Awhile back we were discussing whether or not a photovoltaic module
returns more energy than it took to manufacture.  I posed the question
to the PV Users list and it initiated a discussion that is still going. 
I found this recent posting especially interesting and hope at least a
few ecopathers do, also.  A bibliography on this topic follows.  

Doug Fields

-------------------------------------------------------------------
This subject keeps coming up. Sometimes I think
that the last couple generations have had so many
disappointments in having our beliefs shattered,
from believing in Santa Claus, on down.  People
start to think, "Whatever we do, it is a losing
battle, entropy wins, goodbye Planet Earth."
But no, PV panels really do make a net gain in
energy, and it doesn't take all that long to
recoup the energy expended. It does, however,
depending, not as you might think-- on how you
measure the total inputs--but rather, on which
solar cell technology we are discussing.
We recently had an interesting discussion of this
a couple months ago at a BP Solar sponsored
conference in New Orleans about 3 months ago, and
we had a lot of people from various aspects of the
industry, including Richard Perez of Home Power
magazine, David Katz of Alternative Energy, Dr.
Jean Posbic, Sr. Technical Director for BP Solar,
Gerry Braun, Thin Films Dept., BP Solar, and about
40 or 50 of the who's who in the industry.
OK, enough credentials. Taking worst case
scenario, old method of making silicon ingots from
scratch, in an electric induction furnace,
starting from raw silica sand to liquefy, purify,
cast, and then cool the ingots gradually, slicing
them with a diamond saw (all of which is how
Siemens still makes their cells), it takes maybe 6
years at the most to re-coup the kilowatt hours
that went into production. Remember, the
production process itself is done on a mass scale,
so the costs are spread out by thousands of
kilowatts of PV panels made every year. The PV
manufacturers themselves are always trying to save
steps, save money, and above all, save
electricity. There are a lot of factories out
there whose years production is somewhere on the
order of a couple of megawatts per year.
Next, it is no longer necessary to start from raw
silica oxide sand. A PV cell only requires 1 part
per million purity. Computer chips have to be 1
part per billion impurities. There is now so much
computer chip scrap silicon on the market, that it
has become a major source for PV production.
There are a lot of new technologies which have
brought the energy cost of making PV modules down
to about 3 years.  Some of these include the
extruded ribbon technology, which has many
variations. The most efficient of this particular
method is the production line of ASE, Americas, in
Billerica, MA. They extrude six ribbons in one
continuos six-sided hollow tube. There is no
waste, as they cut the tubes first into 3-inch
high six-sided rings (for lack of a better term),
with a laser. Then they cut the edges of these so
that they end up with six - 3"square thin cells,
with no waste. It is all computer controlled and
robotized.
The thin film people have really got it down, now,
too. Then there are the people like Amonix, who
use only a tiny piece of silicon chip, the size of
your little fingernail, and they make it precisely
like a computer chip, with pins integral to the
back of the chip. Think of what a Pentium chip
looks like, only about 1/4 the size. Then, they
put a huge Fresnel lens on the front and a heat
sink on the back and focus 300+ suns on that
little chip. You don't hear about this technology,
as it isn't available for the home market: it is
only reserved for huge arrays, as they use a
sophisticated tracker with high precision
accuracy. They are not to be confused with Midway
Labs out of Chicago, who are now out of business.
Amonix has put up some 3 and 5-megawatt power
stations, feeding the grid.
Finally, a physicist down in Brazil did and
interesting calculation a couple of years ago,
which was reported in Renewable Energy magazine,
which is a European journal. He compared the
energy it takes to derive, refine, and produce
power from something like 5 grams of plutonium to
the energy produced by 5 grams of silicon, using
the same criteria, and figuring in all the aspects
of manufacture, ganging cells into modules, etc.
He came up with a figure, if I remember correctly,
of around 7 years for the silicon, and something
like 120 years for the plutonium. However, the
waste disposal of the plutonium was not even
entered into the equation, as it wouldn't have
been fair to the silicon, as silicone isn't
poisonous!
- -- Robert Warren
---------------------------------------------------------------------
A number of people on the list requested that I follow through on my
offer,
several weeks ago now, to post a bibliography of papers regarding the
embodied energy in PV modules. My list is not a complete account of
everything ever published on the subject, just the papers I've cobbled
together in spare moments from the resources available at the Georgia
Tech
library (you'll probably need a good university library to find these).
With
that disclaimer, here they are:

K. Kato, A. Murata, and K. Sakuta, "Energy pay-back time and life-cycle
CO2
emissions of residential PV power system with silicon PV module,"
Progress
in Photovoltaics, v. 6, pp. 105-115, 1998.

K. Kato, A. Murata, and K. Sakuta, "An evaluation on the life cycle of
photovoltaic energy system considering production energy of off-grade
silicon," Solar Energy Materials and Solar Cells, v. 47, pp. 95-100,
1997.

A. Inaba, K. Yamada, and H. Komiyama, "An energy evaluation for solar
photovoltaic energy systems," 28th Intersociety Energy Conversion
Engineering Conference, pp. 2.481-2.484, 1993.

W. Palz and H. Zibetta, "Energy pay-back time of photovoltaic modules,"
International Journal of Solar Energy, v. 10, pp. 211-216, 1991.

H.A. Aulich, F.W. Schulze, and B. Strake, "Energy pay-back time for
crystalline silicon photovoltaic modules using new technologies," 18th
IEEE
Photovoltaic Specialists Conference, pp. 1213-1217, 1985.

The two papers by Kato, et al. are very similar, but seem to be very
high
quality work. I hope these will be of interest. I highly recommend that
anybody interested in this subject also dig up the references at the end
of
the paper posted to the list by Richard Corkish on December 21.
___________________________________
Alan Ristow
University Center of Excellence for Photovoltaics
Georgia Institute of Technology
Atlanta, Georgia

Reply via email to