Researchers at the Ulsan Institute of Science and Method created a technology that can estimate both the cost of hydrogen generation and carbon dioxide emissions at the same time (UNIST).
UNIST announced that a research team led by Professor Im Han-kwon (pictured) of the Department of Energy and Chemical Engineering has developed a predictive model that uses artificial intelligence and simulation technology to comprehensively evaluate the performance of the hydrogen production process.
It is a model that uses machine learning, a branch of artificial intelligence, to predict the performance of chemical processes. This model has the benefit of being able to anticipate not just technical performance like yield, but also production costs and CO2 emissions all at once.
The performance of the hydrogen manufacturing process was previously tested using a three-step approach, however, the study team said that utilizing the established model, a value equivalent to that of the three-step evaluation technique was reached up to 99.9%.
This model was used by the researchers to assess the efficacy of a newly built methanol wet reforming process. The suggested process was assessed using a prediction model that changed 12 technological and economic aspects including reaction temperature and labor cost.
When compared to the findings of the previous three-step evaluation, the prediction accuracy of the process’s hydrogen production, carbon dioxide emission, and the hydrogen production cost was 99 percent, 99.9%, and 96 percent, respectively.
The number of reactors, reaction temperature, methanol raw material price, and labor cost was also shown to be the most critical elements in the process’ performance among the 12 technological and economic aspects studied.
