Thermodynamic Properties of Water & Steam

Three tables, one fluid — ordinary water (H₂O):

Saturated — by temperature and saturated — by pressure describe the same thing, the two-phase dome where liquid and vapor coexist, from the triple point (0.01 °C) to the critical point (373.946 °C / 22.064 MPa). They differ only in which variable steps the rows. Each gives the saturation pressure or temperature plus liquid (f), vapor (g) and vaporization (fg) values for specific volume, internal energy, enthalpy and entropy.

Compressed liquid & superheated covers the single-phase regions on a pressure×temperature grid — 96 pressure levels from 0.01 MPa to 1000 MPa, temperatures to 2000 °C — with a phase tag on every row (liquid, vapor, supercritical fluid, and the saturated boundary points). It carries specific volume, density, internal energy, enthalpy and entropy.

Steam-cycle work runs on enthalpy and entropy differences. In a Rankine cycle you read h at the boiler outlet and turbine outlet to get the work and heat per kilogram; entropy fixes the ideal (isentropic) end state before efficiency is applied. Condenser and feed-heater duties are enthalpy balances; relief-valve and trap sizing leans on saturation pressure and specific volume.

In the two-phase region a state isn't fixed by pressure alone — you also need quality, x (the vapor mass fraction). Any property is then the liquid value plus x times the vaporization term, e.g. h = h_f + x·h_fg. That single relationship, built from the f and fg columns, is why saturated tables list both. Specific volume (v) feeds pipe and vessel sizing; density (ρ) in the superheated table feeds mass-flow and velocity checks.

A table is a grid of discrete points, not a continuous function. Real operating states almost always fall between rows, so a value read straight off the nearest line carries a small error unless you interpolate. The grid is also finite: this data stops at 1000 MPa and 2000 °C, and nothing in it authorizes extrapolation past those limits.

Other boundaries to respect: the data is pure water at thermodynamic equilibrium — not brines, glycol mixtures, or metastable (subcooled-vapor / superheated-liquid) states. Internal energy and entropy are referenced to zero for saturated liquid at the triple point, so those numbers are only meaningful as differences. And published values carry the precision of their source; treating the last digit as exact over-states the real accuracy. A table is a high-quality reference, not a substitute for the governing equation of state or for engineering judgment.

Because states fall between rows, interpolation is part of reading a table — not an optional extra. For the saturated tables, one independent variable (T or P) sets the row, so linear interpolation between the two bracketing rows is standard and is exactly what the built-in lookup on this page does, flagging whether a result is an exact node or interpolated, and refusing to extrapolate beyond the published range.

The superheated region needs double (bilinear) interpolation — once in pressure, once in temperature — because two variables move independently. Linear interpolation is an approximation: it is most accurate where properties vary smoothly and least accurate near the critical point and along the saturation dome, where curves bend sharply. When precision is critical there, interpolate on a finer interval or go back to IAPWS-IF97 directly.