Geochemical and hydrologic modeling indicates that geothermal waters in the T > 270°C reservoirs beneath Yellowstone National Park have HCO₃ ≪ Cl and contrast with waters in reservoirs at lower temperatures which attain HCO₃ about equal to Cl. Experiments reacting rhyolite with 0.5 molal solutions of CO₂ at 200° and 350°C were carried out to test the hypothesis of 3 and 4 to explain the chemistry of these springs: that CO₂ is relatively unreactive with volcanic rocks at temperatures >270°C. The experimental results strongly support this hypothesis. Extent of alteration is twenty-seven times greater at 200°C than at 350°C. The dominant process in the experiments appears to be the alteration of the albitic component of the rhyolite by dissolved CO₂ to form a kaolinite-like alteration product plus quartz:

2NaAlSi3O8+2CO2+3H2O=2Na++2HCO3-+Al2Si2O5(OH)4+4SiO2.

Turn MathJax on

CO₂ reacts with water to form H₂CO₃ which dissociates to H⁺ and HCO₃⁻, more so at lower temperatures. Kinetic and thermodynamic considerations suggest that the reactivity of H₂CO₃ with wallrocks is at its maximum between 150° and 200°C, consuming most of the H⁺ and liberating equivalent amounts of cations and bicarbonate. Wallrocks in higher temperature reservoirs are relatively unreactive to dissolved CO₂ which is eventually lost from the system by boiling.

These observations also offer a possible explanation for the change in chemical sediments from chloride-dominated to bicarbonate-dominated salts found in the stratigraphic section at Searles Lake, California, the terminus of the Owens River which derives its dissolved load from hot springs of the Long Valley caldera.