- Email: [email protected]
Further Insights Into Interconnections between the Shallow and Deep Systems from a Natural CO2 Reservoir Near Springerville, Arizona, U.S.A.
Available online at www.sciencedirect.com
ScienceDirect Energy Procedia 63 (2014) 3195 – 3201
GHGT-12
Further insights into interconnections between the shallow and deep systems from a natural CO2 reservoir near Springerville, Arizona, U.S.A. Elizabeth Keatinga, Dennis Newellb, Brian Stewartc , Rosemary Capoc, and Rajesh Pawara a
Earth and Environmental Sciences Division, MS T003, Los Alamos National Laboratory, Los Alamos, NM 87545 b Department of Geology, Utah State University, Logan UT, USA 84322-4505 c Department of Geology and Planetary Science, University of Pittsburgh, Pittsburgh, PA, 15260 USA. ,
Abstract
If carbon dioxide (CO2) sequestration into deep geologic reservoirs is to be accepted by the public and environmental regulators, the possibility of upward leakage into shallow groundwater should be acknowledged and those processes well-understood. Studies of natural CO2 reservoirs and their connection (or lack thereof) with the shallow subsurface is one way to explore these issues. A natural reservoir near Springerville, Arizona, U.S.A. has leaked CO2 to the surface along a fault zone for thousands of years, creating large travertine deposits. In recent times, the CO2 leak rates have declined significantly yet the shallow aquifer is still highly enriched in dissolved CO2. In previous studies, using water level data and simulations we demonstrate that the fault zone provides hydrologic communication between the shallow aquifer and the deeper reservoir. It is reasonable to assume, therefore, that the source of the CO2 in wells completed within the fault zone is the deeper CO2 reservoir. We present water chemistry data to demonstrate the geochemical impact of this CO2 on shallow groundwater quality. Interestingly, arsenic concentrations are elevated, but other trace metals concentrations are not. Previous studies [1,2,3] showed that CO2 originating from magmatic and deep crustal origin is migrating upward along the fault and dissolving into the shallow groundwater. Saline waters are also mixing, to a much lesser degree, but their source was unknown. Here we present strontium isotope data that clearly shows the source to be water/rock interactions with reservoir rocks, which include Paleozoic carbonates, evaporites and shale units. © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
© 2013 The Authors. Published by Elsevier Ltd. (http://creativecommons.org/licenses/by-nc-nd/3.0/). peer-review under GHGT. of GHGT-12 Selection Peer-reviewand under responsibility of responsibility the Organizing of Committee
1876-6102 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Peer-review under responsibility of the Organizing Committee of GHGT-12 doi:10.1016/j.egypro.2014.11.344
3196
Elizabeth Keating et al. / Energy Procedia 63 (2014) 3195 – 3201
Keywords: shallow groundwater impacts of CO2; brine migration
1. Introduction Studies of natural CO2 reservoirs that are leaking to shallow aquifers and the surface provide opportunities to study CO2 leakage at temporal and spatial scales that are difficult if not impossible to achieve with engineered controlled release studies. The natural CO2 reservoir at St. Johns Dome in Arizona, USA has been leaking CO2 to the surface for thousands of years, creating large travertine deposits. At present, travertine deposition has ceased and measurable CO2 flux at ground surface is negligible; however, CO2 continues to enrich the shallow aquifer and springs that discharge to the Little Colorado River. This site has been used to study the physical mechanisms of CO2 leakage and geochemical impacts of this leakage on shallow aquifers [1,2,3]. Insights into CO2 trapping mechanisms in have also been gained here [4]. One of the potential risks to shallow groundwater from geologic CO2 sequestration is brine leakage [4]. Studies of other natural analog sites have shown that saline waters have the potential to be much more damaging to fresh water quality than CO2 alone [5]. Given the stratigraphy of the St. John’s Dome reservoir, which includes marine carbonate and evaporite sequences, it is reasonable to assume that saline water is present at depth associated with these formations. Although there is geochemical evidence of saline water mixing with shallow groundwater at this site [2], the impact of this mixing on groundwater quality of relatively minor. Multi-phase flow simulations [1] showed that the physical characteristics of the fault zone system at this site (Coyote Wash fault) could explain significant upward transport of CO2 without displacing the saline water/fresh water interface. However, additional data would be required to further test this hypothesis. Here we summarize the key elements of previous studies of the site and the significance to geologic sequestration. We also present new strontium isotope data from the site, which we use to constrain the source of the saline water. 