Article pubs.acs.org/est
CO2/Brine Transport into Shallow Aquifers along Fault Zones Elizabeth H. Keating,* Dennis L. Newell, Hari Viswanathan, J. W. Carey, G. Zyvoloski, and Rajesh Pawar Earth and Environmental Sciences Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, United States ABSTRACT: Unintended release of CO2 from carbon sequestration reservoirs poses a well-recognized risk to groundwater quality. Research has largely focused on in situ CO2-induced pH depression and subsequent trace metal mobilization. In this paper we focus on a second mechanism: upward intrusion of displaced brine or brackish-water into a shallow aquifer as a result of CO2 injection. Studies of two natural analog sites provide insights into physical and chemical mechanisms controlling both brackish water and CO2 intrusion into shallow aquifers along fault zones. At the Chimayó, New Mexico site, shallow groundwater near the fault is enriched in CO2 and, in some places, salinity is significantly elevated. In contrast, at the Springerville, Arizona site CO2 is leaking upward through brine aquifers but does not appear to be increasing salinity in the shallow aquifer. Using multiphase transport simulations we show conditions under which significant CO2 can be transported through deep brine aquifers into shallow layers. Only a subset of these conditions favor entrainment of salinity into the shallow aquifer: high aspect-ratio leakage pathways and viscous coupling between the fluid phases. Recognition of the conditions under which salinity is favored to be cotransported with CO2 into shallow aquifers will be important in environmental risk assessments.
■
of aquifers in the arid southwest5,6 suggest hydraulic connections between fresh and saline aquifers do exist and that depressurization of shallow aquifers can cause upward or lateral flow of saline water, and introduce metals and other constituents that degrade water quality.7 In this paper, we focus on a specific leakage scenario at two field sites where a CO2 plume has reached a high permeability conduit and begins to move upward. In this scenario, as will usually be the case at a CCS site, saline water exists either in the deep reservoir, the conduit, or both. The question we explore here is the extent to which upward saline water migration will tend to accompany or even precede CO2 discharge, or if there are conditions that will favor CO2 discharge but not saline water discharge. We first summarize findings at two natural analog sites and then explore possible mechanisms of transport at one of the sites using multiphase flow simulations.
INTRODUCTION There are a limited number of studies that have examined the potential for CO2 injection in deep saline aquifers not only to produce unintended CO2 leakage into shallow layers but also to displace brine upward. The brine leakage scenario is quite important to consider since salts, organic compounds, and trace metals in brine have considerable potential to degrade shallow freshwater quality. A simulation study of CO2 injection into saline aquifers suggested that the risk of upward migration of brine through a stratified aquifer/aquitard system is quite low.1 These authors conclude that only vertical leakage pathways such as faults or leaky boreholes have the potential to transport brine upward. In an examination of density and temperature effects on brine movement, the potential for overpressurization at depth to move brine upward along a high-permeability conduit (e.g., fault or leaky wellbore) was evaluated.2 Depending on the presumed salinity of the brine and the degree of overpressurization in the deep reservoir, the brine could move slightly upward and achieve a new steady-state, or the brine could move upward into the base of a freshwater aquifer and spread laterally. A study of multiphase flow of brine/CO2 discharge through an idealized fault zone illustrated the possibility of immiscible upward displacement of the brine by free-phase CO2.3 These results highlight the strongly transient, nonlinear aspects of flow dynamics and the self-regulating tendency for CO2 discharge to decrease over time. Studies of natural systems also illustrate a range of possible behaviors. Studies of deep saline aquifers in Texas suggest that saline aquifers and shallow fresh water aquifers are compartmentalized by flow-barrier faults, and that upwelling of saline water along faults is unlikely.4 Studies of the salinization © 2012 American Chemical Society
■
FIELD SITES Chimayó, New Mexico. The shallow alluvial-fan sedimentary aquifer at this site has been investigated by several authors.8−10 Briefly, a number of shallow wells penetrating the drinking water aquifer are highly enriched with CO2, including one well which geysers almost pure CO2 daily. As shown in Figure 1, wells located near “Robert’s Fault” tend to be more Special Issue: Carbon Sequestration Received: Revised: Accepted: Published: 290
April 17, 2012 July 12, 2012 July 16, 2012 July 16, 2012 dx.doi.org/10.1021/es301495x | Environ. Sci. Technol. 2013, 47, 290−297
Environmental Science & Technology
Article
Figure 1. Variation of Chimayo groundwater chemistry (modified from ref 8). Blue line is river; red line is Robert’s Fault. Open circles are background (no CO2 impact); gray circles are elevated in CO2 but not brackish water; black circles are heavily impacted by CO2 and brackish water.
Figure 2. Site map of study area at Springerville-St. Johns. Water levels are indicated by blue contours. Red stars show wells sampled. Red box illustrates extent of numerical model. AA′ indicates location of cross-section shown in Figure 8.
moving up along the fault zone and into the aquifer can explain measured pH, pCO2, and major ion concentrations in the shallow aquifer. The inferred CO2 flow rates were 0.4 kg/s distributed along a 3500 × 70 m fault zone. Importantly, the calculated upward flow of saline water was much larger (∼15 kg/s). However, in the previous studies the mechanism causing upward flow of saline water was not investigated. Below we explore possible mechanisms using simplified multiphase models. Springerville-St.Johns CO2 reservoir, Arizona. The Springerville-St. Johns dome is located near the southern margin of the Colorado Plateau straddling the Arizona−New Mexico border. Geologically, the dome is defined by a gently dipping doubly plunging NW−SE anticline that is constructed of Permian-Tertiary Sedimentary strata capped with Quaternary basaltic volcanics and volcaniclastic sediments.12,13 These strata
enriched in CO2 than others in the area, suggesting the CO2 is rising along this structural feature. Interestingly, wells located near the southern end of the fault also show impact of an NaCl brackish water. The TDS in these wells exceeds 6000 ppm, much higher than typical shallow groundwater in the region (
