Operational strategies for predictive dispatch of control reserves in view of stochastic generation

In view of the predictability and stochasticity of wind power generation, transmission system operators (TSOs) can benefit from predictive dispatch of slow and manual control reserves in order to maintain reactive reserve levels for unpredictable events. While scenario-based approaches for stochastic optimization are well suited for this problem, it appears that TSOs are hesitant in adopting this method into their practice of predictive dispatch. Differences in the formulation of constraints and cost functions, the timing and reserve product constraints influence the dispatch result significantly and yield varying results with different practical implications. To support adoption, there is a need to study relevant parameters and trade-offs to be considered in introducing such methods to operation practice, enabling also the investigation of alternate reserve product constraints, e.g., to enable reserve contribution from storage-constrained units. This paper introduces a framework for comparison of operational strategies for system balancing, proposes criteria for performance assessment and exemplifies a systematic evaluation of several operation strategies.