Document aJ5or33bEEGnr8171z7DkNw2Y

. October 20,,99/^Sj^ ------------------------------------------------------------------------------------- To: Don Kerr (416) 920-4905 <p6* I 'litfale Cc: Walter Banas Bob Dickinson Dick Murry Susan Peterson . John Shepherd Dave Sheppard Jack Soule Don Wilke Bill Young Marshall Kem Geoff Granvffle Don Hames Pierre Guimond Gordon Lloyd Deep Khosia Keny Mattila (416) 240-5231 (514)366-5665 (514) 745-2031 (905) 669-3497 (905)454-4916 (519)452-6015 (613)548-5240 (513) 626-3522 (519) 339-8510 (519)339^510 (403)691-2224 (519)339-3417 (613)230-9326 (613)237-4061 (416) 730-4449 (613) 236-4280 Fr: Roger Keefe phone (416) 968-4435 fax (416)968-8415 Re: Cost Impact of NSN Regs - Draft Study Don, Win you request CCPA to set up a conference can for middie to end of this week on the subject study? It is important to review what has happened since the CEPA-ICG meeting and come to agreement on our feedback to Environment and Health Canada on the subject study. After finally getting the document and giving it a cursory review, I think it is imperative we consider as a first option, "atonewaffi/ig"the study. Only as a distant second, should we offer the excellent upgrades suggested by John Shepherd at the last ICG meeting. The underlying methods used by Environment and Health Canada must be stopped before they are used a third time.1 ...J2 CCME. 1995. Cleans' Vehicles and fuels. The draft study vyas heavily criticized by industry which fciled to offer an alternative to the cost/beneftt approach to vehicle emissions. Industry concerns were ignored, and the reports were never finalized (to my knowledge). The petroleum sector is now OCT-20-1996 21:14 ROGER KEEFE P.01 SL 110197 2- - A similar approach was taken by Environment and Health Canada on vehicle emissions. The emissions cost vs. benefits study was just as technical as the one on the NSNR we are considering. In sections dealing with inherent toxicity, exposure, and valuation of benefits, government gave the appearances of fairness, balanced judgement and allowed for uncertainty. When the results of these three components of benefits were integrated, reductions in respirable particulates were predicted (largely by modeling) to be improved a norvdetectable 2-3%, resulting in predicted morbidity and mortality that was simflarty undetectable from background. Nonetheless, the health. benefits were calculatedto be $30B (that's right, billions!). While sttt in draft, the S30B value was leaked to the media from the CCME Ministers' meeting in Yellowknife late last year. I focused on the vinyl chloride chapter in the NSNIWG draft study, and what I expect to see in the headlines very shortly is that the VCM and PVC manufacturers are killing as many as 11 people in the community around their six plants, and an additional 1 to 3 workers, all from cancer. You can see the costs in the relevant tables may exceed $100M. Neither government nor Its contractor bothered to get any facts from industry (that I'm aware) about numbers of people exposed, levels of exposure, or our experience with hemangiosarcoma incidence/rnortality. Making assumptions, the contractor, Calculated an exaggerated cancer incidence at low exposures; Over-estimated exposures; and, Grossly over-valued the health benefits from VCM control / ban.^r' The results do not agree with experience. The number of hemar associated with occupational VCM exposure in the world is ordv 681 the contractor), and I do not believe any of these were in the pare currently in operation. )mas ; (according to i facilities CPPI has been encouraged to investigate two approaches: 1) media, and 2) legal, in the former, the goal would be to grab the initiative, point the finger at governments use of "casual, back-oMhe-envebpe" calculations, surrounding them in very technicalsounding camoflage, and creating fear (when there are virtually zero "risks") - all to justify their continuing existence in an age of budget-cutting. In the second approach, senior bureaucrats would be advised that industry would seek recovery of damages (including punative) for any unfounded allegations of harm. Would ICG consider? % supports a joint effort to get a more sensible balance between cost vs benefits. The cost of the joint initiative is quite large, and the results could still drive a (yet another) major rationalization of the petroleum industry in Canada. OCT--20-1996 21:15 ROGER KEEFE P.02 SL 110198 Draft Report ____ VALUING THE BENEFITS OF THE NEW SUBSTANCES NOTIFICATION REGULATIONS ' Prepared Ion ' Environmental Protection Service Environment Canada Prepared by: Lara B. Sheer, Anne P. Downey, Mark D. Ewen, Timothy B. Petersen (Project Manager) and Brian G. Morrison (Project Director) Industrial Economics, Incorporated 2067 Massachusetts Avenue Cambridge, MA 02140 Project Steering Committee: Elaine DuWon National Wildlife Research Centre Environmental Conservation Service Environment Canada Jim Frehs Water & Habitat Conservation Branch Environmental Conservation Service Environment Canada R.A. (Sandy) Matbeson Commercial Chemicals Branch Environmental Protection Service Environment Canada Danny Guay Regulatory and Economic Assessment Branch Environmental Protection Service Environment Canada Jacqueline Sitwell Bureau of Chemical Hazards Environmental Health Directorate Health Canada Scientific Authority: Paul De Civita Regulatory and Economic Assessment Branch Environmental Protection Service Environment Canada (819) 997-7061 September 1996 SL 110199 TABLE OF CONTENTS EXECUTIVE SUMMARY............................................................................................CHAPTER 1 Introduction........................................................................................................................... *1"1 Selection of Substances Analyzed .........................................................................................1-1 Overview of Methods..............................................................................;........................ -1-2 Identification and Screening of Benefits.......................................... .................1*3 Developing Annual Benefits Estimates.................................................................... 1-7 Developing Total Benefits Estimates..................................................................... 1*11 Results.................................................................................................................................. J-ll Vinyl ChJozide (VC)................................................................................................1-11 Dichloromethane (DCM) ....................................................................................... 1-12 Polychlorinated Biphenyls (PCBs)..........................................................................1-12 Limitations............................................................................................................... 1*16 Organization of Report...................... 1-16 References for Chapter 1 ........................................................... 1-18 Appendix 1A: Supplemental Information on Health Risk Assessment ..................................... 1-19 VINYL CHLORIDE....................................................................................................... CHAPTER 2 Overview of Chapter........................ .2-1 Summary of Benefits................................................................................................ .2-1 Limitations ............................................. 2-1 Organization of Chapter..................... '................................................................... 2-2 Background Information on Vinyl Chloride.........................................................................2-3 Description of the Compound..................................................................................2-3 Information on Manufacture in Canada.................................................................. 2-3 Releases to the Environment.................................................................................... 2-4 Human and Ecological Exposures to Vinyl Chloride ......................................................... 2-6 Human Exposures.....................................................................................................2-6 Ecological Exposures................................................................................................. 2-7 SL 110200 TABLE OF CONTENTS (continued) The Benefits of Controlling Vinyl Chloride Releases: Overview................................... 2-8 Mortality from Cancer.......................................................................... '..................2-8 Morbidity ..............................i..................................................................................2-8 Estimating Human Health Effects.......................................................... ...2-9 Cancer Cases Resulting from Exposure to Vinyl Chloride ............... ................. 2-9 Noncardnogenic Effects Resulting from Exposure to Vinyl Chloride...........................................................................................................2-15 Valuing Human Health Effects: Annual Values.............................................................. 2-18 Carcinogenic Effects: Mortality.............................................................................. 2-19 Noncardnogenic Effects: Morbidity ..................................................................... 2-21 Valuing Human Health Effects: Total Values...................................................................2-22 References for Chapter 2............................................................................................................... .2-25 Appendix 2A: General Population Exposed to VC Emissions from Point Sources ..............................................................................................2-29 Appendix 2B: Estimating General Population Cancer Cases: Detailed Example...................................................................................................2-32 Appendix 2C: Estimates Used to Generate Proposed Distribution of Value of Life .......................................................................................... 2-33 DICHLOROMETHANE................................................................................................. CHAPTER 3 Overview of Chapter.............................................................................................................. 3-1 Summary of Benefits..................................................................................................3*1 limitations of Analysis ............................:.............................................................. 3-1 Organization of Chapter .............................................................. 3*2 Background Information on Dichloromethane .................................................................. 3-2 Description of the Compound.................................................................................. 3-2 Information on Use in Canada ................................................................................3-3 Releases to the Environment.................................................................................... 3-3 SL 110201 TABLE OF CONTENTS (continued) Human and Ecological Exposures to Dichloromethane ................................................... 3-4 Human Exposures.....................................................................................................*3-5 The Benefits of Controlling DCM Releases; Overview.......................... ........................ 3-6 Carcinogenic Effects from Chronic Exposures . 3-7 Noncaidnogenic Effects from Chronic Exposures .......................... ................ 3-7 Acute Health Effects............................................ .. * ^............................................ 3-7 Ecological Effects.....................................................................................................*3-8 Estimating Human Health Effects.........................................................................................3-8 Annual Cancer Cases Resulting from Exposure to Dichloromethane .................................................................................................................. -3-8 Noncarcinogenic Effects Resulting from Exposure to Dichloromethane...................................................................................................... 3-10 Acute Health Effects................................................................................................3-11 Valuing Human Health Effects; Annual Values............................................................. 3-12 Carcinogenic Effects: Mortality..............................................................................3-12 Other Health Effects............................................................................................... 3-13 Valuing Human Health Effects: TotalValues................................................................... 3-14 References for Chapter 3................................................................................................................. 3-16 Appendix 3A: Estimating the Average Daily Exposure of Individuals Exposed to DCM in the Workplace .............................................................................................. .3-17 POLYCHLORINATED BIPHENYLS (PCBs)............................................................ CHAPTER 4 Overview of Chapter..............................................!.............................................................4-1 Summary of Benefits................................................................................................. 4-1 Limitations .................................................................................................................4-1 Chapter Organization ...............................................................................................4-2 SL 110202 TABLE OF CONTENTS (continued) Background Information on PCBs .......................................................................................4-3 Description of PCBs.................................................................................................. 4-3 Information on Use in Canada ................................................................................4-3 Releases to the Environment.................................................................................... 4-4 Ecological and Human Health Exposures .... ........................................................ ....4-5 Ecological Exposure to PCBs.................................... 4-5 Human Exposure to PCBs......................................................................................... 4-9 The Benefits of Controlling PCB Exposures: Overview................................................... 4-9 Benefit Categories................ .4-9 Methods of Estimating PCB-Related Benefits..................................................... 4-11 Annua] Benefit Estimates.....................................................................................................4-13 Extractive Use Value Benefits Non-Extractive Use.............. . Passive Use............................ Human Health ..................... . Total Benefits Estimates References for Chapter 4 Appendix 4A: Distribution of PCBs in Use and Storage in Canada .4-14 .4-27 .4-28 .4-32 4-32 .4-35 4-39 SL H0203 acknowledgements The authors gratefully acknowledge the guidance and assistance offered by Paul De Gvua and the members of the Project Steering Committee in preparing this report. Additional assistance in its development was provided by Stephen MacDonald, Andy Atkinson, Lawrence Fedoruk, Birgit Braune, Francois Lavalee, Arthur Sheffield, Jeff Harris, Art Stehig, Andre Jacquemot, and Chuck Cox. Copies of the draft report were also furnished to John Hflbom and Uwe Shneider. SL 110204 EXECUTIVE SUMMARY CHAPTER 1 INTRODUCTION The New Substances Notification Regulations ("the NSNR" or "the Regulations") were developed to ensure "that no new substances will be manufactured in or imported into Canada on a commercial scale before an assessment has been carried out to determine the risks posed to the environment and human health by the substances (Canada Gazette 1994).