Keith Christman - Sun Aug 7 09:44:03 2005

The Vinyl Institute has developed emissions factors using TRI data for all EDC/VCM/PVC plants as discussed in our response to the TRI dioxin emissions subsequent question. The Task Group’s table on dioxin emissions has been reproduced and will be posted here. In this table is a calculation based on the TRI for emissions in 2003, the most recent year currently available.

More complete data is available at the website http://www.trifacts.org. Data presented there shows that emissions of CDD/CDF to the open environment from the US chlor-vinyl chain has decreased over 70% from 2000 to 2003.

General Comments

The Vinyl Institute appreciates the Task Group’s interest in accurately quantifying releases of dioxin to the environment and VI shares this interest. In 1998 the vinyl industry working with USEPA characterized releases of dioxin to the environment for inclusion in the US EPA’s Inventory of Sources of Dioxin and Dioxin-like Compounds. http://www.epa.gov/ncea/pdfs/dioxin/2k-update/pdfs/Diox in_Chapter_8.pdf This characterization program resulted in dramatic reductions in releases of dioxin to the environment.

Unfortunately, data is still lacking to adequately characterize releases of dioxin from many of the alternative processes being considered by the Task Group. According to US EPA, data are inadequate for inclusion in the dioxin inventory for the following processes associated with iron and steel production; coke production, electric arc furnaces, and foundries. EPA’s best estimate for air emissions, published in their most recent draft, totals 104 grams TEQ/year for these processes including sintering plants. Similarly emissions factors do not exist for primary aluminum production. Air emissions from cement kilns have been adequately characterized and totaled 80 grams TEQ in 2000. Please note that these are emissions to air only and do not include any residuals going to landfills.

For comparison purposes the Task Group may wish to consider some of the available emissions factors for alternative materials considered by the Task Group http://www.epa.gov/ncea/pdfs/dioxin/2k-update/pdfs/Diox in_Chapter_1.pdf.

Air only Iron Ore Sintering 0.62–4.61 ug WHO TEQ/tonne sinter Air only Secondary Aluminum 22.4 ug WHO TEQ/tonne aluminum scrap feed Air only Cement Kilns not burning hazardous waste 0.26 ug WHO TEQ/tonne clinker

These emissions factors appear to be larger than current emissions factors for vinyl manufacturing and yet do not include all of the processes needed to produce the alternative materials. The Task Group should take great care in utilizing dioxin emissions factor data and recognize the lack of data for many alternative processes.

We would like to reemphasize the importance of using only emissions to the open environment for comparisons to alternatives in this study, for two reasons. First, according to US EPA "Properly designed and operated landfills are considered to achieve long-term isolation from the circulating environment (The Inventory of Sources and Environmental Releases of Dioxin-Like Compounds in the United States: The Year 2000 Update (External Review Draft, March 2005; EPA/600/p-03/002A) http://www.epa.gov/ncea/pdfs/dioxin/2k-update/."

Second, US EPA’s Inventory does not have estimates of dioxin in residues from alternative materials being considered by the Task Group. Similarly, the United Nations Environment Programme Standardized Toolkit for Identification and Quantification of Dioxin and Furan Releases indicates that there is no data available for emissions factors for residues from many of the alternative processes being considered. This makes apples-to-apples comparisons difficult.

Regarding the German data, according to the European Council of Vinyl Manufacturers (ECVM) Voluntary Commitment Charter, all European facilities, including EDC/VCM facilities in Germany must emit less than 1 ug TEQ of dioxins /tonne of oxychlorination EDC produced to effluent water. Assuming the balanced process with 50% of the EDC coming from direct chlorination and 50% coming from oxychlorination, this would result in a estimated emission factor to water of 0.5 ug/ tonne of EDC or about 0.7 ug/tonne of VCM . This is comparable to the US reported emissions of 0.2 ug/tonne. The EVCM charter is on the web and can be found in a search for "ECVM charter" see below . We expect that all European facilities are meeting this commitment.

www.ecvm.org/img/db/SPVCcharter.pdf

The Quass study, published in 1997, is based on even earlier information; to our knowledge, the data is older and was not acquired in an organized study. This makes its relevance to 2005 performance marginal. More importantly, however, the Task Group should follow the protocol set by US EPA and focus its analysis and comparisons to alternative materials and processes on releases to the open environment not residue disposed of in hazardous waste or other landfills.

