Environmental Engineering Reference
In-Depth Information
earlier thin-
lm Si cells. To what extent they are limited by the relatively small supply
of ingredients such as Te, In, and Ga, and the toxicity of Cd, is not entirely clear.
The type III cell, at least in thin-
lm form, is purely conceptual at the present date.
It appears that the only cells to operate in or above the limiting range between 30 and
40% ef
ciency are, in fact, multibandgap epitaxial structures, containing crystalline
layers of GaAs, AlAs, InAs, and germanium. These cells are so expensive that they are
credible for large-scale application only in conjunction with concentration factors of
several hundred. Amosaic of such cells would be placed, for example, at the focus of a
parabolic re
ector as shown in Figure 5.5, and would de
nitely need water-cooling to
keep the expensive cells in a safe operating temperature range. The multi junction
concentrating cell systems, to be described later, appear to bemore expensive than the
0.2
-
0.4 $/W suggested in Figure 5.6 as typical of type III cells. Yet, large-scale
concentrating multijunction systems have been presented as capable of grid parity
in favorable locations such as the American southwest or southern Spain. Although
such type III concentrating systems are proposed as competitive with large arrays of
simpler type II cells such as CdTe or CIGS, the fact is that no large facility to date has
been based on the light-concentrating multijunction cell systems. The marketplace
decision to datemay be affected by the higher complexity of the concentrating system
compared to a large array of thin-
lm modules, which do not require tracking and
water-cooling.
The balance on this choice may yet change. The concentrating systems clearly
minimize the volume of semiconductor materials, notably do not require large
amounts of Cd, Te, In, or Se. While performance has been lower, large-scale
concentrating systems based on crystalline silicon (of wide abundance) may yet win
over the multijunction systems, which are based on GaAs, AlAs, InAs, and germa-
nium and have reached 41% ef
ciency. In principle, a single-bandgap cell, for
example, crystalline silicon, with concentration by mirrors or lenses, can reach
40% ef
ciency. The multijunction cells are a proven technology, widely deployed in
space programs going back to the Sputnik Russian space program era. The tech-
nology is available, but its cost and complexity are primary issues, ahead of possible
material supply issues, in adapting to solar farms on earth.
Figure 5.6 suggests that type III cell cost might, in principle, be as low as 0.2 $/W,
although there seem in 2011 to be no examples of thin-
lmcells above 20%ef
ciency.
Martin Green, leading researcher and author [55], stated, There would be an
enormous impact on the economics if these new (Type III) concepts could be
implemented in thin film form, making photovoltaics one of the cheapest known
options for future energy production. Third generation is an appealing description
for any manufacturer who wants to sell his latest product. Following Martin Green,
Figure 5.6, we will reserve the terms Type III or Third Generation, to devices in
the 20
-
60% ef
ciency range. The only existing type III devices, according to this
de
films but concentrating epitaxial crystalline multijunction
devices based on GaAs molecular beam epitaxy, liquid phase epitaxy, or organome-
tallic epitaxy.
We will return to the physics of the ef
ciency limits in the range 30
-
40% for single
junction cells in Chapter 6.
nition, are not thin
