Today I received an e-mail message regarding an upcoming Energy Water Nexus Technical Conference that made this unsupported claim:

Nationally, the largest category for water consumption is electric power generation. Similarly, the largest demand for electricity is water extraction and distribution.

Given the amount of farms and lawns in the U.S., I would expect crop irrigation to be the biggest water user. And, given the population and climate in much of the country, I would expect space conditioning (heating/cooling/ventilation) to be the largest user of electricity.

Is electric generation the largest user of water, and is water infrastructure the largest user of electric energy?

  • 7
    It would not surprise me if there is not some equivocation fallacy behind that claim. Using water for cooling in a power plant only consumes water if it is using evaporation cooling towers, and the water can be both brackish and salt, depending on where the plant is. This is in contrast to drinking and irrigation water, which must be fresh. So something is iffy here, because far from all water that is used for cooling in power generation is drinkable water.
    – user32299
    Commented Nov 12, 2018 at 20:41
  • 3
    This certainly depends on what you consider to be a "category". In the EPA's data at epa.gov/energy/electricity-customers, water extraction and distribution don't appear as a category at all. Commented Nov 12, 2018 at 22:57
  • 7
    Depending on how you defined "used", this could be trivially true for the water-consumption side of things. For example, there are eleven hydroelectric dams on the Columbia River, and the full flow of the river passes through each of them. If you count each dam as a "use", you get a total of about 2 billion acre-feet, a number about equal to 40% of the total rainfall on the Lower 48.
    – Mark
    Commented Nov 13, 2018 at 0:44
  • 1
    As stated, this question does not make a notable claim. A random email message is not notable. Commented Nov 13, 2018 at 8:24
  • 1
    Yup, that USGS report that David quoted confirms it: the claim is flawed because it conflates uses with consumes, and that is wrong.
    – user32299
    Commented Nov 13, 2018 at 10:41

2 Answers 2


No, the claim is faulty on both parts

Using the US Geological Survey report on Estimated use of water in the United States in 2015, we quickly see that the original claim that thermoelectrical power-plants consume the most water is not correct.

Table 2A shows that thermoelectrical power plants do withdraw 133,000 Mgal per day, and that is on par with irrigation, and those two together are the top withdrawers. But 96-97 percent of the water that power plants withdraw is discharged back into the environment and is available for use downstream without further treatment. The consumptive use is only 4,310 Mgal per day (Table 12).

With a consumptive use of only 4,310 Mgal/day, power plants consume much less than Irrigation (Table 7), at 73,200 Mgal/day, which is 17 times as much.

So when the original claim says that:

Nationally, the largest category for water consumption is electric power generation.

...that claim is false.

As for usage of electricity, the claim that making water available is the largest consumer of electricity: no, that is not true either.

As explained by United States Energy Information Agency, the Commercial, Industrial and Residential(**) sectors share about equal parts of electricity use.

Water supply sorts under the "Commercial" sector:

The commercial sector includes retail, office, education, institutional, public, and government facilities, and public services such as water supply, sewage treatment, telecommunications equipment, and outdoor and public street lighting.

...and the commercial sector is listed as using electricity for the following:

  • Refrigeration—14.0%
  • Ventilation—11.2%
  • Lighting—10.6%
  • Space cooling—10.6%
  • Office equipment —7.8%
  • Computers and related equipment—7.5%
  • Space heating—2.6%
  • Cooking—1.8%
  • Water heating—0.6%
  • Other miscellaneous uses combined—33.3%

So water treatment and water transportation do not even have their separate posts, but instead get sorted in under misc. uses, presumably using less electricity than the 0.6% listed — in the commercial sector only, not 0.6% of all use — for water heating.

Hence that part of the claim is wrong too.

Edit: "Use" vs "Consume"

The original claim is the following...

Energy Water Nexus Technical Conference: Driving Savings in Manufacturing

Nationally, the largest category for water consumption is electric power generation. Similarly, the largest demand for electricity is water extraction and distribution. This strong interdependence drives great opportunity in the Energy-Water Nexus (EWN) to impact both energy and water consumption.

So what is an Energy Water Nexus? Turns out Wikipedia has a page on that...

There is no formal definition for the water-energy nexus - the concept refers to the relationship between the water used for energy production, including both electricity and sources of fuel such as oil and natural gas, and the energy consumed to extract, purify, deliver, heat/cool, treat and dispose of water (and wastewater) sometimes referred to as the energy intensity (EI).

Further down the page, we find the following. Emphasis is original, bold is my addition.

Water can either be used or consumed, and can be categorised as fresh, ground, surface, blue, grey or green among others. Water is considered used if it does not reduce the supply of water to downstream users, i.e. water that is taken and returned to the same source (instream use), such as in thermoelectric plants that use water for cooling and are by far the largest users of water. While used water is returned to the system for downstream uses, it has usually been degraded in some way, mainly due to thermal or chemical pollution, and the natural flow has been altered which does not factor into an assessment if only the quantity of water is considered. Water is consumed when it is removed completely from the system, such as by evaporation or consumption by crops or humans. When assessing water use all these factors must be considered as well as spatiotemporal considerations making precise determination of water use very difficult.

