TSAC was charged with preparing a scientific assessment of the evidence on PVCs environmental and health hazards in order to evaluate a credit for PVC avoidance in the U.S. Green Building Councils LEED standard. For more than a decade I have studied the environmental impacts of vinyl, and I have been a participant in the USGBC evaluation of vinyl since its inception. I am the author of the most comprehensive analysis of the PVC lifecycle of which I am aware (Thornton 2002), and a peer-reviewed book on the subject published by a major academic press (Thornton 2000c).
The TSAC report is of extremely poor scientific quality. The report is based on a novel, unvalidated combination of two methodologies that are widely viewed as problematic among environmental scientists lifecycle assessment (LCA) and risk assessment (RA). These questionable methodologies are then applied in a shoddy and incomplete manner. This document would be unsuitable for publication in a quality peer-reviewed journal or for acceptance as a graduate student thesis; it certainly does not provide an adequate basis for a policy decision by a major public organization.
The problems with the report fall into two major categories. First, environmental and health hazards of great consequence were omitted. TSACs method treats as non-existent hazards for which data are inadequate to derive a quantitative estimate of risk and those that are deliberately excluded. Surprisingly, TSAC has chosen to exclude many of the major hazards associated with the PVC lifecycle, and it has completely ignored the most important one the global distribution and accumulation of persistent toxic substances produced in the manufacture, disposal, and accidental combustion of vinyl.
Second, TSACs report was surprisingly unbalanced. A second flaw with RA and LCA is their intense dependence on contestable choices and assumptions made by the assessor, and the ease with which the results can therefore be manipulated. Indeed, it has been shown that the subjective choice of different assumptions can result in a 100 million-fold difference in the results of a toxicological risk assessment (Cothern et al. 1986). Unfortunately, TSACs assessment appears to suffer from a systematic bias that results in a radical underestimation of the hazards associated with PVC.
Due to these problems, which are discussed in more detail below, the TSAC report cannot be accepted as an accurate assessment of PVCs environmental costs compared to alternatives.
1. Failure to conform to scientific standards of reporting and transparency. Any scientific paper or report must provide adequate detail for the methods to be fully understood and replicated. The TSAC report provides some general information on the equations used and the types of hazards considered in the RA and LCA, and it then skips straight to the results. Specific information on the emissions and toxicology input values used for each material -- and where this information came from -- is not provided in most cases. Thus there is no way for the critical reader to know precisely what risks were assessed or whether the emissions and exposure estimates used are credible.
TSAC did supply upon request spreadsheets containing data and calculations relevant to the LCA and RA, but these do not provide adequate information to evaluate the risk calculations made. For example, the LCA spreadsheet contains estimates of the emissions factors for the hundreds of specific pollutants considered in the LCA, but it provides absolutely no explanation of the source of these estimates and the assumptions embodied in them. Some of these emissions figures completely lack credibility: for example, the estimate of dioxin emissions for linoleum and for cork are higher than many of the PVC entries, despite the fact that PVC manufacture and disposal is recognized as a significant source of dioxin, but no dioxin has ever been associated with any specific aspects of the linoleum or cork lifecycl o 'I`4E Asst Professor < A.
