I attended the First Annual MIT Water Summit last week. Since water, unlike hydrocarbons, is essentially neither created nor destroyed, lots of consumption and supply-based metrics can be quite confusing.
First, the usual trivia of interest to water watchers:
The production of 1 cup of Starbucks coffee is tied to consumption of 200 liters of water.
It’s hard to tell whether bottled water (2000x as energy intensive as tap water, but not tested/certified) is safer than tap water (increasing, though small, levels of a variety of pharmaceuticals, including cannabis!).
Water for cooling of power plants accounts for half of all gross consumption – but only 5 % of net consumption after return of the cooling water to source (hopefully after being chilled to near initial levels).
Agriculture consumes over a third of all US water, mostly through evaporation, which eventually becomes water, but somewhere else on earth? And a third of all US agriculture harvested is wasted/lost, either during transport or at the consumer. That’s over 10% of all US consumption!!!
While searching for a level-playing-field analysis of the energy-water nexus, I found a report authored by (among others) my good friend Erik Mielke, while a Research Fellow at the Kennedy School. The publication makes serious effort to present an objective compendium of ‘gallons/mmbtu’ and ‘gallons/kwh,’ of output from a variety for energy sources. It’s not easy, and there is a fair amount of variability, even within the same process, but the report is a good start toward a fair examination of this issue. While primary oil and gas production require little additional water (under 10 gallons/mmbtu), secondary recovery processes, and extractions from shale can require up to 100 gallons/million BTUs. Surprisingly, however, in situ oil sands production (including syncrude), shale oil/gas rank at the low end of that range, as are coal and uranium mining – although shale-related data was obtained at the early stages of commercialization, with more water intensive activities in recent years.
Interstingly, enhanced recovery byCO2 injection – a form of carbon sequestration – is highly water intensive at over 90 gallons/mmbtu (roughly 15 gallons/barrel recovered). And ethanol production from irrigated corn (a quarter of all US corn) is off the charts at 3000 gallons/mmbtu – 20-50x the non-irrigated route. The report also evaluates the energy cost/benefit of various cooling designs for downstream power production.
In short, nearly all energy is water intensive – some processes surprisingly less so than others. And nearly all water is energy intensive. More commentary available on request!