* To characterize the potential benefits of the Regulations, this report analyzes the benefits that would have resulted had the Regulations been implemented in time to control or prohibit the use of three-substances currently or previously manufactured in or imported into Canada: vinyl chloride (VC), dichloromethane (DCM), and polychlorinated biphenyls (PCBs). This approach is analogous to evaluating the human health and environmental damages incurred due to the release of these substances into the Canadian environment, assuming that these damages would not have resulted had the Regulations been in place at the time the substances were introduced. This type of approach is often referred to as an "avoided damages" approach. This executive summary describes the process Environment Canada (EC) and Health Canada (HC) employed to select the substances analyzed. It then outlines the benefits typology used in this report, summarizes the methods used to estimate annual and total benefits, and presents a summary of the results for each substance. It also cites the major limitations of the analysis and provides an overview of the structure of the remaining chapters. SELECTION OF SUBSTANCES ANALYZED In the course of developing the NSNR, EC and HC committed to a mult-stakeholder review of regulatory impacts, including an analysis of the Regulations' environmental and health benefits. This analysis is to be submitted to the New Substances Notification Impact Working Group (NSNTWG), which is comprised of representatives from EC HC Industry Canada, and the Industry Coordinating Group for the Canadian Environmental Protection Act (CEPA). The analysis is to be completed by September 1997, approximately three years after the rules' promulgation. Ideally, the impact analysis would focus on benefits attributable to regulation of new substances under the NSNR. The data required to conduct such an analysis, however, are currently unavailable, and unlikely to become available by 1997. Therefore, EC and HC chose to characterize the Regulations' potential benefits through case studies of three CEPA toxic substances that in all M SL 110205 likelihood would have been prohibited r controlled had they been evaluated through the new substance notification program. EC and HC selected the three substances analyzed in this report using a screening approach to narrow the list of CEPA toodc substances (substances subjected to regulatory control under GEPA and identified as toxic under the Priority Substances Program). First, to ensure a focus on commercially manufactured or imported products, substances that generally would be classified into ne of the following categories were eliminated from consideration: Effluents and wastes. Effluents and wastes are not notifiable under the. Regulations, and therefore are not affected fay them. By-products ofmanufacturing and processing operations. Manufacturing and processing by-products are also not directly notifiable under the regulations. Elements and natutalhK>""*"r TMTppounds. Elements, and naturallyoccurring compounds were eliminated from consideration due to the inherent difficulty in distinguishing anthropogenic impacts from natural impacts. The remaining substances were reviewed to determine: the completeness of data on human health and ecological effects; their suitability as surrogates for the NSNR analysis (i.e., solely or primarily anthropogenic in origin); and whether they had been determined to be CEPA toxic by both EC and HC The result of this effort was the selection of VC, DCM, and PCBs as the most appropriate candidates for analysis. These substances are anthropogenic in origin and have been determined to be CEPA toxic by the two Departments. In addition, the data available on the impacts of these substances is relatively substantial, facilitating the development of a benefits assessment. OVERVIEW OF METHODS We evaluate the benefits of the NSNR using an approach resource economists commonly refer to as the damage function approach.1 As generally applied, this approach includes the following steps: , 1 The term damages refers to negative environmental or human health effects. The benefits of a particular policy (e.g^ a pollution control regulation) are the reduction in damages thr.t rccu!: from the policy's implementation. 1-2 SL 110206 (1) fsritiMTing the to which a regulation would change the quantity of a pollutant r toxic substance released to the environment; (2) Employing dispersion modeling techniques to characterize changes in the concentration of pollutants in environmental media; (3) Applying concentration- or dose-response functions to estimate changes in physical ""p*** (e-g-, human health impacts or impacts on species abundance) attributable to pollution; (4) Valuing the change in physical, impacts, thereby creating an economic benefits estimate; and ; (5) Aggegating benefits estimates across all relevant categories and time periods. The broad scope of this analysis and the limited time and resources available for the effort have necessitated some simplification of this approach. For example, we assume that control or prohibition of a substance under the NSNR would eliminate all associated health or environmental impacts; as a result, the benefits of the Regulations are simply the absolute value of the damages attributed to unregulated manufacture and/or use of each substance. In addition, we have not undertaken any original pollutant dispersion modeling, choosing to rely instead on previously developed information to characterize concentrations of VC, DCM, and PCBs in the environment Nonetheless, the approach employed here is consistent with the underlying theory of scientific injury assessment and environmental impact valuation, and provides a reasonable characterization of the benefits of three hypothetical applications of the NSNR. The following discussion notes key methodological issues in our implementation of the damage function approach to analyze the benefits of controlling VC DCM, and PCBs. As described below, implementation of the methodology includes the following major elements: A preliminary evaluation of each substance to identify major benefit categories and to determine if sufficient data are available to quantify key benefits; Use of this information to develop a detailed evaluation of the effect of each substance on human health and/or the environment; and Valuation of human health and ecological effects, including the incorporation of simulations to address uncertainty in the underlying estimates. Identification and Screening nf A number of different classification approaches have been developed to structure assessments of the benefits of environmental protection. This report employs a benefits classification framework that Environment Canada itself recommends (EC 1996). The Environment 1-3 SL 110207 Canada framework, which is similar to many of the classification schemes in the published literature, is summarized in Exhibit 1*1. Exhibit 1-1 EC* FRAMEWORK FOR CLASSIFYING BENEFITS Category Human Health Benefits resulting from decreased mortality and/or morbidity. Extractive Uses Benefits resulting from the increased use of natural resources that are used directly for commercial, recreational, or subsistence needs (e.g^ increased recreational fishing opportunities). Non-Extractive Uses Benefits resulting from the increased use of environmental services that are not consumed (e^ activities such as hiking and canoeing). Ecological Functions Benefits resulting from a wide range of ecological services that are dependent on a healthy ecosystem (c^, benefits that humans receive from the ability of soil to filter impurities from groundwater). Passive Uses Benefits resulting from the knowledge of the continued existence of natural resources and the desire to leave natural resources to future generations. Built Environment Benefits resulting from reduction m materials damage and soiling. Optional Uses Benefits resulting from potential future uses (in the above categories), including insurance against unknown future events. As the preceding exhibit shows, the benefits of environmental regulations can take a variety of forms. Exhibit 1-2 illustrates the process we have employed to identify the potential benefits of controlling VC, DCM, and PCBs, and to determine which benefits categories to subject to quantitative analysis. As the exhibit indicates, we first conducted an extensive literature review to identify potential benefits. We then screened the available data for each benefits categoiy to determine whether it was sufficient to conduct a quantitative benefits assessment; if so, we judged whether the benefits were likely to be great enough to warrant devoting resources to a quantitative analysis. Benefits categories that passed this screen were evaluated quantitatively, all others were qualitatively assessed. 1-4 SL 110208 Exhibit 1-2 BENEFITS SCREENING PROCESS SL 110209 Exhibits 1-3 through 1-5 give the results f the screening assessment, listing the potential benefits associated with regulation of the three substances of concern and indicating which categories of benefits have been quantitatively or qualitatively evaluated. As the exhibit indicates, the focus of the VC and DCM analyses is on human health benefits, while the PCB analysts addresses a range of environmental benefits, including extractive, non-extractive, and passive uses.2 Rihihii 1-3 POTENTIAL BENEFITS OF NSNR CONTROLS; VINYL CHLORIDE . Benefit Category lnptrti irf flnnewii Form of Analysis Reduced Mortality Studies link VC exposure to angiosarcoma of the liver and Quantitative . cancer of the brain, lung, and digestive tract. Reduced Morbidity VC can cause hepatic, respiratory, cardiovascular, gastrointestinal, hematological, musculoskeletal, renal, dermal, ocular, and other adverse effects. Quantitative Ecological Benefits Byproducts of VC production can adhere to plankton and Qualitative bioaccurouiate in the marine environment. Exhibit 1-4 POTENTIAL BENEFITS OF NSNR CONTROLS: DICHLOROMETHANE Benefit Category Impacts of Concern Form of Analysis Reduced Mortality Chronic exposure to DCM has been linked to liver and lung cancer. Quantitative Reduced Morbidity Chronic exposure to DCM has been linked to Central Nervous System effects such as headaches, dizziness, nausea, and memory loss; effects on liver and kidney function; and benign mammary gland tumors. Quantitative Ecological Benefits Exposure to DCM may adversely affect some aquatic organisms, such as freshwater nematodes. Qualitative , At the instruction of HC, the analysis of PCBs excluded consideration of human health risks. HC considers available data on such risks inconclusive, and therefore insufficient for an analysis of this type. 1-6 SL 110210 * Exhibit 1-5 POTENTIAL BENEFITS OF NSNR CONTROLS: FCBs Benefit Category | Impacts of Concern Extractive Use Recreational Fishing Diminished abundance and diversity of fish species due to reproductive abnormalities lowers quality of the angling experience. Fish consumption advisories (based on human health concerns) reduce the value of angling experience and affect the choice of fishing location. Waterfowl Hunting High tissue concentration levels and potential waterfowl consumption advisories dimmish enjoyment of the hunting experience and affect the choice of hunting locations. Trapping Decline m mink populations due to reproductive abnormalities limits pelt harvest and diminishes trapping experience. Commercial Fishing Diminished abundance and diversity of fish species due to reproductive abnormalities decreases catch. Fish consumption advisories reduce volume of saleable fish. Non-Extractive Use Bird Viewing Other Recreation Passive Use Reductions in bud populations due to reproductive abnormalities and the presence of physical deformities diminishes the enjoyment of birdwatching. Concern over health risks diminishes the enjoyment of swimming, boating, or similar activities and affects the choice of location. Contaminated environment decreases the value individuals derive from a dean natural resource, above and beyond its direct use. Human Health Increased risk of both mortality (cancer) and morbidity (e.g, adverse reproductive effects). Form of Analysis Quantitative Quantitative Quantitative Qualitative Quantitative Qualitative Quantitative Qualitative Developing Annual Benefits Estimates We develop quantitative estimates of annual benefits for each substance by estimating annual human health and ecological effects resulting from exposures to the substance, and by valuing those effects using a "benefits transfer* approach. Further, we explicitly incorporate uncertainty into our analysis by using a statistical technique known as Monte Carlo simulation. A summary of the methods employed to estimate annual benefits is provided below. 1-7 SL 110211 Pjttmmttwg Human Heahb nd Ecological Effects While the particulars of each analysis vary to some extent, we estimate human health and ecological effects by collecting information on: the sources and releases of each substance; the pathways from these sources to human health and ecological receptors; the concentrations of each substance to which human and ecologicalreceptors are exposed; the estimated dose that humans or biota receive; and dose-response relationships. This information is used to quantify likely human health effects and impairment of ecological or natural resource services. For instance, we use health risk assessment methodologies to develop estimates of the number of cancer cases that would be avoided by controlling VC and DCM.3 For PCBs, we assess the effect of the contaminant on the services provided by natural resources, such as the number of recreational fishing trips affected by PCS contamination in fish. Estimating these effects requires careful consideration of the tuning ofharm to human health and the environment, as well as consideration of the effects of other regulatory measures that may mitigate the risks a substance might otherwise pose. For instance, damages that VC emissions cause in Canada today are likely to be quite small, since (1) VC emissions are currently regulated under CEPA and (2) pre-CEPA emissions do not cause lingering problems due to the transient nature of VC in the environment. Thus, to evaluate the benefits of controlling VC as if it were a new substance, we employ 1970s estimates of average atmospheric concentrations of the pollutant in the vicinity of facilities that used or manufactured it; these estimates, developed prior to the regulation of the substance, allow us to estimate the adverse health impacts that would occur in the absense of controls. For PCBs and DCM we use more current information, as each substance continues to cause harm in Canada.4 3 For those readers unfamiliar with health risk assessment methods and terms, we have included a brief primer on these topics in Appendix 1A, which is attached to this chapter. 4 While PCBs have been banned from new uses in Canada since 1980, previous discharges of PQBs continue to cause present day problems due to the persistence of the substance in the environment. DCM is not currently regulated under CEPA; so current rKvis provide a reasonable basis for assessing the benefits associated with regulating this substance. 1-8 SL 110212 Valuing Human Health and Ecological Effects We value human health and ecological changes using die benefits transfer approach. Benefits transfer involves the application ofvalue estimates, functions, data and/or models developed in one contevt to address a rimihr resource valuation question in another context. Benefits transfer in its simplest form involves the application of existing "per unit" values derived at one site (or in one study) to estimate the benefits attributable to an environmental improvement at another site. Such transfers maty include adjustments to account for differences in the characteristics of the activity evaluated in the existing study versus the characteristics of the activity for the "site" (in this case "country") under analysis. Although this technique has been used for many years to evaluate environmental benefits and to assess natural resource damages, it continues to generate some controversy in the environmental and natural resource economics community. This controversy focuses on the applicability of value estimates developed for a particular site in one context to the same or similar site in another context. Determining whether an existing study is appropriate for benefits transfer requires consideration of the overall quality of that study and the degree to which the case examined by the study is similar to the case of interest (typically referred to as the "policy" case). Factors of concern include: (1) fYimmrviity Characteristics The commodity (environmental goods and services or human health effects) valued in the policy case and study case should be the same or similar. The magnitude of change in the commodity between the policy and study case should be similar. The availability of substitute sources of environmental goods and services should be similar. The assignment of property rights should be the same (e^, willingness to pay versus willingness to accept). The temporal perspective (ix, when a commodity is provided and when payment is made) should be the same. (2) Population Characteristics * The characteristics of the populations of interest in the study and policy cases (e.g., age, income, proximity to the site or source of risk and degree of environmental concern) should be the same or similar. Any valuation function employed should include socioeconomic variables, to allow the transfer to account for differences in population characteristics between the study and policy cases. 