Skip to main content

On-line Intelligent Diagnostics and Predictive Maintenance Sensor System Integrated within the Wind Turbine Bus-Bar structure to aid Dynamic Maintenance Scheduling

Final Report Summary - WIND TURBARS (On-line Intelligent Diagnostics and Predictive Maintenance Sensor System Integrated within the Wind Turbine Bus-Bar structure to aid Dynamic Maintenance Scheduling)

Executive Summary:
The Wind Energy Industry has ambitious plans to increase its share of the European electrical energy supply to over 20% within the next two decades. The sector is developing rapidly, with significant investment to increase capacity, reduce variability and improve reliability. Innovation is key to realising these ambitions, particularly in the offshore sector where costs are currently much higher than for conventional energy sources.
Studies on the reliability of wind turbines show that the electrical system is the most common source of failures, with the frequency converter responsible for 50% of these. Yet whilst there are many commercially available condition monitoring systems which can detect impending faults within the drive train of a wind turbine, none are currently available which perform the same function for the converter. The WindTurBars project was created to meet the need for such a system.
WindTurBars was a two year EU FP7 project running from 2012-2014, which brought together 8 partners from research and development centres and small and medium enterprises from across the EU. WindTurBars has developed an On-line Intelligent Diagnostics and Predictive Maintenance System integrated within the Wind Turbine bus bar structure to aid Dynamic Maintenance Scheduling.
A Wind Turbine Unit was developed which measures a number of signals taken from the converter, and which carries out real time processing of the data to detect and identify faults. The results (consisting of raw or processed data and system logs) are packaged and sent to a Base Station (BS). The Base Station is the main information system. Its role is: to store the WTU data in a database; parameterise the WTU through web services and the BS; provide system control to the operator through a dashboard; and to provide an Internal intelligent algorithm to detect, identify and predict possible faults based on the WTU processed and raw data.
A first of its kind prototype system has been developed and subject to validation on a dedicated rig which enabled various electrical faults to be simulated. This was successfully able to detect the following:
• DC link capacitor degradation of ESR or Capacitance due to overvoltage stresses or ageing
• Degradation of the capacitor in the Grid output filter
• Generator-side and Grid-side IGBT degradation

Close to 50 dissemination meetings have taken place within the project, primarily targeting industrial wind turbine OEMs and operators, but also participating at a number of major events including the Wind Energy Exhibition in Hamburg with approximately 33,000 attendees

Project Context and Objectives:
Renewable energy (RE) sources have gained a great importance due to their inexhaustibility, sustainability, ecological awareness and supply of energy security. Among all RE sources, wind energy is currently viewed as one of the most significant fastest growing (at an average annual growth rate of more than 26% since 1990), commercially attractive source to generate electrical energy.

The vision of the wind industry in Europe is to increase wind’s fraction of electrical energy mix to more than 20% within the next 2 decades. To implement this, an average 10-15GW of additional capacity must be manufactured, delivered and implemented every year in Europe. In order to achieve this, further improvements in wind turbine technology are still needed. Wind turbines are not new concepts but still face challenges as stable and reliable sources of energy – issues with efficiency, operations, maintenance and its general costs.

There is a need to reduce the rate of electrical system faults and the corresponding downtime per fault which will contribute significantly to the overall reduction of the operational and maintenance cost associated with current and future wind turbines.

The WindTurBars project provides an On-line Intelligent Diagnostics and Predictive Maintenance Sytem integrated within the Wind Turbine bus bar structure to aid Dynamic Maintenance Scheduling. There are many such systems that monitor mechanical wear. The purpose behind WindTurBars is to look for degradation in the electrical power generation system of the turbine.

The objectives for were defined as:

Scientific objectives

• Definition of the degradation pattern of 2-3 known electrical faults and components. Note: Provisionally including electrical insulation in the coils & electrical connectors.
• Analysis of the evolution of these degradations by means of monitoring the high and low frequency harmonics of the electrical signals using power signal transducers.
• Identify the optimal number and position range of the power signal transducer sensors.
• Characterise 3 types of signal patterns that need to be measured and create a model that enables the characterisation of signal patterns on an ongoing basis for new fault patterns.

Technical objectives

• Select a power signal transducer sensor and adapt by integrating with a wireless sensor node and easily replaceable part of the wind turbine electrical system (such as bus-bars and wiring connectors) in order to be able to deploy the sensor in a wireless communication network within the wind turbine.
• Develop a central wireless communication hub for all the sensors per wind turbine; with a self-learning data acquisition and processing protocol algorithm for dealing with the large amounts of data and sieving through relevant and irrelevant data.
• Develop a software platform with an advanced signal processing algorithm, based on the signal pattern characterisation model (mentioned above), used to correlate the signal patterns to faults. The system will flag up patterns for known faults and store patterns for unknown faults as they occur to be able to flag up for future occurrence.
• Develop a Graphical User Interface (GUI) and Dashboard system for the wind turbine operators to be able to give them a full picture of the status of the wind turbines and possible maintenance scheduling plans based on the information generated by the monitoring system.

