Environmental Engineering Reference
In-Depth Information
$1 per peak watt
, in December 2007. (In the rating $1/Wp, Wp is the power that the
modulewould producewhen illuminatedwith theAM
1.5 spectrumat full intensity
1000W/m
2
. The average intensity in the United States is about 205W/m
2
.) The
founder of the company was quoted to the effect that his
rmwas the
rst to
pro
tably
market cells for less than $1/Wp
. The cells are intended for a 1MW facility in Germany.
It is said that the capacity of the 200 000 square feet facility, where production is based
on roll-to-roll processing, is 430MWp per year. (The company also has plans to
produce roof tiles, which have been a popular product of Energy ConversionDevices/
United Solar, a solar energy
firm of longer standing.) The ef
ciency of the cell is
embedded in the $/Wp but is not stated by the manufacturer. Upper limit ef
ciency
attained in this class of single-junction cell is 20%. Learning curve effects and even
the chance of a tandem version of the mass printed CIGS cell may make 20% a
reasonable eventual efficiency. The First Solar CdTe process does not use nano-ink,
but nevertheless achieves less than $1/Wp cost. It is also stated that a new CdTe
startup, Solextant, will use a nano-ink process (http://Wpww.greentechmedia.com/
articles/read/Thin-FIlm-Solar-Startup-Solexant-and-its-New-CEO/) for CdTe.
Estimates for Nanosolar producing printed CIGS cells depend upon facility cost
about $100million (roughly) for production capacity stated as 430MWp per year, and
cost per watt of the cell (stated as certainly less than $1/Wp in 2007). There is
indication that the real cost in successful production may be as low as 30 cents/Wp.
According to a rule-of-thumb conversion $1/Wp
6 c/kWh, so printed CIGS cells
(based on our estimate of 30 c/Wp) could be used to market power at a rate 2 c/kWh.
A con
rmation of this general scenario is provided by present results for the
similar CdTe cells, for which panel cost has been reported as $0.76/Wp, and of which
the volume production was 1.4GW in 2010 (see Figure 10.2).
¼
10.4.3
Projected Total Capital Need, Conditions for Profitable Private Investment
We can use these approximate numbers (a plant to produce 430MWp per year costs
$100 million) to roughly project the capital costs of a possible solar energy future
(supply total electricity usage in the United States, neglecting costs of land, storage,
and transmission).
If the national average electricity use is 460 GW, and we want to build that capacity
over a 10 year period, then building
¼
522
plants on the Nanosolar, Inc. model could totally meet that need. This gives an initial
facility cost (at $100million per plant) of $52 billion and a solar cell production cost (at
$0.3/Wp) of $0.3
(460 GW/430MW) (1000/205) (1/10)
$67.3 billion per year for 10 years.
Abound Solar (http://Wpww.abound.com/solar-modules/manufacturing) (http://
Www.sustainablebusiness.com/index.cfm/go/news.display/id/21597) has received
a $400 million loan guarantee toward expansion of its production capacity from an
initial 65MWp to 845MWp. From this, we can make an estimate that the capital cost
for a production line is about $400million/800MWp
460/(0.205
10) billion
¼
0.5 $/Wp. If we adopt this,
and ask how much capital would be required to build capacity in solar modules over
10 years to provide 500 GWcontinuous usage, the peak capacity will need to be larger
¼
