Dissolved Oxygen Saturation with Temperature

Dissolved oxygen is the fuel for nearly all corrosion in ordinary water. The metal supplies electrons by dissolving; something has to accept them, and outside acid conditions that something is almost always oxygen being reduced at the surface. Take the oxygen away and the reaction has nowhere to go, which is why deaeration, oxygen scavengers and sealed loops are the first line of defence in every water system that carries steel.

Gases dissolve less readily in warm water, which runs against the intuition built from dissolving sugar. The reason is that dissolution of a gas is exothermic — the molecule gives up energy when it settles into solution — so raising the temperature shifts the equilibrium back toward the gas phase, exactly as Le Châtelier's principle predicts. Fresh water at one atmosphere holds about 14.6 mg/L at the freezing point, 9.1 at 20 °C and 7.6 at 30 °C: nearly halved across the range of an ordinary summer.

The equation given here is the Benson and Krause fit, in the form adopted by APHA Standard Methods and published by the USGS. It is a four-term polynomial in the reciprocal of absolute temperature, and it reproduces the published saturation table to better than 0.01 mg/L across 0–40 °C. It is a correlation, not a theory, and it should not be extrapolated far past that range — for boiler and deaerator work at elevated temperature and pressure, steam-cycle solubility data is the right source and this one will mislead.

Three corrections separate the number this gives from what a meter reads. Salinity lowers it, by about 15–20% at full seawater. Pressure scales it almost proportionally with the barometric pressure, so a lake at 2000 m holds roughly a fifth less than the same water at sea level, and altitude correction is not optional in mountain work. And saturation is a ceiling, not a state: a real system may sit anywhere below it. A closed loop that has run sealed for a week has consumed its oxygen against its own pipe walls and sits near zero; a stream below a weir can briefly exceed saturation from entrained air.

That last distinction resolves a paradox that puzzles people about heating systems. Hot water holds less oxygen, yet hot systems corrode faster — because temperature accelerates the reaction more than the falling solubility slows it, roughly doubling the rate every 20–30 °C. But that is only true while oxygen keeps arriving. A properly sealed closed loop consumes its initial charge in days and then becomes remarkably benign, corroding at rates that would be negligible over decades. A closed loop that keeps corroding is a closed loop that is being fed fresh water from somewhere — a leaking gland, an open expansion tank, a failed air separator, or make-up replacing a leak nobody has found. Chasing the oxygen source is almost always more productive than chasing the chemistry.