Tom Lent - Sun Aug 7 19:29:29 2005

There have been several efforts to estimate dioxin emission factors from manufacture. Pat Costner compiled a series of them in a paper that will be submitted today with this response. Based upon dioxin releases solely as reported by U.S. producers of EDC/VCM/PVC in the 2002 Toxics Release Inventory, Costner obtained a factor equivalent to the 400 ug I-TEQ/tonne of the German study.

While the Task Group is to be lauded for making the effort to ascertain a hard number for dioxin emissions, several issues make calculation of a reliable conservative dioxin emission factor unlikely to be attainable. Calculations relying upon the TRI to be specific are likely to be under representative of actual flows and should only be used as lower bounds.

TRI data generally do not represent actual audited real time measurements of dioxin emissions. Rather they represent an industry self reported modeling exercise estimate. EPCRA Section 313 reporting does not require testing or monitoring. Reporting may be based upon extrapolation based upon spot testing of the facility (e.g., results of a quarterly one day test multiplied to estimate annual flows) or on simple application of emission factors calculated for similar processes in other plants to the annual production of the subject plant. Isolated tests are likely to represent best case operating conditions, not the reality of chemical plant process variability and upset conditions.

The problem noted by the Task Group with mercury wherein large discrepancies have been found between TRI reporting and actual mercury purchases is indicative of the gross level that TRI reporting can vary from actual operations. With an unintentional byproduct like dioxin that will therefore be subject to no alternate reporting for audit, the probability of missing highly significant flows is even greater than with mercury.

The substantial body of evidence collected by Greenpeace and others of high levels of dioxin in waste water outfall and sediments downstream from EDC/VCM/PVC facilities as well as in neighborhood air samples is indicative that facilities emissions are likely much higher that TRI reporting may indicate (see for example, Costner, “PVC: A Primary Contributor To The U.S. Dioxin Burden” in the TSAC database).

It is highly important to include in the analysis all production of dioxins, not just the subset classified as “emissions to the open environment.” The fate of the much larger volumes of dioxins generated is in no way assured. Destruction in hazardous waste facilities is not fully monitored, but rather is industry estimated based upon emission factors negotiated between industry and the EPA. Containment in landfills is also a questionable solution, particularly for such a highly potent persistent bioaccumulative toxicants like dioxin. Precaution informed by real life failures of incinerator and landfill technology dictates that all production should be evaluated, not assumed away based upon supposed normal operating procedure factors. The stakes are too high to ignore.

As for the dioxin emissions associated with alternative materials, we would welcome efforts directed at good analysis of significant dioxin releases from the life cycle of other building materials. Further investigation of some other building material dioxin sources have pointed back at PVC or other chlorinated materials (like solvents). For example copper recycling dioxin releases have proven to be largely related to PVC sheathing on copper wire recyclate. Preemptive removal of PVC sheathing before smelting has helped drive down copper recycling dioxin emissions. Cement kilns fired by hazardous waste incinerators have also been identified as s significant source and have been targeted along with PVC and other chlorinated materials for building guidelines to avoid generation of dioxin. It is important to differentiate between materials whose dioxin emissions can be eliminated rather easily (such as copper through removal or elimination of PVC sheathing) versus those like PVC for which dioxin generation is inherent to the manufacturing process and/or end of life disposition and hence not reasonably avoidable.

It is worth noting again at this stage that landfill fires contribution to dioxin generation may overshadow even the most comprehensive estimates of dioxin emissions from manufacture as noted in our original comments on the Task Group paper. PVC must take the responsibility for this with the available chlorine content for dioxin generation in landfills largely driven by PVC. As noted by the EPA: “Open burning of municipal waste or construction debris containing chlorinated plastics or other chlorine-containing materials would be expected to emit levels of CDD/CDF [dioxins] comparable to or higher than those from uncontrolled MWC emissions. On the other hand, open burning or structure fires in which the materials being burned are low in chlorine content (e.g., wood, yard waste) would be expected to have lower CDD/CDF [dioxin] emissions. Several studies have identified strong correlations between chlorine content and CDD/CDF emissions during combustion tests.” (U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Office of Air And Radiation, 1997. “Locating And Estimating Air Emissions From Sources Of Dioxins And Furans”, EPA-454/R-97-003, Research Triangle Park, North Carolina, May 1997). Reinforcing this, the Best Available Technology guidance from Stockholm Convention includes PVC avoidance

