too much false alarms are generated during the data acquisition period from the healthy wind turbine, the limit definition has to be adapted by including the data which produced false alarms.
During the monitoring process the actual data is first sorted into the referred bin directory after analysing the wind turbine generator rpm. The specified number of data is collected within the bins for averaging (20 measurements). If the bin is complete the OMA parameters are calculated and averaged as described in deliverable D6.1. After comparing the parameter with the baseline limits, a warning is created if the parameters are out of limits. If a specified number of warnings is exceeded, it must be proved if there a damage at the wind turbine is occurred. Nevertheless the algorithm is universal and can be applied to other types of wind turbine. The binning process for the reference data as well as for monitoring is described in deliverable 6.1.
• AE Strategy
The purpose of data analysis is to provide representative AE level across a range of wind speeds or power output for improving the reliability of wind turbines and thereby help define appropriate AE limits.
If the limits are too far from the target value, small deviations from the target value may go undetected but if the limits are too close to the target value there will be a large number of false alarms (that is there will be a signal for action when the process mean is on target and no action is necessary).
It has been convenient to set the warning limits so that, if the mean is on target, 95% of sample means will lie within them. Since we are dealing with sample means it is reasonable to assume they are normally distributed. Choosing 1.96 times the standard deviation only 5% of the values will be outside the limits if the features under study are in the same condition as when the signature was calculated. In general, the following warning limits will be set:
Where μ is the target value and σ is the standard deviation. The standard deviation will be an estimate of the true value and 95% is somewhat arbitrary figure. For these reasons and for simplicity the limits are often set at:
In this specific case, the target value is not known as there are not standards available for establishing Acoustic Emission limits in rotating machinery. When this happens, a large sample should be taken when the process is running satisfactory and the sample mean used as a target value.
In order to inform machine operators or engineers of important structural damages in the wind turbine machinery which is being monitored, an alert strategy has been accomplished. Figure 3 shows the flow which is followed by the software developed in LabVIEW.
Every time an Acoustic Emission (AE) is gathered by the data acquisition system, the software needs to allocate that piece of data in a specific bin as the speed varies constantly. The bin is selected according to the power or the wind speed provided by the SCADA system. Firstly, the software double checks whether or not the bin is full of data. If it is not full, that piece of data is stored within that bin. If the number of files is higher than the maximum number of files per bin allowed, it means that the software has enough data to build up the baseline. That measurement will be used for monitoring purposes.
This AE value can fall inside or outside the limits. It cannot be considered that there is a defect just because there is an outlier. It could just be a false alarm. For this reason, a criteria to establish the alarm limit has to be established. Due to the fact that a new set of data is collected every 30 minutes, the alarm will be activated whether the average of the last three files gathered for that speed falls outside the limits.
The selection of the number of files is completed taking into account that the time interval between measurements is 30 minutes and the power value changes very fast. Therefore, the minimum temporal distance between the last three files stored for a specific bin is 90 minutes. This distance will be usually higher as not many files are stored in the same bin consecutively. If the average of these three files falls outside the limits the software will switch on the alarms to let the operator know about the problem.
• MCSA Strategy
The limits created in the signature mode for the selected features (please see deliverables D2.1 and D2.2) are set as follows. Once a statistically significant sample for one bin has been acquired, all the parameters are averaged and the standard deviation is calculated. The limit selection will affect the sensibility of the system. The lower the alarm limit is selected the smaller size of defects can be detected. More false alarms can be generated though. The equation used to calculate the limits is set as follows:
Where σ is the standard deviation.
The limit selection has been investigated statistically. Considering the distribution of the features extracted from the signal as normal distribution (Gaussian function) only 68.27% of the values lie within one standard deviation of the mean. Therefore, even if the bearings are in good condition 32.72% of the values will be outside the limits. Choosing double the standard deviation, only 4.55% of the values will be outside the limits if the features under study are in the same condition as when the signature was calculated. Thus, the possibility of generating false alarms is reduced significantly.
In order to evaluate real trends and not only one value that can fall off the established limits, the following strategy is implemented.
