Thin Film PV Partnership

1
Thin Film PV Partnership

• Solar Program Review
• Denver, Colorado
• November 7-9, 2005
• K. Zweibel

2
Objective

• The Thin Film PV Partnership focuses on
  subcontracted research in the leading thin film
  technologies in terms of meeting ambitious,
  long-term cost goals for all PV markets.

3
The Best of Times, the Worst of Times

• Strange days
• Great technical progress thin films finally
  coming to fruition
• Lousy budget, mostly due to earmarks cutting into
  subcontracts
• Total emotional schizophrenia
• Present the most upbeat talk about the future of
  the technologies and PV?
• Or, Whine about the budget?
• Lets touch on the budget, because it matters
  but focus on the technology

4
Recent and Predicted Thin Film Module Production
by Technology
Predicted levels from US DOE/NREL Thin Film
Partnership Technology Partners and Sun and
Wind Energy 2/2005. Out-year values (2007 -
2011) conservative.
5
Subcontract out the door has dropped
50 since 2002.
Partnership Subcontract Budgets (out the door)
2002-2005 Actuals 2006 Initial Guidance
6
Partnership Subcontract Budgets (out the door)
2002-2005 Actuals 2006 Needs, negotiated
funding levels
7
Summary of Activities

• The development and improvement of thin film
  module technologies (CdTe, a-Si, CIS) working
  with leading US companies, universities, and
  NREL.
• Module development includes work in
• Cell improvement
• Process development
• Module design and
• Reliability testing.
• Actions are carried out by subcontractors and
  coordinated through National RD Teams with
  In-House NREL researchers

8
High-Level Accomplishments

• Helped create CdTe technology as a key option for
  PV
• Some remaining challenges
• 25 MWp in production at First Solar 50 MWp under
  construction
• Supported ongoing creation of CIS as an option
  for PV
• Only pilot production (sub-10 MWp)
• Not yet accomplished
• Many start-ups, but no money to fund them
• Supported creation of flexible a-Si for rooftops
• 25 MWp in production at UniSolar 25 MWp (or
  more) under construction
• Trying to maintain support of technologies while
  they transition to true commercial success
• Why? Because they can still fail due to risks
  associated with explosive manufacturing growth of
  immature technologies and perhaps unknown outdoor
  reliability issues.
• But funding issues causing us to back away from
  this.
• Also trying to support pipeline of developments
  needed to make long-term PV competitiveness goals
  (6 /kWh or lower higher efficiency modules,
  lower cost processes, stability), but falling
  below critical mass of previous efforts

9
20
NREL
CuInSe2 CdTe Amorphous silicon (stabilized)
NREL
NREL
16
Univ. of So. Florida
BP Solar
EuroCIS
Boeing
Boeing
ARCO
Univ. of So. FL
12
Kodak
Photon Energy
Efficiency ()
AMETEK
Boeing
United Solar
Monosolar
Boeing
Kodak
8
Matsushita
Boeing
ECD
Univ. of Maine
The Best One-of-a-Kind LaboratoryCell
Efficiencies for Thin Films(Standard Conditions)
4
RCA
0
2000
1995
1990
1985
1980
1975
2005
02658722
10
Commercial Thin Film Modules
Data Taken from Websites (prices are estimates)
Compiled by Bolko von Roedern 8/2005 Temperature
coefficients will vary slightly depending on
local spectral content. Company source reports
-0.48/ºC may be more accurate for recent
product. Disclaimer Listing could be outdated or
incomplete (missing manufacturers and/or some
"best" product) prices are estimates for large
quantities.
11
Cost Issues and RD Planning

• Use of System Advisory Model (SAM) for RD
  planning drives us towards making cost and price
  estimates, despite uncertainties
• SAM is primarily a financial and climate
  simulation model input assumptions are critical
  to getting useful output
• To respond to BOS input needs, developing BOS
  cost estimating spreadsheet for handling all
  aspects, including integrators costs and margins
  for all applications, modules, and array sizes
• Spreadsheet is a work in progress, and results
  shown here are only meant to be illustrative.
  However, the challenge to understand relative
  competitiveness and potential is ongoing.
• The results shown here are post-Multiyear Program
  Plan (i.e., they are not part of that document or
  its underlying philosophy). They are part of the
  evolving development of that plan, for the
  future.
• BOS cost estimator is available on request and
  will eventually go up on our website. Feedback is
  most welcome.

