Potato Chips' Water Footprint

By Jeanne Yacoubou, MS

The VRG received an email inquiry from a European reader about the amount of water needed to produce one bag of potato chips listed in our Save Our Water the Vegetarian Way brochure: http://www.vrg.org/environment/water_brochure.php.

Water Footprints and Virtual Water

In the brochure we cited data from page 42 Table 4.2 of a UNESCO-IHE Institute for Water Education document authored by Chapagain and Hoekstra (referred to herein as “Report 16”) giving global averages of the virtual water content of selected foods: http://waterfootprint.org/media/downloads/Report16Vol1.pdf. Appendices of values used to calculate water footprints are found here: http://waterfootprint.org/media/downloads/Report16Vol2.pdf.

Here is visual compilation of Chapagain's and Hoekstra's data. The same numbers appear in Table 3 on page 19: http://waterfootprint.org/media/downloads/Zygmunt_2007_1.pdf

The term “water footprint” as described in the Introduction (p. 11) of Report 16 is defined as follows:

The water footprint of an individual, business or nation is defined as the total volume of freshwater that is used to produce the goods and services consumed by the individual, business or nation. Since not all goods consumed in one particular country are produced in that country, the water footprint consists of two parts: use of domestic water resources and use of water outside the borders of the country. In order to give a complete picture of water use, the water footprint includes both the water withdrawn from surface and groundwater and the use of soil water (in agricultural production).

“Virtual water” in the Summary to Report 16 (p. 9) is defined as “the volume of water required to produce a commodity or service. International trade of commodities implies flows of virtual water over large distances.” On page 13 Chapagain and Hoekstra further elaborate: virtual water varies “...as a function of place and conditions of production.”

A Potato Industry Employee's Perspective: Water Footprint of Potato Chips

The inquirer told us that he had been in the European potato industry for several years responsible for approximately 300,000 t (approx. 661.4 million pounds) of potatoes used for chips annually in a few countries. He thought that the value (185 L = 48.9 gal.) given in our brochure's table for water needed to produce one 200 g (7 oz.) bag of potato chips was too low.

He suggested that the water usage involved in producing a bag of potato chips from start to finish is significantly higher and should include the water involved in all of the following steps:

  • potato seed production (fresh seed used every year);
  • fertilizer and pesticide production;
  • water loss from outgrade potatoes not meeting all quality specifications and thereby discarded from production lines (on farms: 10-20% and in the chip plants: 3-5%);
  • storage from October to June where there is an ongoing water need to humidify and regulate temperature and ventilation;
  • potato (and thereby water) losses when in long-term storage due to inadequate conditions (8-50% of crop);
  • transporting, washing and processing potatoes during chip production;
  • producing, processing and transporting vegetable oils including palm oil used in the final product;
  • packaging materials manufacture;
  • transportation from chip plant to distributors and on to retailers.

Our reader told us:

My opinion is based on my experience in the potato industry. I also cite Pimentel's work which states it takes 500 L of water to grow one kilogram of potatoes. By comparison your brochure states it's only half of that. http://www.kysq.org/docs/Pimentel97.pdf (p. 100)

He related to us a farming scenario when he was an agricultural manager during a drought year:

Can you imagine irrigation equipment working 24/7 on one farm which pumped approximately 5,000 cubic meters (approx.1.3 million gallons) of water per hour onto fields by 8 pumping stations from two rivers and four wells?

...The total area of the farm was more than 6,000 ha (hectares) (23.2 square miles)...potatoes grew on 1,000 ha (3.9 square miles)...The irrigation system was used for vegetables (1/3) and potatoes (2/3)...The biggest unit there was a center pivot supplying 140 ha (0.54 sq. mi.) with a capacity of 600+ cubic meters (158,500+ gallons) of water per hr. used for potatoes only.

Since the irrigation was needed under drought conditions The VRG asked if the system operated during years with normal rainfall. We were informed:

In a “normal” year the irrigation system would work 30-50% below the drought year rate but it depends on the soil type and quality; potato variety; climate at a particular farm location; and weather conditions (rainfall and temperature) in a given year.

In the United States it depends on the state. For example Wisconsin has heavier soil [so irrigation is not needed as much], but in Texas' sandy soil (like in a desert) they irrigate regularly. Here's a picture: pivot irrigation - TX : http://www.gettyimages.com/detail/photo/irrigated-farms-west-texas- usa-high-res-stock-photography/124764883

In a follow up discussion, The VRG learned more about the complexities involved in calculating water footprints for potato chips:

The paradox with potatoes is that the water supplied to fields and later in the factory is the same water that is removed: potatoes retain 78-81% of water but potato chips, only 2.5-3%. In other words from a single one-ton (2,205 lbs.) capacity chip production line the amount of water which must be removed from potatoes exceeds 2.2 t of steam per hour (4,850 lbs. of steam per hour = 9.7 gallons of water per minute).

What happens to the removed water?

