Resizing Electricity's Water Footprint

Among all the resources that go into producing electricity, water often gets lost in the shuffle. Here’s what we know: It can take a lot of water to generate electricity. How much? Well, that’s a complicated question.

Researchers have estimated the often massive volumes of cooling water used by different kinds of power plants, but those estimates don’t take into account the many other thirsty steps required to generate electricity, from extracting fuel from the ground to manufacturing all the parts of a power plant. A new paper takes a crack at creating a water footprint of the entire life cycle of electricity, and in so doing reveals a lot of gaps in our knowledge.

Making these water withdrawal and consumption (make sure you know the difference!) estimates is really tough because there are so many variables. Besides the different electricity generating technologies involved – coal, gas, nuclear, solar, geothermal, wind – there are a large number of questions to consider:

How much water is required in the fuel cycle to extract, process and transport fuels to a power plant? How much water is needed to construct a plant or make all the necessary components? How long is the plant expected to generate electricity? How about the water needed to decommission the plant?

Researchers screened thousands of scientific papers, government reports and corporate sustainability reports in an attempt to answer these questions. So how much water does generating electricity require? That depends. In fact, there is so much variation that your best bet is to review the full results online for specifics. Despite all the variability, there are some important generalizations that can be made to shed light on how the water footprint of electricity includes a lot more than just cooling water.

  • While water used for cooling steam at thermoelectric plants (which create steam to spin turbines and generate electricity) dominates the water footprint of most sources of electricity, the amount of water needed for the fuel cycle – extracting, processing and transporting fuel – and construction of plants can be substantial.
  • For thermoelectric technologies, more water is consumed during the operations phase – and here that primarily means cooling – than any other phase of electricity generation.
  • For non-thermal renewable technologies that have no need for cooling water, like wind and solar photovoltaics, most water is consumed when equipment is manufactured, although this requires very little water.
  • For coal, natural gas and nuclear, the water demand of the fuel cycle may be small when compared to cooling, but still represents a significant water use.
  • Carbon capture and sequestration, a set of technologies that can greatly reduce CO2 emissions from new and existing coal- and gas-fired power plants and large industrial sources, is anything but water-friendly. These CO2-reducing technologies can increase water consumption by 75 percent and water withdrawal up to 97 percent.
  • Hydraulic fracturing can quadruple the amount of water consumed when extracting natural gas from the ground as compared to more conventional methods. Still, the amount of water consumed in natural gas plants, even with hydraulic fracturing, is similar or even less per unit of electricity produced than that consumed in coal refineries or nuclear power plants.

When looking at any of the water withdrawal and consumption figures assembled in this paper, keep in mind that, as the authors say, the gathered estimates remain “few in number, wide in range and many are of questionable original quality.” Yet this paper represents an important first step towards creating a true water footprint of electricity by collecting every estimate we have up to now. 

Why does this matter? The drought in Texas is causing concern for power plant owners. Nuclear power plants were threatened by high floodwaters in Nebraska. In Alabama and Connecticut, power plants were ramped down because cooling water temperatures were too high. The growth in hydraulic fracturing is presenting a new stress on water supplies. Power plants across the nation kill trillions of fish and aquatic life every year because of their cooling water withdrawals. All of these issues arise because of electricity’s reliance on ample supplies of water throughout its life cycle, something that is increasingly threatened in a water-constrained world.

As the paper suggests, there are big data gaps for many of the life cycle stages of electricity, meaning there’s a big opportunity out there for researchers who can begin to answer an increasing number of questions from power executives, water managers and communities, all of whom need to plan ahead for a changing climate and shifting food, water and energy needs.