United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC 27711
EPA-452/R-95-002
July 1995
& EPA
Life-Cycle Impact Assessment:
A Conceptual Framework,
Key Issues, and Summary
of Existing Methods
-------
f!
-------
DISCLAIMER
The information in this document has been funded wholly by the United States
Environmental Protection Agency (EPA) under Contract No. 68-D2-0065 to Research Triangle
Institute. It has been subjected to peer and administrative review, and it has been approved for
publication as an EPA document. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use. Use of this methodology does not imply
EPA approval of the conclusions of any specific life-cycle impact assessment.
-------
-------
PREFACE
Life-cycle assessment (LCA) results can vary depending on how the sponsoring group
defines the goals and scope of the LCA and what methods and data are used to conduct the
assessment. There are an increasing number of organizations using LCA for a wide variety of
internal and external purposes. Conducting an LCA can be complex, and may require
significant data and information depending on the scope and goals of the study. For these
reasons it appears desireable to develop scientifically based guidelines for conducting LCAs.
Also, it is useful to provide technical reports to help users understand the status of LCA,
available methods, sources of data, and other information relevant to conducting LCAs.
The U.S. Environmental Protection Agency (EPA) has responded by supporting a
multioffice LCA program to develop technical information reports and, in some cases,
various guidelines. This multioffice program consists of representatives from the Office of
Research and Development (ORD), the Office of Air Quality Planning and Standards
(OAQPS), the Office of Solid Waste (OSW), and the Office of Pollution Prevention and
Toxics (OPPT). The LCA program uses a consensus-building approach, coordinating closely
with the Society of Environmental Toxicology and Chemistry (SETAC). Through the
organization of a series of workshops, SETAC has laid the groundwork for the development of
a technical framework for conducting LCAs.
The first in a series of EPA LCA methodological guidelines documents, Life Cycle
Assessment: Inventory Guidelines and Principles, and Life Cycle Design Manual:
Environmental Requirements and the Product System were released in early 1993.
Supplementary LCA documents including Life-Cycle Assessment: Public Data Sources for the
LCA Practitioner and Guidelines for Assessing the Quality of Life-Cycle Inventory Data were
released in April 1995. Ongoing EPA LCA projects include life-cycle inventory case studies
on residential carpeting systems, shop towels in industrial laundries, and solvent alternatives;
streamlined LCA methodology development; and product re-design through LCAs.
This document, which is a technical information report, includes the output from a two-
phased research approach on the impact assessment component of LCA. Phase I identified and
discussed key issues in the development of a conceptual framework for conducting an impact
assessment. Phase IX included documenting existing methods that exhibit potential for
application in impact assessment and identifying gaps hi the impact assessment methodology.
This document contains the combined output of Phases I and II.
111
-------
-------
ACKNOWLEDGMENTS
This report was prepared for EPA's OAQPS, under EPA Contract No. 68-W9-0080.
Charles French of OAQPS, Risk Exposure and Assessment Group, served as project officer.
Additional EPA guidance, reviews, and comments were provided by Mary Ann Curran
(ORD/RREL), Eun-Sook Goidel (OPPT), Eugene Lee (OSW), Lynda Wynn (OSW), Tim Ream
(formerly with OAQPS), and Tim Mohin (also formerly with OAQPS). Additional guidance,
reviews andcomments were provided by Bruce Vigon of Battelle. The technical work was
conducted under Research Triangle Institute Project Numbers 35U-5510-10 and 94U-5810-49.
Maria Bachteal, Ramona Logan, Judy Parsons, and Andrew Jessup provided editorial, word
processing, and graphics support for this report.
Peer reviewers included Paul Arbesman, Allied Signal; Derek Augood, Scientific
Certifications Systems; Bob Berkebile, American Institute of Architects; Terrie Boguski,
Franklin Associates, Ltd.; Michael Brown, Patagonia; Frank Consoli, Scott Paper Company;
Gary Davis, University of Tennessee; Richard Denison, Environmental Defense Fund; James
Fava Roy F. Westin, Inc.; Kate Gross, The Body Shop Inc.; Michael Harrass, Amoco
Corporation; Frances Irwin, World Wildlife Fund; Greg Keoleian, University of Michigan; John
Kusz, Safety-Kleen; Dave Mager, Green Seal Inc.; Beth Quay, The Coca-Cola Company; Athena
Sarafides NJ Department of Environmental Protection; Jacinthe Sequin, Environment Canada;
Karen Shapiro, Tellus Institute; Dave Snyder, Allied Fibers; Vincent Stanley, Patagonia; Donald
Walukas, Concurrent Technologies Corporation; and John Young, Hampshire Research Institute.
Additional reviewers included John Wilkens, DuPont; David Wheeler, The Body Shop, Inc.; and
Joel Ann Todd, The Scientific Consulting Group, Inc.
IV
-------
-------
CONTENTS
Page
Chapter
iii
Preface
iv
Acknowledgments "
1-1
1 Introduction
1.1 Key Impact Assessment Terms and Concepts
1 1.1 Inventory Item
1-4
1.1.2 Impact
1-6
1.1.3 Impact Assessment
1.2 Purpose of Life-Cycle Impact Assessment
1.2.1 Relationship Between Impact Assessment and Inventory ^
Analysis
1.2.2 Relationship Between Impact Assessment and Improvement ^
Assessment
1-9
1.3 Applications of Impact Assessment
131 Internal Applications
. . 1-12
1.3.2 External Applications
1 4 Current State of Impact Assessment Practice
1-14
1.5 Scope of This Document
2-1
2 Key Issues Surrounding Impact Assessment
2.1 Standardization of the Impact Assessment Framework 2-1
2.2 Use of Scoping in Impact Assessment
2 3 Communicating Uncertainty in Impact Assessment
2-5
231 Translating Inventory Items to Impacts
2-8
2.3.2 Impact Assessment Results
2.3.3 Methods of Uncertainty and Sensitivity Analysis 2-8
-------
8^CONTENTS (CONTINUED)
Chapter
2.4
Data Availability and Qiality Concerns in Impact Assessment ........ 2-10
2.4.1 Evaluating Data Availability ..................... 2_12
2.4.2 Evaluating Data Quality .............. 0 «',
j ........................ .. £-Lj
2.5 Incorporating Value Judgments into Impact Assessment ............. 2-14
2.6 Transparency ........ I ...........
j ""****...,.. ^i~lO
2.7 Expert Peer Review ____
I .................................... 2-16
2.8 Presentation of Impact Assessment Results ..... ............... 2-17
2.9 Unresolved Issues ..... .................
.......I, Jit~£,^
\
3 A Conceptual Framework for Impact Assessment ... i i
. *«*.«,. O"JL
3.1 Classification ......... | ............
[ ' ......................... 3-3
3.1.1 Developing Impacjt Networks ..................... 3_3
3.1.2 Classifying Inventory Items Within Impact Categories .......... 3.7
3. 1 .3 Example Classification Exercise of High-Density Polyethylene
(HDPE) Production .......... 10
[ ........................... J-o
3.2 Characterization ....... i .............
I *****»., j~y
3.2.1 Determining Assessment Endpoints ..................... 3_10
3.2.2 Selecting Measurerhent Endpoints ............... 3_12
3.2.3 Applying Characterization Models to Develop Impact
Descriptors ...... \ .............. _ , .
1 ........................... J-14
3.2.4 Impact Descriptors ........................... 3 17
3.3 Valuation ..........................
4 Existing Methods for Characterizing Impacts ..................... 4.j
4.1 Checklist Approach ..........................
4.2 Relative Magnitude Appro4ch ..................
i ................ ...... ^r-H-
4.3 Environmental Standards Relation ..................
VI
-------
CONTENTS (CONTINUED)
Chapter
Page
4-10
4.4 Impact Potentials
4.5 Critical Volume Approach
4.6 Environmental Priority Strategy (EPS) 4'16
4.7 Tellus Institute Human Health Hazard Ranking 4'20
4.8 Toxicity, Persistence, and Bioaccumulation Profile (TPBP) 4-25
4 27
4.9 Mackay Unit World Model ^"^
4.10 Canonical Environment Modeling 4"
4-32
4.11 Ecological Risk Assessment
4-36
4.12 Human Health Risk Assessment
5 Resource Depletion: Issues and Characterization Methods 5-1
5.1 Resource Depletion: Key Terms and Concepts
5.2 Sustainable Development and Its Relationship to Resource Depletion .... 5-3
5.3 Resource Depletion Models .
5.3.1 Resource Consumption Ratio 5~"
C Q
5.3.2 Resource Depletion Matrix J"°
6 Methods for Conducting Valuation 6'
6.1 Decision Analysis Using Multi-Attribute Utility Theory (MAUT) ....... 6-1
6.2 Analytic Hierarchy Process (AHP) 6
sr o
6.3 Modified Delphi Technique °"°
(\ 1 *?
6.4 Life-Cycle Costing
6.4.1 Hedonic Pricing 6"14
6.4.2 Contingent Valuation 6"16
6.4.3 Cost of Control Valuation 6"16
vn
-------
- CONTENTS {CONTINUED)
» Chapter
7 Integrated Methods for Impact Assessment 7.1
7.1 Impact Analysis Matrix (IAM) 7_j
7.2 The Environmental Prio:ity Strategy (EPS) Enviro-Accounting
Method 7^7
7.3 Integrated Manuf acturin i and Design Initiative (IMDI)
Envkonmentally Conscious Manufacturing (ECM) Life-Cycle
Analysis 7_g
7.4 Integrated Substance Chun Management 7-12
7.5 ECO-Rational Path Method (EPM) 7-14
8 Key Points and Future Research Needs 0 g-1
8.1 Summary of Key Points \ g_l
8.2 Potential Future Research Needs g_2
Appendix A: National Environmental Protection Policy Act (NEPA)
Environmental Assessment Procedures A-l
Appendix B: Additional Impact Assessment Methods B-l
Appendix C: Key Terms and Definitions C_1
Appendix D: Bibliography
vni
-------
FIGURES
Page
Number
1-1
1-2
2-1
2-2
2-3
3-1
3-2
3-3
3-4
3-5
3-6
3-7
4-1
4-2
5-1
5-2
6-1
6-2
6-3
7-1
7-2
7-3
7-4
LCA Conceptual Framework
Range of LCA Applications
Phased Approach to Impact Assessment Development
Impact Assessment Results Summary Chart
Example of a Possible Impact Assessment Results Format
Conceptual Framework for Life-Cycle Impact Assessment
Key Impact Assessment Decision Points
Example of Basic Network Using CO2
NO Example of Multiple Pathway Impact Network from a Single Inventory
Item
Example of Multiple Inventory Items Leading to Similar Impacts
Possible Impact Categories
Exercise for Choosing Characterization Models
Example Output from the Unit World Model
Conceptual Framework for Ecological Risk Assessment
The Life Cycle of Resources
Resource Depletion Matrix
Details of MAUT Water Pollution Effects Objectives
Example Framework for AHP Applied to Impact Assessment
Modified Delphi Technique
User-Level Impact Analysis Matrix for Ecosystem Impacts
Global-Level Impact Analysis Matrix for Ecosystem Impacts
Options Map for Integrated Substance Cham Management
Conceptual Framework for theEPM
. . 1-1
. 1-10
2-3
2-20
901
3-2
3-4
.. 3-5
. . . 3-6
3-7
... 3-9
3-18
4-30
A-^4
. ... 5-2
. . . 5-10
6-1}
6n
- 1
.... 6-9
7-4
/-H
7 4
. . . . . l~t*
7 1^
. 7-15
IX
-------
-------
TABLES
Number
Page
1-1 Key Impact Assessment Terms: Definitions, Examples, and Issues 1-3
1-2 Potential Internal and External Applications for Impact Assessment 1-11
1-3 Practitioner Survey of Impact Assessment Considerations 1-13
2-1 Possible Approaches, Advantages, and Disadvantages of Sensitivity
Analysis Methods for Impact Assessment 2-!°
2-2 Possible Approaches, Advantages, and Disadvantages of Uncertainty
Methods for Impact Assessment 2-11
2-3 Impact Assessment Data Needs 2'12
2-4 Comparison of Impact Assessment Results Presentation Methods 2-19
3-1 Example Inventory Analysis Data from the Manufacture of HOPE 3-10
3-2 Example Classification of Inventory Items Under Impact Categories for
HDPE Manufacturing
. 3-3 Suggested Criteria for Determining Assessment Endpoints 3-13
3-4 Characterization Models: Tiers of Complexity and Associated Data Needs 3-15
4-1 Summary of Methods to Characterize Impacts 4'2
4-2 Example Checklist for Ecosystem Impacts 4'3
4-3 Hypothetical Example of the Relative Magnitude Approach for Ecosystem
Impacts
4-4 Example Approach for Developing Environmental Standards Relation
Weights 4'8
4-5 State-of-the-Art Impact Potential Functions 4'12
4-6 Ozone Depletion Potential (ODP) of Select Halocarbon Gases 4-14
4-7 Example of the Critical Volume Approach
4-8 Select Environmental Indices Used in EPS
4-9 Example Environmental Load Values
4-10 Carcinogen Potency Factors and Isophorone Equivalents 4-21
4-11 Example RfDS for Noncarcinogenic Ranking 4'22
-------
TABLED (CONTINUED)
.Number
4-12
4-13
4-14
5-1
6-1
7-1
7-2
8-1
Example Human Health Impact Equivalency Ranking .................... 4-24
i
Hypothetical Example of TPBP Approach .......... .................... 4-26
Measures of Risk for Human Hedth Risk Assessment ............... . ..... 4-37
Example Calculations of Generkj Resource Consumption Ratios .............. 5-7
Example Results of Using the Mbdified Delphi Procedure for Comparing
Environmental Areas ........ j ...................................... 6-
TCA Substitute Study Inventory bata ................................... 7.3
i
Leopold Interaction Matrix . . . .| .......................... . ............ 7-6
i
Potential Future Needs for Impact Assessment Research .................... 8-3
XI
-------
CHAPTER 1
INTRODUCTION
Although a wide variety of impact assessment techniques have been mtegm! So various
disciplines, impact assessment in the context of Life-Cycle Assessment (LCA) is in its infancy.
A conceptual framework for conducting impact assessment has been established, but experts
have not yet reached a consensus on specific methods and procedures. This document outlines a
possible framework, discusses key issues, and summarizes existing methods for conducting
impact assessment. This document is not a guidance document, however, but rather a
compendium on the state of practice of impact assessment.
LCA is a holistic concept and methodology for evaluating the environmental and human
health burdens associated with a product, process, or activity. A complete LCA identifies inputs
and outputs; assesses the potential impacts of those inputs and outputs on ecosystems, human
health, and natural resources; and identifies opportunities for achieving improvements. The
basic life-cycle stages covered in LCA include raw materials acquisition, manufacturing,
use/reuse/maintenance, and recycling/waste management. The LCA approach consists of four
interrelated components, including impact assessment. These components are illustrated in
Figure 1-1 and explained below.
Improvement
Assessment
Impact
Assessment
Inventory
Analysis
Figure 1-1. LCA Conceptual Framework
1-1
-------
\» Goal definition and scoping: thp explanation ofithe study's purpose^and objectives;
/, -')' the identificatiomof the, product process, or activity of interest; the identification of the
:-. intended end-use study results; and.the key assumptions and methods employed.
Inventory analysis: theadentific'ation and quantification of raw materials and energy
, ' -i inputs, air emissions, water effluents, solid waste, and other life-cycle inputs:and
outputs. j
Impact assessment: the qualitative or quantitative classification, characterization, and
valuation of impacts of the inventory items to ecosystems, human health, and natural
resources, based on the results of Jan inventory analysis and application of various
methods and models to determine! significance of the inventory items.
Improvement assessment: the identification and evaluation of opportunities to
achieve improvements in product? and/or processes that result in reduced environ-
mental impacts, based on the results of an inventory analysis or impact assessment.
For almost 20 years, a wide variety [of organizations have conducted less-than-complete
LCAs. Most of these LCAs focused on the| inventory analysis component and stopped short of
analyzing impacts. This focus has enabled JLCA analysts to concentrate on developing and
refining procedures for building credible arjd reliable inventories of system inputs and outputs
and using these inventories for identifying possible improvement opportunities.
i
Formal procedures for conducting impact assessments have not yet been established.
The primary purpose of impact assessment in LCA is to assess the potential impacts resulting
from inputs and outputs quantified in the inventory analysis. By providing this information,
impact assessment can enhance the basis fojr evaluating and justifying the trade-offs among a
variety of inputs and outputs, as well as improvement options. As existing LCA and impact
assessment tools are refined and new ones developed, practitioners are expected to include more
impact assessments as part of LCAs.
1.1 KEY IMPACT ASSESSMENT TERMS AND CONCEPTS
In developing procedures for impact assessments, an important step is establishing a
common language. Fundamental terms usefi in impact assessment are often the subject of
confusion. For example, distinguishing between an inventory item and an impact is not always
easy. Although a common practice is to account for the amount of solid waste materials
produced by a system in an inventory analysis, it is not common to account for the amount of
natural habitat consumed to dispose of that solid waste. Some analysts might consider this
consumption of natural habitat an input, while others might consider it an impact. This section
focuses on key terms that distinguish between inventory item and impact and provides a working
1-2
-------
definition of impact assessment. Table 1-1 defines an inventory item and an impact and lists
examples and issues related to each term. Appendix C provides a glossary for other terms and
concepts used in this document.
1.1.1 Inventory Item
An inventory item is defined in this document as a quantitative measure of an energy or
raw material requirement, atmospheric emission, waterborne effluent, solid waste, or other input
or output of a particular product, process, or activity. In the past, an inventory item has referred
to more traditional inputs and outputs. For purposes of impact assessment, however, some more
nontraditional inputs and outputs (e.g., soil compaction, habitat use) associated with a production
system also may be appropriate to consider in the inventory analysis.
TABLE 1-1. KEY IMPACT ASSESSMENT TERMS: DEFINITIONS, EXAMPLES,
AND ISSUES
Definition
Examples
Issues Related to
Impact
Assessment
A quantitative measure of an energy or
raw material requirement, atmospheric
emission, waterborne effluent, solid
waste, or other quantifiable input or
output of a particular product, process,
or activity.
tons of SO2 emissions/year
tons of solid waste per day
biochemical oxygen demand
(BOD) released per unit of
production
tons of oil per unit of output
Interaction between different
releases may create new
substances that increase or mitigate
effects.
« Uncertainty of inventory data can
dramatically affect the results of
impact assessment.
An actual or potential change in an
environmental characteristic resulting
from interactions between the inventory
items, or components of a particular
product, process, or activity and the
environment.
acid precipitation
ozone depletion
soil erosion
habitat consumption
increased risk of cancer
Uncertainty is associated with the
existence, nature, and extent of
impacts in an uncontrolled
environmental setting.
Multiple impact pathways make
allocating impacts difficult.
Qualitative items, such as habitat
consumption and social welfare, are
difficult to determine and quantify.
1-3
-------
;One issue associated with inventory^tem&in the context,of LCA is that the composition
of the inventory is primarily based ontthe goals and scope of theistudy.. Because every input and
.output of a production system cannot typically be included in the inventory analysis^those ? ^
.included and/or excluded from the scope of the inventory analysis should be made transparent to
the user. In addition, practitioners may want to modify the goals and scope of the study to see
how that modification affects not only the composition of inputs and outputs captured in the
inventory analysis, but the overall LCA as well.
A second issue associated with inventory items is the synergistic nature of some
compounds. The synergistic effect of mixed'compounds may increase the concern about the
original compound or create a new compounld(s) that is not captured in the inventory. Such
synergistic compounds may have the potential to create combined impacts greater than those of
the individual releases. For example, the interaction between sulfur dioxide (SO2) and
paniculate matterwhere small particles transport SO2 and sometimes sulfuric acid deep into
the lungscan increase damage (Ott, 1987).; Synergistic compounds do not necessarily need to
be included in the inventory, but practitioners should nonetheless recognize this potential effect
and other factors (e.g., antagonistic effects, assimilative capacity) when drawing conclusions
based on LCA results.
Another issue is distinguishing between an inventory item and an impact. For instance,
should a largely qualitative item such as habitat consumption be included as an inventory item or
should it be treated as an impact? For purposes of this document inventory items are limited to
readily quantifiable "traditional" inputs and outputs (e.g., raw materials and energy use, air
emissions, waterborne effluent, solid waste). [Items such as habitat consumption that are not so
easily quantified and often involve value judgments are treated as impacts.
1.1.2 Impact
In the context of LCA, impact may bej defined as an actual or potential change in an
environmental characteristic that results from interactions between the components of a defined
system and the environment. Impacts relevant to impact assessment are categorized according to
whether they affect ecosystems, human health, and natural resources (SETAC, 1993). Although
they are not the primary focus of impact assessment, social welfare impacts may also be
considered to the extent that they indirectly may cause impacts to ecosystems, human health, and
natural resources. Currently, methods for handling social welfare impacts in the context of
impact assessment are not well developed.
1-4
-------
The Stressor Concept
Although not explicitly used in this document, the stressor concept has provided a
useful means of talking about the relationship between inventory items and subsequent
impacts. A stressor is defined as any physical, chemical, or biological entity that can induce
an impact, and may be characterized by the following attributes:
Type: chemical, physical, or biological
. Intensity: concentration, magnitude, abundance/density
Duration: acute (short term) versus chronic (long term)
. Frequency: single event versus recurring or multiple exposures
Timing: time of occurrence relative to environmental and human health
parameters
Scale: spatial extent and heterogeneity in intensity (EPA, 1992c).
The stressor concept is imbedded (implicitly) in life-cycle impact assessment. In this
context a stressor can be an inventory item that leads to a primary impact(s), or a stressor
can be an impact that leads to a secondary impact(s), and so on. For example, a stressor
could be identified as a quantity of SO2emissions to the airfrom a given product or process
system This SO2 can be linked to primary impacts such as acid precipitation. Ac.d
precipitation is an impact of SO2 emissions as well as a stressor, because it can be l.nked to
secondary impacts such as acidification of water bodies, tree damage, building matenals
corrosion, and the leaching of metals from soils.
Several issues are related to the definition of an impact. First, impact in the context of
impact assessment rarely means an actual impact but instead means a potential impact. The term
"potential" is not meant to minimize concern for those impacts but to point out that impact
assessment does not necessarily provide direct measures of actual impacts, such as the actual
number of dead fish that result from the waterborne effluent X of process A. Instead, impact
assessment might attempt to establish a link between inventory items and potential impacts. For
example waterborne effluent X from process A may be identified in the literature as toxic to
fish above a threshold concentration. Researchers can use this threshold to indicate the potential
for impact and not the actual number of fish harmed or killed. Thus, unless otherwise specified,
the term "impact" in this report implicitly carries the connotation of potential impact.
A second issue is the difficulty in quantifying potential impacts (e.g., estimating the
number of fish mortality resulting from release of waterborne effluent X). Limitations in data
1-5
-------
\; , ayailability, modeling, aridresource limitatiionsandithe complexity of most natural
i systemsoften require a more qualitative djescription of impacts based on some amount of
'>'. quantitative information (e.g., level of pollutants released). This issue, however,-should not
: discourage practitionerstfrom conducting impact assessments. Depending on the goals*and; scope
, / of the LCA, qualitative information may be (adequate, and in some cases preferred, to assist users
in identifying and evaluating opportunities to achieve environmental improvements.
A third issue associated with the term impact is the potential large number of impacts '
associated with any given inventory item, that is, although impact assessment attempts to
establish a link between inventory items anq impacts, a large number of impacts can be
associated with any single inventory item. Ideally, impact assessment would analyze every
potential impact, but that would typically bej infeasible. Therefore, practitioners need to decide
which impacts are within the goals and scopje of the LCA and if those impacts can be estimated
or measured. ;
Finally, the potential for an impact to occur is not easily defined, nor easily captured, in
any impact analysis. The analysis is hindered by a number of uncertainties and a general lack of
knowledge about the natural processes that determine the fate, or impact, of substances or
activities in the environment. The potential for an impact to occur is governed by a number of
different variables, such as those listed in the following function:
Impact = f (location, medium, time, rate of release, routes of exposure, natural processes,
persistence, mobility, accumulation, toxicity, concentration of release,
assimilative capacity, synprgism, antagonism, etc.)
The uncertainty associated with an impact actually occurring is often the subject of considerable
debate. Uncertainty, in this context, focuses Jon the interrelationships between the inventory
items and the associated impacts and between the impacts themselves.
1.1.3 Impact Assessment [
In the LCA literature, impact assessment has different meanings for different people.
The following are a few examples of the multiple interpretations of impact assessment presented
in the context of LCA: |
i
An assessment of the impacts on human health and the environment associated with
raw materials and energy inputs and environmental releases quantified by the inventory
(Tellus Institute, 1992a).
1-6
-------
A system utilized to ascertain the elements and processes involved in translating impact
indicators into the response of environmental receptors and the associated pacts
incurred by the receptors and suffered within the process of transformation (Canadian
Standards Association, 1992).
. A technical, quantitative, and/or qualitative process that characterizes and assesses the
effects of environmental loadings as identified in the inventory stage of the LCA
(SETAC, 1993).
A process that meaningfully relates inventory information into relevant Concerns about
natural resource usage and potential effects of environmental loadings, consistent with
medefined scope, specificity, and technical precision of the life-cycle inventory data
(Procter and Gamble, 1992).
An analysis of the effects of inputs and outputs on the environment, where the effects
are secondary inventory values that are induced to change as a result of the primary
inputs and outputs of an industrial system (Scientific Certification System, 1992).
An underlying theme throughout these descriptions is that impact assessment is a process
of linking the inputs from and outputs to the environment (which are compiled in the inventory)
to potential impacts to ecosystems, human health, natural resources, and possibly social welfare
impacts. For purposes of this report we define impact assessment as follows:
Impact assessment: A systematic process to identify, characterize, and
value potential impacts to ecosystems, human health, and natural
resources based on the results of a life-cycle inventory. .
1.2 PURPOSE OF LIFE-CYCLE IMPACT ASSESSMENT
Impact assessment attempts to take the input and output data compiled in an inventory
analysis and translate that data into either (1) a quantitative and/or qualitative description of
environmental impact, or (2) a description of how each inventory item (per functional unit)
contributes to environmental impacts. A complete impact assessment considers potential
impacts to the full range of environmental media (e.g., air, water, land).
Conceivably, LCA could stop after the inventory analysis. One reason it does not is that
impact assessment makes explicit the methods used to compare and weigh inventory items.
Failing to communicate these methods might convey that all inventory items have relatively
similar magnitudes of impacts. Another reason for continuing past the inventory analysis is to
provide the LCA user with information that is more useful for decisionmaking. For example,
determining the relative overall environmental burden associated with two product systems is
1-7
-------
often difficult.when the emissions-ofione ppllutant/say;SOj, are estimatedto be-higher for one
j production system, while emissionsof adif^rentpollutanys^reactive:;hydrocarbons, are
JS.'Estimated to be higher for the other production system. V:
.-;.? LCA is not necessarily a linear or stbpwis&jprocess: Rather, as suggested by Figure11-1,
information from any of the three components can complement information from the other two
components. For instance, opportunities for environmental and human health improvements do
not necessarily stem from the improvement assessment but can be realized at any stage of the
LCA process. The inventory component alone may be used to identify opportunities for
reducing the amounts of specific inputs and outputs. The impact assessment can provide
additional information about the significance of the inventory items, or identify priorities for the
improvement assessment. The impact assessment may also present important information
suggesting modification of the goals and scope of the LCA, or identify data gaps, research needs,
or significant uncertainties in the LCA. The| following sections discuss the relationships between
impact assessment and the inventory analysis and improvement assessment components of LCA.
I
1.2.1 Relationship Between Impact Assessment and Inventory Analysis
i
Impact assessment focuses on describing potential impacts to ecosystems, human health,
and natural resources through the use of a variety of models. Typically, these models require
supporting data (e.g., environmental or human health information). Therefore, the type of data
collected in the inventory analysis must be qommensurate with the impact assessment model(s).
[
Upgrading inventory data may be necessary to account for the specific data needs of an
impact assessment. While conducting the impact assessment, a practitioner may realize that
additional data (e.g., lexicological, environmental parameters) are needed. For example, to
conduct a detailed impact assessment, the practitioner may need to have information on pollution
speciation or geographic and temporal specificity of impacts. On the other hand, certain
inventory data may not be required given the scope of the overall LCA and/or the impact
assessment.
The importance of making goal statements and determining scope and boundary
conditions prior to developing the inventory [is critical. These activities ensure that the inventory
has the appropriate data needed for conducting the impact assessment or that additional data
collection has been planned, if needed. Inadequate planning can lead to needing additional data
later, which may cause unplanned expenditures or the exclusion of items from the impact
assessment. i
1-8
-------
1.2.2 Relationship Between Impact Assessment and Improvement Assessment
The purpose of the improvement assessment component of LCA is to identify and
evaluate opportunities for reducing or mediating environmental impacts. Opportunities to
achieve improvements may be identified at any stage of the LCA process. Impact assessment
provides a means of identifying improvement opportunities on the basis of impacts. Although
inventory, results can be used to identify opportunities for improvement, impact assessment can
take this information one step further to assess the impacts of the inventory. Also, an impact
assessment supplements the improvement assessment by providing baseline information and
identifying variables that will require further monitoring. Thus, the complexity of the impact
assessment must be matched with the final end use of the resulting information from the
improvement assessment. Once again, scoping plays a large role in maintaining consistency
between the LCA components.
Options identified in the improvement assessment should be evaluated to ensure that the
improvement programs do not create additional, unanticipated impacts. For example, during the
improvement stage the practitioner may discover impacts from proposed improvements
themselves that were not considered in the initial impact assessment. At that point, broadening
the scope of the impact assessment may be necessary to account for the additional impacts.
Although adjusting the scope of the overall LCA or of each LCA component to meet
unforeseen events is possible, maintaining a consistent scope across the components is desirable.
This consistency ensures that the study uses resources and time efficiently and produces results
that are consistent with the goals and objectives of the overall LCA.
1.3 APPLICATIONS OF IMPACT ASSESSMENT
In the context of LCA, impact assessment may be perceived as one tool that
decisionmakers use in the LCA decision development and improvement implementation stages.
As standard procedures and techniques for impact assessment are developed and refined, impact
assessment will enhance the quality of the decision and provide the decisionmaker with a better
frame of reference within which to make the decision.
In the present-day context of LCA, impact assessment may be useful for
characterizing the environmental impacts of inventory items,
uncovering significant cross-media transfers of impacts,
incorporating environmental and human health concerns into the decisionmaking
process,
1-9
-------
evaluating impacts=₯or?,their relevkncKtotpredetermined LCA goals and objectives, and
. translating all impacts and meirrdjetermined importance tosthe LCA audience in a clear
: '* , and concisesmanner (Canadian Standards Association, 1992). :
i .' ''
... Specific applications of impact assessment extend beyond those of inventory analysis.
Although an inventory analysis provides a quantified listing of inputs and outputs, an impact
"assessment relates these items to resulting environmental impacts in a meaningful manner. For
^purposes of this document, two general typjes of LCA applications are distinguished:
1. Internal applications where results are used within an organization and are not
intended for public release; and
2. External applications where results are used, or are intended for use, in a more
public context.
.
As shown in Figure 1-2, the scope and degree of quantification generally increase in
moving from internal to external LCA applications. The broader scope and higher degree of
quantification is often needed for externally applied studies that must withstand widespread
public scrutiny. Table 1-2 provides an ovehdew of a range of internal and external applications
of impact assessment.
APPLICATION
Corporate Strategy/
Internal Communication
Product Design/Modification
Facility Siting/Operation
Public Information/
External Communication
External Policymaking/
Governmental
FORUM
Internal
K>
External
SCOPE
DEGREE OF
QUANTIFICATION
V7
Figure 1-2. Range of LCA Applications
T
Source: SETAC, 1993
1-10
-------
TABLE 1-2. POTENTIAL INTERNAL AND EXTERNAL APPLICATIONS FOR
IMPACT ASSESSMENT
Internal Applications
- Reduce future regulatory liability.
« Compare impacts of generic or raw materials.
Identify materials, processes, or systems that
create significant impacts.
Help develop long-term corporate policy
regarding overall material use, resource
conservation, and reduction of environmental
impacts and risks.
- Forecast potential impacts of new products or
processes.
- Compare alternatives within a particular process
with the objective of minimizing impacts.
Aid in training designers in the use of lower
impact product materials.
- Internally evaluate impacts associated with
' source reduction and alternative waste
management techniques.
Assess industrial process efficiency
External Applications
Provide information that allows consumers or
institutional buyers to evaluate and differentiate
between products.
Provide information to policy makers, :
professional organizations, public-interest
groups, and the general public about the
environmental and human health consequences
associated with a particular product or procesjs
life cycle, the use and release characteristics ,
associated with a particular product or process
life cycle, and potential impacts avoided by :
source reduction and alternative waste
management techniques.
Help develop local, regional, or national long^
term policy regarding overall material use, waste
management, resource conservation, and ;
reduction of environmental impacts and risks.
' Supply information needed for legislative or
regulatory policy that restricts or promotes the
use of specific products, materials, or processes.
13.1 Internal Applications
An internal application of impact assessment is one in which results are never intended to
be, and are never, released to the public (EPA, 1992a). An organization may conduct such an
impact assessment, for example, to determine which production process exposes the organization
to the least current and future regulatory liability.
For internal impact assessments, the sponsoring organization is not required to justify the
methods, data sources, and items included and/or excluded outside the organization. Within the
organization, these aspects of the LCA may or may not require as rigorous a justification 45
needed for an external application. While internal applications may not be required to follow
stringent LCA guidelines, they should nonetheless follow the best practice. However, if the
study results may be used externally at a later date, consideration should be given to conducting
die assessment in the same manner as an external study. ;
1-11
-------
I
Using a holistic, systematiC'approaqhto impact assessment that-considers decisionmaking
factors that were once outside the corporate,sphere may significantly;alter the. corporate
Decisionmaking process. Corporations may-find that performing an impact assessment is.in their
* ^best interest because it may lead to impact reduction? through waste minimization, more;efficient
production processes, and bottom-line cost savings.
I '
1.3.2 External Applications
i
An external application of an impact assessment is one in which results are made
available outside the sponsoring organizatipn (EPA, 1992a). External impact assessment results
may require more rigorous justification because they are open to additional scrutiny and thus
may be faced with more intense peer revie\^ and disclosure of methods and results.
The public may expect external impact assessments to abide by impact assessment
guidelines that represent a wide consensus bf opinion. If guidelines are not followed, the public
may request the sponsoring organization to provide information regarding the deviation from
those guidelines. This request may be the cjase when the impact assessment results are used to
support marketing claims that make producj comparisons and may significantly affect other
external entities, or when the results might Affect public policy.
I
1.4 CURRENT STATE OF IMPACT ASSESSMENT PRACTICE
World Wildlife Fund (1991) recently updated a survey of three LCA
practitionersBattelle, Franklin Associatesl Ltd., and Tellus Instituteto provide an overview
of the state of impact assessment practice. Table 1-3 reports some of these results and describes
the types of environmental and human health analyses currently performed by these three
recognized LCA practitioners, as well as various methodological approaches used in impact
assessment. I
E
In addition, an industry survey by Sijllivan and Ehrenfeld (1992) explored several
companies' uses of analytic tools and programs designed to account for impacts throughout a
product's life cycle. The survey revealed thjit the environmental impacts and life-cycle stages
addressed by companies were fairly consistent. Air, water, soil emissions, and solid waste
generation were addressed by all companies surveyed, and natural resource and energy use were
addressed by eight of ten life-cycle frameworks. Habitat alteration was addressed by four of the
ten frameworks, but biodiversity was rarely addressed.
The survey also found that, although
many of the impact assessment frameworks used
included elements that demonstrate life-cycle thinking, these frameworks are not standardized.
Instead, they range from quantitative assessment techniques (e.g., indexing the importance of
1-12
-------
various
impacts) to more subjective techniques, such as consensus building and professional
judgment (Sullivan and Ehrenfeld, 1992).
TABLE 1-3. PRACTITIONER SURVEY OF IMPACT ASSESSMENT
CONSIDERATIONS
Considerations
I^^M^BB^M^^I^^^^^^^^^^^"'^^"^^^^^^^^
Amount/volume
Toxicity
Exposure
Persistence
Mobility
Global effects (e.g., climate change,
ozone depletion)
Risk assessments
Consumer/worker safety
1. For releases to the environment,
what criteria are used to select
pollutants to measure
a) pollutants covered by
federal/state laws and
regulations
b) pollutants that exceed some
threshold level, regardless of
regulatory controls
c) impact of pollutants (e.g.,
toxicity, etc.)
d) SARA 313 list of toxic
chemicals
2. Are releases assumed to meet
current treatment standards?
Battelle
Yes
Yes
Only where generic pathway
is defined.
degradability.
Via surrogate measures (e.g
water solubility).
Establish equivalency of
various individual
contributions.
No
No
Yes
Yes
Establish impact potential
networks (inventory vs.
impacts).
Sometimes. Prefer actual
releases; treatment standards
used only if no other data are
available.
Franklin
Associates, Ltd.
Yes
Yes
ithway Yes
down or Yes
js (e.g., Yes
of Yes
No
No
Tellus
Institute
Yes
Yes
No
No
No
Yes
No
No
Yes
Yes
Variable
Yes
No
Yes
Only if actual Only if actual
emission data are emission data are not
not available. available.
^^-~"
(continued)
1-13
-------
TABLEvl-3. PRACTITIONER'SURVEY OF IMPACT ASSESSMENT
i ' CONSIDERATIONS (CONTINUED)
Considerations
Battelle
Franklin
Associates, Ltd.
Tellus
Institute
3. Is impact of individual
pollutants estimated?
4. What about relative impacts
within and across media?
5. Is analysis primarily
quantitative or qualitative?
Try to assess: whether
concentration may be a
problem only where a defined
pathway and [threshold level
exist [
Have used valuation by
Analytic Hierarchy Process
(AHP) in streamlined LCA,
but never in a, conventional
LCA. |
Mix of qualitative/
quantitative depends on
product stagej and
environmental pathways.
At least through the
characterization
phase.
Yes
Where comparison Methods developed to
measures can be rank relative impacts,
developed. especially within
media.
Mixdepending Quantitative
on the quality of
data.
Source: Updated in 1994 from World Wildlife Fund, 1991
1.5 SCOPE OF THIS DOCUMENT j
This document outlines a conceptual framework, discusses key issues, and summarizes
existing impact assessment methods. Chapter 2 discusses key issues related to the current use
and future development of impact assessment. These issues include, but are not limited to,
standardization of the impact assessment framework, scoping, uncertainty, data quality, value
judgments, transparency, expert review, and presentation of results.
Chapter 3 outlines a conceptual framework for impact assessment, which includes three
phases: classification of inventory items intja impact categories, characterization of selected
impacts, and valuation of impacts within and between impact categories. This chapter also
discusses the different levels of analysis in the characterization phase, from less detailed loading
assessment to more detailed risk assessment;
Chapters 4 through 7 summarize existing methods that have been presented, discussed, or
used in the context of impact assessment. Chapter 4 profiles existing methods for characterizing
impacts to ecosystems, human health, and natural resources. Chapter 5 discusses issues related
to resource depletion and describes some existing methods for characterizing resource depletion.
Chapter 6 presents those methods that apply to the valuation phase of impact assessment.
Chapter 7 profiles integrated approaches thai combine two or more phases of impact
1-14
-------
assessment typically the characterization and valuation phases. Chapter 8 reiterates key points
regarding impact assessment and discusses potential future research needs to fill gaps in existing
impact assessment procedures and methods as well as to better define the overall role of impact
assessment in the LCA process.
Procedures and experience learned from environmental impact assessment as defined by
the National Environmental Policy Act (NEPA) are included in Appendix A. Untested methods
potentially useful for impact assessment are profiled in Appendix B. Appendix C provides key
terms and definitions, and Appendix D is a bibliography of LCA-related literature.
1-15
-------
-------
CHAPTER 2
KEY ISSUES SURROUNDING IMPACT ASSESSMENT
Much of the current focus in the development of impact assessment is determining how
to apply a wide variety of possible tools and methods within the impact assessment framework.
This chapter discusses key issues related to the future development of impact assessment,
including, but not limited to, standardization of the impact assessment framework, scoping,
uncertainty, data quality, value judgments, transparency, expert review, and presentation of
results.
2.1 STANDARDIZATION OF THE IMPACT ASSESSMENT FRAMEWORK
Increasing use of LCAs has resulted in a broad recognition that some degree of
standardization of methodology is necessary to increase replicability and comparability, as well
as public and peer confidence in external LCA studies (Denison, 1992b). To develop a
standardized impact assessment framework, practitioners must agree not only on the conceptual
aspects of impact assessment but also on other procedural aspects of impact assessment as well.
These aspects might include
a set of steps for the impact assessment practitioner to follow,
a standardized list or checklist of impacts for the practitioner to consider,
a common format for peer/expert review activities,
a code of good practice for impact assessment as part of overall LCA studies, and
a standardized presentation format for impact assessment results.
However, the question remains: Is it possible, and desirable, to develop a standardized
impact assessment framework, or should the choice of framework be left to the practitioner?
Although leaving the choice of impact assessment framework to the practitioner may be
amenable for a wide variety of study scenarios and circumstances, using a standardized impact
assessment framework could provide the following benefits:
Users could make relative comparisons of studies without having to translate LCA
studies to a common denominator.
Potential misuse of impact assessment results to achieve a particular purpose or goal
would be controlled.
Practitioners would have objective guidelines for conducting an impact assessment.
2-1
-------
A standard listing of .impact categories would'remove? some subjectivity in selecting
impacts and wouldi facilitate comparison between studies.
j-
Analysts would be able to incorporate results of other external LCAs into their studies.
