International Journal of Coal Geology

Transcription

1 Accepted Manuscript Impure CO2 reaction of feldspar, clay, and organic matter rich cap-rocks: Decreases in the fraction of accessible mesopores measured by SANS Julie K. Pearce, Grant K.W. Dawson, Tomasz P. Blach, Jitendra Bahadur, Yuri B. Melnichenko, Suzanne D. Golding PII: S (17) DOI: doi: /j.coal Reference: COGEL 2926 To appear in: International Journal of Coal Geology Received date: 8 August 2017 Revised date: 30 October 2017 Accepted date: 13 November 2017 Please cite this article as: Julie K. Pearce, Grant K.W. Dawson, Tomasz P. Blach, Jitendra Bahadur, Yuri B. Melnichenko, Suzanne D. Golding, Impure CO2 reaction of feldspar, clay, and organic matter rich cap-rocks: Decreases in the fraction of accessible mesopores measured by SANS. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Cogel(2017), doi: /j.coal This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

2 Impure CO 2 reaction of feldspar, clay, and organic matter rich cap-rocks: Decreases in the fraction of accessible mesopores measured by SANS Julie K. Pearce 1*, Grant K.W. Dawson 1, Tomasz P. Blach 2, Jitendra Bahadur 3#, Yuri B. Melnichenko 3^, Suzanne D. Golding 1 1 School of Earth and Environmental Sciences, University of Queensland, QLD 4072, Australia 2 Institute for Future Environments, Queensland University of Technology, QLD 4000, Australia 3 Biology and Soft Matter Division, Oakridge National Laboratory, Tennessee, 37830, USA * J.pearce2@uq.edu.au # Current address: Solid State Physics Division, Bhabha Atomic Research Centre, India ^ Deceased March 18 th 2016

3 Abstract During CO 2 geological storage, low porosity and permeability cap-rock can act as a structural trap, preventing CO 2 vertical migration to overlying fresh water aquifers or the surface. Clay and organic matter rich shales, fine-grained sandstones and mudstones often act as cap-rocks and may contain substantial sub-micron porosity. CO 2 -brine-rock interactions can open or close pore throats through dissolution, precipitation or migration of clay fines or grains. This could affect CO 2 migration if the porosity is accessible, with unchanging or decreasing accessible porosity favourable for trapping and integrity. Two cap-rock core samples, a clay and organic-rich mudstone and a more organic-lean feldspar-rich fine grained sandstone, from a well drilled for a CO 2 storage feasibility study in Australia were experimentally reacted with impure CO 2 (+SO 2,O 2 ) and low salinity brine at reservoir conditions. Mercury injection capillary pressure indicated that the majority of pores in both cores had pore throat radii ~ nm with porosities of %. After reaction with impure CO 2 - brine the measured pore throats decreased in the clay-rich mudstone core. Dissolution and precipitation of carbonate and silicate minerals were observed during impure CO 2 reaction of both cores via changes in water chemistry. Scanning electron microscopy identified macroporosity in clays, mica and amorphous silica cements. After impure CO 2 -brine reaction, precipitation of barite, Fe-oxides, clays and gypsum was observed. Ion leaching from Fe-rich chlorite was also apparent, with clay structural collapse, and fines migration. Small-angle neutron scattering measured the fraction of total and non-accessible pores (~ nm radii pores) before and after reaction. The fraction of pores that was accessible in both virgin cap-rocks had a decreasing trend to smaller pore size. The clay-rich cap-rock had a higher fraction of accessible pores (~0.9) at the smallest SANS measured pore size, than the feldspar rich fine-grained sandstone (~0.75). Both core samples showed a decrease in SANS accessible pores after impure CO 2 -water reaction at CO 2 storage conditions. The clay-rich cap-rock showed a more pronounced decrease. After impure CO 2 -brine reaction the fraction of accessible pores at the smallest pore size was ~0.85 in the clay-rich cap-rock

4 and ~0.75 in the feldspar-rich fine-grained sandstone. Reactions during impure CO 2 -brine-rock reaction have the potential to close cap-rock pores, which is favourable for CO 2 storage integrity. Keywords: Sandstone; Shale; Impure CO 2 storage; SANS; Evergreen Formation; CO 2 storage integrity

5 1. Introduction Carbon dioxide storage in siliciclastic reservoirs generally involves injection into a high porosity and permeability reservoir. Low porosity and permeability overlying cap-rocks or seals, which are generally clay-rich shale, mudstone or siltstone, structurally trap the CO 2 preventing or slowing migration (Gaus, 2010). Because CO 2 dissolves in formation water to produce carbonic acid, which reacts with some rock forming minerals(kaszuba et al., 2013), gas-water-rock reactions may increase and subsequently decrease rock porosity; often the clay-rich cap-rock contains minerals that are more geochemically reactive than those in the target reservoir (Farquhar et al., 2015). Over short time periods, mineral dissolution may increase porosity, but over longer time periods, migration of fines and mineral precipitation, including mineral trapping of CO 2 as carbonates, may decrease caprock porosity or permeability (Higgs et al., 2013; Higgs et al., 2015; Watson et al., 2004). Several studies of mineral weathering or natural analogues of CO 2 storage have shown porosity decreases through carbonate and clay precipitation, especially at the reservoir-seal interface (Higgs et al., 2013; Navarre-Sitchler et al., 2013). Recently micro computed tomography (CT) has been used as an automated technique to characterise reservoir and cap-rock porosity down to the micron scale (Farquhar et al., 2015; Golab et al., 2015c; Golab et al., 2010; Pearce et al., 2016b; Wunsch et al., 2014). Experimental CO 2 -water-rock reactions have been shown to result in changes to (micro) CT resolvable porosity, especially through dissolution of calcite cements over the time scale of weeks (Dávila et al., 2016). However micro CT generally does not resolve the estimated 80% of nanometer scale porosity in shales or mudstones (Curtis et al., 2010). Changes in porosity will be most important when pores are open and interconnected or accessible to migrating gas and brine; therefore characterising changes to cap-rock accessible pore fraction in the nanometer size range is needed. Small-angle neutron scattering (SANS) with contrast matching can measure both the total and inaccessible pores in the nanometer size region, and has recently been applied to characterising organic matter rich gas shales and sandstones as well as coal (Bahadur et al., 2015; Clarkson et al., 2013; Melnichenko et al., 2012; Ruppert et al., 2013).

