EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

4175. Dealing with stochasticity in the renewable hydrogen production

Invited abstract in session WB-9: Production Optimization and Supply Chain Management of Green Hydrogen under Uncertainties, stream Energy Markets.

Wednesday, 10:30-12:00
Room: 10 (building: 116)

Authors (first author is the speaker)

1. Patricio Reyes
CASE, BSC
2. Kelly Wagemann
NIC Chile Research Labs
3. Javier Hernanz Zajara
Quantum Advisory Team, Repsol
4. Santiago Fernández
BSC - Repsol Research Center
5. Artur Garcia-Saez
CASE, Barcelona Supercomputing Center
6. Ricardo Enríquez
Quantum Advisory Team, Repsol Technology Lab

Abstract

Energy arbitrage is a common practice in the European energy sector, which consists of taking advantage of differences in energy prices in the market. Companies use arbitrage to buy energy at a certain price, store it in batteries and later sell it at a higher price. Based on the same idea, hydrogen production can be considered renewable if the electricity supply mix contains a minimum amount of renewables. This work models the costs of an energy supply chain, identifying the best moments of energy purchase and sale over a time period, to control an industrial plan producing green hydrogen, taking into account a renewable share. Optimal conditions to operate the plant are unknown, as demand and production are uncertain variables at the moment to participate in the market. Our developed solution evolves from a deterministic model to an improvement with the introduction of stochasticity. The model is tested in realistic conditions with a custom package capable of solving the optimization problem for all the study cases. For a time horizon of 24 hours, the model is solved in less than half a second using state of the art solvers in an HPC cluster. To incorporate uncertainty, the variability of the energy price of the electrical grid is modeled by generating random scenarios. This is achieved by predicting time series using both traditional statistical methods and neural networks. It allows to reliably represent the variability of the operation over the energy market.

Keywords

Status: accepted

Back to the list of papers