Water: the withdrawal/consumption confusion

I’d like to sketch out a personal project – something I think needs doing. And, given that this blog’s modest audience includes some other folks who are, like me, involved in public communication on water issues, maybe it’ll serve as a prod for others.

There is an important confusion in public discussions over two very different meanings of mean water “use”.

The first is the water we take out of the system to do something with – a pump stuck down in the groundwater, a canal diverting water from a river, etc. In water wonkery, that’s called a “withdrawal”.

Once withdrawn, some of that water then gets “used up” completely. The water I put on my garden evaporates or transpires from the leaves of the lovely little desert willow. That water has been “consumed”. But some of the withdrawn water, I put back via a toilet flush that makes its way through the Albuquerque sewer plant, gets cleaned up, and is put back in the river. Six miles south of the sewage treatment plant, an agricultural diversion dam grabs that water and diverts it for “use” by downstream farmers.

If you then follow that water downstream, there’s a “lather, rinse, repeat” discussion to be had as farmers apply water to their fields, some is lost to evapotranspiration, some ends up as tail water in ditches that return to the river, some soaks down into the aquifer. Some of the water withdrawn for power plant cooling evaporates – is “consumed”. Some is returned, warmed, to the river.

When you see numbers thrown around, like “power plant cooling represents 49 percent of water use in the United States”, or “Albuquerque residents use 150 gallons of water per person per day”, you’re most often seeing “withdrawal” numbers. And for some purposes, those are important numbers. As Charles Fishman noted in a twitter discussion this morning, power plants are competing with other users sometimes for the water. Warmed water returned from the power plants, or the stuff that makes it to our sewage outfalls, harms ecosystems. There are times when the withdrawal number matters a great deal.

But there also are discussions in which the consumption number is more important.

Let’s use water conservation here in Albuquerque as an example. For a city built in the desert, we probably “use” too much water. So we need to reduce that. Consider two paths to conservation – the low flow toilet and ripping out a patch of lawn – which might generate an equal reduction in the amount of water the local utility needs to deliver to my house each day. Ripping out the lawn reduces the amount of water lost to evapotranspiration. Switching out my toilet may reduce a like amount of water withdrawn from the aquifer or river, but also reduces the amount of water returned to the river via the sewage treatment plant. With the lawn, I’ve reduced my water “use” by both measures – withdrawal and consumption. With the toilet, I’ve primarily reduced my withdrawals.

The discussions quickly spin into complications – system losses in both the lawn and toilet case, embedded water use at the power plant to pump the toilet water to my house, water quality in the river from the treatment of my toilet water. All important discussions. But the whole thing remains entirely too murky if we’re not clear about which sense of “use” we’re talking about when we try to reduce our water use.

Consider some other examples:

  • The distinction between sewage plants that return their water to a river ecosystem, the ocean, a water source (Lake Mead, for example) or that are turned onto a golf course.
  • Putting gray water on your yard versus sending it to the sewage plant. And do I simply reduce my use of non-gray water? Or do I treat the gray water as bonus water and dump more total water on my yard?
  • Irrigation efficiency improvements that reduce withdrawals but increase crop efficiency, increasing evapotranspiration and yields but reducing ag return flows.
  • Groundwater pumping that depletes aquifers but adds surface flows to a river via sewage plant outfalls.
  • Interbasin transfers that remove water completely from one basin while adding it to another, creating enormous entanglements between the question of “withdrawal” and “use”.

For the serious water wonks, this is complex but comfortable stuff. But when the discussion moves into the public arena, lack of clarity in what we mean by “use” can be extremely confusing, and risks bad policy.

To be clear: there are still important reasons for the low-flow toilet, or making power plant cooling more efficient. But not all reductions in water “use” are the same. We need more clarity in this discussion.


  1. I had that discussion with my 9-year-old the other day. He was learning about water use in school, and wanted to conserve water by flushing less. (“If it’s yellow, let it mellow; if it’s brown, flush it down.”) I tried to figure out how to explain that the issue is different in Durango (near the Colorado River headwaters; how many times will that water be used before it ends up in the Imperial Valley or Phoenix or LA?) versus in the Bay Area (where I learned the saying, on the downstream end). I don’t think I succeeded.

  2. John, thanks for picking up the mantra for these important and much misunderstood concepts. You have managed to explain these in a clearer way than I have ever been able to do. Let’s hope your readership expands significantly and the confused public gets these important words.

  3. Kim – Good you you for the effort! One person at a time…

    Wayne – Thanks. I’d be interested to hear more about how this issue plays out where you are.

  4. Pingback: Clarity on withdrawal versus consumption, San Diego style : jfleck at inkstain

  5. Pingback: Water transfers: the view from downstream : jfleck at inkstain

Comments are closed.