1-9 SL 110213  The population's familiarity with the environmental problem of concern should be similar in the study case and policy case (Boyle 1992; Desvouges 1992; Unsworth 1993). Detailed information on the Canadian and U.S. studies relied upon in this report is provided in Chapters 2,3 and 4. To make the values collected from each study comparable, we use the following approach to adjust for inflation and for the difference in value of the U.S. and Canadian dollars: We adjust- dollar values presented in past Canadian studies by applying the Canadian Consumer Price index (CPI) to inflate, figures to 1995 Canadian dollars. We adjust dollar values presented in past U.S. studies by first translating the US. dollar figures to Canadian dollars using the Purchasing Power Parity (PPP) index, and by then applying the Canadian CPI to inflate figures to 1995 Canadian dollars. This approach was recommended to us by the Scientific Authority in consultation with Statistics Canada (De Gvita 1996). Incorporating Uncertainty In the course of conducting this effort, we were faced with data uncertainties at several stages. For instance, our ability to estimate the cancer cases resulting from human exposures to VC and DCM is limited by a number of factors, including uncertainty regarding the size of the exposed population and uncertainty as to the actual responses of human receptors. Similarly, there is uncertainty associated with the economic unit values identified for particular benefits categories. To generate annual and total benefits estimates deterministically would imply a higher level of certainty than actually exists. As a result, we explicitly recognize uncertainty associated with key independent variables through the use of a statistical procedure commonly known as Monte Carlo simulation. This technique provides information on the variable of ultimate interest (the dollar value of benefits) that is the function of input variables (e.g^ unit economic values) that art probability distributions. The output of this process is a probability distribution reflecting the magnitude and likelihood of various benefit figures.* For instance, in Chapter 2 we provide benefits estimates associated with avoiding VC-induced cancer cases by reporting not only the mean value, but also the 10th and 90th percentile estimates. Although the selection of these figures as our low- and high-end estimates is arbitrary, * We use the computer software program @Risk to conduct our uncertainty analysis. The program offers two sampling techniques: Monte Carlo sampling and Latin Hypercube sampling. We employed Monte Carlo sampling for this analysis. 1-10 SL 110214 their use should provide a mote plausible characterization of the likely range of benefits than would use of the n's extreme upper and lower bounds. Developing Total Benefits Esthnates Sine* the benefits of the NSNR will be realized over more than one year, we also estimate the total benefits of regulating each substance. Hie total benefits estimates are the present value of the stream of annual benefits discounted over a thirty year tune period. We selected a thirty year time period as a common and reasonable long* time horizon (e^, 30 years is commonly used in estimating total operation and maintenance costs at hazardous,waste sites). We-acknowledge that the selection of 30 years in calculating total benefits for this project is somewhat arbitrary; -in reality, the benefits from the NSNR could be longer or shorter than 30 years.* The selection of a discount rate for this type of calculation has been the subject of much debate in the economics profession. Treasury Board, Canada recommends the use of five, 10 and 15 percent real discount rates in conducting analyses of this sort. We have selected lower figures (two, four and six percent), however, as the majority of environmental economists view five, 10 and 15 percent as too high for this type of analysis. The range of discount rates we employ is consistent with the findings of A. Myrick Freeman, a noted resource economist who recently reviewed the literature on this topic and recommends using a discount rate of 2 to 3 percent (Freeman 1993). RESULTS The following discussion summarizes the benefits of controlling or prohibiting the manufacture and/or use of vinyl chloride, dichloromethane and polychlorinated biphenyls in Canada Detailed information supporting these benefits estimates is provided in the body of the report. Vinyl Chloride fVO Vinyl chloride is a chemical compound that has been used in Canada's industrial and commercial- sectors for over 50 years. The primary use of VC is in the production of polyvinyl chloride (PVQ, which is used to manufacture plastic and vinyl products such as electrical wire, food packaging materials, pipe, and other construction products. VC can cause significant adverse6 6 Determining the period of time over which the benefits of the NSNR will accrue depends upon how the NSNR are viewed in relation to other public health and environmental regulations. For - example, in the absence of the NSNR regulations, releases of hazardous substances into the environment will be regulated, but in all likelihood only after human health and environmental problems are discovered. Therefore, the benefits of the NSNR accrue between the introduction of the harmful (new) substance and when it would be regulated under the Canadian Environmental Protection Act. The extent of the damages caused in the interim is largely dependent on the nature of the substance. For instance, unregulated PCB discharges of only a few years can cause 50 or more years of damages, as the substance is persistent in the environment. In contrast, VC is transient in the environment, limiting the time period over which damages are likely to occur. SL 110215 human health effects if production and use of the chemical are not regulated. In the early 1970s, unregulated indoor air concentrations at VC and polyvinyl chloride (PVC) production facilities were extremely high, causing a variety of health problems in plant workers. Further, releases to the ambient air posed significant health risks to those living near VC and PVC production facilities. As a result, the economic benefits of regulating the manufacture and use of VC are significant. As shown in Exhibit 1-6, the annual benefits associated with avoiding worker and general population cancer cayey and occupational morbidity are estimated to range from $15.2 million to $1062 milli n, with a mean estimate of $53.0 million; total benefits over 30 years, using a four percent discount rate and the mean annual benefits figure, are estimated to be $916.4 million. Additional avoided morbidity benefits may also exist, although we have been unable to-quantify these benefits. Dichloromethane (PCM) Dichloromethane is a clear, colorless liquid that is primarily used as a solvent in cleaning or stripping operations (e.g., as a metal parts degreaser), as a solvent and propellant in aerosol products, and as a blowing agent in the production of urethane foams. The benefits of controlling or prohibiting DCM use in Canada are significant. As summarized in Exhibit 1-7, the annual benefits associated with avoiding worker and general population mortality are estimated to range from $12$ million to $582 million, with a mean estimate of $32.6 million; total benefits over 30 years, using a four percent discount rate and the mean annual benefits figure, are estimated to be $563.7 million. Additional morbidity or ecological benefits may also exist, though we were not able to quantify such benefits in this report. Polychlorinated Brphenvh fPCBsl Polychlorinated biphenyls (PCBs) are a family of synthetic chemical compounds used extensively as a dielectric fluid in electrical equipment from the 1940s to the 1970s. PCBs can cause significant harm to the environment and human health if production, use, and disposal of the substance are not properly regulated. The results of our analysis indicate that there are significant economic benefits related to the regulation of substances like PCBs through the NSNR. We summarize the range of annual .benefits and total benefits in Exhibit 1-8. For the extractive use benefit categories that we were able to monetize, we estimate annual benefits of $83 million to $57.6 million, with a mean estimate of $303 million; the present value of these benefits over 30 years, using a four percent discount rate and the mean annual benefits figure, is $524$ million. Our estimate of annual non-extractive use benefits ranges from $5.7 million to $37.4 million, with a mean of $19.1 million; the present value of these benefits, using the four percent discount rate and mean annual benefits figure, equals $3303 million. In addition, our estimate of annual passive use benefits ranges from $21.6 million to $86.4 million, with a mean of $54.0 million; using the four percent discount rate and mean annua) benefits figure for this category yields a present value benefits estimate of $933.8 million. As illustrated by these totals, the direct (extractive and non extractive) use and passive use categories each account for roughly 50 percent of the total estimated benefits of controlling PCBs. 1-12 SL 110216 Exhibit 1-6 SUMMARY OF ANNUAL AND TOTAL BENEFITS: VINYL CHLORIDE (SI995, millions) Annual Benefits Present Value of Total Benefits -- 30 yean Category 1jam Mean High Low @ 6% Mean @ 4% High 2% Mortality General Population | Occupational $ 1(1.3 $4.2 $39.1 $12.1 $80.0 $22.9 $141.8 $57.8 $676.1 $209.2 $1,791.7 $512.9 Morbidity 1 General Population Unknown Unknown Unknown Unknown Unknown Unknown | Occupational $0.7 $1.8 $3.3 $9.6 $31.1 $73.9 ( Ecological \ Total Small > $15.2 Small > $53.0 Small > $106.2 Small > $209.2 Small > $916.4 Small >$23785 M3 Exhibit 1-7 SUMMARY OF ANNUAL AND TOTAL BENEFITS: DIC11LOROMETHANE ($1995, millions) Annual Benefits Present Value of Total Benefits - 30 years 1 Category 1 Mortality Ijow Mean High Low 6% Mean 4% High 2% 1 General Population | Occupational $6.5 $6.3 $16.7 $15.9 $30.0 $28.2 $89.5 $86.7 $288.8 $274.9 $671.9 $631.6 Morbidity General Population Occupational Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown .Unknown Unknown Ecological Small Small Small Small Small Small Total > $118 > $32.6 >$581 > $1761 >$563.7 > $13031 SL 110218 1-14 %% Q Category U Extractive Use N Recreational Fishing 1 Waterfowl Hunting | Trapping Commercial Fishing Subtotal Non-Extractive Use Bird Viewing | Other Recreation | Passive Use | Human Health | Total Exhibit 1-8 .J SUMMARY OF ANNUAL AND TOTAL BENEFITS: PCOs ($1995, millions) Low. , Annual Benefits Mean High Present Value of Total Benefits - 30 years Low @6% Mean @ 4% High @2% | 1 1 $7.3 $(1.6 $0.4 Unknown $8.3 $2ft.9 $1.3 $2.1 Unknown $30.3 $47.3 $2.4 $7.9 Unknown $57.6 $!IKI.5 $8.2 $5.5 Unknown $114.2 . : $465.2 $22.5 $36.3 Unknown $524.0 $1,059.4 $53.8 $176.9 Unknown $1,290.1 | I I $5.7 Unknown $21.6 No! Estimated > $35.6 $19.1 Unknown $54.0 Not Estimated > $103.4 $37.4 Unknown $86.4 Not Estimated > $181.4 $78.5 Unknown $297.3 Not Estimated > $490.0 $330.3 Unknown $933.8 Nol Estimated > $1,788.0 $837.6 Unknown $1,935.1 Not Estimated > $4,062.8 SL 110219 1-15 Limitations The analyses described above are subject to certain limitations. A summary of these limitations for each substance follows. Vinyl Chloride (VQ. To estimate the health effects associated with VC we make a series of assumptions that introduce uncertainty into the analysis. For instance, our analysis of the incidence of VC-induced cancer in the general population is based on standard assumptions regarding the location and mobility of the affected population, which-may over- or under-state total cancer cases. Further, our assessment of occupational health effects relies heavily on incidence data collected in the UJS^ we do not know the extent to which plant conditions may differ between VS. and Canadian facilities. Finally, we have been unable to monetize certain health effects in this assessment (e^ general population health effects other than cancer). Dichloromethane (DCM). Data limitations affect our ability to estimate occupational exposures for DCM, as it is difficult to estimate how many Canadian workers are chronically exposed to significant levels of dichloromethane. While we extrapolate information from the US. to estimate occupational effects, it is difficult to assess how accurately this method approximates the exposed occupational population in Canada. Further, all the quantified benefits for DCM are based on chronic exposures. Since DCM is known to cause serious health problems when individuals receive acute doses, our health benefits are likely to be somewhat underestimated. Polychlorinated Biphenyls (PCBs). The most significant limitation of our analysis of PCBs is the lack of quantitative estimates for all of the possible benefits resulting from a Canadian environment free of PCBs. For instance, we were unable to estimate or monetize the benefits for commercial fishing and certain recreational activities (c.g^ boating and swimming), and at HCs instruction did not attempt to quantify the human health benefits of PCB control In addition, numerous factois simultaneously affect the quality of. the aquatic environment, including toxic chemicals other than PCBs (e.g^ mercury, dioxins), pesticides, and municipal wastewater discharges. As a result, it is difficult to isolate the environmental impacts attributable to PCBs. Further, the long-lived nature of PCBs, along with nearly 20 years of PCB regulation, make it difficult to establish a base year from which to estimate annual benefits. ORGANIZATION OF REPORT The remainder of this report provides detailed information on the benefits of controlling or prohibiting the use and/or manufacture of vinyl chloride (Chapter 2), dichloromethane (Chapter 3) and polychlorinated biphenyls (Chapter A). Each chapter contains: 1-16 SL 110220  Background informed n on the substance, including a desaipti n of the compound, information on its use and/or manufacture in Canada, and information on the releases of the substance into the environment; A description of human and ecological exposures to the substance; An overview ofthe benefits ofcontrolling releases of the substance, using the benefits typology presented above; and A detailed description of the annual and total benefits that result from controlling these substances. ;. Farh chapter also presents the specific limitations associated with these analyses, including key data gaps. 1-17 SL 110221 REFERENCES FOR CHAPTER 1 Boyle 1992. Boyle, Kevin J. and John C Bergstrom. "Benefit Transfer Studies: Myths, Pragmatism and Idealism." Water Resources Research, Vol 28, No. 3, March 1992, pp. 657-663. Canada Gazette 1994. "New Substances Notification Regulations," Canada Gazette, Part II, Vol. 128, No. 7,4 June 1994, p. 1481. De Gvita 1996. Letter from Paul De Gvita, Environment Canada, to Brian Monison, Industrial Economics, Incorporated, 4 March 1996. Desvouges 1992. Desvouges, William, Michael Naughton, and George Parsons. "Benefit Transfer Conceptual Problems in Estimating Water Quality Benefits Using Existing Studies." Water Resources Research, Vol 28, No. 3. March 1992, pp. 675-683. Freeman 1993. Freeman, A. Myrick. The Measurement of Environmental and Resource Values: Theory and Methods. Resources for the Future. 1993. Unswonh 1993. Unsworth, Robert E-, Elizabeth W. Snell and Mark Dickie, The Appicability of a Benefits TransferApproach to Assess the Economic Benefits of Reduced Air Toxics Emissions Under the Clean Air Act. Industrial Economics, Incorporated, 1993. 1-18 SL 110222 Appendix 1A SUPPLEMENTAL INFORMATION ON HEALTH RISK ASSESSMENT 1-19 SL 110223 SUPPLEMENTAL INFORMATION ON HEALTH RISK ASSESSMENT This appendix provides background information on health risk assessment terms and meth ds that are used in Chapters 2 and 3. This appendix is divided into two sections. The first addresses carcinogenic health effects; the second addresses noncarcinogenic health effects. Since the health risk assessment relies on data and techniques developed in large part by the U.S. Environmental Protection Agency (EPA), the appendix focuses on U.S. EPA terms and methods. CARCINOGENIC EFFECTS7 Toxicological assessment of potentially carcinogenic substances often indudes a rating f the "weight of evidence" concerning carcinogenicity. For example, the U.S. EPA assigns evidence from human and animal studies to one of five categories; (1) sufficient evidence, (2) limited evidence, (3) inadequate evidence, (*) no data concerning carcinogenicity, and (5) no evidence of carcinogenicity. U.S. EPA's classification scheme is presented in Exhibit 1A-1. Exhibit 1A-1 U.S. EPA CLASSIFICATION SCHEME FOR CARCINOGENS Group Description A Human carcinogen; sufficient human evidence. B1 Probable human, carcinogen; limited human evidence. B2 Probable human carcinogen; sufficient animal evidence, inadequate or no human evidence, or no human data. C Possible human carcinogen; limited animal evidence, inadequate or no human evidence, or no human data. D Not classifiable as to human carcinogenicity; inadequate human and animal evidence or no data available. E Evidence of noncarcinogenicity for humans. Source: U.S. EPA Guidelines for Carcinogen Risk Assessment, 51 FR 33992-34054,1986 In addition to the qualitative rating assigned to potentially carcinogenic substances, U.S. EPA develops substance-specific cancer slope factors (CSFs) for potential carcinogens. A CSF gives the increase in lifetime cancer risk for an individual continuously exposed to a substance, from birth to , 7 For more detailed information on EPA's method for estimating cancer risks, see U.S. EPA, Risk Assessment Guidance for Superfund, Volume I: Human Health Evaluation Manual, December 1989. 