I
!. v '/ A consistent set of inventory and impact assessment data would be available to all
interested parties. [ ''""' ;
\ - ' -
On the other hand, a standardized framework for impact assessment may have little effect
on LCA practices. For example, despite the development of guidelines for conducting inventory
analysis, a number of significant discrepancies still exist in life-cycle inventory studies. Among
other things, these discrepancies include differences in the definitions of the scope and process
boundaries.
i]
Only a few impact assessments hav^ been conducted, and impact assessment procedures
are still in their formative stages. Any standardized framework will undoubtedly be a function of
future impact assessment research and experience. Therefore, researchers suggested using a
phased approach in which an initial impact assessment framework is developed with presently
available methods. Later, experience and insights derived from using the framework can be used
to refine and/or develop new methods. I
In this approach, experts develop and agree upon general principles and procedures,
analysts begin preliminary case studies usinfe these principles and procedures, and experts use
feedback from preliminary studies to identify areas of need and to refine or redevelop the impact
assessment process. Figure 2-1 provides a conceptual illustration of the phased-approach to
impact assessment development. !
A phased approach would allow researchers to use existing methods available for impact
assessment while methods to fill gaps or analyze more difficult-to-determine impacts, such as
habitat destruction, are developed. In addition, this approach would continue LCA and impact
assessment case studies rather than delaying! them in hopes of establishing the "ideal" approach.
2.2 USE OF SCOPING IN IMPACT ASSESSMENT
Scopingdeciding what will and will not be included in the studyis an integral part of
LCA. In general, the impact assessment should consider all inputs and outputs compiled in the
inventory. The assessment should also include justifiable reasons for any exclusions. Any
justifiable reasons for any exclusions will b£ tied to the goals and scope of the LCA.
2-2
-------
Consensus-
Based
Impact
Assessment
Principles &
Procedures
Refine/Redevelop
Impact
Assessment
Process
Broad Range
of Case
Studies
Figure 2-1. Phased Approach to Impact Assessment Development
On the other hand, in conducting an impact assessment, the practitioner may need to
reevaluate the scope (to identify inventory items or impacts that will need additional data
support), define which impacts are relevant to the LCA, and define the intended application or
end use of the impact assessment results. Currently no set of rules exists that govern the type of
information that can be used in an impact assessment, nor is there a clear need for one.
To define the scope of an impact assessment, the practitioner may find it useful to
consider some primary scoping parameters specific to the impact assessment. These parameters
might include the level of detail of the impact assessment, product system/potential impact
boundaries, and the type of impact information required by decisionmakers. Some generic
scoping parameters that are a function of the overall LCA include the following:
matching the scope of the impact assessment to the goals and objectives of the LCA,
identifying key inventory data that are missing or uncertain,
identifying the variability of inventory data,
identifying the impacts to be studied,
recognizing the purpose(s) for conducting an impact assessment,
determining how the results of the impact assessment are to be used,
2-3
-------
F
providing justification for excluding;any;elements, .
, >\- ,4* determining theilevel of impiactst<> bejstudied,(source,; media,,'orireceptor), ,:
>,Considering the audience .toiwhich results wiU,be;,presented, and ,
.'., ;./" defining spatial andtemporal boundaries (SETAC, 1993).
No one "correct scope" can be assigned to all impact assessments. The scope of an impact
assessment will inevitably be a function of a number of study-specific variables such as goals,
scope, and data limitations.
One important point to consider when initiating an impact assessment, or LCA in
general, is the "nonthreshold assumption." JThe nonthreshold assumption simply says that no
threshold exists for considering environmental loadings in an impact assessment. In other
i
words, despite seemingly insignificant quantities, inventory items nonetheless contribute
cumulatively to impacts and therefore may peed to be considered in the impact assessment. For
example, the energy used to manufacture a Single automobile likely does not release enough SO2
to the atmosphere to cause an appreciable rise in regional acid precipitation. However, when
those SO2 emissions are considered in the context of additional regional emissions, the SO2
emissions may be considered a contributor to the regional acid precipitation and thus the SO2
emissions may warrant consideration in the impact assessment.
The nonthreshold assumption may b$ of greater significance for some inventory items
compared to others. For example, low concentrations of a noncarcinogenic, nonpersistent
pollutant may be below a threshold of concern for human health effects. Practitioners may need
to consider the appropriateness of the nonthreshold assumption for each inventory item with
respect to the potential impact being assessed.
Incorporating the nonthreshold assumption into impact assessment not only provides
justification for considering all inventory items in impact assessment, but also provides impetus
for assessing the relative contribution of those inventory items to specific impact categories. In
other words, the nonthreshold assumption makes it appropriately difficult for LCA practitioners
to eliminate inventory items from further consideration on the basis that the quantity of
inventory items is too insignificant to contribute to impacts.
One concern with using the nonthreshold assumption in the context of impact assessment
is the possibility of misinterpreting the outcdme of the impact assessment to represent actual
causal association between inventory items and impacts. To avoid this situation, practitioners
should make the use of the nonthreshold assumption transparent to users of the impact
assessment. Section 2.8 discusses an approach for summarizing the results of an impact
2-4
-------
assessment, providing a format to clearly communicate specific aspects of the assessment such as
the nonthreshold assumption.
Scoping in impact assessment may draw in part from the scoping process required as part
of Environmental Impact Assessment (EIA) by the U.S. Council on Environmental Quality
(CEQ) in response to NEPA of 1969. The EIA scoping process is described in Appendix A.
2.3 COMMUNICATING UNCERTAINTY IN IMPACT ASSESSMENT
Determining fate and effects of pollutants and substances in the natural environment is
extremely difficult. Uncertainty in the context of impact assessment extends beyond that in the
inventory analysis. Inventory data usually are based on many assumptions, represent aggregated
or averaged measures, contain many gaps, and are broad in nature (e.g., data from different plants
with different levels of technology). Nonetheless these data typically are measurable inputs and
outputs that can often times be bounded with some type of measure (e.g., range) of uncertainty.
Because inventory data are the primary inputs for impact assessment, the range of uncertainty
associated with the impact assessment model is partly dependent on the range of uncertainty
associated with the inventory data. However, additional uncertainties are introduced in the
impact assessment stage of an LCA. Section 2.3.3 discusses specific methods for uncertainty
analysis. Practitioners should describe and discuss the uncertainties associated with LCA impact
assessment methods, data, and results.
2.3.1 Translating Inventory Items to Impacts
A primary issue in impact assessment is the uncertainty surrounding the linking of
inventory items to impacts. Scientific information may indicate that various inventory items are
associated with, or have been shown to cause particular effects. However, it is difficult (if not
impossible) to prove that a specific input or output from a specific LCA causes an actual effect.
Thus the results of the impact assessment will likely not prove that the product system under
consideration actually caused such impacts. None the less, a link can often be made between an
inventory item and a potential impact, or multiple impacts. For example, SO2 emissions have
been linked to the formation of acid precipitation, which in turn can lead to other impacts such as
tree damage, acidification of lakes, corrosion of buildings and materials, and the leaching of
heavy metals from soils.
The causal uncertainty described above is primarily a result of limited understanding of
such concepts as biochemical, physiological, and environmental interactions; fate and transport of
substances released in an environmental setting; and the distribution of nonchemical stresses
(e.g., heat, noise). Factors that may need to be considered to understand impact linkages include:
2-5
-------
^the spatial and >temporal,scales-of [environmental loadings, ""' tH «' y
-.'; it interactions among human-inducejd loadings and natural .loadings,
* natural variability and, the problems of discriminating from "background noise," land
? the different modes of action of the loading on the environment (EPA, 1992d). '
As a result of these factors, impact networks are often diverse, nonlinear, and largely
site-specific, and involve a wide range of potential impacts at various thresholds. Although
increased research in these areas may reduce a substantial amount of uncertainty, certain aspects
of this uncertainty are intrinsically irreduciblefor example, natural climatic variations among
different locations. In addition, inventory data are generally not site-specific, which adds
additional uncertainty to the analysis. [
Clearly, a considerable array of complexities and uncertainties exist when translating
inventory into potential impacts. A key issue is how to account for and communicate this
uncertainty in the context of impact assessment or in the impact networks themselves. Some
possible ways of incorporating uncertainty into impact networks include quantitative approaches
such as incorporating probabilities or measures of compounded uncertainty into the linkages
between inventory items and potential impapts and qualitative approaches such as using a set of
qualitative evaluative criteria.
Quantitative Approaches
Some possible ways of incorporating uncertainty into inventory-to-impact links or
impact-to-impact links include, among other things, incorporating probabilities or measures of
compounded uncertainty into the links. In doing so, the practitioner must bear in mind that the
two concepts have largely different effects on the expression of causal association, as described
below:
1. Joint Probability: When giver} two events, A and B, the probability of both A and
B occurring is the product of the| probability of occurrence of A times the
conditional probability of event B occurring (i.e., the probability of event B
occurring given that event A has! occurred). For example, consider the case where
life-cycle inventory item X leads to potential impact A with a probability of 0.5,
and potential impact A leads to potential impact B with a probability of 0.5 (none of
these statements involves uncertainty). One can then state, on the basis of joint
probability, that the probability of life-cycle inventory item X leading to potential
impact B is 0.25 (0.5 x 0.5), and uncertainty plays no role.
2. Compounded Uncertainty: Using the above format, when given two events,
A and B, each with a given range of uncertainty, the likelihood of both A and B
occurring is the products of the ranges of uncertainty. For example, consider again
2-6
-------
the case where life-cycle inventory item X leads to potential impact A with a range
of probability of 0.2 to 0.8, and potential impact A leads to potential impact B with a
range of probability of 0.2 to 0.8. The effect of compounded uncertainty is that one
can only say that the likelihood of life-cycle inventory item X leading to potential
impact B is between 0.04 and 0.64 (0.2 x 0.2 and 0.8 x 0.8).
The above examples illustrate a key distinction between joint probability and
compounded uncertainty. Unlike joint probability, compounded uncertainty does not make
further potential impacts less likely to occur but instead makes them increasingly more difficult
to predict. This key difference should be kept in mind if either of these two methods are used as
expressions of causal association.
In many impact assessments practitioners would not likely develop quantitative measures
of joint probability or compounded uncertainty for relating inventory items to impacts but rather
would express the inventory data as means and variances or ranges.
Qualitative Approaches
A possible qualitative approach to evaluating causal relationships among inventory items
and impacts in an impact assessment is to use a set of evaluative criteria, such as those suggested
by Hill (1965):
strength (a high magnitude of impact is associated with a particular loading)
consistency (the association is repeatedly observed under different circumstances)
« specificity (the impact is diagnostic of a loading)
» temporality (the loading precedes the impact in time)
presence of a biological gradient (a positive correlation between loading and impact)
a plausible mechanism of action
coherence (the hypothesis does not conflict with knowledge of natural history)
experimental evidence
analogy (similar loadings cause similar impacts)
Although not all of these criteria need be satisfied to support causal association, each will
incrementally reinforce the argument for causality. The presence of refuting evidence does not
necessarily rule out causality. Instead, it may represent an incomplete understanding of the
complex relationships at hand.
2-7
-------
2.3.2 Impact Assessment Results
; Even if'impact assessment were atyle to provide afdirect measurerof-uncertainty, the
concern remains about compounding uncertainties from thei-inventoryiwith uncertainties from an
impact assessment. That is, when the LCJ^. practitioner uses inventory data with much
uncertainty associated with it and then incorporates the uncertainties associated with the impact
assessment process described above, the resulting information is questionable in the
decisionmaking process. Uncertainties surrounding different components of the environmental
impacts evaluation affect the analyst's confidence in making a specific conclusion (EPA,
1994b).
i
Uncertainty clearly plays a large r
-------
Quantitative Methods
. confidence interval/data variability estimation
accuracy, precision, and degree of bias measurement
goodness of fit evaluation
sensitivity analysis
uncertainty analysis
Qualitative Methods
.limitations of life-cycle inventory data for predicting impacts
. vaUdity, accuracy, and limitations of dating inventory items into .mpact categones
. validity, accuracy, and limitations of conversion models used
Techniques such as sensitivity analysis or uncertainty analysis may provide useW
^ of =tt .potential
future applicability to LCA, the reader is referred to EPA (1992b).
analysis can also be used as abasis for decisionmakmg purposes among comparative
assessments.
(1992b).
2-9
-------
,
Tornado diagrams
Dominance considerations
Two-way and three-way
sensitivity analysis
Deterministic sensitivity
analysis
relatively simile to use
wide range of applicability
useful for determining the
dominance of specific
alternatives
allows for evaluations of multiple
variables at the same time
useful for evaluating impacts of
alternatives
applicable to L£A data-quality
evaluation
identifies the most significant
Vnri5ifi1*»c f
' requires the development of a
mathematical model
more applicable for option
selection than for sensitivity
evaluation
does not focus on data
quality per se
requires the development of a
mathematical model
I
I
2.4 DATA AVAILABILITY AND QUALITY CONCERNS IN IMPACT
ASSESSMENT
Recent LCA forums (SETAC LCA Data Quality Workshop, Wintergreen, Virginia
October «. 1992; and SETAC Data Quality Open Forum, Washington, D.C, February 18
1993) recognized that data quality is an integral component of the LCA process LCA must be
able to accommodate varying degrees of data! availability, data types, and data quality. Because
of the multiple and significant ways in whicbJLCA information can be used, identifying and
evaluating data quality and their relationship ^o LCA methodology are important. The following
discussion of data quality issues focuses only on issues that are more specific to impact
assessment.
EPA (1994b) has recently developed guidelines to aid LCA practitioners in assessing the
quality of data used in inventory analyses. Data quality is defined in this document as the degree
of confidence an analyst has in a data source & a data value based on defined data quality goals
data quality indicators, and the role of data quality in the overall context of the LCA (EPA
1994b). These guidelines provide a framework for integrating data quality assessment into' the
inventory analysis process. [
-10
-------
TABLE 2-2. POSSIBLE APPROACHES, ADVANTAGES, AND DISADVANTAGES OF
UNCERTAINTY METHODS FOR IMPACT ASSESSMENT
Analytic
Monte Carlo
Response surfaces
Differential sensitivity
Evaluation of confidence
intervals
ranks contributors to uncertainty
economical
widely applicable
facilitates understanding of
sampling distribution concepts
economical
widely applicable
ranking of uncertainty contributors
widely applicable
ranks contributors to uncertainty
measures uncertainty due to
statistical variability in data
limited applicability
sensitivity to input
assumptions hard to assess
no ranking of uncertainty
contributors
dependence on accurate
information and covariance of
input parameters
accuracy hard to assess
long computation times
possible
large model and code
development costs
limited applicability
no ranking of uncertainty
contributors
Source: Cox and Baybutt, 1981.
Similar to the framework discussed above, a data quality assessment framework is
needed to integrate data quality assessment into the impact assessment process. Currently no
protocol has been developed for assessing the quality of data in impact assessments. In addition
to the quality of data received from inventory, practitioners must also consider the quality of
additional data (e.g., toxicity, bioaccumulation, assimilation, equivalency factors, etc.) needed to
conduct an impact assessment.
The purpose of this section is to outline some of the significant data quality issues facing
impact assessment. With very few impact studies to draw from, pinpointing all the data quality
issues that will be integral to impact assessment is very difficult. However, a recent Tellus
Institute analysis of impacts associated with the production and disposal of packaging materials
found basic problems that were related to data used as input parameters for impact analysis,
including the following:
2-11
-------
A lack of systematic data, on some components of the product system limited the scope
of the analysis as well as the;modeling of significant processes or. activities.
/. -<:-;» Publicly available databasesrofteniontainedrout-ofrdate.data^ellus:Institute, 1992a).
2.4.1 Evaluating Data Availability [ r*
Although the lack of available data required for impact assessment is a primary concern,
many sources of data may be useful for conducting an LCA. EPA (1994a) provides a
comprehensive overview of publicly available data sources for conducting an LCA. Table 2-3
summarizes data needs for impact assessment in terms of a five-tiered system of increasing data
quality and decreasing data availability.
TABLE 2-3. IMPACT ASSESSMENT DATA NEEDS
Conversion Model Tier
Data Needs
Tier 1: Loading Assessment
Tier 2: Equivalency Assessment
Tier 3: Toxicity, Persistence, and
Bioaccumulation
Tier 4: Generic Exposure/Effects
Assessment
Tier 5: Site-Specific Exposure/ Effects
Assessment
Mass, volume, or other units of physical quantity.
Same as Tier 1 plus equivalency algorithms based on hazard
data. Also may include measures for resource stocks and
yields, as well as nonchemical loadings.
\
Same as Tier 1 and 2 plus information on interaction of
chemicals with the environment (i.e., persistence and
bioaccumulation) and toxicity data. Also may include
measures for resource stocks and yields, as well as nonchemical
loadings.
Same as Tier 1 plus generic environmental and human health
data. '
i
Same as Tier 1 plus site-specific exposure and environmental
and human health data.
Source: SETAC, 1993
I
A recent SETAC-sponsored LCA Data Quality Workshop in Wintergreen, Virginia,
recognized that currently available environmental input and output data can only support Tier 2-
to Tier 3-type models. Although many feel that such a method can be improved, others have
recognized the lack of information for Tier 2 !to Tier 5 conversion models.
Advancing to Tier 2- and Tier 3-type conversion models, which require equivalency
factors and chemical properties data, will require the inventory to contain an increased level of
2-12
-------
chemical and site specificity. However, such a level of data quality may be achievable m the
near term Tier 4-type conversion models may use the same data as Tier 3-type models, only in
a different manner. However, in general, to move to Tier 4- and Tier 5-type conversion models,
process- or activity-specific, unaggregated and unaveraged inventory data will be needed.
Inventory data are currently unable to support such models.
In the near term, researchers may be able to develop a database of information that is
specifically designed for use in LCA. It would contain a variety of information on basic
commodities and pollutants that serve as inputs and outputs to many product or process life
cycles, respectively. Such a database could serve as a clearinghouse for generic information for
supporting LCAs and other types of residuals-based analyses.
2.4.2 Evaluating Data Quality
As stated in the beginning of this section, data quality in the context of LCA is defined as
the degree of confidence an analyst has inadata source oradata value (EPA, 1994b). A
primary concern with respect to data quality in this context is the use of less-than-perfect
inventory data in less-than-perfect impact assessment models as described in the previous section
on uncertainty. The resulting information may have questionable usefulness for decisionmakmg
purposes. For example, aggregated secondary data are typically used in inventories. Aggregated
data are not useful for some impact assessment methods, such as fate and transport models or
exposure assessment, that require site-specific data (EPA, 1994b).
A second concern in impact assessment is the quality of additional data needed by
conversion models. The only type of conversion model that does not require any additional data
is loading assessment, where inventory data are used directly (see Table 2-1). Any model
beyond loading assessment requires additional information such as toxicity, persistence,
bioaccumulation, and equivalency factors. Even with the highest quality inventory data, impact
assessment results can be compromised if low quality information (e.g., environmental
characteristics, toxicity measures) is used in the conversion models.
A third concern is the quality of the conversion models themselves. That is, even with
perfect input information, the quality of impact assessment results is governed by the predictive
accuracy of the conversion model(s) used. This issue is not limited to impact assessment but
includes any type of analysis that employs models to transform data into more useful and
meaningful forms. At this stage of impact assessment development, conversion models need to
2-13
-------
be developed and validated as welL>Therefore, one'of the goals of;impact assessment may be to
make the limitations of conversion^models,j as well as any additional infonnatioa'required by the
/models, transparent to the users of .the results.
2.5 INCORPORATING*VALUE1JDPGMENTS INTO IMPACT ASSESSMENT
i - '
Impact assessment is similar to other decisionmaking support systems in that it involves
applying subjective value judgments. Nothing is inherently right or wrong with value
judgments. All individuals and institutions [have subjective values, which they express either
explicitly or implicitly. A key objective of impact assessment is to make subjective value
judgments transparent so that users and others will know the basis from which the assessment
was conducted and any conclusions drawn, j Clearly articulating subjective value judgments lets
the user know the values that guided the impact assessment.
i
I
Both the practitioner and the user of, the resulting impact assessment make value
judgments as the result of such considerations as the presence of uncertainties, data limitations,
and impact assessment model limitations. Because different individuals and institutions have
different values, there is no "correct" set of jvalues to use during the impact assessment.
However, for purposes of impact assessment, and LCA in general, developing a standard
protocol for identifying and evaluating valup judgments may be worthwhile.
In the face of incomplete information and uncertain cause-and-effect relationships, the
practitioner may need to make judgments based on the available evidence. The main problem in
making value judgments about cause-and-efjfect relationships is that directly applicable data are
often insufficient. In such a case, the practitioner must use value judgments to make the best
possible assessment of the relationship given the information at hand. Furthermore, because the
extrapolation of value judgments depends op the practitioner's interpretation of the impact
assessment literature, different people will have different interpretations and, thus, different
value judgments. Unlike other forms of uncertainty (e.g., measurement and sampling error) that
can be generally calculated by means of standard procedures, the type of uncertainty described
above cannot be directly quantified because iof its judgmental nature.
Value judgments occur at varying degrees throughout the impact assessment process.
Within the impact assessment component, value judgments can occur at any of the following
points:
2-14
-------
goal definition and scoping
classification of inventory items to impact categories
determination of impacts of concern
evaluation and selection of models to characterize impacts of concern
interpretation of results obtained from impact characterization efforts
development of assumptions based on logic and scientific principles to fill data gaps
evaluation and selection of ranking or weighting schemes in the valuation phase
In summary, value judgments are integral parts of any decisionmaking system, including
environmental policy decisions. In that regard, impact assessment is no different from any other
public private, or individual decision that can affect the environment or human health. It is
recommended that practitioners clearly articulate those value judgments either qualitatively or
quantitatively and discuss the scientific basis or evidence and any philosophical, cultural, or
intellectual influences for making the judgments. Employing a method such as encoding
probability judgments may provide a means of identifying and quantitatively characterizing
value judgments. Also, this method would enable users of the impact assessment to understand
the frame of reference from which the impact assessment was conducted, even though they may
not personally agree with it.
2.6 TRANSPARENCY
Because assumptions and value judgments are integral parts of impact assessment and
many other decisionmaking systems and they shouldn't be eliminated in LCAs, their use must be
made transparent (i.e., clearly defined). Transparency entails full disclosure of the content and
conduct of the impact assessment process, including assumptions and subjective value
judgments. The practitioner should strive to present the following specific aspects of an impact
assessment in a transparent manner:
goals of the LCA and impact assessment
scope and boundary settings
data sources/data quality/data variabilityuncertainty
models/methods used in the impact assessment process
- assumptions
- limitations
data or methodology manipulations
2-15
-------
% value judgments
exclusions
»i-t lost information'duejtoaggregation, etc.
.*.,/; analyst's interpretation of theijmrilications of all above items on LCAtesults
Transparency in reporting impact assessment results is important for replicability. By
fully disclosing all aspects of an impact assessment, as listed above, the practitioner enables an
external observer or investigator to start with the same original data and reproduce the impact
assessment results. j
Although barriers to full disclosure (e.g., proprietary data, practitioners' self interest in
keeping their methods or databases to themselves) clearly exist in LCA studies, practitioners
should strive to make their studies as replic|able as possible and should fully explain and justify
factors that preclude them from doing so (Ijtenison, 1992b).
Reproducibility is important to support the understanding and credibility of the impact
assessment results. For example, a current {working study comparing two existing LCAs of
corrugated cardboard found that differing results were largely due to differences in study scope
and boundary settings (Ekvall, 1992b).
2.7 EXPERT PEER REVIEW
Scientific data and methodologies used in impact assessment are based on information
that is frequently complex, conflicting, ambiguous, or incomplete. Therefore, EPA supports the
creation of an expert peer review process fojr impact assessment, and LCA in general, to advance
the quality and consistency of LCAs. The desirability of an expert peer review process stems
from four main areas of concern: 1) the lack of understanding of the scope or methodology used
in LCAs, 2) the desire to verify data used and practitioner's compilation of data, 3) questioning
of assumptions used and the overall results, 'and 4) the communication of results (EPA, 1993a).
Practitioners can evaluate the viability and accuracy of impact assessments by
establishing an expert peer review process for impact assessment. Expert peer review can be
integrated into the following stages of impact assessment:
determining the purpose and scope of the impact assessment;
evaluating data sources and the quality of data used in the assessment;
evaluating and selecting assessment and measurement endpoints;
evaluating and selecting conversion models;
2-16
-------
developing assumptions, etc., to fill data gaps; and
interpreting and presenting impact assessment results.
The heightened recognition of the importance and necessity of expert peer review in
LCA studies prompted SETAC to develop an interim expert peer review framework. This
interim framework consists of four main steps:
1. Identify and assemble an expert peer review panel based on specified criteria.
2. Review the purpose, study boundaries, and databases of the LCA.
3. Review the stand-alone data compiled in the life-cycle inventory.
4. Review the draft final report (SETAC, 1992).
The purpose of this discussion on expert peer review is not to recommend a specific
approach for an expert peer review process but rather to identify the reasons for having an expert
peer review process for impact assessment and to discuss some issues for consideration before
establishing an expert peer review protocol.
A related issue is how to conduct these expert peer reviews. For example,
Should the expert peer review process use a standard checklist of review items?
What is the appropriate timing of the expert peer review process with respect to
conducting the impact assessment?
Who should pay for the review with respect to internal and external applications?
How should the expert peer review panel members be chosen?
Because impact assessment is in its developmental stages and involves many concepts
and methods that have yet to be corroborated in practice, the use of an expert peer review panel
will be a key role in shaping the future of impact assessment. The expert peer review panel is
foreseen to consist of a relatively small but diverse group of individuals with experience using
impact assessment methods and/or technical LCA procedures. In addition, although expert peer
review is a critical component of both internal and external applications of impact assessment, a
more stringent level of expert peer review will be required for external applications.
2.8 PRESENTATION OF IMPACT ASSESSMENT RESULTS
One of the more important aspects of impact assessment is the manner in which results
are presented to the intended audience. The results of an impact assessment need to be presented
in an effective manner that facilitates the decisionmaking process. Too much information of too
many different types can result in information overload, whereas too little can hinder the
2-17
-------
decisionmaking process. Practitionere:shoi|ild'strive4o, conduct credible; assessments and present
«, the results objectively. >
! ' :"'c:- ':
>;.; Specific aspects of the;impactasseskmentthat;need;tobevdocumentedand:presented;to
decisionmakers include the following: : "
i
, Content Aspects j
[
Clearly delineate scoping activities, including how the boundaries of analysis were ,
determined.
Report any objective data or results separately from subjective data or results.
i
Express the characteristics of the database, including data sources, uncertainty, and
assumptions. I
i
Clearly delineate analysis of actual versus potential impacts.
Conduct Aspects
Provide justification for all impacts that were excluded from the analysis.
Describe the use of assumptions, including how and by whom they were made.
Describe the use of subjective value judgments, including how and by whom they were
made.
Describe any limitations and/or uncertainties of the valuation method.
Table 2-4 shows some possible mettiods for presenting impact assessment results and
their corresponding advantages and disadvantages. A summary chart can be developed based on
one of the methods described in Table 2-4 to present an overview of impact assessment results.
The chart would provide a variety of information beyond the results of the impact assessment
valuation process, such as
a summary of the data used in the jmpact assessment including measures of
variability/uncertainty, 1
a summary of assumptions and value judgments made in the impact assessment,
a description of the methods/modejs used in the impact assessment,
a description of problems encountered and how they were resolved, and
L
a summary of unresolved issues.
2-18
-------
TABLE 2-4. COMPARISON OF IMPACT ASSESSMENT RESULTS
PRESENTATION METHODS
Presentation
Method
Single Score
Qualitative
Rank
Priori tization
Impact Score
Matrix
Objective
Impact
Assessment
Summary Chart
Provides a single impact
score by aggregating all
impacts based on a
common denominator
Uses symbols or ranges
of subjective impact
scores to provide an
overall ranking of
impacts
1 Prioritizes impacts based
on subjective values that
attempt to identify the
more and less pressing
impacts
Provides a quantitative
or qualitative overview
of impact categories
Provides a variety of
information beyond
impact assessment
results (such as databases
used, methods used,
value judgments, and
limitations), which gives
decisionmakers an
overview of the impact
assessment process
Advantages
Disadvantages
Provides
comparative value
Easy to communicate
Provides
comparative value
Easy to communicate
Provides
comparative value
Identifies high- and
low-priority items
Provides
comparative value
Provides standard
format
Incorporates
quantitative and
qualitative data
Provides
comparative value of
entire impact
assessment process
Provides information
on impact
assessment process
beyond valuation
results
Provides a standard
format
Incorporates
quantitative and
qualitative data
Ambiguous derivation
Does not allow for a relative
comparison of impacts
Difficult to incorporate qualitative
data
Provides no information beyond
valuation results
' Fixes subjective values that cannot
be shared by others.
Provides no quantitative support
Difficult to express relative
comparison of impacts
Too simplistic
Provides no information beyond
valuation results
Encourages focus on only the top
priorities
Priorities are often subjective in
nature
Provides no information beyond
valuation results
Potential for confusion with too
much information
Provides no information beyond
valuation results
Potential for confusion with too
much information
2-19
-------
i Decisionmakers could\use;this information as* a tool to lookMlheiOverall|picture or to
/ focus on a particular aspect of me impact assessment. A standard;presentation!fonnat would
... provide a clear and effective means;of:communicating thevpotentiallyicomplex?arrayof & >,
; information inherent in impactiassessmentjor any.impact analysis. Figure 2-2 provides a
possible framework for the impact assessment summary chart. This kind of chart offers a
relatively objective method of presentation where impact assessment results are presented in the
context of the underlying data, methods, assumptions, and limitations used to achieve those
results.
i
Data
sources
quality^
limitations
uncertainty
Value 3i
dgements
whose valuos
Implications
Unresoh
ed Issues
issues
Implications
Models/Methods
limitations
assumptions
uncertainty
Problems
Encountered
problems
how resolved
Impact
Assessment Results
* see Rgure 2-3
Figure 2-2. Impact Assessment Results Summary Chart
2-20
-------
The impact assessment results portion of the summary chart could become complex and
overwhelming because of the large amount of information and variety of different results. To
help communicate the impact assessment results in a concise and comprehensible manner, a
format can be developed for presenting impact assessment results within the overall impact
assessment summary chart. Figure 2-3 illustrates a possible format for organizing impact
assessment results based on the life-cycle stage and category in which the impact occurs.
Life-Cycle Stage
impact
air
water
land
biodiversity
other
occupational
nonoccupational
other
Resource
flow
other
Raw
Materials
Acquisition
Manufacturing
Processing
MMBIMIMH^HM
Distribution/
Transportation
Use/Reuse/
Maintenance
Waste
Management
..
Note: Several ranking methods could be used in this matrix:
> Pluses (+) for significant impact areas, Minuses (-) for less significant impact areas.
> = High, I = Medium, O = Low.
> Numerical Ranking (e.g., 1 -10) of the significance of impact areas.
Figure 2-3. Example of a Possible Impact Assessment Results Format
2-21
-------
< The impact assessment results>,tabl6icould handleiboth quantitative and qualitative
; information. Qualitative Jirfonnation'coultfcbe expressed'-byj symbols, such as + or vindicating
( more or less significant impacts, or by circles with/varied degrees of shading,.indicating ;the
magnitude of the impact. i . i
2.9 UNRESOLVED ISSUES i
Because impact assessment is still evolving, many issues persist that may play a role in
the future development of impact assessment procedures and methods:
I
1. Although it is recognized that impact assessment is an inherently value-laden
exercise, the following questions remain:
Who makes the value judgments?
Is it feasible to use external expert review for value judgments?
Should guidelines be required for value-laden areas, such as valuation, to help
minimize the level of subjectivity in impact assessments?
2. To control potential misuse of impact assessments, quality standards may be needed
when impact assessments are used for external purposes.
3. Specific evaluation methods (conversion models and impact descriptors) and
valuation methods need to be chosen for analyzing specific impact categories.
4. Although methods such as risk assessment and fate and transport models can be used
for impact assessment, analyzing multiple sites may be overly costly and impractical
because of the requirement for {additional data.
5. Much uncertainty persists in liriking inventory items to impacts. Techniques are
needed to estimate and integrate this uncertainty into impact assessment.
6. The political environment under which the LCA is conducted may affect the scope
and impact considerations. ;
j
7. Practicality of a "cookbook" of impact assessment methods versus a more
streamlined approach
8. Incorporation of economic impact (i.e., cost) information into impact assessment
9. Treatment of chronic versus acute impacts
10. Effects of impacts on future generations
11. Impact distribution equity considerations (e.g., impacts on children or other special
subpopulations)
12. Treatment of human-induced versus naturally caused impacts
2-22
-------
CHAPTER 3
A CONCEPTUAL FRAMEWORK FOR IMPACT ASSESSMENT
A major achievement of the SETAC-sponsored Life-Cycle Impact Analysis Workshop
that took place in Sandestin, Florida, during February 1992, was the development of a three-
phase conceptual framework for life-cycle impact assessment. This three-phase conceptual
framework, illustrated in Figure 3-1, contains the following activities:
1 Classification: the process of assignment and initial aggregation of life-cycle
inventory data to relatively homogeneous groupings of impacts (e.g., photochemical
smog, lung disease, fossil-fuel depletion) within primary impact categories (e.g.,
ecosystem, human health, and natural resources).
2 Characterization: the qualitative and/or quantitative evaluation of potential
impacts The process of identifying impacts of concern (called assessment
endpoints) and selecting actual or surrogate characteristics (called measurement
endpoints) to describe the impacts. Characterization involves using specific impact
assessment models to develop impact descriptors.
3 Valuation: the explicit and collective process of assigning relative values and/or
weights to impacts using informal or formal valuation methods.
In Figure 3-1 the flow from the inventory analysis to improvement assessment is not
necessarily linear because the sequence involves interrelationships and feedback loops among the
major components. This is consistent with the three-component LCA triangle illustrated in
Figure 1-1. For example, not only can opportunities for environmental and human health
improvement be realized at any phase of the LCA, but unplanned modifications may entail
revisiting previously completed components. Each LCA phase is discussed in detail in later
chapters of this report.
Selecting the best-suited approach for conducting a particular impact assessment from a
variety of available methods is important. Practitioners can use the following key decision points
to help select the best-suited approach and shape the assessment:
selecting the goals and scope of the study,
learning stakeholder values and information needs, and
characterizing the desired results.
3-1
-------
^ -/f Life Cycle\
| CLASSIFICATION!
J C CHARACTERIZATION J
H
1
X' Vlnventory^ N
I
^ f Develop Impact |
(^ j Networks J ^X
1
V j Classify Inventory Items |
1 by Injipact Category J ^X,
V 1 Determine Assessment |
1 Endpoints J ^-^
^ 1 Select Measurement |
^ ^ndpoints J , "X
^ ,
V^ [ Apply Cohversion Models to \
(Develop jmpact Descriptors J "X
i
,
V, f Apply Valuation ^
Melliuds lo Synthesize 1 .^
Stakeholder Values and 1
w^^lmpact Descriptors J
i
i '
LIFE CYCLE IMPACT ASSESSMENT
ji.
\
c
Life Cycjle Improvement
Assessment
>
Figure 3-1. Conceptual Framework for Life-Cycle Impact Assessment
3-2
-------
Figure 3-2 illustrates how these key LCA decision points fit in the LCA conceptual
framework, emphasizing the impact assessment component. The decision points would be
constantly revisited throughout the impact assessment and especially in the following impact
assessment activities:
classifying inventory items into impact categories;
determining impacts, or categories of impacts, of concerns;
choosing a model, or models, to characterize impacts; and
valuing impacts, or categories of impacts.
3.1 CLASSIFICATION
When inventory items are taken from, or released to, the environment, they are
considered potential causes of environmental and human health impacts. The classification
phase of impact assessment provides a preliminary link between inventory items and potential
impacts The overall purpose of the classification phase is to organize and possibly aggregate
inventory items into impact categories, which provide a more useful and manageable set of data.
This process is accomplished through the two discrete activities of the classification phase:
using existing or developing new impact networks to identify possible impacts
associated with specific inventory items, and
classifying inventory items within appropriate impact categories.
3.1.1 Developing Impact Networks
The preliminary activity of the classification phase of impact assessment is to
qualitatively associate, or link, inventory items with subsequent impacts. This qualitative link
can be established by reviewing each inventory item in the literature to determine its associated
environmental impact(s). Further review of the literature can identify additional impacts that are
associated with the initial impact. For example, consider that a quantity of SO2 emissions
released into the atmosphere is an item specified in the inventory analysis. A review of the
impact assessment literature might identify the theory that SO2 released into the atmosphere can
lead to the formation of acid precipitation. Acid precipitation, in turn, can be found to lead to a
number of additional impacts, such as the destruction of high-altitude forests, acidification of
water bodies, corrosion of buildings and materials, and leaching of metals from soils. Further
search of the impact assessment literature may reveal that these impacts can induce other
identifiable impacts and so on.
3-3
-------
-<**
Feedback
Loop
Inventory Analysis
J ^
I
Impact Assessment Decision Points
viU
Classification Exercise
Characterization Exercise
f
Valuatibn Exercise
----- 1 I
Improvement Assessment
Improvement
Loop
I J
Figure 3-2. Key Impact Assessment Decision Points
3-4
-------
Associating, or linking, inventory items to their respective impacts is a key issue of
impact assessment because the pathways linking inventory items to their impacts typically are
complex and nonlinear. Practitioners can use existing or develop new impact networks to aid in
mapping out impact pathways. Networks of potential impacts are conceptual diagrams that
illustrate qualitative links between inventory items and potential impacts. As the use of the term
"qualitative links" implies, these networks do not necessarily provide a description of actual
impacts. Instead, networks provide a means of identifying all the various potential impacts that
can be associated with inventory items.
Consider the case of a given quantity of carbon dioxide (CO2) identified in the inventory
analysis. A search of the literature reveals that CO2 is often linked to the greenhouse effect,
which is a buildup of CO2 and other gases that are relatively transparent to sunlight but trap heat
by more efficiently absorbing the longer wave infrared radiation released by the earth (Schneider,
1990). In turn, an enhanced greenhouse effect is linked to other impacts such as global warming,
which in turn is linked to regional climate change. This example, as well as the basic framework
for building an impact network, is illustrated in Figure 3-3.
Developing impact networks can be a difficult task. Pathways from inventory items to
impacts may not yet be fully identified and many factors govern how and what kind of impacts
will result. Because many pathways and impacts can exist, tracing impact networks through a
number of different pathways may be necessary.1
As an example of a multiple pathway impact network, consider the case of nitrogen
oxides (NOX) released from a coal-fired electric plant, as shown in Figure 3-4. In other
situations, multiple inventory items can lead to a similar impact or impacts. As an example of
such a scenario, consider the greenhouse effect.
Greenhouse
Effect
Global
Warming
Regional Climate
Change
Figure 3-3. Example of Basic Network Using CO2
'In this report we do not use the terms primary, secondary, or tertiary to distinguish impact levels because of the
implicit valuation imbedded in those terms and the difficulty of assigning the terms to a complex web of
impacts typical of many impact networks.
3-5
-------
Troposphere
Ozone
Aciijj
Precipitation
1 -Decreased
'"%v Visibility
Respiratory
System Damage
Tree Damage
Acidification of
Water Bodies
Corrosion of
Materials
Leaching of
Metals from Soils
Figure 3-4. NOX Example of Multiple Pathway Impact Network from a
Single Inventory Item
Like many other impacts, a number df different substances may contribute to the
greenhouse effect, as demonstrated in FigureJ 3-5.
[
In summary, linking inventory items |to impacts can take a variety of forms. The linkage
can range from a simple linear one (as shown in Figure 3-3) to one that involves linear and
nonlinear relationships between multiple inventory items and multiple impacts (as shown in
Figures 3-4 and 3-5). It is expected that a "library" of networks will be developed through the
practice of impact assessment, making such kssessments increasingly more feasible and
economical. ;
! 3-6
-------
Increased Risk of
Tropical Disease
Figure 3-5. Examp.e of Mu.«ip.e .nventory Items Uad.ng to Similar impacts
312 Classifying iBventory Items Within Impact Categories
After referring to existing impact networks or developing new ones, practitioners should
" the greenhouse effect. All four of these inventory .terns thus can be
Kgor^
main calories of impacts considered in an impact assessment indude
3-7
-------
Despite the current lackof tools to.analyze] social welfare impacts, practitioners can attempt to
incorporate social welfare impacts by [ . . *
iden^ng me impacts of a product or process life cycle on social^welfare and "
As an example of a social welfare impact, consider the large labor force required to
manufacture automobiles. The immigration of a large labor force into the area may result in '
impacts such as overcrowding and degradation of pristine habitat in nearby recreation areas.
Figure 3-6 provides a generic example of possible impact categories and subcategories as
developed from a hypothetical set of impacj networks. The suggested approach to classification
is to first bodd impact networks and see if t^ey contain any inherent structure for developing
subcategories of impacts rather than starting; with a restructured, and value-laden, list of impact
subcategones. This approach to classification is essentially the same as that used by SETAC
(1993), which groups inventory items into relatively homogeneous problem types, called stressor
categories. Our approach differs only in flu* the term "stressor" is not used because of ongoing
confusion associated with the use of that term.
3.1.3 Example Classification Exercise ofJHigh-Density Polyethylene (HOPE) Production
An inventory analysis of an HOPE production system would likely include numerous
components. Table 3-1 provides examples ohnformation developed in an inventory analysis for
the manufacture of HOPE. y^^i
Ecosystem, human health, and naturaj resource impacts associated with the items listed in
Table 3-1 can be determined by searching the impact assessment literature. For example the
release of SO2 from the manufacture of HDPE, as shown in Table 3-1, can be evaluated for
potential impacts by searching the literature for the effects of SO2 released into the atmosphere
From this search, it will likely be determined 'that SO2 emissions to the atmosphere often
combine with other atmospheric compounds to produce acid precipitation. Thus, as shown in
iable 3-2, SO2 can be categorized under the ecosystem impact category of acidification
3-8
-------
Ecosystem^
Atmosphere
toxicity impacts
ozone depletion
greenhouse effect
visibility changes
smog/fog
ground-level ozone
buildup
climatic change
- micro
- macro
Water
toxicity impacts
contamination
depletion
- surface
ground
thermal changes
turbidity changes
acidification
nullification
eutrophication
chemical alteration
Soil
toxicity impacts
salinity
lateritization
podzolization
acidification
fertilization
Human Health
Chronic Effects
carcinogenic
mutagenic
teratogenic
neurological
damage
reproductive
disorders
major organ
-heart
- lung
- kidney
radiation
ionized
- ultraviolet
-heat
> physiological
anemia
skin disease
sterility
i Acute Effects
accidents
- occupational
- nonoccupational
radiation
- ionized
- ultraviolet
-heat
|« noise
i odor, taste, etc.
microorganisms
Natural Resources
Stock
fossil fuels
minerals
atomic energy
soil
space (e.g.,
landfill)
atmosphere
hydrosphere
aesthetics of
planet
Flow
water resources
forest products
agricultural
products
freshwater
products
saltwater
products
flora and fauna
wind power
ocean tidal power
I solar power
Social Welfare
Others
geomorphic effects
biodiversity effects
habitat alterations
animal welfare
Demographic
migration
morbidity
fertility
mortality
Economic
property value
changes
inflation
opportunity
costs
sectoral
effects
Social
government
relations
regulations
indigenous
people's
rights
quality of life
Community
public
services
. infrastructure
satisfaction
Family
structural
changes
stability
changes
employment
Figure 3-6. Posssible Impact Categories
32 CHARACTERIZATION
classification can provide useful information for reviewing the types of
relationship between specific inventory items and impacts.