6 CO 2 streams injected subsurface from industrial sources such as coal post combustion capture or oxy-fuel firing, cement or steel processing plants may contain ancillary flue gases or impurities including Ar, N 2, CH 4, SO 2, NOx (NO 2, NO), O 2 (Porter et al., 2015; Talman, 2015). In Australian and UK power plants, the relatively low S and N content of coal means historically desulfurization of flue gas has not been necessary. Co-injecting impurities with the CO 2 in geological storage has been suggested to reduce the cost of CO 2 capture by avoiding installing desox and denox technology (Glezakou et al., 2012). Several authors have shown that dissolved SO 2, O 2 and NOx are more reactive than CO 2 to rock through the formation of strong sulphuric and nitric acids (Dawson et al., 2015b; Knauss et al., 2005; Palandri and Kharaka, 2005; Pearce et al., 2015a; Pearce et al., 2015b). However, relatively few studies have investigated the impacts of impurity gases (Gaus, 2010). Coinjection of SO 2 (and O 2 ) with CO 2 has been predicted in geochemical modelling studies and observed in laboratory experiments to result in the precipitation of sulphate and oxide minerals, e.g. hematite or gypsum/anhydrite (Lu et al., 2014; Pearce et al., 2016b; Shao et al., 2014; Xu et al., 2007). Co-injection has the potential to plug pores and decrease accessible porosity effectively selfsealing the cap-rock. The precipitation of Fe-oxides, sulphides, and sulphates such as gypsum or anhydrite with carbonate minerals and clays, has been observed in natural analogue studies of CO 2 accumulation and leakage. For example, gypsum, hematite and pyrite precipitation has been reported in fractures at a natural CO 2 accumulation in Green River, Utah (Chen et al., 2016; Kampman et al., 2014; Wigley et al., 2012). Chopping and Katzuba have shown that anhydrite and pyrite precipitated in pore space and fractures in the presence of natural supercritical CO 2 and dissolved sulphate and sulphide in the Madison Limestone, Wyoming (Chopping and Kaszuba, 2012; Kaszuba et al., 2011). The Evergreen Formation overlies the Precipice Sandstone, a proposed low salinity CO 2 storage reservoir in the Surat basin, Queensland, Australia (Figure 1) (Farquhar et al., 2013; Hodgkinson and Grigorescu, 2012). The Evergreen Formation, the storage cap-rock, consists of quartzoze sandstone interbedded with siltstone, shale and carbonaceous mudstone of variable porosity and permeability,

7 with reported average porosity 15 ± 6%, and average horizontal permeabilities of 87 ± 246 md (Kellett et al., 2012). Several cores from the GSQ Chinchilla- 4 well and West Wandoan 1 well were recently determined to have porosities in the range % and permeabilities of <0.1 md (Farquhar et al., 2015; Golab et al., 2015c). The purpose of this study was to determine if pores in cap-rock cores from a well drilled for a CO 2 storage feasibility study in the Surat Basin, Australia, become opened or closed through the interaction with impure CO 2 brine. Small angle neutron scattering was used to determine the change in accessible pores before and after reaction with impure CO 2 (+ SO 2 and O 2 ) and low salinity brine at reservoir conditions. Complementary scanning electron microscopy with energy dispersive spectroscopy (SEM-EDS), mercury injection capillary pressure (MICP), and experimental water chemistry aid interpretation by detecting mineral habit, macroporosity, mesoporosity, mineral dissolution or precipitation and physical rock changes. 2. Materials and Methods Two whole core depth sections of the Jurassic Evergreen Formation from the West Wandoan 1 well, Surat Basin, Queensland, Australia, were sampled at depths of m and m (well position: latitude , longitude ). The well core has been described elsewhere (Golab et al., 2015a). An interlaminated black carbonaceous silty mudstone and very fine-grained sandstone was selected from m (referred to as 981m) with additional adjacent samples collected for additional analyses. The core at m depth (referred to as 1043m), a fine-grained sandstone with mud rip-up clasts, was also sampled, and exhibited a courser grain size than the 981 m core and was plagioclase feldspar rich. Petrography of the cores has been detailed elsewhere (Dawson et al., 2015a).

8 The mineralogy of the core samples 981m and 1043m is shown in Table 1 from data reported previously (Dawson et al., 2015a; Golab et al., 2015b; Golab et al., 2015c). SEM-EDS performed in this study confirmed that Fe-rich, Mg-poor chlorite was present in both samples. Organic matter and minerals present in minor to trace amounts identified in SEM-EDS include, apatite, Ti-oxide, monazite, zircon, barite, sphalerite, and calcite/ankerite ± Mn. Petrography of the same core section reported elsewhere also identified Fe-oxides (Dawson et al., 2015a). Sample 981m contained a higher amount of organic matter than 1043m as indicated by the higher loss on ignition (LOI) (Table 1) (Ross and Bustin, 2009). In the Evergreen 1043 m sub-sample minor to trace components observed in SEM-EDS performed in this study include organic matter (visibly less than Evergreen 981m), barite, monazite, apatite, zircon, Ti-oxide, sphalerite and pyrite. EDS confirmed the presence of Fe-rich, Mg-poor chlorite. XRD reported elsewhere detected the presence of ~ <5% illite and Fe-oxides (Dawson et al., 2015a). Table 1: Mineral components in the Evergreen Formation cores were determined by QEMSCAN reported previously (Golab et al., 2015b). Major element contents as oxides determined by rock fusion and ICP-OES, and loss on ignition (LOI) is also shown. * Illite was previously identified by XRD to be present at <5% (Dawson et al., 2015a). # Calcite or ankerite was detected via SEM-EDS in trace amounts ~0.1%, occasional barite grains were also observed. Sample 1043m 981m Sample 1043m 981m Depth / m Depth / m Quartz SiO Alkali feldspar TiO Plagioclase Al 2 O Muscovite/Illite Fe 2 O Biotite MnO Illite-Smectite 0* 22.5 MgO Kaolinite CaO

9 Chlorite Na 2 O Calcite # K 2 O apatite P 2 O barite # Unclassified and traces Total Total Clay/Mica LOI The core samples 1043m and 981m were reacted in a batch reactor system at reservoir in situ conditions of 60 C and 12 MPa for 28 days. Core sub-samples were immersed in 100 ml of 1500 ppm NaCl low salinity brine and pressurised with an inert gas, N 2, for 4 days to equilibrate the rock and fluid and provide baseline water samples. A 1500 ppm NaCl low salinity brine was used to replicate a simplified formation water representative of the injection reservoir (Hodgkinson and Grigorescu, 2012). Subsequently, the N 2 was depressurised and a gas mixture consisting of 0.2% SO 2, 2% O 2 and a balance of 97.8% CO 2 was injected into the reactors. The reactions then proceeded for 28 days. This gas mixture is representative of an average possible gas stream from oxyfuel capture (Talman, 2015). The gas stream composition planned to be injected in the field was not known at the time of this study, but is now expected to contain lower concentrations of impurity gases. The batch reactors were based on unstirred Parr reactors with thermoplastic liners to prevent fluid corrosion, the reactor design has been described in detail elsewhere (Pearce et al., 2016a; Pearce et al., 2015a). Virgin and reacted core samples, designated as 1043m and 981m, and 1043mR and 981mR respectively, were prepared for SANS measurements by hand polishing subsplit core samples to 15 mm width discs on the face perpendicular to the bedding plane. Discs were polished to 1 mm thickness +/- 0.2 mm. SEM-EDS images were obtained for pre and post reaction samples. Pre and post reaction images were obtained in the same position where possible on reacted unpolished blocks to directly identify dissolution or precipitation of minerals. SEM-EDS was also performed on the polished discs to identify minor mineral phases, organic matter, mineral habit and visible

