GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted
Energy Solutions

How Much Water Does it Take to Turn on a Light Bulb?

A look at the water intensity (or water footprint) of electricity generation

Image 1 of 3
These cooling towers are at at a nuclear power plant in Byron, Illinois.
Image Credit: Scott Olson - Getty Images
These cooling towers are at at a nuclear power plant in Byron, Illinois.
Image Credit: Scott Olson - Getty Images
The massive 377-MW Ivanpah solar-thermal power plant is located in California's Mojave Desert. BrightSource claims that this technology is 95% more water efficient than other concentrating systems.
Image Credit: BrightSource Energy
This array of dual-axis-tracking 25-kW SunCatcher collectors uses Stirling engine technology, which does not require water.
Image Credit: Stirling Energy Systems

In last week’s blog I took a look at some of the water conservation features in our new house, but I began the blog by addressing the relationship between water and energy. That got me curious, so I’ve been digging deeper into this water-energy nexus, examining the water-intensity of our different electricity sources.

Some of this information is drawn from a 2012 report that I only recently came across: “Burning Our Rivers: The Water Footprint of Electricity,” by the River Network in Portland, Oregon.

Evaporative losses from hydropower plants

Nearly all of our methods for generating electricity involve at least some water consumption, but the differences are huge. Producing electricity with hydropower is the most water-intensive, owing to evaporation from reservoirs. Nationwide, electricity from hydropower plants consumes about 9 gallons of water per kilowatt-hour (kWh) of electricity produced.

In some parts of the world, this evaporation is a big problem. In other areas, not so much. In the arid Southwestern U.S. this evaporation is a huge issue, especially from reservoirs like Lake Mead.

Water use for thermoelectric power plants

Most electricity in the U.S. (about 89%) is produced using thermoelectric power plants. These use a heat source (most commonly coal, natural gas, or nuclear fission) to boil water, creating superheated, high-pressure steam. This steam spins a steam turbine connected to generate electricity. Cooling water is then used to condense the steam back to water.

Depending on the type and age of the power plant, the cooling water is once-through (pulled from a river, for example and then returned to the river at a higher temperature), provided by a cooling pond, or recirculating. The once-through systems use tremendous quantities of water, but the vast majority returns to the water source from whence it was drawn — albeit at a higher temperature (thermal pollution can be a major problem). Some evaporates, however, and is not returned to the river; this is the consumptive use.  

Recirculating cooling systems in power plants use far less water and they don’t add thermal pollution to the body of water from which the water was originally drawn, but they still evaporate considerable water — in fact, typically more than once-through cooling systems — so the consumptive water is very significant.

Comparing coal, natural gas and nuclear relative to water use

Of the three primary fuels used in thermoelectric power plants, natural gas power plants have the lowest water intensity. According to Burning Our Rivers, coal power plants consume 0.69 gallons of water per kWh of electricity produced, natural gas power plants consume 0.17 gallons/kWh, and nuclear plants 0.57 gallons/kWh.

With coal, according to the report, 73% (0.506 gal/kWh) of the water consumption is from evaporation, as described above, while 27% (0.186 gal/kWh) is from upstream sources (mostly mining, and transportation). Once-through cooling of coal plants results in consumptive water use (evaporation) of about 0.3 gal/kWh, while recirculating systems evaporate about 0.7 gal/kWh.

Water consumption from nuclear plants is similar to that of coal though the spread between once-through and recirculating systems is even greater: 0.27 gal/kWh for once-through cooling vs. 0.76 for recirculating systems.

While the water intensity of natural gas power generation is a lot lower than for coal and nuclear, there are significant differences depending on the type of power plant. Combined-cycle plants use much use much less water per unit of electricity output than do single-cycle power plants.

The Burning Our Rivers report shows very low upstream water consumption for natural gas power plants, but the report did not consider hydraulic fracturing (fracking), which results in far greater water use (typically 4-5 million gallons per well) and heavily contaminates that water. An October 2013 report on the water intensity of natural gas extraction from Marcellus Shale in Pennsylvania and West Virginia by researchers at Downstream Strategies and San José University sheds some light on this issue.

