Water and energy - a brief presentation prepared by Vince Gutschick  for the






(Also to be posted on http://gcconsortium.com/academic_page/Water-energy.html ; for more discussion of global change issues, please see the Website of Vince's consulting company at  http://gcconsortium.com , and, for water issues specifically, http://gcconsortium.com/issues_GC_and_water2.html and http://gcconsortium.com/Water_wedges2B.html )


Water use for producing energy (demand side)


   Fossil fuel extraction (modest)

       Volume of water used may be several times the volume of oil produced, with water injection for oil recovery (one may calculate water use for this compared with total human water demand in all activities; it is a small fraction). 

          Let's normalize this to the energy yield of this oil.  Oil has an energy content of about 40 megajoules per kg.  In perhaps more familiar units of kilowatt hours (kWh), this is 11 kWh per kg.  If the oil is used for electric power generation (a lesser use) at 35% efficiency, it is worth about 3.9 kWh per kg. So, the water use is then about 3/3.9 or nearly 0.8 kg per kWh.

          Let's take it to the personal level.  The average American uses 43 kWh per day.  This would require 43*0.8 or about 34 kg = 0.034 tonnes of water per day per person for oil extraction.

   Petroleum refining (modest). From article in wikipedia: "A typical large refinery processing 40,000 metric tonnes of crude oil per day (300,000 barrels per day) circulates about 80,000 cubic metres of water per hour through its cooling tower system"  à 1.9 million tons of water, about 48 times the volume of oil; about 5% (2.4x the volume of oil) is consumed (evaporated). 

          By the same kinds of calculations above, this water use is 80% as much as in the extraction.

          For our average electricity use, this is about 0.027 tonnes of water per day per person. 

   Compare both of these to our M&I use of about 800 L = 0.8 tonnes per person per day, about 13 times greater than the water use in oil extraction and refining.  Of course, not all of this water use is consumptive; relatively little of the waste water evaporates; most returns to rivers or groundwater.  However, pumping of water and wastewater treatment uses significant amounts of energy - about 3% of our national energy use (US EPA).  Thus, the indirect use of water at power plants to supply our pure water is significant.  Water is consumed to supply water.  I haven't worked out the fraction so used.

   Compare the extraction and refining use to our liquid fuel use directly of about 11 L (0.009 tonnes) per person per day, which required about 0.43 tonnes of water to process

   Compare these uses also to irrigation water use for water used to grow our crops.  Directly and indirectly (in meat animal production), we require the growth of about 6 kg dry mass of crop plants per person per day (vs. 0.55 kg direct consumption of dry food mass).  This allows for incomplete harvest index, spoilage, and waste (substantial!).  Crop growth requires about 500 to 2000 kg of water per kg of dry plant matter.  If we use an average of 1000 kg water per kg of plant matter, the plant growth to supply one of us for a day uses about 6000 kg (6 tonnes) of water.  Given that 1/3 of the transpired water is irrigation water, we require about 2 tonnes of irrigation water per person per day for our food supply.

 à fossil energy use requires a significant use of water, but still small compared to water used for our food supply.  However, we keep expanding our energy demand, much more than our food consumption per person (hard to believe the latter if you've been to Wal-Mart)

   Fossil fuel use in power plants, the final water use in this energy production chain

      In using streams or oceans for heat discharge: modest increase in evaporation from hotter water; more concern is thermal pollution affecting aquatic life

      Wet cooling towers - moderate

          A 700 MW coal-fired power plant uses (evaporates) about 1260 tonnes (cubic meters) of water per hour or about 30,000 tonnes per day, while producing 16.8 million kWh per day

          That converts to about 1.8 kg or 0.0018 tonnes of water used per kWh

          An average US resident consumes about 43 kWh per day, which needed about 0.087 tonnes of cooling water - 1/5 of what we flush in toilets, sinks, showers….

      (Also used for nuclear plants)

         (Dry towers à lower cooling efficacy à lower thermal efficiency in energy production)

      Tar sands extraction - huge; even more of a pollution problem

   Capital use of water - constructing facilities

      E.g., water use in steel production, e.g.

         Modest, for wind turbines.  These use a lot more steel than do fossil-fueled power plants, per unit of energy delivered over the facility lifetime.

            A 385 MW coal-fired plant needs about 11,500 tonnes of steel and provides (at 92% capacity factor) 124 billion kWh à steel use is about 93 milligrams of steel per kWh! 

            A 1.5 MW (peak) wind turbine uses about 210 tonnes of steel.  Over its 25-year lifetime, at a typical capacity factor of 30%, this turbine provides about 98 million kWh à steel use is about 2100 mg steel per kWh, or 22x more than the coal-fired plant.

           Steel fabrication requires consumption of up to 25,000 gal (95 tonnes) of water per tonne of steel.

