Noble gases dissolved in groundwater have been used to evaluate characteristics such as origin, residence time and recharge temperature. Although obtaining samples without degassing is highly desired, in certain cases, degassing during sampling is unavoidable. This study investigated the degassing behavior of noble gases during sampling and evaluated the correction methods that can be applied. Groundwater samples were obtained from an underground laboratory and a deep borehole. The amount of exsolved gas was estimated by comparing pressurized and non-pressurized samples. Henry's model, which assumes equilibrium between gas and water, and a diffusion model were applied to the results. In addition, the compositions of gases extracted into an evacuated vessel or hollow fiber contactor were also compared with those of pressurized groundwater samples. The results indicated that significant degassing occurred in unpressurized samples, but equilibrium was achieved between gas and water when the total dissolved gas pressure (TDGP) was relatively small (up to 4.2 atm). Therefore, the d’Ne value, defined as the difference between the Ne concentration of air-saturated water and a sample normalized by the Ne concentration of air-saturated water, could be used in Henry's model to correct exsolution. However, inclusion of the effects of excess air and contamination is problematic. However, when TDGP was high (32 atm in this study), the degassing behavior completely changed and could not be predicted by Henry's model. During extraction of noble gases using an evacuated vessel and hollow fiber contactor, degassing was observed to take place following the diffusion model. As a result, fractionation between Ne vs. Kr and Xe was significant. Thus, although we could extract dissolved Ne with a high recovery ratio (over 0.85), the ratios of Kr/Ne and Xe/Ne were found to be significantly smaller than those in groundwater. The gases extracted by an evacuated vessel or hollow fiber contactor appeared to be applicable only for evaluating He/Ne.
Read full abstract