EPA/600/R-11/107
                                          September 2011
Guidance to Facilitate Decisions for
    Sustainable Nanotechnology
                   Authors
                 Tarsha Eason
                David E. Meyer
               Mary Ann Curran
        Venkata K.K. Upadhyayula (ORISE)
            Systems Analysis Branch
         Sustainable Technology Division
   National Risk Management Research Laboratory
             Cincinnati, Ohio 45268
   National Risk Management Research Laboratory
        Office of Research and Development
       U.S. Environmental Protection Agency
             Cincinnati, Ohio 45268

-------
The U.S. Environmental Protection Agency through its  Office of Research and Development
funded,  managed, and collaborated in the research described  here under EP-W08-010 to Abt
Associates, Inc.   It has been subject to the Agency's review  and has been  approved for
publication as an EPA document.

-------
The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the
Nation's land, air, and water resources. Under a mandate of national environmental laws, the
Agency strives to formulate and  implement actions leading to a compatible balance between
human activities  and  the ability of natural systems to support and nurture  life. To meet this
mandate,  EPA's  research  program is providing  data and  technical  support  for  solving
environmental problems today and building a science knowledge base necessary to manage our
ecological resources wisely, understand how pollutants affect our health, and prevent or reduce
environmental risks in the future.

The National  Risk Management  Research Laboratory (NRMRL)  is the  Agency's  center for
investigation of technological  and management approaches  for preventing and reducing risks
from pollution that threaten human health and the environment. The focus of the  Laboratory's
research  program is on methods  and their cost-effectiveness  for  prevention and  control of
pollution to air, land, water, and subsurface resources; protection of water quality in public water
systems; remediation of contaminated sites, sediments and ground water; prevention and control
of indoor air pollution; and restoration of ecosystems. NRMRL collaborates with both public and
private sector partners to foster technologies that reduce the cost of compliance and to anticipate
emerging  problems.  NRMRL's research  provides  solutions to environmental problems by:
developing and promoting  technologies that protect and improve the environment; advancing
scientific and engineering information to support regulatory and policy decisions; and providing
the technical  support  and information transfer to ensure  implementation  of environmental
regulations and strategies at the national, state, and community levels.

This publication has been produced as part of the Laboratory's strategic long-term research plan.
It is published and made available by EPA's Office of Research and Development to assist the
user community and to link researchers with their clients.
                                Sally Gutierrez, Director
                    National Risk Management Research Laboratory

-------
Products that  incorporate materials manufactured at the nano  scale (i.e., nanoproducts) offer
many potential benefits to society;  however, these benefits  must be weighed against potential
"costs" to the  environment and public health. This document was developed to provide a broad
guidance for assessing the sustainability of nanoproducts and is intended to lay the groundwork
for developing a decision-support framework through continual updates as research in this area
progresses. At the very least, it will aid stakeholders when navigating the various choices that
must  be  made  to foster  the  development of  sustainable nanotechnology. Given the  all-
encompassing nature of sustainability, this work should be of interest to stakeholders in all areas
of nanotechnology, including research, product development, consumer use, and regulation. The
aim of this work is not to make decisions for stakeholders, but to help frame the pertinent issues
that must be addressed to properly assess emerging nanotechnologies and to provide information
on the various tools that may be used to  address them. The foundation of this approach is to
consider existing standards and methods for environmental, economic, and  social assessments
using a life cycle perspective and offer guidance  by relaying first-hand knowledge of applying
assessment  tools  to  nanotechnologies, whenever possible. Brief overviews of the  various
assessment  methodologies are  provided to help  stakeholders make informed  choices when
selecting tools appropriate for their goals. For specific details  of a method, readers are directed to
the referenced standards and guidance documents supporting the application  of these tools. The
key steps  to be included in the evolving framework include: characterizing  a nanoproduct and
identifying potential risks and impacts; identifying relevant stakeholders; defining the goal and
scope  of  an assessment; assessing environmental, economic,  and social impacts; evaluating
sustainability criteria; developing and evaluating alternatives; and selecting and implementing a
decision to support sustainability. Given that the field of nanotechnology is relatively new, there
are significant uncertainties regarding its potential human health and ecological risks and impacts.
Hence, methods developed to address these uncertainties are also explored. Moreover, since the
field of nanotechnology is changing rapidly, this document will be reviewed and updated as
additional  information  becomes  available  to  continue working  towards   the  end goal  of
sustainability.
                                                                                       IV

-------
Notice   	ii
Foreword	iii
Abstract 	iv
Figures and Tables	vi
1. Managing Sustainable Nanotechnology	7
  1.1   Nanotechnology Overview	7
  1.2   Towards Sustainability	8
  1.3   Document Overview	10
  1.4   Key Benefits to the Intended Audience	14
2. Initializing the Path Forward to Sustainability	15
  2.1   Initial Product Characterization and Identification of Potential Risks	16
  2.2   Stakeholder Identification	19
  2.3   Goal and  Scope Definition	20
3. Assessing Environmental, Economic, and Social Impacts	25
  3.1   Environmental Assessment Methods	25
  3.2   Economic Assessment Methods	39
  3.3   Social Assessment Methods	46
4. Assessing Sustainability	51
  4.1   Evaluating Sustainability Criteria	51
  4.2   Decision Theory	54
  4.3   Selecting the Most Sustainable Alternative	58
  4.4   Uncertainty Analysis	63
  4.5   Sensitivity Analysis and Scenario Analysis	63
  4.6   Further Analysis	64
5. Conclusions	65
6. Acknowledgements	65
7. References	66
Appendix: List of Additional Resources	75

-------
Figure 1-1. A Holistic View of Sustainability	8
Figure 1 -2. Overview of a Preliminary Framework for Sustainable Nanotechnology	13
Figure 2-1. Identifying Risks Early in the Process	17
Figure 2-2. Example of Stakeholder Mapping	19
Figure 2-3. Sample Goal and Scope definition: The life cycle of a pair of cotton socks containing
     antimicrobial silver	24
Figure 3-1. General Framework for a Product Life-Cycle Assessment (LCA)	29
Figure 3-2. Nano Risk Framework	31
Figure 3-3. Life-Cycle Costing (as presented in SETAC)	42
Figure 3-4. Social Life-Cycle Assessment	47
Figure 4-1. Mapping the Decision Criteria: Sample Criteria for Assessing Single-Walled Carbon
     Nanotube Synthesis Processes	52
Figure 4-2. Decision Analysis Flowchart	55
Figure 4-3. Modified Decision Analysis (DA) Approach	56
Figure 4-4. Comparison of Single Walled Carbon Nanotubes (SWCNT) Criteria Weightings.... 63

Table 2-1. Traditional Product Development and Sustainable Design proceed	16
     in tandem and entail overlapping steps	16
Table 2-2. Identifying Risks Early in Development	 18
Table 2-3. Examples of Sustainability Criteria	21
Table 3-1. Key Environmental Assessment Methods	26
Table 3-2. Key Economic Assessment Methods	43
Table 3-3. Example Factors Impacting Incumbent Firms	45
Table 3-4. Key Social Assessment Methods	48
Table 4-1. Sample Decision Making Tools	53
Table 4-2. Comparison of Process Elements for Common Decision Making Tools	57
Table 4-3. Summary of Common Multi-Criteria Decision Analysis (MCDA) Methods	59
                                                                                   VI

-------
1. Managing Sustainable Nanotechnology
1.1 Nanotechnology Overview
  The U.S. National Nanotechnology Initiative (NNI) defines a technology as nanotechnology
  only if it involves all of the following:
  • Control or manipulation of matter at dimensions in the 1 to 100-nanometer range.
  • Creation  of structures, devices and systems that have unique properties and functions at
   the nanoscale leading to novel applications.
Nanotechnology relates to the ability to create materials and devices through the manipulation of
individual atoms and molecules (up to  100 nanometers). Further, it involves integrating these
structures  into  larger systems   (Bhushan,   2007).   Similar  to  information  technology,
nanocomponents (the nanoscale building blocks of nanotechnologies) exhibit a diverse array of
characteristics that may be used for a wide range of beneficial applications and have the potential
to generate significant improvements to existing technologies (Palmberg  et al., 2009). These
applications include  medical,  food,  clothing,  defense,
national   security,  environmental   clean-up,   energy
generation,  computing, construction,  and electronics
(Davies,   2009;   EPA-SPC,   2007).    Consequently,
according to a recent publication from the Organization
for Economic Co-operation and Development (OECD),
the  field  of nanotechnology is  rapidly  expanding
(Wiesner et al., 2006) and is anticipated to emerge as a
key engine of growth for the 21st century (Palmberg et
al., 2009).

Throughout this framework, the diverse  products and materials that are made using (engineered)
nanocomponents will be collectively referred to as nanoproducts.
Nanotechnology  is   being  touted  as  a   profound
technological  advancement  capable  of  transforming
society as we know it. Accordingly, the investment in
nanotechnology  initiatives  have  skyrocketed  (Roco,
2009). However, the  potential for explosive  growth in
nanoproduct  markets  should be viewed with cautious
optimism. Without question, knowledge of the physical
world at  the nanoscale opens a world of possibilities to
enable the development of products and  systems  with
great precision and intricate properties that could improve
our overall  quality of life. Systematic control  of the
production of products at the molecular level could enable
the development  of  eco-friendlier  products  and  services.  The  projected  trillion-dollar
nanotechnology market is  expected to provide new markets, create jobs, and greatly increase
profits for businesses.  However, caution must be applied to temper these expectations when
considering  the growing   concerns  regarding  the unforeseen  impacts  of nanotechnology
deployment. For example, with many nanotechnologies requiring energy intensive processes and
rare materials, it is unclear how large-scale deployment will affect the environment. In addition,
the cost needed to  retrofit  production and incorporate nanotechnology may be more than some
"IfI were asked for an area of
science and engineering that will
most likely produce the
breakthroughs of tomorrow, I
would point to nanoscale science
and engineering."

                      Neal Lane
             Former NSF Director
       Assistant to President Clinton
        for Science and Technology

-------
existing companies can afford. This could lead to a radical shift in the make-up of the various
application markets,  resulting in the emergence of corporate monopolies and destabilization of
the global economy. In addition, the more advanced manufacturing  processes will require
workers with greater knowledge and expertise, thereby putting pressure on the existing workforce
to adapt mentally  or risk unemployment, a problem that can quickly alter societal dynamics.
Perhaps the greatest concern with nanotechnology is the potential threat that nanocomponents
pose to human health and ecosystems. NSF's increasing investment in  the safety and societal
implications of nanotechnology (Roco, 2009) stems from recognition that the unique properties
that define a nanotechnology and make it novel are the same properties that could eventually pose
the greatest health risks. These various concerns transcend the science of nanotechnology and
raise the issue of its sustainability.


1.2 Towards Sustainability

 The National Environmental Policy
Act  of 1969,  a precursor  to  the
establishment of the  Environmental
Protection Agency,  formalized  a
growing  understanding  of  the
importance   of  the  relationship
between    humans     and    the
environment. Further its language,

"... to declare a national policy           Figure 1-1. A Holistic View of Sustainability
which     will      encourage
productive and enjoyable harmony between man and his environment; to promote efforts which
will prevent or eliminate damage to the environment and biosphere and stimulate the health and
welfare of man....", foreshadows ideals soon to be of great importance on a global stage. Nearly
two decades later, the World Commission on Environment and Development (WCED) coined the
term  sustainable development as "development that meets the needs of the present without
compromising the  ability of future generations to meet their own needs" (ONGO, 1987).  Thus,
sustainability may be viewed as using resources and developing products  and processes in a way
that  ensures  and promotes a legacy  of economic  viability,  social equity and environmental
responsibility for current and future generations.

Many researchers recognize that traditional growth and development practices are contradictory
to shifting to a sustainability paradigm (2003). While there is great interest in  this area, some
believe  that economic  progress and sustainability are mutually exclusive (Davidson and Julie,
2001). This idea is birthed from the concept that sustainable  development inevitably translates
into benefiting one facet to the  detriment of another. However, in truth,  it is about balance and
optimization and entails evaluating the intricate elements of sustainability (Berkel, 2000; Eason et
al., 2009). Built on a foundation of economic, social and environmental indicators, sustainability-
based decision making is a highly complex challenge and is predicated on the ability to  reconcile
both disparate and integrated aspects of the product, process,  or system under study. Given the
magnitude of its anticipated impact, nanotechnology should therefore be produced and utilized in
a manner that  is environmentally, economically, and  socially sustainable to fully realize its
potential (Helland et  al., 2007; Klopffer, 2008).

Years after the WCED definition was accepted,  researchers have continued to struggle with
reaching a consensus on how sustainability should be measured. However, one key development
and widely accepted  convention was to establish what is termed the three pillars of sustainability:

-------
environment, economy, and society (Figure 1-1). Each of these pillars denotes particular aspects
of the product, process, or system that may be assessed via observable and measurable criteria
(i.e., metrics). While there are tools available that may be used to evaluate characteristics within a
particular pillar, the difficulty lies in making decisions based  on information gathered from
various tools and disparate criteria.

The challenge of sustainability begins with an understanding of the holistic nature of the world in
                                                  which  we  live.   Traditional  Newtonian
                                                  thinking  views  the  world  as  isolated
                                                  subsystems and seeks solutions to problems
                                                  within  each system with  little regard  to
                                                  how  the  various  systems  interact.   In
                                                  contrast, holistic  thinking  examines how
                                                  changes within a subsystem can affect the
                                                  system as  a  whole  through  interactions
                                                  across subsystem boundaries,  recognizing
                                                  that the  world is an integrated sum of its
                                                  parts.  This  holistic approach has inspired
                                                  the field of industrial  ecology, a systems-
                                                  based,   multidisciplinary   approach   to
                                                  understand  the   emergent  behavior   of
                                                  complex integrated human/natural systems.
                                                  With nanotechnology,  it  is important  to
                                                  understand how the various world systems
                                                  (i.e.,  ecosystems,  societies, etc.) will be
                                                  impacted by its existence throughout its full
                                                  life span (i.e., from cradle to grave). The
                                                  life cycle concept identifies five key stages
                                                  where impact may occur, including:

       Raw materials extraction: Activities related to the acquisition  of natural resources,
       including mining non-renewable  material, harvesting biomass, and transporting raw
       materials to processing facilities.
       Materials  processing:   Processing  of  natural  resources by reaction,  separation,
       purification, and  alteration steps  in preparation  for  the manufacturing stage;  and
       transporting processed materials to product manufacturing facilities.
       Product manufacture: Manufacture of product and transport to the consumers.
       Product  use:  Use and  maintenance  activities associated with the  product  by the
       consumer.

       End-of-life disposition: Disposition of the product after its life span, which may include
       transportation,  recycling, disposal, or incineration.
             Systems Thinking

In his landmark book,  The Fifth Discipline,
Peter Senge wrote, "From a very early age, we
are taught to break apart problems, to segment
the world. This apparently makes complex tasks
and  subjects more manageable, but we pay a
hidden  price.  We  can  no  longer  see  the
consequences  of  our  actions; we  lose  our
intrinsic sense of connection to  a larger whole."
For Senge, the  answer lay in systems thinking.
Applying   systems  thinking  helps  creative
individuals    to   see    wholes,   perceive
relationships, uncover connections, expose  root
causes and  master complexity. Senge  argued
that systems thinking  integrates  what might
otherwise be considered separate  management
disciplines,  preventing them from  becoming
"gimmicks  or  the  latest organisation  change
fads." (Smith, 2005)

-------
                            The Daly Rules for Sustainability

 Working from theory initially developed by Romanian economist Nicholas Georgescu-Roegen,
 Herman E. Daly (University of Maryland School of Public Policy professor and former Chief
 Economist for the World Bank), laid out in his 1971 opus "The Entropy Law and the Economic
 Process") suggests the following three operational rules for defining the condition of ecological
 (thermodynamic) Sustainability:

    1. Renewable resources such as fish, soil, and groundwater must be used no faster than
      the rate at which they regenerate.
    2. Nonrenewable resources such as minerals and fossil fuels must be used no faster than
      renewable substitutes for them can be put into place.
    3. Pollution and wastes must be emitted no faster than natural systems can absorb them,
      recycle them, or render them harmless.
Accordingly, each stage during product development should be  optimized  to  minimize the
impacts within the three pillars. However, it is important to remember that changes to one stage
can  lead  to impacts in another  stage. Thus,  the  holistic  approach is needed to achieve
Sustainability. True Sustainability, as described  by Daly (text box  above), may never be fully
achieved during a product's life cycle in the  short term.  Subsequently, the goal  of sustainable
development would be to help identify the most desirable option based on available technologies
and current industrial practices. However, it is beneficial for decision makers to remain open and
continually  explore alternative  solutions for  a product function  by periodically revisiting the
design and manufacture of a product. Potential  product alternatives can provide a comparative
basis to better understand assessment results.  In some cases, the alternative may be an existing
product or process and in others, after the initial design, it may be  worthwhile to generate
alternatives encompassing other available technologies.

1.3
This document presents guidance for integrating Sustainability into  the evaluation, management
and development of nanoproducts. The material is presented while keeping in mind the long term
goal for this work to eventually be refined into  a decision support  framework. It is intended to
support Goal  4  of the  NNI's  Strategic  Plan  "to  support the responsible development of
nanotechnology"(NNI, 2011).  Efforts  in  this  area  are  coordinated by the Nanotechnology
Environmental  and Health  Implications (NEHI) working  group  of the  Nanoscale  Science,
Engineering, and Technology (NSET) Subcommittee  that maintains the Environmental, Health,
and Safety (EHS) research strategy. Since the original EHS research  strategy was written in 2008,
the NEHI has collaborated with stakeholders from government agencies, industry, academia, non-
governmental organizations (NGOs) and the general public to determine how best to update and
improve  EHS research (NNI,  2010).  One of  the primary results of these actions  was the
recognition of the stakeholders to incorporate life cycle considerations for Sustainability into risk
management for nanocomponents and nanoproducts.  Therefore, the preliminary  framework of
this guidance represents a reasonable approach to Sustainability assessment rooted in a life cycle
perspective.

Since anyone involved with the life cycle of  a product may constitute a stakeholder (Freeman,
1984; Mitchell et al., 1997), the  guidance that  follows  is intended to  provide researchers, product
developers, NGOs, policy makers, and consumers with an understanding of the various choices
                                                                                      10

-------
that must be made to ensure the development and sustainable use of nanoproducts. While some
may argue that control of product sustainability is ultimately in the hands of researchers and
product developers,  it is important that stakeholders throughout the product life cycle have a
common understanding of the benefits and challenges associated with sustainability in order to
better work together to achieve it. Further, this guidance helps stakeholders understand the tools
that may be used to assess aspects of sustainability and identifies current approaches  in  the
literature to attempt to integrate data from these disparate evaluations to make quality decisions.
At present, integration of assessment tools is a key challenge for  sustainability assessment in
general with no  clear consensus  choice of method  available. As an  additional  benefit to
stakeholders, discussions of potential pitfalls and  areas  of concern that may  arise  during
assessment are included. Moreover, whenever possible, the authors relay first-hand knowledge of
the application of specific assessment tools to nanoproducts.

The guidance offered in this document may be applied to both new and existing technologies.  An
existing  nanoproduct  can be  a complete  product  redesign,  a modification of an  existing
nanoproduct, or  the incorporation of nanotechnology into  a  traditional consumer  product.
However, the large number of unknowns regarding the use and disposal of nanomaterials  makes
it more challenging  to  successfully address the critical sustainability concerns within the life
cycle of a new nanoproduct or concept prior to full-scale implementation. Therefore, researchers
and product developers must integrate sustainability considerations into subsequent modifications
to correct  for unforeseen  issues as the  nanoproduct evolves and develops.  For the sake of
discussion, guidance is presented for the case  of new technology  development  with  the
understanding that any issues that may arise when dealing with product redesign can be addressed
using the same principles.

Before presenting the preliminary framework that will guide the discussions in this document, it
is important to first recognize a key contribution to life cycle based sustainability assessment in
the open literature to provide readers with an understanding of why this work has chosen to adopt
a different approach  to accomplish decision making for sustainable nanotechnologies. Kloepffler
(2008) has proposed Life  Cycle Sustainability Assessment (LCSA) as a tool to  evaluate  the
sustainability of products. The tool integrates the  existing methods of Life Cycle Assessment
(environment), Life  Cycle Costing (economy)  and Social  Life  Cycle Assessment (society) to
perform  a sustainability  assessment using the existing  indicators of  each method for  the
respective pillars. While this approach is a viable option, it presupposes the ability of these three
tools to account for all potential implications of a product. As will be presented in this document,
there are many challenges associated with assessing nanotechnology that may fall outside of the
scope of these three tools. Therefore, this work will focus on developing a framework that is more
readily adaptable for nanoproducts by allowing  for a customizable selection of indicators, and
therefore tools to be made by those with first-hand knowledge of the product/system to be
studied. This should lead to more relevant decisions regarding the  development of sustainable
nanotechnologie s.

Figure 1-2 demonstrates how the preliminary framework presented in this guidance document can
be deployed during product development. Experts can include risk assessors, economists,  public
health  experts, or other scientists  that may offer important advice in assessing  impacts and
selecting approaches for mitigating these impacts.  The  people who will guide  the  overall
application of the framework are a group of decision-makers typically comprised of experts and
stakeholders. The key steps of the framework include:

-------
        1.  Characterize product and identify potential health and environmental (toxicity)
           risks.   At the outset  of developing the  initial idea or product concept, product
           developers should conduct  an initial characterization of the potential health  and
           environmental (toxicity) risks of the chemicals contained in the product. This would
           allow the team to identify and mitigate potential chemical hazards early on in the
           product design stage.
        2.  Identify stakeholders.  Stakeholders should be identified early and engaged at the
           outset of sustainability assessments, as they provide important input in defining the
           goal and scope of the study, as well as helping to develop appropriate sustainability
           metrics, evaluating and interpreting impacts, and selecting alternative approaches to
           mitigate impacts.
        3.  Define  assessment goal  and  scope.   Establish the goals and  objectives of the
           analysis and determine the appropriate methods and models to meet the assessment
           objectives. This step  involves  translating the broad concept of sustainability into
           concrete, measurable goals, which can be assessed in step 4.
        4.  Assess environmental, economic, and social impacts: Apply methods for assessing
           environmental, economic, and social impacts across the life cycle of the nanoproduct.
           These impacts include, but are not limited to: energy and material use, costs to the
           manufacturer, greenhouse gas emissions, and wages. Uncertainty should be assessed
           and a sensitivity analysis may be conducted as part of this  step.
        5.  Evaluate sustainability criteria.  Normalizing the inventory of impact results into
           pertinent categories related  to the sustainability criteria and eliciting stakeholder
           valuations of criteria to interpret and compare results.
        6.  Develop  alternative  approaches to  mitigate impacts.   The  results of  the
           assessments and evaluations  can help developers identify improvements to  the
           product system that will mitigate  impacts  (e.g., modification to the  manufacturing
           process or use of alternative nanomaterial in an upstream process).
        7.  Assess  the environmental, economic, and social  impacts  of the alternative
           approaches.  Each alternative approach developed should be assessed to determine
           the impacts for each  pillar. The  results  will then be  used to compare different
           alternatives. As  part of this  step, additional alternatives may need to be developed
           and  assessed if the initial designs are  found to be inadequate  for  improving the
           sustainability of the product.
        8.  Select most sustainable  alternative. Once the environmental, social, and economic
           impacts for each alternative are assessed, decision makers work together to select the
           best alternative approach for mitigating impacts and achieving a sustainable product
           design.
        9.  Implement production of  sustainable nanoproduct.  Following selection of the
           preferred alternative, the product should be manufactured and continually monitored
           for further improvements using the sustainability framework.
Although these steps are discussed with  regard to a new product development, we re-emphasize
that the framework may also be applied to products in any development stage, including those
that are already commercialized. As noted above, achieving sustainability is an iterative process
in that products should  be  continually assessed and updated  as  new information  becomes
available. The guidance in this document will be given through detailed discussion of each step of
                                                                                      12

-------
                    Sustainable Nanoproduct Concept
                          Characterize Product and
                           Identify Potential Risks
                                Section 2.1
                           Identify Stakeholders
                                Section 2.2
       Environmental
        Assessment
         Section 3.1
                            Define Assessment
                              Goal and Scope
                                Section 2.3
 Economic
Assessment
 Section 3.2
                           Evaluate Sustainability
                                 Criteria
                                Section 4.1
                          Select Most Sustainable
                               Alternative
                                Section 4.2
                           Implement Sustainable
                               Nanoproduct
   Social
Assessment
 Section 3.3
                                                                      Develop
                                                                    Alternatives
                            Redevelop product
                                concept as
                            technology improves
     Figure 1-2. Overview of a Preliminary Framework for Sustainable Nanotechnology
the  preliminary framework and  is  presented in  Sections  2-4.  Section  5  presents overall
conclusions of this work and how they can be applied to refine the proposed framework. Finally,
the  Appendix  includes  a list  of  additional  resources that  may be useful  in conducting
sustainability assessments of nanoproducts.
                                                                                            13

-------
1.4               to the
As  previously  described, this  document is  anticipated to primarily  assist nanotechnology
researchers,   developers,  engineers,  and  product  manufacturers  in  addressing   potential
environmental, economic, and social implications throughout the life cycle stages of a new or
developing nanoproduct.  It can  also be useful to other decision-makers,  including government
agencies and NGOs  by providing a better understanding of the pertinent issues that must be
addressed to properly assess the sustainability of a nanotechnology.
Industry and government stakeholders can realize many benefits from conducting and supporting
sustainability assessments of nanoproducts. A sustainability assessment can contribute to research
that will aid current efforts to promote health and safety when manufacturing nanoproducts as
well as in their public use or consumption. This will not only help minimize risk, but also reduce
unsubstantiated or  nonscientific  claims  of risks  or  benefits of  nanoproducts. Further,  if
nanoproduct developers can demonstrate  that they  are  serious about developing a sustainable
product that presents minimal long-term risks to human health and the environment, then their
products may be viewed more favorably. Such an effort will help avoid an inaccurate perception
of risks among the general public similar to those that arose for genetically modified organisms
(Bell, 2007). Upon completion, this  process will  aid in  establishing a protocol toward  the
development, management and assessment of sustainable nanoproducts. Additionally, it will
provide life cycle data  (material and energy flows) that may be used as a benchmark for future
sustainability assessments of nanoproducts and technologies,  to measure  improvements, and
evaluate the impacts of possible design changes.
Other benefits include:
    •    Placing human health risk assessment in perspective with other environmental  concerns
        across the life cycle of nanoproducts.
    •    Quantifying energy and resource intensive processes and minimizing their impact.
    •    Identifying cost savings for the manufacturer and consumer.

    •    Developing an  evaluation of impacts and risks to human health, the  environment, and
        society from the local to national and global scales.
    •    Demonstrating  a commitment by manufacturers to stakeholders for the responsible
        development of nanoproducts.
                                     Notice of Use

     Any comparisons discussed in this document refer to internal comparisons within a
     company's own  processes  or  alternatives.  Internal  comparisons can  help a
     company to improve an existing product line, but do not make any claims in
     regards to  other  companies' products.   Comparing products against similar
     products from other companies and making a public statement about results,
     known as a comparative assertion, requires a thorough review process, which
     is not presently addressed.
                                                                                     14

-------
2. Initializing the Path Forward to Sustainability	

Before offering  detailed guidance  based on the preliminary framework,  it is important to
articulate  how sustainability  can be integrated into existing models  of product  deployment.
Product deployment is comprised of two stages, product development and product management.
Traditionally, product development can be broken down into a series of steps including:

     •   Idea Generation: Use brainstorming to identify potential products and markets.

     •   Idea Screening: Eliminate impractical ideas and select most promising alternative.

     •   Concept Development and Testing: Establish engineering and marketing details.

     •   Business Analysis: Use stakeholder feedback to estimate pricing and profitability.

     •   Beta Testing and Market Testing: Produce prototype and evaluate typical use.

     •  Technical Implementation:  Develop quality guidelines for  manufacturing;  compile
        product data sheets; prepare supply chain for product deployment.

     •   Commercialization: Product launch into consumer markets.

At the onset of commercialization, product management is invoked to oversee the product life
cycle. In this case, the term "life cycle"  from a business perspective is different from what is
defined for sustainability and describes the four stages involved with marketing a product: (1)
Introduction, (2)  Growth, (3) Maturity, and (4) Decline. These stages track the market penetration
of a product and  its consumer acceptance. Upon entering the Decline stage, businesses can either
discontinue a product or seek to redevelop it to make it commercially viable.
To achieve development of sustainable technologies and products, careful consideration must be
given to  the issues  of sustainability prior to commercialization  and product  management.
However, this is not  as simple as inserting a "Sustainability Assessment" step into the product
development concept above. Instead, the steps of the sustainability framework described in this
document will have to move in tandem with the steps of product development as the needed data
become available. This will help insure that sustainability is achieved for a product in a manner
that maximizes its benefits to  a company while minimizing its impact on the process of product
deployment.
The  overlap of sustainability with product deployment  is shown in Table  2-1. Sustainability
concerns are  not considered  during the generation  of ideas to avoid stifling creativity. As  a
product idea  develops,  sustainable  design begins with  consideration  of the basic risks  and
eventually grows into full assessment of the  three pillars. However, the various pillars can and
should be considered at multiple times throughout product deployment to maximize the potential
for sustainability. For example, the environmental impacts of manufacturing and distributing  a
proposed product can be examined during concept development and testing using pilot scale data
and engineering  estimations. Once beta testing  is initiated, this assessment can be revisited and
expanded to include the use and disposal phase based on testing results. The data can be updated
using realistic manufacturing  data during technical implementation.  Finally, the environmental
impacts can be reassessed during product management using real market data to  identify any
potential concerns that did not manifest themselves during design. This approach is intended to
provide companies ample opportunity to amend a product to achieve sustainability goals prior to
large scale commercialization  through early detection and action. Thus, sustainable design from a
business perspective  is both iterative and cyclic in nature. By keeping this  in mind, it will be
easier to understand the discussions of assessment and decision-making tools that follow.
                                                                                     15

-------
       Table 2-1. Traditional Product Development and Sustainable Design proceed
                         in tandem and entail overlapping steps
Product Development Sustainable Design
Idea Generation
Idea Screening
Concept Development and
Testing
Business Analysis
Beta Testing and Market Testing
Technical Implementation
Commercialization
Product Management
Initial Risk Screening
Product Characterization, Risk Screening, Stake
Holder Identification, Environmental Assessment
Economic Assessment and Social Assessment
Product Characterization, Risk Screening,
Environmental Assessment
Risk Screening, Environmental Assessment
Economic Assessment and Social Assessment
Risk Screening, Environmental Assessment,
Economic Assessment, Social Assessment
2.1 Initial Product Characterization and Identification of Potential Risks
                                                  Defining the Nanoproduct Concept

                                                1. What is the function of this product?

                                                2. What is the anticipated target market?
                                                3. What are the necessary materials?
                                                4. Is there more than one approach?
One of the first  steps of the framework is  to
appropriately characterize the nanoproduct. This
includes an understanding  of how  the  product
will be used along with a general description of
the intended  materials and characteristics (e.g.,
chemical   composition,  physical  form/shape,
solubility, state of aggregation or agglomeration,
etc.),  and  physical and  mechanical properties.
Characterization  of  the  product  should  be
detailed  enough  to  provide  sufficient  guidance for conducting an  analysis throughout  the
product's life cycle stages. When characterizing the nanomaterials,  the reader can refer to  the
International Organization for Standardization's report on the classification and characterization
of nanomaterials (ISO, 2010).
Before undertaking a full sustainability assessment, it is appropriate to  determine whether easily
identifiable risks are present and whether they can be mitigated (Figure 2-1). This action is  not
intended to be a full risk  assessment as defined in Chapter 3. Instead,  it should be a quick
screening-level identification  of known risks. For example, if a proposed electronic product will
involve lead or other toxic metals as part of the circuitry, the known risks  related to exposure to
these  materials should be addressed before continuing. A benefit of this type of screening-level
risk identification is  that it  can lead to the development of better products that  are more
sustainable. However,  this  may  be more  a  challenging task  when  quickly  considering
nanoproducts because of the general lack of knowledge regarding the associated risks of
nanomaterials. Although the toxicity and risk related to nanomaterials and products continues to
be  investigated,  recent  studies  have found  a  correlation between  certain physicochemical
properties of nanomaterials and their potential toxicities.  For example, the size and shape of a
nanomaterial influences deposition patterns in the human respiratory tract. Nanomaterials that are
deposited in the respiratory tract may cause symptoms similar to those caused by asbestos (e.g.,
                                                                                      16

-------
immune-system  malfunction, lung disease, and cell inflammation) (Olapiriyakul and Caudill,
2008). Other  characteristics  of nanomaterials  that may affect toxicity include  the  chemical
composition, aspect ratio, crystal structure, surface area, surface chemistry and charge, solubility,
adhesive properties, and emergent properties (Klopffer et al., 2007). It has generally been found
that  if nanomaterials  are embedded as part  of larger  objects  (e.g.,  nanocomposites  and
nanocrystalline solids) they are less dispersive and present lower risk than if they are used as free
nanoparticles,  nanorods or nanofibers (Ostertag and Husing, 2008). In addition, recent research
focused primarily on airborne pathways has examined the  hazards associated with exposure to
nanomaterials  in the workplace. These studies have found that some nanomaterials damage lung
tissue after inhalation (Wiesner et al., 2006). The potential risk of such exposure will depend on
how the nanocomponents are produced and handled while being incorporating into nanoproducts.
A study by Sengul, et  al.  (Sengul et al.,  2008)  provides a comprehensive review  of techniques
used to manufacture nanoproducts, a description of the technique, key processes, materials used,
primary energy consumption, and environmentally significant aspects.  These techniques include
top-down approaches where  nano-scale dimensions are achieved through carving or grinding
(e.g.,  lithography, etching,  electro-spinning, and milling), or  bottom-up methods  in  which
nanomaterials  are developed  at the  atomic  scale using vapor-phase,  liquid-phase,  and self-
assembly techniques (Sengul et al., 2008).
                                              Characterize product and identify potential risks
       Characterize product
          and identify
          potential risks
                                                 Initial product
                                                characterization
                                                  identifiable
                                                   potential
Yes
       Implement
      measures to
      mitigate risks
 Consult
regulatory
authorities
                     Figure 2-1. Identifying Risks Early in the Process
As shown in Figure 2-1, if easily identifiable potential risks are found, the product developers
should  consult the  appropriate regulatory authorities  to determine  whether  there  are  any
compliance  issues relevant to the  nanomaterial or nanoproduct. In 2007,  EPA convened a
workgroup to discuss and document the science needs associated with nanoproduct. The resulting
"Nanotechnology White Paper" (EPA-SPC, 2007) includes an overview of environmental statutes
that may be applicable to nanoproducts (e.g., Toxic Substances Control Act (TSCA), Clean Air
Act (CAA),  and Clean Water Act (CWA)). EPA policies and regulations are evolving as research
on nanotechnology  and its impacts continue and additional information becomes available.
However, measures should be taken to  mitigate  any  easily identifiable  risks and ensure
compliance  with  the  current and  appropriate  regulations  before  taking  further  steps in
sustainability assessment. Examples of preliminary data needed to characterize early stage risks
of nanoproducts are given in Table 2-2. Although many  sources  are working to provide an
understanding  of  the  potential toxicity  of nanomaterials,  nanomanufacturers  should  be
encouraged to incorporate toxicity studies if possible into their product development because the
                                                                                       17

-------
                     Table 2-2. Identifying Risks Early in Development
Nanomanufacturing
                                    Are data available on the toxicity of the nanomaterials to
                                    human health or organisms in the environment?
                                    Are   data   available   (pilot   plant)  on  potential  air
                                    emissions,  waste discharges and amounts of solid waste
                                    generated?
                                    Is any risk issues associated with release of these waste
                                    streams? If yes, can the risk be contained  with state of
                                    the   art  treatment  protocols  available  to   treat  air
                                    pollutants/  liquid waste and solid waste?	
     Nanoproduct Use
                                Is there an anticipated release and transport of
                                nanomaterials from the nanoproduct during use?
Nanoproduct End Of Life
           (EOL)
                              • What is the likely frequency of waste generated?
                              • Is nanomaterial encapsulated,  semi dispersed or
                                loosely formulated in the product at its end of life?
                              • What are the effects of nanomaterials on
                                environment when incinerated or landfilled?
A Case Study in Early Intervention
  In the early 1980s, I was working In the plastics industry as the head of health, safety, and environment for a major
business division of a parent company. Researchers at the division had just discovered a thermosetting catalyst that
was unlike any previously available. Epoxy-type resins begin curing at the moment they are mixed with standard cata-
lysts. This new catalyst remained inactive until a very specific and  desirable temperature was reached.
  This catalyst had enormous potential, and the marketing department immediately began distributing samples of
it in order to encourage commercial trials. But there were also indications of problems with the substance. Both my
department and the division's medical department recognized that the catalyst contained small amounts of a material
that potentially could be toxic under certain circumstances.
  Jointly, we formed an internal task force consisting of toxicology, environmental, industrial hygiene, and medical
professionals. Our work was supported by an outside toxicology lab and a prestigious occupational medicine institute.
In relatively short order, we discovered that the material was not just toxic (as suspected}, but in fact so biologically
active that it was attracting interest from external physicians as a possible pharmaceutical because of its radical effect
on the circulatory system,
  Executive management at our division could have attempted to brush  off the preliminary test results as insufficient
and inadequate to justify withdrawal of the material from full-scale commercialization. They could have rationalized
away the risk and limited funding for what eventually became very expensive and sophisticated testing. Instead, they
fully supported our efforts and allowed the toxJcological testing and evaluation to proceed unhindered and in an unbi-
ased manner.
  Ultimately, the division decided to withdraw the catalyst from all  future commercial trials. As a result, no one was
ever seriously injured by the material.
  If the toxicity of the catalyst had been fully understood in advance, the material might have been introduced suc-
cessfully into a highly controlled, specialized niche market where its unique properties could have been utilized fully
and safely. But the mainstream market was not equipped to handle the material. The division considered it too likely
that customers would use the substance as they did other, relatively safe commercial catalysts, without proper safe-
guards.
                                                          By RichardMacLean (MacLean, 2009).
                                                                                                   18

-------
numerous variations that occur from batch to batch and product to product may lead to
unique toxicity characteristics for a given nanomaterial

2.2 Stakeholder Identification
Stakeholder participation is an integral element of sustainable decision making (Kiker et al.,
2005). Given the fact that the sustainability of a product may be viewed differently by different
stakeholder  groups, it is important to obtain their input and value judgments in conducting a
sustainability assessment. In broad terms, a stakeholder includes any group or individual who can
affect or is  affected by any aspect of a product (Freeman, 1984; Mitchell et al., 1997). These
stakeholders may include  those that have  a direct or indirect stake.  For example, direct
stakeholders may include those with a vested financial interest in the company  developing the
product or customers that will eventually use the product. Indirect stakeholders may include those
with the ability to influence the product development (e.g., regulators) (Young, 2008).
All groups or individuals that are likely to be affected, whether positively or negatively, by the
product throughout  its life cycle should be listed through a brainstorming session. When listing
the stakeholders, it is  often helpful to group them into different  categories  (e.g., workers,
management,  customers, community, etc.).  However, depending  on where a  product is in its
development cycle,  the product uses, target market, and associated stakeholders may be difficult
to identify. Therefore, the list of stakeholders may expand and need to be updated and revised as
a product is  further developed and commercialized.
Engaging all the identified  stakeholders is often unrealistic and resource intensive. Accordingly,
decision-makers must narrow the list of stakeholders to those that seem the most appropriate to
engage for the assessment. To assist in this effort, stakeholders may be "mapped" based on their
level of  interest, legitimacy, influence, or other factor relevant for the project (Bryson, 2004;
Young, 2008). Figure 2-2 includes an example stakeholder mapping approach.
                                High
                             Power
                                Low

Keep
Satisfied
Monitor
(Minimum
Effort)

Manage
Closely
Keep
Informed





                                           Interest
        Figure 2-2. Example of Stakeholder Mapping (Bryson, 2004; Young, 2008)
Stakeholders are  typically engaged in the goal definition and scoping phase,  evaluation and
interpretation of impacts and selection of final alternative approaches to mitigate impacts (steps 3,
5, and 8 of the framework). Participation strategies for soliciting stakeholder input may include
individual  surveys, public meetings, workshops, interviews, or a  combination thereof.  The
appropriate participation strategy should match stakeholder preferences, which may be influenced
by the level of trust participants may have of each other, scientific and technical experts, and the
product developers. Furthermore, as the level of trust evolves, the participation strategies may
need  to be  adjusted as appropriate (Anex and Focht,  2002). In  addition to the level of trust
between participants,  the  selected strategy may also reflect other factors such as the decision
context and available resources  for the study  (Kiker  et al.,  2005).  Ultimately, the specific
involvement of individuals or groups will depend on how closely a particular issue is of interest
                                                                                       19

-------
and  may  affect them  (Anex  and Focht, 2002). For  more discussion on  the  identification,
engagement, and analysis of stakeholder input, the reader is referred to Reed and coworkers
(Reed et al., 2009), Cuppen and coworkers (Cuppen et al., 2010), and Aaltonen and coworkers
(Aaltonen, 2011), all of which provide useful insight based on case studies.
                                 Involving Stakeholders

  As indicated in section 1.4, a key step in the decision framework is to identify and engage
  stakeholders who can provide input in defining the goal and developing sustainability metrics,
  evaluating and interpreting impacts and selecting alternative approaches to mitigate impacts.
  This is especially important when working in a collaborative manner,  such as in a private-
  public partnership,  or in  an industry  consortium. As mentioned previously, stakeholders
  include any group or individual who can affect or is affected by any aspect of the nanoproduct
  (Freeman, 1984; Mitchell; 1997), and may also include experts such as LCA practitioners,
  risk assessors,  economists, or  public  health  experts, or other  scientists  that  may  offer
  important advice in assessing impacts and selecting approaches for mitigating these impacts.

  A diversity of perspectives  should be used to  inform the project goals and scope,  identify the
  functional unit and alternatives  (if a comparative assessment is to be conducted), refine the
  methodology, monitor  its implementation, and facilitate use of the information generated by
  the study to make product improvements and, if a comparative assessment is conducted, to
  choose safer materials and processes. Involvement throughout the project helps to ensure that
  stakeholders contribute to, understand and support the outcome, enhancing  credibility and
  promoting product improvements. Stakeholders are drawn from the entire supply chain and
  all life-cycle stages of the product. In addition to  the experts identified  above,  typical
  stakeholders may include: chemical and product manufacturers and suppliers; product users
  and retailers; waste and recycling companies, government agencies;  academics;  and non-
  governmental organizations. Those developing new technologies in the area of study (that may
  be analyzed as potentially safer alternatives) should be included in  the stakeholder group.

  Early  in the study, researchers and product developers should identify, contact,  and inform
  potential stakeholders  of the proposed project goal and scope, methodology, and potential
  benefits, to gain the expertise needed to conduct the study and representation of the broadest-
  possible points of view. Stakeholders can be identified via industry conferences, trade groups,
  academic institutions,  or industry contacts. When proposing a collaborative effort (either by
  sending out an invitation to a group, or making individual  contacts), it is important to point
  out the mutual benefits and potential savings that will result from sharing the cost of the study,
  and from making product improvements that result in the use of less energy and materials and
  provide the competitive advantages associated with generating fewer environmental impacts.
  When  contacting potential stakeholders, it is also important to  consider and explain how
  confidential business information will be protected, and to make clear  that confidential data
  will not be made public.
                                      KathyHart, EPA LCA Project Leader (Hart, 2009)
2.3 Goal and Scope Definition
The goal definition and scoping phase of a sustainability assessment should include information
on the focus of the study, including the research questions to be answered, system boundaries,
and functional unit (quantified reference flow) of the product system. The term functional unit is
most often associated with LCA because it can provide a basis for comparison across the life
                                                                                      20

-------
cycle stages of dissimilar products having  the same  use.  For example,  a study examining
antimicrobial foot treatment for diabetics might be based on the amount of product needed for
one treatment, such as a pair of socks containing nanoscale silver or the prescribed quantity of an
antimicrobial ointment dose. Other assessment tools may require a different reference such as the
mass of a product unit or its monetary value.  These simpler reference flows are also applicable
for LCA  when  considering  a single product or  similar products (i.e. silver socks made  by
different manufacturers). The key point to remember going forward is that once a reference flow
is  established, it can easily be converted into equivalent forms  to meet the needs of the various
assessment tools.
The overall goal of a sustainability assessment is to select and/or develop a product, process, or
service that is sustainable. To make this a more practical goal, the concept of sustainability should
be broken down into criteria for each pillar (environmental, economic, and social), paying special
attention to any  criteria and indicators that  are pertinent  to  the  product in question. If the
manufacture of a nanoproduct is particularly energy-intensive, then  impacts associated  with
energy use might be particularly  critical to the overall sustainability of the product.  Energy-
related criteria and indicators could include, but are not limited to, air and water emissions from
upstream power plants, material resources used for energy  production, electricity costs, power
plant  working  conditions,  and the impact of power  plants on local communities. However,
impacts not related to energy use (ozone depletion, acidification, material costs, etc.) must not be
excluded a priori because they may actually be the more severe  impacts for the product based on
other  phases of  the life  cycle.  Ultimately,  the product  developers  should  work  with the
stakeholders to select relevant criteria that accurately encompass sustainability (Azapagic et al.,
2006). Table 2-3  provides  a list of example sustainability criteria for each pillar. Some criteria
such as noise may not fit uniquely with one pillar. In such cases, the criterion in question should
be assigned to a pillar during goal  and scoping to avoid (if possible) double counting the impact.
The list provided is by no means authoritative  with the understanding that ultimate placement of
criteria within the  pillars may depend on the indicators and metrics needed to measure them and
what tools are available to assess these metrics and indicators.
     Table 2-3. Examples of Sustainability Criteria (adapted from Azapagic et al., 2006)
  Environmental
  Energy use
  Resource use (renewable and
  non-renewable)
  Emissions (air, water, land)
  Global warming
  Ozone depletion
  Acidifications
  Ecotoxicity impacts
  Human toxicity impacts
  Water eutrophication	
Economic
Macro-economic
    Environmental liabilities
    Taxes
Micro-economic
    Capital costs
    Operating costs
    Consumer costs
-   Profitability
Social
Provisions of employment
Health and safety of:
    Employees
    Customers
-   Public
Nuisance
    Noise
-   Odor
Public Acceptability
A list of complete sustainability criteria is too numerous  to include here and would pose  a
daunting challenge if stakeholders tried to account for the entire set. Therefore, stakeholders must
select relevant criteria from each pillar to adequately capture sustainability. The selection process
is not trivial based on the varying preferences of stakeholders with regard to sustainability.  For
                                                                                        21

-------
insight, readers are encouraged to read the work of Hirschberg and coworkers (Hirschberg et al.,
2007)  who have studied the selection process and offer guidance on what to consider when
building a criteria set to  insure  it is scientific, functional and pragmatic.  For  example, a
manageable number of criteria should be selected equally from the three pillars for balance while
capturing essential technological characteristics of products and process to allow differentiation
amongst them. With a defined selection process, stakeholder preference can be incorporated in a
way that minimizes the subjective nature of sustainability. To illustrate the concept of balance, a
sustainability assessment for a given product might examine energy and resource use, global
warming, toxicity impacts, profitability, environmental liabilities, health  and safety, and public
acceptance. It is also possible that certain criteria can only be evaluated qualitatively, such as
social  issues. These  approaches are permissible provided such choices are clearly  defined and
justified in the scope of the assessment. These decisions may be based on the resources available
for the study, data availability, or the product development stage (EPA, 2006).
         Preferences in Sustainability Criteria: The Case of Antimicrobial Textiles

  Consider  the  growing use  of nanoscale  silver in  antimicrobial  textiles.  A group of
  stakeholders for a product in this application might include the product manufacturer,  textile
  groups, government agencies overseeing environmental and public health, consumer advocate
  groups,  and residents of the communities surrounding the  manufacturing site. For the
  manufacturers, profitability is obviously the most important criterion. However, profitability is
  directly related to both operating cost and public acceptance. The operating costs will depend
  on energy and resource use, potential  emissions  and environmental liabilities, taxes,  and
  employee safety. These criteria are all easily quantifiable based on an understanding of the
  manufacturing  process and workplace  risk.  The other concern for profitability,  public
  acceptance,  is  not  readily quantifiable  and may be evaluated  qualitatively  based  on an
  understanding of the market for  the intended product. This might involve simple yes/no
  questions to determine if a large enough consumer base exists or more detailed analyses of
  factors that could create a potential negative public perception.  Government agencies  will
  care more  about the  impact on the  public  and environment.  Criteria for these needs  will
  include climate change (global warming, ozone depletion, acidification) and toxicity impacts
  to humans and ecosystems. Again, these  are all quantifiable criteria. However, as opposed to
  the manufacturer, who might only focus  on its own operation, government agencies will want
  to know about the impacts associated with the entire life cycle of the product. In a similar
  manner, consumers groups and local residents will be  concerned with the potential  health
  effects associated with product manufacture  and use.  For these criteria,  a  quantitative
  measure of health effects will be needed to satisfy the needs of the stakeholders.  This will not
  be easy for most nanoproducts given the lack of available toxicity data for nanocomponents.
  In addition, consumers will also consider product cost when making a decision.  What benefit
  will the product be  to the consumer if it is unaffordable? Again,  until the  manufacturer can
  fully quantify the costs associated with the product, how  can they provide an associated
  consumer cost? This  example should illustrate the complex nature of decision making for
  sustainability. Each stakeholder will not  only have his/her own important criteria that must be
  satisfied,  but will have differing  scopes and boundaries  when  beginning the assessment
  process that must be  reconciled.  Ultimately, the rules for  selection (Hirschberg 2007)  and
  data availability will be detrimental when prioritizing the criteria for assessment.
                                                                                       22

-------
It is critical to recognize that a product is sustainable only if it is sustainable with respect to each
of the three pillars. Accordingly, if one or two pillars are excluded due to practical constraints
(i.e. resources,  data  availability,  etc.),  the resulting study  must be identified as a  partial
sustainability assessment, and the interpretation of results must discuss the excluded pillar(s) and
any known potential impacts associated with the pillar(s). While practical considerations  cannot
be ignored, a lack of data or resources cannot be confused with the absence of impacts. The level
at which  the  criteria will be measured must also be considered.  For  example, the product
developers and  stakeholders may be  interested  in impacts at the  product (micro) level,  an
industrial  sector (meso) level, or an economy-wide (macro) level (Zamagni et al., 2009). These
distinctions are chosen based on the perceived ability to impart change on connected systems
within each level. Thus, little change to space, market, and time is anticipated for perturbations at
the product level while changes at the economy-wide level will affect all systems. Generally, it is
advisable  for  emerging technologies to  start analyses at the product  level and  expand to the
economy-wide level as the product is brought to market  and additional information becomes
available  (Zamagni et al., 2009).  In the  absence  of accurate  data,  possible  macro-level
considerations can be included during the initial assessment if desired using qualitative evaluation
of criteria. Such evaluations can lead to greater insight during interpretation of results.
When defining system boundaries, it  is important to include  every step that  could affect the
overall  interpretation or ability of the  analysis  to address the issues for  which  it is  being
performed. The system boundaries not only refer  to processes, but include the geographical and
temporal boundaries in which the nanoproduct will exist. These last two factors will help identify
suitable data quality criteria and provide a context to interpret results. In determining the system
boundaries, it is helpful to develop a system flow diagram, as shown in Figure 2-3, to depict the
activities  and  direction of flow  of products and materials. This will also be useful later on in
guiding the efforts to gather data for assessment. Each  system step should be represented
individually in the diagram, including the production steps for ancillary inputs or outputs such as
chemicals and packaging.
For a sustainability  assessment, the functional unit has been defined as  a basis to normalize data
using equivalent use (or  service provided to  consumers)  to provide  a  reference for relating
impacts across life cycle stages as part of an improvement assessment. The functional unit is also
useful  for  comparing  alternate product  systems  or technologies.  However,  as  discussed
previously, any  comparisons discussed  in  this framework are internal  to a  company's own
processes  or alternatives. Given  that nanotechnology is an emerging field, the eventual function
of the product and service level to the customer may not be known. As a result, it may  not be
possible to develop  a functional unit, especially if the product is still in  development (Klopffer et
al., 2007).  In such  cases,  the product  developers should determine whether  reasonable
assumptions may be made about the eventual product use and anticipated service provided to the
customer.  If it is not possible to  develop a service-based functional unit and the assessment does
not compare  product  systems,  the product developers should determine  another basis for
organizing the data and results into context.
                                                                                        23

-------
Fi*ld
Preparation
1^
Planting
|
i 1 i I
1 Irrigation Soil Wood Insecl
conservation uoniroi Manager
1
I
Harvesting
1
Seed Cotton Storage
I Ginning
1
Classing
iirzi
I Transportation
I
Marketing
1
[ Transportation
I
Yarn Production
TL,
1 Transportation |
|
Fabric Manufacturing
1





C

i



Ground
[^^r«p»alk>n^^
^
1 Silver Ore Extraetio
	 , i
i
mj Ben«ftc(aUon

1 Transportation
1
Silver Feedstock
manyfacturing

i



Transportation
i
Dislnbutioni
1
NanoscaJe Silver
Manuifaclurln§

Transportat ion

..,,,. 1

^==^r
Transportation j
1
Distribution |
1
Use
Wear and Wash)
1
Disposal




Figure 2-3. Sample Goal and Scope definition: The life cycle of a pair of cotton socks
                            containing antimicrobial silver.
                                                                                 24

-------
3. Assessing Environmental, Economic, and Social Impacts	

The following section presents a summary of methods for assessing the environmental, economic,
and social impacts of a nanoproduct. These assessment methods largely reflect methods that are
commonly used by  industry and LCA practitioners. They were selected because they can be
applied to holistically assess nanoproducts and are based on citable standards and guidance for
application. Qualitative as well as  quantitative assessments may also be used, if needed. For
example, the  EPA's  Comprehensive Environmental  Assessment  (CEA) method  is being
developed as a semi-quantitative tool based on expert judgment to identify information gaps and
research needs (EPA, 2010).

In most cases, it is not necessary (or practical) to apply all of the available methods (Zamagni et
al.,  2009).  The determination of which  methods  are most  appropriate will  depend on  the
applicability of the technique, whether the method will respond to the  research questions and
criteria, and the available resources for conducting the sustainability assessment. In addition,
given the uncertainty associated with nanoproducts, any tool selected should address  uncertainty
and how it may  affect the assessment given the  nature of nanotechnology.   The  work by
Schepelmann et al. (CALCAS, 2008) may be  a helpful resource to aid in selecting methods as it
provides an assessment of the strengths and weaknesses of various model and tools which support
sustainability analysis.

Given the nature  of nanoproducts  and nanomaterials  and the fact that they  are  a  new and
emerging technology,  there are many issues and  data gaps  that may arise when assessing
environmental, economic, and social impacts. Although some issues are applicable  for overall
aspects of a sustainability assessment, certain issues may be relevant for specific life cycle stages
and sustainability pillars. As described in Section 1.2, the key life cycle stages include:

       1.   Raw materials extraction,
       2.   Materials processing,

       3.   Product manufacture,

       4.   Product use, and
       5.   End-of-life disposition.

Some of the  common  issues and data gaps, and possible methods for addressing them,  are
explained in detail below.

3.1  Environmental Assessment Methods

Many of the environmental protection strategies  in which we  have become accustomed can be
viewed as short-term  or quasi-environmental fixes. We now  understand that environmental
problems are rarely contained within a single resource area or within a single product's life cycle.
Instead, they require longer term strategies that extend across geographic regions and timeframes.
It has become  obvious  that a more integrated, systems-based  approach  is required to meet the
needs of today while  maintaining  the prospects  for the  same  quality  of life  for tomorrow's
generation. As a result, many methods and tools are being offered to assess the environmental
impacts of a product throughout its life cycle (see Table 3-1). For  more detailed information
about applying the methods, please consult the references listed on the  right-hand  side of the
table. As discussed in Section 2, the method or tool selected  should be based on the goal and
scope of the sustainability assessment.
                                                                                     25

-------
Table 3-1. Key Environmental Assessment Methods
Method Description/ Scope/ Impacts Measured Reference
Benefits Stage
Life-Cycle
Assessment
(LCA)










Carbon
Footprint










Environmental
ly-Extended
Economic
Input-Output
(EEIO) Life
Cycle Analysis









Life Cycle Risk
Assessment
e.g., Nano Risk
Assessment



Evaluates potential
environmental impacts
associated with a
product, process, or
activity. LCAs consider
multi-media, multi-
attribute impacts by
quantifying energy and
materials used and
wastes released to the
environment from
cradle to grave.

Both GHG Life Cycle
Analysis and Carbon
Footprinting aim to
account for the release
of greenhouse gases
that contribute to global
climate change. The
principal gases are
carbon dioxide,
methane, nitrous oxide,
and fluorinated gases,
such as chlorinated
fluorocarbons (CFCs).
Assesses the economy-
wide environmental
impacts of a product
throughout its life cycle
stages. Note that this
method may also be
used to conduct an
economic assessment
(see Section 4).






Characterizes the
nature and magnitude
of health risks to
humans (e.g., residents,
workers, recreational
visitors) and ecological
receptors (e.g., birds,
Product
to
regional/nati
onal level

All life cycle
stages






All life cycle
stages










Product/
micro level
to economy-
wide level

All life cycle
stages








Product/local
and meso
level

All life cycle
stages

• Natural Resource
Use (e.g., water,
nonrenewables,
etc.)
• Global warming
• Ozone depletion
• Smog Formation
• Acidification
• Eutrophication
• Human Health
• Ecotoxicity
• Land Use
• Etc.
• Carbon
• Greenhouse gases
• Global warming
• Climate Change









• Economic activity
generated
• Natural Resource
Impacts (e.g.,
energy use, fuel
use, ores, etc.)
• Abiotic Ecosystem
impacts (e.g., green
house gas
emissions, ozone
depletion, smog,
etc.)
• Toxic releases by
sector and
chemical
• Health hazards
(e.g.,
neurotoxicity, skin
absorption,
genotoxicity, etc.)
• Environmental
(e.g., aquatic,
(Baumann and
Tillman, 2004;
EPA, 2006;
ISO, 2006;
SET AC, 1992)








(BSI, 2008;
WRI, 2010)










(CMU, 2008a;
Klopfferetal.,
2007; Wiedema,
2010)











(EPA, 2010;
Walsh and
Medley, 2007)




                                                            26

-------
Method Description/ Scope/ Impacts Measured Reference
Benefits Stage











Ecosystems
Services LCA
(ECO-LCA)


Sustainable
Materials
Management







fish, wildlife) from
chemical contaminants
and other stressors that
may be present in the
environment. Risk
assessments are
composed of two sub
assessments: an
exposure assessment
and a hazard
assessment.
Expands upon
traditional LCA and
quantifies ecosystem
services over the life
cycle of a product.
Quantifies the relative
magnitude of material
flows in the global
economy. Methods of
material flow
accounting, such as
Material Flow Analysis
(MFA) and Total
Material Requirements
(TMR), are used.











Product/
local, meso

All life cycle
stages
All life cycle
stages, with a
focus on
material
extraction
and end-of-
life
management
(recycling).

terrestrial, avian,
etc.)

• Safety (e.g.,
explosivity,
reactivity,
corrosivity, etc.)




Ecological services
(e.g., land-use).



Flows (Kg)




















(Zhang etal.,
20 lOb; Zhang et
al., 2010c)


(Fiksel, 2006)










-------
Life Cycle Assessment

The most comprehensive method to assess the environmental impacts of a product, process, or
activity throughout its life cycle stages is environmental life cycle assessment (LCA). A LCA
accounts for the physical flows, i.e. the inputs and outputs, across the full life cycle of a product
system, from materials acquisition to manufacturing, use, and final disposition (see Figure 3-1).
As outlined in the International Standards Organization (ISO)  14040 series, an environmental
LCA study has four major components: goal definition and scoping, life cycle inventory (LCI),
life cycle impact assessment (LCIA), and interpretation of results (ISO, 2006).
Inventory data are subjected to life cycle impact assessment models, which seek to establish a
linkage between a system and the potential, related impacts. The impact models are often derived
and  simplified versions  of more  sophisticated models  within  each of the various impact
categories.

Although work is ongoing  to  reach consensus on which  impact categories to  include,  the
International  Reference Life Cycle Data System (ILCD)  Handbook (JRC, 2010) provides  the
following list of commonly used categories:

   •    Ozone Depletion                             •  Fossil Fuel Use
   •    Global Warming                             •  Land Use
   •    Human Health                               •  Water Use
   •    Ecotoxicity                                  •  Land Use
   •    Eutrophication                              •  Resource Depletion
   •    Acidification
   •    Smog Formation

These simplified models are  suitable for relative comparisons of the potential to cause human or
environmental damage, but are not indicators of absolute risk or actual damage to human health
or the  environment.  For that, a risk assessment is needed. In the  case of a traditional risk
assessment, it is possible to conduct very detailed modeling of the  predicted impacts  of the
chemical on the population exposed and even to predict the  probability of the population being
impacted by the emission. In the case of LCA, hundreds of chemical emissions (and resource
stressors) which are occurring at various locations  are evaluated for their potential impacts in
multiple impact categories.  The sheer number of stressors being  evaluated,  the variety  of
locations, and the diversity of impact categories makes it impossible to conduct the assessment at
the same level of rigor as a traditional risk assessment. Instead, models are based on the accepted
models within each of the impact categories using assumptions and default values as necessary.


LCA is a well-established methodology for  evaluating the  environmental impact  of products,
materials, and processes. By including  the  impacts throughout  the product life  cycle, LCA
provides a comprehensive view of a product's environmental aspects.  It is also valuable in
evaluating the many interdependent processes that are involved in a product system. A change to
one part of this  system may have unintended consequences  elsewhere. LCA  identifies  the
potential transfer of environmental impacts from one medium  to another (e.g., eliminating air
emissions by creating a wastewater effluent instead) and/or from one life cycle stage to another
(e.g., from use and reuse of the product to the raw material acquisition stage). If an LCA were not
performed, the transfer might not be recognized and properly included in the analysis because it is
outside of the typical scope or focus of product design and selection processes.
                                                                                      28

-------
A workshop comprised of international experts from the fields of both LCA and nanotechnology
concluded that the LCA ISO-framework (ISO 14040:2006)  (ISO, 2006) is fully suitable to all
stages of the life cycle of nano-components and nanoproducts (Klopffer et al., 2007). However,
the workshop attendees acknowledged a number of operational issues that need to be addressed,
such as functional unit selection, inventory data  collection and/or estimation,  allocation, and
toxicity assessment.  Similarly,  Bauer and coworkers have also pointed  out  that a suitable
definition of the functional unit and system boundaries for nanoproducts  will be necessary to
facilitate comparative assessments (Bauer  et al.,  2008).   A  complete list of LCA  services,
software tools and databases can be found at http://lca.jrc.ec.europa.eu/lcainfohub/directory.vm.




F


»
roduct System Boundary
Raw Materials
Extraction




,

Inputs
if

Materials
Processing









(materials, energy, resources)

,
T

Product
Manufacture


,
V

Product Use
V )


,
T

End of Life
(EOL)



•9 ¥ f » V

[ Outputs (products, emissions, wastes) ]

        Figure 3-1. General Framework for a Product Life-Cycle Assessment (LCA)
Carbon Footprint

Carbon Footprint (CF) has become widely used in relation to the threat of global climate change.
CF is the measurement of the overall amount of carbon dioxide (CO2) and other greenhouse gas
(GHG)  emissions (e.g., methane, nitrous oxide, etc.) associated with a product, a person, an
organization or an event. For products, the boundaries include the  supply-chain and sometimes
use and end-of-life recovery and disposal. For simplicity of reporting, it is often expressed in
terms of the amount of carbon dioxide (tons or kilograms), or CO2- equivalents.

Two  well-known accounting tools for quantifying  and managing  GHG emissions  are the
Greenhouse Gas  Protocol (GHG Protocol)  (WRI, 2010)  and BSFs "Specification  for the
assessment of the life cycle greenhouse gas emissions of goods and services" (BSI, 2008).
In 2006, the International Organization for Standardization (ISO) adopted the Corporate Standard
as the basis for its ISO 14064-1:  Specification with Guidance at the Organization Level for
Quantification and Reporting of Greenhouse Gas Emissions and Removals.

Several CF calculator tools are available on-line (e.g., http://www.carbonfootprint.com/)

Despite its ubiquitous  appearance  there  seems  to be no clear definition of the term carbon
footprint. There is still much confusion as to what it actually means, what it measures, and what
unit  is to be used. While commonly understood to refer to certain gaseous  emissions that are
relevant to climate change and associated with human production or consumption activities, there
is no agreement on how to measure or quantify a carbon footprint. Questions remain regarding
                                                                                     29

-------
whether the carbon components should be weighted and normalized based on their potential
effect in the atmosphere. Other questions that need to be answered include the following:

    •  Should the carbon footprint include just carbon dioxide (CO2) emissions or other GHG
       emissions as well, e.g., methane?

    •  Should it be restricted to carbon-based gases or can it include substances that do not have
       a carbon atom in their molecule, e.g., N2O which is another powerful GHG?

    •  Should the carbon footprint be restricted to substances with a global warming potential at
       all, since  there are  gaseous emissions that are  carbon-based and  relevant to the
       environment  and health, such as carbon monoxide  (CO) which can convert into CO2
       through chemical processes in the atmosphere?

    •  Should the measure include  all sources of emissions, including those that do not stem
       from fossil fuels, e.g., CO2 emissions from soils?

The Carbon Footprint approach is included in this guidance because of its widespread use in the
management of GHG gas emissions. While developing strategies for mitigating climate change is
indeed important, we must be mindful to not exclusively focus on one factor and discount other
equally important  environmental  aspects,  such those listed under LCA. Decision makers are
encouraged to  build upon the  results  of the carbon footprint  approach  toward a  holistic
examination of environmental  impacts in order to identify potential unintended trade-offs
between different environmental categories.
Environmentally-Extended Economic Input/Output LCA

In conducting an LCA,  the creation of a life cycle inventory often follows a "process-based"
approach in which the resource inputs and the releases to the environment are reported for all the
processes within the life  cycle system. Because finding such process data can be challenging (see
Section 3.1.1)  methodology  developers  created  an approach  that  uses  national  economic
input/output (I/O) models to help estimate the materials and energy resources required for, and
the environmental emissions resulting from, activities in the entire economy (CMU, 2008b).

The  process-based and  the environmentally extended  economic I/O  based approaches have
advantages and disadvantages, and it is possible to create hybrid models that incorporate aspects
of both. As the name suggests, process-based LCAs model processes, such  as the steps for
manufacturing carbon nanotubes. Although process-based LCAs can provide detailed information
about a nanoproduct system, they may necessitate artificial boundaries between the products of
interest and the rest of the economy.

The  disadvantage of input-output-databases is that processes are aggregated at the level of
product groups  rather  than  individual  products. This disadvantage  can be  overcome by
conducting a hybrid analysis,  which essentially links the inputs and outputs  of a process-based
LCA into an input/output-database (Wiedema, 2010). For example, one could use process-based
LCA techniques to model the impacts of the production processes  at a given facility, but use EIO-
LCA to model the supply chain impacts of the electricity purchased by the facility (CMU, 2008a).
Three  main ways of combining process-based LCA and  I/O- based LCA are: tiered  hybrid
analysis, lO-based hybrid analysis and integrated hybrid analysis (CALCAS, 2008).
                                                                                     30

-------
Several I/O databases are available, including a free database from Open IO that covers the USA
in 430 industrial sectors, including emissions relevant to global  warming. Open IO is jointly
administered by the  Sustainability Consortium and the University of Arkansas Sam M. Walton
School of Business.

A well-known LCA-based approach that uses the economic input/output model was developed by
the Carnegie Mellon Green Design Institute and is freely available on-line (www.eiolca.net).
Life Cycle Risk Assessment

Life Cycle Risk Assessment (LCRA) integrates the traditional risk assessment paradigm
with  a life  cycle perspective.  It  attempts  to  examine potential  human  health  and
ecological  impacts  (both  positive  and negative)  in  a broad, systematic  manner by
stepping the decision maker through  the life cycle  of a  material  in identifying the
pertinent exposure pathways and forms  of a substance. This, in turn, can identify the need
for more detailed evaluation at particular  life  cycle stages to characterize  impacts
(Shatkin, 2008).   The life cycle nature of the approach indicates  that it encompasses a
cradle-to-grave framework while accounting for multi-media environmental fate  and
transport, exposure, and effects  on  both ecological receptors and human  health. Other
dimensions such as economic, political, security or societal factors are typically excluded.

In 2005, ED and DuPont entered into a partnership to develop a framework for the responsible
development,  production,  use, and end-of-life disposal or recycling of engineered nanoscale
materials, that is, across a product's life cycle. The resulting "Nano Risk Framework" (Figure 3-
2) develops profiles of nanomaterials' properties,  inherent hazards, and associated exposures
throughout the material's life  cycle (ED-DuPont, 2007).
                                        Iterate
Describe
Material
&
Application



._
-~*==^-
Profile
Lifecycle(s
Properties
Hazards

Exposure

	 !

)







Evaluate Assess
Risks Risk
Management


1 1

Ai&esi-_giioritize_&_geiierate dat

Decide.
Document
& Act



r-
1
Review
&
Adapt



— —
_^-~-~
                  Figure 3-2. Nano Risk Framework (ED-DuPont 2007)

The ED-DuPont Nano-Risk Framework is not intended to be a full-scale life cycle analysis, in
which one pays prominent attention to resource inputs. Instead, it is intended to help users
assess, manage, and report the potential environmental, health, and safety risks associated
with a particular material and  application. It follows a traditional risk-assessment paradigm
                                                                                  31

-------
similar to the one used by the U.S. Environmental Protection Agency (EPA) for evaluating new
chemicals (EPA's New Chemicals Program: http://www.epa.gov/oppt/newchems/index.htm).
However, it does not present a "one size fits all" approach.  Different organizations, depending
their size and structure, will have differing ways of implementing the  framework for maximum
effectiveness.
Ecosystems Services LCA

Natural ecosystems  provide us with a multitude of resources and processes from clean drinking
water,  to processes  such as the decomposition of wastes,  and other benefits such as pleasant
aesthetics.  Ecosystems and the valuable services they  provide  (e.g., soil, pollination,  flood
prevention  and cropland) are often overlooked, but recently attempts are being made to include
these aspects in environmental assessment tools. For example, the Ohio State University Center
for Resilience developed  a  free, online tool called Eco-LCA (http://resilience.eng.ohio-
state.edu/ecolca-cv/). Eco-LCA was developed to complement other LCA tools by showing how
different products and materials have different impacts on nature.

Those  services are  divided into  four  areas: supporting services (soil,  pollination,  sunlight,
hydropotential, geothermal, wind), regulating services  (flood protection,  disease  regulation,
carbon sequestration), provisioning services (fuels, ores, water, timber, cropland), and cultural
services (spiritual and recreational  benefits). Eco-LCA includes various aggregation schemes that
are based on thermodynamic concepts (Zhang et al., 2010a).

Sustainable Materials Management

Sustainable Materials Management (SMM) is an approach to promoting sustainable materials use,
integrating  actions targeted at reducing negative environmental impacts and preserving natural
capital throughout the life  cycle of materials by taking  into account  economic efficiency and
social equity (OECD, 2005). At this point, SMM is more a concept than a single methodology.
Many suggest that the focus of SMM should be on environmental impacts of the materials flows,
rather than simply  on volumes or weights of materials alone (OECD, 2005). Material  Flow
Analysis, or Accounting, (MFA) is a complementary tool that tracks the amounts of a material as
it goes  into multiple products,  as they enter and exit the economy through various  types of
transactions. Although there is no global consensus  on MFA methodology, MFA can provide
important background information  and data for life cycle approaches and SMM.

SSM is especially applicable to nanotechnology because non-renewable metals and rare earths are
often used to manufacture various  nanocomponents. Although the application of nanotechnology
could  lead to significant  reductions  in  the  consumption of critical minerals compared to
traditional technologies, the need to extract more  metals, such  as silver,  gold, titanium and
lithium, and other exotic rare earth metals, such as europium,  cerium,  neodymium, gadolinium,
and terbium, will continue as the demand for nanoproducts grows. These natural resources are in
limited quantities and there may not be enough supply of them in the near future.
                                                                                      32

-------
The Need for Multiple Methods to Understand the Complexity of Nanotechnologies

The primary impacts associated with the development of Carbon nanotube (CNT) nanoproducts
can be attributed to either the  manufacture of CNTs and CNT nanoproducts or the release of
CNTs into the environment throughout the life cycle. Manufacturing impacts are related to both
the selection of raw materials, particularly the metal catalysts and carbon precursors, and the unit
operations involved  during the incorporation of CNTs into nanoproducts. For  example, the
popular chemical vapor deposition (CVD) process used to generate CNTs involves thermal
pretreatment of the gaseous hydrocarbon to increase the feedstock purity (Journet et al.,  1997;
Sinha et al., 2006) and accelerate formation and growth of CNTs (Plata et al., 2009). Although
beneficial to the process, this step can be responsible for the airborne release of greenhouse gases
(GHGs),  volatile organic compounds (VOCs), and polycyclic aromatic hydrocarbons (PAHs)
(Plata et  al., 2009).  Likewise, the formation and release of carbon soot as a byproduct upon
heating the precursor during arc discharge  processes  is possible (Plata et al., 2009). Photo-
reactive VOCs can cause additional impacts like smog  formation and ozone depletion (Plata et
al., 2009; Singh et al., 2008; USEPA, 1976). Certain PAHs can  accumulate and persist in the
environment, posing  a threat to human health as a cancer risk (Plata et al., 2009). Carbon soot
formation not only disrupts the  radiative heat balance (Kauffman and Fraser,  1997) of the
atmosphere but is a serious concern for public health (USEPA, 1977). Furthermore, Plata et al.
(2009) observed a  significant increase in the quantity of greenhouse  gases emitted during CVD
when the reactor temperature is only slightly increased. Similarly, the thermal pretreatment of
ethane as a carbon precursor results in the formation of larger quantities of byproducts including
GHGs such as methane, several photo-reactive VOCs, and toxic compounds such as benzene and
1,3 butadiene (> 36000 ppmv) (Plata et al., 2009). The acid purification step used to remove trace
impurities of the metal catalyst from raw CNTs may result in the discharge of unconventional
liquid waste to wastewater treatment plants (Singh et al., 2008). Some of the waste compounds
such as molybdenum or cobalt chloride (MoCl2, CoCl2) are not completely treated by wastewater
treatment plants and can pose serious risks to  aquatic species if released into fresh-water bodies.
Molybdenum compounds are known to cause anoxic conditions (Arnold et al., 2004; Gooday et
al., 2009) while cobalt compounds support the growth of blue-green algae  (harmful algal blooms)
which can lead to eutrophication (Hansen et al., 1954).

The potential health risk of CNT nanoproducts is a function of their release probability and the
inherent  toxicity  of the CNTs. The potential release of aerosolized CNTs  at CNT  and
nanoproduct manufacturing sites greatly depends on the type of process and work place practices
adapted to handle free CNTs (Kohler et al., 2008).  The aerosolization of CNTs is a function of
their size and rates of diffusion, agglomeration, deposition and re-suspension into the surrounding
environment (Kohler et al.,  2008). For example, the HIPCO method of CNT synthesis releases
larger amounts of CNT aerosols than other  synthesis  methods.  Similarly, gas  phase product
recovery  of CNTs and unit operations such as mechanical milling and dry CNT powder mixing
increase the probability of forming CNT aerosols (Kohler et al., 2008).

The probability of CNT release from nanoproducts during use  and disposal will depend on the
durability of the nanoproduct. For example, a window frame or automotive body panel made with
a CNT-polymer composite material is expected to have a minimal chance of CNT release during
use. On the other hand, products such as CNT textiles  and  CNT coatings may exhibit a larger
potential for release, particularly when subjected to thermal degradation, photochemical oxidation
and other harsh weathering patterns (Kohler et al., 2008). Furthermore, there are  concerns that
CNT-bearing waste will cause problems during conventional disposal processes. Incineration of
CNT-laden waste may generate CNT  aerosols if the operating temperature of the incinerator
facility is less than the decomposition temperature of CNTs (Kohler et al., 2008). The presence of
                                                                                     33

-------
CNTs in landfill leachate may affect the remediation efficiencies for other hazardous pollutants
(e.g. PAH and Pyrene) by altering their fate and bioavailability in the environment (Petersen et
al., 2009; Yang etal., 2006).

The release of CNTs is important because it is the first step in exposure to humans and biological
receptors. Once CNTs are released into the environment, impact of exposure will depend on the
toxicity of CNTs as determined by their physico-chemical properties. The extent of toxic damage
will  be influenced by  both  the bioavailability (uptake)  and bioaccumulation of CNTs by
biological receptors (Linkov  et al.,  2009). Bioavailability and bioaccumulation potentials are
dependent on many factors including the quantity of CNTs released, the physical properties of
CNTs (i.e. size and shape), surface functionality, dispersivity, the presence of impurities (Kushnir
and Sanden, 2008; Linkov et al.,  2009), and the environmental media for exposure (air, water or
soil).  Ultimately, the toxicity potential of CNTs  is not  only varied  based on the physical
properties of  CNTs, but  also based  on the  physico-chemical properties  of the  surrounding
environment and biological nature of cells that are exposed to the CNTs.

If LCA is used exclusively to assess the environmental impacts of a CNT  nanoproduct, it will
adequately capture the issues related to resource management and climate change issues such as
acidification  and eutrophication. Shortcomings  will arise from  the  models underlying  the
characterization of impacts to human and ecosystem health because they have not been proven
with regard to their ability to account for the many factors discussed above that dictate the toxic
risk of CNTs.  These models have been developed using average environmental factors and lack
site-specificity to account for the influence of the factors  on  the behavior of CNTs. Even the
traditional risk assessment approach, which is based on  dose-response studies, may not account
for the cumulative risk of a CNT nanoproduct because it is typically a mass-based assessment and
neglects the influence of chemical properties. However, it does provide the opportunity to better
address the influence of local  environmental factors. Thus, the incorporation of a modified  risk-
based health impact model accounting for chemical factors under site-specific conditions into the
LCA framework would achieve a maximum understanding of the impacts of a CNT nanoproduct
to guide decisions at the local level.

3.1.1
The accuracy of assessment results obtained using the various methods  listed in Table  3-1 will
depend greatly on both the quantity and quality of data used to perform  them. Although the
required data sets vary from method  to method, they generally  fall into two categories, life cycle
processes or chemical risk. Understanding where to obtain this information  can greatly expedite
the assessment process.

Detailed process flow data forms the basis for numerous techniques, including LCA, CF, EIO-
LCA,  Eco-LCA, and  SMM. The  best sources of data for these  assessments  are actual
manufacturers and waste handlers. However, collection of data from primary sources can be time
consuming. In the absence of this data, preliminary Life Cycle Inventories (LCIs) can be built
using any of the  following sources: existing LCI databases  (e.g., U.S. LCI Database, Ecoinvent,
Gabi, etc.), journal articles, patents,  government reports, and manufacturers' websites. This data
will  be directly  applicable  for  LCA, CF,  and SMM. EIO-LCA  simplifies data collection
somewhat because LCI data can  be  taken directly from  pre-existing databases built on the  U.S.
input-output tables. However,  the most current version of this data represents 2002 and may not
accurately capture nanomanufacturing processes.  Eco-LCA requires thermodynamic conversion
factors to express  process data in a comparable form.  The values  should be available  in
engineering and physical chemistry reference books or scientific journals.
                                                                                      34

-------
Chemical risk data is primarily used for LCRA, but is also necessary to some extent when
calculating characterization  factors for the  human health  and ecotoxicity impact categories
tabulated in most  LCA methodologies. LCRA data types include exposure  factors,  chemical
toxicity,  and transport model parameters. These data must  be established  through rigorous
experimentation and require knowledge of site-specific environmental conditions.  Pollutant
release data can be obtained from government reports and databases (e.g., EPA's Toxic Release
Inventory (TRI)). Toxicity,  exposure, and transport data can typically be found in  scientific
journals. For existing chemicals other than nanocomponents, LCA characterization factors might
already be available with  suitable models  such as  USEtox (Rosenbaum et  al., 2008). For
nanocomponents, these  factors must also be  derived experimentally  and  may be  found  in
scientific literature.

3.1.2 Issues Related to Environmental Assessment Methods

Limited life cycle inventory data. A critical issue with  the assessment of any nanoproduct is the
limited availability of inventory data throughout the life cycle of the product. Products still under
development may have limited LCI data for the  manufacturing stage. LCI data for this phase may
also become outdated as the product and  manufacturing technologies evolve (Klopffer et al.,
2007). Because  many nanoproducts are only  beginning to enter the market, LCI  data may
especially be limited for the use and end-of-life (EOL)  stages. Furthermore, LCI databases (e.g.,
those available through Simapro and GaBi) are currently limited in scope and may not include
appropriate secondary data for nanotechnologies, such as material inputs, natural resource inputs,
and emission outputs (Khanna et al., 2007).
Based on the available data, a determination should be made as to whether it is possible to include
all stages of a product life cycle, discussed in further detail below. Developing  a flow diagram of
the product system, such as that in Figure 2-3,  can aid the data collection process. Such a flow
diagram  can help identify the data sources for each category and assist in making reasonable
assumptions when reconciling missing LCI data. These assumptions should be consistent with
those developed as part of the goal  and scoping phase. In addition, the input and output data that
are most likely to  change as the nanotechnology or manufacturing process evolves should  be
identified. For example, there may be a decrease in energy required to manufacture on a per unit
basis as production is scaled up. For existing technologies, data may already exist. However, for
emerging technologies, it may be necessary to gather data from laboratory experiments, models,
databases, or other studies.

Dynamic research environment. Since many nanoproducts are either in prototype development
or pre-production stage, they are constantly subjected to research improvements. Accordingly, the
LCI data with  respect to a particular nanoproduct may lose reliability quickly because the
nanoproducts may undergo a series of improvement iterations within a short period of time. For
instance, Bauer et  al. (Bauer et al., 2008), conducted  an LCA (hypothetical study) on 15 inch
carbon nanotube field emission  display CNT-FED   devices.  In 2007  when the  study was
conducted, it was initially believed  that the deposition of CNTs using chemical vapor deposition
(CVD) presented the most attractive  solution  for fabricating  the  cathode substrate. However,
further investigation indicated that the high  temperatures of CVD did  not allow use of glass
substrates in place of silica, meaning the CNT patterning could not be done using CVD (Fink and
Lee, 2005). Consequently, CNT cathode substrates are built using CNT pastes  (Chu et al., 2006)
or  inks  (Kordas  et al.,  2006).  Since the  process  of manufacturing CNT-FED  changed
significantly, the LCA study conducted by Bauer et al. (Bauer et al., 2008) using CVD grown
CNT cathode substrates lost its significance within three years.
                                                                                     35

-------
Access to confidential business information (CBI) and proprietary data.  LCI data may also be
limited by concerns over confidential business information (CBI) and other proprietary data from
firms upstream and downstream of the product developer. If this data is pivotal to an analysis, it
may be  more effective for a third-party or consultant to work with the companies in conducting
the assessment. This way, the third-party could sign non-disclosure agreements to aggregate and
protect any confidential data from the manufacturers. In addition, the extent to which upstream
and downstream companies could provide aggregated data to the product developers that does not
disclose  confidential information should  also  be explored.  Although this  would  generate
additional uncertainty, it would provide a starting point for generating LCI data from all the  life
cycle stages. As noted above, LCI  databases may  also provide inventory data on upstream
processes.
                    Anticipating the Challenges of Collecting LCI Data

   One of the hardest parts of conducting an LCA is obtaining the necessary data because
   the most accurate data often requires  direct  knowledge  of industrial products  and
   processes.  Unfortunately, this data is typically considered a "trade secret" by industrial
   sources and is not readily available.  So how  does one reach out to companies  and
   persuade them to look beyond their initial fears and contribute their knowledge to a life
   cycle inventory (LCI)? I was forced to ask myself this same question when I decided to
   perform an LCA of a popular nanoproduct. The following is a summary of what I learned
   by going through the process.

   The  current  climate  of  the  nanotechnology  industry  is  extremely cut-throat as
   nanomanufacturers seek to establish themselves as major players  in the nanocomponent
   market. Their livelihood is dependent upon protecting  their  manufacturing processes
   because these processes are what define the company. In  order to get companies to open
   up about their process, you as an LCA practitioner must first gain their trust.  This can be
   a difficult process, but one that ultimately begins with the mindset of the practitioner.

   Instead of viewing companies  as  mere sources for data,  I tried to see the LCA process
   from  their perspective and identify potential outcomes of my work that could benefit their
   day-to-day  business operations  (i.e.  process  optimization, waste  reduction, etc.).  I
   familiarized myself with their goals and achievements to better justify how my project fit
   with their business strategy (i. e. environmental stewardship, community interaction, etc.).
   This allowed me to put together  a unique "sales pitch "  for each company  that invited
   them  to be an active part of an exciting project and not just one of a dozen data sources
   for a technology assessment. In addition, I avoided solicitation through random contact
   and meticulously identified a member of senior management that would best serve as a
   point of contact (i.e. vice president of manufacturing, plant manager, etc.). As a final  step
   in building trust, I offered to protect their data through agreements  of non-disclosure.
   While not every company I contacted was willing to help, these efforts over time did allow
   me to identify enough sources to complete the LCI for my project.

                                        D.  E. Meyer

                                        EPA National Risk Management Research
                                        Laboratory,  Cincinnati, Ohio
                                                                                       36

-------
Uncertainty  regarding  service  level  to the  customer.   The  eventual  function  of  the
nanotechnology and service level to the customer may not be known, which may limit the ability
to develop an appropriate functional unit. As discussed in Section 2.3, reasonable assumptions
may need to be made about the  eventual product use and anticipated service provided to the
customer to develop an appropriate functional unit. If it is not possible to develop a service-based
functional  unit and  different product alternatives  are  not  compared, then another basis for
organizing the data and for putting results into context should  be determined  (e.g., amount of
principle inputs and outputs).

Development of new decision rules.  Decision rules commonly used for other technologies may
not be suitable for nanotechnologies. Typically a mass-based cut-off is used to determine which
materials  to  include  in  the  product system.  However, given the  small  scale  and  mass  of
nanomaterials, this may  not be appropriate. Therefore, decision rules  should also  include
materials that are of known  or suspected environmental and energy significance regardless of
size.

Uncertainty regarding upstream  resource use.  Nanoproducts are frequently high-tech devices
that require rare metals for production.  If there is  reason to believe that scarcity will drive up
material costs, this factor should be included in the sensitivity analysis. If the uncertainty is too
great  or is unknown, then a qualitative  discussion  of the potential impacts  should be included.
Removing  a resource from a region also impacts resource availability for local inhabitants and
typically is further exacerbated by low initial reserves. In such cases, both the impact of mining
operations on the effected population, as well as any impacts caused by the loss of the extracted
materials should be considered.

Limited toxicity data. There is very limited quantifiable data on the toxicity of nanotechnologies
to human health and the  environment (Khanna et al., 2007; Klopffer et al., 2007). There are
several options to consider when addressing this  issue. Toxicity data may be  gathered from
laboratory  experiments, models,  databases, or  other studies (Walsh  and Medley, 2007). For
example,  a study by Sengul (Sengul  et al.,  2008)  provides  a comprehensive review  of
technologies  used  to  manufacture  nanomaterials and  associated  environmental  impacts.
Currently, the Organisation for Economic Co-operation and Development (OECD) is testing a
representative set of manufactured nanomaterials for human health and environmental effects,
including:

    •   Fullerenes (C60),                              • Cerium oxide,
    •   Single-walled carbon nanotubes (SWCNTs),      • Zinc oxide,
    •   Multi-walled carbon nanotubes (MWCNTs),      • Silicon dioxide,Dendrimers,
    •   Silver nanoparticles,                           • Nanoclays , and
    •   Iron nanoparticles,                             • Gold nanoparticles
    .   Titanium dioxide,                               (OECD, 2008)
    •   Aluminium oxide,


When considering data from toxicity studies, it is important to ensure data relevance and accuracy
through the review of important qualifiers, key assumptions, and material properties (Bell, 2007).
In addition, when extrapolating data from laboratory experiments,  it is important to consider the
limitations of the studies and whether they may  be reasonably extrapolated  (e.g., studies
extrapolated from animals to  humans) (Bell, 2007). Depending on the available data, it may also
be  appropriate  to  conduct a threshold  analysis to determine the toxicity level  at which the
nanomaterial becomes a significant contributor to the overall impacts (Klopffer et al., 2007).
                                                                                       37

-------
Similarly, a scenario-based assessment or worst-case scenario assessment may be conducted, in
which the impact potential of the nanomaterials is assumed to be as high as that of the most toxic
materials  (Klopffer et al., 2007).  If data are not available, a qualitative assessment based on
physical and chemical properties that may influence toxicity may be conducted. For example, as
discussed in Section 2.1,  changes  in  the  size  and  shape of a nanomaterial  may influence
deposition patterns in the human respiratory tract.  Excluding human health and/or ecotoxicity
impacts (e.g.,  aquatic ecotoxicity, human health cancer) as part of the assessment and only
focusing  on environmental impacts (e.g., energy use, global warming, ozone depletion) until
additional data become available is also an option (Klopffer et al., 2007), although this exclusion
should be clearly identified in the final report.

Limited exposure data. Given uncertainty regarding the nature and extent of future commercial
use of the nanoproducts, there is limited exposure data with which to conduct risk assessments
(Ostertag and Husing, 2008; Walsh and Medley, 2007). Although toxicity is an important element
of risk assessment, it is only half of the analysis; the analysis of risk requires information about
both toxicity and exposure (Wiesner,  2006). Many  studies have assessed different exposure
pathways  relevant to nanomaterials throughout the life cycle stages (EPA-SPC, 2007). As noted
previously, if nanomaterials are embedded  as part of larger objects (e.g., nanocomposites and
nanocrystalline solids), they are  less dispersive and present lower risk than if they are used as free
nanoparticles,  nanorods  or nanofibers (Ostertag and  Husing, 2008).  Tools  for exposure
assessment  include monitoring, sampling,  and  modeling. However, these  approaches  have
limitations due to the unique characteristics of nanomaterials (EPA-SPC, 2007). For example,
unlike  typical exposure assessment, mass  may not be the appropriate  metric by which  to
characterize exposure. Instead, surface area may  offer a better measure for assessing exposures
(EPA-SPC, 2007).
Given  these limitations, conducting an initial "screening  level" exposure assessment to identify
gaps and "priority starting points"  for a thorough exposure assessment as data becomes available
should be considered (Ostertag and Husing, 2008) . As part of this approach four key steps should
be followed, including (1) defining system boundaries,  (2) identifying relevant product parts,
flows,  stocks, processes and emissions throughout the life cycle,  (3) characterizing the potential
of generation, emission, and exposure for  the areas identified  in  step 2, and  (4) identifying
priority points for the exposure assessment (Ostertag and Husing, 2008).  To identify potential
emission  sources,  relevant  literature  sources  that have  studied likely  areas of exposure  to
nanoparticles throughout the life cycle  should be reviewed (EPA-SPC, 2007; Park, 2009; Sengul
etal., 2008).

In addition, a scenario-based or worst-case  scenario assessment may be conducted (Klopffer et
al., 2007). For example, one study that modeled the  exposure of different nanoparticles  in the
environment looked at a "realistic-exposure scenario" and a "worse-case exposure scenario" to
develop a range of concentrations of the nanomaterials in the environment (Mueller and Nowack,
2008).

3.1.3  Challenges in Conducting Environmental Assessments

Like any other  new technology  under development, early application of nanotechnology is
surrounded by uncertainties about the potential effects on environmental and human health and
safety.  Based on the current state of knowledge, the risk is real for some nanotechnologies, but as
yet unquantifiable.  Current information regarding the environmental effects  and health risks
associated with nanomaterials  is  limited and assessing their  associated  risk  is  not  a
straightforward process. Advances are needed to  advance nanoproduct risk assessment and risk
management. Meanwhile, efforts should be  directed toward a better understanding of tradeoffs
                                                                                      38

-------
and finding superior risk management alternatives instead of setting a goal of estimating exact
risks and benefits.

Nanoproducts have many potential benefits to society with their development and deployment in
science,  engineering and technology. Their benefits, however, need  to be weighed against
potential impacts to environmental and public health. Adequate risk assessment processes  are
needed  to study  these  potential  impacts,  however,  manufacturers  need  to  realize  that
environmental assessments  should not be  limited to  risk of exposure to nano-components.  In
addition, manufacturers need to understand how the manufacture, use, and waste management of
the nanoproducts they make  could  potentially contribute to ongoing environmental problems
during their life cycle from  cradle to grave,  including recycling. For example, is there  the
potential for nanomaterials to impact global warming, stratospheric ozone depletion, acidification
(formation of acid rain) or eutrophication (the increase of chemical nutrients in water)?

LCA is  a powerful tool for  making holistic comparisons among possible or competing systems.
At the same time, not many studies on LCA of nanotechnology have been conducted; therefore
much of the needed information has to be extrapolated through experiences in other similar
industries. Furthermore, the rapid development of nanotechnologies and limited availability of
data makes full LCAs difficult to complete but easy to become outdated.

A two-day workshop on LCA of nanotechnological products (Klopffer et al., 2007) concluded
that the  current ISO-standard on LCA (14040) (ISO, 2006) applies to nanotechnological products
but also that some development is necessary. The following main issues were identified:
       There is no generic LCA of nanomaterials, just as there is no generic LCA of chemicals.
       The ISO-framework for LCA (ISO 14040:2006) is fully suitable to nanomaterials and
       nanoproducts, even  if data regarding the elementary flows  and  impacts  might be
       uncertain and scarce. Since environmental impacts of nanoproducts can occur in any life
       cycle stage, all stages of the life cycle of nanoproducts should be assessed in an LCA
       study.
       While  the ISO 14040 (ISO, 2006) framework is appropriate, a number of operational
       issues need to be addressed in more detail in the case of nanomaterials and nanoproducts.
       The main problem with LCA of nanomaterials and nanoproducts is the lack of data and
       understanding in certain areas.
       While  LCA brings major benefits and useful information, there are certain limits to its
       application and use, in particular with respect to the assessment of toxicity impacts and of
       large-scale impacts.
       Within future research, major  efforts  are  needed to  fully assess  potential  risks and
       environmental impacts of nanoproducts and materials  (not just those related to  LCA).
       There is a need for protocols and practical methodologies for toxicology studies, fate and
       transport studies and scaling approaches.
       International cooperation between Europe and the United  States, together with other
       partners, is needed in order to address these concerns.
       Further research is needed to gather missing relevant data and to develop  user-friendly
       eco-design screening tools, especially ones suitable for use by small and medium sized
       enterprises.
Uncovering and recognizing environmental costs associated with a product, process, system, or
facility  is  important for good management  decisions. Attaining such  goals  as reducing
                                                                                     39

-------
environmental expenses, increasing revenues, and improving environmental performance requires
paying attention to current, future, and potential environmental costs. How a company defines an
environmental cost depends on how it intends to use the information (e.g., cost allocation, capital
budgeting, process/product design, other management decisions) and the scale and scope of the
exercise. Moreover, it may not always  be  clear whether a cost or a saving, such as adopting an
energy efficient process, is classified as partly environmental and partly not. Whether or not a
cost is "environmental" is not critical; the goal is to ensure that relevant costs receive appropriate
attention (White et al., 1995).

Scale. Depending on corporate needs, interests, goals, and resources, environmental accounting
can be applied at different scales which  include the following:
    •   Individual process or group of processes (e.g., production lines)
    •   System (e.g., lighting, wastewater treatment, packaging)

    •   Product or product line
    •   Facility, department, or all facilities at a single location - regional/geographical groups of
       departments or facilities
    •   Corporate division, affiliate, or the entire company.
Scope. Whatever the scale, there is also an issue of scope. Scope refers to the types of costs that
are included. An initial  scope question is whether  the  economic assessment extends  beyond
conventional, internal costs to include potentially hidden, future, contingent, and often intangible
costs, such as image/relationship costs. Another scope  issue is whether a company intends to
consider only those costs that directly affect their bottom line financial profit or loss, or whether
they want to also recognize the costs that results from their activities but for which they are not
directly accountable. These are referred to as external  costs.
Environmental accounting terminology uses terms such as full, total, true and life cycle in order
to emphasize the fact that conventional approaches may be incomplete in that they overlook
important environmental costs  (and potential savings). In looking for and uncovering relevant
environmental costs,  decision  makers may want to apply one or more tools.  As the scope
becomes more expansive, firms may  find it more difficult to  assess  and measure  certain
environmental costs.
Conventional
Costs
Hidden
Costs
Contingent
Costs
Relationship/
Image Costs
Societal
Costs
     Easier to measure
More Difficult to Measure
             Figure 3-2. The Spectrum of Environmental Costs (White et al., 1995)
                                                                                       40

-------
The discussion in this section concentrates on conventional costs (raw materials, utilities, capital
goods, supplies, etc.) as well as environmental costs that are potentially overlooked in decision
making (see Figure 3-2.) This is where companies that are  starting to manufacture products in an
environmentally sustainable way typically begin.
Examples of Environmental Costs Incurred by Companies


Potentially Hidden Costs
Regulatory
Notification
Reporting
Monitoring/testing
Studies/modeling
Remediation
Recordkeeptng
Plans
Upfront
Site studies
Site preparation
Permitting
R&D
Engineering and
procurement
Installation
Training _ . , .
, conventional costs
Inspections
Manifesting
Labeling
Preparedness
Protective equipment
Medical surveillance
Environmental
insurance
Financial assurance
Pollution control
Spill response
Stormwater
management
Waste management
Taxes/fees
Capital equipment
Materials
Labor
Supplies
Utilities
Structures
Salvage value
Bach-End
Closure/
decommissioning
Disposal of inventory
Post-closure care
Site survey
Voluntary
(Beyond Compliance)
Community relations/
outreach
Monitoring/testing
Training
Audits
Qualifying suppliers
Reports (e.g.. annual
environmental reports)
insurance
Planning
Feasibility studies
Remediation
Recycling
Environmental studies
R&D
Habitat and wetland
protection
Landscaping
Other environmental
projects
Financial support to
environmental groups
and/or researchers
Coiningem COSES
Future compliance costs
Penalties/fines
Response to future
releases
linage
Corporate image
Relationship with
customers
Relationships with
investors
Relationship with insurers
Remediation
Property damage
Personal injury
damage
Lega! expenses
Natural resource
damages
Economic loss
damages
and Relationship Costs
Relaiiorsnp witn
professiora staff
Relatiorsnp witn
workers
Relatiorsnp witn
suppliers
Relationship with
lenders
Relationship with
host communities
Relationship with
regulators


(EPA 1995)
Life cycle costing (LCC) can be viewed as building on the ISO standards of LCA as a method for
calculating costs throughout the life cycle of a product.  LCC is based on the same  physical
product system used in LCA, and summarizes all costs associated with a product from inception,
to development, use, and disposition (see Figure 3-3). However, it is pertinent to know the life
cycle  costs from different perspectives, and so costs are typically separated into two segments:
                                                                                       41

-------
those paid by manufacturers and those paid by consumers or society (SETAC, 2009). There are
three types of cost categories to consider when conducting an LCC:
    1.  Internal  costs are those that are associated with real monetary flows. For example, it
       may include costs associated with raw material purchase, transportation, production, or
       incineration. Note that "internal costs" refers to costs that are directly covered by any
       stakeholder in the product system, including producers, consumers, end-of-life recyclers
       and landfill managers, and any other stakeholders involved in the product system.  All
       internal costs should be included in a LCC.
    2.  Anticipated internal costs are costs that do not yet exist, but are anticipated in the future
       (e.g., future disposal costs from anticipated regulations). Depending on the study's goal,
       anticipated internal costs may be included in a LCC.
    3.  External costs are costs outside of the product system boundary, such as  roadways for
       facility construction  or maintenance of a port. These costs should not be  included in a
       LCC.
Resources
Externa
Product
Boundary
Costs
+ 1
| Raw
>• Materials
1 Extraction
I
Revenue
y
Extern;
lities Extern
Costs
1 '
alities Exterr
Costs
r 1 <
alities Exterr
Costs
r i i
alities Exterr
Costs
r i <
C°steJ Materials | C°steJ Product | C°steJ Product | C°StSJ End of Lif
* Processing * Manufacture * Use * tnd-ot-Lit
Revenue^ JRevenue^ JRevenue^ J Revenue
\
Revenue
'
a lities Exterr
I
Revenue
' i
alities Exterr
I
Revenue
p i
alities Exterr
I
Revenue
1 \
alities Exterr
alities
i
'\

Final
Disposition
'
alities
           Figure 3-3. Life-Cycle Costing (as presented in SETAC (SETAC, 2009))
As with an LCA, LCC is a process-based method, accordingly boundaries must be set around the
product system.  To connect the analysis to the broader economy-level, LCC inputs and outputs
may be combined with an economic input/output database. The resulting hybrid analysis allows
the  interconnections between the product system and the rest of the economy to be accounted for
without compromising the level of detail.

Other methods such as total cost accounting (TCA) and total  cost of ownership  (TCO) may also
be used to account for the  costs of a product system. Although these methods do not consider
both internal and external (societal) costs, they may be beneficial for estimating the costs for
certain stages of LCC analysis. For example, total  cost accounting may be used to estimate
production costs for development of a nanoproduct from the manufacturer's perspective and total
cost of ownership may be applied to estimate the costs incurred by a consumer during the use,
maintenance, and end-of-life  stages. As in LCA, LCC can be applied to specific stages of the
product.
  Total cost accounting  (TCA) is concerned  only  with costs  to  the  company itself and
  examines all the costs that go into making a product, including activities from R&D to sales.
  TCA can encompass any combination of business activities.





.. — ^



Manuf



icturmg r







Marketing
1

                 Source: Finding Cost Effective Pollution Prevention Initiative (Global Environment Management Initiative)

-------
Table 3-2 summarizes the economic assessment methods most relevant to nanoproducts. The
table does not include non-hybrid economic input-output models, general equilibrium models, or
partial equilibrium models, because these tools  primarily assess meso- and economy-wide
changes, and are not as applicable to nanoproducts.
                     Table 3-2. Key Economic Assessment Methods
Method Description/ Scope/ Impacts Reference
Benefits Stage Measured
Life Cycle
Costing
(LCC)







Eco-
Efficiency
Analysis










Assesses the comprehensive
costs of a product throughout its
life cycle stages. LCC may apply
techniques such as Total Cost
Accounting, Activity-Based
Costing (ABC), or Total Cost of
Ownership to estimate and
allocate costs between
manufacturers, consumers, and
society.
Combines Environmental Life-
cycle Assessment with Total
Cost of Ownership to draw a
relationship between the
economic value of a product and
its environmental impacts.







Product/
micro level

All life cycle
stages





Product/
micro level

All life cycle
stages








Cost to
manufacturer and
consumer per
functional unit
(service of
product)




Ratio of
economic value
of a product
(from a consumer
perspective) to
the life cycle
environmental
impacts (e.g.,
energy and
material
consumption,
waste, air, and
water emissions).
(Hunkeler et
al., 2008;
SETAC,
2009)






(Saling et
al., 2002;
White etal.,
1995)









3.2.1 Data Sources for Economic Assessments

Life cycle costing requires data for not only the direct cost factors but also indirect costs, as well
as costs related to liability and less  tangible benefits. While it may be easier to obtain data on
direct costs, it may be more problematic to estimate potential future costs. Following are ways by
which these costs may be tracked.

3.2.2 Issues Related to Economic Assessment Methods

High  research and development costs. Nanoproducts are an emerging technology, and involve a
great  deal of research  and development. Even though  material  use during  research and
development may be low, the costs  during this phase often constitute a large amount of overall
product costs.  As a result, the economic assessment may include the  "knowledge" phase, e.g.
research  and development  and  acquisition  via the supply chain. Other  elements, such as
marketing activities, might also be included in the assessment (SETAC, 2009).

Fluctuating manufacturing and capital equipment costs. Costs associated with a nanoproduct
may fluctuate  as the  technology matures,  depending on several factors, including changes in
                                                                                   43

-------
technology design and resources needed for capital equipment. Given the  quickly  evolving
production processes for nanotechnologies, capital equipment may become outdated more rapidly
than for other technologies. Furthermore, some costs may be sensitive to production volume due
to economies of scale, while other costs are rather stable (SETAC, 2009). If capital equipment is
included use  of a hybrid economic input-output model would sufficiently capture impacts from
capital equipment. Alternatively, instead of  EIO data, if more  precise data  is  available, an
evaluation of capital equipment could be nested within the larger assessment when appropriate
and feasible.

Limited consumer and end-of-life cost data. In many cases, there is a high degree of uncertainty
for emerging  technologies regarding future costs during the use and end-of-life stages. If cost data
from the use  and end-of-life stages are limited, it still may be possible to conduct an assessment
solely from the manufacturer's perspective. Such an assessment would not be comprehensive, but
could still provide valuable information for an LCC. Cost data on use and end-of-life  could be
incorporated  into the assessment at a later  date when the product goes to market  or when
additional information becomes available (SETAC, 2009). Identification of costs that may change
over time or by region (e.g., energy costs) are also important as it may impact the economic value
of the product for the consumer (SETAC, 2009). In addition, it is possible to estimate future costs
based on costs for similar products. For example, the cost to recycle a  nano-based lithium ion
battery may  be similar to the cost  of recycling a  lithium-ion  battery that  does not contain
nanomaterials. Any information about the uncertainty surrounding the predictions of future costs
should be incorporated into an uncertainty analysis.

Uncertainty regarding product  lifetime and selecting an appropriate discount  rate.  Given
potential uncertainties in the lifetime of the product and/or when a product may come to market,
it may be difficult to determine an appropriate discount rate (SETAC, 2009). Discounting is used
to account for the time-value  of money and  brings future costs  to a present value. Therefore,
selection of an appropriate discount rate is crucial for estimating future costs (SETAC, 2009).
SETAC  (SETAC,  2009) provides guidelines for selecting  a discount  rate depending on the
stakeholder perspective (e.g., bond rate,  lending rate, or weighted average cost of capital). In
addition, a sensitivity analysis  should be considered to  assess the  impacts of a range of
appropriate discount rates.

Uncertainty regarding economic benefits/costs. Developing nanotechnologies may lead to great
benefits for society, such as improved  military or medical technology, more efficient use of
resources, and more reliable devices. However, these benefits might be accompanied by negative
consequences. For example, improved medical care could increase healthcare budgets, because
money spent  on healthcare ultimately succeeds only in shifting the expenses involved in keeping
individuals alive to a later date.  The longer people live, the more expensive it is to keep them
alive. Moreover, years added to the end of life are likely each to be quite expensive (Sparrow,
2009). There  is a great deal of uncertainty surrounding the potential effect that nanoproducts will
have on  the  economics of medical devices and health care costs, which makes it difficult to
incorporate these concerns into an economic or social assessment.  However, if there is reason to
believe that the nanoproduct in question will  affect a segment of the economy (e.g., healthcare),
it  may be worthwhile to  at least include a qualitative discussion of the  potential impacts.  A
qualitative assessment may be enough to facilitate a discussion  with  stakeholders during the
decision analysis phase of the assessment.

Delineation between external versus internal costs.   Given the high degree of uncertainty
regarding nanoproduct properties, potential environmental impacts, and future regulations, some
costs may be difficult to identify and  estimate, such as future  costs  of remediation, or the
                                                                                      44

-------
reclassification of waste  produced by  the  firm  to  hazardous (SETAC,  2009).  This issue  is
essentially one of externalities. Even though governments and society generally do not directly
cover costs in the product system, sometimes they do  indirectly  cover them  (e.g. costs to
construct and maintain roads near a production facility, impact of air emissions on human health).
These external costs should not be included in the  economic assessment to avoid double counting
impacts. However, if it's anticipated that an external cost may become internalized,  either through
taxes, fees, or new regulations, then it may be appropriate to include such costs in the study, and
to conduct a sensitivity analysis around them (SETAC, 2009).

Uncertainty regarding impacts on other economic sectors. If the nanoproduct has the potential
to develop into a general purpose technology, previous experience suggests that the effects on
productivity and economic growth could be  significant even though these may sometimes come
with a more significant time  lag (Palmberg et al., 2009). Given the high level of uncertainty
regarding the potential of various nanotechnologies to  be transformed  into general purpose
technologies, it may not be possible to  quantitatively assess the broader economic benefits and
impacts associated with a new nanoproduct. In  other words, it is very difficult to  predict if and
how an emerging technology will enable the development of other technologies (Palmberg et al.,
2009).

Even so, if there is some evidence that the nanoproduct in question will enable the development
of other products, it may be worthwhile to at least qualitatively consider the potential economic
implications, such as increased or new demand  for certain materials, job creation,  and economic
growth. For example,  increased demand for high-tech materials might increase the  number of
jobs available to the  high-tech workforce. The  bulk of nanotechnology firms  are based in
developed countries, particularly the United States, (Palmberg et al., 2009). Accordingly, many of
the newly created high-tech jobs might also be located in developed countries.
Uncertainty regarding impacts on incumbent firms.  Some nanotechnology-based advances will
build upon  or be  readily integrated with existing technologies manufactured by "incumbent"
firms.  In such cases, adjustment to the new technology by incumbent firms is expected to be
relatively smooth.  Other nanotechnology  based innovations will involve radically different
processes, in turn destroying the  competencies of existing firms with the potential of reallocation
of economic activity to different firms and regions of the nation or globe (Shea, 2005). While it
can  be difficult to determine  ex-ante the impact of an emerging  technology on existing
competencies, it may be possible to draw qualitative conclusions. Table 3-3 presents a summary
of observations that may aid in providing qualitative assessments regarding whether a nano-based
innovation is more or less likely to  result in decline in economic performance of the incumbent
firm (Shea, 2005).

                  Table 3-3. Example Factors Impacting Incumbent Firms
 Factors more likely to negatively
 impact incumbent firms:
   Nanotechnology-based innovations that
   are complex or where only limited
   information is available.
   Nanotechnology-based innovations that
   constitute a change to a process or
   component that is core to the whole
   product system.
   Bottom-up fabrication techniques (e. g.,
Factors less likely to negatively impact incumbent firms:
 Incumbent firm's existing technologies are reaching the
 limits of their life cycle curve, resulting in increased
 incentive to adopt new technologies.
 Incumbent firm has complementary assets to take
 advantage of the nano-based innovation. Complementary
 assets can range from technical knowledge to
 relationships with distributors and consumers.
 Incumbent firm engages in inter-firm cooperation to keep
                                                                                       45

-------
  based on emerging chemistry
  techniques) as opposed to top-down
  techniques (e. g., use of existing
  lithography techniques).	
abreast of nanotechnology-based developments.
Incumbent firm has a strong ability to exploit external
knowledge relevant to nanotechnology-based innovations.
3.2.3 Challenges in Conducting Economic Assessments

The term 'economic assessment" has many meanings and uses. It can refer solely to the costs that
directly influence a manufacturer's bottom-line (private costs) or it can be expanded to include
the costs to individuals, society and the environment (total costs). Specific economic assessment
issues or challenges vary depending on the scale of the application. Accounting for conventional,
private costs is a fairly straightforward  process. Life  cycle costs which do not have a direct
impact on a company's bottom-line, are harder to model. However, businesses will ultimately
benefit  from moving toward including  probabilistic  and difficult to estimate  costs in their
business decisions.

It is important to ensure that environmental impacts  are not "double counted" when applying the
impact models of LCA and LCC. To avoid double-counting costs, LCC avoids the monetization
of external costs that are environmental impacts not  paid for directly or indirectly by the product
manufacturer (e.g., the potential monetary costs caused by losses due to greenhouse gas emission
and the resulting global warming) (SETAC, 2009) . In addition, when conducting both a LCA and
LCC, it  is important to  use  equivalent system boundaries and  the  same functional unit
(SETAC, 2009). However, it is possible that some components or processes of a product system
would be included in one assessment and not another (or vice versa),  because some aspects could
have high costs but low material flow.  For example, while the cost of research and development
may be high, the material flow and costs may be relatively low once in full production mode. In
this case, research and development costs might be  included in the  LCC, but research and
development materials would not be included in the LCA.
3.3    Social Assessment Methods
While methods for assessing the social impacts of a product are not as mature as methods for
assessing  environmental and  economic  impacts, concerns  about social  responsibility  are
increasing and should not be overlooked (ISO, 2009; Jorgensen et al., 2008). Several initiatives
and organizations, as well as the international community, have given thought to the underlying
core values, principles, and standards that should guide social assessments, and researchers are
actively developing methods for applying these principles and standards to products.
Given the  nascent stage of social sustainability  assessment  and  the  lack of well-developed
methods, this section describes principles and standards relevant to social sustainability, as well
as the currently available methods for applying those standards. In addition, we describe potential
issues and decision factors that may arise when applying these standards to nanoproducts.
The International Organization for Standards (ISO) is in the process of developing the Guidance
on  Social  Responsibility,  which  "provides guidance  on the underlying principles  of social
responsibility" (ISO, 2009).  This document was released as  a draft international standard in
September 2009 and was advanced to  a final draft international standard in March 2010. Other
organizations have also developed internationally recognized standards for social accountability,
which can be used  in social  sustainability assessment. These include  Social Accountability
International's facility level SA8000 standard (SAI, 2008), AccountAbility's AA1000 enterprise
level series of standards (AccountAbility, 2008), and the Global Reporting Initiative's supply
                                                                                     46

-------
chain level reporting guidance (GRI, 2006). Researchers and analysts are currently working to
incorporate these principles and standards into assessment methods, including existing life cycle
assessment methods.
In an effort to create a central framework for adapting ISO 14040 and 14044 to social criteria, the
United Nations Environmental Program (UNEP) together with the Society for Environmental
Toxicology and Chemistry (SETAC) published the Guidelines for Social Life Cycle Assessment
of Products through the Life Cycle Initiative (UNEP, 2009a). The UNEP/SETAC guidelines pull
together and build upon much of the existing  literature to provide a framework for Social Life
Cycle Analysis (SLCA).  These  guidelines recommend that  impact categories be based  on
internationally accepted standards, such as  SA8000  (SAI, 2008) and  AA1000 (AccountAbility,
2008). The UNEP/SETAC guidelines also suggest following the requirements of the Voluntary
Quality Standard for SRI Research (CSRR-QS, 2010) to gather data on upstream suppliers.
The basic components of an SLCA are the same as for an LCA described in Section 3.1. There is
a difference, however, in the type of data collected. As illustrated in Figure 3-4, whereas a LCA
requires data on process and material flows, SLCA requires data on how companies in the supply
chain interact  with stakeholders, such as employees, local  residents, the  broader community,
society at large, and other persons affected by  the company's actions (Dreyer et al., 2006;
Hauschild et al.,  2008). Furthermore,  SLCAs must  typically  include geographic  information,
because social assessments must  take  into  account  conditions  specific  to companies and
geographic areas, such as worker safety, environmental justice, access to resources, etc (Dreyer et
al., 2006; Swarr, 2009). The local/regional nature  of some social impact pathways also creates a
high degree of uncertainty when calculating  impacts with methods built upon aggregate  data,
such as traditional LCA methods (Klopffer, 2008; Norris, 2006). As an alternative, certification
systems, such as fair trade programs, may be used to determine whether materials in the supply
chain meet certain standards,  such as fair wages  and safe and just working conditions (Norris,
2006). Norris introduced the Life Cycle Attribute Analysis (LCAA) to calculate the amount of
output from a supply chain that has an attribute of interest.  Attributes could  be  any quality
relevant to the assessment,  such as fair trade labels (Norris, 2006).


Raw Materials
Extraction
Company
H"
l»

r ~\
Materials
Processing
Company
ir
i^

Product
Manufacturing
Company
T
If

Use and
Maintenance
Company
ir
•»

^ -\
End of Life
(EOL) Actors
ir
Stakeholders (e.g. employees, community members, consumers)
V ^


          Figure 3-4. Social Life-Cycle Assessment (Dreyer and Hauschild, 2010)
Some impacts estimated as part of an environmental assessment may also be considered in a
social assessment, but care should be taken not to double count impacts. For example, human
health impacts (e.g., cancer human toxicity) for both the public and workers may be considered a
social impact. If an impact is included as part of the social assessment, then it should not also be
included as part of the environmental assessment.
Social Impact Assessment (SIA) and Social Assessment (SA) are other methods commonly used
to  analyze  social   impacts.   However,  these  methods  were  developed  to  complement
Environmental Impact Assessment, and apply  more to larger  scale meso  and economy-wide
projects. The United States Department of Agriculture describes  SA as "the basis for identifying
and forecasting consequences of possible projects or policies" (Alan et al., 2003). Similarly, the
                                                                                     47

-------
SIA community describes SIA as a method for "analyzing, monitoring and managing the social
consequences of development" (Vanclay, 2003). While the methods of SIA apply more to larger
scale projects and interventions, the guiding core values and principles are equally applicable to
products. These principles are laid out by Vanclay (Vanclay, 2003). Table 3-4 shows the social
assessment methods most relevant to nanoproducts.
                       Table 3-4. Key Social Assessment Methods
Method Description/ Scope/ Impacts Measured Reference
Benefits Stage
Social Life-
Cycle
Analysis
(SLCA)





Life Cycle
Attribute
Analysis





SocioEco-
Efficiency-
Analysis




Assesses the
social aspects of
products and
their impacts
along their life
cycle.



Calculates the
amount of total
output that has
the attribute of
interest (e.g. fair
trade
certification)

Provides an
assessment of
societal impacts,
as well as
environmental
impacts and
economic costs.
Product/micro
level

All life cycle
stages




Product/micro
level

All life cycle
stages



Product/micro
level
All life cycle
stages


Defined by the stakeholders.
For example:
• Human rights
• Working conditions
• Cultural heritage
• Poverty
• Disease
• Political conflict
• Indigenous rights
Amount or percent of output
that:
a) Has the attribute of
interest;
b) Lacks the attribute of
interest; or
c) Lacks data on attribute
status
• Environmental
• Economic
• Social (e.g., number of jobs,
number of working
accidents occurring during
production).

(Dreyeretal.,
2006; Jorgensen
etal.,2008;
Swarr, 2009;
UNEP, 2009a;
Weidema, 2006)



(Klopffer,
2008; Norris,
2006)





(BASF, 2011;
Klopffer, 2008)




In addition to the methods presented, some computable general equilibrium (CGE) models and
partial equilibrium models (PEMs) incorporate social aspects (Zamagni et al., 2009). However,
these models address the meso and economy-wide impacts, and are not expected to be relevant to
nanoproduct design and manufacture.

3.3.1 Data Sources for Social Assessments

Generally, practitioners of SLCA will need to incorporate a large share of qualitative data, since
models for interpreting the numeric information related to the social issues being addressed are
not well  developed. When numeric  data is  useful—for example,  in assessing the wages of a
particular enterprise—additional data will  still be  needed to address its meaning. For example,
compliance with minimum wage laws does not always mean that the wages are  livable. Often,
                                                                                    48

-------
data may need to be collected to represent site-specific situations, since databases for specific
social and socio-economic impacts are at a minimum (UNEP, 2009a).

The Social Hotspots Database, a project of New Earth, was created in 2007 by Catherine Benoit
and Greg Norris (Benoit and Norris, 2007). The idea began during their work on the Guidelines
for Social Life Cycle Assessment of Products, published by UNEP (UNEP, 2009b), when Benoit
and Norris noticed opportunities for improvement in Life Cycle databases. Most LCA tools lack
the ability to specify  the  geographical  location of production activities -information that is
essential for social impact assessments. The Social Hotspots Database  can play a role similar to
LCA databases in  assessing  product hotspots, but with the  added  benefit of geographical
precision and potential social impacts identification. The development of the database started in
September 2009.

The Social Hotspots Database allows for visibility in the supply chain by:

    •   Providing modeling of product life cycles by country specific sector.
    •   Providing estimates on where the people are in the product's supply chain and what
        specific risks and opportunities might affect them.
    •   Expressing quantitatively the share of a supply chain where specific hotspots are found.

It shows which country specific sectors represent the greatest share of worker hours in a given
supply chain, which ones are the most at risk of human rights and social issues,  and which ones
can represent business opportunities to implement positive changes in livelihood.

3.3.2                 to

Uncertainty regarding social impacts.   Like all emerging technologies, it will be  difficult to
assess the social impacts of nanoproducts before they have been on the market. Nanotechnologies
have the potential to deliver considerable benefits to society, but at the same time these products
may pose great risks (Walsh et al., 2008). Much of the risks that have concerned researchers to
date are issues of toxicity, yet there may be other risks, as well. For example, will disposal and/or
recycling of the nanoproducts be associated with poor working conditions? This raises further
questions  regarding the   scope  of  a social  assessment  done  from  the developer's  and
manufacturer's perspectives.  For example, to what  extent should such  assessments take into
account downstream practices that might be out of the manufacturer's control?  Although the
answers to the questions are not yet fully resolved, they should be considered even as part of a
qualitative  analysis (Dreyer et al., 2006).
Inadequate methods for addressing high  levels of uncertainty.   As described above, for
nanoproducts, there  are  likely to be multiple layers  of uncertainty.  As  a result, traditional
approaches to quantifying uncertainty, such as Monte Carlo analysis, might be inadequate (Seager
and Linkov,  2008). Probability distribution  under conditions of extreme model and boundary
uncertainty lack meaning (at best) or may lead to overconfidence (at worst). To address these
uncertainties  a scenario analysis could be conducted, thereby affording the ability to consider
possible future activities, events or even policy decisions, such as market expansions, resource
availability, use, technology improvements and production caps). This allows product developers
to explore sensitivities and possibilities while gaining a feel for trade-offs (Seager and Linkov,
2008).
                                                                                       49

-------
3.3.3             in

The conduct and implementation of social responsibility assessments is relatively new and poses
complex challenges. The choice of which approach is appropriate will be application specific.
Different contexts will represent different challenges and will need varying levels of assessment.
Some developed countries may already  cover many of Human Rights and Worker Rights
indicators and the application of the law may be well executed. However, developing countries
may not be at this same level. For example, in many cases, enterprises in developed countries are
not allowing freedom of association. Therefore, as part of the assessment, screening for minimum
compliance when thresholds exist, and possibly also to assess performance  beyond compliance
thresholds, is suggested. The elements that should be defined in the goal definition phase of the
assessment and accounted for in the interpretation phase of the study. Comparative studies must
be conducted with the same level of assessment (UNEP, 2009a).

-------
4. Assessing Sustainabilitv
In Section 3, we provided a summary of methods for evaluating the environmental, social and
economic impacts of the product system over its lifetime. The result of these assessments is an
inventory of indicators computed and compiled based upon the established sustainability criteria.
The data alone can provide some insights regarding the nanoproduct to include identifying the
processes with the greatest impacts, prioritizing approaches for impact  reduction and improving
the product system. However, it typically is very difficult to determine  which approach is better
due to complex tradeoffs. For example, a company may be considering three different synthetic
processes for manufacturing silica nanoparticles, each of which are technically and economically
feasible,  environmentally sound and socially responsible. Hence,  while data may be plentiful,
critical knowledge to guide decision making is scarce (Klimberg and Miori, 2010). Further, note
that while making decisions based  on one facet alone is challenging, when coupled with the
unique intricacies of handling multiple factors inherent in sustainability, the difficulty of decision
making skyrockets. In this section, formal decision analysis methods are discussed which may be
used within the context of sustainability to systematically organize information, make sense of the
problem and select the preferred option.
4.1    Evaluating Sustainability Criteria
To illustrate the complexity of decision making, we highlight some of the critical elements that
are important for assessing nanomaterials.   Key  characteristics include  agglomeration  and
aggregation, reactivity, physical form/shape, solubility, surface area, critical functional groups,
contaminant dissociation, bioavailability, bioaccumulation potential and  toxic potential (Tervonen
et al., 2009). These  elements are  primarily related  to  risk, however when assessing the
sustainability  of  nanoproduct production,  information  such  as energy  use, material use,
environmental impacts, human health impacts and  cost should also be considered.  Based upon
preferences, values and objectives, the criteria for comparing each alternative are established by
stakeholders and decision makers.   Figure 4-1 is a mapping of decision criteria considered by
Canis et al. (Cam's et al., 2010) in their comparative assessment of alternatives for single walled
carbon nanotube (SWCNT) synthesis. Note that life cycle impact  assessment (LCIA) score and
health risks are proxy measures which compile multiple indicators related to environmental and
human health impacts.  In the study,  data  were gathered to populate the measures for each
synthesis method by extrapolating from existing studies and expert opinion.   Even with data in
hand, making a decision also requires deciphering the views  of stakeholders who typically
have  diverse goals,  priorities and values. Hence, the question now becomes: How is one
alternative selected over another?
There are numerous studies that evaluate sustainability based on multiple criteria (e.g., (Bailey et
al., 2010; Eason et al., 2009; Halog and Manik, 2011; Hopton et al., 2011; Linkov et al., 2006;
Sikdar and Murray, 2010)), yet not all utilized a formal decision strategy. Further, while decisions
may be made from simplified approaches, such methods  tend to minimize  complexity  and
consequently lose some crucial details (e.g., uncertainty) necessary for making informed, high
quality decisions (Linkov et al.,  2004; McDaniels  et al., 1999). Merkhofer (Merkhofer, 1999)
describes the  characteristics  of  quality  decision making as  decisions that involve  the  key
stakeholders, contains relevant types and amount of information, identifies good alternatives,
provides  logically sound results and properly integrates the preferences of the decision makers.
Accordingly, it is critical that methods be implemented that have the ability to integrate pertinent
information  and help guide decision making.  Table  4-1  is a snippet of the list Merkhofer
(Merkhofer, 1999) presented which displays the scope of various tools that have been developed
to aid in evaluating, refining and selecting desirable alternatives.  The scope is defined by the
                                                                                      51

-------
           Criteria
            Energy
          Consumption
           (GWh/kg)
       Material Efficiency
          (% in mass)
          LCIA Score
          (EcoPoints)
           Cost ($/g)
           Health risks
                                               Alternatives
                                                                  HiPco
                                                                   CVD
Arc
                                                                  Laser
Figure 4-1. Mapping the Decision Criteria: Sample Criteria for Assessing Single-Walled
Carbon Nanotube Synthesis Processes (Canis et al., 2010) HIPCO: High Pressure
Carbon Monoxide; Arc: Arc Discharge; CVD: Chemical Vapor Deposition; Laser: Laser
Vaporization.
                                                                                   52

-------
Table 4-1. Sample Decision Making Tools (Merkhofer, 1999)

Tool
Analytical Hierarchy Process
Cost-Benefit Analysis
Decision Analysis
Environmental Impact
Assessment
Probabilistic Risk Assessment
Structured Voting
Classical Probability Models
Primary Uses
Problem
Definition







Assessing
health or
environmental
risks







Assessing
other risks







Determining
whether
action is
needed







Collecting
information







Screening
alternatives







Identifying
alternatives
	


Evaluating
Alternatives









Selecting
Options



	


Communicating
decisions







                                                                                        53

-------
primary uses within decision making. For example, while cost-benefit analysis may be used for
the full  gamut  of the  decision  making  process from  problem  definition  to communicating
decisions, Analytical Hierarchy Process is typically only used to aid in problem definition,
screening and evaluating alternatives and selecting options. Although, both decision analysis and
cost-benefit analysis cover the  scope of decision support needs, cost-benefit  analysis only truly
accounts for one of the pillars of sustainability, is primarily based on market prices and typically
has very little stakeholder interaction.
Decision theory is  a field of study that aids in formulating hypotheses based on values, risk,
uncertainty and tradeoffs. It relates to techniques to aid decision makers in determining the best
option when the outcomes are uncertain and the decision environment is unpredictable (Parsons
and  Wooldridge,  2002).  Decision  theory is  compartmentalized into two primary genres:
normative  and  descriptive.   While descriptive  decision  theory is rooted  in  experimental
psychology and deals with determining how and why people make decisions, normative decision
theory is prescriptive and relates to finding the best decision (Peterson, 2009). Decision analysis
(DA) is under the banner of decision theory and is a systematic approach to evaluating complex
problems and  enhancing the quality of decisions (Clemen and Reilly, 2001). The basic steps of
DA  include:  problem  identification,  information  gathering,  generating  possible  solutions,
evaluating and selecting solutions. The DA flowchart provided in Figure 4-2 expands these basic
steps to incorporate estimating baseline risk, decomposition and modeling, sensitivity analysis
and  a recursive  step to account for further  analysis when  developing and  analyzing new
alternatives.
Many researchers have noted the benefit of applying decision analysis to life cycle approaches
and  highlight  the similarities between the  two structures (Seppala et  al., 2002).  Accordingly,
when referring to Figure  1-2, note that rudiments of the decision analysis process are dispersed
throughout the framework. Since Sections 2 and 3 contain elements of the initial steps of decision
analysis, once we reach this section of the guidance structure, the product has been characterized
and stakeholders have been identified. In addition, the objectives, goal and scope of the study and
an inventory of indicator data have been compiled for each alternative. Hence, the focus  at this
point is to take the  indicators and use them to make informed decisions. The DA approach was
modified to reflect this  aim and  consists  of evaluating and interpreting impacts, determining
preferences and uncertainty, selecting the  most sustainable alternative and sensitivity analysis
(Figure 4-3).
At its core, DA aids in reconciling conflicting values, objectives and preferences with the aim of
reaching a wise, informed decision  (Keeney and Raiffa, 1976). More accurately, it relates to
handling problems  that  are characterized  by  complexity,  uncertainty, risks and  tradeoffs. As
described in Section 3, there is likely to be a great deal of uncertainty related to the sustainability
of emerging technologies. Characterizing and dealing with these  uncertainties will be a major
challenge in any decision putting a premium on decision analysis methods that handle uncertainty
in an explicit and  transparent  manner. The presence of uncertainty coupled with multiple
attributes describes what Keeny  and Raiffa (Keeney and Raiffa,  1976) denote as "the double
dichotomy of decision problems." Uncertainty is discussed further in Section 4.4.
                                                                                        54

-------
                              Define problem and
                               identify objectives
                             Estimate baseline risk
                              Identify alternatives
                              Collect information
                                  related to
                                 alternatives
                       Decompose and model the problem
                     1.  Model problem structure
                     2.  Model uncertainty
                     3.  Model preferences
                                 Choose the
                                best alternative
                                   Is further
                                   analysis
                                   needed?
                                  Implement
                                  the chosen
                                  alternative
Figure 4-2. Decision Analysis Flowchart (Clemen and Reilly, 2001; Merkhofer, 1999)
                                                                                   JJ

-------
                    Assessing
                  Sustainability
                                 \
                                    \
                                                Evaluate Sustainability
                                                      Criteria
                                       N.
                                                   Select the best
                                                     alternative
                                                  Sensitivity Analysis
                                                                   Yes
No

1


Implement the
chosen
alternative

                   Figure 4-3. Modified Decision Analysis (DA) Approach
Table  4-2  provides  information  (to include stakeholder input) on some  of the most common
decision making strategies. Ad hoc approaches typically have very limited stakeholder input and
contain criteria that often are not explicitly defined. In addition, alternatives are evaluated through
qualitative or semi-quantitative measures and the final selection is often not transparent. Further,
the weighting  schemes are  often  developed by the  decision maker and are not  sufficiently
justified. Cost-benefit analysis  (CBA) and  probabilistic risk assessment (PRA)  have explicit
evaluations schemes, a great deal of stakeholder input and are useful in many cases.  However,
these methods  relate primarily to cost and risk respectively, but are limited in their ability to
                                                                                         56

-------
 Table 4-2.  Comparison of Process Elements for Common Decision Making Tools (CALCAS, 2008; Linkov et al., 2006; Merkhofer, 1999)
 Define problems
 Generate alternatives
 Formulate criteria by which
 to judge alternatives
 Gather value judgments on
 relative importance of
[criteria		
 Rank/select final alternatives
Stakeholder input limited or non-
existent. Therefore, stakeholder
concerns may not be addressed
by alternatives
                              Alternatives are chosen by
                              decision-maker usually from pre^
                              existing choices with some
                              ejgjertjnrjut.	
Criteria by which to judge
alternatives are often not
explicitly considered and
defined.
Non-quantitative criteria
valuation weighted by decision-
maker		
Alternative often chosen based
on implicit weights in an opaque
manner
 Stakeholder input collected after the problem is
defined by decision-makers and experts.
Problem definition is possibly refined based on
stakeholder input.

.Alternatives are generated through formal
involvement of experts in more site-specific
manner.

Criteria and sub-criteria are often defined.
Quantitative criteria weights are sometimes
formulated by the decision-maker, but in a
poqrlyjustified_ma!lS§L	
                                                            Alternative chosen by aggregation of criteria
                                                            scores through weight of evidence discussions
                                                            or qualitative considerations.
Stakeholder input incorporated at beginning of problem
formulation stage. Often provides higher stakeholder
agreement on problem definition. Thus, proposed
solutions have a better chance at satisfying all
stakeholders.	
                                                                           Alternatives are generated through involvement of all
                                                                           stakeholders including experts. Involvement of all
                                                                           stakeholders increases likelihood of novel alternative
                                                                           generation.	
                                            Criteria and sub-criteria hierarchies are developed
                                            based on expert and
                                            stakeholder j udgment.
                                                   Evaluation of total expected costs vs.
                                                   total expected benefits; Criteria often
                                                   based on various economic meausre to
                                                   include: net present value, benefit,
                                                   benefit to cost ratio,  etc.
Quantitative criteria weights are obtained from decision-
makers and stakeholders.
                                            Alternative chosen by systematic, well-defined
                                            algorithms using criteria scores and weights.
Typically defined by decision makers
                                                   Alternatives often generated by a
                                                   limited group of stakeholders and
                                                   decision makers
Preferences are not necessarily made
explicit or considered
                                                   Based upon costs and benefits
 Strength
Simple and low cost
Systematic means of exploring and quantifying
risk; good documentation,quantifies
uncertainty, identifies threats
Ability to handle complex decisions with mutiple
criteria  and stakeholders with multiple viewpoints;
Decision making in concert with stakeholder values and
preferences; strong theorectical foundation; can handle
soft issues (e.g., social) and uncertainty
Strong theoretical foundation with tools
to aid in estimating (cost and benefits);
common unit of measure; helps
managers allocate limited resources;
not everything can be monetized
Weakness
                              Inflexible, can not handle
                              complexity or uncertainty, not
                              reproducible, no logic or audit
                              trail, limited stakeholder
                              involvement; therefore, not all
                              concernsconfiidemd	
                               Difficult, expensive and time consuming;
                               Possible inacuracies due to estimating and
                               assumptions on mechanisms that are not well
                               known leading to large uncertainties and
                               misleading results
                                            Typically time consuming
                                                   Often limited stakeholder interaction;
                                                   deals with net impacts and not who
                                                   pays the costs or reaps the benefits,
                                                   typically based on market prices and
                                                   not true preferences
                                                                                                                                                                                    57

-------
handle the broad aspects of sustainability. On the other hand, multi-criteria decision analysis
(MCDA)  can handle the  diverse  elements  of sustainability and affords the ability to  make
decisions  based upon active  stakeholder input from  the  beginning to  include  developing
objectives, criteria and alternatives.  Moreover, the weights are determined by querying both
decision makers and stakeholders to determine their preferences and the final alternatives selected
are determined systematically.
4,3            the
It is expected that the development of sustainable nanoproducts will involve a finite number of
alternative approaches. Accordingly,  this section provides an overview of methods designed to
handle the  evaluation  of multiple  criteria  (e.g.,  multiple impact  categories)  for multiple
alternatives. These methods are commonly referred to as multi-criteria decision analysis (MCDA)
(Cohon, 2003). The application of MCDA approaches has grown dramatically in the last twenty
years, particularly in the environmental arena and has been used to support decisions in multiple
areas including waste management, energy, sustainable engineering and manufacturing, natural
resources, energy, product comparisons, policy decisions and remediation (Huang et al 2011).
Hence, MCDA may provide a sound framework for assessing nanomaterials and products.

MCDA methods typically synthesize the criteria and select alternatives based upon techniques
such as:  ranking options, identifying  a single  optimal alternative,  incomplete ranking,  or
differentiating between acceptable and unacceptable alternatives (Kiker et al., 2005). The primary
MCDA approaches  are: Multiple attribute decision analysis (MADA) and Multiple Objective
Optimization (MOO). While, MADA is applied when the decision relates to evaluating a finite
set of alternatives, MOO is an operations research approach and is often implemented when the
possible  solution set  is  infinite  and/or contains  continuous  variables.  Examples of these
approaches may be found in Seppala (Seppala, 1999) and Azapagic and Clift (Azapagic and Clift,
1998).  Since, the framework will involve the assessment of a finite number of alternatives, we
will focus on MADA approaches. Further descriptions of these methods may be found in Stewart
(Stewart, 1992), Chen and Hwang (Chen et al., 1992), Norris and Marshall (Norris and Marshall,
1995) and Guitoni and Martel (Guitoni and Martel, 1998).

Some of the most relevant MADA approaches are summarized Table 4-3 and include outranking,
multi-attribute utility theory (MAUT) and analytical hierarchy process (AHP). These approaches
not only aid in selecting an alternative, but all of them except outranking can be used to evaluate
impacts. However, all of the methods are at least partially compensatory which indicates that low
scores  in one criterion can be compensated for by high scores  in another. When  performing a
sustainability assessment, it is important to note that in order for a product to be sustainable, it
must meet the established criteria within each of the three  sustainability pillars -  environment,
economy, and society. Accordingly, some would argue that while components of sustainability
are  integrated,  benefits in one aspect of sustainability do not compensate  for detriments in
another. Hence, care must be taken when using compensatory approaches. A list of some of the
commercially available decision support tools is provided by Azapagic and Perdan (Azapagic and
Perdan, 2005) and Kiker et al. (Kiker et al., 2005). Linkov et al. (Linkov et al., 2006; Linkov et
al., 2004) provide an extensive record of studies that applied MCDA and decision support tools in
environmental decision making.
                                                                                       58

-------
                               Table 4-3. Summary of Common Multi-Criteria Decision Analysis (MCDA) Methods
Method           Description
Elementary       Non compensatory method with no
                   requirement for quantitatively evaluating
                   criteria trade-offs; Ranking may be based
                   upon: the strength of the weakest or
                   strongest link, attributes meeting
                   predetermined thresholds, or best
                   performance on attributes with t
                                                                                                    Reference      Approaches
                                      No weighting is required
                                    Requires attributes to be on
                                    a common scale;
                           (Seppala et al,
                          2002; Yoon and
                           Hwang, 1995)
                Maximin, Maximax,
                Conjunctive, Disjunctive and
                lexicographic
Multi-Attribute
Utility Theory
(MAUT)
Compensatory method in which the
overall score for each alternative is based
on relative weights; Weights typically
determined by surveying stakeholders and
generated by utility functions
(1) Easier to compare alternatives
whose overall scores are expressed as
single numbers. (2) Choice of an
alternative can be transparent if highest
scoring alternative is chosen. (3)
Theoretically sound — based on
utilitarian philosophy (4) Many people
p refer to express net utility in non-
monetary terms.
(1) Maximization of utility
may not be important to
decision makers. (2)
Criteria weights obtained
through less rigorous
stakeholder surveys may
not accurately reflect
stakeholders' true
preferences. (3) Rigorous
stakeholder preference
elicitations are expensive.
  (Baker etal,
 2001; Clemen,
     1996;
  Wolfslehner,
     2008)
Multi-value utility theory
(MAUT), Simple Multi-
Attribute Rating Technique
(SMART)
Outranking
Partially compensatory methods that
determines the extent to which one
alternative dominates another. It allows
options to be classified as "incomparable"
(1) Does not require the reduction of all
criteria to a single unit. (2) Explicit
consideration of possibility that very
poor performance on a single criterion
may eliminate an alternative from
consideration, even if that criterion's
performance is compensated for by very
good performance on other criteria
performance (3) It is easy to  explain.
The algorithms used in
outranking are often
relatively complex and are
often not well understood
by decision-makers.
  (Kiker et al.,
 2005; Linkov et
 al., 2007; Naidu
  et al., 2008a;
   Seager and
 Linkov, 2008;
  Wolfslehner,
     2008)
Preference Ranking
Organization METHod for
Enrichment Evaluations
(PROMETHEE), Elimination Et
Choix Traduisant la Realite
(ELECTRE) (Kangas et al.
2001) and Novel Approach to
Imprecise Assessment and
Decision Environments
(NAIADE) software
Analytical
Hierarchy
Process (AHP)
Compensatory method in which the
overall score for each alternative based on
relative weights. Weights are generated
by a series of pair-wise comparisons. . It
is the most widely used approach of the
MCDA methods.
Surveying pairwise comparisons is easy
to implement
The weights obtained from
pairwise comparison are
strongly criticized for not
reflecting people's true
preferences
  (Huang etal.,
 2011; Kiker et
al., 2005; Linkov
   et al., 2007;
  Saaty, 1988;
   Seager and
 Linkov, 2008)
                                                                                                                                       AHP
                                                                                                                                                            59

-------
Additional details  on applying  MCDA approaches  for assessment of nanomaterials can be
obtained from literature (Cam's et al., 2010; Linkov and Seager, 2011;  Tervonen et al., 2009).
Outranking and aggregated weighted approaches (e.g., multi-attribute utility theory and analytical
hierarchy process) are discussed in the following sections.

4.3.1   Outranking

Outranking is  the  least  compensatory of  the methods  presented and  involves  pair-wise
comparison between alternatives (i.e., one criterion at a time) to determine the extent to which
one alternative is preferred  over the  other. For example, outranking methods may  select the
product alternative that meets the  minimum  requirements and  results  in the greatest impact
reductions  for the greatest number of criteria. Outranking methods are most appropriate when the
criteria are not easily  aggregated,  measurement scales  vary  over wide ranges,  or units are
incomparable (Kiker et  al., 2005; Linkov et al., 2007; Seager and Linkov, 2008). Outranking
methods include  Preference  Ranking  Organization METHod  for Enrichment  Evaluations
(PROMETHEE), Elimination Et Choix Traduisant la  Realite (ELECTRE) (Kangas et al., 2001)
and Novel Approach to  Imprecise Assessment and Decision Environments (NAIADE) software
(Naidu et al., 2008b)). While outranking methods are easy to explain, they  consist of complex
algorithms that are often not well understood by decision makers.
                               Example - Outranking (Naidu 2008)

  Sasikumar Naidu and other researchers from the University of Tennessee published a case study in 2008
  that uses an outranking decision analysis method to select nanoparticle synthesis processes . Naidu, et al.,
  evaluated the tradeoffs between three processes for silica nanoparticle synthesis: Sol-gel, a flame method
  involving tetraethylorthosilicate (TEOS) precursor and a flame method involving hexamethyldisiloxane
  (HMDSO) precursor. To facilitate the decision analysis, Naidu et al. (2008) used the Novel Approach to
  Imprecise Assessment and Decision Environments (NAIADE) software package. The analysis involves
  three steps: (a) pair wise comparison of alternatives, (b) aggregation of all comparison scores, and (c)
  ranking of alternatives.
  a. Pair wise comparisons of alternatives

  The outranking method begins by making pair-wise comparisons between alternatives, one criterion at a
  time. For a criterion j and a pair of alternatives  /' and /', the NAIADE software uses six membership
  functions to quantify the following comparisons:

     •   fi»(i,i')j (i much better than /')

     •   n>(i,i')j (i better than /')

     •   fia(i,i')j (i approximately equal to /')

     •   fi=(i,i')j (i very equal to /')

     •   fi<(i,i')j (i worse than /')

     •   /(«(/',/'), (/' much worse than /')

  These comparisons are scaled from 0 to 1, where 0  is not true at all, 1 is very true, and 0.5 is unclear.
                                                                                          60

-------
                               Example - Outranking (continued)

  b. Aggregation of comparison scores

  The comparison scores for all criteria are aggregated into a single preference intensity index, u*(U')'
                          •(i,i')j-a


  Where * stand for », >, ~, <, or «, and a is the threshold value below which values are not considered
  . Determining the threshold (a) is important to the analysis and will likely involve input from experts
  who are familiar with the given criteria.

  In addition to calculating the preference intensity index, the NAIADE software calculates the entropy of
  the intensity index, C*(/', /''), which indicates the variance in the indices that are above the threshold and
  near 0.5  (maximum fuzziness) . An entropy value of 1 means that all criteria give an exact indication
  (definitely credible or not credible) and an entropy value of 0 means that all criteria give an indication
  biased by maximum fuziness (//.(/,/')=0.5).


  c. Ranking of alternatives
  The intensity indices and associated entropies are combined to rank the alternatives. The final ranking
  is derived from two separate rankings, each one varying from 0 to 1. The first ranking, (|>+(i), indicates
  the extent to which /' is better than all other alternatives,  and the second ranking, (|>~(i), indicates the
  extent to which /' is worse than all other alternatives:
                      ^ (i, n) A Cv> (i, n) + M> (/, ri) A C, (/, n)}
            (0 = -
                    (M>:> (/', ri) A Cy> (i, n) + u> (i, n) A C> (/', n)]
            (0 = - - -i - -i -
4.3.2   Aggregating Weighted Impact Scores
As an  alternative  to outranking  methods,  the  normalized  and  weighted  impacts described
previously may be  aggregated.  Some aggregation methods  are multi-attribute utility  analysis
(MAUT) and analytical hierarchy process (AHP) and allow alternatives to be compared against
each other on a single scale. These methods are essentially  compensatory (Kiker et al., 2005;
Linkov et al., 2007) (MAUT more  so than AHP) and should be used with caution due to  the non-
compensatory nature of sustainability. Further, they can be used to elucidate stakeholder values
and determine weights (Cohon, 2003).

While, both MAUT and AHP rely on stakeholder values to determine the weights, they  differ in
the way that they derive weights for the criteria. Weighting in MAUT relies on utility functions,
while AHP utilizes pair-wise comparisons made by stakeholders. To elicit stakeholder values, the
AHP method asks stakeholders to make pair-wise comparisons between different criteria, which
                                                                                           61

-------
are then  translated  into  a weighting  scheme  via matrix  algebra  techniques.  Stochastic
Multiattribute Acceptability Analysis (SMAA)  is another method for deriving weights from
stakeholder values. Unlike MAUT and AHP, SMAA includes information about uncertainties in
stakeholder values. As a result, SMAA can  help to determine both a weighting scheme and the
sensitivity of results to uncertainties in stakeholder values (Seager and Linkov, 2008). Analytic
Network Process (ANP) is based on AHP, yet allows dependence between decision criteria. Once
the weights are determined, they are then multiplied by normalized impact scores (see Box).
Aggregating Weighted Impact Scores forNano Product X
In this example, an overall sustainability score is generated for three hypothetical synthetic processes by
summing the normalized and weighted scores for all criteria (lower is better). While Synthetic process #2
scores lower than process #3 on economic considerations and process #1 on social considerations, it has
the highest overall score. Note that these values are arbitrary, and for demonstration purposes only.
Selected Criteria Synthetic Process #1 Synthetic Process #2 Synthetic Process #3 1
Environmental
Energy use
Greenhouse gas
emissions
Potential aquatic toxicity
Economic
Consumer cost
Potential environmental
liabilities
Social
Job loss
Worker harm
Overall
0.8
0.25
0.31
0.24
0.53
0.35
0.18
1.34
0.89
0.45
2.67
1.59
0.36
0.25
0.98
1.03
0.46
0.57
0.7
0.46
0.24
3.32
1.11
0.2
0.45
0.46
1.11
0.53
0.58
0.64
0.28
0.36
2.86

Figure 4-4 is a plot of the hypothetical weighting schemes constructed in Canis et al. (2010).  As
illustrated, the weighting systems are quite  subjective and vary greatly based upon values and
priorities. For example, note that while costs are the primary concern for the manufacturer with
minor consideration  of health risks, the environmentalist does not consider cost and assigned
equal weightings to  health  risks, energy consumption and material  efficiency.   When these
weights were applied in  the MCDA, the order of the preferred options changed along with the
reasons for the desirability of the alternative. Consequently,  ranking  of the alternatives is
somewhat different for each scheme.  The  "Manufacturer",  "End User" and "Regulator" all
preferred  (ranked 1st) HIPCO, while the "Environmentalist" had a slightly higher preference for
the Laser alternative.  The rank of the remaining alternatives varied greatly (Canis et al., 2010).
This  exercise  not only  demonstrates the method  but also highlights the  fact  that  different
weighing  schemes may result in quite different decisions. Accordingly, care must be taken when
evaluating alternatives and  additional approaches (e.g., probabilistic  method and/or sensitivity
analysis) may be needed to compensate for stakeholder uncertainty.
                                                                                        62

-------
      100.00%
       90.00%
       80.00%
       70-00°/°
       60.00%
       50.00%
       40.00%
       30.00%
       20.00%
       10.00%
        0.00%
 Health Risks

I Costs

 Energy Consumption

I Material Efficiency

iLCIAScore
Figure 4-4. Comparison of Single Walled Carbon Nanotubes (SWCNT) Criteria Weightings
                               (adapted from Canis et al. 2010)
4.4     Uncertainty Analysis
As described in Section 3, there is a great deal  of uncertainty involved in the assessment of
nanoproducts to include inherent uncertainty related to data quality and availability, costs  and
impacts.  While  some  uncertainties are nearly impossible to estimate (e.g., level of market
penetration), experts may be queried to get a qualitative sense of this variability (Seppala et al.,
2002).  In an attempt to capture uncertainty, researchers  involved in the  SWCNT assessment
developed probabilistic estimations of the performance indicators (Canis et al., 2010). Data were
primarily gathered from literature surveys  and  used to develop triangular distributions for the
parameters. As in the case study, uncertainty is traditionally modeled by assigning probabilities
through expert testimony, theoretical modeling, fitting  empirical data and/or simulation (Clemen
and Reilly, 2001; Keeney and Raiffa, 1976). While some  level of uncertainty is present in  any
assessment, it is of particular importance for new and emerging technologies due to the many
unknowns related to such issues as service level to the customer, upstream resource use, limited
exposure data, market penetration, cost data, systematic impacts on other economic sectors,  etc.
As noted  in  Section 3, due to the multiple layers of uncertainty related  to nanotechnology,
traditional approaches may be  insufficient (Seager and Linkov, 2008). Hence, sensitivity and/or
scenario analysis may be more  appropriate.  Monte-carlo simulation is an approach not only for
investigating the impact of uncertainty, but also exploring various management scenarios. This
simulation approach is  based  upon  altering  variable  values  by  selecting  characteristic
distributions and computing results to determine the impact of changes in underlying variables.
Although sensitivity analysis provides insight on the impact of uncertainty, the range of possible
outcomes  and aids  in determining and  implementing management  decisions, much work is
needed on optimizing these methods (Basson, 1999).
4.5     Sensitivity Analysis  and Scenario Analysis

Sensitivity analysis is a method of determining the effect changes  in controllable variables have
on output response variables. In the context of DA, the aim of sensitivity analysis is to identify
                                                                                       63

-------
the input variables and values that may alter preferences for different alternatives. Accordingly, it
is  crucial that the sources and types of uncertainties are identified  and distinguished. While
probabilities may be defined to characterize some variables, others may require the use of random
sampling simulation and scenario analysis to get a sense of the range of uncertainty in the results.
Using this approach,  the impact of uncertainty in the variable on the viability of a particular
alternative may be explored by varying key controllable variables.

Scenario Analysis is the strategic process of evaluating alternatives by considering possible future
activities and can be used to assess what will happen (predictive), can happen (explorative) or
how a specified target may be reached (normative) (Borjeson et al., 2006; Hojer et al., 2008).
This technique enhances decision making by affording the ability to assess the performance and
impact of alternatives given potential events and outcomes. For example, alternatives may be
evaluated based  upon various levels of market  penetration,  resource availability, production
capacity or changing environmental regulation. Scenario analysis is used effectively in a number
of arenas to include military intelligence, as well  as finance to explore changes in the economy
under several growth scenarios (e.g., rapid, decline or moderate). It is also used in long term
planning and emphasizes the importance of systems thinking  when developing and evaluating
products, processes and systems. Because a system is typically comprised of many factors and
subsequent interactions,  scenario  analysis may highlight outcomes  that are counterintuitive,
previously unknown or unexpected.  The basic steps of scenario analysis include: determining
which factors will be considered and the scenarios to be evaluated for each factor, estimating the
outcomes and assigning the probabilities of various scenarios.  Although estimating outcomes
may involve adapting knowledge of existing systems,  it typically entails some type of modeling
or simulation  activity. The resulting outcomes are compared for the alternatives across the
scenarios. While scenario analysis is a valuable planning tool and is used to explore impacts as a
result of what may happen, there are no assurances of what will happen; of course, this is true of
any forecasting tool. Resources related forecasting and scenario analysis include, among others,
Eriksson and Ritchey (Erikkson and Ritchey, 2002), Konsult and Nilsson (Konsult and Nilsson,
2005), Markham and Palocsay (Markham and Palocsay, 2006)  , Eason et al (Eason et al., 2009),
Hojer et al. (2008), and Borjeson et al. (2006).


4.8

It is evident that the process of determining alternatives that satisfy the sustainability criteria will
require an iterative approach. As such, there is no illusion that ideal options will be determined on
first  pass. Further,   modifications  due  to  technology  enhancements,  regulations,  process
alterations, changing supply horizon or energy security issues may demand the need to adjust
viable options. Moreover, the results of the evaluation provide information  on resources, stages or
processes that may be candidates for redesign or replacement. However,  making improvements
with respect to one impact category of sustainability may adversely impact another category or
pillar. For example, consider the case where a non-renewable  feedstock or material is  replaced
with a renewable material that is processed in an upstream facility with a  history of poor on-site
worker  safety. Although this substitution may improve the environmental sustainability of the
product, it may  have a negative impact on  its social  sustainability.  Therefore, alternative
approaches developed to mitigate impacts must be assessed according to the sustainability criteria
and evaluated. While determining the impact of the substitution outlined above may require a full
re-assessment, if replacing one material with the other does not affect other aspects of the product
system, it may be appropriate to simply assess differences in impacts between the two materials.
Much like an adaptive management  approach, developing more sustainable  alternatives may
occur several times over the course of a product's lifetime and the number  of times this process is
                                                                                       64

-------
repeated depends on the need for and feasibility of developing alternatives. It is important to note
that DA approaches are meant to facilitate decisions, not replace decision makers. As such, after
conducting a decision analysis, it is advisable to discuss the results with stakeholders before
selecting and implementing an approach.
5. Conclusions	

By focusing on sustainability, product researchers, developers, and manufacturers can help ensure
that as nanoproducts advance, they realize their potential benefits to society without jeopardizing
the well-being of humans or the environment, in this generation and beyond. There are many
unknowns surrounding nanotechnologies and nanoproducts, both in terms of performance and
impacts on the environment, economy, and society. The preliminary framework presented in this
guidance should aid in better organizing and understanding the known life-cycle impacts, as well
as help product  developers prioritize new research to better understand the unknowns. This
document is intended to offer a starting point for assessing the sustainability of nanoproducts and
provides a summary of existing methods for assessing various aspects of sustainability. Further, it
highlights the critical elements needed for supporting sustainability  based decision making.
Feedback gathered from this report will be used for enhancement of the work, clarification of the
approach and prioritization of future research. Moreover, given that the fields of nanotechnology
and life cycle approaches are changing rapidly, this document will be reviewed and updated as
additional information becomes available.
6. Acknowledgements	

Our thanks go to Shanika Amarakoon and Brian Segal of Abt Associates who provided us with
the initial material for this document.  The team would also like to thank Kathy Hart (Office of
Chemical Safety and Pollution Prevention:  OCSPP), Jim Alwood (OCSPP), Kristan  Markey
(OCSPP), Tom Seager (Arizona State University), Igor Linkhov (US Army Corps of Engineers),
Mike Davis (National Center for Environmental Assessment) and his post docs: Patricia Gillespie
and Christy Powers) and Alessandra Zamagni (Italian National Agency for New Technologies,
Energy and the Environment) for their valuable comments, suggestions and insight.
                                                                                     65

-------
2003 Sustainable development in dynamic world: Transforming institutions, growth and
       quality of life The World Bank Washington DC, USA.
Aaltonen, K. 2011. Project stakeholder analysis as an environmental interpretation
       process. Int. J. Proj. Mgmt. 29, 165-183.
AccountAbility 2008 AA 1000 Assurance Standard 2008, pp. 1-28, AccountAbility,
       Washington D.C.
Alan, B.D., Ken, C.H., Anne, H.P. and Michael, A.T. 2003 A human dimensions
       framework: Guidelines for conducting social assessments, pp. 1-83, US
       Department of Agriculture, Asheville, NC, ISA.
Anex, R.P.  and Focht, W. 2002. Public transportation in life cycle assessment and risk
       assessment: a shared need. Risk. Anal. 22(5), 861-877.
Arnold, G.L., Abnbar, A.D., Barling, J. and Lyons, T.W. 2004. Molybdenum isotope
       evidence for widespread anoxia in mid proterozic oceans. Science 304, 87-90.
Azapagic, A. and Clift, R. 1998. Linear programming as a tool for life cycle assessment.
       The International Journal of Life Cycle Assessment 3(6), 305-316.
Azapagic, A., Millington, A. and Collett,  A. 2006. A methodology for integrating
       sustainability considerations into process design. Chemical Engineering Research
       and Design 84(6), 439-452.
Azapagic, A. and Perdan, S. 2005. An integrated sustainability decision-support
       framework: Methods and tools for problem analysis, Part II. International Journal
       of Sustainable Development and World Ecology 12(2), 112-131.
Bailey, J.A., Amyotte, P. and Khan, F.I. 2010. Agricultural application of life cycle
       iNdeX(LinX) for effective decision making. Journal of Cleaner Production 18(16-
       17), 1703-1713.
Baker, D., Bridges, D., Hunter, R., Johnson, G., Krupa, J., Murphy, J. and Sorenson, K.
       2001 Guide book to decision making methods, US Department of Energy
       Washington D.C.
BASF. What is the SEEBALANCE® ? 2011 [cited 2011 April 18th ]; Available from:
       http://www.basf.com/group/corporate/en/sustainability/eco-efficiency-
       analysis/seebalance.
Basson, L. (1999) Decision making for effective environmental management: Multiple
       criteria decision analysis and management of uncertainty University of Sydney,
       Sydney, Australia.
Bauer, C., Buchgeister, J., R.Hishier, Poganietz, W.R.,  Schebek, L.  and Warsen, J. 2008.
       Towards a framework for life cycle thinking in  the assessment of nanotechnology.
       Journal of Cleaner Production 16,  910-926.
Baumann, H. and Tillman,  A.M. 2004. The hitch hiker's guide to LCA, Studentlitteratur,
       Lund, Sweden.
Bell, T.E. Understanding risk assessment  of nanotechnology. 2007 [cited 2011 April
       18th]; Available from: www.nano.gov/Understanding_Risk_Assessment.pdf
Benoit, C. andNorris, G. History of the project: The social hotspots database. 2007 [cited
       2011 May 4th]; Available from: http://www.socialhotspot.org/content/history-
       project.
Berkel, R.V. 2000 The sustainable development challenge for technology: will it work in
       time?, pp. 1-17, p2pays.org, Perth, Australia.
                                                                               66

-------
Bhushan, B. (ed) 2007. Handbook of nanotechnology, Springer-Verlag, New York
Borjeson, L., Hqjer, M., Dreborg, K.-H., Ekvall, T. and Finnveden, G. 2006. Scenario
       types and scenario techniques: Towards user's guide to scenarios. Futures 38(7),
       723-739.
Bryson, J.M. 2004. What to do when stakeholders matter Public Management Review
       6(1), 21-53.
BSI2008 Guide to PAS 2050: How to assess the carbon footprint of goods and services,
       p. 58, British Standards Institute, Surrey, UK.
CALCAS 2008 D10 SWOT analysis of concepts, methods and models potentially
       supporting life cycle analysis 037075. Schepelmann, P., Ritthoff, M.,  Santman, P.,
       Jeswani, H. and Azapagic, A. (eds), Wuppertal Institute for Climate,
       Environment, Energy, Manchester, UK.
Canis, L., Linkov, I.  and Seager, T.P. 2010. Application of stochastic multiattribute
       analysis to assessment of single walled carbon nanotube synthesis process. Env.
       Sci. Technol. 44(22), 8704-8711.
Chen, S.J., Hwang, C.L. and Hwang, F.P. 1992. Fuzzy multiple attribute decision
       making: methods and  applications, Springer-Verlag, New York.
Chu, K., Lei, W., Zhang, X., Di, Y., Chen, J. and Yang, X. 2006 Study of normal gate
       CNT-FED using HOP glass, pp. 437-438, IEEE, USA.
Clemen, R.T.  1996. Making hard decisions: An introduction to decision analysis,
       Duxbury Press, Pacific Grove, CA, USA.
Clemen, R.T.  and Reilly, T. 2001. Making hard decisions with decision tools suite,
       Duxbury Press, Pacific Grove, CA, USA.
CMU. Economic input-output life cycle assessment (EIO-LCA), creating a hybrid EIO-
       LCA and process-based  LCA. 2008a [cited 2011 April 18th ]; Available from:
       http://www.eiolca.net/LCAexperts/hybrid-lca-models.html.
CMU. EIO-LCA: Free, fast, easy life cycle assessment. 2008b [cited 2011 April 18th ];
       Available from: http://www.eiolca.net/.
Cohon, J.L. 2003. Multi objective programming and planning, Academic Press, New
       York, USA.
CSRR-QS.  Corporate sustainability and responsible research - Quality standard for SRI
       research. 2010 [cited 2011 April 18th ];  Available from: http://www.csrr-qs.org/.
Cuppen, E., Breukers, S., Hisschemoller, M. and Bergsma, E. 2010. Q methodology to
       select participants for  a stakeholder dialogue on energy options from biomass in
       the Netherlands. Ecological Economics 69, 579-591.
Davidson and Julie, L. 2001. Taking sides: Clashing views on controversial
       environmental issues Goldfarb, T.D. (ed), pp. 294-296, Dushkin/McGraw Hill,
       Hightstown, NJ.
Davies, J.C. 2009 Oversight of next generation nanotechnology pp. 1-48, Woodrow
       Wilson International Center for Scholars, Washington DC, USA.
Dreyer, L.C. and Hauschild, M.Z. 2010. Characterization  of social impacts in LCA. Part
       1: Development of indicators for labour rights The International Journal of Life
       Cycle Assessment 15(3), 247-259.
Dreyer, L.C., Hauschild, M.Z. and Schierbeck, J. 2006. A framework for social life cycle
       impact assessment The International Journal of Life Cycle Assessment 11(2), 88-
       97.
                                                                              67

-------
Eason, T., Hans, C. and Yaw, O. 2009. A systematic approach to assessing the
       sustainability of the renewable energy standard (RES) unhder the proposed
       american renewable energy act (H.R. 890) International Journal of Global Energy
       32(1-2), 139-159.
ED-DuPont 2007. Nano Risk Framework, Environmental Defense-DuPont Nano
       Partnership.
EPA-SPC 2007 Nanotechnology white paper, pp. 1-136, US Environmental Protection
       Agency, Washington DC, USA.
EPA 2006 Life cycle assessment: Principles and practice EPA/600/R-06/060, pp. 1-88,
       Scientific Applications International Corporation, Reston, VA, USA.
EPA. Risk assessment: Basic information. 2010 [cited 2011 April 18th]; Available from:
       http://www.epa.gov/risk/basicinformati on.htm#arisk.
Erikkson, T. and Ritchey, T. 2002 Developing scenario laboratories with computer aided
       morphological analysis pp.  1-8, London, UK.
Fiksel, J. 2006. A framework for sustainable materials management. JOM Journal of the
       Minerals, Metals and Materials Society 58(8), 15-22.
Fink and Lee, R. 2005 Bead blast activation of carbon nanotube cathode. Office, U.S.P.
       (ed), pp. 1-10, Nano Proprietery Inc USA.
Freeman, R.E. 1984. Strategic management: A stakeholder approach Cambridge
       University Press, Cambridge, UK.
Gooday, A.J., Jorissen, F., Levin, L.A., Middelburg, J.J., Naqvi, S.W.A., Rabalais, N.N.,
       Scranton, M. and Zhang, J.  2009. Historical records of coastal eutrophication-
       induced hypoxia Biogeosciences Discuss 6, 2567-2568.
GRI2006 Sustainability reporting guidelines, pp. 1-45, Global Reporting Initiative,
       Amsterdam, The Netherlands.
Guitoni, A. and Martel, J.M. 1998. Tentative guidelines to help choosing an appropriate
       MCDA method. European Journal of Operational Research 109(2), 501-521.
Halog, A. and Manik, Y. 2011. Advancing integrated systems modelling framework for
       life cycle sustainability assessment. Sustainability 3(2), 466-499.
Hansen, H.O., Gerloff, G.C. and Skoog, F. 1954. Cobalt as essential element for blue
       green algae. Physiologia Plantarum 7, 665-675.
Hart, K. Lithium ion batteries and nanotechnology partnership. 2009 [cited 2011 May
       4th]; Available from: http://www.epa.gov/dfe/pubs/projects/lbnp/index.htm.
Hauschild, M.Z., Dreyer, L.C. and  Jorgensen, A. 2008. Assessing social impacts in a life
       cycle perspective - lessons learned. CIRP Annals -  Manufacturing Technology 57,
       21-24.
Helland, A., Scheringer, M., Siegrist, M., Kastenholz, M.,  Wiek, A. and Scholz, R.W.
       2007. Risk assessment of engineered nanomaterials: A survey of industrial
       approaches. Environ Sci Tech 42(2), 640-646.
Hirschberg, S., Bauer, C., Burgherr, P., Dones, R., Schenler, W., Bachmann, T.  and
       Carrera, D.G. 2007 Environmental, economic and social criteria and indicators for
       sustainability assessment of energy technologies 502687/D3.1-RS 2b, pp. 1-31,
       Paul Scherrer Institut, Stuttgart, Germany.
Hqjer, M., Ahlroth, S., Dreborg, K.-H., Ekvall, T., Finnveden, G., Hjelm, O.,
       Hochschorner, E.,  Nilsson,  M. and Palm, V. 2008.  Scenarios in selected tools for
       environmental system analysis Journal of Cleaner Production 16(18), 1958-1970.
                                                                               68

-------
Hopton, M.E., Cabezas, H., Campbell, D.E., Eason, T., Garmestani, A.S., Heberling,
       M.T., Karunanithi, A.T., White, D. and Zanowick, M.A. 2011 A multidisciplinary
       approach to regional sustainability pp. 1-227, US Environmental Protection
       Agency, Washington DC, USA.
Huang, IB., Keisler, J. and Linkov, I. 2011. Multi-criteria decision analysis in
       environmental sciences: Ten years of applications and trends Science of Total
       Environment 409(19), 3578-3594.
Hunkeler, D., Lichtenvort, K. and Rebitzer, G. (eds) 2008. Environmental Life cycle
       costing, CRC Press, Boca Raton, FL.
ISO 2006 ISO 14040: Environmental management - life cycle assessment principles and
       framework pp. 1-91, International Standards Organization, Paris.
ISO 2009 ISO 26000: Guidance on social responsibility, pp. 1-91, International Standards
       Organization, Geneva.
ISO 2010 Nanotechnologies-Methodology for classification and categorization of
       nanomaterials p. 25, International Organization for Standardization, Geneva,
       Switzerland.
Jorgensen, A., Le  Bocq, A., Nazarkina, L. and Hauschild, M.  2008. Methodologies for
       social life cycle assessment. The International Journal of Life Cycle Assessment
       13(2), 96-103.
Journet, C., Maser, W.K., Bernier, P., Loiseau, A., chapelle, M.L.d.l., Lefrant, S.,
       Deniard, P., Lee, R. and Fischer, I.E. 1997. Large scale production of single-
       walled carbon nanotubes by electric arc discharge technique. Letters to Nature
       388,  756-757.
JRC 2010. International reference life cycle data system handbook (ILCD): General
       guidance for life cycle assessment, Luxembourg Publications, Office of European
       Union Ispra (VA) Italy
Kangas, A., Kangas, J. and Pykalainen, J. 2001. Outranking methods as tools in strategic
       natural resources planning. Silva Fennica 35(2), 215-227.
Kauffman, YJ. and Fraser, R.S. 1997. The effect of smoke particles on clouds and
       climate forcing.  Science 277, 1636-1639.
Keeney, R. and Raiffa, H. 1976. Decisions with multiple objectives: Preferences and
       value tradeoffs John Wiley & Sons New York
Khanna, V.,  Bakshi, B.R. and Lee, LJ. 2007 Life cycle energy analysis and
       environmental life cycle assessment of carbon nanofibers production pp.  128-133,
       IEEE, Orlando, FL, USA.
Kiker, G.A., Bridges, T.S., Varghese, A., Seager, P.T. and Linkov, I. 2005. Application
       of multicriteria decision analysis in environmental  decision making. Integr
       Environ Assess Manag.  1(2), 95-108.
Klimberg, R.K. and Miori, V. 2010. Back in business. ORMS Today 37(5), 22-27.
Klopffer, W. 2008. State-of-the-art in life cycle sustainability assessmen (LSCA). The
       International Journal of Life Cycle Assessment 13(2), 89-95.
Klopffer, W., Curran, M.A., Frankl, P., Heijungs, R., Kohler,  A. and Olsen, S.I. 2007
       Nanotechnology and life cycle assessment: A systems approach, pp. 1-37,
       Woodrow  Wilson International Center for Scholars Wasgington DC, USA.
                                                                               69

-------
Kohler, A.R., C.Som, Helland, A. and Gottschalk, F. 2008. Studying the potential release
       of carbon nanotubes throughout the application life cycle. Journal of Cleaner
       Production 16, 927-937.
Konsult, B. andNilsson, J.B. 2005 Scenario analysis including SWOT analysis: Abaltic
       gateway report, pp. 1-107, Tetra Plan A/S Baltic Gateway.
Kordas, K., Mustonen, T., Toth, G., Jantunen, H., Lajunen, M., Soldano, C., Talapatra,
       S., Kar, S., Vajtai, R. and Ajayan, P. 2006. InkJet printing of electrically
       conductive patterns of carbon nanotubes. Small 2, 1021-1025.
Kushnir, D. and Sanden, B.A. 2008. Energy requirements of carbon nanoparticle
       production. Journal of Industrial Ecology 12(3), 360-375.
Linkov, I, Satterstrom, F.K., Kiker, G.A., Batchelor, C., Bridges, T.S. and Ferguson, E.
       2006. From comparative risk assessment to multi-criteria decision  analysis and
       adaptive managment: Recent developments and applications. Environment
       International 32(8), 1072-1093.
Linkov, I, Satterstrom, F.K., Steevens, J., Ferguson, E. and Pleus, R.C. 2007. Multi
       criteria decision analysis and environmental risk assessment for nanomaterials
       Journal of Nanoparticle Research 9(4), 543-554.
Linkov, I. and Seager, P.T. 2011. Coupling multi-criteria decision analysis, life cycle
       assessment, and risk assessment for emerging threats. Env. Sci. Technol. 45(12),
       5068-5074.
Linkov, I, Steevens, J., Chappell, M., Tervonen, T., Figueira, J.R. and Merad, M. 2009.
       Nanomaterials: Risks and Benefits. Linkov,  I. and Steveens, J. (eds), pp. 179-191,
       Springer Science, Faro, Portugal.
Linkov, I, Varghese, A., Jamil, S., Seager, P.T., Kiker, G.A. and Bridges, T. 2004.
       Comparative Risk Assessment and Environmental Decision Making Linkov, I.
       and Ramadan, A. (eds), pp. 15-54, Kluwer Press, Amsterdam.
MacLean, R. 2009. The road to (environmental) hell is paved with good intentions.
       Environmental Quality Management 19(2), 93-99.
Markham, IS. and Palocsay,  S.W. 2006. Scenario analysis in spreadsheets with Excel's
       scenario tool. INFORMS Trans. Ed. 6(2), 23-31.
McDaniels, T.L., Gregory, R.S. and Fields, D. 1999. Democratizing risk management:
       Successful public involvement in local water management decisions. Risk
       Analysis 19(3), 497-510.
Merkhofer, M.W. 1999. Tools to aid Environmental Decision Making. Dale, V.H. and
       English, M.R. (eds), pp. 231-281,  Springer New York
Mitchell, R.K., Agle, B.R. and Wood, D.J. 1997. Toward a theory of stakeholder
       identification and salience: Defining the principle of who and what really counts.
       Academy of Management Review 22(4), 853-886.
Mueller, N.C. and Nowack, B. 2008. Exposure modelling of engineered nanoparticles in
       the environment Environ Sci Tech 42(12), 4447-4453.
Naidu, S., Sawhney, R. and Li, X. 2008a. A methodology for evaluation and selection of
       nanoparticle manufacturing processes based on sustainability metrics.  Environ Sci
       Tech 42(17), 6697-6702.
Naidu, S., Sawhney, R. and Li, X.P. 2008b. A methodology for evaluation and selection
       of nanoparticle manufacturing processes based  on sustainability metrics.
       Environmental Science & Technology 42(17), 6697-6702.
                                                                               70

-------
NNI2010 Risk management methods, & ethical, legal, and societal implications of
       nanotechnology report of the national nanotechnology initiative workshop. 2011.
       Heeter, L. (ed), pp. 1-76, National Science and Technology Council Arlington,
       VA.
NNI. Goal four objectives. 2011 [cited 2011 Sep 6th ]; Available from:
       http://www.nano.gov/goalfourobjectives
Norris, G.A. 2006. Social impacts in product life cycles: Towards life cycle attribute
       assessment. The International Journal of Life Cycle Assessment 11(1), 97-104.
Norris, G.A. and Marshall, H.E. 1995 Multiattribute decision analysis method for
       evaluating buildings and building systems pp. 1-82, NIST, Gaithersburg, MD,
       USA.
OECD 2005 Working group on waste prevention and recycling, OECD, Seoul, South
       Korea.
OECD. List of manufactured nanomaterials and list of endpoints for phase one of the
       OECD testing programme Series on the Safety of Manufactured Nanomaterials
       Number 6 2008 [cited 2011 April 18th ];  Available from:
       http://applil.oecd.org/olis/2008doc.nsf/linkto/env-jm-mono(2008)13.
Olapiriyakul, S. and Caudill, RJ. 2008 A framework for risk management and end-of-life
       (EOL) analysis for nanotechnology products; A case study of lithium ion
       batteries, pp. 1-8, IEEE, San Francisco, CA, USA.
ONGO 1987 Report of the world commission on environment and development, United
       Nations Oxford.
Ostertag, K. and Husing, B. 2008. Identification  of starting points for exposure
       assessment in the post use phase of nanomaterial containing products Journal of
       Cleaner Production 16(8-9), 938-948.
Palmberg, C., Dernis, H.  and Miguet, C. 2009 Nanotechnology: An overview based on
       indicators and statistics, pp. 1-112, OECD, Paris.
Park, J.Y. (2009) Occupational  exposure assessment for nanoparticles University of
       Minnesota, Minnesota.
Parsons, S. and Wooldridge, M. 2002. Game theory and decision theory in multi-agent
       systems. Journal of Autonomous Agents and Multi-Agent Systems 5(3), 243-254.
Petersen, E.J., Pinto, R.A., Landrum, P.P. and Jr, W.J.W. 2009. Influence of carbon
       nanotubes on pyrene bioaccumulation from contaminated soils by earthworms.
       Environ. Sci. Tech 43, 4181-4187.
Peterson, M. 2009. An introduction to decision theory, Cambridge University Press,
       Cambridge, UK.
Plata, D.L., Hart, A.J., Reddy, C.M. and Gschwend, P.M. 2009. Early evaluation
       potential environmental impacts of carbon nanotube synthesis by chemical vapor
       deposition. Environ. Sci. Technol 43, 8367-8373.
Reed, M.S., Graves, A., Dandy, N., Posthumus, H., Hubacek, K., Morris,  J., Prell,  C.,
       Quinn,  C.H. and Stringer, L.C. 2009. Who's in and why? A typology of
       stakeholder analysis methods for natural resource management. J.  Environmental
       Management 90,  1933-1949.
Roco, M.C. Nanoscale based emerging science and engineering. 2009 [cited 2011 April
       18th ]; Available  from: http://www.nanoinstitute.utah.edu/static-
                                                                              71

-------
       content/nanoinstitute/NNI_09-
       1015_Roco@NanoUtah_Emerging%20SE_70sl.pdf
Rosenbaum, R.K., Bachmann, T.M., Gold, L.S., Huijbregts, M.A.J., Jolliet, O., Juraske,
       R., Kohler, A., Larsen, H.F., MacLeod, M., Margni, M., McKone, I.E., Payet, J.,
       Schuhmacher, M., van de Meent, D. and Hauschild, M.Z. 2008. USEtox - The
       UNEP-SETAC toxicity model: recommended characterization factors for human
       toxicity and freshwater ecotoxicity in life cycle assessment International Journal
       of Life Cycle Assessment 13(7), 532-546.
Saaty, T.L. 1988. Decision making for leaders: The analytical hierarchy process for
       decisions in a complex world, RWS Publications, Pittsburgh, PA, USA.
SAL Social Accountability International. 2008 [cited 2011 April 18th]; Available from:
       http ://www. sa-intl. org/.
Saling, P., Kicherer, A., Dittrich-Kramer, B., Wittlinger, R., Zombik, W., Schmidt, I,
       Schott, W. and S., S. 2002. Eco-efficiency analysis by BASF: The method.
       International Journal of Life Cycle Assessment 7(4), 203-218.
Seager, T.P. and Linkov, I. 2008. Coupling multcriteria decision analysis and life cycle
       assessment for nanomaterials. Journal of Industrial Ecology 12(3), 282-285.
Sengul, H., Theis, T.L. and Ghosh, S. 2008. Towards sustainable nanoproducts. Journal
       of Industrial Ecology 12(3), 329-359.
Seppala, J. 1999. LCA Documents. Kloepffer, W. and Hutzinger, O. (eds), Eco-Informa
       Press,  Bayreuth, Germany.
Seppala, J., Basson, L. and Norris, G. 2002. Decision analysis frameworks for life cycle
       impact assessment. J. Ind. Ecol. 5, 45-68.
SET AC (ed) 1992. A conceptual framework for life cycle impact assessment SET AC
       Foundation for Environmental Education Washington DC, USA.
SET AC 2009. Environmental life cycle costing: A code of practice, SET AC Press,
       Brussels.
Shatkin, J.A. 2008. Informing environmental decision making by combining life cycle
       assessment and risk analysis. Journal of Industrial Ecology 12(3), 278-281.
Shea, C.M. 2005. Future management research directions in nanotechnology: A case
       study.  J. Eng. Technol. Manage. 22, 185-200.
Sikdar, S.K. and Murray, D.J. 2010. Energy and water sustainability: What do they mean
       and can we know when we achieved them? Environmental Research, Engineering
       and Management 2(52),  5-13.
Singh,  A., Lou, H.H., Pike,  R.W., Agboola, A.E., Li, X., Hopper, J.R. and Yaws, C.L.
       2008. Environmental impact assessment for potential continuous processes  for the
       production of carbon nanotubes. American Journal of Environmental Sciences
       4(5), 522-534.
Sinha, N., Ma, J. and Yeow, J.T.W. 2006. Carbon nanotube based sensors Journal of
       Nanoscience and Nanotechnology 6, 573-590.
Smith,  H. What innovation is? 2005 [cited 2011 May 4th ]; Available from:
       http://www.innovationtools.com/pdf/innovati on_update_2005.pdf.
Stewart, T.J. 1992. A critical survey on the status of multiple criteria decision making
       theory and practice.  OMEGA International Journal of Management Science 20,
       569-586.
                                                                              72

-------
Swarr, T.E. 2009. Societal life cycle assessment - could you repeat this question? The
      International Journal of Life Cycle Assessment 14(4), 285-289.
Tervonen, T., Linkov, I, Figueira, J., Steevens, J., Chappell, M. and Merad, M. 2009.
      Risk based classification system on nanomaterials Journal of Nanoparticle
      Research 11 (4), 757-766.
UNEP 2009a Guidelines for social life cycle assessment of products, pp.  1-104,
      UNEP/SETAC, Quebec, Canada.
UNEP 2009b Guidelines for social life cycle assessment of products, pp.  1-104, UNEP,
      Quebec, Canada.
USEPA 1976 Vapor phase organic pollutants: Volatile hydrocarbons and oxidation
      products EPA-600/1-75-005, pp.  1-670, United States Environmental Protection
      Agency, Washington DC.
USEPA 1977 Airborne particles: Medical and biological effects of environmental
      pollutants EPA-600/1-77-053, pp. 1-558, United States Environmental Protection
      Agency, Washington DC.
Vanclay, F. 2003. International principles for social impact assessment Impact
      assessment and project appraisal 21(1), 5-11.
Walsh, S., Balbus, J.M., Denison, R. and Florini, K. 2008. Nanotechnology: getting it
      right the first time. Journal of Cleaner Production 16(8-9), 1018-1020.
Walsh, S. and Medley, T. 2007 Environmental defense- DuPont nano partnership, pp. 1-
      87, DuPont, Wilmington, DE, USA.
Weidema, B.P. 2006. The integration of economic and social aspects in life cycle
      assessment The International Journal of Life Cycle Assessment 11(1), 89-96.
White, A.L., Savage, D.E., Brody, J., Cavander, D. and Lach, L. 1995 Environmental
      cost accounting for capital budgeting: A bench mark survey of management
      accountants, p. 70, Tellus Institute Arlington, VA, USA.
Wiedema, B.P. Input-output databases for life cycle assessment 2.0 LCA consultants.
      2010 [cited 2011  April 18th ]; Available from: http://www.lca-net.com/io-
      databases/.
Wiesner, M.R., Lowry, G.V., Alvarez, P., Dionysiou, D. and Biswas, P. 2006. Assessing
      the risks of manufactured nanomaterials Environ Sci Tech 40(14), 4336-4345.
Wolfslehner, B. (2008) Potentials and limitation of multi criteria analysis method in
      assessing sustainable forest management, Institue of Siviculture, University of
      Natural and Applied Life Sciences, Vienna.
WRI. The greenhouse gas protocol initiative. 2010 [cited 2011 April 18th ]; Available
      from: http://www.ghgprotocol.org/.
Yang, K., Zhu, L. and Xing, B. 2006. Adsorption of polycyclic aromatic hydrocarbons by
      carbon nanomaterials. Env. Sci. Technol. 40(6), 1855-1861.
Yoon, K.P. and Hwang, C.L. (eds) 1995. Multiple attribute decision making: An
      introduction.
Young, M. What is stakeholder analysis and why should you do it?" Principal consultant
      with transformed. 2008 [cited 2011 April 18th ]; Available from:
      http://www.docstoc.com/docs/2252199/What-is-stakeholder-analysis-and-why-
      should-you-do-it-What-is-a.
Zamagni, A.,  Buttol, P., Buonamici, R., Masoni, P., Guinee, J.B., Huppes, G., Heijungs,
      R., van der Voet,  E., Ekvall, T. and Rydberg, T. 2009 CALCAS: D20 blue paper
                                                                               73

-------
       on life cycle sustainability analysis, Institute of Environmental Sciences, Lieden
       University, Leiden, The Netherlands.
Zhang, Y., Baral, A. and Bakshi, B. 2010a. Accounting for Ecosystem Services in Life
       Cycle Assessment, Part II: Toward an Ecologically Based LCA.  Environmental
       Science and Technology 44, 2624-2631.
Zhang, Y., Baral, A. and Bakshi, B.R. 2010b. Accounting for ecosystem services in life
       cycle assessment, Part II: Toward an ecologically based LCA. Environ Sci Tech
       44(7), 2624-2631.
Zhang, Y., Singh, S. and Bakshi, B.R. 2010c. Accounting for ecosystem services in life
       cycle assessment, Part I: A critical review. Environ Sci  Tech 44(7), 2232-2242.
                                                                               74

-------
Appendix: List of Additional Resources

Government Organizations
    •   EPA Life Cycle Assessment Research: http://www.epa.gov/nrmrl/lcaccess/
    •   ORD's Nanotechnology LCA website: http://epa.gov/nanoscience/quickfinder/lifecycle.htm
    •   EPA Lean Home: http://www.epa.gov/lean/
    •   ORD Sustainable Technologies: http://www.epa.gov/nrmrl/std/
    •   ORD's Nanotechnology website: http://epa.gov/nanoscience/
    •   EPA Sustainable Futures: http://www.epa.gov/oppt/sf/
Public-Private Partnerships
    •   Woodrow Wilson Center - Nano and LCA
        http ://www. nanotechproj ect. org/file_download/files/NanoLC A_3.07 .pdf
    •   PEN - The Project on Emerging Nanotechnologies (Woodrow Wilson Center)
        http ://www. nanotechproj ect. org/publications/
    •   Life Cycle Initiative: http://jpl.estis.net/sites/lcinit/default.asp?site=lcinit
    •   CALCAS Co-ordination Action for innovation in Life-Cycle Analysis for Sustainability
        http://www.calcasproject.net
Non-Profit and Professional Organizations
    •   International Society for Industrial Ecology: http://www.is4ie.org/
    •   AIChE Sustainable Engineering Forum:
        http://www.aiche.org/DivisionsForums/ViewAll/SEF.aspx
    •   SET AC site on LCA: http://www.setac.org/node/32
    •   American Center for Life Cycle Assessment: http://www.lcacenter.org/
    •   Athena Institute: http://www.athenasmi.org/index.html
    •   Global Reporting Initiative: http://www.globalreporting.org
    •   Ceres : http://www.ceres.org/
    •   Wuppertal Institute: http://www.wupperinst.org/en
Industry Organization
    •   Center for Environmental Assessment of Products and Materials
        http://www.cpm.chalmers.se/links.htm
    •   Nanotechnologies Industry Association: http://www.nanotechia.org/
Academic Institutions
    •   The Center for Environmental Implications of Nanotechnology (CEIN)
        http://cein.cnsi.ucla.edu/pages/
    •   Carnegie Mellon Green Design publications
        http://www.ce.cmu.edu/GreenDesign/publications/index.html
    •   Ohio State Center for Resilience: http://resilience.eng.ohio-state.edu/CFR-site/tools.htm
    •   Arizona State Center for Nanotechnology in Society: http://cns.asu.edu/
    •   International Council on Nano technology: http://icon.rice.edu/
Magazines and Journals
    •   Nanowerk Magazine: http://www.nanowerk.com/
    •   Nanotechnology Now: http://www.nanotech-now.com/
    •   Nano Magazine: http://www.nanomagazine.co.uk/
                                                                                              75

-------