There are two primary ways electricity is generated from solar: utility-scale solar-thermal power plants and either utility-scale or building-scale photovoltaic power generation. Most utility-scale solar thermal is more water-intensive than coal or nuclear power plants.

From the Burning Our Rivers report, parabolic trough systems are shown to consume about 0.80 gal/kWh, while linear Fresnel systems consume about 1.0 gal/kWh, solar power tower systems consume 0.63 gal/kWh, and dish Stirling Engine systems, which are far less common but do not use the heat to generate steam, consume only 0.020 gal/kWh.

Adding to the challenge with large-scale solar-thermal is that these systems want to be located where there is a lot of sunlight, such as the American Southwest, and those places tend to be much drier.

Photovoltaic systems use almost no water in their operation — only 0.002 gal/kWh — with most of that upstream water use for manufacturing.

Finally, wind systems consume less than 0.001 gal/kWh — the lowest of any electricity source — with most of that also upstream.

Low global warming and low water use

It is worth pointing out that the renewable energy technologies for power generation that are growing the quickest in implementation (photovoltaics and wind) are the least water-intensive.

The only measures that do even better from a water-use standpoint are efficiency measures. Using less electricity is the place to start if the goal is to conserve water resources.

Alex is founder of BuildingGreen, Inc. and executive editor of Environmental Building News. In 2012 he founded the Resilient Design Institute. To keep up with Alex’s latest articles and musings, you can sign up for his Twitter feed.


  1. JonathanTE | | #1

    Alex, I don't understand your grammar:

    "Combined-cycle [natural gas] plants are nearly two-and-a-half times as water-efficient as single-cycle power plants."

    Which one is more water efficient?

  2. JonathanTE | | #2

    AJ - whuh?
    AJ, so what's your theory on how the evaporation of the Colorado River helps to reduce cholera in Africa and Asia?

    When the expert(s) says get in the life raft and the "common sense" dude on the internet says "bugger it all, I've been through worse," I'm thinking that expert looks pretty reasonable. Might be a mistake in retrospect, but we're all just playing the odds here.

  3. Alex Wilson | | #3

    Combined-cyle power plants more water efficient
    Sorry for the confusing phrase. Combined-cycle natural gas power plants use a lot less water per unit of electricity output. I think I got that information from the first report I provided a link to: "Burning Our Rivers." I'm not sure if that is solely a function of the combined-cycle power plants being more energy-efficient or if the reduction in water use is greater than the reduction in natural gas use in going from single-cycle to combined-cycle plants.

  4. wjrobinson | | #4

    Evaporation is a good thing too.

    Clouds reflect heat. Rain distributes water.

    "The sky is falling" is not all bad news....

    Experts make mistakes.... If your expert ferry captain says stay put as the ship sinks.... Time to jump ship and use your own common sense.

    2nd point. Compare water quality destruction in the developed world compared to what is done to water where people have little to no infrastructure. Billions trillions of gallons of water filled with human pollutants... Rivers none of us wants to step in let alone drink. And what we need to worry about our power plants creating clean rain via globe cooling cloud makers?

    I'll take the clean clouds, you can have the crap filled rivers.

    Edit for Jonathon: I thought what I posted was clear. More clarity... The blog points to evaporation being a liability, and I am pointing out a point not mentioned that would make evaporation be a positive in that it may balance via reflecting solar while CO2 contains solar. Get it Jonathon??

    As experts, the last two expert ship captains that ended up killing their passengers... told them to stay put when someone like me that has his own brain would laugh at that nonsense and get the hell out. Experts that don't see all the whole picture make expert mistakes.

    Now as to poor countries with pollution, to me raw pollution at that level is a much more urgent and real problem for living and for living any quality of life not evaporation of the Colorado which we should be smart enough in the US to limit water use growth in the southwest. If not smart enough then the southwest gets... the "Darwin award" and deserves it.

    Jonathon, if you don't get me by now... you never will... enjoy your time at GBA.

Log in or create an account to post a comment.


Recent Questions and Replies

  • |
  • |
  • |
  • |