SESO2023

SESO 2023 International Thematic Week

Smart Energy and Stochastic Optimization


M. De Lara, V. Leclère, F. Pacaud

Unfortunately, registration are now closed.

SESO 2023

Monday 9 to Friday 13 October, 2023. École des Ponts Building Coriolis, Room B211.

Here is a link explaining how to get there, or on google maps.

Presentation

We are organizing a scientific workshop to follow the recent trends in the stochastic programming community, in both the academic and the industrial sides. The workshop will take place at École des Ponts, Champs-sur-Marne.

The week is composed of two days of tutorials, and a three-day scientific workshop.

Following SESO 2014, SESO 2015, SESO 2016 SESO 2017, and SESO 2018, the 6th International Thematic Week Smart Energy and Stochastic Optimization (SESO 2023) will be devoted to stochastic optimization, decentralized optimization and their applications to the management of emerging energy systems.

Support:

École_ponts_CMJNLogoGDRROLogo_carre_E4C_bleu

Registration info:

Registration is closed.

Registration fees: Waived for 2023

  • Speakers: free
  • Academic: 67€/day
  • Industrial: 137€/day

Limited support for students and young researchers is available. Please reach out if needed.

Note that a Julia workshop conference is organized the week before this conference.
More information: https://julia-users-paris.github.io/workshop/fr/.

1 Monday 9 October 2023: Tutorial #1

  • Title: Power-to-X: Portfolio Management of Hybrid Power Plants
  • Speaker: Jalal Kazempour Andrea Gloppen Johnsen (DTU)
  • Planning: 10:00 – 17:00

Hybrid power plants comprising renewable power sources and electrolyzers are envisioned to play a key role in accelerating the transition towards decarbonization. However, it is still an open question whether these plants are financially viable. One potential solution to make these plants profitable is to diversify their products, including power, hydrogen, and various frequency-supporting ancillary services, and participate in corresponding markets. This calls for a portfolio management problem under uncertainty. The sources of uncertainty are renewable production, market prices, and activation of ancillary services. In this tutorial with multiple hands-on exercises, we will first start with a deterministic model where the plant only produces power and hydrogen. We will then extend products by adding various ancillary services that the plant can produce, owed by the fast response capability of the electrolyzer. We will eventually model uncertainties, developing a stochastic portfolio management tool, to be used by the plants for optimal bidding strategy in multiple markets. During exercise sessions, we will provide several stylized case studies, work on programming codes, and interpret numerical results obtained.

2 Tuesday 10 October 2023: Tutorial #2

  • Title: JuMP for stochastic programming
  • Speaker: Oscar Dowson
  • Planning: 9:30-17:00

JuMP is a powerful and flexible modelling library that can be interfaced with numerous solver. In this tutorial we will see how to use it to deal with optimization under uncertainty, and in particular with stochastic programming approaches.

The tutorial will use the Julia language with JuMP.jl and SDDP.jl libraries.

Tutorial and instructions can be found here. Please follow the instructions to install Julia and JuMP in advance.

 

3 Wednesday 11 October 2023:  Workshop #1

  • Planning:
    • 9:30-10:00: Welcome
    • 10:00-10:30: Thomas Bittar, RTE, Solving large-scale stochastic optimization programs : application to investment problems for power systems
    • 10:30-11:00: Owen Palmer, Mines ParisTech, Risk-conscious asset sizing and energy procurement planning for an electrolytic hydrogen producer
    • 11:00-11:30: Nam Nguyen, Telecom SudParis, An Efficient Embedded ADMM Algorithm for Quadratically Constrained Quadratic Programming.

