TS 17J0 

155 

1995 



^ <.. 


I a 


<Ja. <fn . ^ 




V' ’ '4^^^s° ‘ “ * 

‘ o^^dk' . a&: "bv-*^ : 


'^O-^ 

■> *^Oys ^ 

“S™Rr "a»" ^^r.c 5 ^ 


n 


<^ON C 



^ V V. 

•;v'“ 0°'%''"“*-"''^^ 

J" O •i' 


c^'^S' 


^ ^ s 


■\'° * ''‘l‘i>^ ^°’‘''« ^<C'‘ ° ’ ‘ V'^^” ""■>\'° ' ‘V'^ c0>-c,f<f. 


■or 

-"•v - • • .'*V ^ * 

o ^ -»r 

i’ ° 

<> '<f> <1 » LI - ^O <» X 

cPV^%"^o 

o . ~ !• ^ 


vV<^ 











. 9 





o 
z 

dV o 

'^/«MW- "V '>^J. 'J 

»p-^, ” 





^ V 

''O. ^^-i'* .'N 

,o'.‘■"'^'''o- > 

%, <i^‘- 

i. ^ 


<r>. 



•0 » ^ 


/‘O^ 

AQ, 


^ V . \ 



V'” v<^ - A;'““ VVa-!V‘- ' v<;;V ^ 

r c^.r- Z. J _ c,.r. Z 



'■* H-i >>"'11 



‘o '^-S' ' 

Z V-V « 

'SI®* <^’'’A, °oOT^7 .'?>'\. •^#” I 

■o' «> <lA A 'V •<5' ■>> -V -QA <K * 

>.V * ’ ' * o°l’ "*' 

' Ad^ w A^a''. 

A®-*’ ’?■>'- 

Oq .^W.” ^o 



’ A » * • ^ \, *=>“°’ X' « o/V"'’ '\>'^ ' • • ^ \ 

y ,r 3ii& ^ V'<^ «SR'' 

•r ^'1 ^< 5 ‘ V#* /-O y^ t ^-\ .-v^ a*^ V 

, *•‘4 <i%sS5^a. <Sj. ^ ^ Mu/Z^ •#>'-' x'‘< 


t- '^J'S^ - z VV « 

J“ / 0 ^^ v>v ■* * -v --^ 

V-'' 

» _r<SB ^.x Mx. 

"bv*" 

■'’ ^ >> * -Z 

[. ■'•oko'’ A ,.„ '*<>*•»,1 Hti ,,, 

: w ^Ad^ "bv^ »'^v>ix. 




'^1 o 'I 

-v ,» 'er's 

v c?i^%'"°o“': 



cTSiR^'. Xrcfs 



N?' 

Ac?” k "'o i''* 'T'^ 

■ : 


- 

;s^-:,..,;V*-o;/..x V* 



A> V Vj o 


'U % 


w aT^ ^ 

” o''^ ° 

r z V-S^ « 

" - cIs \W/W^ 


V ° 

« Lia . 9 jA ^O ^ i. 
/-O v-t- O-A 

# O ■?' Oeft/Z^ -t 

\o^ 


'^ov 

p^X 



<t 

O 

r K : 

'^Ly *■ '-Zy//l\iSlS' V 

X ‘^Xt t- K.^ «■ '^- « 

o ^xy - * M'xxa«^^ 



- ™ . o V 

A'- cc-..y • •' >A' ;AAfV° ” ‘ y'\A—V 5.^ 

w O -i- ^'fTZxhZ’p <i csSJAvs* 

' kd^ "bv^ °‘^^" ''^■'•o 

-y e o %x ^ 

> >f o B ■’' Oa 

r> ff n » 




-—x'rA %'^y 
•a..5'A ■5. 


c^y^yn^ V K^ ^ ^ ^ ^ ^it* ^ ‘ 

'1'' ’ '14^1-?? ‘ “" V-% A>'S' 

o'^fe” \<A : &Jh \ ' 




“ V *. 

V'V’ W ^ 


s * * 



.y;'’' A ' • “ >11;^^*^^°'°''"^^-^°”°' 



r^\ * 




























« LI A 


Of 


C°o ” '>\!r ''" V't'lZ'’ -o 

•i- <J^ Ck _0 * 

O. V a * •» .Q^^ O ■ 


^ .cy 


^o „ K'^ ^ ^ <y. « 1» ^ \a^ 

" \i^ ^-f fO^ 

sC^ .« -to - 


<,* ^ O N o”^ y * r. 

« fife * '^iPr.rJ^^ ■» O 


'to ^ ^ ^ yf" 

❖ o_^ O^ O ''■ 

^ ^ O ^ 


- ^ .<. -t c 







^»./V ' ’ '' 'K' 



' 4’°^ ’ 

\<^ , . . , %, *="*°’n'5- »■■ « o,V/* n ■> 

:!^mi \.y :wmz ’K'C 





o •?• 


9 *■' 





■*'” ‘ Vt‘• “ " o^"/'” " Vt‘”• A *°'“'%'S'^.’ 


« O 

^ ^ tj 







A. O ^ r* .C 

= .* 

O o 


■' r-y A. 

. 


* AO 


C 


o r aV A' - '^^Sil^’ n n r A A “* ^^liiiiif^ n /■‘b'Tr. ^ 



. ^ A\ %W/ A'\ IW/ A A • 


Rd ^ rO^> 

tf_^ - 


^4. ^O ^ 


ea^ Lli^ " - ' « '--fiJ Btp- - - rg pg -• - . fl v«^ I ^rx * 

* A I AA / Jii; 


-oNg, 


*" ^lii^ o AJn o r ■" 

'BIP’ ^ V V, ^ 

<{» ^ ‘•'*^'^^ 0 ''° ’ "‘‘1^ CONC^ 

u> ■ iv ^/*0 o 9 ^V ® v*0 

“"M* fM 

»WW.* A A «,Wt 


^^W/\i^ v A' ^VAA ^ V ‘ <i*. 

^ ,jA r> j- o 

^ ^ "VN^ , s - ° ® 

,r fife <• 

vV^ 



/s>>'" -s^ 

'•^/Oo > ""d 

: ®bv^ 

° t 



\ o^ X^p^yy’ 

. 0 ^ «-'* * Op 

P -<0 ^ .A . '* . (5 51 . O 


^ ^ Ak O ^ ^ 

'a"^A^\o^ 'U Ao v-^*' a 

> ^ ^ s ^ -Lifi. 4a ° A 

vO O ^ 0«, -A^ n ” '<■ ‘■<V /-V" » 

O ■> OeftTZ^P O A V/- O A 

-fyo^ .°'^^- '*'■^0^ 

,'t^ CA yO 


Sr A 
Sr<A 


o'•^',Vn<^^‘•^^t AA oS^' 


O 




^ ftft % y 


> ^<0 ^ 

^ A ^ p 

'o ,t h'^ A 


<^°p 


p A <i W. o -^ ' ''^ 

^ «>AA - ■AAA^'^ y. 7>^ 


"bv 







o ® r ^mmmZ « S®® t 

o c.S'To, o “ <v;^. ^ A<r,. * - 


<■ yO 


O ^ 

O V A O 


- \o^ Ao^ :S^J- <^r.^ 

. '.^#!° n^°A, ' 



y-... -- - 

-i \ A /Jfe; 

cPSA^^-^o ' \. P°'S' —'° 

: .0^ 


r 

» x ' <1 ■ iSgtri !“^ '’ vy » HTSS; * O V „ “\Jxyw 

o ijPtO' »• <A^ * O ^Pt)* >• * -cAPo 


W G 


^ y '^■ 

* As 'j 


ft ft s 


M 



y^\ \^^S A'\ \^^->.* '^% <y A "'’■^ 

'A®o° ' ' ° " ‘A'^.:r * A/' ’' ‘>!s yy 

a" '>-r>^d' o,^»' >>„<< . (^^- oS^Sk' 









's ^ ■' 

■« 

« 


I?; 


> V^. - 

r k 

* * 




i 


’{ , 

•r* 




f 




United States 
Environmental Protection 
Agency 


Office of Air Quality 
Planning and Standards 
Research Triangle Park, NC 27711 


EPA-452/R-95-002 
July 1995 


Air 



A Conceptual Framework, 
Key Issues, and Summary 
of Existing Methods 



























6 


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. 



4 





’ ^ ‘ V r . f 


*1. »i 4n 


-t,’ 




!^:kx 



» i 






iv*’?; Vi'm.. 

1-' " '■ ' "V • iry .-**7 " •' 



f .. 


«4t«fA.r-^rn 


1 

% 


<r» P, ; ^ ill ■*#N\'»''* :'f /cd f 




i .•<-.4':,.(r> I*", 




■\ 


,\ .j^^;, ', •, r .f'H .*.; ^ •*; ’ -v. -t .K^-. ■•f |i jir- iirt^., ^3 

i»0^ rr^vn’i^ii ■-> ^' i r>t ev„^:*^'«*. uvk^ «•♦, ■’ t .f* 

; . I 

^ or^iV. 7 f»/!J«.'tiT,o • ><< .?rf.t» V Wr ;•> |... ;, ,, ^ 

.•'►i vtMIlr,-!* ^./7 U-Jt7...i' n»ot, 


jWV' 


»vfiiv':_ ,,u i^- *. >u^i> ^ •. • ,'i 



i. ''*> 



31 


■/ 



• — ■ ***W' J s. 





i ■*■ 




i !*• ■■■ 


A.‘ ^ 




'1 


•l^- 















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, SET AC 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 II included documenting existing methods that exhibit potential for 
application in impact assessment and identifying gaps in the impact assessment methodology. 
This document contains the combined output of Phases I and IT 


111 


p 


I i 


3ftJ h Of W *44jti Hii'^ (ne / F .») :., yC(fn>t Uttt 

1o i;fkh«iv '^'4/ i \<^} AO-I jjrrifin Ui ‘ii)t)i/j» m 'jr • -vixii ‘It. 

r*’ ■ - •• 

^ ’^rijr btUi .X^<fllKV ■*<* £>fM i r :/Hi<»J*oO {mii tfdTfSJifTi 

^ }i} tj** | "**i' :• ai.'i.si«f“:tCtti tjof i-n^ 

4 AO i hi^r:if- v: *..i-vfvw q» i >. m, i r.AXWs ti 

'f j 'N * ■» > * 

l.'i ^ rtJiatf i .J' ^H|in iiix#HiMQf.i ?F.H/OUll>i ,■; it n^FA 

cA^ g'rjiiaJtitfo^ ;'•’ ■ ^ j > Jm> titw^ 


tt yti ,.jfutr JLUf . i. If-<> 4 ^ 151 / ^ .H II ufT 


i 


*:jmo i/i .|,v. . ax>«|:rj 01 mai^oiq A':> 1 x-mof>li^m i 

10 xillO wi^ rrwr^: -r/^i Jo iejj^krr^ . 1 ?.i<fr .Wful^t 7 /u§ i 

ebT^Oi .r^ la- Vi. ^HtoiO kA W art! ,(f 3 ^ 0 > «Pflv>F 3 V!>a Uu ^ 

fufc nf.( ;jjv• M Uo* ,«V/l^ n 'vfj .t^TOAO) 

' 'Ti 4t*»jh.-W^,tW/>^ t !tfi;T IUa4|t<>1f^ / 3-1^1 .r**'fitl) oUl mjT 

%• 'irtt rt^'f^%,tr) h:Vf .^^otO^ MoV <e ; 7 :v 4 U. .1 i-j. 

^o ;<tvorf< jtfU v.l bitd ^i.\ 1 'Ft^ >o , a '•* i- .. 4 

JJ »i rw iir .'I? a 


. JiT 


0 

’Jt ■ .‘*A -.' 4 ’ w.'‘ n.\A hstii , tA^u f'uuVuV^^'.'O 'ni>v / 1 \ «- n«> - >nK *' 

f V I vr*/^- <r+ t‘OT»..I>yi o^v ' -..i\ l^Pt* IV !W!AUiT,v‘A \nir,ijn:^-xv. v^ 

H .rinv^TiJivi) A’ . vtt -'t*!, v^|fvjw3 

*/ .mm A ‘ V. .1\> ijn-. ^^^^„iubVV KDd ^ 

ea^V'm ^^».v *ioki: rfy, , ^jjj.i^Viq AXI AS3 F;i4i/^ 

;wvi:*4 151),. i'.m ..r^.I.r iT.i ,.* t\nfarri tr. qH* pj2 lt,tm imasi oq 

eAJ \ .1$tiUi£)' .^Ai ii Ckjit »"a) Vj^-/ !)J t •mlr-aoti^ 


<'i»3 


*o*tt $ -'^r! iTj jin}i$p.^'io\M u ^ ti Atiw ,iewiwxib gifiT 

bnu h^OtJo^t J y i.d': V. 1^ 1 4ii5<iii-riU4- i^^Jq 

Xa. 'cm| i. ;iii, »>! w 4^ oilf f%§ v^g tnfliijijiib 

“tol 5 .♦jIj *•}}*■ |.i*»f, Lr'Oy-y.nT^ -tj U » rt^^- 

If* - »flrJ034t*^jcl^F#i Jt^iiCnu **d ur \}ftc lii* n •i.' 4* f^. * * i ni;. .: 

*' b‘«6 ' ‘v;w' ' if 4 >qrr > \^;i» ‘mcia tdi . <41: Wu. / • 


fit? 

I :T 



w 


^ -4. 


J 





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/RJREL), 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, and comments 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 




1 TT 


2Ti13MD03. t A 







fii ,{i;i JamisK.") ASJ Yiuui ,<«It;AO e V-" ? v.i ,v n ,ot 

.iTjrj w issmq M ,q>»otO wona-s^eA ;iae -r..' wo* • ju. n <40 io rfj«i.-n=I «h«,6 

JJ5 4isr.«0 rtfiA -fjiM •(<? tjabivorq «Jiwri.Ti Ni, v/n .ajiuhtyj ASH UnoiJ'bbA 
maafl Wft JV/Wl ,HTfW .'i-ftyj ,fv,t'0> . 1 ,,1 '^O' Istt;*:?' 4ewil-r^^(^^ jXi.^HWUlO) 

,9&/«!tjui5( ftoiiitlbtiA - ?.'TnA'> ftji-.v odRl tnir i.i* ,! 4 %)aC rfnw 

i-^w (wMb,) 1 r. ,r^,7 «I . r.^rtq «W»9,„,I,03 luiB .gW5„v», 

;w-Ol«^U*« OM)'.:?--’-* , . wt ■*w;.r.t -Jottwiail wun IxUaubnca 

trwv ,ten ifUJs baiii/on; ia-s-A .t tur^'i f,. .! .Bi;goJi:nti«u<g .U>»n-J«a *rieM 

suit tH bna ^nifH^octq 


Vi 



adijiwi i >«ia«iA -i'.Jif'J j'Bi.viah-xi.A |,^ t,oh«i(s,ii »->3w9iv»'.ireq 

.i*w|x- ,l wnar -.'iv V. ^o'ip il i»viif..;A rt rfoa .i!'Ti3H^Ze„oijK,nimO 





•■XUijrjDVj'-' 1VJ«:1 uosiiyijviiro'i;lf' V snst‘i .;i.«<.wBU»tJ>'U .,to.I.,-Msixw«Am;*u.- T 
«3#w.t ‘,t»p . B A-.H. I .y-.Atnakru .^omit.ial ^ft-(?itrtv.oJ ^VnO ^i«0 

tkn,iUj*... 'L-.- 'rJ Vyi*M ..jiri qnjia y(joa jrfT ^»oiO Ajs^i ;. 3 :ir .niwyisr -I yofl ,»vgH 
«iiol fr xn.«-,,U biwW j,ivr,f ^lotetn^D ■ 

ernitrA « ..-.••> .r^O .v^»M muX 

.•Ajiri^KTO tr^pi f II! Ziu-yii ' .turt .ttoilaJlort 'KJiiaMtiouviCno ir;ar(: n..<.- a W .lobiTineZ 

.T«l((cifi /w-it. i*»srtM tertlA ,%»l»vrt2 3/ta p-,)a!ij,-Ri ?Mkn .vJinAiiS naviA ' 
.ait«»i>nii .fou-*,;v; /ji'tsctisiM'ijnu-TV rijiainw jwiiKiofpoDsaJgol.jnit^^ .<ioO .xcjluiiW 
fcaii.,3.7 qoi^r <£.,« ypl .1iiwt'V,U:-u(T ;, 7«,qo».\4nB<fW(«M ,sl,uUurw,»r,M 

5>n "pwo ‘^niiiwiwOtiiJtin-).,?. sfO .rtK^T onA bal 





■ I , 


^ ’ ;<• 


IA 



I * 


1 




CONTENTS 


Chapter Page 

Preface.iii 

Acknowledgments .iv 

1 Introduction.1-1 

1.1 Key Impact Assessment Terms and Concepts.1-2 

1.1.1 Inventory Item.1-3 

1.1.2 Impact .1-4 

1.1.3 Impact Assessment.1-6 

1.2 Purpose of Life-Cycle Impact Assessment .1-7 

1.2.1 Relationship Between Impact Assessment and Inventory 

Analysis.1-8 

1.2.2 Relationship Between Impact Assessment and Improvement 

Assessment. 1-9 

1.3 Applications of Impact Assessment.1-9 

1.3.1 Internal Applications.1-11 

1.3.2 External Applications.1-12 

1.4 Current State of Impact Assessment Practice .1-12 

1.5 Scope of This Document.1-14 

2 Key Issues Surrounding Impact Assessment.2-1 

2.1 Standardization of the Impact Assessment Framework.2-1 

2.2 Use of Scoping in Impact Assessment.2-2 

2.3 Communicating Uncertainty in Impact Assessment .2-5 

2.3.1 Translating Inventory Items to Impacts.2-5 

2.3.2 Impact Assessment Results .2-8 

2.3.3 Methods of Uncertainty and Sensitivity Analysis.2-8 


V 
























CONTENTS (CONTINUED) 


Chapter Page 

2.4 Data Availability and Quality Concerns in Impact Assessment .2-10 

2.4.1 Evaluating Data Availability.2-12 

2.4.2 Evaluating Data Quality.2-13 

2.5 Incorporating Value Judgments into Impact Assessment .2-14 

2.6 Transparency .2-15 

2.7 Expert Peer Review.2-16 

2.8 Presentation of Impact Assessment Results. 2-17 

2.9 Unresolved Issues .2-22 

3 A Conceptual Framework for Impact Assessment.3-1 

3.1 Classification.3-3 

3.1.1 Developing Impact 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 .3-8 

3.2 Characterization.3-9 

3.2.1 Determining Assessment Endpoints . 3-10 

3.2.2 Selecting Measurement Endpoints. 3-12 

3.2.3 Applying Characterization Models to Develop Impact 

Descriptors.3-14 

3.2.4 Impact Descriptors.3-17 

3.3 Valuation .3-19 


4 Existing Methods for Characterizing Impacts.4-1 

4.1 Checklist Approach.4-1 

4.2 Relative Magnitude Approach.4-4 

4.3 Environmental Standards Relation .4-6 


VI 

























CONTENTS (CONTINUED) 


Chapter Page 

4.4 Impact Potentials.4-10 

4.5 Critical Volume Approach.4-15 

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 

4.9 Mackay Unit World Model .4-27 

4.10 Canonical Environment Modeling.4-31 

4.11 Ecological Risk Assessment .4-32 

4.12 Human Health Risk Assessment .4-36 

5 Resource Depletion: Issues and Characterization Methods.5-1 

5.1 Resource Depletion: Key Terms and Concepts . 5-1 

5.2 Sustainable Development and Its Relationship to Resource Depletion .... 5-3 

5.3 Resource Depletion Models.5-4 

5.3.1 Resource Consumption Ratio .5-6 

5.3.2 Resource Depletion Matrix .5-8 

6 Methods for Conducting Valuation.6-1 

6.1 Decision Analysis Using Multi-Attribute Utility Theory (MAUT).6-1 

6.2 Analytic Hierarchy Process (AHP) .6-4 

6.3 Modified Delphi Technique.6-8 

6.4 Life-Cycle Costing.6-12 

6.4.1 Hedonic Pricing.6-14 

6.4.2 Contingent Valuation.6-16 

6.4.3 Cost of Control Valuation .6-16 


Vll 
























CONTENTS (CONTINUED) 


Chapter Page 

7 Integrated Methods for Impact Assessment.7-1 

7.1 Impact Analysis Matrix (lAM) .7-1 

7.2 The Environmental Priority Strategy (EPS) Enviro-Accounting 

Method.7-7 

7.3 Integrated Manufacturing and Design Initiative (IMDI) 

Environmentally Conscious Manufacturing (ECM) Life-Cycle 

Analysis.7-8 

7.4 Integrated Substance Chain Management.7-12 

7.5 ECO-Rational Path Method (EPM).7-14 

8 Key Points and Future Research Needs.8-1 

8.1 Summary of Key Points.8-1 

8.2 Potential Future Research Needs.8-2 

Appendix A: National Environmental Protection Policy Act (NEPA) 

Environmental Assessment Procedures. A-1 

Appendix B: Additional Impact Assessment Methods.B-1 

Appendix C: Key Terms and Definitions. C-1 

Appendix D: Bibliography. D-1 


Vlll 















FIGURES 


Number Page 

1-1 LCA Conceptual Framework.1-1 

1- 2 Range of LC A Applications.1-10 

2- 1 Phased Approach to Impact Assessment Development. 2-3 

2-2 Impact Assessment Results Summary Chart.2-20 

2- 3 Example of a Possible Impact Assessment Results Format.2-21 

3- 1 Conceptual Framework for Life-Cycle Impact Assessment .3-2 

3-2 Key Impact Assessment Decision Points.3-4 

3-3 Example of Basic Network Using CO 2 .3-5 

3-4 NOjj Example of Multiple Pathway Impact Network from a Single Inventory 

Item .3-6 

3-5 Example of Multiple Inventory Items Leading to Similar Impacts.3-7 

3-6 Possible Impact Categories .3-9 

3- 7 Exercise for Choosing Characterization Models. 3-18 

4- 1 Example Output from the Unit World Model .4-30 

4- 2 Conceptual Framework for Ecological Risk Assessment .4-34 

5- 1 The Life Cycle of Resources.5-2 

5- 2 Resource Depletion Matrix .5-10 

6- 1 Details of MAUT Water Pollution Effects Objectives .6-3 

6-2 Example Framework for AHP Applied to Impact Assessment.6-7 

6- 3 Modified Delphi Technique.6-9 

7- 1 User-Level Impact Analysis Matrix for Ecosystem Impacts.7-4 

7-2 Global-Level Impact Analysis Matrix for Ecosystem Impacts.7-4 

7-3 Options Map for Integrated Substance Chain Management .7-13 

7-4 Conceptual Framework for the EPM.7-15 


IX 


























’) \ ■ 






i. 


♦ 


: i} S 


t/t 


r 

^ ^. .. ». 'ifWa^ ;ttvO A'jJ 

i f r4;ip»fu.> - .v.r<-v ,,. . 

' *’ ••'•• ... . . .. »i, *.. *t"*.'^a3<lc! ;A A.J J .4 

■ ^ .., *.* - • ' ‘* A >rj.*wjmr| Of (4i«{riu*‘|fSK hiHifTT 

nc c ^ ' • ■.'• ** * 

* •*“^ * ^ .. - < ., V^.. ||t f‘r 


r , * '. 

M 

> ♦»*. * * < 

M ' 

f>r 


J '! 


• •• •♦ •««€»* 

S -F. .. - 



"!^Trr: 


i 


I I • . 


•If’ ! r»%i .r .. -' "t ■ ]ul':*\^^. \^t^ 


J-f . 


d-t . 

r-€ , 


••*►•■•*). » .... . . 4 . . , , . rl't*#*/? '3^ 

. ' • ’ •» * ■ '^. •. J!30 i»qr«£o’ 

. , ‘ U »i.: * > 

•(T^ffii,y«i Air A ,|nf-tt r l b OK 


t-£ 

U 


« 
f 


11 


> « « • « 1 I I i 


I ♦ ^ I • « • . , 


’'•t ' 

l-A 

• - * • « 
l-e 




1 


1 
5 < 


. ' < ♦^^'*< 4.. 5 J**/ ^ K^_ y •V^lllO'.n^llf. ■ I ‘.fjjHI? 

3j » . . ♦ , >^ >, , ». '*'' t (fA f>^.sr^y'' 

0F-i« . , fjitv, M iiiKrW t^U 3<V JtqiNn /:,rst!./il^ 

• ^ . : u,.' » /f»^‘- _ 

ftf ? s ianf^ei t,v 

.• ;j * •.:.*^' * *'j/ • • - " • •*•.••• •. Xr-'^M iU^V 'quCi 

A|nt. ,♦ * #, >*#]» r»f*i-^ .’-» «• U'44 h .... 

' ‘ • •• »• .X ia<tv/ r ;am >o " 

* ‘ % * .? t 

. .....JP(>na/‘.?^#A 'tiiA ivi ilf^ ■' >rr#rrli 3f<j At -4 


. *4 


. » % c C 


• * #> • 


<^V 


. 


« <4 i 


3ur>ifl.hJ0., i/ jfrja i>f6fwV 

:.,<fzr u*.o#fu 

. n?* . Ir^MO 

... v W« 3 ^■'>r,-iriiiZ tu3 qtM motsf^ 

* • « .- • . . *4.# « . ,, ,,! "iH? trril T»^* liBl/lCJtiQ'^TCO 


k-T 

tbt .. 


t-d 

ti 

i-r 

e-r 

KT 


4 



‘ 1*1 ‘ 4;., 

‘ ’ *■».' j'* ■’.'''' 

; ' •■' ’V. 

■■‘■^ . v i,* -. 

' 1(4 


"stCi ..' . 


’>11( 












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.2-10 

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 HDPE.3-10 

3-2 Example Classification of Inventory Items Under Impact Categories for 

HDPE Manufacturing .3-11 

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.4-5 

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.4-15 

4-8 Select Environmental Indices Used in EPS.4-18 

4-9 Example Environmental Load Values.4-19 

4-10 Carcinogen Potency Factors and Isophorone Equivalents.4-21 

4-11 Example RfDS for Noncarcinogenic Ranking.4-22 


X 























TABLES (CONTINUED) 


Number Page 

4-12 Example Human Health Impact Equivalency Ranking. 4-24 

4-13 Hypothetical Example of TPBP Approach .4-26 

4- 14 Measures of Risk for Human Health Risk Assessment.4-37 

5- 1 Example Calculations of Generic Resource Consumption Ratios.5-7 

6- 1 Example Results of Using the Modified Delphi Procedure for Comparing 

Environmental Areas.6-11 

7- 1 TCA Substitute Study Inventory Data.7-3 

7- 2 Leopold Interaction Matrix .7-6 

8- 1 Potential Future Needs for Impact Assessment Research.8-3 


XI 










CHAPTER 1 


INTRODUCTION 

Although a wide variety of impact assessment techniques have been integral to 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 



Inventory 

Analysis 


Figure 1-1. LCA Conceptual Framework 


1-1 



• Goal definition and scoping: the explanation of the study’s purpose and objectives; 
the identification of 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: the identification and quantification of raw materials and energy 
inputs, air emissions, water effluents, solid waste, and other life-cycle inputs and 
outputs. 

• 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 an 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 products 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 LCA analysts to concentrate on developing and 
refining procedures for building credible and reliable inventories of system inputs and outputs 
and using these inventories for identifying possible improvement opportunities. 

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 for 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 used 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 


Inventory Item 


Impact 


Definition 


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. 


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. 


Examples 


Issues Related to 

Impact 

Assessment 


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. 


• 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 items in the context of LCA is that the composition 
of the inventory is primarily based on the goals and scope of the study. 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 compound(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 (SO 2 ) and 
particulate matter—where small particles transport SO 2 and sometimes sulfuric acid deep into 
the lungs—can 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 be 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 (ERA, 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 SOg emissions to the air from a given product or process 
system. This SO 2 can be linked to primary impacts such as acid precipitation. Acid 
precipitation is an impact of SO 2 emissions as well as a stressor, because it can be linked to 
secondary impacts such as acidification of water bodies, tree damage, building materials 
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 




availability, modeling, and resource limitations—and the complexity of most natural 
systems—often require a more qualitative description of impacts based on some amount of 
quantitative information (e.g., level of pollutants released). This issue, however, should not 
discourage practitioners from 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 and 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 be infeasible. Therefore, practitioners need to decide 
which impacts are within the goals and scope 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, synergism, antagonism, etc.) 

The uncertainty associated with an impact actually occurring is often the subject of considerable 
debate. Uncertainty, in this context, focuses on 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: 

• 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 impacts 
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 
(SET AC, 1993). 

• A process that meaningfully relates inventory information into relevant concerns about 
natural resource usage and potential effects of environmental loadings, consistent with 
the defined 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 of one pollutant, say SO 2 , are estimated to be higher for one 
production system, while emissions of a different pollutant, say reactive hydrocarbons, are 
estimated to be higher for the other production system. 

LCA is not necessarily a linear or stepwise process. Rather, as suggested by Figure 1-1, 
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. 

1.2.1 Relationship Between Impact Assessment and Inventory Analysis 

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 commensurate 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., toxicological, 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. 


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 for their relevance to predetermined LCA goals and objectives, and 

• translating all impacts and their determined importance to the LCA audience in a clear 
and concise manner (Canadian Standards Association, 1992). 

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 types 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 overview 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 


FORUM 

Internal 


DEGREE OF 

SCOPE QUANTIFICATION 


3 

O 

(D 

fi> 

». 

5‘ 

(Q 


3 

O 

0 > 

52. 

5' 

(O 


External Policymaking/ 
Governmental 



External 



Figure 1-2. Range of LCA Applications 

Source: SETAC, 1993 


1-10 








TABLE 1-2. POTENTIAL INTERNAL AND EXTERNAL APPLICATIONS FOR 

IMPACT ASSESSMENT 


Internal Applications External 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. 


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 process 
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. 


1.3.1 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 as 
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 
the assessment in the same manner as an external study. 


1-11 






Using a holistic, systematic approach to 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. 

1.3.2 External Applications 

An external application of an impact assessment is one in which results are made 
available outside the sponsoring organization (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 review and disclosure of methods and results. 

The public may expect external impact assessments to abide by impact assessment 
guidelines that represent a wide consensus of 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 case when the impact assessment results are used to 
support marketing claims that make product comparisons and may significantly affect other 
external entities, or when the results might affect public policy. 

1.4 CURRENT STATE OF IMPACT ASSESSMENT PRACTICE 

World Wildlife Fund (1991) recently updated a survey of three LCA 
practitioners—Battelle, Franklin Associates, Ltd., and Tellus Institute—to 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. 

In addition, an industry survey by Sullivan 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 that 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 

Battelle 

Franklin 
Associates, Ltd. 

Tellus 

Institute 

Amount/volume 

Yes 

Yes 

Yes 

Toxicity 

Yes 

Yes 

Yes 

Exposure 

Only where generic pathway 
is defined. 

Yes 

No 

Persistence 

Via mechanical breakdown or 
degradability. 

Yes 

No 

Mobility 

Via surrogate measures (e.g., 
water solubility). 

Yes 

No 

Global effects (e.g., climate change, 
ozone depletion) 

Establish equivalency of 
various individual 
contributions. 

Yes 

Yes 

Risk assessments 

No 

No 

No 

Consumer/worker safety 

1. For releases to the environment, 

what criteria are used to select 
pollutants to measure 

No 

No 

No 

a) pollutants covered by 
federal/state laws and 
regulations 

Yes 

Yes 

Yes 

b) pollutants that exceed some 
threshold level, regardless of 
regulatory controls 

Yes 

Yes 

No 

c) impact of pollutants (e.g., 
toxicity, etc.) 

Establish impact potential 
networks (inventory vs. 
impacts). 

Variable 

Yes 

d) SARA 313 list of toxic 
chemicals 

— 

— 

— 

2. Are releases assumed to meet 
current treatment standards? 

Sometimes. Prefer actual 
releases; treatment standards 
used only if no other data are 
available. 

Only if actual 
emission data are 
not available. 

Only if actual 
emission data are not 
available. 


(continued) 


1-13 






TABLE 1-3. PRACTITIONER SURVEY OF IMPACT ASSESSMENT 

CONSIDERATIONS (CONTINUED) 


Considerations 


Franklin Tellus 

Battelle_Associates, Ltd.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. 

At least through the 

characterization 

phase. 

Yes 

Have used valuation by 
Analytic Hierarchy Process 
(AHP) in streamlined LCA, 
but never in a conventional 
LCA. 

Where comparison 
measures can be 
developed. 

Methods developed to 
rank relative impacts, 
especially within 
media. 

Mix of qualitative/ 
quantitative depends on 
product stage and 
environmental pathways. 

Mix—depending 
on the quality of 
data. 

Quantitative 


Source: Updated in 1994 from World Wildlife Fund, 1991. 


1.5 SCOPE OF THIS DOCUMENT 

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 into 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 that 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 


tnotin m K(i.a lift 07 tbrjn rfmwsMiirti'-fcrtiS'; i.-.xiii^;». )a<nmi gnlbis^M 

\(d fcsofi^: 

*J j-arii lU. A <4 Y . A . 4 . i A. ^ 


Y»?ii 


ibOfJizi^rt .A Xf!>iToqi;/i tti 'A'W*/.> .la'irvfj j JncotifiH ^il 

ow[ 0 /Ibn^fjdA .1^ A'"»r!(j ^j.f TH’* , ft r'.jfcf.y^v ’*r.* fu^o^i 


M ' 

M 



At- '•*•.■■ 


^ • * 


w-Hn<t< •’ll 'ai- **«.-■ • ■ '<11 '* aIa-_ a - ^.t'i', 

wfi -'I.* '‘t-fe.'^, -w 

tf, Ok'r¥iH‘ ^0 

** 


- ^ ' •';<; V.»-' 


‘**y wj^hh* 


««kJ 




Si ^ n/m l‘ «v '■ 


* * '»i 


-I*— k> 


.MU M 

t-vti 

(u 


ft». 

■I ■•«r .iFlllity Oi 





id 0 , • 4 t if * f 


‘-yiS^intttfrssw 


V AW-' 


■..wjimitM 

I » 

H ' '• ■ 


^XHt^ f*uwt : 


« ' ■ 

' 1 '•'. 

;■ 


* ■}. T 



>ri ••» 1 iivSiKj'c^'^rT^ 7 




>3»t * » 

r * 


I 


Sw : ms I nMUvtx a i,i'4 

1 a* ifoitiL-nrxf' ', 


‘■jwy k -a^, 4t»4 %aim! ■ .--jes ■ 

• ■''*“" »r^gn<j|g '^te , uat?yiil 

4VJ *• 4 s*'^To;l>4TJr^>;.»{iAft»»-^ >it* |j^|i itr,. iO. fj 

stjt:ii^Wr liia-Ji^ nwcHl^.'r* • j *); (KiK^n • n»‘,..'itA qustuy, Vnhig 

*1' •» ^ Ml'ti ,1 •■'■■-. **•»'■• '■ • ■ 


A* »V 


•H 

Oi^ptar,' ■■•.*'ai. .. - ^i3fe0R|4|uiJ frill'. *in *. ^ • '.Hii#*»-’vvn-ni.. -.'^inli ipchsdtoi JKf« ♦ 


,.‘'4 


dab*;. *.•'«* f »t;;p ,7 tujiui'Uf'nnjviA v.tf oC^kito^i 

Hn-,i' ah ' ii ' * )QSi v'/tH^n *rjd vv*f<''?' iia*; *^1 1®.$* 1 blial'jo 

s;te^ *h? •>,•'*.■ ^ I•■ ' n th® r'haf < »ieMi fieuilea U vim% 

^ /' ■* 

itjiUtent tii 1.'. ’ 




j Till III ' ' « 

'/aJ ' • ' 

I-i L *’i. MSi *1 ■!. 

*h ' 

ImrtriV tr M 

11 

(o tvv riAi.:v V >1 r.^ir rtitaii^ f««flcsiil^^ < 

CW8ftf^6pffjsoM»l^' ft-.* '-1^. til a»f 



»H OtafKuTi '* Aliii !>•« ^ftadf ^ efiwntii^rrlang 

^, ’'j fiatur*! tn|R9iii'0||| n^our^ dPiCiiWtrCfi nrptc.s i-d 


. n 

I I 


At,': 


tM 






1 M 









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 would facilitate comparison between studies. 

• Analysts would be able to incorporate results of other external LCAs into their studies. 

• 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. 

Only a few impact assessments have 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. 

In this approach, experts develop and agree upon general principles and procedures, 
analysts begin preliminary case studies using 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 

Scoping—deciding what will and will not be included in the study—is 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 be tied to the goals and scope of the LCA. 


2-2 



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 














• providing justification for excluding any elements, 

• determining the level of impacts to be studied (source, media, or receptor), 

• considering the audience to which results will be presented, and 

• defining spatial and temporal 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.” The nonthreshold assumption simply says that no 
threshold exists for considering environmental loadings in an impact assessment. In other 
words, despite seemingly insignificant quantities, inventory items nonetheless contribute 
cumulatively to impacts and therefore may need to be considered in the impact assessment. For 
example, the energy used to manufacture a single automobile likely does not release enough SO 2 
to the atmosphere to cause an appreciable rise in regional acid precipitation. However, when 
those SO 2 emissions are considered in the context of additional regional emissions, the SO 2 
emissions may be considered a contributor to the regional acid precipitation and thus the SO 2 
emissions may warrant consideration in the impact assessment. 

The nonthreshold assumption may be 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 outcome 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, SO 2 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, 

• interactions among human-induced loadings and natural loadings, 

• natural variability and the problems of discriminating from “background noise,” and 

• 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 irreducible—for 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 impacts 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 given 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 able to provide a direct measure of uncertainty, the 
concern remains about compounding uncertainties from the inventory with uncertainties from an 
impact assessment. That is, when the LCA 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). 

Uncertainty clearly plays a large role in impact assessment. This is not to say that LCA 
or impact assessment has no use as a decisionmaking tool, because all decisionmaking tools 
contain some degree of uncertainty whether or not it is explicit. For example, similar problems 
exist in the field of risk assessment where analysts are constantly faced with data-input 
limitations and exposure and effect uncertainties. Like risk assessments, impact assessment can 
be performed with all levels of information, from abysmal to excellent, and can address a variety 
of levels of assessment, from release of individual substances to an environmental media to the 
release of multiple substances to multiple environmental media. 

The key point is that the lower the quality of the information and models used in the 
assessment, the more uncertain the outcome. Therefore, developing a method of identifying and 
communicating uncertainty should benefit the users of impact assessment. Some general areas 
of impact assessment that may be used as indicators of the overall certainty or uncertainty of the 
assessment, and thus affect the quality and usefulness of results, might include the following: 

• quality of input data, 

• structure of impact characterization model, 

• type of model testing, and 

• level of expert review. 

2.3.3 Methods of Uncertainty and Sensitivity Analysis 

Both quantitative and qualitative techniques are available for expressing data quality in 
the context of impact assessment. These methods include the following: 


2-8 


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 

• validity, accuracy, and limitations of classifying inventory items into impact categories 

• validity, accuracy, and limitations of conversion models used 

Techniques such as sensitivity analysis or uncertainty analysis may provide useful 
starting points for impact assessment. Sensitivity analysis is a systematic procedure for 
estimating the effects of data uncertainties on the outcome of a computational model (EPA, 
1993a). It provides a means of determining what does and does not matter in a computational 
model. Researchers have recognized that applying sensitivity analysis in the context of impact 
assessment may be useful in theory only, because it requires the quite difficult process of 
developing mathematical models to evaluate system parameters (EPA, 1994b). However, 
analysts may be able to develop variations of sensitivity analysis methods that better fit the needs 
of impact assessment. Table 2-1 describes proposed methods for sensitivity analysis in the 
context of impact assessment. For further discussions of these methods and their potential 
future applicability to LCA, the reader is referred to EPA (1992b). 

Uncertainty analysis identifies, discusses, and quantifies, to the extent possible, the 
uncertainty in identifying and characterizing potential impacts. The total uncertainty in the LCA 
represents cumulative uncertainties from each phase of impact assessment. Using uncertainty 
analysis, a practitioner can evaluate the effect of uncertainties on the overall impact assessment 
and, when applicable, determine ways for reducing uncertainty. In addition to providing the 
practitioner with insight to the impact assessment’s strengths and weaknesses, uncertainty 
analysis can also be used as a basis for decisionmaking purposes among comparative 
assessments. 

Table 2-2 lists various methods for uncertainty analysis and their advantages. For further 
discussions of the use of these methods in the context of LCA, the reader is directed to EPA 
(1992b). 


2-9 


TABLE 2-1. POSSIBLE APPROACHES, ADVANTAGES, AND DISADVANTAGES 
OF SENSITIVITY ANALYSIS METHODS FOR IMPACT ASSESSMENT 


Method 

Advantages 

Disadvantages 

Tornado diagrams 

• relatively simple to use 

• wide range of applicability 

• requires the development of a 
mathematical model 

Dominance considerations 

• useful for determining the 
dominance of specific 
alternatives 

• more applicable for option 
selection than for sensitivity 
evaluation 

Two-way and three-way 
sensitivity analysis 

• allows for evaluations of multiple 
variables at the same time 

• useful for evaluating impacts of 
alternatives 

• does not focus on data 
quality per se 

Deterministic sensitivity 
analysis 

• applicable to LCA data-quality 
evaluation 

• identifies the most significant 
variables 

• requires the development of a 
mathematical model 


2.4 DATA AVAILABILITY AND QUALITY CONCERNS IN IMPACT 

ASSESSMENT 

Recent LCA forums (SETAC LCA Data Quality Workshop, Wintergreen, Virginia, 
October 4-9, 1992; and SET AC 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 which LCA information can be used, identifying and 
evaluating data quality and their relationship to 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 or 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. 


