Applied Geochemistry (v.30, #C)

Geochemical Aspects of Geologic Carbon Storage by Katherine Romanak; Russell S. Harmon; Yousif Kharaka (1-3).

Hydrogeochemical and isotopic evidence for trans-formational flow in a sedimentary basin: Implications for CO2 storage by Fengtian Yang; Zhonghe Pang; Li Lin; Zhi Jia; Fengna Zhang; Zhongfeng Duan; Zhenhai Zong (4-15).
► We make a hydrogeochemical study on integrity of caprocks at regional scale. ► Groundwater origin, flow pattern, age and mixing processes are investigated. ► Good regional seal and trans-formational flow through caprocks are identified. ► Integrity of caprocks indicates suitability of saline aquifers for CO2 storage.Deep saline aquifers are considered as the most promising option for geologic disposal of CO2. One of the main concerns, however, is the integrity of the caprocks between and above the storage formations. Here, a hydrogeochemical and isotopic investigation is presented, using ionic chemistry, stable isotopes (δ18O, δ2H and 87Sr/86Sr) and radiocarbon dating, on five saline aquifers on a regional scale, namely: Neogene Minghuazhen, Guantao, Ordivician, Cambrian and Precambrian, all found in the Bohai Bay Basin (BBB) in North China. Groundwater recharge, flow pattern, age and mixing processes in the saline aquifers show that the Neogene Guantao Formation (Ng) in the Jizhong and Huanghua Depressions on both of the west and east sides of the Cangxian Uplift is a prospective reservoir for CO2 sequestration, with a well confined regional seal above, which is the clayey layers in the Neogene Minghuazhen Formation (Nm). However, this is not the case in the Cangxian Uplift, where the Ng is missing where structural high and fault zones are developed, creating strong hydraulic connections and trans-formational flow to the Nm aquifer. Comparing storage capacity and long-term security between the various hydrogeologic units, the depressions are better candidate sites for CO2 sequestration in the BBB.

Monitoring groundwater flow and chemical and isotopic composition at a demonstration site for carbon dioxide storage in a depleted natural gas reservoir by Patrice de Caritat; Allison Hortle; Mark Raistrick; Charlotte Stalvies; Charles Jenkins (16-32).
► The CO2CRC injected >65,000 tonnes of CO2-rich gas in a depleted gas reservoir. ► Groundwater levels and composition were monitored for about 5 years in two aquifers. ► Most hydrochemical parameters were statistically similar before and after injection. ► Where water composition did change, trends are not consistent with a CO2 addition. ► Groundwater quality was not measurably affected by the CO2 storage experiment.Between March 2008 and August 2009, 65,445 tonnes of ∼75 mol% CO2 gas were injected in a depleted natural gas reservoir approximately 2000 m below surface at the Otway project site in Victoria, Australia. Groundwater flow and composition were monitored biannually in two overlying aquifers between June 2006 and March 2011, spanning the pre-, syn- and post-injection periods. The shallower (∼0–100 m), unconfined, porous and karstic aquifer of the Port Campbell Limestone and the deeper (∼600–900 m), confined and porous aquifer of the Dilwyn Formation contain valuable fresh to brackish water resources. Groundwater levels in either aquifer have not been affected by the drilling, pumping and injection activities that were taking place, or by the rainfall increase during the project. In terms of groundwater composition, the Port Campbell Limestone groundwater is brackish (electrical conductivity = 801–3900 μS cm−1), cool (temperature = 12.9–22.5 °C), and near-neutral (pH = 6.62–7.45), whilst the Dilwyn Aquifer groundwater is fresher (electrical conductivity = 505–1473 μS cm−1), warmer (temperature = 42.5–48.5 °C), and more alkaline (pH = 7.43–9.35). Carbonate dissolution, evapotranspiration and cation exchange control the composition of the groundwaters. Comparing the chemical and isotopic composition of the groundwaters collected before, during and after injection shows no statistically significant changes; even if they were statistically significant, they are mostly not consistent with those expected if CO2 addition had taken place. The monitoring program reveals no impact on the groundwater resources attributable to the C storage demonstration project.

