Open data

What is the difference between water withdrawal and water consumption?

Short answer

Water withdrawal is the total volume taken from a river, lake or aquifer for use. Water consumption is the part of that volume that does not return to the local water system, mainly lost to evaporation, plant transpiration or incorporation into products. Withdrawal is never smaller than consumption, and datasets such as FAO AQUASTAT report withdrawals, not consumption.

The two words describe different points in the same flow of water. Mixing them up is the most common error when reading water statistics, because the numbers can differ by a large factor.

How are water withdrawal and consumption defined?

  • Withdrawal (also called abstraction or extraction) is water taken out of its source for use. FAO AQUASTAT, the Food and Agriculture Organization's global water database, defines total water withdrawal as the annual quantity of water withdrawn for agricultural, industrial and municipal purposes, and adds up those three sectors. Its figures describe water taken out of the system, whether or not it later returns.
  • Consumption, or consumptive use, is the part of withdrawn water that is not returned to the local system. The US Geological Survey (USGS) defines consumptive use as the part of water withdrawn that is evaporated, transpired by plants, incorporated into products or crops, consumed by humans or livestock, or otherwise removed from the immediate water environment.
  • Return flow is the remainder: water that was withdrawn, used and discharged or drained back to a river, lake or aquifer, where it may be used again downstream.

So consumption = withdrawal - return flow, and withdrawal is always at least as large as consumption.

How do withdrawal and consumption differ for a farm or a power plant?

The following figures are made up to show the arithmetic, not real data. A farm withdraws 100 units of water for irrigation. Crops transpire and soil evaporates 55 units, and 45 units seep back to groundwater or run off to a stream.

  • Withdrawal: 100 units
  • Consumption: 55 units
  • Return flow: 45 units

A thermal power plant with once-through cooling shows the other extreme: a large withdrawal, most of it returned, and a small consumption. The same withdrawal number can therefore mean very different effects on the water source, depending on how much comes back and in what condition.

Where do World Bank freshwater withdrawal figures come from?

The World Bank series Annual freshwater withdrawals, total (billion cubic meters), indicator ER.H2O.FWTL.K3, comes from AQUASTAT. The World Bank defines it as total water withdrawals, not counting evaporation losses from storage basins. It includes desalinated water in countries where that is a significant source, and it splits use into agriculture (irrigation and livestock), industry (including cooling of thermoelectric plants) and domestic use (drinking water and municipal supply).

Two features of the data matter for interpretation:

  • The data are withdrawals, not consumption. A country with a large withdrawal is not necessarily consuming that much, and the data cannot tell you how much returned.
  • Withdrawals as a percentage of internal renewable resources can exceed 100 percent. That happens where nonrenewable aquifers or desalination plants contribute substantially, or where there is significant water reuse.

Can I compare withdrawal data with consumption or footprint data?

  • Do not compare a withdrawal figure to a consumption or water footprint figure without checking definitions.
  • Per-capita withdrawal is a measure of use of the source, not of personal consumption.
  • To estimate consumption you need sector-specific information, such as crop evapotranspiration for irrigation, from a different source than the withdrawal tables.
  • When comparing countries, look at the sector split as well as the total, since the mix of agriculture, industry and domestic use differs widely between them.

The freshwater withdrawals by country sheet holds the withdrawal series in a workbook for sorting and charting, and the blog post on global freshwater withdrawals and what the data shows goes through the patterns. Remember to credit the sources when you publish, as explained in how to cite World Bank data.