2-10 






TABLE 2-2. POSSIBLE APPROACHES, ADVANTAGES, AND DISADVANTAGES OF 
UNCERTAINTY METHODS FOR IMPACT ASSESSMENT 


Method 

Advantages 

Disadvantages 

Analytic 

• ranks contributors to uncertainty 

• limited applicability 

Monte Carlo 

• economical 

• widely applicable 

• facilitates understanding of 
sampling distribution concepts 

• sensitivity to input 
assumptions hard to assess 

• no ranking of uncertainty 
contributors 

• dependence on accurate 
information and covariance of 
input parameters 

Response surfaces 

• economical 

• widely applicable 

• ranking of uncertainty contributors 

• accuracy hard to assess 

Differential sensitivity 

• widely applicable 

• ranks contributors to uncertainty 

• long computation times 
possible 

• large model and code 
development costs 

Evaluation of confidence 

• measures uncertainty due to 

• limited applicability 

intervals 

statistical variability in data 

• 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 databases often contained out-of-date data (Tellus Institute, 1992a). 

2.4.1 Evaluating Data Availability 

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 

Mass, volume, or other units of physical quantity. 

Tier 2: Equivalency Assessment 

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. 

Tier 3: Toxicity, Persistence, and 
Bioaccumulation 

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. 

Tier 4: Generic Exposure/Effects 
Assessment 

Same as Tier 1 plus generic environmental and human health 
data. 

Tier 5: Site-Specific Exposure/ Effects 
Assessment 

Same as Tier 1 plus site-specific exposure and environmental 
and human health data. 


Source: SETAC, 1993 


A recent SET AC-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 in 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 in a data source or a data 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 decisionmaking 
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, as well as any additional information required by the 
models, transparent to the users of the results. 

2.5 INCORPORATING VALUE JUDGMENTS INTO IMPACT ASSESSMENT 

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. Clearly articulating subjective value judgments lets 
the user know the values that guided the impact assessment. 

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 values to use during the impact assessment. 

However, for purposes of impact assessment, and LCA in general, developing a standard 
protocol for identifying and evaluating value 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-effect 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 on 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 of 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 variability—uncertainty 

• models/methods used in the impact assessment process 

- assumptions 

- limitations 

• data or methodology manipulations 


2-15 


• value judgments 

• exclusions 

• lost information due to aggregation, etc. 

• analyst’s interpretation of the implications of all above items on LCA results 

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. 

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 replicable as possible and should fully explain and Justify 
factors that preclude them from doing so (Denison, 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 for 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. Practitioners should strive to conduct credible assessments and present 
the results objectively. 

Specific aspects of the impact assessment that need to be documented and presented to 
decisionmakers include the following: 

Content Aspects 

• Clearly delineate scoping activities, including how the boundaries of analysis were 
determined. 

• Report any objective data or results separately from subjective data or results. 

• Express the characteristics of the database, including data sources, uncertainty, and 
assumptions. 

• 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 methods 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 impact assessment including measures of 
variability/uncertainty, 

• a summary of assumptions and value judgments made in the impact assessment, 

• a description of the methods/models used in the impact assessment, 

• a description of problems encountered and how they were resolved, and 

• a summary of unresolved issues. 


2-18 


TABLE 2-4. COMPARISON OF IMPACT ASSESSMENT RESULTS 

PRESENTATION METHODS 


Presentation 

Method 

Objective 

Advantages 

Disadvantages 

Single Score 

• Provides a single impact 

• Provides 

• Ambiguous derivation 


score by aggregating all 

comparative value 

• Does not allow for a relative 


impacts based on a 
common denominator 

• Easy to communicate 

comparison of impacts 

• Difficult to incorporate qualitative 
data 

• Provides no information beyond 
valuation results 

• Fixes subjective values that cannot 
be shared by others. 

Qualitative 

• Uses symbols or ranges 

• Provides 

• Provides no quantitative support 

Rank 

of subjective impact 

comparative value 

• Difficult to express relative 


scores to provide an 
overall ranking of 
impacts 

• Easy to communicate 

comparison of impacts 

• Too simplistic 

• Provides no information beyond 
valuation results 

Prioritization 

• Prioritizes impacts based 

• Provides 

• Encourages focus on only the top 


on subjective values that 

comparative value 

priorities 


attempt to identify the 

• Identifies high- and 

• Priorities are often subjective in 


more and less pressing 
impacts 

low-priority items 

nature 

• Provides no information beyond 
valuation results 

Impact Score 

• Provides a quantitative 

• Provides 

• Potential for confusion with too 

Matrix 

or qualitative overview 

comparative value 

much information 


of impact categories 

• Provides standard 
format 

• Incorporates 
quantitative and 
qualitative data 

• Provides no information beyond 
valuation results 

Impact 

• Provides a variety of 

• Provides 

• Potential for confusion with too 

Assessment 

information beyond 

comparative value of 

much information 

Summary Chart 

impact assessment 
results (such as databases 
used, methods used, 
value judgments, and 
limitations), which gives 
decisionmakers an 
overview of the impact 
assessment process 

entire impact 
assessment process 

• Provides information 
on impact 
assessment process 
beyond valuation 
results 

• Provides a standard 
format 

• Incorporates 
quantitative and 
qualitative data 



2-19 






Decisionmakers could use this information as a tool to look at the overall picture or to 
focus on a particular aspect of the impact assessment. A standard presentation format would 
provide a clear and effective means of communicating the potentially complex array of 
information inherent in impact assessment or 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. 





Data 


Models/Methods 



• sources 

• quality 

• limitations 

• uncertainty 


• limitations 

• assumptions 

• uncertainty 





Value Judgements 


Problems 

Encoimtered 



• whose values 

• implications 


• problems 

• how resolved 





Unresolved Issues 


Impact 

Assessment Results 



• issues 

• implications 


* see Figure 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 

Raw 

Materials 

Acquisition 

Manufacturing 

Processing 

Distribution/ 

Transportation 

Use/Reuse/ 

Maintenance 

Waste 

Management 

Ecosystem 






air 






water 






land 






biodiversity 






waste 






other 






Human Health 






occupational 






nonoccupational 






other 






Resource 






stock 






flow 






other 







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 table could handle both quantitative and qualitative 
information. Qualitative information could be expressed by symbols, such as + or indicating 
more or less significant impacts, or by circles with varied degrees of shading, indicating the 
magnitude of the impact. 

2.9 UNRESOLVED ISSUES 

Because impact assessment is still evolving, many issues persist that may play a role in 
the future development of impact assessment procedures and methods: 

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 linking 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. 

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 


V 



o 


u 


Develop Impact 
Networks 


Classify Inventory Items 
by Impact Category 


} 

} 


V 


4 - 


Determine Assessment 
Endpoints 


} 


r 

\ 


Select Measurement 
Endpoints 


J 


Apply Conversion Models to 
Develop Impact Descriptors 


} 


i 


UJ 

2 

(/) 

(/} 

LU 

(/) 

(/> 

< 

2 


LU 

-I 

o 

>- 

o 

UJ 

u. 



i 


Apply Valuation 
Methods to Synthesize 
Stakeholder Values and 
v^^Impac^escriptors 




Life Cycle 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 SO 2 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 SO 2 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 



Improvement 

Loop 


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 (CO 2 ) identified in the inventory 
analysis. A search of the literature reveals that CO 2 is often linked to the greenhouse effect, 
which is a buildup of CO 2 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.^ 

As an example of a multiple pathway impact network, consider the case of nitrogen 
oxides 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. 



Figure 3-3. Example of Basic Network Using CO 2 


Tn 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 


















Figure 3-4. NO^ Example of Multiple Pathway Impact Network from a 

Single Inventory Item 


Like many other impacts, a number of different substances may contribute to the 
greenhouse effect, as demonstrated in Figure 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 assessments increasingly more feasible and 
economical. 


3-6 
























Figure 3-5. Example of Multiple Inventory Items Leading to Similar Impacts 


3.1.2 Classifying Inventory Items Within Impact Categories 

After referring to existing impact networks or developing new ones, practitioners should 
review the networks to see if they contain any inherent structure that enables them to establish a 
set of impact categories under which inventory items can be grouped. For example, during the 
review of impact networks, researchers might identify that quantities of air emissions listed in 
the inventory analysis, such as CO 2 , methane (CH 4 ), chlorofluorocarbons (CFCs), and ozone 
(O 3 ), all contribute to the greenhouse effect. All four of these inventory items thus can be 
grouped in the subcategory, greenhouse effect, within the main ecosystem impacts category. 

The three main categories of impacts considered in an impact assessment include impacts 
to ecosystems, human health, and natural resources. Social welfare may be considered as an 
additional impact category, although currently no tools yield a credible analysis of such impacts. 


3-7 
























Despite the current lack of tools to analyze social welfare impacts, practitioners can attempt to 
incorporate social welfare impacts by 

• identifying the impacts of a product or process life cycle on social welfare, and 

• identifying the effect of social welfare impacts on ecosystems, human health, or natural 
resources. 

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 impact networks. The suggested approach to classification 
is to first build impact networks and see if they contain any inherent structure for developing 
subcategories of impacts rather than starting with a prestructured, and value-laden, list of impact 
subcategories. This approach to classification is essentially the same as that used by SET AC 
(1993), which groups inventory items into relatively homogeneous problem types, called stressor 
categories. Our approach differs only in that the term “stressor” is not used because of ongoing 
confusion associated with the use of that term. 

3.1.3 Example Classification Exercise of High-Density Polyethylene (HDPE) Production 

An inventory analysis of an HDPE production system would likely include numerous 
components. Table 3-1 provides examples of information developed in an inventory analysis for 
the manufacture of HDPE. 

Ecosystem, human health, and natural 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 SO 2 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 SO 2 released into the atmosphere. 
From this search, it will likely be determined that SO 2 emissions to the atmosphere often 
combine with other atmospheric compounds to produce acid precipitation. Thus, as shown in 
Table 3-2, SO 2 can be categorized under the ecosystem impact category of acidification. 


3-8 


Ecosystem 

Human Health 

Natural Resources 

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 

• nutrification 

• eutrophication 

• chemical alteration 

Soil 

• toxicity impacts 

• salinity 

• lateritization 

• podzolization 

• acidification 

• fertilization 

• erosion 

Others 

• geomorphic effects 

• biodiversity effects 

• habitat alterations 

• animal welfare 

Chronic Effects 

• carcinogenic 

• mutagenic 

• teratogenic 

• neurological 
damage 

• reproductive 
disorders 

• major organ 
diseases 

- heart 

- lung 

- liver 

- kidney 

• radiation 

- ionized 

- ultraviolet 

- heat 

• physiological 

• anemia 

• skin disease 

• sterility 

Acute Effects 

• accidents 

- occupational 

- nonoccupational 

• radiation 

- ionized 

- ultraviolet 

- heat 

• noise 

• odor, taste, etc. 

• microorganisms 

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 

• solar power 


Social 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. Possible Impact Categories 


3.2 CHARACTERIZATION 

Although classification can provide useful information for reviewing the types of 
impacts associated with specific inventory items, further analysis may be desired to adequately 
describe those impacts. The characterization phase aims to describe and estimate the 
contribution of inventory items to environmental impacts via the use of specific impact 
assessment tools or characterization models. The intended output of the characterization phase is 
a set of impact descriptors, which include data points or other information that describes the 
relationship between specific inventory items and impacts. 


3-9 











TABLE 3-1. 

EXAMPLE INVENTORY ANALYSIS DATA FROM THE 
MANUFACTURE OF HDPE 

Component 

Inventory Item Quantity (mass or volume) 

Resource use 

Crude oil 

Energy demand 

Electricity 

Coal 

Renewable fuel 

Energy in material 

Air emissions 

CO2 

SO2 

NO, 

CO 

Hydrocarbons 

Particulates 

CFC 

Hydrogen 

Water effluents 

Crude oil 

Phenol 

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: 

• determining assessment endpoints, 

• selecting measurement endpoints (if necessary), and 

• applying characterization models to develop impact descriptors. 

3.2.1 Determining Assessment Endpoints 

After identifying the impacts associated with inventory items and grouping them into 
impact categories in the classification stage, the practitioner should review the previously 


3-10 










established goals and scope, and other key decision points, of the overall LCA to identify which 
of the impacts are within the scope of the study, which are referred to as the assessment 
endpoints. These endpoints represent the focus of the characterization efforts. 


TABLE 3-2. EXAMPLE CLASSIFICATION OF INVENTORY ITEMS UNDER 
IMPACT CATEGORIES FOR HOPE MANUFACTURING 


Ecosystem 

1 Impacts 

Human Health Impacts 

Natural Resource Impacts 

Impact 

Category 

Inyentory 

Item 

Impact 

Category 

Inventory 

Item 

Impact 

Category 

Inventory 

Item 

Greenhouse 

effect 

CO, 

CFC 

Particulates 

Carcinogenic 

effects 

Crude oil 

Fossil fuel 
depletion 

Crude oil 

Fossil fuel 

Ozone depletion 

CFC 

Lung damage 

Particulates 

SO 2 

NO, 

Hydrocarbons 

Renewable 
energy use 

Renewable fuel 

Acidification 

SO 2 

NO, 

Odor 

Solid waste 
Ethylene 

Oil 

Phenol 



Smog/fog 

NO, 





Water 

contamination 

Oil 

Phenol 

N 

Organic C 

Solid waste 





Habitat 

alteration 

Fossil fuel 
Renewable fuel 
Solid waste 
Electricity 





Geomorphic 

alteration 

Electricity 

Solid waste 

Fossil fuel 
Renewable fuel 






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 2 emissions quantified in the inventory analysis can be associated with acid 
precipitation, which in turn can lead to a number 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 defined criteria for guiding the determination of 
assessment endpoints. Table 3-3 outlines some 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. 

3.2.2 Selecting Measurement Endpoints 

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 SO 2 emissions can be used as a surrogate to link that quantity of 
SO 2 emissions to impacts such as leaching of metals from soils, tree damage, or fish mortality. If 
the assessment endpoint is directly measurable, then it can be used as a measurement endpoint. 

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 endpoint to the goals and scope of the LCA, 


3-12 


TABLE 3-3. SUGGESTED CRITERIA FOR DETERMINING ASSESSMENT 

ENDPOINTS 


Criteria 

Description 

Study goals 

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. 

Study scope 

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. 

Magnitude of 

environmental 

loading 

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). 

Environmental 

relevance 

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. 

Level 

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 values 

Stakeholder (including societal) values can range from protection of 
endangered species to preservation of environmental attributes for functional 
reasons (e.g., flood water retention by wetlands) or aesthetic reasons (e.g., 
visibility in the Grand Canyon). 

Data availability 

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 an endpoint with the scope and boundaries of the inventory analysis, 

• the intended application or end use of the impact assessment results, 

• data limitations, 

• the availability of impact assessment models, and 

• the ease of characterizing potential impacts (i.e., direct versus indirect impacts). 

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 SET AC Life-Cycle 
Impact Analysis Workshop (see SET AC, 1993) and the October 1992 SET AC 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 
SET AC-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 unaveraged 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 

Inventory data alone are used to evaluate 
on the basis of quantity or volume with 
the assumption that “less is better.” 

Mass, volume, or other units of 
physical quantity of inventory items. 

Tier 2: Equivalency 
Assessment 

Algorithms based on hazard information 
are used to derive impact equivalency 
units to evaluate inventory items within a 
specific impact category. 

Same as Tier 1, plus algorithms for 
equivalency conversions. Also can 
include resource stock and yield, and 
non-chemical loading information. 

Tier 3: Toxicity, 
Persistence, and 
Bioaccumulation 
Assessment 

Interactive properties between a chemical 
and an organism (toxicity) and an 
ecosystem (persistence and 
bioaccumulation) are used to evaluate 
inventory items. 

Same as Tier 1, plus information on 
characteristics of chemical 
interactions with organisms (toxicity) 
and ecosystems (persistence, 
bioaccumulation). 

Tier 4: Generic 

Exposure/Effects 

Assessment 

Generic environmental or human health 
information are used to estimate potential 
impacts of inventory items. 

Same as Tier 1, plus generic 
environmental information and 
regional calibration model. 

Tier 5: Site-Specific 

Exposure/Effects 

Assessment 

Site-specific environmental or human 
health information is used to estimate 
potential impacts of inventory items. 

Same as Tier 1, plus site-specific 
environmental information and a site- 
specific calibration model. 


Source: SETAC, 1993 


Loading Assessment 

Loading assessment is based on the premise of “less is better” and is the simplest type of 
characterization method. In loading assessment, the data generated in the inventory analysis are 
directly used to identify areas where impacts can be reduced through reductions in inputs and 
outputs. Loading assessment does not assess—qualitatively or quantitatively—the impacts of 
those inputs and outputs or the benefits of their reduction. 


3-15 








Equivalency Assessment 

Equivalency assessment includes approaches that translate inventory items into common 
units (via the use of equivalency factors) of impact that can either be evaluated to compare the 
individual contributions of inventory items to impacts or resulting equivalency units to assess the 
collective contribution of items to impacts. Equivalency 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). 

Toxicity, 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 organic 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 is the next higher level of complexity that includes 
approaches that use generic environmental and 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 



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 tropospheric 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 process. 
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 


tocqeii n< ^hboia >\obusnomu'h k> arff 

R ns r^nhh^n ,AD^(Jo 9nniS^m^ixiMst> Joujmi bm \'y'sfns^. ]o 

r’V\T32) 'fc£5 niiiifw r!(r:)hT>Ac,6 -uk? rplqotrg 

ocmio ah5ri(^^c‘qoI^ o; 9£Wjni p:?lf 

I^nargan of ^jLiruil to fwol cni/fbctfin i/-; 

20 vbvftKfdfijitp r^iro «TOiqno;*5l> -j®! <!jAaJLVHjUi\ L%^ m> ,r ’ f»»i!0iJi;ii3(|tH( 

mqtrn o<ho^9 tma jsm^m mw-> - sm nnsrux^rh ^ti/iWiL&up 


-j 




!1 


>fOUAlJAV LK 

bfiM xhid ymr. p !^r*: t^ffji\r;,b cfo-yj ‘ WR? lo ipA « wO 

ttoite..(((,<|i! •'■.j1*!uiai3«n>.n5 atdisaoq ->j 

' ^«>{l t>»(n (toitp0lar lo 


^bvri yURitri^-5 i 

^ -'• ^ --.u V .ild„ SnfeW 

. y M . 

• 1^1 * 




tcsjni'joi^i ant* 


nobix'sc^)x«^.^tnU'f<K*?j ^ ^ 

.^<■■»^Ol(i (i’<liJlfDht;!.j33ll >fpt U^Hc i , 1ftf/i^l'lt. «‘(Ti%V-*t.> u'‘>Ol«V 1i>i)iOltMU;« 

ujv}, -, ,,,# 5rfi ,j,r*-/w4®v 

s^nsibii. oftvti, «rat(i|»fUfi>}V{i jiorffi« »A^.J lAiot lnlv^ln VMjm UiT.Mo>)'f9 93h*iUi<! aifr*' 

b.x..«.h« t. no . ,«qmj 

srti ,„ar..r7,ao=|A:xj 

1«1>.->S «: ■> t=*Tm) ssntf nw JVB^,| Ml' ; ajjf.rj r^titolrv 

f isto 


''■ni 




\ y 


*»ii 



III ii«r,j:-.«i„.v,.„vf«, ,na«l^^:«t» . ,V -**»TS«/1 Mn^h (at»» 94 « a,i;aM 
!»rtl u£>b>a*d ylilM Km ok icttfei-,*,»)«»!',,• .-.f itWnB»!in«Wtetb«d«t^^ 'aomi>^jr-^a 
tjm. loamaj aM aisrsT-*- '< >''« - 'a «: pi/<. r WM «, c«b. n/d laiwaiucia n to TO}al.-.aiq 

Urffsin 9v,i;s>t(<o •(fta n s-’Pdcliv.'rt, jijyeiiifA nn;# f Mblr^Jitc fcflK ,|i-B Auk VjedJ’v.bni/ 
l^c^oioborifad, Plu^,lau.9 n.a . .ot 

►*«-0 









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. SUMMARY OF METHODS TO CHARACTERIZE IMPACTS 



Impact Categories Covered 


Tier of Detail® 




Human 

Natural 




Method 

Ecosystem 

Health 

Resources 

1 

2 3 4 

5 

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. 


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. 


4-2 









TABLE 4-2. EXAMPLE CHECKLIST FOR ECOSYSTEM IMPACTS 


Inventory Items 

Ecosystem Impacts 

CO2 SO2 

O3 

SoUd 

CFC Waste 

Oil 

Efiluent 

Crude Oil 
Use 

Atmosphere 






Toxicity 

Ozone depletion 


/ 




Greenhouse effect 

/ 

/ 




Visibility 

Smog/fog 

/ 

/ 




Ground ozone 


/ 




Climate change 






Water 






Toxicity 



/ 



Contamination 



/ / 

/ 


Depletion 

Thermal 

Turbidity 

Acidification 

Nutrification 

/ 


/ 


/ 

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/or outputs, and 

• comparing levels of inputs and/or outputs between alternative materials, processes, or 
products. 