I witnessed the water just going up in the air; you could see plenty of steam above every chips factory. I raised the subject of wasted energy and water many years ago, but there was no intention to change anything as there was no external pressure at all.

Another aspect of potato chip manufacture came to light during our discussion. The potato industry manager stated:

...In chips there is 30-35% fat from the oil the chips were fried in during processing. So if we add all water used in the supply chain for all ingredients including the vegetable oil...such as local sunflower or canola oil and imported palm oil...the water usage will be much higher than calculated by the authors you cite in your brochure.

Although VRG's research in October 2015 showed that palm oil is not listed as an ingredient in the top ten potato chip brands sold in the US, palm oil is used in some potato chips sold in Europe. We learned that

Palm oil is used in potato chips in some European countries...it may be used as part of a blend of sunflower and palm oils or used alone...I recently saw a bag of chips in the supermarket that contained only palm oil...In other cases the proportion of palm oil is going down from year to year but it is still in usage.

I saw a report showing that European Union countries imported 1,400 tons of palm oil in 2014...The interesting fact is that a large European sunflower and canola oil producer imports palm oil regularly.

Here is a link to a European chip company stating that it uses palm oil in its products: http://www.lorenz-snackworld.com/sustainability/farming/palmoil

Interested readers may learn more about palm oil production and harvest as well as efforts to find alternatives here: http://www.pri.org/stories/2015-09-19/yes-palm-oil-destructive-scientists-are-creating-compelling-alternatives

American Potato Chip Consumption

Recent United States Department of Agriculture (USDA) data place annual US potato chip consumption at 17 lbs. per person: http://www.ers.usda.gov/topics/crops/vegetables-pulses/potatoes.aspx

To calculate how much raw commodity it takes to produce this quantity of chips The VRG used the potato manager's estimate of 0.8 kg (800 g or 1.76 lbs.) of raw potatoes needed to make one 200 g (0.44 lb. or 7 oz.) bag of potato chips. We used 322 million for the US population as of October 2015. http://www.census.gov/popclock/

Based on these estimated values, total annual chip consumption in the US is approximately 5.5 billion lbs. of potato chips. To make this amount of potato chips approximately 21.9 billion lbs. of raw potatoes are used.

Are Chapagain's and Hoekstra's Values Accurate?

We pointed out in our response to the reader that Chapagain and Hoekstra had indicated several sources of error in their report. First the authors stated that their values accounted for water usage to produce only the raw commodity itself but was a global average (meaning values could be significantly more or less in different areas depending on many factors such as climate variability and agricultural practices). See Summary (4th paragraph) and p. 41 Table 4.1 of Report 16 for examples of global averages and water footprints by country.

Second, their values excluded water needed for processing as stated on page 38 of the document cited above:

The volume of process water requirement depends upon the type of product processed and the technology involved. For any specific product the processing water requirement is more or less the same across different countries. There are minor variations, based on the efficiency of water use depending on recycling percentage, cooling processes, etc. As the processing water is only a small part of the virtual water content of a crop..., it will not affect the end results of the study if we assume one constant value for a specific product across the globe...For crop products we have assumed that the processing water requirement is relatively small compared to the virtual water content of the primary crop and we have neglected these values in the subsequent calculation of the virtual water content of the processed crop products.

Third, Chapagain and Hoekstra also made many simplifying assumptions when calculating the water footprint of processed products (i.e., potato chips) derived from primary raw commodities (i.e., potatoes) On p. 26 of Report 16:

The virtual water content of a processed product depends on the virtual water content of the primary crop... from which it is derived. The virtual water content of the primary crop... is distributed over the different products from that specific crop...We have assumed that each individual crop...product p comes from one and only one particular type of primary crop c ... For simplification it is further assumed that a product p exported from a certain country e is actually produced from a primary crop c grown within that country using the domestic resources only.

This last assumption implies that Chapagain's and Hoekstra's potato chip water footprint value does not include the water footprint of the vegetable oil that comprises 30-35% of the chips. The vegetable oil could have been either produced domestically or imported with different water footprints in each case.

Fourth, in Report 16 published in 2004, Chapagain and Hoekstra did not discuss the water footprints of packaging materials and their processing as well as water footprints to run production lines in factories.

In more recent work Hoekstra and researchers have calculated negligible contributions from transportation of materials and energy used during production to total water usage per food product (Tables 2 and 3 on pages 728-29 with explanation given in Sections 3.2.1 and 3.2.2): http://waterfootprint.org/media/downloads/Ercin-et-al-2011-CorporateWaterFootprint-Softdrink_1.pdf

In a similar study, the authors determined that the fabrication of the materials used to package a soy product involved some water but significantly less than that needed to produce the major food ingredients. Processing of the packaging material added only insignificantly to the overall water usage to produce the final food product. See Table 2 on page 395: http://waterfootprint.org/media/downloads/Ercin-et-al-2012-WaterFootprintSoy_1.pdf

Chapagain and Hoekstra have expressed concerns over some of their assumptions and use of certain data sets. They state on pages 70-71 of Report 16:

An important shortcoming is that the estimates of virtual water content of crops are based on crop water requirements, which leads to overestimates in those cases where actual water availability is lower than the crop water requirement. The calculations could be improved by using the actual water use by crops as a basis, which however will require more specific data per crop per country (that we did not have for the current study).