3-9
-------
!TABLE3-1
MANUFACTURED HDPE
THE
Resource use
Energy demand
Air emissions
Water effluents
Inventory Item
^MMi^MHm
Crude oil
Electricity
Coal
Renewable'fuel
Energy in material
C02
SO,
Quantity (mass or volume)
CO
Hydrocarbons
Particulates1
CFC !
Hydrogen
Crude oil >
Phenol I
Nitrogen |
Organic carbon
Solid waste
Source: Ekvall et al. (1992a).
The complete characterization phase, as defined in this document, includes three separate
but complementary activities: j
determining assessment endpoints,
selecting measurement endpoints (if necessary), and
applying characterization models to cjevelop impact descriptors.
3.2.1 Determining Assessment Endpoints
After identifying the impacts associatei with inventory items and grouping them into
impact categories in the classification stage, thj, practitioner should review the previously
i-io
-------
established goals and scope, and other key decision points, of the overall LCA to
of the impacts are within the scope of the study, which are referred to as the assessrassi
endpoints. These endpoints represent the focus of the characterization efforts.
TABLE 3-2. EXAMPLE CLASSIFICATION OF INVENTORY ITEMS UNDER
IMPACT CATEGORIES FOR HDPE MANUFACTURING
2=:^=^=^=
Impact
Category
Greenhouse
effect
Ozone depletion
Acidification
Smog/fog
Water
contamination
Habitat
alteration
Geomorphic
alteration
*Tmmirt« Human Health Impacts Natural Resource Impacts
Inventory Impact
CO2 Carcinogenic
CFC effects
Particulates
CFC Lung damage
SO2 Odor
NOX
NOX
Oil
Phenol
N
Organic C
Solid waste
Fossil fiiel
Renewable fuel
Solid waste
Electricity
Electricity
Solid waste
Fossil fuel
Renewable fuel
Inventory Impact Inventory
Item Category Item
Grade oil Fossil fuel Grade oil
depletion Fossil fuel
Particulates Renewable Renewable fuel
SO2 energy use
NOX
Hydrocarbons
Solid waste
Ethylene
Oil
Phenol
Because LCAs are restricted by their goals and scope, many of the impacts identified may
not be included in the LCA. The only "correct" set of impacts, or assessment endpoints, is that
which satisfies the specific goals and scope of the LCA at hand. Because no single "correct" or
minimum set of assessment endpoints should be included in an impact assessment, practitioners
should make clear, and possibly qualify, the exclusion of any impacts as assessment endpoints.
3-11
-------
For example, SO^emissions'quantifiedin the inventory ^analysis can be associated with acid
precipitation, which in tum;;can lead'to a nujtnber of further impacts as identified in the
classification stage, including the destruction of high-altitude forests, acidification ;of water
bodies, corrosion of buildings and materials, and leaching of metals from soils. The-practitioner
might determine that, for example, acidification of waterbodies is the most pertinent impact
based on the key LCA decision points; thus'acidification of waterbodies will be considered an
assessment endpoint. !
Determining assessment endpoints from a potentially large number and variety of impacts
is by no means a straightforward exercise. The key LCA decision points must continually be
reviewed, possibly in coordination with denned criteria for guiding the determination of
assessment endpoints. Table 3-3 outlines sbme suggested criteria for determining assessment
endpoints. Practitioners should select assessment endpoints that provide useful information for
characterizing potential impacts. The assessment endpoints should be selected in an unbiased,
scientifically objective manner to help ensure that results of the LCA are unbias and credible.
1
I
3.2.2 Selecting Measurement Endpoints
i
If the assessment endpoint is not directly measurable, then the practitioner may opt to
select a measurement endpoint as a surrogate for the assessment endpoint. A measurement
endpoint is a measurable characteristic of an impact that can be related to a specific assessment
endpoint (EPA, 1992b). When selecting measurement endpoints there may be properties of a
specific inventory item, or group of inventory items, for which a surrogate measure (i.e.,
measurement endpoint) of potential impact can be used. For example, the acid deposition
potential of a given amount of SO2 emissions can be used as a surrogate to link that quantity of
SO2 emissions to impacts such as leaching o[f metals from soils, tree damage, or fish mortality. If
the assessment endpoint is directly measurable, then it can be used as a measurement endpoint.
i'
Because a number of possible measurement endpoints may be available, practitioners
need to determine the most appropriate and useful endpoint before beginning the characterization
phase of an impact assessment. Using the key LCA decision points as a guide or a set of
selection criteria may be helpful when choosing measurement endpoints. Some possible criteria
for selecting measurement endpoints that are specific to impact assessment include the following:
the relevance of the measurement e|ndpoint to the goals and scope of the LCA,
3-12
-------
TABLE 3-3. SUGGESTED CRITERIA FOR DETERMINING ASSESSMENT
ENDPOINTS
Criteria
Description
Study goals
Study scope
Magnitude of
environmental
loading
Environmental
relevance
Level
Stakeholder values
Data availability
Good communication between the analyst and the decisionmaker(s) is
important to ensure that the chosen assessment endpoints appropriately meet
and complement the goals and objectives of the study.
Scoping helps to ensure that the goals and objectives of the study are met.
The scope of the study defines not only the spatial and temporal boundaries
of potential impacts considered but also defines such factors as the intended
end use or application of the impact assessment results. If the scope of the
study is defined to consider site-specific impacts of deforestation, then site-
specific impacts would constitute appropriate assessment endpoints.
The magnitude of environmental loadings as quantified in the inventory
analysis could be used to further delimit areas to focus more detailed levels
of impact assessment. However, it would be redundant to use the magnitude
of environmental loadings as a decision point for more simplistic impact
assessment methods (e.g., less is better, relative magnitude).
Environmentally relevant assessment endpoints reflect important
characteristics of the natural environmental system and are functionally
related to other possible endpoints. Changes at higher levels of organization
may be of greater significance because of their potential for causing major
impacts at lower levels of organization.
The most appropriate assessment endpoint is the earliest impact (i.e., nearest
in time to the release of an inventory item to the environment) that allows
one to distinguish between alternative impacts or alternative systems. This
criterion is most applicable to comparative studies.
Stakeholder (including societal) values can range from protection of
endangered species to preservation of environmental attributes for functional
reasons (e.g., floodwater retention by wetlands) or aesthetic reasons (e.g.,
visibility in the Grand Canyon).
Data availability is a limiting factor .that cuts across all fields of research. In
some cases, data may be more readily available for one assessment endpoint
than another, thus making it a more attractive candidate. However, the
convenience of readily available data should not be in lieu of quality. The
quality of the available data should be evaluated against previously
developed data quality goals.
Source: EPA, 1992d.
3-13
-------
the.consistency of amendpoint with the scope and boundaries of the inventory analysis,
the intended application or endvus^ofithftimpactsassessment results, ,;'>,
data limitations, ; , ,. fc
;. «j» the availability of impact assessment models, and
the ease of characterizing potential impacts (i.e., direct versus indirect impacts).
!
I
Ideally, the characterization phase Will quantify the relationship between an inventory
item and an assessment endpoint. When an assessment endpoint can be directly measured, this
process can be relatively straightforward. When it cannot be measured, the practitioner must
establish the relationship between the inventory item and a chosen measurement endpoint. The
practitioner might also use additional extrapolations, analyses, and assumptions to predict or
infer changes in the assessment endpoint. It is critical to make these methods and assumptions
clear in the final impact assessment results.
[
3.2.3 Applying Characterization Models to Develop Impact Descriptors
The ability to characterize measurement endpoints hinges on the availability and use of
specific impact assessment tools, called characterization models, to describe the contribution of
specific inventory items to impacts. The preliminary framework for this characterization activity
is contained in a five-tiered hierarchy of characterization models, as described in Table 3-4.
This five-tiered hierarchy is based on discussions from the February 1992 SETAC Life-Cycle
Impact Analysis Workshop (see SETAC, 1993) and the October 1992 SETAC Life-Cycle Data
Quality Workshop. !
A primary concern of this characterization activity is the lack of available data for
conducting many levels of assessment. At present, data requirements generally increase and data
availability generally decreases moving from Tier 1- to Tier 5-type assessments. A recent
SETAC-sponsored LCA Data Quality Workshop in Wintergreen, Virginia, recognized that
currently available environmental input and output data can only support some Tier 2- to
Tier 3-type models. As shown in Table 3-4,|advancing to Tier 2- and 3-type assessments
requires equivalency factors and chemical-properties (i.e., toxicity, persistence, and
bioaccumulation) data. Proceeding to Tier 4- and Tier 5-type models requires high quality,
process-specific, unaggregated, and unavera^ed inventory data. Data produced from the
inventory analysis are currently unable to support most Tier 4- and Tier 5-type assessments.
Developing a publicly available database specifically designed for use in LCA to serve as a
3-14
-------
clearinghouse for generic information supporting LCAs as well as for other types of residuals-
based analyses is a high-priority item within the LCA community.
TABLE 3-4. CHARACTERIZATION MODELS: TIERS OF COMPLEXITY AND
ASSOCIATED DATA NEEDS
Tier
Description
Data Needs
Tier 1: Loading
Assessment
Tier 2: Equivalency
Assessment
Tier 3: Toxicity,
Persistence, and
Bioaccumulation
Assessment
Tier 4: Generic
Exposure/Effects
Assessment
Tier 5: Site-Specific
Exposure/Effects
Assessment
=====
Source: SETAC, 1993
Inventory data alone are used to evaluate
on the basis of quantity or volume with
the assumption that "less is better."
Algorithms based on hazard information
are used to derive impact equivalency
units to evaluate inventory items within a_
specific impact category.
Interactive properties between a chemical
and an organism (toxicity) and an
ecosystem (persistence and
bioaccumulation) are used to evaluate
inventory items. ,--"
Generic environmental or human health
information are used to estimate potential
impacts of inventory items.
Site-specific environmental or human
health information is used to estimate
potential impacts of inventory items.
Mass, volume, or other units of
physical quantity of inventory items.
Same as Tier 1, plus algorithms for
equivalency conversions. Also can
include resource stock and yield, and
non-chemical loading information.
Same as Tier 1, plus information on
characteristics of chemical
interactions with organisms (toxicity)
and ecosystems (persistence,
^bioaccumulation). ^^^s^^^.
Same as Tier 1, plus generic
environmental information and
regional calibration model.
Same as Tier 1, plus site-specific
environmental information and a site-
specific calibration model.
Loading Assessment-"
Loading assessment is based on the
toctematioB method, to loading assessment, the data generatcdon^einventory analysis are
directly used to identify areas where impacts can be ndooed through reductionists and
outputs. Loading assessment does not assess-^alitatmfly or qaantitatively-the impact^
those inputs and outputs or the benefits of their reduction.
3-15
-------
Equivalency Assessment
i
Equivalency assessmentjncludes approaches that translate inventory items into common
units (via the use of equivalency-factors) of jmpactthat can either be evaluatedto compare the
individual contributions of inventory-items to impactsjor resulting equivalency units to assess the
collective contribution of items to impacts, fequivalency factors are based on mechanisms of
impact that relate groups of inventory items |to specific impacts. Equivalency units can be
aggregated within impact categories to provide an estimate of the total level of impact. This
method essentially consists of multiplying the values for groups of inventory items (e.g.,
greenhouse gases) by the appropriate equivalency factors, thus expressing the inventory items in
equivalency units (e.g., global warming potential).
i
Toxidty, Persistence, and Bioaccumulation Assessment
Toxicity, persistence, and bioaccumulation assessment includes those approaches that are
more comprehensive than the Tier 2 equivalency assessment approaches because they take into
account not only hazard but also ecosystem and organism exposure information. Specifically,
these models often focus on properties such as toxicity as an indicator of hazard and persistence
and bioaccumulation as indicators of exposure. The main premise of these models is to use
information on the inherent properties of substances to assess the potential impacts of chemical
substances on the environment.
Information on the inherent properties of many chemical substances can be found in the
literature (e.g., environmental fate of organi4 chemicals or fate-and-transport literature). It can
also be predicted using computer databases (e.g., Aquatic Toxicity Information Retrieval,
AQUIRE, for water) and models (e.g., Regional Acidification Information and Simulation,
RAINS, for acid precipitation). |
Generic Exposure/Effects Assessment
Generic exposure/effects assessment isrnenext higher level of complexity that includes
approaches that use generjp environmental aiiid human health information to model the potential
impacts of inventory items on a generic level. These generic approaches typically utilize
computer-based models to determine the fate^ transport, and partitioning of substances released
to hypothetical, computer-generated "environments." The computer-generated environments
contain standardized information on the main components of the environment (i.e., atmosphere,
hydrosphere, soil, and biota [plants, animals, [and microorganisms]).
3-16
-------
Site-Specific Exposure/Effects Assessment
Site-specific exposure/effects assessment approaches utilize general to site-specific
environmental and human health information to provide site-specific information on potential
impacts. It should be noted that use of detailed, site-specific information should only be needed
in cases where such information is required to clarify the decision to be made. The necessary
time and resource expense of conducting site-specific studies, as well as data availability limits,
makes their applicability to most impact assessments questionable in most cases.
The use of site-specific approaches may be appropriate for some LCAs. However, for
many LCAs, site-specific approaches may not be necessary or desirable.
Central to the characterization phase is choosing the characterization model that is the
appropriate level of detail to complement the key LCA decision points. The objective at this
phase is to match the available data and resources with the minimum level of detail needed to
distinguish between alternative impacts or systems. Using models that provide more detailed
information is only beneficial if the extra effort provides useful information for decisionmaking.
If data or resources are not available to conduct an assessment of the desired level of detail, then
a less detailed model can be used if it provides useful information. If the less detailed tool or
model does not provide useful information, then the characterization might not be worthwhile.
Figure 3-7 illustrates the decision process through which practitioners choose the
characterization model of appropriate level of detail.
3.2.4 Impact Descriptors
The application of characterization models provides an initial description of impact,
called impact descriptors. When the characterized impact is both the measurement and
assessment endpoint, the practitioner may be able to proceed to the valuation phase of impact
assessment relatively easily, provided the practitioner derived the appropriate information for
satisfying the key LCA decision points. If the measurement endpoint is used as a surrogate
measure for the assessment endpoint, the practitioner may need to relate that measurement
endpoint to the assessment endpoint in some manner. One problem with relating measurement to
assessment endpoints is that the specific type of output produced is not yet clear, because many
models have not been applied in the context of LCA.
3-17
-------
Begin Characterization
Exerdise (n = 0)
i
Select Characterization
Model (n + 1)
Apply model and evaluate result(s). Does the
result(s) satisfy any part of the GOALS to
the degree required by stakeholders.
Recognize and record the result(s). Does
some other part of the GOALS require the
use of a more detailed/complex model?
No
Summarize Results
Yes
Figure 3-7. Exercise for Choosing Characterization Models
3-18
-------
The application of characterization models in impact assessment achieves some
aggregation of the inventory analysis and impact characterizations stages of LCA, resulting in a
simpler set of impact descriptors within each impact category (SETAC, 1993). Impact
descriptors can include quantitative (e.g., numerical level of increase in local troposphenc ozone
buildup) and/or qualitative (e.g., descriptive estimate [high, medium, low] of threats to regional
wildlife populations) information. In other words, impact descriptors can quantitatively or
qualitatively characterize the relationship between specific inventory items and specific impact
categories.
3.3 VALUATION
Once a set of impact descriptors has been developed that as concisely and technically
possible characterizes the relevant environmental impacts being assessed, the explicit application
of valuation methods is appropriate (SETAC, 1993). The valuation phase essentially involves
assigning relative values or weights to impacts based on the integration of stakeholder values and
the associated impact descriptors.
The main objective of valuation is to establish the relative importance (based on
stakeholder values) of multiple impacts to aid in the LCA user's decisionmaking P"ce-
Therefore, the practitioner's primary task is to adequately capture and express to decisionmakers
the full range of potential impacts relevant to the LCA, without overwhelming his/her audience
with information. The practitioner should express these impacts so that determining critical
impact areas on which to focus further research and/or improvement efforts is understood.
Although widely practiced, implicitly and explicitly, in the LCA community, the
valuation stage is the least developed of the three impact assessment stages. In general,
valuation includes the following activities:
» identifying the underlying values of stakeholders,
determining weights to place on impacts, and
applying weights to impact descriptors.
Making successful decisions based on impact assessment requires considering all
assessment results and technical information. In addition, decisions are not solely based on the
precision of measurement but also on how measurements are interpreted in terms of imprecisely
understood study goals and stakeholder values. Although developing a truly objective method
for valuation may be both impossible and inappropriate, several conceptual and methodological
approaches to valuation have been developed (see Chapter 6).
3-19
-------
-------
CHAPTER 4
EXISTING METHODS FOR CHARACTERIZING IMPACTS
This chapter provides descriptions of various types of methods for characterizing impacts
that have been discussed, presented, or used in the context of LCA. The methods described in
this chapter include those that focus on impacts to ecosystems, human health, and/or natural
resources. Methods that have specifically been designed to assess resource depletion have
historically been kept separate from those methods to assess environmental impacts and are
described in Chapter 5. Integrative impact assessment methods that contain a combination of
classification, characterization, and/or valuation activities are presented in Chapter 7. Methods
evaluated for this document that exhibit potential applicability to impact assessment but which
have not been discussed, presented, or used in an LCA context are described in Appendix B.
Although the methods included in this chapter span the three main impact categories (i.e.,
ecosystem, human health, and natural resources) used in impact assessment, some of the methods
are clearly more appropriate for assessing specific impact categories and will be identified as
such. Also, because some methods in this chapter do not fit nicely in the generally established
five-tier hierarchy of detail for impact characterization, as discussed in Chapter 3, they have not
been grouped and/or presented by tier of analysis. The methods however, are presented in the
order of increasing level of detail (i.e., from Tier 1 to Tier 5). Table 4-1 provides summary
information on each of the methods profiled in this chapter.
4.1 CHECKLIST APPROACH
Inventory analysis provides a quantified listing of inputs and outputs at various stages of
the life cycle for a defined product system. The data generated in inventory analysis typically are
provided for the weight or volume of input or output (per unit of production or time) either by
life-cycle stage or by total for the entire life cycle. Such data alone can be used directly to
identify stages in the life cycle where outputs can be decreased. However, the checklist approach
merely compares the data generated in the inventory analysis and does not measure impacts.
More detailed levels of impact assessment may be required to distinguish the relative
environmental importance of various inventory items. Another use of loading data is to compare
the overall output levels between alternative products or production systems.
4-1
-------
TABLE 4-1. SUMMARYOF METHODS TO! CHARACTERIZEIMPACTS
Impact Categories Covered
Tier of Detail"
Method
Human
Ecosystem Health
Natural
Resources
Checklist
Relative Magnitude
Environmental
Standards Relation
Impact Potentials
Critical Volume
Environmental Priority
Strategy
Tellus Ranking
TPBP
Unit World
Canonical Environment
Ecological Risk
Assessment
Human Health Risk
Assessment
"NOTE: Methods are not necessarily confined to any single tier of detail.
I
i
The checklist approach is basically a classification matrix that can be used to correlate
specific inventory items with specific impacts or impact categories. The checklist allows for the
information developed in the inventory analysis to be organized in a meaningful way to provide a
quick overview of qualitative impact information. As shown in Table 4-2, the checklist is
arranged so that the presence or absence of specific impacts can be clearly shown.
Strengths ,
The main strength of the checklist approach is its simplicity. Inventory data alone can be
used directly without modification, and a simplified view of cause/effect relationships is
provided by qualitatively associating impacts and inputs and outputs.
14-2
-------
TABLE 4-2. EXAMPLE CHECKLIST FOR ECOSYSTEM IMPACTS
Inventory Items
Ecosystem Impacts
C02
SO2
03
CFC
Solid
Waste
Oil
Effluent
Crude Oil
Use
Atmosphere
Toxicity
Ozone depletion
Greenhouse effect
Visibility
Smog/fog
Ground ozone
Climate change
Water
Toxicity
Contamination
Depletion
Thermal
Turbidity
Acidification
Nullification
Eutrophication
Chemical change
Soil
Toxicity
Salinity
Laterization
Podzolization
Erosion
Other
Geomorphic
Biodiversity
Habitat alteration
In addition to convenience and ease, other strengths of the checklist approach include the
following:
4-3
-------
:-,? .identifying 'areas for reducing environmental-inputs and/oroutputs, and
' ."r: 'comparing levelso£inputsand/or'outputsbetween;alternativeinaterials,;processes,'or :
products. I
I ' { ">:'
Weaknesses
Although simplicity is the chief strength of the checklist approach, it is also its main
weakness. It is critical to recognize that the checklist approach does not actually assess the
occurrence of potential impacts or their relativ^ magnitudes.
Some additional weaknesses of the checklist approach include the following:
choices for environmental improvement are difficult to justify or defend scientifically,
i
improvements in environmental conditions may not be achieved because potential
impacts are not assessed, }
I;
resources may be wasted on improvepient actions that were not part of the real
environmental issues, and ;
i
opportunities for environmental improvements may have been missed (SETAC, 1993).
\
Relevance to Impact Assessment |
The checklist approach alone can be usjed to identify stages in the life cycle where outputs
can be decreased. The checklist provides a tool to evaluate the data generated in the inventory
analysis. However, the checklist approach does not measure impacts. More detailed levels of
impact assessment may be required to distinguish the relative environmental importance of
various inventory items. Use of the checklist approach would be more appropriate for internal
applications until guidelines are established fo|r the external use of such techniques.
Another use of the checklist approach is to quickly and easily compare the overall input
and output levels between alternative product^ or production systems. Such "quick and dirty"
comparisons may not only help identify someikey differences between alternatives but also help
pinpoint areas to focus more detailed level of Analyses.
(
4.2 RELATIVE MAGNITUDE APPROACH
The relative magnitude approach is another form of loading assessment in which the
input and output data generated in the inventory analysis are associated with specific impact
categories. Within the specific impact categories, inventory items are further grouped into
4-4
-------
subranges based on the level (quantity or volume) of inputs or outputs, thus indicating the
relative contribution of various inventory items to specific impacts.
When using the relative magnitude approach, the assigned subrange values may be either
subjectively or objectively based. Subjectively based subranges could use a scoring range of 1 to
10 for example, where the inventory items with the lowest quantity would receive a score of 1,
and the inventory items with the highest quantity would receive a score of 10. Quantities in
between these two bounds can then be extrapolated. Objectively based subranges would use data
from the inventory analysis directly (i.e., the actual quantities) as subrange values.
A hypothetical illustration of the type of output derived from the use of the relative
magnitude approach is shown in Table 4-3. Although Table 4-3 focuses only on impacts to
ecosystems, it can also be used to assess impacts to human health and natural resources.
TABLE 4-3 HYPOTHETICAL EXAMPLE OF THE RELATIVE MAGNITUDE
APPROACH FOR ECOSYSTEM IMPACTS
Ecosystem Impact Category
Greenhouse Effect
Acidification
Habitat Alteration
Inventory Item
CO2
CH4
N2O
03
CFC
SO2'
NO
NO,
Timber
Coal
Iron Ore
Quantity
(tons)
4.000
0.403
0.173
0.009
0.001
1.380
0.470
0.053
6.000
5.500
0.950
Subrange
Score
10
2
2
1
1
10
4
1
10
9
1
4-5
-------
Strengths
.: The relative magnitude approaches' relatively easy to use; it is based on cause/effect
linkages and takes into account the relative quantities of inventory items. Infaddition, ihe relative
magnitude approach can be helpful for j
i
identifying areas for reducing environmental inputs and/or outputs, and
comparing levels of inputs and/or outputs between alternative materials, processes, or
products.
Weaknesses
The primary drawback of the relative magnitude approach is its limited capability for
comparing different subranges. In addition, as'with many other loading assessment approaches,
the significance of impacts may be misrepresented because impacts are not measured directly.
Some additional weaknesses of the relative magnitude approach include the following:
choices for environmental improvement are difficult to justify or defend scientifically,
improvements in environmental conditions may not be achieved because potential
impacts are not assessed, j
resources may be wasted on improvement actions that were not part of the real
environmental issues, and :
opportunities for environmental improvements may have been missed (SETAC, 1993).
Relevance to Impact Assessment !
The relative magnitude approach can provide a useful screening tool in impact
assessment to quickly evaluate the inventory items and impacts that are most significant to the
LCA. It may prove particularly useful for screening large numbers of inventory items and
impacts. Similar to the checklist approach, however, the relative magnitude approach merely is a
tool to evaluate data generated in the inventory analysis and does not provide measures of
impact. Thus it is more appropriate to use this approach for internal rather than external
application or possibly as a screening tool to pmpoint areas where a more detailed level of
analysis is needed. j
f
4.3 ENVIRONMENTAL STANDARDS RELATION (ESR)
The ESR method is a weighting scheme originally developed by Schaltegger and Sturm
(1993) to evaluate the environmental impacts Of chemical releases in Switzerland. The purpose
4-6
-------
of ESR is to assess chemical releases to air, land, and water based on their relative potential
ecological and human impact. The information produced from applying ESR can be used to
evaluate and compare the relative environmental impacts of alternative products and process or
alternative industries. ESR can also be applied to a single process, a collection of processes, or
entire systems. Although the use of ESR provides a consistent estimate of the environmental
impacts, it does not necessarily preclude the need for additional analyses.
The approach for developing the weights used in ESR is shown in Table 4-4. First, the
approach identifies ambient standards (i.e., target concentrations) established by regulatory
agencies for chemical levels in air, land, and water that are meant to protect ecosystems and
human health. Second, the relationships between the standards were made explicit by converting
the ambient standard concentration for each substance in each medium into milligrams per mole.
This results in substance- and media-specific standards that are directly comparable. The final
step consists of identifying the largest value in all media (in this case substance B, water
standard) and then dividing that value by all other values to derive the individual weighting
factors. This results in substance- and media-specific weighting factors that are relative to every
other chemical in each medium.
The weighting factors have the dimension of pollution units per kilogram (PU/kg) of
substance. The environmental impact of a specific chemical release is thus calculated by
multiplying the quantity of the released substance by its associated substance- and media-specific
weighting factor. The equation for calculating pollution units is as follows:
Pollution Units (PU) = Chemical Emission x Chemical- and Media-Specific
Weighting Factors
Developing an ESR weighting scheme for the U.S. will not be as straightforward as it
was for Switzerland because regulatory standards are much more complex in the U.S. Ideally,
regulatory defined and objectively tested ambient standards would be available for all chemical
releases of interest to all environmental media. However, this ideal situation does not exist.
First, many possible regulatory standards are available in the U.S. for air and water. Second,
many of these standards are available for only a few chemicals under one regulatory framework
(e.g., National Ambient Air Quality Standards, NAAQS, applies to only six chemicals).
4-7
-------
Therefore establishing a:decision rule forpriorjitizing'the standards that shouldibe usedimthe
"weighting scheme is necessary.
k
i '
TABLE 4-4. EXAMPLE&PPROAH FOR DEVELOPINGIENVIRONMENTAL
STANDARDS RELATION WEIGHTS
Ambient Air Standard Ambient Air Standard
Substance (nm/m3) Expressed in (me/mole)
A 1
B 12
C 4
Ambient Land Standard I
(me/kg) i
A 5
B 10
C 8
Ambient Water Standard A
(me/I) 1
0.024
0.28
0.095
.mbient Land Standard
Expressed in (me/mole)
0.28
0.57
0.45
mbient Water Standard
Expressed in (me/mole)
A 2 0.036
B 4 0.72
C 3 0.054
Weighting Factor for Air
Emissions Pollution Units
(PU/ks)
30.0
2.6
7.6 .
Weighting Factor for Land
Emissions Pollution Units
(PU/ke)
2.6
1.3
1.6
Weighting Factor for Water
Emissions Pollution Units
(PU/ta)
20
1
13
Source: Grimstead et al., 1993.
One component of the decision rule is !to attempt to develop a weighting scheme using
standards that were developed using a consistent approach that is protective of ecosystems and
human health and welfare so that the weighting factors for each medium are comparable. For
example, if air standards are designed to protect the atmosphere and human health, but water
standards are designed only to protect aquaticj organisms, then the comparison of pollution units
for each medium is less meaningful. It is possible that the water standards would not protect
human health; therefore, the water factors would underestimate the potential impacts.
4-8
-------
Strengths
Some of the advantages of the ESR weighting scheme include the following:
ambient regulatory standards represent social, political, regulatory, and scientific
opinions and values;
weighting factors used in ESR consider human health and ecological welfare;
weighting factors can be derived for all substances that have ambient regulatory
standards and/or regulatory values;
ESR weighting scheme represents the relative impacts of different chemical releases to
different environmental media; and
ESR approach is flexible and can incorporate state, regional, and local regulations for
location specific assessments.
Weaknesses
Scientific information on toxicity and environmental health effects are generally
considered in establishing ambient standards. However, the ESR weighting scheme's use of
relations between ambient standards is not a thoroughly scientific or ecotoxicological-based
scheme but instead represents a socio-cultural judgment from an ecological perspective (which
relies on ecotoxicological data). No completely objective and undoubtedly valid opinion on the
harmfulness of substances exists because of uncertainties in data. Weights for specific pollutants
are developed in the ESR method according to generally accepted norms and values, which are
theoretically expressed in ambient concentration standards. Such ambient standards may or may
not reflect actual environmental impacts.
In addition, the ESR weighting scheme only considers chemical releases. There is no
way to account for the environmental impacts resulting from raw materials use, energy
consumption, and nonchemical stresses (e.g., noise, heat). It is also critical to recognize that the
number of pollution units derived in the ESR weighting scheme represents only one dimension
of the overall environmental impact, namely those resulting from pollutant releases. For
example, the alteration of pristine habitat, the erosion of fertile top soil, and similar impacts
represent a devaluation of environmental assets that is not captured by the pollution units.
Relevance to Impact Assessment
The information produced from applying the ESR can be used in impact assessment to
evaluate and compare the relative environmental impacts of inventory items where regulatory
4-9
-------
standards'exist. Prelinunaryiworieis being performed to develop pollutant weighting factors
v based on U.S. regulatory standards. .The usex>£;guidelines and reference concentrations, which
are not regulatory standards, mayialsoiprove to-be^useMforithetype of analysis-xpspeciallyiif
, these*reference levels are more strongly based on health and environmentalfeffectsjatherjthan.
technical x>r economic concerns. For example, EPA inhalation reference concentrations (RfCs)
and Maximum Contaminant Level Goals (MCLGs). Although the use of ESR provides a
consistent estimate of the environmental impacts, it does not necessarily preclude the need for
additional analyses.
4.4 IMPACT POTENTIALS
I
For some categories of impacts, it is currently feasible to use algorithms to estimate the
impact potential of various inventory items. These impact potential algorithms provide a means
of converting different types of data generated in the inventory analysis into a common unit for
comparison and/or aggregation within impact Categories. For example, algorithms for
normalizing the contribution of substances to iinpact categories such as the greenhouse effect
have been developed to yield the global warming potential of various substances. Aggregating
theses global warming potentials yields a sum [figure that can then be used to assess the collective
contribution of greenhouse gases to global warming or the contribution of individual greenhouse
gases to global warming.
units:
The formula shown below illustrates the generic method for deriving impact equivalency
Inventory Data x Equivalency Factor = Impact Potential
The inventory data are multiplied by an equivalency factor to yield an impact potential value.
Once calculated, the impact potential values can be aggregated within their respective impact
categories to assess their collective contribution to the impact category or they can be assessed
individually. Table 4-5 describes the state-of-jthe-art impact potential functions that are available
for characterizing specific impact categories, j
i
A major concern of impact potentials ijs developing equivalency factors for all impact
categories that relate inventory data to specifi
-------
categories, it is unclear at this time how equivalency factors would be developed for all
categories of impacts.
Impact Potential Example: Ozone Depletioa Potential (ODP)
Halocarbons, in addition to being a greenhouse gas, also destroy the stratospheric ozone
layer that protects all Me from harmful ultraviolet radiation (Graedel and Crutzen, 1990). An
ozone hole, amounting to a 50 percent reduction in ozone concentration, now appears over the
South Pole in the winter months of the northern hemisphere. Although some features of the
Antarctic ozone hole are not fully understood, there is considerable evidence that CfCs are a
major cause (Graedel and Crutzen, 1990).
Ozone depletion typically is considered in impact assessment and is included as a major
impact category in this document. One way to evaluate ozone depletion in the context of life-
cycle impact assessment is to use equivalency units. In this case, ODP units will be used.
Table 4-6 shows the type of output from using the ODP algorithm for various halocarbons
relative to CFC-11.
Strengths
The primary strength of the impact potentials is that they provide a means of normalizing
the contribution of various substances within specific impact categories. This allows for a direct
comparison of inventory items to determine which inventory items, or groups of inventory items,
contribute most significantly to a specific impact category.In addition, most of the impact
potential algorithms are based on cause-and-effect relationships. Thus unlike the checklist and
relative magnitude approaches, the impact potentials indicate an estimated environmental impact
rather than represent the data generated in the inventory analysis.
Weaknesses
One of the general weaknesses with the impact potentials is that many are based on a
large number of assumptions which makes their scientific credibility questionable. In particular,
the functions for human toxicity, terrestrial toxicity, and aquatic ecotoxicity potentials are based
on a number of debated assumptions which include many inconsistencies. Refer to SETAC
(1994) for a complete discussion of the problems and issues related to these three impact
potentials.
4-11
-------
: TABLE 4-5. STATE-OF-THE-ART IMPACT POTENTIAL FUNCTIONS
Impact Potential Function
Description*
GWP,
ODP(x)
03(x)
O3(CFC-11)
HTP,
icomp
TETP,
DRVDR
Icorop
AJETP.
DR./DFS
tcornp
icomp
The, global wanning potential (GWP) of a gas is the time-
integrated commitment to radiative forcing from the instantaneous
release of 1 kg of a trace gas expressed relative to the radiative
forcing of 1 kg of carbon dioxide (COj): where a; is the
instantaneous radiative forcing due to a unit increase in the
concentration of trace gas I, c;(t) is the concentration of the trace
gas I at time t after its release, and T is the number of years over
whfch the calculation is performed.
Thejozone depletion potential (OOP) is defined as the steady-state
ozone reduction calculated for each unit of mass of a gas emitted
per year (as a continuous release) into the atmosphere relative to
thatfor a unit mass emission of CFC-11: where ODP(x) is the
OOP-value of substance x, O^x) is the change in total ozone at
steady-state per unit mass emission rate of substance x, and
O3(CFC-11) is the change in total ozone at steady-state per unit
mass emission rate of CFC-11.
The Human Toxicity Potential (HTP) is defined as the risk due to
an eijnission flux of 1 kg'year"1 of substance I relative to the risk
due to an emission flux of 1 kg-year'1 of a reference substance:
wherte HTPj comp is the HTP-value for substance I initially emitted
to cobipartment comp, DRj is the change in human risk at a change
of enrtission flux DFj comp ( = Dn^ / Dt) of substance I to
compartment comp, and DR^ is the change in human risk at a
change of emission flux DFref comp (= Drn^ / Dt) are the same
quantities for the reference substance ref.
The Terrestrial Ecotoxicity Potential (TETP) is the risk to
terrestrial ecosystems (Rj) through an emission-flux of 1 kg of
substpnce I to compartment comp (DFphenol ^ = Dm/Dt) relative to
the ripk to terrestrial ecosystems (Rp^,) through an emission-flux
of 1 kg phenol to air (DFphenol ^ = Drn^/Dt).
t
f
The Aquatic Ecotoxicity Potential (AETP) is the risk to aquatic
ecosystems (Rj) through an emission-flux of 1 kg of substance I to
compjartment comp (DF,,,,,.,,,,, & = Dm/Dt) relative to the risk to
aquatic ecosystems (Rp^,,,,,) through an emission-flux of 1 kg
phenol to air (DF^^, ^ =
(continued)
4-12
-------
TABLE 4-5. STATE-OF-THE-ART IMPACT POTENTIAL FUNCTIONS
(CONTINUED)
Impact Potential Function
Description"
POCP =
a/b
c/d
AR =
potential HVrai
1 potential HTsO,/mSO.
'2 3W2
, N
The photochemical ozone creation potential (POCP) is the change
in photochemical oxidant production due to a change in emission
of the particular volatile organic compound (VOC) relative to the
change in photochemical oxidant production due to a change in
emission of ethylene: where a is the change in photochemical
oxidant formation due to a change in a VOC emission, b is the
integrated VOC emission up to that time, c is the change in
photochemical oxidant formation due to a change in ethylene
emission, and d is the integrated ethylene emission up to that time.
The acidification potential (AP) is defined as the number of
potential H* equivalents (H^) per mass unit of substance I (rrij)
compared to the number of potential H* equivalents (H+ref) per
mass unit of reference substance (m^); SO2 is the proposed
reference gas.
The nutrification potential (NP) is the potential biomass in terms of
N-equivalents per unit mass emitted of substance I (m{) relative to
the potential biomass in terms of N-equivalents per mass emitted of
a reference substance (m^); PO4 is the proposed reference
substance.
"Many of the functions list in the table are based on a large number of assumptions that are not discussed here.
Source: Guinee and Heijungs, 1993; Guinee, 1992a; and Guinee, 1992b.
Another general weakness with the impact potentials is that only a handful of impact
categories (i.e., those listed in Table 4-5) can currently be accounted for with this method. In
addition, impact potentials may only be useful for chemical-based inventory items, and not all
chemicals are amenable to the development of impact potentials (such as nutrient and oxygen-
demanding chemicals).
In addition, a common set of impact potentials still needs to be developed in order for the
approach to be used in impact assessment. The applicability of the impact potentials to impact
assessment may also be limited because general environmental features or characteristics vary
4-13
-------
according to geographic location.! This wilHeajd to variation among equivalency units and
diminish the utility of a common database of equivalency ..units. Also, while the-development of
"lequivalency.factors is straightforward in principle/frequently exposure^and effects information ,,
; ' ? f.t
', ton 'which equivalency factors could be based is, lacking. Finally,- the multiple mechanisms
involved in environmental processes are difficult to identify, making their incorporation into
equivalency factors even more difficult. i
TABLE 4-6. OZONE DEPLETION [POTENTIAL (ODP) OF SELECT
HALOCARB0N GASES
Gas
CFC-11
CFC-12
CFC-113
CFC-114
CFC-115
Carbon Tetrachloride
HCFC-22
HC FC-123
HCFC-124
HCFC-125
HCFC-134a
ODP Relative to
CFC-11
1.00
1.00
1.07
0.80
0.50
1.08
0.06
0.02
0.02
o
0 '
Gas
HCFC-141b
HCFC- 142b
HCFC-143a
HCFC-152a
Halon-13012
H-1211
H-1202
H-2402
H-1201
H-2401
H-2311
ODP Relative to
CFC-11
0.11
0.06
0
0
16.00
Source: EPA, 1993a.
Relevance to Impact Assessment |
While impact potentials provide a relatively simple means for relating inventory data to
impact categories, as well as a means for aggregating the data, the delineation of equivalency
factors presents a stumbling block. Currently, equivalency factors are being developed for global
warming, ozone depletion, acidification, photochemical ozone, nitrification, and
biochemical/chemical oxygen demand (BOD/COD).
Detailed examples of the use of impact potentials for determining the GWP and ODP of
various emissions are provided in EPA (1993a).
4-14
-------
4.5 CRITICAL VOLUME APPROACH
The critical volume approach is a variation on the impact potential approach that is
applicable to ecosystem and human health impacts. The critical volume approach is used to
determine the volume of air, water, or soil that is needed to dilute specific substances to a
generally estimated toxicity threshold. For example, if it was known that the threshold
concentration for vegetation was 100 kg of chemical X per 1,000 L of soil volume, and 1,000 kg
of chemical X were released, then the critical volume would be 10,000 L of soil.
The results of calculating critical volumes can be grouped into three categories: critical
volumes of air, water, and soil. Table 4-7 illustrates an example of applying the critical volume
method to a hypothetical set of inventory items.
TABLE 4-7, EXAMPLE OF TEE CRITICAL VOLUME APPROACH
Chemical
Release
Air
A
B
C
D
Water
E
F
Soil
G
H
Quantity Ecosystem Threshold Levels
(kg) (kg/L)
57
88
150
632
126
17
1,000
161
.001
.001
.1
.1
.01
.001
.1
.01
Critical Volume
(L)
57,000
88,000
1,500
6,320
12,600
17,000
10,000
16,100
Strengths
The primary strength of the critical volume approach is that it provides a means of
normalizing a variety of data to a common measure (i.e., critical volume in liters) of
environmental impact. The critical volume approach is relatively simple and convenient to use
4-15
-------
and produces useful results.llnaddition,.:similair to!the;impact potentials described in Section 4.4,
the calculations used in the critical volume approach are basedon toxicity and exposure
" concepts'.'which are already familiar to environmental analysts.