d~w Qi o < < Mate x3E (I University of Oregon S~w d~wQ o &. " ! (I5E) G7P d5EI5EI0F field H5E5E/ 5ENG5EX5EkJ submitterId 5E 6E) %.q 5EvI06E5E pages 5EIK) o XT <6E0I 5E lines a DD $ O P~w & d~wXQ o B. & < A d~wxQ
o < <~w8 ' d~w0Q o P~w ) d~wHQ o BII ) < A~w8 d~whQ o < <. * d~w Q o : a 0G@;E A d~wrI o <
A d~w@GV o <
< d~wtI o
>
> : : p9EG A d~wPqI o < A o d~w F4 o < < d~w sI o > o > o : G8JE
S o d~wpI o S d~wqI o ' S d~wuI o S d~wuI o ' c d~whsI o A < < d~wxuI o
+ # P~wPw d~wpI o B o < A d~wGH o < < d~w@sI o S~wPw d~w8Q o &
P~w d~wȞQ o B < < d~wrI o + " P| d~wQ o B < A d~w o < < d~wptI o d~w`pI o P d~wМQ o B o < A d~w o < < d~wrI o S|
d~wtI o TxDJx
& * * # P d~w rI o > o > =E mE S d~wQ o S d~wQ o ' S d~w(qI o S d~w qI o ' c d~wQ o A < < d~wXrI o
+ " P- d~w@Q o B < A d~w:G o < < d~wpI o S- d~wHQ o &
P| d~wQ o B < < d~wQ o + ! P| d~wQ o B < A d~w o < < d~wQ o P d~wQ o B < A d~w o < < d~wQ o S|
d~wpI o TxDJx
& * * " ! P d~wQ o " > " > # mE
S d~w d~w`Q o ( S d~wQ o ( ' d~w ( S d~wQ o ) S d~w d~wQ o ) ' d~w ) c d~w d~wXQ o * A < * < d~w0Q o *
* + d~w % * S- d~wQ o + T. d~wPQ o + & d~wQ o + P d~wE o , B < , A d~w0mE? o < , < d~wQ o , S
d~w(Q o , & d~w
, P~w d~wQ o - B < - < d~wQ o - + d~w $ - P| d~wQ o - B < - A d~w o < - < d~w d~w0Q o - P d~wȗQ o - B d~w < - A d~w o < - < d~wxQ o - S| d~w Q o - TxDJx - & - * d~w - * % . P d~wQ o / > / > 0 pEeE rs eEH M _aryMaterials x HPeE is_array states (eEpE dbc hfEnE dEeE odbc_result eE eE rs ^ PpExfE _aryReviewStates fE fE d~wodbc_fetch_row eE@fE rs d~w fEfE ! iEpeE eEfE
odbc_exec xfEfE fEeE dbc @fEgE _aryReviewStates
HjEJ d~w commentsd:\apache\pvc.buildinggreen.com\htdocs\includes\pvcdb.inc.php174 c.php8
iEppEiEhfE fEgE odbc_result gEXhE rs hE hE
odbc_exec DJx gEH _aryReviewStates H H dbc M gEhE odbc_result XhEgE rs iEhE odbc_fetch_row hEhE rs d~w hE@iE ! IhfE HhE dbc hEiE _aryMaterialGroups jEiE
d~waryStates w piEiE 0 d~w iE kE No 0lExgE @iEjE odbc_result iEjE rs jE`jE getyesno DJx 8jEH _aryMaterialGroups gEpiE getyesno jEjE odbc_result jE8jE rs iEkE
aryStates w lEhlE@ G HI h
G ;G hlElE ces T . kEHlE 1 HlEH
d~waryStates . HlE sources kE(kE d~w mEiE kEkE Yes (kEkE@ =G H hH kEXH sources XHH < I0lE 8JEWE@ o select impactCategoryId, impactCategoryName from ImpactCategory HQG dbc G 8GmE dbc mEmE pvc mE nE o _aryMaterials mEnE rs @oEHnE odbc_fetch_row nEhnE rs d~w HnEoE $ peEhQ nEnE
odbc_exec nEnE d~w nE@oE dbc @II I I ID G nE nE select materialId, applicationName+' - '+materialName from Material m left join Application a on m.applicationId=a.applicationId hnEpE _aryMaterials PeE(pE dbc pEPpE is_array d~w (pEeE rs xgEpE oEdE xCDodbc_result ppE;G GqE# `tE @I 00 GMT : pE0qE 1;s:P>G qEXqE :3:{ >G3 0qE rE :2;s xtEqE' "Retj| qEhqEqELG dbc :"st
qE9G1 ;a:3 qEI(:GqE g_siteOptions ";i:2;s:0:
12:G8HF XqEPtE :"Stp? 1;s:1:"=";i:2;s:0:"";i:3;s:15:"getReviewStates";}s:17:"g.materialGroupId";a:4:{i:0;s:14:"Material Group";i:1;s:1:"=";i:2;s:0:"";i:3;s:17:"getMaterialGroups";}}comment.filters|a:5:{s:5:"pages";a:4:{i:0;s:5:"Pages";i:1;s:1:"=";i:2;s:0:"";i:3;s:15:"getCommentPages";}s:15:"commentCategory";a:4:{i:0;s:16:"Comment Category";i:1;s:1:"=";i:2;s:0:"";i:3;s:14:"getCommentCats";}s:14:"responseStatus";a:4:{i:0;s:6:"Status";i:1;s:1:"=";i:2;s:0:"";i:3;s:16:"getResponseStats";}s:18:"respondentUserName";a:4:{i:0;s:10:"Respondent";i F rExtE 8:"g0rE 1:"q PtEqE e/Or Domme AGAG :2;scolor }} H uE( 8uE8uEvEIR tE(uE H uEhuE# DuEuEuE (uEuE MG huEuE# uExuEvE8uE uEuE LG uE8vE# vEuE uE uE`vE pvE 8vEvE( wE wE`xEIR `vEvE =G vEvE xLG vE0wE# wEvEhwE vEXwE 8LG 0wEwE# twE@wEwE wE XwEwE LG wE xE# wEwE`xEhwE wE(xE 8xE xEPxE getReviewStates (xExE# lxExE wE PxExE0 W
xEHvEIQH reviewStatus xEyE( yEyE{E(yER xEPyE yEzEzE{E yExyE yE PyEyE Material Group xyEyE# yE`yEzE yEzE XLG yEHzE# $zEyEzEyE zEpzE KG HzEzE# zEXzE{EzE pzEzE zE zE{E getMaterialGroups zEI# ${EzE zE II = ges A @E`E) S d~wI o ^ ? d~w o ^ S~w d~wpI o _ & d~w zI o _ P~w d~wЯI o a Q. d~wI o a S~w d~wI o a T. d~wI o a & a S~w d~w zI o b T. d~wzI o b &
d~w`{I o b S~w d~wzI o c T. d~w {I o c &
d~w {I o c P~w d~w@{I o e Q. d~w{I o e d~wP o e + e S~w d~w@zI o f T. d~w`zI o f & d~wzI o f * g P~w d~wzI o g Q. d~wP|I o g d~w o g + g S~w d~w{I o h T. d~w{I o h & d~w|I o h * i P~w d~w }I o o Q . d~w}I o o d~w@}I o o P~w d~w`}I o o Q . d~w}I o o o S~w d~w0|I o o & ! o P~w " d~w|I o s BD " < s A d~w o < s A d~w o < s < # d~w}I o s $ # d~w|I o s / $ $ 5 s P~w % d~w}I o s B % < s A d~w o < s A d~w o < s = d~wE o { P~w ? d~wE o { @ > ? { S~w 9 d~wI o { & A 9 @ { * _ | P~wPw C d~wE o } D C d~wE o } P~w E d~wE o } F D E } S~w B d~wE o } & G B F } < H d~w{I o + H t 5 I 7 I I d~wx|I o 7 I I d~wE o 7 I I d~wE o 7 I I d~wE o 7 I I d~w0E o 7 I I d~wE o 7 I I d~wE o 7 I I d~wE o P~w J d~w(E o 8 I I J 7 I I d~wHE o P~w K d~whE o 8 I I K 7 I I d~wE o 7 I I d~wE o 7 I I d~wE o ( I * x P~w L d~wPE o M d~w(E
o L A M < < N d~wE o O > cA EE S d~wxE o ? d~w o S d~w0E o ? d~w o S d~wXE o ? d~w o P~w d~wE o + S~w d~wE o T . d~wPE o & d~wE o * P~w d~wpE o S~w d~wE o & P~w
d~wE o P~w d~wE o
- V~w
d~wE o $
F * P~w d~wE o P~w d~wE o S~w d~wE o T 0| & * P~w d~wE o > > c lD(E
S d~wjD o ? d~w o P~w d~wjD o |
Cothern C, Coniglio W, Marcus W. Estimating risk to human health: trichloroethylene in drinking water is used as an example. Environ. Sci. Technol. 20:111-116, 1986.
Melnick RL, Brody C, Huff J. The IARC evaluation of DEHP excludes key papers demonstrating carcinogenic effects. Int J Occup Environ Health. 2003 9:400-2.
Melnick RL. Is peroxisome proliferation an obligatory precursor step in the carcinogenicity of di(2-ethylhexyl)phthalate (DEHP)? Environ Health Perspect. 2001 109:437-42.
Melnick RL. Suppression of crucial information in the IARC evaluation of DEHP. Int J Occup Environ Health. 2003 9:84-5.
Melnick RL. The IARC evaluation of di(2-ethylhexyl)phthalate (DEHP): a flawed decision based on an untested hypothesis. Int J Occup Environ Health. 2002 8:284-6.
Neurath C. PVC's role in dioxin emissions from open burning:
New analysis of US EPA data. Organohalogen Compounds 66:1146-1152, 2004.
Thornton J. Environmental Health Impacts of Polyvinyl Chloride (PVC) Building Materials: A briefing paper for the U.S. Green Building Council prepared under the direction of the Center for Maximum Potential Building Systems and the Healthy Building Network, 2000a.
Thornton J. Final Rebuttal: Environmental Impacts of Polyvinyl Chloride Building Materials - A Briefing Paper for the U.S. Green Building Council. Submitted by the Center for Maximum Potential Building Systems and the Healthy Building Network, December 2000b.
Thornton J. Pandoras Poison: Chlorine, Health and a New Environmental Strategy. Cambridge: MIT Press, 2000c.
Thornton J. Environmental Impacts of Polyvinyl Chloride Building Materials: A Healthy Building Network Report. Washington DC: the Healthy Buiilding Network, 2002. |