Experimental kinetic rate laws for aqueous pyrite reduction at underground hydrogen storage conditions (60–150 °C, up to 150 bar H2)

The reaction kinetics of reductive pyrite (FeS 2 ) dissolution by H 2 is important to predict the long-term aqueous and gaseous sulfide release in underground hydrogen storage and other engineered subsurface sites. We investigated the rates of pyrite reduction based on sulfide formation as a function of temperature (60-150 °C), H 2 partial pressure (0-150 bar), and pH in situ (∼ 4-10) in deoxygenated 0.03 M NaCl solutions. The experiments used natural pyrite powder (ø = 50-100 µm) and were run in hydrothermal batch reactors made of either titanium or Dursan®-coated 316L stainless steel. After experimental durations of ∼ 700-900 h, dissolved S -II (aq) concentrations measured by methylene blue spectrophotometry ranged from 10 -5 to 10 -3 M. Long-term pyrite dissolution and concomitant elevated S -II (aq) concentrations were controlled by the formation of secondary pyrrhotite (Fe 1-x S) and magnetite (Fe 3 O 4 ), which prevented the ion activity product of FeS 2 from achieving rapid saturation. Conversely, in the absence of pyrrhotite and magnetite precipitation, the solutions rapidly equilibrated with respect to pyrite, resulting in low S -II (aq) concentrations. Rates of reductive pyrite dissolution were determined at 12 h from total S -II (aq+gas) concentration vs. time data and were found to increase with temperature and H 2 -partial pressure. The rate dependency on pH at 90 °C, 7 bar P H2 was regressed either with an asymmetric ‘V’-shaped (two rate-pH domains) or a ‘U’-shaped (three rate-pH domains) relation. For this reason, two separate kinetic rate laws were derived, based on the ‘U’-regression ( E a = 35.2 kJ mol -1 ): r = 10 - 5.42 e - 35208 / R T P H 2 0.37 1 - Q K eq or the ‘V’-regression ( E a = 29.4 kJ mol -1 ): r = 10 - 5.13 e - 29370 / R T a H + 0.13 P H 2 0.37 1 - Q K eq The performance of both rate laws was validated by reproducing the experimental S -II (aq) concentrations in kinetic models. This comparison determined their applicability to be valid from 60 to 120 °C, 0-150 bar P H2 , and pH ∼ 6.7-8.6 (‘U’ rate law) or pH ∼ 4.1-8.6 (‘V’ rate law). The rate laws are compatible with geochemical reactive transport codes and will enhance the understanding of geochemical fluid-rock-gas interactions in H 2 -bearing subsurface environments.