Every time a new value is acquired, the software checks whether we are at the baseline stage or at the monitoring stage. If all the values required for the baseline stage had been previously gathered (i.e. the array for a particular bin is full) then we go to the monitoring stage, where each new value acquired is compared against the baseline.
At this point a new array is defined: the ‘OUTSIDE’ array, which is an empty array of a specific size (the number of values, ie the size of the array, can be modified by the user). Every time a new value falls outside the limits established by the baseline it becomes stored in the ‘OUTSIDE’ array. For the ‘OUTSIDE’ array to be populated, the values falling outside the limits must be consecutive, ie if the new value falls within the limits established by the baseline the ‘OUTSIDE’ array is emptied.
Once a number of successive atypical events is gathered, hence the ‘OUTSIDE’ array is completed, the array is analysed. If the trend of the ‘OUTSIDE’ array is crescent (n1
WP7: Development of guidelines for application of developed technology
Partners involved: TWI, AEND, Co-Services, INESCO, SKM, DGZfP, UBRUN, FHF, Innora, AEE, TWEA.
Training Material
Wind energy is presently the fastest growing renewable energy source in the world. However, the industry still experiences premature turbine component failures, which lead to increased operation and maintenance (O&M) costs and subsequently, increased cost of energy (COE). To make wind power more competitive, it is necessary to reduce turbine downtime and increase reliability. Condition monitoring helps by reducing the chances of catastrophic failures, enabling cost-effective operation and maintenance practices, and providing inputs to improve turbine operation, control strategy, and component design.
This WP provides training information of wind turbine drivetrain condition monitoring based on the use of CMSWind system.
Since the gearbox and generator have shown to have the longest downtime and are the most costly subsystems to maintain throughout a turbine’s 20 years of design life, they have been chosen as the main targeted subsystems to be monitored. Installation of CMSWind system and application of Wind turbine drivetrain condition monitoring practices will be addressed in detail along this report.
The custom hardware developed specifically for CMSWind project as well as the development and adoption of the three novel monitoring techniques are explained throughout this training material.
The training material is divided in the four following sections:
- CMSWind hardware architecture and configuration.
- MCSA module description and installation.
- OMA module description and installation.
- AE subsystem description and installation.
The current training manual presents the use and installation of the different CMSWind systems.
The installation process has been developed and explained taking into consideration the test wind turbine installed at TWI for testing purposes during CMSWind project. Due to this fact, additional considerations comprising both health and safety and operational conditions should be taken into account if the system is to be installed in a different wind turbine. However, and thanks to the architecture of the CMSWind system, the applicability of the different techniques and use of the hardware is scalable to different sizes and configurations of wind energy systems.
The training material was developed with a focus on the potential user which does not necessarily need to know technical details about the use of advanced condition monitoring techniques and systems in order to understand the information provided by the CMSWind system. This was fully ensured through the evaluation of professionals not related with the wind energy maintenance or diagnostic techniques sectors in order to ensure that it is fully understandable by the potential user.
WP8: Result of Dissemination and explotation activities
Partners involved: TWI, AEND, Co-Services, INESCO, SKM, DGZfP, UBRUN, FHF, Innora, AEE, TWEA.
Website
As part of the CMSWind Project, Brunel University hosts a website on behalf of the consortium with the domain name www.cmswind.eu.
The website consists of the following tabs:
• Home: Title and summary of the CMSWind Project. The flyers specifically developed for the CMSWind Project can also be found in the home page.
They are in three different languages, for dissemination purposes.
• Project Background: a deeper overview of the project is presented in this section, including more details on project objectives, the proposed solution and work packages.
• Consortium Members: a brief description of each of the CMSWind consortium partners is included in this tab. The logo of each company has also been incorporated.
• News & Events: A very important part of the CMSWind Project is the dissemination activities and events. In this tab, all the conferences and events where information on CMSWind Project is published and disseminated by the consortium members are presented.
• Contact us: contact details for Project Coordinator, Dr. Slim Soua, TWI Limited.
• Private Area: secure member area to upload deliverables, meeting minutes and other documents for the consortium partners.
A snapshot of the website is presented in Figure 21.