2. Previous work relevant to potential leakage at carbon sequestration sites Allis et al [7] developed a conceptual and numerical model of CO2 outflow from the St. Johns Dome reservoir that reasonably-well reproduced the depositional history of travertine along the Coyote Wash Fault Zone. Their simulations showed a transient overpressurization of the reservoir, caused by a pulse of CO2 from depth, leading to a rise in the water table elevation and broad area of water and CO2 discharge. A soil CO2 flux survey showed no evidence of present-day CO2 flux at ground surface. Groundwater chemistry data [8] clearly shows Figure 1. Location of sampled wells (stars), springs, and travertine evidence of CO2 outflow from the reservoir in the deposits (yellow shading). Blue lines indicate the piezometric surface dissolved phase, and the authors conclude that in the shallow aquifer. “dissolved CO2 could be the main source of CO2 leakage from a sequestration site.” Both noble gases [3] and stable isotope ratios of carbon [2] have been used to trace the dissolved carbon dioxide in the shallow aquifer and gas at natural springs to the sources in the deep reservoir and magmatic (mantle) sources below. Figure 1 shows the location of wells and springs sampled in previous studies. Examination of trace element
3197
Elizabeth Keating et al. / Energy Procedia 63 (2014) 3195 – 3201
data (see Figure 2) indicated that metals such as As, B, Li, and Rb were elevated in the CO2-rich groundwater, and concentrations were strongly correlated with Cl. This result suggests that mixing with saline water is controlling trace element chemistry, and that trace elements are not being released in-situ due to CO2 influx and subsequent water-rock reaction. This result is consistent with observations at another natural analog site [6], which suggest that saline water influx affects trace metal concentrations more than in-situ mobilization due to CO2. 1.0 TEP Wells
TEP Wells
0.04
Salado Springs
Salado Springs
0.8
B (mg/L)
As (mg/L)
0.03
0.02
0.01
0.0
0.6
0.4
0.2
0
100
200
300
400
500
0.0
600
0
100
200
0.6
Rb (mg/L)
Li (mg/L)
400
500
600
0.08 0.06
0.2
0.04
0.1
0.02
100
600
200
Salado Springs
0.10
0.3
0
500
TEP Wells
Salado Springs
0.4
0.0
400
0.12
TEP Wells 0.5
300
Cl (mg/L)
Cl (mg/L)
300
Cl (mg/L)
400
500
600
0.00
0
100
200
300
Cl (mg/L)
Figure 2. Relationship between trace metal and chloride concentrations
Application of a Rayleigh degassing model to carbon stable isotope ratio and alkalinity data at this site supported the conceptual model proposed by [3,7] that upward transport in the dissolved phase is the most likely CO2 leakage mechanism at this site. The phase of the CO2 a very important factor in determining the rate of CO2 leakage. Buoyant free phase would migrate upward much more quickly than dissolved phase. In fact, CO2 dissolution will increase the density of water and tend to slow the upward water flow rate. If, then, this site is an appropriate analog for mechanisms of CO2 leakage that might occur at a geologic CO2 sequestration site, one might conclude that leakage will be relatively slow, in the dissolved phase. None of the prior studies addressed the source of the saline water. Here we apply strontium isotope data to discriminate between two hypotheses: one, that the saline water originated in the basement rocks below the reservoir (as did the CO2), or two, that the saline water originated within the reservoir as a result of rock/water interactions. 3. Methods Wells and springs in the vicinity of Coyote Wash fault zone were sampled as described in [2]. Sample locations are shown in Figure 1. Strontium was analyzed on a Thermo Neptune® multicollector ICP-MS (MC-ICP-MS) at the University of Pittsburgh. Strontium separation from the groundwater matrix is detailed in [9]. The separated Sr solution was collected in an acid-cleaned polypropylene centrifuge tube, diluted to 2% HNO3, and the solution analyzed directly for 87Sr/86Sr on the MC-ICP-MS. Instrumental mass fractionation was corrected using an
3198
Elizabeth Keating et al. / Energy Procedia 63 (2014) 3195 – 3201
exponential law, normalizing to 86Sr/88Sr = 0.1194. NIST Standard SRM 987 was run repeatedly with the unknown samples, and the final 87Sr/86Sr value of the sample reflects the offset from this standard over the period of the analysis. To allow comparison of data from this lab over long time periods, the data were normalized to a long-term SRM987 value of 0.710240. Estimated external reproducibility is ±0.000025. Maximum total procedural blanks ranged from 40 to 320 pg Sr (n= 5), resulting in a maximum blank/sample ratio of 0.00016, which is negligible. 4. Results and discussion Strontium concentration and isotope ratios (87Sr/86Sr) are summarized in Table 4.1. Figure 3 is a mixing diagram plotting 87Sr/86Sr versus the inverse concentration of Sr and Cl concentration. Although the range 87Sr/86Sr ratio is small (0.7104 to 0.7111), the variations in Sr (1.13 – 2.78 mg/L) and Cl (138 – 491 mg/L) are appreciable and the data show a strong linear trend. Table 4.1 Sample Location P-7 P-8 P-8 replicate P-9 P-9a P-10 P-16 P-18 P-19 P-20 Salado spring 1 Salado spring 2 Salado 3 - well 82w
