Stochastic planning for feeding a green hydrogen plant into an isolated network

Abstract

In recent years, an electrochemical process called electrolysis has gained prominence. This process uses water and electricity as its main sources, significantly reducing the carbon footprint of hydrogen production. Additionally, colors have been assigned to represent the source of hydrogen production in a simple way. For example, green refers to hydrogen produced by electrolysis using electricity generated from non-conventional renewable energy sources (NCRES). For plants not connected to the national grid, the connection of a green hydrogen plant requires that NCRES be connected to an isolated electrical grid. In these cases, the power supply will depend on the variability of the source. This paper presents the methodology to plan and size the main components of the wind power plant and the battery energy storage system (BESS) to ensure that the electrolyzer constraints can be met during the studied period. Furthermore, it introduces a novel methodology that uses the autoregressive moving average (ARMA) model to generate a sequential Monte Carlo simulation along with dynamic optimization. This approach allows for the sizing of the wind power plant and BESS, considering the stochastic behavior of the wind.