Promotional videos
As part of the CMSWind Project, TWI developed two promotional videos in collaboration with the CMSWind project consortium.
These two videos were developed in parallel by the RTD’s and TWI as coordinator formatted and merged the different clips into a video.
Additionally, the promotional videos were published in the CMSWind Website for public access and dissemination of project results.
The different tasks carried out in order to develop the videos are described as follows:
• Recording and development of clips explaining the system installation.
• Recording and development of clips explaining system configuration.
• Recording and development of clips explaining system operation.
• Provision of clips and pictures related to wind energy sector from project partners.
• Integration and formatting of videos into a common style.
• Development and synchronization of speech text to be embedded in the video.
These videos complement other dissemination materials produced by the CMSWind consortium as well as banners and posters and are being used for project dissemination.
Draft Plan for Use of Dissemination of Foreground
A Plan for Use and Dissemination of Foreground (PUDF) was carried out as set out in the Description of Work for the CMSWind project. The interim PUDF has been reviewed and amended during the course of the project as directed by the Project Steering Committee (PSC) and implemented by the Exploitation Manager, Mr Kilian Rosique from the Asociación Empresarial Eólica, who has been responsible for the planning and coordination of all activities related to adequate and timely dissemination of the project results within the consortium and externally within EU and global NDT and Wind energy community ensuring an appropriate level of disclosure for each action.
The PUDF includes a plan to attract investment for the CMSWind technology beyond the duration of the project.
Final Plan for Use of Dissemination of Foreground
The consortium has looked into the full spectrum of exploitation opportunities of the project results and not just the product development. Therefore, it is intended that exploitable results in the project will arise in many forms. Apart from the technologies developed within the project which will form the basis for commercial products and protected through patents and IPR agreements, exploitation opportunities include: Input to standardization activities, know-how into further EU-sponsored projects, know how into national and industry-sponsored research projects, development of new services based on the prototypes, methods and tools developed by the consortium and finally, the creation of start-up businesses to commercialize the results.
The CMSWind consortium has recognized from the early stages of this initiative the features that will comprise the proposed systems and the industrial, financial, societal and environmental benefits of its successful implementation as a whole.
The role of the industrial partners in the exploitation plan of the project results will be extremely important as they have agreed to proactively pursue the aforementioned exploitation plans. The industrial partners of the consortium will take a leading role in the exploitation of the results since these partners have the greatest commercial experience and play a vital part for the results to reach the appropriate markets.
Furthermore, there are no existing anticipated business agreements which may impose limitations on the subsequent exploitation or information or inventions generated as a result of the project.
The Final version of the PUDF is divided into two sections:
➢ A public one related to results that will be disseminated and the corresponding dissemination activities (specifying the target audience and the applied communication strategy, and presented in a verifiable way to ensure that the EC can keep track of them).
➢ A confidential one describing exploitable results and related planned activities. This section should include:
• A verifiable list of all intellectual property rights that have been applied for or registered (eg a European patent has been applied for).
• A list of all results that may have commercial or industrial applications (eg software, inventions, prototypes, compiled information and data, etc).
• An outline of the owner of each particular element of foreground, whether it is a single participant or several of them (in a situation of joint ownership).
• An explanation of how the foreground has been or is going to be used, in either further research or commercial exploitation activities, including elements such as the following:
Purpose, main features and benefits of each technology or product, derived from the research results: innovative aspects in comparison intended audience.
With technologies and products already available, needs for further R&D activity and implied risks, collaboration needs for exploitation (technology transfer activities).
Customer detection: identification of the potential customers and the factors that affect their purchasing decisions.
Features of the target market: size, growth rate, share that the technology/product could reach, driving factors likely to change the market, legal, technical and commercial barriers, other technologies likely to emerge in the near future.
How the participant (or other entity) entitled to the technology exploitation is positioned (or should be positioned) in the market, competing businesses/applications/technologies.
The PUDF is aimed at the following audiences and respectively at the fulfilment of the following objectives:
➢ European Commission: to communicate the consortium´s strategy and report on dissemination activities.
