Photovoltaics and the RoHS

Photovoltaics and the RoHS
Directive
Mathieu Saurat
Michael Ritthoff
Wuppertal, May2010
Position Paper
Wuppertal Institute 2010       2
Photovoltaics and the RoHS directive
Mathieu Saurat, Michael Ritthoff; Wuppertal Institute, May 2010
Over the last years, different photovoltaic (PV) technologies became commercially available,
while  several  others  are  under  development.  PV  technologies  require  diverse  materials  to
generate  electricity  from  sun  light,  including  in  some  cases  toxic  materials.  To  date,
photovoltaics  are  not  covered  by  the European  Directive  on  the  restriction  of  the  use  of
certain  hazardous  substances  in  electrical  and electronic  equipment (RoHS  directive),  but
there is an ongoing discussion about whether or not the RoHS directive should be extended to
cover photovoltaics.
Against this background, the Wuppertal Institute for Climate, Environment and Energy has
conductedan independent,scientificand open-ended as regards its outcomeposition paper on
the issue of a possible extension of the RoHS directive to photovoltaics.
*
1. Renewable energies and photovoltaics
The European Council endorsed at its Meeting in Brussels on8/9 March 2007 a binding target
of a 20% share of renewable energies in the overall EU energy consumption by 2020. Out of
the  24  GW  of  new  power  capacity  constructed  in  the  EU  in  2008,  19%  (4,700  MW)  were
photovoltaics  (PV)  (Kautto  and  Jäger-Waldau  2009). This  is  more  than  the  target  of
cumulative installed PV system capacities that the European Union had set for itself to reach
in 2006 (3,000 GW). In 2008 the overall installed capacity of solar photovoltaic electricity
was estimated at 9,100 MW.
A varietyof technologies fall under the broadly used term “photovoltaics”. From a material
use  perspective,  two  large  groups  of  such  technologies  can  be  discerned:  silicon-based  and
non-silicon based PV. The former group comprises both crystalline and amorphous (thin film)
PV systems. The latter group is mainly composed of cadmium-telluride (CdTe) and copperindium-(gallium)-selenium  (CI(G)S)  cells,  both  thin  film  technologies.  Since  2006  the
production of thin film PV systems hasexperienced a growth rate higherthan that of the PV
sector as a whole (Jäger-Waldau 2009).
In particular, manufacturers of CdTe PV have rapidly scaled up their production capacities.
The comparatively low costs of thin film PV systems –especially CdTe PV systems–are often
mentioned as a reason for the rapid growth of PV. However, the relatively low price of CdTe
modules  is  partly  compensated  by  higher  installation  costs  because  of  lower  efficiency  of
these thin modules compared to traditional Si-PV.
The  criticality  of  certain  materials used  in  these  technologies  can  be  assessed  considering
their scarcity and toxicity (for humans as well as for the environment). In that respect indium,
gallium and  cadmium  as  “rare  metals”,  and  selenium  and  tellurium  as  “rare  earths”,  are
*
Support by the Non-Toxic Solar Alliance is appreciated.
Wuppertal Institute 2010       3
considered scarce resources. Reserves are limited (e.g. reserves of tellurium reach a mere 22
000 t, USGS 2009) and production capacities are constrained because all these elements are
primarily  mined  as  by-products  of  other  basic  metals  (copper,  zinc,  tin).  Corresponding
production capacities present therefore a very low elasticity and shortages can occur in case of
rising  demand  exceeding  the  production  volume  allowed  by  the  production  of  the  basic
metals, which is mainly determined by non-energy markets.
PV systems can also contain lead used in solders. Cadmium (in CdTe
†
PV) and lead are two
toxic to highly toxic substances whose usage is severely restricted, especially in electrical and
electronic equipment. Photovoltaics, however, have escaped regulation thus far.
2. Photovoltaics and the RoHS directive
The European Directive –referred to as the RoHS directive in the following–on the restriction
of  the  use  of  certain  hazardous  substances  in  electrical  and  electronic  equipment
‡
 (EEE)
forbids  that,  from  1  July  2006,  new electrical  and  electronic  equipment  put  on  the  market
contains  cadmium  and  lead –inter  alia.  However,  a  restricted  number  of  EEE  types  benefit
from exemptions. PV technologies, in particular, are so far excluded from the scope of the
RoHS directive. If it were to change, the immediate consequences would be that PV systems
containing cadmium (CdTe) and lead solders could no longer be imported or manufactured in
the European Union.
There are three possible options concerning the future treatment of PV technology under the
RoHS directive: i) after a clearly defined phase-out period, the same restriction on the use of
hazardous  substances  applies  to  PV  technology  as  to  other  EEE;  ii)  PV  technology  is
permanently excluded from the scope of the RoHS; iii) PV producers are allowed to benefit
from  “grace  periods”  that  are  periodically  (e.g.  every  four  years)  examined  and  possibly
renewed. Of these options, only the first two should be seriously considered. The third option
would  only  perpetuate  the  uncertainty  that  prevails  today  about  the  future  of  specific  PV
technologies  (CdTe  PV)  and  the  use  of  lead  in  solders.  In  a  sector  that  requires  long  term
commitments  both  on  the  production  and  consumption  sides,  this  would  send  the  wrong
signal to investors.
When considering the actual terms of the RoHS directive, and of the underlying precautionary
principle, the regulation of all PV technologies is the only option of the remaining two that
actually  makes  sense.  Strong  principles  command  the  RoHS  directive,  as  well  as  all EU
†
CdTe itself is not as hazardous as cadmium, but as a cadmium compound it is still
considered a hazardous substance.
‡
DIRECTIVE 2002/95/EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL
of 27 January 2003 on the restriction of the use of certain hazardous substances in electrical
and electronic equipment.
Wuppertal Institute 2010       4
directives
§
.  They  prescribe  the  following  order  of  priority  in  occupational  health  and
environmental protection issues: (a) the substitution of hazardous material should be the top
priority;  (b)  if  substitution  is  not  possible,  collective  protection  measures  should  be  put  in
place (e.g. operating below atmospheric pressure in production); (c) ultimately, measures of
individual protection apply(e.g. use of personal protection equipment).
The  arguments  of  the  proponents  of  the  CdTe  PV  systems  who,  of  course,  advocate  a
permanent exclusion of this technology from the RoHS directive may well be valid but they
disregard  the  raison  d’être  of  the  RoHS  directive.  Before  going  into  details,  it  is  worth
reminding these arguments here (Jäger-Waldau, 2009).
(1) First, proponents of the CdTe PV systems cite studies that have found that CdTe used in
PV  is  in  an  environmental  stable  form  that,  under  normal  use  conditions  and  in  case  of
foreseeable  accidents,  does  not leak  into  the  environment.  (2)  Second,  they  note  that LCA
studies have concluded that air emissions of cadmium from the whole life-cycle of CdTe PV
(including mining, smelting and purification) and the potential accidental emissions occurring
during  residential  fires  are  both  orders  of  magnitude  lower  than  cadmium  emitted  into  air
routinely from coal and oil power plants that PV displaces. (3) Third, they use the argument
that every PV technology has some environmental, health, and safety (EHS) issues, but that
the  commercial  viability  of  any  of  the  current  PV  technologies  should  not  be  restricted
because  of  these  issues.  (4)  Fourth,  proponents  of  CdTe  technology rely  on  studies  that
showed that current production of CdTe PV modules have shorter energy pay back times and
lower life cycle CO2emissions than other PV systems, e.g. crystalline silicon (c-Si) or CIGS.
They argue that a low production cost technology like CdTe PV could accelerate PV inroads
in the energy market and that a significant market penetration of any technology would help
the  whole  PV  industry  by  improving  the  installation  infrastructure  and  reducing  the
installation  cost  of  solar  electricity.  (5)  Fifth  and  finally,  CdTe  advocates  echo  the
announcement made by leading CdTe PV producers that they offer to take back end-of-life
modules and recycle them.
Even though the arguments in favour of CdTe PV reminded in the previous section are by and
large  correct,  they  cannot  invalidate  the  proposal  that  all  PV  technologies  ought  to  be
regulated under the RoHS directive, neither can these arguments prove that CdTe PV systems
need to be excluded from the RoHS directive. The directive saw the day in order to enable
reducing the content of hazardous substances (incl. cadmium and lead) in waste, and limiting
the presence of such substances in products and in production processes. Exemptions from the
directive requirements are only permitted if substitution is not possible from the scientific and
technical  point  of  view  or  if  the  negative  environmental  or  health  impacts  caused  by
substitution are likely to outweigh the human and environmental benefits of the substitution.
This alone clearly imposes that cadmium (in CdTe) should be allowed in PV systems if, and
only if, all alternative PV technologies –that do not use any of the substances banned by the
§
Such as the COUNCIL DIRECTIVE 98/24/EC of 7 April 1998 on the protection of the
health and safety of workers from the risks related to chemical agents at work (fourteenth
individual Directive within the meaning of Article 16(1) of Directive 89/391/EEC)
Wuppertal Institute 2010       5
directive–can not replace CdTe PV systems in a way that is satisfactory for human health and
the environment. Therefore, the indication that CdTe in PV is in a stable form under normal
use conditions and in case of foreseeable accidents (see (1) above) is not as such an argument
in favour of CdTe PV systems, as long as substitutive technologies exist, which is the case
(silicon-based PV). As long as CdTe PV systems are allowed, there will be cadmium used in
production  processes,  present  in  consumer  products,  and eventually  in  waste  streams,  all
things that the directive is designed to help avoid (despite (3) above).
Furthermore,  independent  testing  has  shown  that  CdTe  modules  exhibit  a  high  maximum
leaching  potential  of  both  cadmium  and  tellurium  (NGI  2009a,  NGI 2009b).  It  means  that
away from normal use conditions and benign accidents, leaching of cadmium can occur. Tests
have shown that when the CdTe film is exposed to water (for example the protective glass
layers are damaged) CdTe dissolves, thus increasing the risk of leaching into the environment.
While it is true that CdTe PV displacing conventional coal and oil power generation prevents
large amounts of cadmium from being emitted into the air (see (2) above), the same stands for
other  PV  technologies  that do  not  rely  on  CdTe,  such  as  Si-based  PV.  Recent  LCAs  (e.g.
Fthenakis et al. 2008) show that the differences in the indirect emissions of cadmium between
different PV technologies (based on the amount of energy needed for the production of the
PV  system)  are  very  small  in  comparison  to  the  emissions  from  conventional  energy
technologies  that  PV  could  displace. Therefore,  the  life  cycle  performance  regarding
cadmium  emissions  of  different  PV  technologies  are  not  that  far  apart  as  to  justify  an
exemption of CdTe PV from the RoHS directive.
To restrict the use of hazardous substances in EEE and to allow for strategies that stimulate
research  into  substitutes,  are  both stated  aims  of  the  RoHS  directive.  The  exclusion  of
photovoltaics  from  the  RoHS  directive  may have  contributed  to  the  fact  that  CdTe  PV
systems benefited from lower production costs compared to other cadmium-free alternative
technologies (see (4) above) to fuel their strong recent growth. Because cadmium is a wasteproduct  of  zinc,  production  does not  slow  down  even  if  demand  drops.  Consequently,
sufficient amounts of cadmium are available and affordable for the PV industry today, in part
because it was widely banned from electronic products by the RoHS directive and from other
applications like pigment in plastics or glass because of environmental and health concerns
(USGS 2010). Therefore, part of the cadmium that could no longer be used in EEE, plastics
and glass is now to be found in the production of solar electricity from photovoltaics.
There are, however, several PV technology alternatives, all with their particular strengths and
weaknesses,  and  which  all  are  continuously  improved,  requiring  intensive  targeted  R&D
activities. Further delaying permanent ban on cadmium-based PV technologies will send the
wrong  signal  to  producers  and  investors  and  deprive  other  technological  options  from  the
conditions needed to ramp up production capacities and decrease production costs. Moreover,
on  the  consumption  side,  the  stock  of  cadmium-containing  PV  will  grow,  aggravating  the
issue of hazardous waste disposal at the end of their lifetime.
There is to date virtually no experience in the management of end-of-life CdTe PV systems,
or of other PV technologies. PV systems installed today are expected to last for25+ years. By
Wuppertal Institute 2010       6
the time they will need to be disposed of, the company that produced the modules may no
longer exist. The last owner of the installation may very well be different from the first buyer,
ignorant of any take-back system put in place, or reluctant to bare the costs for dismantling its
end-of-life  PV  system.  To  prevent  hazardous  substances  from  ending  up  in  waste  streams
which  nobody  can  assure  that  they  will  be  properly  managed,  the  preferred  option  should
always be to refrain from using such substances in the first place.
Furthermore,  for  the  recycling  of  cadmium  and  lead  contained  in  photovoltaics  to  be
economic 25 years from now, demand will be needed for these metals at that time. However,
bans on cadmium and lead are clearly expected to reach ever further. Even if photovoltaics
were to be the last application where cadmium and lead are allowed, it will probably have
moved away from the CdTe technology by that time –e.g. organic PV may have taken over.
In the end, the problem will be that of toxic waste disposal –not recycling.
The  extension  of  the  RoHS  directive  to  PV  systems  will  not  only  influence  which
semiconductors can be used, it will also restrict the use of lead in soldersin such systems. The
same reasoning as for cadmium apply. Furthermore, the development of lead-free solders for
other  electronic  products  and  components  has  been  demonstrated  and  is  now  established.
There is no principle argument why it should not work with PV.
3. Effects of an extension of the RoHS directive to photovoltaics
3.1. Effects abroad and on international trade
The  RoHS  directive  has  a  strong  influence  outside  of  the  European  Union.  When  the  EU
decides to restrict the use of certain materials in electronic and electrical equipment, thiswill
lead to a worldwide phase-out on these materials in EEE, especially in products exported to
the  EU.  PV  are  exported  from  and  imported  to  the  EU.  The  RoHS  directive  has  value  of
example and has a massive influence on production world-wide.
Furthermore, coverage of PV by the RoHSdirective will not only impose a ban on hazardous
substances in semiconductors for PV but also on lead solders. Both aspects are important for
the production, the use, and –especially–the end-of-life of PV in developing countries.Even
though it might bethat some PV producers have a recycling system, cadmium and lead will
be widely banned when PV produced today reach the end of their lifetime (in 25 to 30 years).
Then, there will be no need for recycling CdTe and lead, but for a safe waste treatment.
The experience  of  electronic  wastes  shows  clearly  that  recycling  concepts  developed  in
industrial countries are not sufficient in a global context. A certain amount of electronic waste
will  be  handled  in  developing  countries  under  inappropriate  conditions,  far  away  from  any
kind of safe working conditions. This is highly relevant because especially cheap PV can be
an  option  for  less  developed  countries  and  regions.  It  can  make  a  basic  electrification
possible, induce  development  in  rural  regions,  and,  to  a  certain  degree,  reduce   rural
depopulation. However, considering typical circumstances in such areas, it seems unrealistic
Wuppertal Institute 2010       7
to  expect  collection  and  recycling  schemes  for  used  PV to  be as  reliable  as  in  Europe.
Therefore, strict European directives forcing thesubstitution of hazardous substances not only
influence Europe but can also ensure additional protection to other regions.
3.2. Effects on competitiveness and employment
Growing and reliable solar markets around the world and especially in Europe have provided
PV  manufacturers  the  opportunity  to  scale  and  reduce  costs  (Gillette  2010).  On  that  basis,
producers of CdTe PV deployed a smart and aggressive expansion of production capacities
(incl. easily duplicable, automated large scale production centres located in countries where
costs of labour are low). This management, rather than a better technology per se, explains the
competitive  price  point  of CdTe  technology  against  silicon-based  alternatives  (Beyer  et  al.
2009). Therefore, a ban on cadmium and lead wouldnot hinder the photovoltaic industry to
continue improving its competitiveness.
The  market  share  of  thin  film  PV  technologies  has  almost  tripled  globally  in  the  past  five
years  (from  5.9%  in  2004  to  16.7%  in  2009).  The  exponential  growth  of  CdTe  PV  was
instrumental in this trend. In 2009, CdTe photovoltaics accounted for 9% of the global PV
market,  for  only  1.1%  in  2004.  Both  alternative  thin  film  technologies  (amorphous  Si  and
CIGS) also captured new market shares, although at a much slower rate. The drastic increase
in the number of CdTe PV installed in the past years means a quick growth of the stock of
cadmium in EEE across the world. This is a concern for the coming 25 to 30 years, but even
more for the years after, when those modules reach the end of their lifetime.
Figure 1: Global market shares of the different thin film PV technologies. CdTe, a-Si/µ-Si,
and CIGS stand for cadmium-telluride, amorphous/ microcrystallinesilicon,and copperindium-gallium-selenium, respectively.Data source: Photon (4-2008), Photon international
(3-2009, 4-2009)
Wuppertal Institute 2010       8
The cost per installed kWp was nearly halved during the last 4 years. With sustained public
incentives,  the  installed  cost  of  photovoltaics  is  on  a  pathway  toward  grid  parity.  High
irradiance solar projects will reach it first. A ban on CdTe photovoltaics will not prevent this
from happening, even though it may delay it.
Figure 2: Cost of PV per installed kWp. Data source:BSW (2010)
For 2009 the employment figures in photovoltaics for the European Union were estimated in
the range of 85,000 to 90,000 (Jäger-Waldau 2009).Of this number, less than 1% correspond
to people  employed  in  CdTe  module  production  in  Europe  (Beyer  et  al.  2009).  About  two
thirds of the employees in the photovoltaic sector in Europe work in the installation of solar
modules.  These  jobs  are  bound  to  demand  in  Europe,  which  is  not  technology  related  but
depends  on  public  incentives.  Hence,  an  extension  of  the  RoHS  directive  will  have  only
negligible effectson European labour market.
3.3. Effects on environmental and climate targets
Analyses  over  the  complete  life-cycle  of  photovoltaics  have  shown  that emissions  of
greenhouse  gases,  air pollutants (SOx,  NOx), and heavy  metals  are  insignificant  in
comparison to the emissions that they replace when introduced in average European and U.S.
grids(Fthenakis et al. 2008). This is true regardless of the PV technology selected.
Among  PV  technology  options,  thin-films  require  lower  material  and  energy  inputs  in  the
production process. Consequently, CdTe PV induce substantially less emissions (GHG,SOx,
NOx, heavy metals) life-cycle wide per KWh produced than non thin-film silicon-based PV
(Fthenakis  et  al.  2008). However,  it  is  not  clear  how  CdTe  PV  performs  in  those  terms
compared to thin-film silicon-based technologies. In any case, the gap must be smaller than
with non thin-film alternatives.
Photovoltaics will continueto play an important role in the rise of renewable energies and in
Wuppertal Institute 2010       9
climate change mitigation. Therefore, supply of PV systems needs to be able to meet demand
even  if  the  scope  of  the  RoHS  directive  is  extended. The  global  market  for  photovoltaics
shows today an oversupply that is larger than the share of CdTe technology in that market
(Beyer et al. 2009). Current competitors of CdTe technology have the capacity to buffer in the
coming  years  any  drop  in  supply  due  to  a  ban  on  cadmium.  Future  new  entrants  with
technologies in R&D today will also increase the offer in the coming decade.
Expected levels of demand will be met with or without CdTe PV. In general, PV technology
will  also  continue  to  improve  towards  lower  per  kWh  impact  than  the  existing  options
(Raugei  and  Frankl  2009).  Climate  targets  that  for  a  part  rely  on  the  development  of
photovoltaics will not be adversely affected by the extension of the RoHS. Overall sustainable
development goals will be positively affected.
4. Conclusions
The conclusion of this position paperis that no matter how low potential environmental and
health  impacts  may  be  under  normal  operating  conditions  of  CdTe  PV,  it  is  not  a  valid
argument against the extension of the RoHS directive to photovoltaics. Cadmium and lead in
market  products  should  be  substituted  when  substitutes  exist,  which  is  the  case  for
photovoltaics.  It  is  the essence  of  the  RoHS  directive.  Recycling  is  not  a  realistic  option
because  it  is  only  a  question  of  time  until  cadmium  modules  and  lead  solders  are  widely
banned  and  it  is  quite  open  which  kind  of  photovoltaics  we  will  use  in  25  or  30  years.
Cadmium and lead should not spread in EEE, but need to be disposed of safely.
CdTe solar cell technology is only one of many PV technologies. A ban on CdTe PV will not
end the development of photovoltaics. The main goal of the application of PV is the reduction
of greenhouse gasemissionsfrom electricity production as an important part of sustainable
development.  But  today’s  discussion  on  climate  change  inadequately  limits  the  focus.  The
best way towards environmental protection and sustainable development is not limited to the
cheapest possible PV.
Producers  of  CdTe  PV  and  users  of  lead-solders  have  preferred  collective  and  individual
protection mechanisms to substitution. This is against the widely accepted order of priority
for  protection  measures.  It  is  inherently  safer  to  prevent  the  use  of  a  hazardous  material,
which  reflects  the  rationale  of  the  RoHS  directive.  If  a  hazardous  material  is  used,  even  if
there is an established recycling concept, there is still a risk of losses during the use phase and
recycling processes, and in countries receiving EU exports the establishment of a functioning
PV  recycling  may  take  still  some  decades.  The  extension  of  the  RoHS  will  ensure  that
existing  cadmium-free  solar  cells  and  lead-free  solders  alternatives  are  used  and  further
developed in photovoltaics.
Wuppertal Institute 2010       10
5. References
Amy C. Tolcin (2010)Cadmium [advanced release], 2008 Minerals Yearbook, U.S.
Geological Survey.
Beyer, M., L. Gataullina and F. Pavel (2009) The effect of the EU RoHS directive on the
European photovoltaic industry. DIW econ Policy Paper 2/2009.
Bundesverband Solarwirtschaft e.V. Statistischen Zahlen der deutschen Solarstombranche
(Photovolataik), April 2010.
Fthenakis, V et al. (2008) Emissions from Photovoltaic Life Cycles. Environmental Science &
technology, 42, 2168–2174.
Gillette, R. (2010) Statement of Robert Gillette, Chief Executive Officer, First Solar before
the U.S. Senate Committee on Environment and Public Works and Subcommittee on
Green Jobs and the New Economy. TEMPE, Ariz., Jan. 28, 2010. Available online from
[accessed 9 Apr. 2010]:
http://www.firstsolar.com/en/news/news_SenateCommitteeEPW012009.php
Ullal, H.S. and B. von Roedern (2007) Thin-Film CIGS and CdTe Photovoltaic Technologies:
Commercialization, Critical Issues, and Applications. Conference paper, 22nd
European Photovoltaic Solar Energy Conference (PVSEC) and Exhibition, Milan, Italy.
September 3–7, 2007.
Jäger-Waldau, A. (200?) Peer Review of Major Published Studies on the Environmental
Profile of Cadmium Telluride (CdTe) Photovoltaic (PV)Systems. European
Commission, DG JRC, Institute for Environment and Sustainability, Renewable
Energies Unit.
Jäger-Waldau, A. (2009) Snapshot on European photovoltaics in world-wide comparison.
European Commission, DG Joint Research Centre, Institute for Energy, Renewable
Energy Unit.
Kautto, N. and A. Jäger-Waldau (2009) Renewable Energy Snapshots 2009. European
Commission, DG Joint Research Centre, Institute for Energy, Renewable Energy Unit.
Norwegian Geotechnical Institute (2010a) Environmental risks regarding the use and final
disposal od CdTe PV modules. Document No. 20092155-00-5-R.
Norwegian Geotechnical Institute (2010b) Leaching from CdTe PV module material –results
from batch, column and availability tests. Document No. 20092155-00-6-R.
Photon -Das Solarstrom Magazin, April 2010.
Photon international –The solar power magazine, March and April 2009.
Raugei, M. and P. Frankl (2009) Life cycle impacts and costs of photovoltaic systems:
Current state of the art and future outlooks. Energy34 (2009)392–399.