By Philippa Jones

There is no shortage of ideas of how to improve component reliability and condition monitoring and some of the best ones were shared at the EWEA 2013 Annual Event. “Lots of companies are making kit, but there is not enough practical real-world validation examples,” he says. “Validation is the key to having compelling evidence that proves the technology is worth investment and not just an ancillary device”. Moreover, “companies need to know they will get a positive return on their investment,” he says. This is particularly important in the current economic climate where companies are looking to produce the best products as cheaply as possible.

Yves Van Ingelgem from the Vrije Universiteit Brussel, Belgium argued the importance of continually monitoring wind turbine foundations for this very reason.

“Early warning of potential corrosion problems will help reduce O&M costs related to offshore wind turbines,” he says. Denja Lekou from the Centre for Renewable Energy Sources and Saving in Greece argues a similar case for using acoustic monitoring for turbine blades. “If the methodology is good, the cost of acoustic emissions sensors and any other related equipment will be offset by the ability to have constant, detailed monitoring of the state of the blades and removing the need for on-site inspections every six to 12 months,” she states.

“At the moment only by stopping a wind turbine and inspecting the blades is it possible to check whether one has developed a crack during the operation of the turbine,” explains Lekou. “By using the acoustic emission technique, the blades can be checked for damage while they are turning.” Acoustic emissions operate at high frequency like ultrasound. “Sensors in the blades detect stress waves that are emitted when, for example, a crack occurs or propagates in the blade,” says Lekou. “Because this monitoring is constant, we can pick up on any cracks much quicker than would be possible with a traditional inspection. Moreover, the acoustic emissions technology will capture not just any local damage, but also monitor larger areas of the blade.”

Balancing the budgets

Rather than focusing on saving money, Steven Buckley from the UK office of consultancy Sinclair Knight Merz, wants to make it easier for companies to see what O&M options are available and how they can budget for some of the associated costs. Likewise, his firm wants to help businesses examine what the impact on energy yield might be by applying statistical techniques based on Mean Time Between Failure rates and Mean Time to Repair rates. “We have produced such analyses for lenders in the past at the development stage and used it in preparing financial models as part of due diligence projects,” says Buckley. “We have typically provided an annual cost estimate which can be used to help build up a ‘maintenance reserve account’ and we have provided ‘sculpted’ availability curves to better reflect the impact of component failures on availability.”

He admits that this is far from being a simple exercise. “The main problem in adopting this approach is differences between O&M contracts, the different technologies and different sites,” states Buckley. “It is a prediction exercise based on a statistical analysis of historic data, but not all turbines are the same and not all sites are the same, so what works for one site isn’t necessarily right for another site.” However, he insists that his company’s approach is “an improvement on using general rules of thumb for O&M costs and availability.” Furthermore, “some elements of the analysis can be adjusted to take into account the differences in technologies and sites and, in the case of operational sites, known failure rates that differ from the norm.” 

According to Buckley, “this approach is viable for all companies, can be used at any stage of development/operation, for projects of different sizes and in different countries.”  He is also confident that it “should be possible” to apply this model offshore wind farms. “I am presently discussing this with my O&M colleagues, who have provided maintenance on an offshore wind farm for many years, to discuss what changes to parameters might be appropriate,” he explains.

Operational offshore

The importance of keeping a close eye on what is happening on offshore turbines is the focus of the paper to be presented by Paul Faulkner, technical sales manager at Strainstall Monitoring. He discussed the design and capability of Structural Health Monitoring Systems (SHMS) on offshore wind turbines together with the management of the tasks and risks in offshore environments. “The application of SHMS in difficult environments is a particularly challenging task,” says Faulkner. In such environments, “ease of installation, ruggedness and reliability of equipment is essential in providing key information about the structural integrity of turbines which is required to evaluate the structural response, status and remaining operational life of the structure”. Faulkner insists that “the installation and commissioning of such systems have a significant focus on safety and access to the structures where onsite retrofitting of sensors and instrumentation are carried out in the field.”

He gives the example of how theory has been put into action by Strainstall. “In April 2010, engineers reported a fundamental flaw in the design of offshore wind turbine foundation structures,” says Faulkner. “The problem affects wind farms across Europe, requiring further investigation into the potential effects on all offshore turbines that have monopole foundations of this design.” He explains that “the grout designed to act as an adhesive between the pile foundation and the transition section of the turbine tower base was failing, and had caused some turbines to slip by up to 25mm. The majority of the, then, 366 UK turbines were built to the same specification and so could potentially develop this fault.”

Supporting brackets used during the construction phase and then subsequently left in place have stopped the turbine towers from slipping further and consequently no safety issues have since been identified. Nonetheless, more detailed analysis of the problem was required. Monitoring systems were therefore installed “as retrofit packages on the monopile towers… this moved on to structural health monitoring systems that were fitted at the onshore construction bases then finally commissioned offshore,” explains Faulkner.

The main challenges in the installation and commissioning of such systems are the environment – weather conditions and accessibility – in which the work has to be carried out and the need to provide a safe environment for engineers. Changing weather conditions can have a significant effect on the ability to stick to an installation programme and companies often have to be extremely flexible in their planning. This can also have an obvious impact on costs, which are also affected not only by the work required but also by issues such as the location of installation and commissioning and whether any work can be carried out onshore before the turbines are installed offshore, for example. The numbers of turbines monitored in a package can also affect the cost. Estimated costs can run from £50-150 000 per turbine, dependent upon complexity, according to Faulkner.

Data sharing

One way of potentially limiting costs would be through the development of a common database, an approach extolled by Stefan Faulstich from the German research institute Fraunhofer IWES. “Extensive national research projects dedicated to reliability analyses on wind turbine failures have been performed during the last years in several countries, including Denmark, Finland, Germany, the UK, the Netherlands, the US and Sweden,” he explains. “There is agreement to a certain degree among the projects on wind turbine failure statistics. However, the same components can react differently in different turbines, for example, because of different technical concepts in use or different site characteristics, while the use of identical and similar components from different manufacturers will lead to different lifetime expectations of the components and thus to a spread of results,” says Faulstich. He believes therefore that “a common database of several or even all wind farm operators is essential to broaden the statistical basis.”

Faulstich accepts that “databases are, of course, only as good as the quality of data they contain.” Hence, “standardisation is vital to ensuring that the information is fit for purpose”. For this reason, the International Energy Agency (IEA) created a taskforce in November aimed at agreeing international standards for reliability data. “The purpose is to bring together the present actors in the industry and research community to create synergies and agreements in the many R&D activities already on-going in the field of statistical failure analysis,” says Faulstich.

Building a useful database is a complicated exercise. “Due to the large amount of parameters influencing reliability behaviour, a large data pool is needed for appropriate analyses,” says Faulstich. “A large database offers the possibility to clearly identify weak points and to make statements on the failure probability of certain components meaningful,” he adds. “It is important to make use of experience by acquiring a lot of information at different locations. This is possible only through semi-automated and highly simplified data management.” This will be the main focus going forward. “There is a need to gain more parameters, data and additional information compared with today, making electronically supported reporting by service teams necessary. The need to gather more and especially more detailed data, while reducing maintenance efforts, should be the long-term goal.”

There are clearly many challenges ahead, but different solutions and approaches do exist.
 
The EWEA 2013 session on ‘Condition Monitoring and Operations’ took place on Wednesday 6 February. See http://www.ewea.org/annual2013/conference/conference-proceedings/ to order proceedings from the session.

Read the current issue of Wind Directions.