. . '
Weaknesses (
To efficiently and successfully use the critical volume approach, a new understanding and
methodology for the impact equivalency approach must be created. However, exposure and
toxicity information is generally lacking for m4ny environmental and human impact areas with
l
which to determine critical volume values. j
In addition, the critical volume approach only takes into account the assimilation of one
chemical at a time. This is, the approach does fiot take into account the interaction between
multiple chemical releases to the same environhiental media. For example, Table 4-7 shows
10,000 L and 16,100 L as the critical volumes of soil needed to dilute 1,000 kg and 161 kg of
chemicals G and H, respectively, to generally accepted threshold concentrations. What is not
provided is an indication of how these critical volumes might be affected as both chemicals are
released to the same medium in the same location.
i
Relevance to Impact Assessment ;
The critical volume approach can provide a relatively familiar framework (i.e., exposure
and toxicity concepts) for normalizing and comparing largely different types of inventory items.
In addition, the critical volume approach can provide a simplified means of normalizing data
generated in inventory analysis by expressing them in terms of volumes, which are then
amenable to aggregation into common impact categories.
In the context of impact assessment, th^ critical volume approach would be most useful
for characterizing chemical releases. However, critical volume algorithms are currently available
for only a limited number of chemicals, and the approach does not lend itself to assessing
nonchemical components. |
4.6 ENVIRONMENTAL PRIORITY STRATEGY (EPS)
The Federation of Swedish Industries and the Swedish Environmental Research Institute
initiated an Environmental Priority Strategies (EPS) system in collaboration with the Volvo Car
Corporation. Although based on implicit value judgments regarding the environmental impacts
of various substances, the EPS system nonetheless provides a means of calculating, in semi-
[
quantitative terms, the overall environmental impact of a product system.
4-16
-------
The EPS system employs environmental indices to convert various material uses and
emissions quantified in the inventory analysis into measures of impacts. These indices are
calculated by carrying out the following steps: (1) each material use or emission being evaluated
is assigned one score for each of the factors listed below, and (2) the six factor scores are then
multiplied together to yield a single score. This single score is expressed in a measure called the
environmental load unit (ELU).
The factors that are assigned scores to calculate indices are the following:
1. Scopethe general impression of the environmental impact
2. Distributionthe extent of the area affected
3 Frequency and/or Intensitythe regularity and intensity of the problem
4. Durabilitythe permanence of the effect
5. Contributionsignificance of 1 kg to the total impact
6. Remediabilityrelative cost to reduce the emission
The higher the ELU of a material, the higher its contribution to an impact and vice versa.
Table 4-8 presents selected environmental indices for raw materials and energy use and for
releases to the air, water, and soil.
Once the indices are determined, the environmental load value (ELY) is determined as a
description of the impacts of the material use or emission in question. The ELY is calculated, as
shown below, by multiplying the quantity of the material use or emission by its environmental
index (typically expressed as ELU per kilogram). Table 4-9 illustrates some generic ELVs using
hypothetical inventory analysis data.
Environmental Load Value = Environmental Index (ELU) Quantity
Strengths
The primary strength of the EPS system is its flexible framework, which allows analysts
to normalize impacts for direct comparison of inventory items either within or between specific
impact categories. A number of environmental load indices have been developed to date, thus
allowing for a relatively comprehensive assessment of environmental impacts. Using the
environmental load indices for specific materials and processes enables the user to calculate
ELVs for individual activities, processes, or an entire system.
4-17
-------
TABLE 4-8. SELECT ENVIRONMENTAL INDICES USED IN EPS
*£. Index
v Raw Materials
Co
,Cr
Fe
Mn
Mo
Ni
Pb
Pt
Rh
Sn
V
Oil
Coal
Land
Arable
Forested
Residual
Energy
Oil
Coal
Electricity
Measure
(ELU/kg)
12,300
22.1
0.38
21
4,200
700
363
42,000,000
42,000,000
4,200
42
0.168
0.1
(ELU/m2)
2.93
1.05
0.98
(ELU/kg)
0.33
0.26
0.014
Index
Air Emissions (ELU/kg)
C02
CO
Nox
N20
Sox
CH
PAC
Aldehyde
Hcl
F
Hg
Cd
0.04
0.04
245
0.6
6.03
10.2
600
20
1E-07
10
.
Soil Emissions (ELU/kg)
As
Cd
Cr
Cu
Hg
Ni
Pb
Sn
Ti
Water Emissions
Suspended matter
BOD
COD
TOC
Oil
Phenol
Phosphorus
Nitrogen
DDT
PCB
Dioxin
Al
As
Cd
Cr
Cu
Fe
Hg
Mn
Ni
Pb
Zn
(ELU/kg)
1E-07
0.0001
0.00001
0.00001
0.00001
1
2
10
10,000
10,000
100
1
0.01
10
0.5
0.005
1E-07
10
1E-07
0.001
0.01
0.00001
Source: Swedish Environmental Research Institute, 1991.
Weaknesses I
I
One of the weaknesses of the EPS systems is in its primary assumption that a linear
relationship between ELVs and increasing or decreasing quantities of inventory items exists. In
reality, this relationship is probably not linear but more complex. Another weakness of the EPS
system is its reliance on value judgments to develop the environmental indices used to calculate
4-18
-------
the ELVs. For example, defining the scope or "general impression of the environmental impact"
of the indices is subject to different interpretations by different stakeholders or by different
geographic locations. Also, EPS does not appear to adequately consider relative toxicity of
pollutants. Thus the robustness of the indices is open to debate. The same point could also be
made for the distribution and remediability components.
TABLE 4-9. EXAMPLE ENVIRONMENTAL LOAD VALUES
Inventory Item
Raw Materials
Oil
Fe
Pb
Ni
Mn
Energy
Oil
Electricity
Air Emissions
CO2
NOX
sox
Water Emissions
Oil
BOD
Pb
Environmental Index
Quantity (kq) (ELU)
1,000
450
123
37
25
1,500
20,000
390
375
248
29
58
5
0.168
0.38
363
700
21
0.33
0.014
0.04
245
6.03
0.00001
0.0001
0.01
Environmental Load
Value
168
171
44,649
25,900
525
495
289
15.6
91,875
1,495
0.00029
0.0058
0.05
Source: Swedish Environmental Research Institute, 1991.
Relevance to Impact Assessment
The ELVs developed through the EPS system are normalized measures of environmental
concerns that can be used to compare inventory items, or they can be aggregated to compare life-
cycle stages or entire product systems. The ELVs can also be used to compare the environmental
impact profiles of alternative materials, processes, or products.
4-19
-------
The EPS system is currently used in some European LCAs and thus provides a practical
methodology for assessing the: impacts;of inventory items on the environment. However, ithe
EPS system is currently set up primarily to assess impacts to ecosystems. If developed and 4
'" refined, modules for assessing impacts to humari health and natural resources would enhance the
overall utility of EPS. I
i
4.7 TELLUS INSTITUTE HUMAN HEALTH HAZARD RANKING
One human health hazard ranking approach, used by Tellus Institute in an assessment of
packaging materials (Tellus Institute, 1992a, 1992b, 1992c), groups inventory items into two
main categories for assessment: carcinogens and noncarcinogens. This section describes the
methods used to assess each of these impact groups.
The Tellus approach assesses the relative human health carcinogenic impact of substances
based on a cancer potency factor, which is meaisured in milligrams/kilogram of body weight/day.
The cancer potency factor is designed to represent the cancer risk associated with various
inventory items. Isophorone was chosen as a bjaseline of comparison for the substances, because
it possesses the lowest cancer potency. The calculated potency factors for various substances are
shown in Table 4-10. !
Noncarcinogenic substances were assessed on the basis of each substance's oral reference
dose. While reference doses (RfDs) can be determined by two routes of exposureoral and
inhalationTellus used oral RfDs because many more oral RfDs are available in the literature.
The oral reference dose provides an estimate of the maximum daily level of exposure that will
not cause harm and is measured in milligrams ^ubstance/kilogram body weight/day (Tellus
Institute, 1992b). i
p
i
The higher the RfD, the less toxic the substance, since a higher dose is needed for an
effect to occur. In Tellus's ranking, the inverse of the RfD was used as the ranking factor in
order for the ranking number to be indicative of lower toxicity. The baseline substance for the
noncarcinogenic ranking was xylene, because it has the highest RfD (i.e., smallest inverse),
which indicates xylene is the least toxic of the [set of pollutants. Thus the inverse RfDs are
compared to xylene to derive "xylene equivalents." Table 4-11 illustrates some of the RfDs and
xylene equivalents for specific substances. >
4-20
-------
TABLE 4-10. CARCINOGEN POTENCY FACTORS AND ISOPHORONE
EQUIVALENTS
Substance
Acrylonitrile
Arsenic
Benzene
Beryllium
Bis(2-ehtylhexl) phthalate
1,3-Butadiene
Cadmium
Carbon tetrachloride
Chloroform
4,4-DDT
1 ,4-Dichlorobenzene
1 ,2-Dichloroethane
1 , 1 -Dichloroethylene
1 ,2-Dichloropropane
1 ,3-Dichloropropene
2,4-Dinitrotoluene
2,6-Dinitrotoluene
1 ,2-Dipenylhydrazine
Ethylene oxide
Hexachlorobenzene
Isophorone
Methylene chloride
Nickel
PAHs
Propylene
Styrene
Tetrachloroethylene
1 , 1 , 1 -Trichloroethane
Trichloroethylene
Vinyl Chloride
Cancer Potency
5.40E-01
5.00E+01
2.90E-02
4.30E+00
1.40E-02
l.SOE+00
6.10E+00
1.30E-01
6.10E-03
9.70E-06
2.40E-02
9.10E-02
6.00E-01
6.80E-02
1.80E-01
6.80E-01
6.80E-01
8.00E-01
3.50E-01
1.60E+00
3.90E-03
7.50E-03
8.40E-01
1.15E+01
2.40E-01
3.00E-02
5.10E-02
5.70E-02
1.10E-02
2.30E+00
Isophorone Equivalents
138
12,821
7
1,103
4
462
1,564
33
2
0.00249
6
23
154
17
46
174
174
205
90
410
1
2
215
2,949
62
8
13
15
3
590
Source: Tellus Institute, 1992b.
4-21
-------
TABLE 4-ll.: EXAMPLE *RfDS FOR NONCARCINOGENIC RANKING
\ ' Substance
Acetone
Antimony
*" Arsenic
Barium
Beryllium
Cadmium
Chromium
Copper
Cyanide
4,4-DDT
Fluoride
Hydrogen Sulfide
Lead
Manganese
Mercury
Napthalene
Nickel
Phenol
Selenium
Tin
Toluene
Zinc
Reference
Dose (oral)
l.OOE-01
4.00E-04
l.OOE-03
5.00E-02
5.00E-03
5.00E-04
l.OOE-fOO
3.71E-02
2.00E-02
5.00E-04
6.00E-02
3.00E-03
1.40E-03
2.00E-01
3.00E-04
4.00E-03
2.00E-02
6.00E-01
3.00E-03
6.00E-01
3.00E-01
2.00E-01
1/RfD
10
2,500
1,000
20
200
2,000
1
27
50
2,000
17
333
714
5
3,333
250
50
2
. 333
2
3
4?
"!Xylene Equivalents
20
5,000
2,000
40
400
4,000
2
54
100
4,000
33
667
1,429
10
6,667
500
100
3
667
3
7
10
Source: Tellus Institute, 1992b.
Once the carcinogenic and noncarcinqgenic rankings have been developed, the analyst
may want to determine the relationship between the two groups of substances. To accomplish
this, Tellus used the Occupational Safety and Health Administration (OSHA) permissible
potency level (PEL) figures. For xylene, the PEL is 100 parts xylene per million parts of air
(ppm), and for isophorone, the PEL is 25 pprh. Converting the ppm units into milligrams, the
100 ppm PEL for xylene translates to 433 mg/m3, and the 25 ppm PEL for isophorone translates
into 141 mg/m3. From these conversions, on£ might deduce that a "safe" dose of xylene is three
4-22
-------
times the "safe" dose of isophorone; thus, isophorone has a xylene equivalent factor of three.
This relationship can be used to compare and determine the combined effect of the carcinogenic
and noncarcinogenic groups as shown in Table 4-12. In this approach, Tellus used isophorone
and xylene equivalents. Other possible equivalent factors for impacts to human health include
the following:
PELsspecify the amount of a pollutant to which a worker can be exposed over an
8-hour work day.
Threshold Limit Values (TLVs)specify the amount of a substance a worker can
be exposed to over an 8-hour work day.
Short-Term Exposure Limits (STELs)only established for a small number of
chemicals and may not be useful for assessing potentially large numbers of substances
found in a typical life-cycle inventory.
Immediately Dangerous to Life and Health (IDLHs)only established for a small
number of chemicals and may not be useful for assessing potentially large numbers of
substances found in a typical life-cycle inventory.
Maximum Concentration Levels (MCLs)used by the Safe Drinking Water Act to
establish regulations for pollutants in public water systems. However, MCLs have
only been established for a few substances.
Strengths
The Tellus approach provides a practical example of impact characterization within the
context of LCA. This approach assesses the relative human carcinogenic and noncarcinogenic
impacts of substances based on cancer potency factors and RfDs, respectively. These techniques
take into consideration well-refined and accepted health effects information to estimate relative
toxicity. The Tellus approach also provides a means of normalizing and evaluating the relative
impact of a variety of different substances.
Weaknesses
The main weakness of the Tellus ranking method is its dependence on a relatively
simplistic approach to comparison of cancer to noncancer, and a corresponding lack of
transitivity in ranking substances. In addition, for many substances, cancer potency factors and
RfDs have not yet been established, and establishing these factors in the near future may not be
feasible. In addition to a lack of key information, the Tellus ranking only considers carcinogenic
and noncarcinogenic human health impacts. It does not consider persistence or bioaccumulation,
and does not look at ecological impacts. For the purposes of impact assessment, it may also be
4-23
-------
useful to have a means of .considering the specific human health.impacts (e.g., mutagenic and
teratogenic impacts).
TABLE :442. EXAMPLE HUMAN HEALTH IMPACT
i EQUIVALENCY RANKING
Substance
Arsenic
Benzene
Beryllium
Cadmium
Chloroform
4,4-DDT
Isophorone
Nickel
Styrene
Toluene
Vinyl Chloride
"The combined ranking assumes 1
Combined Rank - .3 RQueinogcnlc
Carcinogens
Isophorone |
Equivalents
12,821
7
1,103
1,564
2 i
2.49E-03
1
215
8
590
Noncarcinogens
Xylene Equivalents
2,000
400
4,000
200
4,000
10
100
10
7
20,231
22
1,854
4,346
102
2,000
7
373
17
7
1 769
Isophorone Equivalent = 3 * Xylene Equivalent
Equivalents) + (Noncarfcinogenic Equivalents)]
Source: Tellus Institute, 1992b.
Relevance to Impact Assessment
The Tellus approach provides a means fpr normalizing and comparing both human
carcinogenic and noncarcinogenic substances to1 perform cross-substance comparisons of the
potential impact of those substances to human health. In addition, the Tellus ranking methods
allows for determining the aggregate contribution of various life-cycle stages or of alternative
products and processes to human health impacts.
4-24
-------
4.8 TOXICITY, PERSISTENCE, AND BIOACCUMULATION PROFILE (TPBP)
The toxicity, persistence, and bioaccumulation (TPBP) considers the potency (toxicity) as
well as the physical and chemical properties of substances to assess their fate and potential
environmental impacts (SETAC, 1993). The TPBP can be used in two ways:
to construct mobility, persistence, and effects profiles for each of the environmental
loading factors listed in the inventory analysis, and
as a screening tool based on identifiable thresholds to determine whether to proceed to a
more detailed level of assessment (i.e., Tier 4 or Tier 5).
Input information used in the TPBP can be found in generally accepted testing studies,
such as the following, which are readily described in the public and private literature:
acute toxicity testing (LC50, EC50, TD50);
chronic toxicity testing (NOEL);
biodegradation (half life, CO2 evolution); and
bioaccumulation (solubility, octanol/water coefficient, bioaccumulation factor)
(SETAC, 1993).
When input data for TPBP are not accessible or do not exist, using predictive structure
activity relationships from computerized databases may be possible (SETAC, 1993). In addition,
for many substances, this information can be predicted using computer models or databases.
Table 4-13 provides an example of how this information is used to describe potential
environmental impacts.
Strengths
The strengths of the TPBP include the following:
impacts to ecosystems and human health may be considered;
input data for the TPBP are generated from generally accepted testing methodologies
(e.g., acute and chronic toxicity testing);
output from the TPBP may be used to identify priority substances for which more
detailed levels of analysis may be desired; and
» TPBP, unlike the previously described methods, considers information on toxicity,
persistence, and bioaccumulation.
4-25
-------
(TABLE 4-13. HYPOTHETICAL EXAMPLE OE TPBP APPROACH
Life-Cycle
Substance
A
B
C
D
E
F
G
H
I
J
K
Life-Cycle Acute Tori
Quantity (kg) (LCcn)
50
32
246
65
97
32
5
43
Chronic
city Toxicity Biodegrada- Bioaccumiu-
(NOEL) tion (half life) lation Factor
-
-
-
...
...
785
1
324
17
Weaknesses
The primary weaknesses of the TPBPjis that it does not consider environmental exposure
and can only be applied to a limited number of substancesmainly chemicals (SETAC, 1993).
It is not clear how the TPBP would be applied to nonchemical components of ecosystems and/or
human health; althougbit seems possible to identify parameter and thresholds, no attempt has
been made to do so (Vigon and Evers, 1992). I
In addition, there is no consensus about the indices or measures that are best to use, and it
is unclear whether estimated values are of acceptable accuracy for LCA or how this information
should be interpreted in the context of impact assessment (Vigon and Evers, 1992). This
approach could be expanded upon by considering other health effects data, such as cancer
potency values and RfDs, and ecological toxicity information.
Relevance to Impact Assessment
The TPBP provides a further level of detail for impact equivalency type assessments
because it considers not only hazard (toxicity) information but also exposure (persistency and
4-26
-------
bioaccumulation) information. The desired characteristics of the results obtained from impact
assessment may require that such exposure information be considered. In addition, exposure
information may be needed to distinguish among ambiguities in impact equivalency results. The
properties of TPBP make it a good candidate to use in developing priority listings of substances
listed in the inventory that may require assessment at a greater level of detail.
4.9 MACKAY UNIT WORLD MODEL
The Mackay unit world model helps to explain the mechanisms and rates by which toxic
substances are transported into and transformed within the natural environment. This approach
was originally developed as a means of assessing the likely environmental behavior and effects
of newly developed or used chemical compounds before their release into the market and the
environment (Mackay, 1979).
The underlying concept of the unit world model is fugacity. Fugacity can be regarded as
the "escaping tendency" of a chemical substance from a phase. It has units of pressure and can
be related to concentration. Just as temperature (°C) can be related to heat concentrations
(cal/m3) using a proportionality constant, to yield a heat capacity (cal/[m3 x °C]), fugacities (/)
can be related to concentrations using a similar fugacity capacity constant Z, with units of
mol/m3atm by the following equation:
C = Z/
where Z depends on temperature, pressure, the nature of the substance, and the medium in which
it is present. Its concentration dependence is usually slight at high dilution.
The physical significance of Z is that it quantifies the capacity of the phase for fugacity.
At a given fugacity, if Z is low, C is lowthus only a small amount of substance is necessary to
exert the escaping tendency. Toxic substances thus tend to accumulate in phases where Z is high
or where high concentrations can be reached without creating high fugacities.
Example: The fugacity capacity Z for oxygen in water at room temperature is 1.5
moym3 atm (i.e., 0.3/0.2). In air it is 40 mol/m3 atm (i.e., 8/0.2), a ratio of about 27,
Oxygen then adopts a concentration in air 27 times that in water. Conclusion: if we can
find Z for a substance for each environmental phase, we can easily calculate how the
substance will partition. It will reach highest concentrations where Z is highest (Mackay
and Paterson, 1981).
4-27
-------
t Fugacity is used in the-unit world approach tofcalculate partitioning. The,timtworld4s a
hypothetical 1 km3 box that containsSvater,'soil, air, sediment, and aquatic biota. Mathe-
matically, the unit world is represented by a se{ of thermodynamic equations that describe the
partitioning and transformation of a chemical introduced into the box. Chemical-specific
parameters are used to predict the partitioning j>f a given quantity of chemical among the
different components of the unit world. To calculate environmental partitioning in the form of
amounts in each medium, it is necessary to assjime volumes for each medium and an amount of
solute. Medium volumes are based on unit wojrld volumes, consisting of 1 km2 with a 10 km
high atmosphere. In this unit world, 30 percent of the area is covered by soil at a depth of 3 cm,
and 70 percent is covered by water at an average depth of 10 cm (with 3 cm of sediment, 5 ppm
volume of suspended solids, and 0.5 ppm biota). The corresponding volumes of these five
components are as follows:
Atmosphereaccessible volume 1011 m3
Soilaccessible volume 105m3
Wateraccessible volume 106m3
. ,i
Sedimentaccessible volume 10 m ;
t
Aquatic Biotain 106m3 I
I
The fugacity for each of the compounds is calculated as follows:
Pure Substance [
A pure substance (solid or liquid) has a fugacity that is approximately equal to its vapor
pressure (Ps). If its molar volume is km3/g mol), then Z = C/f = l/Psv. The
temperature dependencies of Ps and y (and hence Z) are available in handbooks.
Vapor Phase or in the Atmosphere
Fugacity is usually equal to partial pressure (P); thus, from the gas law, if n is mols and
V is volume, Z = C/f = n/VP = 1/RT}. In the vapor phase Z is independent of the nature
of the substance and is usually about 40 g mol/m3 atm.
Liquid Phase or Water Bodies >
Fugacity or partial pressure is usually related to concentration by the Henry's Law
constant, H as P = PC. It follows that Z is simply 1/H. H is easily calculated as the
ratio of pure substance vapor pressure to solubility.
Sorbed Phases
If the sorption partition coefficient Kp is the ratio of sorbed concentration (g/Mg or
ppm) to water concentration (g/m3 oi ppm), and if the sorbent concentration is S(g/m3),
it can be shown that Z is 10"6 KpS/B
4-28
-------
Biotic Phase
If the biota is regarded as part (fraction y) octanol and the volume fraction of biota is B,
then Z is By Kow/H, where Kow is the octanol water partition coefficient. Sorbed and
biotic phases are the most difficult, but recent work indicates that sorption and
bioconcentration can be related to Kow and organic and lipid contents, thus providing
good estimates for Z (Mackay, 1979).
By estimating the rates of transformation of the chemical (due to photolysis, oxidation,
biodegradation, or other processes), the unit world approach can be used to predict steady-state
concentrations, residence times, and removal rates. An example of the type of output
information derived from the unit world approach is shown in Figure 4-1. As shown in this
figure, the Mackay unit world approach provides the user with information on the partitioning
and concentration of a substance in air, water, soil, sediment, suspended solids, and biota. Thus,
this approach enables the user to determine and compare the impact of various substances on
individual components of the environment. For more detailed explanations of the unit world
approach, see Mackay (1979) and Mackay and Paterson (1981).
Strengths
The unit world model is a relatively simple approach that has been refined and widely
used over the past 15 years to quantify the environmental transformation and fate of chemicals.
Input data for the unit world model, which primarily include data on chemical toxicities, exist for
many different chemicals and are available on readily accessible databases, such as the EPA's
AQUIRE database. However, data on toxicities to terrestrial plants and animals are more scarce
(SETAC, 1993).
Weaknesses
Drawbacks of the unit world model are that it focuses only on the fate of chemicals;
human health effects are not included. Some additional weaknesses of the unit world model
include the following:
results cannot be validated by experimental observation,
data are lacking on many chemical substances, and
currently no practical application exists to serve as a case study example.
4-29
-------
Area « 10« m»
" £
; * ^
' *
§
.
*' /"V. /-'~v-v^
\.) : ^\jf ^*~^~^
'£>\_^sJ^III5~^^
?'"
s$v^ ^
iii§l^?*:^-^-~->^
iilili$r^^^^
^^^ai^
0,5 ppm -
5 pom
3 cm T
^&j₯*rTC§£&&l>'
'^^f^^^w^S
^^*^^^^9
^^ jp*:/
''/
*f*$£
£* /
/
^fe«ft»«feg
^Air
^, Soil
^ Water
/^ Biota
'/ , Suspended
' / solids
j-i, ;
Volum. Z Amount.-"' -? Conee'ntritlon
m* mol/(m* P») mo) 'r mol/m' *<9/8 . ,;
10'° 4.03x10-* 55.0 5.50x10"' 6.96x10"'
9x10> 1.90x10' 2.33 2.59x10-' 2.59 x 10"'
7x10* 0.333 31.8 4.54x10-* 6.82x10-'
3.5 62.8 0.003 8.56x10-' 1.28x10"'
35 37.9 0.018 5.17x10'' 5.17x10''
' .Sediment £.1 x 104 37.9 10.9 5.17x10"' 5.17x10-'
/
1000
Fugaclty = 1.J363 x 10"S Pa Temperature = 25 :C
^ Media densities (kg/m3):
"" Air =
* Biota =
1 19 Soil = 1500 Water = 1000
1000 Suspended solids = 1500 Sediment - 1500
Organic carbon contents: .
Soil » Km Susoended. solids = 4%
.-v*. ' 70.. " Sediment = 4%
Figure 4-1. Example Output from the Unit World Model
Source: Mackay, 1979.
Relevance to Impact Assessment
The unit world approach could provide1 a rational and realistic tool for impact assessment
that enables diverse inventory data to be described in terms of environmental medium
partitioning, concentration ratios, and overall persistence. This approach may also enable the
analyst to determine the sensitivity of each characteristic (e.g., persistence) as a function of the
input data, by altering these data and observing the resulting effect.
t
i
One potential problem with using the dnit world model in LCA applications is that in
many cases the user will not know the correct, proportionate amount of chemical released into
the unit world box (Vigon and Evers, 1992). ^or example, if an inventory analysis revealed that
a system released 100 grams of NOX to the air 'per unit production and this 100 grams was
directly inputted into the model, then the equilibrium partitioning would show the relative
4-30
-------
concentrations of NOX in each of the model components. However, without proportionately
loading the 100 grams of NOX into the unit world box, the concentrations produced would be
meaningless for comparison against lexicological standards (Vigon and Evers, 1992).
An LCA-specific case study using the unit world model is needed to better assess its
relevance to impact assessment.
4.10 CANONICAL ENVIRONMENT MODELING
Canonical environment modeling is similar to, but somewhat more complex and realistic
than, the unit world approach. Instead of using a 1 square kilometer unit world, the canonical
approach uses a simulated reference environment, or a "canonical environment," such as a
generalized stream, lake, pond, or other ecosystem type. Canonical environments do not usually
represent any specific real ecosystem; rather, they are representative of a class of ecosystems
within a general region.
In contrast to the relatively small number of parameters needed by the unit world
approach, canonical environment modeling generally requires a wide variety of environmental
parameters (e.g., soil organic matter content, stream flow). Canonical environment models are
routinely used by EPA and other organizations for ecological risk assessments (see Barnthouse et
al., 1984 and 1985; Suter et al., 1985a and 1985b).
To date, many applications of the canonical environment approach have focused
modeling efforts on aquatic systems. However, similar approaches have been established for
assessing the fate of pollutants in terrestrial systems. Examples of such approaches are found in
Barnthouse et al. (1985a and 1985b) and Suter et al. (1984 and 1985). These models simulate
atmospheric dispersion and deposition of pollutants on soil and uptake of pollutants by biota
(plants and animals).
EPA's Office of Toxic Substances (OTS) also uses the canonical environment concept to
evaluate the fate of pesticides in generic rivers, lakes, and estuaries as part of its Exposure
Analysis Modeling System (EXAMS).
Strengths
Canonical environmental modeling provides information on the fate and transformation
of chemical releases in various environmental media (e.g., air, water, soil, biota). Such
4-31
-------
information enables analysts-to consider not o^ly the^level of pollutants released to various
environmental media (i.e., loading assessment)', but also the ultimate fate of those pollutants.
I
' Canonical environmental models&lso have routinely been used by EPA and other
organizations for ecological risk assessments. JAlthough many of these applications have been
for modeling aquatic systems, similar approaches have been developed for terrestrial and
atmospheric modeling. A wide body of practical examples and experience are available for
potential users to draw upon for guidance. '
i
Weaknesses
One weakness of canonical environment modeling is that there are no means to account
for nonchemical factors and to directly account for impacts to human health. In addition, it
might be (in most cases) uncommon that threshold levels of toxicity, etc., will be exceeded by
the environmental releases of any system acting alone in a given region. It is unclear how the
canonical models would handle cumulative releases from multiple facilities within a given
region. Canonical environment modeling also does not measure impacts per se, but rather the
fate and transformation of pollutants in different environmental media. Although such
information can provide a useful proxy for "inipacts," most environmental components have
some level of assimilative capacity, so assuming that the fate of a pollutant in a specific
environmental media will necessarily impact that component can be misleading.
Relevance to Impact Assessment
In the context of impact assessment, canonical environment models (at their present state
of development) would be most useful for characterizing impacts to ecosystems or resource
supplies (e.g., water bodies, forests). Although applications of canonical models to assess
impacts to animal populations have been performed, using these models to assess impacts to
human health is not clear. j
Canonical environment models also mky be useful in impact assessment when
information (including fate and transformation) on an additional level of detail is needed to
distinguish between a number of different pollutants releases to the same or different
environmental media.
4.11 ECOLOGICAL RISK ASSESSMENT
Within the last 3 years, two independent groupsthe EPA Risk Assessment Forum and
the National Academy of Science (NAS) Coriimittee on Risk Assessment Methodologyhave
i4-32
-------
attempted to develop paradigms for ecological risk assessment. EPA defines ecological risk
assessment as a process that evaluates the likelihood that adverse ecological effects may occur or
are occurring as a result of exposure to one or more stressors. Stressors are defined as any
physical, chemical, or biological entity that can induce adverse effects on individuals,
populations, communities, or ecosystems (EPA, 1992a and 1992b).
The EPA's current framework for ecological risk assessment is conceptually similar to
the risk assessment approach used for human health risk assessment, as outlined in the 1983
NAS report, "Risk Assessment in the Federal Government: Managing the Process." However,
ecological risk assessment can be distinguished from human health risk assessment by three
primary concepts:
Ecological risk assessment can consider effects beyond those on individuals of a single
species and examine entire populations.
There is no single set of ecological values to be protected that can be generally applied.
Rather, these values are selected from a number of possibilities based on both scientific
and policy considerations.
« There is an increasing awareness of the need for ecological risk assessments to consider
nonchemical as well as chemical loadings (EPA, 1992a and 1992b).
The EPA conceptual framework for ecological risk assessment is illustrated in Figure 4-2.
This framework consists of three major phases:
1. Problem Formulation: a planning and scoping process that establishes the goals,
breadth, and focus of the risk assessment. The process of problem formulation
begins with characterizing ecological exposure to loadings and ecological effects,
which includes evaluating loading characteristics, evaluating the ecosystem
potentially at risk, and evaluating the expected or observed ecological effects (EPA,
1992a). The output of the problem formulation is a conceptual model that provides a
qualitative description of how a given loading can affect an ecological component.
2. Analysis: a process of developing profiles of environmental exposure and the
effects of stressors that involve two primary activities: characterization of exposure
and characterization of ecological effects. The outputs of this phase of the risk
assessment are exposure and loading-response profiles that serve as input to the risk
characterization phase described below.
3. Risk Characterization: a process that integrates the exposure and effects profiles
(EPA, 1992a). Risk characterization involves two distinct activities: risk estimation
and risk description. The ecological risk summary provides a summary of risk
estimation and uncertainty analysis results and assesses the level of confidence in the
risk estimates through a discussion of the weight of evidence.
4-33
-------
* : ° »
.iiiv
. ,. ;,>!.<»
: Discussion
Between the
Risk Assessor
and
Risk Manager
(Planning)
.1
Ecological Risk Assessment
PROBLEM FORMULATlOf|l
: i
> f *
':.
A
N
A
L
Y
S
1 .-'
S
Charac
Exp
te
of
OS
*
n
rizatio
ure
i ' Characterization
1 of
. Ecological
1 Effects
1 .''' :'
V
t
, V
1 A , - f H
> ! : - - ,
RISK CHARACTERIZATION
1
^
c
w'
tion; Verification and Monitoring
)
Discussion Between the
Risk Assessor and Risk Manager
(Results)
Risk Me
A
_J
Figure 4-2. Conceptual Framework for Ecological Risk Assessment
Source: EPA, 1992a.
^-34
-------
Models for use in ecological risk assessment are currently in developmental stages,
although for some ecological components models do not yet exist. For an overview and detailed
descriptions of specific models and approaches that may be applicable for use in ecological risk
assessments referr to EPA (1992a and 1992b).
Strengths
Among the strengths of site-specific ecological risk assessment is that it provides the
most ecologically relevant understanding on the existence of or lack of chemical-based impacts
to ecosystems (SETAC, 1993). Models and methods used in ecological risk assessment have
been developed and refined for a number of years in a variety of different fields, and analysts can
draw on the considerable amount of practical experience in the use of these methods.
Weaknesses
In addition to the large resource requirements needed to perform a comprehensive
ecological risk assessment, the results of a comprehensive study may not lend themselves to an
analysis of alternative production systems. In addition, virtually all of the existing studies relate
to specific sites with widespread environmental contamination from past disposal practices or the
potential for future environmental contamination (SETAC, 1993).
Relevance to Impact Assessment
Because of both the technical and resource requirements needed to perform a
comprehensive ecological risk assessment, its use in impact assessment would most likely be
limited to LCAs of a reduced scope or be used to assess critical impact areas identified in a less
detailed analysis after being triggered by the outcome of generic exposure/effects modeling
efforts.
A study to evaluate the applicability of the methods and models used in ecological risk
assessment (see EPA, 1992a and 1992b) would allow for a better understanding of the possible
linkages between ecological risk assessment and LCA, but this issue may not be considered high
priority for the following reasons:
ecological risk assessment methods are already being refined for other purposes,
the use of this level of detail (i.e., Tier 5) would be rare in an impact assessment, and
those people performing risk assessments are already familiar with the basic methods.
4-35
-------
4.12 HUMAN HEALTH RISK ASSESSRtENT
A Site-specific exposure/effects assessment can be accomplished through usingjtraditional
.'risk*assessment methodology, which includes ^he following four components: >.
! . . . ' r;
, 1. ' Hazard Identification: involves fathering and evaluating toxicity data on the types
of health injury or disease that may be produced by a chemical and on the conditions
of exposure under which injury or( disease occurs. It may also involve
characterization of the behavior of a chemical within the body and the interactions it ,
undergoes with organs, cells, or e|en parts of cells. Data of the latter type can be
valuable in answering the ultimate question of whether the forms of toxicity known
to be produced by a chemical agent in one population group or in experimental
settings are also likely to be produced in the human population group of interest.
Note that risk is not assessed at this stage; hazard identification is conducted to
determine whether and to what degree it is scientifically correct to infer that toxic
effects observed in one setting will occur in other settings (e.g., are chemicals found
to be carcinogenic or teratogenic ip experimental animals also likely to be so in
adequately exposed humans?). ;
2. Dose-Response Assessment: invplves describing the quantitative relationship
between the amount of exposure to a chemical and the extent of toxic injury or
disease. Data are derived from animal studies or, less frequently, from studies in
exposed human populations. A chemical agent may have many different dose-
response relationships depending on the conditions of exposure (e.g., single versus
repeated exposures) and the response (e.g., cancer or birth defects) being considered.
3. Exposure Assessment: involves describing the nature and size of the various
populations exposed to a chemical agent and the magnitude and duration of their
exposures. The evaluation could concern past exposures, current exposures, or
exposures anticipated in the future.
4. Risk Characterization: involve^ integrating the data and analyses involved in the
other three steps of risk assessment to determine the likelihood that the human
population of interest will experience any of the various forms of toxicity associated
with a chemical under its known or anticipated conditions of exposure (Environ,
1988).
\
The final step in human health risk assessment, risk characterization, is designed to
generate several types of risk estimates from the results of the first three steps. Since a risk
assessment typically focuses on one or two adverse human health effects, it does not reflect the
full range of adverse effects of the agent or agents in question. Various choices for risk measures
exist, as shown in Table 4-14. The risk measure chosen is based on how the risk assessor
[
collects and organizes information as well as the needs of decisionmakers.
4-36
.
-------
TABLE 4-14. MEASURES OF RISK FOR HUMAN HEALTH RISK ASSESSMENT
Risk Measure
Calculation
Description
Individual Lifetime
Risk
Population Risk
Relative Risk
Standardized
Mortality or
Morbidity Ratio
Loss of Life
Expectancy
dose potency
(individual lifetime risk)
(population exposed)
(incidence rate in exposed group) -f
(incidence rate in non-exposed group)
(incidence rate in exposed group) -f
(incidence rate in general population)
(individual lifetime risk) 36 years
where 36 years = average remaining
lifetime
The excess (or increase in)
probability that an individual will
experience a specific adverse effect as
a result of exposure to a risk agent.
The number of cases resulting from
one year of exposure, or the number
of cases occurring in one year's time.
The risk in the exposed population
compared to the unexposed (or
differently exposed) population.
The number of deaths or cases of
disease observed in an exposed group
divided by the number expected.
The days or years of life lost due to a
particular exposure or activity.
Source: CEQ, 1989.
The risk characterization step of human health risk assessment contains a number of areas
where decisions need to be made. Some key decision areas might include the following:
What are the statistical uncertainties in estimating the extent of health effects? How are
these uncertainties to be computed and presented?
What are the biological uncertainties in estimating the extent of health effects? What is
their origin? How will they be estimated? What effect do they have on quantitative
estimates? How will the uncertainties be described to decisionmakers?
Which dose-response assessments and exposure assessments should be used?
Which population groups should be the primary targets for protection and which
provide the most meaningful expression of the health risk? (National Research
Council, 1983)
4-37
-------
Strengths
Among the strengths of site-specific hiimanuhealth risk assessments that it provides the
, toost relevant understanding of the existence of, or lack of,cchemical-based impacts toshuman
'health. Models and methods used in human health risk assessment have been developed and
refined for a number of years in a variety of different fields, so analysts have a considerable
amount of practical experience in using these methods.
i
Weaknesses |
The assumptions regarding the shape of the dose-response curve (e.g., linear versus
nonlinear) and the existence of thresholds belqw which no impact occurs can have a dramatic
effect on the final impact level. For example, MO impact will be estimated if the ambient
concentration associated with a particular emission source is under the threshold (i.e., the highest
value at which no adverse health impacts can be associated with a pollutant). Dose-response
functions are not always available for some impacts of concern. For instance, human health
impacts associated with regulated air pollutants usually have fairly well-documented dose-
response curves, but other impacts of air pollution (e.g., damage to exposed building materials)
have not been fully investigated.
In addition, the analyst must estimate
the population and/or resources at risk to exposure.
This may be as simple as estimating the number of people living in the locale being used in this
case study. However, the exercise can become more complex if only certain portions of the
human population are affected (e.g., asthmatics, children). For ecosystem and natural resource
impacts, the components at risk are often very! difficult to estimate. In most cases, surveys of
vegetation, aquatic populations, and exposed building materials, for example, are needed. If such
information does not already exist, it must be [gathered from the field, which is a very time-
consuming and expensive task. j
A number of different factors govern the degree of contact, or exposure, a person has with
a toxic agent, including the period of time (duration) a person is exposed to the agent, the route
(inhalation, dermal, ingestion, ocular, injection) of exposure, the amount of agent absorbed into
the body by each route of exposure, environmental concentration of specific agents, and the
tolerance of the exposed population to the agent. In addition to these factors, the risk assessor
must also consider the demographic characteristics of the exposed population to determine the
physiological parameters that affect exposure.
4-38
-------
Relevance to Impact Assessment
Because of both the technical and resource requirements needed to perform a
comprehensive human health risk assessment, its use in impact assessment would most likely be
limited to LCAs of a reduced scope or be used to assess critical impact areas identified in a less
detailed analysis after being triggered by the outcome of generic exposure/effects modeling
efforts.
A study to evaluate the applicability of the methods and models used in human health risk
assessment would allow for a better understanding of the possible linkages between human
health risk assessment and LCA, but this issue may not be considered a high priority for the
following reasons:
human health risk assessment methods are already being refined for other purposes,
the use of this level of detail (i.e., Tier 5) would be rare in an impact assessment, and
those performing risk assessments are already familiar with the basic methods.
4-39
-------
-------
CHAPTERS
RESOURCE DEPLETION: ISSUES AND CHARACTERIZATION METHODS
This chapter includes discussions of issues related to the depletion of natural resources
and describes selected methods for characterizing resource depletion that have been discussed or
presented in the context of LCA. Whereas some of the methods profiled in Chapter 4 account
for the degradation of natural resourcesthat is, impacts to the supplies of natural resources
-------
Replenishment
planting trees
planting grain
Processed Material
metal
refined oil
polymers
starch
Raw Material
mineral ore
crude oil
trees
grain
Manufacturing
Manufacture Object
package
automobile
etc.
Landfill
Mining
Distribution
Consumption
Collection
Used Object
Figure 5-1. The Life Cycle of Resources
Source: EPA, 1992c
Base Consumption Rates: values for the current consumption rates are governed by
how clearly the resource is defined and the spatial and temporal scales within which
rates are calculated.
Economic Factors: levels of natural!resource reserves are governed by the supply of
(e.g., higher resources prices may allow for increased exploration of reserves) and
demand for (e.g., lower resources prices typically lead to higher rates of consumption)
natural resources. ;
Substitutability: the use of substitute materials can preclude or reduce the rate of
depletion of natural resource reservesf
(5-2
-------
Induced Consumption: because the state of depletion of various resources can change
over time, the magnitude and timing of induced consumption should be considered
both before and after the recommendations included in the LCA are implemented.
Intrinsic Renewability Rates: the growth rates for various flow resources change
over time (due to, for example, increased fertilization) and must be characterized and
compared to their maximum growth rates limited by the organism (SETAC, 1993).
5.2 SUSTAINABLE DEVELOPMENT AND ITS RELATIONSHIP TO RESOURCE
DEPLETION
The concept of sustainable development is central to any evaluation of the depletion of
natural resourcesboth stock and flow. The term "sustainable development" came into
widespread use in 1987 when the World Commission on Environment and Development (1987)
released its report Our Common Future, in which "sustainable development was defined as
"development that meets the needs of the present generation without compromising the ability of
future generations to meet their own needs."
Since then, sustainable development has taken on a multifaceted definition embodied in a
process of development that achieves the following goals: 1) a level of per-capita consumption
sustainable for an indefinite period of time; 2) distributional equity; 3) environmental protection,
including protection of biological diversity and the continued functioning of complex natural
systems; and 4) participation of all sectors of society in decisionmaking (Ascher and Healy,
1990).
Although the concept of sustainable development is relatively simple to understand,
translating the seemingly simple concept into practice is still confusing. According to
Ruckelshaus (1989), achieving a state of sustainable development would embody the following
beliefs:
1. The human species is part of nature.
Its existence depends on its ability to draw sustenance from a finite natural world;
its continuance depends on its ability to abstain from destroying the natural
systems that regenerate this world. This seems to be the major lesson of the
current environmental situations as well as being a direct corollary of the second
law of thermodynamics.
2. Economic activity must account for the environmental costs of production.
Environmental regulation has made a start here, albeit a small one. The market
has not even begun to be mobilized to preserve the environment; as a
5-3
-------
...,3 consequence»an increasing amount of the "wealth" we create isdn a sense stolen
from our descendants. ;Vi
3. The maintenance of a livable global environment depends on the sustainable
« development of'the entire hwnai family. ;
If 80 percent of the members of our species are poor, we cannot hope to live in a
world at peace; if the poor nations attempt to improve their lot by the methods we
rich have pioneered, the result will eventually be world ecological damage.
I
Although these beliefs seem well intended (and more or less obvious) they currently are
not incorporated into organizational policymaking, unless it is in the organization's best interest
to do sosuch interests would generally include the realization of some benefit from changing
or averting regulations or sanctions. For interests to be changed, three things are required:
A clear set of values consistent with the consciousness of sustainability must be
articulated by leaders in both the public and private sectors.
Motivations that will support these values need to be established.
Institutions must be developed that \yill effectively apply the motivations (Ruckelshaus,
1989).
From an ecological point of view, a nebessary (but not sufficient) condition for
sustainable development is maintaining an adequate environmental resource endowment. This
endowment constitutes the natural capital (assbts) necessary to provide needed and wanted
environmental servicessuch as climate stabilization, food supply, biological waste disposal,
and materials recycling. \
In the context of LCA, only two long-term fates for the inputs and outputs of a
production system are possible: recycling /reuse or dissipative loss. The more materials that are
recycled, the less dissipation to the environment, and vice versa. Dissipative losses must be
made up by replacement from virgin sources. A sustainable industrial state would therefore be
characterized by minimum use of natural resources and recycling of intrinsically toxic or
hazardous materials or any other materials that cause environmental problems.
1
5.3 RESOURCE DEPLETION MODELS
i
The resource depletion models described in this section are "time-metric" models. Such
models arc based on the basic principle that th'e quantity of stock or flow resource reserves
(Rin units of mass) can be measured at any [ point in time tj. Another class of resource
depletion models is known as "value-metric" (models, which generally attempt to maximize the
; 5-4
-------
net value to society of any given resource consumption scheme. In essence, value-metric models
can be used to estimate a benefit-cost ratio derived from producing a product versus consuming
the resources required to produce the product (EPA, 1992c). Because value-metric models
impress a "value" on the used resources, they may be considered valuation methods. Thus
value-metric models are not discussed in this section. For a description of value-metric models,
see Section 6.4 on economic valuation in this document.
The focus of this section is on time-metric resource depletion models. The key factor in
time-metric models is the resource utilization rate, which is expressed as the rate of resource
replenishment (dR/dt) minus the rate of resource consumption (dRc/dt) at time (tj):
Resource Utilization Rate =
dR
= dR,
rdt;
dt
For stock resources, such as fossil fuels and minerals, the rate of resource replenishment
is considered to be zero because it precludes any replenishment that is relevant to human
societies. With a rate of resource replenishment equal to zero, the resource utilization rate will
be negative, and any level of resource consumption will draw down, or deplete, available
reserves of the stock resource. For flow resources, such as trees, the rate of resource utilization
can be negative or positive, depending on whether resources are being consumed more slowly or
more quickly than their rate of replenishment. When calculating the resource utilization rate, a
negative value represents a net resource depletion, while a positive value represents a net
resource accumulation.
Dividing the resource reserves by the rate of resource utilization yields an estimate of the
time (T) until the reserves are completely depleted.
Time Until Depletion of Reserves = T = R/dR/dtj
A positive t-value represents an accumulation of the resource, and the quantity depends
on the magnitude of the t-value. A negative t-value represents resource depletion, where the
magnitude of t represents the time until the resource reserves are completely exhausted.
Strengths
The majority of existing time-metric models are relatively simple to use and
straightforward to understand. These models also provide a normalizing factor for aggregating
resource depletion within a resource category (e.g., fossil fuels) or for comparing the depletion
5-5
-------
of alternative resources (e.g., natural gas and oil). In addition, the time-metric models can
< ;provide an estimate of the remediability:of the [impact1 (Le., the lower the .magnitude ;of the t-
-..' value, the less tractable the impact). | 5, ... : ; :,
i "*"" ;i';' ' .'-
i
Weaknesses I
-?" : The primary weakness of the time-metric models is that they do not account for whether
the replenishment of a flow resource is equal in quality to the original resource pool. For
instance, although old-growth forest products represent a viable flow resource, replenishment by
managed replanting will not return the forest tp its original level of value or qualityat least in
the near future. In addition, the time-metric models do not account for technological advances
that alter the patterns of resource depletion, or) for the potential substitutability of resources in
the future.
Relevance to Impact Assessment
In the context of impact assessment, the depletion of a stock resource using the time-
f
metric models would involve comparing the rejmaining use years with and without the product or
process system, or with and without specified alternatives. In addition, any evaluation of stock-
resource depletion should consider intergenerational equity or social welfare. For instance,
short-term exhaustion of a stock resource would place a higher value on current populations than
future populations. The analytical approach u^ed in time-metric models allows for a clearer
understanding of the distinction between stock! and flow resources at local and global scales and
provides the analyst with specific units for measuring resource depletion.
5.3.1 Resource Consumption Ratio
The resource consumption ratio approach characterizes the depletion of natural resources
by comparing the magnitude of energy and material consumption to available supplies or
reserves (EPA, 1992a). The resource consumption ratio is expressed by the following equation:
Resource Consumption Ratio =
Consumption per unit of use per unit time
! Supply per unit time
Data on the consumption of natural resources per unit use per unit time may be taken
directly from the inventory analysis. Information on the supply, or reserves, of natural resources
can be obtained from public or private sources (e.g., government reports, nongovernmental
organizations [NGO] publications). The information obtained on the supplies of natural
5-6
-------
resources may need to be normalized by conversion to a standard production time unitusually
annual. In addition, data on the supply of natural resources can have various measures for
yields, as well as for resource reserve use rates. Different measures for yields may also need to
be normalized. Table 5-1 provides examples of the application of the resource consumption
ratio to various generic data.
In addition to providing a means for comparing the use of natural resources to existing
supplies, the resource consumption ratio may also be used to assess the degradation of natural
resources resulting from outputs or pollutants. Assessing the degradation of natural resource
supplies could be accomplished by comparing the level of exposure to a pollutant to the
assimilative capacity of the natural resource supply. For example, if the level of exposure of
resource stock A to chemical X is 10,000 kg/year and the assimilative capacity of chemical X to
resource stock A is 7,500 kg/year, then the resource consumption ratio would be 1.33. Used in
this manner, a resource consumption ratio that is greater than 1 signifies that exposure to a
pollutant is greater than the assimilative capacity of the resource stock and is thus a net resource
degradation. A ratio that is less than 1 signifies that the resource is able to assimilate the
pollutant completely. (However, this ratio does not account for exposure to multiple pollutants.)
The resource consumption ratio provides a simple means of normalizing product or
process input data. The normalized figures may serve as indicators of unsustainable resource use
or degradation and/or may be used to compare alternative input materials to identify those that
yield minimal natural resource impacts. In addition, data on the consumption of natural
resources generated in the inventory analysis may be used directly. Information on the supply,
or reserves, of natural resources can be obtained from public or private sources (e.g.,
government reports, NGO publications).
TABLE 5-1 EXAMPLE CALCULATIONS OF GENERIC RESOURCE CONSUMPTION
RATIOS
Natural Resource
Timber
Oil
Coal
Natural Gas
Iron Ore
Input Quantity
(tons/annum)
150,000
2,500
200
575
1,350
Supply/Reserves Resource Consumption
(tons/annum) Ratio
2,600,000,000
150,000,000
500,000,000
37400,000
450,000
5.7E-05
1.7E-05
4.0E-07
1.5E-05
3.0E-03
5-7
-------
r
Significant efforts may be required to lievelop resource supply and exposure information
I .for this approach, and it is not clear whether calculating resource consumption ratios for
'individual products or processes or thedncremental total demand for the resource willibe .,;,;,;
" necessary. In addition, the significance of thd resource consumption ratio is unclear.
5.3.2 Resource Depletion Matrix
The resource depletion matrix is a variation of the time-metric model that provides a
conceptual framework for evaluating both the! local and global depletion of stock and flow
resources. This approach provides a more analytical characterization of stock- and flow-
resource depletion than that obtained from inventory analyses. The more analytical approach
used in this resource depletion matrix allows for a clearer understanding of the distinction
between stock and flow resources at the local and global scales and provides the analyst with
specific units for measuring resource depletion.
For stock resources (e.g., fossil fuels r minerals), measures of depletion are reflected by
their rate of use, or exhaustion, measured in units of time. This concept is expressed by the
following equation: \
(M) = T
where M is mass and T is time. M represent^ the supply of the stock resource and theoretically
has units of time. However, because the rate |of production of stock resources covers such a long
time span, it is assumed that the rate of production is zero.
In the depletion of flow resources (e.g., forest products or water), two attributes must be
considered: (1) the size and rate of consumption of the resource "pool" and (2) the rate of
replenishment (both natural and managed replacement). These two attributes are incorporated in
the following equation: [
(M/T)
I
where M is mass and T is time. The first term in the above equation could be used as a
comparison to the depletion of stock resources because the flow resource whose current rate of
consumption is greater than the rate of replenishment could be depleted in a finite period of time
'< 5-8
-------
if not redirected by management intervention. For example, over-harvesting of certain species
of trees (e.g., mahogany in tropical forests), where the rate of consumption exceeds the rate of
replenishment, will result in the depletion of the resource in a measurable period of time.
The framework for a resource depletion matrix is illustrated in Figure 5-2. This matrix is
divided into four quadrants based on four categories of resources: stock, flow, local, and global.
The cells corresponding to stock resources yield a measure of the time until the resource stock is
depleted. The cells corresponding to flow resources provide a measure of resource depletion,
which may be used to determine the sustainability of the resource use.
Characterizing flow resources is somewhat more complicated because the rate of
replenishment must be considered. In addition, it is not clear whether the replenishment of a
flow resource is equal in quality to the original resource pool. For instance, although old-growth
forest products represent a viable flow resource, replenishment by managed replanting will not
return the forest to its original level of value or qualityat least in the near future.
In the context of impact assessment, using the resource depletion matrix would involve
comparing the remaining use years with and without the product or process system, or with and
without specified alternatives. In addition, any evaluation of stock-resource depletion should
consider intergenerational equity or social welfare. For instance, short-term exhaustion of a
stock resource would place a higher value on current populations than future populations. The
more analytical approach used in this resource depletion matrix allows for a clearer
understanding of the distinction between stock and flow resources at the local and global scales
and provides the analyst with specific units for measurement of resource depletion.
5-9
-------
I
Local
Global
Stock
SL/CL=dL
SG/CG=UG
iFlow
CL/RL = DL
if DL < 1 , then CL - RL = EL
and QL/EL = UL
if DG < 1 , then CG - RG = EG
and QQ / EG = UG
C - consumption rate (amount/unit of time)
S- stock (amount)
U - use-years (time)
D - depletion index (dimensionless)
R- replenishment rate (amount/unit of time)
E- excess consumption rate (amouht/unit of time)
Q- standing quantity (amount)
Figure 5-2. Resource Depletion Matrix
i
Source: SETAC, 1993
$-10
-------
CHAPTER 6
METHODS FOR CONDUCTING VALUATION
The valuation phase of impact assessment involves assigning relative values or weights
to impacts based on their associated descriptors as derived in the characterization phase and
stakeholder values. The primary objective of this valuation exercise is to integrate information
on environmental impacts with stakeholder values to establish the relative importance of impacts
or categories of impacts. Thus the challenge to practitioners is to adequately capture and express
to decisionmakers the full range of potential impacts relevant to the LCA and to the stakeholders
without overwhelming their audience with information.
Making successful decisions based on impact assessment requires considering all
assessment results and technical information. In addition, decisions are not solely based on the
precision of measurement but also on how measurements are interpreted in terms of imprecisely
understood goals and values. Although developing a truly objective method for valuation is both
impossible and inappropriate, several conceptual and methodological approaches to valuation do
exist Those approaches that have been used, presented, or discussed in the context of LCA are
described in this chapter. In addition to the approaches described in this chapter, several
integrated approaches, as discussed in Chapter 7, also contain implicit or explicit valuation
components.
6.1 DECISION ANALYSIS USING MULTI-ATTRIBUTE UTILITY THEORY
(MAUT)
Simply stated, decision analysis is a method that breaks down complex decisions
involving multiple issues into constituent parts or individual attributes to provide a better
understanding of the main factors guiding the decision. Decision analysis using MAUT is useful
when deciding between largely different types of considerations. In addition, it provides a
logical structure for analyzing complex weighting issues.
The first step in decision analysis is to identify all important objectives and attributes.
While this step may seem obvious, it is necessary to ensure that the valuation focuses on the
right problem. The objectives and attributes of the decision at hand may be identified by using
tools such as an objectives hierarchy (Keeney and Raiffa, 1976). Developing an objectives
hierarchy may proceed in either a top-down or bottom-up fashion:
6-1
-------
Top-down:
Bottom-up:
The decisionmaker(s) is asked to identify overall objectives. For
LCA these might be to minimize overall environmental and human
health impacts or t6 maximize public opinion.
, "'4
An exhaustive list of specific attributes of concern to the.
decisionmaker is initially identified. Items in the Initial list of
attributes may then [be aggregated, eliminated, or redefined in the
determination of a final set of attributes. For example, an initial LCA
attribute list might include acid deposition impacts, solid waste
impacts, corporate image, waste disposal cost minimization, etc. The
decisionmaker(s) may decide that acid deposition is not a significant
problem in the region and thus eliminate it from the list, resulting in a
streamlined set of attributes.
Whether the objectives and attributes are determined through a top-down or bottom-up
approach, the final set of attributes should haye certain characteristics. An overall objective
would be at the top and a comprehensive set o'f issue-specific objectives are then derived that are
consistent with the overall objective. Finally,
attributes that are meaningful, measurable, and
predictable are derived for each specific objective. According to Keeney and Raiffa (1976), who
describe the entire MAUT process in detail, the set of attributes should be
i
comprehensive, !
as small as possible in number,
nonoverlapping,
I
judgmentaUy independent, and j
operational.
Decision analysis with multiple issues ior objectives, such as impact assessment, would
include the following steps: [
1. Break the issue or decision down into single objectives and attributes.
2. Utilize the attributes to measurk the degree to which an objective is achieved by a
management option (attributes should be relevant to the issue, measurable,
predictable, comprehensive, anjl nonoverlapping).
3. Identify objectives and attributes that build consensus about the nature of the
issue at hand.
4. Estimate the effects of various actions (decisions) on the attributes.
6-2
-------
An example decision tree outlining objectives and measurable attributes of water
pollution effects as part of the overall objective of environmental improvement is shown in
Figure 6-1. The attributes (e.g., predicted effect on human health) as shown in Figure 6-1
provide a foundation upon which analysts can estimate the effects of various actions.
In the context of impact assessment, where tradeoffs between impacts to ecosystems,
human health, and natural resources must be made, employing decision analysis does not
necessarily require following the above-outlined steps. Decision analysis in impact assessment
would likely include employing a model to predict ecosystem, human health, and natural
resource impacts and associating each impact with a unit of measure or value.
Strengths
The multi-attribute analysis capabilities of MAUT allow for an evaluation of cross-sector
and/or multi-media issues. For example, in using comprehensive environmental assessment
Environmental Improvement
Minimize Air Pollution Effects
Minimize Water Pollution Effects
Minimize Land Pollution Effects
i
I
Minimize the Effect of Discharged
Pollutants on Ecosystems and
Human Health
T
Results of effluent
toxfcity tests
Predicted effect
on human health
Results from biocriteria
assessment downstream
of effluent discharges
Figure 6-1. Details of MAUT Water Pollution Effects Objectives
Source: Modified from SETAC, 1993
6-3
-------
techniques, such as LCA, decisionmakers are pften faced with decisions that can cut across
multiple environmental media (e.g., air pollution, water pollution, solid waste,3esource use).
MAUTprovides a framework for breaking such multi-attribute decisions into,a1 set of «,*
-'*;/-''
measurable attributes from which the analyst
-------
decomposing a problem into its constituent parts and then using simple pakwise compa-EKBffis, tsr
develop priority rankings in each hierarchy.
Steps to follow when using the AHP are described below. Particular steps may be
emphasized more in some situations than in others, and as noted, interaction is generally
necessary.
1. Define the problem and determine what you want to know.
2. Structure the hierarchy from the top (the objectives from a general viewpoint)
through the intermediate levels (criteria on which subsequent levels depend) to
the lowest level (which usually is a list of the alternatives).
3. Construct a set of pairwise comparison matrices for each of the lower levelsone
matrix for each element in the level immediately above. An element in the higher
level is said to be a governing element for those in the lower level. In a complete
simple hierarchy, every element in the lower level affects every element in the
upper level. The elements in the lower level are then compared to one another
based on their effects on the governing elements above. This yields a square
matrix of judgments. The pairwise comparisons are made based on which
element dominates another. These judgments are then expressed as integers. If
element A dominates element B, then the whole number integer (or exact value
with decimals if known) is entered in row A, column B, and the reciprocal
(fraction) is entered in row B, column A. If element B dominates element A, the
reverse occurs.
4. N(n-)/2 judgments are required to develop the set of matrices in Step 3.
(Reciprocals are automatically assigned in each pairwise comparison.)
5. Having made all pairwise comparisons and having entered the data, the
consistency is determined using the eigen value. (Aw = lmax w is determined.
The consistency index uses the departure of lmax from n compared with
corresponding average values for random entries to yield the consistency ratio
CR).
6. Steps 3,4, and 5 are performed, for all levels and clusters in the hierarchy.
7. Hierarchical composition is used to weight the eigen vectors by the weights of the
criteria and the sum is taken over all weighted eigen vector entries corresponding
to those in the next lower level of the hierarchy.
8. The consistency of the entire hierarchy is determined by multiplying each
consistency index by the priority of the corresponding criteria and adding them
together. The result is then divided by the same type of expression using the
random entry corresponding to the dimensions of each matrix weighted by the
priorities as before, so that the CR is about 10 percent or less. If the CR is not 10
percent or less, the quality of the judgments should be improved, perhaps by
6-5
-------
f** revising the manner in which questions are asked in making :pairwise /
comparisons. If this fails to improve consistency, it is Ukely: that the problem
'* x! should be more accurately structured by grouping similar elements under more
: meaningful criteria. A return to! Step 2 would be required, although only the
problematic parts of the hierarchy may need revision.
i* ₯"' 9. To perform absolute measurementthat preserves the rank of the alternatives and
j r satisfies expectations and prior [commitments, each lowest level subcriterion is
":r divided into a complete set of intensities so that an alternative always reflects one
of these intensities, i Then the intensities are pairwise compared according to
perceived importance or priority with respect to that criterion. Finally, the
alternatives are rated one at a time. The intensities for each criterion and the
weighted ratings are added to obtain an overall rank on a ratio scale. Unlike
paired comparisons, the procesls to rate intensities requires expert knowledge. In
most decision problems about the future, there is no such expert knowledge.
Also, experts have been known to have biased and misjudged the importance of
the intensities. In such cases paired comparisons must be used (Saaty, 1992).
i
Applying the AHP approach to the valuation phase of impact assessment is relatively
straightforward. In the AHP example illustrated in Figure 6-2, the overall goal of the LCA
(environmental improvement) is at the top of fhe hierarchy; factors affecting this goal are on the
next level. These factors would probably be the impact descriptors formed in the
characterization phase. Subcriteria at the next level might include economic considerations,
uncertainty, assumptions, judgments, etc. |
t
Strengths
The main strength of the AHP is that ijt provides an efficient framework and procedure
for making individual or group decisions on single or multiple attribute problems. Some
additional strengths of the AHP include the following:
relatively simple and straightforward to use,
available AHP computer software package (called Expert Choice),
overall view of complex relationships inherent in multi-faceted problems and in the
judgment process, and !
flexible enough to handle a wide variety of problem types.
6-6
-------
Environmental
Improvements
Natural Resource
Impact Descriptors
Human Health
Impact Descriptors
Ecosystem
Impact Descriptors
Relative Contribution of
Inventory Items to Impacts
Relative Contribution of
Inventory Items to Impacts
Relative Contribution of
Inventory Items to Impacts
Value Judgments
Value Judgments
Value Judgments
Figure 6-2. Example Framework for AHP Applied to Impact Assessment
Weaknesses
One weakness of the AHP results from the pairwise comparison process. This process
requires expert knowledge to rate the intensities (see Step 9 in the AHP process outlined above)
of the attributes being compared. In the case of most future problems, there is no such expert
knowledge. In addition, the possibility exists that the experts can have a bias and/or might
misjudge the importance of particular attribute intensities. At any rate, because of its reliance on
the values and judgment of a select group of individuals, it is unlikely that the results of an AHP
study could be replicated.
6-7
-------
Relevance to Impact Assessment ,, ;iV s
TherAHP may provide a useful tool for evaluating multi-attribute,;complex problems.
Such problems typify those encountered in the valuation phase of impact assessment where,
aforeseeably, a wide variety of multi-media and/or cross-sector environmental impacts must be
tjconsidered. The AHP also provides a useful-framework for integrating stakeholder values with
-.environmental impacts.
3 MODIFIED DELPHI TECHNIQUEJ
The Delphi Technique is a procedure originally developed by the Rand Corporation for
eliciting and processing the opinions of a group of experts knowledgeable in the various areas
involved. The Delphi Technique addresses thef need to structure a group communication process
to obtain a useful result for a given objective. In essence, the Delphi Technique attempts to
create a structured format to elicit collective knowledge.
In response to a number of shortcomings associated with the Delphi Technique (see
Unstone and Turoff, 1975), a modified Delphi technique has been developed. This modified
Delphi technique provides a systematic and controlled process of queuing and aggregating the
judgments of group members and stresses iteration with feedback to arrive at a convergent
consensus. The weighting system discussed in the following section does not include all the
elements of the original Delphi Technique. In Addition, results of these ranking sessions need
further study, feedback, and substantive input from field data before using.
The weighting procedure can be simply employed. A deck of cards is given to each
person participating in the weighting. In this example each card names a different technical
specialty. Each of the participants is then asked to rank the technical specialties according to
their relative importance to explaining changes in the environment that would result from a
particular system. Then each individual is asked to review the list and make pairwise
comparisons between technical specialties, beginning with the most important specialty. The
most important technical specialty is compared with the next important specialty by each
individual, and the second technical specialty with respect to the first. For example, the first
technical area might receive a weight of 100 percent, and the second most important technical
area might be considered only 90 percent as important as the first. The second and third most
important technical specialties are compared, and the third most important is assigned a number
°ff°r example, 95 percentbased on its relative importance compared to the second most
important technical specialty. A sample diagram of the comparison is presented in Figure 6-3.
6-8
-------
Groundwater
Sociology
This participant judged noise to be
55% as important as ecology in tl~
context of this proposed action.
1
Ecology
.85
Noise
.55
This participant judged ecology to
be 85% as important as sociology in
explaining the environmental effects of
the system. This proportion is based on
the experience and judgment of the
participant.
Figure 6-3. Modified Delphi Technique
Source: Modified from Jain etal., 1993.
The formula for weighting the technical specialties is
V..
v..
IJ
(i = 1,2,3,.- -,n)
x =
ijxu (i = 2,3,...,n)
where
Wr = weight for the ith technical specialty area by the jth scientist,
n = number of technical specialties,
6-9
-------
P = 1,000: total numberof points to be distributed among the technical specialties,
XM «= the jth scientist's assessment bf the ratio of importance of the ith^technical
specialty in relation to the (i-]j)th technical specialties, and ... ,
Vy = measure of relative weight for the3th technical specialty, area by the jth
scientist , j
lis technic
To accomplish the second part of this technique (i.e., to rank attributes within a technical
specialty), each participant or group independently ranks attributes in his or her own specialty.
The information from these pairwise comparisons can then be used to calculate the relative
importance of each of these specialty areas; a fixed number of points (e.g., 1,000) is distributed
among the technical specialties according to individual relative importance.
i
t
After the weights are calculated from the first round of this procedure, the information
about the relative weights is presented again to the experts, a discussion of the weights ensues,
and a second round of pairwise comparisons is made. The process is repeated until the results
become relatively stable in successive rounds.
r
In a demonstration of this method, an interdisciplinary group of college graduates with
very little training was asked to rate the following areas according to their relative importance in
environmental impact analysis and to distribute a 1,000-point total among these categories:
air quality
ecology
water quality
aesthetics
economics
transportation
earth science
sociology
natural resources and energy
* health science
land use
noise
6-10
-------
After a thorough group study of all 12 areas, the group was asked to rate the areas again.
The results, shown in Table 6-1, indicate that although some relative priorities changed, the
points allocated to each category remained essentially the same. Similar ratings may be
developed for attributes within each group.
Strengths
This modified Delphi technique provides a systematic and controlled process of queuing
and aggregating the judgments of group members and stresses iteration with feedback to arrive
at a convergent consensus. Attributes within a technical specialty are ranked by an expert in that
technical specialty and aggregated over the expert panel, thereby creating a structure for ranking
alternative impact areas (see Table 6-1).
TABLE 6-1. EXAMPLE RESULTS OF USING THE MODIFIED DELPHI
PROCEDURE FOR COMPARING ENVIRONMENTAL AREAS
Before Interdisciplinary Study
Water
Air
Natural Resources
Health
Ecology
Land Use
Earth Science
Economics
Sociology
Transportation
Aesthetics
Noise
Average Point
125
122
109
100
97
81
79
62
60
56
54
53
1,000
After Interdisciplinary Study
Area
Water
Air
Natural Resources
Ecology
Health
Earth Science
Land Use
Sociology
Noise
Economics
Transportation
Aesthetics
TOTAL
Average Point
Distribution
128
126
105
93
88
87
78
64
62
62
61
46
1,000
NOTE: The numeric values in this table are particular to a specific case study. A different group would
certainly arrive at different decisions, and any application directed toward comparison between
attributes should be made in the context of a specific planning situation.
Source: Jain et al., 1993.
6-11
-------
Weaknesses : ' \
i
One of the weaknesses of the modified Delphi technique isionelthat typically plagues
most valuation toolsnamely the requirement of expert knowledgehvith which to rate ^ I*
, ^^environmental attributes. In many cases, there is no such expert knowledge. Inr addition, the
possibility exists that the experts can have a bias and/or may misjudge the importance of
particular attribute intensities. !
t ' *
Another weakness of the modified Delphi technique is in the ranking process. This
process requires a wide variety of technical specialists to rank attributes within their respective
technical specialty area. In addition, the results of the ranking sessions may require further
study, feedback, and substantive input from field data before using. Conceivably, a large
amount of time and resources could be spent on such follow-up analysis.
Relevance to Impact Assessment
The information generated from the rriodified Delphi technique may provide a useful
procedure for calculating the relative importance of each specialty (i.e., environmental attributes
or impacts) area. From this, a fixed number o|f points (e.g., 1,000) may be distributed among the
technical specialties, thus indicating the relative importance of individual specialty areas.
However, the level of technical expertise and time required to conduct a thorough evaluation of
each specialty area may limit the application of the modified Delphi technique to valuation.
6.4 LIFE-CYCLE COSTING1 j
A life-cycle inventory would address environmental inputs and outputs of a production
system, while the impact assessment would address the environmental impacts associated with
those inputs and outputs. Life-cycle costing extends impact assessment by taking an additional
step (i.e., placing a dollar value on impacts). Methods for assigning costs are described below.
i
Monetary values for environmental imbacts can be determined for certain types of
impacts. The market value, for instance, of crop loss or damages caused by air pollutants can be
valued directly by assessing the market value of the lost output. However, quantifying an impact
chain leading to revenue loss may be difficultj For example, translating NOX emissions from the
production of a glass bottle into an incremental change in ambient ozone concentration, and
quantifying crop loss from that increment is highly uncertain. In addition, placing monetary
'Portions of this section were summarized from White et al. (forthcoming).
6-12
-------
values on many impacts (e.g., adverse human health effects) is difficult from an economic and
ethical perspective.
For the purpose of analysis, different types of value that individuals place on the
environment have been distinguished: use value, option value, and existence value. Use value is
based on the utility people derive from the "consumption" of the environment for recreational
purposes, such as boating, fishing, and other sporting activities. The option value is the use
value in the presence of uncertainty. People may not consume the environment at present but
may want to do so in the future. Having the option for future use is assumed to be valued by
consumers. Finally, the existence value is the value people assign to the environment for
"altruistic" reasons; it is the utility they derive from the knowledge of the existence of the
environment.
Several methods are available for indirectly valuing impacts by estimating the use,
option, and/or existence values that individuals place on environmental amenities or the
devaluation resulting from environmental harm. These methods involve the following:
1) examining behavioral responses that are, or might be, influenced by an externality;
2) assuming or creating a fictitious market to elicit the value that individuals might assign to an
externality; or 3) analyzing the implicit value placed on pollution abatement by society through
the actions of its regulatory agencies. Methods in each of these three categories are briefly
described below. For detailed descriptions of these methods and their corresponding strengths
and weaknesses, the reader is referred to White et. al. (Forthcoming), Tellus Institute (1992a,
1992b, 1992c), Desvouges et al. (1991 and 1989), and the Organization for Economic Co-
operation and Development (OECD) (1989).
Strengths
One of the main strengths of life-cycle costing is that the basis for measurement (i.e.,
dollars) is a metric that most people can readily understand. Monetary values for environmental
attributes also enable analysts and decisionmakers to directly compare environmental and
economic considerations, whereas environmental and economic decisionmaking have generally
been treated as separate, unrelated entities.
Another strength of life-cycle costing is that the valuation methods and techniques have
been refined over a long period of time, are applicable to a wide variety of impact types, and
offer much practical experience for analysts to draw upon.
6-13
-------
Weaknesses I y;
! **, t
* Life-cycle costing is open to criticism for using economic valuation methods to "price"
environmental attributes (e.g., the extinction of species, loss of pristine forest habitat, or adverse
human health effects). For example, aJcomprenensive estimate of society^willingness to accept
the loss of the spotted owl in the Western United States can easily surpass the GNP of most, if
not all, countries. Some additional criticisms of using monetary values to assess environmental
impacts are the following:
large gap between rich and poor in te|ms of disposable resources for environmental
care, !
. needs of today often.outweigh the needs for tomorrow,
insufficient knowledge to value environmental impacts because the full consequences
of impacts are not fully understood, and
monetary valuation focuses on human needs.
In addition, methods of life-cycle costing often rely on a set of assumptions that may or
may not accurately reflect reality. Some of these assumptions are outlined in the discussion of
specific methods below.
I '
Relevance to Impact Assessment |
Life-cycle costing methods may be useful in the context of impact assessment for
translating impacts into a common metric (i.e., dollars) for direct comparison of impacts within
and between impact categories. The presentation of impacts in monetary terms also can
facilitate decisionmakers' consideration of tradeoffs between environmental and economic
issues. :
One integrated approach to impact assessmentthe EPS Enviro-Accounting Method
outlined in Chapter 7provides an example of the use of economic valuation in the context of
LCA.
i
6.4.1 Hedonic Pricing
Hedonic pricing attempts to identify a surrogate for the nonexistent market for the
environment. Markets that qualify as surrogate markets for the environment are those in which a
private good is traded that may bear some relationship to the public environmental good. The
notion underlying the concept of hedonic prices is that people derive utility from various
-------
attributes of a product. A product has many attributes, some of which can relate to tie presence
of a public good. A house, for example, can have features individual consumers value
differently. Each of these common features commands a price; however, this price is implicit:
individual features of a house are not sold separately. One attribute of the house is the
environment in which it is located.
In theory, one can construct demand functions that depend on these individual
characteristics, and one can derive the amount of money consumers are willing to spend to
obtain one more unit of q, the environmental quality feature. (If q is air quality, then "one more
unit of q" would refer to "one unit less of pollutant," where the "pollutant" could refer to an
index of air pollution.) One would expect to observe differentials in housing prices, depending
on the quality of the specific environment in which they are located.
The notion of a good embodying many characteristics implies that a job, too, has many
characteristics in addition to the wage that it pays. One important characteristic is the risk to the
health and life of the worker. It is argued that workers will only accept a job with high risk
when given a "compensating wage differential." The hedonic wage method relates the size of
wage differentials for various jobs to their lives.
For this approach to measure what it intends to measure, several assumptions must be
made pertaining to the aggregability of individual preferences (see OECD, 1989). In addition, it
is subject to many sources of bias (see OECD, 1989), for example, strategic bias. Because
environmental quality is a public good (once It is provided, people cannot be excluded from its
consumption), people have an incentive to understate their preference (if they are held to pay),
counting on the fact that other people will provide for the supply of the good. This is the free-
rider problem. Also several sources of bias are based on individual rationality. It has been
observed that people respond to the starting value that is quoted to them (source for the "starting
point bias"). In addition, there is also concern about whether the hypothetical markets
correspond well enough to real markets.
Apart from various technical problems (see OECD, 1989), the obvious flaw of this
approach is that it only targets the value of an area for a very specific narrow use. Surely people
value natural resources for more than the amenities they offer. And again, there is no way this
method would allow the contribution of a single pollutant to environmental degradation to be
evaluated.
6-15
-------
*» The derivation of an implicit price for an environmental characteristic from an ideal type
demand function is rarely a straightforward calculation. Estimating these implicit processes
from observable market data, however, requires strong assumptions and is not without problems.
Apart from the usual assumptions about the structure of individual utility functions relating to
aggregability, it has to be assumed that peoplejhave a wide enough array of choices to make their
.decisions on the basis of all characteristics. This is obviously hardly ever the case. Often, one
characteristic overrides all others; proximity to the place of work often takes this role. People do
not usually have a choice about where they find work; thus, they may move into an environment
that they would not move into otherwise. '
Another problem is that finding a sample with sufficient variation (i.e., enough houses
that exhibit different characteristics) is not easy. The specific environment of houses varies
together with other factors, and it is very hard to isolate the influence of one variable when they
vary together. And, as stated above, in the absjence of a wide array of choices, people are likely
to base their decisions on characteristics other than the environment.
I
One problem with this method is that it> presupposes information about job characteristics
on the part of workers and researchers. Workers often do not have sufficient information about
the risks to their health and life posed by their jobs. Also, unless a job exposes one to specific
pollutants, establishing a worker's dislike for a specific pollutant is not possible. This method
also involves the problem of measurement. Dz[ta on specific pollution at work are not readily
available; data usually only exist on the consequences of hazards, such as accidents, morbidity,
and mortality. Hedonic wage studies would bej more useful in damage cost studies if they could
indicate the value that people ascribe to their lives.
i
6.4.2 Contingent Valuation >
Contingent valuation assumes hypothetical (contingent) markets. In essence, it consists
of experiments in which people arc asked to express their valuation for a specific environmental
commodity. These experiments can be designed as bidding games, questionnaires, and so forth
(see Freeman, 1982 and Mitchell and Carson, ]j991).
Understanding the change in environmental conditions consumers arc asked to evaluate is
important Two concepts arc suggested in the literature: willingness to pay (WTP) and
willingness to accept (WTA). Loosely speaking, the former is the amount of money a consumer
would be willing to spend to secure an environmental benefit, and the latter is the compensation
that the consumer would demand to accept an environmental cost. However, both concepts can
6-16
-------
be applied to similar changes in environmental conditions. For example, consider a policy to
clean up 90 percent of sulfur oxides emissions. WTP then is the maximum amount of imoney an
individual would give away to have 90 percent of the sulfur oxides emissions abated, while
maintaining his or her utility level, and WTA is the amount of money he or she would haw to be
given to accept the pollution while maintaining the utility level corresponding to the absence of
90 percent of the present pollution.
Economic theory suggests that these two values do not really differ. However, empirical
studies assessing the magnitude of WTA versus WTP have consistently produced far greater
amounts for WTA than for WTP. There has been ongoing discussion about this apparent
discrepancy. It was long known that the greater the difference between WTA and WTP, the
greater the income elasticity of demand. WTP is obviously limited by an income constraint,
whereas WTA is not.
6.4.3 Cost of Control Valuation
The cost of control valuation method enjoys increasing popularity as utility companies
attempt to internalize the environmental cost of energy production. Some states (e.g., California,
Massachusetts, Nevada, New York, Wisconsin, Oregon) have proposed or adopted this approach
to incorporate the environmental costs of electricity production into their energy planning
processes.
This approach infers that the cost society attributes to pollution may be derived from
government regulations for specific pollutants. Complying with standards set for pollutant
emission is costly; thus, there must be a perceived benefit to pollution abatement. Two concepts
are central to this approach: the marginal cost of pollution abatement and the marginal benefit
of pollution abatement.
Marginal Cost of Pollution Abatement is an increasing function of the amount of
pollutant being controlled. Increasing marginal cost also means that the unit cost of
abatement (the cost of abatement per unit of pollutant) rises as more and more pollution
is abated. To remove the first unit of pollutant, one would choose the cheapest
technology available. The most expensive technology would only be employed if the
potential of cheaper technologies was exhausted.
Marginal Benefit of Pollutant Abatement is a decreasing function of the amount of
pollutant being removed. For example, the benefit from preventing one more ton of
SO2 to enter the atmosphere is smaller the more SO2 has already been controlled. The
6-17
-------
t negative side of this relationship is tl|at the marginal damage function of pollution is
-' generally increasing; that is, the damage that one unit of pollutant causes is greater, the
higher .the overall pollution levels. ;
. '.The optimal emission standard for a particular pollutant is that level of pollutant at which
"* «4 the marginal cost of abatement equals the marginal benefit of abatement. Setting such a standard
*1 would require an efficient allocation of resources for pollution abatement activity. To do more
l would cost society more than the benefits that would result from implementing the standard.
Several problems are associated with the pollution abatement approaches described
above. First, no emission standard exists for each individual pollutant Controlsnot
standardsare administered for some pollutanis; others are not regulated at all. Controls present
the problem of "joint cost of pollution control": where several pollutants causing different
environmental impacts can be captured with onje-and-the-same device. The problem lies in how
the cost of that device should be allocated to individual pollutants. In addition, a value for the
pollutant the device is intended to capture can only be inferred because the regulation implies a
certain value for this pollutant.
I
Another problem is that regulations for Sail pollutants may not exist. A case in point is
the emission of greenhouse gases. One could value the costs of these emissions through the
costs of the measures that would offset the emissions (e.g., afforestation). It also seems
legitimate to assume that society holds consistent preferences, and that for some pollutants,
regulations addressing different but similar ones can be used. For example, the banning of lead
acid batteries from incinerators reveals the regulator's (representing society's) preference that
heavy metals should not be emitted. It seems legitimate to assume a regulation banning other
heavy-metal products of similar toxicity. ''
6-18
-------
CHAPTER?
INTEGRATED METHODS FOR IMPACT ASSESSMENT
Integrated methods have been developed, or are being developed, to include some
combination of classification, characterization, and/or valuation activities of impact assessment.
This chapter profiles some of these integrated methods. Some of these methods integrate data
developed from an inventory analysis with expert decision or economic valuation methods to
yield information that is relevant to not only environmental decisionmaking but also to overall
business decisionmaking, which includes a number of factors (e.g., profitability, product quality)
in addition to environmental performance.
7.1 IMPACT ANALYSIS MATRBI (IAM)
The IAM is an exploratory, qualitative, expert-based approach to impact analysis that
builds on the results of an inventory analysis. The IAM approach is described below by
explaining its development and initial use.
The IAM approach was developed as part of a broader assessment of source-reduction
potential for halogenated solvents, which included an assessment of alternatives to such solvents
in specific applications. The IAM allowed for the direct evaluation of the relative environmental
burdens of a particular application of a halogenated solvent and its alternative(s) and made the
tradeoffs between them explicit. Two specific applications involving substitution systems for
TCA (1.1.1 -trichloroethane) were evaluated:
substitution of a caustic aqueous cleaner for TCA vapor cleaning of metal parts and
substitution of supercritical CO2 paint spraying for TCA-based paint spraying.
Comparisons of these two TCA substitute systems were conducted on two different
levels: user (or shop) level and global level. User-level impacts referred to ecosystem impacts
that emanated from a boundary drawn around a particular facility using the TCA substitute
system. For example, only waste disposal activities associated with using the substitute were
considered. Global-level impacts took into account all of the traditional life-cycle stages,
including raw materials acquisition, manufacturing, use/reuse/maintenance, and recycle/waste
management Analysis at these two levels allowed for identifying additional tradeoffs between
the two systems. That is, it allowed options that appeared favorable from the user's point of
view but unfavorable from a global point of view, or vice versa, to be identified and evaluated.
7-1
-------
The IAM process entails convening a group of experts to carry out the following steps:
1. Identify appropriate impact categories. The IAM study of the two TCA
substitutes consisted of jfive columns of inventory data (inputs and outputs) and
seven rows of ecosystem impact categories. These impact categories were
selected by expert judgment and included
global warming
ozone depleting potential
nonrenewable resource utilization
air quality :
water quality
land disposal i
transportation effectsj
(It should be noted that|in applying the IAM approach in other settings, impact
categories that match the particular characteristics of each comparison should be
used. The identification and exclusion of various impact categories should be
transparent and sufficient justification should be provided.)
2. Determine which cells in the LAM represent either double counting or
meaningless comparisons. For instance, in the case of the two TCA substitutes it
was determined that aqueous wastes had no significant impact on global
warming; thus the correjsponding cell in the LAM was eliminated. As a result of
this step, 17 of the 35 IA.M cells were eliminated.
t
3. Assign unweighted "scores" to each viable cell in the LAM. Scores for the TCA
study were assigned in relation to a particular option chosen as the base. In this
study, a "+" was used t<|> signify a larger ecosystem impact than the base option
(TCA), and a "-" was used to signify a lesser impact A "0" can be used to
signify little or no perceived difference in impact. Determination of scores was
based on a combination of life-cycle inventory data and expert knowledge of
associated ecosystem impacts.
4. (Optional). Apply weights to the initial unweighted scores to determine if the
results will change significantly. The weighting scheme used in the TCA study
assigned a "++" to relatively strong ecosystem impacts and a ""to relatively
large reductions in impact.
(It should be noted thadthe weighting may or may not be restricted to a^single
impact category, depending on the views of the expert panel. However, the basis d
for assigning weights and the scope of comparison within and across impact ;'.'.,' ,
categories should be made .transparent.) , j >
7-2
-------
5. Sum the individual cell scores (pluses and minuses) to derive overall scores for
each row and column and, if appropriate, for the entire matrix. Unweighted
scores in the TCA substitute IAM ranged from +18 to -18, and weighted scores
ranged from +36 to -36.
Table 7-1 shows data gathered for the two TCA substitutes. For each substitute, the data
were broken down into user-level and global-level items. The corresponding lAMs for the TCA
substitutes are shown in Figures 7-1 and 7-2. As an example of the type of information that may
be derived from the IAM approach, compare and contrast the scores in the energy-inputs column
evaluated at the user versus the global levels. From the perspective of the user, impacts derived
TABLE 7-1. TCA SUBSTITUTE STUDY INVENTORY DATA
Vapor Degreasing
Aqueous Cleaning
Parameters
User
Global
User
Global
Amount Used (tons)
TCA
Aqueous cleaner
Material Inputs
Trona deposits, salt, sand
Crude, natural gases
Energy Inputs
Power or Fuel (per million BTU)
Atmospheric Emissions (tons)
26.6
0
0
0
520
0
0
1.2
2.9
1,530
0
2.7
0
0
1,730
0
0
4.5
0
1,800
Cl-HC, HC/particulates, C12
TCA
Particulates
Water vapor
Aqueous Wastes (tons)
Solid Wastes (including spent catalyst,
solids/ sludge, used oil, and shale in tons)
TCA and oil (from OTVD)
0
21.6
0
0
0
0
6.5
2
21.6
0
0
682
4.9
6.5
0
0
0
288
1,822
0
0
<0.1
0
<0.1
288
1,822
0.3
0
Source: Source Reduction Research Partnership, 1991.
7-3
-------
User-Level Impact Analysis Matrix
Ecosystem Impact
TOTAL
1 Impactinq Parameters
Material
s I !\0\ ^ ^
-1
' »' -
'^\-f- , ;
- '
-1
-2
Energy
Inputs
+1
,^Nf"-' ^
+1
i +1
I +1
+1
+1
\ +6
Air
Emissions
-1
-1
'-s- <-J"'*< '
-1
~V^ "
, " >,
; % ,?*; '~
-3
Aqueous
Wastes
ffjt fffff fSf s ^ ]
t ' i
+1
+1
+1
'< , -;; /^.
+3
Solid
Wastes
,
',* * 'y ""'
I , ?'' '
''%&*^ '-
-1
-1
"1
-3
TOTAL
0
-1
0
0
+2
+1
-1
+1
Notes: 1. Shaded cells signify no basis for impact.
2. A rating of "-1" represents decreased impact, "0" represents the same impact, and "+1" represents an
increased impact. i
Figure 7-1. User-Level Impact Analysis Matrix for Ecosystem Impacts
Source: Source Reduction Research Partnership, 1991.
Global-Le>
Ecosystem Impact
Categories
Global Warming
Nonrsnewable Resource Use
Water Quality
Land Disposal
Transportation Effects
TOTAL
Material
Inputs
t f V ss
i '^ "
-1
:% . "
}C₯- X ' %^ '
?% f!/?*-^
-1
-2
te\ Impact Analysis Matrix
Impacting Parameters
Energy
Inputs
0
.- , f
0
0
0
o
0
0
Air
Emissions
-1
-1
i -' >\₯-
-1
* ','
, ' X'% V ' '",
^ s *,>'-"
-3
Aqueous
Wastes
/ <« f
' '"'' '' , '< '"','>
:, ' , ',"
-",',{>«':
+1
+1
+1
' ' '' ,,','
+3
Solid
Wastes
,"' -%-
<'>' , "'
-1
,ff * f f
-1
-1
-3
TOTAL
-1
-1
-1
-1
+1
0
-2
-5
Notes: 1. Shaded cells signify no basis for impact.
2. A rating of "-1" represents decreased impact, "0" represents the same impact, and "+1" represents an
increased impact. j ' :
i
Figure 7-2. Global-Level Impact Analysis-Matrix for Ecosystem Impacts; ;?
Source: Source Reduction Research Partnership, ?l 991. >: ;
7-4
-------
from energy input requirements were a dominating category and were much higher for the
aqueous substitute relative to the TCA system because of the high pumping and heating
requirements of the aqueous substitute. In contrast, global-level impacts derived from energy
requirements were found to be essentially the same for the two systems.
Strengths
The IAM is relatively simple and convenient to use, is flexible enough to account for a
wide variety of impacting parameters (i.e., life-cycle components) and environmental impact
categories, and can be used at different levels of analysis (e.g., global versus shop level). The
IAM also does not require any additional data beyond that which is generated in the inventory
analysis and uses a relatively objective technique (i.e., less is better) to evaluate the associated
environmental consequences.
Weaknesses
One weakness of the IAM is that it does not measure impacts. Appropriate impact
categories are chosen by expert judgment, and inventory items from two alternatives are merely
compared according to a "less is better" ranking for their contribution to their associated impact
categories. However, this process does not provide insight into how impact categories relate to
one another. For example, in Figure 7-2 both the totals for global warming and nonrenewable
resource are -1. From this, the reduction in global warming and nonrenewable resource use
appear to be "equal" from the use of aqueous cleaners. However, the method does not indicate,
for example, how better or worse a 1-ton reduction in global warming gases is compared to a
1-ton reduction in nonrenewable resource use.
One possible variation to the IAM matrix that may help to better express the relationship
of impact categories to one another is the Leopold interaction matrix. The cells in the Leopold
interaction matrix contain the ratio of the magnitude of impact (M) to the importance of the
impact (I). M expresses the extensiveness or scale of the impact, and I expresses the importance
of the impact (to stakeholders). The basic framework for the Leopold interaction matrix is
shown in Table 7-2.
7-5
-------
TABLE 7-2. LEOPOLD INTERACTION MATRIX
Life-Cycle Staee
Raw Materials Manufac- Use/Reuse/ Recycle/Wasf
Impact Category Acquisition turing Maintenance Manaeemen
Global Wanning
Ozone Depleting Potential
Nonrenewable Resource
Utilization
Air Quality
Water Quality
Land Disposal
^Transportation Effects
Scale Ranges: M - 1 to 10 1 = low
I - 1 to 10 10 = hi|
M/I M/I
M/I M/I
M/I M/I
M/I M/I
M/I M/I
M/I M/I
M/I M/I
M/I
M/I
M/I
M/I
M/I
M/I
M/I
M/I
M/I
M/I
M/I
M/I
M/I
M/I
:st magnitude of impact, or lowest level of importance.
best magnitude of impact, or highest level of importance.
Another weakness of the IAM is that its use in a noncomparative study, which includes
only a single set of data from one alternative and no set of data against which to evaluate the
alternative, is not clear. For example, in the case study example outlined above, an IAM for
aqueous cleaning alone would be meaningless without the baseline of vapor cleaning against
which to compare aqueous cleaning; With just one set of data, the IAM could possibly be
modified to provide a general indication of the impact categories and/or impacting parameters
that are most significantly affected. ;In this case the pluses and minuses in the matrix cells would
be used to represent the relative significance of particular impact categories or impacting
parameters. i
Relevance to Impact Assessment
The IAM approach may provide a relatively simple, quick, and useful means of.' v. .
qualitatively comparing the environmental implications of two or more alternative systems .
without having to characterizing impacts. The more qualitative nature of the IAM would make.
it more appropriate for internal applications or as a screening tool to identify impact categories
or life-cycle components that require a more detailed level of assessment. ;. ?t>. ,
7-6
-------
7.2 THE EPS ENVIRO-ACCOUNTING METHOD
Prepared for the Swedish Environmental Research Institute, the EPS Enviro-Accounting
method describes impacts on the environment in terms of one or several safeguard subjects using
the EPS method described in Chapter 4 arid then places a value on changes in the safeguard
subjects according to the WTP to restore them to their normal status.
The five safeguard subjects included in the EPS Enviro-Accounting method are the
following:
human health,
biodiversity,
production,
resources, and
aesthetic values (Swedish Waste Research Council, 1992).
Impacts are characterized and valued on a relative scale using ELUs according to the
WTP for avoiding negative effects on the safeguard subjects. Environmental indices are then
calculated for the materials and processes being studied. Background information is derived
from LCA-based inventories of the materials and processes under review. The values are not
absolute figures but rather points of reference for further analysis and refinement.
Environmental impact valuation is described as a subjective matter that can be given some
degree of objectivity by studying decisions made in society or by surveying people's opinions.
Contingent valuation is cited as a method for generating a relative rating of various
environmental effects. Contingent valuation is used in the EPS Enviro-Accounting approach to
determine individuals' WTP to avoid certain environmental effects. To date, EPS indices for a
wide range of environmental impacts have been developed using such WTP figures.
The output from the EPS Enviro-Accounting system is a value, based on a common
metric, for different environmental impacts. The value may be broken down into its individual
components for further analysis, and the user can determine the level of detail desired.
Strengths
The EPS Enviro-Accounting method is strong in that it accounts for a wide variety of
impacts within five main impact categories: human health, biodiversity, production, resources,
and aesthetic values. Impacts within and between these main categories are characterized and
valued on a relative scale allowing for a direct relative comparison of impacts. In addition, the
7-7
-------
information required in the EPS Enviro-Accounting methods is derived from LCA-based
inventories and readily available environmental valuation studies.
Weaknesses
One weakness of the EPS Enviro-Accounting method is that environmental impact
valuation is a highly subjective matter. Although monetary units provide an easily understood
value metric, using monetary values to assess environmental impacts has been criticized for
several reasons: j
!
a large gap exists betweeij rich and poor in terms of disposable resources for
environmental care,
the needs of today often 6utweigh the needs for tomorrow,
insufficient knowledge exists to value environmental impacts because the full
consequences of impacts are not fully understood, and
monetary valuation focuses on human needs.
Relevance to Impact Assessment
Because the EPS Enviro-Accounting method was developed in the context of LCA, it is
readily applicable for impact assessment. The five safeguard subjects may be used to categorize
inventory items into impacts categories, and environmental valuation studies using WTP may be
used to estimate costs and develop coefficients expressing the relative environmental impact (in
economic terms) of alternative items. However, environmental valuation studies are sometimes
controversial in their own field of economic research. The use of such valuation studies and/or
techniques for impact assessment irjay be similarly problematic.
7.3 INTEGRATED MANUFACTURING AND DESIGN INITIATIVE (IMDI)
ENVIRONMENTALLY CONSCIOUS MANUFACTURING (ECM) LIFE CYCLE
ANALYSIS
i
As part of a Sandia National Laboratory program, IMDI selected Department of Energy
(DOE) stakeholders (e.g., designers, manufacturers, Environmental Safety and Health personnel,
environmental technology staff, industry, EPA, and academicians) and surveyed;;them to
establish a basis for defining environmental impact metrics. A panel discussion was also
conducted. The survey asked for tv^o primary responses:
1) "Identify environmental impacts of activities related to manufacturing, use and i
disposal;" and i
7-8
-------
2) "... list the criteria that might be used to assess one product or process against
one another with respect to minimizing those impacts" (Watkins, 1993).
The panel used the AHP process, supported by Expert Choice software, and group
decisionmaking. The panel developed an IMDI Environmental Impacts Model that builds on
earlier SET AC work.
The panel discussed the possibility of using Colby's (1991) five environmental
management paradigms as a basis for assigning weights to environmental impacts. The
environmental impacts associated with the entire life cycle were included in the group's
proposed model (i.e., the group developed an extensive list of environmental impacts). The
"costs" associated with these impacts were not evaluated.
A weighting method of cost estimation based on Colby's five environmental
management paradigms was discussed. It was suggested that rather than deriving or assigning
absolute weights, the weighting system could be used for sensitivity analysis over a range of
values for the different impacts (Watkins, 1993). Colby's paper (1991) discussed the
distinctions, connections, and implications for the future of environmental management by
describing the changing strategies and the related philosophies of the following broad
environmental management paradigms:
frontier economics,
environmental protection,
resource management,
eco-development, and
deep ecology.
Associated with each paradigm are differing philosophies of human-nature relationships.
The paradigms are overlapping and encompass several schools of thought. Colby's paper does
not explicitly detail methods for evaluating environmental costs; however, it suggests that
environmental costs would be treated differently according to the prevailing environmental
management paradigm. The following is a description of the possible environmental costing
methodologies under each of the five paradigms.
Frontier Economics
Property owners and the public at large pay environmental costs (not necessarily the
polluter).
7-9
-------
Production is limited by manmade factors. Natural factors are not accounted for.
Analytic modeling and planning methodologies include net present value,
maximization, and cost/benefit analysis of tangible goods and services.
t
Economic analysis is based on the neoclassical model of the closed economic system.
En vironmental Protection \
Taxpayers (public at large) pay environmental costs.
Analytic modeling and planning methodologies include environmental impact
assessment after design, optimum pollution levels, equation of WTP and compensation
principles. \
Economic analysis is based on the neoclassical model of the closed economic system.
Ecological benefits are difficult to quantify, so environmental management in this
paradigm is treated strictly as an added cost.
Resource Management '
"Polluter" (producers and consumers) pays environmental costs.
Analytic modeling and planning methodologies include natural capital; true (Hicksian)
income; maximization of tlnited Nations System of National Accounts; ecosystem and
social health monitoring; and linkages between population, poverty, and environment.
Economic analysis based on an extension of neoclassical economics that incorporates
all types of capital and resourcesbiophysical, human, infrastructural, and
monetaryinto calculations of national accounts, productivity, and policies for
development and investment planning.
Pollution can be considere<^ a "negative resource" (causing natural capital degradation),
rather than an externality. \
The concern for nature stems from the fact that hurting nature is beginning to hurt
economic man. Environmental expenditures are considered necessary to avoid "more"
costs. i
I
Eco-Development
A "pollution prevention pays" concept rewards those that do not pollute. The economy
is structured to reduce pollution as a throughput.
Analytic modeling and planning methodologies include ecological economics; open ;
system dynamics; socio-teqhnical and ecosystem process design; integration of social,
economic, and ecological criteria for technology; and trade and capital flow based on
community goals and management.
The relationship between spciety and nature can be considered a "positive sum game."!:
Human activities are organized to be synergistic with ecosystem processesiand'services;
7-10
-------
Emphasis is placed on efficient, clean, renewable energy sources; environmental
information; community consciousness; and experiential quality of economic activity.
An example of the eco-development paradigm is the International Joint Commission
between U.S. and Canada, which explicitly uses a stakeholder, positive-sum approach.
DeepEcology
Environmental costs avoided by foregoing development.
Analytic modeling and planning methodologies include grassroots regional planning,
multiple cultural systems, and conservation of cultural and biological diversity.
Strengths
One strength of the approach used by IMDI for assessing environmental impacts is that it
provides a framework and methodology for breaking complex problems down into constituent
parts. The method provides a framework for organizing complex issues into a more easily
manageable format that defines goals, objectives, subobjectives, and criteria relating to
environmental quality. The criteria may then be assessed individually against expert knowledge
and stakeholder values to gain a better understanding of the problem at hand. Through the use
of the IMDI methodology, coefficients can be established for various substances that indicate the
relative environmental impact of those substances. Such coefficients may be directly compared,
allowing for a relative comparison of individual substances or the evaluation of the
environmental profile of an entire system.
Weaknesses
The primary weakness of the IMDI approach is that the process for developing weights
for individual substances is highly subjective. It is not clear how weights developed by different
groups could be compared against one another in a meaningful way. The AHP pair-wise
comparison process is largely a subjective process requiring expert knowledge to rate the
intensities of the environmental impacts being compared. In the case of most future problems,
including potential environmental impacts, there is no such expert knowledge. Thus the
weighting factors developed by different groups would not be very meaningful. In addition, the
possibility exists that the experts can have a bias and/or misjudge the importance of particular
attribute intensities. Because the IMDI approach relies on the values and judgment of a select
group of individuals, the results from this approach probably could not be replicated.
7-11
-------
The IMDI approach also concentrates
impacts and thus is somewhat limited
manufacturing because additional factors
contribute to decisions
solely on environmental quality and environmental
in its intended application to environmentally conscious
(e.g., cost, functional requirements, performance) also
affecting product design (Watkins, 1993).
Relevance to Impact Assessment
The concept of assigning weights in the IMDI, particularly for a range of values, is
particularly noteworthy to impact assessment because it attempts to provide a common metric
for valuing environmental impacts. jBecause of the subjective nature of weighting process used
in the IMDI approach, its use wouldj be more appropriate for internal impact assessment
[
applications. The IMDI approach requires further testing before results can be supported in
external applications.
7.4 INTEGRATED SUBSTANCE CHAIN MANAGEMENT
Developed by VNCI (an association of the Dutch Chemical Industry), integrated
substance chain management (also called the VNCI process) was designed to evaluate a
substance throughout its entire life Cycle (Canadian Standards Association, 1992). Integrated
substance chain management was aljso designed to encourage the use of environmentally
preferable substitutes and recycling alternatives and the identification and closure of leaks.
To include all environmental issues, each link in the substance chain is checked against a
comprehensive list of environmental themes, including global warming, ozone depletion,
acidification, eutrophication, photochemical ozone formation, dispersion of toxic substances,
disposal of wastes, and disruption/depletion of natural resources.
Based on rough estimates of product system inputs and outputs and their associated
environmental issues, options for process improvement can be proposed. The selection of
options for detailed analysis is based on
environmental impact,
'f , *
cost effectiveness, and
relevance to decisionmakers (Canadian Standards Association, 1992)J ,' <
The output of the detailed analysis is a two-axis (environmental impact/economic impact)
options map. The options map is developed by determining the environmental;and economic,,;
profiles of the substance in question! and positioning the various options (see'.Eigure. 7-3).
7-12
-------
Environmental Impact
Environmental Costs
Economic Savings
Environmental Costs
Economic Costs
i
0
Environmental Benefit
Economic Savings
^-
^
Environmental Benefit
Economic Costs
Quadrant where system
options would ideally be
mapped
Environmental Yield
Quadrant where most
system options will
be mapped
Figure 7-3. Options Map for Integrated Substance Chain Management
Source: Canadian Standards Association, 1992
The environmental profile provides a comprehensive overview of the relevant
environmental impacts associated with each process option. Impacts are quantified in terms of a
single unit of measurement for each impact category (e.g., tonnes of CO2 equivalent as an agent
of global warming) and shown in terms of a fraction of the total national emission of that
environmental impact theme. Exact changes of inputs and outputs associated with each process
option are calculated and a checklist is used to determine the extent of changes in other
input/output factors. The data are then converted to scales for measuring the impact on each
environmental theme, and sensitivity analysis examines how changes in the underlying database
of inputs, outputs, and conversion factors affect the results of the analysis.
The economic profile evaluates the economic impact of the proposed process options.
When the environmental and economic profiles are completed, all quantitative figures and
qualitative comments in each profile are combined to arrive at a final conclusion concerning the
total environmental and economic impacts.
7-13
-------
I
Relative weights are then assigned to each environmental theme to enable the
aggregation of environmental impacts associated with each process option. The environmental
i
and economic impacts are then combined to represent a single point on the options map. The
origin of theoptions map represents the "do nothing" option. Options that represent both
environmental and economic improvements will be plotted in the upper right-hand quadrant.
Options that represent both environmental and economic setbacks will be plotted in the lower
left-hand quadrant.
r
Strengths
The main strength of the integrated substance chain management approach is that it
provides a framework and methodol9gy for integrating environmental concerns, economic
concerns, and stakeholder values. The resulting options map portrays the environmental and
economic differences between proceks options, allowing for relatively easy and objective
decisionmaking. In addition, the development of relative weights for each environmental theme
enables the environmental impacts associated with each process option to be aggregated to yield
an overall environmental profile for the system.
Weaknesses
The main weakness of the integrated substance chain management approach is that it
employs a relatively simplistic weighting scheme and may not be applicable to in-depth
assessments of impacts. Additionally, it is not clear how life-cycle economic costs would be
developed for use in the options mapi.
i .
Relevance to Impact Assessment
The integrated substance chain management approach would be most applicable to
internal impact assessments where a number of different factors (e.g., environmental protection,
economic well being, public image) affect the environmental decisionmaking process and a less
detailed level of assessment provides adequate information to make the decisions at hand.
i
7.5 ECO-RATIONAL PATH IVfrETHOD (EPM)
EPM represents a procedure that builds on the ESR method described in Chapter 4 to
integrate environmental and economic informationtwo of the primary dimensions of
environmental decisionmaking. The process for integrating these two dimension .comprises ;
three main stepsrecording, judgment, and decision-;as shown in Figure,7-4, . i
7-14
-------
Ecological Dimension
Economic Dimension
Module 1 ;
Pollution-Added Account
Module 11 ;
Costs/Revenues
Account
E
8
Module HI
Pollution Units (PU)
Module IV
Contribution Margin (CM)
per unit product
i
0)
D>
i
i
Module V
O
0)
Q
PU per CM
Economic-Ecological
Efficiency
Figure 7-4. Conceptual Framework for the EPM
Source: Schaltegger and Sturm, 1993
7-15
-------
Looking at the ecological dimension in Figure 7-4, the first step is to collect and record
information on environmental releases. Releases in the context of EPM include inputs, desired
output, and undesirable outputs. Although labeled "pollution-added account" (see Module I in
Figure 7-4) the information for this step may be generated through using traditional inventory
analysis procedures. Sometimes the|data developed in the inventory analysis are sufficient for
evaluating environmental improvement options, but often inventory data alone are insufficient.
For example, when one system releases more CO2 and another system releases more NOX, then
no obvious and objective judgments jare possible. To weigh one pollutant against another, a
preference ranking is needed. This step, termed judgment, is a procedure for developing weights
or "pollution units" (see Module III in Figure 7-4) for releases according to their environmental
relevance (based on ambient concentration standards for various media). Developing these
relative weights is accomplished using the ESR method as described in Chapter 4.
I
With regards to the economic dimension as shown in Figure 7-4, the first step (see
Module II) is to collect and record information on economic costs/revenues including
environmental compliance costs and|earnings. This information is typically generated in
traditional accounting practices but iWy need to be broken out of an aggregate account (e.g.,
overhead) and appropriately allocated to a specific product or process. After all the necessary
cost/revenue information is recorded, the contribution margin (see Module IV in Figure 7-4) is
calculated as a measure of economic efficiency.
i
i
The integration of the economic and ecological information is shown in Module V in
Figure 7-4. In this module, the data produced from Modules III and IV are integrated by
calculating the quotient pollution units per, for example, created dollar contribution margin of a
product or process. This calculation I provides a measure of the economic-ecological efficiency
of specific products and processes. In general, the most preferable products or processes are
those with low pollution units and high contribution margin (i.e., small PU/CM ratio).
Strengths
i
Some of the strengths of the EPM include the following:
[
framework and methodology are provided for integrating environmental-and economic
considerations;
weighting factors using ambient regulatory standards represent social, political,
regulatory, and scientific opinions and values;
T
7-16
-------
. weighting schemes used in EPM represent the relative environmental and economic
impacts of different chemical releases to different environmental media;
. EPM is flexible and can be used at a variety of different spatial levels (e.g., state,
regional, and local).
Weaknesses
One weakness of the EPM is that the ESR weighting scheme's use of relations between
ambient standards is not a natural scientific or ecotoxicological based scheme, but instead
represents a socio-cultural judgment from an ecological perspective (which relies on
ecotoxicological data). However, no objective and undoubtedly valid opinion on the
harmfulness of substances exists. ESR develops weights according to generally accepted norms
and values, which are theoretically expressed in ambient concentration standards. Such ambient
standards may or may not reflect actual environmental impacts.
Another weakness of the EPM is that the method for characterizing the economic impacts
of pollutants is somewhat simplistic. It is not clear whether the results of the economic impact
assessment component would be useful to decisionmakers.
Relevance to Impact Assessment
The information produced from applying the EPM can be used in impact assessment to
evaluate and compare the relative environmental and economic impacts of inventory items
where regulatory standards exist. Although using the ESR approach provides a consistent
estimate of the environmental impacts, it does not necessarily preclude the need for additional
analyses. EPM requires additional testing in the context of LCA to better gauge its applicability
to impact assessment.
7-17
-------
-------
CHAPTERS
KEY POINTS AND FUTURE RESEARCH NEEDS
The purpose of life-cycle impact assessment is to translate the results of an inventory
analysis into a description of environmental impacts, providing users with additional information
to discern between alternatives (e.g., inventory items, systems). Impact assessment also makes
explicit the methods used to compare and weigh alternatives. This document covers a wide
variety of issues related to impact assessment and outlines existing methods that have been used
or presented in the context of impact assessment. Again, it should be kept in mind that this
document is not a guidance document, but rather a compendium on the state of practice of
impact assessment.
This final chapter summarizes some of the key points discussed throughout this
document and provides a listing of potential research needs for the future research and
development of impact assessment techniques and methods. While impact assessment is still in
its infancy, this document illustrates promise for the current applications and future development
of impact assessment techniques and methods.
8.1 SUMMARY OF KEY POINTS
This document covers a broad range of material which cuts across a variety of research
areas. Some of the key points that can be drawn are summarized below.
Impact assessment has been conceptually defined to include three phases: classification
of inventory items into impact categories, characterization of potential impacts, and
valuation of impacts. However, formal procedures and methods for conducting impact
assessment have not yet been established.
Impact assessment may be useful for a variety of both internal and external applications
(see Chapter 1). Although internal applications may not be required to follow stringent
LCA guidelines, they should nonetheless follow the best practice.
Practitioners may not need to complete a full impact assessment to obtain useful
information. In some cases, merely classifying inventory items into impact categories
may provide adequate information for users to identify improvement options. In other
cases, a more detailed impact assessment information may be needed.
A wide variety of methods are available for use in impact assessment (see Chapters 4
through 7), ranging from simplistic checklists to complete risk and economic impact
assessments. A rule-of-thumb for choosing the appropriate method(s) is to choose the
method(s) that provides adequately detailed information to make the decision at hand
8-1
-------
(usually to discern the relative impact of different substances). A more complex
method(s) is used only when the resulting information is needed to advance the
decision to be made.
There is a general lack of methods for assessing the impacts of nonchemical loadings
(e.g., habitat alteration, heat, noise) to the environment. Many of the existing methods
for characterizing impacts |(see Chapter 4) are based upon chemical exposure and
toxicology data and cannojt readily be used to assess the impact of nonchemical
loadings. ,
i
There are a number of plapes in the impact assessment where value judgments may
play a significant role. It ijs critical that practitioners document points in the impact
assessment process where [value judgments were employed, the set of values used! to
make the judgment, and hpw those judgments may affect the outcome of the impact
assessment; !
A significant level of uncertainty is associated with impact assessment (e.g., linking
inventory items to impacts). Uncertainty, however, is a fact of life for virtually all
areas of research. Although there are currently no formal procedures for evaluating
uncertainty in impact assessment, practitioners should nonetheless document and
evaluated sources of uncertainty and appropriately qualify impact assessment results.
As with other LCA components, it is critical that practitioners clearly communicate the
content and conduct aspects of the impact assessment in the final LGA report. This
includes, but is not limited|to, the goals and scope, data sources used and their quality,
models used and their assumptions and limitations; and data or methodology
manipulations; value judgments employed, and the analyst's interpretation of these
aspects on the overall LCA| results.
I
8.2 POTENTIAL FUTURE RESEARCH NEEDS
i
Potential research needs (adapted from Vigon and Evers, 1992) identified by the LCA
community regarding the future development and application of impact assessment tools and
procedures are listed in Table 8-1.
8-2
-------
TABLE 8-1. POTENTIAL FUTURE NEEDS FOR IMPACT ASSESSMENT
RESEARCH
Research Needs
Effort
Relate EIS scoping process to impact assessment
Define impact descriptors for LCA applications
Determine basis for defining stock resource pool
Develop impact category equivalency factors
Identify ecohazard profile parameters, thresholds
Develop method for evaluating depletion of water resources
Develop method for linking resource development with depletion
Achieve international consensus on impact assessment
Assess feasibility of nonchemical impacts matrix
Prepare impact analysis technical support document
Develop and validate of streamlined impact assessment methods
Develop a reference data base of generic impact assessment information
Develop library of impact networks
Prepare broad range of impact assessment case studies
< Develop methods for factoring uncertainty into impact assessments
Determine feasibility of resource management/economic models for LCA
Develop methods for estimating biodiversity change and habitat alteration
Develop models to assess susceptibility due to health stress
Develop better human exposure models within an LCA
Evaluate ecological risk assessment models/methods
Develop/validate ecological hazard matrix approach
Fill data gaps in the following areas:
Health exposures
Short-term and long-term bioassays
Effects of unintended product use
Exposure from nonmanufacturing
Nonpoint sources of pollution
Low
Low
Low
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
High
High
High
High
High
High
High
High
High
High
High
Source: Modified from Vigon and Evers, 1992
8-3
-------
-------
Appendix A
National Environmental
Policy Act (NEPA)
Environmental Assessment
Procedures
-------
-------
Under Section 102 of the National Environmental Policy Act (NEPA), federal agencies
are required to make a full and adequate analysis of all environmental effects of implementing
its programs or actions (Jain et al., 1993). In the context of NEPA, an environmental impact
assessment (EIA) is used for determining if a more detailed environmental impact statement
(EIS) is required. EIAs utilize a list of environmental "attributes" for which baseline values are
compared against actual or expected values to determine the level of potential impact. After the
environmental "attributes" are determined, the EIA scoping process is used to evaluate and
streamline a comprehensive list of "attributes" or impacts.
The comprehensive list of environmental attributes considered in an EIA and the scoping
process used to streamline that comprehensive list to a reference project may be useful in the
context of impact assessment where a wide variety of impacts require consideration. These
components of EIA are described in further detail in the following sections.
A.1 ENVIRONMENTAL ATTRIBUTES ADDRESSED IN EIA
Environmental attributes are variables that represent characteristics of the environment
(see Table A-1). The environment is difficult to characterize because it contains numerous
attributes exhibiting complex interrelationships. However, anticipated changes in the attributes
of the environment and their interrelationships are defined as potential impacts. All lists of
environmental attributes are a shorthand method for focusing on important characteristics of the
environment. Because of the complex nature of the environment, any such listing is limited and,
consequently, may not capture every potential impact. The more complete the listing is, the
more likely it will reflect all important effects on the environment, but this list may be expensive
and cumbersome to apply.
Table A-2 summarizes possible environmental attributes in eight categories that comprise
the biophysical and socioeconomic environment at a generalized level. While this list of
attributes represents a reasonable breakdown of environmental parameters, it is likely to require
modification or supplementation depending on the type of action to be assessed. For a more
complete description of these attributes, the reader is referred to Jain et al. (1993).
A.2 EIA SCOPING PROCESS
When EIAs were first introduced, decisionmaking based on EISs was being
compromised by their inclusion of what many considered to be insignificant factors. These
insignificant factors were considered to be background noise, while significant factors were in
A-l
-------
TABLE A-l. ENVIRONMENTAL ATTRIBUTE CATEGORIES USED IN EIA
Environmental Attributes
Air
Diffusion factor
Particulates
Sulfur oxides
Hydrocarbons
Nitrogen oxide
Carbon monoxide
Photochemical oxidants
Hazardous toxicants
Odors
Water
Aquifer safe yield
Flow variation
Oil
Radioactivity
Suspended solids
Thermal pollution j
Acid and alkali ;
Biochemical oxygen demand (BOD)
Dissolved oxygen (DO)
Dissolved solids ;
Nutrients i
Toxic compounds I
Aquatic life ;
Fecal coliforms '
t
Land
Soil stability
Natural hazard
Land-use patterns
Ecology
Large animals (wild and domestic)
Predatory birds
Small game
Fish, shellfish, and waterfowl
Field crops
Threatened species
Natural land vegetation
Aquatic plants
Sound
Physical effects
Psychological effects
Communication effects
Performance effects
Social behavior
Human Aspects
Lifestyles
Psychological needs
Physiological systems
Community needs
Economics
Regional economic stability
Public-sector review
Per capita consumption
Resources
Fuel resources
Nonfuel resources
Aesthetics
Source: Jainetal., 1993.
danger of being concealed and possibly overlooked. "Scoping", was introduced in EIA as a? '«- ?
process used to determine the range |(i.e., scope) of issues to be addressed. CEQ regulations >
require using the scoping process early in the planning stages, as soon as practicable Rafter agency
decision to prepare an EIS.
A-2
-------
TABLE A-2. ENVIRONMENTAL ATTRIBUTES USED IN EIA
Attribute
Air
Diffusion Factor
Particulates
Sulfur Oxides
Hydrocarbons
Variables to be
Measured
Data Sources
Mitigation of
Impact
« Stability
Mixing depth
Wind speed
Precipitation
Topography
The concentration of all
solid and liquid particles
averaged annual arithmetic
mean of all 24 h paniculate
concentrations at a given
location.
The 24 h annual arithmetic
mean concentration of SO2
present in the ambient air.
The 3 h average annual
concentration of ambient
hydrocarbons, expressed in
ppm, and measured
between 6 and 9 a.m. (peak
hydrocarbon concentration
time).
Primary sources of data are
the National Weather
Service and the United
States Geological Survey
(USGS).
Data sources include state
pollution control
departments, county air
pollution control offices,
multi-county air pollution
control offices, or city air
pollution control offices.
Data are generally
compiled and published
annually by air quality
monitoring programs
established by state
pollution control agencies;
the EPA; and county,
regional, multi-county, or
city air pollution control
agencies.
Data are generally
available from state air
quality monitoring
programs. Other potential
sources include the EPA
and city or county
monitoring agencies.
Mitigation techniques have
not been adequately
defined.
Source reduction
Reduction or removal of
receptors from the area
Paniculate removal
devices
Use of protected,
controlled environments
Source reduction
Reduction or removal of
receptors from polluted
areas
Gas removal devices
using absorption,
adsorption, and catalytic
converters
Use of protected,
controlled environments
Control of motor
vehicle emissions
Control of stationary
source emissions
Reduction or removal of
receptors from area
> Use of a controlled
environment
(continued)
A-3
-------
TABLE A-2. ENVIRONMENTAL ATTRIBUTES USED IN EIA (CONTINUED)
Attribute
Variables to be
Measured
Data Sources
Mitigation of
Impact
Air (continued)
Nitrogen Oxide
The average Annual
concentration of nitrogen
oxides in the ambient air,
measured in ppm.
Carbon Monoxide
The maximu4i 8 h and 1 h
concentration of carbon
monoxide measured in
micrograms per cubic
meter.
Photochemical Oxidants The maximunTi hourly
average concentration
measured in micrograms
per cubic meter.
Sources of data include
state pollution control
departments and county,
multi-county, or city air
pollution control offices.
Sources of data include the
state pollution control
department, the county air
pollution control office, or
the city air pollution
control office.
Sources of data include the
state pollution control
department, the county air
pollution control office, or
the city air pollution
control office.
Control of motor
vehicle emissions
Control of stationary
source emissions
Reduction or removal of
receptors from area
Gas removal devices
using absorption,
adsorption, and catalytic
converters
Use of a controlled
environment
Control of motor
vehicle emissions
Control of stationary
source emissions
Reduction or removal of
receptors from area
Control of motor
vehicle emissions
Control of stationary
source emissions
Reduction or removal of
receptors from area
Gas removal devices
using absorption,
adsorption, and catalytic
converters
Use of a controlled
environment
(continued)
A-4
-------
TABLE A-2. ENVIRONMENTAL ATTRIBUTES USED IN EIA (CONTINUED)
Attribute
Air (continued)
Hazardous Toxicants
Odors
Variables to be
Measured
The variable to be
measured varies with the
toxicant.
Data Sources
Only a few city, county,
regional, and state agencies
monitor hazardous
toxicants and emissions.
Data on toxicant
monitoring are available
from state and local air
pollution control agencies
when collected.
The average annual
concentration of
selected odor
contaminants in ppm by
volume.
The odor intensity, rated
from 0 (no odor) to 4
(strong odor) by a
panel.
No systematic monitoring
and data collection are
done by state and local
agencies.
Mitigation of
Impact
Use of materials that do
not generate hazardous
toxicants
Use of processes that do
not generate hazardous
toxicants
Source reduction
Control, removal
devices
Moving people from
contaminated areas
Dilution of odorant
Odor counteraction
Odor masking
Source reduction
' Removal or receptors
from polluted areas,
and/or downwind odor
path fatigued olfactory
odor perception
Water
Aquifer Safe Yield
Flow Variations
The amount of water
withdrawn in a unit of
time, usually expressed as
thousands of acre-feet of
water per annum.
The typical unit of flow
measurement is cubic
feet per second.
Velocity as measured in
feet per second.
Sources of data include
local USGS offices and
state water agencies.
Data sources include local
Army Corps of Engineers
offices and state water
agencies.
All activities likely to
change the physical nature
of the aquifer, land surface
runoff, and percolation.
Water availability to the
aquifer should be carefully
controlled.
All activities such as land
use projects and water
impoundment and
operation should be
considered minimize flow
variations from the mean
natural flow.
(continued)
A-5
-------
TABLE A-2. ENVIRONMENTAL ATTRIBUTES USED IN EIA (CONTINUED)
Attribute
Water (continued)
Oil
Radioactivity
Suspended Solids
Thermal Pollution
Acid And Alkali
Variables to be
Measured
Data Sources
Mitigation of
Impact
Quantitative?
milligrams of oil or
grease per liter of water
Qualitative: I
visible oil slick
oily taste/odor
coating of banks or
bottom i
i
The quantity of any
radioactive material in
which the
disintegrations per
second ar^ 3.7 x 1010,
expressed las Curie (Ci)
Microcuri^ (10~6Ci)
Picocurie (10'12Ci)
Readily se;ttleable
suspended solids are
measured in milliliters
per liter of settled water.
Water temperature
measured in degrees
Centigrade or Fahrenheit.
Data sources include local
Army Corps of Engineers
offices and state water
agencies.
Data may be obtained from
the Nuclear Regulatory
Commission (NRC) and
state water agencies.
Sources of data include
local USGS offices, local
Army Corps of Engineers
offices, and state water
agencies.
Sources of data include
local USGS offices, local
Army Corps of Engineers
offices, and state water
agencies.
Sources of data include
local USGS offices, local
Army Corps of Engineers
offices, and state water ,
agencies. .
Controlling all direct
discharge
Treatment of surface
runoff for oil separation
Restrict lagooning of oil
wastes to prevent
potential groundwater
contamination
Waste containing
radioactivity should be
treated separately by
means of dewatering
Monitoring and control
of radiation facilities
Controlling/treatment of
discharge, including
sanitary sewage and
industrial wastes
Minimize activity that
increases erosion or
contributes nutrients to
water
Use of cooling towers in a
closed-loop water cooling
system.
; ^Neutralization of acidic
, or alkaline waters by. <
incorporation of alkaline
or acid wastes,
respectively
Source reduction of acid
or alkaline wastes
(continued);.
A-6
-------
TABLE A-2. ENVIRONMENTAL ATTRIBUTES USED IN EIA (CONTINUED)
Attribute
Variables to be
Measured
Data Sources
Mitigation of
Impact
Water (continued)
Biochemical Oxygen
Demand (BOD)
The amount of oxygen
consumed (mg/L) by
organisms during a five-
day period at 20°C.
Dissolved Oxygen (DO) Milligrams of oxygen per
liter of water.
Dissolved Solids
Nutrients
Toxic Compounds
Aquatic Life
Total dissolved solids,
determined after
evaporation of a sample of
water and its subsequent
drying at 103°C.
Includes measurement of
phosphorus, nitrogen,
carbon, iron, trace metals
hi their appropriate terms.
The spectrum of toxic
materials is extremely large
and highly diverse in terms
of effects. Measurement
can be expressed as pg/L
for specific compounds.
Field observations
Sources of data include
local USGS offices, local
Army Corps of Engineers
offices, and state water
agencies.
Sources of data Include
local USGS offices, local
Army Corps of Engineers
offices, and state water
agencies.
Sources of data include
local USGS offices, local
Army Corps of Engineers
offices, and state water
agencies.
Sources of data include
local USGS offices, local
Army Corps of Engineers
offices, and state water
agencies.
Sources of data include
local USGS offices, local
Army Corps of Engineers
offices, and state water
agencies.
Data may be obtained from
local Fish and Wildlife
offices.
Treatment of all wastes
containing organic
material:
- biological
- chemical
- packaged units
Treatment of all wastes
containing organic
material:
- biological
- chemical
- packaged units
Controlled landfilling to
avoid possible leaching
Deep well injection of
brine
Control and treatment of
surface runoffs
Waste water treatment
Natural assimilation
Monitor and control of
all toxic wastes
Dilution
Control and reduction of
all water quality attributes
listed
(continued)
A-7
-------
TABLE A-2. ENVIRONMENTAL ATTRIBUTES USED IN EIA (CONTINUED)
Attribute
Variables to be
Measured
Data Sources
Mitigation of
Impact
Water (continued)
Fecal Coliforms
Coliform density, reported Sources of data include
in terms of coliform per local Army corps of
100 mL. [ Engineers offices and state
water agencies.
Treatment of all wastes
containing organic
material:
- biological
- chemical
- packaged units
Land
Soil Stability (Erosion)
Natural Hazard
Land-Use Patterns
Soil comppsition
Degree of slope
Length of [slope
Nature and extent of
vegetative! cover
Intensity/frequency of
exposure to eroding
forces
I
Specific to each type of
hazard. \
Data are generally
available from local U.S.
Soil Conservation Service
offices.
Compatibility of use
between parcels as
indicated by such variables
as: |
type and intensity of use
noise '
transportation pattern
prevailing Wind
direction ;
buffer zones
aesthetics
Sources of data include the
Corps of Engineers,
USGS, U.S. Forest
Service, National Weather
Service, state geologists,
and local universities.
Municipal land use plans,
county land use planning
commission, regional land
use council, Bureau of
Land Management,
National Park Service,
Bureau of Reclamation,
Corps of Engineers,
Tennessee Valley
Authority, and the
Department of Energy.
Erosion control devices:
- ground cover
- tile drainage
- grassed waterways
- terracing steep slopes
- catch basins
Specific to each type of
hazard.
Inclusion of buffer
zones
Use of zoning and land
use ordinances
Community
participation in the land
use planning process
(continued)
A-8
-------
TABLE A-2. ENVIRONMENTAL ATTRIBUTES USED IN EIA (CONTINUED)
Attribute
Variables to be
Measured
Data Sources
Mitigation of
Impact
Ecology
Large Animals (Wild
and Domestic)
Predatory Birds
Small Game
Fish, Shellfish, And
Waterfowl
Field Crops
Population
Number of species
Habitat (in hectares)
Human intrusion/noise
Population
Number of species
Habitat (in hectares)
Human intrusion/noise
Population
Number of species
Habitat (in hectares)
Human intrusion/noise
Population
Number of species
Habitat (in hectares)
Human intrusion
pH
BOD
DO
Coliform bacteria
Pesticide concentrations
Acres of land
Percent farmed
Type of crop
Natural habitat (in
hectares)
Human intrusion
Data sources include the
U.S. Fish and Wildlife
Service, wildlife experts,
and universities.
Data sources include the
U.S. Fish and Wildlife
Service, wildlife experts,
and universities.
Data sources include the
U.S. Fish and Wildlife
Service, wildlife experts,
and universities.
Data sources include the
U.S. Fish and Wildlife
Service, wildlife experts,
and universities.
Minimize human
intrusion/noise
Creation of National
Parks, National Wildlife
Areas, or other
protected areas of
habitat
Minimize human
intrusion/noise
Creation of National
Parks, National Wildlife
Areas, or other
protected areas of
habitat
Minimize human
intrusion/noise
Creation of National
Parks, National Wildlife
Areas, or other
protected areas of
habitat
Minimize human
intrusion/noise
Creation of National
Parks, National Wildlife
Areas, or other
protected areas of
habitat
Minimize human
intrusion/noise
Minimize use of
pesticides and
herbicides
(continued)
A-9
-------
TABLE A-2. ENVIRONMENTAL ATTRIBUTES USED IN EIA (CONTINUED)
Attribute
Variables to be
Measured
Data Sources
Mitigation of
Impact
Ecology (continued)
Threatened Species
Natural Land
Vegetation
Aquatic Plants
Population
Number of species
Habitat (in hectares)
Human intrusion
Acres of native
vegetation'
Number and types of
species |
Human intrusion
Data sources include the
U.S. Fish and Wildlife
Service, wildlife experts,
and universities.
Data sources include the
U.S. Fish and Wildlife
Service, wildlife experts,
and universities.
Population
Number of species
Habitat (inlhectares)
Human intrusion
pH |
BOD 1
DO
Coliform bacteria
|
Pesticide concentrations
Data sources include the
U.S. Fish and Wildlife
Service, wildlife experts,
and universities.
Minimize human
intrusion/noise
Creation of National
Parks, National Wildlife
Areas, or other
protected areas of
habitat
Breeding programs
Minimize land
conversion
Restrict vehicular
intrusion
Creation of National
Parks, National Wildlife
Areas, or other
protected areas of
habitat
Minimize waste and
nutrient inputs
Restrict drainage of
wetlands
Creation of National
Parks, National Wildlife
Areas, or other
protected areas of
habitat
(continued)
, i. ?
A-10
-------
TABLE A-2. ENVIRONMENTAL ATTRIBUTES USED IN EIA (CONTINUED)
Attribute
Variables to be
Measured
Data Sources
Mitigation of
Impact
Sound
Physical Effects
Psychological Effects
Communication Effects
Loudness, measured in
decibels (dB)
Duration
Frequency
Loudness, measured in
decibels (dB)
Duration
Frequency
Psychological stress
Loudness, measured in
decibels (dB)
Duration
Frequency
Ambient noise levels
Distance between
speaker and listener
Under the Noise Control
Act of 1972, EPA promul-
gates noise-emission stan-
dards for construction and
transportation equipment,
motors/engines, and elec-
trical equipment. Data for
construction noise are pro-
vided by the General
Services Administration.
OSHA provides noise ex-
posure criteria for occupa-
tional health.
Under the Noise Control
Act of 1972, EPA
promulgates noise-
emission standards for
construction and
transportation equipment,
motors/engines, and
electrical equipment. Data
for construction noise are
provided by the General
Services Administration.
OSHA provides noise
exposure criteria for
occupational health.
Under the Noise Control
Act of 1972, EPA
promulgates noise-
emission standards for
construction and
transportation equipment,
motors/engines, and
electrical equipment. Data
for construction noise are
provided by the General
Services Administration.
OSHA provides noise
exposure criteria for
occupational health.
Source reduction
Dampening
Dissipation
Deflection
Ear protection
Sound enclosures
Removal of receptors
from high noise areas
Source reduction
Dampening
Dissipation
Deflection
Ear protection
Sound enclosures
Removal of receptors
from high noise areas
Source reduction
Dampening
Dissipation
Deflection
Ear protection
Sound enclosures
Removal of receptors
from high noise areas
Use of headsets
(continued)
A-ll
-------
TABLE A-2. ENVIRONMENTAL ATTRIBUTES USED IN EIA (CONTINUED)
Attribute
Variables to be
Measured
Data Sources
Mitigation of
Impact
Sound (continued)
Performance Effects
Social Behavior Effects
Loudness, measured in
decibels (dB)
Loudness, measured in
decibels (dB)
Duration
Frequency
Ambient noise levels
Distance between
speaker and listener
Under the Noise Control
Act of 1972, EPA
promulgates noise-
emission standards for
construction and
transportation equipment,
motors/engines, and
electrical equipment. Data
for construction noise are
provided by the General
Services Administration.
OSHA provides noise
exposure criteria for
occupational health.
Under the Noise Control
Act of 1972, EPA
promulgates noise-
emission standards for
construction and
transportation equipment,
motors/engines, and
electrical equipment. Data
for construction noise are
provided by the General
Services Administration.
OSHA provides noise
exposure criteria for
occupational health.
Source reduction
Dampening
Dissipation
Deflection
Ear protection
Sound enclosures
Source reduction
Dampening
Dissipation
Deflection
Ear protection
Sound enclosures
Removal of receptors
from high noise areas
Human Aspects
Lifestyles
Variables to be measured
for this attribute cannot be
precisely defined. The
objective is to identify
general changes in social
activities thai will be
caused by the proposed
action. ;
Data for this attribute may
be generally obtained from
the predictions by
community social leaders,
local political leaders,
academics, etc.
Although impact to this
attribute cannot be
completely mitigated, the
effect of anticipated
impacts could be lessened
by forewarning
participants.
(continued)
A-12
-------
TABLE A-2. ENVIRONMENTAL ATTRIBUTES USED IN EIA (CONTINUED)
Attribute
Variables to be
Measured
Data Sources
Mitigation of
Impact
Human Aspects (continued)
Psychological Needs
Physiological Systems
Community Needs
Although no specific
variables are identified for
this attribute, a general
feeling of the degree to
which the psychological
needs of individuals and
communities are being met
can be obtained.
No variables can be
measured for this attribute.
The detailed activities and
implications of those
activities must be carefully
examined.
Population
Demographics
Available housing
Capacity of public
services
Characteristics of land
use
Data on this attribute can
be generally obtained from
psychologists, personal
surveys, local counselors,
clergy, and law
enforcement officials.
Data on this attribute can
be generally obtained from
psychologists, personal
surveys, local counselors,
clergy, and law
enforcement officials.
Data may be obtained from
public surveys, local
planning agencies, police
and fire departments, local
officials.
Including an action plan
that would provide
assistance for affected
individuals
Consultation
Social programs
Taking precautionary
measures to avoid the
impact.
Employing specific
safety practices
Using protective devices
Including a plan for
providing public services
to accompany the proposed
activity.
Economics
Regional Economic
Stability
Percentage of total regional
economic activity affected.
Data sources include local
and regional business and
employment statistics.
Increase the demand for
the output of highest
growth industries in the
region
Change the distribution
of demand for the
output
(continued)
A-13
-------
TABLE A-2. ENVIRONMENTAL ATTRIBUTES USED IN EIA (CONTINUED)
Attribute
Variables to be
Measured
Data Sources
Mitigation of
Impact
Economics (continued)
Public Sector Review
Per Capita Consumption
Annual average
revenues arid
expenditures of the
relevant government
agencies
Expenditures necessary
to provide adequate
public services without
the project'
i
Average amount that will
be spent in each future year
throughout the life of the
project by affected
individuals.
Data sources include State
and Local Finances, and
the Statistical Abstracts of
the United States.
Design project activities to
either reduce social costs
or increase payments to the
local government.
Data may be obtained from
State and Local Finances.
Establish direct linkages
with area industries,
businesses, or other
economic activities to
encourage inflows of
money.
Resources
Fuel Resources
Nonfuel Resources
Aesthetics
Rate of fuel
consumption (in Btu)
Useful energy output
derived from fuel
consumption
Heat content of fuels
Types of fuel
Points of resource
consumption
Consumption rates
Quantities pid content
of wastes ffom resource
acquisition' activities
Individual perception and
values for defining beauty
make it difficult to quantify
aesthetic impacts.
Data sources includes the
Gas Engineer's Handbook
Mining Statistics, Energy
Information
Administration, and State
and Local Statistics
Data sources include the
Gas Engineer's Handbook
Mining Statistics, Energy
Information
Administration, and State
and Local Statistics.
Data may be obtained from
surveys, and other specific
measurements.
Alternate fuel selection
Conservation of fuel
resources
Economizing on
resource requirements
Development and use of
substitutes
Recycling programs
Public participati on ;in :
planning processes
Designation of natural
areas
Source: Jain et al., 1993.
A-14
-------
In the first part of the EIA scoping process, a comprehensive list of impacts is
streamlined to a particular study to minimize the proliferation of insignificant items (Jain et al.,
1993). Impacts cannot be eliminated from this comprehensive list without first evaluating the
significance or relevance of those impacts to the proposed project. For example, it would be
inappropriate for a proposed project to consider impacts to a timber resources category if the
project does not utilize, or produce an adverse impact on, timber resources. Thus, not only does
this scoping process reduce inefficient use of time and resources, but it also helps to pinpoint the
most critical impacts for analysts and decisionmakers to consider.
The second part of the EIA scoping process entails the tiering of impacts. Tiering comes
into play when some of the impact categories on the "long list" are potentially affected by a
project, but they are of fairly insignificant consequence. Such impacts are tiered to a lower level
of importance and not initially evaluated in the study (although they may be evaluated during the
study if necessary). The EIAs used tiering to organize the comprehensive list of impacts in a
more manageable and meaningful manner, by differentiating relatively insignificant and
significant impacts. The same problem may exist in impact assessment where a practitioner may
need to evaluate potentially large numbers of impacts in the classification phase and streamline
the list in the characterization phase.
From these two activities, a comprehensive list of environmental impacts may be tailored
to a specific reference project to help analysts and decisionmakers pinpoint and address the
critical impacts associated with the project.
In summary, the EIA scoping process requires an early analysis of potential impacts with
reference to a specific project. The scoping process strives to
1) eliminate inappropriate impact categories from the analysis,
2) tier less important impact categories to a lower level of analysis, and
3) identify the critical impacts that must be addressed in the analysis.
A-15
-------
-------
Appendix B
Additional Impact
Assessment Methods
-------
-------
This appendix contains descriptions of methods that have been evaluated for applicability
to impact assessment but have not been tested or presented in the context of LCA. As in the
presentation of methods in Chapters 4 through 7, the methods in this Appendix are presented in
the order of increasing level of detail.
B.I GREEN INDICATORS
Green indicators are calculated characteristics of a product or process that may be used to
evaluate the environmental compatibility of the product or process by identifying indicators that
are undesirably high or low. The ultimate goal of the green indicators is to give environmental
concerns equal weight with other more traditional concerns, such as manufacturing and
reliability, as part of an overall approach to green engineering design (Navinchandra, 1991).
Table B-l lists some green indicators that may be useful for a simple impact assessment (i.e.,
loading-type assessment).
Strengths
The primary strength of using the green indicators is that they provide a multi-
dimensional view of a product system that can enable decisionmakers to simultaneously address
a wide variety of issues and concerns. For example, most environmental assessment techniques
only provide information on environmental effects. Green indicators provide information not
only on environmental effects, but also on product performance, recyclability, useful life, cost,
etc. Such information is integral to making high quality decisions concerning tradeoffs between
alternatives products, processes, and materials as well as between environmental, economic, and
production concerns.
Some additional strengths of the green indicators include the following:
relatively convenient and easy to calculate,
limited amount of external data is required,
can be used as part of an overall green product design program, and
involves a life-cycle perspective.
Weaknesses
One possible weakness of the green indicators is that they do not estimate environmental
impacts per se. Rather, the indicators are merely proxies that can be related to environmental
impacts. For example, although the degradability indicator provides an estimate of the portion
of material in a product that is degradable, it does not indicate how harmful the degradable and
B-l
-------
TABLE B-l.
EXAMPLE GREEN INDICATORS
Indicator
Percent Recycled
Degradability
Life
Junk Value
Separability
Potential Recyclability
Possible Recyclability
Useful Life
Utilization
Net Emissions
Total Emissions
Total Hazardous Fugitives
Source: Navinchandra, 1991.
Description
The percentage of recycled material in a product.
The ratio of the volume of degradable material in a product to the total
volume of the product.
The time it takes for the degradable portion of a product to degrade. A
curve showing the expected volume of reduction over time is used to
determine life.
This js a measure of the total time a product will take to degrade into
the environment. It is calculated as the area under the life curve (above)
and expressed in units of cubic inches per year.
A measure of what materials can be separated from a product. It is the
ratio pf the volume of separable materials to the total volume of the
product. (The notion of separability is different from disassembly.)
i
The ratio of the volume of recyclable materials to that of unrecyclable
materials.
Composites and glued materials are potentially recyclable but cannot be
recycled because they are inseparable. This indicator must be measured
on a part-by-part basis and must take into account the available
recycling methods and their economic viabilities.
When a material leaves the environment and enters the human world it
is being used. Useful life is defined as the time an item spends in the
activity for which it was designed.
The rptio of the useful life of a product or material to the time it takes to
"retufti" to the environment.
The respective sums of solid, gaseous, and waterborne emissions from a
partiqular product or process life cycle.
The s^im of all solid, gaseous, and waterborne emissions taken together
from ja particular product or process life cycle.
A measure of the weight of hazardous fugitives, expressed as the ratio
of the weight of hazardous emissions per unit weight of product.
nondegradable portions may be to thb environment;' The approach merely assumes; that less-'-
nondegradable material is necessarily- "better" for the;environment. MU. ' ,
B-2
-------
Some additional weaknesses of the green indicators include the following:
too simplistic,
does not account for impacts to human health, and
unclear how some indicators (e.g., life-cycle cost) would be calculated.
Relevance to Impact Assessment
The green indicators approach would likely be most suitable for a less detailed Tier 1-
type assessment of environmental impacts. Although somewhat simplistic, the green indicators
would enable decisionmakers to consider a wide variety of factors in addition to emission levels
that can be integral to making decisions concerning tradeoffs between alternative products,
processes, and materials as well as between environmental, economic, and production concerns.
B.2 POLAROID'S ENVIRONMENTAL ACCOUNTING AND REPORTING SYSTEM
(EARS)
Polaroid's Environmental Accounting and Reporting System (EARS) was developed as a
tool to help measure the progress of its Toxic Use and Waste Reduction (TUWR) Program goals.
EARS is a centralized database that allows Polaroid to track virtually every one of the 1,400
materials the company uses, from office paper to chlorinated solvents (Nash et al., 1992). Each
material is classified into one of five toxicity categories to reflect the degree of potential
environmental harm it poses (see Table B-2). With EARS, Polaroid records the quantities and
treatment methods of materials in all five categories at several points along the process line.
Use, waste, and by-products are measured and recorded per unit of production.
Strengths
EARS has turned out to be a beneficial program because it
provides employees with information needed to assess the environmental quality of
their actions;
provides incentives for making continual improvements in environmental performance;
provides an effective Total Quality Environmental Management (TQEM) tool,
fulfilling several different functions throughout the company;
allows employees to predict the environmental impacts of new chemicals before the
company makes a commitment to their use; and
translates complex environmental data into a simple index that has meaning throughout
the company (Nash et al., 1992).
B-3
-------
TABLE B-2. POLAROID'S EARS CATEGORIZATION OF CHEMICALS
Category
I&II
III
IV
Number of
Chemicals
Category I -38
Category H- 65
All remaining
chemicals
Examples
ammonia
benzene
CFCs
acetic acid
pyridine
kyrene
I.
acetone
butanol
i
Environmental
Impact
Most severe environmental
impact; highly toxic; human
carcinogens
Moderately toxic; corrosive;
suspected animal carcinogens
Least environmental impact
Reduction
Emphasis
Minimize use
Recover and reuse
onsite
Reuse onsite
following on or
offsite recycling
bardboard Depletes natural resources
paper during manufacture and
olastic
disposal
Maximize
recycling and reuse
onsite
Source: Modified from Nash et al., 1992.
Weaknesses
The primary weakness of EARS is that it does not measure environmental releases nor
does it estimate environmental impacts. EARS is essentially a classification system in which
chemicals may be grouped according to their known environmental toxicity.
t
In addition, many complain that EARS data requirements are too time consuming and
that EARS is cumbersome to,use (sefe Nash et al., 1992). Accuracy is also a persistent concern.
People responsible for computing EARS numbers and recording the data have varying levels of
skill and familiarity with the materials of interest. In addition, EARS is not linked with the
f
company's financial system. Thus, the company is unable to readily assess the financial benefits
of environmental improvements to its operation.
! .&* . *A,- --'
Relevance to Impact Assessment ' ' r
It is unclear how EARS could be used in the context of impact assessment. Perhaps ;at a
i
most basic level, inventory items coiild be grouped into EARS-like categories based on their
relative environmental toxicity. This would result in a listing of the most critical inventory items
B-4
-------
and their respective quantities that possibly could help analysts and decisionmakers pinpoint
improvement opportunities and/or areas that require a more detailed level of analysis. Such an
approach would likely be more appropriate for internal rather than external applications.
B.3 JUDGMENT PROBABILITY ENCODING
Judgment probability encoding was developed by Argonne National Laboratory to
provide a means of quantifying subjective probabilities for impacts. The main objective of this
approach is to reduce divergence among expert judgment through an encoding process in
estimating the probability of impact(s) resulting from exposure to substances.
Encoding in this context ensures that the questions used to derive judgment probabilities
are always phrased identically, that specific assumptions and definitions are always the same,
and that the encoding process proceeds similarly for each of the participants (Argonne National
Laboratory, 1991). Thus, any differences in judgment probabilities can be attributed to true
differences in values or opinions and not to differences in assumptions, understanding, or
procedures.
The output of the judgment probability encoding approach is a range of probabilities
regarding a specific function, (i.e., the likelihood of impact X resulting from pollutant A). The
encoded judgment probability may then be communicated in a variety of waysas a
distribution, a range, a mean, or a median. For example, consider the scenario where five
experts were solicited for judgment probabilities regarding the likelihood of X tons of CFCs
being linked with stratospheric ozone depletion. Each expert is provided exactly the same
information, assumptions, understanding, and procedures in exactly the same manner. For the
purposes of this example, generic probabilities are provided in Table B-3. These judgment
probabilities are then used to derive further quantitative characterizations of value judgments.
For instance, the analyst may decide to use a mean (0.25) to express the probability judgment
values or a range (0.15 to 0.35).
Strengths
The primary advantage of the judgment probability encoding approach is that it can take
into account the normalization (via impact probability values) of a wide variety of potential
impacts. By normalizing impacts, this approach enables decisionmakers to choose alternatives
from a subjective point of viewby relying only on the probability figures as impact
descriptors. In addition, the judgment probability encoding approach is easy to conduct.
B-5
-------
TABLE B-3. GENERIC ENCODED JUDGMENT PROBABILITIES EXERCISE
Expert
Median
Mean
Range
1
2
3
4
5
= 0.20-0.25
= 0.18-0.25
= 0.10-0.35
Judgment Probabilities for the Occurrence of Impact A
0.20
0.10
0.20
0.15
0.25
-0.25
-0.15
-0.30
-0.20
-0.35
Weaknesses |
The disadvantages of the judgment probability encoding approach are that it measures
impacts indirectly, in terms of judgment probabilities, and it may be too simplistic for impact
assessment. It would also, for all practical terms, be impossible to replicate the results of a
judgment probability encoding study. However, results from similar studies could be used to
i
verify and support the results of a judgment probability encoding study.
i
A code of good practice will ^ieed to be established for selecting and conducting the
expert encoding process to elicit judgment probabilities. Some questions that may need to be
considered in this respect include the following:
Who chooses the expert paijiel?
How many experts are required to conduct the approach?
From which fields should the experts be chosen?
.4?
Who approves the selection of experts and monitors the judgment probabiMtytericoding >
process? ;
Relevance to Impact Assessment
The judgment probability encoding process may. be useful in the context of impact
assessment as a simplified impact.characterization approach based upon expert judgment. Beingt
an entirely subjective approach, itwbuld be more appropriate for internal that external: ,>.'
B-6....
-------
applications. In addition, the judgment probability encoding approach may be useful in cases
where data on environmental conditions are not available or where nontraditional impact
categories are involved (e.g., species loss, habitat destruction, aesthetic loss).
B.4 HUMAN EXPOSURE DOSE/RODENT POTENCY DOSE INDEX
The Human Exposure Dose/Rodent Potency Dose (HERP) Index provides a common
factor for measuring the potency of various carcinogenic substances. The HERP Index is
calculated by determining the ratio of TD50 to human exposure. TD50 is the daily dose rate (in
milligrams per kilogram) needed to halve the percentage of tumor-free animals at the end of a
standard lifetime (Ames et al., 1987). Analogous to LD50 the lower the dose rate, or TD50
value, the more potent the carcinogen. Some example HERP Index values for specific
carcinogens are shown in Table B-4. Because the rodent data are calculated on the basis of
lifetime exposure at the indicated daily dose rate, the human exposure data are also expressed as
lifetime daily dose rates despite the notion that human exposure may likely be less than daily
over a lifetime.
TABLE B-4. EXAMPLE HERP INDEX VALUES
Potency of
Carcinogen
HERP
(%)
0.001
0.004
0.0004
0.0002
0.0003
0.008
0.6
0.004
2 1
Daily Human
Exposure
1 liter (tap water)
1 liter (well water worst)
1 liter (well water best)
1 hour (pool)
14 hours (A/C conventional
home)
14 hours (A/C mobile home)
Carcinogen Dose Per
70-kg Person
Chloroform, 83 pg
Trichloroethylene, 2,800
Pg
Trichloroethylene, 267
Pg
Chloroform, 12 pg
Tetrechloroethylene, 21
PS
Chloroform, 250 pg
Formaldehyde, 598 pg
Benzene, 155 pg
Formaldehyde, 2.2 pg
Rats
(119)
(-)
(-)
(119)
101
(119)
1.5
(157)
1.5
Mice
90
941
941
90
(126)
90
(44)
53
(44)
References
96
97
948
99
100
28
Source: Ames et al., 1987.
B-7
-------
Strengths
Using the HERP Index to assess carcinogenic impacts provides a means of normalizing
carcinogenic substances and allows for different types of carcinogens to be directly compared for |
their carcinogenic potential. In addition, the HERP Index allows for different types of
carcinogens to be aggregated so that the total contribution of inventory items to cancer can be
assessed. In addition, a TD50 database already exists but is quite extensive.
Weaknesses '
On the downside, using the HERP Index values as direct estimates of impact would be
inappropriate. Many uncertainties and assumptions are associated with extrapolating from
experimentation on rodents to valued for human carcinogenicity. Another problem with using
the HERP Index is that information is lacking on natural carcinogens and their relationship to
man-made carcinogenic substances.;
In addition, the HERP Index is based on the assumption that dose-response relationships
are linear, but this assumption may not be correct. Dose responses that are not linear but
quadratic or hyperbolic would yield HERP Index values much lower than those obtained by
using a linear dose response mechanism.
Relevance to Impact Assessment i
The HERP Index may be useful in the context of impact assessment for characterizing,
comparing, and/or aggregating the carcinogenic impact of inventory items. It should be stressed
that this method is only applicable for assessing carcinogenic impacts. However, because the
HERP index is highly controversial {within its own field of human health research, it should not
be used in impact assessment. :
B.5 ENVIRONMENTAL INDICES
i
A wide variety of environmental indices have been developed to provide an estimate
i-
ambient pollutant levels in different environmental media. These ambient levels of pollutants
are used as a proxy for estimating environmental impacts. These indices are, in essence,
equivalency functions that may be u|sed to compare the relative impact of a variety of different?
substances released into the environment. This section discusses two main groups,of ;
indicesair pollution indices and water pollution indices. : . "
-------
Strengths
A main strength of the environmental indices described in this section is that they have
been developed, refined, and used in practice for a number of years. There is a large body of
experience to draw upon for using and interpreting such indices.
Weaknesses
A primary weakness of the indices included in this section is that they account for only a
small subset of possible pollutants. In addition, the indices provide measures of ambient
concentrations for a region as a whole. Thus using the indices to estimate the contribution of a
single source of pollution to overall regional levels would be difficult.
Relevance to Impact Assessment
Although the indices described in this section do not measure impacts per se, they may
be used to compare "before" and "after" scenarios for the releases of a proposed project or used
as baseline information for conducting a detail impact assessment. Beyond providing an
indication of ambient pollutant concentrations in regional air and water sinks, the use of
environmental indices in the context of impact assessment is unclear.
B.4.1 Air Pollution Indices
A number of air pollution indices have been proposed in journals, conference
proceedings, and research reports. Additional indices have been developed by state and local air
pollution control agencies and have been implemented to routinely report air quality data to the
public. In the mid-1970's, so many different reporting schemes were in use that the government
found it necessary to adopt a national air pollution index, the Pollutant Standards Index (PSI)
(Ott, 1987). The PSI and other air pollution indices are summarized in Table B-5.
B.4.2 Water Pollution Indices
Indices have also been developed that can be used with data available from current water
quality monitoring activities to provide an estimate of water pollution (see Table B-6). There
are two basic types of water pollution indices: increasing scale indices and decreasing scale
indices. Increasing scale indices refer to "water pollution" indices while decreasing scale indices
refer to "water quality" indices (Ott, 1987). Water pollution indices may also be grouped into
five main categories:
B-9
-------
TABLE B-5. CLASSIFICATION OF AIR POLLUTION INDICES
Variables"
Index
Classification*
CO NO, OX TSP COH SO, OthJ
2A3C
2CC
3C3C
iA3B
;
7C3C
nA3A
(n=l to 5)
nA3A
(n=l to 5)
nA3A
(n=l to 4)
Green's Index
Combustion Products Index
(CPI)
Measure of Undesirable
Respirable Contaminates
(MURC)
Air Quality Index (AQI)
Ontario Air Pollution Index
(API)
PINDEX
Oak Ridge Air Quality Index
(ORAQI)
MITRE Air Quality Index
(MAQI)
Extreme Value Index (EVI)
Short Time Averaging
Relationships to Air
Quality Standards
(STARAQS)
(PSD
* Classification is based on the Thom-Ott air jpollution index classification system. The first digit indicates the number of
pollutants the index addresses. The first letter indicates the calculation method used, where A = nonlinear, B = segmented
linear, C = linear, and D = actual concentrations. The subindex number to the calculation method indicates the type of,'
calculation model used, where 1 = individual, 2 = maximum, and 3 = combined. The last letter indicates the type of :
descriptor categories used by the index, where A = standards, B = standards and episode criteria, and C = arbitrary, v
b CO, carbon monoxide; NO2, nitrogen dioxide; OX, photochemical oxidants; COH, coefficient of haze; SP, total
suspended particulars; SO2, sulfur dioxide.
c Fuel burned and ventilating volume.
d Hydrocarbons and solar energy. .
e Visibility and industrial emissions. '
,
Source: Ott, 1987. -
ental Quality Index
air)
Standards Index
8A3A
5B2B
[Total
»
756759
B-10
-------
TABLE B-6. CLASSIFICATION OF WATER POLLUTION INDICES
Index
^^MM
General Indices
Quality Index (QI)
Water Quality Index (WQI)
Implicit Index of Pollution
River Pollution Index (RPI)
Social Accounting System
Specific-Use Indices
Fish and Wildlife (FAWL)
Index
Public Water Supply (PWS)
Index
Index for Public Water
Supply
Index for Recreation
Index for Dual Water Uses
Number of
Variables
10
13
11
13
11/13
12
31
Scale
decreasing
decreasing
increasing
increasing
decreasing
decreasing
decreasing
decreasing
decreasing
decreasing
Variables Used
DO, alkalinity, chlorides, CCE, pH,
temperature, specific condition, total
coliforms, other biological.
DO, BOD, nitrates, phosphates, pH,
temperature, turbidity, total solids,
fecal coliforms.
DO, BOD, COD, iron, manganese,
ammonia, nitrates phosphates, ABS,
CCE other chemical, pH, suspended
solids.
DO, BOD, COD, phosphates, other
chemical, temperature, specific
condition, total coliforms.
DO, BOD, alkalinity, hardness,
chlorides, pH, temperature, specific
condition, total solids, fecal
coliforms, total coliforms.
DO, ammonia, nitrates, phosphates,
phenol, pH, temperature, turbidity,
dissolved solids.
DO, alkalinity, hardness, nitrates,
chlorides, fluorides, sulfates, phenol,
pH, turbidity, dissolved solids, color,
fecal coliforms.
DO, BOD, hardness, iron, nitrates,
fluorides, phenol, pH, temperature,
turbidity, dissolved solids, color,
fecal coliforms.
DO, nitrates, phosphates, oil and
grease, pH, temperature, turbidity,
suspended solids, color, other
physical, total coliforms.
iron, manganese, ammonia, nitrites,
chlorides, fluorides, sulfates, phenol,
other chemical, pH, specific
condition, color, fecal coliforms.
(continued)
B-ll
-------
TABLE B-6. CLASSIFICATION OF WATER POLLUTION INDICES (CONTINUED)
Index
Number of
Variables
Scale
Variables Used
Specific-Use Indices (continued)
Index for Three Water Uses
Planning Indices
Prevalence Duration Intensity
(PDI) Index
Nation al Planning Priorities
Index (NPPI)
Statistical Approaches
Composite Pollution Index 18
(CPI)
Index of Partial Nutrients 5
Index of Total Nutrients 5 [
Principal Component Analysis b '
Harkins* Index b
Beta Function Index b
increasing DO, alkalinity, hardness, iron,
manganese, chlorides, sulfates, pH,
temperature, turbidity, suspended
solids, total solids, color, other
physical, fecal coliforms.
increasing
Note: Because of their flexibility and
special-purpose nature, the planning \
indices and statistical approaches
do not lend themselves to detailed
comparison.
increasing
Priority Action Index (PAI)
Environmental Evaluation
Systems (EES)
Canadian Pollution Index
(CPI)
Potential Pollution Index
(PPI)
Pollution Index (PI)
b
78J
\
b
3
b
increasing
decreasing
increasing
increasing
increasing
increasing
decreasing
decreasing
N/A
increasing
increasing
* Water quality variables account for 14 of the 78 variables used in this system.
b Any number of variables can be included.
Source: Ott, 1987. '
B-12
-------
general water quality indices,
specific-use indices,
planning indices,
statistical approaches, and
biological indices (Ott, 1987).
Table B-6 summarizes these indices (with the exception of biological indices not
amenable to classification). Three general types of biological water quality indices evaluate
water quality on the basis of its impact on aquatic lifetypes and quantities of certain indicator
organisms, mathematical properties of populations of organisms, and physiological or behavioral
responses of certain organisms to pollution.
B.6 DEGREE OF HAZARD EVALUATION
The degree of hazard evaluation system was developed as a scientifically sound and
consistent way to deregulate the tracking of non-RCRA special wastes that pose low or
negligible hazard. The degree of hazard evaluation ranks wastes according to their respective
degrees of hazard and is based on five characteristics of a waste stream:
weighted accumulative toxicity of constituents (as modified by environmental fate),
disease potential (infectious waste),
fire (ignitability),
leaching agents (pH), and
biological hazard (biodegradability) (Plewa et al., 1986).
The degree of hazard evaluation places primary emphasis on toxicity to rank potential
hazard. Thus toxicology data are used to generate a numerical score for a substance's equivalent
toxicity (Plewa et al., 1986). The calculation of equivalent toxic concentration of each life-cycle
waste component (C^) is as follows:
Equivalent Toxic Concentration = C = A? ((
where
= the concentration of component i as a percentage of the waste by weight,
= a measure of the toxicity of component i,
B-13
-------
A = a constant equal to 300 used to allow entry of percent values for C{ and to adjust
the results so that a reference material, 100 percent copper sulfate with an oral
toxicity of 300 mg/kgj, achieves an equivalent toxicity of 100, and
i '
Bj = a conversion factor u|ed to convert toxicities (tj) to equivalent oral toxicities.
Table B-7 shows conversion factors (Bj) for various toxicity measures.
For carcinogens and mutagens, a TD50 oral rat dose is used if available. Otherwise
carcinogens are assigned a Tj 0.1 mg/kg, and mutagens are assigned a T} of 0.6 mg/kg.
Toxicities are converted to equivalent oral toxicities as specified in Table B-7. Oral rat toxicity
values are preferred, followed by inhalation rat, dermal rabbit, aquatic toxicity, and other
mammalian toxicity values. If there! is more than one value for the toxicity from the best
available source, the lowest (most tojxic equivalent oral toxicity value) is used. If a carcinogen
or mutagen is assigned a value for T^ in the absence of a TD50, Bj is assigned a value of 1.
The relative toxic amount, M,, of the entire waste stream mixture is calculated as follows:
Relative Toxic Amount = M = S CEQ
f
where S = the maximum size (kg) of waste output produced in a month. The result of these
calculations will be an estimate of the relative toxic amount (M) for each waste output evaluated
that takes into account the comparative toxicity and amount of each component. For each waste
output, the number calculated for Mjcan range from 0 to greater than 10,000. The relative toxic
amount is then converted into categories of hazard: negligible, low, moderate, or high.
TABLE B-7. TOXICITY CONVERSION FACTORS
Conversion Factors
Toxicity Measure
Oral-LD50
Carcinogen/mutagen - LD50 I
Aquatic - 48 or 96 hr LC50 j
Inhalation -LC50
Dermal -ID
For The Equivalent Oral
Units
mg/kg
mg/kg
ppm
mg/1
mg/kg
Toxicities (Bj):
BI
1.00
1.00:
5.00: . i
25.oa i ; i , ,
0.25"'"*
Source: Thomas and Miller, 1992.
B-14.;
-------
The results of an actual degree of hazard evaluation conducted to evaluate two types of
sand wastes produced by an iron foundry are illustrated in Tables B-8 and B-9. For a more
complete description of the degree of hazard evaluation, refer to Reddy (1985), Plewa et al.
(1986), and Plewa et al. (1988).
Strengths
The degree of hazard evaluation method may be used to normalize chemical substances
in a manner that allows the analyst to compare not only the equivalent toxicity of various
chemicals but also other inherent characteristics of those hemicals. In addition, the degree of
hazard evaluation method has been used in practice and refined for a number of years.
Weaknesses
One problem with using the degree of hazard evaluation in impact assessment is data
availability. Out of over 5,000 RCRA and non-RCRA waste streams analyzed, over 70 percent
were ranked as "unknown" hazards due primarily to the following data deficiencies:
information that was required on waste streams but was missing,
TABLE B-8. DEGREE OF HAZARD EVALUATION OF IRON FOUNDRY
MOLDING SAND WASTE #1
Sand Waste #1
Component Name
Chromium
Barium peroxide
Arsenic pentoxide
Lead monoxind
Cadmium
Selenium dioxide
Total Equivalent Toxicity
Overall Hazard Ranking
Concentration (5)
0.000002
0.000012
0.000002
0.000005
0.000000
0.000002
Equivalent Toxicity
0.00006
0.000003
0.00000008
0.000000001
0.000000000
0.000000000
0.000063
Negligible
Source: Thomas and Miller, 1992.
B-15
-------
TABLE B-9. DEGREE OF HAZARD EVALUATION OF IRON FOUNDRY
MOLDING SAND WASTE #2
Sand Waste #2
Component Name
Nickel
Phenol
Cadmium
Chi oroform
Barium peroxide
Fluorine
Chromium oxide
Lead monoxide
Xylenes, total
Arsenic pentoxide
Methylene chloride
Toluene
2-butanone
Acetone
Silver dioxide
Mercury oxide
Selenium dioxide
Silica
Total Equivalent Toxicity
Overall Hazard Ranking
Concentration (%)
000171
0,001544
0;00008
0,000039
0.00028
0,09
0,00017
0[000074
OJ000002
01000002
0|00003
0.000044
0;00022
0100042
0.000035
ojooooooooo
0,000003
99.9049
i
Equivalent Toxicity
0.0513
0.00772
0.0024
0.00117
0.000067
0 .000058
0.000006
0.000000964 3
0.0000000888
0.000 000075
0.00000005389
0.0000 000528
0.00000004074
0.0000000252
0.00000000372
0.00000 0000
0.000000000
0.000000000
0.0068
High
Omtw%A* TlmAVMAf* «n*4 Xlf^lln..
data necessary for many toxicity hazard calculations that were not available in the
public literature, and
[ ^ ,;.'"]
vague names for wastes or chemicals that were often used rather than trade names..
Relevance to Impact Assessment I
The degree of hazard evaluation method may be used to normalize a varietyiiof chemical-
based inventory items in a manner that allows the analyst to compare not onlyi the equivalent . v
B-16
-------
toxicity of various inventory items but also other inherent characteristics of those items as well.
In addition, degree of hazard evaluation projects have been used in practice for a number of
years and thus may be currently applicable to impact assessment.
B.7 HAZARD RANKING METHODS
A number of hazard ranking methods have been developed for a variety of different
purposes. Hazard ranking methods, much like Tier 2- and Tier 3-type characterization models,
rank the relative risk of substances released to the environment based on hazard (e.g., toxicity)
and sometimes exposure (e.g., persistence, bioaccumulation) information. The following
sections describe some of the primary hazard ranking methods.
Strengths
Hazard ranking methods have several identifiable advantages. Most are relatively easy to
use, they do not require extensive data, and three major routes (groundwater, surface water, and
air) are considered. In addition, factors have been carefully selected for consistency and to avoid
redundancy, and they often are built upon previously developed models (including the JRB
Associates, Inc. model).
Weaknesses
A number a criticisms have been raised about using hazard ranking methods:
The score for hazard potential is based on only the most hazardous substance rather
than on a composite of all constituents.
Low population areas tend to receive lower scores than higher population areas.
The use of distance to population as a weighting factor is used even in situations where
there is no evidence of release.
Few provisions exist for incorporating additional technical information into the models.
Individual factor scores are often aggregated into a composite total score.
Relevance to Impact Assessment
The hazard ranking methods described in this section are most similar to the Tier 2- and
Tier 3-type characterization models described in Chapters 3 and 4 (see the toxicity, persistence,
and bioaccumulation assessment). Most of these methods have scoring systems in which the
relative "hazard" associated with a variety of substances is estimated. Because most of the
hazard ranking methods focus on impacts to human health, it is not clear whether they would be
useful for estimating impacts to ecosystems or natural resources.
B-17
-------
B.7.1 EPA's Hazard Ranking System (HRS)
The HRS was developed by tjie MITRE Corporation to meet Comprehensive
Environmental Response, Compensation, and Liability Act (CERCLA) requirements mandating
that ranking systems be based on relative risks. In this context, relative risk takes into account
the population at risk, the hazardous potential of releases, the potential for contamination of
drinking water supplies (for both ecosystem and human health impacts), and other appropriate
factors. HRS ranks facilities in term$ of the potential threat they pose by describing the manner
in which hazardous wastes are contained, the route by which they are released, the characteristics
and amount of the hazardous substance, and the likely ecosystem and human health targets (see
Table B-10). :
TABLE B-10. OVERVIEW OF RATING FACTORS
Category
Groundwater
t
Route
Factors
Surface-Water Route
Air Route
Route Characteristics
Containment
Waste Characteristics
Targets
depth tp aquifer of
concern
net precipitation
permeability of
unsatupted zone
physical state
containment
toxicity/persistence
hazardous waste
quantity
groundwater use
distance to nearest
well/pdpulation
served
facility slope and
intervening terrain
one-year 24-hour
rainfall
distance to nearest
surface water
physical state
containment
toxicity/persistence
hazardous waste
quantity
surface water use
distance to sensitive
environment
population
served/distance to
water intake
downstream
reactivity
incompatibility
toxicity
hazardous waste
quantity
land use> «
population within 4-|
mile radius
distance to sensitive |
environment
Source: Federal Register, 1988.
B-18
-------
The HRS assigns three scores to a hazardous facility:
1. The potential for harm to humans or the environment from migration of a
hazardous substance away from the facility by routes involving groundwater,
surface water, or air.
2. The potential for harm from substances that can explode or cause fires.
3. The potential for harm from direct contact with hazardous substances at the
facility (Sandia National Laboratories, 1986).
Scores for each hazard mode are determined by evaluating a set of factors that
characterize the potential of the particular facility to cause ecosystem and human health impacts!
Each factor is assigned a numerical value on a scale of 0 to 3, 5, or 8, according to prescribed
guidelines. The assigned value is then multiplied by a weighting factor to yield the individual
factor score. The individual scores may then be aggregated within each factor category, and
then the aggregated scores for each factor category are multiplied together to develop scores for
migration (groundwater, surface water, air), fire and explosion, and direct contact.
Use of the HRS requires information about the facility in question, its surroundings, the
hazardous substances present, and the geological characteristics of the surrounding area. When
there are no data for a factor, it is assigned a value of zero. However, if a factor with no data is
the only factor in a category, then the factor is given a score of 1.
B.7.2 Modified Hazard Ranking System (MHRS)
The Modified Hazard Ranking System (MHRS) was developed by Battelle Pacific
Northwest Laboratory (PNL) for DOE to rank sites that contain both chemically hazardous and
radioactive wastes. MHRS was developed to work within the framework of EPA's HRS, and
the overall scoring system is the same for both methods. The modifications to the HRS for sites
containing radioactive wastes were restricted to the waste characteristics category of the ground-
water, surface-water, air, fire and explosion, and direct-contact routes.
In developing a scoring system for radioactive wastes in MHRS, the concentration and
the type of radiation emitted by the radionuclides were factored into the ranking. The scoring of
the radionuclides is based on an estimate of the potential radiation dose to a maximally exposed
individual (the product of dose factor times concentration is estimated).
The MHRS splits the waste characteristics categories into chemical wastes and
radioactive wastes. The scoring system for chemical wastes is the same as that of EPA's HRS.
The hazards of the radioactive and nonradioactive wastes are evaluated separately and the score
B-19
-------
is assigned over the same range of values. The higher score of the two is the value assigned to
the site. The site ranking is based on the maximum score (chemical or radioactive) from each
route and is calculated as described in HRS. Scoring for radioactive wastes through each route
is described below. \
For the air route, information [on the maximum observed concentration of radionuclides
in air at the site is required. If no concentration of atmospheric radioactivity significantly above
background has been observed, then ihe waste characteristics score for the air route is zero. If
release of radionuclides has been observed, then the total concentration for each radionuclide
group is calculated. A matrix table for the air route is then used to determine the waste
characteristics score by selecting the largest value among the groups.
For the surface-water route, if release has been observed, the total surface-water
concentration for each nuclide group [is determined and the highest resulting waste characteristics
score among the groups is selected. The largest score among nuclide groups derived from the
maximum potential surface-water releases is then compared with that from observed release.
The greater of the two is recorded in the surface-water route.
In the groundwater route, if release has been observed, the highest waste characteristics
score among the nuclide groups resulting from the observed releases is used. This score is then
compared with the score calculated from the maximum potential release. The maximum
potential concentration for each radionuclide is determined by multiplying the amount disposed
of at the site by the transport coefficient. The total potential groundwater concentration
associated with each nuclide group is calculated by summing all radionuclides within the group
(see Table B-l 1). The waste characteristics score for each nuclide group can then be determined
from a matrix table (see Table B-12). The largest value among the groups is compared with that
from the observed release. The greater of the two values is recorded for the groundwater route.
The fire and explosion route ^nd the direct-contact route are usually of less importance
than other routes for hazardous waste sites. Therefore, a detailed description for scoring these ,
two routes is not provided here.
B.7.3 U.S. Air Force (USAF) Hazard Assessment Rating Methodology
I '
The US AF has sought to establish a system to "develop and maintain a priority listing of
contaminated installations and facilities for remedial action based on potential hazard to public
health, welfare, and environmental impacts" (Sandia National Laboratories, 1986);
this system priorities are to be set for taking further actions at sites. - Thus the Hazard
B-20
-------
TABLE B-ll. RADIONUCLIDE GROUPS
Group
Nuclides
A
B
C
D
E
F
226+DRa, unidentified alpha emitters
129I, 210+DPb, 90+DSr, 229Th, 233+DU, unidentified beta and gamma emitters
241 Am, 243Am, 134Cs, 237+DNp, 230+DTh, 232+DTTi
243Cm, 244Cm, 60Co, 135Cs, 17+DCs, 152Eu, 154Eu, 22Na, 94Nb, 73Ni, 238Pu, 239Pu,
240Pu, 234U, 244Pu, 225Ra, 151Sm, 99Tc, 22*+DTh, 234U, 238+DU, 235+DU
225Ac, 14C, 55Fe, 93Mo, 59Ni, 239Np, 241Pu, 125+DSb, 240U
3H
Source: Sandia National Laboratories, 1986
TABLE B-12.MATRIX TABLE FOR GROUNDWATER ROUTE WASTE
CHARACTERISTICS SCORE
Maximum Ground- Water Concentration (pCi / L)
Nuclide 1(T3
A 0
B
C
D
E
F
10'2 10'1
1 3
0 1
0
10°
7
3
1
0
101
11
7
3
1
0
102
15
11
7
3
1
0
IO3
21
15
11
7
3
1
104
26
21
15
11
7
3
10s
26
21
15
11
7
106 107
26
21 26
15 21
11 15
io8 io9
26
21 26
Source: Sandia National Laboratories, 1986
Assessment Rating Methodology was developed to provide a relative ranking of sites that are
suspected of having been contaminated from hazardous substances.
The Hazard Assessment Rating Methodology considers four aspects of the hazard posed
at a specific site:
B-21
-------
the possible receptors of the; contamination,
t
the waste and its characteristics,
potential pathways for wast£ contaminant migration, and
any efforts to contain the cqntaminants.
Each of these categories contains a number of rating factors that are used in the overall
hazard rating. For example, the wastje characteristics category is scored in three steps. First, a
point rating is assigned based on an assessment of the waste quantity and the hazard (worst case)
associated with the site. Second, the score is multiplied by a waste persistence factor, which
reduces the score if the waste is not Very persistent. Finally, the score is modified according to
the physical state of the waste. Scores for liquid wastes are unchanged, while scores for sludges
and solids are reduced. The scores % each of the three categories are then added and
normalized to a maximum possible score of 100.
The US AF Hazard Assessment Rating Methodology is based on the same JRB model as
the EPA HRS method and is similar ^n many respects. The best way to highlight the strengths
and weaknesses of the US AF ranking method is to identify those components that differ
significantly from the HRS approach. These differences are found in the areas of:
Waste Quantity: the US AF method deals more realistically with the quantities of toxic
substances by having "quantity" indicate the total amount of chemicals in a particular
hazard classification.
Persistence: values for persistence in the USAF method are used to modify the waste
characteristics score (based on toxicity and quantity). This may be inappropriate
because different types of chemicals contribute to the overall waste characteristics
score (i.e., it is better to combine toxicity and persistence considerations for individual
chemicals as done in the HJRS than to apply one persistence score to a diverse class of
chemicals).
Air Releases: these are not considered in the USAF ranking method, thus the potential
risks associated with a siteicould be underestimated.
B.7.4 Relative Hazard Ranking System
Hazard evaluations for toxic [chemicals and low-level radioactive wastes have generally
been performed independently of one another and without a means of comparison. Exposure of
ecological systems to ionizing radiation usually results in nonspecific damage, while exposure to
a chemical can produce specific damage to a specific biologic activity. Comparing radioactive
hazards with chemical hazards is difficult because of differences between the, underlying ..??,
B-22,
-------
mechanisms of radiation and chemical effects. In ranking waste disposal sites that contain a mix
of chemical and radioactive wastes, a relative rating of chemical hazard and radiation effects is
necessary. A few approaches have been suggested that might be useful in comparing relative
hazards. Four of these approaches are summarized in Table B-13.
B.8 THE AMOEBA APPROACH
AMOEBA is the Dutch acronym for "a general method of ecosystem description and
assessment." The AMOEBA approach is based on the concept of sustainable development and
was developed for and applied to the Dutch Water Management Plan (see Kuik and Verbruggen,
1991; and Udo de Haes, Nip, and Klijn, 1991). AMOEBA is a conceptual model for the
development of quantitative and verifiable ecological objectives, and it provides a means for
quantitatively describing and assessing ecosystems.
The AMOEBA approach employs "ecological values," which are defined as desired
states of ecological components as predetermined by decisionmakers and/or stakeholders. In
order to establish precisely these ecological values, the most fundamental values humans
attribute to plant and animal life are examined. Three categories of ecological characteristics are
used in deriving ecological values:
TABLE B-13. ALTERNATIVE APPROACHES TO RELATIVE HAZARD
RATINGS
Approach
Applicability
Limitations
Rem-Equivalent Chemical
MPC/EPC-Air and Water
Equivalents
Equivalent Hazard Categories
Site-Specific Risk Management
Committee
carcinogenic
mutagenic
teratogenic
substances
performance criteria
disposal volumes
offsite concentration limits
general toxic effects
based on definitive data
local conditions
credible
easily understood
The significance of dose-
response and safety standards is
undefined and depends on levels
of acceptable risk.
Depends on validity of
MPC/EPC limits subject to
change.
Database is usually acute rather
than chronic toxicity.
Potentially subjective changes in
value judgments with time or
committee members.
B-23
-------
I
Production and Yield: These characteristics are valuable for functional reasons. This I
category is a prerequisite for human existence (e.g., fisheries). These values are closely]
associated with the abundance of species, the production of oxygen, and the self-
purifying capacity.
Species Diversity: This is| valuable for ethical and aesthetic considerations. It
involves concepts such as the preservation of species, rarity, and completeness.
Self-Regulation: Self-regulation has ethical, aesthetic/recreational, and economic
considerations that are closely related to concepts such as naturalness, stability,
intactness, authenticity, ancl visual integrity. Moreover, self-regulating ecosystems
have low management costs (Kuik and Verbruggen, 1991).
AMOEBA-type approaches typically present three values for each study parameter:
reference (baseline) values, target (objective) values, and current (measured) values. The
relationship between these three values is shown in Figure B-l. These values are plotted on a
circular figure for each parameter (see Udo de Haes, Nip, and Klijn, 1991). Determination of
these three values is integral to the AMOEBA approach.
i
Reference values are obtained by a reference system, which has been only slightly
influenced by human activities or not at all. Such a system contains the conditions for the
evolution and survival of organisms^ including humans, living in and around it. The
introduction of a reference system provides a standard against which an assessment of the
ecological condition of a system can be made. The closer one can come to mimicking the
reference system, the larger the chance of ecological sustainability. The overall ecological
objective, however, does not necessarily have to coincide with the reference system.
f t
Decisionmakers and/or stakeholders must decide on the maximum acceptable distance
from the reference point to establish a verifiable ecological objective. This distance is the target
value. Target values may both exceed or fall short of the reference values, depending on the
parameter. The compromise between the ecological quality objective (the target value) and the
reference value is evidenced by the Discrepancies between the two values. *
i ;> .
Current or measured values represent the actual state of the system. Currentevaluesimay
i
be determined by direct measurement, modeling, or through secondary data sources; The ;; ;
i,
difference between the target values and the current value indicates the extent to which the. '.
ecological objective has fallen shortjof or been surpassed by either an existing or .proposed.: j
activity. I .
B-24
-------
Current
Objective
Bad
measures
0 Reference
Figure B-1. Relationship Between AMOEBA Components
Source: Udo de Haes, Nip, and Klijn, 1991.
A case study example using the AMOEBA approach is provided in Udo de Haes, Nip,
and Klijn (1991).
Strengths
The primary strength of the AMOEBA approach is that it uses actual environmental
conditions as baseline information against which to estimate the environmental implications of a
project. Such an assessment can provide a more realistic study of the effects of a project on the
environment.
Weaknesses
A weakness of the AMOEBA approach is that, although reference environmental values
are determined relatively objectively, establishing target values is a highly subjective process
that would involve consulting experts from a variety of different fields of expertise. In addition,
a method for integrating and interpreting the effects of environmental releases from other
projects is not clear.
Relevance to Impact Assessment
In the context of impact assessment, the AMOEBA approach may be useful for
estimating the extent to which impacts are within or exceed, the stakeholder-determined
maximum acceptable values (the target values) from environmental reference points.
B-25
-------
In addition, because the current values represent the actual state of the system, the
difference between the target values and the current value indicates the extent to which the
ecological objective has fallen short Of or been surpassed by either an existing or proposed
activity. This information may be useful for highlighting to decisionmakers the compromise
I
between the ecological quality objective (the target value) and ambient environmental conditions
(the reference value).
B-26-
-------
Appendix C
Key Terms and Definitions
-------
-------
Assessment Endpoint
An impact of concern identified from a variety of potential impacts
resulting from any given inventory item. Assessment endpoints are
determined based on the goals and scope of the LCA. Groundwater
depletion, for instance, may be an assessment endpoint associated
with a quantity of groundwater used to manufacture of paper
products.
Classification
Conversion Models
Characterization
Direct Impact
Goal Definition and
Scoping
Impact
Impact Assessment
The process whereby inventory data are assigned to impact
categories (e.g., photochemical smog, lung disease, fossil-fuel
depletion) under primary impact groups (e.g., ecosystem, human
health, and natural resources). For example, CO2 emissions may be
classified into the greenhouse effect category under primary
ecosystem impacts group.
Models that help to characterize environmental impacts based on the
data obtained from an inventory analysis. An example of a
conversion model is the Mackay Unit World Model, which uses a
generic computer fate-and-exposure model to characterize the
partitioning and transformations of chemical substances introduced
into a hypothetical 1 km3 "ecosystem box."
The assessment and possible estimation of the magnitude of
environmental impact. Characterization involves the use of specific
impact assessment tools, known as conversion models and impact
descriptors.
A potential impact that is directly attributable to an inventory item.
A direct impact associated with ozone emissions could be photo-
chemical smog.
A discrete activity in the LCA process, which may be reevaluated or
modified at any point, that involves defining the study purpose and
objectives; identifying the product, process, or activity of interest;
identifying the intended end-use of study results; and key
assumptions employed.
A potential ecosystem, human health, or natural resource effect
associated with an inventory item. Acid deposition, for instance,
may be an impact to the natural environment associated with X tons
of SO2 emissions identified in the inventory analysis.
A quantitative and/or qualitative process to classify, characterize,
and value impacts to ecosystems, human health, and natural
resources based on the results of an inventory analysis.
C-l
-------
Impact Network
Impact Descriptor
Improvement
Assessment
Indirect Impact
Input
Inventory Analysis
Life-Cycle Assessment
(LCA)
The conceptual, qualitative linking of inventory items to potential
direct and indirect impacts. For instance, NOX emissions listed in
the inventory may be linked to acid precipitation, which in turn may
be linked to tree damage, acidification of lakes, soil leaching, and
corrosion of materials.
A measure or set of significant environmental attributes associated
with a particular impact or impact category. For example, a CO2
emissions value from an inventory could be run through the
appropriate conversion model to yield the potential level of
greenhouse gas build up or global warming.
A process to identify and evaluate opportunities for achieving
improvements in products and/or processes that result in reduced
environmental effects, based on the results of an inventory analysis
or impact assessment.
A potential impact that is not directly attributable to an inventory
item, but rather stems from another impact. Human respiratory
damage, [for instance, could be indirect impacts of photochemical
smog, which is a direct impact of ozone emissions.
A raw material, energy, or other resource requirement of a product
system, [inputs may include the amount of timber required to
produce 1 ton of paper, the amount of natural gas required per unit
of plastic production, or the amount of soil erosion per activity.
A process of identifying and quantifying, to the extent possible,
resourceland energy inputs, air emissions, waterborne effluents,
solid waste, and other inputs and outputs throughout the life cycle
of a product system. The inventory may include such items as the
tons of CJ:O2 released per unit of production, the amount of coal per
unit of production.
A holistic approach to evaluating the environmental burdens
associated with a product system by identifying inputs from and
outputs to the environment; assessing the potential impacts of those
inputs and outputs on the ecosystem, human health, and natural
resources; and identifying and evaluating opportunities for
achieving improvements. LCA consists of four complementary
componentsgoal definition and scoping, inventory analysis,
impact assessment, and improvement assessment.
C-2
-------
Measurement Endpoint
Nonthreshold
Assumption
Output
Production System
Valuation
A measurable response to an environmental loading that may act as
a surrogate measure, quantitative or qualitative, for a related
assessment endpoint. For example, acid precipitation could be a
possible measurement endpoint for the assessment endpoint of lost
recreation revenue at lake X that is indirectly attributable to NOX
emissions. A different measurement endpoint for this scenario
could be the lost recreation revenue at lake Y due to NOX emissions.
The concept that although recognizing any single inventory item
within a given product system as a significant contributor to specific
impacts is difficult, that inventory item nonetheless contributes to
impacts when placed in the context of other product systems, and
may therefore need to be considered in impact assessment.
Air emissions, waterborne effluents, solid waste, or other releases to
the environment associated with the life cycle of a given product
system. Outputs can include the quantity of CO2 released per unit of
production, the volume of solid waste per unit of time, and the level
of noise or odor associated with a particular activity.
An operation or group of operations associated with the production
of a product or service that has clearly delineated input and output
boundaries and includes operations associated with each life-cycle
stage. The product system associated with polyethylene production,
for instance, includes not only the company manufacturing the
polyethylene, but all of the intermediate companies that produce the
materials for the polyethylene production, such as the oil refinery
and a natural gas transportation company.
The explicit and collective process of assigning relative values
and/or weights to potential impacts of concern (assessment
endpoints). Analytic methods, for example, such as the Analytic
Hierarchy Process (AHP) may be used to estimate the relative
importance (value) of various impacts or impact categories to
multiattribute decisions.
C-3
-------
-------
Appendix D
Bibliography
-------
_
-------
Ames, Bruce N., Renae Magaw, and Lois Swirsky Gold. 1987. "Ranking Possible
Carcinogenic Hazards." Science 236:211-219.
Argonne National Laboratory, Environmental Assessment and Information Sci
1991 Assessing the Risk of Chronic Lung Injury Attributable to Long-Term Ozone
Exposure. ANL/EAIS-2. Argonne, IL: Argonne National Laboratory.
Ascher, William, and Robert Healy. 1990. Natural Resource Policymaking in Developing
'Countries. Durham, NC: Duke University Press.
Ayres, Robert U. 1989. "Industrial Metabolism." In Technology and Environment, pp. 23-49.
' Washington, DC: National Academy Press.
Bailey, Paul E. 1990-91. "Life-Cycle Costing and Pollution Prevention." Pollution Prevention
Review Winter:27-39.
Barnthouse, L.W. 1991. The Implications of Risk Assessment for Product Life Cycle Analysis.
Oak Ridge, TN: Oak Ridge National Laboratory.
Barnthouse L W G W Suter II, C.R. Baes HI, S.M. Bartell, M.G. Cavendish, R.H. Gardner
B^th^'^^'^dAER'osen .1985a .Environmental Risk Analysis for Indirect Coal
Liquefaction. ORNI/TM-9120. Oak Ridge, TN: Oak Ridge National Laboratory.
Barnthouse L W G W. Suter n, C.R. Baes III, S.M. Bartell, R.H. Gardner, R.W. MUlemann,
Barnthouse, L.W., O. W ^ ^^ ^ ^^ ^ ^ ^ ^ ^^ m5b Umt
Release Risk Analysis for Environmental Contaminants for Potential Concern in
Synthetic Fuels Technologies. ORM/TM-9070. Oak Ridge, TN: Oak Ridge National
Laboratory.
Beanlands, Gordon E., and Peter N. Duinker. 1983. An Ecological Framework for
Environmental Impact Assessment in Canada. Halifax, Nova Scotia: Dalhousie
University.
Berube, M., and S. Bisson. 1991. Lifecycle Studies: A Literature Review and Critical Analysis.
Quebec, Canada: Ministere De L'Environment Du Quebec.
Bober Thomas W Norman D. Bornstein, Daniel S. Dixler, Dale K. Humbert, James B Knaak,
'"A !M^WS, DavidP.Richardson, and Frank Vincent. 1993. "Proto^^
Identification of Health and Environmental Effects of Packaging Materials in Municipal
Solid Wastes." New York: Academic Press, Inc.
Boustead, Ian. 1992. "The Relevance of Reuse and Recycling Activities for the LCA Profile of
Products." U.K.: The Open University.
D-l
-------
Bunn, Derek W. 1984. Applied Decision Analysis. New York: McGraw-Hill Book Company.
Canadian Standards Association. 1992. Environmental Life Cycle Assessment. Draft report.
Ontario (Toronto), Canada: Canadian Standards Association.
Cascone, Ronald W. 1992. "Life-pycle Assessment on PVC Packaging Systems." Paper
presented at the Society of Environmental Toxicology and Chemistry Life Cycle Impact
Assessment Workshop, Saridestin, FL, February 1-7.
Colby, M.E. 1991. "Environmental Management in Development: The Evolution of
Paradigms." Ecological Economics 3:193-213.
.
Council on Environmental Quality^ 1989. Risk Analysis: A Guide to Principles and Methods
for Analyzing Health and Environmental Risks, eds. John J. Cohrssen, and Vincent T.
Covello. pps. 8-25.
l
Council for Solid Waste Solutions, > June 1990. Resource and Environmental Profile Analysis of
Polyethylene and Unbleached Paper Grocery Sacks. Final report prepared by Franklin
Associates, Ltd., Prairie Village, KS.
Cox, David C, and Paul Baybutt. 1981. "Methods for Uncertainty Analysis: A Comparative
Survey." Society for Risk Analysis 1(4):251-258.
Denison, Richard A. 1991. "Whither Impact Analysis?" Washington, DC: Environmental
Defense Fund.
Denison, Richard A. 1992a. "Relationship of Social Welfare Impact Category to
Environmental Categories." j Paper presented as part of the Society of Environmental
Toxicology and Chemistry Life-Cycle Impact Assessment Workshop, Sandestin, FL,
February 1-7. ;/
Denison, Richard A. 1992b. "Toward a Code of Ethical Conduct for Lifecycle Assessments."
Paper presented at the Institute of Packaging Professionals Life-Cycle Analysis
Conference. Alexandria, V/L September 30 to October 2.
Desvousges, William H., F. Reed Jojinson, Josephine A. Mauskopf, Stephen A-Johrkoin, Joel S.
Smith, K. Nicole Smith, K. rjlicole Wilson, and Maria G. Benerofe. 199,1. Accounting
for Externality Costs in Electric Utility Planning in Wisconsin. Final report prepared for?
the Task Force on Externality Costing under RTI Project Number 35U-5198-01.
Desvousges, William H., Richard W. Dunford, and Jean L. Domanico. 1989. Measuring
Natural Resource Damages:
An Economic-Appraisal. Final report prepared for the
American Petroleum Institute, Washington, DC under RTI Project Number 35U-3981.
D-2,
-------
Ekvall Tomas, Henrikke Baumann, Goran Svensson, Toman Rydberg, and Anne-Marie
TUlman. 1992a. "Life-Cycle Assessment: Pilot Study on Inventory Methodology and
Data Bases." Sweden: Gothenburg (Chalmers).
Ekvall Tomas 1992b. "Life-Cycle Analyses of Corrugated Cardboard: A Comparative
'Analysis of Two Existing Studies." Draft working paper. Sweden: Gothenburg
(Chalmers).
Environ. 1988. Elements of Toxicology and Chemical Risk Assessment. Washington, DC:
Environ Corporation.
Evers David P., and Bruce W. Vigon. 1992. Resource Depletion Analysis in Life Cycle Impact
Analysis. Draft Final Report. Prepared for the U.S. Environmental Protection Agency,
Office of Air Quality Planning and Standards. Columbus, Ohio: Battelle.
Fava, James A., Richard Denison, Tim Mohin, and Rod Parrish. 1992- "Life Cycle Assessment
Interim Peer-Review Framework." Paper prepared for the SETAC LCA Advisory
Group.
Federal Register. 1988. Hazard Ranking System (MRS) for Uncontrolled Hazardous Substance
Releases. 53(247), Friday, December 23, 1988.
Franklin Associates, Ltd. 1989. Comparative Energy and Environmental Impacts for Soft Drink
Delivery Systems. Prairie Village, KS: Franklin Associates, Ltd.
Franklin Associates, Ltd. 1991. Resource and Environmental Profile Analysis of High-Density
Polyethylene and Bleached Paperboard Gable Milk Containers. Paper prepared for the
Council for Solid Waste Solutions. Prairie Village, KS: Franklin Associates, Ltd.
Freeman, A.M. 1982. Air and Water Pollution Control: A Benefit-Cost Assessment. Wiley,
New York.
Funtowicz, Silvo O., and Jerome R. Ravetz. 1990. Uncertainty and Quality in Science for
Policy. The Netherlands: Kluwer Academic Publishers.
Graedel, Thomas E., and Paul J. Crutzen. 1990. "The Changing Atmosphere " In
Planet Earth: Readings from Scientific American Magazine, pp. 13-24. New York:
W.H. Freeman and Company.
Greenland, David. 1983. Guidelines for Modern Resource Management. Columbus, Ohio:
Charles E. Merril Publishing Company.
D-3
-------
Grimstead, Brad, Stefan Schaltegger^ Christopher H. Stinson, and Chris Waldron. 1993.
"Methodology for Assessing the Environmental Impacts of Chemical Emissions in the
United States." Kirkland, ^VA: Pioneer Environmental Consulting.
Grodsky, Jamie A. 1993. "Certified Green: The Law and Future of Environmental Labeling."
The Yale Journal on Regulation 10:147.
Guinee, Jeroen B., and Reinout Hefejungs. 1993. "A Proposal for the Classification of Toxic
Substances within the Framework of Life-Cycle Assessment of Products." Chemosphere\
26(10): 1,925-1,944.
Guinee, Jeroen B. 1992a. "Headirigs for the Impact Analysis." SETAC European Workshop or
Environmental Life Cycle j*oialysis of Products. Leiden, Sweden: Centre of
Environmental Science, Leijden University.
I
Guinee, Jeroen B. 1992b. LCA Classification. Draft report. Leiden Sweden: Center of
Environmental Science, Leiden University.
Hill, A.B. 1965. "The Environment and Disease: Association or Causation?" Proceedings of
the Royal Society of Medicine 58:295-300.
i
Hunt, Robert G., and William E. Franklin. 1975. "Resource and Environmental Profile
Analysis of Beer Containers." Chemtech August:474-481.
I
Hunt, Robert G., Jere D. Sellers, arjd William E. Franklin. 1992. Resource and Environmental
Profile Analysis: A Life Cycle Environmental Assessment for Products and Procedures.
Prairie Village, KS: Franklin Associates, Ltd.
Husseini, Ahmad. 1992. "Canada: From Framework to Standard For Life Cycle Assessment."
Paper presented at the Society of Environmental Toxicology and Chemistry Life-Cycle
Impact Assessment Workshbp, Sandestin, FL, February 1-7.
i
Jain, R.K., L.V. Urban, G.S. Stacey, and H.E. Balbach. 1993. Environmental Assessment. New
York: McGraw-Hill, Inc. j
Keeney, Ralph L., and Howard Raifra. 1976. Decisions with Multiple Objectives: Preferences
and Value Trade Offs. New! York: John Wiley & Sons, Inc. ;
Kuik, Onno, and Harmen Verbruggjsn. 1991. In Search of Indicators of Sustainable, ;
Development. The Netherlands: Kluwer Academic Publishers. ; , I-
D-4
-------
^^^^
tfTSde Impact Assessment Workshop, Sandestin, PL, February 1-7.
Linstone Harold A., and Murray Turoff. 1975. The Delphi Method: Techniques and
AppLtions. Reading MA: Addison-Wesley Pubhshmg Company.
^*^ Aworking paper- International
Institute for Applied Systems Analysis, Laxenburg, Austria.
Mackay Donald. 1979. "Finding Fugacity Feasible." In Environmental Science and
Technology 13: 1,218-1,223.
Mackay, Donald, and Sally Paterson. 1981. "Calculating Fugacity." Environmental Science
and Technology 15:1,006-1,014.
Mekel, O.C.L., and G. Huppes. 1990. Environmental Effects of Different Packaging Systems
'for Fresh Milk. Printed in the Netherlands.
Mitchell RC
Carson. 1991. An Experiment in Determining Willingness to Pay for
Quality Improvement. Draft Report to the U.S. Environmental
Protection Agengy, Washington, D.C.
Measurement Tool." Total Quality Environmental Management Autumn.
National Research Council. 1983. Risk Assessment in the Federal Government: Managing the
Process. National Academy Press, Washignton, D.C.
Navinchandra,D. 1991. "Design for Environmentability." Technical report in American
Society of Mechanical Engineers, Vol. 31.
Organization for Economic Co-operation and Development (OECD) 1989. Environmental
g PoTcy Benefits: Monetary Valuation. Paris: OECD Publications Service.
Ott,WayneR. 1987. Environmental Indices: Theory and Practice. Ann Arbor, MI: Ann
Arbor Science Publishers, Inc.
D-5
-------
Phaneuf, Yves. 1990. "EIS Process and Decision Making." A background paper prepared for
the Canadian Environmental Assessment Research Council. Quebec (Hull), Canada.
i
Plewa, et al. 1986. Assessing a Degree of Hazard Ranking To Illinois Waste Streams. \
Hazardous Waste Research juid Information Center, No. RR-013. Illinois Department of
Energy and Natural Resources.
Plewa, et al. 1988. Refining the Degree of Hazard Ranking Methodology for Illinois Industrial
Waste Streams. Hazardous Waste Research and Information Center, No. RR-029.
Illinois Department of Energy and Natural Resources.
Proctor and Gamble Company. 1992. "Considerations for Impact Analysis in Life Cycle
Studies." Paper presented at! the Society of Environmental Toxicology and Chemistry
Life-Cycle Impact Assessment Workshop, Sandestin, FL, February 1-7.
Ream, Timothy J. 1992. "Life Cycle Assessment: A Tool for Addressing Environmental
Problems." Paper presented iat the 85th Annual Air and Waste Management Association
Meeting and Exhibition, Kansas City, MO, June 21-26.
Reckhow,K.H. 1991. "Decision Theory Applied to Product Life Cycle Assessment." Paper
presented at the Society of Environmental Toxicology and Chemistry Life-Cycle Impact
Assessment Workshop, Sandpstin, FL, February 1-7.
I
Reddy,K,R. 1985. Special Waste tategorization Study: Volume 1. Hazardous Waste
Research and Information Center, No. RR-005. Illinois Department of Enerev and
Natural Resources. '
Ruckelshaus, William D. 1989. "Toward a Sustainable World." In Managing Planet Earth
Readings from Scientific American. New York: W.H. Freeman and Company.
Saaty, Thomas L. 1992. "How to Make Environmental Decisions." Paper presented at the
Institute of Packaging Professionals Life Cycle Analysis Conference, Alexandria VA
September 30 to October 2.
Sandia National Laboratories. 1993. "Environmental Conscious Manufacturing Life Cycle
Analysis." Draft report prepared by Randall D. Watkins and Adra Baca for the
Integrated Manufacturing and Designing Initiative program. Albuquerque, NM: Sandia
National Laboratories.
I
Sandia National Laboratories. 1986. \Risk Assessment and Ranking Methodologies for
Hazardous Chemical Defense^Waste: A State-of-the-Art Review and Evaluation Final
report, SAND86-0530. Albuquerque, NM: Sandia National Laboratories * ' «
D-6
-------
Schaltegger, Stefen C, and Andreas J. Sturm. 1993. "Eco-Controlling: An Integrated
Economic-Ecological Management Tool." In Green Business Opportunities. Financial
Times, Pitman Publishing.
Schaltegger, Stefen C. 1993. "Strategic Management and Measurement of Corporate
PollutionEcological Accounting: A Strategic Approach for Environmental
Assessment." Discussion Paper #183. Strategic Management Research Center,
University of Minnesota, Minneapolis, MN.
Schneider, Stephen H. 1990. "The Changing Climate." In Managing Planet Earth: Readings
from Scientific American Magazine, pp. 25-36. New York: W.H. Freeman and
Company.
Scientific Certification Systems, Inc. 1992. Life Cycle Inventory and the Environmental Report
Card. Oakland, CA: Scientific Certification Systems, Inc.
SETAC Foundation for Environmental Education, Inc. 1993. A Conceptual Framework for
Life-Cycle Impact Assessment. Pensacola, FL: Society of Environmental Toxicology
and Chemistry.
SETAC Foundation for Environmental Education, Inc. 1992. Guidelines for Conduct ofLCA
Peer Review. Draft paper. Pensacola, FL: Society of Environmental Toxicology and
Chemistry.
SETAC Foundation for Environmental Education, Inc. 1991. A Technical Framework for Life-
Cycle Assessments. Pensacola, FL: Society of Environmental Toxicology and
Chemistry.
Source Reduction Research Partnership. 1991. Source Reduction and Recycling of Halogenated
Solvents: Lifecycle Inventory and Tradeoff Analysis. Report prepared for the Source
Metropolitan Water District of Souther California and the Environmental Defense Fund.
Prepared by Jocabs Engineering Group, Inc. Pasadena, CA.
Stanley, G.B. et al. 1979. "Radionuclide Migration from Low-Level Waste: A Generic
Overview." Technical report in Management of Low-Level Radioactive Waste, eds.
M.W. Carter et al. Vol.2.
Sullivan, Michael S., and John R. Ehrenfeld. 1992. "Reducing Life-Cycle Environmental
Impacts: An Industry Survey of Emerging Tools and Programs." Total Quality
Environmental Management, Winter.
Suter, G.W. II, L.W. Barnthouse, C.R. Baes m, S.M. Bartell, M.G. Cavendish, R.H. Gardner,
R.V. O'Neill, and A.E. Rosen. 1984. Environmental Risk Analysis for Indirect Coal
Liquefaction. ORNL/TM-9074. Oak Ridge, TN: Oak Ridge National Laboratory.
D-7
-------
Suter, G.W. n, L.W. Barnthouse, SrR. Kraemer, M.E. Grismer, D.S. Durnford, D.B.
McWhorter, F.R. O'Donnelj, C.R. Baes IE, and A.E. Rosen. 1985. Environmental Risk
Analysis for Oil from Shale. ORNL/TM-9074. Oak Ridge, TN: Oak Ridge National
Laboratory. I
Swedish Environmental Research Institute. 1993. Recent Developments in Methodologies for
Impact Assessment in the Context of Life Cycle Assessment. Paper presented at the 14th
annual meeting of SETAC iii Houston, TX, November 14-18,1993.
[
Swedish Environmental Research Institute. 1992. The EPS Enviro-Accounting Method.
Technical report prepared for the Swedish Waste Research Council, Goteborg Sweden.
Swedish Environmental Research Institute. 1991. "A System for Calculating Environmental
Impact." Technical report prepared for the Swedish Waste Research Council, Goteborg
Sweden.
i
Tellus Institute. 1992a. CSG/Telluk Packaging Study, Report #5: Executive Summary. Report
prepared for the Council of State Governments, Lexington, KY, and U.S. Environmental
Protection Agency. Boston,'MA: Tellus Institute.
Tellus Institute. 1992b. CSG/Telluls Packaging Study, Volume I. Report prepared for the
Council of State Governments, Lexington, KY, and U.S. Environmental Protection
Agency. Boston, MA: Tellus Institute.
i
Tellus Institute. 1992c. CSGITellus Packaging Study, Volume II. Report prepared for the
Council of State Governmen!ts, Lexington, KY, and U.S. Environmental Protection
Agency. Boston, MA: Tellps Institute.
i
Thomas, David L., and Gary D. Miller. 1992. "Using Existing Hazardous Waste Databases:
Limitations and Future Needs." Hazardous Wastes and Hazardous Materials 9(1).
Pages 97-111.
U.S. Congress. 1992. Green Products By Design: Choices for a Cleaner Environment. Office
of Technology Assessment, OTA-E-541. Washington, DC: U.S. Government Printing
Office.
U.S. Environmental Protection Agency. 1994a. Life-Cycle Assessment: Public Sources of Data
for the LCA Practitioner. Final report. EPA, Office of Solid Waste, Washington, B.C.
Prepared by Battelle, Columbus, OH.
U.S. Environmental Protection Agency. 1994b. Guidelines for Assessing the Quality of Life
Inventory Data. Final report. EPA, Office of Air Quality Planning and Standards,
Research Triangle Park, NC.
Park, NC.
Prepared by Research Triangle Institute,- Research Triangle
D-8
-------
US Environmental Protection Agency. 1993a. Life Cycle Design Manual: Environmental
Requirements and the Product System. EPA/600/R-92/226. Prepared by the National
Pollution Prevention Center, University of Michigan, Ann Arbor, ML
US Environmental Protection Agency.' 1993b. Life-Cycle Assessment: Inventory Guidelines
and Principles. EPA/600/R-92/245. Prepared by Battelle Institute, Columbus, OH, and
Franklin Associates Ltd., Prairie Village, KS.
US Environmental Protection Agency, Risk Assessment Forum. 1992a. Framework for
Ecological Risk Assessment. EPA/630/R-92/001. Washington, DC.
U.S. Environmental Protection Agency, Risk Assessment Forum. 1992b. Report on the
Ecological Risk Assessment Guidelines Strategic Planning Workshop.
EPA/630/R-92/002. Washington, DC.
U S Environmental Protection Agency. 1992c. Resource Depletion Analysis in Life Cycle
Impact Assessment. EPA Contract No. 68-CO-003. Prepared by Battelle Institute,
Columbus, OH.
U S Environmental Protection Agency. 1992d. Summary of Selected New Information on
Effects of Ozone on Health and Vegetation. EPA/600/8-88/105F. Washington DC.
U S. Environmental Protection Agency, Risk Assessment Forum. 1991. Summary Report on
Issues in Ecological Risk Assessment. EPA/625/3-91/018. Washington, DC.
U S Environmental Protection Agency, Office of Policy, Planning, and Evaluation. 1990.
Hazardous Substances in Our Environment. EPA 230/09/90/81. Prepared by Research
Triangle Institute, Research Triangle Park, NC.
U.S. Environmental Protection Agency. 1987. The Risk Assessment Guidelines of 1986. Final
Report. EPA/600/8-87/045. Washington, DC.
U S Environmental Protection Agency. 1985. Methodology for Characterization in Exposure
Assessments. Final Report. EPA/600/8-85/099. Prepared by Research Triangle
Institute, Research Triangle Park, NC.
U.S. Environmental Protection Agency, Office of Health and Environmental Assessment. 1984.
Selected Approaches to Risk Assessment for Multiple Chemical Exposures. Final Report.
EPA/600/9-84-014a. Cincinnati, Ohio.
Udo de Haes, H.A. "A General Framework for Environmental Life-Cycle Assessment of
Products." SETAC European Workshop on Environmental Life Cycle Analysis of
Products. Centre of Environmental Science, Leiden University, Leiden, Sweden.
D-9
-------
Udo de Haes, H.A., Maarten Nip, and Frans Klijn. 1991. "Towards Sustainability: Indicators
of Environmental Quality." Technical report in In Search of Indicators of Sustainable
Development. Editors Onno Kuik and Harmen Verbruggen, 1991. Boston MA: Kluwer
Academic Publishers. !
j
Vigon, B.W., and Allan A. Jensen. 1992. "Lifecycle Assessment Data Quality and Databases
Practitioner Survey." Discussion initiation paper presented at the Society of
Environmental Toxicology and Chemistry Lifecycle Assessment Data Quality
Workshop, Wintergreen, Virginia, October 4-9.
Watkins, Randall D. 1993. IMDI ECM Life Cycle Analysis Part I: Environmental Impact
Metrics Definition, Stakeholder Survey and Panel Review. Draft report. Sandia National
Laboratories, Albuquerque,
White, Allen, Deborah Savage, and |Karen Shapiro. Forthcoming. "Life-Cycle Costing:
Concepts and Applications." In Life-Cycle Assessment, ed. Mary Ann Curran. New
York: McGraw-Hill, Forthcoming.
j
i
Wolf, C.P. 1992. "Impact Assessment and Life Cycle Assessment." Paper presented at the
Society of Environmental Toxicology and Chemistry Life-Cycle Impact Assessment
Workshop, Sandestin, FL, February 1-7. New York, NY: Social Impact Assessment
Center. j
t
Wolf, C.P. 1988. "A Systems Approach to Impact Assessment." Paper presented at the
Proceedings of the International Workshop on Impact Assessment for International
Development. British Columbia (Vancouver), Canada. New York, NY: Social Impact
Assessment Center.
World Commission on Environment and Development. 1987. Our Common Future. Oxford:
Oxford University Press.
i
World Wildlife Fund. 1991. Gettirig at the Source: Strategies for Reducing Municipal Solid
Waste. Island Press: Washington, DC.
D-10,
-------
TECHNICAL REPORT DATA .
(Please read Instructions on the reverse before completing)
T-
1. REPORT NO.
EPA-452/R-95-002
4. TITLE AND SUBTITLE
Life-Cycle Impact Assessment:
A Conceptual Framework, Key Issues
of Existing Methods
and Summary
r. AUTHOR(S)
I. PERFORMING ORGANIZATION NAME AND ADDRESS
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
15. SUPPLEMENTARY NOTES
RECIPIENT'S ACCESSION NO.
. REPORT DATE
July 1995
. PERFORMING ORGANIZATION CODE
. PERFORMING ORGANIZATION REPORT NO.
0. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NU.
68-D2-0065
13. TYPE OF REPORT AND PERIOD COVERED I
14. SPONSORING AGENCY CODE
EPA/200/04
16. ABSTRACT i j.
Life-Cycle Assessment (LCA) is a holistic concept and approach for evaluating
the environmental and human health impacts associated with a product, process, or
activity. A complete LCA looks upstream and downstream, identifies inputs and outputs,
and assesses the potential effects of those inputs and outputs on ecosystems, human
health, and natural resources.
This report presents a conceptual framework for conducting a life-cycle impact
assessment (LCIA), discusses major issues, and summarizes existing methods. It also
identifies some of the advantages, and disadvantages of various methods.
KEY WORDS AND DOCUMENT ANALYSIS
COSATl Field/Group
b.lDENTIFIERS/OPEN ENDED TERMS
DESCRIPTORS
19 SECURITY CLASS (ThisReport)
18 DISTRIBUTION STATEMENT
20. SECURITY CLASS (Thispage)
EPA Form22a(WB
PREV.OUS ED.TION is OBSOLETE
-------
------- |