Natural analogue monitoring to estimate the hydrochemical change of groundwater by the carbonating process from the introduction of CO2

Abstract

This study monitored four geothermal groundwater (TW) wells (Suanbo, Yuseong, Deoksan, and Seokmodo), four carbonated groundwater (CW) springs (Chojeong, Bugang, Shin, and Bangadari) and shallow groundwater (GW) in order to interpret the carbonating process from the seepage of CO2. In a bid to figure out the carbonated reservoir where CO2 mixing occurs, the CW samples represent the condition of the final carbonated water as natural analogues. The initial fresh water can be assumed to be the following depth: the deep aquifer with 300–700 m below the surface level (b.s.l) and the shallow aquifer to be below 300 m b.s.l. The pairs of hydrogen (δ2H) and oxygen (δ18O) isotopes indicated that all the groundwater samples had been recharged from meteoric water at various heights. The Δ14C data of the TW samples present residence time from 1980 ± 30 to 6400 ± 30 years BP (before present) and the δ13C data of dissolved inorganic carbon in the CW distributed from −3 to −10‰ range implying the mantle degassing effects. Regardless of seasonal variations, the TW samples and the CW samples were shown to be the Na-HCO3 type and the Ca-HCO3 type, respectively, indicating these aquifers remain in closed conditions. Based on the chalcedony geothermometer results, the carbonated reservoir of the CW samples were estimated to range between 43.08 and 83.13 °C, which were paired with the average temperature of the TW sites. In turn, the reservoir depth estimations taken by multiplying temperatures by geothermal gradients were estimated to range between 1.34 km and 3.90 km. Meanwhile, the in-situ data of the CW sites were estimated to range between 3.8 and 18.7 °C and between 263.7 and 488.7 mV of redox potential of oxidizing condition, which means there are the possibilities of shallow carbonated reservoir. In order to delineate the hydrochemical changes of the fresh groundwater during carbonation at a deep reservoir and a shallow reservoir, the PHREEQC modeling was conducted with 0.0001–10 mol L−1 of CO2 mixing. As the CO2 concentrations increased, the TW and the GW displayed different ionic variations but the inland groundwater identically moved to the Ca-HCO3 type. The coastal TW can be reflecting the carbonation process of the coastal AZMI section, containing large quantities of salts. The carbonation of these samples is triggered by 0.01 mol L−1 of CO2 introduced into the reservoir but it still retained the Na-Cl water type despite a Log PCO2 of over −0.5. The study results imply that there can be deep reservoirs and shallow reservoirs responsible for developing carbonated groundwater. It will be useful to estimate hydrochemical variations of groundwater resulting from the natural-artificial CO2 leakage.