Sr (mg/L) 1.13 1.47 2.75 2.78 1.55 1.17 2.57 1.66 1.42 2.64 2.56 2.60
87
Sr/86Sr 0.710391 ±0.000005 0.710481 ±0.000008 0.710489 ±0.000008 0.711103 ±0.000007 0.711100 ±0.000007 0.710655 ±0.000006 0.710316 ±0.000007 0.710984 ±0.000006 0.710517 ±0.000006 0.710546 ±0.000008 0.711062 ±0.000007 0.711073 ±0.000007 0.711081 ±0.000008
The excellent linear fit (R2= 0.94 and 0.96) to the data strongly suggests mixing between groundwater with higher Sr/86Sr ratios and high Sr and Cl (low 1/Sr and 1/Cl) and groundwater with lower 87Sr/86Sr ratios and lower Sr and Cl (high 1/Sr and 1/Cl). This is consistent with the conclusions of [2], that showed strong mixing relationships between major and trace elements suggesting influence of saline fluids in the shallow groundwater. [2] also showed that carbon stable isotope ratios in the shallow groundwater indicate a component of dissolved CO2 that had migrated from depth with these saline fluids. Based on carbon stable isotope ratios and 3He/4He ratios in the gas reservoir, the deeper CO2 component is interpreted to be of magmatic and deep crustal origin, migrating into the Springerville-St. Johns dome along basement penetrating faults [3] The saline component is likely derived from the sedimentary units, but could also be of deep crustal origin traveling into the basin with the CO2. The strontium isotope ratio mixing relationship presented here allows for this possibility to be tested. Even though the end-member groundwater compositions for this mixing trend are not known, both the dilute (low 87Sr/86Sr) and saline (high 87 Sr/86Sr) end-members must fall along this trend. Based on the linear regression of the data in Fig. 3, the intercept with the 87Sr/86Sr axis at very high salinity (1/Sr and 1/Cl approaching zero) is 0.7116, only slightly higher than the maximum measured in the groundwater If the saline end member entered the basin along with the CO2 from the basement (Precambrian crystalline rocks), the fluid should carry high Sr and Cl with a strong radiogenic signal. The 87Sr/86Sr of basement-derived fluids could be on the order of 0.73 - 0.75 as observed in Grand Canyon springs [10], which is not consistent with these observations. 87
Elizabeth Keating et al. / Energy Procedia 63 (2014) 3195 – 3201
Figure 3. 87Sr/86Sr vs. 1/Sr (A) and 1/Cl (B) for TEP wells and Salado Springs. The dashed line in each figure is a two-component mixing relationship (R2 for linear fit shown). End-member compositions are unknown, but must fall on the mixing trend. State well 11-21 is projected on (B) based on reported Cl concentrations in [8]. . To further illustrate, Fig. 4 shows an alternative mixing line that includes a saline, radiogenic end-member, and this line is significantly different than the observed data. This indicates that the saline end-member is derived from groundwater-rock interaction within the CO2-reservoir host rocks, which include Paleozoic carbonates, evaporites and shale units [8]. 5. Conclusions The natural CO2 reservoir at St-John’s Dome provides an opportunity to study physical and geochemical mechanisms governing upward migration of brine and CO2 into a shallow aquifer. The new data presented here, combined with previous studies [1-4,7,8], suggest that CO2 is rising from basement rocks into the gas reservoir, and then up into the shallow aquifer along a relatively permeable fault zone (Coyote Wash). Although in the past CO2 may have been transported to the surface in the gas phase, at present it appears to be rising in the dissolved phase. Trace metal concentrations in the shallow aquifer are elevated, but this appears to be caused by mixing with saline waters from below rather than from in-situ mobilization of trace elements caused by CO2. The strontium isotope data presented here suggests, however, that the saline waters have a different source than the CO2. The conceptual
3199
3200
Elizabeth Keating et al. / Energy Procedia 63 (2014) 3195 – 3201
model of gaseous CO2 entering a deep saline water reservoir, dissolving into the saline water, then slowly moving upward in the dissolved phase and mixing with fresh groundwater is consistent with available hydrologic and geochemical data. The method we applied to use carbon and strontium isotopes to trace the origins of both the salinity and the carbon should be readily applied to CCS sites.
