[SGVLUG] Should we use biofuels? Was ...Tom & his Prius....

[David Lawyer]

I want to find out more on this topic, but haven't done much yet.
When I do, I'll let this group know the results.  But I thought that
at least I should write you about what I know about it now as it's an
important topic.

The posts that I've looked at indicate that some people may have been
seduced by the biofuel scam.  While I could be wrong, from what
evidence I've looked at so far, it seems to be a scam.  An example of
a biofuel is ethanol or biodiesel, made from corn, soybeans, and/or
switchgrass.  David Pimentel, a professor at Cornell, has analyzed the
energy costs of producing ethanol and biodiesel and claims that it
takes more energy to produce than is returned when you burn it as a
fuel.  This like trying to recover oil from a depleted oil field where
it would take more energy to extract the oil than you would get from
the oil.  Not a very economic or environmentally sound prospect.  I
think Pimentel is right and that he may have even undercounted the
energy used to produce biofuels.  

See:
http://www.news.cornell.edu/stories/July05/ethanol.toocostly.ssl.html"

I'll talk about ethanol, but I think that the situation is similar for
biodiesel.  If one checks the Internet, one finds that most reports
erroneously show an energy gain from using corn, etc. to produce
ethanol.  This is mainly because the government is subsidizing such
production to the tune of a few billions of dollars each year and
there is a lot of money to be made from erroneous science.  Another
reason is that to get the "correct" figures for the fossil fuel energy
needed to produce ethanol is very difficult, if not impossible as I'll
explain later.

Pimentel's most recent report is named "Ethanol Production Using Corn,
Switchgrass, and Wood.  A comparison of Pimentel's calculations with
the US Dept. of Agriculture may be found at:

"http://72.14.203.104/search?q=cache:NL4jjalUAOkJ:www.mathproinc.com/pdf/2.1.6_Ethanol_NEV_Comparison.pdf+%22David+Pimentel%22+%22ethanol+production+using+corn%22&hl=en&gl=us&ct=clnk&cd=10&ie=UTF-8"

Upon looking at this a little I'm convinced the Pimentel is more
correct than the US Dept. of Agriculture which claims an energy gain.
I need to also look at the original report which I'll do later at a
library.

Strange as it may seem, I think that the fossil energy use may be even
higher than Pimentel estimates.  He only covers about 14 sources of
inputs to make ethanol.  I don't think he uses input-output matrices
which should result in higher values for fossil fuel energy.  For
example, one needs to count the energy used to supply the
infrastructure to farming such as putting in roads, utilities, etc.

EMBEDDED ENERGY and the LEONTIEF MODEL:

All this has to do with the embedded (or embodied) energy in
commodities.  To find this one needs an input-output matrix for a
closed economy (the world?).   It probably doesn't exist but when the
US was more self sufficient, we had such data for the US and still
have this type of data.  But I don't know how useful they are today
with all the manufacturing of the vast volume of imports to the US
being done in foreign countries.  When I started in to explain it
below, I thought it would be short; just one simple equation and its
inverse.  But it didn't turn out this way.

Here's my explanation of the Leontief model and how it's used to
determine embodied energy content.  Let x be the vector of commodities
(say produced in one year) x1 might be kwh of electricity x2 might be
tons of coal, x3 might be the number of automobiles, x4 the amount of
rugs as valued in dollars, etc.  There will be thousands of such
commodities and the units of measure will often be different (tons,
number of items, dollars).  There is also a lot of double counting.
For example electricity may be generated from coal, automobiles made
from steel, etc.  But in spite of all this the model, works.

In order to produce each commodity, other commodities are required and
consumed in the production process.  This includes depreciation of
plant and machinery.  So let the dimension of the commodity vector x
be n.  From knowledge of the production technology, form the n x n
matrix A such that Ax is the vector of commodities consumed in order
to produce  x amount of commodities.  Ax is a vector and each element of
it is for the same commodity as the corresponding element of x.  So
since it takes commodities to make commodities how much net
commodities do we produce?  x - Ax.  This is because from the
production vector x, we must subtract what it took to create x.  Let y =
x - Ax.  y is the net production earmarked for consumption by people
(the final output vector).  Since some elements of y will be
intermediate commodities, they will be equal to zero (or nearly so).
Example: structural steel is an intermediate commodity except that a
museum might get some I-beams to display in it's exhibit which would
make this tiny amount of steel a final consumption commodity.

