Vattenfall, one of the biggest offshore wind developers in Europe, wants to make fossil free living possible within one generation and is driving the transition to a sustainable energy system.
Georgios Leontaris, Senior Data Analyst at the Planning, Analysis and Resource Management team at Vattenfall supports all offshore wind asset and product projects. Georgios and his team are responsible for creating robust schedules that reflect the project strategy and provide guidance and support. Moreover they enable data-driven decisions for the installation of offshore wind assets in all phases of the projects from early development to construction.
Vattenfall wants to become the front-runner
Georgios explains that Vattenfall is one of the biggest offshore wind developers in Europe. However, as Vattenfall has the ambition to enable fossil free living within one generation, we should become the front-runner within a very competitive sector he says and continues:
-Therefore, we are constantly seeking ways to improve the current ways of doing our business and we are eager to explore state-of-the-art tools and methods with a strong focus on increasing the safety of our operations.
Driven by a complex industry in a challenging environment
Georgios describes the offshore operations and specifically installation and maintenance operations for offshore wind assets as complex and expensive operations which are taking place in a changing, uncertain and challenging environment.
-The use of AI (and machine learning more specifically) seemed promising to address these challenges by creating a wave spatial model which can enable better informed decisions for our operations and potentially improve their safety.
Variability of wave height
-In this proof of concept (PoC) we were looking for a model which could represent the variability of wave height that is one of the most important limiting metocean parameters for the vessels within the offshore wind farm site. As these sites are very large with different characteristics along these, it is often the case that there is difference in the wave heights between the different locations in the offshore wind farm site. This difference can be significant especially for certain operations for which the wave height limit for a safe operation is low.
The developed machine learning model takes as input measurement of wave height from the installed measuring devices at few locations and infers the wave height at different locations within the site where measuring equipment is not present, Georgios concludes.
This can reduce the cost of installing additional measuring equipment, increase safety of operations by identifying cases where conditions at location farther from the measuring device are worse and also avoid unnecessary trips which could have been initiated based on the measurements from the few installed devices.“
The PoC showed promising results
The expertise of Tenfifty in AI and data science appeared to be the right choice to support in the investigation of this innovative approach to optimize our operations and improve the safety which is the highest priority for Vattenfall.
-I also want to pay tribute to my colleague Dr. Robert Mawdsley, Senior Lead Metocean Analyst (Metocean team att Vattenfall) who contributed with knowledge and expertise that were crucial for a successful outcome of this PoC.
Georgios describes that further validation is needed to be used live during operations, but he is positive that this application has potential to add significant value to Vattenfalls’ business and increase safety of their operations and potentially reduce costs from unnecessary trips.
“The team at Tenfifty, consisting of very knowledgeable and skillful individuals, created a PoC solution using machine learning that showed the potential added value of utilizing this technology in our operations.”
The competence and expertise of Tenfifty met our already high expectations
-This tailored AI solution was built by Tenfifty according to our specific requirements. We were impressed by the ability of the team to quickly understand the complex process of offshore operation and Tenfifty’s problem solving ability based on a variety of different mathematical methods which they used to efficiently and effectively tackle problems that arose during modeling and validation. Thus, it was possible for them to create a useful model for us which can be used in real life.