➢ Consortium partners: to inform about participants ´rights and obligations, as well as notify to other participants partners´ intentions in order to enable them to exercise their objection right in case their legitimate interest could be impaired.
Potential Impact:
CMSWind will allow several environmental benefits to be drawn by supporting the growth of the wind energy industry. It will assist in the minimisation of failures and associated costs while it will considerably improve the reliability of wind turbines. Moreover, this project will contribute to the optimisation of the efficiency of preventive maintenance thus extending the operational lifetime of assets and reducing the need for new components which contribute to the depletion of valuable natural resources.
The rise in greenhouse gas emissions from energy is unsustainable. By 2030, global greenhouse emissions could more than double due to rising use of fossil fuels, notably in developing countries. At the same time most climate experts suggest that carbon dioxide emissions need to be halved if the worst impacts of climate change are to be avoided. The IPPC states that CO2 emissions will need to reach their peak by 2015 at the latest and immediately start declining for the world to stay below a two degree Celsius increase in average temperature. New nuclear fusion and carbon capture will not be available within that timeframe. Therefore, the need for a long-lasting solution that is environmentally benign, economically sound and can be put quickly and efficiently into place is more urgent than ever. Wind energy fills all these criteria and it is one of the most important tools that Europe possesses in decarbonising power generation while maintaining the economic growth and prosperity of the region. Some member states already acquire a significant proportion of their electricity production from wind energy, eg Denmark has 21%, Spain 11% and Germany 7%.
Wind energy, apart from the fact that it does not emit any carbon dioxide emissions, does not deplete natural resources in the way that fossil fuels do, nor does it cause environmental damage through resource extraction, transportation, or waste management. It even operates without water consumption and can be deployed in sites where there is shortage of this resource, a critical factor for the decades to come particularly for the Mediterranean countries of Europe. The EU’s total installed wind energy capacity in 2007 avoided the emission of 91 million tonnes of CO2; the equivalent of taking 46 million cars off the road and equal to 26% of the EU’s Kyoto obligation. An overall production of 180 GW through wind energy by 2020 would achieve a saving of 328 million tonnes of CO2 which is the equivalent to taking 165 million cars off the road, accounting for three quarters of the EU’s overall car fleet.
The impact for Europe as a whole from the CMSWind project is summarised next:
• To support the REA’s 2007 Renewable Energy Road Map in building a more sustainable future.
• To support the Directive 2001/77/EC of the European Parliament and the Council of 27th September 2001, on the promotion of electricity from renewable energy sources in the internal electricity market.
• To support Decision No 646/2000/EC of the European Parliament and of the Council of 28th February 2000, adopts a multiannual programme for the promotion of renewable energy sources in the Community (Altener1998 to 2002), by encouraging private and public investment in the production and use of energy from renewable sources.
• To support the strategic initiatives of many member states (Germany, the UK, Denmark, Spain, etc.) in substantially increasing the market share of wind energy and promoting the decarbonisation of their economies.
• To support the binding measures of the Kyoto Protocol and the decisions made during the 2007 Bali Convention for the protection of the environment and the substantial reduction in the emission of global warming gases.
The CMSWind consortium intends to contribute profoundly in the improvement of reliability within the wind power generation industry by delivering the technology required in order to substantially reduce unexpected wind turbine failures and unnecessary costs that result from them. By increasing the reliability of the European wind turbine fleet the wind farm operators will be able to improve their maintenance strategies, minimise operating costs leading to a reduction of the cost per MWh produced, improve their efficiency, increase public confidence in renewable energy sources. Moreover it will allow the use of the financial resources recovered as a direct result of the reduction of operating costs for the improvement of existing infrastructure and the construction of new wind farms throughout Europe. The long-term implication is that it will be possible to plan the economic future of the region on the basis of known and predictable cost of electricity, derived from an indigenous energy source free of all the security, economic and environmental disadvantages associated with oil and gas. The improvement in the efficiency of maintenance procedures and operational reliability of wind turbines will also play a significant role on the way towards the decarbonisation of the European economy and achieving energy independence from oil and gas imports.
List of Websites:
www.cmswind.eu