So how do we find energy from y = x - Ax ?  First we solve for x:
By definition x - Ax = [I - A]x, where I is the identity matrix.
Invert I - A and call the result matrix B, otherwise known as the
Leontief inverse.  Premultiply y = x - Ax = [I-A]x by B resulting in 
x = By.  So for any desired consumer output y we can find how much each
commodity in vector x needs to be produced.

So what is the embedded energy content of one unit of final output of
commodity y[k] (an element of y).  Simple.  Just let y = 0 0 0 0 1 0 0
0 ... where the 1 is in position k.  This is just an economy which
produces only one unit of the k th commodity and nothing else.
Substitute this y into x = By and you get a vector x of all the amounts
of all the other commodities required to create one unit of k.
Obviously, this vector is just the k th column of matrix B.  Now some
of these commodities to make one unit of k will be fossil fuels: coal,
oil, etc and we know how much of them we need to get a unit of the k
th commodity.  Once we know this we are about done, except to sum up
the energy by using the heat content of each fuel (BTU/gallon etc.).

So if you think about it, this is the correct way to do it and there
are a lot of technical papers written about this and related concepts.
But where is the data to create matrix A?  And just how big will
matrix A be?   We have to consolidate commodities that are slightly
different under the same commodity designation to keep A from being
too large so that it can be rapidly inverted and also due to detailed
data not being available.  For inverting an n x n matrix, the number
of steps by conventional methods is ~ n^3.

One simple way to create A is to measure all quantities in dollars and
use only the output of large economic sectors as commodities.  The BEA
(Bureau of Economic Analysis) has such data based on surveys taken
every 5 years.  I just read something regarding that they don't want
to disclose the details of this data collection and aggregation due to
things such as inaccuracies.  Perhaps it's the Bureau of the Census
that actually collects the data known as I-0 accounts but the BEA
publishes it (and it's on the Internet in Excel format).

I remember well in the 1970's the low energy efficiency reported for a
bicycle due to use of this method, based on $.  The energy per dollar
of tires was found and since bicycle tires cost a lot more per unit
weight than auto tires, this method grossly overestimated the imbedded
energy in bicycle tires and hence the energy intensity of bicycling.

But if one says that to avoid the complexities and inaccuracies of the
Leontief matrix method, one should just trace energy flows thru the
economy directly, it becomes a nearly impossible task.  To grow corn
may take fertilizer, to make fertilizer requires paper-work, to make
paper requires rail transportation, to provide rail transportation
requires railroad signals, to make the signals requires
semiconductors, and on and on so that this chain reaches an infinite
depth.  Don't say that each added energy component here is very small
and can thus be neglected.  Since there are an infinite number of
tiny components in each chain and also an infinite number of chains, a
very large number of them need to be counted to get accurate results.
In fact this type of accounting would visit each component (like
paper) and infinite number of times.  At each stage, energy is used.
Adding up the energy by this method consists of summing an infinite
number of infinite series, which of course must converge since no
commodity has infinite energy embedded in it.  The way to avoid all
this accounting is to use the Leontief model.  So I think that any
reasonably accurate analysis of embedded energy should use the
Leontief method.

There is also the question as to what are the final outputs.  Is all
human food a final output?  Perhaps not.  For example, if we use
animal power in production and feed the animals, then their food and
care would be considered an intermediate product  The same would hold
true for slaves doing the work.  If you don't need a slave any longer
you can breed less slaves and not have as many slaves to feed and
clothe.  But what about the working class?  If they are just wage
slaves, then their food, clothing, shelter, automobiles, etc. are just
intermediate goods and the final outputs are only the final consumer
goods for the leisure class.  How to define this class?  Does it
include homeless people who don't work?  One could divide some people
into two components, part of such a person would be the labor class
and represent an intermediate input.  Anther part would be the leisure
class type.  This might be the case for someone doing Linux
programming who gets laid off a lot and takes a vacation when not
working.  When unemployed they are part of the leisure class and their
energy and material use is a final output.

