Chemical exchange between carbonate and internal water populations represents the lowest possible water/rock fraction in natural materials: with molar ratios possibly lower than one percent. Despite their low abundance, internal water masses in fossils and mineral specimens can play an important role mediating exchange reactions: lowering the energetic barriers for resetting clumped isotopes, as well as supplying oxygen to reset carbonate δ18O values. In heating experiments below 375°C, these reactions do not reach full clumped isotope equilibrium at the experimental temperature. Applying the dual-clumped isotope system to heated and unheated bivalve and coral carbonate reveals that this partial equilibration is due to a mixture of fully equilibrated and unaltered endmembers, as opposed to the carbonate uniformly reaching equilibrium at a cooler temperature. This interpretation appears to contradict the behavior of fluid inclusion δ18O values, which appear to have equilibrated at a cooler temperature. We attribute this discrepancy to the existence of multiple populations of internal water, such as organic-bound water and “true” fluid inclusions of liquid water, which could have different exchange kinetics with the carbonate. Dual-clumped analyses reveal the possibility for calculating peak heating temperature of thermally altered minerals, and future work will focus on finding unambiguous indicators for thermal alteration. An unambiguous indicator for heating will enable the use of new climate archives that otherwise would be disregarded due to the possibility of heating by ancient humans (shell middens), or fossils geothermally heated during burial.