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 relative magnitudes. 

Some additional weaknesses of the checklist 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, 

• 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 checklist approach alone can be used 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 for 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 products or production systems. Such “quick and dirty” 
comparisons may not only help identify some key 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 

Inventory Item 

Quantity 

(tons) 

Subrange 

Score 

Greenhouse Effect 

co^ 

4.000 

10 


CH 4 

0.403 

2 


NjO 

0.173 

2 


O3 

0.009 

1 


CFC 

0.001 

1 

Acidification 

SO 2 

1.380 

10 


NO 

0.470 

4 


NO, 

0.053 

1 

Habitat Alteration 

Timber 

6.000 

10 


Coal 

5.500 

9 


Iron Ore 

0.950 

1 


4-5 







Strengths 

The relative magnitude approach is relatively easy to use; it is based on cause/effect 
linkages and takes into account the relative quantities of inventory items. In addition, the relative 
magnitude approach can be helpful for 

• 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, 

• 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 pinpoint areas where a more detailed level of 
analysis is needed. 

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 for prioritizing the standards that should be used in the 
weighting scheme is necessary. 


TABLE 4-4. EXAMPLE APPROACH FOR DEVELOPING ENVIRONMENTAL 

STANDARDS RELATION WEIGHTS 


Substance 

Ambient Air Standard 
(me/m^) 

Ambient Air Standard 
Expressed in (mg/mole) 

Weighting Factor for Air 
Emissions Pollution Units 
(PU/kg) 

A 

1 

0.024 

30.0 

B 

12 

0.28 

2.6 

C 

4 

0.095 

7.6 


Ambient Land Standard 
(mg/kg) 

Ambient Land Standard 
Expressed in (mg/mole) 

Weighting Factor for Land 
Emissions Pollution Units 
(PU/kg) 

A 

5 

0.28 

2.6 

B 

10 

0.57 

1.3 

C 

8 

0.45 

1.6 


Ambient Water Standard 
(mg/I) 

Ambient Water Standard 
Expressed in (mg/mole) 

Weighting Factor for Water 
Emissions Pollution Units 
(PU/kg) 

A 

2 

0.036 

20 

B 

4 

0.72 

1 

C 

3 

0.054 

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 aquatic 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. Preliminary work is being performed to develop pollutant weighting factors 
based on U.S. regulatory standards. The use of guidelines and reference concentrations, which 
are not regulatory standards, may also prove to be useful for the type of analysis. Especially if 
these reference levels are more strongly based on health and environmental effects rather than 
technical or 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 

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 impact 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. 

The formula shown below illustrates the generic method for deriving impact equivalency 

units: 


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-the-art impact potential functions that are available 
for characterizing specific impact categories. 

A major concern of impact potentials is developing equivalency factors for all impact 
categories that relate inventory data to specific impacts. While it is generally agreed that 
equivalency factors should be based on impact mechanisms directly related to the impact 


4-10 


categories, it is unclear at this time how equivalency factors would be developed for all 
categories of impacts. 

Impact Potential Example: Ozone Depletion Potential (ODP) 

Halocarbons, in addition to being a greenhouse gas, also destroy the stratospheric ozone 
layer that protects all life 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 SET AC 
(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” 

GWPj = 

T 

f a.Cj(t)dt 

0 

The global warming 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 

T 

j^co2'^co2 

0 

forcing of 1 kg of carbon dioxide (CO 2 ): where a^ is the 
instantaneous radiative forcing due to a unit increase in the 
concentration of trace gas I, Cj(t) is the concentration of the trace 
gas I at time t after its release, and T is the number of years over 
which the calculation is performed. 

ODP(x) 

OjCx) 

03 (CFC- 11 ) 

The ozone depletion potential (ODP) 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 
that for a unit mass emission of CFC-11: where ODP(x) is the 
ODP-value of substance x, 03 (x) is the change in total ozone at 
steady-state per unit mass emission rate of substance x, and 
03 (CFC -1 1) is the change in total ozone at steady-state per unit 
mass emission rate of CFC-11. 


DR,/DF.p 

The Human Toxicity Potential (HTP) is defined as the risk due to 
an emission flux of 1 kg*year'' of substance I relative to the risk 
due to an emission flux of 1 kg»year‘’ of a reference substance: 
where HTPj is the HTP-value for substance I initially emitted 

to compartment comp, DRj is the change in human risk at a change 
of emission flux DFj ( = Dm; / Dt) of substance I to 

compartment comp, and DR^f is the change in human risk at a 
change of emission flux DF^^ ( = Dm^f / Dt) are the same 

quantities for the reference substance ref. 


TETPuon^ = 

DR/DF. 

1 icomp 

The Terrestrial Ecotoxicity Potential (TETP) is the risk to 
terrestrial ecosystems (Rj) through an emission-flux of 1 kg of 
substance I to compartment comp ( DFp(,eno, ^ir = Dm/Dt) relative to 
the risk to terrestrial ecosystems (Rphenoi) through an emission-flux 
of 1 kg phenol to air (DFphenoi air = Dmphenoi/Dt). 

^^phenol^^Ephenol jjr 

AETP. = 

DR/DF 

1 icomp 

The Aquatic Ecotoxicity Potential (AETP) is the risk to aquatic 
ecosystems (Rj) through an emission-flux of 1 kg of substance I to 
compartment comp (DFphe„j,, = Dm/Dt) relative to the risk to 
aquatic ecosystems (Rphenoi) through an emission-flux of 1 kg 
phenol to air (DFphe„oi air = Dmph,„o/Dt). 

DR ^ ,/DF ^ , 

phenol phenol air 


(continued) 


4-12 










TABLE 4-5. STATE-OF-THE-ART IMPACT POTENTIAL FUNCTIONS 

(CONTINUED) 


Impact Potential Function 


POCP 


a/b 

71 ^ 


potential H^i/m 
APj = -i—!- 

potential 


NP. = 


/ HI 

equivaknisi' **i 
N«,u.valentsP04/mPO, 


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 equivalents per mass unit of substance I (m^) 
compared to the number of potential equivalents P®''' 

mass unit of reference substance (m^^f); SOj 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 (mj) relative to 
the potential biomass in terms of N-equivalents per mass emitted of 
a reference substance (nij^f); PO 4 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 will lead to variation among equivalency units and 
diminish the utility of a common database of equivalency units. Also, while the development of 
equivalency factors is straightforward in principle, frequently exposure and effects information 
on 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. 


TABLE 4-6. OZONE DEPLETION POTENTIAL (ODP) OF SELECT 

HALOCARBON GASES 


Gas 

ODP Relative to 
CFC-11 

Gas 


ODP Relative to 
CFC-11 

CFC-11 

1.00 

HCFC-141b 


0.11 

CFC-12 

1.00 

HCFC- 142b 


0.06 

CFC-113 

1.07 

HCFC-143a 


0 

CFC-114 

0.80 

HCFC-152a 


0 

CFC-115 

0.50 

Halon-130l2 


16.00 

Carbon Tetrachloride 

1.08 

H-1211 



HCFC-22 

0.06 

H-1202 



HC FC-123 

0.02 

H-2402 



HCFC-124 

0.02 

H-1201 



HCFC-125 

0 

H-2401 



HCFC-134a 

0 

H-23 11 




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 THE CRITICAL VOLUME APPROACH 


Chemical 

Release 

Quantity 

(kg) 

Ecosystem Threshold Levels 
(kg/L) 

Critical Volume 

(L) 

Air 

A 

57 

.001 

57,000 

B 

88 

.001 

88,000 

C 

150 

.1 

1,500 

D 

632 

.1 

6,320 

Water 

E 

126 

.01 

12,600 

F 

17 

.001 

17,000 

Soil 

G 

1,000 

.1 

10,000 

H 

161 

.01 

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. In addition, similar to the impact potentials described in Section 4.4, 
the calculations used in the critical volume approach are based on 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 many environmental and human impact areas with 
which to determine critical volume values. 

In addition, the critical volume approach only takes into account the assimilation of one 
chemical at a time. This is, the approach does not take into account the interaction between 
multiple chemical releases to the same environmental 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. 

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, the 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. Scope—the general impression of the environmental impact 

2. Distribution—the extent of the area affected 

3. Frequency and/or Intensity—the regularity and intensity of the problem 

4. Durability—the permanence of the effect 

5. Contribution—significance of 1 kg to the total impact 

6 . Remediability—relative 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 (ELV) is determined as a 
description of the impacts of the material use or emission in question. The ELV 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 

Measure 

Index 

Measure 

Index 

Measure 

Raw Materials 

(ELU/kg) 

Air Emissions 

(ELU/kg) 

Water Emissions 

(ELU/kg) 

Co 

12,300 

CO 2 

0.04 

Suspended matter 

lE-07 

Cr 

22.1 

CO 

0.04 

BOD 

0.0001 

Fe 

0.38 

No, 

245 

COD 

0.00001 

Mn 

21 

N 2 O 

0.6 

TOC 

0.00001 

Mo 

4,200 

So. 

6.03 

Oil 

0.00001 

Ni 

700 

CH 

10.2 

Phenol 

1 

Pb 

363 

PAC 

600 

Phosphorus 

2 

Pt 

42,000,000 

Aldehyde 

20 

Nitrogen 

10 

Rh 

42,000,000 

Hcl 


DDT 

10,000 

Sn 

4,200 

F 

lE-07 

PCB 

10,000 

V 

42 

Hg 

10 

Dioxin 

100 

Oil 

0.168 

Cd 


A1 

1 

Coal 

0.1 



As 

0.01 

Land 

(ELU/m^) 

Soil Emissions 

(ELU/kg) 

Cd 

10 

Arable 

2.93 

As 


Cr 

0.5 

Forested 

1.05 

Cd 


Cu 

0.005 

Residual 

0.98 

Cr 


Fe 

lE-07 

Energy 

(ELU/kg) 

Cu 


Hg 

10 

Oil 

0.33 

Hg 


Mn 

lE-07 

Coal 

0.26 

Ni 


Ni 

0.001 

Electricity 

0.014 

Pb 


Pb 

0.01 



Sn 


Zn 

0.00001 



Ti 





Source: Swedish Environmental Research Institute, 1991. 


Weaknesses 

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 

Ouantitv (ke) 

Environmental Index 
fELUl 

Environmental Load 
Value 

Raw Materials 

Oil 

1,000 

0.168 

168 

Fe 

450 

0.38 

171 

Pb 

123 

363 

44,649 

Ni 

37 

700 

25,900 

Mn 

25 

21 

525 

Energy 

Oil 

1,500 

0.33 

495 

Electricity 

20,000 

0.014 

289 

Air Emissions 

CO 2 

390 

0.04 

15.6 

NO, 

375 

245 

91,875 

SO, 

248 

6.03 

1,495 

Water Emissions 

Oil 

29 

0.00001 

0.00029 

BOD 

58 

0.0001 

0.0058 

Pb 

5 

0.01 

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, the 
EPS system is currently set up primarily to assess impacts to ecosystems. If developed and 
refined, modules for assessing impacts to human health and natural resources would enhance the 
overall utility of EPS. 

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 measured 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 baseline 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 exposure—oral and 
inhalation—Tellus 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 substance/kilogram body weight/day (Tellus 
Institute, 1992b). 

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 

Cancer Potency 

Isophorone Equivalents 

Acrylonitrile 

5.40E-01 

138 

Arsenic 

5.00E+01 

12,821 

Benzene 

2.90E-02 

7 

Beryllium 

4.30E+00 

1,103 

Bis(2-ehtylhexl) phthalate 

1.40E-02 

4 

1,3-Butadiene 

1.80E+00 

462 

Cadmium 

6.10E+00 

1,564 

Carbon tetrachloride 

1.30E-01 

33 

Chloroform 

6.10E-03 

2 

4,4-DDT 

9.70E-06 

0.00249 

1,4-Dichlorobenzene 

2.40E-02 

6 

1,2-Dichloroethane 

9.10E-02 

23 

1,1 -Dichloroethy lene 

6.00E-01 

154 

1,2-Dichloropropane 

6.80E-02 

17 

1,3-Dichloropropene 

1.80E-01 

46 

2,4-Dinitrotoluene 

6.80E-01 

174 

2,6-Dinitrotoluene 

6.80E-01 

174 

1,2-Dipenylhydrazine 

8.00E-01 

205 

Ethylene oxide 

3.50E-01 

90 

Hexachlorobenzene 

1.60E+00 

410 

Isophorone 

3.90E-03 

1 

Methylene chloride 

7.50E-03 

2 

Nickel 

8.40E-01 

215 

PAHs 

1.15E+01 

2,949 

Propylene 

2.40E-01 

62 

Styrene 

3.00E-02 

8 

Tetrachloroethylene 

5.10E-02 

13 

1,1,1 -Trichloroethane 

5.70E-02 

15 

Trichloroethylene 

l.lOE-02 

3 

Vinyl Chloride 

2.30E+00 

590 


Source; Tellus Institute, 1992b. 


4-21 





TABLE 4-11. EXAMPLE RfDS FOR NONCARCINOGENIC RANKING 


Substance 

Reference 

Dose (oral) 

1/RfD 

Xylene Equivalents 

Acetone 

l.OOE-01 

10 

20 

Antimony 

4.00E-04 

2,500 

5,000 

Arsenic 

l.OOE-03 

1,000 

2,000 

Barium 

5.00E-02 

20 

40 

Beryllium 

5.00E-03 

200 

400 

Cadmium 

5.00E-04 

2,000 

4,000 

Chromium 

l.OOE+00 

1 

2 

Copper 

3.71E-02 

27 

54 

Cyanide 

2.00E-02 

50 

100 

4,4-DDT 

5.00E-04 

2,000 

4,000 

Fluoride 

6.00E-02 

17 

33 

Hydrogen Sulfide 

3.00E-03 

333 

667 

Lead 

1.40E-03 

714 

1,429 

Manganese 

2.00E-01 

5 

10 

Mercury 

3.00E-04 

3,333 

6,667 

Napthalene 

4.00E-03 

250 

500 

Nickel 

2.00E-02 

50 

100 

Phenol 

6.00E-01 

2 

3 

Selenium 

3.00E-03 

333 

667 

Tin 

6.00E-01 

2 

3 

Toluene 

3.00E-01 

3 

7 

Zinc 

2.00E-01 

5 

10 


Source: Tellus Institute, 1992b. 


Once the carcinogenic and noncarcinogenic 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 1(X) parts xylene per million parts of air 
(ppm), and for isophorone, the PEL is 25 ppm. Converting the ppm units into milligrams, the 
100 ppm PEL for xylene translates to 433 mg/m^, and the 25 ppm PEL for isophorone translates 
into 141 mg/m . From these conversions, one 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: 

• PELs—specify 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 4-12. EXAMPLE HUMAN HEALTH IMPACT 
EQUIVALENCY RANKING 


Substance 

Carcinogens 

Isophorone 

Equivalents 

Noncarcinogens 
Xylene Equivalents 

Combined Ranking^ 

Arsenic 

12,821 

2,000 

20,231 

Benzene 

7 


22 

Beryllium 

1,103 

400 

1,854 

Cadmium 

1,564 

4,000 

4,346 

Chloroform 

2 

200 

102 

4,4-DDT 

2.49E-03 

4,000 

2,000 

Isophorone 

1 

10 

7 

Nickel 

215 

100 

373 

Styrene 

8 

10 

17 

Toluene 


7 

7 

Vinyl Chloride 

590 


1,769 


“The combined ranking assumes 1 Isophorone Equivalent = 3 * Xylene Equivalent 

Combined Rank - 3 [(Carcinogenic Equivalents) + (Noncarcinogenic Equivalents)] 
om me - ^ 

Source: Tellus Institute, 1992b. 


Relevance to Impact Assessment 

The Tellus approach provides a means for normalizing and comparing both human 
carcinogenic and noncarcinogenic substances to 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, CO 2 evolution); and 

• bioaccumulation (solubility, octanol/water coefficient, bioaccumulation factor) 

(SET AC, 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 OF TPBP APPROACH 


Life-Cycle 

Substance 

Life-Cycle 
Quantity (kg) 

Acute Toxicity 
(LC^n) 

Chronic 

Toxicity 

(NOEL) 

Biodegrada¬ 
tion (half life) 

Bioaccumu- 
lation Factor 

A 

50 


... 


... 

B 

32 


... 

... 


C 

246 





D 

65 





E 

97 





F 

32 





G 

5 





H 

43 





I 

785 





J 

324 


... 

... 


K 

17 



... 



Weaknesses 

The primary weaknesses of the TPBP is that it does not consider environmental exposure 
and can only be applied to a limited number of substances—mainly chemicals (SETAC, 1993). 

It is not clear how the TPBP would be applied to nonchemical components of ecosystems and/or 
human health; although it seems possible to identify parameter and thresholds, no attempt has 
been made to do so (Vigon and Evers, 1992). 

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/m^) using a proportionality constant, to yield a heat capacity (cal/[m^ x °C]), fugacities (/) 
can be related to concentrations using a similar fugacity capacity constant Z, with units of 
mol/m^atm by the following equation: 

C = Zf 

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 low—thus 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 
mol/m^ atm (i.e., 0.3/0.2). In air it is 40 mol/m^ 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 


Fugacity is used in the unit world approach to calculate partitioning. The unit world is a 
hypothetical 1 km box that contains water, soil, air, sediment, and aquatic biota. Mathe¬ 
matically, the unit world is represented by a set 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 of 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 assume volumes for each medium and an amount of 
solute. Medium volumes are based on unit world volumes, consisting of 1 km 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: 

• Atmosphere—accessible volume lO^^m^ 

• Soil—accessible volume lO^m^ 

• Water—accessible volume lO^m^ 

A 

• Sediment—accessible volume 10 m 

• Aquatic Biota—in lO^m^ 

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 (P^). If its molar volume is v(m^/g mol), then Z = C// = 1/P^v. The 
temperature dependencies of and v (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// = nA^P = 1/RT. In the vapor phase Z is independent of the nature 
of the substance and is usually about 40 g mol/m 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/m^ or ppm), and if the sorbent concentration is S(g/m^), 
it can be shown that Z is 10'^ KpS/H. 


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 toxicides, 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 
(SET AC, 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* 


E 

h 


i 

o 



30«< 


70* 


Volume, 


Z. 


P«) 


Amount, 

mol 


Concentration 


mol/m’ 


^9/g 


Suspended 

solids 


Sediment 


lO’o 

4.03 X 10-* 

55.0 

5 50x i0-» 

6 96x 10-* 

9x 10J 

1.90 X 10’ 

233 

2.59 X 10-* 

2,59 X 10-* 

7x 10« 

0.333 

31.8 

4.54 X 10-* 

6 82x 10-* 

35 

62.6 

0.003 

8.56 X 10-* 

1 28x 10-’ 

35 

37.9 

0.018 

5.17X 10-* 

5 17x 10-* 

2 1 X 10* 

37 9 

109 

5.17X 10-* 

5 17 > 10-* 


1000 


Temperature = 25 'C 


Fugacity = 1.363 x 10'* Pa 

Media densities (kg/m*): 

Air = 1 19 Soil = 1500 

Biota = 1000 Suspended so.'ids - 1500 

Organic carbon contents: . 

Soil = 2®* Suspended solids = 

Sediment = 4®o 


Wate' = 1000 
Sediment - 1500 


Figure 4-1. Example Output from the Unit World Model 

Source: Mackay, 1979. 


Relevance to Impact Assessment 

The unit world approach could provide 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. 

One potential problem with using the unit 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). For example, if an inventory analysis revealed that 
a system released 100 grams of NO^ 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 NO^ in each of the model components. However, without proportionately 
loading the 100 grams of NO^ into the unit world box, the concentrations produced would be 
meaningless for comparison against toxicological 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 Bamthouse 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 
Bamthouse 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 only the level of pollutants released to various 
environmental media (i.e., loading assessment), but also the ultimate fate of those pollutants. 

Canonical environmental models also have routinely been used by EPA and other 
organizations for ecological risk assessments. Although 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. 

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 “impacts,” 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. 

Canonical environment models also may 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 groups—the EPA Risk Assessment Forum and 
the National Academy of Science (NAS) Committee on Risk Assessment Methodology—have 


4-32 


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 



Figure 4-2. Conceptual Framework for Ecological Risk Assessment 

Source: EPA, 1992a. 


4-34 




































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 ASSESSMENT 


Site-specific exposure/effects assessment can be accomplished through using traditional 
risk assessment methodology, which includes the following four components: 

1. Hazard Identification; involves gathering 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 even 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 in experimental animals also likely to be so in 
adequately exposed humans?). 

2. Dose-Response Assessment: involves 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: involves 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 

dose • potency 

The excess (or increase in) 
probability that an individual will 
experience a specific adverse effect as 
a result of exposure to a risk agent. 

Population Risk 

(individual lifetime risk) • 

(population exposed) 

The number of cases resulting from 
one year of exposure, or the number 
of cases occurring in one year’s time. 

