Mercier, Thomas
[UCL]
Jomaux, Julien
[UCL]
De Jaeger, Emmanuel
[UCL]
This paper introduces a novel modelling of the economics of energy storages providing frequency containment reserves. The focus lies both on the practical operation of day-ahead markets and the stochastic and autocorrelation characteristics of the grid frequency as they are practical concerns for any energy storage operator. Based on yearly series of grid-frequency measurements we put in light seasonal trends in the grid frequency. After having removed these trends we build an ARMA-GARCH model with fat-tail feature to explain the remaining autocorrelation. The overall model can be used to generate scenarios of 15-minute time-step frequency deviations and we explain how these scenarios can be used along with stochastic programming to dispatch and compute the revenues of a storage providing frequency containment reserves. By considering two different cases, battery storage and variable-speed pumped-storage, the paper emphasizes the importance of the technical characteristics of energy storages providing frequency containment reserves.
Bibliographic reference |
Mercier, Thomas ; Jomaux, Julien ; De Jaeger, Emmanuel. Stochastic programming for valuing energy storage providing primary frequency control.2016 IEEE International Energy Conference (ENERGYCON) (Leuven, Belgium, du 04/04/2016 au 08/04/2016). |
Permanent URL |
http://hdl.handle.net/2078.1/175826 |