Presentations

Abstract

Recent trends in offshore wind have revealed persistent economic challenges in the North Sea basin, but also political commitment to drive its development. Some key challenges facing offshore wind are grid congestion and electricity price erosion. In Germany for instance, grid curtailment amounted to 18% for offshore wind in 2024 and the capture rates are regularly displaying monthly values below 90%. Further offshore wind and inland solar deployments will exacerbate those structural challenges and affect the cost-efficiency of our future energy system. In that context, energy storage presents a promising solution by mitigating the mismatch between supply- demand or supply-grid availability.  The present study investigates the potential of co-locating storage with an offshore wind plant. It forms the first output of the OESTER project, a three-year initiative involving major energy industry players such as RWE, Vattenfall, and TNO. In this study, we present results of the initial sizing, conducted by DMEC, of the offshore storage system, that encompasses the different storage technologies considered in the project. Our focus is on the system-level value that storage assets can bring to co-located wind farms. The following specifications are assumed for a baseline wind plant: a planted capacity of 1045 MW and a 1 GW HVDC export infrastructure, without overloading allowed. Wind production is simulated by Vattenfall and accounts for wake losses and wind turbine control. Netherlands electricity price and system-level metrics are obtained from Montel and used for a target year 2035, that emulates 2019 weather data and assumes given deployments in renewables.  A specific care has been given in simulating various types of curtailment: production-side curtailment (avian, operation and maintenance) is modelled statistically using industry availability figures;  export-side curtailment, related to grid congestion, is modelled by relating it to the residual load, following an approach derived by TNO and implemented for our study case. Finally, it is assumed that the baseline wind plant does not export energy during episodes of negative day-ahead prices, which are accounted for in the Montel dataset.  Then, we identify two sizing use cases relevant system-level metrics that should be improved by the addition of an energy storage system. Two sizing use-cases are identified: the first should have energy storage increase the Annual Energy Exported (AEEx). The second is to improve the alignment between offshore power export and system demand. Day ahead prices are identified as a relevant metric to weight export and are used to formulate a Value-Weighted Energy (VWE) metric. For those two use cases, we want to find the minimal storage CAPEX required to enable an improvement in the associated metric, respectively VWE and AEEx, over the baseline wind plant. That sizing is conducted by parametrizing and solving a non-linear constrained optimization problem. A forward Euler numerical approach is chosen to represent the quasi-steady evolution in storage state variables. The non-linear control problem is solved, over a whole year with a time discretization of one hour, with the optimization toolbox CasADi in combination with the solver IPOPT.