Offshore wind turbines face higher wind speeds than onshore turbines and face strong ocean currents, requiring more robust designs and significantly higher capital costs. While they generate more energy due to stronger winds, these increased costs result in a higher levelized cost of energy (LCOE). The HIPERWIND project has developed new design simulation models that reduce the LCOE by up to 9%, thus making offshore wind turbine construction and operation more cost-effective and reliable.
This year, wind energy reached a cumulative installation capacity of 1TW. The capacity is expected to grow up to 10TW by 2050. On this scale, reducing costs by 9% is monumental.
Hiperwind Logo“HIPERWIND set out to achieve a significant reduction in the LCOE by understanding how to deal with uncertainties in the wind turbine design modelling chain,” says Project Coordinator Nikolay Dimitrov from DTU Wind. “We examined how to quantify and identify various uncertainties, ranging from environmental conditions to loads and wind turbine reliability. With this information, we focused on reducing material use by better understanding model performance and reducing uncertainty. This approach helped minimize material use and lower energy costs. This methodology has demonstrated the feasibility of designing more efficient systems.”
At the core of HIPERWIND is managing uncertainties. Uncertainties translate into higher safety margins, adding materials to components, shorter maintenance cycles, and increases in the cost of financing wind farms. Uncertainty management is consequently a driver in reducing costs and risk – thereby improving the production reliability, and ultimately, the value of offshore wind.
Game changer
“HIPERWIND could be a game changer,” Clément Jacquet from EPRI Europe says. “We delivered a significant reduction of the LCOE of up to 9% – and even 10% is achievable if we consider the most optimistic case we have. In the least optimistic case, the reduction will still be 5%.” EPRI assessed the impact of HIPERWIND technologies on LCOE, requiring both a holistic approach and detailed analysis of offshore wind farm costs. This work resulted in a new, adaptable framework that EPRI will use in future projects to improve the economic efficiency of both onshore and offshore wind farms.
The project used a real-world case study involving the Teesside offshore wind farm off the coast of England, owned by project partner EDF. Data and models specific to the wind farm were used to identify and quantify uncertainties in turbine tower and foundation design. The team then assessed whether the improved knowledge could reduce costs if the wind farm were rebuilt.
HIPERWIND thereby demonstrated that using less material in turbine construction can reduce upfront costs (capital expenditure), which make up about 30% of the overall cost of energy. Additional cost reductions were achieved by scheduling maintenance during low energy price periods, boosting both cost savings and operational efficiency.
Exploitation
Leveraging the measured data and advanced physics-based and data-driven models, this uncertainty management and reduction philosophy was applied throughout the offshore wind turbine design modelling chain – and beyond.
IFP Energies Nouvelles (IFPEN) is also already applying HIPERWIND results, improving chain modelling by accurately quantifying wind turbine fatigue loads. “The project has produced some significant reliability design procedures that are market-ready and thereby go beyond the research domain,” explains Martin Guiton from IFPEN. “Taking uncertainties into account, we obtain a reduction of 21% of the mass of the wind turbine structure, which is a lot.”
Likewise, ETH Zurich is now using these methodologies not just to solve wind-related problems but also earthquake-related problems, such as the seismic fragility of buildings in complex environments and the design of high-rise buildings under random wind excitation.
“The project required us to develop a new methodology from scratch to handle uncertainties in high-dimensional inputs and responses,” says Senior Scientist Stefano Marelli, Chair of Risk, Safety and Uncertainty Quantification at ETH Zürich. “Our work on surrogate modelling techniques, which accelerated algorithm development and enabled cross-partner collaboration, proved to be successful.”
