Dynamic dimensioning of the FRR needs

Study on dynamic dimensioning of the FRR needs (October 2017)

In this study, Elia presents a new method to ‘dynamically’ size the balancing reserves needs close to real-time based on day-ahead predicted system conditions, including offshore and onshore wind power, solar photovoltaics, electricity demand and schedules of power plants and transmission assets.

In contrast, the current ‘static’ method dimensions the balancing reserves needs year-ahead based on historic system imbalances and renewable prediction errors, after which these needs are fixed for the entire year. In such a method, extreme events - which are expected to occur more frequently in the future, but only during exceptional conditions - are setting the reserve needs for the entire year. This results in oversizing during lower risk periods, while lacking capacity in the periods with real high imbalance risks.

In the study, it is demonstrated that the proposed alternative methodology is improving reliability and cost efficiency of balancing generation an off-take, particularly in future systems with increasing renewable generation. Advanced statistic solutions, including machine learning techniques, are used to determine the total needs for these balancing reserves, i.e. FRR needs, based on an analysis of historic system imbalances and the corresponding predicted system conditions.

This study investigates the potential of such dynamic dimensioning approach and contains two parts. The first part contains the results of an Analysis (including a Cost and Benefit Analysis) of six potential methods for the dynamic sizing of FRR needs. The objective of this first part is to give recommendations towards a set of methodologies which need to be further analysed in a proof of concept.

The second part contains the final selection of the most promising methods for implementation, accompanied by an implementation plan. Besides the proof of concept, this part also includes a financial implementation impact assessment. The report concludes that the implementation of a dynamic sizing method would result in:

  • a better reliability management and robust behaviour in the future electricity system;
  • a positive business case following average balancing reserve needs reductions ;
  • an enduring solution compared to static sizing towards the future;
  • and a technical feasibility and transparency of the results.

The implementation trajectory for an operational tool is foreseen at 9 to 12 months. The effective application of dynamic dimensioning is subject to a follow-up study on dynamic (‘daily’) procurement of mFRR (manual Frequency Restoration Reserves, also known as tertiary reserves) in 2018.