The integration of renewable energy sources, especially offshore wind farms, poses a challenge to the stability of the electricity

grid. As the proportion of wind power grows, the variability and uncertainty of wind power generation leads to frequent grid

overloads, which require curtailment and result in significant energy losses. In addition, the increase in renewable generation

leads to higher intra-day and day-ahead market volatility. Both challenges can be addressed with storage, which allows for

economically optimized portfolio management and has the potential to alleviate congestion by acting as a buffer. However,

current control strategies often neglect the inherent uncertainties in both wind power forecasts and grid load dynamics.

This study tackles the critical issue of how to effectively manage battery storage systems in wind park-grid setups under uncer-

tain conditions to reduce energy losses caused by frequent grid overloads. We explore the potential of incorporating probabilistic

forecasts into battery management strategies, which may improve the mitigation of overload-induced energy losses compared to

conventional deterministic methods. In addition, our studies analyze current market effectiveness, concerning energy losses, and

how regulations need to be adapted under changing revenue scenarios.

Tested on real world wind park data, our findings suggest how probabilistic forecast-based optimization control methods

outperform traditional deterministic forecast and rule-based algorithms, particularly in scenarios with high variability and uncer-

tainty in wind energy generation and grid load. This finding indicates that accounting for forecast uncertainty not only improves

grid stability, but also enables more efficient use of wind energy, challenging the prevailing reliance on deterministic models in

battery dispatch, while allowing market participants to achieve greater financial benefits through probabilistic forecasting in the

future development of energy systems.

These findings have important implications for the design of future renewable energy systems and regulatory adaption, where

flexible storage operation will be critical to manage growing shares of variable generation. By providing a robust simulation

model validated with real-world data, this work contributes to the advancement of predictive control in energy systems, supporting

the broader goal of increasing renewable energy integration while maintaining grid reliability.