Abstract Engineered geological porous media hydrogen storage must be designed to ensure secure storage, and use appropriate monitoring, measurement and verification tools. Here, we identify and characterize 60 natural hydrogen seeps as analogues for potential leakage from engineered storage reservoirs to consider implications for monitoring. We report and compare the geological and environmental setting; seepage mode (dry gas/associated with water); co-released gases; seep rates and areal fluxes; temporal variation; seep structure; gas source; and composition. Seep characteristics are determined by local geological and hydrological conditions, specifically whether hydrogen gas is seeping through soils and unconsolidated sediments, fractured bedrock or into water. Hydrogen is typically co-emitted with other gases (CO 2 , CH 4 , N 2 ) with CH 4 , the most common co-emitted gas. The structural controls on seep location and characteristics are similar between hydrogen and CO 2 seeps. However, compared to CO 2 , hydrogen is more readily dispersed when mixing with air and hydrogen is more prone to being consumed or transformed via biotic or abiotic reactions, and so the quantity of leaked hydrogen can be greatly attenuated before seeping. Monitoring approaches should therefore be tailored to the local geology and hydrological conditions, and monitoring approaches to detect hydrogen and associated gases would be appropriate.

Abstract: The study of natural analogues can inform the long-term performance security of engineered CO 2 storage. There are natural CO 2 reservoirs and CO 2 seeps in Italy. Here, we study nine reservoirs and establish which are sealed or are leaking CO 2 to surface. Their characteristics are compared to elucidate which conditions control CO 2 leakage. All of the case studies would fail current CO 2 storage site selection criteria, although only two leak CO 2 to surface. The factors found to systematically affect seal performance are overburden geopressure and proximity to modern extensional faults. Amongst our case studies, the sealing reservoirs show elevated overburden geopressure whereas the leaking reservoirs do not. Since the leaking reservoirs are located within <10 km of modern extensional faults, pressure equilibration within the overburden may be facilitated by enhanced crustal permeability related to faulting. Modelling of the properties that could enable the observed CO 2 leakage rates finds that high-permeability pathways (such as transmissive faults or fractures) become increasingly necessary to sustain leak rates as CO 2 density decreases during ascent to surface, regardless of the leakage mechanism into the overburden. This work illustrates the value of characterizing the overburden geology during CO 2 storage site selection to inform screening criterion, risk assessment and monitoring strategy. Correction notice: The original version was incorrect. This was due to an error in the Acknowledgements and Funding section, which omitted to list the funding bodies of RSH.

Abstract The Upper Rhine Graben (URG) is a seismically active tectonic structure in intraplate Europe. Large and moderate earthquakes have occurred along the URG in the past but no coseismic surface faulting has been reported so far. We investigated active faulting along the western edge of the northern URG and identified the 25 km-long linear Riedseltz–Landau normal fault scarp as a major tectonic structure affecting late Pleistocene and Holocene deposits. The fault zone is exposed in the Riedseltz quarry where it affects Pliocene sand and gravels and overlying late Pleistocene (Wurm) units. These units have not been buried deeper than a few tens of metres and yet the fault zone contains cataclastic deformation textures. Cataclasis is demonstrated by spalling and transgranular fractures in quartz grains concentrated in deformation bands with reduced grain size. The observed microstructures suggest multiple phases of deformation with cataclasis followed by emplacement of a prominent Fe-oxide matrix into deformation bands, and later emplacement of a clay-rick matrix into fractures. Previous geological and geophysical studies along the fault show late Pleistocene (Wurm) loess deposits ( c. 24–10 ka before present) and early Holocene sand–silty deposits with individual or cumulative 1.5 and 0.7 m surface slip, respectively. Field observations and previous results from shallow geophysics provide a minimum 0.15 mm a −1 time-averaged slip rate. The Riedseltz fault parameters integrated in a dislocation model suggest a minimum Mw 6.6 earthquake as a plausible scenario in the northern URG. The observations of cataclasis in shallowly buried sediments coupled with observations of the late Quaternary fault scarp call for palaeoseismic studies that may document the occurrence of a larger earthquake on the western edge of URG. Surface faulting of young, shallowly buried sediments associated with cataclasis provides new evidence for assessing the occurrence of large earthquakes and seismic hazard assessment in the northern URG.

Abstract The fluid flow properties of faults are highly variable and spatially heterogeneous. We use numerical simulation of flow through field maps of detailed fault zone architecture to demonstrate that flow across the fault zone is controlled by connected high-permeability pathways, which are highly tortuous in mapped fault outcrops. Such small-scale, geometrically complex, fault zone architectural features can never be resolved for subsurface faults. Consequently, the key to prediction of subsurface bulk fault zone hydraulic properties is a statistical characterisation of the likelihood and frequency of such connected pathways. We demonstrate for a single architectural feature, the fault core, that thickness variation along strike can be well described by a spatially correlated random field with a spherical covariance structure. These data are from a single site in a specific lithology. To enable such statistics to be used to make predictions at other sites, a large number of similar datasets must be pooled. This will enable us to relate such spatial statistics to gross properties such as host rock lithology and fault throw, which are measurable for subsurface faults.

Abstract Analysis of fault zone structure and composition of two intermediate-displacement faults in the Colorado Plateau reveal how fault structure varies as a function of lithology, and how faults impact fluid flow. The Little Grand Wash fault cuts Jurassic Summerville through Cretaceous Mancos Shale rocks, and consists of a complex set of interweaving fault strands. Fault relays are developed where sandstone and shale are cut by the fault in roughly equal amounts. Ancient and modern fluid flow is documented by the presence of travertine and tufa deposits, an oil seep, and CO 2 gas seeps. Analysis of the water, travertine, and gas composition indicate that the CO 2 emanates from a reservoir 1.5–2 km deep, and charges a shallow aquifer. Cross-fault flow is inhibited, and the gas and water flows in the footwall damage zone of the fault. Analysis of the Bighole fault in the San Rafael Swell shows how fault structure varies with displacement. The fault is exposed entirely in the aeolian Jurassic Navajo Sandstone, and consists of a dense fault core interpreted to be a densely packed set of deformation bands bounded by a narrow slip surface. The fault zone consists of conjugate deformation band sets in the hanging wall and footwall of the fault, and the fault core thickness does not vary significantly with net slip.