I have been simmering a couple of weeks over the headline “Wine’s Mammoth Water Footprint: 120 Liters To Make One Glass?” (Lewis Perdue’s Wine Industry Insight March 6th, 2009). In the context of the current ongoing drought here in California, I suggest that characterization of water use by the wine industy as “mammoth” is well, um… unhelpful. The engaged reader should ask: is the characterization even accurate? “Mammoth” by what standard? Compared to what? Or was this just a sensationalist headline — a bit of journalistic bombast?

Valid question I think, since the article cited in the WII piece (“Thirsty Work,” Economist.com Feb. 25th, 2009) used no such value-laden qualifiers, only going so far as to state “Consumers… might be surprised to learn…” Enquiring minds want to know.

I give Mr. Perdue credit that in his article, under the sub-head “Site Short On Substantiation” he asks “Do any readers have an idea if this is accurate?” and gets some answers.

From my own experience: the Permit Resources Management Divisions in both Napa and Sonoma Counties use a figure of 16 gal water per case as a baseline winery production water footprint. A number of years ago a friend who was water manager for Fetzer figured through simple conservation measures they had reduced this to under 8 gallons/case. A recent article in Wine Business Monthly (“Water Use In The Winery” by Paul Franson, Dec. 15th, 2008) cited that Fetzer has further reduced their water footprint to 5 gal/case, while more agressively conservative producers are using as little as 3.6 gal/case. SO, for wine production the range of water use is something like 0.2-2.0 liters per glass. Where do the other 118 liters per glass come from?

Um… growing the grapes. Again, from my own experience at our Estate vineyard: since the vines went into production in 2005 we have applied no water for frost control and approximately 60,000 gallons of irrigation per acre annually. Our average yield has been just under 3 tons/acre, and very conservatively our press yield is something like 165 gallons of new wine per ton of grapes crushed. So our water use during the growing season is about 120 gallons per gallon of new wine, or 15 liters per glass. Hmmm… at Westwood we are using something less than 17 liters of water to make a glass of wine, not 120 liters. Where’s the discrepancy?

This 120 liter figure cited in the WII and Economist articles comes from data compiled by MacArthur “genius grant” recipient Dr. Peter H. Gleick and his staff at the Pacific Institute, their work also diseminated through the Water Footprint Network. Digging into some of the online information Gleick and co-workers have compiled, I found the following general statement:

These kinds of data are fraught with problems and uncertainties, and users should be extremely careful about using them for other than the most simple comparisons. When we can, we like to use ranges to try to bracket many of the uncertainties, but other sources rarely mention uncertainties or provide ranges of estimates. For example, the Water Footprint reports that 15,500 kg of water are required to produce beef, but work from the Pacific Institute reports a range of 15,000 to over 70,000 depending on diet, climate, the amount of product from each cow, and other variables. Similarly, the Water Footprint reports single estimates for the production of a range of vegetable and feed crops, but actual water requirements will vary dramatically with climate, soils, irrigation methods, and crop genetics.

Equally, if not more complicated, is evaluating the water required to produce manufactured items. For example, the water required to produce a liter of a soft drink may be as low as 2 to 4 liters per liter of product. But vast quantities of water are also consumed to produce the feedstocks, such as sugar or corn syrup, used in the same product. There are no consistent rules for where to draw the “supply chain” boundaries in such estimates, making it critical that users understand the assumptions that go into these values. This table, for example, lists 125 liters of water to make a kilogram of sheet paper, but it seems likely that this is the value for producing paper alone, and excludes the water required to grow the tree itself. Similarly, fewer than ten liters of water are required to process milk, but as many as 1,000 liters may be required if the water to produce the cow itself is included [all emphasis mine - JMK].

Wow — now that’s a caveat, one entirely missing from the downstream discussions. And one that can more than cover the range of 17 to 120 liters of water per glass of wine.

From another article (“Water Use In The Vineyard: The West” in the same issue of WBM cited above) I find that our rate of water use at the Estate is at the low end of the reported range – consistent with the cool climate and abundant ground water present at our site. From the high end of the data presented by Mark Greenspan in the article I calculate vineyard water usage of 110 liters per glass — within striking distance of Gleick’s 120 liters. OK.

But it is important to look at all the data Dr. Gleick’s group presents. Grapes are among the least thirsty crops, comparable to other tree fruits but way less thirsty than water hogs such as cotton, rice and corn. And all these crops use way, way less water than livestock.

None of this even touches what I consider to be the more important factors in the water use discussion: the environmental costs. Runoff, pollution, erosion, costs of wastewater treatment, deferred costs of environmental degradation and mitigation. By all of these measures, grape growing — as most of us practice it here in the North Coast — has an extraordinarily small footprint.