10 porosity (JEOL 6460LA environmental, and Hitachi TM3030). Pre and post reaction Helium pycnometry (Micrometrics AccuPyc II 1340), unstressed brine permeability, and mercury intrusion (Micromeritics AutoPore IV 9500) were also performed, with the methods described elsewhere (Dawson et al., 2015a; Massarotto et al., 2010). Pore throats of 1.5 nm to 150μm radius, mercury porosities, and total accessible porosities were obtained by MICP and He pycnometry respectively (calculated from grain and bulk densities). During the high pressure and temperature CO 2 O 2 SO 2 - brine reactions, water samples were taken from the Parr reactors at time 0, 120, 216, 288, 384, 456, 552 and 648 h, with ph and conductivity measured immediately with probes. Water samples were then filtered (45μm), diluted, and acidified with ultrapure nitric acid for detection of major and minor ions with Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES) ( Perkin Elmer Optima 8300DV, error < 5%) and Inductively Coupled Plasma Mass Spectrometry (ICP-MS)( Agilent 7900, error <5%) to aid interpretation of dissolution and precipitation processes. Additionally unfiltered water samples were analysed for total element contents. Virgin and reacted polished discs, 1043m, 981m, 1043mR, and 981mR, were dried in a vacuum oven and inserted into custom built high pressure and temperature cells for SANS analysis using the general purpose SANS beamline (GP SANS) at Oak Ridge National Laboratory s (ORNL) High Flux Isotope Reactor (Fig. 2) (Wignall et al., 2012). Measurements were performed with a Q range of A -1, and neutron wavelengths 4.72, and 6 A with a / resolution of The four core polished discs were first measured at atmospheric conditions (room T = 23 C), and then again at zero average contrast (ZAC) conditions using high pressure deuterated methane (CD 4 ) as a contrast matching fluid. The pressure of CD 4 was increased stepwise to reach ZAC conditions for each sample to find the minimum scattering at resulting gas pressures of psi. At ZAC conditions the scattering length density or scattering power of the accessible pores filled with the gas under pressure become equal to the scattering power of the rock matrix, so that the net scattering is from the inaccessible pores. Further details on the contrast matching technique have been published elsewhere (Bahadur et al., 2015; Melnichenko et al., 2012). Source to sample

11 distances of 18.5m, and 0.3 m were used with conditions generally similar to previous studies (Melnichenko et al., 2012; Ruppert et al., 2013). Data were corrected for background, exact sample thickness, and detector efficiency, and pre-calibrated standards were used to put the data on an absolute scale. An approximate relationship of scattering vector to pore radius r 2.5/Q was used to relate data to pore radii. 3 Results and Discussion SEM-EDS pre reaction 981m Evergreen Formation sample 981m is a dark mudstone with light sandstone laminations (Fig. 2). The sandy layers of 981m appeared to contain more of the Fe-chlorite, trace carbonates and zircon grains than the mudstone layers (Fig. 3a). Fe was identified in illite-smectite, and Fe-rich Mg-poor chlorite and biotite were observed (Fig. 3b,c). Evergreen 981m exhibits a fine-grained clay-rich matrix: Slit like porosity is present in kaolinite, and in mica booklets up to ~ 100 µm in length (Fig. 4a). Macroporosity was also observed in chlorite, illite, and barite cements (Fig. 4c,d). Large patches (up to 200 μm length) of apparently pore filling kaolinite rimmed by chlorite were present in sandy layers of 981m (not shown). In 981m macro porosity was also observed in calcite cement, Ti oxide cements, and apatite via SEM images of polished thin sections of core (not shown). Although pores are observed in the SEM images, pore connectivity cannot be determined from a 2D image. Pieces of organic matter were also present in 981 m (Fig. 4b). Organic matter was generally in the size ranges 40 x 5 µm, 20 x 20 µm, and 100 x 50 µm, however occasional large pieces up to 1.5 mm x 50 µm were present (not shown). Some organic matter also contained layers of pyrite and clays (not shown).

12 SEM-EDS post reaction 981mR Post reaction 981mR had visible brown precipitated material, with some bleaching of the dark mudstone layers in (Fig. 2 d). The Fe content of chlorite decreased (appeared leached) after reaction (Fig. 3d), and ankerite and Ca-plagioclase also appear corroded. Fe-bearing fine-grained precipitates of oxides and clay/amorphous material and fine-grained barite covered mineral surfaces post-reaction (Fig. 3d,e): Ca-sulphate with a gypsum morphology was observed post reaction on 981mR (Fig. 3d-f). After reaction, SEM-EDS of 981mR showed areas of chlorite exhibiting altered or collapsed appearances (Fig. 4e). Fe-oxide precipitates were observed on organic matter, along with silica grains which potentially physically migrated onto the organic matter during reaction (Fig. 4f). The alteration of chlorite was consistent with the observation via micro CT images that X-ray density decreased in chlorite, and Fe had apparently been leached (not shown) (Golab et al., 2015b). These observations are consistent with the observation of ion leaching and structural collapse of clays illite and chlorite after reaction under acidic conditions, (Baker et al., 1993; Brandt et al., 2003) and after CO 2 or CO 2 SO 2 O 2 brine reaction of reservoir rock core (Erickson et al., 2015; Renard et al., 2014) SEM-EDS pre reaction 1043m The fine-grained sandstone, 1043m, was light in colour with Fe-oxide laminations (Fig. 2). The 1043m sample has a coarser and more homogeneous grain size than981m (Fig. 5a), and feldspar, zircon and monazite grains are more abundant. K-feldspar and Ca-plagioclase (labradorite) grains in 1043m show some diagenetic alteration (Fig. 4c). A small amount of calcite cement was also present mixed with pore filling clays. Small pieces of dispersed organic matter (< 30 µm) were present (Fig. 6b), however 1043m was less organic-rich than 981m in agreement with the lower measured LOI (Table 1). SEM images of the 1043m polished discs pre-reaction show pores are observable in clays

13 and between colloidal silica grains (Fig 6). Pores were present between amorphous silica and chloritized grains (Fig. 6). Amorphous or colloidal silica was more prevalent in 1043m, and irregular shaped and slit shaped pores were observed between colloidal silica grains and illite, kaolinite, and chlorite (Fig. 6c,d, inset). Chlorite and kaolinite exhibited a pore filling morphology (Fig. 6c), and Illite-smectite appeared to rim pores (Fig. 6h) Plagioclase feldspar diagenetic alteration to clay leaving inter-grain contacts was also observed (Fig. 6a): Golab et al. (2015b) also observed this in other sections of this core SEM-EDS post reaction 1043mR K-feldspar, calcite and plagioclase in 1043mR showed corrode surfaces post reaction (Fig. 5c,d,e). Fine-grained Fe-rich clays, barite, and Ca-sulphate precipitated and covered mineral surfaces (Fig. 5 b,d,f), and some pre-existing mixed clays appeared altered after reaction (Fig. 6f). Precipitates on K- feldspar surfaces had a kaolinite morphology and appeared crystalline (Fig. 5d). Organic matter occasionally contained precipitated material or potentially migrated fines (Fig. 6g). Illite-smectite and kaolinite were observed both pre and post-reaction: illite-smectite appeared to rim pores prereaction (Fig. 6d) after reaction a clay, either illite-smectite or kaolinite appeared to also bridge pores (Fig. 6h). The observation of pore bridging clay post reaction in 1043mR could indicate it had precipitated in pores or physically moved during reaction. However it is possible that this morphology was present pre-reaction but not observed in SEM images of the surface MICP Porosity pre reaction

