Jean Mura ¹ Magali RANCHOU-PEYRUSE ^1,2,3 Marion GUIGNARD ² Marion DUCOUSSO ^1,3 Marie LARREGIEU ² Marie-Pierre Isaure ² Isabelle LEHECHO ^2,3 Guilhem Hoareau ⁴ Marie POULAIN ^1,3 Mateus De Sousa Buriti ¹ Pierre CHIQUET ^3,5 Guilhem Caumette ^3,5 Anélia Petit ⁶ Pierre Cézac ^1,3 Anthony Ranchou-Peyruse ^2*

• ¹ Universite de Pau & Pays Adour, E2S-UPPA, LaTEP, Pau, France
• ² Université de Pau & Pays Adour, E2S-UPPA, CNRS, IPREM, Pau, France
• ³ Joint Laboratory SEnGA, UPPA-E2S-Teréga, Pau, France
• ⁴ Universite de Pau & Pays Adour, E2S-UPPA, CNRS, TOTAL, LFCR, Pau, France
• ⁵ Teréga, Pau, France
• ⁶ Storengy, Bois-Colombes, France

If dihydrogen (H2) becomes a major part of the energy mix, massive storage in underground gas storage (UGS), such as in deep aquifers, will be needed. The development of H2 requires a growing share of H2 in natural gas (and its current infrastructure), which is expected to reach approximately 2% in Europe. The impact of H2 in aquifers is uncertain, mainly because its behavior is site dependent. The main concern is the consequences of its consumption by autochthonous microorganisms, which, in addition to energy loss, could lead to reservoir souring and alter the petrological properties of the aquifer. In this work, the coinjection of 2% H2 in a natural gas blend in a low-salinity deep aquifer was simulated in a three-phase (aquifer rock, formation water, and natural gas/H2 mix) high-pressure reactor for three months with autochthonous microorganisms using a protocol described in a previous study. This protocol was improved by the addition of protocol coupling experimental measures and modeling to calculate the pH and redox potential of the reactor. Modeling was performed to better analyze the experimental data. As in previous experiments, sulfate reduction was the first reaction to occur, and sulfate was quickly consumed. Then, formate production, acetogenesis, and methanogenesis occured. Overall, H2 consumption was mainly caused by methanogenesis. Contrary to previous experiments simulating H2 injection in aquifers of higher salinity using the same protocol, microbial H2 consumption remained limited, probably because of nutrient depletion. Although calcite dissolution and iron sulfide mineral precipitation likely occurred, no notable evolution of the rock phase was observed after the experiment. Overall, our results suggested that H2 can be stable in this aquifer after an initial loss. More generally, aquifers with low salinity and especially low electron acceptor availability should be favored for H2 costorage with natural gas.