Carbon and oxygen isotope indications for CO2 behaviour after injection: First results from the Ketzin site (Germany)

https://doi.org/10.1016/j.ijggc.2010.02.005Get rights and content

Abstract

Stable carbon isotopes may serve as useful tracers to monitor the fate and migration of injected CO2 gas in the subsurface. This is true as long as the already present dissolved inorganic carbon (DIC) has a different isotopic composition13CDIC) than the injected CO2 (δCCO213). After dissolution and turnover of the injected CO2 to DIC, mixture of both sources (δCCO213 and δ13CDIC) enables isotope and mass balance calculations to quantify the degree of dissolution of the injected CO2 (geochemical trapping). Furthermore, δ13CDIC may be used as a sensitive tool to determine its source, as well as to indicate the arrival of CO2 at observation wells. Isotope measurements from before and after injection of CO2 at the Ketzin site near Berlin are presented here. Before injection bacterial activity, especially in the injection well, was observed. After injection, most of the DIC present originates from injected CO2 and, generally, an inverse trend exists between the expected increase in DIC contents and its corresponding, decreasing isotope signal. The stable isotopic composition of the water (δOH2O18), however, remained relatively homogeneous throughout the monitoring period. In order to shift the δOH2O18 signal, large amounts of CO2 are expected to be required. In order to verify the results, and to see how the observed trends evolve, further monitoring, including more isotope measurements combined with further biogeochemical data, will be necessary.

Introduction

Reducing CO2 emissions into the atmosphere has become a topic of increasing attention in recent years (Clemens and Wit, 2002). One of the proposed methods of doing so is by CO2 injection into deep aquifers (Bachu, 2000, Metz et al., 2005). The most promising methods include CO2 injection into either depleted hydrocarbon reservoirs or saline aquifers (Gunter et al., 2004, Zhang et al., 2007). Monitoring this process is of major importance in order to understand, amongst others:

  • storage volumes and efficiencies,

  • long-term behaviour of the injected CO2 in geological formations as well as to monitor its safety.

Isotopes have proven to be powerful tracers of crustal fluid processes, such as mineral and solubility trapping that act on subsurface CO2 (Gilfillan et al., 2009). So far few isotope data are available on tests where CO2 is actively injected into the subsurface (Assayag et al., 2009, Emberley et al., 2004, Emberley et al., 2005, Gilfillan et al., 2009, Kharaka et al., 2006a, Kharaka et al., 2006b, Johnson et al., 2009, Raistrick et al., 2006). This manuscript focuses on stable isotope monitoring of δ13CDIC and δOH2O18 in conjunction with geochemical data from the European pilot CO2 storage project, CO2SINK (CO2 Storage by Injection into a Natural Saline Aquifer at Ketzin) in NE-Germany.

The Ketzin site is the first on-shore CO2 injection pilot test site in Europe. CO2 injection began in June 2008 (Borm and Förster, 2005, Schilling et al., 2009, Würdemann et al., in this issue). It is planned to inject a maximum of 60,000 t of CO2. The data presented here are based on the first sampling campaigns from before and during injection.

Chemical processes contributing to CO2 storage include dissolution of CO2 and interaction with the surrounding geosphere (Metz et al., 2005). Solubility trapping occurs when injected CO2 dissolves in in situ fluid to form carbonic acid (H2CO3) and ionic trapping when HCO3 is formed (Eq. (1)). Solubility and ionic trapping, termed here as geochemical trapping, depend on factors including pressure, water–CO2 ratios, salinity, temperature and pH. The formation of bicarbonate (HCO3) according to Eq. (1), is expected to be the dominant form of storage within 50 years after injection (Raistrick et al., 2006), whereas mineral trapping, the fixing of CO2 as carbonates, for example by anorthite dissolution (Bachu et al., 1994) (Eq. (2)), is expected to show significant effects only beyond 500 years (Xu et al., 2002).CO2(aq) + H2O = H2CO3 = H+ + HCO3CaA12Si2O8 + CO2 + 2H2O = Al2Si2OH5(OH)4 + CaCO3

