Computational model improves planning of low-emission hydrogen plants

The advancement of hydrogen as an energy alternative for industrial decarbonization depends not only on technological innovations in production but also on methods capable of optimizing plant operations in response to the variability of renewable sources. In this context, researchers at the University of Campinas (Unicamp) have developed a computational model that reduces costs and ensures greater robustness in the planning of plants dedicated to the production of this clean fuel. The study, led by Luis Oroya, from the Department of Systems and Energy of the School of Electrical and Computer Engineering (FEEC-Unicamp), was published in the *International Journal of Hydrogen Energy*.

Applied Mathematics to Reduce Costs and Address Energy Uncertainties

The proposed model seeks to minimize total capital and operating expenses while ensuring stability in the face of fluctuations inherent in solar and wind generation, the main sources of electricity used in water electrolysis to obtain low-carbon hydrogen. To address these uncertainties, the methodology selects representative and extreme scenarios, enabling greater precision in plant design and future performance.

To make the process computationally feasible, the researchers reformulated the problem in terms of mixed-integer linear programming (MILP), a mathematical technique that allows solving complex systems with multiple variables and constraints. The approach uses decomposition algorithms that divide the challenge into smaller subproblems, which are solved iteratively until refined results are achieved. Using MILP makes it possible to simultaneously consider integer variables, such as the number of equipment, and continuous variables, such as energy volumes or costs, enabling more realistic analyses for the strategic planning of green hydrogen plants.