     

    • 12:00-12:30: Michel Marot, Présentation d’E4C
    • 12:30-13:00: Frédéric Babonneau, KEDGE, Modeling uncertainty in long-term energy models
    • 13:00-13:30: Daphné Tuncer, Towards the design of a data layer for the management of the E4C smart buildings

     

    • 13:30-15:00 Lunch

     

    • 15:00-15:20: Victor Spitzer, Lhyfe, Optimization of green hydrogen production under wind power forecast uncertainty
    • 15:20-15:40: Zoé Fornier, Metron, Accommodating fairness in a shared-energy allocation problem with uncertainties
    • 15:40-16:00: Ruiwen LIAO, Univ. Paris-Saclay, A two-stage stochastic programming model for lot-sizing with onsite generation of renewable energy

     

    • 16:30-16:50: Bianca Marin Moreno, EDF, Reimagining Demand-Side Management with Mean Field Learning
    • 16:50-17:10: Camila Martinez Parra, RTE,  A Two-timescale Decision Hazard Decision Formulation for Prospective Studies in Energy Systems under Uncertainties
    • 17:10-17:30: Mariam Sangare, A distributed and parallel scheduling method for demand response in energy communities with distributed generation and storage
    • 17:30-17:50: Nikolaus Houben, Optimal Microgrid Control: The limits of deterministic optimization for demand charges

4 Thursday 12 October 2023: Workshop #2

  • Planning:
    • 9:30-10:00: Ari-Pekka Perkkiö, LMU, Convex stochastic optimization
    • 10:00-10:30: Teemu Pennanen, Kings College, Optimal Operation and Valuation of Electricity Storages via Convex Optimization
    • 10:30-11:00: Wim Van Ackooij, EDF, Seasonal storage valuation for large and accurately modelled systems
    • 11:00-11:30: Eduardo Moreno, UAI, Benders Adaptive cuts for 2-Stage linear stochastic problems

     

    • 12:00-12:30: Adrien Lefranc, LAAS, Sparse Moment Sums-Of-Squares Hierarchies for Optimal Power Flow
    • 12:30-13:00: Nils Lohndorf, Univ of Luxembourg, Dynamic Hedging of Energy Portfolios with Stochastic Dual Dynamic Programming
    • 13:00-13:30: David Wozabal, Vrije Univ., Nash-Type Bargaining for Risk-Averse Agents with Applications in Energy

     

    • 13:30-14:40: Lunch

     

    • 14:40-15:00: Michel de Lara, Ecole des Ponts, Handling risk in decision-making: some experiences
    • 15:00-15:20: Nadia Oudjane, EDF, Optimizing new flexibilities in power systems
    • 15:20-15:40: Tristan Rigaut, Schneider Electric, Controlling Microgrids Without External Data: A Benchmark of Stochastic Programming Methods
    • 15:40-16:00: Pasqual Quach, Centrale-Supelec, A comparative study on robust optimal design of microgrids under random contingency

     

    • 16:30-16:50: Jonathan Dumas, RTE, Stochastic unit commitment applied to the security of supply in the dynamic margin monitoring strategy context
    • 16:50-17:10: Olivier Beaude, EDF, Sharing decarbonization efforts between regions: from studies done the “engineers’ way” to a game-theoretic formulation
    • 17:10-17:30:
    • 17:30-17:50:  Raian Lefgoum, PERSEE, Multistage stochastic optimization of a hydrogen supply chain
      • 17:50-19:30: Cocktail

5 Friday 13 October 2023: Workshop #3

  • Planning:
    • 9:30-10:00: Bernardo Pagnoncelli, SKEMA, Ranking and Contextual Selection
    • 10:00-10:30: Welington de Oliveira, Mines ParisTech, A derivative-free outer-approximation method for chance-constrained optimization
    • 10:30-11:00: Lucas Merabet, Metron, A primal-dual approach for risk-averse stochastic dual dynamic programming
    • 11:00-11:30: François Pacaud, Mines ParisTech, TBA

     

    • 12:00-12:20: Vitor Luiz Pinto de Pina Ferreira, TotalEnergies, Stochastic and Robust Unit Commitment for industrial Microgrid
    • 12:20-12:40: Jonathan Hornewall, EDF, Exact Optimization Methods for a Vehicle to Grid Smart Charging Problem
    • 12:40-13:00: Ingrid Sanchez Jimenez, Weather impact on cost recovery variability. A cosimulation of an investment and operational agent-based model.