The geochemical effects of O2 and SO2 as CO2 impurities on fluid-rock reactions in a CO2 storage reservoir
Introduction
Global warming related to the release of greenhouse gases and particularly CO2 into the atmosphere has in recent years been a subject of major worldwide concern (Cox et al., 2000, IPCC, 2007, Manabe and Wetherald, 1980). One approach of mitigating this problem is to capture and safely store large amounts of CO2 in geological reservoirs, thereby reducing CO2 emissions and the negative impacts arising from climate change (Oelkers and Cole, 2008, Tol, 2009, Williamson, 2016). Carbon Capture and Storage (CCS) has been recognised and recommended as one technology that can significantly reduce CO2 emissions while continuing to pursue the use of fossil fuel resources (IPCC, 2005, Oelkers and Cole, 2008, Schrag, 2009). Key impediments for commercial-scale CCS projects are the costs of carbon capture and uncertainties relating to the long-term containment of CO2 within a designated geological storage reservoir. CO2 captured from fossil fuel power plants contains impurities, such as N2, NOx, Ar, O2 and SO2 of varying concentration (IEAGHG, 2011, Porter et al., 2015). Co-injection of these impurities might have an impact on well injectivity, wellbore integrity, cap rock integrity, CO2 containment and the water and mineral composition in the reservoir. The removal of CO2 impurities from captured CO2 is costly (DiPietro et al., 2011), thus there would be significant reductions in CO2 capture cost if CO2 impurities could be co-injected.
The impact of SO2 and O2 at concentrations of 50–2900 ppm and 45–5200 ppm, respectively, on the performance of geological CO2 storage have been examined in previous geochemical simulations (Gunter et al., 2000, Knauss et al., 2005, Koenen et al., 2011, Waldmann and Rütters, 2016, Wang et al., 2012, Wang et al., 2011). These works have put emphasis on the formation of sulphuric acid (H2SO4) due to the co-injection of SOx, which can lower the pH of the reservoir water and subsequently induce dissolution of minerals such as carbonates (Gunter et al., 2000, Knauss et al., 2005, Koenen et al., 2011, Wang et al., 2011). Coinciding with this acid production, the precipitation of secondary minerals such as anhydrite and gypsum have been reported (Gunter et al., 2000, Knauss et al., 2005, Koenen et al., 2011, Waldmann and Rütters, 2016, Wang et al., 2011), thus reducing porosity in a reservoir and well injectivity (Waldmann and Rütters, 2016, Wang et al., 2011). Additionally, modelling has shown that the co-injection of O2 may cause the dissolution of pyrite and siderite (Koenen et al., 2011) in a siliciclastic reservoir and decrease the CO2 trapping capacity due to a decrease in the density of the CO2 mixture (Wang et al., 2012, Wang et al., 2011).
The influence of SO2 and O2, at concentrations of 100–19895 ppm and 500–8000 ppm, respectively, have also been investigated in laboratory experiments using single minerals (such as hematite) (Garcia et al., 2012, Palandri et al., 2005); reservoir rocks or cap rocks (Cantrell et al., 2012, Erickson et al., 2015, Jung et al., 2013, Renard et al., 2011, Renard et al., 2014, Wilke et al., 2012); and coupled with reactive transport modelling (Bolourinejad and Herber, 2015, Corvisier et al., 2013, Pearce et al., 2015). These studies confirmed the dissolution of silicates and carbonates (Bolourinejad and Herber, 2015, Corvisier et al., 2013, Pearce et al., 2015, Renard et al., 2011, Renard et al., 2014, Wilke et al., 2012) and release of cations (Cantrell et al., 2012, Jung et al., 2013) as predicated in the geochemical simulations mentioned above. It has also been reported that SO2 could disproportionate to form H2S and H2SO4 acid, which resulted in the reductive dissolution of iron-bearing minerals (e.g., hematite) and formation of siderite (Garcia et al., 2012, Palandri et al., 2005). Occasionally, sulphur-bearing minerals (pyrite, elemental S and anhydrite) were observed as the products of SO2 disproportionation or oxidation (Corvisier et al., 2013, Palandri et al., 2005, Pearce et al., 2015). In addition, O2 was reported to cause oxidative dissolution of pyrite leading to the precipitation of sulphate bearing minerals (e.g., gypsum) (Cantrell et al., 2012, Jung et al., 2013, Renard et al., 2011, Renard et al., 2014).
