COMMENTS ON THE ADVANCE NOTICE OF PROPOSED RULEMAKING "OZONE - DEPLETING CHLOROFLUOROCARBONS PROPOSED PRODUCTION RESTRICTION" BY THE ENVIRONMENTAL PROTECTION AGENCY (FEDERAL REGISTER; VOL. 45, NO. 196 - TUESDAY, OCTOBER 7, 1980, PAGES 66726-66734; EPA [OPTS-62009 (TSW-FRL 1606-5)] SUBMITTED BY E. I. DU PONT DE NEMOURS & COMPANY (INC.) WILMINGTON, DELAWARE JANUARY 5, 1981 ------- PREFACE On October 1, 1980, the United States Environmental Protection Agency (EPA) published an Advance Notice of Proposed Rulemaking (ANPR) titled, "Ozone - Depleting Chlorofluorocarbons: Proposed Production Restriction" (45 Federal Register 66726- 66734). EPA requested submission of comments. E. I. du Pont de Nemours & Company (Du Pont) manufactures and markets chloro- fluorocarbons (CFCs) under the trademark "Freon". Du Pont herein presents comments for the record on: Inappropriateness or use of CFCs at of restricting this time. domestic production Infeasibility and impact of. implementing CFC controls as proposed. Questions or requests for additional clarification may be addressed to: information or Paul W. Halter Environmental Manager "Freon" Products Division E. I. du Pont de Nemours & Company Wilmington, Delaware 19898 (302/774-6484) ------- Table of Contents and Conclusions PREFACE EXECUTIVE SUMMARY A. Problem Background B. Du Pont Assessment C. Du Pont Position D. Problems with EPA's Approach to Issue E. Recommendations for Action by EPA I. INTRODUCTION II. CHLOROFLUOROCARBON USES AND ESSENTIALITY A. Introduction B. Refrigeration and Air-Conditioning 1. Description of Use 2. Essentiality and Benefits 3. Alternatives and Limitations C. Mobile Air-Conditioning 1. Description of Use 2. Essentiality and Benefits 3. Alternatives and Limitations D. Solvents 1. Description of U.se 2. Essentiality and Benefits 3. Alternatives and Limitations E. Blowing Agent for Rigid Polyurethane Foam 1. Description of Use 2. Essentiality and Benefits 3. Alternatives and Limitations F. Blowing Agent for Flexible Polyurethane Foam 1. Description of Use 2. Essentiality and Benefits 3. Alternatives and Limitations ES-1-37 3 8 23 27 31 1-1-9 II-1-45 2 4 4 4 5 7 7 B 10 10 11 12 13 13 15 16 17 17 18 19 11 ------- Table of Contents Paqe II. CHLOROFLUOROCARBON USES AND ESSENTIALITY (CON'T) G. Blowing Agent for Polystyrene, Polyethylene and Phenolic Foam 21 1. Description of Use 21 2. Essentiality and Benefits 22 3. Alternatives and Limitations 23 H. Food Freezant 24 1. Description of Use 24 2. Essentiality and Benefits 24 3. Alternatives and Limitations 25 I. Sterilant Gas 28 1. Description of Use 28 2. Essentiality and Benefits 29 3. Alternatives and Limitations 30 J. Intermediate for Fluoropolymer Production 32 1. Description of Use 32 2. Essentiality and Benefits 32 3. Alternatives and Limitations 34 K. Chlorofluorocarbons and Energy Conservation 36 1. Summary 36 2. General Conclusions and Methodology 39 3. Specific Applications 41 4. U.S. Department of Energy Standards 42 L. Summary 44 III. LEGAL ISSUES III-1-45 A. Introduction 2 B. Authority to Regulate 4 1. Findings to Support Regulation May Not Be Made 4 2. International Concerns 10 3. Scope of Proposed Regulation 14 4. Regulatory Focus 16 111 ------- Table of Contents Paqe III. LEGAL ISSUES (CON'T) C. Economic Disincentives Regulation 17 1. Authority 19 2. Implementation and Operation 22 a. Direct Allocation 23 b. Auction 24 3. Competitive Impacts 26 D. Rulemaking Procedures 29 1. General-Clean Air Act 29 2. General-Toxic Substances Control Act 30 3. Research 33 4. Economic and Regulatory Impact Analyses 38 a. Clean Air Act 38 b. Toxic Substances Control Act 41 c. Executive Order 12044 42 d. Regulatory Flexibility Act 43 E. Conclusion 45 IV. THE SCIENCE IV-1-63 A. Introduction 2 B. The Chlorofluorocarbon/Ozone Depletion Theory 7 1. General Description of the Theory 7 2. Model Calculations - What They Are; Why They Are Needed 9 3. Previous Scientific Assessments of the Theory 10 C. Ozone Measurements and Ozone Trend Analysis 15 1. Overview 15 2. Detail 17 D. EPA's ANPR Assessment of the Theory 22 E. Present Status of the Theory 32 1. Production and Release of CFCs 32 2. Lower Atmospheric (Tropospheric) Processes 32 IV ------- Table of Contents P1? v-^"' """"" Page i If'SO - 4 til ST., S1" ,-; 7.Yi WASHIWGTO-j. ' IV. SCIENCE (CON'T) a. CFC-11 and CFC-12 Lifetimes 33 b. CFC-21 34 c. CFC-22 36 3. Transport to the Upper Atmosphere 36 4. Chemistry in the Lower and Upper Atmosphere 37 a. Hydroxyl Radical Reactions 39 b. Pressure and Temperature Dependencies 41 c. Alternative Reaction Products 42 d. Chlorine Nitrate 43 5. Atmospheric Models 44 a. 2-D Calculations 45 b. C02/N20 Effects 46 c. Volcanoes 47 6. Stratospheric Measurements 48 a. Chlorine Species 48 b. Nitrogen Species 50 F. Resolution of Uncertainties 51 1. Atmospheric Measurements 51 2. Modeling 54 3. Chemistry 55 G. Summary 56 V. THE QUESTION OF RISK V-l-64 A. Introduction 2 B. Impact of Uncertainties in the Under- lying Science of the Theory of Ozone Depletion on Risk 5 1. Introduction 5 2. Major Current Uncertainty Sources in the Atmospheric Science . 6 3. Errors Made by EPA In Treatment of Uncertainties 9 a. EPA Over Relies on the "Key Findings" of the NAS Report 9 ------- Table of Contents Paqe V. THE QUESTION OF RISK (CON'T) b. EPA Places Sole Reliance on the NAS Report 10 c. EPA Does Not Acknowledge Conflict Between the NAS Report and More Recent Reports 11 d. EPA Relies on an Out-of-Date Report, While Ignoring Recent Critical Develop- ments In the Science 13 C. Impact of Uncertainties in the Potential Effects of Ozone Depletion on Risk 14 1. Introduction 14 2. Human Skin Cancer Effects 18 a. Melanoma Skin Cancer 18 b. Nonmelanoma Skin Cancer 19 3. Natural Variations in Normal Background Radiation, Its Simulation and Its Measurement 21 4. Crop Effects 22 5. Marine Effects 24 6. Climatological Effects 25 D. Probability and Timing of Reducing Uncertainties 27 1. Introduction 27 2. Du Pont/Fluorocarbon Project Panel Estimates 27 3. SRI Workshop Conclusions 28 4. Conclusion 30 E. Risk In Waiting - Risk Versus Time 32 1. Introduction 32 2. Conclusions from 1980 Du Pont Submission 34 3. Conclusions from University of Maryland Study 35 4. Conclusions from Systems Control, Inc. Study ' 37 5. Summary 37 F. Is the Risk Developing As Predicted? 39 1. Introduction 39 2. Opposing Trends 40 3. Reliability of the Theory 41 VI ------- Table of Contents Page V. THE QUESTION OF RISK (CON'T) G. Availability and Significance of an Early Warning System 43 H. The Relationship of the International Aspects of the Issue to Risk 46 I. Risk Created by Regulation and the Need for Risk - Risk Comparison 48 J. Approaches Which Are Inappropriate for Assessment of Risk on the Chlorolfuorocarbon/ Ozone Issue 53 1. Preoccupation with Extreme Future Extrapolation 54 2. Conviction that Immediate Decisions Are Necessarily Required, and Are Necessarily Better than Deferred Decisions 57 3. Excessive Emphasis on Political Action Over Objective Scientific Decision-Making 59 K. Summary and Conclusions 60 VI. INTERNATIONAL ASPECTS VI-1-37 A. Introduction 2 B. Differences In National Approaches to the Issue 4 C. Illogic and Limitations of U.S. Unilateral Response 7 1. Proposed U.S. Cap on Production Will Have Inconsequential Direct Environmental Impact 7 2. Why U.S. Production Cap Will not Result in EPA's Goal of Worldwide Regulatory Action 9 D. Consequences of U.S. Unilateral Response 13 1. Potential for Counter-Productive Results 13 2. Imbalance Between Costs and Potential Environmental Benefit 14 3. Loss of Political Option 15 E. Need for a True Global Assessment, Consensus and Resolution of Issue 17 F. The Leadership Role - Suggestions on How to Proceed 20 VII ------- Table of Contents Page VI. INTERNATIONAL ASPECTS (CON'T) G. International Trade Implications of Proposed Controls 23 1. Exports 23 a. Inclusion of Exports Under Domestic Production Cap Would Eliminate Exports 23 b. Restriction of Exports Would Have No Net Environmental Benefit 24 c. EPA's Defense of Proposed Policy to Include Exports Under a Domestic Production Cap is Weak 25 d. EPA Expresses More Concern for Foreign Exporters to the U.S. Than for U.S. Exporters to Other Countries 28 2. Imports 29 a. Imports Should Be Treated"the Same as Exports 29 b. If U.S. Production Is Capped, Imports Should Be Capped Separately and On the Same Basis 30 c. Taxing of Imported Finished Goods Made With CFCs 31 d. Potential for Illegal Imports Has Not Been Addressed 32 H. Summary 34 VII. ECONOMIC CONSIDERATIONS VII-1-71 A. Introduction 2 B. Economic Significance of Chlorofluorocarbons 4 C. Regulation of Chlorofluorocarbons via Economic Incentives 10 1. Introduction 10 2. Impact on CFC Prices 12 3. Inflation 19 4. Economic Growth 20 5. Employment 21 6. CFC Substitutes 22 7. Energy 29 8. Financial Markets 30 9. Impact of Uncertainty 31 Vlll ------- Table of Contents VII. ECONOMIC CONSIDERATIONS (CON'T) D. Regulation of Chlorofluorocarbons via Command and Control 35 E. Inadequacy of the Rand Report to Support a Regulatory Decision 37 1. Introduction 37 2. Data Base 40 3. Study Assumptions 43 a. No Regulatory Restriction of Alternatives 43 b. Discount Rates 43 c. Time Delay of Emissions 44 d. Transfer Payments Not Inflationary 44 e. Scope of Economic Incentives Regulatory Options 44 f. Legal Issues 45 4. Limitations on Use of Report Findings 46 a. Introduction 46 b. Analytical Conclusions Must Be Extrapolated With Care 47 c. Regulatory Cost Remains Uncertain 49 d. Criteria for Benchmark Controls Too Restrictive 49 i. Enforceability 50 ii. Adequacy of Information 50 iii. Immediacy of Emission Reduction 51 e. Designs of Economic Incentives Options Are Too General 52 f. Transfer Payment Concerns Are Not Resolved 53 g. Inadequate Attention Is Given to Market Structure Effects of Regulatory Option Design 55 5. Needed Further Work 55 a. Consider Mixed Regulatory Options 55 b. Risk Trade-Off Analysis Needed 56 c. Develop Alternative Approach to Policy Evaluation 56 IX ------- Table of Contents Page VII. ECONOMIC CONSIDERATIONS (CON'T) d. Broaden Analysis 57 i. Time-Frame 57 ii. Technical Assessments 58 iii. Design 58 iv. Option Implementation and Administration 59 v. Legal Issues 60 e. Expand and Add Detail to Economic Incentives Option Structures 60 F. Miscellaneous ANPR Points Having Economic Implications 61 1. Choosing a Regulatory Strategy 61 2. Cost/Benefit Analysis 61 3. EPA's Long-Term Regulatory Strategy 63 4. Product/End Use Bans 63 5. Economic Incentives or Disincentives? 64 6. Tax or Surcharge on CFC Use 64 7. Base Year 64 8. Term of Permits 65 9. Direct Allocation of Permits to Manufacturers 66 10. Direct Allocation of Permits to Users 66 11. Government Auction of Permits 67 G. Summary 69 VIII. THE SEARCH FOR ALTERNATIVES VIII-1-7 A. Introduction 2 B. Criteria for Alternatives 2 C. Scope of Program 4 D. Program Status and Plans 5 E. Timetable 5 F. Summary 7 ------- Table of Contents Paqe IX. CONCLUSIONS AND RECOMMENDATIONS IX-1-18 A. Conclusions 2 B. Recommendations 13 X. APPENDICES A. Description of Major Pertinent Reports and Submissions on the Chlorofluorocarbon/Ozone Issue A-l-13 B. The Du Pont Development Program on the Alternatives to Commerical Chlorofluorocarbons B-l-15 C. The Energy Consequences of Chlorofluoro- carbon Regulation (Battelle Report) C-l-28 D. A Comparison of Some of the Principal Findings of the November 1979 National Academy of Sciences' Report and the October 1979 United Kingdom Department of the Environment's Report D-l-5 E. Chlorofluorocarbons and Ozone - The Science E-l-69 F. Effects of Ozone Depletion 1. Human Skin Cancer; Review by Professor Frederick Urbach, M.D. F-l (1-199) 2. Measurement and Instrumentation; Review by Dr. William H. Klein. F-2(l-19) 3. Agricultural Crops; Review by Professor R. Hilton Biggs. F-3(l-13) 4. Aquatic Ecosystems; Review by Dr. David M. Damkaer. F-4(l-36) G. Ranking Compounds by Potential for Ozone Depletion - "Permit Pounds" G-l-9 H. Scope of Proposed Regulation H-l-13 I. Economic Incentives Regulatory Options 1-1-39 J. Chlorofluorocarbon Production and Emissions J-l-16 XI ------- Table of Contents Page X. APPENDICES (CON'T) K. Industry Funded Fluorocarbon Research Program - Effect of Chlorofluorocarbons on the Atmosphere (CMA/FPP) K-l-77 L. Uncertainties - Chlorofluorocarbon Effects and Stratospheric Ozone (SRI Report) L-l-8 XI. BIBLIOGRAPHY XI-1-47 NOTE: The Executive Summary and Sections I-IX appear in Volume 1. Section X (Appendices A-F) appear in Volume 2. Section X (Appendices G-L) and Section XI-Bibliography appear in Volume 3. xn ------- EXECUTIVE SUMMARY COMMENTS ON THE ADVANCE NOTICE OF PROPOSED RULEMAKING "OZONE-DEPLETING CHLOROFLUOROCARBONS: PROPOSED PRODUCTION RESTRICTION" BY THE ENVIRONMENTAL PROTECTION AGENCY SUBMITTED BY E. I. DU PONT DE NEMOURS & COMPANY. (INC.) WILMINGTON, DELAWARE JANUARY 5, 1981 ES-1 ------- EXECUTIVE SUMMARY Table of Contents Paqe A. PROBLEM BACKGROUND 3 B. DU PONT ASSESSMENT AND CONCLUSIONS 8 C. DU PONT POSITION 23 D. PROBLEMS WITH EPA'S APPROACH TO ISSUE 27 E. RECOMMENDATIONS FOR ACTION BY EPA 31 ES-2 ------- Executive Summary Background EXECUTIVE SUMMARY A. PROBLEM BACKGROUND Uses - Chlorof1uorocarbons or CFCs, also commonly known as fluorocarbons, are used worldwide because of their safety, energy efficiency, high stability and performance attributes. Uses include: Commercial and residential refrigeration and air-conditioning. Automotive air-conditioning. Expanding agents used to manufacture plastic foams, including thermal insulating foams. Cleaning agents for precision electronic and electrical equipment and also military hardware. Fireproofing of sterilizing gas for hospital and industrial use. Freezing of food. Intermediate for fluoropolymer production. Aerosol propellants (although not in the United States except for a few specific exceptions). ES-3 ------- Executive Summary - Background CFCs are emitted to the atmosphere at the point of their use or during the lifetime of products which contain them. Their inherent stability, so necessary in these uses, also means that CFCs do not contribute to photochemical oxidant levels in smog, a major national concern with most volatile organic compounds. Theory - In 1974, scientists theorized that because of this stability in the lower atmosphere (troposphere), essentially all CFCs emitted eventually diffused unreacted into the upper atmosphere (stratosphere). In the stratosphere, the CFC molecules would be subjected to high energy radiation from the sun and dissociate, liberating chlorine atoms from the molecules. A compound known to occur naturally in trace amounts in the stratosphere is ozone (CO , a form of oxygen. Ozone is generated in the stratosphere from the interaction of sunlight and oxygen. This ozone serves the important function of limiting the amount of solar ultra-violet (UV) light which penetrates to the earth's surface. Several natural processes are thought to convert ozone back to oxygen. Chlorine has been suggested as a contributor to one of these processes. The theoretical concern over CFCs is that chlorine liberated from the photodissociation of CFCs may add to the natural chlorine in the stratosphere, and hence result in a lowering or depletion of the natural balance of ozone, thus allowing more UV to penetrate to ground level. The underlying concern is that a large increase in UV could increase the incidence of skin cancer, affect the produc- tivity of crops and marine life and, conceiveably, alter the climate. ES-4 ------- Executive Summary - Background It is important to note that these changes were theorized to occur gradually over a 70-100 year period, presuming that CFC emissions continued at a constant rate worldwide. Industry Position - From the theory's inception, industry has maintained: The theory warrants serious concern and should be investigated, Scientific measurements and evaluations -- not hypothesis should decide the issue, Experimental evidence can be obtained to quantitatively verify or disprove the theory, and There is time to perform these necessary experiments without undue risk to the health and welfare of society or the world's ecosystems. To these ends, industry launched a major research effort to prove or disprove the theory. Other research efforts were initiated by various government bodies. Global Realities - Also from the inception it has been clear that due to the widespread global use of CFCs, any effective solution to the theorized problem would have to be premised on two global political realities: The potential problem of future depletion of ozone by CFC emissions is global in nature, requiring global assessment and coordinated action, and ES-5 ------- Executive Summary - Background Such global action as may be appropriate will not occur until and unless there is a proper global resolution of the status of the underlying science, the quantitative validity of the theory, and the related risks. EPA Action - In response to the theory, the United States Environmental Protection Agency (EPA), in conjunction with the Consumer Product Safety Commission (CPSC) and the Food and Drug Administration (FDA), essentially banned in 1978 all domestic use of CFCs as aerosol propellants. Prior to the initiation of this regulatory process, this use of CFCs accounted for approximately half of U.S. CFC production. As a consequence of this ban, the U.S. share of total world CFC production has fallen from approximately 50 percent down to approximately 37 percent. World Action - In spite of major political efforts by EPA to convince other countries to impose like bans, to date only Canada, Sweden and Norway have followed suit. None of these countries was a major producer of CFCs. More recently, the European Economic Community (EEC) agreed to a 30 percent reduction (from 1976 levels) in the use of CFC aerosol propellants to be effected by 1982. However, to date rto country in the world except the United States has proposed to regulate the non-aerosol uses of CFCs. Recent EPA Action - On October 7, 1980, EPA issued an Advance Notice of Proposed Rulemaking (ANPR) in which it proposes to further regulate domestic production and use of CFCs by imposing a cap on ES-5 ------- Executive Summary - Background total CFC production at current levels, regardless of product or application. (In April, 1980, EPA had announced its intent to propose such a regulation as the first step in an eventual planned phase-down of 50 to 70 percent in world CFC production). EPA's stated preference is for the cap to be implemented through some yet to be determined system of production or use allocation or auction. It is then stated that this artificial supply limitation would cause prices to increase, thus creating an economic incentive to use substitutes in place of CFCs or to force CFC users to better conserve CFCs during their use. The Agency acknowledges that such an action would not have any significant impact per se^ on either U.S. or world emissions of CFCs but the action is justified on the basis that it will achieve the Agency's goal of stimulating international cooperation, and will enable EPA to retain leadership on the issue. It is against this scientific and political background that the Du Pont Company submits its comments on EPA's ANPR on CFCs. EG-7 ------- Executive Summary - Assessment & Conclusions B. DU PONT ASSESSMENT AND CONCLUSIONS In response to the ANPR, the Du Pont Company has reviewed the status of the Chlorofluorocarbon (CFC)/Ozone Depletion Theory, its implications, and the array of options available to industry and government. Upon completion of this review, we reassessed our position and program, and the position, program and plans of EPA. We conclude: 1. There remains no scientific justification for any further regulation of CFCs at this time. 2. A unilateral approach to a solution of the CFC/Ozone Issue through domestic regulation, without an international consensus, is seriously flawed. 3. Continuing assessment of the science and surveillance of the ozone layer is the only sensible option open to the world's governments. 4. EPA has done inadequate work to support its regulatory proposals and to apprise the public of exactly what its proposals are. 5. The proposed economic incentives regulatory options have many serious problems. 6. EPA has not provided sufficient evidence to support a conclusion that continued release of CFCs represent an unreasonable risk to human health and the environment; nor that the potential environmental benefit of the proposed regulation is justified by the risk and cost of such regulation. ES-8 ------- Executive Summary - Assessment & Conclusions Findings leading to these overview conclusions are discussed in depth in the body of the submission. A summary of key points follows: 1. There remains no scientific justification for any fur- ther regulation of CFCs at this time. The CFG/Ozone Depletion Theory remains an unverified theory. There is major disagreement within the world's scientific community as to the validity of the theory. There remain numerous discrepancies between what is assumed, estimated, and predicted in the theory and actual atmospheric measurements. Analysis of actual measurements of stratospheric ozone concentration over the last 20 years, (analysis sensitive to an approximate change in concentration of plus or minus 1 to 1.5 percent) does not detect any depletion of ozone -- in sharp contrast to predictions (based on computer model calculations) made in 1979 by the National Academy of Sciences (upon which EPA relies) that over 2 percent depletion of the ozone has already occurred. One may conclude from this key discrepancy that either: i. The theory is wrong in that it signifi- cantly overstates potential depletion of ozone by CFCs, or ii. There is an equal and offsetting positive effect on ozone generation. ES-9 ------- Executive Summary - Assessment & Conclusions In either case, ozone is not exhibiting a net decrease as predicted. This technique of analysis of ozone concentrations, called "time-trend analysis" is sufficiently sensitive to serve as an early warning system of any developing problem. Under this umbrella, the research programs needed to resolve the key discrepancies between theory and measurement, and to reduce the key uncertainties, can be allowed to proceed with confidence that their continuance in the absense of further regulation does not result in undue risk. Most of these needed research efforts already are underway. Many are specifically targeted to resolve key uncertainties and discrepancies. Even if these studies should confirm that the theory is quantitatively accurate, the risk in waiting for these results is small. As an example, the maximum calculated long-term incremental change in depletion from a 5-year U.S. regulatory postponement would, even in the extreme case of a ban, be approximately only 0.2 percent. This should be compared to the known natural variations in stratospheric ozone of approximately 10 percent occurring over periods of a decade or so. EPA cites as its justification for further regula- tion a 1979 report by the National Academy of Sciences (WAS). There are serious discrepancies between this report and other reports issued imme- diately prior to, and subsequent to, the WAS report. Further, EPA does not incorporate recent develop- ES-10 ------- Executive Summary - Assessment & Conclusions ments in the science in its assessment develop- ments which render portions of the NAS report incorrect and reduce the NAS predicted depletion numbers by approximately one-half. In addition, EPA continues to ignore the utility of ozone time-trend analysis as an early warning system without making any attempt to formally assess the technique in light of recent rapid advances. 2. A unilateral approach to the CFC/Ozone Issue through domestic regulation, without an international consensus, is seriously flawed. The United States produces and uses more CFC than any other single country, yet the U.S. share is approximately only one-third of world CFC production. The problem, if it exists, is global in nature. Thus, no country can solve the potential problem through unilateral regulation. Other countries generally have not regulated CFCs at all, or have employed restrictions on aerosol uses that are much less stringent than U.S. measures, due to their assessments that serious questions exist on the validity of the ozone depletion theory and that most of these questions can be resolved without serious risk to human health or the environment. EPA's strategy of obtaining worldwide regulation of CFCs through the setting of a U.S. regulatory example is ill-conceived and doomed to failure. The underlying disagreement on this issue between countries of the world is over the validity of the theory. More information and objective, shared ES-11 ------- Executive Summary - Assessment & Conclusions assessments of the science are needed to resolve this disagreement before a world consensus and a solution can be forged. Regulatory example-setting by EPA will not advance this goal. Such action relies on political effort, not scientific facts. There is rio evidence to support EPA's conclusion that its unilateral proposals will meet its stated goal of a world major phasedown of CFC uses. In light of the failure of the 1978 U.S. aerosol pro- pellant ban to stimulate significant action abroad of equivalent scope, it is difficult to see how further U.S. regulation will meet with any greater success. Further, a unilateral cap on U.S. production will not per ^£ have a significant potential environ- mental benefit to the U.S. or the world should the theory prove to be valid, but will have a large negative impact on the U.S. economy. In short, no one potentially gains but the United States pays. 3. Continuing assessment of the science and surveillance of the ozone layer is the only sensible option open to the world's governments Given: the great scientific uncertainties and discrepancies between theory and measurement, that research programs are underway to resolve these uncertainties and discrepancies, the low risk in waiting for this needed resolution, ES-12 ------- Executive Summary - Assessment & Conclusions the existence of ah early warning system for any developing problem, the need for, but lack of, an international resolu- tion on this issue, and the very questionable utility of EPA's excessively political approach to obtaining this resolution, we believe that a strategy of diligent testing and assessment, coupled with constant surveillance, and prompt action if a problem is found to be developing, is the only supportable policy available. We do not propose that the world's governments wait until actual human or environmental harm occurs or is determined to be inevitable. 4. EPA has done inadequate work to support its regulatory proposals and to apprise the public of exactly what its proposals are The potential risks created from restricting CFC availability, resulting in use of currently available CFC substitutes, have not been addressed by EPA. In comparison to currently available alternatives or substitutes, CFCs have the more desirable combina- tion of characteristics: safety (nonflammability, low toxicity)y energy efficiency, material compat- ability and value-in-use. EPA does not propose to eliminate major segments of U.S. industry, such as refrigeration. Therefore, it must evaluate the availability of substitutes and what the risks, costs, etc., of these substitutes will be. ES-13 ------- Executive Summary - Assessment & Conclusions Any proper regulation must balance the potential risks which the regulation is projected to reduce versus the risks which such regulation would create, e.g., the risks from forced use of substitute pro- ducts or processes. The low potential for future availability of safe alternatives for CFCs has not been addressed. It has been more difficult to develop suitable alternatives for the current commercial CFCs than initially believed. Such compounds, meeting EPA's standards of environmental acceptability and industry's standards of safety, utility, and cost, ji_f available aj; all, are at best 7-10 years away from commercialization. EPA has not assessed the energy consequence of its proposed regulation, neither restriction on current CFC uses nor what would be sacrificed from unavailability of CFCs for new uses. EPA has prematurely concluded that economic incen- tives regulatory options are preferable. The reasons given for the stated preference are based on theoretical economic relationships and projected responses. But EPA's preference is not based on detailed assessments of the impacts of these options on the actual producers and consumers of CFCs and CFC-dependent goods. Additionally, the primary ES-14 ------- Executive Summary - Assessment & Conclusions competing option, emission reduction, has not been adequately assessed, either technically or economically. The Rand study cited by EPA in support of its pre- ference is not an adequate work on which to base a decision. The study is deficient opposite the use to which it is being put by EPA because it a) is based on an obsolete data base, b) does not include all CFCs being proposed for regulation, c) does not assess all uses of CFCs which would be impacted by the proposed regulation, and d) only compares regulatory options under artificial study para- meters, several of which were selected to meet budget and time constraints on the study. However, most importantly, the Rand study is an empirical comparison of regulatory options under artificially bounded study conditions; not an economic impact study of the consequences of the options being applied to the real world uses of CFCs (all CFCs, all uses). The economic incentives regulatory options proposed by EPA are not adequately developed. Only the theory and skeletal structures are presented. No "how to's" are suggested. Left unanswered are questions on "How would the options be put in place?" and "How would they work?". The Agency has not yet done its homework on these concepts. Industry's ability to provide meaningful comment is severely constrained until such time as it is apparent what exactly is to be commented on. No determination has been made on the impact of the production cap proposal. No determination has been ES-15 ------- Executive Summary - Assessment & Conclusions made on which specific CFC-dependent end products and services would be impacted by the proposal, what the cost would be, or whether such cost could be justified by the potential environmental benefit to the ozone layer of giving up or curtailing these specific products or services. There has been no quantification of these trade-offs. 5. The proposed economic incentives regulatory options have many serious problems There are substantial questions concerning the authority of EPA to promulgate a regulation per- mitting it to auction permits for the right to use or produce CFCs. Such a system would result in revenue generation by EPA, an authority vested in the Congress. Both auction and allocation of use or production rights under a production limitation would create complicated economic problems. In addition, anti- competitive concerns created by these options cannot be ignored. Any system of redistribution of limited resources, particularly if done more than once, e.g., by yearly auction, will create tremendous uncertainty on the part of both producers and users. Business cannot plan or operate effectively under uncertainty of this magnitude. To reduce or eliminate this uncertainty many businesses will attempt to do one of two things: 1) hoard permits to insure supply or 2) precipitously eliminate CFCs from their product lines. Both ES-16 ------- Executive Summary - Assessment & Conclusions actions will have adverse consequences consequences not addressed to date by EPA. Hoarding will create shortages elsewhere, possibly in some "essential" use areas. Precipitous phaseout of CFCs will result in performance problems, possible end- product shortages, loss of jobs, and an increase in industrial and consumer risk from exposure to less safe alternatives. The consequences of both actions will mean significantly more economic impact than the smooth transition case presumed by EPA. It cannot be stressed enough that the economic incentives implementation schemes, in concert with a production cap, have the potential for creating massive uncertainty, which in turn will result in major unaddressed impacts. EPA seems to assume that ownership of production or use rights should be in its hands. For all the discussion about economic incentives options allowing the free market to operate, such schemes represent more, not less, government intrusion into the marketplace. EPA ownership means that the distribution of production and use evolved over the years would be disassembled, to be replaced by the imposition of a new distribution system controlled by government, with unknown consequences. Unilateral imposition of a production cap will have negative impacts on U.S. industry and the balance of trade. Higher prices created by the unilateral U.S. restriction will reduce the competitiveness of U.S. CFCs and CFC-dependent products abroad, but will ES-17 ------- Executive Summary - Assessment & Conclusions have no net environmental benefit. This is because U.S. regulation will have no impact on demand abroad, only on the ability of U.S. industry to meet this demand competitively. What U.S. companies lose will be gained by foreign companies. The control of imports as proposed is equally inequitable as foreign firms would have an unfair advantage in the domestic marketplace. This is because they would be competing from an unrestricted production base opposite domestic producers, thus giving them an unfair cost advantage. EPA's system of permit pounds is inaccurate. A correct evaluation of "potential environmental risk" or "depletion potential" for a CFC (according to the theory) must include weight percent chlorine and the altitude at which chlorine is released to the atmosphere. Below we compare EPA's relative permit pound ranking to a correctly developed ranking: EPA Du Pont CFC-11 1.0 1.00 CFC-12 1.27 1.19 CFC-113 1.30 1.22 CFC-114 2.04 1.64 CFC-115 5.00 2.86 CFC-22 5.56 34.00 ES-18 ------- Executive Summary - Assessment & Conclusions This means that it would take approximately 34 pounds of CFC-22 to equal the potential environ- mental concern of 1 pound of CFC-11, not the roughly 6:1 ratio presented by EPA. The inclusion of CFC-22 in the proposed cap is unjustified and counterproductive. EPA does not have studies or data to support any finding of environmental risk from the use of CFC-22. The fact is that its potential risk factor and its total production makes it less of a potential or theoret- ical problem to the ozone than other compounds, such as methyl chloroform, which are not addressed by EPA. Further, to the extent that CFC-22 represents one of the solutions to the potential problem of release of more environmentally suspect CFCs, its inclusion under the proposed restriction is counterproductive. Conversion from CFC-12 to CFC-22 will be restricted by the inclusion of CFC-22 under the regulation due to the attendant business uncertainties. It is the emission of CFCs that should be of con- cern, not the use. The use of CFCs cannot possibly impact on stratospheric ozone, only emissions to the atmosphere. Consequently, uses of CFCs resulting in no emissions must be exempted. The use of CFC-22 as a raw material for the production of fluoropolymers is a prime example. EPA illogically makes no exemption for such non- emitting CFC uses in its regulatory proposals. ES-19 ------- Executive Summary - Assessment & Conclusions EPA has not provided sufficient evidence to support a conclusion that continued release of all CFCs represents an unreasonable risk to human health and the environment nor that the potential environmental benefit of the proposed regulation is justified by the risk and cost of such regulation. There is no evidence that ozone is being depleted as predicted. EPA incorrectly bases its case on the theoretical impact of CFCs exclusive of other factors. EPA does not address other potential depleters of ozone, such as methyl chloroform nor, more importantly, the potential off-setting effect of C02 releases. EPA does not demonstrate that the theorized risk from CFC emissions can justifiably be considered in isolation from other competing risks, or from off-setting factors which reduce net risk to the environment. EPA presents an "either-or" choice of regulating now or waiting for some distant future date at which time it will be too late to stave off harm if the theory should prove to be valid. However, research programs targeted at the remaining key uncertainties and discrepancies are underway. And this research can be done under the umbrella of an existing early warning system ozone time-trend analysis. EPA has not addressed the incremental risk in waiting for these research results in conjunction with the use of the early warning system. ES-20 ------- Executive Summary - Assessment & Conclusions A related point is that EPA has not quantified the theorized benefit of regulating now versus deferral, again in conjunction with the existing early warning system. EPA has not demonstrated that there would be any significant environmental benefit, per se from its proposed unilateral restrictions. Further, no support is offered for EPA's proposition that further unilateral regulation of CFCs by the U.S. will achieve EPA's goal of a world phaseout of CFCs, nor is any probability of success given. The law supports regulation by EPA if there is an environ- mental benefit not to set an example with the unsupportable expectation that other countries will then "see the light" and respond with their own regulations. EPA has not studied how the individual uses of each of the CFCs would be affected by the proposed rule how much would emissions be reduced?, to what degree would this potentially benefit the environment?, would the cost be justified by this incremental environmental benefit?, and would this reduction in risk be justified by the increase in risks incurred through the use of substitute products and processes? EPA proposes regulating all CFCs, yet only CFC-11 and CFC-12 have been studied to any extent. The potential impact on the ozone of CFC-113, CFC-114, CFC-115 and CFC-22 has not been adequately studied or assessed. These latter CFCs should be studied, not only to determine if they are of potential risk to the environment (and if so to what degree), but ES-21 ------- Executive Summary - Assessment & Conclusions also to determine to what extent they could contri- bute to a solution should evidence show that CFC-11 and CFC-12 are depleting stratospheric ozone. ES-22 ------- Executive Summary Du Pont Position C. DU PONT POSITION Given the preceding conclusions, it is the position of the Du Pont Company that: Time-trend analysis of actual ozone measurements provides an early warning system for any developing problem with stratospheric ozone. It should be immediately adopted and efforts undertaken to refine the technique even further. The existence of this system permits deferring any further regulation while the needed remaining research is performed. Regulation should be based on scientific facts, not unconfirmed theory. Imperfect computer calcula- tions, premised on numerous questionable assump- tions, many in conflict with actual measurements, are not an adequate basis on which to undertake major regulation. The key uncertainties and discrepancies surrounding the theory can be resolved with very low potential risk to health or the environment. Regardless of the results from ozone trend analysis, major research efforts should continue to resolve the underlying science definitively. Du Pont wilj, continue to support these efforts. As the potential problem of ozone depletion is global, solutions must be global. Global solutions hinge on objective global assessment and resolution of the science. Maximum effort should be given to ^obtaining such assessment rather than the political efforts being made by EPA. Further unilateral U.S. ES-23 ------- Executive Summary Du Pont Position regulation as proposed by EPA will not help obtain the needed global assessment and resolution. The United States already has taken actions far beyond those of other major industrialized nations. Further unilateral regulation as proposed by EPA will be exceedingly costly and unfair to U.S. industry and consumers, but will not have any major potential environmental benefit for the U.S. or for the world. Therefore, absent the demonstration of any risk, especially any meaningful short-term risk, absent a resolution of the science leading to a confirmation that CFCs will deplete stratospheric ozone to an extent harmful to public health and the environment, and absent an international resolution of what restrictions, if any, are appropriate, there should be no further regulation of CFCs. In the face of these conditions, if EPA continues to proceed with a regulatory program on CFCs, there should be a Congressional Oversight Hearing on the entire issue and EPA's actions. After a proper resolution of the science, should regulatory restrictions prove necessary to protect stratospheric ozone, such regulation should focus on the net change to ozone from all anthropogenic causes, not on just one potential perturbation in isolation. Should further regulation be necessary, it should be limited to CFC-11 and CFC-12 (approximately 90 percent of the theorized problem), thus allowing ES-24 ------- Executive Summary Du Pont Position less potentially harmful CFCs, such as CFC-22, to contribute to the solution through their use as alternatives. Should further U.S. restrictions on CFCs prove necessary, they should be based on a balanced assessment of risk and benefit from CFC use, and on the risk of continued CFC use versus the risk from the use of available alternative products and processes. Any further regulations determined to be necessary should consider the energy impact of reduced availability of CFCs. Any further regulation should consider that, excepting other currently available CFCs, there are no suitable, safe alternatives for most CFC uses at this time. Du Font's program to develop suitable commercially viable fluorocarbon alternatives is continuing but we are no longer optimistic of success. Should any further U.S. regulation of CFCs prove necessary, the regulation should be reviewed opposite the ongoing justification on a prescribed periodic basis. The economic incentives regulatory proposals currently advanced by EPA are not well thought-out. There are significant problems in the areas of legality, economic impact, and balance of trade which must be addressed. Performing an untried regulatory experiment on such an important industry, ------- Executive Summary Du Pont Position absent the necessary thorough thought and analysis, is unwise. The ability of industry to provide meaningful comment on EPA's economic incentives proposals is limited by EPA's failure to present a description of how these concepts would specifically be designed and applied, and how they would function. Any further consideration of a production cap in conjunction with production or use allocation or auction of permits should provide for: 1) exemptions for non-emitting uses (such as intermediates for production of fluoropolymers), 2) exemption for exports, 3) a restriction on imports parallel to domestic restriction, 4) a minimum of 10 years advance notice of allocation system change to reduce uncertainty, 5) yearly reviews of the effectiveness of the cap as a control measure, both environmen- tally and economically, 6) yearly reviews of the justification for the level of the cap based on the most recent depletion measurements and model calculations, and 7) a clear statement of the guidelines to be used to decrease oj: increase the permissible production level. ES-26 ------- Executive Summary - Problems with EPA's Approach D. PROBLEMS WITH EPA'S APPROACH TO ISSUE EPA has limited its assessment of atmospheric science to one report [NAS, 1979a] , and has not considered other reviews, for example the United Kingdom Department of the Environment's report [UK DOE, 1979] or the European Economic Community Council's report [EEC, 1980] which reach conclusions at variance with NAS. EPA's ANPR assessment of the science does not include developments since the 1979 NAS report developments which cut the NAS predicted future depletion estimate in half. Scientific knowledge relevant to this issue is changing rapidly. Yet EPA indicates no plans for updating its assessment of the science or for arranging for periodic reviews by qualified outside bodies such as NAS. Major efforts should be made to include current information in the decision-making process and to review periodically such decisions opposite ongoing developments. Ozone depletion, should it occur, would be a global problem requiring a global response. Such response has not occurred due to scientific assessments by other nations that the problem does not warrant regulatory response at this time. Therefore, the key need is for a global assessment and resolution of the science leading to a global response policy. However, EPA seems to be concentrating on advancing ES-27 ------- Executive Summary - Problems with EPA's Approach" its views through unilateral regulation, rather than working to meet the underlying need for inter- national scientific assessment. EPA consistently focuses on an extreme worst case scenario. This focus grossly overstates the potential risk involved in continued emissions of CFCs. The worst case scenario has no reasonable probability because it would require business to operate as if there were no environmental or regulatory concern. Further, this scenario ignores recent CFC production history, and ignores the effect of CFC aerosol phase-downs now occurring in Europe. A further problem with EPA's treatment is the failure to assign any ranking or probability to the scenarios. Worst case scenarios are treated with equal (and in some cases higher) emphasis than the most probable scenarios indicated by the facts and common sense. In EPA's treatment, uncertainties are ignored and low probabilities become "fact". The problem is theorized to evolve gradually over 70-100 years. However, EPA limits its discussion to an "either-or" choice either action must be taken now or we will have to suffer the consequences of major ozone depletion. This "either-or" choice ignores the facts that: 1) time-trend analysis of ozone measurements already is available as an early warning system, 2) no ozone depletion has been detected to date, and 3) even if the theory should ES-28 ------- Executive Summary - Problems with EPA's ApproacE prove to be quantitatively correct, a regulatory deferral of even five years would not result in a significant incremental increase in risk. EPA has not yet addressed the risks associated with a limitation on the availability of CFCs, for example, the flammability and toxicity of the available non-CFC alternatives which would be forced into use. No attempt has been made to balance speculative long-term risks from continuing use of CFCs versus known risks from alternatives. EPA has not adequately developed the regulatory options it proposes to use. The theory of economic incentives options has been presented, but as yet there has been no fleshing-out of how the options would look, how they would be implemented or how they actually would function. The ANPR is the second time (the Draft Rand Report [Rand, 1979] being the first) that industry has been asked to comment specifically on the same non-specific proposals. EPA has not adequately assessed the economic impacts which would result from its proposed restrictions. The economic work done to date is a limited comparative study of regulatory options, not an impact assessment of these options applied to the real world of CFC production and use (of all pro- ducts and all applications). Nor has EPA yet addressed the impact of its pro- posals on the nation's energy use, present and future. ES-29 ------- Executive Summary - Problems with EPA's Approach The Agency has prematurely decided in favor of economic incentives options prematurely because alternative, options, such as emission reduction, have not been adequately examined by EPA, and because EPA has not given adequate attention to the impacts of the economic incentives proposals. EPA's stated preference for economic incentives options seems to be based on the Agency* s perception the options would be easier to design, implement and enforce; not upon whether the the incentives options would be better for the affected industries and consumers. Experimentation on critical industrial segments of the nation's economy is unwise. ES-30 ------- Recommendations for Action By EPA E. RECOMMENDATIONS FOR ACTION BY EPA The following are Du Font's recommendations for actions to be taken, or at least initiated, by the EPA, which we believe will lead to a proper resolution of the CFC/Ozone Controversy. Should this resolution dictate the need for further regulation, such actions as outlined herein will help ensure a balanced and cost-effective regulation. EPA should promptly arrange for an updated assess- ment of ozone trend analysis by a qualified outside body, such as the NAS. If an outside review body cannot be employed, a joint industry/government/ academia symposium should be held to review objectively the method opposite the questions: How sensitive is it? What is the confidence range? How and when can it be further improved? A companion recommendation would be for EPA to support the further development of trend analysis. EPA should arrange for an objective, thorough review of the science (both the theory itself and effects of ozone depletion) by an international panel of qualified scientists. A joint NAS/UK Royal Society effort would be a logical starting point. Inter- national political organizations such as the Organi- zation for Economic Cooperation and Development (OECD) are not adequate for this assessment due to the limited participation of scientists and the political pressures present in such groups. Even if an international review cannot be promptly arranged, EPA should recontract with NAS for an updated review of the science, followed by a yearly ES-31 ------- Recommendations for Action By EPA reassessment. The predicted problem is a long-term problem. The science is changing rapidly. Any regulatory decision based on the science at any point in time must be reassessed as the scientific justification for that decision changes. Between NAS reports, EPA should meet quarterly with the Chemical Manufacturers Association (CMA) Fluorocarbon Project Panel (FPP) , and other appropriate scientific advisors, to stay current with the broad spectrum of scientific developments. The Agency needs to publish the parameters of its decision making on the issue: i. What specific level of ozone depletion does EPA consider to pose an unreasonable risk to health and the environment? ii. What will it take to convince EPA there is or is not a serious problem, e.g., a. What sensitivity of ozone trend analy- sis is accepted (and on what basis)?; What ozone trend analysis results would be viewed as a significant indication of a developing problem? b. What other science developments would be viewed as significant? c. What criteria does EPA use to judge the credibility of sources and reported developments, and which sources meet these criteria. ES-32 ------- Recommendations for Action By EPA d. What will EPA do to ensure staying abreast of developments? e. What is the process EPA uses to get developments assessed and to the attention of the regulatory decision- makers? f. What must happen internationally to convince EPA of the need or lack of need for further U.S. regulation? By whom? In what time period? iii. How does the specific proposed regulation result in reduction of risk on this issue and what is the magnitude of this reduc- tion? If in the periodic reviews of the science, it is determined the risk has decreased significantly, what are the parameters of regulatory response? The Agency needs to redefine the problem of ozone depletion generically and then determine and justify whether CFCs should be treated in isolation from other potential depleting compounds and in isolation from potential ozone increasing compounds. The charge to EPA of the 1977 Clean Air Act Amendments is protection of stratospheric ozone, not the regulation of CFCs. What is the justification for including CFC-22 under the regulation when methyl chloroform represents a greater total potential problem? Conversely, modelers now include the CC^/ ozone augmentation effect. This needs to be factored into EPA's assessment. ES-33 ------- Recommendations for Action By EPA On the international level, EPA should abandon its excessively political strategy in favor of an effort to help obtain the needed global scientific assess- ment and resolution. The Agency should publish its plans for furthering the scientific resolution of this issue. As pointed out in previous sections, EPA must complete a significant -body of work before it can support the proposed regulatory options. Further assessment and study are needed in the areas of: i. Emission reduction and alternatives. What is achieveable, in what time-frame and at what cost? ii. Impact of economic incentives options. What would be the actual impact to industry and consumers if the options were applied to all CFCs and all CFC uses as proposed? iii. Energy penalty of regulations. iv. Risk from alternatives substituted for CFCs, and a risk-risk assessment of continued CFC use versus use of alter- natives. v. A detailed fleshing-out of the incentives options for comment specifying exactly how they would be structured, how they would be implemented and how they would function. ES-34 ------- Recommendations for Action By EPA We recommend that EPA hold a series of informational exchange meetings around the country to discuss its proposals, hear concerns and gather information to help its studies of i.-iv. above. EPA needs to employ a more realistic timetable. The current timetable shows a completion date of Janu- ary, 1981 for the final draft of the proposed rule. The ANPR comment period closes January 5, 1981. Further, we question how EPA can digest and evaluate ANPR comments and submissions in time to publish a formal proposed rule in March, 1981. The proposed timetable appears unrealistic unless EPA has no interest in the ANPR comments and has already made up its mind on how to proceed. Given the magnitude of the issues which remain to be addressed, parti- cularly on the economic incentives options, we fail to see how a reasonable proposal can be finalized in this period. We ask for clarification. Due to the untried nature of the economic incentives options, the numerous questions and concerns which have yet to be answered, and the total lack of experience with these regulatory options in the real world, if EPA elects to proceed with their use, we would strongly urge that a pilot test first be undertaken. The options should first be applied to a carefully monitored industry or industry segment, and the impacts thoroughly evaluated, rather than immediately applying this theoretical regulatory approach to the CFC industry which has such broad and major impacts on the total economy. ES-35 ------- Recommendations for Action By EPA If EPA decides to promulgate a rule, (regardless of which regulatory option is selected) the Agency should issue an annual report detailing: i. Results of actual ozone measurements. Has depletion been detected? If so, how much and at what rate? ii. Computer calculated or estimated ozone depletion based on best current infor- mation. What are the current model calculations? iii. Status of U.S. versus world regulatory situation. Has the U.S. regulation achieved the Agency's goals? Have other countries followed EPA's lead or is the U.S. example being ignored? iv. The continued need for the regulation as promulgated. v. The economic impact of the promulgated regulation, particularly if new regulatory concepts are involved. This should be compiled by major market segment and business size, as well as a summary report. ES-36 ------- Recommendations for Action By EPA vi. The identity of substitutes employed in place of CFCs by use category. For all substitutes (and especially new substi- tutes) safety data, toxicity data, energy efficiency, development cost of replacement substances and redesign cost for manufac- turers should be monitored for a period of 10-20 years to determine the true cost of regulation for guidance in future rule- making efforts. ES-37 ------- I. INTRODUCTION 1-1 ------- Introduction In 1974 a theory was advanced [Molina and Rowland, 1974] that the family of chemical compounds termed chlorofluorocarbons (CFCs), also commonly known as fluorocarbons, upon their release into the lower atmosphere (the troposphere), were eventually transported to the upper atmosphere (the stratosphere), where they entered into a complex series of reactions which resulted in the gradual decrease (depletion) of ozone. A decrease in ozone is expected to result in an increase in solar ultraviolet (UV) radiation reaching the Earth's surface. Over time, such an increase in UV is theorized to result in various adverse effects, including an increase in human skin cancer, possible damage to certain crops and marine species, and potentially even a small modification in the climate. In 1978, the Environmental Protection Agency (EPA), together with the Consumer Product Safety Commission (CPSC) and the Food and Drug Administration (FDA) , acting on this theory, promulgated a rule which prohibited the use of CFCs in aerosol propellants in all but a few essential applications (43 Federal Register 11301 ejt seq., March 17, 1978, 40 C.F.R. Subsection 762 e_t seq.) . Although there were many uses of CFCs in the United States, the aerosol propellant use was singled out because it represented approximately half of the consumption of CFCs, resulted in prompt and complete release of the CFCs to the atmosphere, was generally considered to be a "non-essential" use, and alternative products were available. In April of 1980, EPA announced [Jellinek, 1980a; EPA, 1980a] that it intended to proceed with further regulation of CFCs in the United States. In October of 1980, EPA published an Advance Notice of Proposed Rulemaking (ANPR) to outline its regulatory thinking and to solicit comment on various approaches to regulation of 1-2 ------- Introduction non-aerosol propellant uses of CFCs (45 Federal Register 66726 et seq.) . All non-aerosol uses are covered in the proposal regardless of essentiality of use. EPA cites as justification for this further regulation reports released in November and December 1979 by the National Academy of Sciences [NAS, 1979a; 1979b]. No scientific studies published more recently than the 1979 reports are cited by EPA or listed as supportive of the Agency's position. EPA plans to promulgate CFC regulations under the authority of Section 157(b) of the Clean Air Act (42 U.S.C. Subsection 7457) or under Section 6(c)(l) of the Toxic Substances Control Act (15 U.S.C. Subsection 2605). In choosing a regulatory strategy for non-aerosol CFC regulation, EPA stated that a primary concern will be the "effect the choice [of strategy] will have on other nations' decisions in this area" [45 Federal Register 66728]. Within this overall concern, EPA identified three different strategies for regulating CFCs: 1. Wait-and-See 2. No Growth 3. Substantial Emissions Reduction Under the Wait-and-See approach, EPA would take no action until better evidence of the ozone depletion theory is obtained. EPA rejected this approach because it believes that were it adopted, the great majority of other producing and using nations will follow suit. Therefore, they believe that a domestic strategy of wait-and-see is likely to be equivalent to a world strategy of wait-and-see (45 Federal Register 66728-66729). 1-3 ------- Introduction Under the No Growth scenario, the EPA states that it would limit CFC production to present levels and would take no further action until warranted by international conditions or further evaluation of the credibility of the theory. This strategy is acceptable to the Agency on a short-term because it will "convince other nations to agree to concerted international action" and because it is a "signal to other CFC-producing and using nations that the United States is concerned enough about the risks entailed in the depletion of the ozone layer that it is willing to take serious action on the basis of present knowledge." On a long-term basis, however, a no growth strategy is considered unacceptable because it "would still result in an unacceptable level of risk" (45 Federal Register 66729). Under the Substantial Emissions Reduction strategy, the United States would reduce its production to some fraction of the level predicted to be necessary on a worldwide basis to achieve an acceptably low level of ozone depletion. As a short-term option, this strategy was rejected by EPA because "it could strengthen the Wait-and-See attitude abroad by creating the impression that other nations could afford to wait before taking action." However, on a long-term basis, this is considered the only acceptable option because of the "extreme caution" which must be exercised on account of the "substantial" evidence that ozone depletion is occurring (45 Federal Register 66729) . Du Pont believes that none of the three options above are truly adequate at this time to deal with the problem of potential ozone depletion. We reject the wait-and-see approach because it ignores the fact that CFCs may present a problem and implies that nothing need be or will be done to learn more about the potential problem and means of making responsible regulatory decisions concerning it. And, at this time, we reject both the 1-4 ------- Introduction No Growth and Substantial Emissions Reductions alternatives because both entail costly regulation at a time when the scientific basis to support regulation is questionable. Instead, we have identified a fourth option, not mentioned by EPA, but one which we believe is the only appropriate one at this time. This option may be called "Assessment and Surveillance". Under the Assessment and Surveillance approach, the research currently being funded by the Chemical Manufacturers Association Fluorocarbon Project Panel (CMA/FPP), individual CFC producers, government and others will continue. This research is directed toward reducing the uncertainties in the ozone depletion theory, and toward a more accurate and thorough evaluation of the quantitative validity of the theory. The research is being conducted under the umbrella of analyses of actual ozone measurements called ozone time-trend analysis which is capable of providing the surveillance necessary to warn of any developing problem in stratospheric ozone levels. The availability of ozone time-trend analysis as an early warning system in ozone changes, coupled with periodic reassessments of the need for regulatory action, provides EPA with an assurance that regulation of CFCs can be deferred, without unreasonable risk developing. Should the results of ozone trend analysis or ensuing research results indicate that ozone is being depleted, the wisdom of further deferral can be reassessed. It is important to note that at this time ozone time-trend analysis is not capable of proving or disproving the theory of CFC catalyzed ozone depletion per se. It can, however, indicate whether ozone is being impacted from any source. Regulation appropriately should focus on protection of the, ozone, not just on what one anthropogenic effect might be. To date, no net depletion has been detected. 1-5 ------- Introduction Our comments throughout are directed toward support for the Assessment and Surveillance option. Specifically, we note that the uncertainties in the science make regulation at this time unwise and unwarranted. As pointed out, the availability of the early warning system allows EPA to monitor stratospheric ozone while the research and assessment is being conducted. The risk of waiting until the science is more completely resolved is therefore minimal. In addition, if ozone depletion is indeed a problem, it is international in scope and unilateral action by the United States will have no appreciable environmental impact. Given the Agency's goal to stimulate foreign action with this regulation, it is difficult to see how non-aerosol regulation in this country will stimulate foreign regulation in light of the failure of the U.S. aerosol regulation to stimulate action abroad of equivalent scope. (As a result of the U.S. aerosol ban, total domestic use of CFCs was reduced by approximately 50%. No major country has to date announced equivalent regulation). Throughout the comments, therefore, we provide support for the Assessment and Surveillance approach. The submission body is divided into nine sections: In Section II we discuss the uses and essentiality of CFCs. Here we note the various major applications in which CFCs are used and the reasons for their use. We also point out the limitations of the currently available substitute products and processes. In Section III we discuss the legal issues of CFC regulation. We note that EPA has no authority to regulate at this time. We also point out the many problems of economic incentives regulation, and we. suggest additional studies which must be conducted before regulations can issue. 1-6 ------- Introduction In Section IV we discuss the science of the ozone depletion theory. We note the uncertainties surrounding the 1979 calculations and the advances that have been made since those calculations. Particular attention is given to recent developments (and their significance) in efforts to detect a trend in actual ozone concentration measurements. Section V evaluates the risk from continued production and use of CFCs and from deferring regulation for several years. The bottom line, of course, is that the risk is not developing as predicted, and with the availability of an early warning system, deferral of costly regulation to obtain better information is a defensible option. We also discuss the risk which would be associated with regulation of CFCs at this time and the need for risk-risk evaluation. In Section VI we discuss the international ramifications and implications of CFC regulation. Ozone depletion, if it occurs, is an international problem and unilateral action by the United States will have no appreciable environmental benefit. We also discuss the need for a true global assessment, resolution and cooperative action plan, and point out that EPA's proposed program will not advance such a goal. In Section VII we discuss economic considerations of EPA's proposed regulation of CFCs. We note that the impact of economic incentives regulation are far greater than EPA has realized, and that, therefore, the Agency must conduct additional studies to support such regulatory initiatives. Section VI I.I contains a discussion of the search for alternatives to the currently used CFCs. We note that research 1-7 ------- Introduction has been on-going almost since the inception of the theory but that for the most part, no viable alternatives have emerged. Section IX contains a general summary and conclusions,, including our recommendations for action by EPA. Section X contains appendices which in general are more detailed and more technical discussions of portions of the main body of comments. Where appropriate, we have indicated in the text those places where an appendix provides additional information. Section XI lists references. Throughout the submission, extensive use is made of references, both to previous reports and submissions and to numerous articles, papers, etc., by outside researchers. Copies of all cited references are submitted as part of Du Font's formal response to the ANPR. In addition, due to the fact that the CFC/Ozone issue has been on-going since 1974, a large body of reports has been generated by CFC producing and using companies and their contractors. Most of these have been submitted to EPA or have formed the basis for previous submissions to EPA by industry. However, due to changes in personnel at the Agency and to a desire to issue an up-to-date and complete official record on which decisions can be made, we are incorporating as part of our response to the ANPR those previous reports and submissions which remain pertinent. We also include copies of a number of reports prepared by EPA contractors which we believe contain important data and analyses not as yet given adequate attention by EPA. A list of the above reports and submissions with a brief descrip- tion of each may be found in Appendix A. 1-8 ------- Introduction Last, Du Pont is an active member of the Alliance for Responsible CFC Policy. We participated extensively in the formulation of the Alliance's comments which are being forwarded to the Agency. We incorporate the Alliance's comments as our own and they should be considered as part of the Du Pont submission on the ANPR. 1-9 ------- II. CHLOROFLUOROCARBON USES AND ESSENTIALITY Paqe A. INTRODUCTION 2 B. REFRIGERATION AND AIR-CONDITIONING ' 4 C. MOBILE AIR-CONDITIONING 7 D. SOLVENTS 10 E. BLOWING AGENT FOR RIGID POLYURETHANE FOAM 13 F. BLOWING AGENT FOR FLEXIBLE POLYURETHANE FOAM 17 G. BLOWING AGENT FOR POLYSTYRENE, POLY- ETHYLENE AND PHENOLIC FOAMS 21 H. FOOD FREEZANT 25 I. STERILANT GAS 28 J. INTERMEDIATE FOR FLUOROPOLYMER PRODUCTION 32 K. CHLOROFLUOROCARBONS AND ENERGY CONSERVATION 36 L. SUMMARY 44 II-l ------- CFC Uses and Essentiality A. INTRODUCTION Historically, discussion of the Chlorofluorocarbon/Ozone Depletion Issue has centered on three questions: 1. To what extent, if any, will emissions of the family of chemicals known as chlorofluorocarbons (CFCs) lead to an eventual depletion of the earth's stratospheric ozone layer? 2. If this were to occur, what would be the consequences to human health and the environment?, and 3. If there would be adverse consequences, to what extent should the production or use of CFCs be curtailed? Often forgotten in these discussions is the key fact that CFCs play a wide, and in many instances essential, role in modern society. Perspective is needed on: 1. The nature of these chemicals. 2. Where they are used. 3. Why they are used. 4. Why they can't be readily replaced with something else. Consequently, we believe it is appropriate to begin our submission with a review of the major uses of CFCs. For each II-2 ------- CFC Uses and Essentiality use, we provide a brief description, a highspot of the essen- tiality or benefits of the use and a review of the limitations of existing alternatives. For more detailed information on the uses of CFCs (where used and why) and the competing available alternates, we refer the reader to the March, 1978 Du Pont Submission to EPA [Du Pont,. 1978], a copy of which is attached. Although this submission was based on 1976 data (the most recent full-year for which data was available at the time of the 1978 submission preparation), it remains pertinent in its descriptions of CFC uses and non-CFC alternatives. (Details on Du Font's program to develop suitable alternative fluorocarbons appears in Section VIII and Appendix B). In addition, because energy conservation has become such an important national objective, and because CFCs generally provide a significant energy advantage over their alternatives, we discuss CFCs and energy conservation in Section II-K. (De- tailed support for our energy discussion appears in Appendix C). II-3 ------- CFC Uses and Essentiality B. REFRIGERATION AND AIR-CONDITIONING 1. Description Chlorofluorocarbons (CFCs) are used widely as refrigerants in mechanical refrigeration and air-conditioning systems. In these systems, for instance in a home refrigerator, the refrigerant alternatively is expanded and compressed to dissipate heat from a cooling chamber. In 1979 approximately 270 million pounds of CFC were used in refrigeration and air-conditioning applications, or very roughly 33 percent of total domestic CFC production. For simplicity, the ensuing discussion will be limited to refrigeration. 2. Essentiality and Benefits Refrigeration is essential to today's way of life. Meat, poultry, dairy products, fruits, and vegetables - from processing to storage, to transportation, to the consumers' tables - require refrigeration whether the product is sold fresh or frozen. More than three-fourths of all food .consumed by Americans is processed, shipped, or marketed under some degree of refrigeration. Nearly all households in the United States have one or more refrigerators and freezers. Bulk shipments of food products are made in 178,000 refrigerated vans and 27,000 refrigerated freight cars for eventual distribution through over 40,000 supermarkets and nearly 180,000 other food stores, as well as 250,000 restaurants and other commercial and institutional eating establishments - all of which require refrigeration equipment. II-4 ------- CFC Uses and Essentiality This vast array of refrigeration equipment and facilities helps ensure that food reaches the nation's tables with minimum loss of food value and minimum risk to health because of spoilage and contamination. Refrigeration also is critical in other activities which are esential to public health. Blood, bone, and tissue are stored under refrigerated conditions in most of the 8,000 hospitals in the United States. In other medical applications, refrigeration of biological matter is used either to preserve or to destroy the viability of the material and to prevent degeneration. The manufacture and storage of lifesaving Pharmaceuticals require refrigeration. Virtually all such refrigeration equipment is designed for, and exclusively uses, CFC refrigerants. Other refrigerants cannot be substituted in this equipment. In fact, other refrigerants generally are considered too hazardous for use, even if suitable equipment to use them were available. Therefore, there is considerable concern for the public safety in the use of non-CFC alternatives as indicated below. More recently, energy concerns have resulted in many consumer appliances receiving energy efficiency ratings. All the high-efficiency refrigerators and freezers developed under this energy-conserving program utilize CFC refrigerants and CFC-blown thermal insulating foam (see Section II. - E) . 3. Limitations of Alternatives Prior to 1931, refrigeration equipment used refrigerants such as methyl chloride, ammonia and sulfur dioxide. These materials are toxic and some are flammable or explosive. II-5 ------- CFC Uses and Essentiality CFCs were developed specifically to overcome these serious hazards. Their widespread use attests to how well they have met the requirements for safety. Incidents of refrigerant toxicity and fires are almost unknown in CFC refrigeration systems. On the other hand, the literature contains numerous reports of deaths, injuries, and fires attributed to ammonia, sulfur dioxide, methyl chloride, and other non-CFC refrigerants. Efforts to date to develop safe alternatives for CFC refrigerants have been unsuccessful. (See Section VIII and Appendix B). II-6 ------- CFC Uses and Essentia1ity C. MOBILE AIR-CONDITIONING 1. Description Chlorofluorocarbon 12 is the refrigerant used in mobile air-conditioning in automobiles, trucks, and farm tractors, etc. the refrigerant alternately is expanded and compressed to dissipate heat from the inside of the vehicle. In 1979 approximately 105 million pounds of CFC were used in mobile air-conditioning applications, roughly 13 percent of total domestic CFC production. 2. Essentiality and Benefits Air-conditioning plays a key role in reducing the debilitating effect of heat stress on human activity. It contributes to improved safety, health and productivity. For example, driving under the heat stress conditions commonly experienced in many parts of the United States during the summer months has been shown to adversely affect driver alertness. Lack of alertness has been ascribed a key factor in many of the 47,000 fatalities and 1,800,000 disabling injuries experienced annually in traffic accidents in the United States. A recent study of the effect of heat stress in driving perfor- mance conducted for the National Highway Traffic Safety Adminis- tration concluded that "suitable air-conditioning equipment or other effective countermeasures should be available to drivers who will be exposed for extended periods of time to [heat stress] conditions." Virtually all such mobile air-conditioning equipment is designed for the exclusive use of CFC-12. Other refrigerants cannot be substituted in this equipment. CFC air-conditioning Ii-7 ------- CFC Uses and Essentiality refrigerants are thus an important asset in current programs to increase passenger safety in transportation. 3. Limitations of Alternatives Chlorofluorocarbon 22, the refrigerant used in residen- tial and commercial air-conditioning, although now under regula- tory consideration by the EPA, is not as great an environmental concern as CFC-12 because most of it is removed naturally in the lower atmosphere. Chlorofluorocarbon 22, therefore, has been considered as an alternative refrigerant for use in auto air-conditioning; however, it necessitates higher equipment operating pressures to achieve proper operation. To obtain these higher pressures, stronger and heavier equipment than that currently used is needed for safety reasons. Since automobile makers have identified weight reduction as an important means to help achieve mandated energy conser- vation goals by 1985, weight increases to accomodate air- conditioning would waste fuel. Further, the development of this equipment would be an expensive and time-consuming task. For example, one auto company reports the development of necessary new compressors would require an investment of $150 million and five to seven years of development effort. Systems which use air as the refrigerant (air cycle systems) are being investigated for auto air-conditioning but major questions concerning their energy efficiency, effective- ness, and reliability still must be answered. Air cycle systems also are likely to add considerable weight to the vehicle and II-8 ------- CFC Uses and Essentiality require more power to operate. Further, there is no assurance that the current concerted development effort - which also would require five to seven years - will result in a workable air cycle system for use in mobile air conditioners. II-9 ------- CFC Uses and Essentiality D. SOLVENTS 1. Description of Use Chlorofluorocarbon solvents are based on CFC-113. They are used mainly as high-quality cleaning solvents in special equipment that purifies and recycles the solvent for multiple reuse. The equipment contains the vapor from the boiling solvent. Parts to be cleaned are lowered into the vapor. Freshly distilled solvent condenses on the parts, rinsing off the contaminants. The vapor heats the cleaned parts so they dry rapidly as they are removed from the vapor. Solvents serve many purposes, including the removal of soldering fluxes from electronic components, the cleaning of metal, plastic items, and glass, and the drying of parts by displacing water. Chlorofluorocarbon solvents are most fre- quently used where high-reliability cleaning is essential, such as in the manufacture of semiconductor, aircraft, computer, medical and military devices. For certain special uses, CFC-113 is blended with other solvents. Such formulations preserve the low toxicity and non- flammable characteristics of the CFC component. In 1979 approximately 130 million pounds of CFC were used in solvent applications, roughly 16 percent of total domestic CFC production. 11-10 ------- GFC Uses and Essentiality 2. Essentiality and Benefits The nonflammability and low toxicity of CFC-113 solvent provide a major contribution to worker safety. No implications of carcinogenicity, mutagenicity, or teratogenicity have been found after extensive toxicity testing and after many years of use. No organic solvent has a lower toxicity rating that CFC-113. It has high chemical stability and does not require the addition of stabilizing chemicals. This permits repetitive recovery and reuse without danger of decompo- sition leading to acid formation which can damage the items being cleaned. Many other nonflammable solvents do require such stabilizing additives. This same chemical stability also means that CFC-113 is photochemically stable, and hence, does not contribute to photochemical oxidant levels in smog - a major national air quality concern with many alternative solvents. Other important advantages result from the thermodynamic properties of CFC-113. The low boiling point and low heat of vaporization result in unusually low energy requirements for vapor cleaning equipment operations, thus contributing to energy conservation. The low temperature and high density of the vapor in vapor cleaning equipment also permit unusually efficient recycle, typically well in excess of 99 percent. The low boiling point also permits safe handling of parts as soon as they are taken from the vapor cleaning equipment. 11-11 ------- CFC Uses and Essentiality 3. Limitations of Alternatives Other solvents might be used in some applications, but at a penalty. Replacement of CFC-113 by hydrocarbon solvents, which are flammable, would require major plant investment in explosion-proof equipment. Chlorinated solvents are more toxic, and can damage many plastics and elastomers that are compatible with CFC-113. Vapor emissions of hydrocarbons and chlorocarbons, with few exceptions, result in elevated levels of photochemical oxidant in smog, and are implicated in concerns over the safety of the work environment. Water is an alternative for some cleaning operations, but critical cleaning of electronic, medical, space program, and defense items require extremely low levels of residual soil. Water cleaning systems can leave behind trace amounts of the chemical surfactants necessary in the water process. On electronic circuit boards and components, these residual surfactants can cause electrical leakage, equipment malfunctions, and reduced reliability. Solvent cleaning processes also concentrate removed soil for proper disposal as the solvent is recycled, while water cleaning disperses the soil in large volumes of water. Additionally, there are increasingly stringent requirements for discharge of contaminated water. Cleaning up water to meet the standards generally is quite energy intensive. 11-12 ------- CFC Uses and Essentiality E. BLOWING AGENT FOR RIGID POLYURETHANE FOAM 1. Description of Use Rigid polyurethane foams are formed by reaction of suitably formulated chemicals. The inclusion of chlorofluoro- carbon 11 as a "blowing agent" permits the formation of a light foam by "inflating" the reacting chemicals. The foam formulation and reaction conditions can be controlled so that the walls of each minute bubble or cell remain intact, trapping the blowing agent in the foam. This "closed-cell" structure also contributes to the rigid character of the foam. Formulation changes permit the manufacture of open-cell or flexible polyurethane foams for other uses. (See Section II-F). Rigid polyurethane foams are used primarily for thermal insulation, for example, in home construction paneling and roofing, and between the walls of refrigerators and freezers. Efficient insulation uses trapped gas to keep heat or cold where needed, as a down-filled jacket keeps the wearer warm by trapping air. Because of its low thermal conductivity, CFC vapor is much more effective than air as an insulating gas. Actually, CFC-blown polyurethane foams provide the best insulation possible using today's materials and technology. The energy savings In the insulation industry, insulation may be measured as the "K factor." Lower K factors signify lower heat loss and thus better insulation. For example, K factors for chlorofluoro- carbon-blown polyurethane: 0.12; for fiberglass: 0.25 btu/hr/in/degrees F/ft . Therefore, the polyurethane foam is twice as effective per unit thickness. (See Section II-K). 11-13 ------- CFC Uses and Essentiality associated with this use are extremely large (See Section II-K). Additionally, compared with other blowing agents, CFC blowing agents permit a lighter weight, more uniform, high quality foam, with greater adaptability to many insulating tasks. Rigid polyurethane foam is manufactured in several forms. Approximately one-quarter of the production is boardstock, used in residential, commercial, industrial, and transportation construction. Alternatively, foam may be prepared in a preformed cavity. Such "poured-in-place" foam constitutes almost half of rigid polyurethane foam production and is particularly important in refrigerator and freezer manufacture. Additionally, the foam can be sprayed on the surface to be insulated. Tanks and pipelines can now be insulated economically by such techniques; whereas the high cost of alternative insulating procedures could not be justified by the energy savings realized. The same spray-on technique is replacing the expensive paper felt and hot asphalt technique for sealing and insulating roofing on industrial and commercial buildings. Spray-on application accounts for about one-quarter of rigid polyurethane foam production. The application of sprayed foam typically is performed by hundreds of small, local busi- nesses. Minor additional uses are in packaging and marine flotation devices. In 1979 approximately 75 million pounds of CFCs were used in these applications, roughly 9 percent of total domestic CFC production. 11-14 ------- CFC Uses and Essentiality 2. Essentiality of C F C B ,lo w lng__Ag_ e^n JL^_AJl_C lg.jj.gjj. _C e_ l_ 1 Polyurethane Foam Rigid polyurethane foam insulation made using CFC blowing agents has .physical advantages which are unattainable by other insulation materials. Rigid polyurethane foam using CFC blowing agents: Provides excellent insulation in thin amounts,thus conserving energy and increasing usable space for household refrigerators and freezers. Provides the most energy-efficient, commerical refrigeration display and storage facilities available. Imparts structural integrity as well as insulation for walk-in refrigeration storage, refrigerated railroad cars, refrigerated delivery trucks, refrigerated truck trailers, and public and commercial roofing and paneling - both interior and exterior. Permits large foam sections and filling applications to be made without overheating and charring the foam core, by moderating the effect of the heat from the chemical reaction in which the polyurethane resin itself is formed. Gives excellent adhesion to metal surfaces. Promotes uniform density throughout the entire foam structure. Il-ib ------- CFC Uses and Essentiality Permits the application of spray-on insulation over a wide range of weather conditions. 3. Alternatives and Limitations With present technology and materials, there is no available substitute for CFC blowing agents in rigid polyurethane foams.. If, due to regulation, rigid polyurethane foams were to be replaced with such alternatives as fiberglass, other foams made without CFC blowing agents, or wood pulp products, then deficiencies must be accepted in such factors as insulation effectiveness, weight, cost of materials and application, structural integrity, and energy conservation. These are, in fact, the reasons for selecting rigid polyurethane foams over competitive technology. 11-16 ------- CFC Uses and Essentiality F. BLOWING AGENT FOR FLEXIBLE POLYURETHANE FOAM 1. Description of Use Flexible polyurethane foams are formed by reaction of suitably formulated chemicals. The manufacturing processes are varied from those used for rigid polyurethane foam manufacture (See Section II-E) to ensure breakage of the polymer walls which initially separate the minute bubbles or cells created by the "inflating" action of the blowing agent. The result is a three-dimensional network of cells, open to each other, which contribute to the flexibility of the foam. Water, methylene chloride and chlorofluorocarbons (CFCs) are used as blowing agents to control some of the physical properties of the final foam. These three agents are partly, but not fully, interchangeable. The extent of interchangeability depends on the formulation and foam characteristics desired. Flexible polyurethane foam is prepared either as a large "bun" (which has the appearance of an enormous loaf of bread) , which is subsequently cut into "slabstock," or in a mold of its ultimate design shape. Flexible polyurethane foam is utilized in padding for furniture, and for seats and interiors in transportation. It is used for bedding, textile laminate, carpet underlay, gasketing, sound deadening, and packaging. In 1979 approximately 50 million pounds of CFC .were used in these applications, roughly 6 percent of total domestic CFC production. 11-17 ------- CFC Uses and Essentiality 2. Essentiality and Benefits Flexible polyurethane foam products essentially have replaced the cotton/steel spring construction used in furniture manufacture. The use of foam contributes to the comfort of the user resulting in high customer demand. Use of polyurethane foam also simplifies furniture construction, reducing costs. This combination of advantages has held down furniture costs to the homeowner and has resulted in almost complete conversion of the industry to this improved construction. The automotive industry uses substantial quantities of polyurethane foam, both molded and slab, not only to offer the customers the benefits of greater comfort and superior styling, but also to achieve the weight reductions necessary to meet mandatory government specifications on fleet fuel consumption. Soft foams with high resilience also provide lightweight padding to satisfy mandated crash protection requirements in automobile interiors. There are a number of properties of CFCs which make them ideally suited as blowing agents. They are odorless, nonflammable, have low toxicity and are nonreactive. These properties are particularly significant since blowing agents are released relatively rapidly from open-cell foams into the workplace enviroment during foam preparation, cutting, and curing. These properties are a major asset in maintaining a safe working environment. CFC blowing agents also help assure a consistent high quality in foam production. 11-18 ------- CFC Uses and Essentiality 3. Alternatives and Limitations The extent to which additional water may be used as an alternative blowing agent is severely restricted since it impairs the essential flexibility of the foam. Numerous studies have been undertaken to find other substitutes for CFC blowing agents in the manufacture of polyurethane foams. Methylene chloride has been used as a substitute in a number of polyurethane foam formulations. However, concessions must be made in the physical properties of the manufactured foam. Methylene chloride also requires special polyols and amine catalysts in almost all foam formulations. Users have said CFCs are "forgiving" blowing agents, meaning that less strict control is necessary versus methylene chloride, or alternatively, that less off-specification foam product is produced than is the case with methylene chloride. There is not the toxicity concern with CFCs that there 2 is with methylene chloride. The substitution of methylene chloride for chlorofluorocarbons, by some foam producers, has required additional ventilation for the curing tunnels in which the foam bun is formed to ensure safe working conditions for the foam-line operators. 2 The Threshold Limit Value (TLV) is a conventional measure of the maximum acceptable average exposure during a workday. CFC blowing agents with a TLV of 1,000 ppm have the highest (least toxic) value assigned to any chemical except carbon dioxide (a natural product of respiration) and simple asphyxiants such as nitrogen. The TLV for methylene chloride is 100 ppm or onetenth the value for CFC blowing agents. Current toxicological concerns could result in the allowable exposure to methylene chloride being lowered further in the future. 11-19 ------- CFC Uses and Essentiality Some polyurethane foam manufacturers have been able to substitute methylene chloride for CFC in certain high-density foam formulations, but not in all low-density formulations. (Low-density ("supersoft") foams are finding growing use in quality furniture manufacturing). Low-density polyurethane foams are substantially more difficult to make using methylene chloride and require extensively-modified formulations compared with foams using CFC blowing agents. Therefore, a direct across-the-board substitution of methylene chloride for CFC cannot be made with existing technology. 11-20 ------- CFC Uses and Essentiality G. BLOWING AGENT FOR POLYSTYRENE, POLYETHYLENE, AND PHENOLIC FOAMS 1. Description of Use When chlorofluorocarbon (CFC) blowing agent is blended with molten polystyrene resin and extruded through a die, the blowing agent vaporizes and a sheet of foamed polystyrene forms. After a short aging period, during which air permeates into the cells, the sheet is "thermoformed" (shaped by heat) in hot presses into the final product form. The thermoformed products are everyday items such as egg cartons, meat trays, produce and fast-food containers, disposable dinnerware, and many industrial containers. The containers are sanitary, nonabsorbent, esthetically attractive and readily decorated. Modifications of the process give molded and loosefill packaging, insulation products, and even a filler for lightweight concrete. Analogous processes use the same blowing agents to manufacture polyethylene foams, which are unmatched by any other packaging material for lightness and high compressive strength. Comparable techniques yield hard, porous foams from phenolic resins. The foams are used in diverse applications, from pipe insulation (where the inherent flame retardancy of the resin is required), to packaging for fresh-cut flowers. In 1979 approximately 40 million pounds of CFC were used in these applications, roughly 5 percent of total domestic CFC production. 11-21 ------- CFC Uses and Essentiality 2. Essentiality and Benefits CFC blowing agents are nonf lammable and have an ex- tremely low toxicity rating. Since substantial blowing agent is emitted during manufacture of these foamed plastics, these two properties are a major asset in maintaining a safe working environment. Ventilation requirements are reduced with attendant heating and cooling energy savings. CFC blowing agents have a unique combination of addi- tional properties. They are: Compatible with, and soluble in, resins for easy formulation and excellent control of the charac- teristics of the final foam. Inert, which provides nonreactivity with the resin and equipment during high-temperature extrusion phases. Excellent contributors to needed physical charac- teristics in the extruder and die, such as surface tension and viscosity. Inherently good nucleation agents - a term describ- ing the ready formation of numerous and uniform small bubbles throughout the expanding foam. Provide the excellent insulation properties needed in thermal insulation uses. (See Section II-K). 11-22 ------- CFC Uses and Essentiality 3. Alternatives and Limitations Two possible alternatives to CFC for blowing polystyrene foam are pentane and chemical blowing agents. Pentane blowing agent is used by some manufacturers. However, the flammability and explosion hazards of pentane must be minimized by explosion-proofing all equipment and rigorous elimination of static electricity. Costs of $500,000 per plant are estimated for ventilation and roll storage area alterations required for pentane use. Heating costs are estimated thereafter to increase 25 percent due to the higher ventilation required to limit fire and explosion hazards. The hazard to worker and property cannot be eliminated, as the history of pentane use clearly shows. Although pentane is cheaper than CFCs, basic material costs are nevertheless increased by pentane use since more resin per article must be used to match the properties of CFC-blown polystyrene foam products. Alternatively, chemical blowing agents could be used which decompose to give off nitrogen under the hot extruder conditions. Chemical blowing agents currently are expensive and limited to certain specific uses. The chemical blowing agent concentration necessary to produce light foam would raise product costs, making the product noncompetitive. Chemical residues are left from chemical blowing agents, and high residue levels create stability problems in the product. Information provided to EPA by Du Pont [Du Pont, 1978]. 11-23 ------- CFC Uses and Essentiality Low-density polyethylene foam and phenolic foams with high thermal insulation value cannot be produced today with other blowing agents. Elimination of these foam products would force a return to such alternatives as wood pulp products for packaging. Wood pulp products largely have been superseded due to the improved sanitary and esthetic properties of plastic foams. Wood pulp food containers absorb moisture and grease from food contents, a sanitary and esthetic disadvantage. One area where costs would clearly increase is in hospi- tals where the replacement of sanitary and germ-free disposable foodware with chinaware would add labor and energy costs assoc- iated with adequate washing and cleaning for reuse. 11-24 ------- CFC Uses and Essentiality H. FOOD FREEZANT 1. Description The liquid food freezant (LFF) process of freezing food consists of spraying the food with,or immersing the food directly in, liquid food-grade CFC-12, which boils at -22 degrees F (-30 degrees C) . This results in very rapid freezing. It is used only to freeze fragile, difficult-to-freeze, and relatively expensive food products. The primary products frozen in LFF are cob corn, french-sliced green beans, seafood (primarily shrimp and clams), berries, and small portions of meat. Vapor from the freezing bath is efficiently condensed and recycled by a secondary refrigeration unit. In 1979 approximately 10 million pounds of CFC-12 were used in this application, very roughly 1 percent of total domestic CFC production. 2. Essentiality and Benefits Freezing of food is an important and growing method of food preservation and an important alternative to canning and preserving with chemical additives. Approximately 7 percent (about 20 billion pounds per year) of all food consumed in the United States is frozen at some time in the distribution channel from production to storage, to transportation, to retail display cases, to consumers' tables. 11-25 ------- CFC Uses and Essentiality There are three basic methods of food freezing: Percent of Total Method Frozen Food Air Blast 88 Cryogenic (Liquid nitrogen or carbon dioxide) 10 LFF 2 Since LFF and cryogenic freezing of food are more expen- sive than air-blast freezing, they are used only when considera- tions of product quality, processing, or other advantages dictate their use. An example is frozen corn-on-the-cob. The recent retail availability of this product is due to the high-quality product made possible by the LFF process. 3. Limitations of Alternatives For specialty products where considerations of product quality, processing ease, and yield are important, either LFF or cryogenic systems can be used. Air blast could be used, but the quality of the frozen food product is reduced. Therefore, the LFF process always is compared and evaluated opposite cryogenic systems. The major advantages of LFF over cryogenic systems are: Due to the ability of LFF systems to recycle the freezant for repeated use (which is not practical in cryogenic freezing), there is a bottom-line savings in cost of freezing when using LFF. Typically, this amounts to approximately a three cent savings per 11-26 ------- CFC Uses and Essentiality pound of food frozen, depending upon the specific frozen food. Due to the recycle of the refrigerant, the use of LFF is more energy-efficient than cryogenic freez- ing. Products are crust-frozen almost instantly and, consequently, product dehydration is lower (almost zero) for LFF. This results in a cost savings and a quality advantage over cryogenic freezing. Products also are completely frozen more rapidly in LFF which results in a further quality advantage. (Slow freezing allows the formation of large ice crystals which damage the cell structure of the food; this results in the food becoming mushy when thawed). Small food items, such as shrimp or berries, can be "individually quick-frozen," since they are separated automatically when dropped into the boiling freezant. Other processes result in frozen clumps of the food items unless separated by hand prior to freezing - an expensive, labor-intensive step. 11-27 ------- CFC Uses and Essentiality I. STERILANT GAS 1. Description of Use Ethylene oxide (EtO) is an extremely effective sterilizing agent. However, its flammability, in the pure form, severely limits its use. Consequently, where the sterilizing ability of EtO is required but its flammability cannot be tolerated, EtO is mixed with an inerting agent, such as a chloro- fluorocarbon or carbon dioxide. The inerting agent most often used is CFC-12. The most common mixture of EtO and CFC-12 is 12 percent EtO and 88 percent CFC-12. The principal markets for the mixture are hospitals and manufacturers who prepare specially sterilized equipment and devices for pharmaceutical and medical use. It also is used as a fumigant and pesticide for granaries, warehouses, and ship cargo holds. In gas sterilization, a specific concentration of the gas is introduced into a specially constructed sterilizing chamber wherein temperature, humidity, and time of exposure can be readily controlled. These parameters must be predetermined for each sterilization cycle and are dictated by the nature of the items to be sterilized. Items commonly sterilized by the CFC/EtO blend include catheters, gloves,, syringes, tubing for anesthetic and respiratory units, and anesthetic and respiratory equipment (nebulizers, humidifiers, mouthpieces, manifolds). Sterilization of inhalation therapy equipment is a major use of the blend as this equipment has the potential for patient-to-patient contamination and infection. 11-28 ------- CFC Uses and Essentiality In 1979 approximately 15 million pounds of CFC were used for this application, very roughly 2 percent of total domestic CFC production. 2. Essentiality and Benefits Gaseous sterilization, in particular methods utilizing the CFC/EtO blend, has grown to be increasingly important in recent years. Materials of construction (such as plastics and elastomers) in many medical devices are not compatible with steam autoclave conditions, radiation, or chemical sterilization, thus creating a need for gaseous sterilization. Additionally, the use of sterilizing gas enables the sterilization of medical supplies and Pharmaceuticals after they have been packaged, thus eliminating any danger of contamination caused by handling in packaging. Other benefits of using the CFC/EtO mixtures for gas sterilization are: Flammability and explosion hazards are eliminated. CFC-12 is of such low toxicity that the toxicity of the blend is only one-twentieth that of pure EtO, an additional safety benefit for workers. Much lower pressure (60 pounds per square inch) in shipping containers than blends with carbon dioxide (750 pounds per square inch). The CFC/EtO blend can utilize lighter weight, less costly and more easily handled shipping and storage containers. 11-29 ------- CFC Uses and Essentiality Sterilizing cycles are shorter than those needed for blends with carbon dioxide. A constant composition is obtained whether the supply cylinder is full or almost empty. For blends with carbon dioxide, some fractionation occurs, which can lead to inconsistent sterilization. Sterilization at lower chamber pressures results in the need for less costly sterilizing chambers than 4 is the case for carbon dioxide/EtO blends. 3. Alternatives and Limitations Nonflammable blends of carbon dioxide and EtO can be prepared but have the major practical disadvantages discussed above. The use of CFC-12/EtO has grown rapidly because previous sterilizing techniques employing heat, steam or chemicals are now limited due to deleterious effects on certain instruments and equipment. Today, medical devices contain electronic, plastic, or easily damaged components. Thus, the gas sterilizer is becoming common in hospitals alongside autoclaves which previously handled most sterilizations. 4 The composition of the nonflammable blend with carbon dioxide is 10 percent EtO. Sterilizing effectiveness is proportional to the number of EtO molecules present. To achieve comparable sterilizing rates and effectiveness, the pressure necessary using the carbon dioxide blend is three times as high as the pressure using the CFC blend (due to the difference in partial pressures of CFC-12 and carbon dioxide). 11-30 ------- CFC Uses and Essentiality Aside from irreparably damaging some items, autoclave sterilization conditions (moisture and high temperature) may cause premature deterioration of items constructed from plastics and elastomers, resulting in higher replacement costs. Cold chemical sterilization (glutaraldehyde) has the limitations of long exposure times, inability to destroy all organisms, and poor efficacy. An item sterilized in a chemical bath must be removed, dried, and packaged. The rinsing and handling necessary in packaging can recontaminate the item. Radiation sterilization is limited by the cost of the apparatus (which may be hundreds of thousands of dollars) , the adverse effect upon certain elastomers and plastics, and the lack of information on exposure time and dosage for specific sterilization problems. Presently manufactured alternative fluorocarbons (not implicated in the ozone depletion theory) do not have suitable flame suppression properties or pressure characteristics. They cannot be substituted for the presently used CFC-12. 11-31 ------- CFC Uses and Essentiality J. INTERMEDIATE FOR FLUOROPOLYMER PRODUCTION 1. Description of Use In this application, CFC-22 is first chemically con- verted to the monomers, tetrafluoroethylene and hexafluoropro- pylene, which, in turn, are reacted to form fluoropolymers and fluoroelastomers. Both steps in this process are carried out in tightly sealed process equipment, from which emissions are negligible. The final products are non-volatile polymers from which CFCs cannot be regenerated. Neither the monomers nor the final polymers have any potential for ozone depletion. Fluoropolymers and fluoroelastomers are ultra-perfor- mance polymers which are used widely in industry in very harsh environments or very demanding service. Use of these materials is growing rapidly to meet new high standards for safety, pollu- tion control, equipment life, equipment utility and energy conservation. In 1979 approximately 60 million pounds of CFC-22 were used in this application, very roughly 7 percent of total domes- tic CFC production. 2. Essentiality and Benefits Fluoropolymers have the outstanding chemical, electri- cal, and high temperature properties, plus resistance to burning, needed by industry to meet new high standards for safety, pollu- tion control, and energy conservation, and to minimize cost through better utility and long equipment life. They are also used as replacement parts in the human body, in energy exploration and production, and have space program and military applications. In some cases fluoropolymers are irreplaceable. In many cases fluoropolymers could only be replaced by exotic 11-32 ------- CFC Uses and Essentiality metals or other special materials, or by a substantial sacrifice in equipment life, utility, safety and/or energy consumption. Since fluoropolymers are used so widely in industry, the overall cost to society of replacing fluoropolymers would be very large. The following list of Du Pont products provides an indication of the wide range and importance of fluoropolymers made from CFC-22: "Teflon" and "Tefzel" Resins Hexafluoropropylene Monomer "Viton" Fluoroelastomers "Kalrez" Perfluoroelastomers "Teflon" Film & Tubing "Teflon" Heat Transfer Products "Nafion" Products "Teflon" FEP Coated "Kapton" Film "Delrin" A/F Acetal/Fluoro- polymer "Armalon" Felts and Fabrics "Dulite" Finishes "Silverstone" Non-Stick Finishes "Teflon" Non-Stick Finishes "Zepel" Fabric Fluoridizer "Zonyl" Fluorochemical Surfactants "Krytox" Oils and Greases "Vydax" Fluorocarbon Telomers "Teflon" Fibers "Tefzel" Film Some uses for these products follow; Valve and pipe liners Packing Bellows Bearing pads Seals Thermoplastic compounds, industrial greases. Fibers, metal coatings, and impregnates (such as packings and glass cloth). II-3; ------- CFC Uses and Essentiality Rings Insulators Tape Thread seal tape Film, tubing. Heat exchangers Chemical equipment liners Moldings for semiconductor industry High performance wire insulation for the air- craft, computer, utili- ties, rapid transit, and nuclear industries. Membranes for the chlorine industry Roof structures Filtration (anti-air pollution) equipment 3. Limitations of Alternatives There are no acceptable alternatives for fluoropolymers in many applications, because fluoropolymers are chemically inert to virtually all chemicals, can be used at very high tempera- tures, and have good electrical properties. Where they could be replaced, the cost would generally be very high due to higher initial cost of available substitutes, reduced utility, decreased safety, or shorter equipment life. For example, wire insulated with "Teflon" resin is acceptable as a fire alarm cable and for use in plenums, but wires insulated with non-fluoropolymer must be put in conduits because they are not as heat-resistant. These savings make it easy to retrofit buildings to improve fire safety. Another example is the use of fluoropolymer lined pipe and vessels to replace glass-lined equipment which is more expensive and easy to break. In many cases, fluoropolymers have replaced much more expensive parts machined from exotic metals, with considerable reductions in energy consumption and cost. 11-34 ------- CFC Uses and Essentiality Despite extensive research on alternative routes to tetraf1uoroethylene and related fluoromonomers, there is no practical process for manufacture of tetrafluoroethylene and the related monomers other than the route based on CFC-22. 11-35 ------- CFG Uses and Essentiality K. CHLOROFLUOROCARBONS AND ENERGY CONSERVATION 1. Summary The use of chlorofluorocarbons (CFCs) in air-condi- tioning, refrigeration, insulating foam and other uses results in a large energy savings compared with systems using non-CFC alternatives. Based on a study by Battelle Columbus Laboratories for Du Pont, [Battelle, 1980 See Appendix C] if such uses of CFCs were banned and the next-best non-CFC alternative technology was forced progressively to replace CFC technology, the energy penalty would grow. Calculated in terms of gallons of fuel equivalent, the first year penalty of 847 million gallons increases to over 9.5 billion gallons at the tenth year and the total for the first decade is almost 50 billion gallons (See Table 1) . Prior to this study the value of CFCs in conserving energy resources had been neither thoroughly assessed nor quantified. The study concluded that a ban on the use of CFCs would have "an adverse and serious impact on an already serious energy problem." Although the study examines the effect of a ban on the use of CFC-11 and CFC-12, and a ban is not now contemplated, the Environmental Protection Agency has announced its intention to limit CFC production to current levels, and a longer-term interest in a phase-down of United States CFC production by 50-70 percent. Such regulatory action would affect all uses of CFCs, including 11-36 ------- CFC Uses and Essentiality Table 1 SUMMARY OF ENERGY PENALTIES FROM A CHLOROFLUOROCARBON BAN (MILLIONS OF GALLONS FUEL EQUIVALENT) [FROM BATTELLE, 1980] Using Next Best Non-CFC Alternative 1981 1990 Decade Automotive Air-Conditioning Home & Store Refrigeration Insulating Foams Liquid Food Freezing TOTAL (a) 161 446 231 9 847 1,070 5,267 3,171 12 9,520 6,870 27,461 15,365 106 49,802 (a) Includes losses due to outdoor compressor when using ammonia, and incremental effect of elimination of CFCs in both refrigerant and foams. 11-37 ------- CFC Uses and Essentiality those examined in the study. The energy penalty, to a first approximation, may be expected to be proportional to the percentage phase-down of CFC production. The above energy penalty would stem solely from cur- tailing the availability of CFCs for existing uses and the projected growth of these uses. The penalty as calculated did not include the impact of expanded CFC use in current applica- tions without growth in the application itself. As an example of this later type of penalty we note that the U.S. Department of Energy (DOE) recently published rule proposals [45 Federal Register 43976-44086] setting forth energy efficiency standards for household appliances. In its support documents [DOE, 1980], DOE concludes that meeting these standards would entail an increase in the use of CFCs. (This is discussed in more detail in section 4). So the energy penalty of CFC regulatory restriction would be tied not only to the impact of restricting CFCs from current applications but also from future expanded CFC applications. A related energy impact would stem from the unavail- ability of CFCs for new applications uses of CFCs that are only in their infancy or, in some cases, only on the drawing board. Several of these are discussed in Section VII-C. 11-38 ------- CFC Uses and Essentiality 2. General Conclusions and Methodology For each use studied, various alternative technologies were evaluated. Since only uses of CFC-11 and CFC-12 were studied, incorporating other uses and other CFCs (as EPA intends to do) would increase the penalty. As an example, domestic air-conditioning, which uses CFC-22, was not included. Liquid Food Freezing, which uses CFC-12, was not included as an example of the energy conservation for even minor CFC uses. The study noted that the energy penalty would undoubt- edly continue to grow for many decades beyond the first since displacement by obsolescence of the more efficient CFCs would not be complete for several decades in such uses as thermal insulating foam. In conducting the study, Battelle compared the energy consumption of CFC-using systems and equipment currently in use versus the best of several alternatives, even if such alterna- tives have not yet been demonstrated as commercially practical. Predetermined criteria were used, Such as equivalent safety-in- use, comparable cooling values and continued compatibility with existing space parameters. The actual penalty would be a composite of increases in gasoline, fuel oil, coal, nuclear and hydroelectric power, etc. To total the penalty, Battelle converted each penalty to a common denominator, equivalent gallons of fuel, assumed to have an energy content of 140,000 BTU/gallon. The tenth year total, 9.5 billion gallons, is almost incomprehensibly large. To aid in understanding, equivalents in national energy sources and consumption terms were calculated by Battelle (Table 2). 11-39 ------- CFC Uses and Essentiality Table 2 HOW MUCH IS 9.5 BILLION GALLONS? [FROM BATTELLE, 1980] The Battelle report describes this energy penalty of a hypothetical ban on CFCs in the tenth year as equivalent to: The fuel required to drive 12 million average cars (about 10 percent of all autos on United States roads) for one year. About 45 percent of current annual oil production from Alaska's North Slope. The energy (excluding gasoline) required to supply 11 cities of 500,000 population each for one year. The energy output of 29 nuclear plants. Eighteen times the petroleum savings envisioned by use of gasohol in 1978. 11-40 ------- CFC Uses and Essentiality Subsequently, in a report to the Environmental Protec- tion Agency (EPA), on the economic implications of CFC emissions, Rand Corporation [Rand, 1980] estimated the 1990 energy penalty for avoiding CFC use just in insulating foam at 6.4 billion gallons, or twice that estimated by Battelle, possibly reflecting the conservative assumptions used throughout the Battelle study. 3. Specific Applications a. Refrigeration and Automotive Air-Conditioning Ammonia or propane were selected as the non-CFC alternatives with least energy penalty. In each instance, complete redesign of equipment is necessary: heavier equipment is needed for the higher pressures, and special design consid- erations must avoid otherwise unacceptable toxicological and flammability hazards to building or vehicle occupants. Air cycle and absorption systems were disqualified due to even greater potential energy penalties. The non-CFC alternatives may prove economically imprac- tical, warned Battelle, even if the energy penalty could be tolerated. b. Insulating Foams CFC-blown closed cell foams provide the most efficient insulation available from today's technology. The low thermal conductivity of the trapped CFC vapor in the foam yields an insulation about twice as efficient as non-CFC alternatives, for instance fiberglass, for a given thickness. Foams blown with alternative blowing agents such as pentane or carbon dioxide may be comparable in appearance, but not in insulating performance. 