Integration Objectives

• Integrate power transducer sensors, wireless sensor nodes, and easily replaceable wind turbine components (such as bus-bars and wiring connectors) to produce a complete subassembly.
• Integrate all the developments into a robust kit with an adequate control system that can automate the monitoring process from the sensors of the wind turbine (possibly offshore) to the wind turbine operators’ station remotely.

Project Results:
During the first Reporting Period of the project, the consortium concentrated their efforts in defining the types of electrical faults that occur in wind turbines and developed desk-based models to enable the overall the Wind TurBars system to be defined, components selected to meet the project operational specification and key components purchased. Work was also carried out developing data processing approaches that would be needed for the Wind TurBar system.

Key project achievements were:

• Extensive literature review to determine the key electrical faults within wind turbines
• Investigation of the different wind turbine models and their key electrical components
• Data modelling of the normal operation of the selected wind turbine type
• Evaluation of initial failures to be characterised and simulated
• Simulation of key faults and introduction of these into the simulation
• Investigation of data processing methods and an assessment of their suitability for within the Wind TurBars system
• Initial economic evaluation of the target price for the final system based on lost following electrical failure
• Initial component selection

The major work to develop hardware, software, data processing, integration and validation was carried out in Reporting Period 2. A sensor array has been developed which is minimally intrusive and retro fittable into existing operational turbines. This targets the voltage and current waveforms at the bus bar sections of the Wind Turbine.

The faults identified for analysis are:

• DC link capacitor degradation of ESR or Capacitance due to overvoltage stresses or ageing.
• Degradation of the capacitor in the Grid output filter.
• Generator-side and Grid-side IGBT degradation and gate misfiring
• Electrical degradation of the generator – winding faults, reduction in permanent magnet field strength.

Parallel algorithms continually run on an intelligent Data Acquisition System at the wind turbine looking for degradation in these areas. These algorithms process the raw input data to detect degradations and transient instability, raising warnings or alarms at configurable thresholds. Having multiple thresholds allows tighter monitoring of the Turbine to understand the inherent system stability. Processing the data at the turbine significantly reduces the communications network loading, a block of data only being passed back for further expert analysis if an event threshold is exceeded. An additional scheduled summary of the system status is also returned to the base station for trending, modifying of event threshold levels and improvements to the algorithms. This scheduled reporting is configurable so that early life is monitored more closely than a mature turbine – the real time algorithms on the turbine are of course run continuously.

A breakdown of the Base Station and the Wind Turbine Unit functionality is attached.

Each Wind Turbine Unit monitors the detailed electrical parameters on the bus bars around the inverter circuit of the Wind Turbine looking for degradation in the electrical systems.

The sensors are connected to sbRIO DAQ via a Physical Interface (PHY) board, which also contains some wide band filtering and protection circuitry. The National Instruments RIO base design was chosen because it is a common design environment across the research consortium partners and the sbRIO is specifically used by one of the consortium partners for an existing Wind Turbine monitoring system based on mechanical sensing. This compatibility will give us a greater chance of placing equipment on a field trial and also gives an optional path forward by combining the technologies to give a product with enhanced capabilities. The sensors and DAQ input stages are scalable to be of use with Prototype models, test systems, and real Wind Turbines.

The expert base station can request a raw data dump at any time to assist with any investigations or background checking of the system. At the Base Station, these network messages are stripped out and reformatted to JSON format to allow a standard interface for the Base Station Expert application and data base functions.

Scheduled maintenance can be planned from this data and the configurable thresholds can be modified to monitor for further degradation whilst the maintenance is queued in case the degradation accelerates.

There is an additional local wireless link between neighbouring Wind Turbines connecting through the sbRIO board. This notifies neighbouring Wind Turbines of an event, so that they can look for matching irregularities in their electrical parameters and report this data if seen, to the base station for additional analysis to determine if the event has been caused by local transient event or a real degradation on a specific Wind Turbine.

Local environmental conditions can also be monitored if this data is not available at the Turbine to complement the raw data dump for expert analysis.

The consortium have successfully built, commissioned and validated the integrated Wind TurBar Condition Monitoring system. The final Intelligent Data Acquisition System (Wind Turbine Unit) was produced and is capable of capturing data to monitor the electrical signals for the degradation conditions specified in the first Reporting Period and communicating this to the ‘Base Station’. This incorporated the final sensor configuration and a sbRIO PCB with embedded software and algorithms. Main data communication was achieved via an Ethernet cable and the system also has the wireless capability to link a DAQ system with its neighbouring units. Advanced Signal Processing Software algorithms and database system has been produced which carried out expert analysis of the data and is capable of categorising faults and prompting action.

