• Energy Research
• Mines ParisTech close

AbstractWe evaluate the impacts of DR in future electricity systems and to what extent they can facilitate the spread of high shares of renewable energy while maintaining a given level of reliability. We propose an approach based on long-term planning models, the MARKAL/TIMES models and our analysis is demonstrated by the case of Reunion Island, which aims to produce electricity using 100% renewable energy sources by 2030. We demonstrate that higher shares of intermittent sources weaken reliability of supply and can be counter balanced through demand response solutions.

AbstractIn order to address significant share of intermittency in the generation mix, a dynamic reliability constraint on kinetic energy was endogenized in the technical optimum TIMES model. Dedicated to La Réunion island: (i) The potential contribution of electrochemical storage technology, especially NaS, to the power dynamics and the reliability has been demonstrated; and (ii) A high share of variable renewable plants (around 50%) can be considered without jeopardize power transmission, provided appropriate investments in storage are made.

The reliability of power supply, defined as the ability to recover a steady-state condition after a sudden disturbance, is crucial for operating power systems. It is usually ensured by controlling voltage and frequency deviations and involves events occurring from a few milliseconds to a few hours. However, reliability requirements are largely ignored when dealing with long-term issues. To reconcile such contrasting timescales, it seems logical to rely on energy considerations based on thermodynamics. Two reliability indicators, assessing the magnetic and kinetic energy reserves of a power system, are derived from this approach. They enable to quantify the reliability of a given production mix and make it possible to choose between increasing shares of intermittent sources and maintaining an expected level of reliability. Since the indicators tackle reliability issues without focusing on a specific timescale, they are effective for both discussing the long-term evolution of reliability and improving the real-time management of a power system.