4 Presentations

• 
  10:30 AM - 10:55 AM

  Modified and Augmented Nodal Analysis-based Optimal Power Flow

  • Abraham K. N'Zi, presenter, Department of Electrical Engineering, Polytechnique Montréal, Mila & GERAD
  • Nasim Rashidirad, Hydro-Québec
  • Jean Mahseredjian, Department of Electrical Engineering, Polytechnique Montréal
  • Antoine Lesage-Landry, Department of Electrical Engineering, Polytechnique Montréal, Mila & GERAD, QC, Canada

  The optimal power flow (OPF) method focuses on the operational efficiency of electric power systems. It aims to optimally dispatch generation to minimize, e.g., production costs or losses, while ensuring that system constraints and operational limits are met. Modern grids include a variety of devices that are often omitted in traditional OPF formulations, namely, different types of generators, devices such as transformers or power electronics-based devices, because it can be difficult to model them efficiently while using the conventional mismatch equations employed in most traditional OPF problems. The modified and augmented nodal analysis (MANA) extends the traditional nodal analysis by expressing circuit equations in a generic sparse matrix representation. The MANA approach can seamlessly accommodate constraints from various types of devices commonly encountered in modern networks. This, in turns, allows their straightforward inclusion as constraints in the OPF, resulting in a more accurate representation of modern power grids. This work proposes a new OPF formulation based on MANA and the corresponding power flow constraints, which we refer to as MANA-OPF. The non-convex MANA-OPF is solved using a nonlinear solver, namely, IPOPT in Julia. The formulation is tested on a 3-bus case, the IEEE 14, 30 and 118 bus, and the 2736sp case from the PowerModels library. The results are compared to a standard approach used by PowerModels. We observe an average gap observed of 0.68% between the proposed approach and the nonlinear program’s solution, thus attesting their high similarity in terms of optima. MANA-OPF offers a significant advantage in modelling devices that are challenging to integrate into standard OPF formulations, such as transformers, switches, breakers, and voltage sources. Its current-voltage formulation enables the straightforward and efficient inclusion of these components by directly using their current and voltage equations.

• 
  10:55 AM - 11:20 AM

  Task Mapping Strategies for Electric Power System Simulations on Heterogeneous Clusters

  • Julie Durette, presenter, Department of Electrical Engineering, Polytechnique Montréal, Mila & GERAD, QC, Canada
  • Gunes Karabulut Kurt, Poly-Grames Research Center, Department of Electrical Engineering Polytechnique Montréal, QC, Canada
  • Antoine Lesage-Landry, Department of Electrical Engineering, Polytechnique Montréal, Mila & GERAD, QC, Canada

  In this work, we propose improved task mapping strategies for real-time electric power system simulations on heterogeneous computing clusters, addressing both heterogeneous communication links and processing capacities, with a focus on bottleneck objectives. We approach the problem through two complementary models: the bottleneck quadratic semi-assignment problem (BQSAP), which optimizes task configuration for a fixed number of computing nodes while minimizing communication and computation costs; and the variable-size bin packing problem with quadratic communication constraints (Q-VSBPP), which minimizes the required number of computing nodes, particularly valuable for resource provisioning scenarios. Our contributions include extending the PuLP library to solve both problems with explicit incorporation of communication costs and processing constraints, and formalizing the nomenclature and definitions for bottleneck objectives in graph partitioning. This formalization fills a gap in existing literature and provides a framework for rigorous analysis and application of task-mapping techniques to real-time electric power system simulation. We also compare results with SCOTCH partitioning library.

• 
  11:20 AM - 11:45 AM

  Approximate Stochastic Dynamic Programming Framework for Hydropower Optimization

  • Luckny Zephyr, presenter, Laurentian University
  • Bernard F. Lamond, Université Laval
  • Kenjy Demeester, Rio Tinto

  We developed an approximate stochastic dynamic programming framework for hydropower optimization in which the value functions are approximated by sampling the reservoir level space over well-defined simplices (not necessarily full-dimensional). The latter are chosen iteratively by combining Monte Carlo simulation, linear programming and queuing strategy. This hybrid strategy is adopted to mitigate the complexity of simplicial decomposition, popular in global optimization and that we introduced in approximate dynamic programming. The proposal is tested on a real-world power system and benchmarked against an internal software package developed and used by an industrial partner.

• 
  11:45 AM - 12:10 PM

  A Two-Phase Optimization Framework for Profile Block Bidding in Short-Term Hydropower Scheduling

  • Mohammad Jafari Aminabadi, presenter,
  • Sara Séguin, Université du Québec à Chicoutimi
  • Stein-Erik Fleten, Norwegian University of Science and Technology
  • Ellen Krohn Aasgård, Norwegian University of Science and Technology

  This paper presents a two-phase model for short-term hydropower scheduling in the day-ahead market using profile block bids. The first phase generates feasible production profiles through a nonlinear optimization model that considers startup costs, opportunity costs, and hydrological constraints. In the second phase, a stochastic linear program selects the most profitable profiles under different price scenarios. The method is efficient due to structural properties like total unimodularity and is validated through a case study in Norway. For validation purposes, the proposed model has been compared with an hourly bidding strategy. Due to reduced complexity, the model has a very short solution time and either outperforms hourly bidding or produces results that are very close.