14 Mercury intrusion profiles for 1043m showed that the average pore throat radius was 8.05 nm. Pore throats up to 200 nm radius were measured in the sample, the majority of pore throats were in the ~ nm range, and the porosity of the sample was 5.45 % (Fig. 7). Sample 981m also contained pores with throat radius up to ~200 nm, however the average pore radius (8 nm), and majority of pore throats (~ nm) were smaller than 1043m. The measured porosity of 981m was higher (8.36 %) than 1043m. Mineralogically, 981m had a higher clay content and lower quartz content than 1043m (Table 1) and a unimodal pore throat distribution in the mesopore range, suggesting that the pores may be concentrated in the clay minerals and the inorganic matrix (Ross and Bustin, 2009). Total accessible porosities (calculated from He skeletal and Hg bulk densities using the method of Massarotto et al., 2010) were 17.1 and 24.1 % for 1043m and 981m respectively, potentially indicating the presence of micropores (Massarotto et al., 2010). Note that helium porosities in the range of % have been reported elsewhere for Evergreen Formation cores from different depths of the same well, indicating significant variability through the formation MICP post reaction Post-reaction the number of pore throats with radius up to 200 nm appeared to have decreased in 981mR (Fig. 7). The porosity determined by mercury intrusion was 1.62 % (verses 8.36 % for 981m), with a slightly higher average pore radius of 9.8 nm (versus 8 nm for 981m). The calculated total accessible porosity of 981mR was also lower after reaction at 12.4% (versus 24.1 % for 981m). Mercury intrusion data provide a good indication of pore throat sizes which control permeability. Since mercury intrusion is a destructive technique, the use of adjacent core material before and after reaction may have introduced some heterogeneity to the porosity changes observed. Some authors have noted a peak shift to smaller pore throat size or closing of pores in mercury intrusion

15 owing to grain compression (Clarkson et al., 2013; Kuila and Prasad, 2013). Additionally measured pore throats may be smaller than pore bodies. Kuila and Prasad associated illite pores in the size range of nm diameter with clay platelets, and ~ 3 nm diameter with clay interlayer spaces, and correlated micro and mesopores in shales with illite-smectite clays. More recently several authors have also noted the importance of pores in organic matter, especially in mature organic-rich gas bearing shales e.g. Marcellus or Barnett (Bahadur et al., 2015; Ruppert et al., 2013). The peak in pore throats at ~ 10 nm radius measured in sample 981m corresponds with the clay interlayer size expected for clays such as illite-smectite. Illite-smectite is a common component of many cap-rocks and a significant mineral phase in 981m. However, pores of ~10 nm radius could be present in organic matter and other minerals. Since the cap-rocks studied here are relatively clay-rich and organic matter lean and immature (relative to gas bearing shales) the majority of the porosity is likely associated with mineral matter (Dong et al., 2017). Ross and Bustin (2009) also associated the majority of nanometer size pores in low maturity clay-rich Jurassic shales with inorganic matter, mainly the clays illite, kaolinite and chlorite; whereas they associated the majority of pores with organic matter in thermally mature shales. Using the method of Rezaee et al., (2012), calculated air permeabilities corresponding to the measured mercury porosities were 13.9 md for 1043m, 20.6 md for 981m, and 4.3 md for 981mR after reaction (Rezaee et al., 2012). These values are however somewhat higher than expected for cap-rocks. Measured horizontal unstressed brine permeabilities of both 1043m and 981m were sub milidarcy (< 1 md), however, no measurable changes were observed in 1043mR after CO 2 -brine reaction (981mR could not be measured for technical reasons). Measured ambient air permeabilities have been reported elsewhere for 4 different depth sections of the Evergreen Formation: permeabilities range from md, much lower than the underlying Precipice Sandstone reservoir rock (Golab et al., 2015a). The pore throat size distributions in the clay-rich Evergreen Formation cap-rocks 1043m and 981m reported here are lower than those measured in

16 the underlying quartz rich Precipice Sandstone (not shown), consistent with the permeability measurements (Golab et al., 2015a), and indicating that the Evergreen Formation is a good seal SANS SANS profiles of 981m at ambient conditions showing scattering from all pores, and scattering from inaccessible pores (ZAC conditions) are shown in Fig. 8a. The scattering profiles for 981mR post impure (O 2 SO 2 CO 2 ) reaction at ambient and ZAC conditions are also shown. The samples 981m, 1043m and 981mR, 1043mR all contain pores accessible to CD 4 before and after reaction. Scattering intensity corresponding to all pores and inaccessible pores was lower for the reacted core samples of 981mR and 1043mR. The fraction of accessible pores (ɸ AC) in each sample as a function of Q was calculated using the method of Melnichenko et al., (2012) (Melnichenko et al., 2012; Ruppert et al., 2013). Data for Q < 0.01 A -1 (~< 10 nm pores) were disregarded owing to the effect of high pressure CD 4 condensation in pores. Stefanopoulos et al., (2017) have shown that the majority of pores <2.5 nm in shale may be inaccessible to CO 2 (Stefanopoulos et al., 2017). The fraction of accessible pores for each sample is shown in Figure 9, where the error bars correspond to the maximum error (assuming 5% error in the measurement and 5% error in determining the ZAC pressure). In 1043m, the fraction of accessible pores decreases gradually from ~0.95 at small Q (larger pore diameter) to ~0.75 at larger Q or smaller pore radius (Figure 9). A similar behaviour was observed for Barnett shale samples (Clarkson et al., 2013; Ruppert et al., 2013). After reaction in 1043mR the number of accessible pores had decreased slightly over all pore sizes with the fraction of accessible pores decreasing from ~0.90 to 0.75 with decreasing pore radius. The fraction of accessible pores in the clay and organic-rich 981m is overall slightly higher than in 1043m, decreasing to 0.9 as pore radius decreases (Figure 9a). After reaction, the fraction of accessible

17 pores decreased and Q increased across all of the pore sizes, with a greater change after reaction in 981mR than for 1043mR. The greatest reduction in the fraction of accessible pores for 981mR was at the smallest pore size which is consistent with the slight increase in average pore throat radius measured via MICP. After reaction, the fraction of accessible pores in 981mR decreased from ~0.95 to 0.85 as pore radius decreased. We have also observed a consistent decrease in accessible mesopores after impure CO 2 brine reaction (at higher temperature and pressure) of three clay-rich Roseneath-Murteree-Epsilon (REM) shales from the Cooper Basin, Australia, via SANS measurements (Pearce et al., in prep). Accessible porosity has a control on permeability (Dong et al., 2017). Other studies of CO 2 storage cap-rock or reservoir rock cores have characterised changes in accessible porosity or more often permeability before and after pure CO 2 and water or brine reaction by various techniques and over different pore size ranges. A study on two Polish cap-rocks and six reservoir rocks showed that porosity decreased in both cap-rocks (in agreement with our results) and five of the six measured reservoir rock cores after reaction with CO 2 -brine (Tarkowski and Wdowin, 2011). In addition they reported that the N 2 permeability of the cap-rocks showed no change before and after reaction at < 0.01 MD, but decreased in three reservoir rocks and increased in three reservoir rocks. Mouzakis et al. (2016) used water chemistry, SEM, and SANS with D 2 O as a contrast matching fluid to measure connected porosity in clay-rich Marine Tuscaloosa shale and carbonate-rich Gothic shale cap-rocks before and after reaction with CO 2 brine or brine (Mouzakis et al., 2016). They observed carbonate mineral dissolution and an increase in the number of connected < 200nm pores in the Gothic shale after brine reaction; however, connected porosity only slightly increased after CO 2 -brine reaction. SEM images showed carbonate mineral dissolution and precipitation of anhydrite in the CO 2 -brine reacted sample. In contrast, pore connectivity decreased in the clay-rich Marine Tuscaloosa shale after CO 2 brine reaction which was attributed to the precipitation of minerals, including anhydrite, in pore throats and to hydration and alteration of clays. The decrease in SANS pore connectivity of the clay-rich Marine Tuscaloosa shale after CO 2 -brine reaction is in agreement with our results. The

18 alteration of clays and mineral precipitation in the reacted Marine Tuscaloosa shale is in agreement with our SEM observations and water chemistry (described in section 3.4 below). Mouzakis et al. (2016) however measured an increase in pore connectivity in the Gothic Shale after CO 2 -brine reaction. This is likely owing to the high carbonate mineral content of 20-30% which dissolved during reaction increasing pore connectivity. In contrast 981m and 1043m contained < 1% carbonate content. This suggests cap-rocks with high quantities of reactive carbonate minerals could have an increase in pore connectivity after CO 2 -brine reaction. This demonstrates the need to test and understand cap-rock with a range of mineralogies. Several studies using fractured cores have also reported changes in porosity or permeability by CO 2 - brine reaction and fines migration. Calcite dissolution resulting in illite and chlorite detachment and migration increased the porosity on a μm scale but decreased the permeability of a fractured reservoir sandstone by pore throat clogging (Pudlo et al., 2015). A decrease in fracture permeability of a carbonate-rich cap-rock through calcite dissolution and dolomite fines migration has also been reported (Ellis et al., 2013). Calcite dissolution from a fractured marl cap-rock has also been reported to result in a porous zone (micron scale porosity observed via SEM) on CO 2 -brine reaction, although the permeability did not significantly change (Dávila et al., 2016). However, when reacted with sulphate rich brine and CO 2 the marl fracture permeability decreased through gypsum precipitation. A carbonate-rich rock from France showed a decrease in porosity from anhydrite precipitation after SO 2 -O 2 -CO 2 reaction, and clay alteration (Renard et al., 2014). In contrast, carbonate-rich cap-rock and reservoir core from the Netherlands showed an increase in permeability after reaction with SO 2 -CO 2 and brine due to strong carbonate mineral dissolution (Bolourinejad and Herber, 2015). In a different application, a permeability increase was observed in coal and attributed to calcite dissolution in coal cleats after reaction with HCl; however, permeability subsequently decreased in coals where kaolinite was also present in the cleats via kaolinite fines migration and pore clogging (Turner and Steel, 2016). These various changes in accessible porosity and permeability indicate that several factors including mineral content, organic matter content,

19 solution ph, the onset of precipitation, and migration of clays could all play a role in opening or closing pores. This points to the importance of combined geochemical and petrophysical testing of actual core samples of different mineralogy to improve our understanding and ability to predict longer term effects applicable to field injection scenarios Experimental water chemistry During experimental O 2 -SO 2 -CO 2 brine reactions of the Evergreen 981m and 1043m cap-rocks at simulated reservoir conditions, after O 2 -SO 2 -CO 2 gas injection, the solution ph initially decreased owing to the dissolution of the gases and generation of carbonic acid and also stronger sulphuric acid due to the dissolution of SO 2 into the brine. For reaction of 1043m ph initially decreased to 1.8 (sampled after 120 h of reaction) and subsequently increased to 2.3 (by the end of the experiment, 648 h), for 981m ph initially decreased to 1.92 and subsequently increased to 2.3, indicating ph buffering by mineral dissolution. Solution conductivity during reaction of 1043m increased to 7.8 ms/cm and subsequently decreased to 5.27 ms/cm; with 981m it increased to 6.42 ms/cm and then decreased to 3.29 ms/cm. The concentration of dissolved ions (Fe, Si, Al, Ca, Na, K, S) were measured in solution by ICP-OES after the O 2 -SO 2 -CO 2 gas injection. Unfiltered experiment waters were also analysed for Fe, Si and Al. As shown in Figure 10, the concentration of dissolved Fe, Si and Al (and others including Ca, Na, K, S) during reaction of the cap-rocks increased, indicating initial release by mineral dissolution of carbonates and silicates or clay ion exchange. Initial carbonate and silicate mineral dissolution is consistent with other short term experimental studies on CO 2 -brine reactivity of cores (with or without impurities) or CO 2 injection field studies, while some authors also observed kaolinite or siderite precipitation (Carroll et al., 2013; Farquhar et al., 2015; Horner et al., 2014; Kaszuba et al., 2005; Kharaka et al., 2006; Shevalier et al., 2013; Smith et al., 2013; Wdowin et al., 2014).

20 Subsequently dissolved ion concentrations from both cores 981m and 1043m decreased or stabilised (Fig. 10), indicating subsequent incorporation into the precipitated minerals observed via SEM. The subsequent precipitation of minerals including oxides, clays and sulphates is consistent with other experimental studies especially those including impurity gases SO 2 and O 2 in the injected gas stream (while some have also reported siderite precipitation) (Garcia et al., 2012; Palandri and Kharaka, 2005; Pearce et al., 2016b; Pearce et al., 2015a; Pearce et al., 2015b; Renard et al., 2014; Shao et al., 2014; Wilke et al., 2012). Several geochemical modelling studies have also predicted mineral precipitation over longer time scales with or without impurity gases including Ca-sulphate mineral precipitation and cap-rock self-sealing (Gaus et al., 2005; Gherardi et al., 2007; Tian et al., 2014; Xu et al., 2007). Decreases in cap-rock or low porosity formations have also been observed in natural analogue studies of CO 2 accumulation via kaolinite and carbonate mineral precipitation over geologic timescales (Higgs et al., 2013). Concentrations of several metals in the experiment waters were also determined without filtered to give total metals. The concentration of Fe was slightly higher indicating some clay fines (e.g. chlorite) may have been released from the rock, or that precipitates < 45μm were forming in solution during the experiments. 4. Conclusions A feldspar rich, and a clay and (relatively) organic matter rich cap-rock core from a well drilled for a CO 2 storage site feasibility study were reacted at reservoir PT conditions with impure CO 2 (containing SO 2 and O 2 ) and low salinity brine. Macro porosity was observed in clays and other mineral phases. Mineral dissolution and subsequent precipitation were observed after impure CO 2 -brine reaction, along with fines migration and clay alteration via SEM EDS and water chemistry.

21 The majority of pore throats were in the ~ nm radius range. After reaction the clayrich cap-rock porosity was measured (MICP and total accessible), porosity decreased and average pore size increased slightly. The fraction of accessible mesopores measured by SANS has a decreasing trend towards smaller pore size in both cap-rocks (i.e. less pores were open at the smaller pore size). After impure CO 2 brine reaction, the fraction of accessible SANS mesopores decreased in both cap-rocks. More pores were observed to become closed at the smaller SANS measured pore size (~10 nm). A decrease in accessible pore size is favourable for CO 2 storage cap-rock integrity. Future studies on the changes in accessible porosity after impure CO 2 reaction with a range of cap-rock core mineralogies including calcite cemented, calcite and clay cemented, and quartz rich cores are suggested. 5. Acknowledgments The authors acknowledge assistance from the UQ School of Earth and Environmental Sciences Environmental geochemistry laboratory, and we acknowledge the facilities, and the scientific and technical assistance, of the Australian Microscopy and Microanalysis Research Facility at the Centre for Microscopy and Microanalysis. Alison Law, Anastasia Dmyterko, and Dean Biddle are also thanked for technical assistance, along with Victor Rudolph for lab access. Luc Turner is thanked for providing a map. The research at Oak Ridge National Laboratory s High Flux Isotope Reactor was sponsored by the Laboratory Directed Research and Development Program and the Scientific User Facilities Division, Office of Basic Energy Sciences, U.S. Department of Energy. This research was supported in part by the ORNL Postdoctoral Research Associates Program, administered jointly by the ORNL and the Oak Ridge Institute for Science and Education. Part of this study was funded by a

22 UQ CIEF FirstLink Grant ( ), and a UQ new staff research start-up fund. This manuscript was improved by the comments of two anonymous reviewers. JKP would like to dedicate this manuscript to both Yuri, who is now scattering neutrons in the sky, and to my mother, for everything you did for me, may you both rest in peace. 6. References Bahadur, J., Radlinski, A.P., Melnichenko, Y.B., Mastalerz, M., Schimmelmann, A., Small -Angle and Ultrasmall-Angle Neutron Scattering (SANS/USANS) Study of New Albany Shale: A Treatise on Microporosity. Energy & Fuels 29, Baker, J.C., Uwins, P.J.R., Mackinnon, I.D.R., ESEM study of authigenic chlorite acid sensitivity in sandstone reservoirs. Journal of Petroleum Science and Engineering 8, Bolourinejad, P., Herber, R., Chemical effects of sulfur dioxide co-injection with carbon dioxide on the reservoir and caprock mineralogy and permeability in depleted gas fields. Appl. Geochem. 59, Brandt, F., Bosbach, D., Krawczyk-Bärsch, E., Arnold, T., Bernhard, G., Chlorite dissolution in the acid ph-range: a combined microscopic and macroscopic approach. Geochimica et Cosmochimica Acta 67, Carroll, S.A., McNab, W.W., Dai, Z., Torres, S.C., Reactivity of Mount Simon sandstone and the Eau Claire shale under CO2 storage conditions. Environmental Science and Technology 47, Chen, F., Turchyn, A.V., Kampman, N., Hodell, D., Gázquez, F., Maskell, A., Bickle, M., Isotopic analysis of sulfur cycling and gypsum vein formation in a natural CO2 reservoir. Chem. Geol. 436,

23 Chopping, C., Kaszuba, J.P., Supercritical carbon dioxide brine rock reactions in the Madison Limestone of Southwest Wyoming: An experimental investigation of a sulfur-rich natural carbon dioxide reservoir. Chem. Geol , Clarkson, C.R., Solano, N., Bustin, R.M., Bustin, A.M.M., Chalmers, G.R.L., He, L., Melnichenko, Y.B., Radliński, A.P., Blach, T.P., Pore structure characterization of North American shale gas reservoirs using USANS/SANS, gas adsorption, and mercury intrusion. Fuel 103, Curtis, M.E., Ambrose, R.J., Sondergeld, C.H., Structural Characterization of Gas Shales on the Micro- and Nano-Scales. Society of Petroleum Engineers, doi: / ms. Dávila, G., Luquot, L., Soler, J.M., Cama, J., Interaction between a fractured marl caprock and CO2-rich sulfate solution under supercritical CO2 conditions. Int. J. Greenh. Gas Control 48, Dawson, G.K.W., Biddle, D., Farquhar, S.M., Gao, J., Golding, S.D., Jiang, X., Keck, R., Khan, C., Law, A.C.K., Li, Q., Pearce, J.K., Rudolph, V., Watson, A., Xing, H., 2015a. Achieving Risk and Cost Reductions in CO2 Geosequestration through 4D Characterisation of Host Formations. University of Queensland, ANLEC R&D. Dawson, G.K.W., Pearce, J.K., Biddle, D., Golding, S.D., 2015b. Experimental mineral dissolution in Berea Sandstone reacted with CO2 or SO2-CO2 in NaCl brine under CO2 sequestration conditions. Chem. Geol. 399, Dong, T., Harris, N.B., Ayranci, K., Twemlow, C.E., Nassichuk, B.R., The impact of composition on pore throat size and permeability in high maturity shales: Middle and Upper Devonian Horn River Group, northeastern British Columbia, Canada. Marine and Petroleum Geology 81, Ellis, B., Fitts, J., Bromhal, G., McIntyre, D., Tappero, R., Peters, C., Dissolution-Driven Permeability Reduction of a Fractured Carbonate Caprock. Environmental Engineering Science 30, Erickson, K.P., Lempp, C., Pöllmann, H., Geochemical and geomechanical effects of scco2 and associated impurities on physical and petrophysical properties of Permotriassic Sandstones (Germany): an experimental approach. Environ. Earth Sci. 74,

24 Farquhar, S.M., Dawson, G.K.W., Esterle, J.S., Golding, S.D., Mineralogical characterisation of a potential reservoir system for CO2 sequestration in the Surat Basin. Australian Journal of Earth Sciences 60, Farquhar, S.M., Pearce, J.K., Dawson, G.K.W., Golab, A., Kirste, D., Biddle, D., Golding, S.D., A fresh approach to investigating CO2 storage: Experimental CO2-water-rock interactions in a freshwater reservoir system. Chem. Geol. 399, Garcia, S., Rosenbauer, R.J., Palandri, J., Maroto-Valer, M.M., Sequestration of non-pure carbon dioxide streams in iron oxyhydroxide-containing saline repositories. Int. J. Greenh. Gas Control 7, Gaus, I., Role and impact of CO2-rock interactions during CO2 storage in sedimentary rocks. Int. J. Greenh. Gas Control 4, Gaus, I., Azaroual, M., Czernichowski-Lauriol, I., Reactive transport modelling of the impact of CO2 injection on the clayey cap rock at Sleipner (North Sea). Chem. Geol. 217, Gherardi, F., Xu, T., Pruess, K., Numerical modeling of self-limiting and self-enhancing caprock alteration induced by CO2 storage in a depleted gas reservoir. Chem. Geol. 244, Glezakou, V.-A., Peter McGrail, B., Todd Schaef, H., Molecular interactions of SO2 with carbonate minerals under co-sequestration conditions: A combined experimental and theoretical study. Geochimica et Cosmochimica Acta 92, Golab, A., Arena, A., Khor, J., Goodwin, C., Young, B., Carnerup, A., Hussain, F., 2015a. Milestone 1.4 Final report of RCA and SCAL data on plugs from West Wandoan-1 Well. FEI Lithicon, Report for ANLEC R&D. Golab, A., Arena, A., Sommacal, S., Goodwin, C., Rajan, P., Dodd, N., Khor, J., Deakin, L., Zhang, J., Young, B., Carnerup, A., 2015b. Milestone 2.9: Final report of digital core analysis results for plug samples from West Wandoan-1 Well. FEI-Lithicon, Report for ANLEC R&D.

