Dispa-SET 2.0: unit commitment and power dispatch model
Details
- Publication date
- 1 January 2015
- Author
- Joint Research Centre
Description
Most analyses of the future European energy system conclude that in order to achieve energy and climate change policy goals it will be necessary to ramp up the use of renewable energy sources.The stochastic nature of those energies, together with other sources of short- and long-term uncertainty, already have significant impacts in current energy systems operation and planning, and it is expected that future energy systems will be forced to become increasingly flexible in order to cope with these challenges. Therefore, policy makers need to consider issues such as the effects of intermittent energy sources on the reliability and adequacy of the energy system, the impacts of rules governing the curtailment or storage of energy, or how much backup dispatchable capacity may be required to guarantee that energy demand is safely met.Many of these questions are typically addressed by detailed models of the electric power sector with a high level of technological and temporal resolution.
This report describes one of such models developed by the JRC's Institute for Energy and Transport: Dispa-SET 2.0, a unit commitment and dispatch model of the European power system. The new model is an updated version of Dispa-SET 1.0, in use at the JRC since 2009. The aim of this new version is to represent with a high level of detail the short-term operation of large-scale power systems. To that purpose the model considers that the system is managed by a central operator with full information on the technical and economic data of the power plants, the demand, and the transmission network.
The model is formulated as a tight and compact mixed-integer program, implemented in GAMS and solved with CPLEX.This report describes the formulation of the model and explains how it is implemented in GAMS. To illustrate its capabilities, a simulation has been run using historical data for Belgium. The comparison between historical data and simulation outputs indicates a fairly good agreement. The report concludes with the main conclusions of this work and presents forthcoming improvements to the model.