4 Presentations

• 
  03:30 PM - 03:55 PM

  La prévision de la demande pour la gestion du réseau électrique québécois: nouveaux enjeux et défis

  • Rachel Bazile, presenter, Contrôle des Mouvements d'Énergie de la division TransÉnergie d'Hydro Québec
  • Olivier Milon, TransEnergie

  Après un bref survol de l’approche paramétrique utilisée pour la prévision de la demande chez Hydro-Québec, cette présentation traitera de différentes avenues de développement actuellement à l’étude pour garantir des prévisions de qualité dans un contexte de transition énergétique et de contraintes de gestion plus importantes.
• 
  03:55 PM - 04:20 PM

  Profitability for power system planning

  • Cheng Guo, presenter, University of Toronto
  • Merve Bodur, University of Toronto

  Power systems capacity expansion models have traditionally taken a centralized planner’s perspective to find cost-optimal generation capacity to reliably meet load. Unfortunately, such models do not ensure individual generators are adequately remunerated. We present an expansion model that determines optimal generation/storage capacity investment decisions, while ensuring individual units achieve profitability.

  Keywords: power system economics, power system planning, complementary modeling
• 
  04:20 PM - 04:45 PM

  Heating network optimal dimensioning

  • Violette Berge, presenter, Artelys Canada Inc.

  PLANHEAT is a collaborative research project aiming at developing an integrated simulation tool for heating and cooling systems. The final goal is to support local authorities in the selection, simulation and comparison of alternative low carbon and economically sustainable scenarios for district heating and cooling. Within this project, Artelys developed an algorithm to dimension a heating or cooling network, taking into account the optimal route given all the city’s structural constraints. The techno-economic optimization considers both the investment cost and the operations cost.
• 
  04:45 PM - 05:10 PM

  Intelligent decision-making algorithm for energy storage systems.

  • Ysaël Desage, presenter, DIRO, Université de Montréal
  • Fabian Bastin, DIRO, Université de Montréal
  • François Bouffard, Department of Electrical and Computer Engineering, McGill University

  As the diversity and efficiency of energy resources continues to grow, the energy storage on a medium and large scale is one of the major challenges in today's energy sector. With this in mind, the present project aims to implement an intelligent decision-making policy based on dynamic stochastic programming and deep reinforcement learning. The goal is to optimize the behaviour of any agent with energy storage capacity, regardless of its nature, operating objective or technology. The changes resulting from the optimisation of the behaviour of all these agents will allow improvement in certain large-scale problems related to the production and distribution of energy on the electricity grid, a decrease in the individual bill for consumers, and also a direct benefit for the environment and sustainable development.

  Energy Storage; Smart Grid; Battery; Automated Learning; Reinforcement Learning; Stochastic Dynamic Programming.