1-20 SL 110224 death, per unit increase in the substance's concentration.1 CSFs provide estimates f risks associate^ with low-dose exposure to suspected r known caidn gens. They are estimated from data on high dose exposures and/or high dose laboratory exposure to animals. The CSF for an agent with an "A" rating win be based - at least in part - on human data, while one with a "B2" rating win be based entirely on animal data. Generating a Qmeer Stone Factor In deriving a CSF, U.S. EPA evaluates available toxicological information about a chemical and selects an appropriate data set. In choosing appropriate data sea, human data of high quality are preferred to animal data. If animal data are used, the species',that responds most similarly to humans is preferable. If there is no dear choice among species for which data exist, data for the most sensitive speries is given the greatest emphasis. Occasionally, when no single study is judged to be most appropriate, U.S. EPA may derive a CSF based on the geometric mean of slope factors from several studies that collectively support the estimate. Concept of Nonthreshold Effects Carcinogens are assumed to have no threshold, Len there is no level of exposure that does not pose some probability, however small, of generating a carcinogenic response. Therefore, the development of a CSF usually involves extrapolating from the high doses administered to experimental animals to the lower exposure levels expected for human exposure in the environment. Because the extrapolation of the dose-response is to a range where data are not available, the actual CSFs are uncertain. Hence a large margin of safety is built into the slope factors. Generally, US. EPA takes a conservative approach in estimating a CSF by calculating the slope factor as the highest value that can be justified with 95 percent confidence. Thus, given the data, it is unlikely that the CSF is greater than that reported. Assessing Carcinogenic RkIk A substance-specific CSF is multiplied by the substance-specific daily dose to estimate the incremental probability of a person developing cancer over a lifetime (70 years) as a result of exposure to the potential carcinogen. A cancer risk of 1 x 10* implies a probability of one in a million of developing cancer. * EPA Integrated Risk Information System, 1995. 1-21 SL 110225 NONCARC3NOGENIC EFFECTS' Unlike cudnogenic effects, noncarcinogenic effects are assumed to occur only at exposures above a threshold concentration below which protective mechanisms and function reserve capacities prevent reactions to exposure. Therefore, a range of exposures exists from zero to some finite value that can be tolerated by the individual without the manifestation of an adverse effect The reference dose (R2D) is an estimate of a daily exposure level for humans, including sensitive subgroups, that is not likely to cause adverse effects during a lifetime of exposure. To develop an RfD, it is necessary to identify the upper bound of a tolerance range for the most sensitive populations. Derivation of a Reference Dose In developing an RfD for a substance, U.S. EPA all available toxicity studies. If adequate human data are available, this information is used as the basis of the RfD. Otherwise, animal study data are used. When choosing the appropriate human or animal study to use in deriving the RfD, U.S. EPA must make judgements regarding the relevance and quality of the available experimental studies. In addition, U.S. EPA must identify the effect characterized by the lowest observed adverse effect" (LOAEL), which is referred to as the critical toxic effect, and serves as the basis for the RID. After the study and critical toxic effect have been selected, U.S. EPA identifies the exposure level that represents the highest exposure level tested in the scurfy at which the critical effect was not observed. This highest "no observed adverse effect lever (NOAEL) is used to develop the RfD. Applying Uncertainty Factors The RfD is derived from the NOAEL by consistent application of the following uncertainty factors (UF): * A factor of 10 is used to account for variation in biological response to exposure among the general population and is intended to protect sensitive subgroups, such as the elderly, from adverse effects. * A factor of 10 is used when extrapolating from animate to human* to account for interspecies variability. * A factor of 10 is used when a NOAEL derived from a subchronic rather tha" a chronic study is used as the basis for a chronic RfD. * A factor of 10 is used when a LOAEL rather than a NOAEL is used as the basis for an RfD. For more detailed information on EPA's method for estimating noncancer risks, see U.S. EPA, Risk Assessment Guidance for Superfund Volume 1: Human Health Evaluation Manual, December 1989. 1-22 SL 110226 In addition to the UFs listed above, a modifying factor (MF) ranging from greater than zero to 10 is used to reflea a qualitative professional assessment of additional uncertainties in the critical study and in the entire toxicity database for the substance not specifically addressed by the preceding UFs. The default value for the MF is 1.0. Calculation of the Reference Dose To calculate the RfD, the appropriate NOAEL (or LOAEL if 8 suitable NOAEL is not available) is divided by the product of all of the applicable uncertainty faaors and the modifying factor. The calculation is as follows: -- RfD NOAEL or LOAEL/(UF, X UF,.. JC MF) RfDs are expressed in units of mg/kg-day. Assessing the Potential for Noncammoeenic Effects The Rfd assumes that there is an exposure threshold below which adverse effects will not occur. If the exposure level (dose) exceeds this threshold, adverse effects could potentially occur. This potential for adverse systemic effects is charaaerized by the hazard quotient (HQ). The HQ is calculated according to the following equation: HQ = Dose/Rfd The greater the value of the HQ above one, the greater the potential for adverse effects. However, the HQ is not a statistical probability, a ratio of 0.001 does not mean that there is a one in one thousand chance of the effect occurring. Furthermore, the potential for adverse effects does not increase linearly as the RfD is approached or exceeded. 1-23 SL 110227 VINYL CHLORIDE CHAPTER 2 OVERVIEW OF CHAPTER Vinyl chloride (VC), a commercially important chemical compound used in the plastics industry, can cause significant adverse human health effects if production and use of the chemical are not regulated. In this chapter, we characterize and value the health effects that could have been avoided if the New Substances Notification Regulations (NSNR) had been promulgated at the time VC was introduced into Canada. As described below, the New Substances Notification Regulations would have produced significant economic benefits by reducing individual and worker exposures to VC. Summary of Benefits In the early 1970s. unregulated indoor air concentrations at VC and polyvinyl chloride (PVC) production facilities were extremely high, causing a variety of health problems in plant workers. Further, releases to the ambient air posed significant health risks to residents living proximate to VC and PVC production facilities. As a result, the economic benefits of regulating the manufacture and use of VC are significant As shown in Exhibit 2-1, the mean annual benefits associated with avoiding worker and general population cancer cases and occupational morbidity are estimated to be $53.0 million dollars per year, total benefits over 30 years are estimated to be $916.4 million, using a four percent discount rate. Further, additional avoided morbidity benefits may exist although we have been unable to quantify these benefits for purposes of this report limitation*; To estimate the health effects associated with VC, we make a series of assumptions that introduce uncertainty into the analysis. For instance, our analysis of the incidence of VC-induced cancer in the general population is based on standard assumptions regarding the location and mobility of the affected population, which may over- or under-state total cancer cases. Further, our assessment of occupational health effects relies heavily on incidence data collected in the U.S.; we 2-1 SL 110228 do not know the extent t which plant conditions may differ between U.S. and Canadian facilities.1 Finally, we have been unable to monetize certain health effects in this assessment (e^, general population health effects other than cancer). Specific data limitations are discussed in later sections. Edubit2-1 SUMMARY OF ANNUAL AND TOTAL BENEFITS: CONTROLLING VC MANUFACTURE AND USE IN CANADA * ($1995, tnQUoos) . Category Annual Tuifiu Present Value of Total Benefits <30 yen, 4%) Mortality General Population Occupational $39.1 *$12.1 $676.1 $209.2 Morbidity Genera] Population Occupational Unknown $1.8 Unknown $31.1 Ecological Small Small Total >$53j0 >3916.4 Orpnm.tir.p nf This chapter is oxganized into six sections. The first section provides a general description of the VC and PVC industries, including information on VC releases into the environment. We then summarize available data on human health and ecological exposures to VC in uncontrolled settings. The third section describes the health effects that commonly result from VC exposures, covering both carcinogenic and non-carcinogenic effects. The fourth and most lengthy section of this chapter provides a detailed description of the methods and assumptions used to estimate the incidence of cancer and other health effects in workers and the general population. The final two sections present annual and total benefits estimates, respectively, based on the incidence levels developed in the fourth section. As discussed later in this chapter, we have developed a range of benefits estimates to reflect ihe uncertainties present in the benefits categories we quantify 2-2 % SL 110229 BACKGROUND INFORMATION ON VINYL CHLORIDE Description of the Compound VC is a rhrmie*i compound that has been used in Canada's industrial and commercial sectors for over 50 years. It is a synthetic chemical with no known natural sources. A colorless prtTrvherpiw.i gas, VC has a mild, sweet odor, is slightly soluble in water, and is highly soluble in fats and organic solvents. In addition, VC has a low boiling point, a high vapor pressure, and is a known human carcinogen (US. EPA 1994b; VC in Drinking Water 1992). The primary use of VC is in the production of PVC PVC is used to manufacture plastic and vinyl products such as electricalwire, insulation and cables, industrial and household equipment, medical supplies, food packaging materials, piping, and other building and construction products. PVC is also used as a raw material in the paper, glass, rubber, and automotive industries, and in the production of 1,1,1 trichloromethane (US. EPA 1994b; VC in Drinking Water 1992). In addition, VC was once used as a refrigerant in cooling equipment and as an aerosol propellant in some cosmetics, drugs, pesticides, and other consumer products (Doniger 1978); these uses were discontinued when VCs carcinogenic properties became known. Information on Manufacture in Canaria Unlike PCBs and dichloromethane, VC and PVC are manufactured in Canada. The nation's major production facilities - including two VC plants and four PVC plants - are located in Alberta, Ontario, and Quebec. The locations of these plants are presented in Exhibit 2-2. Expansion of these facilities over the past 20 yean has increased domestic production of both VC and PVC These facilities, however, are highly mechanized, requiring a relatively small number of workers. Fabrication plants, which manufacture end-use consumer products using PVC resins, are more labor intensive (Doniger 1978). &faibit2-2 LOCATIONS OF VC AND PVC MANUFACTURING PLANTS Type of Plant VC and PVC VC and PVC PVC PVC Region Fort Saskatchewan, Alberta Sarnia. Ontario Niagara Falls, Ontario Sbawinigan, Quebec ( Source: EC 1986 ________j 2-3 SL 110230 Two primary processes are used to produce VC One involves the pyrolysis f ethylene dichloride (EDC): ethylene is chlorinated to EDC 'which is then thermally- dehydrochlorinated, producing VC and HO (EC 1986). The other process involves the hydrochlorination of acetylene. In the presence of a catalyst, a vapor-phase reaction occurs between acetylene and hydrogen chloride, producing VC (Lowenheim 1975). To produce PVC, the monomer VC is polymerized. The concentration of VC entrapped in PVC depends upon the production process; it can range from (U to 8,000 ppm (U.S. EPA 1975). In concentrations ofVC monomer in PVC were reduced drastically between 1973 and 1975, and range from 1 to 10 ppm (VC in Drinking Water 1992). Releases to the Environment Emissions and effluents from VC and PVC processing plants axe the primary sources of VC in the work place and in the environment There are, however, other pathways by which VC can enter the environment including accidental spills and the outgas of VC from plastic products. Other release routes include the disposal of VC wastes in landfills; the biodegradation of trichlorethylene, tetrachloroethylene and 1,1,1-trichloromethane; the evaporation of chemical wastes; and tobacco smoke (ATSDR 1993). Once VC enters the environment it can affect a variety of environmental media. Because VC is highly volatile, it vaporizes easily. Therefore, most VC releases end up in the atmosphere. There is potential for the contamination of other environmental media, however, such as water and soil The following subsections provide a detailed description of atmospheric, water, and soil contamination from VC releases, as well as potential release routes. Atmospheric Contamination The primary sources of VC in the atmosphere are VC and PVC plant emissions. PVC manufacturing facilities account for approximately 85 percent of VC emissions, while VC facilities account for approximately ten percent2 Data on total VC emissions to the atmosphere in the early 1970s from uncontrolled plants are limited; it was estimated, however, that a total of 4,500 tons of VC were emitted to the ambient air in 1973. After imposing controls on VC emissions from VC and PVC plants, emissions were reduced by about 95 percent (EC 1986). VC emissions from VC and PVC plants can be categorized by source; regulated point sources, incidental releases, and fugitive emissions. Regulated point sources include equipment such as the reactor, which may release VC when it is opened for Honing, maintenance and inspection; the VC condenser, which discharges VC to the atmosphere after treatment by air strippers or other pollution control devices; equipment downstream from the stripper, and process vents. Emissions from incidental releases generally come from emergency venting due to the malfunction or breakdown of equipment. Fugitive emissions originate from sources other than the vent, stack, flue, 3 The remaining five percent comes from the plastics fabrication industry and other minor sources. 