A matter for future consideration is the issue of including or excluding irrigation losses from the water footprint definition. In the current study we have not included them mainly for the practical reason that data on irrigation losses are generally not specified per crop, so that they cannot be included in the calculation of the virtual water content of specific crops. But it can be argued that it is indeed right to exclude irrigation losses, because these losses largely return to the system again. The counter-arguments in favour of including the losses are that the withdrawal in itself has an impact already that should not be neglected (the return flows do not return to the precise place where they were withdrawn), that a fraction of the total loss really gets lost for further use through evaporation, and that return flows are often polluted and cannot be reused without treatment or dilution.

A second shortcoming in the current assessment of water footprints is that we have focused on expressing the impact of human activities on the quantitative use of water resources. The water footprint concept has been defined in this study as the quantity of water required to fulfill human’s demand for goods and services. Further development of the water footprint concept would expand the water footprint definition in order to include impacts of human activities on water quality as well.

Based on the information presented by the potato industry manager, another shortcoming of Chapagain's and Hoekstra's work could be expressed by modifying the first sentence quoted directly above from Report 16 (page 70) like this (added/changed words bolded below):

An important shortcoming is that the estimates of virtual water content of crops are based on crop water requirements, which leads to underestimates in those cases where actual water availability and water used or removed to make processed products is higher than the crop water requirement.

With respect to Chapagain's and Hoekstra's second shortcoming concerning the effect of human activity on water quality, it would be helpful to include: raw materials' water footprints as well as processed products' water footprints; and the environmental and human health effects caused by the production and use of synthetic fertilizers and pesticides in agriculture.

In line with observations made here concerning Chapagain's and Hoekstra's work, a September 2015 article published at Bakeryandsnacks.com reiterates the main points: http://www.bakeryandsnacks.com/Manufacturers/Chip-brand-sustainability-level-has-a-long-way-to-go-Rank-a-Brand/?utm_source=newsletter_daily&utm_medium=email&utm_campaign=16-Sep-2015&c=yazB/DHFv2UpQjNc84mPIg==&p2

See also: http://rankabrand.org/chips-crisps


Despite these and other limitations and qualifying assumptions to all water footprint calculations, Chapagain and Hoekstra have produced a vast body of knowledge about water footprints serving as a basis for future refinement by researchers just as Chapagain and Hoekstra also continue to develop ways to calculate water footprints more accurately. Comprehensive data and calculations that include all the water needed to produce particular food products are becoming more readily available. For example, many more recently published articles estimate the total amount of water used to produce many typical food items. Here are two: http://waterfootprint.org/media/downloads/Ruini-et-al-2013_1.pdf http://waterfootprint.org/media/downloads/Jefferies-et-al-2012_1.pdf

Interested readers may attempt to calculate such values of specific food products by using assessment manuals like this one: http://waterfootprint.org/media/downloads/TheWaterFootprintAssessmentManual_2.pdf

The Water Footprint Network website also includes an interactive tool for individuals to calculate their unique water footprint based on their lifestyle choices and consumption patterns: http://waterfootprint.org/en/resources/interactive-tools/personal-water-footprint-calculator/

Some readers may prefer to listen to Dr. Hoekstra speak about water footprints and virtual water: https://www.youtube.com/watch?v=9xCE0t73-nk

Post Script: Net-Zero Chips

Some potato chip companies are redesigning their production methods at pilot sites to achieve a “net-zero” snack chip: http://www.foodprocessing.com/articles/2012/pepsi-near-net-zero/ http://2012books.lardbucket.org/books/entrepreneurship-and-sustainability/s09-03-frito-lay-north-america-the-ma.html

Frito-Lay's® system is described in this document (p. 21): http://www3.epa.gov/chp/documents/wbnr031810_hedman.pdf

The VRG asked the potato industry manager if European companies are also piloting net-zero technologies and received this reply:

In some parts of Europe companies don't have customers with great environmental awareness and since competition is not a factor, no changes are in process that I know of...

Post Script: Ugly Vegetables

An innovative way for individual consumers to conserve water used for food production is to purchase the outgrades (i.e., cosmetically blemished fruits and vegetables that are discarded from production lines).

For more information: http://thinkprogress.org/climate/2015/08/19/3692594/ugly-fruit-and-vegetable-food-waste-campaign/ http://www.npr.org/sections/thesalt/2014/12/09/369613561/in-europe-ugly-sells-in-the-produce-aisle http://www.npr.org/sections/thesalt/2013/09/21/222082247/trader-joes-ex-president-to-turn-expired-food-into-cheap-meals

Interested consumers may find out more about ugly grocers near them: http://www.endfoodwaste.org/united-states.html