Figure
4. 87Sr/86Sr vs. 1/Sr showing mixing trend defined by data (dashed line) and a fmixing line with a radiogenic end-member representative of fluids exchanged with basement rocks. This mixing line is assumed to have a radiogenic end member with a 87Sr/86Sr of 0.73 and a Sr concentration of 10 mg/L (0.1 L/mg) and a lower end-member based on the low end of the Springerville data set. The groundwater data presented here cannot be fit a significantly more radiogenic end-member.
6. References
[1] Keating, EH, Newell, DL, Viswanathan, H, Carey JW, Zyvoloski, G., and Pawar, RJ. CO2/Brine Transport into Shallow Aquifers along Fault Zones. Environmental Science & Technology; 2012; 47: 290-297. [2] Keating, EH, Newell DL, Dempsey, DE, Pawar, RJ. Insights into interconnections between the shallow and deep systems from a natural CO2 reservoir near Springerville, Arizona, International Journal of Greenhouse Gases. 2014; 25: 162-172. _http://dx.doi.org/10.1016/j.ijggc.2014.03.009. [3] Gilfillan SMV, Wilkinson M, Haszeldine RS, Shipton ZK, Nelson ST, and Poreda RJ, 2011, He and Ne as tracers of natural CO2 migration up a fault from a deep reservoir: International Journal of Greenhouse Gas Control, 2011; 5: 1507-1516. [4] Gilfillan SMV, Sherwood Lollar B, Holland G, Blagburn, D, Stevens S, Schoell M, Cassidy M, Ding Z, Zhou Z., Lacrampe-Couloume G, and Ballentine C J. Solubility trapping in formation water as dominant CO2 sink in natural gas fields. Nature. 2009; 458:614-617. [5] Carroll S, Keating EH, Mansoor K, Dai, Z, Sun, Y, Trainor-Guitton, W, Brown, C. Key Factors for Determining Groundwater Impacts Due to Leakage from Geologic Carbon Sequestration Reservoirs. International Journal of Greenhouse Gas Control. 2014;29:153̽168. [6] Keating EH, Fessenden JE, Kanjorksi N, Koning DJ, Pawar RJ. The impact of CO2 on shallow groundwater
Elizabeth Keating et al. / Energy Procedia 63 (2014) 3195 – 3201
chemistry: observations at a natural analog site and implications for carbon sequestration. Environmental Earth Science 2010; 60:521-536. [7] Allis RG, Moore J, White, S. Reactive Multiphase Behavior of CO2 in Saline Aquifers Beneath the Colorado Plateau.2005. Final report to US Department of Energy. 23pp. [8] Moore J; Adams, M, Allis, R, Lutz, S, Rauzi, S. Mineralogical and geochemical consequences of the long-term presence of CO2 in natural reservoirs: An example from the Springerville-St. Johns Field, Arizona, and New Mexico, U.S.A. Chemical Geology. 2005; 217: 365-385. [9] Wall AJ, Capo RC, Stewart BW, Phan TT, Jain JC, Hakala JA, Guthrie, GD. High throughput method for Sr extraction from variable matrix waters and 87Sr/86Sr isotope ratios by MC-ICP-MS. Journal of Analytical Atomic Spectrometry. 2013; 28:1338-1344. [10] Crossey LJ, Fischer TP, Patchett, PJ, Karlstrom KE, Hilton DR, Huntoon P, Newell D, Reynolds, A. 2006. Dissected hydrologic system at the Grand Canyon: Interaction between deeply derived fluids and plateau aquifer waters in modern springs and travertine. Geology. 2006; 34: 25-28.