Despite efforts to study the effect of SO2 and O2 in laboratory experiments and in geochemical models only field experiments can demonstrate the actual conditions and processes under in situ conditions (Wei et al., 2015). Indeed, O2 has been injected (as part of a CO2 or air stream) into reservoirs to enhance oil recovery in the past few decades (Taber, 1985). It was reported that injected O2 caused issues with material corrosion and it was possibly consumed by oxidation of hydrocarbons in the reservoir (Taber, 1985). Interestingly, no oxidative dissolution of minerals was identified (Taber, 1985). However, it is noted that pyrite oxidative dissolution might occur due to water flooding in enhanced oil recovery projects, and this would produce SO42− (Hutcheon, 1998). The presence of SO42− could enhance the activity of sulphate reducing bacterial and cause sour reservoirs (Gieg et al., 2011, Hutcheon, 1998, Vance and Thrasher, 2005). In cases of CO2 storage field experiments, O2 was co-injected at a concentration <7% into pyrite-containing Rousse reservoir, but geochemical effects relating to the O2 were not reported (Monne and Jammes, 2015). The impact of the co-injected O2 in the field study, however, was evaluated using geochemical models that simulate the field experiments or data from laboratory experiments, (Chiquet et al., 2013, Corvisier et al., 2013, Girard et al., 2013, Monne and Jammes, 2015, Renard et al., 2011, Renard et al., 2014). These modelling and laboratory studies suggested that injection of O2 and SO2 can cause dissolution of minerals (e.g., carbonates and pyrite) and the formation of sulphate-bearing minerals (e.g., anhydrite and barite) (Chiquet et al., 2013, Corvisier et al., 2013, Girard et al., 2013, Monne and Jammes, 2015, Renard et al., 2011, Renard et al., 2014).
Recently, Wei et al. (2015) conducted a field study investigating the geochemical impact of O2 and N2 in a shallow and open aquifer at a depth of 180–250 m in Tongliao, China. While these conditions are not representative for a CO2 storage reservoir, the study identified the oxidative dissolution of pyrite due to the co-injection of O2 and highlighted the generated SO42− as an indicator of CO2 and O2 migration.
In this study we investigate the impact of O2 and SO2 as CO2 impurities (from an oxyfuel system as a part of the Callide Oxyfuel project) on mineral-fluid interactions in a siliciclastic reservoir at the CO2CRC Otway site, Otway Basin (Australia). Two push-pull experiments (Otway Stage 2B Extension Project, 2BX) were carried out. In first test (Test 1) CO2-saturated water, with trace amounts of impurities in the CO2 (16, 0, <0.1 and 5 ppm vol/vol for NOx, N2 SO2 and O2, respectively), was injected and allowed to react with the reservoir rock for up to three weeks. The second test (Test 2) was conducted in a similar manner with the concentrations of impurities are 9, 1100, 67 and 6150 ppm vol/vol for NOx, N2, SO2 and O2, respectively. Dissolved impurities hereafter are referred to as co-injected gases. In addition, Test 1 and Test 2 will be referred to as tests conducted without and with CO2 impurities, respectively. The test without CO2 impurities (Test 1) is defined as the case of which CO2 impurities have insignificant concentrations (less than 20 mg/L). Contrarily, the test with CO2 impurities (Test 2) is defined as the case of which CO2 impurities have concentrations of higher than 20 mg/L. N2 was present in Test 2 but its geochemical impact on mineral-fluid interactions has not been observed (IEAGHG, 2011). In addition, the experiment included the co-injection of NO2 at a very low concentration, but the results are not discussed here due to three reasons. First, the injected concentration of NOx in Test 2 (CO2 with impurities) was 9 ppm, much smaller than 16 ppm in Test 1 (CO2 without impurities). This made it difficult to evaluate the impact of this impurity. Second, measured concentration of NO3− in back-produced water was ≤1 mg/L, well within the range of NO3− created by the oxidation of NOx in both tests (1.2 and 0.6 mg/L, respectively). Third, it was intended to inject 50 ppm of NO2, but there was an apparent loss of this impurity during transport, possibly oxidation in storage vessels or partitioning from the liquid CO2 within the storage vessel. Water was back-produced and sampled to determine changes in water composition and associated geochemical processes in the reservoir. A set of tracers (fluorescein, Na+, Cl−, Br−, Sr2+ and Li+) were added to the injection water to be able to trace the mixing of injection and formation water. Aqueous speciation, redox state and reaction pathway modelling were used to assist data interpretation.
It is noted that the Otway 2BX data set has been split into three parts in order to address three separate questions:
- 1.)
Results on the tracer performance were presented in Black et al. (2017) to evaluate the suitability of different soluble tracers (fluorescein, Na+, Cl−, Br−, Sr2+ and Li+) in geological reservoirs. It was showed that Br− and Li+ performed very well, and they can be used as conservative tracers in siliciclastic reservoirs to monitor water mixing (Black et al., 2017).
- 2.)
