ROMEO helps reduce LCOE thanks to reliable O&M support tools
ROMEO (Reliable OM decision tools and strategies for high LCoE reduction on offshore wind) is a European collaborative project that gathered 12 partners from 6 countries. From 2017 to 2022, they investigated on innovations and technologies allowing to drive down the costs of operation and maintenance (O&M) of offshore windfarms.
Currently O&M accounts for up to 30% of the overall cost of energy. This figure is subject to raise with the development of bigger farms in increasingly technical environments, further away from shore, and in deeper water involving the use of floating foundations. To maintain control of the costs, it is essential to be able to anticipate maintenance actions, plan them in advance, and allocate resources optimally.
In ROMEO, algorithms have been developed to predict failures of wind turbine components based on measurements from Supervisory Control And Data Acquisition (SCADA) systems and expert knowledge regarding the failure modes involved.
Indeed, knowing when a specific component is going to fail pave the way to an optimal allocation of the resources available. Specifically, spare parts can be procured ahead of the failure, reducing the need for important and expensive stocks and warehouse management costs associated. Similarly, logistic and maintenance work can be planned, and necessary vessels or contractors can be booked in advance. This increased predictability of operations helps to limit downtime and subsequent energy production loss. It also keeps rush costs down. Moreover, by acting proactively, we increase our chance of fixing issues at an early stage, thus avoiding more severe damage to the equipment that would call for heavier and more expensive maintenance.
As part of the project, an assessment of the technical and economic performance expected from the implementation of such condition-based maintenance policy has also been conducted. Although the result of this assessment is not completely satisfactory for all the algorithms developed, the overall conclusion is that the approach gives encouraging results. It is then worth improving further the data management, the algorithms, and our capabilities to evaluate the economic performance associated with their use.
As the project is ending, some learnings can be drawn. First, trying to anticipate events requires beyond sophisticated algorithms, a sturdy base of knowledge. That base of knowledge needs to be continuously reinforced and must take the form of reliable data (technical and economic) on one hand and sharp technical expertise on the other hand. The combination of these two assets allows for a deeper understanding of the failure mechanisms and symptoms associated. Secondly, one must take advantage of that stronger base of knowledge to develop top performing algorithms. Finally, maintenance policies need to integrate the use of predictive tools from design stage. This demands a good collaboration between different teams including experts at R&D. The UK Center Renewable Team is already discussing the next steps with EDF-Renewables. R&D will continue to help operative teams to leverage on the increasing availability of data and develop state-of-the-art processing capabilities.