Water and Transportation Fuels

View my white paper on Sigma Xi regarding the water consumption required for various types of alternative fuels. See: http://www.sigmaxi.org//programs/issues/King.pdf. (also see below for full text of article if link).

In this white paper I refer to an article I wrote in Environmental Science and Technology on this same subject of the Water Intensity of Transportation. In this paper I calculate the "gallons of water per mile" that are embodied via consumption (meaning mostly evaporated) and withdrawal (taken from a water source and returned to the source) during farming, mining, processing, and refining of feedstocks into fuels.

Feedstocks studied are petroleum, coal, natural gas, sun, wind, biomass (corn and soy).
Fuels studied are ethanol, gasoline, diesel, biodiesel, electricity, hydrogen, and natural gas.


FULL SIGMA XI ARTICLE:
On the Water Consumption for Transportation Fuels
Carey W. King, Ph.D.
University of Texas at Austin
Center for International Energy and Environmental Policy

Driving light duty vehicles (LDV) on most alternative fuels and energy sources will
consume more water per mile driven than by driving on gasoline and diesel1. The
exceptions are using compressed natural gas with natural gas powered pumps, electricity
derived from non-thermal renewable electricity (e.g. solar PV and wind), and hydrogen
derived from either electrolysis of water using non-thermal renewable electricity or steam
methane reforming. To effectively plan for the environmental consequences of moving
from high energy density petroleum to lower quality fossil fuels and biomass, we must
not unduly distribute fuels with low energy return on investment. Water consumption is
just one environmental attribute for focus, but an important one from a quantity and
quality perspective.

In 2003 the average fuel efficiency of the U.S. LDV fleet was 20.5 mpg of gasoline.
These gasoline vehicles consume, via embodied water in mining and refining, 0.1-0.2
gallons of water per mile (gal H2O/mile). Using tar sands, coal, and oil shale converted to
liquids consumes 0.3-0.5 gal H2O/mile. If using electricity from the average U.S.
generation mix, driving a car using an electric motor from a battery consumes 0.2-0.3 gal
H2O/mile. If the grid electricity is used for electrolysis of water to create hydrogen, using
that hydrogen in a fuel cell vehicle results in consumption of 0.4-0.5 gal H2O/mile. Using
non-thermal renewable electricity for electric and fuel cell vehicles consumes less than
0.05 gal H2O/mile, and obtaining the hydrogen from steam methane reforming of natural
gas consumes just under 0.1 gal H2O/mile.

The other major category of potential LDV fuels is biofuels. If the biomass feedstock is
irrigated, using so-called “blue water” from aquifers and reservoirs, the water
consumption for corn-based ethanol (E85) and soy-based biodiesel is orders of magnitude
higher than other fuels with U.S. averages of 28 and 8 gal H2O/mile, respectively. Note
that only 10% of soy and 15-20% of corn bushels are irrigated in the U.S. However, some
highly irritating regions of the U.S. could embody over 100 gal H2O/mile by growing
irrigated corn converted to E85. Cellulosic-based irrigated grasses could result in 1-9 gal
H2O/mile of consumption. Without irrigation, fueling today’s fleet of LDVs using E85
ethanol and soy biodiesel would consume 0.1-0.4 gal H2O/mile, comparable to
unconventional fossil conversions.

For almost any product or application, but certainly for agriculture, the embodied water is
essentially transferred to drier regions. For instance, in providing food aid to developing
nations, the U.S. essentially becomes a net exporter of water – that is water embodied in
the exported food. If a given region does not have the climate, or irrigation water source,
1 King, C. W. and Webber M. E. Water Intensity of Transportation. Env. Sci. & Tech. Online
9/24/08 at: http://pubs.acs.org/cgi-bin/abstract.cgi/esthag/asap/abs/es800367m.html.
for growing a certain crop, importing that crop may be a good option. However, should
the irrigation water be drawn from a fossil aquifer, the same crop grown under a different
circumstance will not be sustainable in the long term and should be considered a fossil
fuel itself simply based upon its supply chain, or life cycle. The irrigation of biomass
represents the growing of that biomass in areas with insufficient rainfall. This
relationship itself does not necessarily imply a lack of sustainability. For instance,
diverting nearby river water into a reservoir can create a stable supply source that can
withstand the normal climactic variations in rainfall.

The embodied water concept for agriculture can also be applied to biofuels but with
additional focus upon thermodynamics and energy return on investment (EROI). Many
regions may import some sugar cane-based ethanol from Brazil, yet don’t have the
climate to grow the sugar cane. Because of the relatively high EROI of cane ethanol over
corn ethanol, international shipping can make sense. Thus, the history, or supply chain, of
the biofuel is important, not just its final properties. The supply chain of fossil fuels will
also become more important over time as lesser quality resources are extracted. There
was no need to pay attention to EROI from the early coal beds and oil wells because they
so clearly allowed increased lifestyle and leisure relative to the world before the
industrial revolution.

In the concept of moving to non-petroleum fuel sources, the Renewables Fuels Standard
(RFS) of the Energy Independence and Security Act (EISA) of 2007 has been both good
and bad from a policy perspective:

Bad in the sense that the RFS initially pushes a feedstock-fuel combination, corn
ethanol, that has detrimental environmental consequence in terms of nutrient runoff
and low EROI. But, the consequences of runoff can be minimized by not overfertilizing,
more widespread use of better tilling practices, and buffer strips next to
major rivers.

Good in the sense that the RFS has pushed forward the scrutiny of how we use energy
resources, biomass included, and focused much attention on how we can create better
biomass to fuel conversions. It has also raised worldwide awareness on the ethics of
how agricultural land should be used: food, fuel, or both?

Today, the struggle for new fuel supplies is clouded because it is not obvious what
options best provide for increased or even continued levels of lifestyle and leisure. What
is more obvious is that there are geographic regions that can sustainably grow certain
kinds of biomass that other regions simply cannot. Unconventional fossil resources such
as tar sands and oil shale have lower EROI and higher water consumption needs than
conventional petroleum. The need for more water in industrial fuel systems is an
indicator of moving to lower energy efficient sources.

Lawmakers creating future policy regarding agriculture and energy need to be cognizant
that the tie between energy and water will only increase into the future. Generally, energy
sources with lower energy density tend to require more water for mining, farming,
refining, and processing. Certainly an increase in vehicle fuel efficiency decreases the
“gallons of water per mile” traveled, but in then end, fresh water sustainability is
measured on only one “per” basis: “gallons of water per Earth”.

Ethics: Allocation factors for renewable energy systems

I recently wrote for Worldchanging about how allocation factors can possibly be representative of our ethics. Allocation factors are the fraction of the energy input into a renewable system, usually for analyzing biofuels, that is associated (or allocated) to each of the products.

For example, the main product from corn ethanol is the ethanol, and coproducts are distillers grains for cattle feed. Because these allocation factors can be based upon the energy, mass, or economic content of the coproducts, different analyses of the same process results in different outcomes in terms of the sustainability or renewability of the process. One of the major issues is that what is economically most attractive is often not the most energetically efficient. A possible policy goal could be to guide these two concepts together.

Click allocations factors and ethics to go to Worldchanging website for the article, or read text below:

**********************

Moving toward a sustainable, or renewable energy-based economy, stresses the views of how people value their time and exertion. Our system of economics puts value on products and services that allow people to spend less time and/or exertion while performing a task. This value system is exactly why fossil fuels have been the driving factor for increases in accumulation of material goods and leisure time over the course of the industrial revolution.

Historically, fossil fuels have had such high energy density (and energy return on that invested to mine them) that we haven't worried too much about how to allocate the energy invested. When a barrel of oil is refined or a cubic foot of natural gas is burned, it has been obvious that we can produce more products and spend more time in leisure or progressive work. It is because of the concerns of fossil resource scarcity together with environmental effects (air pollution, greenhouse gases and climate change, etc.) that alternatives to fossil resources are sought.

By contrast, when it comes to renewable energy products, particularly biofuels, we've applied intense scrutiny to figuring out the energy return on total energy (or fossil energy) invested because the returns are not as easily determined as being sufficiently greater than one. Part of this scrutiny is because of the inherently lower energy density of carbohydrates (i.e. biomass) versus hydrocarbons (i.e. fossil fuels). Another part of the scrutiny derives from the knowledge that fossil fuels currently permeate the vast majority of the manufacturing and agricultural practices of the industrialized world, and understanding the optimal manner in which to deal with their reduced presence, and possible absence, is not obvious.

Allocation factors are an example of the struggle of society to understand the value of output of renewable energy systems and processes. The allocation factor is a term used to describe how much of the total energy input to a renewable energy system should be “allocated” to, or associated with, both the primary product output (e.g. ethanol, biodiesel, biocrude, etc.) as well as any process coproducts. These allocation factors are also used to assign greenhouse gas quantities to compare competing energy systems. Renewable energy systems that output electricity, such as photovoltaic solar panels and wind turbines, are fairly straightforward in giving an allocation factor of one. That is to say, all of the energy and material inputs that go into manufacturing, operation, and maintenance of the system are used to produce the only output: energy in the form of electricity. There is no product other than the electricity.

Assigning an allocation factor for biofuel production is more difficult. Biofuels originate from some form of biomass (e.g. corn, soybeans, cellulose, etc.) that can be used for multiple purposes (e.g. food and fuel) and the extracting them creates output products besides the fuel itself, termed coproducts. For example, in the typical processing of biodiesel from soybeans, the major outputs are the primary product of biodiesel plus the coproducts of soy meal and glycerin [1]. Fossil fuels have similar product/coproduct distinctions (e.g. natural gas for fertilizers and petroleum for plastics), but because we know there is no long term sustainable use of them, there has been no need to scrutinize how we derive their various products.

So a question arises: for every unit of energy input from field to fuel, how much of that input should be responsible for each product? To answer this question, there are multiple proposed allocation concepts. The different allocation methods for coproducts are three non-energy methods and three energy-based methods [2] that are designated by whether the energy consumption of the processes is allocated according to:

Non-Energy Methods
• the 100% principle such that all energy consumed is allocated to the primary product (e.g. biofuel).
• the mass fraction of each of the products,
• the economic market value of each of the products,

Energy-based Methods
• the energy content (calorific value) of each of the products,
• the energy displaced by each of the products with respect to an existing or customary way of producing the product, or

100% Principle
Allocating 100% of energy inputs to renewable energy systems is the most simplistic and uninformative. There are no decisions to be made, and it removes the capacity for society to learn how to use all available resources and technologies while reusing and recycling as much as possible. On the other hand, its simplicity easily allows policymakers and consumers to understand the impacts and benefits of renewable systems. Essentially, the 100% principle is the extreme case that assumes no useful coproducts are possible, or that coproducts are free in terms of monetary or energy input.

Mass Fraction
Allocating by mass fraction is very straightforward and easy to understand. Techniques that minimize coproducts should be viewed as positive since otherwise, they would not be coproducts but instead the primary product. We can likely assume the primary product is the most market viable, at least at the time the renewable energy project is begun.

Market Value
Using market value to allocate coproducts is the method most akin to the free market principles. Brazil’s past and continued focus upon sugar cane as a cash crop theoretically enables their companies to decide how much sugar versus ethanol to produce from the same crop. If one price is up, they can focus on that product versus the other. Currently, the ethanol price is up as a group of Brazilian companies has arranged the first “practical application of verified sustainable ethanol” trade with Sweden [3]. Thus, a market value of coproducts potentially allows a producer to tune his process according to the rather short time scales of commodity fluctuations. The main drawback of this method is that market prices change, and what could be a good energy balance one day could be a poor one a week later [1].

Energy Content (calorific value) of Products
Focusing upon the energy content of the products seems like a fundamental method because the purpose of renewable energy systems is to produce a product with high energy content. The primary product should in fact contain more energy than the coproducts, otherwise the system may have to be reanalyzed in terms of thermodynamic efficiency. This suboptimal energy content ratio could possibly occur if there is pressure to tailor a biofuel to existing infrastructure (which would be a pressure from the market). The difficulty with this method is that it does not indicate the effort required to achieve the energy intensive fuel or product. For instance, lasers contain high power concentrated in a tight beam, but much power is required to get the energy in that form.

Process Energy Input
Allocation due to the energy input into the renewable system seems like a logical choice because we are, after all, trying to figure out how to allocate the energy consumed in the renewable energy process. However, this allocation method can be somewhat confusing when the primary product and one or more coproducts results from the same subprocess. For example, if there is an unavoidable coproduct that results from the feedstock processing steps, how much input energy went into that unavoidable byproduct? What if the coproduct has no use, meaning it is actually a waste? Nonetheless, this method can often be more straightforward as in the case with wet-milling corn ethanol since during pre-treatment the starch (used for ethanol) is separated from the grain (used for coproducts), and thus subsequent energy used for processing the grain can easily allocated to the coproducts.

Energy Displaced (energy for replacement coproduct)
Allocation due to the energy displaced is an inherently comparative methodology. It requires diligent astute knowledge of the field of the product in order to know the energy input into replacement products. Also, an equivalent replacement product must exist. Here, the energy required for producing the primary product is reduced by the amount of energy required for the replacement product. For instance, Shapouri assumes that animal feed products (e.g. DDG) produced from corn ethanol processing can directly replace soybean meal. A difficulty arises if soy meal, itself a possible coproduct from biodiesel production, might use corn-based animal feeds as a replacement product as well. They can’t both replace each other. So there can be multiple choices of replacement products that can provide a range of answers for the primary product.

Can these allocation factors reveal something about culture, society, and how we value our energy and time? Is there a correct or more ethical method?

Pradhan et al. suggest that the correct method depends upon the question being asked. If renewability is the question, they say the mass fraction should be used, but if economic sustainability is to be determined, then the market value allocation approach should be used [1]. For philosophers who like to find the ultimate truth, this solution is rather non-satisfactory, and it avoids the question of whether the market should recognize that energy return on energy invested (EROI) is the major driver for economic growth or if economic growth potential is the driver for the choice of energy resources. The tail can’t wag the dog, but hopefully with enough flow of accurate information the EROI and economic return will continuously feedback to each other and arrive at the same solution.

[1] Pradhan, A.; Shrestha, D. S.; Van Gerpen, J.; and Duffield, J. 2008. The Energy Balance of Soybean Oil Biodiesel Production: A Review of Past Studies. Transactions of the American Society of Agricultural and Biological Engineers. 51 (1): 185-194.

[2] Larson, E. A review of life-cycle analysis studies on liquid biofuel systems for the transport sector. Energy for Sustainable Development. June 2006, Vol. X, No. 2: 109-126

[3] Guardian, UK. June 25, 2008. Brazil signs deal to export sustainable ethanol. http://www.guardian.co.uk/business/feedarticle/7609299.

Hold back the flow ... of false claims on water for transportation

This past week my colleague Michael Webber presented some preliminary results (currently under review for publication at Environmental Science and Technology) at the 2008 annual meeting of the American Association for the Advancement of Science. A journalist from the Toronto Star reported about the results of our work. And Gordon Quaiattini, the president of the Canadian Renewable Fuels Association (CRFA), put in his 2 Canadian cents worth of comment to the Star editor.

First, our work on this is under review so I won't comment too much on the methodology until it is accepted and published, but I can clarify some aspects of the table in the Toronto Star article as well as the comment by Mr. Quaiattini.

As far as Mr. Quaiattini is concerned, let me assure him that neither me nor Michael are against biofuels. What we are for is understanding the impacts of all fuels. That is why we presented information that compares a variety of fuels, and future work can focus on additional biofuels and alternative fuels.

Mr. Quaiattini claims that 85% of U.S. corn is non-irrigated. This is fairly consistent value as in 1998 we show approximately 1.9 billion bushels irrigated (see http://www.nass.usda.gov/census/census97/fris/fris.htm Table 22) out of about 9.8 billion bushels of US corn grain (see http://www.nass.usda.gov/ and select 'US corn grain' stats for 1998 in the pull down menu) - this gives 15.6% irrigated. He also states the numbers of 3 gallons of water to process the corn into a gallon of ethanol, and this is at the lower end of the range of values we used.

The data presented in the aside in the Star article lists ethanol water 'use' (note in this case consumption and withdrawal are roughly equivalent) as 40-130 gallons per mile driven on E85. This is close, but not quite accurate as noted. We calculate 12-136 gallons per mile driven on E85 derived from irrigated corn in the U.S. The range exists because not all regions that grow corn need the same amount of irrigation. Obviously some regions get more rain than others. We have made no claim (yet!) on the total water consumed and withdrawn for travel in light duty vehicles in the US.

NOTE: when considering ethanol derived from non-irrigated corn, the values for consumption and withdrawal are less than 0.5 gallons/mile. This shows you that the vast majority of the water of concern is for irrigation.

IMPORTANT:
Does this mean we should not use biofuels? ABSOLUTELY NOT!!!

What it does mean is that we need to understand the limits of our water resources while considering the tradeoffs that that the "biofuels vs. fossil fuels" debate entails. Fossil fuels are essentially really old biomass as nature has done a lot of work for us in growing the plants and storing them in the ground (over 100s of millions of years) for us to now use. Biofuels are essentially really young fossil fuels.

When planning for growing crops either for food or fuel, we need to use both the land and water resources responsibly. I applaud the efforts of the Canadian Renewable Fuels Association and other similar organizations that are helping promote alternatives to fossil fuels for transportation or stationary applications. I believe we can avoid a water conundrum, and our work is providing information to help society do just that.