Register for a free trial
Offshore Wind Journal

Offshore Wind Journal

Drones, AI and IoT to make on-demand detection and monitoring a reality

Wed 07 Nov 2018 by David Foxwell

Drones, AI and IoT to make on-demand detection and monitoring a reality
Lab tests at Fraunhofer IWES provided convincing results prior to real-world tests on the Meerwind Süd and Meerwind Ost

Service companies and technology providers are planning to use drones, artificial intelligence, advanced sensors, wireless networks, the internet of things and smartphone apps to automate fault detection in turbines

Concepts for inspecting the rotor blades of offshore wind turbines using thermography and acoustic monitoring to detect potential issues with turbines before they become potentially damaging were presented by experts from Deutsche WindGuard and the University of Bremen, Bremen Institute for Metrology, WindMW Service, Fraunhofer IWES and the Bremer Institut für Messtechnik, Automatisierung und Qualitätswissenschaft (BIMAQ) in Germany at a recent conference.

They described a concept developed and tested on the Meerwind Süd and Meerwind Ost offshore windfarms off the coast of Helgoland in northern Germany designed to reduce the need to use rope access climbers to inspect turbines. “Their use either as independent systems or in combination with unmanned aerial vehicles (UAVs) will enable more efficient, safer monitoring,” they said in a poster presentation at the Global Wind Summit in Hamburg in September.

The objectives of a project they recently completed were to apply non-destructive testing and structural health monitoring techniques for enhanced rotor blade monitoring. Thermography was applied to detect potential structural issues; acoustics integrated into the turbine were used to detect potential problems within the structure.

The acoustic emission measuring system integrated in the rotor blade served as an early warning system by detecting internal damage, for example at the root of the rotor blade. The thermal imaging camera, on the other hand, detects surface damage, such as that caused by erosion due to rain. Acoustic emission and piezoelectric sensors were attached to the inner surface of the rotor blades in structurally relevant areas – especially at known weak points. A processor that collected and analysed the sensor data was integrated in the rotor hub.

Structural defects cause friction which in turn generates heat, and the heat flow in the material can be detected by thermal imaging. The project used a passive thermography technique, which detects differences in temperature. Deutsche WindGuard Engineering and BIMAQ have been applying thermal surveying to visualise heat flow in onshore windfarms for several years. The challenge in this particular project was to adapt this proven technique to the requirements of offshore installations.

The sensors work in much the same way as microphones. If the tensile forces in a specific area of the rotor blade suddenly change, the structure releases energy in the form of heat and surface waves that can be measured by the sensors. The sound waves captured by each sensor have different signal delays. By analysing their arrival times, it is possible to pinpoint the source of the damage. The acoustic measuring system delivered convincing results during lab tests at Fraunhofer IWES and was tested in a real-world scenario, along with the thermographic system, in early 2018 on Meerwind Süd and Meerwind Ost.

By attaching thermal imaging cameras to UAVs or ‘drones’ it is possible to detect subsurface defects in composite materials, including delamination, inclusions, faulty bonding in the loadbearing web-flange joints, and shrinkage cavities. Under operational loads, defects deep inside the rotor blade, if not detected and dealt with in good time, can lead to more serious structural damage and eventually lead to a total breakdown.

The UAV-compatible infrared system was tested against a high-end camera and found to be able to ‘visualise’ structures below the surface of the blade. The acoustic emissions system was able to evaluate large amounts of data in a relatively short timeframe while detecting potential damage and eliminating acoustic noise.

During static rotor blade and fatigue tests, the researchers identified damage including adhesive and inter-fibre fractures, damage to web-flange joints, cracks in the trailing edge of rotor blades, and faulty bonding in the blade root area.

Esvagt has teamed up with Wind Power Lab to offer windfarm owners an ‘on demand’ turbine monitoring service

Turbine monitoring using AI and drones

Without using innovative techniques such as the ones being pioneered in Germany, accessing offshore wind turbines in winter can be particularly difficult. To address this issue, Wind Power Lab, which specialises in using artificial intelligence to assess blades, and service operation vessel owner Esvagt, have teamed up to apply drone technology and artificial intelligence (AI) techniques to turbine monitoring.

Together they aim to develop a ‘service train’ concept to detect faults on offshore wind turbines using drones. Under this concept, windfarm owners could book blade data acquisition and automated blade defect assessments. “The process is that simple,” said Esvagt. “You book your desired number of turbines for a blade inspection and decide on a deadline for your blade defect assessment.”

Together, Esvagt and Wind Power Lab have formed a joint venture, EWPL Ocean, with Anders Røpke as chief executive. “EWPL Ocean will be able to deliver a unique product portfolio that will help offshore windfarm owners and operators into a more industrialised world,” said Mr Røpke.

“For an agreed, fixed fee, customers can have high quality images taken by drones and analysed with a best-in-class AI setup delivered by Wind Power LAB. It means that windfarm owners no longer need to worry about weather risks or logistics. Our AI technology can detect flaws as small as 1 mm wide. EWPL Ocean then provides a detailed report in a timely manner, allowing turbine owners to generate data-driven repair campaigns for the upcoming repair season.”

EWPL Ocean was launched in September 2018. Its blade assessment service is available this coming winter season, a time of year usually characterised by very limited offshore inspections due to the weather. The company summarises the advantages of the concept thus: no weather downtime when manned access is not possible; no access required to turbines; assessments of turbine condition ‘on demand’; service is available year-round.

iWindCr project

Avonwood Developments, with partners Avanti Communications and University of Portsmouth in the UK used the wind summit to present a poster about the intelligent real-time corrosion monitoring and detection on offshore wind turbines (iWindCr) project they are working on.

The project is co-funded by the Energy Catalyst and aims to develop cost-effective technology comprising sensors and advanced software plus end-user applications for detecting and monitoring corrosion and surface damage on wind turbines. The outcome of the project is expected to improve turbine lifespan, output and reliability by reducing the O&M costs from unplanned or unscheduled maintenance caused by corrosion and surface damage.

The real-time remote sensing technology they are developing is designed to provide autonomous detection and monitoring, providing exhaustive and detailed data on corrosion processes. At the heart of the iWindCr project is the concept of using technology that can measure changes in the electrochemical states – such as potential, current or resistivity – of components.

The iWindCr proof-of-concept system consists of a wireless sensor network of smart miniaturised sensors fixed to internal or external turbine structures. It utilises the concept of the internet of things to integrate the wireless network with satellite and terrestrial communications networks, providing a guaranteed IP for data backhaul from remote sites to a control room. Power for the system would be provided by solar panels.

The system will use a cloud-based solution for predictive analyses of corrosion/surface damage trends and notify authorised stakeholders through web and smartphone apps.

Recent whitepapers

Related articles





Knowledge bank

View all