So, in the context of talking about Energy Water Nexus / water-energy nexus, there is indeed a vital difference between claiming water has been "used" or if it has been "consumed".

The only way the claim about thermoelectric plants and water can be true is if whoever wrote that flyer made a mistake, and erroneously wrote "consumption" when they actually meant "usage". I personally find that to be either unlikely, or really embarrassing on part of someone that supposedly calls to a conference on the subject and thus should know the difference.

And no matter that, it still does not make the second part of the claim — about electricity being used the most for producing and moving water — true.

The following graphic, from the same Wikipedia page...

enter image description here

Water / Energy usage and respective dependencies. Image Source: US Department of Energy - Bauer, D., Philbrick, M., and Vallario, B. (2014). "The Water-Energy Nexus: Challenges and Opportunities." U.S. Department of Energy. Public Domain

...shows that electricity usage to produce water and waste waster treatment is 0.1 and 0.1 Quads per year (0.1 Quad = 29 TWh) respectively, while other usage of electricity is substantially more.

The same graphic also shows a very tiny portion of water leaving the "Thermoelectric Cooling" box and going into the "Consumed Water" box. The sums of other connectors means that — according to that graphic — water is consumed at a rate of 2,000 Mgal per day out of 196,000 Mgal/day used for thermoelectric cooling.

(**) Anyone else get SimCity vibes from that? :D


Apparently the claim is mostly correct, and hydroelectric dams have little to do with it. Impounded water that falls through a turbine doesn't count as consumption. Thermoelectric consumption of water instead involves turning (somewhat) usable liquid water into one of the following: Water vapor for cooling that goes into the atmosphere, somewhat unusable waste water that is treated as sewage, or quite unusable toxic waste water that needs significantly more treatment than mere sewage. The vast majority of the consumed liquid water is converted to vapor.

The non-parenthetical "mostly" and the parenthetical "somewhat" is because some of the water consumed to generate electricity is saline water. Consumption of water to generate electricity has been dropping for the last few decades (an 18% decline since 2010 alone) while irrigation has been steadily increasing. Apparently irrigation has finally overtaken electricity generation as the number one consumer of fresh water.

Since 1950, the US Geological Survey has produced a report on estimated water use in the US once every five years. The most recent is Estimated use of water in the United States in 2015. This document lists hydroelectric generation as the number one consumer, followed closely by irrigation. The following graphic is from that document.

Graphic depicting various consumers of water in the US. Thermoelectric power is #1, using 41% of the supply, followed closely by irrigation at 37%.

Some of the 41% of the total water supply consumed by thermoelectric electricity production is saline water. Excluding that saline water, thermoelectric consumption came in at number two in 2015, just behind irrigation. This turnaround is very recent. The document sites an 18% reduction in water consumption for generation of electricity. Up until 2015, thermoelectric was the number one consumer in the US of fresh water, well ahead of irrigation consumption.

  • 3
    I have to down-vote this answer until there is an edit because it fails to distinguish between use and consumption. In the report quoted, thermoelectric powerplants use 95,100 (fresh) + 37,800 (saline) = 133,000 million gallons per day (Table 2A), but they only consume 3,760 (fresh) + 547 (saline) = 4,310 million gallons per day (Table 12). The original claim is about consumption, i.e. water that gets withdrawn and then not discharged again for further use down the line, but this answer equals use with consumption, and that makes for a factually faulty answer.
    – user32299
    Commented Nov 13, 2018 at 10:33
  • 2
    The image describes "withdrawals", but "withdrawal" is not the same as "consumption". Most of the water (~96-97%) that is withdrawn by thermoelectric power-plants is redeposited and can be used downstream; that portion is not consumed. So it appears the original claim is indeed deceiving us by claiming the water is consumed.
    – user32299
    Commented Nov 13, 2018 at 10:39
  • @MichaelK - IMNHO, you are reading the document incorrectly. That said, the cited document isn't that great. Commented Nov 13, 2018 at 10:59
  • 1
    @MichaelK - Moreover, the vast majority of the water consumed by most other processes can, after treatment, be used further downstream. This is why we have sewage treatment plants. This also applies to irrigation. A good chunk of the water used for irrigation goes toward transpiration, a tiny bit goes into the plant, but most of the water consumed for irrigation comes out of the system as water that is somewhat more polluted than the water that went in to irrigate the crops. Commented Nov 13, 2018 at 11:03
  • Does thermoelectric cooling water which is withdrawn from a river or the sea and then returned to the same river or sea a few degrees warmer even need further treatment?
    – Henry
    Commented Nov 13, 2018 at 11:12

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .