# How much energy is released by decompressing water?

How much energy would be released if 2L of (relatively pure) water compressed to 1L at room temperature was suddenly decompressed to atmospheric pressure? Even a first-order approximation would be helpful.

As far as I can tell, said water would end up sitting on the border between Ice-VII and Ice-X, as according to here ice-VII has a density of ~1.50g/cc and ice-X has a density of ~2.51 g/cc.

I am aware that increasing the density of water by 2x would require an rather high pressure (to put it mildly). It's not impossible, though - based on the chart here it's "only" about 7*10^10 Pa - and we've achieved higher sustained pressures in a laboratory setting before.

I'm not sure how to calculate this, though, as I have little data on either the density of the various phases as a function of pressure, or the energy of each phase change.

• How much energy does it take to form the Ice-VII? Remember that work=force*distance. For a first approximation, assume the density of the phases doesn't depend on the pressure. – Peter Shor Jul 25 '16 at 1:57
• There are "issues" with this question. The device that could compress water to the degree indicated, and the device that could hold the resultant pressure, probably does NOT exist for a volume of 2L. If you try this with a diamond-anvil cell, you might get useful results (for 0.1 ml of volume, or less). – David White Jul 25 '16 at 2:13
• @DavidWhite - Consider it in the same vein as this question, then, if you wish. – TLW Jul 25 '16 at 3:33
• @PeterShor - it actually goes from water to ice-VI to ice-VII to partially-to-ice-X, as far as I can tell. Quick estimate works out to something like 21MJ. – TLW Jul 25 '16 at 3:45
• I've never worked with pressures as high as you are suggesting, but do know that at least with SCUBA - sized pressures I would rather prefer having a hydraulic line rupture than a gas line. While gas ruptures tend to release lots of energy, hydraulic ruptures are mild in comparison. Doesn't $\int{PdV}$ apply, and isn't dV small? – docscience Jul 25 '16 at 3:47