► A geochemical modeling was studied for CO2 geological storage. ► Three types of formation water composition were examined. ► Higher salinity made an arrival time of CO2 at an observation well earlier. ► Equilibrated water with a reservoir can estimate contributions of CO2 trapping. ► Actual formation water provided better understanding of chemical composition change.Carbon dioxide capture and storage (CCS) is one of the options for reducing CO2 emissions to the atmosphere from large-scale point sources. Stored CO2, initially present mainly in a supercritical state, reacts with formation water and minerals in saline aquifers and, some of it alters to dissolved CO2 (e.g., CO2aq, HCO 3 - ) and to mineral CO2 (e.g., CaCO3) over the years. Geochemical reactions play an important role in increasing the security of CO2 storage. Understanding the processes of CO2 trapping is important in evaluating the risks involved and in obtaining public acceptance of its geological storage. In this geochemical modeling study, formation water composition is used in sensitivity analyses of CO2 behavior in a saline aquifer. Three types of formation water compositions were examined; (1) a NaCl solution at an assumed salinity, (2) a solution in equilibrium with the minerals in a reservoir and (3) actual formation water collected from the reservoir. Higher salinity water resulted in an earlier arrival time of CO2 at an observation well. Equilibrated water was used to estimate contributions of CO2 trapping mechanisms in long-term predictions. The estimated proportion of trapped CO2 phases (gas, aqueous and mineral phase) was almost the same as when actual water was used for the initial water composition in the calculations. However, using actual formation water will provide better understanding of chemical composition changes, especially in short-term prediction.

Predictive modeling of CO2 sequestration in deep saline sandstone reservoirs: Impacts of geochemical kinetics by Victor N. Balashov; George D. Guthrie; J. Alexandra Hakala; Christina L. Lopano; J. Donald Rimstidt; Susan L. Brantley (41-56).
► Model baseline 8 components sandstone reservoir was developed. ► Reactive CO2 diffusion from supercritical phase into the sandstone pore brine was modeled. ► 97% of the maximum value (34.5 kg CO2 per m3 of sandstone) was sequestered by 4000 y. ► Varying mineral kinetics changes reaction paths to the equilibrium sequestration. ► To predict sequestration within 1000 y requires accurate Olg, Ab, Smct kinetic data.One idea for mitigating the increase in fossil-fuel generated CO2 in the atmosphere is to inject CO2 into subsurface saline sandstone reservoirs. To decide whether to try such sequestration at a globally significant scale will require the ability to predict the fate of injected CO2. Thus, models are needed to predict the rates and extents of subsurface rock–water–gas interactions. Several reactive transport models for CO2 sequestration created in the last decade predicted sequestration in sandstone reservoirs of ∼17 to ∼90 kg CO2 m−3. To build confidence in such models, a baseline problem including rock + water chemistry is proposed as the basis for future modeling so that both the models and the parameterizations can be compared systematically. In addition, a reactive diffusion model is used to investigate the fate of injected supercritical CO2 fluid in the proposed baseline reservoir + brine system. In the baseline problem, injected CO2 is redistributed from the supercritical (SC) free phase by dissolution into pore brine and by formation of carbonates in the sandstone. The numerical transport model incorporates a full kinetic description of mineral–water reactions under the assumption that transport is by diffusion only. Sensitivity tests were also run to understand which mineral kinetics reactions are important for CO2 trapping.The diffusion transport model shows that for the first ∼20 years (20 a) after CO2 diffusion initiates, CO2 is mostly consumed by dissolution into the brine to form CO2,aq (solubility trapping). From 20 to 200 a, both solubility and mineral trapping are important as calcite precipitation is driven by dissolution of oligoclase. From 200 to 1000 a, mineral trapping is the most important sequestration mechanism, as smectite dissolves and calcite precipitates. Beyond 2000 a most trapping is due to formation of aqueous HCO 3 - . Ninety-seven percent of the maximum CO2 sequestration, 34.5 kg CO2 per m3 of sandstone, is attained by 4000 a even though the system does not achieve chemical equilibrium until ∼25,000 a. This maximum represents about 20% CO2 dissolved as CO2,aq, 50% dissolved as HCO 3 ,aq - , and 30% precipitated as calcite. The extent of sequestration as HCO 3 - at equilibrium can be calculated from equilibrium thermodynamics and is roughly equivalent to the amount of Na+ in the initial sandstone in a soluble mineral (here, oligoclase). Similarly, the extent of trapping in calcite is determined by the amount of Ca2+ in the initial oligoclase and smectite. Sensitivity analyses show that the rate of CO2 sequestration is sensitive to the mineral–water reaction kinetic constants between approximately 10 and 4000 a. The sensitivity of CO2 sequestration to the rate constants decreases in magnitude respectively from oligoclase to albite to smectite.