3201
ScienceDirect Energy Procedia 63 (2014) 3195 – 3201
GHGT-12
Further insights into interconnections between the shallow and deep systems from a natural CO2 reservoir near Springerville, Arizona, U.S.A. Elizabeth Keatinga, Dennis Newellb, Brian Stewartc , Rosemary Capoc, and Rajesh Pawara a
Earth and Environmental Sciences Division, MS T003, Los Alamos National Laboratory, Los Alamos, NM 87545 b Department of Geology, Utah State University, Logan UT, USA 84322-4505 c Department of Geology and Planetary Science, University of Pittsburgh, Pittsburgh, PA, 15260 USA. ,
Abstract
If carbon dioxide (CO2) sequestration into deep geologic reservoirs is to be accepted by the public and environmental regulators, the possibility of upward leakage into shallow groundwater should be acknowledged and those processes well-understood. Studies of natural CO2 reservoirs and their connection (or lack thereof) with the shallow subsurface is one way to explore these issues. A natural reservoir near Springerville, Arizona, U.S.A. has leaked CO2 to the surface along a fault zone for thousands of years, creating large travertine deposits. In recent times, the CO2 leak rates have declined significantly yet the shallow aquifer is still highly enriched in dissolved CO2. In previous studies, using water level data and simulations we demonstrate that the fault zone provides hydrologic communication between the shallow aquifer and the deeper reservoir. It is reasonable to assume, therefore, that the source of the CO2 in wells completed within the fault zone is the deeper CO2 reservoir. We present water chemistry data to demonstrate the geochemical impact of this CO2 on shallow groundwater quality. Interestingly, arsenic concentrations are elevated, but other trace metals concentrations are not. Previous studies [1,2,3] showed that CO2 originating from magmatic and deep crustal origin is migrating upward along the fault and dissolving into the shallow groundwater. Saline waters are also mixing, to a much lesser degree, but their source was unknown. Here we present strontium isotope data that clearly shows the source to be water/rock interactions with reservoir rocks, which include Paleozoic carbonates, evaporites and shale units. © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
© 2013 The Authors. Published by Elsevier Ltd. (http://creativecommons.org/licenses/by-nc-nd/3.0/). peer-review under GHGT. of GHGT-12 Selection Peer-reviewand under responsibility of responsibility the Organizing of Committee
1876-6102 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Peer-review under responsibility of the Organizing Committee of GHGT-12 doi:10.1016/j.egypro.2014.11.344
3196
Elizabeth Keating et al. / Energy Procedia 63 (2014) 3195 – 3201
Keywords: shallow groundwater impacts of CO2; brine migration
1. Introduction Studies of natural CO2 reservoirs that are leaking to shallow aquifers and the surface provide opportunities to study CO2 leakage at temporal and spatial scales that are difficult if not impossible to achieve with engineered controlled release studies. The natural CO2 reservoir at St. Johns Dome in Arizona, USA has been leaking CO2 to the surface for thousands of years, creating large travertine deposits. At present, travertine deposition has ceased and measurable CO2 flux at ground surface is negligible; however, CO2 continues to enrich the shallow aquifer and springs that discharge to the Little Colorado River. This site has been used to study the physical mechanisms of CO2 leakage and geochemical impacts of this leakage on shallow aquifers [1,2,3]. Insights into CO2 trapping mechanisms in have also been gained here [4]. One of the potential risks to shallow groundwater from geologic CO2 sequestration is brine leakage [4]. Studies of other natural analog sites have shown that saline waters have the potential to be much more damaging to fresh water quality than CO2 alone [5]. Given the stratigraphy of the St. John’s Dome reservoir, which includes marine carbonate and evaporite sequences, it is reasonable to assume that saline water is present at depth associated with these formations. Although there is geochemical evidence of saline water mixing with shallow groundwater at this site [2], the impact of this mixing on groundwater quality of relatively minor. Multi-phase flow simulations [1] showed that the physical characteristics of the fault zone system at this site (Coyote Wash fault) could explain significant upward transport of CO2 without displacing the saline water/fresh water interface. However, additional data would be required to further test this hypothesis. Here we summarize the key elements of previous studies of the site and the significance to geologic sequestration. We also present new strontium isotope data from the site, which we use to constrain the source of the saline water. 