Optimization Days 2014
Including an Industrial Optimization Day
HEC Montréal, May 5 - 7, 2014
JOPT2014
HEC Montréal, 5 — 7 May 2014
 
      MD6 Planification de la production d'électricité 2 / Electricity Production Planning 2
May 5, 2014 03:30 PM – 05:10 PM
Location: Nancy et Michel-Gaucher
Chaired by Pascal Côté
4 Presentations
- 
                 03:30 PM - 03:55 PM 03:30 PM - 03:55 PMShort-Term Hedging for an Electricity RetailerA dynamic global hedging procedure making use of futures contracts is developed for retailers of the Nord Pool electricity market. Statistical models are developed for the electricity load, the day-ahead spot price and futures prices. Backtests show that the proposed procedure provides considerable risk reduction when compared to hedging benchmarks. 
- 
                 03:55 PM - 04:20 PM 03:55 PM - 04:20 PMThe Impact of Complexity in Stochastic Dynamic Programming Model for Hydropower OptimizationIn the field of reservoirs management, a very popular way to solve the optimisation problem is to use stochastic dynamic programming with a lag-one model to represent the inflows. We show in a case study on Kemano system, managed by Rio Tinto Alcan and located in B.C., how increasing modelisation complexity of inflows and constraints can lead to dramatic performance gains. 
- 
                 04:20 PM - 04:45 PM 04:20 PM - 04:45 PMTwo Approaches to Aggregate Smart Grid’s Energy Systems’ Production PlanIn this paper, we propose a mathematical programming formulation and a Markov chain Monte Carlo approach in order to compute the production plan of a smart grid by aggregating its energy systems’ production plans and considering the physical constraints of the grid. 
- 
                 04:45 PM - 05:10 PM 04:45 PM - 05:10 PMAnalyse du comportement de méthodes d'optimisation pour le calage efficace de modèles hydrologiques coûteux en temps de calculLa recherche porte sur l'analyse du comportement de trois algorithmes d'optimisation (SCE-UA, DDS et MADS) lorsqu'employés pour le calage d'HYDROTEL, un modèle hydrologique distribué et à base physique coûteux en temps de calcul. L'étude présentera également la configuration adéquate de MADS quant au calage d'HYDROTEL et l'impact du type de modèle hydrologique sur le comportement des méthodes d'optimisation sélectionnées. Les résultats et conclusions de l'étude seront présentés. 