Due to difficulties in finding a suitable wind turbine site for field trial validation, a test rig was designed and manufactured by the consortium to fully validate the Wind TurBar system in an in-house test environment. Functional and performance tests were carried out on the Wind TurBar system to ensure it met the system specification.

Potential Impact:
The Wind TurBars technology will enable wind turbine operators to reduce down time of the wind turbine system and hence enable them to reduce their loss of earnings for electricity production (LE-EP). In addition the technology allows operators to minimise their operational expenditure (OPEX) for repairs in two ways:

• By enabling the operator to exchange components that start to fail early in order to avoid catastrophic and costly knock-on effects (CC-KE) of the total system.
• By enabling operators to conduct preventative maintenance as part of the routine maintenance and thereby prevent costly unscheduled repairs that may require the use of expensive maintenance equipment (EME) like cranes or helicopters.

The proposed value of the Wind TurBars technologies for operators is that for a small increase in the capital expenditure (CAPEX) for the purchase of the Wind TurBars technology (estimated system costs are €7,000) will generate a significant reduction of the OPEX by minimising LE-EP, reducing CC-KE and minimal use of EME. The consortium currently estimates that the Wind TurBars technology will enable operators to save about €13,000 per annum per offshore wind turbine. Thus the consortium estimates the return of investment (ROI) for wind turbine operators to be about 1 year.

The growth in wind powered electricity generation around the world has grown exponentially over the past 20 years. From a very low base in 1991, the installed capacity passed 10GW in 1998, it increased 10 fold in under 10 years, and in 2011 reached 237GW. Current estimates are that it will reach approximately 318GW at the end of 2013.

Within Europe, the EC has set itself a binding target that, by 2020, 20% of total energy consumption is to be generated from renewable sources. On this basis the European Commission’s Joint Research Centre estimates that member states will need to generate 35-40% of total electricity (3,200-3,500 TWh) from renewable energy sources by this time. The European Wind Energy Association (EWEA) estimates that this will require a total of 180 GW of installed wind capacity, including 35 GW offshore.

Total installed wind power capacity for EU27 passed the 100GW mark in 2012. Assuming an average turbine capacity of 2MW, this equates to around 50,000 turbines currently in operation in the EU. The total market size for new turbines is approximately 9000 per annum.

The US targets 20% wind-based electricity generation equating to over 300 GW by 2030. China aims for 15% renewable power generation by 2020. From these figures it is clear that wind is now an established component of the world’s electricity generating capacity supply, and will continue to increase in importance over the coming decades. Consequently, the issue of wind turbine reliability is also likely to become increasingly important over coming years.

When components in a wind turbine fail, the ripple effect can be staggering. Operators must deal with crane mobilization expenses (as high as €100,000 per incident), lost energy production, excessive costs per kW-hr, and untimely delays in obtaining replacement parts in a burgeoning industry where the demand for necessary components routinely outstrips supply. The costs from unplanned shutdowns and maintenance fixes can further be compounded by accessibility issues, particularly when the nacelle of a wind turbine is 100 m off the ground or situated offshore miles out at sea. Worker safety, too, is always a cause for concern.

It is estimated that costs of operation, maintenance and parts for a wind turbine with 20 year life is 10-15% of the total income. For offshore turbines the cost is estimated to be 20-25% of the total income. The inexactitude of the science of calculating O&M costs begins with the variety of ingredients that make up the whole in addition to scheduled and unscheduled maintenance. An analysis four years ago from Windstats, shows that the costs of O&M ranged from €15-26/MWh. The latest data from the International Energy Agency, reporting from 12 different countries, gives a similar range of €7-26/MWh.

Throughout the life of the project the consortium has monitored the commercially available condition monitoring landscape and has found none for detecting electrical system failures on wind turbines. Part of the reason for this is probably the fact that individual electrical system failures do not introduce long periods of downtime, and hence are not particularly costly to the operator. However, taken together, the downtime associated with these failures is clearly very significant. Moreover, these problems become more acute still when the turbines in question are located offshore. The commercialisation of the Wind TurBar will have a significant impact on the O&M of wind turbines.

To expand on our response to question 18 (page 17) in relation to increase in employment in small and medium-sized enterprises:

HVW – will seek follow-on funding and would require additional Business Development Staff (1-2 people) to promote the technology and manage this new business stream for HV Wooding. This is subject to securing necessary funds.

Ventech/ Mikrosay – Reactive increase in staff to meet supply chain needs including aftersales service. This is subject to achieving full market readiness (not viable yet) and subsequent commercial sales. Ventech may also require additional technical staff to work on the next phase of the project i.e. developing new performance features (1 person)

List of Websites:
The project website can be found at: