September 2011 - I-Micronews

I S S U E N ° 1 3 S E P T E M B E R 2 0 1 1
Inkjet has high potential for non-contact deposition
of etchants, dopants and metals, though the process
can be slow, and the fast-drying materials can easily
clog the tiny nozzles to limit performance, presenting
needs for new inks, new head designs and cleaning
processes, and perhaps new solutions to recognize
and correct print areas where blocked nozzles
have impacted deposition quality. Lasers also have
considerable potential for innovative improvements
to solar manufacturing, as their highly-localized
heating allows precise control of micromachining
with little thermal stress or mechanical damage for
fine geometry features on delicate thin wafers. Both
inkjet and laser production processes are already
well established in other high volume industries,
which could accelerate the development of solar
applications.
Higher quality cells require purer,
higher-performance materials, more
controlled processes
Improving solar efficiency will also require purer
materials and more careful control of processes
for contamination. As cell efficiency starts to get
closer to theoretical limits, after 18% or so, eking
out the next round of gains will mean paying
more careful attention to the small stuff. The PV
industry will need to evolve beyond its focus on
high speed automation modeled on that of the
automotive industry, to start to look more like the
semiconductor industry - with its obsession with
contamination control and process monitoring, and
its practices of filter use, cleaning technologies, and
keeping humans away from the processing. Gases,
de-ionized water and other standard materials
for PV production will all move to higher levels of
purity and quality. On the cost reduction front,
replacing silver metallization with copper could gain
a significant advantage, but some reliability issues
with corrosion and Cu migration into the silicon
remain, necessitating additional buffer and capping
steps that add back costs.
Demand will move to higher purity
polysilicon as well
PV makers are increasingly demanding higher
quality polysilicon as well, for higher efficiency cells.
Yole Développement projects that demand will
move to semiconductor grade material—9 nines to
11 nines purity—essentially displacing solar-grade
poly almost entirely in the market. With a backlog
of solar module supply built up in inventory, and
slowing demand from the cutbacks in government
incentives, demand for poly this year looks likely
to come in considerably below last year’s 160,000
MT, leaving an oversupply in the sector. The four
leading suppliers Wacker Chemie (Germany),
Hemlock Semiconductor (Michigan, US), OCI
(Korea) and GCL Poly (China), who are all providing
semiconductor-grade material, together have enough
P V M a n u f a c t u r i n g
capacity among them to supply the entire PV
industry demand this year. With no constraints
on availability, it will be the higher purity material
from these leading suppliers that sells, and enough
demand for it to maintain poly prices stable at
around $50/kg this year. Many of the new entrants
who more recently added poly capacity have found
it harder than expected to optimize this complex
chemical process to match the leading pure-play
players in quality and cost, and many will likely
quietly get out of the business.
HCPV efficiency could provide an edge,
if sector and supply chain can scale to
volume
Long term trends towards efficiency could bode
well for the future of HCPV, and this still barely
commercial technology has wide margin to bring
down costs if it moves to volume production. HCPV
now reaches 29% efficiency at the module level at
leading suppliers, compared to 17%-20% for the
top flat panels, and it has even more of an edge in
hot climates, as its performance degrades less at
temperature.
Plenty of hurdles remain, however, for this
complicated new technology. Appropriately sunny
locations remain limited, long term reliability remains
unproven and actual commercial production so far
remains insignificant. Though HCPV suppliers claim
that the technology can deliver the lowest levelized
cost of energy with the least use of land on some
sites, the actual results can vary significantly for
particular location, climate and topography, so even
big users first want to try small test installations on
their sites, and this small installation stage could
continue for some time. Meanwhile, the large flat
panel PV sector is continually investing its far bigger
resources into improving its efficiency and bringing
down its costs. While HCPV may have big potential
with enough investment, with the market still so
small, the sector has limited resources to put into
development.
The supply chain of key components on which
HCPVsystems depend has even less to invest, as the
total markets for each of the various components,
from epi wafers to lenses to trackers, remain even
smaller still. Progress will depend on getting to big
projects to get volumes up to bring down costs,
by spreading the fixed costs of the 25MW-50MW
capacity HCPV plants over megawatt and not just
kilowatt-scale projects, and fully automating both
component production and system assembly - and
finding the significant capital to invest to do so.
Perhaps even more important will be finding the
solutions for the most effective business models or
working partnerships among the parts of the supply
chain. As volumes grow, HCPV systems makers
17
“Lasers also have
considerable
potential for
innovative
improvements
to solar
manufacturing,
as their highlylocalized
heating
allows precise
control of
micromachining
with little
thermal stress
or mechanical
damage,”
says Milan Rosina,
Yole Développement