25 Golab, A., Knuefing, L., Goodwin, C., Sommacal, S., Carnerup, A., Dawson, G., Pearce, J.K., Golding, S.D., 2015c. Milestone 5.7: Final report on geochemical reactivity studies of core material using ScCO2. Lithicon FEI, Report for ANLEC R&D. Golab, A.N., Knackstedt, M.A., Averdunk, H., Senden, T., Butcher, A.R., Jaime, P., D porosity and mineralogy characterization in tight gas sandstones. The Leading Edge 29, Higgs, K.E., Funnell, R.H., Reyes, A.G., Changes in reservoir heterogeneity and quality as a response to high partial pressures of CO2 in a gas reservoir, New Zealand. Marine and Petroleum Geology 48, Higgs, K.E., Haese, R.R., Golding, S.D., Schacht, U., Watson, M., The Pretty Hill Formation as a natural analogue for CO2 storage; an investigation of mineralogical and isotopic changes associated with sandstones exposed to low, intermediate and high CO2 concentrations over geological time. Chem. Geol. 399, Hodgkinson, J., Grigorescu, M., Background research for selection of potential geostorage targets case studies from the Surat Basin, Queensland. Australian Journal of Earth Sciences 60, Horner, K.N., Schacht, U., Haese, R.R., Characterizing long term CO2-water-rock reaction pathways to identify tracers of CO2 migration during geological storage in a low-salinity siliciclastic reservoir system. Chem. Geol /j.chemgeo Kampman, N., Maskell, A., Chapman, H.J., Bickle, M.J., Evans, J.P., Purser, G., Zhou, Z., Gattacceca, J., Schaller, M., Bertier, P., Chen, F., Turchyn, A.S., Assayag, N., Rochelle, C., Busch, A., Drilling and sampling a natural CO2 reservoir: Implications for fluid flow and CO2-fluid-rock reactions during CO2 migration through the overburden. Chem. Geol. 369, Kaszuba, J.P., Janecky, D.R., Snow, M.G., Experimental evaluation of mixed fluid reactions between supercritical carbon dioxide and NaCl brine: Relevance to the integrity of a geologic carbon repository. Chem. Geol. 217,

26 Kaszuba, J.P., Navarre-Sitchler, A., Thyne, G., Chopping, C., Meuzelaar, T., Supercritical carbon dioxide and sulfur in the Madison Limestone: A natural analog in southwest Wyoming for geologic carbon-sulfur co-sequestration. Earth and Planetary Science Letters 309, Kaszuba, J.P., Yardley, B., Andreani, M., Experimental Perspectives of Mineral Dissolution and Precipitation due to Carbon Dioxide-Water-Rock Interactions. Reviews in Mineralogy and Geochemistry 77, Kellett, J.R., Radke, B.M., Ransley, T.R., J.G., B., Stewart, G.A., Chapter 5: hydrogeological framework, in: Smerdon, B.D., Ransley, T.R. (Eds.), Water Resource Assessment for the Surat Region, A report to the Australian Government from the CSIRO Great Artesian Basin Water Resource Assessment, CSIRO Water for a Healthy Country Flagship, Australia. Kharaka, Y.K., Cole, D.R., Hovorka, S.D., Gunter, W.D., Knauss, K.G., Freifeld, B.M., Gas-waterrock interactions in Frio Formation following CO2 injection: Implications for the storage of greenhouse gases in sedimentary basins. Geology 34, Knauss, K.G., Johnson, J.W., Steefel, C.I., Evaluation of the impact of CO2, co-contaminant gas, aqueous fluid and reservoir rock interactions on the geologic sequestration of CO2. Chem. Geol. 217, Kuila, U., Prasad, M., Specific surface area and pore-size distribution in clays and shales. Geophysical Prospecting 61, Lu, J., Mickler, P.J., Nicot, J.-P., Yang, C., Romanak, K.D., Geochemical impact of oxygen on siliciclastic carbon storage reservoirs. Int. J. Greenh. Gas Control 21, Massarotto, P., Golding, S.D., Bae, J.S., Iyer, R., Rudolph, V., Changes in reservoir properties from injection of supercritical CO2 into coal seams -- A laboratory study. International Journal of Coal Geology 82, Melnichenko, Y.B., He, L., Sakurovs, R., Kholodenko, A.L., Blach, T., Mastalerz, M., Radliński, A.P., Cheng, G., Mildner, D.F.R., Accessibility of pores in coal to methane and carbon dioxide. Fuel 91,

27 Mouzakis, K.M., Navarre-Sitchler, A.K., Rother, G., Bañuelos, J.L., Wang, X., Kaszuba, J.P., Heath, J.E., Miller, Q.R.S., Alvarado, V., McCray, J.E., Experimental Study of Porosity Changes in Shale Caprocks Exposed to CO2-Saturated Brines I: Evolution of Mineralogy, Pore Connectivity, Pore Size Distribution, and Surface Area. Environmental Engineering Science 33, Navarre-Sitchler, A.K., Cole, D.R., Rother, G., Jin, L., Buss, H.L., Brantley, S.L., Porosity and surface area evolution during weathering of two igneous rocks. Geochimica et Cosmochimica Acta 109, Palandri, J.L., Kharaka, Y.K., Ferric iron-bearing sediments as a mineral trap for CO2 sequestration: Iron reduction using sulfur-bearing waste gas. Chem. Geol. 217, Pearce, J.K., Dawson, G.K.W., Law, A.C.K., Biddle, D., Golding, S.D., 2016a. Reactivity of micas and cap-rock in wet supercritical CO2 with SO2 and O2 at CO2 storage conditions. Appl. Geochem. 72, Pearce, J.K., Golab, A., Dawson, G.K.W., Knuefing, L., Goodwin, C., Golding, S.D., 2016b. Mineralogical controls on porosity and water chemistry during O2-SO2-CO2 reaction of CO2 storage reservoir and cap-rock core. Appl. Geochem. 75, Pearce, J.K., Kirste, D.M., Dawson, G.K.W., Farquhar, S.M., Biddle, D., Golding, S., Rudolph, V., 2015a. SO2 Impurity Impacts on Experimental and Simulated CO2-Water-Reservoir Rock Reactions at Carbon Storage Conditions. Chem. Geol. 399, Pearce, J.K., Law, A.C.K., Dawson, G.K.W., Golding, S.D., 2015b. SO2-CO2 and pure CO2 reactivity of ferroan carbonates at carbon storage conditions. Chem. Geol /j.chemgeo Porter, R.T.J., Fairweather, M., Pourkashanian, M., Woolley, R.M., The range and level of impurities in CO2 streams from different carbon capture sources. Int. J. Greenh. Gas Control 36, Pudlo, D., Henkel, S., Reitenbach, V., Albrecht, D., Enzmann, F., Heister, K., Pronk, G., Ganzer, L., Gaupp, R., The chemical dissolution and physical migration of minerals induced during CO2 laboratory experiments: their relevance for reservoir quality. Environ. Earth Sci. 73,

28 Renard, S., Sterpenich, J., Pironon, J., Chiquet, P., Randi, A., Geochemical effects of an oxycombustion stream containing SO2 and O2 on carbonate rocks in the context of CO2 storage. Chem. Geol. 382, Rezaee, R., Saeedi, A., Clennell, B., Tight gas sands permeability estimation from mercury injection capillary pressure and nuclear magnetic resonance data. Journal of Petroleum Science and Engineering 88, Ross, D.J.K., Bustin, R.M., The importance of shale composition and pore structure upon gas storage potential of shale gas reservoirs. Marine and Petroleum Geology 26, Ruppert, L.F., Sakurovs, R., Blach, T.P., He, L., Melnichenko, Y.B., Mildner, D.F.R., Alcantar-Lopez, L., A USANS/SANS study of the accessibility of pores in the barnett shale to methane and water. Energy and Fuels 27, Shao, H., Kukkadapu, R.K., Krogstad, E.J., Newburn, M.K., Cantrell, K.J., Mobilization of metals from Eau Claire siltstone and the impact of oxygen under geological carbon dioxide sequestration conditions. Geochimica et Cosmochimica Acta 141, Shevalier, M., Nightingale, M., Mayer, B., Hutcheon, I., Durocher, K., Perkins, E., Brine geochemistry changes induced by CO2 injection observed over a 10 year period in the Weyburn oil field. Int. J. Greenh. Gas Control 16, Supplement 1, S160-S176. Smith, M.M., Sholokhova, Y., Hao, Y., Carroll, S.A., Evaporite Caprock Integrity: An Experimental Study of Reactive Mineralogy and Pore-Scale Heterogeneity during Brine-CO2 Exposure. Environ. Sci. Technol. 47, Stefanopoulos, K.L., Youngs, T.G.A., Sakurovs, R., Ruppert, L.F., Bahadur, J., Melnichenko, Y.B., Neutron Scattering Measurements of Carbon Dioxide Adsorption in Pores within the Marcellus Shale: Implications for Sequestration. Environmental Science and Technology 51, Talman, S., Subsurface geochemical fate and effects of impurities contained in a CO2 stream injected into a deep saline aquifer: What is known. Int. J. Greenh. Gas Control 40,

