What amount of energy would it take to replace the entirety of America's water consumption with desalinated water? And how does it compare with the country's current energy generation capacity?

By "water consumption" I generally mean things that one usually thinks of when thinking of using water, including but not limited to household, commercial, agricultural, mining, and industrial uses. I'm not talking about filling reservoirs for recreation, hydroelectricity generation, or outflows meant to preserve fisheries. If there is another case worth considering, feel free to ask, or simply note it in the answer.

By "amount of energy", I mean to take seawater and turn it into usable water, not to pump such desalinated water from, say, San Diego to Denver. This would clearly be a non-trivial amount, but I think it would be too hard to calculate with much precision. (I guess the lower bound would be the current hydroelectric generation capacity of the USA, and a reasonable upper bound would be 10x that amount.)

  • 1
    $\begingroup$ Do you count the trillions of $ in energy to build the plants and pipelines ? $\endgroup$ May 18 at 19:00
  • $\begingroup$ You should be able to look up the average water consumption per person, etc., and add the data to your question. Similarly, desalination energy consumption figures should be available. Where did you get stuck? Note that desalination plants have to return the brine to the ocean and small changes in salinity cause stratification and possible death of marine life in the saltier water. Plenty on YouTube on the subject. $\endgroup$
    – Transistor
    May 18 at 19:55
  • 1
    $\begingroup$ The wikipedia page on the subject actually mentions the question being asked here, under the "design aspects" section. Claims +10% of total domestic energy use could thus supply all domestic water use by desalination, citing an industry publication of the membrane technology people. So take that with a grain of salt (no pun intended). The even bigger catch has to be the fresh water used in industry and agriculture. $\endgroup$
    – Pete W
    May 18 at 23:56
  • $\begingroup$ I think you've asked an extremely broad question that really needs to have better boundaries set in order to make this remotely answerable. For starters, you ask, "how much energy would it take to replace America's water consumption with desalinated water," and therein include, "household, commercial, agricultural, mining, and industrial uses." Can you advise on what data sources you want to be considered? Over what time frame are you referring to (daily average, annual, summer)? Note that industrial uses may be fine with non-potable water, which changes the energy a lot. $\endgroup$ Jun 15 at 17:51

Amount of Water Used

The amount of water used in the USA in 2015 for all uses other than thermoelectric power (1) (read: to cool power plants) was 189B gal/day, or 715M m^3/day.

Amount of Energy Needed

The amount of electrical energy it takes to desalinate ocean water is in the range of 3-5.5 kWh/m^3.

Therefore the amount of electrical energy it takes to desalinate a day's worth of water in the USA is at least 2.15B kWh/day.

Comparison with Amount of Energy Produced

The amount of electrical energy produced in the USA in 2020 was 4.01T kWh, which works out to 10.93B kWh/day. This means the USA would need to increase electrical energy production by 19.7% in order to derive the fresh water for the USA's daily water needs from ocean water (2).

Electrical energy production has historically increased at the rate of 4% per year (4), meaning that this energy need could be met in about five years, not accounting for changes in energy and water use.

Suppose we wanted to derive the energy solely from renewables. The amount of electrical energy produced from renewable sources (not including hydropower) in the USA for 2020 was 501B kWh/year or 1.37B kWh/day. By increasing the electrical energy production of non-hydro renewables by 157%, America could completely replace its water needs via desalination.


  1. I omit power plant cooling because either this often already comes from seawater, or the use of this water can be sharply reduced by converting to closed-loop systems. ("Powerplants that used once-through cooling systems accounted for 96 percent of all thermoelectric-power withdrawals.")
  2. Deriving this fresh water via waste water processing would take much less energy.
  3. This compares 2020 energy production against 2015 water consumption, but this shouldn't keep us from getting to a pretty close solution.
  4. Yes I referenced Engineering Explained directly rather than derive the growth rate directly. Fwiw we use the same source for USA electrical energy production.
  • $\begingroup$ You would think people would want that wastewater distilled anyways if everything else was being distilled too. Heh. $\endgroup$
    – DKNguyen
    Aug 23 at 21:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.