---------------------------------------------------------------------
Back to biofuels.  Suppose that there is no energy gain from use of a
biofuel.  That is, it takes the same amount of energy from fossil
fuels to make it, included use of embedded energy, as one gets back
when one burns the biofuel.  I'm not sure how the ethanol studies
accounted for embedded energy and it seems that the more optimistic
(and erroneous) estimates don't account for it in farm machinery.  How
the pessimistic estimate by Pimentel accounted for it I have yet to
find out.

With no energy gain, is there anything wrong with using crops for
fuel?  Of course there is a lot wrong with it.  One is that it costs
more and currently the government is wasting billions on such subsidy.
This waste, and many others, as well as a negative savings rate could
lead to financial collapse of the US.  For every $2 of imports we buy
we only export about $1 to pay for it and other $1 is either borrowed
from foreigners or we sell our corporations to foreigners help pay for
the deficit in foreign trade (foreigner buy our stocks).  Part of this
is due to government borrowing to cover it's wasteful practices such
as ethanol subsidy (and there are many more).  Financial collapse of
the US could result in a depression far worse than the one in the
1930s because our natural resource base such as oil are much depleted
and our industrial base has been in large part been destroyed by
imports.  We've also lost the know-how of the engineers and technicians
that used to work in these industries.  This wasn't the case in the 1930's

Another effect is to increase the price of food and take land out of
cultivation that was once being used for food production and convert
it to fuel production.  The means more deaths from malnutrition
(perhaps in some cases even starvation).  Of the worlds 6.5 billion
people, about half are malnourished.  Now one can argue that killing
people is a good thing because it reduces population and we are
overpopulated.  But starving people isn't a very humane way to reduce
population.  Birth control is more humane.  Furthermore, starvation
often kills someone whom society has already invested a lot in, such
as care as a child and education.  It makes no economic sense to make
such an investment in a person and then kill them prematurely.

A third problem is soil erosion and aquifer depletion (where water is
pumped from underground).  It's claimed that over the past 40 years,
about 30% of cropland has been taken out of production worldwide due
to soil erosion and shortages of water.  So if more cropland is needed
for growing fuels, it could result in even more erosion and
underground water depletion.

On the Internet, you can find many that have been tricked into
supporting the biofuels scam.  But there are some that haven't and
have published angry message opposing it.  One person equates biofuels
with genocide, saying that it will kill poor people due to less food
being available at low prices..

Actually, biofuels are what we used when the earths population of
humans was much smaller and each person used far less energy than
today.  Wood was used directly a a fuel, for cooking and heating.
In 1850 in the US, fuel wood consumption exceeded that of coal by a
factor of 10.  (Source "Energy in the American Economy", Johns Hopkins
Press, 1960. p.45+ "Fuel Wood" section).  Most railroad locomotives in
1850 were powered by wood.  But the amount of motorized travel today
is 200 times what it was in 1900 (See my website
www.lafn.org/~dave/trans/energy/fuel-eff-20th.html).  Now what was
the volume of motorized travel in 1850?  Assuming travel directly
proportional to rail mileage it would only be 5% of the value in 1900.
But there was also very significant steamboat travel in the US in 1850
so assume 10%.  This implies that we now travel about 2000 times as
much as we did in 1850 when we were using biofuels for transportation.
We stopped using wood because it was becoming scarce and more
expensive.  Note that using fuel wood likely resulted in a positive
gain as no energy was expended growing it.

So can we go back to biofuels now that we are doing 2000 times more
motorized travel than we did in 1850 using mostly biofuels?

So biofuels will work if we had far less population than we do today
and also traveled much less per person.  Biofuels might be feasible if
we could say reduce population by 90% and reduce travel per person
also by 90%.  This would mean reducing travel to only 1% of the
present volume which would be easy to support using renewable energy.

I just don't think there is a technical fix for our energy problem.  I
heard all kinds of tecnical fix proposals during the energy crisis of
the early 1970s.  They don't seem to have worked out as we now use
about 1/3 more petroleum as we did in 1970 and imports have about
quadrupled since then in spite of the proposals in the 1970's to
decrease imports.  

			David Lawyer