12
Next few slides

• Will compare thin film technologies with x-Si to
  get a feel for the status of thin films today,
  pointing the way towards future needs.
• It will also establish that thin films ARE FULLY
  COMPETITIVE RIGHT NOW IN ALMOST ALL PV MARKETS.
  This has never before been true.
• That is why these are the best of times

13
Residential Roofs 2005
Adjusted to same roof area as the 4 kWp x-Si
system
14
UNI-SOLAR Shingle Roofing Products A Solar
Shingle Installation
15
Observations about Residential Roofs

• One-time soft integrator costs for small
  systems play a big negative role (marketing,
  design, inspection, permit, transportation,
  management, profit).
• High OM costs imply the need for systems with
  the most output per unit area and for all
  residential systems, OM can actually affect
  payback.
• Limited roof space favors x-Si, which produces
  the most W/area.
• Rack-less, thin film laminates can offset most of
  this x-Si advantage.
• The cost difference necessary for thin film on
  glass modules to compete is substantial if roof
  space is limited (e.g., about 1.5/Wp)
• All systems get hit hard by smaller roof space
  can any be economical below 2 kWp? (In the
  future, yes see below.)
• Mortgage deduction is critical to economics
  retrofits are worst case economics
• Creative approaches may alter the economics by
  minimizing one-time front-end costs and by
  mandating new construction to include PV.
• This is a unique market with its own drivers
  observations based on the residential market
  should not be over-used to characterize other
  major PV markets. However, since the residential
  market is so politically important, these
  idiosyncrasies will continue to play a role in
  the evolution of PV acceptance.

16
Flat, Commercial Roof 2005
Conventionally accepted LCOE definition ignores
loss of business fuel cost tax deduction and
makes these LCOEs cash flow negative with equal
fuel rates add 40 to LCOE to get approximate
cash-flow-neutral adjusted LCOE. Use of the
conventional LCOE approach makes these and the
MYPP LCOEs for this application seem artificially
low.
17
UniSolar and Solar Integrated Technologies
18
UniSolar Rooftop System, Beijing Capital Museum
19
First Solar CdTe Rooftop
Katzenbach Juwi
Memmingen SAG
20
Observations on Flat, Commercial Roofs

• Flat commercial roofs are large enough to avoid
  the impacts of one-time soft costs and OM on
  low-efficiency modules.
• Adjusting the array size downward to fit limited
  roof space does not alter the economics
  significantly.
• All module technologies are competitively very
  similar, even the lowest efficiency thin film
  modules on glass, given normal module price
  assumptions.
• Laminate modules have a clear advantage that
  allows them to be priced somewhat higher and
  still be lower at the system level.
• Thin films with normal pricing are fully
  competitive in this market with x-Si.
• Even with an adjusted LCOE 40 higher than given
  in the prior slide, PV is in the range of 17-20
  /kWh, which is becoming directly competitive
  without subsidy in some high-priced retail
  electricity regions RIGHT NOW (in contrast with
  the residential market, which is about a dime/kWh
  higher due to small array sizes).

21
Large, Ground-Mounted (nontracking) 2005
A pay-as-you-go approach like Tucson Electric
favors avoids loan costs and would reduce LCOE by
as much as 50 from those given here.
22
Mainbernheim
Dimbach
Rost Gschwend
System Integrator Beck Energy
23
First Solar Array at Tucson Electric, Arizona
24
Observations on Ground Mounted

• Array size further reduces impact of first costs,
  OM, and other BOS impacts on low-efficiency thin
  films.
• Due to their low module costs, thin films appear
  to be gaining a significant competitive
  advantage, which should grow going forward
• With pay as you go financing early utility
  adopters will only need another halving of costs
  to make PV cost-competitive with conventional
  energy sources. This will make meeting state
  portfolio standards and other incentives less
  onerous during the transition to full
  cost-competitiveness.
• There is surprising opportunity in utility-scale
  and distributed ground-mounted PV, even on the
  wholesale side of the meter.

25
Robust Takeaways

• Thin films are a lot more competitive today than
  almost anyone realizes this is the big news of
  2005.
• Except for the smallest systems, even
  low-efficiency thin films may be priced low
  enough to be competitive
• Certain technologies have clear competitive
  advantages
• Laminates for roof markets (e.g., UniSolar a-Si)
• Reasonable to high-efficiency thin films with
  moderate to low manufacturing costs (e.g., First
  Solar CdTe)
• Maybe the US hasnt lost the international PV
  wars quite yet
• And why? Because DOE had the foresight to fund
  RD, not just demonstrations and markets.

26
Back to the Future Why should thin films become
price leaders?

• Several thin films competitive today in various
  PV markets
• Todays thin films are manufactured on a much
  smaller scale than x-Si (future economies of
  scale are easier for thin films)
• Scale-up and continued tech development of thin
  films take advantage of low-hanging fruit,
  since thin films are less technically mature
  (progress is not close to asymptotic)
• Another way of saying it Thin films will come
  down their learning curve more quickly than x-Si

27
Experience Curve Impact of Smaller Cumulative
Production and Higher Growth Rates
If a disruptive technology can once become cost
competitive, the market will try to equilibrate
purchases, resulting in a faster growth rate and
a faster decent of the experience curve for the
new technology.
28
Future Potential

• Lets look at the future when we combine
  technical improvement in thin films with BOS cost
  reductions and market aggregation and volume
  increases the DOE 2020 goals.

29
2020 Residential (Kansas City) All _at_ 40 m2 size
Kansas City has average US sunlight, about 1700
kWh/m2-yr 25 less than Phoenix but
residential and commercial roofs will be
ubiquitous, not confined to the Sunbelt.
30
Residential 2020

• Toughest market due to small size systems
  (upfront costs and fewer pieces to absorb
  overheads)
• Still, thin film laminates look cost-effective,
  and others are in an acceptable range
• Looks adequately low to be cost-effective even in
  average locations especially laminates.
• Assumes some smart new approaches to new
  housing and standardization of costs like taxes,
  inspection, and permitting to lower overheads and
  one-time-costs.

31
2020 Commercial Flat Roof Phoenix
Thus unless the same deduction is allowed,
LCOEs would only be cashflow neutral for
utility rates 40 higher than those given here.
32
2020 Commercial Flat Rooftop

• Attractive, large market
• Even if hit by 40 loss of tax deduction for
  avoiding fuel costs, still attractive
• Low enough that it remains quite attractive in
  all US solar locations (another 25-50 cost
  addition)
• Even greater societal value if tax deduction of
  virtual fuel costs retained why not? The fuel
  costs are being avoided!
• Possible incentives (like the above tax
  avoidance) to make on-site storage or fuel
  production attractive, could also open new
  markets.

33
2020 Ground-mounted Phoenix (IPP financing)
34
Ground Mounted 2020

• Largest advantage for thin films due to smallest
  one-time and BOS costs
• So inexpensive that utilities may be able to pay
  as they go like Tucson Electric does, reducing
  costs well below IPP (loan) levels maybe
  30-50!
• Could open up a serious new avenue for PV sales
  as utilities try to meet portfolio standards from
  cash flow
• Attractive everywhere in the US
• Possibly low enough to sustain storage/dispatchabi
  lity costs and fuel-making costs.

35
A few other thoughts

• These arent even the ultimate potentials of
  these thin film technologies possible another
  25-60 further cost reductions.
• PV will be coming down in cost throughout the
  21st century as technology improves a boon to
  humanity.
• Once cost-effectiveness and size of solar
  resource is understood, need for storage and fuel
  production RD will come into focus to meet
  climate change and oil depletion.

36
Technology Maturity Stages
37
Approximate Chronology of Thin Film Maturity
Stages
38
Stages and Gates

• In the US, CdTe and a-Si have turned the corner
  and are both present in the marketplace and
  directly competitive with established x-Si
  technologies for key applications
• Flexible a-Si for PV laminates on large, metal
  roofs
• CdTe/glass for large fields and large flat roofs
• CIS remains in pilot production, and progress to
  market presence is hesitant (due to greater
  complexity)
• Non-US Wurth Solar is in that transition many
  new start-ups attracting attention and money
• If CIS can make this transition, likely will be
  quite competitive in all markets
• Do not compare technologies at different stages
  based on the same criteria CIS is not in the
  same stage as a-Si and CdTe or x-Si and needs
  further development (assuming its worth doing
  which it is).

39
Recent Past, Near-Term and Mid-Term Future

• Thin films came off the mat in 2004, US
  production doubling and doubling again in 2005
  (and maybe in 2006-07, too)
• US leads in thin films
• Technological readiness was as critical as the
  silicon shortage without it, thin films would
  not even have been able to respond.
• Near-term looks fabulous possibly more
  doublings, rapidly bringing thin films to center
  stage and pressuring silicon producers who want
  refiners to expand feedstock capacity
• Mid-term future depends on success of fielded
  systems reliability and relative rates of
  successful (?), explosive expansion possible
  thin film majority in the marketplace much sooner
  than people realize, unless thin films stumble
  due to embedded risks.

40
Future Directions

• Support thin film technologies towards lower
  costs by keeping the pipeline full of innovation,
  assuring companies stay aggressive
• Foresee and address issues that are outside
  immediate corporate horizons
• Indium and tellurium rarity (i.e., The TW
  Challenge thinner cells)
• Cadmium ESH issues and perceptions
• Push towards even more aggressive cost goals
  suitable for PV fuel production and dispatchable
  electricity (below current DOE EERE goals)

41
Some Concerns

• Partnership subcontract funding is dropping
  rapidly, so US technical lead in thin films may
  dissipate.
• Infrastructure built up over twenty years at
  universities is already being lost.
• Ability to fund start-ups is almost negligible,
  yet there are a large number of start-ups.
• National Teams may be dropped.
• Center of Excellence may be dropped.
• Hot and Humid Testing may be dropped.
• Good progress at leading companies may not be
  supported.
• Very high growth scenarios depend on sustained
  move of technical accomplishments to
  manufacturing robust ability to transfer early
  manufacturing to much larger volumes and
  near-perfect reproducibility of making reliable
  modules that last outdoors threatened by lack
  of our support.
• Partnership Program faces possible shut down mode
  over next half decade if budgets do not reverse.
• But thin films may succeed, and US PV leadership
  may be restored, anyway!
• The best of times, the worst of time

42
Website

• To review this talk in its entirety, go to
• http//www.nrel.gov/ncpv/thin_film/