Numerical simulation of large-scale seasonal hydrogen storage in an anticline aquifer: A case study capturing hydrogen interactions and cushion gas injection

Abstract

In response to the Paris Agreement, the transition from fossil fuels to renewable energy resources (RER) is one of the most crucial measures in responding to climate change. Under this context, hydrogen storage with the concept of ‘power to gas’ represents a possible solution for the mismatch problem between fluctuating RER generation and downstream demand in a seasonal time scale. Compared to conventional low storage capacities of salt caverns, widely spread aquifers provide a possibility for safe, cost-effective, and environmentally friendly long-term hydrogen storage. However, current knowledge of hydrogen storage in these aquifers is only limited to a few impure hydrogen projects; understanding their trapping mechanisms as well as how to recover their trapped hydrogen is urgent. In order to fulfil the knowledge gap, not only are the complex underground interactions of hydrogen with rock or a fluid (residual trapping or dissolution trapping) modelled and validated against reliable experimental data, but also multiple cycles and cushion gas injection are integrated to enhance hydrogen recovery in this study. Through powerful numerical studies, the results indicate that: (a) at a seasonal storage time scale, residual trapping has the most effect in reducing hydrogen round-trip efficiency; (b) multiple cycles are helpful for hydrogen storage in an aquifer reservoir with round-trip efficiency increasing from 50.9% in the first cycle to 78.6% at the end; (c) nitrogen as a cushion gas for injection has a better performance compared with carbon dioxide with a capability to unlock residual trapped and dissolution trapped hydrogen; (d) nitrogen as a cushion gas for injection can effectively alleviate insufficient hydrogen production capacity in the first cycle with the hydrogen round-trip efficiency increased from 50% to 70%; and (e) the hydrogen extraction phase ends up with a very high recovery efficiency (around 85%) in a nitrogen injection scenario.