29 Tarkowski, R., Wdowin, M., Petrophysical and Mineralogical Research on the Influence of CO2 Injection on Mesozoic Reservoir and Caprocks from the Polish Lowlands. Oil Gas Sci. Technol. Rev. IFP Energies nouvelles 66, Tian, H., Xu, T., Wang, F., Patil, V.V., Sun, Y., Yue, G., A numerical study of mineral alteration and self-sealing efficiency of a caprock for CO2 geological storage. Acta Geotechnica 9, Turner, L.G., Steel, K.M., A study into the effect of cleat demineralisation by hydrochloric acid on the permeability of coal. Journal of Natural Gas Science and Engineering 36, Watson, M.N., Zwingmann, N., Lemon, N.M., The Ladbroke Grove-Katnook carbon dioxide natural laboratory: A recent CO2 accumulation in a lithic sandstone reservoir. Energy 29, Wdowin, M., Tarkowski, R., Franus, W., Determination of changes in the reservoir and cap rocks of the Chabowo Anticline caused by CO2 brine rock interactions. International Journal of Coal Geology 130, Wigley, M., Kampman, N., Dubacq, B., Bickle, M., Fluid-mineral reactions and trace metal mobilization in an exhumed natural CO2 reservoir, Green River, Utah. Geology 40, Wignall, G.D., Littrell, K.C., Heller, W.T., Melnichenko, Y.B., Bailey, K.M., Lynn, G.W., Myles, D.A., Urban, V.S., Buchanan, M.V., Selby, D.L., Butler, P.D., The 40 m general purpose small -angle neutron scattering instrument at Oak Ridge National Laboratory. Journal of Applied Crystallography 45, Wilke, F.D.H., Vásquez, M., Wiersberg, T., Naumann, R., Erzinger, J., On the interaction of pure and impure supercritical CO2 with rock forming minerals in saline aquifers: An experimental geochemical approach. Appl. Geochem. 27, Wunsch, A., Navarre-Sitchler, A.K., Moore, J., McCray, J.E., Metal release from limestones at high partial-pressures of CO2. Chem. Geol. 363, Xu, T.F., Apps, J.A., Pruess, K., Yamamoto, H., Numerical modeling of injection and mineral trapping Of CO2 with H2S and SO2 in a sandstone formation. Chem. Geol. 242,

30 Figure 1. Map of the West Wandoan 1 well in Queensland, Australia, and generalised stratigraphy. The West Wandoan 1 well location is shown in green. Figure 2. Photos of samples for SANS analysis (left) a) 1043mR. b) 1043m. c) 981m. d) 981mR. Right: The 981m sample in the high pressure SANS cell. Figure 3. SEM images of the unpolished cube surface of 981m and 981mR. a) 981m - fine grained layer (top) and course grained sandy layer (bottom), organic matter (org) and traces of carbonates and chlorite (Ch). Inset Illite-Smectite pre-reaction (50 µm image width). b) Sample 981mR (similar view in a) after reaction, view slightly shifted down) with corroded carbonates and fine-grained precipitates of Fe-oxides. c) Sample 981m - Fe-Mg chlorite/chloritized mica. d) Sample 981mR (similar view as c post-reaction) fine-grained Fe-containing precipitates and barite. e) 981mR - Ferich smectite-like clay precipitates. f) 981mR - precipitated gypsum. Figure 4: High resolution SEM images of 981m pores in the polished discs before reaction a-d: a) 981m - pore filling kaolinite (Ka), plagioclase (Pl), muscovite/ illite (Mu) and quartz (Qz). b) 981m - Organic matter (org). c) 981m - Plagioclase and chlorite (Ch). d) 981m - K-feldspar (K-F) and barite and apatite cement (Ba, Ap). Inset colloidal silica. 981mR after reaction e-h: e) 981mR - altered chlorite next to kaolinite. f) 981mR - Silica migrated grains, and Fe-oxide precipitates on organic matter. g) 981mR - organic matter and illite-smectite matrix. h) 981mR - Mg-containing illitesmectite.

31 Figure 5. SEM images of sample 1043m unpolished block pre and post-reaction with precipitates. a) 1043m - Surface view with bright zircon. b) 1043mR similar view in a) but post-reaction with the surface covered by fine precipitates. c) 1043m - K-feldspar grain pre-reaction. d) 1043mR - K-feldspar grain in c) post-reaction with surface corrosion and precipitated booklets. e) 1043m - labradorite with mixed calcite and clay surrounded by kaolinite pre-reaction. f) 1043mR - post-reaction surface view with Ca-sulphate precipitated on the surface. Figure 6: Images of 1043m polished disc before reaction a-d: a) 1043m - grains cemented by porefilling kaolinite and colloidal silica, with a bright zircon (Zr). b) 1043m - Organic matter. c) 1043m - Mixed chlorite clay and inset colloidal silica. d) 1043m - Colloidal silica and illite-smectite. 1043mR after reaction e -h. e) 1043mR - Quartz, plagioclase and silica, inset magnified amorphous silica cement. f) 1043mR - Organic matter and altered clay. g) 1043mR - Organic matter with silica grains. h) 1043mR - Amorphous silica with illite-smectite rimming and bridging pores. Figure 7. Log differential mercury intrusion with pore throat radius (displayed on a log scale) for virgin 1043m (red squares), virgin Evergreen 981m (blue circles), and post-reaction 981mR (green triangles). Figure 8. Small-angle neutron scattering profiles for a) 981m and 981mR, b) 1043m and 1043mR: Ambient condition, pre-reaction virgin sample shown in black, ambient post-reaction sample shown in red, ZAC pressure pre-reaction virgin sample shown in green, ZAC pressure post-reaction sample shown in blue.

32 Figure 9: Fraction of SANS accessible pores (ФAC) as a function of scattering vector for core samples pre-reaction (blue diamonds) and post reaction (R, red squares). a) 981m and 981mR. b) 1043m and 1043mR. From left to right the data correspond to pore radii of ~ nm. Error bars correspond to a maximum error as described in the text. Figure 10: Dissolved concentrations of Fe, Al and Si in waters from the reactors during O 2 -SO 2 -CO 2 - brine reactions of the two cap-rocks 981m and 1043m. Note the decreasing or stabilising concentrations after ~200h indicating mineral precipitation. Time zero corresponds to the concentrations after an initial N 2 -brine equilibration.

33 Figure 1

34 Figure 2

35 Figure 3

36 Figure 4

37 Figure 5

38 Figure 6

39 Figure 7

40 Figure 8

41 Figure 9

42 Figure 10

43 Graphical abstract