Relative Risk 

(incidence rate in exposed group) ^ 
(incidence rate in non-exposed group) 

The risk in the exposed population 
compared to the unexposed (or 
differently exposed) population. 

Standardized 
Mortality or 
Morbidity Ratio 

(incidence rate in exposed group) -r 
(incidence rate in general population) 

The number of deaths or cases of 
disease observed in an exposed group 
divided by the number expected. 

Loss of Life 
Expectancy 

(individual lifetime risk) • 36 years 
where 36 years = average remaining 
lifetime 

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 human health risk assessment is that it provides the 
most relevant understanding of the existence of, or lack of, chemical-based impacts to human 
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. 

Weaknesses 

The assumptions regarding the shape of the dose-response curve (e.g., linear versus 
nonlinear) and the existence of thresholds below which no impact occurs can have a dramatic 
effect on the final impact level. For example, no 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. 

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 



iji^ancl oi sonay^hii 

8 riDcilu^ h»i*j '^r4 arit ;;viJCiaB : rov,'r^ 

^ ;l4fitil’'^irr.'^ iLfirx 4 •. b^aDir^mooMin 

aid t<» fco^iirfchnd 

?inrt»MW rti '*M "K •> ;ku ♦ -' v k . .eftoft^ 


a 


' ^*4 S i# 

t r 


;lan riiU:^«Axai4‘^l»^ bf^u Mrshv^in hnu .ij%c>-iJ'm >*1 yiii %iii :i«oUv^ <y ybtnJ* A 

eTui ivi ■^-: !■« ..'!• .’4 )^§.ffL, ■ ‘ .■• .4.1 ' ■ V'.TiirrjitfcW-ytt 



•(* V •^' ' Ml 


.7 :■ 






biifi jcl5w^r?^•5^^ui lb*' • ii H.*>iir4*-V NvaJ x-h • ’ 

•*; (wv. A.. ‘ * -.*. 5. •■- *♦. ■ V -4 -:r.t, T^irni.Ui;KhciJtJ-' 

.?t>ofUiri * 'Xw * y iii vy- ?.:? rz ♦♦i-HfK '>rvf»i8ovnirT»oii:so9f<sti * ^ ^ 

nipw;. «. • ** • V v'. . , > a'i^*Tfe.'4; 


■Mt' 


f_*|i«w,c t». Pr.-tiiJ 

IMj# t’n »*:f frl'T i** < “ '• •»C 






I' 


I 

f 

^ * 




U '', flo^. -‘0 v‘' t.iPi* .,#/«. 4^ tf riN!{ ',*• 

Tui r!>y ! -'nf,i»W 4*. r- fMiV* UviM Hi iWKfri/:: as ^ . Hi.< 

<-‘ - ‘L'-:.kv * a‘,t=-. h. -bnMilpfrn«'* ^X' rw Cf4Uift t*Oftigf>5^ 

■ Sjf, 

- f 4’i ‘laye-'a^i \t |4, m»i,/ * ’I'orrcv'^ v i^c^vr/id fittu f -1 .v^orc^t .(' 

' ' * »*» .#1 #' * ^ ^ 


.iHfV^iB, *t. M*.v;, ,8^. sumyi. ,;ij 

k. 

# fkis, <#N f'smj.'.., If ^uch 


.;i 

ijt 


irt^M, uX » . . * 4 

c<^Atin n'>;i uf ; <v^ 



^ W 4 rt ■> Or *1 ■ '-v. %bidl) U *1 vt«y l 



*. loxic iigP ' I * - dir fd • 

fii'V *’n 

CmUi^ i'> A *• •- ' MMiMlf 

,/i .'r *4 / f 

W fiT «Wl' *''.' .<M I . . ■ -I 


Mlnfic/ I N# • '-4 .rv.i^.- tu uTpc^U'^. s ^'<•4 Wn ^\i 

\ 4 ^*->71 .» 4 \^»»i/*r4 IV <hr »l,tnU t ^ H 

i 

I# %tfr m ^ i8|toa:^i isfgm^ llwodaCC b»i 
iiXTiimMCiaii *4 frrciii.; .j . ^.->'X 

<»i 3 bc«r ritit «v 



oiJise 




1 * f 

,. ™ * ’ • *• . ■ ■ <1 

*L. '*. -.'-L 

/•■•i;L(Li..:a IV'J ■ " 



'■flj 









CHAPTER 5 

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 resources—that is, impacts to the supplies of natural resources 
(e.g., contamination of water supplies)—this chapter includes those methods to characterize 
natural resource depletion only. In the context of LCA, resource depletion has traditionally been 
addressed in a different manner than other types of impacts and thus has been separated from 
other methods of impact assessment as described in Chapter 4. 

5.1 RESOURCE DEPLETION: KEY TERMS AND CONCEPTS 

The term natural resources, in the context of LCA, refers to any component directly or 
indirectly derived from the natural environment (EPA, 1992c). Natural resources provide the 
basic raw materials for most production systems. They can be used as inputs to production 
systems in the form of raw materials or energy, or they can be altered or their value diminished 
as a result of outputs from a production system (EPA, 1992c). The complete life cycle of 
resources is illustrated in Figure 5-1. 

Natural resources are typically classified as either stock or flow resources. This section 
discusses the distinction between stock and flow resources as it relates specifically to impact 
assessment. This distinction between stock and flow resources is important to consider because 
the procedures for characterizing impacts to each of these categories may be slightly different. 

Stock resources include those that cannot be replenished through natural processes on 
time scales relevant to human societies (SETAC, 1993; EPA, 1992c). Examples of stock 
resources might include fossil fuels, mineral ores, surface and ground water, and soil. Stock 
resources are typically considered finite. In contrast, flow resources include those that can be 
readily replenished either by natural or artificial processes (EPA, 1992c). Examples of flow 
resources include most flora and fauna (e.g., trees, fish, wildlife). 

Key issues associated with considering stock and flow resources in impact assessment 
include, but are not limited to, the following: 


5-1 


Replenishment 

planting trees 
planting grain 



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 reserves. 


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 resources—both 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 


consequence an increasing amount of the “wealth” we create is in a sense stolen 
from our descendants. 

3. The maintenance of a livable global environment depends on the sustainable 
development of the entire human 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. 

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 so—such 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 will effectively apply the motivations (Ruckelshaus, 
1989). 

From an ecological point of view, a necessary (but not sufficient) condition for 
sustainable development is maintaining an adequate environmental resource endowment. This 
endowment constitutes the natural capital (assets) necessary to provide needed and wanted 
environmental services—such 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. 

5.3 RESOURCE DEPLETION MODELS 

The resource depletion models described in this section are “time-metric” models. Such 
models are based on the basic principle that the quantity of stock or flow resource reserves 
(R—in 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 at time (q): 

dR dR 

Resource Utilization Rate = — = dR ^ 

dl, 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 impact (i.e., the lower the magnitude of the t- 
value, the less tractable the impact). 

Weaknesses 

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 to its original level of value or quality—at 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- 
metric models 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 analytical approach used 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 ConsumpUon Ratio = Consumpdon 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 unit—usually 
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 ou^uts 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 

Input Quantity 
(tons/annum) 

Supply/Reserves 

(tons/annum) 

Resource Consumption 
Ratio 

Timber 

150,000 

2,600,000,000 

5.7E-05 

Oil 

2,500 

150,000,000 

1.7E-05 

Coal 

200 

500,000,000 

4.0E-07 

Natural Gas 

575 

37,500,000 

1.5E-05 

Iron Ore 

1,350 

450,000 

3.0E-03 


5-7 





Significant efforts may be required to develop resource supply and exposure information 
for this approach, and it is not clear whether calculating resource consumption ratios for 
individual products or processes or the incremental total demand for the resource will be 
necessary. In addition, the significance of the 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 or 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) _ ^ 

(MfT) 

where M is mass and T is time. M represents 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) 

(M/T) 


+ (MTT) = (T) 



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 quality—at 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 



Stock 

Flow 

Local 

Sl / Cl = Ul 

Cl / Rl = Dl 

if Dl < 1, then Cl - Rl = 
and Ql / El = Ul 

Global 

Sq/Cq =Uo 

Cq / Rq ” 

if Dq < 1, then Cq - Rq = Eq 
and Qq / Eq = Uq 

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 (amount/unit of time) 

Q - standing quantity (amount) 


Figure 5-2. Resource Depletion Matrix 

Source: SETAC, 1993 


5-10 








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: The decisionmaker(s) is asked to identify overall objectives. For 

LCA these might be to minimize overall environmental and human 
health impacts or to maximize public opinion. 

Bottom-up: 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 have certain characteristics. An overall objective 
would be at the top and a comprehensive set of 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 

• comprehensive, 

• as small as possible in number, 

• nonoverlapping, 

• judgmentally independent, and 

• operational. 

Decision analysis with multiple issues or 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 measure the degree to which an objective is achieved by a 
management option (attributes should be relevant to the issue, measurable, 
predictable, comprehensive, and 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 



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 often faced with decisions that can cut across 
multiple environmental media (e.g., air pollution, water pollution, solid waste, resource use). 
MAUT provides a framework for breaking such multi-attribute decisions into a set of 
measurable attributes from which the analyst can develop a multi-attribute utility function. This 
multi-attribute utility function can, under favorable conditions (see Keeney and Raiffa, 1976), be 
broken down into single attribute utility functions, which can then be combined in a 
multiplicative or additive manner according to the values of estimated scaling coefficients 
(SETAC, 1993). 

Weaknesses 

The primary weakness of MAUT is that it is very difficult to implement because of some 
of the following characteristics: 

• determination of the appropriate utility function to employ, 

• decomposition of the multi-attribute utility function, 

• derivation and use of multiplicative functions to combine the single attribute utility 
functions, and 

• estimation of scaling coefficients. 

Because of limiting characteristics such as those listed above, the MAUT process has 
been simplified and refined in “spin-off’ methods such as the Analytic Hierarchy Process (AHP) 
described in Section 6.2. 

Relevance to Impact Assessment 

In the context of impact assessment, MAUT could be used to scale predicted impacts on 
a 0 to 100 utility scale, multiplied by the importance weights, summed, and then compared to 
identify the maximum utility management strategy (SETAC, 1993). However, the subjective 
nature of the scaling process is open to considerable debate, and analytic difficulties of the 
scaling process may limit whether scaling may be accomplished at all. Therefore, the most 
practical application of MAUT for purposes of impact assessment may be for decisionmakers to 
consider, separately, the importance weights and the impacts evaluation (SETAC, 1993). 

6.2 AHP 

The AHP is a systematic procedure for demonstrating a problem in a hierarchical 
structure, based on the values of the decisionmaker(s). The AHP organizes basic reasoning by 


6-4 


decomposing a problem into its constituent parts and then using simple pairwise comparisons to 
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 levels—one 
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 = 1^^^ w is determined. 

The consistency index uses the departure of 1^^ 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 


revising the manner in which questions are asked in making pairwise 
comparisons. If this fails to improve consistency, it is likely that the problem 
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. 

9. To perform absolute measurement that preserves the rank of the alternatives and 
satisfies expectations and prior commitments, each lowest level subcriterion is 
divided into a complete set of intensities so that an alternative always reflects one 
of these intensities. 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 process 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). 

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 the 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. 

Strengths 

The main strength of the AHP is that it 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 




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 

The AHP may provide a useful tool for evaluating multi-attribute, complex problems. 
Such problems typify those encountered in the valuation phase of impact assessment where, 
foreseeably, a wide variety of multi-media and/or cross-sector environmental impacts must be 
considered. The AHP also provides a useful framework for integrating stakeholder values with 
environmental impacts. 

6.3 MODIFIED DELPHI TECHNIQUE 

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 the 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 
Linstone 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 
of—for example, 95 percent—based 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 





Ecology 




.85 


I I iiw uwi^oi IV I vv/ 

55% as important as ecology in th 
context of this proposed action. 



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 e)^erience and judgment of the 
participant. 


Figure 6-3. Modified Delphi Technique 

Source: Modified from Jain et al., 1993. 


The formula for weighting the technical specialties is 



(i = 1,2,3,...,n) 





(i = 1) 

(i = 2,3,...,n) 


where 


Wjj = weight for the ith technical specialty area by the jth scientist, 
n = number of technical specialties, 


6-9 






















P = 1,000: total number of points to be distributed among the technical specialties, 

Xjj = the jth scientist’s assessment of the ratio of importance of the ith technical 

specialty in relation to the (i-l)th technical specialties, and 

Vjj = measure of relative weight for the ith technical specialty area by the jth 
scientist. 

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. 

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. 

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 

After Interdisciplinary Study 


Average Point 


Average Point 

Area 

Distribution 

Area 

Distribution 

Water 

125 

Water 

128 

Air 

122 

Air 

126 

Natural Resources 

109 

Natural Resources 

105 

Health 

100 

Ecology 

93 

Ecology 

97 

Health 

88 

Land Use 

81 

Earth Science 

87 

Earth Science 

79 

Land Use 

78 

Economics 

62 

Sociology 

64 

Sociology 

60 

Noise 

62 

Transportation 

56 

Economics 

62 

Aesthetics 

54 

Transportation 

61 

Noise 

53 

Aesthetics 

46 

TOTAL 

1,000 

TOTAL 

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 


One of the weaknesses of the modified Delphi technique is one that typically plagues 
most valuation tools—namely the requirement of expert knowledge with which to rate 
environmental attributes. In many cases, there is no such expert knowledge. In addition, the 
possibility exists that the experts can have a bias and/or may misjudge the importance of 
particular attribute intensities. 

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 modified 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 of 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 COSTING* 

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. 

Monetary values for environmental impacts 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 difficult. For example, translating 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 


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, a comprehensive estimate of society’s 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 terms 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. 

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 assessment—the EPS Enviro-Accounting Method 
outlined in Chapter 7—provides an example of the use of economic valuation in the context of 
LCA. 

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 


6-14 


attributes of a product. A product has many attributes, some of which can relate to the 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 people have 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 absence of a wide array of choices, people are likely 
to base their decisions on characteristics other than the environment. 

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. Data 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 be more useful in damage cost studies if they could 
indicate the value that people ascribe to their lives. 

6.4.2 Contingent Valuation 

Contingent valuation assumes hypothetical (contingent) markets. In essence, it consists 
of experiments in which people are 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, 1991). 

Understanding the change in environmental conditions consumers are asked to evaluate is 
important. Two concepts are suggested in the literature: willingness to pay (WTP) and 
willingness to accept (WTA). Lx)osely 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 money 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 have 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 fu'st 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 
SO 2 to enter the atmosphere is smaller the more SO 2 has already been controlled. The 


6-17 


negative side of this relationship is that 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 
the marginal cost of abatement equals the marginal benefit of abatement. Setting such a standard 
would require an efficient allocation of resources for pollution abatement activity. To do more 
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. Controls—not 
standards—are administered for some pollutants; 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 one-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. 

Another problem is that regulations for all 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 7 

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 MATRIX (lAM) 

The lAM is an exploratory, qualitative, expert-based approach to impact analysis that 
builds on the results of an inventory analysis. The LAM approach is described below by 
explaining its development and initial use. 

The lAM 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 lAM allowed for the direct evaluation of the relative environmental 
burdens of a particular application of a halogenated solvent and its altemative(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 CO 2 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 lAM process entails convening a group of experts to carry out the following steps: 

1. Identify appropriate impact categories. The lAM study of the two TCA 
substitutes consisted of five 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 

• transportation effects 

(It should be noted that in applying the lAM 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 corresponding cell in the lAM was eliminated. As a result of 
this step, 17 of the 35 LAM cells were eliminated. 

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 to 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 that the weighting may or may not be restricted to a single 
impact category, depending on the views of the expert panel. However, the basis 
for assigning weights and the scope of comparison within and across impact 
categories should be made transparent.) 


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 lAM 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 lAM 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 

26.6 

0 

0 

0 

Aqueous cleaner 

0 

0 

2.7 

0 

Material Inputs 

Trona deposits, salt, sand 

0 

1.2 

0 

4.5 

Crude, natural gases 

0 

2.9 

0 

0 

Energy Inputs 

Power or Fuel (per million BTU) 

520 

1,530 

1,730 

1,800 

Atmospheric Emissions (tons) 

Cl-HC, HC/particulates, CI 2 

0 

2 

0 

<0.1 

TCA 

21.6 

21.6 

0 

0 

Particulates 

0 

0 

0 

<0.1 

Water vapor 

0 

0 

288 

288 

Aqueous Wastes (tons) 

0 

682 

1,822 

1,822 

Solid Wastes (including spent catalyst. 

0 

4.9 

0 

0.3 

solids/ sludge, used oil, and shale—in tons) 

TCA and oil (from OTVD) 

6.5 

6.5 

0 

0 


Source: Source Reduction Research Partnership, 1991. 


7-3 







User-Level Impact Analysis Matrix 


impacting Parameters 

Ecosystem Impact 
Categories 

Material 

inputs 

Energy 

Inputs 

Air 

Emissions 

Aqueous 

Wastes 

Solid 

Wastes 

TOTAL 

Global Warming 


+1 

-1 



0 

Ozone Depleting Potential 



-1 



-1 

Stock Resource Use 

-1 

+ 1 




0 

Air Quality 


+1 

•1 

+1 

-1 

0 

Water Quality 


+1 


+1 


+2 

Land Disposal 


+1 


+ 1 

-1 

+1 

Transportation Effects 

-1 

+1 



-1 

-1 

TOTAL 

-2 

+6 

-3 

+3 

-3 

+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. 

Figure 7-1. User-Level Impact Analysis Matrix for Ecosystem impacts 

Source: Source Reduction Research Partnership, 1991. 


Global-Level Impact Analysis Matrix 


Impacting Parameters 

Ecosystem Impact 
Categories 

Material 

Inputs 

Energy 

Inputs 

Air 

Emissions 

Aqueous 

Wastes 

Soiid 

Wastes 

TOTAL 

Global Warming 


0 

-1 



-1 

Ozone Depleting Potential 



-1 



-1 

Nonrenewable Resource Use 

-1 

0 




-1 

Air Quality 


0 

-1 

+ 1 

-1 

-1 

Water Quality 


0 


+ 1 


+ 1 

Land Disposal 


0 


+ 1 

-1 

0 

Transportation Effects 

-1 

0 



-1 

-2 

TOTAL 

-2 

0 

-3 

+3 

-3 

-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. 

Figure 7-2. Global-Level Impact Analysis Matrix for Ecosystem Impacts 

Source: Source Reduction Research Partnership, 1991. 


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 lAM 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 
LAM 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 lAM 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 lAM 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 Stage 


Impact Category 

Raw Materials 
Acquisition 

Manufac¬ 

turing 

Use/Reuse/ 

Maintenance 

Recycle/Waste 

Management 

Global Warming 

M/I 

M/I 

M/I 

M/I 

Ozone Depleting Potential 

M/I 

M/I 

M/I 

M/I 

Nonrenewable Resource 
Utilization 

M/I 

M/I 

M/I 

M/I 

Air Quality 

M/I 

M/I 

M/I 

M/I 

Water Quality 

M/I 

M/I 

M/I 

M/I 

Land Disposal 

M/I 

M/I 

M/I 

M/I 

Transportation Effects 

M/I 

M/I 

M/I 

M/I 


Scale Ranges: M - 1 to 10 1 = lowest magnitude of impact, or lowest level of importance. 

I - 1 to 10 10 = highest magnitude of impact, or highest level of importance. 


Another weakness of the lAM 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 lAM 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 lAM 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. 

Relevance to Impact Assessment 

The lAM approach may provide a relatively simple, quick, and useful means of 
qualitatively comparing the environmental implications of two or more alternative systems 
without having to characterizing impacts. The more qualitative nature of the lAM 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. 


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 and 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: 

• a large gap exists between rich and poor in terms of disposable resources for 
environmental care, 

• the needs of today often outweigh 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 may be similarly problematic. 

7.3 INTEGRATED MANUFACTURING AND DESIGN INITIATIVE (IMDI) 

ENVIRONMENTALLY CONSCIOUS MANUFACTURING (ECM) LIFE-CYCLE 

ANALYSIS 

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 two primary responses: 

1) “Identify environmental impacts of activities related to manufacturing, use and 
disposal;’’ and 


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. 

• Economic analysis is based on the neoclassical model of the closed economic system. 

Environmental 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 United 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 resources—biophysical, human, infrastructural, and 
monetary—into calculations of national accounts, productivity, and policies for 
development and investment planning. 

• Pollution can be considered 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. 

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-technical 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 society and nature can be considered a “positive sum game.” 
Human activities are organized to be synergistic with ecosystem processes and 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. 

Deep Ecology 

• 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 solely on environmental quality and environmental 
impacts and thus is somewhat limited in its intended application to environmentally conscious 
manufacturing because additional factors (e.g., cost, functional requirements, performance) also 
contribute to decisions 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. Because of the subjective nature of weighting process used 
in the IMDI approach, its use would 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 also 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, 

• cost effectiveness, and 

• relevance to decisionmakers (Canadian Standards Association, 1992). 

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 Figure 7-3). 