11-41 ------- CFC Uses and Essentiality c. Liquid Food Freezing Certain delicate foods, for instance shrimp and berries, are too fragile for normal air blast freezing and must be frozen by direct contact with liquid CFC-12 or cryogenically with liquid carbon dioxide or liquid nitrogen. Since the CFC-12 freezant is efficiently recycled, substantial energy savings accrue compared with the cryogenic alternatives. Although liquid freezing is a minor application for CFC-12, it too plays its part in energy conservation due to CFC use. 4. U.S. Department of Energy Standards In June of 1980, the U.S. Department of Energy (DOE), responding to an act of Congress (PL 95-619), published proposed rules [45 Federal Register 43976-44086] setting forth energy efficiency standards for nine household appliances. These standards which will become final in February of 1981, necessi- tate improved insulation efficiencies, as well as other changes. Specifically, DOE states that replacement of fiberglass insulation with polyurethane foam is "cost-effective" and "technically feasible" in freezers and refrigerators; design option No. 1 states: "Fiberglass insulation is replaced with polyurethane foam insulation. Since polyurethane foam has a thermal conductivity of about one-half that of fiberglass, this option greatly reduces cabinet heat leak." [DOE, 1980, p. C-2] . For hot water heaters, DOE's No. 1 option is: 11-42 ------- CFC Uses and Essentiality Improved insulation - Improved insulation involves using thicker fiberglass insulation, denser fiberglass insulation, or substituting polyurethane foam insulation for fiberglass. This option may require changing jacket sizes." [DOE, 1980, p. C-6]. As stated earlier, CFC-11 is the only blowing agent used in producing insulating polyurethane foams. Some 10-15% of refrigerators, 10% of freezers and essentially all hot water heaters presently being manufactured use fiberglass insulation, so conversions will be necessary. In its accompanying Environmental Assessment Document DOE concludes: "An increase in the use of chlorofluorocarbons (CFCs) is expected as manufacturers seek to improve the insulating characteristics of refrigerators, refrigerator-freezers, freezers and water heaters. Compared to projected U.S. consumption of CFCs in 1990, however, these increases are expected to be small, representing less than 4% of projected U.S. consumption of CFCs in 1990." [DOE, 1980, p. S-4]. While the 4% figure cited above, of itself, is not large, it constitutes another specific instance of an upward pressure on CFC demand which has not been taken into account by EPA in its proposed production cap. Furthermore, it creates a classic "Catch 22" situation, wherein one Federal regulation runs head-on into another, with business firms caught right in between. 11-43 ------- CFC Uses and Essentiality L. SUMMARY Six major chlorof luorocarbons (CFCs) are manufactured in commercial quantity in the United States: CFC-11 (CC13F) CFC-12 (CC12F2) CFC-22 (CHC1F2) CFC-113 (C2C13F3) CFC-114 (C2C12F4) CFC-115 (C2C1F5) These compounds play an important role in the welfare of society and in the national economy. The most common CFC uses include: The heat transfer fluid in residential and almost all commercial refrigeration and air-conditioning. The heat transfer fluid in all automobile and truck air-conditioning. The foaming agent used to manufacture plastic foams, used for thermal insulation, cushioning and packaging. Cleaning agents for precision electronic and mechan- ical equipment, and also for military hardware. Fire and explosion suppressant for sterilizing gas in hospital and industrial uses. Liquid food freezant. Intermediate for f luoropolymer production 11-44 ------- CFC Uses and Essentiality An approximate distribution of total 1979 U.S. CFC production by end use follows: Use CFC (Million Pounds) Percent of total U.S. Production Refrigeration and Ai r-Cond it ioning Auto Air-Conditioning 270 105 33 13 Solvents 130 16 Blowing Agent for Polyurethane Rigid Foam 75 Blowing Agent for Polyurethane Flexible Foam 50 Blowing Agent for Other Foams Liquid Food Freezant 40 10 5 1 Sterilant Gas 15 Other (Miscellaneous Uses, Export and Use as Chemical Intermediate) 120 15 TOTAL 815 100 11-45 ------- CFC Uses and Essentiality CFCs are ideal for these uses because they have a unique combination of physical and chemical properties. These proper- ties result in very important benefits, including high safety- in-use, low energy consumption and high compatibility with other materials. Because the uses are dependent upon the specific proper- ties of the individual CFC employed, the compounds generally are not interchangeable among applications. Unlike the case with CFC aerosol propellants, the uses of CFCs currently at issue generally cannot be replaced with other compounds or processes without creating severe and often unacceptable tradeoffs. The most important limitations of avail- able alternatives are toxicity, flammability, energy efficiency, performance and, of course, economics. Consequently, any deli- beration on the need for, and degree of, potential restriction of CFCs should be undertaken in the context of their critical importance and the problems which would be created by their unavailability. 11-46 ------- III. LEGAL ISSUES Page A. INTRODUCTION 2 B. AUTHORITY TO REGULATE 4 C. ECONOMIC DISINCENTIVES REGULATION 18 D. RULEMAKING PROCEDURES 29 E. CONCLUSION 45 III - 1 ------- Legal Issues A. INTRODUCTION In the October 7, 1980 Federal Register, EPA published an Advance Notice of Proposed Rulemaking (ANPR) to outline its regulatory thinking and to solicit comment on the various approaches to regulation of non-aerosol uses of CFCs. To support its current initiative against CFCs, EPA is relying upon reports prepared in 1979 by the National Academy of Sciences [NAS, 1979a; 1979b] . No studies more recent than the 1979 NAS re- ports are cited by the Agency or listed as supportive of EPA's action. EPA's ANPR indicates that the CFC regulations will be promulgated under the authority of Section 157 of the Clean Air Act (42 U.S.C. §7450-7459), unless the Administrator deter- mines that it would be in the public interest to proceed under Section 6(c) (1) and Section 9(b) of the Toxic Substances Control Act (TSCA) (15 U.S.C. §2601). This regulatory effort by EPA raises matters of serious concern to Du Pont. This portion of our comments is divided into three different sections. In the first section, we question the Agency's legal authority to proceed with further regulation of CFCs. Specifically, we question whether there is an adequate scientific basis to regulate and whether EPA has sufficiently considered all of the scientific evidence before moving ahead with regulation. In the second section, we share with the Agency some of our concerns and perceived problems with the Agency's preference for economic incentives regulation, or more accurately, economic disincentives regulation. We point out areas where the Agency's legal authority is questionable and areas where it appears as though its analysis has not been very carefully thought out. We also discuss some very complicated questions related to the economic and competitive impact of the economic disincentives regulation outlined by the Agency. In the third section, we outline for the Agency the procedures it must follow before proposing a formal rule. Both the Clean Air III - 2 ------- Legal Issues Act and TSCA impose upon EPA the requirement to set up a rulemaking docket. In addition, there are certain further scientific studies which EPA must conduct before proposing a rule. And, there are considerable economic impact analyses which must be performed before proposing a rule. Particularly if EPA goes forward with economic disincentives regulation, these economic impact analyses become very important. For that reason, they must be extensive and they must be complete. Ill - 3 ------- Legal Issues B. AUTHORITY TO REGULATE 1. Findings to Support Regulation May Not Be Made In its ANPR, EPA singled out two statutes under which it could proceed should it determine that further regulation of CFCs at this time is desirable. One is Section 157 of the Clean Air Act, as amended (42 U.S.C. §7450-7459). The other is Section 6(c)(l) of the Toxic Substances Control Act (TSCA). (15 U.S.C. §2601 e_t seg.). Section 157(b) of the Clean Air Act provides, in pertinent part: [A]fter consideration of the research and study under Sections [153 and 154] of this title, ... the Administrator shall propose regulations for the control of any substance, practice, process or activity (or any combination thereof) which in his judgment may reasonably be anticipated to affect the stratosphere, especially ozone in the stratosphere, if such effect on the stratosphere may reasonably be anticipated to endanger public health or welfare. It would appear, therefore, that once the requisite studies under Sections 153 and 154 have been completed, and the results analyzed, EPA may regulate CFCs under 157 (b) jj: CFCs may reasonably be anticipated to affect the ozone in the strato- sphere, and if^ such effect may reasonably be anticipated to endanger public health or welfare. On the other hand, Section 6 (a) of TSCA provides, in pertinent part: If the Administrator finds that there is a reasonable basis to conclude that the manufacture, process, distribution in commerce, use, or disposal of a chemical substance or mixture, or that any combination of such activities, presents or will III - 4 ------- Legal Issues present an unreasonable risk of injury to health or the environment, the Administrator shall [regulate such chemical substance or mixture]. Thus, under TSCA, the Administrator has authority to regulate CFCs, if CFCs now present an unreasonable risk of injury, or if they definitely will present an unreasonable risk of injury to health or the environment. This finding is to be distinguished from the "may present an unreasonable risk" finding necessary to support testing under Section 4 of TSCA. Whereas, EPA need only find that a chemical may present an unreasonable risk to subject it to testing requirements, it must find that a chemical actually does or will present an unreasonable risk before subjecting it to a control regulation. As shall be demonstrated below, it is clear that, given the available scientific evidence, the findings necessary to support further CFC regulation under either Section 157(b) of the Clean Air Act, or Section 6 of TSCA, may not be made. Knowledge of the stratosphere is a science in its infancy. Although much already is known about the chemistry and physics of the stratosphere, and new results are being added at a rapid pace, much work remains to be done. As a result, any study of the stratosphere must not be viewed as the "last word" on stratospheric knowledge. Rather, it must be viewed as a "snapshot" a freezing of the stratosphere and what is known about it when the study was conducted.^ The NAS report of November, 1979 [NAS, 1979a] , was such a snapshot. Regardless of the validity of criticisms leveled against the report upon its release (overstatement of conclusions, under-estimates of uncertainties, missing chemis- try) , the fact remains there have been significant advances in the science since publication of the report advances which throw into further question the utility of the NAS report in III - 5 ------- Legal Issues supporting a regulatory determination. As an example, recent studies of the reaction of key chemical species indicate that the predictions of future ozone depletion made by the NAS were overstated by at least a factor of two. Another consideration is the computer models used in an attempt to simulate the atmosphere and to calculate the future ozone depletion numbers. It should be emphasized that the models employed by the NAS, and relied on to this day by EPA, are simplified (one-dimensional) mathematical representa- tions. Similar models were used to predict damage to the strato- sphere from SST aircraft. Such predictions were proved to be wrong based on subsequent scientific information. A one-dimensional model was also used by the research team employed by the United Kingdom's Department of the Envir- onment in its assessment of the issue an assessment roughly concurrent with the NAS effort. In the UK report the authors concluded: "These findings, together with other discrepancies between model calculations and measurements bring into question the validity of the models presently used to predict ozone perturbations." [UK DOE, 1979, p.5] A comparison of some of the principle findings of the November, 1979 National Academy of Science's report and the October, 1979 United Kingdom Department of the Environment's report appears as Appendix D. More recently (June, 1980) , a report by the Commission of the European Economic Community (EEC) concluded in part: "As they [the models] are simplified, they cannot fully describe the behavior of the atmosphere and its minority constituent parts." [EEC, 1980, p.8] III - 6 ------- Legal Issues We further note that at a recent (September, 1980) meeting between EPA, EPA's scientific consultants and representa- tives from the industry scientific research program, EPA was warned against placing too much reliance on the predictions of the one-dimensional models and using these predictions to make regulatory decisions. In the past year advances have been made toward developing more sophisticated two-dimensional models. These models more accurately represent the conditions found in the real atmosphere. The NAS report suggested that some of the known discrepancies between the one-dimensional models and 'atmospheric measurements would be eliminated by two-dimensional models. However, this has not been the case. Discrepancies remain. EPA should exercise extreme caution when basing regulatory decisions on models known to be incomplete and at variance with actual measurements. (These discrepancies are discussed in more detail in Section IV - The Science and in Appendix E). These recent changes in the chemistry, and continuing discrepancies between models and measurements, provide evidence that the theory of depletion of stratospheric ozone by CFCs has not been proved. The changes and discrepancies also indicate that there is substantial uncertainity surrounding the whole issue of the reliability of computer models as predictive tools. A further problem is that EPA has not founded its regulation upon actual measurements of the stratosphere. It is clear that Congress intended that any regulation promulgated be designed to protect the over-all stratospheric ozone, and that the regulation be based upon actual measurements of the ozone layer. in fact, it directed the Administrator to: III - 7 ------- Legal Issues rely upon reputable scientific and medical data and measurements from both the laboratory and the field (see H. Rep. No. 94-575, 94th Cong., 2nd Sess., part 1 at 14) before proposing regulations. Industry and its contractors have for some time been analyzing the measurements of stratospheric ozone taken over the past twenty years. This statistical and analytical technique is commonly known as "ozone time-trend analysis" or "ozone trend analysis". This technique, sensitive to an approximate change of + 1 to + 1.5 percent ozone concentration over a period of a decade, has detected no ozone depletion. EPA's rejection of these results, while failing to provide any countering analysis of its own, is in direct conflict with Congress1 intent that the scientific basis for CFC regulation be predicated on reputable measurements from the field. The above arguments illustrate the uncertain scientific basis of EPA's regulatory decision-making. As will be shown, it is clear that substantial uncertainties remain in the scientific theory. The computer models used by EPA have too much uncer- tainty associated with their calculations to make them reliable predictors of actual ozone depletion. (A more detailed discus- sion of the scientific uncertainties surrounding ozone and the 1 Time-trend analyses of ozone measurements are being per- formed by a number of independent groups. The studies and their results are discussed in detail in Section IV and Appendix E. These results show that there has been a slight increase in average ozone levels over the last ten years. The 95 percent confidence limits of these analyses are approximately + 1 percent to + 1.5 percent. In other words, with 95 percent confidence, a decrease or increase in ozone of 1 to 1.5 percent over a decade can be detected by these analyses. The range of sensitivity, i.e., the limits of the ability to detect a change in ozone, reflect the slightly different results from the applicable individual studies. Ill - 8 ------- Legal Issues impact of CFCs on stratospheric ozone is presented in Section IV and Appendices E and F). Moreover, CFCs are extremely useful chemicals. They are noncorrosive, nonflammable, virtually nontoxic and they have a unique combination of properties which make them ideally suited for use in a wide variety of products and processes. For these reasons, they touch the life of every American. They are used in food freezing, food transportation and storage, home, commercial and vehicle cooling, furniture, packaging, insula- tion, cleaning of precision components and fire retarding. It is apparent, therefore, that the findings necessary to support regulation under Section 157(b) of the Clean Air Act or Section 6 of TSCA, may not be made. A summary of the evidence shows that: a. The computer models used by EPA to predict depletion have too many uncertainties associated with them to be reliable predictors. Recent developments in many scientific areas throw into question the computer models' data base and conclusions drawn from it. There is simply too much uncertainty for the requisite finding of Section 157 or Section 6 to be made. b. The availability of analysis of actual strato- spheric ozone measurements (ozone trend analysis) provides an early warning system which would, if the results so indicated, give EPA the information necessary to proceed with regulation. EPA has not considered the ozone trend analysis results, in direct contravention with its statutory mandate to consider ozone measurements. (See Section IV and Appendix E for a more detailed discussion of ozone trend analysis and its significance.) Ill - ------- Legal Issues c. The evidence that increased solar ultraviolet radiation should ozone depletion occur as predicted -- would cause harm to human health or the environment is too uncertain to make conclusive inferences. (See Section V and Appendix F for a more detailed discussion of the effects of solar ultra- violet radiation). Again, there is too much uncertainity to declare conclusively that CFCs may reasonably be anticipated to endanger public health or welfare, or that CFCs present an unreasonable risk to health or the environment. Thus, EPA is without statutory authority to proceed with this rulemaking. 2. International Concerns EPA's expressed intent witn respect to exports is that exports will be subject to the conditions of regulation under the proposed rule. Under Section 12(a) of TSCA, EPA may not regulate chemicals which are manufactured for export from the United States. The only exception to this rule is if the EPA can find that the chemical will present an unreasonable risk of injury to health within the United States, or to the environment of the United States. Accordingly, before EPA may regulate CFC exports it must take an affirmative finding that these exports present an unreasonable risk of injury to health or the environ- ment in the United States. The issue of exports cannot be considered in a vacuum. It bears more broadly on the overall issue of the international concern surrounding ozone depletion and CFC regulation. Ozone depletion, if it occurs, is truly an inter- national problem. CFC emissions anywhere around the world contribute equally to the potential for ozone depletion. This ozone depletion in turn would affect all areas of the globe. As a result, to be effective in reducing potential ozone depletion, III - 10 ------- Legal issues CFC regulation must be international in scope. For reasons outlined below, EPA's unilateral action in the United States is unlawful under the Clean Air Act, TSCA and general principles of administrative law. (Non-legal aspects of the international character of this issue are taken up in Section VI). Congress recognized the international significance of the stratosphere and ozone when it passed the Clean Air Act amendments. Section 156 of the Amendments reflects this Congressional concern and provides that EPA and the President must "develop [international] standards and regulations which protect the stratosphere." It is thus apparent that EPA should regulate CFCs in this country only when the international consensus and basis for regulation is established. The requirement that regulation to limit ozone depletion be internationally based is corroborated in the legislative history of the Clean Air Act: In view of the worldwide impact of any ozone depletion and the fact that half of the world's halocarbon use is outside the United States, research efforts must be established and a base established for international or regional regulation, if it becomes necessary. (emphasis added) Sen. Rep. No. 95-127, 95th Cong., 1st Sess. at 64 (1975). In addition, it is questionable whether the Agency has the authority to promulgate regulations which have as much, if not more, benefit abroad as they do in the United States, yet at the same time force the United States industry to shoulder the entire burden of regulation. That is, it is questionable whether, in the absence of international consensus and agreement on regulatory steps, TSCA or the Clean Air Act gives EPA the authority to promulgate a regulation whose potential benefit is only minimally aimed at protecting U.S. citizens but whose burden will fall entirely on U.S. citizens. Ill - 11 ------- Legal issues Furthermore, promulgating regulations in this country (under the pertinent sections of TSCA or the Clean Air Act) which are designed to stimulate action by other nations is not valid regulatory action. The regulations must have as their purpose and effect the protection of humans and the environment against unreasonable risk. This is required by the legislative history of the Clean Air Act: [CFC regulation must] be necessary to assure protec- tion for health and the environment and to protect the stratosphere. H. Rep. No. 95-29, 95th Cong., 1st Sess. at 102, n. 2. Courts construing other environmental and health statutes have come to similar conclusions regarding an admini- strative agency's authority to regulate absent a showing of substantial or significant risk, and absent a showing tnat the proposed regulation is reasonably necessary to protect against the risk. In Industrial Union Department, AFL-CIO, et. al. , v. American Petroleum Institute, et. al. 48 U.S.L.W. 5022, US , (June 24, 1980) (hereinafter cited as the Benzene case), the Supreme Court held that the Occupational Safety and Health Administration (OSHA) is required to find that a significant risk to worker health exists before promulgating a standard seeking to reduce the risk, and that the standard will indeed reduce the risk to acceptable levels. At issue in Benzene was the interpretation of Section 3(8) and 6(b)(5) of the Occupa- tional Safety and Health Act. Section 3(8) permitted OSHA to promulgate standards that are "reasonably necessary or appro- priate to provide safe or healthful employment." OSHA attempted, under Section 3(8), to lower the workplace exposure threshold limit value (TLV) for benzene from 10 ppm to 1 ppm. Nowhere in the rule-making record was there an explicit finding that expo- sure to 10 ppm benzene presented a significant risk to human healtn, and likewise there was not a finding that a new standard 111 - 12 ------- Legal Issues at 1 ppm was reasonably necessary to protect against a signifi- cant risk. In striking down OSHA's standard, the Court said: We agree ... that [Section] 3(8) requires [OSHA] to find, as a threshold matter, that the toxic substance in question poses a significant risk in the workplace and that a new, lower standard is therefore 'reasonably necessary or appropriate to provide safe or healthful employment and places of employment' ... Congress in- tended, at a bare minimum, that [OSHA] find a significant risk of harm and therefore a probability of significant benefits before establishing a new standard. 48 U.S.L.W. at 5024 (1980). In Aqua Slide 'n' Dive v. Consumer Product Safety Commission, 569 F.2d 831 (5th Cir. 1978), the Fifth Circuit Court of Appeals held that the Consumer Product Safety Commission (the Commission) was required to show substantial risk and protection against the risk before regulating under Section 9 of the Consumer Product Safety Act (CPSA). Section 9 of the CPSA requires a finding that a new standard or regulation be "reasonably necessary to eliminate or reduce an unreasonable risk of injury" before the Commission may promulgate a new regulatory standard. The Commission had sought to require manufacturers of swimming pool sliding boards to post signs around the boards warning of possible harm from improper sliding into pools. Plaintiff sued, claiming that the warning signs were not, under Section 9 of the CPSA, reasonably necessary to protect against an ostensible unreasonable risk. The court agreed with the plaintiff, holding: In evaluating the 'reasonable necessity1 for a standard, the Commission has a duty to take a hard look, not only at the nature and severity of the risk, but also at the potential the standard has for reducing the severity or frequency of the injury.... 569 F. 2d at 844. 'Of course, no standard would be expected to impose added costs or inconvenience to the consumer unless there is reasonable assurance that the frequency or severity of III - 13 ------- Legal Issues injuries or illnesses will be reduced.1 Id., at 839 (citing H.R. Rep. No. 1153, 92nd Cong., 2nd Sess., 33 (1972)). As with the Aqua Slide situation, the EPA admits that the proposed CFC regulations will have little or no environmental benefit (45 Federal Register 66729). 3. Scope of Proposed Regulation The ANPR states that the regulation will cover all chemicals with the following general formula: CnC1xFyH2n + 2-x-y where x and y are each greater than zero (0) . (45 Federal Register 66728) . Under this general formula, EPA plans to include all alkanes that contain at least one chlorine and one fluorine atom. The proposed formula is too broad for three reasons: (1) it includes many CFCs which, were ozone depletion to occur as predicted, are considered part of the solution and not the problem; (2) it excludes some non-CFC compounds which, were ozone depletion to occur as predicted, are greater con- tributors to stratospheric chlorine than many CFCs; and (3) it includes even high molecular weight polymeric compounds with no potential for CFC emissions. The NAS report on the science of the ozone depletion theory [NAS, 1979a] was limited to the study of CFC-11 and CFC-12 and reached no supportable conclusions on the potential effect on the ozone of other CFCs. In fact, none of the assessments of potential future stratospheric ozone depletion has included any CFCs except CFC-11 and CFC-12. Furthermore, CFC-113, CFC-114, and CFC-115, all included under the Agency's proposed regulatory formula, have never been studied thoroughly opposite their potential for stratospheric ozone depletion. And III - 14 ------- Legal Issues CFC-22 has been given only a cursory examination. This was acknowledged by EPA in the ANPR (45 Federal Register 66728) . The Agency stated: "NAS also briefly considered CFC-22 (chlorodifluoro- methane) but excluded it from analysis because as a par- tially halogenated compound, its likelihood of reaching the stratosphere before dissociating is much less than that of fully halogenated compounds." (emphasis added). 45 Federal Register 66728. Not only is there little or no evidence to support including all CFCs in a regulatory formula, but by so attempt- ing, the Agency will hinder and possibly stop altogether the development of potential alternatives to those commercial CFCs theorized to create the greatest risk. For example, CFC-22, CFC-141b and CFC-142b all show promise as replacements for CFC-11 and CFC-12 in refrigeration equipment and blowing agent applications. Due to a different chemical structure, these three compounds all show less calculated potential than CFC-11 and CFC-12 for stratospheric ozone depletion. However, by including these in the regulatory scheme, EPA will severely reduce incentives for conducting the design and development necessary to make those chemicals viable alternatives to CFC-11 and CFC-12. So long as the potential alternatives are included in the cap proposal, users will be extremely reluctant to convert to their use lest those chemicals be more severely restricted in the future. (See Section VIII and Appendix B for a more detailed discussion of the research and development of CFC alternatives). The second problem with EPA's regulatory formula is that it excludes some non-CFC compounds which, were ozone depletion to occur as predicted, are greater contributors to stratospheric chlorine, and hence potential ozone depletion, than many CFCs. A good example of this is methyl chloroform. Ill - 15 ------- Legal Issues With the molecular formula, CH3 CC1, (1,1,1-trichloro- ethane), methyl chloroform does not fit within EPA's proposed formula. Yet under a true application of the permit pound concept (see Appendix G for a detailed discussion of the relative ranking of compounds by their potential for strato- spheric ozone depletion the permit pound concept), methyl chloroform is a greater potential ozone depletor than CFC-22. The NAS recognized this: "Atmospheric measurements indicate that methyl chloro- form is contributing between a quarter and a half as many chlorine atoms to the stratosphere as are CFC-11 and CFC-12. If it gains increased usage as a substitute for other solvents in degreasing and coating operations, it may well become the largest source of stratospheric chlorine." [NAS, 1979b, p. 45] Moreover, information on the potential effect of methyl chloroform on stratospheric ozone is available from an EPA sponsored conference [EPA, 1980b] specifically on this subject. The third problem with EPA's regulatory formula is that it includes even high molecular weight polymeric compounds if they contain any trace of chlorine (as defined, x >0 and n is unlimited), e.g., fluoropolymers. This is true even though such materials have no potential for CFC emissions. Accordingly, the formula proposed by EPA for this regulation is so broad that its proposal would exceed the Agency's statutory authority. (The question of regulatory scope is taken up in more detail in Appendix H). 4. Regulatory Focus In the ANPR EPA states that the problem associated with CFCs results from their emissions into the lower atmosphere III - 16 ------- Legal Issues which eventually make their way to the stratosphere (45 Federal Register 66726). As such, any control features implemented by the Agency must be designed to eliminate or reduce in some way CFC emissions. However, lacking from the ANPR's discussion is any consideration of those CFCs which, although manufactured, are never emitted into the atmosphere. The most notable example of this is that portion of CFC-22 production used as a monomer intermediate in the manufacture of fluoropolymers. These polymers have a variety of beneficial uses. (See Section II-J). Because the monomeric CFC used in their manufacture does not result in any emission, EPA is without authority to include this use of CFC-22 in any regulation. By including CFCs used as monomers in a proposed rule, EPA will have greatly magnified the impact of the regulation. Not only will the CFC using and producing industries be im- pacted, but also the fluoropolymers industries manufacturers and users. Thus, the Agency's regulation will impact the paint industry (where fluoropolymers are used as dispersants) , the coating industry (where fluoropolymers are used as coating insulation for wire and for plumber's tape), the cookware industry, the hardware and tools industry, the industrial fabrics industry, and a host of other industries where the unique characterisitics of fluoropolymers provide excellent products. Hence, EPA must exclude from regulation those CFCs which are used as monomers in the manufacture of fluoropolymers. Ill - 17 ------- Legal Issues C. ECONOMIC DISINCENTIVES REGULATION In the ANPR, EPA discusses several control options available to it for regulating CFCs. Several are traditional "mandatory control" options which include technology-based controls, as for example, the recovery and recycle of CFCs from flexible urethane or rigid non-urethane foam production using carbon absorption techniques, performance standards on solvent degreasing equipment, conversion to CFC-502 in retail food store refrigeration and the use of CFC-22 or helium as a test gas in refrigeration equipment manufacture. Other control options considered are selected product bans and a use ceiling combined with a significant new use rule (SNUR). The Agency also discussed what it calls "economic incentives regulation". And it is clear from the ANPR and from subsequent discussions with Agency personnel that some form of economic incentives regulation is the preferred mode for further CFC regulation. Economic incentives regulation, or more appropriately, economic disincentives regulation, would function by directly restricting CFC production. The restriction of production would come via some sort of "cap" on manufacture or on use. As demand for CFCs outstrips the available supply, the price of CFCs will rise. As the price rises, firms must decide whether to pay the higher price of CFCs, to switch to a less attractive alternative, to better conserve CFCs during use, or, barring ability to do any of these, to go out of business. According to the Agency, it is the forced use of these control options which may be avail- able to companies which makes economic disincentives regulation so attractive. (45 Federal Register 66730). However, there are very serious problems attendant to using economic disincentives regulation. (These problems are III - 18 ------- Legal Issues discussed in detail in Section VII and Appendix I) . Below we discuss legal concerns with economic disincentives regulation: 1. There is questionable authority under TSCA for EPA to implement such regulations; and the authority under the Clean Air Act is not at all clear or unambiguous; 2. EPA has not provided direction as to how it would implement such a program or how such a program would operate once implemented. We foresee significant equitable and implementation issues which must be addressed in such a program; 3. There are very complicated questions relating to the economic and competitive impacts of the proposals which must be resolved before EPA proposes a final rule. 1. Authority Section 157 (b) of the Clean Air Act authorizes EPA to promulgate "regulations for the control" of any chemical which ostensibly may harm the ozone. It would appear, therefore, that an economic disincentives approach would be authorized under the Clean Air Act if it could be considered a "regulation for the control" of CFCs. However, no cases have been decided under Section 157(b) and therefore we have no judicial interpretation which would shed any light on this issue. The relevant legisla- tive history, however, indicates that Congress did not intend to limit EPA to traditional mandatory control technology when implementing regulations to protect the stratosphere. See H. Rep. No. 95-294, supra, at 102, n. 2. Under Section 6 of TSCA, once the requisite finding of unreasonable risk is made (which, as noted above, cannot be made III - 19 ------- Legal Issues in this case) , EPA may then choose among the following regulatory options: 1. A ban on manufacturing; 2. A limitation on the amount manufactured; 3. A ban or limitation on manufacture of the chemical for a particular use; 4. A requirement to label the chemical and instruct customers of possible health hazards; 5. A requirement that manufacturers maintain records of manufacture and process; 6. A prohibition on any method of use of the chemical; 7. A prohibition on any method of disposal; and 8. A requirement that manufacturers give notice to distributors and to the public of unreasonable risk associated with the chemical. To be lawful, any economic disincentives regulation must fall under one of the above eight categories. In addition, any regulatory options under TSCA must be the least burdensome of the above options, but the option shall nonetheless be sufficient to "protect adequately against such [unreasonable risk]". Since economic disincentives regulation is essentially a limitation on the amount manufactured or used, a "cap" could conceivably fall under option No. 2 or No. 3 of the above eight. However, the cap, as proposed in the ANPR, is also a quota. Earlier draft bills of TSCA specifically authorized EPA to set quotas for regulating chemicals. Significantly, these provisions for a quota were deleted from the final bill. TSCA, therefore, does not provide explicit authority for EPA to impose quotas, and Congress1 deletion of the authorization for a quota from the final bill could mean that they did not intend for EPA to have the authority. Ill - 20 ------- Legal Issues It is interesting to note that the provision in the Senate bill providing for the establishment of quotas was removed at the recommendation of the EPA itself: Another difficulty we have with [the Senate bill] concerns the requirement that the Administrator provide for the assignment of quotas in any regulation limiting the amount of a substance which may be manufactured, imported, or distributed. The mandatory requirement of a quota system would make the regulatory process vastly more cumbersome and difficult to administer. Thus, we recommend that the quotaprovisionBedeleted. The act already provides that when it is necessary to adopt a rule with respect to a chemical substance to protect against an unreasonable risk, the Administrator shall select the least stringent requirement practicable, consistent with protection of health and the environment. In our view, restrictions limiting the amount of a substance that may be manufactured would be the most stringent requirement, other than a total ban, and the establishment of quotas would seldom be necessary. Nevertheless, we strongly recommend against becoming involved in the establishment of quotas Tor various manufacturers, even in such limited situations. (emphasis added).LetterfromJohnRT Quarles, Jr., Acting Administrator of EPA, to Sen. Warren G. Magnuson, Chairman of the Senate Committee on Commerce, June 23, 1975, reprinted in BNA, The Legislative History of the Toxic Substances Control Act, at 367. Though this letter from Administrator Quarles was referring to TSCA, its rationale and logic are equally appli- cable to EPA's authority under the Clean Air Act. A quota would make the regulatory process as cumbersome and difficult to ad- minister under the Clean Air Act as it would under TSCA. And the recommendation against becoming involved in the establish- ment of quotas under TSCA is equally compelling to becoming involved in the establishment of quotas under the Clean Air Act. Aside from the legal questions of whether EPA has the authority to implement economic disincentives regulation, there are other very serious legal problems surrounding the whole concept of this sort of regulation. One such problem is with the "permit pound concept" as described in the ANPR. Ill - 21 ------- Legal Issues The permit pound concept involves a complex series of calculations designed to establish a system which would allow those CFCs considered by the Agency to have the greatest potential for ozone depletion to be more heavily regulated than those with less potential for depletion. EPA's calculations allow that for every pound of one CFC (CFC-11) that can be manufactured or used, up to 5.6 pounds of different CFCs may be manufactured or used. The relative ranking advanced by EPA is correct. However, the absolute values assigned by EPA are in- correct and misleading as to the relative depletion potential of the various CFCs, and reflect the Agency's basic lack of knowl- edge about the ozone depletion potential of CFCs other than those specifically studied by the NAS. (See Appendix G for a more detailed discussion of the permit pound concept). As noted earlier, EPA is without statutory authority to regulate CFCs based on the current state of scientific knowledge. And, if evidence validating the theory is obtained, EPA is without authority to regulate any CFC other than CFC-11 and CFC-12. However, should EPA persist in this regulation and should it persist in including in the regulation CFCs other than CFC-11 or CFC-12, it must consider implementing some form of the permit pound concept. But the Agency must use a more accurate permit pound ranking. (See Appendix G for a more accurate ranking). 2. Implementation and Operation There are also legal questions surrounding the design, implementation and operation of an economic disincentives regulation as outlined in the ANPR. According to the Agency's ANPR, economic disincentives regulation would be implemented by means of a "cap", either directly on CFC users or on CFC producers. That is, the Agency III - 22 ------- Legal Issues would place a ceiling on the total amount of chlorofluorocarbons that may be used by an individual CFC user, or on a total amount that may be produced by a producer. The Agency outlined two different methods of distri- buting the rights to available CFCs under a "cap". One is by direct allocation to the users and producers. The second is by means of an auction, whereby permits to produce or use CFCs are auctioned by the government to the highest bidder. EPA will encounter serious problems with both approaches. a) Direct Allocation. Under a direct allocation program, EPA would be required to promulgate a TSCA Section 8 (a) rule or use some other information gathering tool to gain data from the CFC producing and CFC using firms. This data would then be used to establish a government allocation system. In the ANPR, EPA discusses production as the benchmark by which this allocation will be made. EPA must also consider whether it would be more equitable to use another factor such as production capacity, total investment in CFCs, or some other measure by which to determine an allocation program. A direct allocation program will essentially freeze producers' or users' market shares to the historical base used in the allocation program. How will the Agency justify the freezing of market shares perhaps the most drastic form of market interference available to the government? Another problem with a direct allocation program concerns new entrants into the market. How does the Agency propose to deal with them? Will it take one company's produc- tion or use rights from it and give them to the new entrant? On what basis will this determination be made? Will the Agency allow new production or use rights for the new entrant? Or, as yet another alternative, will EPA merely allow the new entrant III - 23 ------- Legal issues to purchase rights from existing companies? The Agency must provide guidance to the industry on this issue. EPA must also consider the problem of a company dropping out of the market. What will happen to the drop-out's production or use rights? Will they be redistributed among the existing producers or users? On what basis will this be done? Will the withdrawing company be permitted to sell its leftover rights? If so, will that sale be binding on the government during the next round of distributing CFC rights? Or suppose a company has a major problem (such as a plant breakdown) and is unable to produce all of its allocation. Will it be entitled to carry over its production rights until the next permit period? Finally, under a direct allocation program EPA must consider the length of time a distributed production or use right will be effective. To allow for rational planning, businessmen will need a minimum of ten years. If it chooses a shorter time span, EPA must explain why its choice is preferable, legally and economically, over one which minimizes the impact on business planning. b) Auction. A second option to EPA is the auctioning of available CFC production or use rights. Under such an auction EPA would establish a limited quantity of CFC rights and sell them to the highest bidder. At this auction, only producers could be invited, only users, or both producers and users, depending on the system EPA sets up. The price EPA would obtain for the rights would be the highest price it could possibly obtain. The price would therefore reflect what businesses thought it was in their own economic interest to pay for the CFC rights. We note first of all that a true auction, where the price paid for the CFC right is a reflection of the III - 24 ------- Legal Issues businessman's profit incentives, would be unlawful under the Independent Office Appropriation Act (31 U.S.C. §483a (1970)) and cases decided under it. There are other legal problems attendant to an auction. How often would the auction be held? Will it include foreign companies as well as U.S. companies? An auction would create property rights in the permits auctioned off. A question arises as to who owns these permits. Does the government own them? Do the producers own them? Do the users own them? If the government owns them, will this constitute a taking without due compensation in violation of the Fifth Amendment? What will EPA do with the money it receives in an auction? How will the Agency monitor participation at the auction? As with a direct allocation program, these questions on the auction must be answered before EPA proposes a formal rule. EPA must also consider that a futures market may develop for auction permits. If such a market does develop, how will EPA deal with it? Will it require registration with the Securities and Exchange Commission? Will it be subject to coverage under the 1933 and 1934 Securities Acts? What is the nature of this coverage? If it is not subject to such coverage, will it be subject to scrutiny under another securities law? Is the creation of a futures market with the potential for attrac- tion of non CFC-producer or user speculators consistent with the premise that the auction will allow the more essential users to obtain required amounts of CFCs? These are all questions which EPA must answer in a proposed rule. In addition, there is no guarantee that under a cap those industries in which CFCs are essential will get all the product they need. The Agency merely assumes that the essential industries will be able to afford the higher prices for CFCs. Yet, EPA provides no support for its assumption, nor III - 25 ------- Legal Issues has it analyzed the price elasticity for the various uses to determine if the assumption is valid. For example, has EPA considered the national security implications of CFC regula- tion? CFCs are used in a variety of ways in our national defense effort, for example, in cleaning of missile guidance systems. Has the Agency considered a means of assuring con- tinued availability of CFCs to the defense industries, parti- cularly if it turns out that those industries cannot obtain what they need, when it is needed? 3. Competitive Impacts As the largest domestic producer of CFCs, Du Pont is quite concerned about the potential for anticompetitive effects from economic disincentives regulation. EPA may not dismiss these concerns. Economic disincentives regulation on the scale being proposed by EPA may force a disruption in the marketplace in a way never seen before by American business. As mentioned earlier, under a direct allocation system there is a problem in distributing production or use rights to any potential new market entrant. In its ANPR, EPA states that it will handle this by allowing existing producers or users to sell their shares to new entrants (45 Federal Register 66730). Guidelines from EPA are needed to insure that such a system would operate equitably. Further, under such a system a means to adequately compensate producers who would be required to decrease production would have to be established. Under an auction system, would non-CFC producing or using companies be allowed to purchase permits and thereby disrupt the planning of the CFC industry? What factors will EPA consider in balancing the benefits of competitive bidding, while at the same time insuring that existing producers and users of CFCs are able to obtain and utilize allowed CFC production most efficiently? Ill - 26 ------- Legal Issues A further problem with a cap and its accompanying artificial shortage is that spot shortages of CFCs are likely to arise. Has the Agency considered a way to handle spot shortages, especially if a critical industry is affected? Under a no growth in demand market, businesses would not be likely to maintain excess capacity. Thus, the problem of spot shortages would become very real if no excess in capacity is maintained. EPA should provide an incentive for the maintenance of suffi- cient capacity and should establish procedures for any short- term increase in CFC capacity needed to meet essential demands. EPA must also be aware of the antitrust decree to which the five producers are now subject. This decree provides, among other things, that: a. Each CFC manufacturer is required to sell to anyone who is engaged in the business of reselling refri- gerant gas and who satisfies the manufacturer's customary credit requirements. b. In the event a manufacturer has insufficient refri- gerant to supply all resellers who seek to purchase the refrigerant, the manufacturer "shall, in such circumstances, determine unilaterally and without consultation with any other [manufacturer] or any groups of purchasers of refrigerant gas, in a manner in which demand shall be met on a basis of any allocation, reasonable and equitable under all the circumstances ...." This decree, of course, will have a very substantial impact upon any governmentally-imposed allocation system. It directly limits a producer's discretion in designing his own system to allocate scarce refrigerant gas. This decree runs contrary to any allocation system which would be designed by EPA to imple- III - 27 ------- Legal Issues ment an economic disincentives regulation. Has EPA consulted with the Department of Justice on this issue and on other issues raised in the comments to determine the scope of its authority in regard to these sensitive areas of its regulatory proposals? Ill - 28 ------- Legal Issues D. RULEMAKING PROCEDURES An ANPR is a relatively new procedure of administra- tive law. It is designed to allow the Agency to solicit comments in a forum where the Agency has not yet committed itself to one particular regulatory action. Since an ANPR is essentially an information-gathering tool, extensive rulemaking procedures need not be followed before publishing one. However, before EPA publishes a proposed rule, there are certain procedural requirements it must follow. Many of these are outlined below. 1. General - Clean Air Act Section 307 (d) of the Clean Air Act, which applies to subtitle B of subchapter I (relating to stratosphere and ozone protection), as noted in Section 307(d) (1) (h) , requires publica- tion of a notice of proposed rulemaking, accompanied by a statement of basis and purpose. This statement of basis and purpose must, at a minimum, include a summary of: a. The factual data on which a proposed rule is based; b. Methodology used in obtaining the data and in analyzing the data; and c. The major legal interpretations and policy considerations underlying the proposed rule. This statement must also contain a summary of the findings, recommendations and comments of EPA's scientific review committee and the National Academy of Sciences. If EPA's proposed rule differs in any important respect from the recommendations of these scientific bodies, an explanation of the reason for such differences must be included in the III - 29 ------- Legal Issues statement. All data, information and documents referred to and relied upon by the Agency shall be included in the docket as of the date of publication of the proposed rule. The docket should begin with the publication of the ANPR and should include all documents underlying it. The docket should also include all comments on the ANPR which the Agency receives as well as letters from Congress and from other interested citizens concerning EPA's proposed Phase II regulation. During rulemaking, EPA must afford interested parties the opportunity for the oral presentation of data, views and arguments in accordance with Section 307(d)(5) and must keep such oral record open for thirty (30) days after the hearing to allow an opportunity for submission of rebuttal and supplementary information. Implicit in this is the opportunity for any interested party to cross-examine any witness. 2. General - Toxic Substances Control Act Under Section 6(c) of the Toxic Substances Control Act (TSCA), before EPA may issue a proposed rule, it must publish a statement with respect to: a. The effects of CFCs on health and the magnitude of exposure of human beings to CFCs; b. The effects of CFCs on the environment and the magnitude of the exposure of the environment to CFCs; c. The benefits of CFCs for the various uses and the availability of substitutes for such uses; and d. The reasonably ascertainable economic consequences of the rule, after consideration of the effect on III - 30 ------- Legal Issues the national economy, the environment, and public health. EPA's proposed rule must contain the following information: a. A draft finding that there is a reasonable basis to conclude that the manufacture, processing, distribution in commerce, use or disposal of CFCs, or any combination of such activities presents or will present an unreasonable risk of injury to health or the environment; and b. A document which states with particularity the reasons for the proposed rule, together with a statement of why the proposed rule protects adequately against the risk(s) involved using the least burdensome requirements authorized by TSCA. 40 C.F.R. &750.2. In light of the "international motivation" of EPA's pending regulation, and the fact that unilateral action by the United States will have little or no measurable environmental benefit, EPA must explain why a cap or other regulation "ade- quately protects against the risk" involved, and why adoption of the Assessment and Surveillance strategy is not a viable alter- native. The Agency must set forth its reasons for regulating, and if the reason is stimulation of foreign regulation, EPA must explain why stimulation of foreign regulation is a proper motivation for regulation in this country. And finally, the proposed rule must contain a state- ment with respect to the four issues mentioned in Section 6(c) of TSCA, i.e., the effects of CFCs on health, the effects of CFCs on the environment, the benefits of CFCs for the various III - 31 ------- Legal Issues uses and the availability of substitutes for these uses, and the reasonably ascertainable economic consequences of the rule after considering the national economy, small business, innovation, the environment and public health. The proposed rule must also contain an analysis of the impacts of alternative courses of regulatory action. 40 C.F.R. &750.2. If the Administrator of EPA decides to proceed with Phase II regulation under TSCA instead of the Clean Air Act, it must in the proposed rule include a brief statement describing such findings. This statement must discuss, at a minimum: a. All relevant aspects of the risk; b. A comparison of the estimated cost of complying with actions taken under TSCA and under the Clean Air Act; and c. The relative efficiency of actions under TSCA and the Clean Air Act to protect against the risk. 40 C.F.R. &750.2(5). These statements may be combined in the same narrative for efficiency of exposition, but they must contain a discussion of the factual, analytical, policy and legal considerations behind the Agency's decision to issue the proposed rule in the form chosen. All factual materials and each analytical metho- dology seriously considered shall be fully disclosed. Signifi- cant areas of uncertainty known to the Agency under each heading shall be identified and the manner in which the Agency intends to deal with them shall be specified. Ill - 32 ------- Legal Issues In light of the substantial uncertainties surrounding the ozone depletion theory, EPA must address in its proposed rule the research projects it has undertaken to help resolve some of the uncertainties. Compliance with the research provisions of the Clean Air Act may suffice in this respect. This is discussed in Section 3 on the next page. Under TSCA, EPA must in its proposed rule provide for an informational hearing at which any person may submit oral comments. This hearing shall come not less than two weeks after close of the public comment period. Under Section 6(c)(3) of TSCA, if there are disputed issues of material fact which will require resolution, as is certainly the case when dealing with the highly uncertain ozone depletion theory, EPA is required to allow cross-examination of witnesses. This is necessary for a full and true disclosure with respect to the controversial ozone depletion theory. Finally, any rulemaking docket in any proposed rule must contain the results of the additional research which EPA must conduct in accordance with the Clean Air Act (discussed below) and the results of the further economic and regulatory impact studies which EPA must do (discussed below). 3. Research A reading of part B, subchapter I, of the Clean Air Act amendments leaves one deeply impressed with the Congres- sional concern that the stratospheric ozone problem be researched thoroughly. Both the statute itself and its legislative history indicate quite clearly that EPA is commanded to look at the stratospheric ozone problem more broadly than just the potential effects of CFCs. For example, in Section 150 of the Act, the Congressional Declaration of Purpose, Congress delineated four purposes in passing the Clean Air Act amendments relative to the stratosphere: III - 33 ------- Legal Issues a. To provide for a better understanding of the effects of human actions on the stratosphere, especially the ozone in the stratosphere; b. To provide for a better understanding of the effects of changes in the stratosphere, especially the ozone in the stratosphere, on the public health and welfare. c. To provide information on the progress of regula- tion of activities which may reasonably be antici- pated to affect the ozone in the stratosphere in such a way as to cause or contribute to endanger- ment of the public health or welfare. d. To provide information on the need for additional legislation in this area, if any. The legislative history echoes this general concern for the stratosphere as a whole, and not just the effects CFCs may have on it: [It] should be pointed out that the strato- spheric ozone layer is also threatened by other sub- stances [than halocarbons]. Of course, the prob- lem of emissions of pollutants from high flying air- craft is well-known. There could also be ozone deple- tion caused by: bromine compounds, other sources of chlorine than halocarbons, or oxides of nitrogen from fertilizers or combustion of fossil fuels. H. Rep. No. 94-1175, 94th Cong., 2nd Sess., at 76. While it is true that halocarbons (halocarbons is a term which includes, but is not restricted to, CFCs) were in- cluded in the potential problems listed by Congress, it is clear that EPA was directed to study the issue broadly. Ill - ------- Legal Issues Other sections of part B of the Clean Air Act amend- ment confirm this Congressional concern. For example, in Section 153(a), EPA is directed to study the effects not only of fluorocarbons on the ozone, but also of: a. The release into the ambient air of other sources of chlorine; b. The uses of bromine compounds; and c. The emissions of aircraft and aircraft propulsion systems employed by operational and experimental aircraft. Section 153(a) further directs EPA to study any physical, chemical, atmospheric, biomedical or other research and monitoring data which may be necessary to ascertain any direct or indirect effects upon the public health and welfare of changes in the stratosphere, especially ozone in the stratosphere. Under Section 153(b), EPA is required to undertake research on: a. Methods to recover and recycle substances which directly or indirectly affect the stratosphere, especially ozone in the stratosphere; b. Methods of preventing the escape of such substances; c. Safe substitutes for such substances; and d. Other methods to regulate substances, practices, processes, and activities which may reasonably be III - 35 ------- Legal Issues anticipated to affect the stratosphere, especially ozone in the stratosphere. Section 153 (d) requires EPA to direct the National Academy of Sciences to study, not just CFCs, but the effects of all substances, practices, processes and activities which may affect the stratosphere, especially ozone in the stratosphere. Under Section 154, EPA is directed to receive infor- mation from the National Oceanic and Atmospheric Administration on detection of changes in the stratosphere and the climatic effects of such changes. Section 154(b) requires EPA to obtain from the National Aeronautics and Space Administration infor- mation on the physics and chemistry of the stratosphere for early detection of potentially harmful changes in the ozone in the stratosphere. Section 154(c) directs EPA to ask the National Science Foundation for information which will increase scientific knowledge of the effects of changes in the ozone layer in the stratosphere upon living organisms and ecosystems. Section 154 (d) requires EPA to obtain information from the Department of Agriculture which will increase scientific knowledge of the effects of changes in the ozone in the stratosphere upon animals, crops and other plant life. Section 154(e) requires EPA to obtain information from the Department of Health and Human Services (formerly the Department of Health, Education and Welfare) which will increase scientific knowledge of the effects of changes in the ozone in the stratosphere upon human health. It would have been helpful for EPA to have discussed these studies in the ANPR. At this time, it would be beneficial for the Agency to publish a supplement to the ANPR discussing these studies. This supplement could also explain how these reports were used in EPA's decision-making process. Ill - 36 ------- Legal Issues And very importantly, Section 156 requires EPA, in conjunction with the President, to undertake to enter into international agreements to foster cooperative research which will complement or augment studies or research done in this country. To date, EPA's compliance with these various sections of the Clean Air Act has been notably inadequate. The Agency has received studies only on CFCs and those studies have been flawed, as discussed in Section IV and Appendices E and F. There have been no thorough studies on the effects other chlorine-containing compounds may have on the ozone. The Agency has done insufficient research on methods of recovery and recycling CFC emissions; they have done insufficient research on methods of preventing the complete emission of CFCs; the Agency's research on substitutes for CFCs has been woefully inadequate. Indeed, if EPA had done adequate research on sub- stitutes, it would have realized that many are harmful chemicals and the harm resulting from them is more immediate to persons potentially exposed than is the harm theorized to result from CFCs. The Agency's international activities have concen- trated on proclaiming the issue resolved and underscoring the need for regulation. There has been little or no effort by EPA to obtain an international consensus on the science or an inter- national resolution to conduct more research to more completely resolve the scientific uncertainties, as required under Section 156. Accordingly, before EPA may publish a proposed rule, it must undertake efforts to obtain international resolution of the science so that it will have before it a more complete data base upon which to make responsible regulatory decisions. Ill - 37 ------- Legal Issues 4. Economic and Regulatory Impact Analyses As noted earlier, EPA's preferred form of regulation is an economic disincentives approach, i.e., an allocation or auction of production or use rights in conjunction with a cap on CFC production or a cap on CFC use. Because such a system is an untested form of regulation, its effects upon the economy and individual businesses are extremely difficult to predict at this time. Therefore, EPA must conduct an extensive and detailed economic analysis to help answer, among others, the questions posed throughout these comments. In particular, EPA must study how a cap would be implemented and how a cap would operate under practice. Vague generalities in the ANPR are insufficient. In addition, EPA must also conduct an analysis of the long-range impacts of a cap, both in terms of impacts upon the economy and in terms of impact upon the affected industries. These requirements in Sections 157 and 317 of the Clean Air Act, and in Section 6 of TSCA, are discussed in more detail below. a) Clean Air Act Section 157 of the Clean Air Act requires EPA to "take into account the feasibility and the costs" of regulations promulgated under its authority. The legislative history elaborates on this requirement: By [requiring EPA to take into account the feas- ibility and costs of regulation, Congress] intends to assure that any such [regulation] is undertaken only with adequate awareness of its costs and its other economic impact and social impacts. This informal awareness is necessary for [EPA] in determining what combination of stratospheric pro- tection measures are most appropriate. (emphasis added) H. Rep. No. 94-1175, 94th Cong., 2nd Sess., at 81. Ill - 38 ------- Legal Issues The language of the bill which House Report No. 94-1175 accompanies, H.R. 10498, (referring to EPA's requirement to consider the feasibility and costs of regulation) was adopted in toto in the bill finally passed by Congress and signed by the President. Thus, Section 157 requires not only an economic impact analysis designed to allow more informed decision-making on the part of EPA and more informed participation by the public, but also an analysis of the social impacts of CFC regulation. The social impacts cannot be underestimated. CFCs touch the life of every American. Despite the Agency's assertions to the con- trary, regulation will affect some very basic elements of American life, such as refrigeration and air-conditioning. Given EPA's long-term goal to cut CFC production by 50-70% [Jellinek, 1980a] , it is likely that in the near future U.S. citizens may have to do without some of the luxuries and even necessities we have come to enjoy in the twentieth century. Hence, EPA's Section 157 "feasibility and costs" analysis must include an analysis of the social impacts the regulation will have upon American life. This requirement of a social impact study is reinforced in other sections of the relevant laws. (See e.g., Sections 2 (c) and 6 (c) (1) of TSCA.) Section 317 of the Clean Air Act requires EPA to conduct an economic impact assessment before publication of the proposed rule, and this assessment must be made part of the public record. Included in the assessment must be an analysis of: 1. The cost of compliance, and the cost of alternative regulations (one alternative is no regulation at this time); 2. The inflationary or recessionary potential of regulation; in - 3y ------- Legal Issues 3. The effects on competition with respect to small business; 4. The effects on consumer costs; and 5. The effects on energy use. The importance of the Section 317 economic impact analysis cannot be overstated. It strikes at the very heart of the essential nature of CFCs and at the reasons CFCs present no unreasonable risk to health or the environment. One does not have to be an economist to recognize that as CFC prices rise, inflationary pressures on the economy will result. The smaller businesses which cannot afford the higher costs of CFCs will be forced out of business. Consumers will have to pay higher prices for those articles which contain CFCs. (These issues and others are discussed in more detail in Section VII Economic Considerations and Appendix I) . The economic impact studies are important also because of the revolutionary, landmark form of regulation EPA is planning to adopt. Were the economic studies incompletely done and the impacts of the proposed economic incentives regulation not fully appreciated, EPA could cause a great deal of unintentional harm to the U.S. economy. Finally, CFCs are great energy savers. Their regulation will result in what at times may amount to drastic energy penalties. (See Appendix C for details on the energy penalty, and Section II Uses and Essentiality for a discussion of the energy saving properties of CFCs). It is pertinent to note that the recent Department of Energy (DOE) energy saving guidelines [DOE, 1980] demand an increase in the use of CFCs. Energy conservation is, of course, a high priority national concern. The DOE's requirements guidelines run contrary to EPA's proposed CFC regulation. The III - 40 ------- Legal Issues conflict between the two is very real. Given the Agency's long-term strategy to cut CFC production by 50 to 70 percent, there can be no denying that CFC regulation will result in a substantial energy penalty. EPA merely assumes that those industries with the most energy efficient use of CFCs will be able to pay the higher price for CFCs. Yet without having done a price elasticity study of the CFC-using industries, EPA cannot support its assertion. EPA must, therefore, as part of its social and economic impact analysis, quantify the energy penalty from CFC regulation and must explain how this penalty is offset by benefits of at least a comparable magnitude. We again emphasize the importance of the Section 317 analysis containing an assessment of the social impacts of CFC regulation. It must be made clear to the public exactly what they are giving up in return for this ostensible protection of health and the environment. Congress's purpose in adopting Section 317 was: to assure that the Administrator and the public would have before them adequate analyses of the economic impacts of alternative courses of action or inaction under the Act. The availability of such information is expected to help shape wiser policy and to permit the public to participate in rulemaking in a more informed and effective way. H. Rep. No. 95-294, 95th Cong., 1st. Sess. at 51 (1977). The American people must be made aware of the cost of CFC regulation. b) Toxic Substances Control Act Section 6(c) and (d) of TSCA requires EPA to consider the benefits of CFCs, the feasibility of substitutes, and the economic impact of CFC regulation. In particular, EPA must III - 41 ------- Legal Issues study the feasibility and the consequenses of using substitutes for CFCs. TSCA Section 6(c) and (d) also direct EPA to consider the impacts CFC regulation will have on small business and on technological innovation. Furthermore, Section 2(c) requires that EPA consider the economic and social impact of any action the Agency takes. Again, as in Section 317 of the Clean Air Act, the requirement of a social impact analysis - of how regulation will affect American life - is required. There are other regulatory impact analyses which EPA must conduct. Some of these are discussed below. c) Executive Order 12044 Under Executive Order 12044, signed by the President on March 23, 1978, and appearing in 43 Federal Register 12661, EPA is required to conduct a regulatory impact analysis of CFC regulation. Under 12044, CFC regulation must be designed in such a way so as not to impose unnecessary burdens on the economy, on individuals, on public or private organizations, or on any state or local government. To comply with Executive Order 12044, EPA must publish a statement which contains, at a minimum, findings that: 1. The proposed regulation is needed; 2. The direct and indirect effects of the regulation have been adequately considered; 3. Alternative approaches to the regulation have been considered and the least burdensome of the acceptable alternatives has been chosen; and 4. Public comments are being considered and an adequate response has been prepared. Ill - 42 ------- Legal Issues In addition, the regulation must be written in plain English and must be understandable to those who must comply with it. It must contain an estimate of the new reporting burdens necessary for compliance with the regulation and it must contain a plan for evaluating the regulation after its issuance. The statements must also include an analysis of the effects regulation will have on competition, and a detailed explanation of the reasons for choosing one method of regulation over others. In conclusion, EPA must explain why CFC regulation is necessary, in light of its negligible environmental impact, and why stimulation of foreign regulation is an appropriate goal of U.S. regulation. d) Regulatory Flexibility Act EPA must also conduct an impact analysis in compliance with the recently passed Regulatory Flexibility Act, Public Law 96-354. The regulatory impact analysis required by this Act must include, at a minimum: 1. A description of the reasons why action by the Agency is being considered; 2. A succinct statement of the objectives of, and the legal basis for, the proposed rule; 3. A description of, and an estimate of, the number of small entities to which a proposed regulation will apply; 4. A description of the proposed reporting, record- keeping and other compliance requirements of the proposed rule, including an estimate of classes of III - 43 ------- Legal Issues small entities which will be subject to the requirement and type of professional skills necessary for the preparation of the report or record; and 5. An identification, to the extent practicable, of all relevant Federal rules which may duplicate, overlap or conflict with the proposed rule. The regulatory impact analysis must also discuss any significant alternatives to the proposed rule which would accomplish the objectives of the underlying statute but which would minimize any "significant economic impact" of the proposed rule. This analysis of the alternatives must include, at a minimum, whether the Agency should consider: 1. The establishment of differing compliance or reporting requirements for small entities; 2. The clarification, consolidation and simplifica- tion of compliance and reporting requirements for small entities; 3. The use of performance rather than design stan- dards ; and 4. The possible exemption from the rule of small entities. Ill - 44 ------- Legal Issues E. CONCLUSION It is clear from this section that further CFC regulation at this time is neither appropriate nor supportable. Not only has EPA failed to fulfill its legal responsibility before regulating, but regulating such an essential chemical in the face of uncertain science is unwise policy. Moreover, the type of regulation EPA is contemplating has been neither well thought out nor well conceived. Before any proposed rule can issue, there must be substantial additional studies and research done to support such an action. Ill - 4b ------- IV. THE SCIENCE A. INTRODUCTION 2 B. THE CHLOROFLUOROCARBON/OZONE DEPLETION THEORY 7 C. OZONE MEASUREMENTS AND OZONE TREND ANALYSIS 15 D. EPA'S ANPR ASSESSMENT OF THE THEORY 22 E. PRESENT STATUS OF THE THEORY 32 F. RESOLUTION OF UNCERTAINTIES 51 G. SUMMARY 56 IV-1 ------- Science A. INTRODUCTION The Chlorofluorocarbon/Ozone Depletion Theory postulates that CFCs, which are emitted to the atmosphere after use in a variety of applications, are eventually transported to the upper atmosphere (stratosphere). The high stability of the compounds has been assumed to preclude reaction in the lower atmosphere (troposphere). At higher altitudes, however, the intensity of ultraviolet light increases and the compounds are broken down by photolysis, presumably releasing chlorine atoms. Chlorine is involved in the very complicated series of reactions comprising stratospheric chemistry. Those reactions include one pair which constitutes a cycle by which chlorine can catalyze the destruction of ozone (a form of oxygen). Ozone shields the surface of the earth from ultraviolet radiation which might lead to deleterious effects if the amount reaching the earth's surface were significantly increased (see Section V). Other chemical reactions lead to temporary "holding tanks" for chlorine and to removal from the stratosphere. Scientists have developed computer programs to model the processes involved in order to estimate the net effect of CFCs. The models are used to calculate the time-varying chemical com- position of the atmosphere. The results are then often compared with actual measurements of the concentration of various chemical species at different vertical and horizontal locations. The use of models to "predict" the effect of CFCs on ozone has been accepted widely because ozone depletion, if it occurs, is a slow, gradual process extending decades into the future. Accordingly, any change or trend in ozone over the near-term would be small and, presumably, difficult to detect. Most models used to calculate or simulate the atmosphere are one-dimensional (1-D). In such a model, the earth's atmo- sphere is averaged and the only movement of chemical species in IV-2 ------- Science the model is in the vertical direction. 1-D models are limited inherently by their treatment of atmospheric dynamics, their inability to deal with seasonal atmospheric variations, and other simplifications of the atmosphere during their construction. Accordingly, it is essential to verify that measured and calculated profiles of chemical species agree in the present-day atmosphere to give assurance that any assumptions and simplifi- cations incorporated into the model do not distort reality. By its very nature, the theory is a rapidly changing compilation of state-of-the-art science. The relatively young field of atmospheric science is hard-pressed to eliminate the many large uncertainties involved. Nonetheless, the potential for significant changes in the amount of ozone has prompted the Environmental Protection Agency to consider further regulation of CFCs in addition to the 1978 ban on the use of CFCs as aerosol propellents. Justification is cited as the 1979 assessment of the science by the National Academy of Sciences [NAS, 1979a]. Unfortunately, this assessment came during a period of rapid changes in the science, which continues to develop at a very fast pace. Current assessments indicate a large decrease in calcu- lated potential effects, and continue to emphasize the many uncertainties underlying the issue. In the following discussion of the science, we first provide an historical review of the theory and the several scientific assessments which have been done. Next, we present a detailed discussion of an area which has been consistently overlooked by EPA and which argues strongly and convincingly against the need for immediate regulatory action-the real-world experimental information provided by over 20 years of ozone mea- surements. IV-3 ------- Science Unlike the theoretical treatment provided by models, available measurements provide a record of 'actual ozone concen- tration. Statistical analysis of this record is capable of iden- tifying even very small (increases or decreases) in the average ozone concentration. The use of these analyses as an early warning of actual changes in ozone provides the ideal complement to the theoretical treatment in the models, and permits continued investigation of the theory in the absence of detectable change in ozone concentrations. The sensitivity of the technique provides confidence that changes, if they occur, will be detected sufficiently early to allow for effective responses by the appropriate governmental bodies. Even allowing for the calculated "overshoot", before any potential ozone changes were reversed, potential effects of ozone change can (even if the theory proves to be correct) be limited to very small levels. The importance of such an early warning system cannot and should not be underestimated. As noted above, the science of the atmosphere is rapidly evolving, and contains major uncertainties. Therefore, following the discussion of ozone measurements, we present an analysis of the approach to the science chosen by EPA in its Advance Notice of Proposed Rulemaking. The theory is inadequately and often inaccurately discussed. Uncertainties and caveats concerning these uncertainties are downplayed or ignored. The progress of "Overshoot" is the term applied to the theoretical maximum ozone depletion calculated by the models to occur after CFC emissions cease. The theory assumes that CFCs already present in the lower atmosphere at the time emissions were to cease would still be transported to the stratosphere, causing ozone depletion to increase to a peak value before slowly returning to normal levels. This peak value is calculated to be about 1.5 times any existing ozone depletion at the time CFC emissions cease. IV-4 ------- Science the science beyond that included in the NAS report is entirely neglected, treating one of the most active years of recent research as if it were unimportant. The EPA view is then contrasted, with a discussion of recent developments and the remaining key uncertainties in the theory. These uncertainties were consistently deemphasized or underestimated by the NAS, and virtually ignored by the EPA. A variety of recent evidence suggests that CFCs may not, in fact, be completely transported to the stratosphere with no destruction in the lower atmosphere. Any destruction represents a decrease in ozone depletion from that currently calculated by atmospheric models. The transport mechanisms themselves are not well understood and are treated only in an empirical fashion in atmospheric models. Yet calculated ozone depletion is sensitive to the rate and mechanism of transport. Stratospheric chemistry is also critical to the theory. As an example, recent reaction rate changes alone reduce esti- mates of calculated depletion to about half that reported by the NAS. Those changes represent large advances in our knowledge, but at the same time have greatly magnified perceptions of the inherent uncertainty in model results. Finally, we show that the models themselves are sources of uncertainty. Previous assessments have relied almost exclusively on one-dimensional models which neglect known atmospheric variations with latitude and season of the year. Further, the models have been largely restricted to consider the CFC effect in isolation from other current changes, such as increases in carbon dioxide (C02) When tested against other known perturbations, such as volcanic eruptions, the models fail IV-5 ------- Science to duplicate measured effects. Furthermore, the models do not calculate an accurate picture of the present stratosphere. Major discrepancies remain even in the most up-to-date treatments, and these discrepancies involve the chlorine species which are of so much current concern. Even in the extensive discussions below, many points may not be discussed in sufficient scientific detail for the well-informed reader. Appendix E will provide a useful reference source with a much more detailed discussion of the scientific evidence. IV-6 ------- Science B. THE CHLOROFLUOROCARBON/OZONE DEPLETION THEORY 1. General Description of the Theory The earth's atmosphere to an altitude of about 50 km is divided into two regions the troposphere or lower atmosphere and the stratosphere or upper atmosphere. The two regions are separated by the tropopause, which varies in altitude from about 8 km at the poles to about 16 km at tropical latitudes. In contrast to the troposphere, where turbulence and rapid vertical mixing occurs, the stratosphere is relatively quiescent. Ozone (03) is the most important trace constituent of the stratosphere. It is formed predominately at altitudes between 25 km and 35 km in the tropics, where short wavelength solar radiation dissociates molecular oxygen atoms which combine with molecular oxygen to form ozone [Chapman, 1930] . °2 + nv (solar radiation) *0 + 0 (1) 2 [0 + 02 + M*-*03 + M] (2) Net: 3 02~>2 03 *M is any other molecule Although ozone is produced in the tropics, highest ozone concentrations are found in polar regions at altitudes of about 15 km, as a result of air motions in the stratosphere. The production of ozone is currently assumed to- be relatively insensitive to man's activities. Ozone strongly absorbs solar radiation in the longer wavelength radiation region 240-320 nm. °3 + hv * °2 + ° IV-7 ------- Science It is this absorption that shields the earth from ultraviolet radiation. In this process, ozone is not destroyed since nearly all the oxygen atoms produced recombine with molecular oxygen to produce ozone (Reaction 2). Although the reactions just described are the predominant stratospheric production process for ozone, several competitive destruction processes exist. The amount of ozone in the stratosphere is maintained by a dynamic balance between production and destruction processes. The most important destruction process, one that destroys nearly 70 percent of the ozone produced, is a catalytic cycle involving nitric oxide (NO and nitrogen dioxide (NO-). NO + 03> N02 + 02 (4) N02 + 0» NO + 02 (5) Net: 0 + 03 * 2 02 Because jet aircraft engines exhaust oxides of nitrogen, the N0-N02 catalytic cycle formed the basis for concern in the early 1970s that supersonic transports flying in the stratsophere would deplete the ozone layer [Johnston, 1971]. Stolarski and Cicerone [1974] and Wofsy and McElroy [1974] suggested that the chlorine cycle may be a destruction process for stratospheric ozone, Cl + 03 » CIO + 02 (6) CIO + 0 » Cl + 02 (7) 0 + 03 * 2 02 but only natural sources of chlorine, e.g., volcanoes, were considered. The same year, Molina and Rowland [1974] published IV-8 ------- Science their suggestion that chlorofluorocarbons (CFCs) provide a significant chlorine input to the stratosphere. In its simplest elements, the suggestion was that: CFC have a long atmospheric lifetime, CFCs diffuse to the stratosphere, CFCs are decomposed by ultraviolet radiation to produce chlorine atoms, and Chlorine atoms gradually reduce the concentration of ozone in a catalytic cycle (Reactions 6 & 7). The authors did not make an estimate of the amount of ozone depletion by CFCs. 2. Model Calculations - What They Are; Why They are Needed If Molina and Rowland's suggestion the ozone depletion theory were true, significant effects of CFC on the ozone layer only would occur decades into the future. As a result, scientists have tried to estimate the magnitude of the effects by use of time-dependent, one-dimensional (1-D) diffusion models of the atmosphere, which are really complicated computer programs that attempt to represent mathmatically the atmosphere of the earth. These use chemical and photochemical reactions as input data and largely ignore detailed atmospheric dynamics. Some important points to note in reference to computer model calculations are: Ozone depletion has not been measured. All estimates of ozone depletion are computer model IV-9 ------- Science calculations which depend on the model methodology and assumptions, and the accuracy and completeness of the input data. High estimates of ozone depletion, e.g., 16 percent, are calculated to occur over a hundred years in the future and are based on working assumptions described blow which are not likely to occur, even if the theory were true. Computer models do not yet represent or simulate adequately the present atmosphere. The predictive reliability of computer models is unknown. These points and others will be discussed more fully in section E, "Present Status of the Theory". 3. Previous Scientific Assessments of the Theory A number of scientific reports by both U.S. and European scientific bodies have appeared which reviewed the status of the Chlorofluorocarbon/Ozone Depletion Theory. The most striking aspects of the reports is that they do not agree as to the validity of the theory, i.e., a scientific consensus that the theory is valid does not exist. On the contrary, they emphasize the dynamic and uncertain nature of the science as viewed at different times by different people. 1976 The first in-depth scientific assessment of the possible effects of CFCs on the ozone layer was the National Academy of Sciences (NAS) report, "Halocarbons: Effects on Stratospheric IV-10 ------- Science Ozone" [NAS, 1976]. The report concluded that continued release of CFCs to the atmosphere at 1973 levels would produce gradual ozone depletion in the range of 2 percent to 20 percent about 100 years in the future with a most probable value of about 6 percent to 7.5 percent. However, the report acknowledged a substantial lack of information, and concluded: "Additional improvements in our knowledge of the atmo- sphere and of stratospheric chemistry are essential to permit more accurate assessments to be made of the extent of potential reductions in the stratospheric ozone." (emphasis in the original) [NAS, 1976, p. 20]. It should be emphasized that these values of ozone depletion were calculated with a 1-D model. They were not measured values. 1977 A subsequent report by the National Aeronautics and Space Administration [NASA, 1977] was published the following year. One significant finding of the report was that a preliminary (now confirmed) fast rate for the reaction H02 + NO increased calculated ozone depletion by CFCs from about 7.5 percent to 15.0 percent (Lawrence Livermore Laboratory model) .[NASA, 1977, p. 192]. The report emphasized the need for additional research to better understand stratospheric science. 1979 Two important reports were published in 1979 within one month of each other the NAS report, "Stratospheric Ozone Depletion by Halocarbons: Chemistry and Transport", [NAS, 1979a] and the United Kingdom Department of the Environment report, "Chlorofluorocarbons and Their Effect on Stratospheric Ozone" [UK DOE, 1979]. Although the two reports used similar models, IV-11 ------- Science similar input data, and stated very similar values of calculated ozone depletion, they reached opposite conclusions with regard to the validity and rejLjLa^bjLjLj.t.yo^tjigc^a^jL£Lil.^^ted depletion vis-a-vis actual, real-world ozone depletion. The United Kingdom report placed low confidence in the reliability of the calculations, while the NAS report expressed high confidence in them. A brief quotation from each report is illustrative and representative: U.K. Report [UK DOE, 1979, p. 194 and p. 6] "It is not, therefore, realistic to assign overall uncertainty limits to our calculated ozone pertur- bations; deficiencies in our basic knowledge of the processes establishing the composition of the stratosphere and in the modelling technology cast doubts on their validity.... The report concludes that present understanding of ozone depletion is limited and is based on model assumptions which have not been adequately validated." NAS Report [NAS, 1979a, p. 1] "The uncertainty range means that for the case of continued release of CFMs [chlorofluoromethanes, e.g., CFC-11 and CFC-12] at the 1977 level there is a 1 chance in 40 that the ozone depletion will be less than 5 percent and 1 chance in 40 that it will be greater than 28 percent.... Although there are a few exceptions the comparisons between the models and measurements of substances in the present stratosphere is considered to be satisfactory within the uncertainties of the measurements. We, therefore, believe that the projections for ozone depletion are valid within the stated uncertainty ranges." A more detailed comparison of quotations from the NAS and U.K. reports appears as Appendix D. IV-12 ------- Science 1979 A report published in 1979, "The Stratosphere: Present and Future", [NASA, 1979] is a detailed review of stratospheric science. Two items in that report bear comment. First, the results of a modeling exercise conducted by NASA are given in this report. Ten modeling groups were asked to perform several calculations using standard input data. Although the models differed in some respects, "most models employed similar values for many of these parameters, and this [was] reflected in the agreement between the different model results." Not surprisingly, the different groups all calculated about the same amount of ozone depletion at steady state [NASA, 1979, p. 340], It must be emphasized that this agreement among the several modeling groups does not enhance or increase the validity of the calculations as claimed often by EPA. All the models are constructed basically the same way, and all used very similar input data. That the results are similar was an expected result, and demonstrates primarily the absence of computational errors. Consensus on the model output does not imply consensus on either model completeness, correctness of the input data, or validity of the results. The second key item in the report is the discussion on the potential for the detection of ozone changes or trends in the record of ozone measurements. The report states [NASA, 1979, p. 325] that the "percent thresholds for detecting changes in ozone globally in 10 years using the best statistical methods and Dobson network data" are: "+3.6 For detecting a true change in global average total ozone after allowing for uncertainty due to statistical error, instrument drifts and spatial sampling biases." IV-13 ------- Science "+5.7 For detecting a CFM effect after allowing for errors above plus possible variations due to other anthropogenic changes." However, the above thresholds were obtained by subjective esti- mate of the likely size of the uncertainty. It is important to be aware of the subjectivity of these estimates when one compares them with more recent results based on acutal ozone measurements. NASA acknowledges the very high uncertainty of the estimates with the statement [NASA, 1979, p. 321] (original emphasis retained), "It should be emphasized that the values given for the individual standard errors in estimating the threshold for trend detection are, for the most part, quite soft." 1980 The most recent scientific assessment of the ozone depletion theory is the report of the Commission of the European Communities to the Council of the European Communities, "Chlorofluorocarbons on the Environment" [EEC, 1980]. This report was based, in part, on the analysis of the NAS and U.K. reports by Brasseur [1980], The main conclusion of the Commission report [EEC, 1980, p. 7] is noteworthy: "The foregoing analysis shows the need for further research. But if the requisite decision is delayed, the likely effect of CFCs may be greater and the con- sequences more serious. The British report shows that by extending the date of an assumed total cessation of CFC emissions from 1 January 1979 to 1 January 1983, the maximum amount by which the total quantity of ozone would be depleted would increase from 0.5 percent to 0.6 percent since, according to the model, the interval between the cessation of emissions and the maximum 0., depletion lies in the range 7 to 15 years. Con- sequently, a delay of 5 years before any decision is taken on CFCs can be reasonably accepted." IV-14 ------- Science C. OZONE MEASUREMENTS AND OZONE TREND ANALYSIS 1. Overview Ozone concentration has been measured regularly at a number of locations (Dobson stations) around the world for varying lengths of time. Several stations have been in operation for over 20 years. Most estimates up to and including the 1979 NAS report concluded that although the system of measuring stations provided useful information regarding ozone concen- tration, the natural and experimental fluctuations precluded detection of small trends, unless the accuracy, precision, and internal consistency of the system of stations were improved. Such improvements were expected to require several years. o The statistical method of time series analysis had been applied to the measurements, but historically lacked the scope and treatment necessary to verify and reduce the sensitivity of the method. Nonetheless, the work was very promising, requiring only the development time necessary for any new application. This development has been pursued with fruitful results -- only now receiving the attention they merit as an important contribution to the science concerning ozone depletion. The current status of the technique is summarized below: 2 A time series is a set of measurements in time such that each measurement may be related to the previous or several previous measurements, e.g., daily temperatures. Time series analysis is a statistical technique to analyze the time series for patterns, cycles, or trends and to evaluate them quantitatively. IV-15 ------- Science Time series analyses of ozone measurements (varyingly called time-trend analysis, ozone trend analyses, or OTA) completed after publication of NASA [1979] and the NAS [1979a] reports conclude that ozone increased by a small, but statistically insignificant amount over the period 1970-1978. The very high ozone concentrations measured in 1979 raise that increase to a statistically significant level, at the same time demonstrating the sensi- tivity of the technique to the trend in the data. The studies establish 95 percent confidence limits of +_ 1.0 percent to 4; 1.5 percent on the trend estimate. The 95 percent confidence limits will narrow with each subsequent year of data. The studies imply ozone trend analysis can serve as an early warning system. It is important to understand exactly what is meant by these results. Clearly they could be explained if the ozone depletion theory is erroneous. The absence of a trend can also be explained if the net ozone trend from all recent effects is less than the threshold of detection, a situation in which two or more effects partially offset each other. Statistical trend ana- lysis, therefore, does not unequivocally contradict the ozone depletion theory. For example, increases in atmospheric carbon dioxide may cause increases in ozone levels. The rapidly increasing sensitivity of ozone trend analyses must be included in the regulatory decision-making process. The results of trend analysis should be given as much IV-16 ------- Science importance as the model predictions of ozone depletion them-_ selves. Ultimately, the regulatory issue is whether ozone is being depleted or not, not what is predicted by computer models programmed around numerous assumptions. Throughout the Chlorofluorocarbon/Ozone Depletion Issue, scientists and regulators have focused attention on calculated theoretical steady-state ozone depletion rather than experimental ozone measurements themselves. The reasons for this unusual situation are: a). If the theory is correct, significant ozone depletion will occur only gradually decades in the future. Accordingly, it was considered necessary to use computer models to estimate the magnitude of the depletion that might occur. b) . The theoretical depletion calculated to have occurred already was thought to be about 2 percent (recently revised to about 1 percent). The scientific community virtually assumed that such a change could not be detected by the Dobson network. 2. Detail Hill and Sheldon [1975] first applied time series analysis to Dobson measurements taken from 1932-1970 at Arosa, Switzerland, and used this technique to predict ozone values at Arosa in the period 1971 to 1974, and make qualitative estimates of ozone levels later in the decade. Subsequently, Hill _et al. [1977] and Pagano and Parzen [1975] applied ozone trend analysis (time series analysis applied to ozone records) to a larger ozone data record. In their most recent work, Hill et al. [1977] and Tiede et al. [1979] pointed out that no evidence of ozone depletion exists, and concluded that a trend of 1.5 percent per IV-17 ------- Science decade could be detected by this technique. The trend could be due to "one or a combination of man-related activities, long-term natural trends, or instrument drifts." More recent work, since the publication of the NAS report, also agrees that no ozone depletion has occurred. Reinsel ejt al. [1980] analyzed Dobson ozone measurements from 36 stations worldwide. Statistical, instrumental, and geograpical errors were considered. They found that the global trend estimate for total ozone change was (0.28 _+ 1.4) percent for the period 1970-1978. The uncertainty limits of this analysis (+ 1.4 percent) are much smaller than those given in NASA [1979]. St. John [1980a; 1980b] used a somewhat different approach and obtained results very similar .to those of Reinsel et al. St. John reports that his analysis of 14 Dobson stations with records from 1958 through 1978 show a change of (0.3 + 1.2) percent through 1978. More recently, St. John e_t al. , [1980] have extended the analysis of 14 Dobson stations to include 1979 data. An average trend of (+ 1.5 + 1.0) percent is found for the period 1959-1979, with the method responding as it should to the high 1979 values. (See below). Neither the Reinsel et al. nor the St. John study identifies the cause of the observed effect, i.e., no ozone depletion. At this time, any trend -- up or down detected by the method could be caused by natural forces, man-made perturbations, or a combination of the two. In addition, Angell and Korshover's [1980] studies have also shown the lack of ozone depletion. "The 1979 data suggest the highest global total-ozone value since 1970, or a value a significant 2 percent above average." IV-1S ------- Science Angell [1978] has reported on the analysis of ozone measurements in the 32 km - 46 km region, a part of the so-called Umkehr data, where the greatest percent change in ozone is calculated to occur. His analysis showed a 12 percent increase in ozone from 1964-1972. Photochemical theory at the time suggested about a 5_ percent decrease should have occurred. More recently, Angell and Korshover [1980] updated the report saying: "Thus, there is still no evidence of an anthropogeni- cally-induced decrease ^n ozone in t,his sens^t^ve layer." (emphasis added). In effect, the Dobson total column ozone measurements and the Umkehr stratospheric layer ozone measurements indicate that ozone has increased in concentration. An important difference is that the Dobson data have been subjected to time series analysis while the Umkehr data have not. It is important to note that the statistical analyses of Dobson measurements have uncertainty limits (95 percent confidence) for detection of any trend in the range of + 1.0 percent to + 1.5 percent. For example, St. John £t al. , [1980; St. John, 1980a; 1980b] shows that above the 14 stations used in the analysis, the best trend estimate for ozone from 1958-1979 is + 1.5 percent. The 95 percent confidence limits of +_ 1.0 percent mean there is only 1 chance in 40 the increasing trend is less than + 0.5 percent(+1.5 -1.0), and 1 chance in 40 it is greater than 2.5 percent(+1.5 +1.0). Umkehr measurements record ozone concentration in several distinct altitude layers of the stratosphere rather than the more common total column ozone measurements which measure the ozone in a column of air extending from the surface of the earth to the upper limits of the earth's atmosphere. IV-19 ------- Science The sensitivity of the method suggests it appropriately may be used as an early warning system. If a small ozone depletion can be detected, e.g., -1.5 percent +_ 1.0 percent (a range of -0.5 to -2.5 percent change in ozone), steps can be taken quickly to reduce emissions so that, in this example, maximum depletion would not exceed 2.25 percent to 3.75 percent, largely below the 3.6 percent detection threshold value suggested in NASA [1979]. (These values follow from the "overshoot" concept. The 1.5 "overshoot" factor times -1.5 percent and -2.5 percent gives -2.25 percent and -3.75 percent, respectively). Finally, Tiao [1980] has reported on preliminary studies to show that the Dobson network does represent a true global average of ozone measurements. The study involves analysis of the 1970-1977 Nimbus 4 satellite ozone data. The global average ozone measurements from the satellite show a negative trend of 0.5 percent per year which has been attributed to drift in the satellite's instrumentation [Cunnold, 1980; Stolarski, 1980]. The known but unquantified drift in the instrument prevents derivation of a trend in actual ozone from the satellite data. However, Tiao also shows that the average trend in ozone measured by Nimbus 4 for the specific geographical regions near the 36 Dobson stations used in his ozone trend analysis is also -0.5 percent per year. Therefore, the trend in average ozone levels above the 36 Dobson stations is identical to the trend in average global ozone levels as measured by Nimbus 4 during an 8-year period, which leads one to conclude reasonably that the Dobson network measures true global average ozone. To summarize, ozone trend analysis is a sensitive method to monitor ozone levels in the stratosphere. Ozone depletion has not occurred. It is unreasonable to suggest that highly uncer- tain computer model calculations are more reliable than time series analyses of actual ozone measurements, when the latter; IV-20 ------- Science Show increase in ozone of more than 1 percent, rather than depletion, Are sensitive enough (+_ 1 percent to +_ 1.5 percent) to detect quickly any excursion of ozone outside normal levels, Become more sensitive with time. IV-21 ------- Science D. EPA's ASSESSMENT OF THE THEORY The justification for proposed rulemaking on CFCs by EPA is stated in the opening sentence of the ANPR: "Because of the destructive effect of chlorofluoro- carbons (CFCs) on stratospheric ozone, EPA is considering restricting their production such that the potential for ozone depletion does not increase over present levels." Clearly, the assumption is that ozone is decreasing. Just as clearly, the tacit assumption is that CFCs, in and of themselves and regardless of other natural and man-made long-term changes, will cause such a decrease unless emissions are controlled. For a number of reasons, the analysis which has led to these assumptions is inadequate. A very serious problem appears in defining the scope of the problem. Unlike most regulatory issues facing the Agency, in which a single chemical species might produce a deleterious effect .regardless of other circumstances (e.g., certain toxic substances), changes in ozone are controlled by a variety of coupled chemical reactions. In order to understand changes in ozone, it is necessary to understand all such effects simultaneously. EPA acknowledges in the ANPR that it "has relied primarily on the scientific analysis of the National Academy of Sciences [NAS, 1979a; 1979b] for support of this theory as well as for an assessment of the potential hazards posed by continued world emissions of CFCs." The now already outdated report of the Panel on Stratospheric Chemistry and Transport [NAS, 1979a, p. 14] notes that it is "difficult to project accurately the effect of increased halocarbon release when the release of other man-made pollutants may also be increasing in a undetermined way." However, the Panel does not consider other perturbations in either the scenarios for calculations into the future or in the error analysis of their results. The entire approach treats IV-22 ------- Science CFCs as if they were the only chemical substances which might produce effects on ozone, i.e., as if the CFC perturbation were separable from other effects occurring simultaneously. With regard to ozone and protection of the environment, the problem is not that simple. The possibility of CFC effects combined with C02 or N20 changes is well documented by NASA in another of EPA's cited references [NASA, 1979], Numerous technical papers [e.g., Miller e_t al., 1980c; Penner, 1980a] further document this situation. The narrow view implicitly ignores the real world issue in favor of one which seems more tractable. As discussed below and in Appendix E, serious problems remain even in this narrower approach. Lack of attention to the real world is also evident in the treatment afforded to ozone measurements themselves. Scientifically, it is undoubtedly important to understand causes of ozone changes, and furthermore to detect such changes with sufficient sensi'tivity to assign a cause. However, this very strong demand is not immediately applicable to the regulatory problem. Given a theory about possible changes in ozone, any changes which might be demonstrations of that theory are cause for concern. This remains true even if such changes are not unambiguously related to the theory. Thus, the important regulatory question becomes: "How large must a trend in ozone be to be detected, regardless of its cause?" In the ANPR EPA states it has been motivated to take action now rather than to wait for better information since "validation of the ozone depletion theory through environmental monitoring is limited because a minimum ozone change of approximately 5 percent over a period of ten years would be required before a depletion could be observed with statistical confidence." The analysis is incorrect for several reasons. First, the figures quoted represent "best guesses" [NAS, 1979a, p. 93] and "estimates" [NASA, 1979, p. 286] made without IV-23 ------- Science analysis to determine actual statistical significance of the data. Furthermore, the quoted numbers referred to the specific scientific question of demonstrating cause and effect, rather than to the immediate regulatory question of identifying changes. The EPA analysis is misdirected; it also ignores available information. Statistical analyses of the actual ozone measurements have been made, and were available to EPA [CMA, 1980a]. A small trend would be detectable, but it is not present [St. John e_t al. , 1980; Reinsel jit al. , 1980; Watson, R. T., 1980]. Rather than 5 percent, a trend of only 1 to 1.5 percent would be detectable over 10 years. Any "overshoot" would similarly be reduced. Calculations based on current theory imply an overshoot of no more than half the amount of depletion existing at the end of emissions. Thus, even if the theory proves correct, responsible action is capable of limiting ultimate depletion to and below EPA's 5 percent [EPA, 1980a, Jellinek, 1980a]. The fact is that depletion which is calculated to have occurred already has not been observed. On the contrary, in several recent sophisticated analyses of ozone data, slight increases in ozone have been noted, as discussed immediately preceding this section. This fact has been ignored. (Still more details are presented in Appendix E). EPA's inattention to the capabilities of ozone measurements and analysis has allowed many other misleading points to be made in the ANPR. Large steady-state ozone depletion is simply not a viable possibility. Even the largest estimates of trend detectability would allow for early action if it becomes necessary. Large calculated numbers based on unrealistic future situations do not justify immediate action. Concern over growth in production is likewise misplaced. In the first place, large-scale growth of 7 percent or 9 percent per year is not reasonable because of its requirements for new production capacity. Furthermore, capacity has already IV-24 ------- Science been limited for European producers by actions of the European Economic Community, and significant production capacity has been dismantled in the U.S. To describe 9 percent future growth based on the recent historical record of production [CMA, 1980b] is equally unsupportable. In addition, a point which is neglected in the EPA statement regarding such growth is the effectiveness of mere concern as a deterrent to increased production (see Section VII-Economics). A sensible businessman simply would not risk a large amount of investment capital in a threatened industry for what could be a very short-term return. In short, the projected growth represents a naive view of business sensibilities, and concerns over such growth are misdirected as are the ANPR arguments which depend on growth. (For further discussion on this point, see Appendix J). Realistic considerations are equally missing from the EPA discussion of atmospheric modeling. To quote a basic text, "An Introduction to Scientific Research" [Wilson, 1952]: "A successful scientist knows that all models are somewhat defective and that certain aspects of his visualization do not apply to the problem in hand." One must realize that atmospheric models do not predict a realistic future. Models may, however, be used to calculate the consequences of a set of assumptions about the future. Thus, a modeler honestly describing his work on the CFC problem must qualify his "predictions" (a more appropriate word is "calculations") with a number of caveats. The results will obtain only, if 1) the model is complete, 2) input data are accurate, and 3) approximations are justified. Furthermore, the "future" as calculated in atmospheric models bears little resem- blance to the future as we expect it to be. Models assume that all inputs to the stratosphere remain constant except the varia- ble(s) under consideration. Simple examples of expected depart- ure from the CFC scenarios abound: CO- is increasing; N20 is IV-25 ------- Science thought to be increasing; other chlorine-containing compounds are possibly increasing; natural events may have unpredictable long- term effects; aircraft and space shuttle fleets are likely to continue injecting increasing amounts of nitrogen oxides, etc. In short, models "predict" only what would happen in a simplified idealized atmosphere with a number of qualifications. The limitations of models do not imply that they are unimportant or useless, but merely that they are imperfect research tools. (The imperfections are described in considerable detail in Appendix E). The continued emphasis in the ANPR on "predicted effects", without acknowledgement of the many serious actual and potential differences between the model "future" and reality, is particularly dangerous in that the political decisions will ultimately be made not by the scientists who take such restric- tions for granted but by layment who may not be properly informed. Along with the capabilities of the models themselves, the input information necessary for such calculations is inaccurately discussed in the ANPR. As an example: "The chemical reaction rate coefficients and the photolysis rate coefficients (measures of the speed with which reactions occur), the vertical diffusion coefficients .(a measure of the speed of the vertical transport), and other parameters needed to produce a numerical solution have generally been measured either in the field or in the laboratory. A few are estimates." The chemical reaction and photolysis rate coefficients have indeed in most cases been measured, but only under laboratory conditions. A number of them have been measured at more than one temperature and a smaller number at more than one pressure, yet both temperature and pressure often influence the rates of IV-26 ------- Science reactions. However, only a very small number have been measured over the full range of temperature and pressure conditions occurring in the stratosphere. In the absence of appropriate measurements, modelers use extrapolations where limited data are available. But in the case of measurement at a single set of conditions, 'temperature or pressure dependencies generally are assumed to be nonexistent. Such extrapolations and assumptions are sometimes justified, but almost equally often wrong. (Specific examples will be discussed below). In many cases, the values chosen for models represent a compromise between disagreeing measurements, again contributing to overall uncertainty. Diffusion coefficients are another matter entirely. Representation of vertical transport by diffusion coefficients is itself an unproven hypothesis a simplification of the three-dimensional circulation patterns of the atmosphere to "model" vertical transport. Such coefficients are derived, rather than measured, by fitting an assumed mathematical expression to measured species concentrations. They are a mathematical representation of a much more complex phenomenon, and are chosen to produce reasonable agreement between calculated and measured results for atmospheric species which appear to be well understood. The applicability to other chemical species again involves an assumption. (A discussion of eddy diffusion theory and its use limitations in modeling is included in Appendix E). The implied satisfaction with these input parameters in the ANPR belies the dynamic nature of atmospheric science. New information (much of which is discussed in detail in Appendix E) continues to arise from ongoing research, modifying earlier model inputs and demonstrating that, in effect, far more than "a few" of the input parameters "are estimates". As the parameters IV-27 ------- Science change, so will the calculated depletion. As our understanding of atmospheric processes evolves, so will the models used to simulate them. The ANPR acknowledges that "simplifications" have been introduced into these models, but cites only the limitation to one-dimensional (1-D) models. Other simplifications are numerous and potentially at least as significant. Despite apparent concern over the 1-D approximation, no reference whatsoever is made to the two-dimensional modeling now being developed, or to the results of 2-D models made available during the past year since the NAS report. In the ANPR, EPA has chosen to rely on a 1979 report on possible ozone depletion, its conclusions, and its discussion of uncertainties. The above-mentioned rapid pace of the science makes the construction of such a report a difficult task, and furthermore quickly outdates a summary assessment made at any given time. The year since the issuance of the report has been marked with a number of new developments, some of which have dramatic impacts on the perceived severity of potential CFC effects. Those developments have arisen from the many ongoing research programs in academia, industry, and government. Each such development is discussed thoroughly in the review which follows in this section and in Appendix E. An additonal risk in reliance on any single review is the possibility of false indications of scientific consensus. Several documents made available to the EPA before and after the issuance of the NAS reports, and before the publication of the ANPR, point out the variety of opinion which exists [UK DOE, 1979; Du Pont, 1980a; 1980b; CMA, 1980a; EEC, 1980]. That these are rejected summarily and without explanation is puzzling. One cannot overestimate the importance of understanding the uncertainties involved. The NAS attempted to quantify these uncertainties [NAS, 1980a] , and their efforts have been IV-28 ------- Science criticized [Du Pont, 1980a; 1980b]. But throughout the exercise, the NAS made clear that the uncertainty estimates were, in general, subjective a "best professional judgment". Such caveats cannot be ignored. To report such confidence limits as objectively derived science without qualification is to ignore something basic to the scientific process. In referring to the role of human bias in science, Wilson [1952] wrote: "No human being is even approximately free from these subjective influences; the honest and enlightened investigator devises the experiment so that this own prejudices cannot influence the results. Only the naive or dishonest claim that their own objectivity is a sufficient safeguard." The NAS panel, of course, was careful to point out the subjectivity surrounding many of its estimates. However, in spite of additional cautions on that point [Du Pont, 1980a] , summary statements which neglect to mention such restrictions have now been taken by EPA to provide the basis for regulation. The detailed discussions in the next section and Appendices E and F reiterate the claim that those estimates of an uncertainty range were far too narrow, and that the year-old assessment of the issue is an overstatement of its severity. Overstatement of severity extends also to the second of the NAS reports which have provided the basis for EPA's proposed actions [NAS, 1979b] . The primary danger of ozone depletion is cited as risk of greater exposure to ultraviolet radiation at the earth's' surface. It is well known both that different wavelengths of ultraviolet radiation are attenuated in varying amounts by ozone and that any biological effects are also likely to be wavelength dependent. In combining these variations to discuss the effective change, the NAS Committee on Impacts of Stratospheric Change [1979b] chose to gauge effectiveness at each wavelength by the potential for DNA damage, which is, perhaps coincidentally, the procedure which results in the greatest IV-29 ------- Science increase in "damaging ultraviolet" (DUV) for any given decrease in ozone. While the report makes clear the hazards of assuming a given spectrum (i.e., a particular gauge of effectiveness for different wavelengths of UV) without examining each individual effect, it is the defined DUV change which is cited by EPA as posing a danger to humans, other animals, and plants. To summarize briefly, the scientific rationale advanced by EPA as the basis for proposed rulemaking suffers from many shortcomings, each of which may impact an ultimate regulatory decision: The chosen scope of the problem is too narrow. The scientific basis for results discussed is not understood. New information is ignored. Uncertainties and weaknesses in science are not clearly stated or acknowledged. Scientific consensus is gauged by a single document in the face of several others with disparate conclusions. Little attention is paid to the distinction between scientific exercises and the real world. Conclusions are overstated. It is essential to clarify the actual state of the science at the time of regultory decisions. In the face of uncertainty, such decisions are extremely difficult; in the face of inadequate, incorrect or misleading information, they will IV-30 ------- Science often be wrong. In the section which follows and in more detail in Appendix E, each aspect of the atmospheric science involved in potential ozone depletion is reviewed thoroughly, with reference both to earlier reviews and to more recent information. This review also advances what is felt to be both a more realistic and a more complete assessment of the current state of science surrounding the Chlorofluorocarbon/Ozone Theory. IV-31 ------- Science E. PRESENT STATUS OF THE THEORY This section discusses key uncertainties in the science of the ozone depletion theory and new information which became available after the NAS [1979a] and NASA [1979] reports were published. Each step of the theory is assessed and its current status is summarized. In paticular, the differences with respect to the expressed view of EPA are noted. 1. Production and Release of CFCs The historical CFC production and emissions data used for input to models are well established, and are discussed more thoroughly elsewhere in this document (Appendix J). However, CFC releases for the future are in all cases merely scenarios. There is not, nor can there be, any general consensus of what releases might be more than a few years from now, due to the very high connection between business decisions and the regulatory decisions which might be made during those years. Thus, the major uncertainty here lies in the interpretation of model results for a given scenario and in the likelihood of that scenario ever occurring. 2. Lower Atmospheric (Tropospheric) Processes The primary role the troposphere may play in the ozone depletion theory is in the possible destruction of CFCs-11 and 12. Research has so far failed to demonstrate the significance of any possible mechanism, but several efforts continue. The emphasis of recent research results has been in the use of tropospheric trace species measurements to provide indication of sink processes for the CFCs. IV-32 ------- Science a. CFC-11 and CFC-12 Lifetimes The atmospheric lifetimes of CFC-11 and CFC-12 have been believed generally to be long, e.g., 50-100 years. The reasons are: Systematic analysis of known tropospheric destruc- tion processes indicated none are likely to remove any significant fraction of CFCs from the tropo- sphere. Measured values of CFCs in the troposphere are consistent with the conclusion that CFC-11 and CFC-12 have long atmospheric lifetimes. Clearly, unknown processes which destroy CFCs have not been evaluated, if they exist. The first item above says the processes that have been evaluated do not destroy significant amounts of CFCs. It does not establish that CFCs are not destroyed in the troposphere. The second item ignores errors in the measurements and calculations of the total amount of CFCs in the atmosphere. Even a slow removal process of 2 percent per year, would be important to the theory, and would not yet be detected by the measurements. Nonetheless, the model calculations assume all CFC-11 and CFC-12 released to the atmosphere reaches the stratosphere, and all the chlorine content of these CFCs participates in the ozone destruction cycle. Obviously, any fraction of chlorine which does not reach the stratosphere reduces calculated ozone depletion proportionately. An experiment called the Atmospheric Lifetime Experiment (ALE) has been in progress since 1978 to measure the atmospheric IV-33 ------- Science lifetime of CFC-11 and CFC-12. The methodology of the experiment appears in the scientific literature [Cunnold et. al. , 1978], The objective of the experiment is to measure CFC-11 and CFC-12 concentrations for 3 years or longer, several times each day, at 4 or more remote sites throughout the world. The measured increase in concentration of CFCs versus time, i.e., the trend in CFC concentrations, is compared to that calculated from known .releases of CFCs over the same time period. The lifetimes of 'CFC-11 and CFC-12 then are calculated by comparing the two trend estimates. Preliminary results from 18 months of measurements indicate the lifetime of CFC-11 is about one-half that assumed in the NAS report [1979a] and elsewhere, i.e., about one-half the calculated 50-year lifetime from stratospheric destruction alone. At this time the value still has large uncertainty limits which Include the longer lifetimes, but the limits will become rapidly smaller as more measurements are recorded. If this lifetime is confirmed, the effects of CFC-11 on ozone, if any, will be halved. Similarly, any process that removes CFC-11 is likely to remove CFC-12 but at a slower rate. It, therefore, is reasonable to expect that some significant fraction of CFC-12 may also be removed in the troposphere, with a proportional reduction in calculated ozone depletion. b. CFC-21 CFC-21 has been detected in the troposphere by several research groups [Rasmussen ^t al., 1977; Penkett et al., 1980; Crescentini and Bruner, 1979; 1980; Singh _et al., 1977; Cronn and Harsch, 1979] at concentrations ranging from 1 to 15 parts per trillion by volume, (pptv.). An early suggestion that it is an artifact of the analysis (absorption of the CFC-21 standard onto "Teflon" parts of the analytical apparatus and desorption during analysis of air samples) has not been supported [NASA, IV-34 ------- Science 1979, p. 92]. At these measured levels, the amount of CFC-21 exceeds that produced globally by about a factor of 100-1000. That being the case, it has been suggested that CFC-21 is formed in the atmosphere from CFC-11. CFC13 ^ CHFC112 (12) If that is the case, the production of CFC-21 . from CFC-11 must, in fact, be occurring at a faster rate than the destruction of CFC-21 by reaction with OH radical, a relatively fast process. In short, process (12) would be a significant tropospheric sink for CFC-11. The destruction of CFC-11 and CFC-12 on sand surfaces has been studied [Ausloss e_t al. , 1977; Ausloss and Rebbert, 1980], In one experiment, in the absence of air, CFC-21 was formed from CFC-11 in the presence of sand, but not when air (oxygen) was present. More recently, two research groups have obtained widely different results from the analysis of air samples collected at different times and places. One group [Penkett et al. , 1980] consistently found CFC-21 in the 1-3 pptv concentration range, while the other [Crescentini and Bruner, 1979; 1980] found low values (1-3 pptv) both in rural and industrialized areas (in one case, no CFC-21 was found near a CFC production plant) , but very high amounts (up to 100 pptv) of CFC-21 in air masses that passed over the Saraha Desert. In summary: The ALE research suggests an unknown tropospheric sink exists for CFC-11, and possibly CFC-12. CFC-21 may be a degradation product of CFC-11, presumably formed on sand surfaces. IV-35 ------- Science Background levels of CFC-21 are low but may be high in air masses that pass over desert areas. c. CFC-22 CFC-22 is a commercial product and is released to the atmosphere. Recent analyses [Rasmussen et at., 1980] suggest that the quantity of CFC-22 in the atmosphere is much larger than that from estimated releases. Several possible explanations are: The measurements are in error. The release estimates are in error. A production process for CFC-22 exists in the atmosphere. If the latter were true, CFC-12 would be a likely source of CFC-22 CF2C12 ^ CHC1F2 (13) by analogy with the discussion for CFC-ll/CFC-21. If so, reduced quantities of CFC-12 would reach the stratosphere, with a resultant reduction in calculated ozone depletion. 3. Transport to the Upper Atmosphere In an atmospheric model, transport is the movement of chemical species. In a 1-D model, transport describes the movement of species in the vertical direction; in a 2-D model, movement is described in the vertical and latitudinal (north-south) directions. The diffusive movement of species is described mathematically, and the altitude-dependent rate of transport is defined by an eddy-diffusion coefficient. These IV-36 ------- Science coefficients are chosen to produce agreement with measurements of long-lived species. Two-dimensional models also contain an average air circulation in addition to diffusion. The NAS report [1979a, p.11] states, "It is felt that the uncertainty in the eddy-diffusion coefficent at a given altitude is about a factor of 2, although it is admitted that this uncertainty estimate is somewhat subjective." Other studies show calculated ozone depletion is very sensitive to the choice of transport. Derwent [UK DOE, 1979, p.158] made several ozone depletion calculations, changing only the transport from one run to the next. He used transport parameters reported by five well-recognized modeling groups throughout the world. The calculated ozone depletion in this study ranged over nearly a factor of 4 depending on the transport description used. Based on this study, calculated ozone depletion is uncertain by as much as a factor of 4, whereas NAS limited its uncertainty estimate to a factor of 2. It is just this sort of underestimation which can lead to overconfidence in model results. 4. Chemistry in the Lower and Upper Atmosphere The year since the release of the NAS [1979a] report has brought several major changes in the accepted set of reaction rates [Chang, 1980], Also included in the most recent revisions are many minor changes. The net effect has been a reduction in calculated steady-state ozone depletion to roughly 7 to 10 percent rather than the much higher numbers calculated by most modelers in 1979 [NASA, 1979; NAS, 1979a] . At least as important as the changes in ozone depletion calculations is the increased awareness of uncertainty in the reaction rate data used to derive those results. Many of the new IV-37 ------- Science rate results have come as a surprise to kineticists. The importance of accurate temperature and pressure dependence measurements has surfaced as a major issue, particularly for the chemistry of the hydroperoxyl (HO-) radical. This radical reacts with most of the major stratospheric chemical species, and strongly influences their concentrations. Reaction product identification, likewise, has been a major issue, with the realization that assumed products may not always be correct and that alternative products (i.e., different reaction "channels") may lead to very different calculated ozone depletion. The uncertainty, of course, extends in both directions. Resolution of the uncertainties may lead to either increases or decreases in calculated ozone depletion. However, they reinforce the contention that stratospheric chemistry is simply not well enough understood to permit confidence in current model calcula- tions. The major developments may be summarized briefly as follows: Recent studies have shown several reactions of the hydroxyl (OH) radical may be faster than thought previously. The faster rates substantially reduce calculated ozone depletion. Several other changes in kinetic and photochemical data contribute to further reductions in calculated depletion. They reflect the major uncertainties still present in the model input data set, particularly with respect to temperature and pressure dependencies. For several important chemical reactions, alter- native products to those assumed are possible. The IV-38 ------- Science alternative products, if formed, would reduce substantially calculated ozone depeltion. The rate of formation of chlorine nitrate and its fate in the stratosphere, are not well defined. Since this species has a large influence on calculated ozone depletion, the calculations remain highly uncertain,, and could vary in either di rection. a. Hydroxyl Radical Reactions i. Hydroxyl Radical and Nitric Acid In a recent study [Wine et al., 1980], the rate of the OH + HN03 reaction has been found to be significantly faster at stratospheric temperatures than that measured previously. The effect is to increase the rate of removal lower the calculated average concentration of OH in the lower stratosphere. This, in turn, slows the conversion of "inactive" HC1 to "active" Cl OH + HC1 ^ H20 + Cl which may participate in the ozone depletion cycle (Reactions 6 and 7) on p. IV-8, which, in turn, lowers the predicted ozone depletion calculations. These results agree with several atmospheric measurements (discussed in section 6 below) which suggest indirectly that calculated stratospheric OH concentrations are too high. No direct measurement of OH in the stratosphere is available. The effect of the new reaction rate depends on the reaction products IV-39 ------- Science OH + HN03 - » H20 + N03 (8a) - ^ H0 + N0 (8b) The Lawrence Livermore Laboratory model calculated steady-state ozone depletion of 9.5 percent if Reaction 8a occurs and 12.3 percent if Reaction 8b occurs [Wine ^t al. , 1980] . Products (8a) seem more likely to form since the reaction path is simplier and these products are more stable. The reactions are under study. Kurylo [1980] has recent preliminary data confirming the rate, although new values [Marinella ert al. , 1980] are somewhat lower. Marinella ^t al. , [1980] also identify the major product channel as (8a) . ii. Hydroxyl Radical and Peroxynitric Acid Recently, the photolytic lifetime of peroxynitric acid, H02NC>2, in the stratosphere was shown to be much longer than thought previously [Molina and Molina, 1980a; 1980b] . As a result, other destruction processes, e.g., reaction with OH, have become important. Two new measurements [Littlejohn and Johnston, 1980; Barker _et al. , 1980] of the OH reaction indicate a relatively fast rate, which might be expected, by analogy to the OH + HNO-, and OH + H202 reactions. The effect of a fast rate is similar to that for the nitric acid reaction -- reduced stratospheric OH concentration and reduced calculated ozone depletion. Use of the average of the new rates with that of Wine et al. , for OH + HN03 reduces calculated depletion still further to about 6.0 percent. iii. Hydroxyl Radical and Hydroperoxyl Radical Several measured values [Hack ejt al_. , 1978; Burrows £t_ al. , 1978a; 1978b; Chang and Kaufman, 1978; Demore, 1979; and IV-40 ------- Science Hochanadel e_t al., 1972] of the rate of the OH + H02 reaction vary over a range of a factor of 10. However, a slow rate for the reaction is used in computer models. If the fast rate is correct instead, or a pressure dependence exists, the effect is similar to that of nitric acid and peroxynitric acid reduced stratospheric OH concentration and reduced calculated ozone depletion. It should be noted that reduced calculated ozone deple- tion is more nearly consistent with the results of ozone trend analyses. b. Pressure and Temperature Dependencies Several minor revisions have contributed to the recent reductions in calculated ozone depletion, and reflect the continuing uncertainty involved in model input data. Much of this uncertainty lies in the lack of adequate data concerning the dependencies of rate constants on pressure and temperature. Even where data are available, as in this example of H02 + H02, individual studies are often incomplete. An attempt will be made in the new NASA recommendations [Chang, 1980] to gather this rate data for H02 + H02 into an altitude-dependent form. For most reactions, pressure dependence has not been measured or has been measured over a very limited range. Temperature dependence data are considerably more satisfactory, but are far from complete. The necessary assumptions made (usually that the reaction rate does not vary with pressure or temperature) to permit modeling of a reaction with incomplete data confound attempts to adequately estimate uncertainty. Other dependencies noted for some reactions, such as the variation of the H02 + H02 rate with water vapor concentration [e.g., Lii et al, 1979] have not been investigated for most IV-41 ------- Science reactions, and add to the caution necessary in viewing model results. c. Alternative Reaction Products For each of the following reactions, the assumed products (reaction a) and alternative products (reaction b) are given. H02 + CIO HO + CIO HOC1 + h HOC1 + 0, HC1 + 0 HO 3 + Cl HC1 + 0. HO + Cl HC1 + 0 (9a) (9b) (10a) (10b) (Ha) (lib) If reaction b occurred to the extent of 10 percent for any of the above reactions, the increased production of HC1 removes chlorine from the catalytic cycle and calculated ozone depletion would be reduced significantly [Howard, 1980], Multiple reaction possibilities (channels) for the reactions are theoretically possible and experimental obser- vations have been explained in terms of alternative reaction pathways at stratospheric temperatures. Stimpfle et al. , [1979] mentioned channel 9b as a possible explanation for the unusual curvature of the rate-versus-temperature data they measured for the H02 + CIO reaction. Until the products of these reactions have been determined quantitatively at stratospheric temperatures and pressures, the uncertainty exists that ozone depletion is vastly overestimated because HC1 channels are not included in model calculations. IV-42 ------- Science d. Chlorine Nitrate Chlorine nitrate is a very important species in the stratosphere. It is formed by the reaction CIO + N02 > C10N02 The reactants, CIO and N02, are both part of ozone destruction cycles (Reactions 6 and 7, Reactions 4 and 5 - p. IV-8), while chlorine nitrate is inactive. The rate of formation of chlorine nitrate in the stratosphere and its subsequent fate can have direct effects on calculated ozone depletion. To evaluate the effects, the following must be known: The rate of formation of C10N02. The destruction (photolysis) rate of C10N02. The subsequent reactions of the photolysis products. All the above for any isomers, e.g., OC10NO or C100NO of C1N03. Since this information is not available, the effects of ClON02 and/or its isomers are estimated only with a high degree of uncertainty. Recently, Molina e_t al. , [1980] suggested that other isomers are important. Should they be formed and be less stable than C10N02, calculated depletion could be increased. IV-43 ------- Science 5. Atmospheric Models As in the area of chemistry, research in modeling has continued at a rapid pace during 1980. The models represent an attempt to synthesize the mass of experimental and theoretical information into interpretable results. As such, use of state- of-the-art calculations is imperative. All information must be considered, and the prognostic capabilities of models must be tested whenever possible. The key research of the past year has produced several results: Two-dimensional (2-D) models (including both latitude and altitude variation) permit more detailed calculations of the atmosphere than one-dimensional calculations (which consider only altitude) and allow more realistic comparisons of calculated values with actual measurements. A 2-D model has shown that calculated ozone depletion occurs mainly at polar regions in winter, and the absolute ratio of the percentage increase in ultraviolet radiation to the percentage decrease in global average ozone concentration is closer to 1 rather than the commonly accepted value of 2. In other words, even if ozone is being depleted, the projected effects will not be as deleterious as previously predicted by the 1-D models. The most recent chemistry reduces the latitudinal variation somewhat, but the effect remains. Increased concentrations of carbon dioxide and nit- rous oxide in the atmosphere reduce calculated ozone depletion. Accordingly, these chemicals must be IV-44 ------- Science considered in any realistic assessment of anthro- pogenic effects on ozone. It has been suggested recently [Johnston, 1980] that volcanoes inject large quantities of chlorine into the stratosphere, which should decrease ozone severly near the point of injection. These effects are not evident in ozone measurements. a. 2-D Calculations The principal advantage of 2-D model calculations is that they more accurately represent the real world. A complete 2-D model would include a full description of chemistry, latitudinal transport with its seasonal changes, seasonal and latitudinal variations in solar flux and atmospheric temperature profiles. Profiles for chemical species are calculated as a function of latitude, altitude, season, and day or night. By way of comparison, 1-D models include only average values of chemical species concentrations as a function of altitude and do not include seasonal variations. As an example, the Oxford University 2-D model has been used to examine variations in ozone depletion with latitude and season [Pyle and Derwent, 1980]. The maximum depletion is calculated to occur at polar regions in winter. This is the time and location of lowest population, lowest biological activity, and minimum incident ultraviolet solar flux. The Oxford modelers find the global average "physical amplification factor", i.e., the percent increase in erythemally-weighted (i.e., weighted based on its effectiveness in inducing sunburn) UV-B dose associated with a one percent decrease in global average ozone concentrations, is closer to 1 rather than 2. This, in turn, reduces by that same factor any estimates of biological effects IV-45 ------- Science related to a given change in ozone (provided they are related by the chosen weighting). Another example is that of N2°' The biosphere produces this gas naturally, but the source strength (the amount of N^O produced with time) varies with latitude by about a factor of 10 from the equator to 40 degrees N. A 2-D model can deal with this real-world situation, while a 1-D model cannot, since latitude does not exist in a 1-D model. It is the assumptions, uncertain input data, and simpli- fications -- perhaps over-simplifications of 1-D models that bring their calculated values into question -- that and the serious discrepancy between measured and calculated profiles for important chemical species. b. Carbon Dioxide/Nitrous Oxide Effects i. Carbon Dioxide It is well-documented that the amount of atmospheric carbon dioxide is increasing. This is projected to warm the troposphere and cool the stratosphere the greenhouse effect. Since the rate of ozone destruction decreases in the stratosphere with decreasing temperature, the expected doubling of atmospheric CO- concentration [NAS, 1979c] will increase ozone levels several percent. A group in England recently reported [Groves and Tuck, 1979] that the calculated combined effect of C02 and CFCs on the ozone layer is less than the calculated CFC effect alone. Although the two separate effects are not additive, calculated ozone depletion from CFCs is reduced by about 3-5 percent in a model scenario in which the atmospheric CC>2 concentration is IV-46 ------- Science doubled by about the year 2030 [Penner, 1980b]. As an example, if CFCs alone produce calculated depletion of 13.9 percent, including CCU increases in the model would reduce this value to 8.9 to 10.9 percent. ii. Nitrous Oxide Agricultural fertilizer has been suggested as an addditonal source of N-O to the stratosphere. Using 1979 chemistry, doubling atmospheric N~0 concentration by itself has little calculated effect on ozone concentrations. However, the combined ^0 and CFC effect reduces calculated ozone depletion by about a factor of 2 compared to the CFC effect alone [NAS, 1979a, p. 181]. The reduced calculated depletion results primarily from increased chlorine nitrate formation as calculated by the models. Updated chemistry complicates the situation by introducing a significant effect from N-O alone. However, the calculated combined effect of N-O and CFCs is still less than the sum of the individual calculated effects. c. Volcanoes Johnston [1980] recently suggested that volcanoes annually inject directly into the stratosphere amounts of chlorine equivalent to 15 to 35 percent of the chlorine content of 1975 global CFC production. If the ozone depletion theory is correct, such massive injections of chlorine over a short period of time would be expected to produce severe local depletion of ozone. Long-term effects would not induce a trend, although the model calculated "normal" ozone concentrations might be somewhat reduced. Du Font's 2-D model calculations [Steed et al. , 1980] show that even for low estimates of chlorine from the Mt. Agung eruption in 1963, ozone near the equator should have been depleted by as much as 10%, with the effect gradually spreading over the entire globe. However, no evidence is found in actual IV-47 ------- Science ozone measurements for such a decrease. One conclusion that could be drawn from this information is that the models overestimate the effects of chlorine on stratospheric ozone. 6. Stratospheric Measurements One test of the reliability of computer models is to compare calculated altitude profiles of key chemical species with measured profiles for the present-day atmosphere. Agreement between the two would lend some support to the reliability of model calculations. Conversely, disagreement must indicate error in the model calculations. An analysis of this type is particularly important in the case of chlorine species, which are central to the ozone depletion theory. a. Chlorine Species i. Hydrogen Chloride The amount of hydrogen chloride (HC1) calculated by the models to be present in the stratosphere is less than that measured in the upper stratosphere and greater than that measured in the lower stratosphere. The difference suggests that the model improperly distributes chlorine among the various atmospheric chlorine species. The NAS report expressed a belief that the problem was an artifact of latitudinal transport not accounted for in 1-D models, and that two-dimensional (2-D) calculations would remove the discrepancy [NAS, 1979a, p. 161]. However, the 2-D model calculated profile at the latitude (30 degrees N) where these measurements were made is similar to the 1-D calculated profile and its slope continues to disagree with the measurements [Miller e_t al. , 1980a] . Recent revisions in chemistry exacerbate the problem. IV-48 ------- Science i i. Chlorine Monoxide (CIO) Radical Although calculated and measured CIO values using NASA [1979] chemistry agree reasonably well at about 38 km, the average measured values below 30 km in the lower stratosphere are about one-fourth calculated values [NASA, 1979, p. 177], Some measurements are one-tenth the calculated value. Since the amount of CIO in the stratosphere governs calculated ozone depletion, the latter has been clearly overestimated. The measurements also suggest indirectly that the stratospheric concentration of the OH radical calculated by the model is too high. As in the case of HC1, the NAS report suggested that 2-D model calculations might remove this discrepancy. Two-dimensional calculations show, the discrepancy persists [Miller et a_l. , 1980a]. Recent revisions in chemistry improve the agreement in the amount of CIO, but a significant discrepancy in slope remains which still leads to overestimates of CIO in the lower stratosphere, and points to problems in partitioning of chlorine. On July 14, 1977, J. G. Anderson [Anderson et al. , 1980a] measured exceptionally high values of CIO in the stratosphere, and simultaneously measured normal values of ozone. The results are inconsistent with the ozone depletion theory. Even though high CIO values were not recorded in subsequent measurements, there can be little doubt the 1977 measurements were correct, i.e., the measured amounts of CIO and ozone were present in the stratosphere. The results can be explained in several ways, including the possible existence of an ozone generating cycle catalyzed by chlorine, which would imply the ozone depletion theory is wrong. Other measurements of CIO also differ from calculated values [Parrish e_t £l. , 1980; Menzies, 1979]. One feature of CIO measurements is great atmospheric variability. For example, Anderson has recorded CIO concentrations that differ by a factor IV-49 ------- Science of 2 over an altitude difference of only 400 meters. Computer models are incapable of dealing with this experimental reality. b. Nitrogen Oxides Relative to actual measurements, models calculate too much odd nitrogen in the stratosphere including very high concentrations of NO and HNO, in the upper and lower stratosphere, respectively [NASA, 1979, p. 171]. The discrepancies have not been explained. The calculated HN03/N02 ratio is also higher than the measured ratio [Miller £t al. , 1980c]. Using the faster rates for several OH reactions narrows this discrepancy somewhat. IV-50 ------- Science F. RESOLUTION OF UNCERTAINTIES This section describes a number of scientific studies 4 either in progress or planned for the near future . The results of these studies will increase our knowledge of stratospheric science and should reduce the degree of uncertainty associated with the effects of chlorine, if any, on the stratospheric ozone layer. 1. Atmospheric Measurements An important series of atmospheric measurements is planned for 1981 which may constitute a direct test of the validity of the Ozone Depletion Theory. Whether CFCs produce a net decrease in stratospheric ozone can be determined by measuring the concentrations of the postulated ozone-destroying species and ozone at the same time and place in the stratosphere. This simultaneous measurement of the key species 0-j, 0, C10,H02 and N02 takes into account the interactions of these radicals with one another. It can be demonstrated in the laboratory that CIO reacts with 0 and Cl reacts with 0-,. The rate constants for the reac- tions have been measured. However, in the earth's stratosphere, there are other radicals, e.g., certain NOV and HO species, A A which also react with ozone and atomic oxygen. The critical question is this: "When chlorine is added to the real strato- sphere, does it increase, decrease, or not affect the net concen- tration of 0-,? Due to the coupling of the chlorine, nitrogen and 4 Many of these studies are being funded by the CFC producer's research program under management of the Chemical Manufacturers Association (CMA) Fluorocarbon Project Panel (FPP). A summary of this program appears as Appendix K. IV-51 ------- Science hydrogen cycles, the presence of chlorine can actually increase the concentration of ozone in the stratosphere if it reacts preferentially with NOV or HOV to reduce their effectiveness in A X destroying ozone. An analogous situation demonstrated the effect of interacting active species. Prior to 1977, models calculated that the injection of NO exhaust from the jet engines of supersonic aircraft would result in net ozone depletion. However, when the rate constant for the reaction of NO and H02 was remeasured and used as input data, the model calculated that in the lower stratosphere, where the ozone concentration is greatest, the added oxides of nitrogen resulted in a net increase in ozone. Two experiments which are designed to measure the effect of chlorine on stratospheric ozone are in progress and should produce data in 1981. J.G. Anderson (Harvard) will simultaneously measure CIO and other species which react with ozone (H02f N02, OH and 0) and, also, 03 itself. It will be done in the so-called "reel-down" experiment. This equipment is a variation of Anderson's proven CIO measurement technique. The probability of success- ful operation is high. These "in situ" measure- ments, taken in a single balloon flight, will produce several concentration profiles for key species over 10 km altitude intervals in the stratosphere. Between 28 and 38 km, where transport is slow compared with the chemical interactions, the measurements can be interpreted unequivocally. IV-52 ------- Science Below 28 km, the measurements may require a correction for transport. A valid method of accomplishing this is being established. P. Solomon (State University of New York) has developed and tested a ground-based ultrasensitive millimeter-wave detector for measurement of CIO and 03 [Parrish e_t al., 1980]. The total vertical column amounts of CIO and 0, are measured with it. A concentration profile of these species with altitude can be calculated from the shape of the measured absorption lines. This unit has already been used to make over 30 CIO measurements which are consistent with the rapid falloff of CIO in the lower stratosphere, observed repeatedly by Anderson using a different measurement technique. This equipment, which can simultaneously measure CIO and 0-j, will be moved to a location in New Mexico where clear skies will allow almost daily measurements. A significant body of data will be generated in 1981. The Atmospheric Lifetime Experiment (ALE) to measure the lifetime of CFC-11 and CFC-12 in the atmosphere is still in progress. A preliminary estimate is that the atmospheric lifetime of CFC-11 is about half that assumed to be the case from solely stratospheric destruction processes. The estimate was made with only 18 months of data and, accordingly, the error bars are very wide, i.e., 7 years to infinity. In 1981, 24 months and 30 months of data will become available which should better define the CFC-11 lifetime and reduce the error bar range substantially. The CFC-11 lifetime should be defined in 1982 and a preliminary estimate of the CFC-12 lifetime should also be available. If the preliminary lifetime for CFC-11 is confirmed, IV-53 ------- Science and CFC-12 behaves similarly, i.e., its lifetime is about half that assumed to be the case from solely stratospheric destruction processes, calculated ozone depletion would be halved. In relation to the ALE program and the question of tropospheric sinks for CFC-11 and CFC-12, a program is in progress to measure the concentrations of CFC-21 in the atmosphere at several sites and to correlate them with CFC-11 concentrations. These studies may establish a mechanism for the conversion of CFC-11 to CFC-21 in the troposphere. The fluorocarbon industry program is actively seeking [Upper Atm. Programs Bull., 1980] to fund experiments to measure total chlorine in the stratosphere. Such measurements would undoubtedly provide increased understanding of stratospheric chlorine chemistry since the chlorine content of the stratosphere is an assumed value. Two methods are under development. The first involves collecting samples on activated charcoal filters followed by neutron activation analysis [Berg, 1980] . The second technique, which is funded by the industry research program, involves decomposition of the sample by plasma or microwave discharge, followed by measurement of the chlorine atomic emission lines [Howard, 1977]. Initial results from at least one method should be forthcoming in 1981 or early 1982. 2. Modeling Ozone trend analysis studies continue, and will include recent ozone measurements from the Dobson stations and satellites, Umkehr measurements, attempts to correlate ozone trends with other meterological variables, and methods to establish that the Dobson stations do measure average global ozone, as suggested by preliminary studies. An important part of the program will be to calculate the trend in ozone with the most recent ozone measurements from the Dobson station network. IV-54 ------- Science Another important aspect of modeling is to refine and use 2-D modeling capability along with atmospheric measurements to validate model simulations of the present atmosphere. Other 2-D model studies will be directed to better understanding of the latitudinal distribution of effects induced by atmospheric perturbations, e.g., volcanoes, anthropogenic pollutants, etc. The 2-D studies will complement 1-D calculations and will help to quantify some of the averaging assumptions necessary in 1-D models. The effects of increasing concentrations of C02, and possibly N20, in the atmosphere are of concern, and these effects must be better defined. A number of modeling groups are studying this important question as it relates to the greenhouse effect and its effect on calculated ozone depletion. 3. Chemistry Major uncertainties to be addressed in chemistry include the products of important chemical reactions, and the pressure and temperature dependence of reaction rates. These three items are obviously related since it is necessary to measure reaction products and rates over the range of temperature and pressure encountered in the stratosphere. This significant undertaking is in progress for several important reactions, e.g., OH + CIO, OH + HN03, and results should be forthcoming over the period 1981-1983. .Most of the reaction product studies have the potential to reduce calculated ozone depletion, or to leave the value unchanged, rather than to increase ozone depletion estimates. IV-55 ------- Science G. SUMMARY In the preceding discussions, a number of major scientific aspects of the CFC/Ozone Issue have been addressed, and conclusions drawn should impact directly on the need for possible regulation. Each of these issues has been inadequately dealt with by the EPA in the ANPR. Briefly, we have shown: Statistical analyses of real-world ozone measure- ments (ozone trend analysis) from the worldwide Dobson measuring network show a slight increase in ozone has occurred during 1970-1979. The analyses are sufficiently sensitive to detect an increasing or decreasing trend in ozone concentration of +_ 1.0 to + 1.5 percent over this period. In contrast, NAS [1979a] stated 2.1 percent depletion should have occurred according to model calculations. The results suggest some combination of the- following possibilities: 1. Computer model calculations significantly overestimate ozone depletion by CFCs, or 2. An opposite trend natural or man-made -- is offsetting any ozone depletion caused by CFCs. Recent results in chemistry and modeling support the former option. The Lawrence Livermore Laboratory (LLL) recently reported that their central value for calculated future potential ozone depletion has been reduced from 18.6 percent to 13.9 percent through revisions in model input data. This change has been made since publication of the NAS report, which reported IV-56 ------- Science the 18.6 percent value. The 18.6 percent calculated value was "adjusted" in the report to 16.5 percent the value cited by the ANPR. As the NAS panel relied on the LLL model for its analyses, the 13.9 percent value should now be taken as the NAS base case, not 18.6 percent. To provide a clear relation between new results and the NAS report, we employ 13.9 percent as the base case in the discussion that follows. One-dimensional computer model calculated concen- trations of CIO (the critical chemical intermediate in the ozone depletion theory) in the lower strato- sphere (made using NASA [1979] recommended chemis- try) exceed the "normal" range of measured values by about a factor of 4. This discrepancy remains in two-dimensional model calculations, suggesting basic errors in understanding of stratospheric processes. If calculated values of CIO concentrations are arti- fically reduced in the model to agree with the measured values, calculated ozone depletion is reduced from 13.9 percent by more than a factor of 2^. New chemical data have led to some improvement of CIO agreement, and a concomitant reduction in calculated ozone depletion. The rates of chemical reactions used in model calculations are of major importance. For example, several reactions of OH radical, e.g., with HNOj, have been recently reported to have rate constants larger than those recommended by NASA [1979] . The effect on model results is a reduction in calculated OH radical and CIO in the lower stratosphere. The larger rate constants reduce calculated ozone deple- tion to about 6 percent - 9 percent. These figures IV-57 ------- Science represent the model result for tentative 1980 NASA recommended chemistry along with the recent first measurement of a rate for OH + HCUNCU. Preliminary results from the Atmospheric Lifetime Experiment indicate (although with large uncer- tainty) that only about half of the CFC-11 released at ground level is transported to the stratosphere. Presumably, CFC-12 would behave similarly. Since calculated ozone depletion is proportional to the amount of chlorine injected into the stratosphere, these results would further reduce all calculated depletion values by up to a factor of 2. Unexpectedly high measured values of CFC-21 and CFC-22 in the troposphere suggest they may be degradation products of CFC-11 and CFC-12, respectively. High levels of CFC-21 seem to correlate with the movement of air masses over the Sahara Desert. This could indicate a destruction process for CFC-11 and CFC-12 before they reach the stratosphere, and thus lead to a like reduction in the amount of calculated stratospheric ozone depletion. Significant discrepancies exist between measured and calculated values of several species in the atmosphere, in addition to CIO, among which are the HCl/HF ratio, and the HN03/N02 ratio. These discrepancies remain in recent model calculations and lend further support to the suggestion that basic errors exist in our understanding of strato- spheric processes. IV-58 ------- Science One-dimensional models, such as used by NAS, average, and hence ignore, the known latitudinal (north to south) and seasonal variations in stratospheric ozone concentrations. For this and other reasons, one-dimensional models are useful only diagnostically. Alone, they are not adequate for prognosis. Since models provide only global averages, actual measurements, taken at specific locations, cannot be compared with model calculations. Two-dimensional models are being actively developed to reduce such uncertainties. A two-dimensional model at Oxford University shows that calculated ozone depletion is greatest at high latitudes (the polar regions) and in the winter season. Since ozone is at its maximum, and ultraviolet flux at its minimum, at that time and location, any possible biological effects of ozone depletion are minimized. In at least three important chemical reactions, alternative products to those now assumed have been suggested. In every case, HC1 is an alternative product. This is very important because the formation of HC1 removes active chlorine from the theoretical catalytic ozone depletion cycle. (Even a 10 percent channel to HCl in the reaction of OH + CIO would reduce calculated ozone depletion signifi- cantly) . Studies are in progress to identify quantitatively any HCl production channels from the reactions. Other isomers of ClNO^ may be produced along with ClONOj. Less stable isomers could reduce the effectiveness of this holding tank (a reservoir of IV-59 ------- Science inactive chemical species which may be converted to an active ozone depleting form), and increase calculated depletion. Likewise, more stable isomers could reduce calculated depletion, but the situation is very uncertain. General uncertainty in reaction rates, solar flux, transport, etc., could easily vary the calculations over a wide range of values. When the effects of increased concentrations of CC^ and/or ^0 in the atmosphere are considered simultaneously with those of CFCs, calculated ozone depletion by CFCs is reduced. Volcanoes may annually inject chlorine into the stratosphere equivalent to 35 percent of that in a year's global production of CFC-11 and CFC-12. The local effects of these injections would be extremely large according to theory, but have not been observed. The major conclusions one may draw from these points are: Ozone depletion has not been detected. Calculated ozone depletion is now about one-half or less that stated in the NAS report. Preliminary scientific developments, if confirmed, will substantially further reduce calculated ozone depletion. IV-60 ------- Science It is useful to summarize the changes in the theory as demonstrated by model calculations of steady-state depletion. We present below several of the developments and the corresponding theoretical depletion taking each effect into account: Calculated Ozone Depletion 1979, NAS calculated steady- state depletion (from LLL Model). (This number was ad- justed downward by NAS to 16.5% to account for possible tropo- spheric sinks and feedbacks. That adjustment will not be made below). 18.6% (16.5%) Summer 1980, LLL model results were revised to include several minor changes in rate constants. 13.9% Summer 1980, results revised to include Wine et al., [1980] rate for OH + HN03 -* H20 + N03. Fall 1980, results revised to include all tentative NASA recommendations. 9.5% 7.5% Fall 1980, results revised to include new measurement of OH + H02N02 [Littlejohn and Johnston, 1980]. 6.0% IV-61 ------- Science Other preliminary results may be considered with respect to their individual effects on this last result: Calculated Ozone Depletion If C1N03 isomers are very unstable photolytically If atmospheric lifetimes of CFC-11 and CFC-12 are reduced by a factor of 2. 6%-> 7-8! 6% -» 3% If effects are considered 6%-» 1-3% and finally, If all three of the above effects are considered together (in the above order) 6% 4% IV-62 ------- Science Stated simply and directly, the potential for ozone depletion is currently perceived to be far less than it was only one year ago. While major remaining uncertainties could easily affect that conclusion in either direction, most of the movement is expected to be in the direction of reduced calculated ozone depletion. But regardless of how the calculated depletion numbers are affected by resolution of some of the uncertainties^ the bottom line or "fail-safe" is the ability of ozone trend analysis to detect even small trends in actual ozone concen- tration. This ability, coupled with the slower calculated rate of depletion implied by current models, provides a sophisticated, capable early warning system, with greatly reduced numerical overshoot values if and when depletion is detected. The insistence of EPA on ignoring all these developments and rushing into regulation justified with an out-of-date science assessment is unwarranted. IV-63 ------- V. THE QUESTION OF RISK Page A. INTRODUCTION 2 B. IMPACT OF UNCERTAINTIES IN THE UNDERLYING SCIENCE OF THE OZONE DEPLETION THEORY ON RISK C. IMPACT OF UNCERTAINTIES IN THE POTENTIAL EFFECTS OF OZONE DEPLETION ON RISK 14 D. PROBABILITY AND TIMING OF REDUCING UNCERTAINTIES 27 E. RISK IN WAITING - RISK VERSUS TIME 32 F. IS THE RISK DEVELOPING AS PREDICTED.-> 39 G. AVAILABILITY AND SIGNIFICANCE OF AN EARLY WARNING SYSTEM 43 H. THE RELATIONSHIP OF THE INTERNATIONAL ASPECTS OF THE ISSUE TO RISK 46 I. RISK CREATED BY REGULATION AND THE NEED FOR RISK - RISK COMPARISON 48 J. APPROACHES WHICH ARE INAPPROPRIATE FOR ASSESSMENT OF RISK ON THE CFC/OZONE ISSUE 53 K. SUMMARY AND CONCLUSIONS 60 V-l ------- Risk A. INTRODUCTION The key underlying issue in the CFC/Ozone controversy is risk how to assess it; how to manage it. This is a com- plex issue because it is made up of so many interlocking questions, for example: What is the current or future risk to human health and the environment from continuing release of CFCs? How certain is the risk? Does the projected risk necessitate acting immediately? If not, what is the risk in waiting? Will the benefits from waiting, e.g., better information, be worth the risk in waiting? What is the risk associated with acting immediately? What is the balance between the risk associated with waiting and the risk associated with immediate action? Although the risk question is complex, the objective of risk assessment and management may be stated simply: to determine the course of action which will be recognized as wise at a future time. EPA has stated that: "If [EPA waits] until better data are in on whether chlorof luorocarbons deplete the earth's ozone layer, either the theory will be wrong and there will be no harm done or the theory will be right and it will be too late to do anything about it". [Jellinek, 1980b] V-2 ------- Risk We believe that this approach to the complex issue of ozone depletion risk is not supportable by the available facts. On the one hand, based on currently available information, EPA greatly overstates the risk from not acting immediately; and, on the other hand, EPA ignores the substantial potential risk which could be incurred from premature or unnecessary regulation. In order for a risk assessment to be of use to the policy maker, and credible to all participants in a rule-making process designed to manage the risk, it must be accurate and thorough. All important elements of the issue which bear on risk must be addressed and integrated into the whole. A risk assessment which ignores critical components, for example, the question of uncertainty or the relationship of projected deve- loping risk versus time, will not provide an accurate view of the true situation. Regulation based on incomplete or inaccu- rate risk assessment may be bad regulation and subject to legal challenge. In this section of our ANPR response, we discuss important risk components which must be included in a properly balanced analysis of the CFC/Ozone Issuecomponents which to date have not been adequately addressed. These include: The impact of uncertainties both in the underlying science and in estimates of the potential effects of ozone depletion. The importance, probability and timing of reduc- tions in these uncertainties. The relationship of these uncertainties to the U.S. versus world regulatory status and potential actions. The question of risk versus time, which encompasses questions such as, "Is the risk developing as pro- jected?", "What is the risk in waiting?", and "Is there any way to provide an early warning system?" V-3 ------- Risk The need to develop an assessment of the risks created by regulating CFCs and to balance these risks with the risks from not regulating. The last part of this section is a discussion of a number of approaches to the question of risk which we have identified in the ANPR and in previous Agency statements and documents approaches to the question of risk which we believe are not conducive to the obtaining of the thorough and proper risk determination needed for the CFC/Ozone Issue. V-4 ------- Risk B. IMPACT OF UNCERTAINTIES IN THE UNDERLYING SCIENCE OF THE OZONE DEPLETION THEORY ON RISK 1. Introduction In an ideal rule-making situation, the risk from a chemical under assessment is known from experimentation and testing, its benefits are also known, and a determination can be made as to what extent it is reasonable to give up the benefits in order to lower the risk from its continued use. When the degree of risk is not well-known or when the accuracy or vali- dity of the stated risk is highly uncertain, this determination becomes significantly more complex and subject to major error. Such is the situation with CFCs and the theory of stratospheric ozone depletion. In this instance, there are predictions and estimates of risk from continued use of CFCs, but there is substantial debate over how accurate or "certain" these estimates really are. The underlying importance of these risk uncertainties is their impact on the regulatory decision processAre estimates of risk likely to prove to be sufficiently accurate or "certain" that supportable decisions based on them may be made? The more uncertain the data base on which risk assessments are made, the less certain the accuracy of the risk assessments and, thus, the less certain it is that regulatory decisions will prove to be correct. V-5 ------- Risk Consequently, we believe the question of uncertainty as it applies to the underlying science and data base of the Chlorofluorocarbon/Ozone Issue is critical. Therefore, we review below in section 2 the major existing uncertainties and their significance to risk assessment. In section 3 we discuss errors made by EPA in its treatment of uncertainties in the ANPR. These deficiencies in EPA's treatment of uncertainties are critical. Their existence throws into question the validity of the risk assessment being used by the Agency to support its regulatory decision. The uncertainties in atmospheric science are discussed in detail in Section IV and Appendix E; uncertainties in effects of ozone depletion are discussed in Appendix F. 2. Major Current Uncertainty Sources in the Atmospheric Science To review from Section IV, the major existing uncer- tainties in atmospheric science are: Quantification of CFC emissions actual spheric removal rates. tropo- "Scientif ic uncertainties are introduced both in estimates of ozone depletion and in estimates of the effects of ozone deple- tion. The importance of the former to risk determination is discussed in this section; the importance of the latter in sec- tion V-C. The effects of these uncertainties are cumulative. At each successive step, it is difficult to intelligently discuss a given step and its uncertainties without tacitly or explicitly assuming "best guess" conclusions of each preceding step. As an example, the preface of the NAS/CISC report, on effects of ozone depletion [NAS,1979b], acknowledges dependence on the preceding NAS/PSCT report on the underlying science [NAS,1979a]. The cumulative nature of the uncertainties must be kept prominently in mind. V-6 ------- Risk Quantification of actual transport rates of impor- tant species in the troposphere and stratosphere, vertically and laterally. Lack of determination that relevant atmospheric chemistry is either accurately or completely understood. Lack of demonstration that the computer simulations ("models") used can provide accurate predictions of future atmospheric conditions. Interpretation of model results and their relation- ship to the present and future atmosphere. Quantification of ozone changes into changes in UV-B and then into changes in damaging UV (DUV) using appropriate action spectra. (An uncertainty that bridges the atmospheric science and the effects of ozone change.) Discrepancy between calculated ozone depletion and observed trends in ozone. It is commonly argued about the nature of uncertainty that the high range is as likely as the low range and since the high range, if correct, implies greater risk, it should figure disproportionately in risk assessments. However, this should not be true if the predicted risk occurs gradually over time, and methods exist for detecting the effect. Under these condi- tions, the larger predicted environmental effects in the uncer- tainty range clearly should be detected fairly early. If they are not detected, one may conclude that the high range can be adjusted downward. The CFC/Ozone Issue is just such a case. In the CFC/Ozone Depletion Issue, ozone trend analysis already provides an early warning technique for changes in ozone. The absence to date of any detectable downward trend in ozone concentrations indicates that the greater risks associated with the high range of ozone depletion estimates may be discounted. V-7 ------- Risk In addition, a monitoring system for gathering and reporting CFC production and emissions has been in place for several years, so a high range of ozone depletion as a result of unknown 2 emissions is not a concern. The uncertainties which exist surrounding the detection of ozone depletion (about +1 to + 1.5 percent depletion) and quantification of CFC emissions (about + 5 percent) are subject to further reduction but are already small in proportion to the uncertainties surrounding our understanding of atmospheric pro- cesses and our ability to model those processes. Ideally, the uncertainties are reducible to the point where measurements and modeling are consistent with each other. As discussed in Section IV, the opportunity for major progress in this direction exists in the coming years. (Meanwhile it would be folly to reject that which is measured with relatively great certainty in favor of that which is calculated with rela- tively little certainty). In fact, already there is evidence that the uncertainty ranges assigned for the modeling of some atmospheric processes have been underestimated (See Appendix E); more recent values of certain factors lie outside the correspond- ing "confidence limits" adopted by NAS [NAS, 1979a]. Introduction of recent revisions into modeling has reduced the calculations or estimates of future ozone depletion. Other current research and better model simulation are expected to reduce depletion estimates further. Therefore, the most This is separate from growth. Growth in emissions, should current trends be reversed, would be known and not an addi- tional uncertainty. Uncertainty develops if emission levels are not known. V-8 ------- Risk likely scenario is a progressive reduction of ozone depletion estimates towards the shrinking uncertainty range, which is being established through continuing improvements in the ozone data base and in the technique of ozone trend analysis. And the combination of no measurable ozone depletion and model calcula- tions of declining theoretical ozone depletion reduces the risk associated with a regulatory postponement while the research .progresses. The following section discusses EPA's treatment of uncertainties in the ANPR. Because of the cited deficiencies, EPA's risk assessment substantially overstates the risk relative to a risk assessment based on an up-to-date treatment of uncer- tainties (as highlighted above). 3. Errors Made by EPA in Treatment of Uncertainties a) EPA Over Relies on the "Key Findings" Section of the NAS Report [NAS, 1979a] The first major error in EPA's treatment of uncertain- ties is that the Agency overestimates and misrepresents the cer- tainty of future ozone depletion calculations, apparently due to overreliance on the "Key Findings" Section of the NAS Report [NAS, 1979a], while ignoring qualifying statements in the .body of the report. It is clear from a careful reading of the full NAS Report that the panel was hazarding a rough estimate of the probability of stratospheric ozone depletion by CFCs based on information available to them at the time of writing (summer 1979) . As examples, the NAS Report acknowledges (p. 17): "There are two possible sources of error that, inherently, cannot be quantified." and, V-9 ------- Risk "It is obviously impossible to estimate the unknown with any precision." Such qualifications do not appear in EPA's ANPR discussion of uncertainties in ozone depletion estimates. b) EPA Places Sole Reliance on the NAS Report [NAS, 1979a] A more basic error on the part of EPA is its exclusive reliance on a report which, even at the time of issuance, did not treat adequately the question of uncertainty and which was in conflict with other assessments available at that time. The adequacy of the NAS treatment of uncertainties itself is discussed in greater detail in Du Font's submission [Du Pont, 1980a] , where it is noted that the NAS Report's conclusions: * Conflict strongly with considered viewpoints of much of the world's scientific community, for example, the October 1979 report by the United Kingdom's Department of the Environment [UK DOE, 1979] . * Are not substantiated by the data used in the NAS Report. Are based, in part, upon serious inconsistencies and omissions in the body of the NAS Report. Attempt to quantify fully the uncertainties in the calculated predictions by including subjective and unsupported assignments of precise uncertainty ranges due to factors which the body of the NAS Report itself describes as "unquantifiable." V-j.0 ------- Risk A comparison of the treatment of uncertainty by the NAS and the UK reports is attached as Appendix L, but two quotations from each will serve to illustrate the differences: NAS, 1979a UK DOE, 1979 "The uncertainties in the chemical rate coefficients, in atmospheric transport, and in the use of one-di- mensional models have been combined to give an overall uncertainty range of a fac- tor of 6 within a 95 per- cent confidence level." "There have been consider- able improvements in the computer model and in the laboratory and atmospheric measurements which have reduced the uncertainty range." "It is not therefore realistic to assign over- all uncertainty limits to our calculated ozone per- turbations; deficiencies in our basic knowledge of the processes establishing the composition of the stratosphere and in the modeling technology cast doubts on their validity." "The STRAC [The UK Strat- ospheric Research Advisory Committee] report deals extensively with the uncer- tainties in the model re- sults. Not all of them could be assessed quanti- tatively and it is not possible to assign error ranges to these estimates that allow for all the un- certainties. These have, however, widened rather than narrowed since Pollu- tion Paper 5 was published [1976]." c) EPA does not Acknowledge Conflict Between the NAS Report and More Recent Reports A related error in EPA's presentation of uncertainty in the ANPR is in not weighing the NAS Report's treatment of uncertainties against reports which have become available since issuance of the NAS Report, for example, the European Economic Community, June, 1980 analysis of the science [EEC, 1980]. V-ll ------- Risk The key findings of the EEC report, detailed below, provide sharp contrast with the NAS assessment: There is now much more information available about the photochemical theory of ozone in the strato- sphere than there was ten years ago. There are still more uncertainties, however; The models have helped to improve knowledge of the stratosphere; As they [models] are simplified, they cannot fully describe the behavior of the atmosphere and its minority constituent parts; In the next few years more sophisticated models must be developed which can take into account simultaneously the chemical, thermal and dynamic aspects of atmospheric processes; This is a task which cannot be completed within five years but steady efforts must be made in this direction; Permanent observation and monitoring of ozone are therefore particularly important; At present there is nothing to indicate that CFCs have had a genuine effect on the ozone layer (emphasis added); Observation facilities should therefore be deve- loped i.e., both satellite measurements, which supply a large number of observations, and ground measurements, which are easier to calibrate; The examination of the balance sheets of the minority constituent parts should be continued in order to detect natural and artificial sources or sinks of these compounds in the atmosphere; It is vital to study simultaneously all the effects of human activities on atmospheric ozone. V-12 ------- Risk The problem of ozone and its vulnerability to compounds of human origin has now become a per- manent problem. The figures now advanced will have to be revised frequently to take into account the development of knowledge, the degree of sophistication of the models and the observations of the minority constituent parts. d) EPA Relies on an Out-of-Date Report While Ignoring Recent Critical Developments in the Science A last failing by EPA is in not weighing the accuracy of the NAS [1979a] treatment of uncertainty against recent developments in the science. As an example, the NAS rather confidently assessed that it was quite unlikely there would be any major changes forthcoming in the area of reaction kinetics. Yet, work within the last year has produced results which sig- nificantly change a number of reaction rate constants, with the consequence that the depletion prediction made by the NAS of 16.5 percent now is reduced by approximately half. Refinements in chemistry and atmospheric modeling could lead one to conclude the uncertainties are being narrowed and therefore the ability to make a proper regulatory decision has been enhanced. However, to the contrary, the refinements have led to an increased awareness of the large existing uncer- tainty. Moreover, the results of ozone trend analysis further throw into question the utility of the NAS1 assessment. These critical actual measurements (which have been available to the policy maker) are the only measurements which reflect what actually is happening to stratospheric ozone. They indicate that no problem is developing. As more measurements become available, the certainty of what actually is occurring increases, and the measurement series can better test the validity of calculations of ozone depletion based on the theory. V-13 ------- Risk C. IMPACT OF UNCERTAINTIES IN THE POTENTIAL EFFECTS OF OZONE DEPLETION ON RISK 1. Introduction The underlying risk issue actually is not whether, or to what extent, CFCs may deplete stratospheric ozone, but rather, what would be the consequences to human health and the environment should such depletion occur. More attention needs to be given to these possibilities and to the likelihood of their occurring. Too often, discussion has focused on a numbers game between various computer calculations of a hypothetical, far in the future, depletion of ozone (which assumes continuing emissions at current levels ad infinitum). A more realistic assessment would be: "If ozone depletion were to occur at x% per year, what would be the unavoidable future consequences for each ongoing year of emissions at current levels, and to what extent would these consequences justify curtailing CFCs now from their current uses?" Critical factors in this evaluation are the confidence with which these projections for the future are made and the time it will take to improve the confidence level. In other words, what is the incremental risk of delay? A related question has to be: "How much depletion could occur before the attendant increase in UV caused a problem, i.e., what is the danger threshold of ozone change?" Our observations here are almost in parallel to those presented in the previous section: a) the more uncertainty surrounding these estimates, the less likely a correct regula- tory decision can be made, and b) EPA's assessment of these uncertainties is grossly out of phase with the best currently available information. In fact, the uncertainties associated with.the predicted effects of ozone depletion out-weigh even the very significant uncertainties associated with the issue of whether ozone depletion is occurring as predicted. This is V-14 ------- Risk because the data are more sparse and the relationships more tenuous. Yet a regulatory decision should be based more on whether any harm is developing, or is likely to develop, than on whether certain chemical or physical changes may occur in the stratosphere. Thus, the question of uncertainties in the area of potential effects of ozone depletion are of paramount importance. Because EPA again bases its assessment almost ex- clusively on the NAS Report, we begin our discussion with a critique of the NAS Report by the Committee on impacts of Stratospheric Change (CISC) [NAS, 1979b] . In Du Font's earlier critique [Du Pont, 1980b] of the CISC report we commented: "The CISC Report shows that there has been a di- verse, but neitner extensive nor definitive, research effort into the possible impacts of increased damaging ultraviolet light (DUV) on the world's plant and animal (including human) eco-systems. That effort has demon- strated a number of potential interactions between in- creased DUV flux to the earth's surface and those eco- systems. However, the significance to the real world of those potential interactions (with the possible ex- ception of nonmelanoma skin cancer) has not been adequately demonstrated. Virtually all of the exper- iments have been exploratory or preliminary in nature. The body of the CISC Report and its associated appendices generally provide a rational discussion of all relevant experimental work to date, taking parti- cular care to point out a variety of experimental uncertainties and failings that may affect any conclu- sions to be drawn. Most conclusions found in the body of the report also include relevant qualifications. However, at least three of the "Key Findings" (concern- ing melanoma, damage to crops, and damage to aquatic organisms) go beyond the evidence presented in the report. Each describes as fact something which is explicitly stated in the report as an unverified possibility. Since this report describes a portion of the evidence on which EPA will make its regulatory decisions, it is imperative that the Agency consider the whole report, avoiding reliance 'on the overly conclusive "Key Findings." V-15 ------- Risk And we indicated in the letter of transmittal to EPA [Halter, 1980] that we considered: "...the subject matter of the CISC Report necessitates the assistance of outside consultants in order to prepare more complete comments. Should we be successful in obtaining more in-depth critiques, these will be forwarded to you." The critiques have been obtained and are listed below: Predicted Effect/Concern from Ozone Depletion Human Skin Cancer (Appendix F-l) Measurement and Instrumentation (Appendix F-2) Agricultural Crops (Appendix F-3) Aquatic Ecosystems (Appendix F-4) Author/Affiliation Professor Frederick Urbach, M.D, Center for Photobiology Temple University School of Medicine Philadelphia, PA Dr. Wilj.iam H. Klein, Director Smithsonian Radiation Biology Laboratory Rockville, MD Professor R. Hilton Biggs Institute of Food and Agricul- tural Science University of Florida Gainesville, FL Dr. David M. Damkaer University of Washington Seattle, WA There is striking unanimity in the appended reviews. Each reviewer acknowledges that the body of the NAS report and its appendices provide a reasonably good status report, as of 1979, of the several areas of knowledge, and point to the numerous caveats and acknowledgements of inadequate data bases for conclusions contained therein. The reviewers differentiate this status report, which is what it was intended to be, from V-16 ------- Risk any form of final report, which it was not intended to be. The reviewers express substantial concern over the summary and Key Findings sections where the state of knowledge is oversimplified and where the multitude of necessary qualifications are largely omitted. We note that it is these very sections of the report which seem to predominate EPA's statements. There is further unanimity that in all effects areas there is no data base to predict quantitatively the effects of ultimate depletion and no basis to predict an imminent catas- trophe . The need for research is emphasized strongly, and fre- quently specifically. The consultants conclude that the data base identifies possibilities which should be the basis for such further research. And it is acknowledged that there is time for research and that the hazard in waiting for research results for a limited period is negligible. Specifically, Dr. Urbach comments: "Finally, calculations, using worst case assumptions based on NAS data, strongly suggest that a 5 year delay in regulation of CFC will not have a dis- cernable effect on increases in incidence of [nonme- lanoma skin cancer] or [malignant melanoma]." And Dr. Biggs concludes: "...the degree of uncertainty that is associated with the possibilities would seem to indicate that the best course of action would be to proceed for a limited period of time to mount a good research effort to reduce (a) the uncertainties associated with knowing the degree of stratospheric changes expected in relation to time, say five years when some verifi- cation of whether stratospheric ozone changes predicted by atmospheric scientists is actually occurring, and (b) those uncertainties associated with biological effects of UV-B radiation on plants." V-17 ------- Risk Highspot summaries of the consultants' findings follow in sections 2-6. The full reports appear in Appendix F. 2. Human Skin Cancer Effects In the ANPR, EPA quotes NAS' [1979b] estimates of increased skin cancer incidence without mention of the cautionary statements in the NAS reports or the many conflicting conclusions on the causes of skin cancer, melanoma in particular. In this section, we will very briefly present some qualifications which should be made on EPA's and the NAS/CISC Report's statements on skin cancer. These statements are based on the review recently performed by Dr. Frederick Urbach for the Du Pont Company. The most important conclusion by Dr. Urbach, from the point of view of the regulatory decision-maker, is that no dis- cernable effect on the incidence of malignant melanoma or non- melanoma skin cancer is to be expected due to a 5-year postpone- ment of regulatory action by the United States. (Dr. Urbach's full review appears in Appendix F-l). a. Melanoma Skin Cancer For most forms of malignant melanoma (the rare but often fatal type of skin cancer), medical data show that inci- dence is not related to chronic repeated damage from accumulated doses of UV-B. There is, thus, no possibility of assuming any reasonable dose-response relationship between UV-B dose and changes in malignant melanoma incidence. The use by CISC of essentially the same "model" used to calculate changes in nonmelanoma skin cancer incidence from projected ozone depletion is unsupportable. V-18 ------- Risk What can be said about malignant melanoma is that a real worldwide increase in incidence has occurred in the absence of any ozone depletion and that it is middle class males and females who show this increase, not those habitually or occupationally exposed to UV-B. No discernable effect on malignant melanoma incidence is to be expected from a 5-year postponement of any regulatory action on CFCs. The above conclusions should be contrasted with EPA's ANPR statements: "Assuming continuation of present patterns of sun- light exposure, NAS predicts a 16 percent ozone deple- tion would result in several hundred thousand addi- tional cases of nonmelanoma skin cancer annually, and, with somewhat less certainty, in several thousand addi- tional cases of melanoma skin cancer (often fatal) annually in the United States alone." and: "For melanoma, this statistical relationship [between increased incidence and increased ozone depletion] is less certain but appears to be about two to one." The semiquantitative estimates of malignant melanoma referred to in the ANPR are speculative. b. Nonmelanoma Skin Cancer Nonmelanoma skin cancer is the common but rarely fatal skin cancer. It has the best prognosis of any cancer. While a correlation between most, but not all, nonmelanoma skin cancer incidence and solar UV-B exposure can be reasonably inferred, existing "models" for quantifying the dose-response relationship need considerable refinement. V-19 ------- Risk Present methods for estimating changes in nonmelanoma skin cancer incidence following a calculated change in UV-B clearly overestimate effects. Present methods, for instance, uncritically use the worst relationship between calculated changes in UV-B and the presumed damage that such changes could produce, and they make no allowance for attenuation of UV-B in the outer skin cells. These outer skin cells are not capable of being transformed into cancer cells since they typically are not capable of division. Dr. Urbach's review concludes that the NAS/CISC [NAS, 1979b] projections of increased skin cancer are based on all nonmelanoma skin cancer, while in fact about one-third of basal cell carcinoma occurs on sites and under conditions which suggest this subset has no relationship to UV exposure. Further conclusions are that: Existing models, in- cluding the calculations made by CISC, need considerable refine- ment before realistic estimates of changes in nonmelanoma skin cancer incidence can be made for projected depletions of ozone. Present models, including the techniques used by CISC, clearly overestimate the risk. No discernable effect on nonmelanoma skin cancer incidence is to be expected from a 5-year postponement of any regulatory action of CFCs. These conclusions should be contrasted with EPA's emphatic statements in the ANPR that: "A relationship has been epidemiologically estab- lished between increased DUV exposure and incidence of nonmelanoma skin cancer. For nonmelanoma skin cancer approximately a four percent increase in incidence can be expected for every one percent increase in ozone depletion on average." V-20 ------- Risk 3. Natural Variations in Normal Background Solar Radia- its Simulation and its Measurement Comparison of data between Rockville, MD (39 N) and Panama (9°N) shows approximately a 340 percent natural increase in ultraviolet radiation across this 30° latitude band, or approximately 0.16 percent average increase per mile (see Appendix F-2). The variation is almost 8 times larger than the 44 percent ultraviolet increase which NAS/CISC [NAS, 1_9 7_9_b_j ca 1 cu 1 ated J:o_ result from an ultimate ozone depletion of 16.5 percent. This observation provides some needed perspective. It is illuminating to compare these numbers to the 0.2 percent potential maximum incremental ozone depletion difference, corre- sponding to a 0.6 percent increase in damaging ultraviolet radiation, between a U.S. ban in 1980 versus a ban in 1985 [Du Pont, 1980b] (Details of this calculation appear in Appendix E) . The 0.6 percent difference may be compared to existing ultraviolet increments over a north-south movement of 4 miles. However, more importantly, it indicates the existing environment provides ultraviolet differentials much in excess of our current concerns. It also suggests that the natural envir- onment readily could be used for experiments on the effects of varied amounts of ultraviolet radiation on representative crops. Such experiments would avoid the need for growth chamber exper- iments. This is important because there are critical physical deficiencies in our ability to simulate changes in ultraviolet radiation using experimental growth chambers. Yet, it is these growth chambers on which most data are presently based and on which EPA reaches its conclusions. This is discussed in more 340 percent * 30° * 69 miles/0 latitude = 0.16 per- cent/mile ------- Risk detail in the following section. (The above points are deve- loped more fully in the review by Dr. William H. Klein, which appears in its entirety in Appendix F-2). 4. Crop Effects With regard to the effect on crops from a 16 percent depletion of the ozone, EPA states in the ANPR: "Other significant effects of increased DUV may include reduced crop yields from many important agricultural species, including tomatoes, sugar beets and corn..." This statement exaggerates the NAS conclusion, which was: "Key Finding 12 - Crop yields from several kinds of agricultural plants are likely to be reduced as a result of a 16 percent to 30 percent ozone depletion. Present data does not permit a quantitative estimation of the expected production losses but do show differences in the ultraviolet sensitivities of different plants cultivated in the United States. Since non-agricultural plants show ultraviolet sensi- tivities in the same range as do agricultural plants, effects of ozone depletion on wild and cultivated plants should be similar." The Key point here is that the NAS limited itself to con- cluding effects may occur from depletion in the range of 16-30 percent, a depletion range in excess of the calculated ultimate potential depletion of 16 percent associated with 1977 pro- duction. Additionally, the body of the NAS report strongly emphasizes the paucity of good data and the numerous uncer- tainties. These factors have been ignored by EPA in its summary position in the ANPR. V-22 ------- Risk Specific to the findings of the NAS/CISC report [NAS, 1979b] , Dr. R. Hilton Biggs (See Appendix F-3 for his full review) reports that "No studies of plant responses per se^ [in controlled environmental growth chambers] have ever been used successfully to quantitatively predict crop yield under field conditions." and "[controlled] environment growth chamber studies cannot be used to extrapolate to field condi- 4 tions." Yet, almost all the evidence cited in the ANPR for predicting adverse effects of increased ultraviolet radiation on crops is based on growth chamber studies. The discrepancy between growth chamber and field experimental results may be illustrated by comparing two findings which appear in the NAS Report: * From growth chamber studies: "...soybeans tend to be generally sensitive to UV-B radiation..." [NAS, 1979b, p. 284] But from field studies, a contradictory conclusion is offered: "Several crop species such as....soybeans... exhibited no detectable response to UV-B radiation supplements as large as two to three times the present solar DUV for summer conditions at 30°N under field conditions." [NAS, 1979b, p. 285] Dr. Biggs' review discusses in some detail the pertinent physical and biological reasons why growth chambers are unsuccessful models for quantitative predictions of crop yield under field conditions. V-23 ------- Risk The misleading high sensitivity of soybeans grown in growth chambers has been prominently and prolifically featured in EPA statements [EPA, 1980a; 1980c; 1980d; 1980e; 1980f; 1980g; Jellinek, 1980a; 1980c; Wellford, 1980] while the results from field experiments have been ignored. One can conclude that EPA does not appreciate the relative significance of the differing results. 5. Marine Effects EPA's key statements in the ANPR on the effect of a 16 percent ozone depletion on marine life are: "Other significant effects of increased DUV may include...; significant larval and juvenile killings of certain seafood species including anchovies, mackeral, shrimp and crab; and adverse effects on the microorganisms constituting the base of the marine food chain." This should be contrasted to the full NAS Conclusion which was: "Key Finding 13 - Larval forms of several important seafood species, as well as microorganisms at the base of the marine food chain, would suffer appreciable killing as a result of a 16 to 30 percent ozone depletion. Present ignorance of ultraviolet penetration into the waters that they inhabit and of the depth distribution of the organisms precludes an estimate of actual losses." [NAS, 1979b, p. 7] Again it is pertinent that the NAS limited itself to concluding effects may occur from depletion in the range of 16-30 percent, a depletion range iji excess of the calculated ultimate potential depletion of 16 percent cited by EPA. Additionally, the body of the NAS/CISC Report strongly emphasizes the paucity of good data and the numerous uncertain- V-24 ------- Risk ties. These factors have been ignored by EPA in its summary position in the ANPR. In many of the experiments cited by the NAS Reports, and subsequently by EPA in the ANPR, the experimental conditions, such as ultraviolet radiation dose-rates, temperature and water depth were not realistic simulations of natural conditions. For some studies cited, the radiation dosimetry and calculations of the radiation level are open to substantial question. (These and other points are discussed in detail in Dr. David M. Damkaer's review which appears in its entirety in Appendix F-4). The bottom line to the uncertainties surrounding the results from these experiments is somewhat analagous to the situation with regard to crops (discussed in the previous section)the data represent the necessary first step of ex- ploratory research, but are not suitable as a final base on which future quantitative predictions can be made. 6. Climatological Effects EPA states in the ANPR: "In addition, continued accumulation of CFCs in the lower atmosphere (troposphere) may induce a slight warming of the mean global surface temperature, but this is less than the warming predicted for a doubling of atmospheric carbon dioxide caused by fossil fuel combustion." The text infers these are part of the NAS conclusions. What the NAS does say about potential temperature change is that: V-25 ------- Risk i. "...the warming due to...CFMs is expected to be an order of magnitude smaller than that expected from the increased CO?." [NAS, 1979b, p. 118] ii. The [uncertainty] is so large, in fact, that the net warming due to... [CFMs] has an uncertainty equal to the expected mean." [NAS, 1979b, p. 118] iii. ...a change between successive nonoverlapping 20-year averages of surface temperature at 60°N must exceed 0.4° to be statistically significant..." [NAS, 1979b, p. 106] One could equally well state that the potential temperature change from CFC release is so uncertain that it may not occur at all. But even if a temperature change from CFC release were to occur it would be an insignificant increase relative to that possible from C02- V-26 ------- Risk D. PROBABILITY AND TIMING OF REDUCING UNCERTAINTIES 1. Introduction As discussed in section B, Section IV and Appendix E, the science underlying the CFC/Ozone Issue is developing rapidly. In the year since the NAS reports, there has been a substantial increase in knowledge in a number or critical areas, which, in turn, has served to narrow several of the key remaining uncer- tainties. The narrowing of uncertainties is a continuous pro- cess. If research continues as expected, the reduction of uncertainties likely will extend into the foreseeable future as the data base increases and understanding of atmospheric pro- cesses improves. However, one cannot confidently predict exactly to what extent, over time, this will happen. Yet the very question often asked is: "When will the uncertainties be reduced sufficiently to confidently make a determination as to the validity of the theory?" Consequently, attempts have been made to assign to the key uncertainties time estimates for their resolution. We review two of them [Ward, 1979; SRI, 1980] in the following sections. We end with a discussion on why it does not matter how long it will take to resolve the uncertainties so long as an early warning system for any developing problem is available. 2. Du Pont/Fluorocarbon Project Panel Estimates At the request of EPA, Du Pont prepared, in coopera- tion with the Chemical Manufacturers Association (CMA) Fluoro- carbon Project Panel (FPP) a submission reviewing uncertainties in the ozone depletion theory [Ward, 1979] The submission's summary included the comment: V-27 ------- Risk "This submission identifies several important uncertainties in the ozone depletion theory, explains the significance of the uncertainties/ reviews the industry-supported research targeted to the uncertain- ties, and gives the results expected and anticipated timing. The program is consistent with the recommendations of the National Academy of Sciences and cooperative with government agency and academic research. The industry believes that there is time to verify or disprove the theory experimentally without undue risk. The issue should be decided on such scientific measurements and evaluations, not just upon unverified theory." The estimates of timing were arrived at in consultation with the investigators funded by FPP and with the FPP project coordinators. A copy of the tabulated contents from the submission appears on the following page. In reviewing this estimated, timetable however, it is important to recognize that, although in each case the expected results are anticipated in 5 years or less, the anticipated time required to obtain results can change as research progresses. 3. SRI Workshop Conclusions Subsequently and independently, EPA sponsored a "work- shop" at SRI International in March, 1980. None of the FPP/CMA coordinators, and no CFC-producing industries were invited to attend the workshop. The conclusions of the workshop partici- pants, as reported in the report on the proceedings [SRI, 1980] seem to have been that: i) There are many key uncertainties remaining which are critical to making a correct regulatory decision, V-28 ------- Uncertainty Significance Research Estimated Time Required Are there trope-spheric sinks? Is all pertinent chemistry known and quantitatively included? Are 1-D models adequate? Ozone Trend Analysis 10 year sink, not currently excludable, would reduce ozone depletion 10-fold. Uncertainties associated with reaction rates, photolysis, and reaction products could together reduce ozone depletion 13- fold. Missing chemistry could reverse sign of ozone depletion (ozone augmen- tation) . Transport uncertainties could reduce ozone depletion by a factor of 0.7. Latitude and seasonal dis- tribution for comparison of measurements and calculations. Latitude distribution of calculated depletion reduces impact. Atmospheric Lifetime Experiment Research on silica-catalysed decomposition and PC-21. Studies on reaction rates and temperature dependence. Absorption cross-section measure- ments. Studies on reaction pathways. Studies on hypothetical missing chemistry suggested by discre- pancies between measurements and calculations. Anderson's "reel-down" experiment Methods for, and measurements of, total stratospheric chlorine. Not included 2-D modeling studies. 2-D model development and utili.zation. Basic test of whether CPCs Statistical Trend Analysis deplete ozone. (Mote that Establishment of Detection this is not the same as Threshold whether ozone is being depleted from any sourcea detection ability already in hand.) 3-5 years 1 year for measurements 3 years 3 years 3 years 5 years 2 years 2 years Not applicable 2-3 years 2-3 years 2 years or more depending on actual detection threshold. < VD P. tn ------- Risk ii) These uncertainties generally are "researchable", and iii) Most are researchable within 1-5 years, depending on the uncertainty. The summary tables on uncertainties from this report appear as Appendix L. The full report is attached as part of our submission. It is noteworthy that of a broad range of 37 issues classified as "high" or "moderate" in importance and identified as "researchable", 28 were listed as requiring 5 years or less and an additional 4 were expected to require 5-10 years. Of the remaining 5 researchable issues requiring 10 years, and the 6 "unresearchable" issues, none involved CFC releases and none involved transport or atmospheric chemistry and modeling. 4. Conclusion The research needs described in the two reports above are completely compatible with research recommendations made by the NAS [NAS, 1979a; 1979b] and NASA [NASA, 1979], although these later reviews did not make specific estimates of the time required. However, the NASA report did conclude: "The uncertainties associated with estimating the long-term impact of several perturbing influences to the stratospheric ozone layer continue to be large. However, prospects appear good for improving the situation in the near future." (Emphasis added.) [NASA, 1979, p. 362] The conclusion is clear and inescapable. The uncer- tainties concerning the effects of CFC on stratospheric ozone, and the effects of ozone changes should they occur, are so large V-30 ------- Risk that a reasoned and logical decision-making process for further CFC regulation cannot presently be mounted. There has been a tendency on the part of EPA and others to acknowledge the uncertainties, but, understandably lacking confidence in their very near-term resolution, to con- clude that an indefinite wait for the needed resolution will incur unreasonable risk. Therefore, it is argued that regulation must occur now. What this analysis neglects, however, is the existence of an early warning systemozone time-trend analy- sis. From a regulatory standpoint, it should not matter whether the underlying scientific uncertainties will be resolved in 1, 5 or even 20 years, so long as it is apparent that an unreasonable risk to human health or the environment is not developing during the period. Trend analysis provides such an ability. Under the umbrella of an early warning system based on actual ozone obser- vations, the time it will take to resolve the uncertainties becomes of academic interest. Research to ultimately resolve all the discrepancies and unanswered questions can proceed with confidence, no matter how long it might take, so long as no risk is developing. The trigger for regulation should be "developing risk", not concern that it may take x years to get the final answers, and that if these answers are negative it will be too late to act. These points are discussed in more detail in the ensuing sections E-G. V-31 ------- Risk E. RISK IN WAITING-RISK VS. TIME 1. Introduction In the United States many of the studies [NAS, 1979a; 1979b] have concentrated on calculations of ultimate or long-term depletion and on attempts to predict the effects of such long-term depletion should it in fact occur. The reports note that such effects occur gradually over a period of about a hundred years. However, exclusive consideration of the long- term calculations and estimates is largely inappropriate for decisions as to what action should be taken now. Such long- range estimates are mainly relevant to a decision as to whether action on the issue should be postponed for a long period, say 50 years. The relevant decision that is facing us now is whether regulation of the remaining nonaerosol uses of CFCs in the United States has to be decided upon and initiated now (1981). The alternative to immediate regulation is postponement for limited sequential periods. We are not recommending that postponement or deferral be for a single pre-set specific V-32 ------- Risk period. Rather, we believe deferral should be on a sliding scale, say a year at a time, in conjunction with periodic formal assessments of whether the theorized risk is developing ("Assess- ment and Surveillance"). As discussed elsewhere, ozone trend analysis permits this to be done. The deferral should not be just to postpone regulation but rather, in order to: a) continue to work to reduce uncertainties in the ability to make stratospheric predictions, b) reduce uncertainties in our ability to estimate the effects of ozone depletion, c) insure the list of available regulatory options is complete, and A pre-set period of 5 years has been often mentioned. Several analyses [Ward, 1979; SRI, 1980] of the work required to reduce the critical uncertainties suggest that most can be materially reduced during periods varying between one and 10 years, with a period of about 5 years being most often stated. Reduction of uncertainties is a continuous process but the period of 5 years is selected as one within which major improvements in knowledge and re- ductions in uncertainties are expected (see previous section). More recently, the EEC reviewed the state of the science [EEC, 1980] and the need for further CFC controls and concluded: "...a delay of 5 years before any decision is taken on CFCs can be reasonably accepted." The key point underlying all these assessments is not that a 5 years deferral in regulation has some special significance but rather, given the current state of knowledge, studies in reducing the uncertainties probably can be made within this period of time, and the risk in waiting is acceptable. With the recent improvement in ozone trend analysis, we have the ability to assess developing risk on a yearly basis and to rethink the wisdom of continuing the deferral according- ly. No committment is required that deferral be for any pre-set period. V-33 ------- Risk d) complete and quantify the economic impacts of such regulatory actions. We stress that the relevant choice should be between regulating now or waiting for limited sequential periods while monitoring on an ongoing basis whether any risk is developing. It should be kept in mind that as our knowledge improves and more trend analysis of actual ozone concentrations becomes available, the decision to continue to wait may be altered at any time. The decision is not, as EPA tries to paint it, a simple choice of regulating now or having to wait until it is too late to head off major deleterious efforts should the theory prove to be correct. The risk basis for such a decision is not long-term or ultimate ozone depletion estimates but estimates made for incre- mental depletion and corresponding estimates of effects for the postponement period under consideration. Such analyses are conspicuously absent in the NAS Reports [NAS, 1979a; 1979b], although not from all the reports EPA has received or spon- sored. Fortunately, it is relatively simple to make the calculations, and the anticipated effects can be prorated. Below we review conclusions reached in a Du Pont sub- mission [Du Pont, 1980b], a study from the University of Maryland [Bailey, 1980] and a study by Systems Control, Inc. [SCI, 1979] which are pertinent to the question of risk from limited periods of delaydelay taken in order to reduce the uncertainties. 2. Conclusions from 1980 Du Pont Submission The Du Pont submission [Du Pont, 1980b] examined the calculated environmental difference between two extreme hypothetical regulatory scenarios: V-34 ------- Risk A ban on all CFC production in the U.S. in 1980. A ban on all CFC production in the U.S. postponed until 1985. The difference in potential ozone depletion between these two scenarios would reach a maximum of 0.2 percent in the year 2010 and subsequently decline. The scenarios assume con- tinued use at 1978 rates by the rest of the world. The 0.2 percent maximum incremental depletion corresponds to a 0.6 percent maximum incremental increase in ultraviolet reduction, also occurring in 2010 and subsequently declining. Such incremental changes in ozone and ultraviolet radiation would be insignificant and undetectable (See Appendix F). Similar calculations could be repeated for a five-year postponement in any regulatory scenario. Any regulatory scenario less severe than a production ban would necessarily result in even smaller calculated incremental effects. Although it is recognized that a total ban is not being considered as a practical regulatory action in 1980 or 1985, the extreme regulatory scenario was chosen to emphasize that the incremental effects from a 5-year U.S. postponement of even extreme action were insignificant. Details appear in Appendix E and [Du Pont, 1980b] (attached). 3 Conclusions ^f_rgm^University of Maryland Study The University of Maryland Study for EPA by Professor Martin J. Bailey [Bailey, 1980] is an in-depth evaluation of the risks vs. the benefits of not regulating. The "centerline" estimates are found to: "give the surprising result that the unregulated release of CFCs [if the theory of ozone depletion ------- Risk proves to be valid] would produce benefits ranging from equal to almost double the costs." [Bailey, 1980, p.ij An important part of the report is a section on "Immed- iate Decision versus Deferral", which discusses the value of reducing the existing uncertainties before making a regulatory decision. It is also pointed out that, if the U.S. prefers prompt regulation of CFC emissions, it can only present a mixed and inconclusive case to other countries (a key point, because one of EPA's stated objectives for regulating now is to obtain international cooperation on regulation of CFCs worldwide.) The report notes: "Because the growth path of CFC emissions is exponential, the next 10 years of production and use will commit the economy to only about 13 to 14% of the eventual risk of damage from CFCs if the supplies of Fl [fluoride] ores are exhausted, as expected within the next 10 years or so. During the next 10 years, scientific knowledge of all aspects of the problems should improve markedly. Hence, even if worldwide restriction of CFCs production and use fails to develop quickly, nearly all the risk can be avoidedthe risk that the case for restriction may become conclusion, as a result of new knowledge, after several years." [Bailey, 1980, Abstract of report] "All these perspectives on the problem, combined with the long time periods involved, imply a high value to improvements in knowledge, which can reduce the choice of a damaging or needless regulatory strategy." [Bailey, 1980, p.4] "Clearly, there is a high payoff to improving our knowledge and narrowing the uncertainties." [Bailey, 1980, p. 80] "...further regulation can safely be deferred until more knowledge accumulates." [Bailey, 1980, A-31 V-36 ------- Risk and concludes: ...the need is for a narrowing down of the major uncertainties within the next 10 to 15 years. The emphasis should be on well-verified projections and predictions not on quick [regulatory] responses. The rush to regulate has outgrown our knowledge in this area, apparently quite needlessly; there is ample time to reach reliable findings on whether an environmental hazard exists in the use of CFCs or whether instead they are beneficial." [Bailey, 1980, C-3] We find it noteworthy that the report has been omitted from EPA's ANPR discussion. 4. Conclusions from Systems Control Inc. Study The SCI study, [SCI, 1978] prepared for the National Science Foundation, evaluated internal and external costs associated with several alternative strategies for controlling CFC emissions. Among its conclusions was that, in most of the cases examined, the minimum cost alternative is to wait until ozone depletion is detected. It should be noted that at the time of the study (late 1978) the lowest detection "threshold" considered was 2.5 percent, while in 1980 the threshold is already down to approximately 1.5 percent or less (See Section IV and Appendix E). In the SCI study the cost advantage of the "Wait and See" approach increases as the detection threshold is lowered. Consequently this advantage of the Wait and See approach has increased as a result of improvements in ozone trend analysis. 5. Summary The conclusion to be reached from the above analyses is that the most appropriate regulatory policy is a cautious V-37 ------- Risk approach of "Assessment and Surveillance" for an ongoing series of limited periods, with suitable scientific review during each period, so that the wisdom of that policy is frequently reassessed. There are two further factors which support limited postponement of action: * The economic and other risks of unnecessary or premature regulation are sufficient to warrant great regulatory caution. (This is discussed in greater detail in section I.) There is an independent techniqueozone trend analysiswhich, apart from periodic scientific reviews, permits us to monitor our environment and which provides an early warning system. Ozone trend analysis provides a desirable redundancy in environmental monitoring. (This is discussed in greater detail in section F, Section IV - Science and Appendix E). We believe the factors discussed above are the factors which should be considered in reaching regulatory decisions on the need for short-term or immediate actions. They have been a major part of the reasoned approach adopted by the United Kingdom and the EEC. The reasoning followed by EPA (as described in the ANPR) in reaching its conclusions on regulatory action ignores these factors, concentrating instead on arguments which are inappropriate for reaching prudent and unemotional short-term regulatory decisions. (Section J discusses these arguments and approaches being used by EPA). V-38 ------- Risk F. IS THE RISK DEVELOPING AS PREDICTED? 1. Introduction Statistical analyses of ozone measurement data con- sistently point to recent increases in both total column ozone and upper Umkehr layer ozone concentrations. The uncertainties imply detestability of a long term trend at or below the 1.5 percent level. In contrast, 1979 models calculated current de- pletion to be 2.1 percent, with approximately 1.5 percent to have occurred in the seventies from CFC-11 and CFC-12. Other steadily increasing chlorocarbon emissions (e.g. methyl chloroform) would increase those numbers by approximately half (2.1^3 percent). Model calculations using current data revise these figures down- ward by as much as a factor of two, to as low as 1.5 percent - still outside the 95 percent confidence limits placed on the observations. This conflict between measurement and models is not unique. Model predictions of other effects based on best available data for volcanic eruptions, nuclear explosions and the eleven-year solar cycle are also not reproduced in actual ozone observations. Two explanations are possible for these continued differences between model calculations and ozone observations: i) In each case, an opposing trend may have masked the calcu- lated change. ii) Models may be consistently over-estimating depletion due to inadequacies in chemical treatment or to unwarranted assumptions. In other words, the theory is not quantitatively correct. These possible explanations are examined in more detail in the following sections. V-39 ------- Risk 2. Opposing Trends Ozone measurements have not verified the several concentration fluctuations calculated to have occurred from a variety of perturbations, including the CFC effect, either in magnitude or in qualitative features. Intuitively, it is unlikely that opposing but unsuspected trends in each case have countered the calculated trends just sufficiently to obscure detection. Nevertheless, such intuition is perhaps insufficient reason to reject the possibility. Even if one assumes for the moment that a CFC effect is occurring but being cancelled by other effects, the simple reality is that ozone levels themselves are not currently being reduced below "normal" by CFCs. From the statistical analyses, one also knows that postulated opposing trends must themselves be long-term effects and immediate catastrophic reversal is unlikely. Thus, in this scenario of opposing trends, for the near term at least (say up to twenty years) , any CFC effect is mitigated by the unknown opposing effect. Potential depletion is less, secondary effects are less, the risks are reduced. One must also consider what else will take place in the next twenty years. First, one may expect continual refine- ments in both model calculations and ozone measurements. Both will improve the regulator's ability to make decisions. Further- more, the assumption of opposing trends reduces the risk asso- ciated with delaying a regulatory decision until it can be made on a sounder basis. Measurements to date show no current threat and imply a reduction of future threat for a period long enough to encompass regulatory deferral for further research and even- tual action (if the evidence justifies it) and all of this at considerably lower risk than that perceived just one year ago. V-40 ------- Risk As this discussion points out, perceived risk is not a static quantity. During any period of regulatory postponement, ozone measurements would continue to be updated. As the data set lengthens, the statistical analysis improves in its ability to separate the CFC trend from unknown intermediate length trends, and only very long term trends could counteract the CFC trend without being detected. To the extent the depletion is still undetected in successive years, confidence in the wisdom of regulatory postponement increases. Even the time scale for possible future depletion increases. A final consequence of continued absence of detectable depletion is that opposing trends become less likely to mimic the calculated increasing CFC depletion curve. Eventually the postulate of opposing trends become highly unlikely, leading to the second possible explanation for current results: Theoreti- cal calculations may overestimate any CFC effect. 3. Reliability of the Theory The discrepancy between ozone measurements and model calculations may well lie in the theoretical models themselves. Having discussed the possibility of opposing trends, we now consider the alternative. The detailed discussion of transport, chemistry, tropospheric processes and models appearing in Appendix E define a large number of uncertainties associated with model calculations. The 1979 calculations by the Lawrence Livermore Laboratories' model were certainly overestimates relative to the currently accepted chemical scheme. Other preliminary developments seem likely to reduce calculated depletion still further. If one accepts several such develop- ments as being accurate, calculated depletion through 1978 is reduced and may then fall within the uncertainty range for trend analysis of the ozone measurements. Hence, in the absence of opposing trends, ozone trend analysis provides a gauge for V-41 ------- Risk confidence in model calculations. With current trend results, a lower calculated ozone depletion is more likely to be correct than a high one. Thus, the regulator obtains a great deal of direct information from ozone measurements. First, ozone is not currently decreasing as earlier model calculations would imply. Therefore, the immediate concern for stopping a trend which may be already underway is alleviated. Second, large (even 16 per- cent) calculated steady state depletion implies rather larger (approximately 2 percent) present day calculated depletion which is simply not consistent with the current data. The wis- dom of continuing to rely on 1979 model calculations in spite of this evidence is suspect. To the extent that the calculated steady state depletion is smaller, the calculated annual change in ozone is also smaller. Therefore, ozone measurements tell us that the need for precipitous immediate action is also removed. To put it simply, the smaller the potential effect, the more time available to study it before making regulatory decisions. A logical conclusion to this analysis is that both current uncertainties and recent analyses of ozone measurements support a deferral of regulatory action in favor of further research. The question then becomes: "How long a deferral is reasonable?" To answer that, one must consider the early warning capabilities of ozone measurements and how those capabilities improve with time. V-42 ------- Risk G. AVAILABILITY AND SIGNIFICANCE OF AN EARLY WARNING SYSTEM The feasibility of an early warning system based on ozone measurements depends not on the ability of the system to confirm the theory, but rather on the ability of the system to detect change in ozone. The ozone measurement analyses dis- cussed here imply that a net long-term change of approximately 1.5 percent is detectable with confidence, regardless of its cause. Calculations suggest that immediate cessation of pro- duction at that time would allow overshoot to half-again that value, i.e. approximately 2.3 percent. The practicalities of the regulatory process would entail some further delay, so a more conservative estimate of overshoot might imply an eventual depletion of 3 percent if the regulatory process were begun when ozone measurements indicate a decrease in ozone of 1.5 percent below "normal". However, this analysis of overshoot is essentially a worst case. It assumes a case calculated using 1979 model imputs. To the extent these model calculations are over- estimates of current depletion, (as seems to be the case based on new information) they are also overestimates of the rate of depletion and hence of the overshoot level. Each additional year of measurements which show no trend reduces both overshoot and potential steady state depletion still further. That is, the perceived problem gets smaller, and response becomes easier and more effective. A deferral is justified by the current data - ozone has not decreased as model calculations indicate it should Ozone levels fluctuate above and below some average value. Ozone depletion is a decrease in the average of the latest values compared to the long-term average of older ozone concentrations, i.e., the "normal" concentrations. V-43 ------- Risk have. Each additional year of measurements with no downward trend provides even stronger support for further deferral since a decreased potential effect can be more easily dealt with by the regulator. Additionally, a deferral in regulation will permit research to continue, and any subsequent decision to be reached on a more firm basis. A number of critical projects in progress (discussed in Appendix K) will be completed in the near future. Confidence in model calculations is likely to be increased as laboratory work and atmospheric measurements continue to refine the current picture of the stratosphere. Uncertainties will likely be reduced, and a proper decision will be facilitated. Finally, the early warning system itself is expected to improve with time for several reasons. A longer data base reduces the statistical uncertainty associated with detection of a long-term trend. The shape of the calculated CFC trend itself is better contrasted with other long-term trends as more years are included. The ability of a data set to separate a relatively long periodic trend from a very long term monotonic trend in- creases with the length of the data set, and the possibility of masking by opposing trends is reduced. Thus, the capabilities of the early warning system will increase with timebetter sensitivity and a more accurate search for a particular kind of trend. Careful attention to a well organized early warning system based on ozone measurements will indicate, with better precision each year, the actual changes taking place in ozone. Current precision is sufficient to permit speedy observation of change well in advance of large depletion. To the extent that ozone change is not detected, maximum likely CFC induced ozone depletion has increasingly smaller upper limits placed upon it. Regulatory deferral contingent on periodic reevaluation has the V-44 ------- Risk significant advantage of allowing for a better decision if and when it must be made. In the face of an uncertain theory and conflicting measurement of both 1) the quantity of interest - ozone concentrations - and 2) the crucial link in the theoretical reaction chain - chlorine oxide concentrations (See Section IV .and Appendix E) - any present day decision to regulate would be made without adequate scientific justification. V-45 ------- Risk H. THE RELATIONSHIP OF THE INTERNATIONAL ASPECTS OF THE ISSUE TO RISK CFCs and the potential depletion of the ozone is truly an international issue. This is so for these reasons: CFCs are produced and used worldwide, with the largest producer nation (U.S.) accounting for only approximately one-third of the total. If CFCs prove to deplete ozone, CFC emissions from all uses, in all countries, will contribute equally, i.e. a pound of emissions from a developing country will be just as bad as a pound from the U.S. A country cannot unilaterally protect its own overhead ozone. If ozone is depleted, it is depleted world-wide, regardless of whether a given country emits CFCs or not. Therefore, the problem, if it exists, can be solved only through a global effort. To this end, the question of risk must be addressed and managed internationally. Unilateral efforts will have little impact on the overall problem if it exists. The issue of risk as it applies to individual coun- tries' populations and to the world community, and as it applies to assessments and actions by individual countries, is an exceedingly complex one. Even the decision by an individual country that a chemical of known benefit, but uncertain risk to the population of that country, should be controlled is a complex one. In the instance of CFCs, it is ambiguous whether such a unilateral decision would have any benefit in reducing the perceived risk, regardless of the degree of any eventual certainty that the use of the chemical poses a risk to a country's population. V-46 ------- Risk Because of this complexity, discussions of risk and the international situation branch off into the science (how it is differently viewed around the world), international politics, and how an international solution to the problem can be forged. Individual facets of the problem, like risk management, cannot be discussed in isolation from related topics. Therefore, these and other parts of the international aspect to the CFC/Ozone Issue will be discussed together in Section VI. V-47 ------- Risk I. RISKS CREATED BY REGULATION AND THE NEED FOR RISK - RISK COMPARISON If the currently at issue uses of CFCs were "frivo- lous", that is, of no consequential redeeming value to society, or if there were readily available alternative products or pro- cesses, which at little cost to anyone could be used in place of CFCs and provide the same end products, with the same degree of safety, etc....then the need for addressing the trade-offs from regulation may not be critical. Under this circumstance the only major question before the policy-maker would be: "Is there an unreasonable risk from the continuing use of CFCs?" However, this is not the case with CFCs. CFCs are highly essential compounds to a vast variety of products and processes considered to be highly beneficial, and in some cases essential, to today's way of life. CFCs fill needs for society, else they would not be in demand. An analysis of the needs filled (See Section II) indicates that the needs are suffi- ciently basic that they must be filled regardless of the avail- ability of CFCs. Thus, restraint on CFC supply will result in pressures to meet the underlying needs in some manner. The chief manner will be the use of the best available alternative product or processes. But as also discussed in Section II, the use of the currently available alternative products which would be employed (if only because there were no other choices) would result in an increase in risk to workers and in some cases, consumers, over the situation now with CFCs. Also, to the extent that more energy is required due to unavailability of CFCs (See Section II-K and Appendix C) , there would be an increase in health and environmental risks associated with energy production and use. And last, as CFC regulation would create unemployment in the industries supplying precursors for CFC manufacturing, CFC manufacturing itself, and the numerous manufacturing, trans- V-48 ------- Risk portational and service industries associated with CFC use and products dependent upon CFCs, there is a need to address the social and welfare risks induced by unemployment. A few examples of the above risks are discussed below. Specific human health and welfare risks would be associated with a regulated reduction in the availability of CFCs due to the unique combination of physical properties CFCs provide. The most important is the combination of nonflamm- ability and low toxicity. Consider flammability. If problems of equipment re- design and conversion are ignored, hydrocarbons can be substi- tuted for CFCs in some uses, for instance pentane can be used as a blowing agent in the manufacture of certain polystyrene foams, and propane technically can be used as a refrigerant. Should such substitutions occur however, new fianunability and explosion hazards are created: 1) in the transportation of the hydrocarbon by rail or roadway, and at the storage point in the manufactur- ing plant site, 2) in the use of the hydrocarbon substitute in the manufacturing plant (This is particularly true if the pro- cess involves the release of the hydrocarbon in the plantas is the case with pentane/polystyrene foam production), and 3) if the substitute hydrocarbon is not released as a part of the manufacturing process, as would be the case with a propane refrigerant, new flammability and explosion hazards are created in the transportation and warehousing of the finished product, at the point of consumer sale, and during the life of the product in the consumer's hands. 7This last hazard is perhaps the most significant since main- tenance of adequate safety standards for flammable materials in consumers' hands is notoriously difficult. So serious is such a potential hazard in the case of refrigerants, as an example, that propane is banned [ANSI, 1971] along with other hydro- carbon refrigerants from use in institutional, public assembly, residential and commercial occupancy applications. Ammonia is under similar restriction. V-49 ------- Risk A parallel set of additional hazards is created when the substitute material is substantially more toxic than the CFC presently used. As examples, methylene chloride may be used as a alternative blowing agent for certain types of polyurethane foams, and ammonia or methyl chloride technically may be used as refrigerants. But toxicity hazards are created in the workplace if the substitute is emitted (as is the case with flexible polyurethane foams), or hazards are created for the consumer if the substitute, meant to be retained in the product, should escape (e.g. from a refrigerator or air-conditioner). Other problems can develop from the use of alternative solvents. Again flammability and toxicity problems may be created in the workplace where the alternative solvent is used. More insidious effects involve the potential for contamination of products with more toxic and less inert solvents. Sucn contamination could lead to direct toxic effects (for instance in the cleaning of medical equipment), or to hazards from failure or improper operation of containment equipment. There inevitably would be unforeseen risks. As a CFC-using practice is discontinued, novel and relatively untested technology is forced in with its own risks. Such novel technology may or may not be called for in the regulation, and any risks which may develop from the use of new products or processes may not be foreseeable, but the fact remains that if these risks develop, they are a cost of regulation. If CFCs are severely restricted, unemployment will result. Even if such unemployment is only temporary, amounting to dislocation of individuals from CFC industries to industries geared to replace CFC uses, there will be social and welfare costs associated. These costs are not just loss of income, loss of tax revenue or increased cost of social services. A recent study [Brenner, 1976] found that a one percent unemployment V-50 ------- Risk (about a million jobs) over a 6-year period is statistically related to: 36887 total deaths, including 20240 deaths from cardiovascular disease, 920 suicides, 648 homicides, and 495 deaths from cirrhosis of the liver. Additionally, the study found the social impact includes: 4427 first admissions to state mental hospitals, and 3340 admissions to state prisons. CFC dependent employment amounts to roughly 780,000 (See Section VII - Economic Considerations). Obviously, all of these jobs p would not be lost due to a cap on CFC production. But as losses would be expected, especially in the small business sectors (See Section VII), the above statistics should be kept in mind. Therefore, in conclusion, we believe it is necessary in any consideration of regulation of CFCs to carefully assess the consequences of reduced CFC availability in terms of the risks which such reduction might generate. This is true even if CFCs are found unequivocally to deplete ozone. And after the risks from reduced availability of CFCs have been determined, Q °Industry has not been able to estimate even approximately how many jobs might be lost were EPA's proposals to be enacted because EPA has not provided sufficient detail on its proposals (suchaicaplevel, lead-time,implementation,etc.)toenable any meaningful quantitative analyses. (See Section VII and Appendix I). V-51 ------- Risk they must be weighed against the risks from continued CFC use. Only a decision which minimizes total risk is in the best interest of society. There already has been some effort to address some of these risks and to balance them against the potential risks of CFCs [Du Pont, 1978; NAS, 1979d; Rand, 1980], And in EPA's Development Plan for CFC regulation [EPA, 1980e] it is stated: "Regulatory actions related to CFCs will cause increased use of substitutes which may have different but still undesirable effects of their own. To what extend do we attempt to do risk/risk tradeoffs?" So we conclude that EPA acknowledges this problem and tne need to do something about it. It is strange, therefore that the ANPR contains neither discussion of these risks nor recognition of the need to take them into consideration, nor does the ANPR solicit comment on them. The only mention is a passing reference in a dis- cussion of criteria for banning a CFC useno mention is made opposite the Agency's preferred option of a cap as an economic incentive option. We find a rule-making decision on CFCs which focuses solely on the potential long-term risks from continued CFC use, ignoring known risks (e.g. flammability and toxicity of alter- natives) which would be incurred from a restricted availability of CFCs, to be imbalanced, short-sighted and certainly not in the overall best interests of the public. V-52 ------- Risk J. APPROACHES WHICH ARE INAPPROPRIATE FOR ASSESSMENT OF RISK ON THE CHLOROFLUOROCARBON/OZONE ISSUE A growing world population, using advancing techno- logy, in a finite environment of which we have limited under- standing, will generate pressures and problems of extraordinary complexity. There are numerous examplesthree involving the atmosphere are the increasing concentration of carbon dioxide, acid rain, and the concern over possible depletion of ozone by CFCs. It is immediately obvious that the concern can be removed by banning or severely limiting the products in question, and the environmental risk will be removed. However, it is inappropriate to stop the analysis at that point, for now other risks emerge from this corrective action. It may be claimed, of course, that blame for the new risks lies with the entrepreneur attempting to fill the need created by the regulation. Whether or not this assignment of fault is legally sound does not eliminate the new risk. Place- ment of fault is not the point. The "fall-out" from any new risks is inevitably society's cost or society's risk. Arguments can be generated that, for one reason or another, seek to expedite regulatory decisions on environmental matters by suggesting society cannot afford the time to analyze risk thoroughly. In fact, society cannot afford not to make the analysis thorough. In the absence of thorough analysis, inappro- priate decisions will be made, forcing regulatory costs higher and net environmental benefits lower. In the ANPR and in previous Agency statements and documents, we have identified a number of approaches to the question of risk which we believe are not conducive to the obtaining of a proper risk determination for the CFC/Ozone Issue: V-53 ------- Risk 1. A preoccupation with extreme future extrapolation. 2. A conviction that immediate decisions are necessarily required, and are necessarily better than deferred decisions. 3. An excessive emphasis on political action over objective scientific decision-making. These factors are separately considered below: 1. Preoccupation with Extreme Future Extrapolation Unfortunately, the publication of, and incessant reference to, extreme risk scenarios seems to have been the rule, and not the exception, in the CFC/Ozone Issue [EPA, 1980d; .i.y80e; Jellxnek, I980a; iy80c]. These scenarios generally are predicated on two unrealistic assumptions: a) That world production and use of CFCs will continue to increase over the next several decades (EPA varyingly has used 7 percent and 9 percent as estimates of annual CFC growth for the next several decades) . Of interest to us is that the NAS calculated potential ozone depletion for four release scenarios (ATD) provided by EPA {NAS, 1979aj . Yet, the scenario most often cited by i£PA is case D, which assumed a 7 percent per annum growth in CFC emissions from 1980 to 2000--the only scenario which assumed uninterrupted growth. Two points bear making. First, such a scenario assumes there will be no further regulation anywhere in the world. In fact, ongoing regulatory impacts such as the EEC 30 percent CFC-11 and 12 propellant reduction are ignored by this scenario. Second, and even more telling, these scenarios fly in the face of what actually has been transpiring with V-54 ------- Risk total world CFC productioninformation which was provided to EPA well before publication of the ANPR [Hasten, 1980; Block, 1980]. As developed more fully in Apprendix J, the rate of production worldwide for CFCs has been essentially constant or slightly downward since 1975. This is the trendline on which current need for regulatory action should be based, recognizing that a substantial change in that trendline would be justification for reassessing the need for action. The trendline since 1975 corresponds to NAS Case A (constant) or possibly Case B (some reduction) but not Case D (7 percent per annum growth which EPA cites). Case D can be considered as an intellectual exercise, but its use to describe likelihood is grossly misleading. Further, as has been pointed out elsewhere [Du Pont, 1980a; Hasten, 1980; CMA, 1980a], Case D implies a quadrupling of production capacity by 2000, representing numerous business commitments which will not be made given the present environmental and regulatory uncertainties, even if increasingly scarce capital for 9 such an undertaking were available. Even if EPA were to totally discount information on use and growth provided by industry, it should consider the findings of its own contractors: 9In 1979, the world production of CFC-11 and CFC-12 was approximately 1600 million pounds. Assuming, very approximately, that production was 75% of manufacturing capacity, capacity is estimated at roughly 2150 million pounds. Case D assumes a 7% growth in production per annum which equates to a need for a capacity by the year 2000 of around 8700 million pounds, a 6750 million pound increase. Using a conservative figure of $0.85 capital for each additional annual production pound of new capacity, investment to meet case D growth would total very approximately $5.7 billion not a very likely scenario given the environmental and regulatory uncertainties. V-55 ------- Risk In an EPA sponsored workshop it was recognized that estimates of CFC release "even beyond 10 years are problematic" [SRI, 1980, p. 29], because technical obsolescence and technical innovation are highly important and unquantifiable uncertainties. And Rand stated: "We emphasize that the estimated growth rates to 1990 cannot be projected to continue beyond that year. The CFC applications most respon- sible for the current high growth rates are in a phase of increasing market penetration, as a result of either increased use of final pro- ducts or increased use of CFCs in manufactur- ing those products. By 1990, penetration should be complete in most existing markets, so the CFC use growth rate should slow to approximately that of the GNP, unless signifi- cant new uses of CFCs are developed in the next decade. Moreover, easily extracted fluorine is expected to become scarce toward the end of this century, which will increase the prices of CFCs and provide incentives to develop new technologies that are less CFC-oriented." [Rand, 1980, p. 6] b) The second suspect assumption made for the extreme risk scenarios is that there will be no ability to improve the available scientific information over time and react accordingly. Obviously, our understanding of the quanti- tative validity of the ozone depletion theory will improve as the research programs already underway produce results. Ideally, should the validity of the theory strengthen, so should further regulations, and vice versa. A related point is that virtually all reports on this subject [e.g. NAS, 1979a; 1979b] have focused almost exclusively on calculated long-term depletion. While the NAS V-56 ------- Risk calculated some interim depletion values in the PSCT report, no attempt was made in the NAS/CISC Report, or to our knowledge since by EPA, to consider the significance of the small calcu- lated ozone depletion which would be associated with a limited postponement of regulation while uncertainties are reduced. We are presented with an "all or nothing", approach a choice in no way supported by the scientific or political realities. In conclusion, major increases in capacity (as would be necessitated to meet such growth forecasts) will not occur, given current environmental and regulatory concerns, prior to a drastic reduction or invalidation of current computer calcula- tions of future ozone depletion. Equally unrealistic is the presumed regulatory inability to respond adequately in future years in response to evidence of any developing problem. The possibility of this combination of assumptions is economically, scientifically and politically inconceivable. Preoccupation with extreme future extrapolation or future disaster scenarios distracts attention from the reality of the slow theorized development of this potential environ- mental problem, and the redundant opportunities which exist for frequent objective reassessments, and for measured reaction over a substantial time period. 2. Conviction that Immediate Decisions Are Necessarily Required, and Are Necessarily Better than Deferred Decisions As a leading country in technology and quality of life, it is inevitable that the United States will be among the first to encounter, recognize, and be forced to deal with major envi- ronmental concerns. However, such early recognition generally should not be translated into early calls of "crisis". In the case of CFCs and the ozone, a "crisis" approach is an over- V-57 ------- Risk reaction to a theoretical problem which throws into question our understanding of our environmnt. It is recoil rather than informed problem-solving and decision-making, and is a p_arti- cularly inappropriate reaction in the case of the CFC/Ozone concern since hasty decisions are not only unwarranted but unnecessary. We believe a crisis approach to analysis of the CFC/Ozone Issue can be seen in EPA's statements. For example, in dismissing the "Wait-and-See Strategy", EPA considers a single scenario, a hands-off policy for 10 years which would allow an increase in emissions by 1990 "commensurate with an eventual equilibrium depletion of 32 percent". This scenario has assumed the unrealistic global 7 percent growth of the NAS Case D scenario. EPA further states, without justification, that: "little research and development on alternatives to CFCs would occur [during a postponement]...." [EPA, I980d, p. 4] There is no reason to select an entire decade as the time for postponement. There is no reason to assume, given the ongoing scientific research into the theory and recently devel- oped ability to monitor the ozone layer, that were it to be determined that ozone depletion is occurring at a rate which would be harmful, nothing could be done about it until 10 years had expired. Equally, there is no evidence to support a pre- diction that a postponement automatically would result in a 7 percent annual growth during the postponement. And there is no evidence to support a conclusion that the current major research and development on alternatives would be eliminated as a result of any postponement. V-58 ------- Risk 3. Excessive Emphasis on^ Political Action Over Objective Scientific Decision - Making Due to the subject matter, it is inevitable that the political process will be heavily utilized in major environmental issues. However, relying on political process to advance a view in place of performing necessary research and assessment is a short-sighted and eventually counter-productive shortcut, eliciting like response from those with dissenting views, regardless of their interest in having the issue be decided on the scientific merits. The EPA Decision Memorandum repeatedly evaluates regulatory strategies in terms of the international political impact. For instance, the "Wait-and-See Strategy" is dismissed by the comment "Finally, the rest of the world may be convinced that U.S. does not regard ozone depletion as a serious concern" [EPA, 1980d, p. 4]. A more balanced approach would recognize that: * International conviction will stem from scientific knowledge developed through research, in the leadership of which the U.S. can effectively demonstrate its concern and commitment. The objective of regulation to protect strato- spheric ozone should be to protect stratospheric ozone by optimized regulatory strategies which balance risks and benefits, and risks and risks. The objective of the regulation should not be merely to place a U.S. regulatory agency in the "forefront of international attempts to reduce CFC emissions" [EPA, 1980d, p. 4], or to signal other nations that "the U.S. is seriously committed to reducing CFC emissions". [EPA, 1980d, p. 4]. V-59 ------- Risk K. SUMMARY AND CONCLUSIONS The current unconfirmed concerns over calculated potential future depletion of stratospheric ozone require the management of a complex set of conflicting data and uncertain- ties, and risk and benefit assessments. Yet EPA officials have claimed that the management choice may be reduced to an "either- or" propositioneither the theory eventually will be proved to be incorrect and no harm will have occurred from regulation, or the theory eventually will be proved to be correct and at that time it will be too late to head off serious harm if regulation had not been previously enacted. We disagree with both sides of this proposition. An assessment that unneeded regulation will cause no harm can only be made by focusing exclusively on the potential risk of not regulating, and ignoring the environmental and economic harm potentially caused by regulating. The idea that no harm will have been done by premature or unnecessary regula- tion does not have support among scientists, economists, indus- try or the consumer (See Section VII). The other side of EPA's "either-or" proposition (that regulation must occur now or we will have to wait to act until it is too late) simply is not consistent with the up-to-date facts available on the CFC/Ozone Issue. We believe that proper risk management of the CFC/ Ozone issue lies in the middle ground deferral of a final regulatory decision while attempting to improve the information base through an extensive international research effort, supported by industry and government, and continuous close monitoring of the situation for evidence of any developing risk. Such a strategy of "Assessment and Surveillance" is supportable because: V-60 ------- Risk 1) The uncertainties surrounding essentially all aspects of this issue (the science, the effects of depletion, the regulatory cost and the risk of regulation) remain so great as to almost guarantee that any "either-or" regulatory decision will prove to be wrong. 2) Currently available information indicates that not only is the risk not as great as predicted a year ago, but also there is no evidence it is developing as predicted, and 3) We now have the ability to monitor the situation opposite any developing risk (ozone trend analysis), and rethink the wisdom of a deferral strategy as may be suggested by the results of this monitoring. This period of deferral certainly should not be just to postpone regulation, but rather a period in which efforts are made to perform work in a number of inadequately studied critical areas which bear on the question of risk to the ozone from CFCs. Study areas should include: 1) Increasing the number of actual measurements of ozone and improving the sensitivity of the. ozone monitoring system. 2) Reducing the uncertainties surrounding atmospheric science and modeling, especially in areas in which there exists conflict between measurements and model calculations. (In risk management, distinc- tion is necessary between apparent precision and reality. The apparent precision of calculations of future potential ozone changes must not hide V-61 ------- Risk the fact that real confirmation of these predictions is lacking.) 3) Reducing the uncertainties associated with the potential effects of a depletion of stratospheric ozone. 4) Working toward an international concensus on the science and the risk, and a coordinated program on how to manage the risk. (The question of risk to the ozone from CFCs is a global problem. A single nation or unilateral approach to risk management will both be ineffective and create major inequities [See Section VI]. Accordingly, we believe there is a preeminent need to develop an international scientific and regulatory con- sensus.) International cooperation is needed to reduce uncertainties and to perform calm and objective analyses of the consequences of various courses of action with regard to risk-benefit and risk-risk weighing. 5) Properly focusing the question of ozone depletion risk. The overall protection of stratospheric ozone should be the objective, not just preventing possible effects from CFCs. Evaluation should consider the predicted effects from natural causes, e.g., volcanoes, and from the anthropo- genic chlorine containing compounds, e.g., methyl chloroform, and most importantly, the predicted augmentation of ozone by carbon dioxide. 6) Studying the risks which would be incurred by regulation of CFCs, and balancing these against the potential risks which regulation would be de- V-62 ------- Risk signed to reduce. We believe that proper management of risk requires evaluation and consideration of all subjects which bear on risk. Failure to take into account all critical components to the risk equation will result in a risk assessment which will not provide an accurate view of the situation. Regulation based on an incomplete or inaccurate risk assessment may be bad regulation for everyone and subject to legal challenge. In conclusion: Postponement of CFC regulation could marginally increase the risk to future generations should the theory prove to be quantitatively valid. However, in the time frame expected to be required for resolution of the uncertainties, it is difficult to comprehend how such an increase could be significant opposite the major advantages of postponement: 1) Reduced risk of unnecessary regulation. 2) More time to identify and test safe technological alternatives. 3) More time for analysis and selection of the best regulatory plan. The need to reduce the uncertainties is great, and there is near consensus among groups as diverse as industry consultants (See Appendix F), the European Economic Community [EEC, 1980], other countries' environmental agencies [UK DOE, 1979], EPA contractors [SRI, 1980] and EPA consul- tants [Bailey, 1980], that there is time to reduce the uncertainties without incurring unreasonable risk and that such effort overall will be cost-effective. V-63 ------- Risk The conclusions in this regard from a workshop sponsored by EPA in March, 1980 [SRI, 1980] to evaluate "...the critical issues that are hindering EPA's ability to make a fully supportable decision on future CFC regulations" are especially noteworthy: "...the degree of uncertainty in the balance between costs and benefits of control is so great that the making of control decisions is seriously complicated; the decision to control cannot always be unambiguously proven correct. Consequently, further research to settle major issues by reducing uncertainties was seen as highly important and cost-effective." [SRI, 1980, p. 9] * "The author [of the workshop proceedings report] concludes that critical uncertainties in the issues seriously complicates some control decisions even when they are justified on the basis of the best available information, and therefore, that further research is needed and cost-effective." [SRI, 1980 p. iii] "All [Workshop participants] agreed, however, that further research would be cost-effective in making a better decision." [SRI, 1980, p. 52] These conclusions are all the more interesting in that: * The workshop and report took place in 1980, after the release of the NAS reports. * The workshop was limited to participants invited by EPA. Participants included EPA contractors, con- sultants and EPA personnel. Industry was excluded. No mention of the workshop findings appear in EPA's ANPR discussion of its regulatory proposals. V-64 ------- International Aspects VI. INTERNATIONAL ASPECTS A. INTRODUCTION B. DIFFERENCES IN NATIONAL APPROACHES TO THE ISSUE C. ILLOGIC AND LIMITATIONS OF U.S. UNILATERAL RESPONSE 7 D. CONSEQUENCES OF U.S. UNILATERAL RESPONSE 13 E. NEED FOR A TRUE GLOBAL ASSESSMENT, CONSENSUS AND RESOLUTION OF ISSUE 17 F. THE LEADERSHIP ROLE - SUGGESTIONS ON HOW TO PROCEED 20 G. INTERNATIONAL TRADE IMPLICATIONS OF PROPOSED CONTROL 23 H. SUMMARY 34 VI-1 ------- International Aspects A. INTRODUCTION Since the inception of the Chlorofluorocarbon/Ozone Depletion Theory, it has been clear that the problem if it exists, is global because: CFCs are produced and used worldwide, and It ultimately makes little difference where CFCs are used and emitted geographically because they mix in the atmosphere such that any potential de- pletion of the ozone layer impacts all countries to varying degrees. And, therefore, solutions to the potential problem must be global because: No country can specifically protect its ozone layer even if all CFC use within its boundaries is terminated, and No country can have much direct impact on the over all potential problem of ozone depletion through unilateral regulation. EPA and its contractors have acknowledged these real- ities as evidenced by the following quotations: "...no single nation accounts for a large enough portion of world use or production to be able unilaterally to control ozone depletion." (ANPR) "Even the most stringent restrictions on U.S. emissions can have only a modest payoff in ozone protection in the absence of regulatory action by other countries that contribute to worldwide emissions." [Rand, 1980, p. 251]. Yet, strangely, the actions EPA proposes in the ANPR are not in step with the current global perspective on this issue, nor do VI-2 ------- International Aspects the proposals work toward insuring the needed global agreement on the issue. In the ensuing sections, we first discuss the differ- ences in logic and approach to this global issue being taken by the governments of the world. This is followed by an analysis of the illogic and limitations of EPA's unilateral program and proposals, and the consequences we foresee. We then underscore the need for a global assessment and resolution of the science as a necessary first step towards an effective global solution to the issue, and suggest how EPA, by changing its approach, can take a leadership role in obtaining this solution. Last, we examine the international trade implications of the control proposals advanced in the ANPR. However, before moving to these topics a further per- spective should be reiterated. EPA has addressed its analysis and proposals on the premises of a) the validity of the 1979 NAS ozone depletion estimate of 16.5 percent (based on 1977 CFC release rates), and b) world use growing at 9 percent/year. Both of these are faulty. As discussed at length in Section IV (Science), V (Risk) and Appendix F, the incorporation of current information into the models reduces calculated depletion by ap- proximately half or more from the 1979 number, and analysis of actual ozone measurements indicate the problem is not developing as predicted. Secondly, as discussed in Appendix J, EPA's fore- casted CFC growth is not consistent with the declining world production figures for 1974-1979 and the reality of the dampen- ing effect on growth from ongoing regulatory uncertainty. ^ri short, the potential problem is neither as severe as EPA claims it to be, nor could it conceivably worsen to the extent EPA claims it would. These realities eliminate the need to act now, whether unilaterally or internationally. As we will see, the actions of other countries are more consistent with these real- ities than is EPA's program. VI-3 ------- International Aspects B. DIFFERENCES IN NATIONAL APPROACHES TO ISSUE National responses to the CFC/Ozone Depletion Issue generally can be categorized in one of two ways. One response has been to focus on worst case scenarios, minimize the uncer- tainties and immediately place the problem into the regulatory and political machinery. This approach has been embraced by EPA and has led the Agency to issue the ANPR to achieve the Agency's goal of "stimulating international cooperation in this [control of CFCs] area." [EPA, 1980h, p. 1]. The second response, being followed by the United Kingdom, the Commission of the European Economic Communities (EEC) , Japan and others has been to review the developing science periodicallya process of monitoring the level of cer- tainty and degree of risk from the concernand then decide when, and to what extent, to involve political and regulatory machinery. The U.S. approach resulted in essentially a total ban of CFC aerosol propellants in 1978. The European approach has resulted in an agreement to implement a voluntary 3^ percent cut back in CFC-11 and CFC-12 aerosol propellant use from 1976 levels, by the end of 1981, while continuing to monitor the science and periodically review the need for further regulation. The U.S. approach has been to call for a total cap on U.S. production of all CFCs for all the remaining CFC uses, re- gardless of essentiality, partly in an attempt to pressure by example other countries into taking further regulatory action. Almost all comparisons are made between EPA and the EEC countries because by EPA's ANPR calculations (ANPR table 2) and ours, the U.S. & EEC account for between 75 and 80 per- cent of world CFC production and use. VI-4 ------- International Aspects The European approach has been to put a cap on CFC-11 and CFC-12 plant capacity and to continue to study the science in conjunc- tion with periodic reviews of the need for further regulation. No nonaerosol rules have been proposed so far anywhere except in the United States. The U.S. approach may be summarized by the following quotations from EPA documents: "Worldwide regulation of CFC emissions is needed." (emphasis added) [EPA, 1980e, p. 2] Yet, elsewhere, EPA states: "The [recommended] decision [is] to initiate addi- tional [U.S.] regulation of CFCs without experimental proof of the ozone-depletion theory " (emphasis added) [EPA, 1980h, p. 2] "The decision of the Agency [is] to initiate further regulation to address the as_ yet unproven risk p_f ozone depletion despite the apparent lack of such a determination abroad " (emphasis added) [EPA, 1980h, p. 2] In contrast, the European approach may be summarized by the following quotations: "...strict regulation is not warranted." [UK DOE, 1979] "...a delay of 5 years before any decision is taken on CFCs can be reasonably accepted." [EEC, 1980] Given these two simultaneous reactions to the same set of data and scientific uncertainties, two consequences emerge: 1) Countries, including the U.S., which adopt the "U.S. response" place themselves at an economic VI-5 ------- International Aspects and international trade disadvantage opposite those adopting the "scientific" response. 2) Countries adopting the "scientific" response have implicitly rejected the "U.S. response" approach and consequently are likely to be swayed only by scientific developments, improved modeling, actual measurements, etc. They are unlikely to be swayed by entreaties to embrace the "U.S. response", or by publicity surrounding regulatory plans under which the EPA might further emphasize its commit- ment to the "U.S. response". VI-6 ------- International Aspects C. ILLOGIC AND LIMITATIONS OF U.S. UNILATERAL RESPONSE In the CFC Development Plan EPA states: "The present regulatory initiative to limit [U.S.] production so that the potential for ozone depletion is limited to present levels, is part of the effort to stimulate coordinated worldwide action." [EPA, 1980e, p. 1] We interpret this to mean that the Agency believes its proposed U.S. production cap will have an impact on eventual ozone deple- tion (if it occurs) and succeed in getting other countries to regulate CFCs beyond their current plans. We believe that both conclusions are wrong and, further, that the proposal would create a significant imbalance in regulatory costs vs. potential benefits to the United States. 1) Proposed U.S. Cap on Production Will Have Inconsequen- tial Direct Environmental Impact In 1975 the Federal Task Force on Inadvertant Modifi- cation of the Stratosphere, [IMOS, 1975] identified the single largest national use of CFCs as the aerosol industry in the United States. At that time, approximately a quarter of world CFC production was utilized in this market. Whether sub- sequent regulation of that market was justified remains highly debatable but it represented the only single use and single nation market that when eliminated would make a significant impact (25%) on global CFC emissions. Any futher unilateral national response, whether of single CFC use such as aerosol propellants or all uses in any country, will not have a significant impact on global CFC emissions. This may be seen from an analysis of numbers pre- sented in the ANPR. VI-7 ------- International Aspects EPA estimates that 35.8 percent of total world use in 1977 occurred in the United States. EPA also estimates that eventual depletion of the ozone will reach 16.5 percent should emissions continue at 1977 levels. Therefore, one may conclude that even if all U.S. uses were banned immediately, but other countries were not to act, the maximum impact on the eventual ozone depletion numbers would be 35.8 percent of the predicted 16.5 percent depletion, resulting in an eventual depletion of 10.6 percent. In the ANPR, EPA also presents estimates of ultimate depletion, assuming that instead of emissions remaining constant at 1977 levels, they grow worldwide at 9 percent annually until 1990. Without debating here the reasonableness of this forecast (see Appendix J) , under this scenario EPA concludes that the U.S. contribution to eventual ozone depletion would be 7.5 per- cent (absolute) of the calculated 32.0 percent total eventual depletion, or only 23 percent (7.5/32.0) of the problem. Clearly then, a total U.S. ban under this scenario would only have a small impact on the eventual problem. Of course, EPA is not now proposing a ban, but is proposing a cap on U.S. production at current levels. Under this scenario, EPA's numbers show that the U.S. contribution to ozone depletion would be 4.5 percent (absolute) of the calcu- lated 29.0 percent total eventual depletion, or 16 percent (4.5/29.0) of the problem. So the direct potential environ- mental gain (calculated by EPA) from capping U.S. production in 1980 versus letting U.S. production increase until 1990 (at the assumed world growth rate of 9 percent) is only a net decrease in eventual ozone depletion from 32 percent to 29 percent, a 9 (32-29) percent ( 32 ) relative decrease in the potential eventual problem. But, under this scenario, the U.S. contribution to the potential eventual problem would decline from 7.5 percent VI-8 ------- International Aspects (7.5-4.5) (absolute) to 4.5 percent (absolute), a 40 percent ( 7.5 ) relative decrease. In conclusion, the proposed U.S. cap on production would not directly result in a significant reduction in ultimate ozone depletion, were it to occur. In fact, if the theory is correct, even a total immediate U.S. ban would not result in the eventual problem being significantly reduced. Therefore, EPA's statement, quoted at the beginning of this section, that the present regulatory initiative to limit (cap) U.S. production is "so that the potential for ozone depletion is limited to present levels" simply is not supportable even by EPA's numbers. As EPA also can draw the same conclusions from these numbers, we have to assume that the major objective of the Agency's cap proposal is political, i.e. "to stimulate coord- inated worldwide action" [by pressuring other countries to regulate through the setting of U.S. example] 2) Why U.S. Production Cap Will Not Result in EPA's Goal of Worldwide Regulatory Action EPA has stated that: "Worldwide regulation of CFC emissions is needed. However, worldwide action does not appear to be forthcoming." [EPA, 1980e, p. 2] And as discussed previously, one of the stated objectives of the U.S. cap is to stimulate other countries to regulate. The Agency then supports this action with the convoluted logic that: VI-9 ------- International Aspects "...while action by other nations is not assured if the United States acts at this time, inaction by the United States would almost certainly assure inaction by the rest of the world." [EPA, ]980e, p. 2] These few statements raise so many issues it is difficult to know where to begin. However, the following points seem pertinent: a) We would agree with EPA that major worldwide regulatory action is not forthcoming at this time but we differ sharply with EPA as to the "why". The above statement, when taken with the Agency's stated justification that a cap will "stimulate coordinated worldwide action", seems to suggest there is a lack of ability on the part of other countries to under- stand the problem, or that foreign nations are indifferent to possible environmental threatsand therefore, the only way to resolve this problem is through the political tactic of the U.S. setting an example. We believe that worldwide regulatory action is not forthcoming because assessments of the issue (the science, the uncertainties, the risk) by others lead to the conclusion that further regulatory action is not needed at this time. Upon a reading of the United Kingdom's [UK DOE, 1979] report and the ECC Commission's [EEC, 1980] report there can be little doubt that the science has been reviewed thoroughly, but found wanting opposite the justification for further regulation. b) The position of EPA that further regulation is needed has been quite well known by other countries for some time [Blum, 1980; EPA, 1980a]. Therefore, we must question the logic that only through further U.S. regulation will other nations be stimulated to regulate. If they have not chosen to act as EPA feels necessary, it surely is not because EPA has not made its position known. Perhaps it is because others do not VI-10 ------- International Aspects agree with the Agency's analysisin which case we fail to see how further U.S. regulation or "example setting" by the U.S. will resolve the underlying differences in the assessments. A U.S. cap on production will not be an effective signal to others that there is an urgent need to regulate absent the availability of more persuasive facts that the problem is indeed sufficiently serious to require further immediate action. It seems unlikely that the rest of the world will rush to decimate its CFC indus- try in response to EPA's example, and urgings for regulation, when such urgings are based on a tenuous theory. c) It can be argued that the United States already has taken substantial action, which if "leading by example" is a viable tactic, should have resulted in substantial worldwide regulatory action. Specifically, the U.S. showed a willingness to promulgate a ban on half of its use of CFCs aerosol propellantsin 1978. We find it significant that, although a few small or nonproducing countries followed suit, no major industrialized nation has. If a 50 percent cutback by the United States was not effective in obtaining the international regulatory cooperation EPA believes is needed, what probability is there that further U.S. regulation, such as a production cap, will, in and of it- self, have the desired effect? Discussion of logic, support and probability of success for the proposition that unilateral U.S. regulation will force regulation by others, is notably lacking from the ANPR. For EPA just to say its policy will be effective is hardly adequate. The burden is on the Agency to demonstrate that its proposals will result in environmental benefit in excess of costs. d) Perhaps the most questionable aspect of EPA's logic is the statement that although further U.S. action may not in and of itself, assure action by others, inaction by the U.S. VI-11 ------- International Aspects would assure inaction by the rest of the world [EPA, 1980e, p. 2]. A number of points are pertinent here: i) An arrogance is expressed that unless the EPA regulates no one else will take any action. We presume this means EPA believes that others will be incapable of making risk assessments and reaching a regulatory determination on their own without EPA's regulatory example. ii) The Agency is narrow-minded in its view of what constitutes "action". The clear implication is that the only consequential action is regulatory action. Yet, the EEC, for example, can hardly be accused of taking no actiona commitment has been made to study the science, to periodically review the results, and to periodically reassess the need for further regulatory action. We view this as a significant and wise course of action. Likewise, EPA's assessment of potential actions available to EPA is limited. The choice of action does not have to be between regulating now or not regulating now. An alternative action by EPA would be to work to obtain the needed inter- national resolution on the science and to pursue time-trend analysis. A firm commitment to a resolution of the science could not be viewed by anyone as a lack of action. VI-12 ------- International Aspects D. CONSEQUENCES OF THE U.S. UNILATERAL RESPONSE 1) Potential for Counterproductive Results The political approach being taken by EPA could have a number of results which would be counterproductive to any even- tual resolution of this issue. a) It can be argued that any further U.S. regulation could be viewed by the rest of the world as affording more time before having to seriously consider evaluating whether the problem is real and, if it is, to what extent action is called for. This probably will not be true of countries in the EEC (who already are engaged in periodic reviews and assessments) , but certainly could be valid for those nations not yet so involved. And even in Europe, a knowledge that the U.S. is continuing to act to limit the potential for any eventual problem could lead to a perception that this creates more of a safety margin for evaluation before any further decisions would be necessary. EPA expressed concern in the ANPR for this possible outcome, yet then concluded the cap proposal would minimize its possibility. Once again we question the logic and support for a conclusion that the potential for a problem has been eliminated just because EPA has said that it has. b) Another possible result would be more serious. The political approach taken by EPA could create a backlash. The CFC/Ozone Issue is a complex scientific problem. Others, for example, the UK [UK DOE, 1979] and the EEC [EEC, 1980], are attempting to manage the problem by first attending to the science and, only then, following with political or regulatory actions as may be needed. EPA's continuing political push for further regulation, absent attempts to resolve the underlying differences in the scientific assessments between the large producing countries, could create the perception that EPA VI-13 ------- International Aspects is being extremewanting regulation regardless of costs or the facts of the matter. To the extent that EPA's approach is perceived this way, and results in offending other governments, the approach will be counterproductive to obtaining the inter- national assessment and cooperation that most agree are needed. The issue would become further polarizedexactly opposite to what is needed. This threat is not academic. 2) Imbalance Between Costs and Potential Environmental Benefit In section C we raised questions as to the potential effectiveness of EPA's unilateral regulatory approach to obtaining regulation by other nations. To the extent we are correct in our assessment that the Agency's plan as presently proposed will not achieve its goal, the possibility is created for a gross imbalance between the potentially large cost of this plan to the U.S. but only marginal theoretical environmental benefits. Others, including EPA contractors, have also examined this question, concluding: "In the absence of control actions by other nations, the benefits accruing to the United States alone from stringent domestic control measures are less than the costs of those measures." [NAS, 1979b, p. 259]. "Moreover, if the U.S. pursues a regulatory program while most of the rest of the world does not, the con- tinued unregulated emissions abroad will limit the effects of the U.S. program to a modest worldwide effect, with a hopelessly unfavorable balance between costs and benefits to the U.S." [Bailey, 1980, p.3]. The cost to the United States of EPA's proposed cap and economic incentive control options is discussed in detail in Section VII. VI-14 ------- International Aspects 3) Loss of Politial Option In the middle 1970s, before the U.S. aerosol pro- pellant ban, the U.S accounted for roughly 50 percent of world CFC production and use. The EPA then banned U.S. production and use of CFCs as aerosol propellants in 1978. Prior to the initiation of this regulatory process, U.S. aerosol propellant uses of CFCs amounted to approximately half U.S. production, or 25 percent of the world total. If one accepts the "leading by example" logic, it is most surprising that this large "signal" has not resulted in significant regulatory actions by the rest of the world. The situation today is that the U.S. uses approx- imately 36 percent of total CFC produced, and assuming EPA's present growth/year, by the year 1990 the U.S. contribution to the eventual ozone depletion problem would be only 23 percent (7.5% depletion from U.S. contribution )_ (32.0% depletion - total world contribution) (ANPR, Table 3) . However, EPA is proposing to cap U.S. production at current levels, and absent action by other countries, the Agency calcu- lates (ANPR, Table 3) this will result by 1990 in the U.S. con- tribution to the eventual ozone depletion problem falling to only 16 percent of the problem ( 4.5% depletion from U.S.) (29.0% depletion - world) The question we ask is this: If the major U.S. action of cutting production by 50 percent did not result in major regulation worldwide, and, if as we predict, the current pro- posed action (to cap U.S. production) also will not have the result of major regulation worldwide, what conceivable chance will the U.S. have to influence the eventual problem in 1990, when at that time the U.S. will only contribute 16 percent to the total potential problem? What pressure will be available at that time to the EPA, especially when through preceding VI-15 ------- International Aspects excessive reliance on the political approach it 'may have discredited its input into the international scientific arena? Clearly, the U.S. must not allow itself, through premature reliance on a heavily political approach, to end up in a situation where little has been gained but the political approach tool is no longer viable. It is much better to use the scientific approach now (which sooner or later has to be addressed anyway) and reserve the political approach to such time and place where it may make a meaningful difference. VI-16 ------- International Aspects E. NEED FOR A TRUE GLOBAL ASSESSMENT, CONSENSUS AND RESOLUTION OF ISSUE In the previous sections we have presented the global nature of the CFC/Ozone Depletion issue and seen that no one country unilaterally can have a significant impact on the outcome through direct regulatory action, even including total bans. We have discussed the differences in the approach to this issue between EPA and other nations, principally those of the EEC. Using EPA numbers, we also have shown how EPA's unilateral cap proposal will not have a significant direct environmental benefit absent action by other countries. And we discussed our reasons for believing that the unilateral approach by EPA will not result per se_ in other countries moving to regulatea repudiation of the "leading by example" premise. The main reason for the projected failure of EPA's unilateral approach is that it is a political answer to what remains a complex scientific question. That other major coun- tries accept the problem as a scientific one, and are working to resolve it accordingly, has been demonstrated through numerous quotations taken from recent official documents such as the UK DOE [1979] and EEC Commission [EEC, 1980] reports. Whatever actions the U.S. EPA follows, we believe most foreign countries are likely to continue to seek and react to a scientific resolution. Outside the U.S. there is little incen- tive or movement to adopt an excessively political posture opposite what so obviously are scientific concerns, nor is there the disregard for the economic penalty which marks the U.S. regulatory process. The NAS [NAS, 1979b] clearly recognized the need for a global scientific consensus, sensing that only through such a consensus can the problem, if real, be dealt with. In VI-17 ------- International Aspects fact, it appears that throughout the world the U.S. is one of a handful of countries that do not place a scientific consensus clearly at the top of their list of priorities for handling the concern over stratospheric ozone depletion by CFCs. We come, therefore, to the inescapable conclusion that to effectively influence, in any major way, the outcome of this issue, a cooperative global effort is required and that effort must begin with the basics a global assessment and consensus of where the science stands, the degree of uncertainty, and the risk of waiting for better information in light of the results from ozone trend analysisa consensus which presently is clearly absent. If the problem is real, the sooner this assess- ment is undertaken, the sooner a resolution can be reached on the need for (and degree of) action, and the sooner appropriate, coordinated, equitable international control action can be undertaken. If the problem develops not to be quantitatively significant, the sooner individual countries, e.g., the U.S., can cease unilateral actions which place their industries and economies at an international disadvantage. If regulation is necessary, the response must be essentially global and be based on appropriate scientific and economic investigation. The extent to which the EPA reacts without the knowledge from such investigations is largely immaterial in terms of the global environment, although not to the U.S. economy. Indications are that political lobbying of foreign governments by EPA is unlikely to produce significant further regulatory results, but will continue to further polarize world scientific opinion, directly counter to progress towards a scientific consensus. There is time to obtain this consensus without unreasonable risk to the world's populations and environment (See Section V). The UK [UK DOE, 1979], the EEC [EEC, 1980], VI-18 ------- International Aspects and industry's assessments all find, for example, that up to 5 years could be taken without substantial risk. And the existence of ozone trend analysis provides an additional margin of safetya margin which could be reviewed annually to assess the wisdom of further regulatory deferral. VI-19 ------- International Aspects F. THE LEADERSHIP ROLE - SUGGESTIONS ON HOW TO PROCEED In its CFC Development Plan, EPA states: "The United States as the largest producer and user of CFCs, must take a leading role in this effort [to obtain worldwide regulation of CFCs]." [EPA, 1980e, p. 2] We believe the Agency's view of what constitutes appropriate leadership on the CFC/Ozone Issue is skewed. The Agency's view seems to be that leadership means being the first to regulate. We believe, on the other hand, that the leadership role should be in obtaining the needed global resolution of the science and then, _if_ it is needed, to lead a global coordinated effort to obtain the appropriate equitable degree of regulation. It is likely that virtually all major producing countries would appropriately regulate CFCs i£ current concerns over ozone depletion were validated throughly. It is also evident that U.S. political pressure, in the absence of such adequate validation, has had very limited success, and it is doubtful that much further regulation elsewhere will occur unless an accepted scientific justification is developed. There is a critical opportunity for U.S. leadership. That opportunity is to organize an international program, involving government, industry and academia, to resolve uncertainties and to work towards generating the objective scientific information needed. However, as discussed in the previous section, if for no other reason than the fact that other key nations view the science as unresolved, EPA cannot lead simply by proceeding with, and pressuring for, further regulation. Such "leadership" will be rejected. VI-20 ------- International Aspects There is another major advantage in the U.S. leading a "learn-before-regulating" effort. The ozone depletion concern of the EPA (if it has not already been adequately demonstrated by the essentially complete ban on aerosol uses), can be demon- strated on a continuing basis by technical leadership in an international scientific program, and, at the same time, the risks from hasty or unnecessary regulation in the U.S. can be mitigated. We suggest elsewhere in this submission how such leadership might be effected, but basically the recommendation is that the National Research Council and the United Kingdom Royal Society jointly convene with corresponding scientific societies in France, Germany, Italy and such other countries as are appropriate, to form a scientific review committee to 2 produce an international review of the issue. We further recommend that EPA postpone both regulatory action in the U.S. and political lobbying for CFC regulation abroad until such reviews are made. By supporting such an attempt to reach a truly objective assessment, and refraining from domestic and inter- national efforts to lead the conclusions of the committee to any preconceptions of what that conclusion should be, EPA would demonstrate both its concern for reaching sound conclusions on the CFC/Ozone Issue and its leadership abilities. A complete review of the science, taking into account all the most recent developments, would be the ideal. How- ever, if such a broad undertaking cannot be effected, there are more limited areas for review which would maximize the paybackozone trend analysis being the foremost among these. VI-21 ------- International Aspects In summary, the United States can more effectively demonstrate its concern and leadership by coordinating a research effort which will stimulate the reaching of an international scientific consensus, than by international political pressure. VI-22 ------- International Aspects G. INTERNATIONAL TRADE IMPLICATIONS OF PROPOSED CONTROLS The following discussion focuses on the trade impli- cations of the proposed economic incentives options. Legal issues relating to regulatory authority over imports and exports are reviewed in Section III. Detailed comments on the economic implications of these regulatory options appear in Section VII and Appendix I. 1) Exports From a reading of the ANPR, the basic export question is whether (under a cap on U.S. CFC production, in conjunction with some sort of allocation scheme or any other regulatory op- tion) exports of CFCs should be considered part of the domestic limit or excluded altogether. We will demonstrate why exports should be excluded. However, regardless of the regulatory outcome, special care should be taken to insure that the U.S. world trade position is not unilaterally and unfairly penalized. a) Inclusion of Exports Under Domestic Cap Would Eliminate Exports In discussing its thinking that a U.S. production cap should cover both domestic use and export, EPA states in the ANPR: "...it appears unlikely that firms would continue exporting because they would be placed in an unfavor- able pricing position in the foreign CFC market." We agree with EPA's analysis. The CFC business, both domestic and abroad, is extremely competitive, manifesting itself primarily in extreme price sensitivity. A small increase in price by one supplier will result in dramatic volume shifts to lower priced suppliers. VI-23 ------- International Aspects Were export pounds to be incorporated under a domestic production cap, with the resultant increase in price (projected and desired by EPA in order to drive down CFC demand), U.S. producers would be in the position of trying to compete in export markets with a commodity chemical at a price higher than available from foreign firms. This would be so because foreign firms would not be subject to the artificially higher prices brought on by the cap and allocation system. Under such condi- tions, U.S. firms would lose essentially all their export business. Given the current concern over U.S. trade deficits, such a scenario created by unilateral U.S. regulations seems totally out of place, particularly in that, as we shall see below, such action will produce no net environmental benefit. b) Restriction of Exports Would Have No Net Environmental Benefit While it may be argued that a domestic production cap imposes no direct restrictions on exports per se, the antici- pated higher prices resulting from a production cap and alloca- tion scheme (discussed at length in Section VII and Appendix I) will result in a loss of the U.S. CFC export market. What is particularly appalling to us is that even if all U.S. exports of CFCs were eliminated, there would not be any decrease in total world CFC use or emissions. As EPA acknowledges in the ANPR: "Under [an] approach of including exports in the [domestic U.S.] production ceiling, foreign firms could conceivably increase their production to offset any decrease in U.S. exports, which would run counter to the long term goal of reducing global CFC emissions." VI-24 ------- International Aspects We agree with this EPA observation except the words "could conceivably" should be replaced with "would". There are four ways to meet CFC demand in a given foreign country: 1) U.S. exports, 2) exports from other countries, 3) production by a U.S. owned manufacturing facility in that country, and 4) production by a foreign owned manu- facturing facility in that country. Any restriction, direct or indirect, for example through pricing, of the ability of U.S.