           So, steel for a coal-fired power plant requires about 8.8 millionths of a tonne of water per kwH produced - essentially irrelevant.  For a wind turbine, it is about 0.00021 tonnes per kWh, or about 1/10 the consumption of water for any of the 3 stages in fossil fuel use (extraction, refining, combustion in a power plant).

      Semiconductor fabrication for solar photovoltaic (PV) cells uses significant amounts of water.  The consumptive (non-recycled) use is largely indirect - the water use in generating the copious amounts of electricity for silicon refining.  About 1/10 as much energy is used in making PV panels as they will deliver in their 20-year lifetimes.

         Also a groundwater contamination issue (e.g., Intel in Abq.)

   Biomass growth for biofuels (liquid or used in power plants (enormous) - for transpiration of plants

       Transpiration ratio of vascular plants (corn, switchgrass…) is ca. 500-2000 kg water per kg dry biomass, which contains about 20 MJ/kg of energy; at 50% conversion rate (we're far from that with corn ethanol, might get to the 30% or so with cellulosic ethanol)  (See interesting side note below.)

           So, suppose we used renewables for our fuel and our electric power production (i.e., all our energy needs, of about 312 Gigajoules per person per year).   We'd need to grow plants with an energy content at least 2x greater, or 624 GJ.  This is about 31 tonnes of plants per person annually, requiring about 31,000 tonnes of water.  If irrigation supplies 1/3 of that water, it's about 10,000 tonnes per year à over 30 tonnes per day, per person.  This dwarfs all other water use.  Biofuels also have a huge impact on land use (we'd have to stop growing almost all food crops, or degrade habitat and marginal crop land completely).

       In addition to direct use of managed supplies for irrigation, there is also alteration of rainfall recharge (more rainwater intercepted by energy crops, less going to recharge aquifers)

       Somewhat less impact on water, but still a big impact on land use: algal biofuels

           Evaporation from open ponds (can't cover them; capital energy use in making plastic covers would be about 15 years of energy production, by my estimates)

               Somewhat less than transpiration by higher plants

               May use brackish water, in part


Converse: impacts of energy use on surface water renewal (supply side)


   Linkages: fossil energy use à greenhouse gas emissions à climate warming

      Subsequent effects

         Altered precipitation amount: globally up by about 6% per deg. C rise, but unevenly distributed (moisture convergence occurs) à greater droughts and greater floods (and floods can't be captured, as well as causing material damage)

         Altered snowpack and glacier dynamics

            Less average snowpack predicted for NM supplies (Colorado Rockies)

            Earlier snowmelt à timing to irrigation needs might be offset - though if crops can also be planted earlier, is this a "wash"? No: longer growing season with only slightly more precip. globally à more evaporation, less use by crops in transpiration

              à China has gained 2 weeks of wheat growing season in last 20 y and 8% in yield, but this is reversing now from the above effects

              à Big crop losses are predicted in the tropics à global social unrest

         Will crops need to transpire more water per amount of growth? 

            Higher air and plant temperature à more transpiration

            Higher CO2 à stomatal closure à less transpiration

            Overall: probably slightly less water use per unit of plant growth

         Sea level rise (about 2 mm/y currently) à saltwater intrusion into some coastal water supplies.   Might accelerate if polar ice caps melt/dislodge (mixed results so far)

   Bigger impacts on water availability: human population growth, plus rising standards of living in China, India, Brazil, … à more food production, and more than proportional to population (meat production: need feed crops in amount 3 to 7 times the meat mass)


Key ideas:


   We have been using (profligately) an energy technology (fossil fuels) that had quite low impacts on water and land use, among other things, though a big impact on climate.

   We must go low-carbon in energy production.  However, we need be aware that the renewable low-carbon technologies have big impacts on land and water use (and thus, also, on other species).  Nuclear energy is most benign here, while bringing in other issues (arms proliferation, waste management) that are beyond the scope here.  It will have to be a big part of our energy future pragmatically, and its reinviograted development is already being pursued in many nations.

   Biomass energy is a stopgap only (terrible water-use efficiency, poor energy efficiency, and most schemes increase greenhouse gas emissions from land clearance, direct or indirect), until we get low-carbon energy technologies fully in place (wind, solar thermal, solar PV, nuclear)

   No free lunch - all energy and water-use technologies have "footprint" / downsides / unintended (but foreseeable) consequences

       à We need to pick the least adverse mix of technologies

       à We need to limit population growth in a rational and humane manner…and reduce our consumption of meat (in the US, esp.; give China, India, others a bit of room to increase theirs…not that we can really stop them)



Side note on the transpiration ratio of plants: one reason it is so high is that CO2 is now so dilute in the air…and it is 5 to 10 times more dilute than it used to be about 400 million years ago because plants were so successful in:

   1) doing photosynthesis - making carbon compounds for their own bodies, and

   2) making products, esp. lignin, that were so hard for pests, disease organisms, and soil recycling organisms to break down that a lot of the carbon they captured got buried, cumulatively….as a good part of our coal, our major fossil fuel reserve.