Evaluation of the pressure-volume-temperature (PVT) data of water from experiments and molecular simulations since 1990

Since 1990, many groups of pressure-volume-temperature (PVT) data from experiments and molecular dynamics (MD) or Monte Carlo (MC) simulations have been reported for supercritical and subcritical water. In this work, fifteen groups of PVT data (253.15-4356 K and 0-90.5 GPa) are evaluated in detail with the aid of the highly accurate IAPWS-95 formulation. The evaluation gives the following results: (1) Six datasets are found to be of good accuracy. They include the simulated results based on SPCE potential above 100 MPa and those derived from sound velocity measurements, but the simulated results below 100 MPa have large uncertainties. (2) The data from measurements with a piston cylinder apparatus and simulations with an exp-6 potential contain large uncertainties and systematic deviations. (3) The other seven datasets show obvious systematic deviations. They include those from experiments with synthesized fluid inclusion techniques (three groups), measured velocities of sound (one group), and automated high-pressure dilatometer (one group) and simulations with TIP4P potential (two groups), where the simulated data based on TIP4P potential below 200 MPa have large uncertainties. (4) The simulated data but those below 1 GPa agree with each other within 2-3%, and mostly within 2%. The data from fluid inclusions show similar systematic deviations, which are less than 2-5%. The data obtained with automated high-pressure dilatometer and those derived from sound velocity measurements agree with each other within 0.3-0.6% in most cases, except for those above 10 GPa. In principle, the systematic deviations mentioned above, except for those of the simulated data below 1 GPa, can be largely eliminated or significantly reduced by appropriate corrections, and then the accuracy of the relevant data can be improved significantly. These are very important for the improvement of experiments or simulations and the refinement and correct use of the PVT data in developing thermodynamic models of water or water-bearing fluids.