2. Previous work relevant to potential leakage at carbon sequestration sites Allis et al [7] developed a conceptual and numerical model of CO2 outflow from the St. Johns Dome reservoir that reasonably-well reproduced the depositional history of travertine along the Coyote Wash Fault Zone. Their simulations showed a transient overpressurization of the reservoir, caused by a pulse of CO2 from depth, leading to a rise in the water table elevation and broad area of water and CO2 discharge. A soil CO2 flux survey showed no evidence of present-day CO2 flux at ground surface. Groundwater chemistry data [8] clearly shows Figure 1. Location of sampled wells (stars), springs, and travertine evidence of CO2 outflow from the reservoir in the deposits (yellow shading). Blue lines indicate the piezometric surface dissolved phase, and the authors conclude that in the shallow aquifer. “dissolved CO2 could be the main source of CO2 leakage from a sequestration site.” Both noble gases [3] and stable isotope ratios of carbon [2] have been used to trace the dissolved carbon dioxide in the shallow aquifer and gas at natural springs to the sources in the deep reservoir and magmatic (mantle) sources below. Figure 1 shows the location of wells and springs sampled in previous studies. Examination of trace element
3197
Elizabeth Keating et al. / Energy Procedia 63 (2014) 3195 – 3201
data (see Figure 2) indicated that metals such as As, B, Li, and Rb were elevated in the CO2-rich groundwater, and concentrations were strongly correlated with Cl. This result suggests that mixing with saline water is controlling trace element chemistry, and that trace elements are not being released in-situ due to CO2 influx and subsequent water-rock reaction. This result is consistent with observations at another natural analog site [6], which suggest that saline water influx affects trace metal concentrations more than in-situ mobilization due to CO2. 1.0 TEP Wells
TEP Wells
0.04
Salado Springs
Salado Springs
0.8
B (mg/L)
As (mg/L)
0.03
0.02
0.01
0.0
0.6
0.4
0.2
0
100
200
300
400
500
0.0
600
0
100
200
0.6
Rb (mg/L)
Li (mg/L)
400
500
600
0.08 0.06
0.2
0.04
0.1
0.02
100
600
200
Salado Springs
0.10
0.3
0
500
TEP Wells
Salado Springs
0.4
0.0
400
0.12
TEP Wells 0.5
300
Cl (mg/L)
Cl (mg/L)
300
Cl (mg/L)
400
500
600
0.00
0
100
200
300
Cl (mg/L)
Figure 2. Relationship between trace metal and chloride concentrations
Application of a Rayleigh degassing model to carbon stable isotope ratio and alkalinity data at this site supported the conceptual model proposed by [3,7] that upward transport in the dissolved phase is the most likely CO2 leakage mechanism at this site. The phase of the CO2 a very important factor in determining the rate of CO2 leakage. Buoyant free phase would migrate upward much more quickly than dissolved phase. In fact, CO2 dissolution will increase the density of water and tend to slow the upward water flow rate. If, then, this site is an appropriate analog for mechanisms of CO2 leakage that might occur at a geologic CO2 sequestration site, one might conclude that leakage will be relatively slow, in the dissolved phase. None of the prior studies addressed the source of the saline water. Here we apply strontium isotope data to discriminate between two hypotheses: one, that the saline water originated in the basement rocks below the reservoir (as did the CO2), or two, that the saline water originated within the reservoir as a result of rock/water interactions. 3. Methods Wells and springs in the vicinity of Coyote Wash fault zone were sampled as described in [2]. Sample locations are shown in Figure 1. Strontium was analyzed on a Thermo Neptune® multicollector ICP-MS (MC-ICP-MS) at the University of Pittsburgh. Strontium separation from the groundwater matrix is detailed in [9]. The separated Sr solution was collected in an acid-cleaned polypropylene centrifuge tube, diluted to 2% HNO3, and the solution analyzed directly for 87Sr/86Sr on the MC-ICP-MS. Instrumental mass fractionation was corrected using an
3198
Elizabeth Keating et al. / Energy Procedia 63 (2014) 3195 – 3201
exponential law, normalizing to 86Sr/88Sr = 0.1194. NIST Standard SRM 987 was run repeatedly with the unknown samples, and the final 87Sr/86Sr value of the sample reflects the offset from this standard over the period of the analysis. To allow comparison of data from this lab over long time periods, the data were normalized to a long-term SRM987 value of 0.710240. Estimated external reproducibility is ±0.000025. Maximum total procedural blanks ranged from 40 to 320 pg Sr (n= 5), resulting in a maximum blank/sample ratio of 0.00016, which is negligible. 4. Results and discussion Strontium concentration and isotope ratios (87Sr/86Sr) are summarized in Table 4.1. Figure 3 is a mixing diagram plotting 87Sr/86Sr versus the inverse concentration of Sr and Cl concentration. Although the range 87Sr/86Sr ratio is small (0.7104 to 0.7111), the variations in Sr (1.13 – 2.78 mg/L) and Cl (138 – 491 mg/L) are appreciable and the data show a strong linear trend. Table 4.1 Sample Location P-7 P-8 P-8 replicate P-9 P-9a P-10 P-16 P-18 P-19 P-20 Salado spring 1 Salado spring 2 Salado 3 - well 82w