From surface water and near surface groundwater studies we know that isotope and geochemical data of dissolved inorganic carbon (DIC) can be used to distinguish between their various sources, including the atmosphere, soil respiration and carbonates (Barth et al., 2003). Such isotope information can also be applied to CO2 injection in order to distinguish between already present and injected inorganic carbon (Raistrick et al., 2006). This approach is effective if the isotopic value of the injected CO2 differs from that of the dissolved inorganic carbon (δ13CDIC) that is present in the aquifer. Injected CO2, which is formed during combustion of former C3 plant material is expected to have δCCO213 value ranging between −23 and −33‰ (Sharp, 2007). The baseline δ13C value of DIC of an aquifer, on the other hand, is expected to be around −10‰ (Clark and Fritz, 1997), although it may vary depending on the isotope composition of naturally available CO2 in the subsurface and the ratio between carbonate and silicate dissolution. For instance marine carbonates usually have δ13C values of ∼0‰ (Clark and Fritz, 1997). Thus, dissolution of CO2 that originates from combustion decreases the initial δ13CDIC value of the groundwater. δ13CDIC that derives from silicate mineral reactions (Eq. (2)) depends solely on the δ13C value of the CO2 source (Clark and Fritz, 1997). If sufficient information of the isotope values of the end members is available, isotope and mass balance calculations can be used to quantify geochemical trapping of CO2 as DIC in groundwater (Raistrick et al., 2006).

Other factors that may be affected by CO2 injection and that may influence the DIC concentration and δ13CDIC value of the fluid, include carbonate mineral dissolution, for example in the form of calcium carbonate (Eq. (3)), and dissimilatory sulfate reducing bacteria (SRB) that metabolise organic carbon (Eq. (4)) (Emberley et al., 2005, Raistrick et al., 2006). The magnitudes of these two reactions are determined by the carbonate concentration, organic matter content and bacterial conditions of the aquifer. The dissolution of carbonates will cause an increase in DIC concentrations and an isotope signal shift towards more positive values (δCHCO3130), where as SRB would cause an isotope shift towards more negative values (δCHCO31320to30) (Emberley et al., 2005, Raistrick et al., 2006).H2CO3 + CaCO3 = Ca2+ + 2HCO3SO42− + 2CH2O = 2HCO3 + H2S

All these reactions may have played a role, before, during and after the injection of CO2 at the Ketzin site and typical isotope fingerprints can be expected as a response. Here we present δ13CDIC data from the Ketzin site. Our objectives were to test the validity of the isotope method and to define isotope and mass balance calculations of geochemical trapping. These results are helpful to find out the percentage of CO2 dissolution at the Ketzin site and to outline frequency and design of sampling campaigns in future monitoring programs, at this and other sites with similar plans of CO2 injection.

Section snippets

Geological setting/materials and methods

The Ketzin site is located in the northeast German Basin (NEGB) and lies on the eastern part of the Rostow-Ketzin anticline. Triassic (Buntsandstein, Muschelkalk and Keuper) and lower Jurassic formations constitute the structurally deformed overburden above a salt pillow (Förster et al., 2006).

At Ketzin, CO2 is injected into the 80 m thick and lithologically heterogeneous Stuttgart Formation of the Middle Keuper (upper Triassic) (Förster et al., 2006). The caprock is approximately 210 m thick and

Results

δ13C values of the injected CO2 gas varied between −26.6 and −30.2‰ (Table 1), with a mean value of −28.3‰. The δOCO218 values varied between −8.7 and −29.3‰, with a mean value of −29.2‰. Baseline δ13CDIC values of the three wells ranged between −4.2 and −20.1‰, DIC concentrations were between 54 and 155 mg L−1 and δOH2O18 values ranged between −5.2 and −5.6‰, at sampling depths from 625 to 760 m bgs (Table 2).

