Long duration energy storage is an enabling technology to unlock the full potential of wind energy generation to support the green transition. Firming wind and solar energy enables enhanced value and revenue generation in electricity markets and unlocks new applications for wind-to-x including hydrogen and e-fuel production or serving fast-growing demand for data centers. Sperra has been developing a deep subsea pumped storage hydrogen (SPSH) system through a U.S. Department of energy grant where individual units of 2 MW with duration of up to 14 hours or more can be deployed and co-located with deep floating offshore wind (FOW) - in depths of around 1000 m or more. In this case study based on California met-ocean conditions and markets, the combined operation of FOW-SPSH system is modeled with DTU's HyDesign software to optimize the overall system design and operation for different offtake pathways including conventional market participation as well as a novel use case for powering an offshore floating data center. The integration of offshore floating solar energy is considered in both cases along with the potential value of battery storage integration as well. Furthermore, the analysis for data centers explores both grid-connected as well as off-grid scenarios. Different constraints are applied to represent the operating/availability requirements of data centers. Sensitivity analysis is applied to key cost and performance parameters to understand the impact of uncertainties on the optimal system design and its economic performance. The results demonstrate the technical feasibility of the FOW-SPSH system to meet data center operational requirements but at a cost premium - though integration of solar energy and battery technologies can alleviate this premium and reduce the requirements for overprinting of the wind capacity. Operating in an electricity market context, the FOW-SPSH integrated system can improve the economics substantially - particularly when considering participation in ancillary service markets.

With a more volatile and flexible energy system, the operation of wind turbines and wind farms will need to adjust with more flexible operation. On the one hand, down-regulation of wind farms is sometimes required by grid operators when excess power is in the grid. On the other hand, operators can react to spot-market prices by de- or uprating of turbines depending on the current price. When applying such strategies, the influence on the turbine’s lifetime needs to be considered for an optimal operation of the farm in the long-term.
The optimization process for such problems, surrogate models for wind turbine damage equivalent loads (DELs) are used, created from aero-elastic simulation. With an identical setup of underlying models, different optimization approaches for finding operational plans can be used. A trade-off between maximizing the income form power production and fatigue damage of different components needs to be found. In previous studies, different approaches have been applied for different use cases (e.g. gradient-based optimizers with continuous setpoints or combinatorial methods for discrete values for each condition). The methods differ in complexity, computational effort and accuracy. For example, a continuous problem formulation with freely variable power setpoints has the advantage of higher flexibility in the choice of setpoint, which might lead to better optimization outcomes. However, using a number of discrete power setpoints can enable the inclusion of complete turbine shutdown into the optimization, which is necessarily discontinuous. Additionally, the differences in computational effort are unclear and need to be investigated. In most of the studies, simplifications and assumptions have to be made, in order to solve the problem with the available resources. Found solutions are optimal under those assumptions but a comparison or verification is often difficult. Nevertheless, they all clearly indicate an improved performance of turbine operation at different optimal trade-offs on a Pareto front. A reduction of lifetime fatigue damage by about 10% is possible while revenue is kept constant compared to nominal operation. Even higher reduction in fatigue damage - resulting in longer turbine lifetimes - can be achieved when accepting small losses in energy production or revenue.
Within this work, we optimize the operational plan of a wind turbine with different optimization approaches . We setup a problem which can be solved with both, combinatorial and continuous optimization. We compare the results in terms of dedicated KPIs such as computational effort, objective function and constraint handling. Advantages and disadvantages for each of the approaches are discussed and the potential for further improvement is investigated.

Much attention has been given to the need for energy-storage solutions to handle so-called Dunkelflaute events - periods with significant shortage in renewable power output. These events of energy droughts are one of the major criticisms against electricity systems primarily based on solar and wind resources. To reduce the high, short and long-term storage capacities needed for handling the longest and most intensive periods with energy droughts, a complementary measure is to activate large shares flexible loads. In this work, we analyze process flexibility of energy-intensive industries as a demand-side measure to cope with Dunkelflaute events. We demonstrate how the inherent flexibility characteristics of different type of industries can be utilized through coordinated scheduling to enable sizable load flexibility to the grid. For a case study of offshore wind in the North Sea and industries located in southern Norway, we quantify the reduction in needed storage capacities for handling Dunkelflaute events through activation of industrial process flexibility.