Southeastern Section - 66th Annual Meeting - 2017

Paper No. 6-1
Presentation Time: 8:00 AM-12:00 PM

COMPARISON OF METHODS USED FOR CO2 LEAK DETECTION DURING PULSED RELEASE TESTS INTO POTABLE AQUIFER WELLS


EDENBORN, H.M., Geological & Environmental Systems Directorate, Research & Innovation Center, National Energy Technology Lab; U.S. Department of Energy, Pittsburgh, PA 15236, ZOU, Youquin, Bureau of Economic Geology, The University of Texas at Austin, 10100 Burnet Rd., Bldg 130, Austin, TX 78758 and YANG, Changbing, Bureau of Economic Geology, The University of Texas at Austin, 10100 Burnet Rd, Bldg 130, Austin, TX 78758, edenborn@netl.doe.gov

One approach proposed to reduce the release of anthropogenic CO2 to the atmosphere is deep geological storage. Monitoring for potential upward migration of CO2 to surface and near-surface aquatic environments is therefore necessary. Conventional alkalinity titration methods used to measure aqueous CO2 in the field have the significant shortcomings of being tedious, potential interference due to non-carbonate alkalinity, and being highly susceptible to error in pH measurements. In this study, we compared the effectiveness of a novel CO2 detection method using multiple volume expansion with conventional pH/alkalinity titrations. Shallow (ca. 6 m) groundwater monitoring wells in a siliclastic-dominated alluvial aquifer were injected with 10, 60 and 100% volumes of CO2 gas and geochemical responses were monitored during the injection and retreat phases. A CarboQC beverage carbonation meter was used to measure the concentration of CO2 in water by monitoring temperature and pressure changes and calculating the PCO2 from the ideal gas law. For each pulsed test, gas from a CO2 tank was bubbled into the bottom of the test well. Water samples were collected from various depths in the well by peristaltic pump for alkalinity titration, chemical analysis, and direct CO2 analysis. Results showed that the CarboQC method provided rapid (< 3 min) and repeatable results under field conditions within a measured concentration range of 1.6 - 31 mM CO2. Good correlation was observed between CarboQC and alkalinity/pH CO2 analysis, but 10 times the number of samples could be processed in the field with the CarboQC, and CO2 based on alkalinity titration required the further analysis of anions and cations. The ability to analyze water samples quickly and frequently allowed for a more accurate assessment of CO2 concentration and distribution in the well under field conditions. Depth analysis indicated that a strong vertical gradient of dissolved CO2 developed, either due to the natural rise of different-sized gas bubbles, or the influx of low-CO2 groundwater at depth.