Keith Christman - Mon Aug 8 18:41:44 2005

Mr. Lent references Pat Costner’s (Greenpeace) analysis of TRI data for 2002. The Task Group should note that Ms. Costner includes dioxin-containing wastes going to treatment in her analysis not releases. This dioxin is destroyed through high temperature combustion processes and NEVER released to the environment. Dioxin and furan air emissions from these processes are regulated under the Clean Air Act and are included in air emissions under the Toxics Release Inventory. The Task Group should focus on releases to the open environment air, water and LAND SURFACE following EPA's inventory since this is the information most relevant to risk assessment and LCA.

Mr. Lent raises concerns about the use of TRI data for analyzing dioxin emissions. We agree that there are some limitations to the use of TRI data for dioxin quantification. These concerns may be valid for processes associated with alternative materials which have not been well quantified according to US EPA. Limitations have been noted by EPA (see page 1-8 http://www.epa.gov/ncea/pdfs/dioxin/2k-update/pdfs/Diox in_Chapter_1.pdf ) We have generally used the US EPA Inventory of Sources of Dioxin as our reference for dioxin quantification in the US. In EPA’s Dioxin Source Inventory, US EPA has characterized vinyl industry data for dioxin emissions as “adequate for quantitative estimation with medium confidence in the emission factor and at least medium confidence in activity level.” EPA’s confidence level in dioxin data for the vinyl industry is far higher than other materials considered by the Task Group including cement and metal categories (see table 1-7 http://www.epa.gov/ncea/pdfs/dioxin/2k-update/pdfs/Diox in_Chapter_1.pdf ). As shown in our comments above, TRI data and EPA’s Inventory match up reasonably well for the vinyl sector because of the work done on quantifying dioxin emissions by our industry in cooperation with USEPA.

On landfill fires Mr. Lent references a US EPA paper that is nearly a decade old. On May 6, 2005, US EPA released a new comprehensive review of dioxin emissions in the United States. The Task Group should refer to this EPA document if it needs more information on landfill fires or other issues raised here by Mr. Lent (The Inventory of Sources and Environmental Releases of Dioxin-Like Compounds in the United States: The Year 2000 Update (External Review Draft, March 2005; EPA/600/p-03/002A http://www.epa.gov/ncea/pdfs/dioxin/2k-update/pdfs/Diox in_Chapter_1.pdf) Mr. Lent also suggests that Best Available Technology for the Stockholm Convention includes PVC avoidance. We suggest the Task Group directly review Stockholm Convention BAT requirements. The US vinyl industry practices BAT as outlined by the Stockholm Convention see page 260 http://www.pops.int/documents/meetings/cop_1/meetingdoc s/en/inf_7/INF-7.pdf The Convention does NOT promote PVC avoidance. Best Available Techniques for open burning includes a number of measures that may marginally decrease emissions of dioxins, furans and countless other pollutants. The only real answer is to eliminate the practice.

And realistically, how much open burning of the materials we are discussing actually occurs, given that it is illegal in most of the country?

On the topic of overall incineration the Stockholm POPs Secretariat has actually stated: "Given the current state of the research, the suggestion that the BAT/BEP guidance 'advocate' the elimination of PVC as a best practice or technique has not been incorporated". Footnote: Secretariat Note to BAT/BEP Expert Group regarding Best Available Techniques for waste incineration. http://www.pops.int/documents/meetings/bat_bep/3rd_sess ion/EGB_3_infs/EGB3_INF_5_note_Muniwaste.pdf The suggestion on PVC that Mr. Lent refers to applies only to the intentional open burning of waste which occurs in some developing countries. The BAT document states “Open burning is an environmentally unacceptable process that generates chemicals listed in Annex C of the Stockholm Convention and numerous other pollutant products of incomplete combustion. Consistent with Annex C, Part V, section A, subparagraph (f) of the Stockholm Convention, the best guidance is to reduce the amount of material disposed of via this method with the goal of elimination altogether.” We support this approach. Recognizing that this may not be immediately feasible in some developing countries the BAT document goes on to say “Other techniques which may effect improvement include, with respect to the materials burned: avoid including non-combustible materials, such as glass and bulk metals, wet waste and materials of low combustibility; avoid waste loads containing high chlorine content, whether inorganic chloride such as salt, or chlorinated organics such as PVC; and avoid materials containing catalytic metals such as copper, iron, chromium and aluminum, even in small amounts. Materials to be burned should be dry, homogeneous or well blended, and of low density, such as non-compacted waste.” (see page 49 and 50 http://www.pops.int/documents/meetings/cop_1/meetingdoc s/en/inf_7/INF-7.pdf )

As one of the most abundant chemical entities in the crust of the Earth, chloride is found in numerous organic (e.g., PVC) and inorganic compounds (e.g., salts), both natural and synthetic. Many of these compounds are present in waste. Consequently, efforts to starve chloride out of the waste stream via restrictions on vinyl would not be effective. In fact, ordinary waste contains more than a million times more chloride than is necessary to generate dioxins and furans. Cconsequently, policies focused on removing the chloride contributed by PVC would not impact dioxin and furan formation when actual mixed waste is burned, and this has been shown in numerous scientific studies (see VI comments April 2004).

This is further reinforced in that the BAT/BEP guidance regarding high-chloride specifically notes that there is no difference between inorganic and organic sources of chloride.

The Task Group should use dioxin emissions factors to the open environment for vinyl from TRI. Air 0.07 ug WHO TEQ/tonne of EDC and Water 0.19 ug WHO98TEQ/tonne EDC.

Keith Christman - Thu Aug 11 12:40:09 2005

Dioxin the Big Picture

In all of this continuing discussion about dioxin and vinyl's role in dioxin emissions, let's not forget the fact that dioxin levels in the environment have been falling for decades (Hagenmaier H, Walczok M, Organohalogen Compd. 1996;28:101-104. Alcock RE, Jones KC. Environ. Sci. Tech. 1996;30:3133-3143) and US EPA now estimates that dioxin emissions from quantified sources have been reduced by 89 percent between 1987 and 2000. ( http://www.epa.gov/ncea/pdfs/dioxin/2k-update/pdfs/Diox in_Chapter_1.pdf ). During this time vinyl production increased by about 60 percent.

The Vinyl Institute agrees that decisions should be driven by data and not speculation. Mr. Lent quotes Costner, 1995: PVC: A Primary Contributor to the US Dioxin Burden, and notes that samples of internal plant waste and river sediment were taken and analyzed for dioxin. In fact, in this paper, only four samples were analyzed for CDD/CDF. Three were stolen samples of in-process waste material that was awaiting destruction, but was not emitted. The fourth sample was of sediment taken downstream from the discharge of the then-Geon LaPorte facility. CDD/CDF fingerprints for the sediment sample and for the heavy ends waste samples are shown in the attached Figure 1 (see posted file fingerprints from costner paper). Note that the congener distribution for the authentic waste samples is virtually all furans. The sediment sample is virtually all hepta- and octa-chloro dioxins. Clearly, the sediment and the waste came from two different sources.

Moreover, comparison of the sample taken downstream from the discharge of the Geon facility with the USEPA Background Sample for Sediments, indicative of an average for the US, shows that the two are indistinguishable (Figure 2).

Lent is incorrect; there is nothing unusual about the sediment found by Greenpeace downstream of an EDC/VCM/PVC facility. This Greenpeace sediment sample is indistinguishable from USEPA background sample for sediments but clearly distinguishable from vinyl waste samples. THE FINGERPRINTS DO NOT MATCH.

Even the title of this reference cannot be correct. If PVC is a "Primary Contributor to the US Dioxin Burden" then the two should be directly correlated; it has been shown repeatedly that production of PVC and dioxin in the environment are inversely correlated as discussed above.

Tom Dietsche - Thu Feb 16 16:23:17 2006

This posting by Tom Dietsche, USGBC LEED Program Manager, is on behalf of Tom Lent, who originally posted it on 8/11/05 (and a technical problem caused the text to disappear).

Mr. Keith Christman restates a frequently made assertion of industry that “dioxin levels in the environment have been falling for decades… and US EPA now estimates that dioxin emissions from quantified sources have been reduced by 89 percent between 1987 and 2000. … During this time vinyl production increased by about 60 percent.”

We would not expect a direct correlation between vinyl production and dioxin levels in a period era when other previously major sources of dioxin were undergoing dramatic reductions. As the EPA notes in the document Christman cites: “By 1995, EPA had proposed or promulgated regulations limiting CDD/CDF emissions from municipal waste combustors (MWCs), medical waste incinerators (MWIs), hazardous waste incinerators, cement kilns burning hazardous waste, and pulp and paper mill facilities using bleached chlorine processes.” (http://www.epa.gov/ncea/pdfs/dioxin/2k-update/pdfs/Dioxin_Chapter_1.pdf ) In each case this, has led to dramatic reductions in emissions from those sectors. For example medical waste incinerators have been reduced from a high of 5000 down to about 100 today and hospitals have substantially reduced the quantity of PVC medical equipment going in to those remaining incinerators, leading to a precipitous decline in the EPA’s estimates of dioxin emissions from that sector from 2440 g/TEQ in 1987 to 378 in 2000. A close look at the graphs the Vinyl Institute offers in making this point indicate that the decline they cite has flattened off. If the Institute is correct in assessing that the bulk of PVC production has yet to hit the landfills we can expect significant upward pressure on those levels as more PVC begins to feed landfill fire contribution.

Christman makes various references to fingerprint issues with data from Pat Costner’s analyses which we could inquire with her about. If you will forward to me the figures that he references, I will be happy to forward this information on to Ms Costner for reply.

Christman states that “dioxin in wastes going to treatment “is dioxin is destroyed through high temperature combustion processes and NEVER released to the environment” I find this claim to be rather incredible. I don’t know of an incinerator manufacturer who can truthfully guarantee zero releases of dioxins to the atmosphere. They all claim very high levels of destruction (but not zero) and even these assertions are substantiated only by modeling from static controlled condition tests, not from continuous real time monitoring of actual dynamic operating conditions with the variable feeds and upset conditions that are inevitable in real life operations.

Christman notes that the “TRI data and EPA’s Inventory match up reasonably well for the vinyl sector because of the work done on quantifying dioxin emissions by our industry in cooperation with USEPA.” I should hope so since they are both industry self reported data sources.

Christman notes “On landfill fires Mr. Lent references a US EPA paper that is nearly a decade old. On May 6, 2005, US EPA released a new comprehensive review of dioxin emissions in the United States. The Task Group should refer to this EPA document if it needs more information on landfill fires or other issues raised here by Mr. Lent (The Inventory of Sources and Environmental Releases of Dioxin-Like Compounds in the United States: The Year 2000 Update (External Review Draft, March 2005; EPA/600/p-03/002A http://www.epa.gov/ncea/pdfs/dioxin/2k-update/pdfs/Diox in_Chapter_1.pdf)” Inspection of the May 2005 EPA document shows that their estimate of landfill fire contribution remains unchanged from the earlier document and they conclude that landfill fires have the “potential to significantly increase the present inventory if preliminary release estimates are confirmed.”

Christman suggests that Stockholm BAT PVC avoidance recommendations only apply to “intentional open burning of waste which occurs in some developing countries.” Reference to the EPA Inventory estimate for backyard barrel burning makes clear that the EPA does not consider this a practice that is restricted to developing countries.

Christman further asserts that “efforts to starve chloride out of the waste stream via restrictions on vinyl would not be effective… Consequently, policies focused on removing the chloride contributed by PVC would not impact dioxin and furan formation when actual mixed waste is burned, and this has been shown in numerous scientific studies (see VI comments April 2004).” You will remember that this assertion and the study that they frequently cite that was purportedly carried out by the ASME have been thoroughly rebutted several times over the years that this debate has been carried out at the USGBC. For both reviews of the politics of the ASME study and analysis of its data and of a large body of relevant studies that have found strong correlations between PVC content and dioxin formation, refer to testimony that Dr. Joe Thornton submitted at the very beginning of this process and to his “Environmental Impacts of Polyvinyl Chloride Building Materials,” published by the Healthy Building Network in 2002 (http://www.healthybuilding.net/pvc/Thornton_Enviro_Impacts_of_PVC.pdf) and submitted to the committee in the past.