The Solsilc Process

1
The Solsilc Process

• Energiseminar
• June 11th - 2008
• Lars Nygaard

2
World electricity production from 2000 to 2040
TWh
100000
10000
World Electricity from PV in TWh
1000
World Electricity Generation in TWh
100
2020 PV 1 world electricity 2040 PV 26
world electricity
10
27
34
15 Growth
1
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
2024
2026
2028
2030
2032
2034
2036
2038
2040
Source Solar Generation and IEA-PVPS
Market and Applications
3
Silicon to the Photo Voltaic (PV) Industry

• The PV industry has historically used by-products
  from the electronic industry
• But the PV industry is growing faster than what
  can be delivered of by products from the
  electronic industry
• The consequence is a lack of solar grade silicon
  SoG-Si
• The future competitors we expect to be
• Optimized chemical processes with fluid bed
  technology
• Metallurgical refining processes under
  development
• Solsilc mean to have a competitive concept in
  this future market
• Let us look at the alternatives

4
The three ways to Solar Grade Silicon
Chemical
Metallurgical 98,5 Silicon
Refine liquid Si with slag gas and solidify
React Si with Cl and/or H to new compounds
Decompose to precipitate the Si and relief Cl/H
5
The three ways to Solar Grade Silicon
Chemical
Metallurgical
Metallurgical 98,5 Silicon
Metallurgical 98,5 Silicon
Refine liquid Si with slag gas and solidify
Refine liquid Si with slag gas and solidify
React Si with Cl and/or H to new compounds
Crush and leach out Fe, Al, Ca Ti
Remelt
Decompose to precipitate the Si and relief Cl/H
Directional solidification
6
The three ways to Solar Silicon Feedstock
Solsilc
Chemical
Metallurgical
Metallurgical 98,5 Silicon
Metallurgical 98,5 Silicon
Metallurgical 99,5 Silicon
Refine liquid Si with slag gas and solidify
Refine liquid Si with slag gas and solidify
Refine/Filter the Si in liquid state
React Si with Cl and/or H to new compounds
Crush and leach out Fe, Al, Ca Ti
Remelt
Decompose to precipitate the Si and relief Cl/H
Directional solidification
Directional solidification
7
Competitors down the metallurgical way

• Carsten Körnig, leader of Bundesverband
  Solarvirtshaft believe that in 2015 will the
  metallurgical way may have conquered 50 of
  the market for feedstock to solar cells
• Elkem Solar, Dow Corning, Ferro Atlantica,
  Becancour Silicon and others are working on this
  route as long as we know
• Test scale production the metallurgical way are
  in operation, but it is too early to appoint a
  single winner or some winners
• Solsilc Process is to our knowledge the leading
  process based on high purity raw materials

8
Competitors down the metallurgical way

• Carsten Körnig, leader of Bundesverband
  Solarvirtshaft believe that in 2015 will the
  metallurgical way may have conquered 50 of
  the market for feedstock to solar cells
• Elkem Solar, Dow Corning, Ferro Atlantica,
  Becancour Silicon and others are working on this
  route as long as we know
• Test scale production the metallurgical way are
  in operation, but it is too early to appoint a
  single winner or some winners
• Solsilc Process is to our knowledge the leading
  process based on high purity raw materials

The Solsilc concept saves several refining
steps This will compensate for higher raw
material costs
9
The Solsilc Conception

• Two Whitbread sailor men met in Cape Town
  February 1998
• Benno Wiersma, a Dutch business man with a strong
  belief in Solar energy
• Steinar Lynum, an engineer involved in the
  development of a pure carbon black process
• They soon found out that they needed some quartz
  to fulfill Bennos dream
• They contacted SINTEF and Ola Raaness got the job
  to find a suitable quartz
• The Solsilc consortium was formalized July 19th
  1999

10
The Solsilc development 1999 to 2006

• The development of the Solsilc concept were
  organized in two EU-supported projects and the
  main partners are
• Benno Wiersma who financed most of the
  development
• SINTEF contributing with knowledge and other
  resources
• ScanArc, a Swedish company with plasma furnace
  knowledge
• ECN, a Dutch energy research institution
• Then in 2004 the partners started a search for a
  possible partner who could contribute in the
  development and finally industrialize the Solsilc
  concept
• This search lead in 2004 to an informal
  co-operation with FESIL, a Norwegian ferroalloy
  company having produced and refined silicon for
  many years
• In 2006 a formal agreement was signed between the
  Solsilc partners and FESIL and a new company was
  established

11
Organisation
Structure map
Fesil Holding AS
Delta N.V.
Sunergy BV
SINTEF
100
55
45
20
ScanArc AB
Fesil Venture AS
Sunergy Investco BV
SHA
10
49
60
10
ECN
51
Transfer of technology
Fesil Sunergy AS
SDC AS Solsilc Development Company
12
Important raw materials Quartz

• Quartz is the most crucial raw material
• The Solsilc process require very pure quartz and
  our entrance will to a high degree influence this
  very limited market of to day
• FS has contacts with a number of quartz producers
  
• FS will during Q2-2008 sign contracts to secure
  future supplies of quartz with sufficient
  cleanness
• From early 2010 we will need about 15.000 t per
  year
• Expansions are planned in steps of 15.000 t
  quartz per year
• Each step of 15.000 t quartz will be a
  substantial volume in the world market for our
  quartz quality

13
The Quartz

• The earth crust is rich in Silicon repre- senting
  60 of its metallic elements
• Silicon is bound as oxide (SiO2), but most of it
  is found in combination with aluminium, iron,
  calcium etc
• Fortunately substantial amounts of very pure
  quartz (SiO2) are found widely spread around in
  the world
• This means

14
The Quartz

• The earth crust is rich in Silicon repre- senting
  60 of its metallic elements
• Silicon is bound as oxide (SiO2), but most of it
  is found in combination with aluminium, iron,
  calcium etc
• Fortunately substantial amounts of very pure
  quartz (SiO2) are found widely spread around in
  the world
• This means

In the global competition no company or
technology can substantially benefit from their
own quartz mine. In the long run the best
technology will win.
15
Important raw materials Carbon Black

• Carbon black (CB) is also a crucial raw material
• We need very clean carbon black. Therefore it has
  to be produced from natural gas
• Fortunately there are much clean gas around the
  world
• One carbon black producer cover 70 of this high
  purity market called thermal carbon black
• An agreement with this company will be signed in
  Q2-2008
• This producer can expand and cover all our future
  needs
• We have been able to obtain fair/nice prices
  because
• Very clean natural gas is available 120 km from
  Trondheim
• Our research partner ScanArc has the necessary
  technology to erect a carbon black plant if we
  should want our own production

16
Important for the process Pure electrodes

• The reduction process consumes an substantial
  amount of electrodes
• Therefore the electrodes must contain a low
  content of aches
• Low ashes electrode quality is normally produced
  with a limitation concerning the diameter
• Electrodes suited for the Solsilc process have an
  unusual combination of big diameter and a very
  low ash content
• Fortunately more than one of the electrode
  producers are able to produce electrodes with
  this wanted combination

17
The Solsilc Process in 7 steps

• The Solsilc Process can be described in 7 steps

18
Step 1 Agglom.

• 1. Agglomeration of quartz, carbon black and
  SiC-fines
• Function
• Normal production of silicon uses lumpy quartz,
  coal and woodchips, but those materials are not
  clean enough
• The quartz and carbon black are received as fine
  powders
• Since it is high gas velocities in the reduction
  furnace the raw materials must be agglomerated
  into pellets/lumps
• Status
• Clean binders are identified
• The agglomeration is optimized with a mix of
  selected binders
• An agglomeration plant can be erected by using
  standard industrial technology

19
Step 2 Silicon Carbide production

• Silicon Carbide
• Formation of silicon carbide (SiC) is an
  intermediate process in the reduction furnace.
  This is an energy consuming process and we want
  to relieve the reduction furnace from a part of
  this process

Here is a picture of SINTEFs rotating SiC
furnace, a very interesting process for the
Solsilc concept
20
Step 3 Reduction Furnace

• The picture shows the 100 kW pilot furnace
• The charge is agglomerates made of quartz, SiC
  and carbon
• It is a big challenge to succeed to get a stable
  operation with such special raw materials
• No wood chips or coal can be used caused by their
  content of impurities
• In June 07 we produced very pure metal we now are
  testing through the next process steps
• In June 2008 it is planned a new production
  campaign with very pure silicon

21
Step 4 Holding Furnace

• 4. Holding furnaces Tapped metal are collected
  and processed
• Function
• Get dissolved carbon to precipitate as SiC
  particles
• Stir the metal to get slag/oxides out of the
  silicon
• Status
• A practical method for transfer of tapped metal
  from the reduction furnace to the holding furnace
  is under evaluation
• We use induction furnaces, a well known technique
• We have good contacts with experienced furnace
  suppliers
• Some development will be needed concerning
  selection of optimal lining materials and
  controlling of the precipitation process

22
Step 5 Filtering

• 5. Filter Si with one or more filters
• Function
• Stop precipitated SiC and oxides from following
  the metal into the casting furnace
• Status
• Positive small scale results obtained November
  2007 at SINTEF
• Positive full scale results obtained in December
  2007 at FESILs Holla plant

23
Step 6 Casting
SINTEFs Crystalox furnace where pilot produced
silicon have been tested with very promising
results

• The casting will be done in a directional
  solidification furnace
• The solidification starts in the bottom and the
  slowly raising freezing front squeezes most of
  the impurities into the liquid silicon
• When the top of the silicon bath finally freeze,
  most of the impurities are collected here in the
  ingot top
• By cutting away the ingot top the analyses of the
  remaining part are substantially improved

24
Step 7 Clean Cut

• 7. Cleaning and cutting of the ingot
• Function
• The planned directional solidification accumulate
  impurities in all rim zones and where the metal
  freezing front finally stops
• The final ingot will be thoroughly analyzed
• A program for optimum cutting of the ingot will
  be developed
• Status
• Standard equipment for cleaning and cutting will
  be selected
• Rotating steel blades covered with diamantes is
  the standard
• Zones with much iron etc, will be sold as
  metallurgical silicon
• Zones only enriched on carbon will be returned to
  the reduction furnace where carbon is a necessary
  raw material
• Re-melting of silicon in this way will only cost
  us 0,5-1 /kg

25
The Solsilc development 2006 to 2007

• FESIL went into the development of the Solsilc
  concept with its knowledge and long experience
• Some strategies were changed and the development
  activities were expanded
• A new project was established (Solsilc Demo) with
  support from the European Union
• The Norwegian Research Council supports the
  Solsilc development through its BIA-program from
  2006 to 2010

26
Important Reached Milestones

• June 2007
• In pilot scale (100 kW) probably the most pure
  silicon ever was produced in a reduction furnace
• The produced pure metal have been refined by
  directional solidification. The resulting
  feedstock are remelted for production of ingots
  for wafer cutting. This work is going on just now
  
• November 2007
• The first half technical test ( 1 MW 10 x pilot
  scale) was a success
• The Solsilc process could be controlled in this
  scale. (Here we did not invest in ultra pure
  raw materials)

27
Important Reached Milestones

• June 2007
• In pilot scale (100 kW) probably the most pure
  silicon ever was produced in a reduction furnace
• The produced pure metal have been refined by
  directional solidification. The resulting
  feedstock are remelted for production of ingots
  for wafer cutting. This work is going on just now
  
• November 2007
• The first half technical test ( 1 MW 10 x pilot
  scale) was a success
• The Solsilc process could be controlled in this
  scale. (Here we did not invest in ultra pure
  raw materials)

We have demonstrated a reduction capacity at a
level of 1000 t/year of solar grade silicon
feedstock
28
The Quality Challenge

• As mentioned before the wafer producers for many
  years used only very pure by-products from
  electronic industry
• But insufficient supplies of this pure silicon
  forced the wafer producers to evaluate use of
  less pure silicon
• Fortunately tests have now showed that the wafer
  producers to some degree have been spoiled by
  clean by-products from electronic silicon
• This is now opening a market for metallurgical
  processes that never will reach the purity
  obtained with normal Siemens process.
• Typical analyses for acceptable metallurgical
  based feedstock were in Munich last month roughly
  indicated to be
• Phosphorous around 1,3 ppm
• Boron --- --- 0,3 ppm
• FESIL Sunergy has not yet announced any analyses
  of our expected quality, but the shown analyses
  are within our reach

29
Further plans

• We prepare for erection of different test
  facilities from the autumn 2008 at FESILs moth
  balled Lilleby plant in Trondheim
• Our letter to the Norwegian Pollution Control
  Authority was mid May picked up by the press
• Here a possible plant in Orkanger is taken up for
  discussion
• This does not mean that Orkanger is selected, but
  our board of directors want to evaluate the
  terms/support here compared to other locations
• To day two people are working full time in FESIL
  Sunergy, but early February one person with a PhD
  will join our team
• It is very important to FESIL Sunergy to hire
  skilled people from the FESIL organisation. In
  this way we more or less double our working
  capacity
• SINTEF, ScanArc and ECN contribute with
  absolutely necessary RD competence and capacity
  to the project
• FESIL Sunergy will soon go out to search for more
  clever people to join our small, but efficient
  staff

30
Reduction Process Challenge

• When you can not use any of your normal raw
  materials, you have to change your habits
• This experience felt strange, perhaps like
  replacing a golf ball with a football. We managed
  that new game
• We are very glad that the ball did not feel like
  a curling stone. Then the project had stopped