Sr (mg/L) 1.13 1.47 2.75 2.78 1.55 1.17 2.57 1.66 1.42 2.64 2.56 2.60
87
Sr/86Sr 0.710391 ±0.000005 0.710481 ±0.000008 0.710489 ±0.000008 0.711103 ±0.000007 0.711100 ±0.000007 0.710655 ±0.000006 0.710316 ±0.000007 0.710984 ±0.000006 0.710517 ±0.000006 0.710546 ±0.000008 0.711062 ±0.000007 0.711073 ±0.000007 0.711081 ±0.000008
The excellent linear fit (R2= 0.94 and 0.96) to the data strongly suggests mixing between groundwater with higher Sr/86Sr ratios and high Sr and Cl (low 1/Sr and 1/Cl) and groundwater with lower 87Sr/86Sr ratios and lower Sr and Cl (high 1/Sr and 1/Cl). This is consistent with the conclusions of [2], that showed strong mixing relationships between major and trace elements suggesting influence of saline fluids in the shallow groundwater. [2] also showed that carbon stable isotope ratios in the shallow groundwater indicate a component of dissolved CO2 that had migrated from depth with these saline fluids. Based on carbon stable isotope ratios and 3He/4He ratios in the gas reservoir, the deeper CO2 component is interpreted to be of magmatic and deep crustal origin, migrating into the Springerville-St. Johns dome along basement penetrating faults [3] The saline component is likely derived from the sedimentary units, but could also be of deep crustal origin traveling into the basin with the CO2. The strontium isotope ratio mixing relationship presented here allows for this possibility to be tested. Even though the end-member groundwater compositions for this mixing trend are not known, both the dilute (low 87Sr/86Sr) and saline (high 87 Sr/86Sr) end-members must fall along this trend. Based on the linear regression of the data in Fig. 3, the intercept with the 87Sr/86Sr axis at very high salinity (1/Sr and 1/Cl approaching zero) is 0.7116, only slightly higher than the maximum measured in the groundwater If the saline end member entered the basin along with the CO2 from the basement (Precambrian crystalline rocks), the fluid should carry high Sr and Cl with a strong radiogenic signal. The 87Sr/86Sr of basement-derived fluids could be on the order of 0.73 - 0.75 as observed in Grand Canyon springs [10], which is not consistent with these observations. 87
Elizabeth Keating et al. / Energy Procedia 63 (2014) 3195 – 3201
Figure 3. 87Sr/86Sr vs. 1/Sr (A) and 1/Cl (B) for TEP wells and Salado Springs. The dashed line in each figure is a two-component mixing relationship (R2 for linear fit shown). End-member compositions are unknown, but must fall on the mixing trend. State well 11-21 is projected on (B) based on reported Cl concentrations in [8]. . To further illustrate, Fig. 4 shows an alternative mixing line that includes a saline, radiogenic end-member, and this line is significantly different than the observed data. This indicates that the saline end-member is derived from groundwater-rock interaction within the CO2-reservoir host rocks, which include Paleozoic carbonates, evaporites and shale units [8]. 5. Conclusions The natural CO2 reservoir at St-John’s Dome provides an opportunity to study physical and geochemical mechanisms governing upward migration of brine and CO2 into a shallow aquifer. The new data presented here, combined with previous studies [1-4,7,8], suggest that CO2 is rising from basement rocks into the gas reservoir, and then up into the shallow aquifer along a relatively permeable fault zone (Coyote Wash). Although in the past CO2 may have been transported to the surface in the gas phase, at present it appears to be rising in the dissolved phase. Trace metal concentrations in the shallow aquifer are elevated, but this appears to be caused by mixing with saline waters from below rather than from in-situ mobilization of trace elements caused by CO2. The strontium isotope data presented here suggests, however, that the saline waters have a different source than the CO2. The conceptual
3199
3200
Elizabeth Keating et al. / Energy Procedia 63 (2014) 3195 – 3201
model of gaseous CO2 entering a deep saline water reservoir, dissolving into the saline water, then slowly moving upward in the dissolved phase and mixing with fresh groundwater is consistent with available hydrologic and geochemical data. The method we applied to use carbon and strontium isotopes to trace the origins of both the salinity and the carbon should be readily applied to CCS sites.