7-12 


Environmental Impact 
+ 

i t 


Environmental Costs 
Economic Savings 


Environmental Benefit 
EcorK>mic Savings 


Quadrant where system 
options would ideally be 
mapped 


Envrironmental Yield 




Environmental Costs 
Economic Costs 



Environmental Benefit 
Economic Costs 




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 CO 2 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 






















Relative weights are then assigned to each environmental theme to enable the 
aggregation of environmental impacts associated with each process option. The environmental 
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. 

Strengths 

The main strength of the integrated substance chain management approach is that it 
provides a framework and methodology for integrating environmental concerns, economic 
concerns, and stakeholder values. The resulting options map portrays the environmental and 
economic differences between process 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 map. 

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. 

7.5 ECO-RATIONAL PATH METHOD (EPM) 

EPM represents a procedure that builds on the ESR method described in Chapter 4 to 
integrate environmental and economic information—two of the primary dimensions of 
environmental decisionmaking. The process for integrating these two dimension comprises 
three main steps—recording. Judgment, and decision—as shown in Figure 7-4. 


7-14 


Decision Judgement Recording 



Ecoiogicai 


Dimension 


Economic Dimension 



Figure 7-4. Conceptuai 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 CO 2 and another system releases more NO^, then 
no obvious and objective judgments are 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. 

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 may 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. 

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 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 

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; 


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 


f i' 


m^' 




W 


. « ■ ‘ 


r . t. 


\\ 

I 


*S‘ 




'V 

. V. 


fV 


ii; 

If*, 

;’j 


f', <r- 






til 


monous hfl*>isp.-i!«ifn»riigHvM! ;ti m! it..oPrt 

*r. ' ■" » 

H^?(r f ♦ .1> intcwma rMw* invc!:.-ury 

nSl>W)M Mara ^Hwwctnj^aTTw^icin. 


^nnon 



Wyjqmi aJ/rtr. 


.lm..i'SS‘*oMif.''M|ij,'.i, ^•i;;<'i?''J t ”' ■' 'i lUtt ■*“4, (he fust Step (scc_''< 



.v,.4 ,;f<« % .A'^ ' ^"] 

. ,V I - j 


oi m^Jrf^ 


^ . r-^ - y — mr-v-'o^w - ,| - , . f * a«aeo n Tiw Tjm^yiyjr*^*^ ^ 


if* 'UxXi-iii* T!u 4 v\iiiCtuU.ijL<o«4 pt^tyrri^ii i itvjxtt^ <if »it: cU«af«aN^«>i<jifrji,^^ i 

of i/p, * r ' K;rti 4tw^, !'i‘5lS*«c>i. tr, t;(W!rhit tij^- ^MelAr&^ilc ; AxJ<f0i pfooe^ \rc 
<1v.w %^iA »fnv rfiiiii$mj<A ■; t m^f|gi(i e., fn‘l{ mj/tjM r«tio>. 


SlnrfffM 



I 1 


-i 


$(im (if thr tiU , 'g’iii of .iKb«4r Ottf l*iMwi^<. 


•j 


•* km’ntVf ■f' $l^^'*iv:.^,pa>^' 'rX;f m pi: mmprrr* C4T‘i- imrkfi\mi ’jyL-iXfiUt adc 


• ,gk uwa 'v-i 'i' f ■' ktti itj£a^--v-fjt‘'Kten^.i4 fixifTir' j ‘o.-ijii. poIuStaL 




ftpihi&y, tpit w r(0»iu 




■■ vJ i 


11 


n-r 


;.i.«^ j^,,. 


[fc., 

. li 


'.(li 




rjtlfi 



.. &• '^- I JS'l 








CHAPTER 8 

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 cannot readily be used to assess the impact of nonchemical 
loadings. 

• There are a number of places in the impact assessment where value judgments may 
play a significant role. It is 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 how 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 LCA 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. 

8.2 POTENTIAL FUTURE RESEARCH NEEDS 

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 Low 

• Define impact descriptors for LCA applications Low 

• Determine basis for defining stock resource pool Low 

• Develop impact category equivalency factors Moderate 

• Identify ecohazard profile parameters, thresholds Moderate 

• Develop method for evaluating depletion of water resources Moderate 

• Develop method for linking resource development with depletion Moderate 

• Achieve international consensus on impact assessment Moderate 

• Assess feasibility of nonchemical impacts matrix Moderate 

• Prepare impact analysis technical support document Moderate 

• Develop and validate of streamlined impact assessment methods Moderate 

• Develop a reference data base of generic impact assessment information High 

• Develop library of impact networks High 

• Prepare broad range of impact assessment case studies High 

• Develop methods for factoring uncertainty into impact assessments High 

• Determine feasibility of resource management/economic models for LCA High 

• Develop methods for estimating biodiversity change and habitat alteration High 

• Develop models to assess susceptibility due to health stress High 

• Develop better human exposure models within an LCA High 

• Evaluate ecological risk assessment models/methods High 

• Develop/validate ecological hazard matrix approach High 

• Fill data gaps in the following areas: High 


— Health exposures 
— Short-term and long-term bioassays 
— Effects of unintended product use 

— Exposure from nonmanufacturing 

— Nonpoint sources of pollution 

Source: Modified from Vigon and Evers, 1992 


8-3 












waJ 

“fOj 

alfirtsboK; 

^KtnR^jKiM 

r 

3J«ni>bcM 


'1 


ckClsiCHl to 1^ nudfi. 

in Vi^JiyT ^T,r • r •y?y -»<■ i if w *' '■ "^l y " " ••n* Tjr'^T^*''''TPtai io«jScj;.j 

iS.'i’jaa&ia* mrfwid.». 

ftJT cJijir- ieh.:^f i! !• "’•' wjii.w* 

i<x.’j»s«tR<ih^t'j|0<ul-4i)<Aii|»iid9niun<.»na • • 

• nmrcafi»ijim ,.< ini.i^ i--. r>.ikhH^^‘ 

ifUy a j*: ^. Ic ’ wq! m the kiiTOfCt. 

iMC &p5ictr'' 

Ji Niaia,'ftoiicl<y5b Wiv/ Ifi9(^u|0k^ ^nnlaif yn?l hoHitun f\<ih>f^ • 

• A MAiimciMU 

to inT|VWDii^i'»3k;i?g^l# Iti viruu^iK ill >!( 

«V4* ri^mtdrch Al^hcsapiv;Ji':ritbny ^ 



» 

» 


I 

c ihe 


ff^tii L,CA %o/r>j>>’n>cf»i^^ :i js ^j 

ftiuH »; aiKl c«rtoct p • tnme?' :‘.%Hiiv^t th i>:. ..--.il i^ reoi> jii« • 

/IviH 

rfa»w 


imhjc^ bet.. f .u*-- a,a I 

n«j:>i* ■‘■'I'. li},f U^«; wJ^fvfS tr- ■.. ._ 

;{frtt'V<^ WirtBwir^^Tiwn uf 




TOTKNrUL 





K', 


J^Mr-Axt^} t /-: 

A jJ ns Mo»-^x)jf» os/bi^'ieuiui ' 4 'A^v*.<l 

iiSiH «*w h W««*» the t < A 

tIsW(Ttipnisdlj' >«j(ati]int »>« fta-^T ifciM^s-itaiqrtixipVHt'tyXw w... ■ ^»*W» <t' twJs vnd 
•isiijn^, .)■ wmc timed jo fw^te jt-i uMf* jin »f(>i(oi vti m »mv nH • 

?a‘<rjn<j.,;> irU-jH — 
s vmf^inhiji Ui* —* 

%'tt i^hUj »? b^besum*! >0 _ 

fltr/u — 

. noin/ltm 1-1 •4ga<tt^ Iftwnon — 

»’^l ,rt9v3 \,nt iioj(iV nwrft bdjtieuM :sc/v/oi 


^ Ic : 

, » 

/ .- t ‘ 

<V' 


1 












Appendix A 

National Environmental 
Policy Act (NEPA) 
Environmental Assessment 

Procedures 



f. ' 

I’ '".f * : ' 


-h 



* ■'■*•• 
v' ^ ■'■• 



>t 

1 

Ijfr 


‘ I 


i;.t 




1 









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.l 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-1 


TABLE A-1. 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 

• Acid and alkali 

• Biochemical oxygen demand (BOD) 

• Dissolved oxygen (DO) 

• Dissolved solids 

• Nutrients 

• Toxic compounds 

• Aquatic life 

• Fecal conforms 

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: Jain et al., 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 after agency 
decision to prepare an EIS. 


A-2 










TABLE A-2. ENVIRONMENTAL ATTRIBUTES USED IN EIA 


Attribute 

Variables to be 

Measured 

Data Sources 

Mitigation of 

Impact 

Air 




Diffusion Factor 

• Stability 

Primary sources of data are 

Mitigation techniques have 


• Mixing depth 

the National Weather 

not been adequately 


• Wind speed 

• Precipitation 

• Topography 

Service and the United 
States Geological Survey 
(USGS). 

defined. 

Particulates 

The concentration of all 

Data sources include state 

• Source reduction 


solid and liquid particles 

pollution control 

• Reduction or removal of 


averaged annual arithmetic 

departments, county air 

receptors from the area 


mean of all 24 h particulate 

pollution control offices. 

• Particulate removal 


concentrations at a given 
location. 

multi-county air pollution 
control offices, or city air 
pollution control offices. 

devices 

• Use of protected, 

controlled environments 

Sulfur Oxides 

The 24 h annual arithmetic 

Data are generally 

• Source reduction 


mean concentration of SO 2 

compiled and published 

• Reduction or removal of 


present in the ambient air. 

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. 

receptors from polluted 
areas 

• Gas removal devices 
using absorption, 
adsorption, and catalytic 
converters 

• Use of protected, 
controlled environments 

Hydrocarbons 

The 3 h average annual 

Data are generally 

• Control of motor 


concentration of ambient 

available from state air 

vehicle emissions 


hydrocarbons, expressed in 

quality monitoring 

• Control of stationary 


ppm, and measured 

programs. Other potential 

source emissions 


between 6 and 9 a.m. (peak 

sources include the EPA 

• Reduction or removal of 


hydrocarbon concentration 

and city or county 

receptors from area 


time). 

monitoring agencies. 

• Use of a controlled 
environment 


(continued) 


A-3 






TABLE A-2. ENVIRONMENTAL ATTRIBUTES USED IN EIA (CONTINUED) 


Attribute 


Variables to be 
Measured 


Mitigation of 

Data Sources Impact 


Air (continued) 
Nitrogen Oxide 


Carbon Monoxide 


Photochemical Oxidants 


The average annual 
concentration of nitrogen 
oxides in the ambient air, 
measured in ppm. 


The maximum 8 h and 1 h 
concentration of carbon 
monoxide measured in 
micrograms per cubic 
meter. 


The maximum 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 

Variables to be 

Measured 

Data Sources 

Mitigation of 

Impact 

Air (continued) 




Hazardous Toxicants 

The variable to be 
measured varies with the 
toxicant. 

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. 

• 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 

Odors 

• 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. 

• 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 

The amount of water 
withdrawn in a unit of 
time, usually expressed as 
thousands of acre-feet of 
water per annum. 

Sources of data include 
local uses 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. 

Flow Variations 

• The typical unit of flow 
measurement is cubic 
feet per second. 

• Velocity as measured in 
feet per second. 

Data sources include local 
Army Corps of Engineers 
offices and state water 
agencies. 

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 


Variables to be 
Measured 


Mitigation of 

Data Sources Impact 


Water (continued) 
Oil 


Radioactivity 


Suspended Solids 


Thermal Pollution 


Acid And Alkali 


Quantitative: 

• milligrams of oil or 
grease per liter of water 

Qualitative: 

• visible oil slick 

• oily taste/odor 

• coating of banks or 
bottom 

• The quantity of any 
radioactive material in 
which the 
disintegrations per 
second are 3.7 x 10*®, 
expressed as Curie (Ci) 

• Microcurie (10'®Ci) 

• Picocurie (10‘*^Ci) 

• Readily settleable 
suspended solids are 
measured in milliliters 
per liter of settled water. 


Water temperature 
measured in degrees 
Centigrade or Fahrenheit. 


pH 


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 uses offices, local 
Army Corps of Engineers 
offices, and state water 
agencies. 


Sources of data include 
local uses offices, local 
Army Corps of Engineers 
offices, and state water 
agencies. 

Sources of data include 
local uses 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 


• ControllingAreatment 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. 

Sources of data include 
local uses offices, local 
Army Corps of Engineers 
offices, and state water 
agencies. 

• Treatment of all wastes 
containing organic 
material: 

- biological 

- chemical 

- packaged units 

Dissolved Oxygen (DO) 

Milligrams of oxygen per 
liter of water. 

Sources of data include 
local uses offices, local 
Army Corps of Engineers 
offices, and state water 
agencies. 

• Treatment of all wastes 
containing organic 
material: 

- biological 

- chemical 

- packaged units 

Dissolved Solids 

Total dissolved solids, 
determined after 
evaporation of a sample of 
water and its subsequent 
drying at 103°C. 

Sources of data include 
local uses offices, local 
Army Corps of Engineers 
offices, and state water 
agencies. 

• Controlled landfilling to 
avoid possible leaching 

• Deep well injection of 
brine 

• Control and treatment of 
surface runoffs 

Nutrients 

Includes measurement of 
phosphorus, nitrogen, 
carbon, iron, trace metals 
in their appropriate terms. 

Sources of data include 
local uses offices, local 
Army Corps of Engineers 
offices, and state water 
agencies. 

• Waste water treatment 

• Natural assimilation 

Toxic Compounds 

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. 

Sources of data include 
local uses offices, local 
Army Corps of Engineers 
offices, and state water 
agencies. 

• Monitor and control of 
all toxic wastes 

• Dilution 

Aquatic Life 

Field observations 

Data may be obtained from 
local Fish and Wildlife 
offices. 

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 
in terms of coliform per 

100 mL. 

Sources of data include 
local Army corps of 
Engineers offices and state 
water agencies. 

• Treatment of all wastes 
containing organic 
material: 

- biological 

- chemical 

- packaged units 

Land 

Soil Stability (Erosion) 

• Soil composition 

• Degree of slope 

• Length of slope 

• Nature and extent of 
vegetative cover 

• Intensity/frequency of 
exposure to eroding 
forces 

Data are generally 
available from local U.S. 
Soil Conservation Service 
offices. 

• Erosion control devices: 

- ground cover 

- tile drainage 

- grassed waterways 

- terracing steep slopes 

- catch basins 

Natural Hazard 

Specific to each type of 
hazard. 

Sources of data include the 
Corps of Engineers, 
uses, U.S. Forest 

Service, National Weather 
Service, state geologists, 
and local universities. 

Specific to each type of 
hazard. 

Land-Use Patterns 

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 

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. 

• 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 


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. 


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 


Aquatic Plants 

• 

Population 

Data sources include the 

• Minimize waste and 


• 

Number of species 

U.S. Fish and Wildlife 

nutrient inputs 


• 

Habitat (in hectares) 

Service, wildlife experts. 

• Restrict drainage of 


• 

• 

• 

• 

• 

• 

Human intrusion 
pH 

BOD 

DO 

Coliform bacteria 

Pesticide concentrations 

and universities. 

wetlands 

• Creation of National 
Parks, National Wildlife 
Areas, or other 
protected areas of 
habitat 


(continued) 




A-10 






TABLE A-2. ENVIRONMENTAL ATTRIBUTES USED IN EIA (CONTINUED) 


Data Sources 


Mitigation of 
Impact 


Attribute 


Variables to be 
Measured 


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-11 








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 defmed. The 
objective is to identify 
general changes in social 
activities that 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 

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. 

Data on this attribute can 
be generally obtained from 
psychologists, personal 
surveys, local counselors, 
clergy, and law 
enforcement officials. 

• Including an action plan 
that would provide 
assistance for affected 
individuals 

• Consultation 

• Social programs 

Physiological Systems 

No variables can be 
measured for this attribute. 
The detailed acdvities and 
implications of those 
activities must be carefully 
examined. 

Data on this attribute can 
be generally obtained from 
psychologists, personal 
surveys, local counselors, 
clergy, and law 
enforcement officials. 

• Taking precautionary 
measures to avoid the 
impact. 

• Employing specific 
safety practices 

• Using protective devices 

Community Needs 

• Population 

• Demographics 

• Available housing 

• Capacity of public 
services 

• Characteristics of land 

use 

Data may be obtained from 
public surveys, local 
planning agencies, police 
and fire departments, local 
officials. 

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 

• Annual average 
revenues and 
expenditures of the 
relevant government 
agencies 

• Expenditures necessary 
to provide adequate 
public services without 
the project 

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. 

Per Capita Consumption 

Average amount that will 
be spent in each future year 
throughout the life of the 
project by affected 
individuals. 

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 

• Rate of fuel 
consumption (in Btu) 

• Useful energy output 
derived from fuel 
consumption 

• Heat content of fuels 

• Types of fuel 

Data sources includes the 
Gas Engineer’s Handbook 
Mining Statistics, Energy 
Information 

Administration, and State 
and Local Statistics 

• Alternate fuel selection 

• Conservation of fuel 

resources 

Nonfuel Resources 

• Points of resource 
consumption 

• Consumption rates 

• C^iantities and content 
of wastes from resource 
acquisition activities 

Data sources include the 

Gas Engineer’s Handbook 
Mining Statistics, Energy 
Information 

Administration, and State 
and Local Statistics. 

• Economizing on 
resource requirements 

• Development and use of 
substitutes 

• Recycling programs 

Aesthetics 

Individual perception and 
values for defining beauty 
make it difficult to quantify 
aesthetic impacts. 

Data may be obtained from 
surveys, and other specific 
measurements. 

• Public participation 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 fu’st 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 


o 


'1 





Vb ^i;ftcilBV;l ndl rr r' i iJi;u»r./-’ y1 %. 

. t » , . ^ . r ■■• •“ 1 •'. T;-?*»» 'ni IQ'^^V* v.' “*-* 

:<o 4 4 ji <• r^j ^yy.^’y j^r a fen no yafttdv^'^* ---— 

li 4r,Viai♦} 1: Oi ?t3W|K^ »rf )3<»j|pl ^ •*(5«0'*^Tq it tol dt«f«fCl(jqj5<lf 


'y 

^riJ j;rfDqrii<T QJ <qlod 0«U Ji M t-rd Sirtp Id :>W i* vA(by! 

r lui . «* A ‘ I ■ #<; . h - > i^ • j . ^ t V« i ^ 


• ,M ,f , .‘ ‘‘it L, ... 

rtoHznoQ, "‘t. 


■Bk»t V' 


U# iAf;^ 


^9 


riu^ ^ 


>VT»oy ,7bi*‘4/y»» V? ^^‘mV. ^ihtlnttnw h^ik . .rR jjifU ^0 ne*^ •mo’m ^dT 

fi teiooH*)® ru **wi rawf^yd' a;, a jnuu 

bv3ii!»wol 4 wljOTf- * 9v. /Vdtilo ^ia rswU )u^ J^^oiq 

» t \. \ f , '■ “SL ‘' '.>uc;w‘*y*.i|i k-f 

feftfc, <4 ,iarj«»m }i;l^fi!}:;ib')fn bna u-^ 

yflfn (1 -rwif^jy jm.tTiwt«i^t'J8^t »j * 21 ^ yo^ti mo^do»q wn*?* 5 HT 

• W»i . ^k ^ 

6nxfrr£0'^#f:. icru^xr^.c^tr'^mya ^rjr^mr^»:nsffmo> ^uno^f^^f$m^87Jv^r 

9(> 

,»fliiq noafei;a5Jot.T(|f(fti 9i<» <11 H .’ 



V-* -2 ‘Ovn-ft • J*.|tji<? ri fmt ti.tr* M ' '’ .'. i>" ♦ AfU^i’•‘u- fiiti • tulfOa 

Urw'?«J :x5 (*m *j;y6qiiii ii8i?»mricriry*i^ 7. iOfia^-rs ., r Aw*» 

2>fi; .:t 7thfia ipna mkx^ 3k.;. m •iiff^:i^h (A 

»j u> y) S' ».*‘-i • 



■*' ' t ?».<» »J t *.' 

diTf? 2}3cq<Mi bbfwstc;^ itraltxbnA yCt/r^ am . A13 uu , ORmir*!;# ril 

^•. «>- -- . . ^;47iq V'q-->q«» d ot *• wtji 

i , *if^..T - ■- If«i. x-'iiti 

.. *c«>v r'» w , 

^{?yJwii ;xii sli br-s^ 'in^ifVflVfd^krU 1 / .•rtqlr* tpoiii u *iij y^li-56l (€ 

.♦*■■ • iflir*.* >> . u ■ j ''liu x>4n>ci||||^ 

’W^ t- I •■|■>'4;^ • ft. >t»U *JI^.’'l >4lM'i,*~H Ci2^n||j| 

• Di'4<i>lJ^u‘io of Oi ,1 ' 

‘i ■>!*.• itii%*** •'*m*-* 



I 

Ul 




'tyW-'Uf .’ V,r ■ * *(ll‘V •• mvrtij.^ini|,. 

!«/ :«?» ■ • -•; 


* -.«jH»-. < 


HtwA n ii m a m 


• *'■• r ,1. ; 


as: 


t I A A»* 4.4 










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.l 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-1 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-1 


TABLE B-1. EXAMPLE GREEN INDICATORS 


Indicator 

Description 

Percent Recycled 

The percentage of recycled material in a product. 

Degradability 

The ratio of the volume of degradable material in a product to the total 
volume of the product. 

Life 

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. 

Junk Value 

This is 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. 

Separability 

A measure of what materials can be separated from a product. It is the 
ratio of the volume of separable materials to the total volume of the 
product. (The notion of separability is different from disassembly.) 

Potential Recyclability 

The ratio of the volume of recyclable materials to that of unrecyclable 
materials. 

Possible Recyclability 

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. 

Useful Life 

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. 

Utilization 

The ratio of the useful life of a product or material to the time it takes to 
“return” to the environment. 

Net Emissions 

The respective sums of solid, gaseous, and waterborne emissions from a 
particular product or process life cycle. 

Total Emissions 

The sum of all solid, gaseous, and waterborne emissions taken together 
from a particular product or process life cycle. 

Total Hazardous Fugitives 

A measure of the weight of hazardous fugitives, expressed as the ratio 
of the weight of hazardous emissions per unit weight of product. 


Source: Navinchandra, 1991. 

nondegradable portions may be to the environment. The approach merely assumes that less 
nondegradable material is necessarily “better” for the environment. 

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 

Number of 
Chemicals 

Examples 

Environmental 

Impact 

Reduction 

Emphasis 

I&II 

Category 1-38 
Category 11 - 65 

• ammonia 

• benzene 

• CFCs 

Most severe environmental 
impact; highly toxic; human 
carcinogens 

Minimize use 

III 


• acetic acid 

• pyridine 

• styrene 

Moderately toxic; corrosive; 
suspected animal carcinogens 

Recover and reuse 
onsite 

IV 

All remaining 
chemicals 

• acetone 

• butanol 

Least environmental impact 

Reuse onsite 
following on or 
offsite recycling 

V 


• cardboard 

• paper 

• plastic 

Depletes natural resources 
during manufacture and 
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. 

In addition, many complain that EARS data requirements are too time consuming and 
that EARS is cumbersome to use (see 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 
company’s financial system. Thus, the company is unable to readily assess the financial benefits 
of environmental improvements to its operation. 

Relevance to Impact Assessment 

It is unclear how EARS could be used in the context of impact assessment. Perhaps at a 
most basic level, inventory items could 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 ways—as 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 view—by 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 

Judgment Probabilities for the Occurrence of Impact A 

1 

0.20 - 0.25 

2 

0.10-0.15 

3 

0.20-0.30 

4 

0.15-0.20 

5 

0.25-0.35 

Median = 0.20 - 0.25 

Mean =0.18-0.25 

Range =0.10-0.35 



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 
verify and support the results of a judgment probability encoding study. 

A code of good practice will need 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 panel? 

• How many experts are required to conduct the approach? 

• From which fields should the experts be chosen? 

• Who approves the selection of experts and monitors the judgment probability encoding 
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. Being 
an entirely subjective approach, it would 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 TD^q to human exposure. TD 5 Q 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 LD 5 Q the lower the dose rate, or TD^q 
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 

Daily Human 

Carcinogen Dose Per 




(%) 

Exposure 

70-kg Person 

Rats 

Mice 

References 

0.001 

1 liter (tap water) 

Chloroform, 83 pg 

(119) 

90 

96 

0.004 

1 liter (well water—worst) 

Trichloroethylene, 2,800 
Pg 

(-) 

941 

97 

0.0004 

1 liter (well water—best) 

Trichloroethylene, 267 

Pg 

(-) 

941 

948 

0.0002 


Chlorofomi, 12 pg 

(119) 

90 


0.0003 


Tetrechloroethylene, 21 

Pg 

101 

(126) 


0.008 

1 hour (pool) 

Chloroform, 250 pg 

(119) 

90 

99 

0.6 

14 hours (A/C conventional 
home) 

Formaldehyde, 598 pg 

1.5 

(44) 

100 

0.004 


Benzene, 155 pg 

(157) 

53 


2.1 

14 hours (A/C mobile home) 

Formaldehyde, 2.2 pg 

1.5 

(44) 

28 


Source: Amesetal., 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 TD 5 Q 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 values 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 

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 

A wide variety of environmental indices have been developed to provide an estimate 
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 used to compare the relative impact of a variety of different 
substances released into the environment. This section discusses two main groups of 
indices—air pollution indices and water pollution indices. 


B -8 


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-y 

Other 

Green’s Index 

2 A 3 C 





• 

• 


Combustion Products Index 
(CPI) 

2 CiC 







c 

Measure of Undesirable 

Respirable Contaminates 
(MURC) 

lAjC 





• 



Air Quality Index (AQI) 

3 C 3 C 

• 



• 


• 


Ontario Air Pollution Index 
(API) 

2 A 3 B 





• 

• 


PINDEX 

7 C 3 C 

• 

• 

• 

• 


• 

d 

Oak Ridge Air Quality Index 
(ORAQI) 

nA 3 A 
(n=l to 5) 

• 

• 

• 

• 


• 


MITRE Air Quality Index 
(MAQI) 

nA 3 A 
(n=l to 5) 

• 

• 

• 

• 


• 


Extreme Value Index (EVI) 

nAaA 
(n=l to 4) 

• 


• 

• 


• 


Short Time Averaging 
Relationships to Air 

Quality Standards 
(STARAQS) 

6 B 3 A 

• 

• 

• 

• 

• 

• 


Environmental Quality Index 
(EQI-air) 

8 A 3 A 

• 

• 

• 

• 

• 

• 

e 

Pollution Standards Index 
(PSD 

5 B 2 B 

• 

• 

• 

• 


• 



Total 

7 

5 

6 

7 

5 

9 



® Classification is based on the Thom-Ott air pollution index classification system. The first digit indicates the number of 
pollutants the index addresses. TTie 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. 

^ CO, carbon monoxide; NOj, nitrogen dioxide; OX, photochemical oxidants; COH, coefficient of haze; SP, total 
suspended particulates; SOj, sulfur dioxide. 

' Fuel burned and ventilating volume. 

** Hydrocarbons and solar energy. 

^ Visibility and industrial emissions. 

Source: Ott, 1987. 


B-10 








TABLE B-6. CLASSIFICATION OF WATER POLLUTION INDICES 


Index 

Number of 
Variables 

Scale 

Variables Used 

General Indices 

Quality Index (QI) 

10 

decreasing 

DO, alkalinity, chlorides, CCE, pH, 
temperature, specific condition, total 
coliforms, other biological. 

Water Quality Index (WQI) 

9 

decreasing 

DO, BOD, nitrates, phosphates, pH, 
temperature, turbidity, total solids, 
fecal coliforms. 

Implicit Index of Pollution 

13 

increasing 

DO, BOD, COD, iron, manganese, 
ammonia, nitrates phosphates, ABS, 
CCE other chemical, pH, suspended 
solids. 

River Pollution Index (RPI) 

8 

increasing 

DO, BOD, COD, phosphates, other 
chemical, temperature, specific 
condition, total coliforms. 

Social Accounting System 

11 

decreasing 

DO, BOD, alkalinity, hardness, 
chlorides, pH, temperature, specific 
condition, total solids, fecal 
coliforms, total coliforms. 

Specific-Use Indices 

Fish and Wildlife (FAWL) 
Index 

9 

decreasing 

DO, ammonia, nitrates, phosphates, 
phenol, pH, temperature, turbidity, 
dissolved solids. 

Public Water Supply (PWS) 
Index 

13 

decreasing 

DO, alkalinity, hardness, nitrates, 
chlorides, fluorides, sulfates, phenol, 
pH, turbidity, dissolved solids, color, 
fecal coliforms. 

Index for Public Water 

Supply 

11/13 

decreasing 

DO, BOD, hardness, iron, nitrates, 
fluorides, phenol, pH, temperature, 
turbidity, dissolved solids, color, 
fecal coliforms. 

Index for Recreation 

12 

decreasing 

DO, nitrates, phosphates, oil and 
grease, pH, temperature, turbidity, 
suspended solids, color, other 
physical, total coliforms. 

Index for Dual Water Uses 

31 

decreasing 

iron, manganese, ammonia, nitrites, 
chlorides, fluorides, sulfates, phenol, 
other chemical, pH, specific 
condition, color, fecal coliforms. 


(continued) 


B-11 






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 

14 

increasing 

DO, alkalinity, hardness, iron. 


manganese, chlorides, sulfates, pH, 
temperature, turbidity, suspended 
solids, total solids, color, other 
physical, fecal coliforms. 


Planning Indices 




Prevalence Duration Intensity 
(PDI) Index 

b 

increasing 

Note: Because of their flexibility and 
special-purpose nature, the planning 
indices and statistical approaches 
do not lend themselves to detailed 




comparison. 

Nation al Planning Priorities 

Index (NPPI) 

b 

increasing 


Priority Action Index (PAI) 

b 

increasing 


Environmental Evaluation 

Systems (EES) 

78a 

decreasing 


Canadian Pollution Index 
(CPI) 

b 

increasing 


Potential Pollution Index 
(PPI) 

3 

increasing 


Pollution Index (PI) 

b 

increasing 


Statistical Approaches 




Composite Pollution Index 
(CPI) 

18 

increasing 


Index of Partial Nutrients 

5 

decreasing 


Index of Total Nutrients 

5 

decreasing 


Principal Component Analysis 

b 

N/A 


Harkins’ Index 

b 

increasing 


Beta Function Index 

b 

increasing 



* Water quality variables account for 14 of the 78 variables used in this system. 
** Any number of variables can be included. 

Source; Ott, 1987. 


B-12 







I 


• 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 life—types 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 = = A^ (Cj/BjTj) 

where 

Cj = the concentration of component i as a percentage of the waste by weight, 

Tj = a measure of the toxicity of component i. 


B-13 


A = a constant equal to 300 used to allow entry of percent values for Cj and to adjust 
the results so that a reference material, 100 percent copper sulfate with an oral 
toxicity of 300 mg/kg, achieves an equivalent toxicity of 100 , and 

= a conversion factor used to convert toxicides (tj) to equivalent oral toxicides. 
Table B-7 shows conversion factors (Bj) for various toxicity measures. 

For carcinogens and mutagens, a TD 5 Q oral rat dose is used if available. Otherwise 
carcinogens are assigned a 0.1 mg/kg, and mutagens are assigned a Tj of 0.6 mg/kg. 

Toxicides are converted to equivalent oral toxicides 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 toxic equivalent oral toxicity value) is used. If a carcinogen 
or mutagen is assigned a value for T- in the absence of a TD^q, B^ 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 CgQ 

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 M can 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 For The Equivalent Oral Toxicities (B;): 


Toxicity Measure_Units_B: 


Oral - LD 50 

mg/kg 

1.00 

Carcinogen/mutagen - LDjq 

mg/kg 

1.00 

AquaUc - 48 or 96 hr LC 50 

ppm 

5.00 

Inhalation - LC 50 

mg /1 

25.00 

Dermal - LD^n 

mg/kg 

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 

Concentration (5) 

Equivalent Toxicity 

Chromium 

0.000002 

0.00006 

Barium peroxide 

0.000012 

0.000003 

Arsenic pentoxide 

0.000002 

0.00000008 

Lead monoxind 

0.000005 

0.000000001 

Cadm ium 

0.000000 

0.000000000 

Selenium dioxide 

0.000002 

0.000000000 

Total Equivalent Toxicity 


0.000063 

Overall Hazard Ranking 


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 

Concentration (%) 

Equivalent Toxicity 

Nickel 

0.00171 

0.0513 

Phenol 

0.001544 

0.00772 

Cadmium 

0.00008 

0.0024 

Chi oroform 

0.000039 

0.00117 

Barium peroxide 

0.00028 

0.000067 

Fluorine 

0.09 

0 .000058 

Chromium oxide 

0.00017 

0.000006 

Lead monoxide 

0.000074 

0.000000964 3 

Xylenes, total 

0.000002 

0.0000000888 

Arsenic pentoxide 

0.000002 

0.000 000075 

Methylene chloride 

0.00003 

0.00000005389 

Toluene 

0.000044 

0.0000 000528 

2-butanone 

0.00022 

0.00000004074 

Acetone 

0.00042 

0.0000000252 

Silver dioxide 

0.000035 

0.00000000372 

Mercury oxide 

0.000000000 

0.00000 0000 

Selenium dioxide 

0.000003 

0.000000000 

Silica 

99.9049 

0.000000000 

Total Equivalent Toxicity 


0.0068 

Overall Hazard Ranking 


High 


Source: Thomas and Miller, 


• 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 

The degree of hazard evaluation method may be used to normalize a variety of chemical- 
based inventory items in a manner that allows the analyst to compare not only the equivalent 


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 the 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 terms 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 





Factors 




Groundwater 



Category 


Route 

Surface-Water Route 

Air Route 

Route Characteristics 

• 

depth to aquifer of 

• facility slope and 




concern 

intervening terrain 



• 

net precipitation 

• one-year 24-hour 



• 

permeability of 

rainfall 




imsaturated zone 

• distance to nearest 



• 

physical state 

surface water 





• physical state 


Containment 

• 

containment 

• containment 


Waste Characteristics 

• 

toxicity/persistence 

• toxicity/persistence 

• reactivity 



hazardous waste 

hazardous waste 

• incompatibility 



quantity 

quantity 

• toxicity 

• hazardous waste 





quantity 

Targets 

• 

groundwater use 

• surface water use 

• land use 




• distance to sensitive 

• population within 4- 


• 

distance to nearest 

environment 

mile radius 



well/population 

• population 

• distance to sensitive 



served 

served/distance to 
water intake 
downstream 

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 the 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-11). 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 and 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 

The USAF 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). As part of 
this system priorities are to be set for taking further actions at sites. Thus the Hazard 


B-20 


TABLE B-11. RADIONUCLIDE GROUPS 


Group 

Nuclides 

A 

226'*’i>Ra, unidentified alpha emitters 

B 

i29j 210 +Dpjj 22^Th, 2^^"*^^, unidentified beta and gamma emitters 

C 

^«Am. '^^Cs, 

D 

^«Cm, “Co. "’Cs, ‘”Eu, '^^Eu, “Na, ’^Nb, ”Ni, “*Pu, “’Pu, 

"^“Pu. ^”U, 2”Pu , “’Ra, ‘’'Sm, ’’Tc, "“U, 

E 

^“Ac. ‘^C, ”Fe, ”Mo, ’’Ni, ^^’Np, ^^'Pu, 

F 



Source; Sandia National Laboratories, 1986 


TABLE B-12.MATRIX TABLE FOR GROUNDWATER ROUTE WASTE 

CHARACTERISTICS SCORE 


Maximum Ground-Water Concentration (pCi / L) 

Nuclide lO'^ lO'^ 10* 

10“ 

10^ 

10* 

10* 

lO"* 

10* 

10“ 

10^ 

10* 10“ 

A 0 13 

7 

11 

15 

21 

26 





B 0 1 

3 

7 

11 

15 

21 

26 




C 0 

1 

3 

7 

11 

15 

21 

26 



D 

0 

1 

3 

7 

11 

15 

21 

26 


E 


0 

1 

3 

7 

11 

15 

21 

26 

F 



0 

1 

3 

7 

11 

15 

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, 

• the waste and its characteristics, 

• potential pathways for waste contaminant migration, and 

• any efforts to contain the contaminants. 

Each of these categories contains a number of rating factors that are used in the overall 
hazard rating. For example, the waste 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 for each of the three categories are then added and 
normalized to a maximum possible score of 100. 

The USAF Hazard Assessment Rating Methodology is based on the same JRB model as 
the EPA HRS method and is similar in many respects. The best way to highlight the strengths 
and weaknesses of the USAF 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 USAF 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 HRS 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 site could 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 

• carcinogenic 

• mutagenic 

• teratogenic 

• substances 

The significance of dose- 
response and safety standards is 
undefined and depends on levels 
of acceptable risk. 

MPC/EPC-Air and Water 
Equivalents 

• performance criteria 

• disposal volumes 

• offsite concentration limits 

Depends on validity of 

MPC/EPC limits subject to 
change. 

Equivalent Hazard Categories 

• general toxic effects 

• based on definitive data 

Database is usually acute rather 
than chronic toxicity. 

Site-Specific Risk Management 
Committee 

• local conditions 

• credible 

• easily understood 

Potentially subjective changes in 
value judgments with time or 
committee members. 


B-23 







• Production and Yield: These characteristics are valuable for functional reasons. This 
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 amd 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, and 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-1. 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. 

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. 

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. 

Current or measured values represent the actual state of the system. Current values may 
be determined by direct measurement, modeling, or through secondary data sources. 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. 


B-24 


Current 


Objective 


Bad 


^ Good 


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 
between the ecological quality objective (the target value) and ambient environmental conditions 
(the reference value). 


B-26 


Appendix C 

Key Terms and Definitions 



tflaU#rr!*. -^Bitif* ‘I. 

dil^Cfcn re Vi'VfCft f * i/jrr‘ t ir ■ *’ v .' 


,u ♦ a ’Mtr 

A'i -; 


etoJogh;iU ol>/“wVi*-' h taU<*r. • ,t: o'* i K*:;n hy »*' .Sr» <.• ■•» t.' 

tfcli .liy T^iiJ f^»(fij, ;t £^14 Vv U*!i;i!.i:?nriL \ 


ft'twwMi Ilk eil'tilfica.' 


vihe ro^ncpirrrTC 




ufir^ 41 '» 


iV* C. • X 

m »0m4 _ - -- 


0 xIbneqqA 


snoi/jfiiftaQ bnii ainioT yoM 




H Hy yMj 

il: ^ 



m;^' i ! 



i-;.A 

^'T *1 

L“i»r J 


H 

'<'i 


'» ♦ 



I'.ij «y.<(r*n. Mib 
.'-'iJ W VftKh thft 
!{,^'.ft mopci^^Kl 'Jl 
f9 iIk: c.^JPTnironufo 
ryi>-^cic<it\i1 cosrxJjti 

'Vj^ 


■I' 



















Assessment Endpoint 


Classification 


Conversion Models 


Characterization 


Direct Impact 


Goal Definition and 
Scoping 


Impact 


Impact Assessment 


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. 


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, CO 2 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 km^ “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 SO 2 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-1 


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, NO^ 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 CO 2 
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, 
resource and 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 CO 2 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 
components—goal 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 
emissions. A different measurement endpoint for this scenario 
could be the lost recreation revenue at lake Y due to 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 CO 2 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 


. 6^?&lcr. % .r^vhUBijUkfjr^^'wi loj^ ^ rio ‘0‘ ^ 'Iv(*s* i 

« od tin;, , .'i/rf«ff^ > fi rj • < ^ ' ux>^»a. > fnr*. ;.r'ri ni4iy 

lijof Jo j(T9n»ii‘,3^tifii to' .; *;tn 

OJ ^rdiLUidhiM i(fKmiibn« ^! if -«rX iu< :i 4t, r. rr^ j-ri twc- 

W^QriUOtifUlSj 4‘ ' 

eiORHttvy fDT ciiifT]pto;*i :^)2 

«ittati YM4I *^»i u •,. jfh>y '’fIStjry ot»uJii t* WiWrtWJMfcJ*'- 

3fti»»5siMioftK!n»ruw;.'».<1-0,, ' tj. ■ .iu;<j*ci<w< v>pl<ittetioiqK«,i^4W^!/^ : .' 

' qi R-'tfxKifWiO^ « aH' v70r»".vti T*jis <^ru ' 'T .Jj^i W4fn^;.^ 


oh£;aa« aijUifii m-rns^u/^r, n 'i'oU^i, i 


ima **m»ny?! J:>. li* tti xsn^t'O f» ;xii, '' idj ! > t'u ^ 

wwnf^f^^j riti^nu n* b^( vSi <ir., . •• ,.' hm yhu: , 

• b.’ ■■• ■*’ J ‘.UlA h a^- » 1 f ivr^r .n(f% tHttl if 



rararhicWrJ 

_ pc nr ifnu, ff, .;jjt in reiu^iC _ 

01 i»i>4a!>|4i wUd hri;?? ,r.,n7(rfii i ja^f. rjj<i , /: Kd!iikdid«l|} rr>akv n' m» j* 

J3Mi>0tq jrYi^ <. 1 « '^K't •/' r* 3 

lU tiOiJ TrV , . O'' )ti - 

bv9l aili ix^*ir ’trf^».''\rr ‘ '’Kr riC,| 3iru»w 

'. '*ni u - ■' '"iiinSkv. < Mui * <;v, : j; al 



I 



'I 


?£i 


Ifo e#j ' m? , ii«<wwn» 

. »' uO"^-*ViJp'‘.;r,rr^ «4i(i^nr.or/l«V4 n f?.c r<U-*T.i 

svinilm-irt) Uf' jvai.. ;i-./B I .,^'a} *'''' .. •‘""^' 

m ■ ^- •> r- .. . . . 1- 

I.i'e- Ai •’ *.• ^ A 

- '■ ^ ■ 


«al5y t^vr^^v 1 fuf i i/c Jo ;r 

to r? i3i»? V j iw 

S^tt^^feiiA.'i)t/-u ‘ '^i. ■^^:ji - »f< • 


inilm -.rt, Uf. JV, p. ;.-./B I v*« ‘''T»" '* o<< r««^ 

itoho^ii*' . V, 77 ^ •/..nq«tf - 0 ’facul-} ;! *1 




I.-- • 1 - 


^•ri' t( <u :. ‘«r.v.i\lwmi / 4 .‘» U% ,n>n,i UtiUl ft. 

,1 »Y5H-IJ! ^ N* KVwwf)'Wiir ♦ ipw-'i' . * »aj j4 

MH'Ui .u H##* ,f: /-r, r^itiiji||( 

^ p' •< * 4 #*''* ““ I 5 j»^ru>$yirtf?ttt ’t«unt.l OnNti -.'I 

.r «* •»#» 4? . ’.". Op|F,M Tr- aNTI £g 

ii- «•. ^ i.(, A .'<>11 .ivc> u ^u’totjtj.'ticfn - uy 

i idKt^ Ju> 1 6^.7ptn|^ i ffiw '-v., ^ 

^ .--M.'.- u .;.>'«r7t:oewl 414<i4r. ..’m, ' * 








Appendix D 
Bibliography 



4 



•.«T 





If 

’ ’ 


I i . 


f ■ 


fr 

i 



T^ f 




-S, 


™. I-*-;. 






^ amm m m v i ■ •.,»« 


. , 4i?4 ? 
, Vo I-*- 


. 

ya.., 


^ i.i 


m: 


*vi fc^:- 


ll'g* 




< 

' • ^ 








■■ .* ...;*■ ^^^' 
. ti 

■i 


% 


i. 


:y>.- 


\ i- 


'I . . . ■ "■ 


..ii 


•-*1 


4.1 


j 



•. A'!! 




.i«r- I, 


















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 Sciences Division. 
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. 

Bamthouse, L.W. 1991. The Implications of Risk Assessment for Product Life Cycle Analysis. 
Oak Ridge, TN: Oak Ridge National Laboratory. 

Bamthouse, L.W., G.W. Suter II, C.R. Baes HI, S.M. Bartell, M.G. Cavendish, R.H. Gardner, 
R.V. O’Neill, and A.E. Rosen. 1985a. Environmental Risk Analysis for Indirect Coal 
Liquefaction. ORNL/TM-9120. Oak Ridge, TN: Oak Ridge National Laboratory. 

Bamthouse, L.W., G.W. Suter 11, C.R. Baes III, S.M. Bartell, R.H. Gardner, R.W. Millemann, 
R.V. O’Neill, C.D. Powers, A.E. Rosen, L.L. Sigal, and D.S. Vaughan. 1985b. Unit 
Release Risk Analysis for Environmental Contaminants for Potential Concern in 
Synthetic Fuels Technologies. ORNL/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. 

Bembe, 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. Bomstein, Daniel S. Dixler, Dale K. Humbert, James B. Knaak, 
Robert A. Mathews, David P. Richardson, and Frank Vincent. 1993. “Protocol for the 
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-1 


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-Cycle Assessment on PVC Packaging Systems.’’ Paper 

presented at the Society of Environmental Toxicology and Chemistry Life Cycle Impact 
Assessment Workshop, Sandestin, 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. 

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.’’ 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, VA. September 30 to October 2. 

Desvousges, William H., F. Reed Johnson, Josephine A. Mauskopf, Stephen A. Johnston, Joel S. 
Smith, K. Nicole Smith, K. Nicole Wilson, and Maria G. Benerofe. 1991. 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 

Tillman. 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 (HRS) 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 Managing 
Planet Earth: Readings from ScitniWic Amtncdxi 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, WA: 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 Heijungs. 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. “Headings for the Impact Analysis.” SETAC European Workshop on 
Environmental Life Cycle Analysis of Products. Leiden, Sweden: Centre of 
Environmental Science, Leiden University. 

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. 

Hunt, Robert G., and William E. Franklin. 1975. “Resource and Environmental Profile 
Analysis of Beer Containers.” Chemtech August:474-481. 

Hunt, Robert G., Jere D. Sellers, and 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 Workshop, Sandestin, FL, February 1-7. 

Jain, R.K., L.V. Urban, G.S. Stacey, and H.E. Balbach. 1993. Environmental Assessment. New 
York: McGraw-Hill, Inc. 

Keeney, Ralph L., and Howard Raiffa. 1976. Decisions with Multiple Objectives: Preferences 
and Value Trade Offs. New York: John Wiley & Sons, Inc. 

Kuik, Onno, and Harmen Verbruggen. 1991. In Search of Indicators of Sustainable 
Development. The Netherlands: Kluwer Academic Publishers. 


D-4 



Lawrence Berkeley Laboratory. 1992. “Product Life Cycle Analysis: The Missing Social 

Dimension.” Paper presented at the Society of Environmental Toxicology and Chemistry 
Life-Cycle Impact Assessment Workshop, Sandestin, FL, February 1-7. 

Linstone, Harold A., and Murray Turoff. 1975. The Delphi Method: Techniques and 
Applications. Reading MA: Addison-Wesley Publishing Company. 

Lubkert, Barbara, Yrjo Virtanen, Manfred Muhlberger, Jyrki Ingman, Bruno Vallance, and 
Sebastian Alber. 1991. “Life-Cycle Analysis IDEA: An International Database for 
Ecoprofile Analysis, A Tool For Decision Makers.” Working 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 Ejfects ofDijferent Packaging Systems 
for Fresh Milk. Printed in the Netherlands. 

Mitchell, R.C. and R.T. Carson. 1991. An Experiment in Determining Willingness to Pay for 
National Water Quality Improvement. Draft Report to the U.S. Environmental 
Protection Agengy, Washington, D.C. 

Nash, Jennifer, Karen Nutt, James Maxwell, and John Ehrenfeld. 1992. “Polaroid’s 

Environmental Accounting And Reporting System: Benefits and Limitations of a TQEM 
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 Awmca/i 
Society of Mechanical Engineers, Vol. 31. 

Organization for Economic Co-operation and Development (OECD). 1989. Environmental 
Policy Benefits: Monetary Valuation. Paris: OECD Publications Service. 

Ott, Wayne R. 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. 

Plewa, et al. 1986. Assessing a Degree of Hazard Ranking To Illinois Waste Streams. 

Hazardous Waste Research and 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 at 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, Sandestin, FL, February 1-7. 

Reddy, K, R. 1985. Special Waste Categorization Study: Volume I. Hazardous Waste 
Research and Information Center, No. RR-005. Illinois Department of Energy 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. 

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 
Pollution—Ecological 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 Halo genated 
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. Bamthouse, C.R. Baes HI, 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. II, L.W. Bamthouse, S.R. Kraemer, M.E. Grismer, D.S. Dumford, D.B. 

McWhorter, F.R. O’Donnell, C.R. Baes IB, and A.E. Rosen. 1985. Environmental Risk 
Analysis for Oil from Shale. ORNL/TM-9074. Oak Ridge, TN: Oak Ridge National 
Laboratory. 

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 in 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. 

Tellus Institute. 1992a. CSGITellus Packaging Study, Report US: 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. CSGITellus Packaging Study, Volume I. Report prepared for the 
Council of State Governments, Lexington, KY, and U.S. Environmental Protection 
Agency. Boston, MA: Tellus Institute. 

Tellus Institute. 1992c. CSGITellus Packaging Study, Volume II. Report prepared for the 
Council of State Governments, Lexington, KY, and U.S. Environmental Protection 
Agency. Boston, MA: Tellus Institute. 

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, D.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. Prepared by Research Triangle Institute, Research Triangle 
Park, NC. 


D-8 


U.S. 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, MI. 

U.S. 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. 

U.S. 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-C0-003. 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 Haimen Verbruggen, 1991. Boston MA: Kluwer 
Academic Publishers. 

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. IMDIECM Life Cycle Analysis Part I: Environmental Impact 

Metrics Definition, Stakeholder Survey and Panel Review. Draft report. Sandia National 
Laboratories, Albuquerque, NM. 

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. 

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. 

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. 

World Wildlife Fund. 1991. Getting 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) 

1. REPORT NO. 2. 

EPA-452/R-95-002 

3. RECIPIENT'S ACCESSION NO. 

4. TITLE AND SUBTITLE 

Life-Cycle Impact Assessment; 

A Conceptual Framework, Key Issues , and Summary 
of Existing Methods 

5. REPORT DATE 

July 1995 

6. PERFORMING ORGANIZATION CODE ^ 

7. AUTHOR(S) 

8. PERFORMING ORGANIZATION REPORT NO. 

9. PERFORMING ORGANIZATION NAME AND ADDRESS 

Office of Air Quality Planning and Standards 

U.S. Environmental Protection Agency 

Research Triangle Park, NC 27711 

10. PROGRAM ELEMENT NO. 

11. CONTRACT/GRANT NO. 

68-D2-0065 

12. SPONSORING AGENCY NAME AND ADDRESS 

Office of Air Quality Planning and Standards 

U.S. Environmental Protection Agency 

Research Triangle Park, NC 27711 

13. TYPE OF REPORT AND PERIOD COVERED 

14. SPONSORING AGENCY CODE 

j 

i 

EPA/200/04 1 

15. SUPPLEMENTARY NOTES 

16. ABSTRACT 

Life-Cycle Assessment (LCA) is a holistic concept and approach for evaluating i 

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 


a. DESCRIPTORS 

b.IDENTIFIERS/OPEN ENDED TERMS 

c. COSAT 1 Field/Group 




18. DISTRIBUTION STATEMENT 

19. SECURITY CLASS (This Report) 

21. NO. OF PAGES 

20. SECURITY CLASS (This page) 

22. PRICE 


EPA Form 2220-1 (Rev. 4-77) 


PREVIOUS EDITION IS OBSOUETE 






























1 j 


ATAO 


__kV'tu^l^o^1r(c\^4 >w*¥itm ''wxVi<r«^MH»*% 

■ 'tif finVirfMiincj.yf/a. * Teci ’ikjii T^Vi- \ nt jn ■:•*! 


'^>rs Onj-v Jlid* <<?v *• j9vT 7 iLT/w>»ttv?^^#3iiv®5k| 


3000 




t >JM , t *V'4ZT .jjTOViSJBT^ !C'f3 ^-IDOOi:) A 






fYw.*tSt' *.)»S''-'fisi.': (1 i'nui/ ^^pe«r p-r ; ,.* f<<'??'’.^>^'4'/erf 


i*1 


i bns-jz hnr, ^^Jl-.j/p yfA io doJ^iiO { 


WMkirif, RttaCto!'V) . r' . />40. n -i , a'Vt'.- 

__< ,virw> or.'.' ..«iW^ '45i‘ (FteiaMSa 


0#l*' VoH iOA ViiM^*1o'^ .If 
n* /»8i«ft ja •?«a jiAinnsll ^fJ;A to , 


v/lfiitr, !>3 »i 51 V«|<^, gi^ 4 ‘ Sfua-Kl.>rirw^ .3.U 

jpA^unitt. 

"■' ^.■l^'Wi» **tS~ i ft^‘ ■*‘*l ii ' -I* ■ ! I .< I I - '~ 

oT^a-c vn ij •tui Ary 


f . * 


i 

C‘ > ’1) «rt Ai * .^ifMfW'. aiiJLI^a i vcir Aftc^y )c.i>t.' orfsoiit * jI ty^r I 


Scr'c.i*' "'f i n' -■'t’l .'■"717 ri^Kibrf'K'Ky 
^rrtjBur*7tt Mi kVf)$Iih< 

«b«>% ., .« — _ — *** ..a _ o ..o ^ 


JO r- Uv K>«»» n30M<fBi rnwnjKl Jj«. «s^n^wit, 'j :wno erfj 


•«u3iAJiV »» iiiill aj|»l#Aqw ijiocl VvV A .3lVJl30B 


«B‘lu4i5uo bi/0 »;wq.< i#±tHo 

I** tcmlifi;' f- f V 4 I p 

y34.qar> --.t 

08 A« Jl .•Uo»*rt^‘!jf'r.‘‘*:^r 1 >tr'Jl^iijA*./"iOt rv'^ueol '-.'^lua A^kiOdj J idansq^Bof 

.«i'M.i(»» Buc.^iBv *10 aiany^qvl^tomih &»tb .avtAiii^VDa ^t<3 lo va^n i»ill> 5 t>bl. 
H <'!d Ti '* «', fi«>vi;r> v. : ^ifl0i^ol<]qiint:llt i W? .. /,wr?io/j f .* OikfoCilt 

.V. A "- taijy Hw .>fe ' 


Wu»‘'<-W'V^Uif" I90l ^ffnH0 Jtt (WSi>B^cr- !i:fr(c$iet fotM^tttiJ'it:}, h(i*^ 'lfmi 


-^■'"" • • -Xi^" . 

aM 8 9 V»« « 0 i xa ' X i/^%:;i 



IV : Aftri* PC 

1/ ^ 


J j»JjPi*'* t »o i4 >^}rm 04^/ ao,- '^ ylJk 

-«r' -•-. W?-. . , i — 


fc*nir 


auQt^iiiottaa 


‘^•- ‘. ri 


1*1 


V ' i" 


tt'iM 


JsaxvTr .o;r?q ,^r.'>^s.yr.[j <i7* -., 


*»♦ 


’ iWSsiTmTiVH * Tir .C7 fliT 


'■ ■'^r 



rtx^^iica'o'hyj At 


j 


#«AwV.W^t. .tr- a* »W«#VSW^ f^#CtC 


^ A-^% 











































































































« 



■ lA^ -- - 



9 


V ><; 


sr 




♦ 










^ -V Oft 
* <r 


rSSN\ .. J> O 

'^z*o5; 


if 0 no'^ ^ ' 

A ^f35lO^-v ^'<4, 

«. o ^ 

o C^'^-v o 


H It J. 





Q- '»' 



V»-^:^'^co~o.v 

^O ^ ..^0- t 



" o® ❖ 



O o 

^ A^ - LI« ^At ‘'o 
«/, 








^ tO"”* ^■^Cx ''■”*«' 

0° °0 A^ Cp ■>'■ 

Vo^ .'•^^t •^ov^ ^ 

- - --- •' 

o o':^;\o^^‘•w.^ .X - ^ 

D- A.O^ o 


C"V 

o •' ’ 

« 








« 

K O J' o "*■, ' 

V*»«°’^p ,. o>V" * V’ 



^ V-, rf- 



*.;,.»V .. . 'V* 

;^. ■’ v'> o •_• o^ Vy » ^ 

^<^’'’“%%^p/ %^p*° 

'» • '' 00N = ,/<i>/^ • • -,\'° ' ‘V' . 

» '^O' <• "I ° 


r> OV ^ 
-.0^ %./. 


^ J> T^VVVO^ 

CV, ^ i 

9j/ ifOKO”^ 


<1 LI a . - fx 




'*> x'A' 


* ^ -• 

* V > • • r "'t'o 

« P c,^ 




■> olS*^ 


<t> V. AT</^\>.-,- 0? 

*^A ^ ry » -i 4.^ 

^ov ° 


^ - \KSn5- a f 1 ^ 

V° ,. V' 

.'% 4- |sf\^^/V, o A> 

%%“’ o' 

‘>'-^ ■'. aiK“ .‘i>-^s-^ I 

^ vj, o 


\ -.W/v 

j. 1^ o 

; ; 

• <^°X. “ 

V CL J* 




'■'»3’”V’,. 'V‘ 

X -O^ >X 

‘>’ -L . S« >? <i,‘’‘’'V °o‘''(J|OT^" ^ 

-C>^ ... ^<.>T7>'V . 

3 A^S!^*^'''t cPV 

; .‘■ 


'^'' ''o”."!.'*'' <^’ '<x'‘ 

^y ^oNCa 

0 4,"^ 




Lia, 


^ A 
& 



* . 

' A*^ 4 >'' ’'f. O ^ 

c„^^' Ad^ : 

»AOx » 




LI a 





*- '*y >. ^ 

^ <t '' 

a„- ^O. 


% ‘'o ,^ y*‘* ^ „ 
% ^O > ,■> 

: '^ov^ ° 

° t 

^ o^ 

.O 4. 


•y ■A''^ 

^VoO» = ,. ^ J 

9-x.x^ 



^ I» « y 
^4> ^o ^ s 





M*» ^ ^ 

',^0 « 

rA'^^Ai n A a."*^ J' '^>6 4r ^i>■* 


g IT 


.V 5 * * 4> 


.>' t <« * V 'I 

V % « ❖ 

.■<b > ^ 
















» H 

: '^O-^ ; 

.'•■■' * • • vV*°“°’o^ *’ ' ■>,V'” • V 4 * • ^>o *o.V"-’ * Vt‘ ** ^ 

^ .Vv'^ ® 7 li!l!lllK *^\ o 


o 




- . X ^.4 „ * 

^<!-'’ x.^ c>„ '. *w.* O \ \’^iw^ ^ ft, --•,’»w=".’- „o' '^x V 

* v'^o * * >■,. \j j- 4»”“’'!v'" * '>^4* “\''°“°v°’■’ * "'V 

o “ ^iB r ^*=3^ - z ^'C' « 

” o * vV^'^ "1 ° 

’ ^°o ■ '"' ” 

V-: W -f^o^ 

^ n 


° cONC;, <? ^OV 0-1 

^ 4 . vD * _rS^«^^ X ^<r ^ C> ❖ 4 > 'J 

^ r. OV ®l5^y^^Si» ‘/*0 <■ » 0> 

^ /-\ ^ <1 

o 


4 ‘/*0 <■ 

Or- * 





X 4 >5 e» ^ 

\ ‘'o^ 


w < K >• ^ A' ^ *^^U\Vss:^ ^ ^ 

^ Z n Z ~ '■ 

'*’ -Vx 

V^>OV A» <V^ %<V 

V-'O 4 S^\<^ 


^ M^/rk^r. ^T Si® ^ 


4 ^'' 4' X ^ ^x ^ ■‘^W^,'^ X^’ 

\'° • ‘ V'^ cO"".. V * ’ ' d^ x* '•' 


'1^ O^ /f 
ACK » 

, ^ ft, o ■ 

^°\' ° ‘’A°%V. 

l^p/ /\\^p/ ' 

W . C^ .-i- iSsOTZ^ -f ^ << X ^ ^ O ^ 


/xv 

' ^ '■> ■ 0 ’ 

<^x ^ 

4^ coNG-?x 




% o xl . 

: '^o^ » 

.*. 0 ^ V 

= \/ / 

4-^ 4 XX ^ 4.''x n^ "f V'i 

4 /-O 



s-/^ 


“S®®" ''^"'^‘= 3 '’'^ -■ 

^ 4/^0 4 »>‘^''d^ 


v-^ xV 9>y 

® V'V o ^ 


4-^ 


' °Ww- .a®.” «•’“’% °.W¥* 





^ 4 V-V <3 ^ 


^ bx ■ ^ < 



•; '^ov^ . 
» >?-^x ” 

.'xO ‘^' * 


. --/vX<i;^>°o“ ’ vV • • A<' 

4 ^ Mix ^ ^ S&(up2^ -f ^ . A <c'I' Wx x^J 

o «?5^Bs . ^^cy 



x*^Q, 


■ r> “'j.'^'*^1^ 4” o' '^j <iy Oxx •I’x 

''44 ^4^ ■'^ONO^ .<.0 ° 

^ ^yv Cv ^ V ^ 

M^<flT)»e». V ^x x"o- *. ,x^y9/i. X xV 






Ov 


M>, « 

4 '^y^O <■ 

<* 


.•\'*y>o:°'‘' 



% o X 

: '^ov^ 

•=^07/l\}^ V '' XX •h V 

0_^ S' 4 i 

»«• /A ; A>/ 

' ‘kA^-'-xV •' AA^A°o° ' coy=,.V' 

•im^- <^04" * 





I> 

e„ 'y7'W.°o'3" %.**• 
ft -. ■ " • ft\ °“°%-^ y’'»Av 

N ■f V *9 ._xxx “^X yy6 A ^ ^ X » - 

o 0-0 


4 4 ^ 


: '^o^ ® 
O ^Oy. t 





^ ^f'-3^ ‘^^^M/A o 4^x0^ cT SH%c,^ r 

Ao 0°Al^Oo C?^. 

* - ^ ^ - ,4,0 


^ ^<ok 4 4 

0 ^-o u 



O^ W 

A,.o V'-’* 

. 0 ^ *.’• X A ^4 O . 


xy x'V -o 

o V O'*' -"- 


/ft '^' O.'' ’’^ xO^ M XX Oft 

’(' '^ y*4 <L '*' W x O <• As/f//72_ -f '-J 

: "bv^ 



"bv^ 


■'AWMiS