Reactive transport simulation study of geochemical CO2 trapping on the Tokyo Bay model – With focus on the behavior of dawsonite by Yasuko Okuyama; Norifumi Todaka; Munetake Sasaki; Shuji Ajima; Chitoshi Akasaka (57-66).
► A TOUGHREACT simulation is carried out modeling CO2 saline aquifer storage. ► CO2 solubility trapping within the plume is followed by convective mixing. ► Carbonate rind is formed in the plume front by mixing of evolved and intact water. ► Dawsonite is a transient phase stably presents only under high CO2 saturation. ► Dawsonite does not fix CO2 permanently in the reservoir with low salinity water.A long-term (up to 10 ka) geochemical change in saline aquifer CO2 storage was studied using the TOUGHREACT simulator, on a 2-dimensional, 2-layered model representing the underground geologic and hydrogeologic conditions of the Tokyo Bay area that is one of the areas of the largest CO2 emissions in the world. In the storage system characterized by low permeability of reservoir and cap rock, the dominant storage mechanism is found to be solubility trapping that includes the dissolution and dissociation of injected CO2 in the aqueous phase followed by geochemical reactions to dissolve minerals in the rocks. The CO2–water–rock interaction in the storage system (mainly in the reservoir) changes the properties of water in a mushroom-like CO2 plume, which eventually leads to convective mixing driven by gravitational instability. The geochemically evolved aqueous phase precipitates carbonates in the plume front due to a local rise in pH with mixing of unaffected reservoir water. The carbonate precipitation occurs extensively within the plume after the end of its enlargement, fixing injected CO2 in a long, geologic period.Dawsonite, a Na–Al carbonate, is initially formed throughout the plume from consumption of plagioclase in the reservoir rock, but is found to be a transient phase finally disappearing from most of the CO2-affected part of the system. The mineral is unstable relative to more common types of carbonates in the geochemical evolution of the CO2 storage system initially having formation water of relatively low salinity. The exception is the reservoir-cap rock boundary where CO2 saturation remains very high throughout the simulation period.

A geochemical clogging model with carbonate precipitation rates under hydrothermal conditions by Seung-Youl Yoo; Yoshihiro Kuroda; Yoshitada Mito; Toshifumi Matsuoka; Masami Nakagawa; Akiko Ozawa; Kazutoshi Sugiyama; Akira Ueda (67-74).
► We propose a geochemical clogging model with filtration and advection for mineral trapping. ► Column tests were performed to evaluate precipitation rates and permeability in porous media. ► Filtration coefficient (λ) and fluid pore velocity (u) are applied to the numerical calculation. ► The smaller λ/u, the more uniform secondary mineral deposition. ► Controlling value of λ/u becomes the key parameter for sustainable CO2 fixation.A step-wise numerical calculation method was developed to provide predictions of when and where carbonate deposits might be found through reservoirs during CO2 sequestration. Flow experiments through porous media using a supersaturated carbonate fluid were also performed in order to observe flow rates. In order to evaluate precipitation rates and permeability change in the formation, calculated flow rates based on the proposed geochemical clogging model were compared with the experimentally observed data. Both high and low temperature cases were studied to understand how hydrothermal conditions can affect precipitation rates of carbonate. According to chemical kinetics, growth rates of minerals are generally proportional to the saturation index (S.I.) that depends on temperature. Thus, a supersaturated fluid has the advantage of improving the filtration and the amount of C fixation (σ). However, when the ratio of filtration coefficient (λ) to pore fluid velocity (u) increases, the permeability around the injection point tends to be significantly reduced by carbonate accumulation, and thus, this might result in insufficient injection of CO2. Therefore, it is essential to understand how to control both λ and u so that the precipitation of carbonate can be located as far away from the inlet as possible.

