New algorithm spells more efficient underground hydrogen storage: A team of researchers from Khalifa University (KU), King Fahd University of Petroleum and Minerals, and University Teknologi Petronas have developed an algorithm that can predict where and how much hydrogen can be stored underground, according to a statement. The researchers aimed to address the challenge of finding space to store hydrogen as demand for the green fuel grows. A potential solution is underground storage in sedimentary formations, such as shale gas reservoirs. The sedimentary rock has a high adsorption rate and is safe for hydrogen storage, but estimating the amount of hydrogen that can be adsorbed in shale is a complex and time-consuming task.

What they did: The researchers used gradient boosting regression, a machine learning technique, to create a data-driven model that can predict hydrogen adsorption in kerogen — the organic material in shale — based on various parameters like pressure, temperature, and kerogen density. Compared to other machine learning methods, the model proved to be accurate and fast for different types of shale paving the way for quicker identification of the best locations and conditions for storing hydrogen in formations.

Tags: