ALBERT

Description: Core samples of the Upper Triassic (Keuper) Stuttgart Formation from the Ketzin pilot storage site have been experimentally treated with CO2. The major objective of the experimental program was to investigate the effects of long-term CO2 exposure on the mineralogical and petrophysical properties of the target formation.

Description: In order to characterize the chemical reactivity and hence the mineralogical CO2 sequestration potential of carbonate minerals, laboratory fluidsiderite/ankerite CO2 exposure experiments have been performed. Experiments were conducted on a hydrothermal rocking autoclave system equipped with flexible Titanium cells allowing for isobaric and isothermal fluid sampling. Crushed siderite/ankerite separates have grain sizes between 100 and 200 μm. The minerals were reacted together with pure CO2 and 2 M NaCl brine at 80 °C and 20 or 30 MPa, respectively; run durations were one week. Approximately 120 ml NaCl brine and 5.5 g of powdered carbonate separate were filled into the Titanium cells to yield a brine to mineral weightratio of 20 to 1. Experiments were supersaturated with CO2 during entire run durations. Successive fluid samples were taken after several different time steps to determine time dependent dissolution behavior. Solid samples were recovered upon completion of each run and analyzed by XRD as well as SEM. Fluid sample compositions were determined by ICPAES.

Description: In order to investigate and characterize single fluid-mineral interactions we successfully implemented a new hydrothermal laboratory. CO2-exposure experiments using separates of rock-forming minerals were performed on a hydrothermal rocking autoclave. The system is equipped with flexible Titanium cells allowing for isobaric sampling. Experiments were run for one week at 80°C and 20 MPa/30 MPa. Rietveld refined XRD data reveal that the initial siderite separate is composed of 69.6±1.3 wt% siderite, 26.7±1.2 wt% ankerite and 3.8±0.8 wt% quartz, respectively. 0ver time, siderite abundances increase and ankerite abundances correspondingly decrease, while quartz abundances are constant within error. Fluid data show rapid increases for Ca2+, Mg2+, Mn2+ and Fe2+. After these rapid increases, Ca2+ and Mg2+ reveal slight decreases that are followed by subsequent rises to maximum concentrations at the end of the experiments, while Mn2+ and Fe2+ decrease continuously after the initial maxima. SEM micrographs of CO2-exposed samples indicate dissolution of ankerite, while siderite and quartz are mainly unaffected. The experiments on the siderite separate clearly show that ankerite is dissolved and siderite is stable. We conclude that siderite is a potential CO2 trapping phase in iron-bearing reservoirs.

Description: Two sets of laboratory fluid-mineral experiments have been performed in order to analyze the effect of SO2-NO2 impurities in theCO2 stream on the chemical reactivity of (A) siderite and (B) illite.

Description: The injection of CO2 into deep saline aquifers is the most promising strategy for the reduction of CO2 emissions to the atmosphere via long-term geological storage. The study is part of the CO2 SINK project conducted at Ketzin, situated 40 km west of Berlin. There, food grade CO2 has been pumped into the Upper Triassic Stuttgart Formation since June 2008. The main objective of the experimental program is to investigate the effects of long-term CO2 exposure on the physico-chemical properties of the reservoir rock.

Description: In order to investigate CO2-brine-rock interactions occurring at the Ketzin pilot storage site, core samples of the siliciclastic reservoir rock were exposed to pure CO2 and synthetic reservoir brine at simulated in-situ P-T conditions of 5 MPa and 40 °C. Autoclaves were opened and rock and fluid samples taken after 15, 21, 24 and 40 months, respectively. The samples were analysed for mineralogical and chemical composition and compared to baseline data of untreated samples. XRD data with Rietveld refinement show decreasing weight percentages for analcime, chlorite, hematite and illite. While plagioclase as well as K-feldspar both do not reveal a coherent trend over time, quartz exhibits increasing weight percentages in the same interval. On freshly broken rock fragments corrosion textures were found on plagioclase, K-feldspar and anhydrite surfaces of CO2-treated samples. BSE images of the respective samples indicate (intensified) alterations of feldspar minerals. EMPA data display a change in plagioclase composition from intermediate to sodium-rich and albite endmember compositions during CO2 exposure. Compared to the synthetic brine used for the experiments, sodium, magnesium and chloride concentrations increased slightly, while potassium, calcium and sulfate concentrations significantly increased. Potassium and calcium even exceed reservoir brine concentration levels. Experimental observations were reproduced using the reactive geochemical modeling code Phreeqc-2. The mineralogical and geochemical measurements imply preferred dissolution of calcium out of plagioclase next to dissolution of K-feldspar and anhydrite. Petrophysical data show tendentially increasing porosities and permeabilities that also suggest mineral dissolution during the experiments. Due to the heterogeneous character of the Stuttgart Formation it is often difficult to distinguish between natural, lithostratigraphic variability and CO2-related changes. Assuming thermodynamic equilibrium preliminary reactive geochemical modeling of the observed CO2-fluid-rock interactions shows that the measured evolution of fluid composition is consistent with precipitation of albite and dissolution of anhydrite and illite, respectively. In a next step, kinetic data have to be included into the model to determine changes over time. Based on experimental data, the integrity of the Ketzin reservoir is not significantly affected by CO2.

Description: Sandstone samples of the Stuttgart Formation at Ketzin have been experimentally treated with CO2 and synthetic reservoir brine in high-quality steel autoclaves at simulated in situ P–T conditions (5.5 MPa, 40 °C). In order to observe mineralogical changes induced by CO2, untreated samples are compared to CO2-treated ones. Most samples show an analogous petrography of mainly quartz and plagioclase. Heterogeneities are related to minor mineral phases, such as K-feldspar, hematite, muscovite, biotite, illite, chlorite and opaque phase(s). These are attributed to the variability of the fluvial reservoir sandstone. The samples are weakly consolidated. Analcime, dolomite and anhydrite are only found as cement phases. During the experiments dissolution of anhydrite and plagioclase is observed. SEM micrographs of CO2-treated samples show corrosion textures on mineral surfaces of intermediate plagioclase, as well as precipitation of euhedral albite crystals. Overall, the data indicate preferred dissolution of calcium-rich plagioclase, K-feldspar and anhydrite and stabilization or precipitation of albite.

Description: Rock core samples of the Upper Triassic Stuttgart Formation from the Ketzin pilot CO2 storage site were exposed to pure CO2 and synthetic reservoir brine at simulated reservoir P-T conditions of 5 MPa and 40 °C. Autoclaves were opened and samples were taken after 15, 21, 24 and 40 months, respectively. The samples were analysed mineralogically and geochemically and compared to baseline data of untreated samples. XRD analyses with Rietveld refinement show no significant mineralogical changes for the studied intervals. On freshly broken rock fragments of the CO2-treated samples, corrosion textures were found on plagioclase, K-feldspar and anhydrite surfaces. BSE images of the respective twin samples show (intensified) alterations of feldspar minerals. EMPA data display a change in plagioclase composition from intermediate to almost pure albite endmember compositions after CO2 exposure. Inorganic fluid data show, besides others, highly increased calcium, potassium and sulfate concentrations [1]. The experimental observations were reproduced using the reactive geochemical modeling code PHREEQC. The mineralogical-chemical measurements imply preferred dissolution of calcium out of plagioclase next to dissolution of K-feldspar and anhydrite. Due to the heterogeneous character of the Stuttgart Formation, which formed in a fluvial environment [2], it is often difficult to distinguish between natural variability and CO2-related changes. Additional data is needed to interconnect the indicated changes during the experiments and to better understand CO2-brine-rock interaction occurring within the Ketzin reservoir.

Description: To evaluate mineralogical-geochemical changes within the reservoir of the Ketzin pilot CO2 storage site in Brandenburg, Germany, two sets of laboratory experiments on sandstone and siltstone samples from the Stuttgart Formation have been performed. Samples were exposed to synthetic brine and pure CO2 at experimental conditions and run durations of 5.5 MPa/40 °C/40 months for sandstone and 7.5 MPa/40 °C/6 months for siltstone samples, respectively. Mineralogical changes in both sets of experiments are generally minor making it difficult to differentiate natural variability of the whole rock samples from CO2-induced alterations. Results of sandstone experiments suggest dissolution of the anorthite component of plagioclase, anhydrite, K-feldspar, analcime, hematite and chlorite + biotite. Dissolution of the anorthite component of plagioclase, anhydrite and K-feldspars is also observed in siltstone experiments. In an inverse modeling approach, an extensive set of equilibrium simulations was set up in order to reproduce the experimental observations of the sandstone experiments. Simulations generally show fairly good matches with the experimental observations. Best matches with measured brine data are obtained from mineral combinations of albite, analcime, anhydrite, dolomite, hematite, illite, and kaolinite. The major discrepancies during equilibrium modeling, however, are reactions involving Fe2+ and Al3+. The best matching subsets of the equilibrium models were finally run including kinetic rate laws. These simulations reveal that experimentally determined brine data was well matched, but reactions involving K+ and Fe2+ are not fully covered. The modeling results identified key primary minerals as well as key chemical processes, but also showed that the models are not capable of covering all possible contingencies.