2-4 % SL 110231 or other openings that are specifically designed for the purpose of emitting VC Fugitive emissions can be rm-rf by valves, flanges and seals, open-ended lines, drains, vessel opening losses, building ventilators, and open ponds. Individual occurrences of fugitive emissions may be small, although total fugitive VC releases can be large when emissions from all possible sources are summed (EC 1986). Transportation accidents are another source ofVC releases to the atmosphere, as not all VC is polymerized into FVC at its initial point of production. VC is typically shipped to PVC factories under pressure as a liquefied gas. In the US., for example, about 95 percent of VC is shipped in rail tank cars; the rest is shipped in tank trucks, barges, and tank vessels. An accidental leak, puncture or explosion of these containers can release the liquefied gas, which is then likely to evaporate into the atmosphere and pose a potential health risk to transportation workers, nearby residents, travellers, and safety officers (Doniger 1978). Other potential sources of VC in the atmosphere include the outgas of VC fronanew plastic products and the volatilization of VC products in municipal aihd hazardous waste landfills, In addition, VC is released in the indoor air of VC and PVC plants. As a result, VC and PVC workers without proper protection can be exposed to potentially significant concentrations of the compound. Water Contamination VC releases into the environment can also affect surface water, groundwater, and drinking water. Surface Water. The most likely source of surface water contamination is wastewater from VC and PVC manufacturers. In Canada, the effluent from the production of PVC is discharged into a variety of waterways, including a waterway in Quebec and the Welland River and St. Clair River in Ontario (EC 1988). VC reaching surface water via contaminated effluent is likely to migrate rapidly, on the order of hours or days, to the air (ATSDR 1993). The presence of materials such as fumates, surfactants, and particulates, however, may extend the residence time of VC in water (U.S. EPA 1980). Groundwater. VC can leach into groundwater as a result of spills, leachate from landfills and hazardous waste sites, and effluents from the plastics industry (ATSDR 1993). In addition, the biodegradation of trichloroethylene and perchloroethylene in groundwater can form VC For a variety of reasons, including resistance to microbial degradation, VC can remain in groundwater from months to yean (VC in Drinking Water 1992). Drinking Water. Small amounts of VC may be present in public water supplies as a result of industrial wastewater discharges (U.S. EPA 1980) or as a result of groundwater contamination (see above). The occurrence of VC in potable water, however, is primarily associated with the conveyance of water via PVC pipes manufactured with incompletely polymerized VC monomer. These extremely small amounts of entrapped VC monomer can leach from PVC pipes (VC in Drinking Water 1992). Since the level of 2-5 SL 110232 residual VC that is not molecuiariy bound in PVC pipe is extremely low, cigwifirant exposure to VC. via this pathway is considered unlikely (Coiner 1996). Soil Contamination The primary sources of VC releases to soil are spiDs, eDuent discharges, and leachate from disposed VC materials. VC is likely to volatilize rapidly from dry soQ surfaces, however, due to its high vapor pressure. VC placed ten centimeters deep in dry soil, for example, is expected to have an effective half-life of about 12 hours. In addition, VC has' a low soD organic- carbon, adsorpti n coefficient, and is highly mobile in soil, suggesting a strong potential to migrate' from soil to groundwater (ATSDR 1993). HUMAN AND ECOLOGICAL EXPOSURES TO VINYL CHLORIDE This section describes the concentrations of VC to which humans and biota were likely to be exposed in the early 1970s. The focus of this section (and the remainder of the analysis) is on human health, as ecological impacts are a relatively minor concern for this compound. The following subsections describe the most common pathways for human exposure to VC, and the concentrations at which VC was found in different subpopulations. We also provide a brief summary of potential ecological effects of VC exposure. As discussed in the first chapter of this report, we focus on emissions from uncontrolled VC sources to estimate human exposures in this analysis. This approach follows logically from the ultimate goal of this project, to evaluate the benefits of controlling or prohibiting new substances. As a result, much of the information we present in this section is based upon data from the 1970s, when VC emissions were largely uncontrolled. Human Exposures VC and PVC workers and individuals living in close proximity to VC and PVC production facihties were exposed to the greatest concentrations of VC Exhibit 2-3 summarizes the typical concentrations to which various subpopulations were exposed in uncontrolled conditions. 2-6 SL 110233 2-3 TYPICAL EXPOSURE TO VINYL CHLORIDE VIA INHALATION IN THE EARLY 1970s Subpopntation ^QQOCQttltoOD General population Oppb Individuals living in dose proximity to VC and PVC plants 0 to 814 ppb Average VC and PVC workers 1 to 100 ppm Maximally exposed VC and PVC workers 250 to 300 ppm VC concentrations in ambient air were generally fairly low. Background levels of VC are often expressed in pans per billion (ppb), whereas indoor air concentrations of VC in VC and PVC plants are expressed in parts per million (ppm). The average concentration of VC to which populations living in the vicinity of emissions sources were exposed ranged as high as 814 ppb (ATSDR 1993). For individuals not living in the vicinity of point sources, exposure to VC via inhalation was expected to be essentially zero. Inhalation is also the primary route for occupational exposure. Workers responsible for cleaning reactor vessels, which are the chambers in which VC is convened to PVC in polymerization plants, were frequently exposed to airborne concentrati ns on the order of 250 to 300 ppm (Amdur 1991). In addition to inhalation, humans can be exposed to VC via ingestion. As discussed above, VC can be found in drinking water at very low concentrations due either to the contamination of underground sources of drinking water or to the migration of VC from PVC pipes. In addition, VC can migrate from packaging material containing VC into food, since the monomer is soluble in alcohols and mineral oil. The presence of VC in drinking water or food, however, was relatively uncommon, and the average daily intake of VC through diet was essentially zero (ATSDR 1993). The only potentially significant exposure to VC via ingestion was found in PVC production plants, where workets could have been exposed to VC through the ingestion of PVC dust containing entrapped VC monomer (U.S. EPA 1975). Ecological Exposures Data on the ecological effects of VC contamination are limited. It is unlikely, however, that VC releases to the environment posed a threat of significant ecological damage, due to the quick volatilization of VC into the atmosphere. The potential for biomagnification in the food chain does not appear to be important because of the high volatility of VC and the fact that it is readily metabolized by higher-trophic-level organisms (ATSDR 1993). 2-7 SL 110234 One of the by-products of VC production may cause ecological damage if discharged int aquatic environments, as this by-product (EDC tar) can adhere to substances such as plankton. In addition, marine animals rapidly accumulate EDC tar from contaminated sea water. Since these tars have a relatively short half-life, ecological effects from exposure were likely to be much less severe than from exposure to other types of chlorinated hydrocarbons (UJS. EPA 1975). THE BENEFITS OF CONTROLLING VINYL CHLORIDE RELEASES: OVERVIEW This section provides an overview of the benefits associated with the prohibition or control f VC in Canada. The primary benefits of controlling VC emissions are reductions in mortality and morbidity due to VC exposures in the general population and among VC and PVC workers. Once VC is inhaled or ingested, it can cause a variety of adverse health effects. In this section, we summarize the considerable amount of information available on the health problems related to VC exposure. Mortality from Cancer In the early 1970s, the cancer-causing capacity of VC was recognized. The available epidemiological studies tend to focus on the incidence of angiosarcoma of the liver, a rare form of cancer, because of the ability to draw a causal relationship between exposure to VC and the development of tumors. Exposure to VC by inhalation is also suspected to increase the risk of cancer of the brain, lung, and digestive tract (US. EPA 1994b). The types of cancer associated with VC are often fatal. Studies demonstrate, for example, a significant association between exposure to PVC dust and the risk of lung cancer mortality (Wagoner 1983). The average life expectancy (post-diagnosis) for an individual who has contracted cancer via exposure to VC is estimated to be 0.4 to 8.9 yean (Sax 1981). Due to the nature of these VC related cancers, we value the cancer cases estimated from VC exposures (later in this chapter) as fatalities. Morbidity In addition to cancer, VC can cause a variety of other deleterious health effects. For instance, exposure to VC can cause hepatic, respiratory, cardiovascular, gastrointestinal, hematological, musculoskeletal, renal, and dermal/ocular problems. Specific examples of common VC-related problems are listed in Exhibit 2-4. A number of studies also suggest that exposure to VG^nay*affect sexual function and hormonal levels in humans-and animals (ATSDR 1993). In addition, some studies suggdrtaiSOfrHtically significant increase*'in-chromosomal aberrations in humans exposed to VC, and a variety of studies show that members of communities with nearby VC 2-8 SL 110235 polymerization facilities have greater incidences of some f rms of developmental toxicity.3 For example; some of these studies show an excess of fetal loss ydicn. fathers,have been.exposed to VC These studies and others, however, fail to demonstrate a statistically significant correlation between developmental toxicity and either parental occupation or proximity to the facility (ATSDR 1993). Exhibit 2-4 NONCARC3NOGENIC HEALTH EFFECTS FROM VINYL CHLORIDE EXPOSURE Type of Exposure Health Effects Acute exposure to high levels of VC Dizziness, headaches, giddiness.- ligbt-beadedness, nausea, poor memory, tingling sensations, weight loss, chronic bronchitis Chronic exposure to lower levels of VC Liver damage, Raynaud's syndrome, dnwlatoiy disturbances m extremities, thrombocytopenia, dermatitis, scleroderma-like skin changes, lytic lesions of the terminal phalanges in hands and feet, pseudo dubbing of fingers, thyroid insufficiency, damage to the spleen and lungs, functional disturbances of the central nervous system Source: ATSDR 1993 ESTIMATING HUMAN HEALTH EFFECTS The following discussion provides specific information on the health-related benefits of controlling VC in Canada. In particular, we describe the detailed methodologies used to estimate general population and occupational cancer cases, as well as morbidity effects in workers. The information developed in this section (e.g^ estimated annual cancer cases) serves as the basis for the benefits estimates developed in the following section. Cancer Cases Resulting from Exposure to Vfnvl Phlyriri,. General Population In 1973, the U.S. Environmental Protection Agency conducted a health risk assessment for community exposure to VC, in support of regulating air emissions around VC and PVC plants (Kuzrnack 1975). To estimate potential exposures, SPA divided the area surrounding each plant into four bands ranging up to five miles from the plant. Through standard diffusion modeling of over 40 plants, the annual average ambient VC concentrations in each band were calculated. Detailed information for each individual plant was weighted to reflect differences in the type of plant, size of plant, and meteorological and topographic conditions at the plant site. The results were then used to estimate average VC exposures for individuals living in the vicinity of a "model" VC or PVC plant. Developmental toxicity is the occurrence of adverse effects that may result from exposure to a chemical prior io conception, during prenatal development, or postnatallv to the time of sexual maturation. 2-9 SL 110236 Anrnn.1 fVw We use a fanr approach to estimate potential VC exposure for the population living close i VC and PVC plants. First, we estimate the size of the affected population. Because the VC and PVC industry is based in only four Canadian cities, the size of this subpopulation is relatively small. Exhibit 2-5 presents estimates of the population that resides within specified distances of Canada's two VC and four PVC plants. We developed these estimates using information on the total size and denary of the population in each of the cities of interest. A detailed description of bow these estimates were developed is provided in Appendix 2A Exhibit 2-5 ESTIMATED POPULATION LIVING WITHIN A FIVE-MILE RADIUS OF VC OR PVC PLANTS Distance from Plant VC Plants PVC Plants Band 1 (0-0.5 mi) Band 2 (05-1 mi) Band 3 (1-3 mi) Band 4 (3-5 mi*) 1,444 4331 38.142 42340 2376 8329 61313 107389 | | | To estimate the concentrations of VC to which individuals within each band might be exposed, we use the ambient air concentrations provided in the U.S. EPA study, which are presented in Exhibit 2-6. These figures represent average annual VC concentrations in ambient air near an uncontrolled VC and PVC plant under typical meteorological conditions.4 As shown in this exhibit, VC concentrations are higher around PVC than VC plants. Further, as one would expect, concentrations drop substantially with distance from the facility. Exhibit 2-7 summarizes our high-end estimates of the number of cancer cases associated with each type of facility. As shown by the rows labelled "VC Concentration" and "Dose", we estimate the annua] dose of VC to which the average person in each band could be subjected based on the estimated annual average concentration of VC to which the individual is exposed, adjusted for average adult inhalation rates and body weight5 *We calculate the individual lifetime cancer risk by multiplying the dose by the cancer slope factor for VC exposure via inhalation.* Finally, we 4 The ambient air concentration estimates presented in the U.S. EPA study are a result of two independent studies, the results of which differed by less than 25 percent The data presented in the EPA assessment included variations in meteorological conditions from location to location (Kuzmack 1975). 5 The average adult inhalation rate is assumed to be 20 mVdav, and the average adult body weight is 70 kg (U.S. EPA 1989). ` The inhalation cancer slope factor for vinyl chloride isi03i(mg/kg-day)^(U.S. EPA 1994a). 2-10 SL 110237 estimate the number of annua} cancer cases by multiplying the population by the cancer risk and dividing fay the avenge lifespan (70 years). The rows labelled "Annual Cancer Cases" in Exhibit 2*7 provide a summary of our best higb-end estimates of the number of cancer cases likely to be prevented under the Hew Substances Notification Regulations for those individuals living near a VC or PVC plant7 As shown at the bottom of the exhibit, we estimate that 1.7 cancer cases would occur around VC facilities and 95 cases around PVC facilities, for a total of 112 avoided annual cancer cases. Exhl*2-6 AVERAGE ANNUAL VINYL CHLORIDE CONCENTRATIONS IN AMBIENT AIR Vinyl Chloride Concentration (ppb) Distance from Plant Band 1 (0-1X5 mi) VC Phut . - 113 PVC Plant 323' Band 2 (05-1 mi) 20 57 Band 3 (1-3 mi) 5 15 Band 4 (3-5 mi) 2 6 Source: Kuzmack 1975 An on-going U.S. EPA review of VCs cancer slope factor suggests that the number of cancer cases estimated above may be high, Thejrew slope factor wiHincxrrporate pharmacokinetics, which will?reduce the dope factor and thus the indmduaiJifctime oncer risk (Guth 1996).* To reflect thisipfoirmation, we have calculated low-end estimates by dividing our high-end estimates by a factor oflO. As a result, we value the general population mortality benefits in the next section by using annual case estimates that range from 1.1 to 112 annual cancer cases. 7 A detailed example of how these estimates are developed is provided in Appendix 2B. * Pharmacokinetics represents the movement of a substance within the body, which is affected by the uptake, distribution, elimination and transformation of the compound within the body. Thus, a pharmacokinetic model develops an estimated internal dose of the contaminant that is a more accurate measure of exposure than the applied dose. SL 110238 Exh&it2-7 GENERAL POPULATION CANCER CASES; HIGH-END ESTIMATE | Band 1 ((MLS mi) Band 2 ((15-1 mi) Band 3 (1-3 mi) Population VC CooccntntioD Date CunStt Annual Gubtt Cases Populatioo VC Concentration Dote Cmuw BiriAnnual Cinrrr Cases Dose Oncer Risk Annual Band 4 (3-5 mi) Population VC Concentration Dose Oncer Ride Annual TOTAL ANNUAL CANCER CASES Cam VC Plant 1,444 0394 mg/m* 0.084 (mgfcg-day) 23X10* 05 4331 0052 mg/m* 0.015 (mg/kg-day) 43X10* 03 38,143 0014 mg/m* 0.004 (mg/kg-day) 13X10* 06 42340 0005 mg/m* 0.001 (mg/kg-day) 43X10* 03 1.7 FVC Plant 2376 0.840 mg/m3 0340 (mg/kg-day) 73X1(H 3.0 8,629 0.148 mg/m3 0.042 (mg/kg-day) 13X10* 13 61313 0.039 mg/m3 0.011 (mg/kg-day) 33X10* 23 107389 0.015 mg/m* 0.004 (mg/kg-day) 13X10* 10 93 2-12 SL 110239 The methodology described above requires a senes of. assumption^thatXmay introduce uncertainty into the results ofthis analysis;. For instance, weassume that the population is evenly rfjitm'hiititrt throughout the repcm! analHSrthe VC or FVC plant is located in the center of the region. Further, our approach assumes that the potentially affected population breathes the ambient air within their respective band 24 hours a day, seven days a week, 365 days a year. Our methodology does not consider the travel patterns and daily mobility of each population, and thus the variation in levels of VC exposure. These simplifying assumptions may lead to an over* or under-estimate of cancer among the subpopulation surrounding VC and FVC plants. Occupational Before much attention was given to the monitoring of VC levels in the work place, indoor air concentrations exceeded what health authorities now consider to be acceptable exposure levels* In the absense of regulations and subsequent controls on VC-related activities, it is possible that VC and PVC workers could be subject to potentially significant cancer risks.*10 Fsrimatmp Annual Cams To estimate the annual cancer cases expected within the population of VC and PVC workers under unregulated conditions, we must first estimate the size of the affected occupational population. We do so using data from US. VC and PVC facilities. In the early 1970s, 940 production workers were engaged in VC production at 17 US. plants, for an average of 55 workers per plant At PVC facilities, approximately 5,600 workers were engaged in production at 40 facilities, for an average of 140 workers per plant (US. EPA 1975). In Exhibit 2-8, we use these average figures, adjusted upward and downward fay 50 percent to account for uncertainty, to project the average number of Canadian workers potentially exposed to high concentrations of VG,UJ As shown in this exhibit, we estimate that between 336 and 1,006 Canadian production workers could be exposed, in the absence of occupational and environmental safety standards, to high concentrations of VC. ' As of 1978, high levels of VC in the workplace were believed to be responsible for angiosarcoma of the liver in at least 68 workers worldwide (Doniger 1978^ 10 For the purpose of this analysis ;we. ignore the ;potential.mitigation of health risks by -`occupational health and'safety requirements,', which may only be applied after a problem is discovered. n For instance, the "low" estimate of VC workers per plant is simply 55*0.5, or 28 workers. u The adjustments for uncertainty reflect the fact that we do not have adequate information on labor intensity and plant size across U.S. and Canadian facilities. 2-13 S-L 110240 &Ubit24 ESTIMATED VC AND PVC WORKER POPULATION IN CANADA VC PVC Total Wodcen per plant 28-83 70-210 -- Total, aO ph"** 56-166 280 - 840 336 1,006 Hie remaining data we use to estimate annual cancer cases among VC and PVC workers is derived from published epidemiologic studies of health effects among workers who were exposed to high concentrations of VC before 1974. We take this approach for two reasons. First, although these studies do not include specific estimates of worker exposure levels, they do provide the most direct indication of the health risks associated with working in these pints. .Second, the available cancer slope factor for VC is applicable to exposures up to only several hundred pg/ro} (approximately 0.1 ppm). It has been well documented that in the U.S. before 1974, exposure to VC concentrations of 250 to 300 ppm was common for many job categories in VC and PVC industries (Doniger 1978).11 * *In short, the expected ambient concentration of VC in. the air of uncontrolled VC and PVC plants significantly exceeds the upperbound concentration for which the cancer slope factor is appropriate. Based on a review of four different studies, the US. EPA estimated that the incidence rate of liver angiosarcomas among highly exposed VC and PVC workers (i-e., working in areas with VC concentrations of roughly 200 to 500 ppm), adjusted for the average latency period, is approximately 3.1 x 104 cases per worker-year of exposure (Kuanack 1975). Assuming an average career length of 20 yean for these workers, their individual lifetime cancer risk is estimated to be 62 x 104. In ther words, approximately six percent of all highly exposed VC and PVC worken would be expected to contract liver angiosarcoma during the course of their lifetime. If we assume that all of the VC and PVC worken in Canada were exposed to high levels f VC in the early 1970s and would have continued to be subject to these exposures in the absence f the Regulations, the incidence rate of cancer among VC plant worken would range from 3.4 to 10.0 cases and among PVC plant worken would range from 16.8 to 50.4 cases, for a total ranging from 20.2 to 60.4 cases. An incidence rate of 20.2 to 60.4 cancer cases is equivalent to approximately one to three total cancer cases per work year for a work force of the size presented in Exhibit 2-8.14 It was also not uncommon for peak exposures to reach up to 4,000 ppm for short amounts f time. Worken in other occupations, before the control of VC levels in the work place, are likely to have been exposed to lower and less sustained concentrations. M ^DC t0 **iree canccr cases per work year is high for a relatively small work force. Therefore, it^is unlikely that this situation would continue unabated year after year once the problem was discovered. The benefits associated with even a few years of improved control, however, are significant. 2-14 SL 110241 limitations In -ctimaTm off-npatinnai effects, we use incidence information from U.S. workers. To the extent that r*nriin and U.S. worker exposures are different, our estimates could be under or over estimates. Further, we do not take into account the incidence of other types of cancer in the above agreement (U, cancers other than angiosarcoma of the liver), as we only have qualitative ;nfnn*tinn cm the incidence of other types of cancer. To the extent these cancers would occur, the estimates presented above could be under-estimates. Noncardnogenic Effects Resulting from Exposure to Vmvi Qiloride General Population: Point Sdtnce Exposures Because uncontrolled VC and PVC plants emit high concentrations of VC to the ambient air, individuals living within a five mile radius of these plants could experience noncardnogenic health effects related to VC emissions. We determined the likelihood of risk of noncardnogenic effects by calculating a hazard quotient for potentially exposed populations. The hazard quotient is the ratio of the expected dose to a reference dose (RfD) or reference concentration (RfC); a hazard quotient greater than one is an indication that some noncardnogenic effects may occur.15 Generally, an RfD or RfC is used to determine the level below which there are no observed adverse impacts in humans due to daily exposure to a particular toxic substance. The U.S. Environmental Protection Agency, however has not established an RfD or RfC for VC (ATSDR 1993). Instead, a minimal risk level (MRL) for VC of 0.002 ppm has been derived by ATSDR, based on an animal study." Exhibit 2-9 presents the hazard quotients for each band surrounding a VC or PVC plant As indicated in this exhibit, the hazard quotients for individuals living within each band equals or exceeds one in all cases. This information suggests that there is a considerable likelihood that some noncardnogenic effects could occur in the general population due to VC exposures. We n te, however, that lS; EPA is m the process of developing an RfC for.VC Although this value has not been released as of yet, it is likely to be higher than the MRL established by ATSDR (Guth 1996), which would produce lower hazard quotients. As a result, we believe that the figures presented in Exhibit 2-9 should be employed with caution. " See Appendix 1A for more information on this topic. u The results of a study conducted by Bi, et aL (1985) indicate that exposing rats to 10 ppm of VC for ax hours a day, six days a week, for six months of the year, results in increased fiver weight. The iowestbbseiVed advene effect level (LOAEL) is lO ppm/ ATSDR used the results of this study to establish the MRL of 0.002 ppm for humans, using a series of adjustment factors. The adjustment factors account for (1) the difference between intermittent and continuous dosing, (2) the uncertainty surrounding the difference between a LOAEL and a NOAEL, (3) the uncertainty sunounding the extrapolation from animal data to humans, and (4) the uncertainty surrounding hliman variability. The MRL is meant to be used for screening purposes only. The MRL is not used for regulatory action. 2-15 SL 110242 Exhibit 2-9 GENERAL POPULATION NONCARCINOGENIC EFFECTS VC Plant Band 1 (0-05 mi) Dose m (mg/kg-dav) Hazard Quotient 0084 57 Band 2 (05-1 mi) Dose in (mg/kg-day) Hazard Quotient 0015 io Band 3 (1-3 mi) Dose in (mgAeg-day) Hazard Quotient 0004 4 Band 4 (3-5 mi) Dose m (mg/kg-day) Hazard Quotient 0001 1 Note: MRL 0.002 ppm * 0.001 (mg/kg-dsy) PVC Ham 0240 162 0.042 -29 . O011 8 0.005 3 General Population: Other Exposures The general population as a whole may face health risks from VC exposure due to sources of release other than emissions from VC or PVC plants, such as transportation accidents, drinking water contamination, or direct use and disposal of VC under unregulated conditions. In the following subsections, we briefly describe these other sources of VC release and their potential contributions to adverse health effects in the general population. The discussion focuses on noncarcinogenic health effects, since the incremental cancer risks associated with these sources of release are likely to be dwarfed by the cancer risks associated with emissions from VC and PVC production facilities. Transportation Accidents VC produced in Canada is shipped by pipeline to neighboring PVC plants and by railroad to other plants in Canada, the U.S. and Mexico. VC was also once shipped in tanktrucks, but this practice was discontinued in 1975 due to the unacceptable risk of highway accidents (EC 1986). Shipment of VC by rail reduces the risk of accidental release, but does not totally eliminate it. In 1980, for example, a train derailment in MacGregor, Manitoba resulted in a spill of 49.7 tons of VC; accidental releases of VC from tank cars have also occurred in the U.S. Nonetheless, the safety record for VC transport in recent years has been good, suggesting that the risk of noncarcinogenic health effects in the general population resulting from transportation accidents is likely to be small 2-16 SL 110243 priifrmg Water In the absence of regulatory controls, the general population might be exposed to elevated concentrations of VC in drinking watcQttuc primarily(to the leaching of*yC from PVC pipes. The available data, however, are insufficient to allow us to estimate tire extent to whichsuch exposures ought occur. For example, the leaching ofVC from PVC pipes depends on several factors, including dwell time, which is the period of time water stands in the pipes; leaching rate, which is determined by the amount and distribution of VC in the pipe and in the water, the equilibrium partitioning of VC between the pipe and the water, and the diffusion of leachate out of the pipe and into the air. Few data exist on pipe/water partitioning coefficients or plastic pipe diffusion coefficients, making the prediction of VC concentrations in water difficult (Podoll 1986). Despite these limitations, the available data suggest that the risks associated with VC in drinking water are likely to be several orders of magnitude below the risks associated with inhalation of VC production emissions. For example, a drinking water study conducted by the U.S. Environmental Protection Agency examined VC concentrations in the drinking water of five water distribution systems using PVC pipes. The highest VC concentration was 0.54 ppb, which came from the most recently installed and longest pipe system. Traces of VC (0.012 to 0.023 ppb) were found in two other systems nine years after installation (VC in Drinking Water 1992). These concentrations are so low that the risks of adverse health effects appears minimal. Othar vr 11m* and ftrieaw* In the absence of regulation, VC might be used in a variety of consumer products. The associated adverse health effects would depend on the type and extent of use, manufacturing processes, and subsequent disposal practices. Since the range of potential practices is highly uncertain, we have not attempted to analyze the health risks they might present. Occupational As discussed earlier in this chapter, VC is known to cause a number of noncarcinogenic health effects in workers exposed to high concentrations of the substance. Information on the incidence of various effects, based on research reported by U.S. EPA in the 1970s, is summarized in Exhibit 2-10. As shown in this exhibit, the incidence rates for a number of these health effects are quite high, particularly for exposed workers. While they do not present detailed exposure data for these workers, these studies do provide a dear indication of the range of potential noncarcinogenic effects for workers exposed to VC To provide an indication of the quantitative benefits of reducing or eliminating these health effects, we assume that 10 percent of all highly exposed workers would suffer from one or more noncarcinogenic effects that would impair their daily activities. Based on a worker population of between 336 and 1,000 (see Exhibit 2-8), we estimate that between 34 and 101 workers would be thus affected if VC emissions were uncontrolled. 2-17 SL 110244 Exhibit 2-10 INCIDENCE RATE OF OCCUPATIONAL NONCARC3NOGENIC EFFECTS Study Health Effect Incidence Literature Review of 12,724 PVC Workers Acroostaofysis1 Raynaud's Syndrome Symptoms2 Kirin 1 >tinm 0.9% 0.8% 03% Liver Disturbances* 0.6% Clinics] Observations of50 Highly Exposed PVC Workers Raynaud's Syndrome Symptoms2 Liver Disturbances* 16% 76% Liver Function Screening Programs for 1,183 PVC Workers Liver Function Abnormalities4 43% Source: Kuzmack 1975 Notes: 1. Gradual erosion of bone at fingertips 1 Cold bands and feet 3. Changes in liver function suggestive of cellular liver damage 4. Abnormalities considered serious enough by plant physicians to justify moving workers to areas where there is no exposure to hepatic toxins VALUING HUMAN HEALTH EFFECTS: ANNUAL VALUES In this section, we describe the process of valuing the health effects analyzed above. As indicated in Exhibit 2-11, we estimate the mean annual benefits associated with avoided morbidity and mortality to be S53.0 million. Below we discuss our approach and calculations. Eah2iit2-ll SUMMARY OF BENEFITS ESTIMATES Category Mem Annual Benefits (1995$, millions) Mortality General Population Occupational Morbidity General Population Occupational Ecological $39.1 Sill Unknown $13 Small Total >$533 2-18 SL 110245 Carcinogenic Effects: Mortality For purposes of this analysis, we assume that all cancers attributable to VC exposure would be fatal. This assumption is based on the particularly lethal nature of angiosarcoma of the liver, the type of cancer most often contracted from VC exposures.17 To value these effects, we utilize recent research into the statistical "value of life" conducted by XEc for the U.S. EPA's Office of Air and Radiation. As part of this effort, IEc evaluated three types of value of life estimates: wage-risk studies, which estimate the additional compensation individuals demand in the labor market for tairing ridrier jobs; contingent valuation studies, in which individuals are asked to state their wfllingness-to-pay to avoid additional increments of statistical mortality risk; and consumer behavior studies, which examine situations other than the labor market that involve risk-dollar tradeoffs. Based on this research and analysis, we have developed a lognormal distribution of the value of a statistical life, using a mean value of $63 million and a standard deviation of $4.4 million (S1995). Appendix 2C presents a list of the studies used to develop this estimate, which draws largely from a survey article authored by one of the field's leading researchers (Vbcusi 1992)." In Exhibit 2-12, we summarize the benefits associated with avoided cancer cases from VC exposure in the general population surrounding VC and PVC plants. As shown in this exhibit, the majority of the benefits come from reducing VC exposures to the general population living close to PVC plants. This is not surprising, considering that, under uncontrolled conditions, ambient concentrations of VC around these plants are estimated to be more than twice those around VC plants. Note that the values presented in this table are the tenth percentile, mean, and 90th percentile values from a 1,000 iteration Monte Carlo analysis. We generated the distribution using the lognormal distribution described above to characterize the statistical value of life, coupled with a uniform distribution to represent the estimated number of annual cancer cases, which range from 1.1 to 11.2 per year." 17 We note that other researchers have valued carcinogenic health effects using a weighted average of estimates of the "value of life" (for mortality) and "cost of illness" (for non-fatal cases) (CCME 1995). While this approach may be appropriate for forms of cancer that are potentially treatable, we believe that using a weighted average that includes "cost of illness" estimates in our analysis would result in an underestimate of the benefits of avoiding a generally fatal form of cancer. '* The studies cited in Appendix 2C include analyses of compensating wage differentials for fatal injury risk in several countries, including Canada; the Values are reported in the source document in American dollars, and have been converted to Canadian dollars using the method outlined in Chapter 1. We note, however, that the list of studies does not include four wage-risk studies for Canadian labor markets (Vodden et al. 1994, Martinello and Meng 1992, Meng and Smith 1990, and Meng 1989). We have not reviewed these four studies in detail, but have reviewed a summary of the studies contained in a report prepared for the Canadian Council of Ministers of the Environment (CCME 1995). The value of statistical life estimates reported for these four studies are comparable to those presented in Appendix 2C, with means ranging from $4.7 million to $7.6 million. The mean of the distribution of values upon which we rely ($63 million) would be unchanged if these studies were added to those cited in the appendix. w A uniform distribution in this context means that there is an equal chance that any value between 1.1 and 11.2 is the true number of cases. 2-19 SL 110246 Exhibit 2-12 ANNUAL BENEFITS ASSOCIATED WITH AVOIDED DEATHS FROM CANCER IN THE GENERAL POPULATION ($1995, ndfioos) T>pe of Plant Low Estimate (10th Percentile) MGQ TXUfuI^LD jGtP^^n^BC (90th Percentile) VC PVC Totals S1.5 S&8 $103 S&jO S33.1 $39.1 $12.0 S6&.0 $80.0 I | In Exhibit 2*13, we summarize the benefits associated with avoided cancer cases from occupational exposures to VC We used the lognormal distribution to represent the statistical value of life, as described above, and a uniform distribution to represent the likely number of annual cancer cases, which we estimate to range from one to three per year. As shown in this exhibit, the majority of the benefits estimated come from avoided cancer cases among the PVC worker population, which is substantially larger than the population of workers in VC plants. Eriribit2-13 ANNUAL BENEFITS ASSOCIATED WITH AVOIDED DEATHS FROM CANCER IN THE OCCUPATIONAL POPULATION _____($1995, millions) Type of Plant VC PVC Totals Low Estimate (10th Percentile) $0.8 $3.4 $4.2 Mean $2.1 $10.0 $12.1 High (90th Percentile) $3.9 $19.0 $22.9 To provide a sense of the uncertainty associated with the values presented in this table, Exhibit 2-14 presents a distribution of potential benefits associated with occupational exposures from PVC plants. As shown in this graphic, total benefits estimates could potentially range from S1.4 million to $41.4 million, based on this particular Monte Carlo run of 1,000 iterations. 2-20 SL 1X0247 Exhibit 2-14 DISTRIBUTION OF ANNUAL BENEFITS FROM AVOIDED PVC WORKER CANCER CASES 0.19 0.15 0.11 < H -- r " C 3 > W O S 5 `9 0.07 0.04 Co It* 0.00 too 1.4 8.0 14.7 21.4 28.0 34.7 41.4 CO Values In millions of dollars Noncareinogenic Effects: Morbidity As discussed above, there are a number of noncardnogenic effects that may result from acute and/or chronic exposures to high concentrations of VC The benefits associated with reducing morbidity can include the following. Avoidance of direct medical costs, including the expenditures to combat a particular illness, such as doctor's fees, hospitalization costs, and pharmaceutical purchases. Indirect costs, including lost wages and leisure time due to Qlncss. -1 Pain and suffering, which encompasses the physical and psychological discomfort associated with morbidity effects (Unsworth 1993). Several estimation techniques have been developed to assign economic benefit estimates to reductions in air pollution related morbidity effects, although these techniques generally address only a portion of total morbidity benefits. Much of the morbidity literature associated with air pollutants focuses on respiratory effects (e^, asthma and bronchitis symptoms caused by air pollutants). While VC is known to cause some respiratory problems (e.g., chronic bronchitis), more commonly cited noncardnogenic effects are f the type cited in the previous section, such as Raynaud's Syndrome and acroostaolysis. As a result, we value morbidity effects using information collected in the CCME (1995) study on restricted activity days (RADs). RADs are a measure of illness defined as a day on which illness prevents an individual from engaging in some or all of his or her usual activities, including days spent in bed and days with minor activity restrictions due to illness. As part of valuing morbidity in the CCME study, the authors analyzed a number of studies estimating RADs. Based on this review, they provide a central estimate of $70 (1994$) per RAD, with a low estimate of $35 per RAD and a high estimate of $105 per RAD. For purposes of our analysis, we have adjusted these figures to annual values by multiplying by 365 days, and have adjusted the figures to 1995 dollars.3 As a result, we value morbidity effects assuming the annual benefits of avoiding restrictions on activity range from $13,000 to $39300 per year, with a mean value of $26,100. In Exhibit 2*15, we summarize the benefits of avoided morbidity from occupational exposures. To develop these estimates we use a discrete probability distribution to represent the estimated annual RAD values, and a uniform distribution for the number of cases, which are estimated to range from 34 to 101 per year. As shown in this exhibit, we estimate total benefits to range from $0.7 million to $33 million per year, with a mean value of $13 million, based on a Monte Carlo run of 1,000 iterations. 3 Implicit in this approach to developing an annual benefits estimate is the assumption that the activity of affected VC and PVC workers is restricted 365 days a year. Many of the health effects f interest (e.g., acroostaolysis) could have such persistent effects; however, to the extent these health effects would not impair activity year-round, the methodology lends an upward bias to the benefits estimate. 2-22 SL 110249  Exhibit 2-15 ANNUAL BENEFITS ASSOCIATED WITH AVOIDED MORBIDITY IN VC AND FVC WORKERS (in minima) Low Estimate (lOtb Percentile) Mean Estimate (90th Percentile) S0.7 SO $33 VALUING HUMAN HEALTH EFFECTS: TOTAL VALUES The results of the previous section indicate that the annual benefits of controlling VC are considerable. In addition, these benefits presumably would continue to accrue for a number of yean, since, in the absence of controls, health effects would continue unabated. In Exhibit 2-16, we summarize our estimates of annual benefits and the present value of total benefits attributable to an environment free of VC As illustrated in this exhibit, we estimate that annual benefits (f r quantifiable benefits categories) range from $152 million to $1062 million, with a mean annual benefits estimate of $53.0 million. Further, the present value of benefits ranges from $02 billion to $2.4 billion over a 30 year period, depending upon the annual benefits estimate (low, mean, or high) and discount rate employed. See the discussion on discounting in Chapter 1. 2-23 SL 110250 exhibit 2-16 SUMMARY OF ANNUAL AND TOTAL BENEFITS: VINYL a ILOR1DE ($1995, millions) I Annual Benefits Present Value of Total Benefits - 30 yean ] H Category Low Mean High Low @ 6% Mean @ 4% High @ 2% 1 | Mortality I General | Population | Occupational $10.3 $4.2 $39.1 $12.1 $80.0 $22.9 $141.8 $57.8 $676.1 $209.2 $1,791.7 I $5119 1 I Morbidity | General I Population y Occupational | Ecological 1 Total Unknown Unknown Unknown Unknown $0.7 Small > $15.2 $1.8 Small > $53.0 $3.3 Small > $106.2 $9.6 Small > $209.2 Unknown $31.1 Small > $916.4 Unknown 1 $73.9 I Small | >$23783 | 2-24 references for CHAPTER 2 ATSDR 1993, Agency for Toxic Substances and Disease Registry, US. Public Health Service. Toxicological Profile for Vinyl Chloride (Draft). Apnl 1993. Amdur 1991. Amdur, Mary On John DouU, and Curtis D. Klaassen. Casarett and Doull's Toxicology: The Basic Science of Poisons. Fourth Edition. New York: Pergamon Press, 1991. Boettner 1982. Boetmer, EA^ GX. Ball, Z. Hollingsworth, and R. Aquino. "-Organic and Organotin Compounds Leached From PVC and CPVC Pipe," Univenatyof Michigan, Ann Arbor School of Public Health, 1982. Cash 199Z CarL "Shattering PVC roof membranes," Progressive Architecture, VoL 73, No. 2, p. 43, February 1992. CCME 199S. Environmental and Health Benefits of Cleaner Vehicles and Fuels. Summary Report and Supplemental Reports prepared for the Canadian Council of Ministers of the Environment Prepared by Carolyn Lang, Hagler Bailly Consulting; Greg Yarwood, ENVIRON International, Inc; Francois Lalonde, Environment Canada; and RoberrBloxara, Ontario Ministry of Environment and Energy. October 1995. Coiner 1996. Persona] communication with Ron Coiner, National Sanitation Foundation. Committee 1987. Committee on the Evaluation of Carcinogenic Substances, National Health Council of the Netherlands, "A Scientific Basis for the Risk Assessment of Vinyl Chloride," Regul Toxicol Pharmacol. VoL 7, No. 1, pp. 120-127, March 1987. CPI 1991. CPI Profiles, July 1991. Received from Tedd Brien, Use Patterns Section, Commercial . Chemicals Evaluation Branch, Environment Canada. DoE 1992. Department of the Environment. "Vinyl Chloride Release Regulations, 1992," Extract Canada Gazette, Pan II. December 2,1992. Doll 1988. Doll, R. "Effects of Exposure to Vinyl Chloride. An Assessment of the Evidence," Scandinavian Journal of Work and Environmental Health. Vol. 14, No. 2, pp. 61-78, April 1988. Doniger 1978. Doniger, David D. The Law and Policy of Toxic Substances Control: A Case Study ofVmyl Chloride. Resources for the Future, 1978. D&B 1996. Dun &. Bradstreet - Canadian Duns Market Identification Dialog File, 1996. EC 1986. Environment Canada. Environmental Status Repon 1979-1984: Vinyl Chloride Industry. September 1986. Report EPS 1/AP/l. 2-25 SL 110252 EC 1988. Environment Canada. Environmental Status Report 1985*1986: Vinyl Chloride Industry. July 1988. Report EPS l/AP/2. EC 1992. Environment Canada. Environmental Status Report 1987-1990: Vinyl Chloride Industry. August 1992. Report EPS l/AP/4. Fisheries 1978. Fisheries and Environment Canada, Environmental Protection Service. Air Pollution Emissions and Control Technology: Vinyl Chloride Industry. March 1978. Econ mic and Technical Review Report EPS 3-AP-77-4. Fisheries 1977. Environment Canada 1977. Environment Canada, Environmental Protection Service. Emissions of Vinyl Chloride to the Ambient Air Around Manufacturing Facilities in Ontario. December 1977. Surveillance Report EPS 5-AP-77-14. Gutb 1996. Personal communication with Dan Guth, U.S. Environmental Protection Agency Infante 1976. Infante, Peter F,, Joseph K. Wagoner, Anthony J. McMichael, Richard J. Wasweiler, and Henry Falk. "Genetic Risks of Vinyl Chloride," The Lancet, pp. 734-5, April 3, 1976. Kuzmack 1975. Kuzmack, Arnold M. and Robert E. McGaugby. Quantitative Risk Assessment for Community Exposure to Vinyl Chloride. Prepared for the U.S. Environmental Protection Agency. December 5,1975. Lowenheim 1975. Lowenheim, Frederick A. and Marguerite K. Moran. Faith, Keyes. & Clark's Industrial Chemicals. Fourth Edition. New York: John Wiley & Sons, 1975. Martinello and Meng 1992. Martinello, F. and R. Meng. "Workplance Risks and the Value of Hazard Avoidance." Canadian Journal ofEconomics, Volume '25, pp. 333-345,1992. Meng 1989. Meng, R. "Compensating Differences in the Canadian Labour Market." Canadian Journal of Economics, Volume 22, pp. 413-424, 1989. Meng and Smith 1990. Meng, RA and DA. Smith The Valuation of Risk of Death in Public Sector Decision-Making." Canadian Public Polity, Volume 16, pp. 137-144,1990. NPRI 1993. National Pollutant Release Inventory, 1993, Canaria Podoll 1986. Podoll, T. "Impact of Leaching by Plastic Pipe, Fittings, and Joining Compounds," Plumbing Materials and.Drinking Water Quality. New Jersey: Noyes Publications, pp. 68-76, 1986. Saz 1981. Sax, N. Irving. Cancer Causing Chemicals. New York: Van Nostrand Reinhold. 1981. As dted in memorandum from Jim Neumann and Bob Unsworth, "Addenda to Mortality Valuation Methodology." September 28,1993. 2-26 SL 110253 Selleck 1991. SeUeck, RE. and BJ. Marinas. "Analyzing the Permeation f Organic Chemicals Through Plastic Pipes," Journal oftheAmerican Water Works Association, VoL 83, No. 7, pp. 92-97, July 1991. Theriault 1983. Theriault, G,, H. Inina, and S. Gingras. "Evaluation of the Association Between Birth Defects and Exposure to Ambient Vinyl Chloride," Teratology. VoL 27, No. 3, pp. 359-370,1983. Unsworth 1993. Unsworth, Robert E and James E Neumann. "Review of Existing Value of Morbidity Avoidance Estimates; Draft Valuation Document," memorandum prepared by . Industrial Economies, Inc. for the US. EPA, Office of Air and Radiation, Washington, D.C, September 1993. US. EPA 1975. US. Environmental Protection Agency. Scientific and TechnicalAssessment Report on Vinyl Chloride and Polyvinyl Chloride. December 1975. EPA/600/6-75AXM. US. EPA 1980. U.S. Environmental Protection Agency, Office of Water Regulations and Standards, Criteria and Standards Division. Ambient Water Quality Criteriafor Vinyl Chloride. Washington. DC. October 1980. EPA 440/5-80-078. US. EPA 1988. US. Environmental Protection Agency, Office of Health and Environmental Assessment. Evaluation of the Potential Cardnogeniatv of Vinyl Chloride (75-01-4) (Final Report). June 1988. EPA/600/8-91/199. US. EPA 1989. U.S. Environmental Protection Agency. Risk Assessment Guidance for Superfund Volume J: Human Health Evaluation Manual (Part A) (Interim Final). December 1989. EPA/540/1-89/002. US. EPA 1994a. US. Environmental Protection Agency, Office of Solid Waste and Emergency Response. Health Effects Assessment Summary Tables: Annual Update. March 31,1994. VS. EPA 1994b. US. Environmental Protection Agency, Air Risk Information Support Center (Air RISC). Health Effects Notebook for Hazardous Air Pollutants (Draft). December 1994. VC in Drinking Water 1992. Vinyl Chloride in Drinking Water Guideline for Canada (Draft). November 1992. Viscusi 1992, Fatal Tradeoffs: Public and Private Responsibilities for Risk. New York: Oxford University Press. 1992. Wagoner 1983. Wagoner, JJC "Toxicity of Vinyl Chloride and Poly (vinyl chloride); A Critical Review," Environmental Health Perspective. Vol. 52, pp. 61-66, October 1983. Wu 1989. Wu, W., R.S. Roberts, Y.G Chung, WJL. Ernst, S.C Havlieek. "Extraction of Organotin Compounds from Polyvinyl Chloride Pipe," Archives of Environmental Contamination and Toxicology, Vol. 18, No. 6, pp. 839-843, November 1989. 2-27 SL 110254 Vodden et aL 1994. Vodden, 1C, D. Smith, R. Mcng et aL The Social Cost ofMotor Vehicle Crashes in Ontario. Prepared for the Government of Ontario, Safety Research Office, Safety P licy Branch. 1994. 2-28 SL 110255 Appendix 2A GENERAL POPULATION EXPOSED TO VC EMISSIONS FROM POINT SOURCES Overview Applying ambient air concentration estimates developed in the 1975 EPA study requires that we estimate the number of individuals living proximate to VC and PVC facilities. Specifically, we need to estimate the number of individuals living within four bands located around each facility. The bands have the following distances: -> Band 1: 0 to 0.5 miles. Band 2: 0.5 to 1 miles. Band 3:1 to 3 miles. Band 4:3 to 5 miles. The purpose of this appendix is to describe the methodology we employ to estimate the size of the affected population within each band. We use 1991 Canadian census data on the total populati n and size of the four Canadian cities in which VC and PVC plants are located. Method We develop the estimates of the total affected populations (presented in Exhibit 2*5 of the body of the chapter) by summing affected populations across each of the four cities with VC or PVC manufacturing facilities. We employ the following method to estimate the affected population within each city. Step 1. Based on the area of each city, determine if the city has a radius of less than 5 miles (assuming the city is "round"). If so, determine the number of bands covered by the city (see Exhibit 2A-1 for an illustration for Samia, Ontario). Step 2. Calculate the area of each band within the border of each city. Step 3. Assuming a uniform distribution of population within each city, apply population density information for each city to estimate the number of individuals within each band. 2-29 SL 110256 Exhibit 2A-1 EXAMPLE OF GENERAL POPULATION SURROUNDING A VC OR PVC PLANT IN SARNIA, ONTARIO Ontario is 4.5 miles. This dotted line represents the outer boundary of the town. SL 110257 Implicit in this approach are the f Mowing assumptions: (1) the VC or PVC plant is located in the center of each city, and (2) the population of the areas immediately outside f each city is zero.2 Exhibit 2A-2 summarizes our results by city and population band. Exfai>it2A-2 ESTIMATED POPULATION BY CITY AND BAND Qty Sarnia, Ontario Niagara Falls, Ontario Shawinigan, Quebec Ft. Saskatchewan, Alberta 1 884 246 1,186 559 2 2453 739 3459 1478 Band 3 28301 7485 15,186 9442 4 42340 65450 0 0 Total 74378 73320 19,932 12,079 " While this second assumption is an over-simplification, it does not have a large impact on our results. 2-31 SL 110258 Appendix 2B ESTIMATING GENERAL POPULATION CANCER CASES: DETAILED EXAMPLE Overview The purpose of this appendix is to provide a detailed example of the calculations performed to estimate foe total annual cancer cases in foe general population living near VC and 1PVG plants in foe absence of controls on VC emissions. The following example estimates foe annual number of cases for foe 1,444 individuals estimated to live in Band 1 around VC plants. Step 1: Calculate the Ambient Air Concentration of Vinyl Chloride. From the U3. EPA study, we apply an ambient air concentration of 113 ppb to Band 1 (0 to 0.5 miles from a VC facility). We converted this concentration from ppb to mg/mJ: (113 ppb). * (1 pgm/1000 ,ppb) * 0.113 ppm (0.113 ppm) V^^gte^'/f^ppm) 0.294 mg/m3 Step 2: Calculate the Dose. To estimate foe dose to which each resident is potentially exposed, we multiplied foe ambient air concentration of VC by foe inhalation rate and foe inverse of foe average weight of an adult: * (0294 rog/m3) (20 mVday) (1/70 kg) 0.084 mg/kg-day Step 3: Calculate foe Cancer Risk. To estimate foe individual lifetime risk of cancer, we multiplied foe dose by the cancer slope factor for VC: (0.084 mg/kg-day) * (03 mg/kg-day)'1 = 23 x 10'J Step 4: Calculate foe Annual Cancer Cases, To estimate foe annual number of cancer cases in a given population, we multiplied foe cancer risk by the population and divided by the average adult lifc$p*w- (23 x 1C1) (1,444 people) / (70 years) = 032 people/year % % 1 Appendix 1C B 1 ESTIMATES USED TO GENERATE PROPOSED DISrRIDUTION OP VALUE OF LIFE | |' Author tmf Year | Kncknet and Uclh (1991) (US) Value of life Rest Estimate (J1995, millions) 0.8 Full Reference T.J. Kneisner and J.D. Lecth. 