Water Footprint: a stumbling block?

As India starts to debate water use efficiency, various tools are being considered. The clear and present danger is blind adoption of a tool without examining its limitations and suitability to Indian conditions. The Water Footprint (WF) is one such tool under consideration of the government to assess water efficiency.

In 2002, Arjen Hoekstra, professor of water management at University of Twente, The Netherlands, introduced the concept. WF is an indicator of freshwater use that looks at the direct and indirect water use of a consumer or producer. It can be considered a comprehensive indicator of freshwater resource use instead of merely a measure of water withdrawals.

The water footprint of a product is the volume of fresh water used in its production, measured over the full supply chain, says the Water Footprint Assessment Manual, Earthscan, 2011. It is a multi-dimensional indicator, showing water consumption volumes by sources and polluted water volume by the volume of pollution. The components are specified geographically and temporally and divided into blue, green and grey.

Some definitions are in order: consumption under WF is the loss of water from the total volume of water within a catchment area due to evaporation, transfer to another catchment or the sea, or its incorporation into a product; blue water is surface and groundwater; green water refers to rainwater; and grey water is the volume of fresh water needed to assimilate the pollution. WF differs from conventional measurements of water use in three ways, says Ruth Mathews, executive director of the Water Footprint Network (WFN), and, therefore, it offers a better perspective of how a consumer or producer relates to the use of water. 

It is a volumetric measure of water use and pollution but does not measure the severity of the local environmental impact of either. It does not include blue water use insofar as this water recycles to where it came from and includes green and grey water. WF also includes indirect water use in processes—for example, water used in the supply chain. Thus, the footprint of a product is the total of water used in growing, transporting and manufacturing it, and the water required to assimilate pollution.

A related concept is virtual water that was defined by John Anthony Allan, professor at King’s College in London, as the volume of fresh water used to produce a product measured at the place where the product is actually made. Like WF, it encompasses all the water used in the supply chain and, therefore, the production process.

Both these have two practical uses, say Vijay Kumar and Sharad Jain of the National Institute of Hydrology, Roorkee. Writing in Current Science, they contend they can be a way to regional water security by helping decide which industries to allow in a region, or which crops to grow that are suited to local water resources. Additionally, the virtual water contained within a product is indicative of the environmental impact of producing it. Water footprints can be as specific or general, as needed, ranging from a single product to an entire country.

Calculating the footprint

The most common use of the water footprint is to calculate the quantity of water used to grow a unit (kilogram) of food or produce a specified weight of a product. Calculations by WFN show it takes 822 litres of water to grow a kilo of apples. For a kilo of beef, it is 15,415 litres, and 5,521 litres for a kilo of mutton. Rice needs 2,497 litres while potatoes need 287 litres a kilo. However, these are global averages.

To go beyond superficial global averages, it is necessary to look at temporal and geographical specifics. For example, for a cotton shirt, the raw material may be grown in China, manufactured in Malaysia and sold in Germany. These are national averages and production circumstances and processes differ; the place of production of each component will influence size and composition of the shirt’s water footprint. The footprint will change with the production process.

According to WFN’s Water Footprint Manual, there are two ways to calculate a footprint: the Chain Summation Approach and the Stepwise Accumulative Approach. The former is used for particular cases while the latter is a more generic approach. In the Chain Summation Approach, water footprints associated with the process can be fully attributed to the product. The water footprint is the sum of water consumed by each individual process that constitute the production without double accounting. 

In the Stepwise Accumulative Approach, the calculation is based on the water footprints of the inputs needed at the last processing step. In other words, if there are several input products, the water footprint of the final product can be calculated by adding up those of the inputs and the process. This is a more rapid assessment.

To calculate the water footprint of a business, the water used directly and polluted by the business has to be added to what is similarly used by the business’ supply chain. This business water footprint is divided into the blue, green and grey components. This is useful from a business’ perspective as businesses can control their operational water footprints directly, and those of their supply chain partners indirectly. Thus, a business can make an informed choice about water policy and decide where to site a production unit. Similarly, policy makers can use water footprints to decide cropping patterns based on the amount of water available.

The criticism

However, WF is a uni-dimensional tool as it only considers water as an input without accounting for other factors. Critics point out the footprint is not politically neutral as it influences a country’s agriculture, industry and livelihoods. In fact, the use of just one tool can impact food security if it is the only way to measure if a country should or should not grow a particular crop. For example, a country facing shortages of water and food will grow or import the kind of food its population requires regardless of its water footprint.

The water footprint, says Denis Wichelns of the Institute of Water Policy at Lee Kwan Yew School of Public Policy in Singapore, is the ratio of the estimated water used, diverted or consumed to the crop yield or industrial output. It focuses on water without considering other factors of production such as geography, opportunity costs of land, employment, energy use and resource constraints. For example, in the supply chain, many other factors are at work such as employment, energy use and efficiencies that are not captured by the water footprint.

In other words, water footprints do not describe the role and relative importance of water in production; it is a quantitative measure. They do not cover water scarcity or sustainability of the source of water nor the productivity or livelihoods. “Water footprints are a one-dimensional ratio that cannot provide helpful insights into complex policy issues,” says Wichelns.

Water Footprints assign the same value to water regardless of how critical it is to the product and other variables that affect productivity. For example, in agriculture many other inputs determine the output but the water footprint ignores these. The ratio remains the same: water (m3) / output (tonne). But the calculation does not factor in land, soils, fertilizers, pests and pesticides, labour and weather. Wichelns says the difference in outputs may be due to non-water inputs but the footprint of the same crop grown in two farms will vary. Thus, comparing them is meaningless.

For example, so-called dry land crops that are believed to be less water guzzling have a higher water footprint as there is a gap of almost 300 per cent between potential and actual productivity, says Mahtab Bamji from the Dongoria Trust, Hyderabad. The main reasons are sub-optimal irrigation and poor soils. This perhaps explains why the farmers prefer to cultivate paddy as long as there is water to grow it even in dry areas. In other words, water is a small part of the total cost of cultivation and other factors determine where crops are grown and consumed.

The same logic applies to industry where machines, chemicals, manpower, power and organisational efficiencies determine productivity; the water footprint does not capture these. A more efficient process will turn out a product with a lower water footprint but the footprint calculation will not reflect this. Equally, a production process with a low water footprint may use more energy and, therefore, be more environmentally damaging. Therefore, in industry also comparing outputs from different companies is meaningless. 

The variations are magnified at the lowest level. For example, farm A may grow wheat using 1,100 cubic metre (cu. m) of water per tonne and farm B using 900 cu. m. However, A may be more mechanized while B may have got more timely rainfall. Thus, there are farm-to-farm fluctuations that have little to do with water use. The same applies at the regional and national levels.

Temporal variations also reduce the Water Footprint’s usefulness. In a given year, a farmer may grow a tonne of apples using 600 cu. m of water, but may need 800 cu. m the next year on account of a delayed flowering caused by spring frost. In sum, water footprints do not tell us anything about productivity, values or livelihoods. They only tell us about the amount of water used for a particular item in a particular context.

Another problem, according to Wichelns, relates to the components of the footprint – blue, green and grey water. There is no scientific basis for this separation and the opportunity cost for soil moisture (green water) can be quite large. The grey water does not give any information about the kind of pollution.

Despite its limitations, the water footprint can possibly be one tool to measure water consumption by farming or industrial processes, but not the only one. India has other compulsions – livelihoods and health – that may outweigh a uni-dimensional and context-specific approach. While it is desirable to have a yardstick to assess water use, it needs to be wielded in the local development context and not in isolation that may lead to erroneous conclusions.