Document qdK3QgzyqgbY74Qv2e6dKKE9k
An LCA on PVC Packaging A New Look at Methodology and Use of LCAs
Ron Cascone Senior Consultant September 1991
CHM SYSTEMS
Ckea Syslewi lac.
303 South Broadway Tarrytown, New York 10591 Tslephona: (914) 631-2828 Telex: 221844 Facsimile (914) 631-8851
SPI-07501
CHM SYSTEMS
There have been for a long time some glaring flaws in LCAs, such as the tendency of presenters or users to attempt to add together the weights of emissions of vastly
dissimilar materials, with or without "weighing factors , to derive glib, but meaningless'
"bottom lines' fo r comparisons. Also, seldom was the issue o f data uncertainty broached. We proposed an approach that would address these and many other, sometimes
SPI-07502
co
<
_Oi
u.
O
Ul
</>
oz
I> o
5
z
<
>
oo
<a
O > a
o
a O
z o
X h
UJ
<2
o-i S
z
<
*
Oo
UzJ <
ac> O
c
6
O ra
MCO J3J ,,J2 yz
uwE '
u.. Oe U E|
TJ O >> S
g | E la S S *
>. w O 0)
E0)
6e
10
to
XJ
.
Ul ai
DOwQ).
a> h_
ao
E
8
--a--> E
k. C4O
c>o cto c S?
d) CL
52
t> fi s 2S a.
fI
CO x: 3~ o
> -J CcD
I_ (0
~ OT
SI
X(D3
5--3 EQm)
9 to xj
> >* _ to iS
t
o0k.)
<Ul q
CO
X0J)
0>) c n<u
:I
Ul
E
Ul
'0 s
c
.a22>:
X)
CO
CO
co
C3O a c oo CJ) c 'B
c oo ffi
8ca3o)
oco
CD X)
3
qxa>
xcoCzo c0)
CO
~o
CO
o
XI 3
a.
CD JCDO
q
c o
X> CO CO X)
C<O0
CO
CaO> c C3O
XI
>. x>
CO
c o 'CoO
XJ
w
CO
CakO>.
>
CD CO 3
oCO
XI
XcI co
XJ
CD XZ .CO
2)
3 CL
c
CD XI
C>D xCzO
CO
o
CO
JUXl. O O
X) CD XZ
C JC o o
c o
CD
o
CO
CD XZ
3
O
CL co
2
CJ) *2
E
CD
o
(/) T,_o
0) CD
T3 C
a (A
CD CD
Q> O
l/) 0) x:
S 3 CL
c o |E
XI
CD CO CO XI
CO c o
44 c CD
E c o
4-*
g. *k. o
> c CD
CO
CkD. X4Z4
CL k_
4O--' 4o--
ID CJ
to CD
c XJ
CO
4-*
(0 *</) aI.>.
CkD. CO CO E
o D
4C--D Ui >x
E (0
CD XI
1.
o
->* o 13
XI
So
>
CD =
CO
a
* CD
CD XZ
CO
E
_c
2zz g x: MJ
XI XJ CC CD cd
o (SQ a)
-
CD
(/> ?
E
JD
XI CD
XI
o
-4-4
CO
as
XI
c co
CD X)
CckO. _>*
CoD oco
CkaD.)
XI
a>
ai
CD
*s cco
.co* CO 44
ca>
CO o
(D CO
co CO
CD
CO
xCzO
ID CO
O3
c
CD
XcI
O
co
CO
xk;t
CO
0>)
>
CD
nai
CD
X)
CD
>
E
CO
o
X* CL CD XX CO X)
tco co
CO 3
c
o 15
3 CL CD
XJ CD
CO CO CD k_
CL X CD
S'
CO CD XI 3
tk-_
k. CO
0
3 XI
CO Co o Q
a> ra
co
92.
X) 3 co CD
xz
C
o c o o tr oCO c*-4
CD O CkD. CO
c
CD
CHM SVSTEMS
All of these 'shafts' and 'shaft nots' pointed Chem Systems to the developm ent of an automatically-solving Lotus 1,2,3 (Version 3.1) spreadsheet model with open access to assumptions and key parameters to enable changes over time or for sensitivity studies. Figure 1 illustrates the overall architecture of the model. Figure 2 indicates the structural relationships and information flows among the m odel's process modules and information
SPI-07503
co
3io sc3
E E
8u liCD
TJ tOko. tj
CL TJ o JM3OC o
o
kC.L
c
o
o E
E ToJ
o .ti
TmJ 1o
TJ
SO3inZ icon itno E
3_ a J!l
O c3
CM
1T0ocJ)
c0) c>
3 5 3**b
to
s 3in
Tcoo s
TJ CO
ts
1CD
T3 C CO
in
3 -j
*o S
5
o
g 2
CD CD 5
D> |
E oo (0
*3 .2 cm
8.tj s 2 (0 o
O TCJ o cn _
S CE
8o
c ^ a- E 5 s
to o
>
t/5 =>
0) (0 ti)
<ckI--0* O3
W X) <o3
CD
o
in
7X33 | D) S" -Q) 20?)
in >
.1
CO D
to c attoo) ca>>
CD --
Cy3*) to S CO
OL Q) LI.
to *o3
ro
E o
tj
to
E to cooo
.t!
in
tco _3 ao
8-2 II* 1- s OT
ok_ oW3
0) (30
C0D)
<D
$
to
TJ 0)
in ID
COD)
3 V)
O
V)
73 to '> to
a
o
<D
CD
2in
S
o E x+E-:* a> i cc o to c
oa
> TJ
< <OD c o
CO
^
ID
-S
o o Ci--L o O) CT"D
tco ooco
c 0) i0n) Ca)L
c
3
E
oo
TJ
o !-cE- ow nTOL-- 2? ZSLH3_ Ek.
cr <u
S3 ~L
C
(o0
~
m
-2 Q-
sz
CD
!c in Pc iinn FL. 8 oX
7c .KS2 oo 3t
TJ
C
CO -Q m a> *=
a0>) w
b) .2
o CO
CO
E
TJ
TokJ.
X3) sz
c
tj c
in o
sz
to $2 eg
TJ C -h
CO ^ ^
_ in O
to *-
O {a
O CL CO
>-
Eo
a
Oo O co
3
iOn
- L_
3
m TJ
Oc/j ^rr Cto
P in
in
c
c
itno
=^
>
V2.
oo,,
0 3-
L O Q
t2 .5m2
3
.ts
o
TcJ
O CinL o
l/l
X KJ
CO
IE, -
eo
TJ S
II(A
ID J8 Qu-)
= SJ c
z O
Qo. o(A
10 a 3
E
<D
w9 Eoc 8
to
Wa 3
T3 C
tn 3 cn IE
,, < {3
T)
5 | s TJ
eO ="
w o
t3--
3
1-
c
oo>
Mk.
o wOl
tCo (3 t3
O) c o
58
^O)
C
By-8a
O)
E
?
x: o
<0 <3 o
3 s
Jo<Q0
E E 4w-* CD
TrJ (0 k~i . o TJ 5 ccn
g E _c o cc
o Q. `6
g> ;o c E
' =133 5
.2 `Oc 17] 8J
TJ ffi TJ W C
1
TJ
c
I
tJ
a o 0o Sw
pc *- 13
TkJ. >
k> 8
at
II
We made our models, and exercised them on cases of PVC detergent and oil bottles, thermoformed sheet blister pack and cartons, and plasticized film meat wrap, compared to alternatives in glass, HDPE, PET, PETG, SBS paperboard, and PE-coated kraft "butcher p a p e r, respectively. there are many qualifications. The m ost important is that the differences among facilities tend to be greater than among products. Figure 7 illustrates this point for key toxic
E
t:|3coa3
oe
* sa.
c
(O l3o_> IT
c o5 c
O 3 -Q
EE S- VOTI 1=
O .9 3
is?
oc I 3 to
M TJ o
L_
>
3 E
'OC3OT
>
5
co w
Qoo.
o O3
>
c " c
e o
>,
xj
"
ac <jl> *3 |
o>
C
TJ c
& c
sz
oc XJ E s T-l
cgO) '
fc~\ I
"35 cn Sf
E *=_
JfJt
TJ O
C CL
O
L
PC
s
{2 .o9 E -D 8
w XroJ c
8
k.
m
TkJ. o
JO
c o xo:
t On CCO *3f-
c
ffi
_ t2_o
_u. E03J {oL2
_ .2
>
TOJ E oc
to c
TJ _
.1oQ3.
.t-f Dk. --uo .2Q2-xj tJ E c oo
o
X
c
coa. o>
g 6 5 O O- o9 c
8 3?Eo
> 13
TJ
C
cd>
2
o 13
a
,, U
is
to 3
UJ =
.c9
ks
o
to
=y5
-- O
.9
CL
oE
Si m
o
ii
c oc
TaJ to: CL c
+3
XJ r T3j . TOkJ.
CL
Oc) O E c 2ro ::
CL CL k. .
o . c k.
TJ
E o
in
I c
TJ C'
ffi !`
J)
8
oX
<:
X
tcaj
};
m
wa sw
ro
t2
9
g "g w o
^ ? -2
o o S a> gQ- _19 ,r a*
ra "
3O(0
w
KJ
3SffS
n in iinn
-*g S. 11C
"o
.55=5 >Oo >aJ.
a
ra o
toQ:. k_
Mc 8 a>
TJ 8 .<0Q40-)T
emissions data, does not extend to mining, transportation or to < jmmercial, military, public works, nuclear, or power utility facilities; it only covers m anu cturing operations. Figure 8 illustrates the conceptual degradation of confidence towa Is the raw material supply and end use parts of the life cycle. Fortunately, the key naterial conversion processes are covered in the TRI.
SPI-07505
2
Ssat
8 0) *
rijQ
tf
%
* <D
4Cia-.*
ccn oaufi
C O)
wcotn
e
CD ni
o Eo
o t(c0sA.) E
L. o E ao3 E
a(0
TrJ*
ra <
8a
c
?
oCVCCO)L *- 4-*
co
Q. E3
5 .t;
(0 a<-< </) oc
tsD)
TCJ
ca Eoo Mac
8'S
o
OT
E
C4i>-nO^
E .c O
w n>. 4C-
Ecoic>
.*o-- c X)
> Eoc
cn
8 E
8o
E
Q. -
EOO S3T*'
73 c
c
cCOi>=
o -or-
< q
3
(A
OCi m3
i3n oc
oc
-oWO
v_ o>
> a E
in
E O)
sf w^ oI- co
-2 o
8S CP 2<A
E to> too O 5 co O
a
5
5
oEoco-
8
a o
cn
*o*- "J
O3 t> cE _
J3
x*o0*i a~C
X)
C 0 g- a 2 J3 52
4= O
cn ..-25 ""O3 C *8 >cn, a
r_a
25 cn 5
E
n
.ts
ts au
co
oc
iinn
E
ac E
a *to=r
L. c0)
d)
4c-*
8
2
ccn
w
d) c
ToJ E
2 5a> c m *2
Si
un *- O c/) 2
c la *
| .9 W C< 8y
ao
in
8
3
X) c co TJ
(L0. 40-)
0) (0
8
S3
8 aS Q.
a o
co
S3 3
o TJ
TJ C
E oE
o
TJ S TJ
r(H0 T>
5Cu
CO
a
CkO. 0>)
$
k.
O
c
3
Flaws of LCAs
Sources and uses of data Poor quality, obscure models System definition - scope limitations Selection of effluent Inventory species Presentation of results
SP/-07506 CHM SVSTMS
Goals of Chem Systems/VI LCA
Scope - Complete systems analysis; full breadth of environmental concerns Credibility - Involve peer reviewers, avoid facile assumptions Flexibility - Lotus computer model Accessibility - Discuss assumptions, disaggregate data Objectivity - Avoid weighting factors or "bottom lines"
SPI-07507 CHM SVST6MS
FIGURE1
CSI LCA MODEL ARCHITECTURE
OUTPUT MATRIXES FOR COMPARATORS
CHM SVST6MS SPI-07508
HUMIOI fVMCI
FIGURE 2
MODEL COMPONENT RELATIONSHIPS
SPI-07509
FIGURE 3
EPA ESTIMATES OF 1988 EMISSIONS OF CRITERIA POLLUTANTS BY SECTOR
Total transportation
Stationary fuel combustion Electric utilities Industrial Commercial/institutional/residential
Industrial processes
Solid waste Incineration
Forest fires
Other miscellaneous
Total of all sources
SO
1.0
15.0 2.3 0.8 3.7 0.0 0.0 0.0
22.8
Million Short Tons/Year
NO.
PM
CO
8.9 1.5 45.3
7.8 3.4 0.6 0.7
0.1 0.2
0.0
21.8
0.5 0.3 0.2 0.7 1.1 7.4
2.9 5.2
0.3 1.9 0.9 5.9
0.1 0.7
7.6 67.3
VOC
6.7
0.0 0.1 0.8 9.4 0.7 0.8 1.9
20.5
rru trirrrtnmmT
SPI-07510
Emission data from the TRI
Public data base - designed for accessibility Manufacturing industries are compelled to report (fines up to $25K/day) Industry reports annually Basis constantly reviewed, updated, challenged, debated Serves as the benchmark for U.S. industrial performance
Provides a consistent, credible, and relevant basis, but has some deficiencies
SPI-07511
CHM SVST6MS
CHM SYSTMS
> Q 3 H V)
z
Q UJ OC UJ Q
CzO
z8
o u!
c OT
UJ J
UccJ
OC
o
UJ Q. O O (O
</)
oOC
(UOJ
o
oooz
UJ
s
8oz
gUI 5 i ifUWtI
E
ICoICI Iufci
8
<
X
a 3Z
a 5
a
I.UI1II
az
I2
Saco2
2
8
K B!
2
J____ I____ I____ L
FIGURE 5
ENVIRONMENTAL DISTRIBUTION OF TRI RELEASES AND TRANSFERS
TRI Environmental Distribution
Release to air (fugitive and point source) Release to surface waters Release to land disposal Public sewage treatment Off-site treatment or disposal Transfer to underground (deepwell) injection Total
Percent
38.90 5.79 9 00 9.14 17.70 19.47 100.00
FIGURE e
TRANSPORTATION VOLUMES AND MODES OF MAJOR CHEMICAL GROUPS
88838
esei `snoi & SNomm
V#
IS
CL
(/)
FIGURE 7
ANALYSIS OF SELECTED TRI EMISSIONS AMONG ETHYLENE PLANTS
(ON A SAMPLE OF 10 OF 35)
~i -i
lie
e> +
^1
ui s a
-HI t;
~ CO
8 PI
UI S
U00J
ui 3
*-*1 tt
(U0I
11EI
3
5
X
uzi 4
mh
| 88 * " " - 3 2 S M
Nouonaoud jo saNnod Nonun U3d saNnod
3
<
0 1
UJ
I o
VPO" ho>E (/>
& I
9LSZ0-ldS
KNNWeilOMMmH
viva Q3znvnuoN
svoi jo suwn aoNHauNoo ivnidaoNoo
e 3unou
Conclusions
PVC generally has some energy advantages over most alternatives Conclusions on environmental issues are difficult, uncertainties overwhelm differences Don't "micro-manage" economy to achieve environmental improvement Use LCAs to help improve each product or system individually LCAs are better than anecdotal approaches Need much more work on LCA methodology and data base
SPI-07517 CHM SVSTMS
Peer review feedback
General approval of methodology - TRI use, computer program Past LCAs accurate and useful on energy analysis, but not on environment Differences between U.S. and European analyses - feedstocks, processes, transportation, data
available
SPf-07518 CHM 5VST6MS