Data on Ca2+, total Fe and alkalinity were presented in a previous work (Vu et al., 2017) to deal with the impact of changes in formation water composition (due to storage and treatment at surface), and the implications for the re-injection of this produced water. Several factors, including an increase in pH, decreases in temperature and pressure (from 60 °C and 140 bar to ambient temperature and pressure), a change in redox state (from reducing to oxidising), residence time in holding tanks and the circulation of the produced water within the holding tanks caused precipitation of carbonates, silicates and iron oxides/hydroxides. Consequently, the stored water became depleted (with respect to several elements such as Fe and Ca2+), and subsequently undersaturated or further undersaturated with respect to some minerals such as siderite and dolomite. The re-injection of the Fe depleted water, therefore, led to dissolution of Fe-bearing minerals (e.g., siderite) to restore existing equilibrium (Vu et al., 2017).
- 3.)
Data on Mg2+, K+, Na+, Si, SO42− and total dissolved solid are reported for the first time in this manuscript to interpret the impact of dissolved CO2 impurities on fluid-mineral reactions.
Section snippets
Field experiments
Push-pull tests (Otway Stage 2B Extension Project) were carried out using the CRC-2 well at the CO2CRC Otway site (Fig. 1). Initially, over 500 t of formation water was produced and stored at surface in three holding tanks. 100.2 and 102.0 t of O2 depleted water were mixed with 5.20 t of CO2 (Test 1) and 4.52 t of CO2 with impurities (NOx, N2, SO2 and O2, Test 2) at a depth of 846 m through a gas mandrel at a ratio to achieve CO2-saturation under reservoir conditions. The CO2-saturated water was
Tracer results
The concentrations of Br− and Li+ in back-produced water in Test 1 and Test 2 decreased at different levels (Supplementary Material Fig. 1). The tracers are expected to behave conservatively, specifically the injection water concentration is expected to be observed in the first 40 t of back-produced water (Haese et al., 2013). The decrease in concentration of tracers indicates that the injection waters were diluted by the formation water. In other words, the back-produced water was a mixture of
Fluid-mineral reactions promoted by dissolved CO2
The consistency between the measured and estimated (by the Br− mixing model) K+, Si, SO42− and TDS suggests that there was not any or very minor acid-induced dissolution of reactive minerals (such as K-feldspar) in the reservoir as the result of the CO2-enriched water injection. However, within the first 25 t of production, the measured Mg2+ (and Ca2+ and total alkalinity, see also Vu et al., 2017) exceeded the expected values, suggesting that carbonates such as dolomite and calcite were
Conclusions
A field experiment in a siliciclastic reservoir was carried out in order to study changes in water composition and respective geochemical processes relating to co-injection of SO2 and O2 as CO2 impurities. During the experiment, CO2 with 6150 ppm O2 and 67 ppm SO2 was dissolved in the injection water. The dissolution and oxidation of the SO2 led to small increases in dissolved SO42− concentration. However, the impact of SO2 on mineral dissolution reactions was found to be minimal due to the high
Acknowledgment
Thanks to graduate and undergraduate students Syed Anas Ali, Cesar Castaneda Herrera, Scott Ooi for their help conducting the fieldwork. The authors thank Lawrence Berkeley National Laboratory for access to the U-tube system, the site operator (Upstream Production Solutions), Mr Rajindar Singh (CO2CRC Ltd.) and Dr Chris Spero (COSPL) for their invaluable logistical and on site support. We acknowledge funding provided by the CO2CRC Ltd and Callide Oxyfuel Services Pty Ltd (COSPL) as a part of
References (74)
- et al.
An experimental study of calcite and limestone dissolution rates as a function of pH from −1 to 3 and temperature from 25 to 80 °C
Chem. Geol.
(1998) - et al.
Modeling carbon dioxide sequestration in saline aquifers: significance of elevated pressures and salinities
Fuel Process. Technol.
(2005) - et al.
Rates of mineral dissolution under CO2 storage conditions
Chem. Geol.
(2015) - et al.
Aqueous phase tracers for monitoring fluid mixing in geological reservoirs: results from two field studies
Int. J. Greenh. Gas Control
(2017) - et al.
Chemical effects of sulfur dioxide co-injection with carbon dioxide on the reservoir and caprock mineralogy and permeability in depleted gas fields
Appl. Geochem.
(2015) - et al.
Development of scaling parameters to describe CO2–rock interactions within Weyburn-Midale carbonate flow units
Int. J. Greenh. Gas Control
(2013) - et al.
Geochemical assessment of the injection of CO2 into Rousse depleted gas reservoir – part II: geochemical impact of the CO2 injection
Energy Procedia
(2013) - et al.
Kinetic geochemistry comparative study of the kinetics and mechanisms of dissolution of carbonate minerals
Chem. Geol.
(1989) - et al.
Impact of co-injected gases on CO2 storage sites: geochemical modeling of experimental results
Energy Procedia
(2013) - et al.
Short-term effects of impurities in the CO2 stream injected into fractured carbonates
Int. J. Greenh. Gas Control
(2016)