Figure
4. 87Sr/86Sr vs. 1/Sr showing mixing trend defined by data (dashed line) and a fmixing line with a radiogenic end-member representative of fluids exchanged with basement rocks. This mixing line is assumed to have a radiogenic end member with a 87Sr/86Sr of 0.73 and a Sr concentration of 10 mg/L (0.1 L/mg) and a lower end-member based on the low end of the Springerville data set. The groundwater data presented here cannot be fit a significantly more radiogenic end-member.
6. References
[1] Keating, EH, Newell, DL, Viswanathan, H, Carey JW, Zyvoloski, G., and Pawar, RJ. CO2/Brine Transport into Shallow Aquifers along Fault Zones. Environmental Science & Technology; 2012; 47: 290-297. [2] Keating, EH, Newell DL, Dempsey, DE, Pawar, RJ. Insights into interconnections between the shallow and deep systems from a natural CO2 reservoir near Springerville, Arizona, International Journal of Greenhouse Gases. 2014; 25: 162-172. _http://dx.doi.org/10.1016/j.ijggc.2014.03.009. [3] Gilfillan SMV, Wilkinson M, Haszeldine RS, Shipton ZK, Nelson ST, and Poreda RJ, 2011, He and Ne as tracers of natural CO2 migration up a fault from a deep reservoir: International Journal of Greenhouse Gas Control, 2011; 5: 1507-1516. [4] Gilfillan SMV, Sherwood Lollar B, Holland G, Blagburn, D, Stevens S, Schoell M, Cassidy M, Ding Z, Zhou Z., Lacrampe-Couloume G, and Ballentine C J. Solubility trapping in formation water as dominant CO2 sink in natural gas fields. Nature. 2009; 458:614-617. [5] Carroll S, Keating EH, Mansoor K, Dai, Z, Sun, Y, Trainor-Guitton, W, Brown, C. Key Factors for Determining Groundwater Impacts Due to Leakage from Geologic Carbon Sequestration Reservoirs. International Journal of Greenhouse Gas Control. 2014;29:153̽168. [6] Keating EH, Fessenden JE, Kanjorksi N, Koning DJ, Pawar RJ. The impact of CO2 on shallow groundwater
Elizabeth Keating et al. / Energy Procedia 63 (2014) 3195 – 3201
chemistry: observations at a natural analog site and implications for carbon sequestration. Environmental Earth Science 2010; 60:521-536. [7] Allis RG, Moore J, White, S. Reactive Multiphase Behavior of CO2 in Saline Aquifers Beneath the Colorado Plateau.2005. Final report to US Department of Energy. 23pp. [8] Moore J; Adams, M, Allis, R, Lutz, S, Rauzi, S. Mineralogical and geochemical consequences of the long-term presence of CO2 in natural reservoirs: An example from the Springerville-St. Johns Field, Arizona, and New Mexico, U.S.A. Chemical Geology. 2005; 217: 365-385. [9] Wall AJ, Capo RC, Stewart BW, Phan TT, Jain JC, Hakala JA, Guthrie, GD. High throughput method for Sr extraction from variable matrix waters and 87Sr/86Sr isotope ratios by MC-ICP-MS. Journal of Analytical Atomic Spectrometry. 2013; 28:1338-1344. [10] Crossey LJ, Fischer TP, Patchett, PJ, Karlstrom KE, Hilton DR, Huntoon P, Newell D, Reynolds, A. 2006. Dissected hydrologic system at the Grand Canyon: Interaction between deeply derived fluids and plateau aquifer waters in modern springs and travertine. Geology. 2006; 34: 25-28.
3201