1991. 'Compensating Wage Differentials for Fatal Injury Risk In Australia, Japan, and the United Stales,' Journal of Rirk arid Uncenomly 4(1):75*90. 1 f1 | Smith and Gilherl (I9B4) I Dillingham (1985) 1 Duller (198.1) | Miller and Ourla (1991) | Moore and Viscusl (1988a) in V.K. Smith and C. Gilbert. 1984. The Impticll Risks In Ufe: A Comparative Analysis,' Eronomfer Leffrrr, 16: 393-399. 1.2 A. Dillingham. 1985. "The Influence of Risk Variable Definition on Value of Life Estimates,' Economic Inquiry 14: 277-294. 1.5 RJ. Duller. 1983. "Wage and Injury Rale Responses to Shifting Levels of Workers* Compensation,' In John D. Woerall, ed. Softly end the llbrfc Font. Ithaca: Cornell University. ILR Press. 1.6 T. Miller and J. Gurla. 1991. The Value of Statistical Ufe In New Zealand." Report to the New Zealand Ministry of Transport, Land Transport Division. J.4 MJ. Moore and W.K. Vtscoal. 1988. `Doubling the Estimated Value of Ufe: Results Using New Occupational Fatality Data,* Jorumot of Policy Anofysb and Management 7(3): 476-490. Viscini, Magat, and Huber (1991b) 3.1 W.K. Vlseusl, W.A, Magat, and J. Hatter. 1991. 'Pricing Environmental Health Risks: Survey Assessments of RiskRisk and Rhk-Dottsr Tradeoffs,* Joumot of fjidnaitmul Economkt and Monogemenl 201:32-57. Gegax cl at. (1985) * B Marir. nod Psacharnpoufos (1981) | Knelrner and tjeelh (1991) (Australia) | Getting. de Ifaan, and Schulte (1988) B Couiineau. Lacroix, and Oltard (1988) 9.5 No full reference available 3.8 A. Morin and G. Psacharopoukts, 1982. The Reward for Risk In the Labor Market: Evidence from the United Kingdom end a Reconciliation with Other Studies,* Journal of Politico! Economy 90(4): 827-853. 4 5 TJ. Kneisner and J.D. Leeth. 1991. Xnmpemating Wage Differentlafs for Palal Injury Risk In Australia, Japan, end the United Stales* Joumot of Risk ond Urtccuointy 4(t):75-90. 4.6 S. Gerfclng, M. de ffean, and W. Schufre. 1988. The Marginal Value of Job Siftly: A Contingent Valuation Study,* Journal of Rid and Unccrtointy 1(1): 185-200. 4.9 J. Couslneau, R. Lacroix, and A, Girard. 1988, "OeeupMkmal Hatard and Wage Compensating Differentlats,* University of Montreal Working Paper. Jones-Lee (19B9) Dillingham (1985) CD 5.1 M.W. Jones-Lee. 1989. The Economict of Softly and Phytical Risk, Oxford: Basil Blackwell. 5J A. Dillingham. 1985. The InRuenec of Risk Variable Definition on Value of Life Estimates * Economic Inquiry If: 277-294 L 110260 2-33 Appendix 2C (continued) 1 Author and Yew 1 Viscusi (78. 79) R.S. Smith (1976) V.K. Smith (1976)* 1 Olson (1981) 1 Vljtiul (1981) ESTIMATES USED TO GENERATE PROPOSED DISTRIOUTION or VALUE OP LIFE Value of Ufe Beal Estimate (SI99S, millIons) Pull Reference | 5.6 W.K. Viscusi. 1978. "Labor Market Vahtalkms nf Ufe and l.inth: Emphlesl Eslimalet and FoHcy Impttcnlirms,' ftthfic Polity 26(3): 359-386. ----- --. 1978.'Wealth Efiecls and Earnings Prenffnms for Joh Hatards* Ret-few of Economics amt Statistics 60(3): 4118-416. --------- 1979. Employment Hetnnis: An inicuiyntum of Mattel Performance, Cambridge: Harvard University Press. --------- 1979, 'Job Hazards and Worker Ouil Rales: An Analysis of Adaptive Worker Behavior* inttmatkmad Eeommtir Review 20(1): 29-58. ------ 1979. "The Impact of Occupational Safety and Iterdlh Regulation.*. Beff Jotrmnf of Economies 10(1): 117140. 6.3 R.S. Smith. 1976. The Occupaitonat Safety ami ileaith Act: In (ioats amt /fcftfrtvmmfg Washington: American Enterprise Institute. 6.4 No full reference available 7.1 CA. Olson. 1981. 'An Analysis,of Wage Differential Received by Worker* nn Dangerous Jobs' Journal of Hitman Resources 16: 167-185. 8.9 W.K. Viscusi. 1981. *Occupathmal Safety and Health Regulation: In Impact and Policy Alternatives,' J. Credne, ed., Research in ftiMf Tutky Analysis ami Ifmapatog vut 2, Greenwich, Cornu JAI Press, pp. 281-299. R.S. Smith (1974) | Moore and Vhctni (I988i) | Knehncr and leeth (1991) (Japan) 9.8 R.S. Smith. 1974. 'The FeasibiHly of an 'Injury Tas* Approach to Occupational Safety* Law ami Contemporary Problems 38(4): 730-744. 9.9 M.J. Moore and W.K. Viscusi. 1988. *00*110111 the Estimated Value of Ufe: Remrflt Using New Occupational Fatality Data," Journal of PoRcy Analysis ami Management 7(3): 476-490. 10.4 TJ. Knctmcr and J.D. Lecth. 1991. "Compensating Wage Differentials for Fatal Injury Risk In Australia, Japan, and the United Stale*,* Journal of Risk ami Uncertainty 4(l):75-90. I Hereof and ScMollman (1987) 12.4 H.W. Hereof, Jr. and A.M. Schlollntan. 1987. "Valuing Risk bt the Workplace: Market Price, Wlittngnets t Pay, and the Optimal Provision of Safety* University of Tennessee Working Paper. 3 Leigh and Poison (1984) 13.2 J.P. Leigh and R.N. Poison. 1984. 'Etlhnalei of the Value of Accident Avoidance at the Job Depend on Concavity of the EtptaNflng Differences Curve,* The Quarterly Review of Economics and Business 24(1): 56-66. 1 Leigh (1987) 14.2 J.P. Leigh. 1987. `Gender, Firm Size, Industry and Estimates of the Value-of-Ufe* Jotmtaf of Ifrahti Economics 6: 255-273. I Onren (1988) 18.4 J. Gwen. 1988. "Compensating Wage Differentials and the Endogeneity of Job Riskiness,* The Review of Economics and Statistics T0{ty. 9-lb. 1 Source: W.K. Vlscosl. 1992. Fotai Tradeoffs: PUWir and frtvatt Responsibilities for fork. New York: Oxford Unlvcriily Press. | * Added after consultation wHh Vtscnsl. SL 110261 DRAFT TOXICOLOGICAL PROFILE FOR VINYL CHLORIDE Prepared by: Sciences International, Inc. Under Subcontract to: Research Triangle Institute Under Contract No. 205-93-0606 Prepared for U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES Public Health Service Agency for Toxic Substances and Disease Registry August 1995 "DRAFT FOR PUBLIC COMMENT" SL 110262 4. PRODUCTION, IMPORT, USE, AND DISPOSAL 4.1 PRODUCTION Production of vinyl chloride monomer in the United States was approximately 13.75 billion pounds in 1993 (C&EN 1994; USITC 1994), an increase of nearly 22%. This exceeds the latest available U.S. estimated total production capacity figures (beginning of 1993) of 12.79 billion pounds at the beginning of 1993 (SRI 1993) and represents a substantial recent increase in production and capacity, even if vinyl chloride production is at full capacity. Previously, vinyl chloride production had grown only 1-5% per year over the last 5 years (C&EN 1994). Over the last 10 years, vinyl chloride production has grown at an average rate of about 7%. not including the large increase in 1993, although production volumes were fairly volatile (C&EN 1994). Vinyl chloride is currently produced in the United States by 10 companies at 12 facilities, which are as follows (SRI 1994); Westlake Monomers Corporation in Calvert City. Kentucky; Borden Chemicals and Plastics in Geismar, Louisiana; Dow Chemical in Oyster Creek. Texas, and in Plaquemine, Louisiana; Georgia Gulf Corporation in Plaquemine, Louisiana; PPG Industries in Lake Charles. Louisiana; Vista Chemical Company in Lake Charles, Louisiana; B.F. Goodrich Company in LaPorte, Texas; Formosa Plastics Corporation in Baton Rouge, Louisiana, and in Point Comfort, Texas; Occidental Chemical Corporation in Deer Park, Texas; and Oxymar in Ingleside, Texas. Table 4-1 summarizes the facilities in the United States that manufacture or process vinyl chloride. This information was obtained from the 1992 Toxic Chemical Release Inventory (TRI92), and it summarizes the reported release data for 1992 (TRI92 1994). Table 4-1 also lists the maximum amounts of vinyl chloride that are present at these sites and the end uses of the vinyl chloride. Vinyl chloride was first produced commercially in the 1930s by reacting hydrogen chloride with acetylene. Currently, vinyl chloride is produced commercially by the chlorination of ethylene through one of two processes, direct chlorination or oxychlorination. Direct chlorination reacts ethylene with chlorine to produce 1,2-dichloroethane. Similarly, oxychlorination produces 1.2-dichloroethane, but this is accomplished by reacting ethylene with dry hydrogen chloride and oxygen. After both processes, the 1.2-dichloroethane is subjected to high pressures (2.5-3.0 megapascals) and temperatures '"DRAFT FOR PUBLIC COMMENT"' SL 110263 . PRODUCTION. IMPORT, USE. AND DISPOSAL DRAFT FOR PUBLIC COMMENT-- Table 4-1. Facilities That Manufacture or Process Vinyl Chloride Facility Location* Range of maximum amounts on site in pounds Activities and uses KEYSOR-CENTURY CORP. UNION CARBIDE CORP. FORMOSA PLASTICS CORP. GEORGIA GULF CORP. WESTLAKE PVC CORP. UNION CARBIDE CHEMICALS BF GOODRICH CO. BORDEN CHEMICALS & PLASTICS BF GOODRICH CO. AIR PRODS. & CHEMICALS INC. WESTLAKE PVC CORP. WESTLAKE MONOMERS CORP. VULCAN MATERIALS CO. BORDEN CHEMICALS & PLASTICS PPG IND. INC. SAUGUS, CA TORRANCE, CA DELAWARE CITY, DELAWARE CITY, PACE, FL TUCKER, GA HENRY, IL DE DE 1,000,000-9,999,999 10,000-99,999 1,000,000-9,999,999 1,000,000-9,999,999 1,000,000-9,999,999 100,000-999,999 1,000,000-9,999,999 ILLIOPOLIS, IL LOUISVILLE, KY CALVERT CITY, KY CALVERT CITY, KY CALVERT CITY, KY GEISMAR, LA No Data 1,000,000-9,999,999 10,000-99,999 1.000.000-9,999,999 1.000.000-9,999,999 100,000-999,999 GEISMAR, LA No Data LAKE CHARLES, LA No Data CERTAINTEED CORP. VISTA CHEMICAL CO. FORMOSA PLASTICS CORP. SULPHUR, LA WESTLAKE, LA BATON ROUGE, LA 1,000,000-9,999,999 No Data 1,000,000-9,999,999 RHONE POULENC BASIC CHEMICALS BATON ROUGE, LA DOW CHEMICAL CO. PLAQUEMINE, LA 10,000-99,999 No Data As a reactant As a reactant As a reactant As a reactant As a reactant As a reactant As a reactant; As a product component As a reactant As a reactant As a reactant As a reactant Produce; For sale/distribution Produce; For on-siteuse/processing; As a by-product; As a reactant Produce; For on-site use/processing; For sale/ distribution; As a reactant Produce; For sale/distribution; As a by-product; As an impurity; As a reactant As a reactant Produce; For sale/distribution Produce; For on-site use/processing; For sale/distribution; As a by-product; As a reactant Ancillary uses Produce; For sale/distribution; as a reactant; As a manufacturing aid; Ancillary uses Cfi f HO NJ (Ti 4 i CO t* o tsj U1 . use. AND DISPOSAL DRAFT FOR PUBLIC-COMMENT-- Facility Table 4-1. Facilities That Manufacture or Process Vinyl Chloride (continued) Location* Range of maximum amounts on site in pounds Activities and uses BF GOODRICH CO. GEORGIA GULF CORP. PLAQUEMINE, LA PLAQUEMINE, LA 100,000-999,999 No Data UNION CARBIDE CORP. MARINE SHALE PROCESSORS INC. OCCIDENTAL CHEMICAL CORP. DOW CHEMICAL USA MILES INC. VISTA CHEMICAL CO. OCCIDENTAL CHEMICAL CORP. BF GOODRICH CO. GOODYEAR TIRE & RUBBER CO. VYGEN CORP. BF GOODRICH AVON LAKE VISTA CHEMICAL CO. OCCIDENTAL CHEMICAL CORP. GREENWOOD PLATING DOW CHEMICAL CO. HAHNVILLE, LA AMELIA, LA ADDIS, LA MIDLAND, MI KANSAS CITY, MO ABERDEEN, MS BURLINGTON, NJ PEDRICKTOWN, NJ NIAGARA FALLS, NY ASHTABULA, OH AVON LAKE, OH OKLAHOMA CITY, OK POTTSTOWN, PA GREENWOOD, SC FREEPORT, TX No Data 1,000-9,999 100,000-999,999 100,000-999,999 100,000-999,999 No Data 1,000,000-9,999,999 1,000,000-9,999,999 100,000-999,999 100,000-999,999 1,000,000-9,999,999 1,000,000-9,999,999 1,000,000-9,999,999 1,000,000-9,999,999 No Data SHINTECH INC. FORMOSA PLASTICS CORP. FREEPORT, TX POINT COMFORT, TX No Data No Data UNION CARBIDE CHEMICALS & UNION CARBIDE CHEMICALS OCCIDENTAL CHEMICAL CORP. RHONE-POULENC BASIC CHEMICALS GARLAND, TX TEXAS CITY, TX LA PORTE, TX HOUSTON, TX 100,000-999,999 1,000,000-9,999,999 1,000,000-9,999,999 10,000-99,999 As a reactant Produce; For on-site use/processing; For sale/distribution; As a reactant As a by-product As a reactant As a reactant As a reactant; Ancillary uses As a reactant As a reactant As a reactant As a reactant As a reactant As a reactant As a reactant As a reactant As a reactant As a product component Produce; As a by-product; As an inqpurity; As a reactant; As a chemical processing aid Import; For on-site use/processing; As a reactant Produce; For on-site use/processing; As a by-product; As a reactant As a reactant As a reactant Ancillary uses Ancillary uses Table 4-1. Facilities That Manufacture or Process Vinyl Chloride (continued) Facility OCCIDENTAL CHEMICAL CORP. BF GOODRICH CO. OCCIDENTAL CHEMICAL CORP. GEON CO. OCCIDENTAL CHEMICAL CORP. BONDCOTE CORP. i Location* PASADENA, TX DEER PARK, TX DEER PARK, TX LA PORTE, TX GREGORY, TX PULASKI, VA Range of maximum amounts on site in pounds No Data 1,000,000-9,999,999 No Data No Data No Data 100,000-999,999 Activities and uses As a reactant As a reactant Produce; For sale/distribution Produce; For sale/distribution Produce; For sale/distribution As a formulation component Source: TRI92 1994 * Post office state abbreviation used >. p r o d u c t io n , im p o r t , u s e , a n d d is p o s a l DRAFT FOR PUBLIC COMMENT!-- CD IT* CMHOHl cn 4. PRODUCTION. IMPORT. USE. AND DISPOSAL (550-550C). This causes the 1,2-dichloroethane to undergo pyrolysis, or thermal cracking, which forms the vinyl chloride monomer and hydrogen chloride. The vinyl chloride monomer is then isolated (Cowfer and Magistro 1985). The technical grade product is available in 99.9% purity (HSDB 1994). Efforts are being made to minimize by-product formation in 1,2-dichloroethane pyrolysis (Cowfer and Magistro 1985). 4.2 IMPORT/EXPORT Imports of vinyl chloride totaled 119 million pounds in 1992 and 164 million pounds in 1991 (CPS 1993). Imports have been steadily declining from a high of 302 million pounds in 1989. prior to which they had been increasing (CPS 1993). Exports of vinyl chloride were 1.63 billion pounds in 1992. down from the all time high of 1.78 billion pounds in 1991. Over the past 20 years, expons of vinyl chloride have fluctuated fairly widely, but have been generally increasing an average of about 17% per year since 1975. 4.3 USE Vinyl chloride is an imponant industrial chemical because of its wide variety of end-use products and the low cost of producing polymers from it. Furthermore, polyvinyl chloride (PVC) is one of the most efficient construction materials available when analyzed on an energy-equivalent basis (Cowfer and Magistro 1985). Major end-use products include PVC products, such as automotive pans and accessories, furniture, packaging materials, pipes, wall coverings, and wire coatings, as well as vinyl chloride-vinyl acetate copolymer products, such as films and resins (Cowfer and Magistro 1985; Eveleth et al. 1990). End-use data for 1992 indicate that 98% of vinyl chloride monomer production is for making polyvinyl chloride and various polyvinyl chloride copolymers; the other 2% is for miscellaneous uses (CPS 1993). Vinyl chloride has been used in the past as a refrigerant, as an extraction solvent for heat-sensitive materials, and in the production of chloroacetaldehyde and methyl chloroform (IARC 1979). In the United States, limited quantities of vinyl chloride were used as an aerosol propellant and as an ingredient of drug and cosmetic products; however, these practices were banned by the EPA in 1974 (IARC 1979; HSDB 1994). "'DRAFT FOR PUBLIC COMMENT*" SL 110267 4.4 DISPOSAL 4. PRODUCTION, IMPORT, USE. AND DISPOSAL Since vinyl chloride has been identified by EPA as a hazardous material, its disposal is regulated under the Federal Resource Conservation and Recovery Act (RCRA) (EPA 1993d). The transportation of hazardous materials for disposal is regulated by the Department of Transportation in compliance with this act (DOT 1993). The recommended method of disposal is total destruction by incineration. The temperature of the incinerator must be sufficient to ensure the complete combustion of the vinyl chloride in order to prevent the formation of phosgene. The recommended temperature range for incineration is 450-1,600C, with residence times of seconds for gases and liquids, and hours for solids (HSDB 1994). If in solution, the vinyl chloride product may need to be adsorbed onto a combustible material prior to incineration. Recommended materials include vermiculite. sawdust, or a sand-soda ash mixture (90/10) covered with wood and paper (OHM/TADS 1985). The vinyl chloride can also be dissolved in a flammable solvent prior to incineration. An acid scrubber should be used in conjunction with the incinerator in order to remove any hydrogen chloride that is produced by the combustion process (HSDB 1994; OHM/TADS 1985). Alternatively, chemical destruction may be used, especially with small quantities. One-to-two days is generally sufficient for complete destruction (HSDB 1994). Aqueous by-product solutions from the production of vinyl chloride are usually steam-stripped to remove volatile organic compounds, neutralized, and then treated in an activated, sludge system to remove nonvolatile organic compounds remaining in the waste water (Cowfer and Magistro 1983). "DRAFT FOR PUBLIC COMMENT"* SL 110268