Table 3 presents the monitoring data (δ13CDIC, DIC and δOH2O18) of the wells. After the

Discussion

Whereas the baseline δOH2O18 values (∼−5.5‰) did not show significant changes with depth in any of the wells, the δ13CDIC baseline data of the three wells were not homogeneous. A negative δ13CDIC shift with depth was observed in all three wells, though to a much greater degree in injection well Ktzi 201, where δ13CDIC values ranged from −8.4 to −20.1‰ at 647 to 720 m depth, respectively (Fig. 1). At the observation wells, values ranged between −4.2 and −9.2‰ at comparable depths. At Ktzi 202,

Conclusions

At the CO2 injection pilot test area in Ketzin, stable isotopes (δ13CDIC and δOH2O18) in conjunction with geochemical data were used to monitor the behaviour of injected CO2 in the subsurface. The first 9 months of injection at the Ketzin site showed rapid stable isotope and chemical interaction of carbon between the injected CO2 and the reservoir water. Oxygen isotope shifts of the formation water after CO2 injection, however, could not be observed and may require larger masses of CO2. Isotope

Acknowledgements

We would like to thank the CO2SINK group for supplying and permitting us to run samples from the CO2SINK project. We are indebted to Peter Pilz and Mashal Alawi from the Geoforschungszentrum Potsdam for their help during fieldwork. Invaluable help was provided during laboratory analyses by Bernd Steinhilber, Annegret Walz, Ellen Struve and Tom Wendel. Further constructive advice was provided by Peter Grathwohl at the University of Tübingen. We would also like to thank the two anonymous

References (34)

  • D. Morozova et al.

    Monitoring of the microbial community composition in saline aquifers during CO2 sequestration by fluorescence in situ hybridisation

    International Journal of Greenhouse Gas Control

    (2010)
  • B. Prevedel et al.

    The CO2SINK boreholes for geological CO2-storage testing

    Energy Procedia

    (2009)
  • F. Schilling et al.

    Status report on the First European on-shore CO2 storage site at Ketzin (Germany)

    Energy Procedia

    (2009)
  • B. Wiese et al.

    Hydraulic Aquifer Characterisation of the Ketzin Test Site

    International Journal of Greenhouse Gas Control

    (2010)
  • T. Zhang et al.

    Experimental investigation on the carbon isotope fractionation of methane during gas migration by diffusion through sedimentary rocks at elevated temperature and pressure

    Geochimica et Cosmochimica Acta

    (2001)
  • Y. Zhang et al.

    System-level modeling for economic evaluation of geological CO2 storage in gas reservoirs

    Energy Conversion and Management

    (2007)
  • N. Assayag et al.

    Water–rock interactions during a CO2 injection field-test: implications on host rock dissolution and alteration effects

    Chemical Geology

    (2009)
  • Cited by (48)

    • Assessing geochemical reactions during CO<inf>2</inf> injection into an oil-bearing reef in the Northern Michigan basin

      2019, Applied Geochemistry
      Citation Excerpt :

      The difference between the gas composition in the Charlton 19 Reef (dominated by alkanes) and the Dover 33 Reef (dominated by CO2) suggests that the geochemistry of the Dover 33 reef has been dominated by the long-term injection of CO2 that occurred for approximately 17 years prior to the current study. The concentration and isotopic composition of DIC can be used to evaluate the extent of CO2 dissolution into the brine and subsequent water-rock interactions (Becker et al., 2011; Mayer et al., 2015; Myrttinen et al., 2010). Five fluid samples from L-M 1–33, L-M 2–33, and L-M 5–33 wells, obtained in October and November of 2012 (before the start of the injection tests), and L-M 2–33 swab samples from October and December 2013, were analyzed several times under different conditions to determine δ13C of DIC.

    • The inherent tracer fingerprint of captured CO<inf>2</inf>

      2017, International Journal of Greenhouse Gas Control
    • Tracking the interaction between injected CO<inf>2</inf> and reservoir fluids using noble gas isotopes in an analogue of large-scale carbon capture and storage

      2017, Applied Geochemistry
      Citation Excerpt :

      A wide range of geochemical tracing techniques have been developed to track the movement and storage of injected CO2 (Humez et al., 2014), yet there is an outstanding need to develop a robust tool which can resolve the amount of CO2 dissolved into the formation water within a storage site. The isotopic composition of CO2 (δ13Cand δ18O) has been widely used to trace injected gas and track its interaction with reservoir rock and formation water as well as constrain the mechanism of storage (e.g. Assayag et al., 2009; Johnson et al., 2011, 2014; Lu et al., 2012b; Myrttinen et al., 2010; Serno et al., 2016; Raistrick et al., 2006). Both δ13C and δ18O measurements have also been used as a leak detection tool (e.g. Mayer et al., 2015).

    View all citing articles on Scopus
    View full text