Evaluation of the geochemical impact of biomethane and natural gasmix injection in sandstone aquifer storage

EU is seeking to rapidly replace part of the natural gas usually stored in deep saline aquifers
by its renewable analogue, i.e. biomethane. While natural gas is depleted in oxygen, up to
10 000 ppm of O2 can be measured in biomethane. But to date, the geochemical impact of
the injection of a gas mix containing natural gas (thus CH4 and CO2) and biomethane (thus
oxygen) in deep saline aquifer has never been studied. The objective of this study is to
evaluate the resilience of geological storage to oxygen injection and predict the evolution
of the quality of the formation water, the gas plume and the rock formation. To do so,
multiple injection scenarios with various gas quality were tested using multiphase reactive
transport modeling. The gas mixtures were injected in two different deep saline aquifers in
the Paris Basin (France) whose petrophysical and geochemical characteristics were
reproduced during the simulation. Site 1 was a felspathic sandstone with clay cement and
site 2 was a sandstone with clay-calcareous cement.

One dimensional radial flow model evidenced that in both sandstones, the injection of gas
mix induced an acidification of the solution from pH~8 to pH~6. The injected oxygen
originating from biomethane was quickly consumed during pyrite oxidation and
contributed to the acidification of the formation water close to the injection point. However,
the injection of 1% mol fraction of CO2(g) contained in the gas mix was the main acidification
factor. Model demonstrated that the sandstones pH buffering capacity relied upon three
geochemical processes (i) calcite dissolution, (ii) feldspar dissolution and (iii) clays
dissolution and sorption capacity. These three pH-buffers were efficient for oxygen contents
from 10 to 1000 ppm. Models predicted that the gas mix injection could induce minimal but
long-term change in the nature of mineral phases but without significantly impacting the
porosity. Overall, the gas quality was preserved in both sandstones. This result was
evidenced with the modeling of entire gas injection and withdrawal scenario. CO2(g)
exsolution, and to a lesser extent H2S(g) exsolution, could occur during these cycles. In
addition, this study through modeling results concluded that the injection of biomethane
did not significantly change gas-water-rocks interactions compared to those of natural gas
injection. This study also evidenced that the detailed results were largely site-dependent.