► Three batch experiments on alkali feldspars–CO2–brine interactions. ► Partial equilibrium not attained between secondary minerals and aqueous solutions. ► The form of rate laws affects the steady state feldspar dissolution rates.In order to evaluate the extent of CO2–water–rock interactions in geological formations for C sequestration, three batch experiments were conducted on alkali feldspars–CO2–brine interactions at 150–200 °C and 300 bars. The elevated temperatures were necessary to accelerate the reactions to facilitate attainable laboratory measurements. Temporal evolution of fluid chemistry was monitored by major element analysis of in situ fluid samples. SEM, TEM and XRD analysis of reaction products showed extensive dissolution features (etch pits, channels, kinks and steps) on feldspars and precipitation of secondary minerals (boehmite, kaolinite, muscovite and paragonite) on feldspar surfaces. Therefore, these experiments have generated both solution chemistry and secondary mineral identity. The experimental results show that partial equilibrium was not attained between secondary minerals and aqueous solutions for the feldspar hydrolysis batch systems. Evidence came from both solution chemistry (supersaturation of the secondary minerals during the entire experimental duration) and metastable co-existence of secondary minerals. The slow precipitation of secondary minerals results in a negative feedback in the dissolution–precipitation loop, reducing the overall feldspar dissolution rates by orders of magnitude. Furthermore, the experimental data indicate the form of rate laws greatly influence the steady state rates under which feldspar dissolution took place. Negligence of both the mitigating effects of secondary mineral precipitation and the sigmoidal shape of rate–ΔGr relationship can overestimate the extent of feldspar dissolution during CO2 storage. Finally, the literature on feldspar dissolution in CO2-charged systems has been reviewed. The data available are insufficient and new experiments are urgently needed to establish a database on feldspar dissolution mechanism, rates and rate laws, as well as secondary mineral information at CO2 storage conditions.

A novel high pressure column flow reactor for experimental studies of CO2 mineral storage by I. Galeczka; D. Wolff-Boenisch; T. Jonsson; B. Sigfusson; A. Stefansson; S.R. Gislason (91-104).
► Detailed description of a high pressure column flow reactor set-up. ► Description of sampling procedure, technical problems and potential improvements. ► Presentation of the plug testing with CO2–charged water and basaltic glass.The objective of this study was to design, build, and test a large scale laboratory high pressure column flow reactor (HPCFR) enabling experimental work on water–rock interactions in the presence of dissolved gases, demonstrated here by CO2. The HPCFR allows sampling of a pressurized gas charged liquid along the flow path within a 2.3 m long Ti column filled with either rock, mineral, and/or glass particles. In this study, a carbonated aqueous solution (1.2 M CO2(aq)) and basaltic glass grains was used. Given the pressure and temperature rating (up to 10 MPa at 90 °C) of the HPCFR, it can also be used with different gas and/or gas mixtures, as well as for supercritical fluid applications. The scale of the HPCFR, the possibility of sampling a reactive fluid at discrete spatial intervals under pressure, and the possibility of monitoring the evolution of the dissolved inorganic carbon and pH in situ all render the HPCFR unique in comparison with other columns constructed for studies of water–rock interactions. We hope that this novel experimental device will aid in closing the gap between bench scale reactor experiments used to generate kinetic data inserted into reactive transport models and field observations related to geological carbon sequestration. A detailed description and testing of the HPCFR is presented together with first geochemical results from a mixed H2O–CO2 injection into a basalt slurry whose solute concentration distribution in the HPCFR was successfully modelled with the PHREEQC geochemical computer code.

Laboratory experiments and field study for the detection and monitoring of potential seepage from CO2 storage sites by Giorgio Caramanna; Yang Wei; M. Mercedes Maroto-Valer; Paul Nathanail; Mike Steven (105-113).
► Detection of CO2 seepage in sediments and water. ► Laboratory experimental reactors for the study of CO2 impact on soil and water. ► Panarea Island as natural submarine CO2 seepage analogue by scientific divers. ► pH is the main parameter to be monitored in order to detect the presence of CO2. ► Acidification of soil and water is the main consequence of CO2 seepage.Potential CO2 seepages from geological storage sites or from the injection rig may affect the surrounding environment. To develop reliable detection techniques for such seepages a laboratory rig was designed that is composed of three vertical Plexiglas columns. The columns can be filled with sediments and water; CO2 can be injected from the bottom. Two columns are used to simulate the impact of CO2 on soils; while the third one, which is larger in size, simulates CO2 seepage in aquatic environments. The main results of the laboratory experiments indicate that increased levels of CO2 generate a quick drop in pH. Once the seepage is stopped, a partial recovery towards the initial values of pH is recorded. The outcomes of the laboratory experiments on the aquatic seepage are compared with observations from a submarine natural emission of CO2. In this natural underwater seepage multi-parametric probes and laboratory analysis were used to analyze the composition and the chemical effects of the emitted gas; basic acoustic techniques were tested as tools for the prompt detection of CO2 bubbles in water.

► We develop equipment for detecting and monitoring CO2 leakage from seafloor. ► We verify effectiveness of the equipment through sea-going observation. ► These equipment show high ability for detection and monitoring of leaked CO2. ► We propose a strategy for detecting and monitoring CO2 leakage in sub-seabed CCS.Carbon dioxide capture and storage (CCS) in sub-seabed geological formations is currently being studied as a potential option to mitigate the accumulation of anthropogenic CO2 in the atmosphere. To investigate the validity of CO2 storage in the sub-seafloor, development of techniques to detect and monitor CO2 leaked from the seafloor is vital. Seafloor-based acoustic tomography is a technique that can be used to observe emissions of liquid CO2 or CO2 gas bubbles from the seafloor. By deploying a number of acoustic tomography units in a seabed area used for CCS, CO2 leakage from the seafloor can be monitored. In addition, an in situ pH/pCO2 sensor can take rapid and high-precision measurements in seawater, and is, therefore, able to detect pH and pCO2 changes due to the leaked CO2. The pH sensor uses a solid-state pH electrode and reference electrode instead of a glass electrode, and is sealed within a gas permeable membrane filled with an inner solution. Thus, by installing a pH/pCO2 sensor onto an autonomous underwater vehicle (AUV), an automated observation technology is realized that can detect and monitor CO2 leakage from the seafloor. Furthermore, by towing a multi-layer monitoring system (a number of pH/pCO2 sensors and transponders) behind the AUV, the dispersion of leaked CO2 in a CCS area can also be observed. Finally, an automatic elevator can observe the time-series dispersion of leaked CO2. The seafloor-mounted automatic elevator consists of a buoy equipped with pH/pCO2 and depth sensors, and uses an Eulerian method to collect spatially continuous data as it ascends and descends.Hence, CO2 leakage from the seafloor is detected and monitored as follows. Step 1: monitor CO2 leakage by seafloor-based acoustic tomography. Step 2: conduct mapping survey of the leakage point by using the pH/pCO2 sensor installed in the AUV. Step 3: observe the impacted area by using a remotely operated underwater vehicle or the automatic elevator, or by towing the multi-layer monitoring system.

Tracers – Past, present and future applications in CO2 geosequestration by Matthew Myers; Linda Stalker; Bobby Pejcic; Andrew Ross (125-135).
► Chemical tracers are used in CCS to understand the behaviour of injected CO2. ► They can be used to assess plume migration, storage mechanisms and potential leaks. ► The high solubility and complex phase behaviour of CO2 are challenges for tracers. ► Tracer techniques complement geophysical measurements such as seismic. ► The subsurface behaviour of tracers is relatively unknown limiting their potential.Chemical tracers have been used in various C capture and storage (CCS) projects worldwide primarily to provide information regarding subsurface migration of CO2 and to verify CO2 containment. Understanding the movement and interactions of CO2 in the subsurface is a challenging task considering the variety of states in which it exists (i.e. gas, liquid, supercritical, dissolved in water) and the range of possible storage mechanisms (i.e. residual or capillary trapping, dissolved in water, structural trapping or incorporation into minerals). This paper critically reviews several chemical tracer applications and case studies for CCS projects. In many instances, there are parallels (e.g. tracer classes and applications) between tracers in the oil and gas industry and in CCS. It has been shown that chemical tracers can complement geophysical measurements (e.g. seismic) in understanding the formation behaviour of CO2. Although tracers have been successfully used in many CCS projects, some fundamental information, for example partitioning and adsorption, about the behaviour of tracers is still lacking and this can be an issue when interpreting tracer data (e.g. determining leakage rates). In this paper the deployment and recovery of chemical tracers and their use on various CCS projects are described.

CO2 leakage impacts on shallow groundwater: Field-scale reactive-transport simulations informed by observations at a natural analog site by Elizabeth H. Keating; J. Alexandra Hakala; Hari Viswanathan; J. William Carey; Rajesh Pawar; George D. Guthrie; Julianna Fessenden-Rahn (136-147).
► Use natural analog for studying CO2 impacts on groundwater. ► Transport simulations reproduce key field geochemical data. ► Water quality impacts may be local in extent but persistent.It is challenging to predict the degree to which shallow groundwater might be affected by leaks from a CO2 sequestration reservoir, particularly over long time scales and large spatial scales. In this study observations at a CO2 enriched shallow aquifer natural analog were used to develop a predictive model which is then used to simulate leakage scenarios. This natural analog provides the opportunity to make direct field observations of groundwater chemistry in the presence of elevated CO2, to collect aquifer samples and expose them to CO2 under controlled conditions in the laboratory, and to test the ability of multi-phase reactive transport models to reproduce measured geochemical trends at the field-scale. The field observations suggest that brackish water entrained with the upwelling CO2 are a more significant source of trace metals than in situ mobilization of metals due to exposure to CO2. The study focuses on a single trace metal of concern at this site: U. Experimental results indicate that cation exchange/adsorption and dissolution/precipitation of calcite containing trace amounts of U are important reactions controlling U in groundwater at this site, and that the amount of U associated with calcite is fairly well constrained. Simulations incorporating these results into a 3-D multi-phase reactive transport model are able to reproduce the measured ranges and trends between pH, pCO2, Ca, total C, U and Cl at the field site. Although the true fluxes at the natural analog site are unknown, the cumulative CO2 flux inferred from these simulations are approximately equivalent to 37.8E−3 MT, approximately corresponding to a .001% leak rate for injection at a large (750 MW) power plant. The leakage scenario simulations suggest that if the leak only persists for a short time the volume of aquifer contaminated by CO2-induced mobilization of U will be relatively small, yet persistent over 100 a.

A review of continuous soil gas monitoring related to CCS – Technical advances and lessons learned by S. Schloemer; M. Furche; I. Dumke; J. Poggenburg; A. Bahr; C. Seeger; A. Vidal; E. Faber (148-160).
► We present a review of continuous soil gas monitoring. ► Achievements are based on the world wide first continuous monitoring program. ► Guidelines and recommendations are given.One of the most vigorously discussed issues related to Carbon Capture and Storage (CCS) in the public and scientific community is the development of adequate monitoring strategies. Geological monitoring is mostly related to large scale migration of the injected CO2 in the storage formations. However, public interest (or fear as that) is more related to massive CO2 discharge at the surface and possible affects on human health and the environment. Public acceptance of CO2 sequestration will only be achieved if secure and comprehensible monitoring methods for the natural habitat exist. For this reason the compulsory directive 2009/31/EG of the European Union as well as other international regulations demand a monitoring strategy for CO2 at the surface. The variation of CO2 emissions of different soil types and vegetation is extremely large. Hence, reliable statements on actual CO2 emissions can only be made using continuous long-term gas-concentration measurements. Here the lessons learned from the (to the authors’ knowledge) first world-wide continuous gas concentration monitoring program applied on a selected site in the Altmark area (Germany), are described.This paper focuses on the authors’ technical experiences and recommendations for further extensive monitoring programs related to CCS. Although many of the individual statements and suggestions have been addressed in the literature, a comprehensive overview is presented of the main technical and scientific issues. The most important topics are the reliability of the single stations as well as range of the measured parameters. Each selected site needs a thorough pre-investigation with respect to the depth of the biologically active zone and potential free water level. For the site installation and interface the application of small drill holes is recommended for quantifying the soil gas by means of a closed circuit design. This configuration allows for the effective drying of the soil gas and avoids pressure disturbance in the soil gas. Standard soil parameters (humidity, temperature) as well as local weather data are crucial for site specific interpretation of the data. The complexity, time and effort to handle a dozen (or even more) single stations in a large case study should not be underestimated. Management and control of data, automatic data handling and presentation must be considered right from the beginning of the monitoring. Quality control is a pre-condition for reproducible measurements, correct interpretation and subsequently for public acceptance. From the experience with the recent monitoring program it is strongly recommended that baseline measurements should start at least 3 a before any gas injection to the reservoir.

► Carbonation proceeds via a dissolution–precipitation mechanism. ► Carbonated cement has heterogeneous porosity at a micro-scale. ► Initially-formed rinds of carbonate minerals may hinder complete cement carbonation.It is crucial that the engineered seals of boreholes in the vicinity of a deep storage facility remain effective for considerable timescales if the long-term geological containment of stored CO2 is to be effective. These timescales extend beyond those achievable by laboratory experiments or industrial experience. Study of the carbonation of natural Ca silicate hydrate (CSH) phases provides a useful insight into the alteration processes and evolution of cement phases over long-timescales more comparable with those considered in performance assessments. Samples from two such natural analogues in Northern Ireland have been compared with samples from laboratory experiments on the carbonation of Portland cement. Samples showed similar carbonation reaction processes even though the natural and experimental samples underwent carbonation under very different conditions and timescales. These included conversion of the CSH phases to CaCO3 and SiO2, and the formation of a well-defined reaction front. In laboratory experiments the reaction front is associated with localised Ca migration, localised matrix porosity increase, and localised shrinkage of the cement matrix with concomitant cracking. Behind the reaction front is a zone of CaCO3 precipitation that partly seals porosity. A broader and more porous/permeable reaction zone was created in the laboratory experiments compared to the natural samples, and it is possible that short-term experiments might not fully replicate slower, longer-term processes. That the natural samples had only undergone limited carbonation, even though they had been exposed to atmospheric CO2 or dissolved HCO 3 - in groundwater for several thousands of years, may indicate that the limited amounts of carbonate mineral formation may have protected the CSH phases from further reaction.

► Assessing the potential consequences of CO2 leakage to freshwater resources. ► Batch-reaction experiment with isotopic monitoring program. ► Reactive behavior of the CO2 combining both mixing after dissolution and CO2–water–rock interactions. ► The multi-isotope approach (δ13C, 87Sr/86Sr, δ11B, etc.) show changes in isotopic ratios with time.The assessment of the environmental impacts of CO2 geological storage requires the investigation of potential CO2 leakages into fresh groundwater, particularly with respect to protected groundwater resources. The geochemical processes and perturbations associated with a CO2 leak into fresh groundwater could alter groundwater quality: indeed, some of the reacting minerals may contain hazardous constituents, which might be released into groundwater. Since the geochemical reactions may occult direct evidence of intruding CO2, it is necessary to characterize these processes and identify possible indirect indicators for monitoring CO2 intrusion. The present study focuses on open questions: Can changes in water quality provide evidence of CO2 leakage? Which parameters can be used to assess impact on freshwater aquifers? What is the time scale of water chemistry degradation in the presence of CO2? The results of an experimental approach allow selecting pertinent isotope tracers as possible indirect indicators of CO2 presence, opening the way to devise an isotopic tracing tool.The study area is located in the Paris Basin (France), which contains deep saline formations identified as targets by French national programs for CO2 geological storage. The study focuses on the multi-layered Albian fresh water aquifer, confined in the central part of the Paris Basin a major strategic potable groundwater overlying the potential CO2 storage formations. An experimental approach (batch reactors) was carried out in order to better understand the rock–water–CO2 interactions with two main objectives. The first was to assess the evolution of the formation water chemistry and mineralogy of the solid phase over time during the interaction. The second concerned the design of an isotopic monitoring program for freshwater resources potentially affected by CO2 leakage. The main focus was to select suitable environmental isotope tracers to track water rock interaction associated with small quantities of CO2 leaking into freshwater aquifers.In order to improve knowledge on the Albian aquifer, and to provide representative samples for the experiments, solid and fluid sampling campaigns were performed throughout the Paris Basin. Albian groundwater is anoxic with high concentrations of Fe, a pH around 7 and a mineral content of 0.3 g L−1. Macroscopic and microscopic solid analyses showed a quartz-rich sand with the presence of illite/smectite, microcline, apatite and glauconite. A water–mineral–CO2 interaction batch experiment was used to investigate the geochemical evolution of the groundwater and the potential release of hazardous trace elements. It was complemented by a multi-isotope approach including δ13CDIC and 87Sr/86Sr. Here the evolution of the concentrations of major and trace elements and isotopic ratios over batch durations from 1 day to 1 month are discussed. Three types of ion behavior are observed: Type I features Ca, SiO2, HCO3, F, PO4, Na, Al, B, Co, K, Li, Mg, Mn, Ni, Pb, Sr, Zn which increased after initial CO2 influx. Type II comprises Be and Fe declining at the start of CO2 injection. Then, type III groups element with no variation during the experiments like Cl and SO4. The results of the multi-isotope approach show significant changes in isotopic ratios with time. The contribution of isotope and chemical data helps in understanding geochemical processes involved in the system. The isotopic systems used in this study are potential indirect indicators of CO2–water–rock interaction and could serve as monitoring tools of CO2 leakage into an aquifer overlying deep saline formations used for C sequestration and storage.