- based firms and facilities to export would absolutely have no impact on the underlying demand for CFC in the subject country. The demand would still be met - but by foreign export or foreign production facilities. The net result: loss of U.S. export market; gain for foreign producers; no net change in amount of world CFC use; no net change in potential environmental problem; so no environmental benefit from this unilateral U.S. regulatory option. If a regulatory policy results in no benefit to the environment, it should not be implemented, particularly when such a policy would severely penalize U.S. industry and exacerbate a major U.S. problembalance of trade. c) EPA's Defense of Proposed Policy to Include Exports Under a Domestic Production Cap is Weak The Agency obviously anticipated the above arguments against the inclusion of exports under a domestic production cap because several paragraphs in the ANPR are dedicated to the Yet another consideration is that the U.S. would still be penalized even if the CFC/Ozone Depletion Theory is even- tually totally invalidated. This is so because a cessation of exports by the U.S. will result in a dismantling of export organizations which were years in the making. These could not be replaced very readily or rapidly. VI-25 ------- International Aspects discussion of why, even in the face of the above situation, exports should not be exempted. We find these arguments to be weak. Consider the argument in the ANPR given opposite the fact that foreign producers would increase their production to replace no longer competitive U.S. exports: "However, some nations have already taken action to limit their capacity to produce CFCs and have indicated that they will consider taking additional actions." The facts are that: i) Only countries of the EEC (nine) have limited capacity and this is only on CFC-11 and CFC-12, whereas all U.S. CFC exports to countries around the world would be eliminated by the production cap induced higher prices. ii) The bulk of exports are CFCs other than CFC-11 and CFC-12 anyway so the capacity cap in Europe is irrelevant. iii) A limit on capacity is hardly the same as a limit on production. (Given the European phasedown of CFC-11 and CFC-12 aerosol propellants there is excess capacity which could be utilized to produce the CFCs to fill the void created by the U.S. loss of exports). iv) The statement that other nations have indicated they will consider taking additional action requires explanation. What nations? What action (to what degree and with what effect)? When? How would this balance out the penalty to U.S. business? These are all unanswered questions. VI-26 ------- International Aspects The ANPR presents a second argument as to why exports cannot be exempted: "The global problem may be exacerbated if the United States is in the position of encouraging other nations to take further actions to control CFC emissions while at the same time not controlling United States CFC exports." Several questions are pertinent opposite this statement: i) What global problem will be exacerbatedthe problem of other nations not agreeing with EPA's assessment that regulation is needed? ii) How does this bear on the question of whether EPA should undertake any action which would eliminate U.S. exports without any potential offsetting environmental gain? iii) Are not other nations able to cut off imports from the U.S. if they believe them to be harmful? Is EPA the only body which is capable of deciding whether CFCs are harmful? If, as we suggest in sections E and F, EPA were to concentrate its efforts on obtaining an international assessment of the science, followed by a resolution between countries on what was the risk and what, therefore, needed to be done, the Agency's concern over exacerbating the global problem would be eliminated. Further, such a resolution, leading to a global cooperative program of regulation (should it be needed), would eliminate the unfair impacts which would be brought on by EPA's unilateral approach. VI-27 ------- International Aspects d) EPA Expresses More Concern For Foreign Exporters To The U.S. Than For U.S. Exporters To Other Countries The ANPR states: "...a mechanism must be devised to regulate imports of CFCs and CFC containing articles so that imports are neither given an advantage nor placed at a disadvantage in comparison to domestic manufacturers." We already have seen how the Agency's proposed in- clusion of exports under a domestic production cap will elimi- nate U.S. exports of CFCs. The only concern expressed by EPA in this regard is that if exports are not so controlled, it may create problems for EPA with foreign governments. No concern is expressed by EPA for the economic consequences to U.S. industry or the U.S. economy from such action. Yet on the import side of the equation, EPA is concerned that imports not be placed at a disadvantage. It would seem that it is acceptable to penalize U.S. industry but not to compensate it. In conclusion, we believe that any U.S. cap on domestic production should exclude exports. Failure to exempt exports will result in an elimination of exports. Loss of U.S. exports will be filled by foreign producers, so there will be no net environmental change. Loss of exports unfairly penalizes U.S. industry and the U.S. economy. EPA's defense of this policy proposal is weak. VI-28 ------- International Aspects 2) Imports The agency presents three options for treatment of CFC imports: a) freezing imports at current levels; b) capping imports at current levels in conjunction with a permit system; and c) putting imports under the total domestic cap and allowing foreign producers to bid for production rights. Before commenting on some of the implications of these options, we suggest that a fourth option be added to the listtreating imports exactly the same as exports. a) Imports Should be Treated The Same As Exports As policy, the United States is opposed to trade barriers, which any of the above options would amount to. However, given the anticipated effect of EPA's proposed policy on exports, we believe it is grossly unfair to U.S. industry and to the nation to not attempt to balance out the import/export equation. Simply put, if U.S. industry will not be able to export CFCs to foreign markets, neither should foreign producers be allowed to export their product to our markets, especially under conditions of a limited domestic market. Since inclusion of U.S. exports under a domestic cap does not ban these exports as such, but only renders them non-competitive to foreign suppliers due to higher price, foreign suppliers (with a price advantage due to not having to meet the imposed U.S. regulatory requirements) should not be permitted to expand their sales in the U.S. market at the expense of U.S. industry. If, as EPA states, the objective is to devise controls on global CFC usage, the U.S. government must, at a minimum, be concerned with a combined export-import policy which does not place the U.S. industry at a competitive disadvantage to foreign producers. VI-29 ------- International Aspects b) If U.S. Production Is Capped, Imports Should be Capped Separately And On The Same Basis Developing the above thoughts opposite the options presented for the treatment of imports, we believe that, at a minimum, it would be necessary to either freeze imports at the level in the base period proposed for domestic production, or cap imports at this level in conjunction with a permit system. We strongly oppose the option of permitting foreign producers unlimited access to a capped U.S. market as it would mean significant loss of U.S. market to foreign producers for the reasons detailed below. Under a cap, the cost to a U.S. manufacturer producing CFCs would have 3 elements: 1) raw materials, 2) cost of manufacture, transportation and sales other that raw materials, and 3) the cost of the production permit. A foreign producer entering into the capped U.S. market also would have all three costs except that item 2 would be lower. . This is true for the following reason. Under a cap, U.S. producers, not being able to export competitively and having a ceiling on how much could be made domestically, would find their cost per CFC pound- produced increasing at a rate faster than that of foreign producers. The difference is in fixed cost. Fixed cost items such as overhead will go up yearly, but the total amount produced under a cap is held constant, so the cost per pound- produced increases. Foreign producers, on the other hand, not being limited to how much can be produced, will be able to spread these cost increases over an increased volume of production, and keep the fixed cost per pound steady or maybe even lower it. Over time, the difference between the U.S. cost per pound and the foreign cost per pound could become great. Assuming that all other costs, e.g., raw materials, labor, etc., remain roughly comparable, the net result will be the ability of foreign producers to offset higher permit pound bids by the VI-30 ------- International Aspects lower cost per pound of production. Therefore, foreign producers could purchase permits at a price in excess of what could be justified by U.S. producers, given the restricted demand for the end product. Therefore, over time, foreign penetration of a U.S. market would increase at the expense of existing U.S. industry. EPA states in the ANPR: "...this option [the option of including imports under the domestic production ceiling and allowing importers to compete in the permit market with domestic firms]...would remove the possibility of importers enjoying a cost advantage over domestic manufacturers." Economic realities indicate that this conclusion is wrong. A policy which penalizes the cost structure of U.S. industry over foreign producers, prices U.S. producers out of foreign markets, and then permits foreign producers unlimited access to U.S. markets is extremely inequitable and unwise. If the need arises for U.S. production to be capped, import levels also must be capped but on a separate basis. If permits are to be sold for U.S. production, separate permits also must be sold for imports. c) Taxing of Imported Finished Goods Made With CFCs There would be another problem for U.S. industry which would develop under a unilateral domestic CFC production cap. This would be the competitive advantage created for imported final products which were produced or processed by, or which contained, CFCs. Domestic manufacturers would have to pay a premium for CFCs under a cap. However, foreign manufacturers producing goods abroad which are dependent upon CFCs, for VI-31 ------- International Aspects example, frozen shrimp or berries, would only have to pay "normal" CFC prices. Upon importation into the United States, these foreign goods would be at a cost advantage to domestically produced goods. The same would be true for products containing CFCs, such as auto air-conditioners. In fact, it is conceivable that certain products dependent upon CFCs now produced in the U.S. would have their manufacturing exported in order to assure CFC availability at a reasonable pricefollowed by the finished goods then being imported back into the U.S. Obviously, all of these situations would penalize domestic manufacturers without resulting in any net potential environmental gain from lowered consumption of CFC. EPA's contractor, The Rand Corporation, examined this problem and concluded: "Under either economic incentives or mandatory controls [if they were to be imposed], both of which increase the costs of producing domestic final products, imported final products made with CFCs should be taxed." [Rand, 1980, p. 247] d) Potential For Illegal Imports Has Not Been Addressed Given the likely high cost of CFCs under a domestic production cap, there would be temptation for some to attempt to obtain "unpermitted" CFC at a lower price. To the extent this situation evolves, EPA's stated objective of limiting CFC production would be undermined and those U.S. users not involved would be placed at a competitive and product price disadvantage. Rand also touched on this point, concluding: " As noted later in this section, one possible enforcement problem raised by economic incentives policies might be prevention of illegal CFC imports." [Rand, 1980, p.242] VI-32 ------- International Aspects " Unlike mandatory controls on the behavior of users, economic incentives policy requires enforcement to prevent illegal imports of CFCs." [Rand, 1980, p.246] The ANPR makes no mention of this potential problem. VI-33 ------- International Aspects H. SUMMARY A paramount fact of the CFC/Ozone Depletion Issue is that the problem, if it exists, is global in nature. This fact bears both on any assessment of the problem and importantly in consideration of any solution. This cannot be stressed enough, for even if CFCs eventually are proved to deplete ozone to the degree calculated by the models, and the potential effects are as projected by the NAS, the fact remains that without global acceptance and commensurate action the problem cannot be solved. Unilateral action by any country cannot significantly reduce the risk should the theory and all its elements be valid. This reality has been addressed throughout the evolution of the issue by most participants, including EPA and its contractors: "It should be noted that effective control of CFC caused ozone depletion cannot be achieved by one nation acting alone." [EPA, 1980e, p. 2] Yet EPA's analysis in the ANPR of the international aspects of this issue does not seem to us to reflect an understanding or proper perspective of this global reality. In the first place, EPA's approach to the issue seems to be one which emphasizes political over scientific aspects. We find that EPA has determined the problem is real (despite the acknowledged uncertainties) and has concluded, therefore, that it requires immediate worldwide regulatory attention. Yet the very countries the Agency's program is focused on, the EEC, continue to view the problem as an unresolved scientific issuean issue which must continue to be studied, but one not requiring immediate regulatory attention: ". . .a delay of 5 years before any decision is taken on CFCs can be reasonably accepted." [EEC, 1980] VI-34 ------- International Aspects Given these sharp differences between the assessments and programs, we believe the Agency's proposed unilateral response is inappropriate, inefficient and not likely to obtain the stated goal of "coordinated worldwide action." [EPA, 1980e]. Specifically, the proposed production cap will not have any consequential direct environmental effect if other countries continue to produce CFCs as forecast. (Indeed, even a unilateral total U.S. ban would not reduce the problem enough, if the theory proves to be real.) Further, given the underlying reasons for the differences between the regulatory programs of the U.S. and other countriesdisagreements over the science--we fail to see how the effort being advanced by EPA, with a stated objective ". . .to stimulate coordinated worldwide action," can have much chance of success if EPA ignores the need to resolve the science assessments. The U.S. already has taken action well beyond that of any other major producing country, yet apparently this "example" has not been sufficient stimulus to obtain worldwide regulatory action. The question must be asked then: "On what basis does EPA expect such contrasting success for the proposed production cap?" EPA fails to provide any support in the ANPR for such success, much less document any cost effec- tiveness. In fact, EPA's own contractors [NAS, 1979b; Bailey, 1980] found that absent control actions by other nations, the U.S. would be faced with ". . .a hopelessly unfavorable balance between costs and benefits." [Bailey, 1980] A recent editorial in the New York Times touched on all of the above points: "If countries around the world continue for a decade to expand their use of chlorofluorocarbons, the American limits proposed by the E.P.A. would have only a trivial effect. So E.P.A. justifies its plan largely as a diplomatic signal, to show other nations that the problem requires international attention. But that rationale deserves further scrutiny." VI-35 ------- International Aspects "A domestic freeze would inevitably drive up the prices of many productshow much remains to be spelled out; some say only a little. But a nation already burdened with rising costs should not take another costly regulatory step for insufficient gain. If America cannot succeed along at this task, the E.P.A.'s proposals need to be measured by their diplomatic value in persuading other nations to cooperate. So the wisest course might be to make further American restrictions contingent on inter- national action. The need is for more global controls, not unilateral disarmament." (emphasis added) [N.Y. Times, 1980]. There must ultimately be global resolution on this issue, followed by an appropriate coordinated program to deal with the problem, if it is found to exist. Since EPA's proposed program will not advance this objective (in fact, a case may be made that the Agency's program will be counterproductive), what then? We come to the inescapable conclusion that there must be a return to basicsa global assessment and consensus of where the science stands, the degree of uncertainty and the risk of waiting for better information. If the problem exists, the sooner this is undertaken, the sooner a coordinated global program can be effected to deal with it. If the problem develops to be insignificant, the sooner the individual countries like the United States can cease unilateral activities which place their industries and economy at a global disadvantage. There is an opportunity for the United States to take a leadership role to obtain the above assessment. Several suggestions have been made on how to proceedbeginning with a discontinuation by EPA of the currently favored political approach of attempting to effect world regulation by example. Last, we have examined the international trade implications of the proposed U.S. domestic production cap in conjunction with allocation or auction schemes. We find that, VI-36 ------- International Aspects as proposed, these options would place U.S. industry at a competitive disadvantage, both abroad and at home, yet not result in any meaningful potential environmental gain. In particular, any option incorporating exports under a domestic production cap would result in U.S. exports being priced out of foreign markets. These markets, however, would be met by foreign producers who are not restricted, with the result that although U.S. industry is penalized, there would be no gain towards EPA's environmental goal. EPA mounts no legitimate defense in the ANPR for such a consequence. On the other side of the coin, we find that foreign producers, operating under an unrestricted climate, could, over time, develop a substantial cost advantage to U.S. producers who would be operating in a severely limited market. Unless any potential regulation takes specific steps to place imports to the U.S. under the same constraints and penalties as mandated for domestic production and export, we foresee a loss of the U.S. market to foreign producers. VI-37 ------- VII. ECONOMIC CONSIDERATIONS Page A. INTRODUCTION . 2 B. ECONOMIC SIGNIFICANCE OF CHLORO- 4 FLUOROCARBONS C. REGULATION OF CHLOROFLUOROCARBONS 10 VIA ECONOMIC INCENTIVES D. REGULATION OF CHLOROFLUOROCARBONS 35 VIA COMMAND AND CONTROL E. INADEQUACY OF THE RAND REPORT 37 F. MISCELLANEOUS POINTS IN THE ANPR 61 WHICH HAVE ECONOMIC IMPLICATIONS G. SUMMARY 69 VII-1 ------- Economic Considerations A. INTRODUCTION In the ANPR, EPA considers two basic approaches to further regulation of CFCs: (1) the standard "command and control" approach, including product and/or end use phase-downs or bans, and technology-based standards, and (2) a so-called "economic incentives" approach which would limit aggregate CFC production or use but provide a theoretical flexibility to vary product mix, end uses, etc., over time based on market considerations. Because EPA states a preference for the economic incentives approaches, our discussion below focuses principally on the potential micro- and macroeconomic consequences of these regulatory options. We also comment on: (1) the lack of attention by EPA to command and control regulatory options for CFCs, (2) the inadequacy of the Rand Report [Rand, 1980] (the major economic study performed to date) to support a regulatory decision , and (3) miscellaneous points in the ANPR which have economic implications. Our commentary is largely qualitative, necessitated by the very short response period for the ANPR relative to the time typically required to develop quantitatively supported arguments. The Rand Report was of little help for this purpose, principally because Rand's data bases have not been made available. Nevertheless, our viewpoints, as expressed below, rely heavily on conventional wisdom in their development. As such, although rigorous analysis is lacking at this point, such viewpoints should be accepted as valid until proven otherwise by We include a critique of this report in this section (and more detail in Appendix I) because it is the only study performed to date on economic incentives options as they would apply to CFCs, and because EPA specifically cites the findings of this study in the ANPR in support of the Agency's stated preference for the- economic incentives approach. VII-2 ------- Economic Considerations standard techniques of economic analysis. In other words, the burden of proof is on EPA to show, through such analyses, that the conventional wisdom is wrong in this instance and that its CFC cap proposal can be accomodated without severe adverse eco- nomic consequences. To help place all the ensuing discussion in perspective, we begin this section with a highspot summary of the economic significance of chlorofluorocarbons. Regulatory decisions on CFCs will affect major industries, with large employment, and a wide array of consumer goods. Should regulation prove to be needed, great care will have to be taken to insure that the regulatory option selected will be the most cost effective to society. The fact that Du Pont is offering serious commentary on the relative merits of some of the regulatory options under consideration by E?.\ in no way should be interpreted as an indication that we accept the need for, or the inevitability of, further regulation of CFC uses. As discussed in detail elsewhere in our submission, the Du Pont position remains: Since the subject and effects of potential stratospheric ozone depletion by chlorof luorocarbons, and potential actions to deal with the per- ceived problem, are matters of international consequence, we believe there needs to be an international resolution of the underlying scientific differences prior to further regulatory action. V1I-3 ------- Economic Considerations B. ECONOMIC SIGNIFICANCE OF CHLOROFLUOROCARBNS Table 1 summarizes the economic significance of CFCs-11, 12, 22, 113 and 114 by their major end uses. The numbers were taken from published information where possible -- where not, the numbers are Du Pont estimates. The base year varies depending upon the reference. CFC production and use are for 1979. VII-4 ------- Economic Considerations Note; References cited in these footnotes are not cited in Section XI Bibliography NE Not estimated. (a) Du Pont estimates based on sales information and Du Font's estimated market share. (b) Employment and other estimates were made by Du Pont in a series of white papers on the Chlorofluorocarbon/Ozone Depletion Issue, August, 1979. Employment in refrigera- tion and air-conditioning has since been increased to include CFC-22 related employment. (c) Report of Federal Task Force on Inadvertent Modification of the Stratosphere (IMOS)" "Fluorocarbons and the Environment", June, 1975, p. 98. Employment data is for 1974 except as noted. (d) Includes service locations and companies. (e) Air-Conditioning and Refrigeration Institute: "The Ozone Controversy", June, 1978, p. 21, estimated 97,400 establishments for 1972. An annual growth rate for 1972-1979 of 1.75 percent is assumed, leading to 110,000 establishments for 1979. (f) Bureau of Domestic Commerce, Department of Commerce: "Economic Significance of Fluorocarbons", December, 1977, p. 21. (g) Source as for (e), reports $20.6 billion for 1972. (h) Source as for (c) . IMOS data is assumed to include CFC-22. The employment in CFC manufacture has been subtracted from the IMOS total for this industry segment. (i) IR&T Review Draft IRT-20000/1: "The Use and Emissions of Chlorofluorocarbons in Mobile Air-Conditioning", October, 1978, p. 53. Total service facilities. (j) Motor Vehicle Manufacturers Association: "Motor Vehicle Facts and Figures - 1978." (k) Source as for (i). Replacement value: $500/unit (p. 4). 1976 installed units: 64,498,000, growing at 4.7 percent per annum (p. 19). 1979 replacement value = 64,498,000 x (1.047) x $ 500 = $37 billion. Du Pont's lower estimate of $33 billion is listed the difference is within the uncertainty of the estimates. VII-6 ------- Economic Considerations (1) IMOS report, p. 98 [see footnote (c)] shows employment in plants of major uses of CFC solvents. (m) Du Font's solvent equipment manufacturers (SEMs) sell approximately 600 units per year and we estimate these SEMs represent 50 percent of the market for such equipment. Total sales are thus 1200 units per year with average life of 8 years. Total in-service units approxi- mately 10,000. (n) Du Pont estimate of direct foam line operation, maintenance and supervision only [source as for (b)]. (o) Society of the Plastics Industry (SPI): "The Importance rof Chlorofluorocarbons and Polyurethane Foams", Urethane Division Bulletin U-109, March, 1980, p. 12. (p) Rand Corporation: "Economic Implications of Regulating Chlorofluorocarbon Emissions from Nonaerosol Appli- cations", R-2524-EPA, June, 1980, p. 92. Total CFC-blown foam production for 1979 is 489 million pounds at average value of $0.75/lb. (Du Pont estimate) = $367 million. (q) Replacement value is difficult to estimate. Alternative blowing agents for insulating foam are not available so the total replacement value is listed. Our estimate of $2500 million is contained in a Du Pont memorandum from R. M. Kitchens to file, dated December 12, 1980. (r) IMOS Report, p. 98 [see footnote (c)]. Estimate includes employment for the production of the foam raw materials and the foam itself, but not the products made from the foam. This reference does not disaggregate employment by foam type or use [cf footnote (v) ] . (s) Source as for (p) , pp. 47, 48. Fifty companies, 70-130 plants (median 100). (t) Source as for (p) . Table 3.A.3 (p. 46) shows total flexible urethane foam production at 1,275 million Ib. for 1977, and 1,420-1,690 million Ib. for 1980. We interpolate 1400 million Ib., for 1979. Table 3.A.I (p. 45) indicates 42 percent is blown with CFC (=588 million Ib., foam) and the product has an average value of $1.00/lb. (u) Du Pont estimates costs for flexible polyurethane plants to convert to alternative blowing agents at $400,000 per plant for 100 plants = $40 million. Du Pont memorandum from R. M. Kitchens to file, dated December 12, 1980. [see footnote (y) estimates for comparable changes for polystyrene foam plants]. VII-7 ------- Economic Considerations (v) IMOS Report, p. 98 [see footnote (c)]. Larger estimate includes employment in raw material, foam and end product production. This reference does not disaggregate employment by foam type or use [see footnote (r)]. (w) Rand Corporation Working Note: "Interim Report: The Use and Emissions of Chlorofluorocarbons in Urethane Closed-Cell Foams", WN-10401-EPA, December, 1978. Appendix A, pp. 66-67, identifies 29 nonurethane foam producers. Du Pont estimates, based on market intelligence, that there are 101 producing locations. (x) Source as for (w), p. 16 (Table 3), estimates 473,600,000 Ib., for foam produced in 1979. Du Pont estimates average value at $1.30/lb., for a total value of $616 million. (y) Rand Corporation Working Draft: "Economic Implications of Regulating Chlorofluorocarbon Emissions from Nonpropellant Applications", WD-348-EPA, September, 1979, p. 201. Plant conversions from CFC-12 to pentane for polystyrene foam production is estimated at $460,000 structural changes plus $80,000 per extruder line. Labor costs increase by $90,000 annually; energy costs increase 12 cents/lb.; insurance costs 2 percent of capital. Total first year cost = [(29 x 460,000 + 101 x 80,000) x 1.02] + 90,000 + (0.12 x 473,600,000) = $79 million. Annual production see footnote (x). (z) Du Pont estimate of companies and locations based on liquid food freezant machines sold by franchised manufacturers. (aa) Du Pont estimate based on market intelligence. Average annual (season) use of equipment (1,000 hours) x average freezing capacity (13,000 Ib./hour) x number of machines (30) = 390 million Ib., rounded to 400 million Ib., valued at $1.00/lb. (bb) Du Pont estimates 30 machines at average estimated cost of $300,000 each = $9 million. (cc) Rand Corporation Working Note: "Interim Report: The Use and Emissions of Chlorofluorocarbons in Sterilization Applications", WN-10275-EPA, September, 1978, pp. 9, 10. Rand estimates 3950 units based on Du Pont estimates provided in 1978. A subsequent assessment by Du Pont [see footnote (b)] indicates Hospital and Institutional units were underestimated by 700 units, and Industrial units were underestimated by 20, for a revised total of 4670 units. VI1-6 ------- Economic Considerations (dd) Source as for (p), pp. 194-195. Market for sterilant gas is estimated at 11.7-14.3 million pounds in 1976, growing at 9.5 percent annually, which corresponds to 15.4 - 18.8 million pounds in 1979. Taking the lower end of this range, and Du Font's estimated price to users of 90 cents/lb., annual value is $14 million. (ee) Source as for (cc). The Rand estimate for installed value is $58 - $84 million (pp. 9-11) based on Du Pont 1978 estimates. Using the revised estimates [footnotes (b)], and market growth rates, a value of $89 million is derived for 1979. (ff) Bureau of Domestic Commerce, Department of Commerce: "Economic Significance of Fluorocarbons", 1975. Fluorocarbon-dependent employment in the United States estimated at approximately 600,000, or about 0.7 percent of the total U.S. employment. In addition, this source estimated approximately 900,000 indirectly-dependent employees, for a total of 1.5 million. ------- Economic Considerations C. REGULATION OF CHLOROFLUOROCARBONS VIA ECONOMIC INCENTIVES 1. Introduction EPA indicates a preference in the ANPR for an economic incentives approach to further CFC regulation, presumably on the grounds that such an approach minimizes adverse economic consequences. However, depending on the degree of CFC emissions reduction sought thereby, this regulatory approach might involve grave economic impacts for certain industry sectors and for the U.S. economy as a whole. There are three interrelated factors which suggest this would be the case: Regardless of the precise regulatory approach, the essence of its impact is to restrict CFC avail- ability to a level below the market demand prior to regulation. CFCs are "essential" in most important end uses, i.e., as a public policy matter, the needs which CFCs currently satisfy will have to be satisfied in some alternate way. Satisfactory substitutes for CFCs (CFCs or products depending upon them) generally are not available. Those substitutes which would have to be used would create significant safety, energy and economic penalties to producers, users and consumers . 2 CFC uses and their essentiality are discussed in detail in Section II. Footnote 3 appears on following page. VII-10 ------- Economic Considerations Accordingly, even under a regulatory approach involving aggregate production limits, with the market determining which former end users do without CFCs, the following type scenario would eventuate: needs currently served by "regulated" CFCs would be met at a cost which escalates substantially over time -- as CFCs become more and more scarce relative to their demand and/or higher cost substitutes displace CFCs. The eventual consequences are predictable: higher inflation rate, slower rate of economic growth, increased business failures, higher unemployment, deteriorating international competitiveness and an ever-expanding cycle of economic dislocations growing out of the continued uncertainty. These and other economic themes are developed in more detail below. (A more detailed analysis of economic incentives options appears in Appendix I). But first, to set the stage for this discussion, it is appropriate to review briefly earlier comment on CFC essentiality (See Section II). As a class of compounds, they are truly remarkable. Available commercially for some 50 years, they still have no important functional competition in many of their original end uses. Refrigeration, air-conditioning, certain other heat transfer applications and thermal insulating foam are prime examples. This market scenario is unusual; product life cycles are typically much shorter because consumer preferences and/or underlying relative cost positions tend to be constantly changing, with our competitive free enterprise system fueling the process. CFCs have endured basically for one reason -- their price has always been significantly lower than their perceived value-in-use. The economic implications of this opposite EPA's intent to further regulate CFCs are particularly pertinent, as will be seen below. A discussion of currently available substitutes to CFCs may be found in Section II. Discussion of fluorocarbon alternatives appears in Section VIII and Appendix B. VII-11 ------- Economic Considerations In the following discussion of specific potential economic impacts from further CFC regulation, our approach, at this time, is of necessity qualitative, indicating the likely direction of economic events rather than attempting to quantify their magnitude. For purposes of this discussion, the assumption is that CFC production is "capped" at current levels and this cap is gradually reduced over time to eventually achieve a 50-70% reduction in CFC use. Although the ANPR does not specifically propose additonal CFC regulation following imposition of a production or use cap, it is appropriate to consider economic consequences in this light because: EPA has stated its intent to push CFC emission levels significantly below (i.e., 50-70% range) that which would be achieved by a cap [Jellinek, 1980a]. From the standpoint of CFC producers and users, the strong, continuing threat of further CFC regulation would tend to have the same effect as further regulation itself, due to the necessity for business to plan its activities opposite a 10 to 15 year time horizon (this point is elaborated on under "Uncertainty", below). 2. Impact on CFC Prices EPA does not dispute that CFC prices would rise in response to further regulation. In fact, EPA counts on this to be the engine which drives reduction in CFC use over time. However, EPA is apparently assuming this process will take place in an orderly manner with only minor, if any, economic dislocation. Unfortunately, EPA's logic is a gross oversimpli- fication of a highly complex, dynamic process. In our view, vII-12 ------- Economic Considerations under the regulatory scenarios proposed by EPA, CFC prices would rise very rapidly, at times in a totally uncontrolled manner, and, in their wake, create substantial economic dislocation. The following analysis illustrates why this would happen. Figure 1 (page VII-14) portrays a typical supply/demand relationship and the resulting so-called "equilibrium" price. The upward sloping supply curve means producers will be willing to supply more of the product as its price rises; conversely, the downward sloping demand curve means users will demand less of it as the price rises. Assuming these relationships reasonably reflect economic behavior, the system can readily adapt to change. For example, if demand patterns were to shift due to a technological change creating new end uses for the product, the demand curve would shift to the right, meaning that more of the product would be consumed at each level of price. This causes the equilibrium price to rise in the short run (Figure 2 - page VII-15). Producers, noticing the demand shift and reaping the benefits through higher prices, might then be disposed to add production capacity, thus shifting the supply curve to the right at each level of price. This causes the equilibrium price to decline (Figure 3 - page VII-16). What this demonstrates is that the natural tendencies of the system permit an orderly process of change that, absent general inflation, causes reasonable price stability over time. ------- Typical Supply/Demand Relationships Price Supply Curve Demand Curve Quantity PE = Equilibrium Price QE = Equilibrium Quantity Figure 1 w o o o 3 P- O o o 3 CO I-1- PJ n- p- o 3 in ------- Pro-Forma Supply/Demand Relationship After Demand Increase EO D1 Q Figure 2 w o o 3 O 3 H- O O o 3 tn ro H (U rl- H- o 3 cn ------- Pro-Forma Supply/Demand Relationships After Supply Increase < H H I Po p, Q, M a o o H- O O o 3 CO H» a 0> Figure 3 O 3 CO ------- Economic Considerations However, under the subject regulatory proposal, the supply and demand relationships for CFCs would present a vastly different picture. Four points are pertinent: Absent regulation, the general level of demand for CFCs would increase over time, i.e., the demand curve would shift to the right; Demand for most CFCs is relatively inelastic, i.e., demand is not particularly sensitive to price -- this would be represented by a steeply sloping demand curve; Regulations, as proposed, will have the impact of freezing supply at some level--a supply curve perpendicular to the X-Axis would depict this; then as reductions in the cap are implemented, this supply curve would shift to the left; and finally, General inflation will likely continue at current rates, say 10% per year, over the long-term -- this also moves the demand curve to the right. Figure 4 (page VII-18) is an illustration of these CFC supply/demand relationships and their probable impact on price. As indicated, these relationships complement each other in creat- ing much higher CFC prices. Moreover, with the supply curve ver- tical and the demand curve steeply sloped, very small changes in CFC availability or demand portend very large CFC price in- creases. Obviously, such a situation would be extremely volatile. For example, prices would be affected, not only by the VII-I7 ------- Pro-Forma Supply/Demand Relationship After Production Cap. 00 1 QEO Figure 4 td o o 3 O 3 H- O O o 3 CO H- D. CD rt H- o 3 (0 ------- Economic Considerations forces just described, but also by artificial forces specu- lators, hoarding, rumors, etc. The volatility that has existed in exchange markets for many years and more recently in the bond markets bears witness to this. Another price characteristic worth noting is that after a rapid increase, prices seldom come down as far as they went up, even if most of the increase was unrelated to underlying economic forces. The foregoing scenario of vastly higher CFC prices and an extremely unstable price environment would have many unfavor- able economic consequences; these are described in later sec- tions. The point to be made here is that EPA misreads the supply/demand forces affecting CFCs. Whereas EPA sees only a gradual increase in prices over time reducing CFC use at a pace industry can manage, we believe the more likely outcome would be a dramatic rise in CFC prices, particularly as further proposed restrictions are imposed, to the point where a chaotic market environment might ensue. The findings of Rand are worth noting: "Our cautious assumptions about CFC demand imply that a very high tax or permit price would have to be set throughout the entire period to achieve a cumulative emissions reduction equivalent to zero growth." [Rand, 1980, p. 221] . 3. Inflation Rand indicates that regulation-induced increases in CFC prices will not contribute to a higher general inflation rate because such increases will merely create transfers of funds within the economy and will not have a fundamental influence on the money supply [Rand, 1980]. This logic is tenuous in the context of current economic realities. In particular, a fundamental consequence of a CFC production cap would be a reduction in aggregate supply. Economists generally agree that VII-1S ------- Economic Considerations such actions are inflationary. Witness the public policy move toward so-called "supply-side" economics where the intent is to increase aggregate supply as a principal means of reducing inflation. Also, it is extremely unlikely the government would permit the market to fully determine to whom and in what magnitude these transfer payment funds should accrue. Rather, it is likely these funds would flow to the government for retransfer in some discretionary manner. This, in turn, is more likely to stimulate demand than supply, an inherently inflationary action. Other factors would also cause inflationary impacts: (1) the greater magnitude of CFC price increases than envisioned by EPA or Rand would create a larger flow in inflation-inducing transfer payments; (2) since CFCs are necessary in the production of so many products, including basics like housing and food, the downstream effects of higher CFC prices would contribute to inflation; (3) capital requirements of current CFC users to support shifts to substitute products could cause an increase in the level of interest rates, another source of inflation. Another aspect of the inflation question needs to be kept in focus. Inflation is viewed as the nation's foremost domestic problem and efforts to reduce it or, at least, to control its growth dominate public policy initiatives. In this environment, it would seem that government actions which would have the effect of accelerating the rate of inflation, as would the proposed CFC regulation, should be judged against particu- larly exacting standards as to their impact on the public welfare. 4. Economic Growth a) A generally accepted fact of economic life is that there is an inverse proportionality between the rate of VII-20 ------- Economic Considerations inflation and the rate of economic growth. Accordingly, further CFC regulation, with its attendant inflationary consequences, would tend to slow the overall rate of economic growth. More particularly, the principal determinant of economic growth is the aggregate level of business investment in productive capacity. The flow of transfer payments which would likely occur following further CFC regulation (as described earlier) would create a demand bias that would have the effect of reducing funds available for business investment. b) On the microeconomic level, a firm's ability to grow is determined principally by its profitability. CFC user, and probably also producer, profitability could be adversely impacted by CFC regulation. With the very large number of firms involved, the impact on overall economic growth could potentially be fairly large. c) Small business considerations are pertinent as well. In most CFC user industries, small businesses predominate. For these firms, the more appropriate consideration may not be their continued ability to grow but rather their survivability. A characteristic of such firms is a general inability to withstand significant adverse changes in their cost structure, capital requirements, or final demand. Under the economic scenario of further CFC regulation developed thus far, it is apparent many of these firms would be hard pressed to survive. Obviously, this downside possibility involves negative consequence not limited to the economic arena. 5. Employment Continuing to pursue these economic interrelationships, we see that, since high inflation and low growth mean higher unemployment, further CFC regulation can lead to an increase in unemployment. Potential examples include the following: VII-21 ------- Economic Considerations Failure of small firms as a result of CFC regulation would cause an uncompensated loss of jobs; Many larger companies, whose survivability is not the issue, might nevertheless face the necessity to close down plants. Since many of these plants are located in small towns and are a significant factor in the local economy, ripple job loss effects might also be a prospect. Rand notes: "Plant closures are an extreme manifestation of a more general consequence of regulation. Fixed investments have been made in the past in equipment, structures, and human skills that cannot be easily adapted to the new regulatory environment. Under regulation, these investments are devalued. In the extreme case, a plant is closed down, some if its equipment might be sold, but the rest is scrapped. Workers are laid off, and while they eventually find other jobs, they cannot use certain skills specific to their earlier employment. But even if a plant does not close, returns to fixed capital, both physical and human, are less under regulation than had been anticipated when the investments were made." [Rand, 1980, p. 235]. 6. CFC Substitutes In assessing potential economic impacts of CFC regulation via a production cap, a critical area is the avail- ability and cost of functional substitutes for CFCs. In particular, if safe, cost effective substitutes were generally available and overall CFC demand were not growing, CFC emissions could be capped and probably also reduced at some reasonable pace over time without the grave economic consequences described above. As indicated earlier, CFCs have always sold at a price which is significantly less than value-in-use. This is basically VII-22 ------- Economic Considerations the result of the interaction of two factors: (1) from an end-use standpoint, CFCs represent a unique and highly desirable combination of physical and chemical properties, including high energy efficiency, low toxicity and low chemical reactivity; and (2) from a production viewpoint, manufacturing processes are not. particularly complex, necessary raw materials have been available, and it has been possible to expand CFC production capacity at a rate necessary to support demand growth, all of which have contributed .to the business being highly competitive at the producer level. One of the consequences of this is that the process of developing functional substitutes for CFCs is in its infancy, simply because there has been no economic incentive to pursue it prior to the threat of CFC regulation. Accordingly, across all the important applications of CFCs, end users face relatively unattractive options were CFC availability to be restricted: Less satisfactory (safety, performance, value) but available functional substitutes; Large investments in new plants; retooling of production lines; redesign of products to avoid or reduce CFC use; or Investment in development of a new product or process to replace CFCs. Which option a particular end user chooses is largely an economic decision but one importantly influenced by several qualitative factors, such as legal restrictions, safety, environmental concerns, confidence in new technology, lead time necessary to implement major investment decisions, and risk of failure. In the end, the typical end user would probably face a rather narrow range of options to CFCs, and the narrower the range, the more VII-23 ------- -conomic Considerations likely the options would involve significant cost and consequent adverse economic impact. Section II, on CFC uses and essentiality, reviews in detail the lack, or limitations, of currently available substitutes for CFCs in their major end uses. And in Section VIII and Appendix B we cover in detail the process of seeking and developing commercially suitable new substitutes for CFCs the pitfalls, risks, long lead times and status of these efforts. However, for perspective, this section will review one specific example of the process of developing a suitable alternative with which Du Pont is familiar the search for an alternative for auto air-conditioning. This is included here to provide a feel for the magnitude of the job involved and the economic consequences, and because we believe the realities presented herein for this one CFC application are fairly typical for CFC uses in general. Approximately 100 million pounds of CFC-12 are consumed annually in the auto air-conditioning market (original equipment and after market). This is considered a significant CFC emission source by EPA and one, therefore, that presumably would be impacted by future regulations. Ever since the potential for CFC regulation became apparent, the auto companies have expended considerable effort looking at options to current systems charged with CFC-12. Du Pont, for its part, has mounted a significant effort, involving many technical man years, looking for an alternative to CFC-12 in auto air-conditioning. The current status of this -effort is as follows: After screening hundreds of compounds, Du Pont has identified only one candidate refrigerant substitute for CFC-12 FC-134a which appears to have physical and chemical properties that would permit it to substitute for CFC-12 without major equipment modifications. VII-24 ------- Economic Considerations However, the chemistry of this compound is significantly more complex than that of CFC-12, such that we have not yet been able to demonstrate a process for commercial scale manufacture, nor do we feel we are particularly close to such a breakthrough. Moreover, were this breakthrough to come, it would then still require perhaps 6-10 years before commercial production could begin, considering the lead times for toxicology (2-3 years), bench scale and pilot plant work (1-2 years), design (1 year) and plant construction (2-4 years) . Finally, even assuming FC-134a could successfully be brought to market, its price would probably have to 4 be initially in the $10-$20 per pound range (1980 dollars). Such a high price would be necessary to justify: (1) the high development cost, (2) the high construction cost for a new manufacturing facility (especially given the high cost of money), (3) the risk in proceeding (e.g., if CFC-12 were to be vindicated from concerns over ozone depletion, it would continue to be used and there then would be no market at all for a high priced replacement); and to support profitable operation at relatively low initial production rates. This compares with a current price for CFC-12 of approximately 5$0.50 per pound. A new manufacturing facility would be sized to provide adequate production for demand many years into the future. However, initial demand would come nowhere near filling this capacity for two reasons: (1) it probably would take manufacturers several years to phase out CFC-12 in their equipment lines and replace it with FC-134a, and (2) the vast majority of demand for CFC-12 in the auto air-conditioning application is not in original factory equipment but rather in the sector of industry which services this equipment. And the reality here is that even when all new auto air-conditioning systems begin to use FC-134a in a given year, all automobiles manufactured in preceding years would have to continue to use CFC-12. Thus, it could take 5-10 years before the demand for FC-134a reached a level which would permit the operation of the new manufacturing facility at high enough rates to bring the cost per pound down as a result of economics of scale. VII-25 ------- Economic Considerations If a direct replacement for CFC-12 cannot be made available, the auto companies indicate the development and retooling costs associated with implementing an alternative equipment design using CFC-22 (a less severe theoretical ozone depletor) would be in the billions of dollars. There is really no other long-range solution other than limiting auto air-conditioning in general. EPA itself concludes emission control has only minor potential in this end use. As discussed more fully in Section VIII and Appendix B, generally similar circumstances for other large volume CFC end uses can be cited: Solvents Substitutes are environmentally suspect and also face potentially stringent regulation. A large investment in new equipment is required in most instances when an alternate solvent is introduced. Blowing Agents In insulation foams, there are no substitutes for CFCs which can produce a foam with the same insulating properties. - Alternate blowing agents (methylene chloride and pentane) in other applications involve environmental or flammability concerns. The recycle option requires capital expenditures. VII-25 ------- Economic Considerations Refrigeration - There are no practical substitutes for CFCs in important end uses. Low toxicity and nonflammability are properties of CFCs which have been key factors in their growth over time. CFC regulation will, therefore, necessarily involve safety trade-offs. In addition, a number of the compounds which have been cited as potential CFC replacements in certain end uses are themselves environmentally suspect. These factors raise some very critical questions: a. Is there the possibility that further CFC regulation will, on balance, have a negative impact on the environment? We quote from Rand: "In some product areas - most notably flexible foams, solvents, and sterilants - a significant opportunity for reducing CFC emissions lies in substituting other chemicals for CFCs. The alternative chemicals may impose environmental or worker health hazards of their own. In the absence of controls on the alternative chemicals, policies that work well in reducing the ozone depletion risk from CFCs will increase the risk of other hazards." [Rand, 1980, p. 248]. b. As a regulatory option, is the economic incentives approach likely to be more or less effective than command and control options in minimizing overall environmental impacts? Again, from Rand: The safety and environmental limitations of currently available substitutes for CFCs are discussed in detail in Section II. VII-27 ------- Economic Considerations "Although even mandatory controls will encourage some firms to use chemical substitution to avoid costly compliance with CFC regulations, the degree of chemical substitution should be far greater under economic incentives than under mandatory controls. To the extent that the substituted chemicals are found to be hazardous to worker health or the environment, this greater substitution is a disadvantage of economic incentives policies." [Rand, 1980, p. 18]. c. What happens if chemicals which EPA is counting on as CFC substitutes are themselves regulated at some future date? Rand offers two noteworthy conclusions on this point: "Economic incentives rely more heavily on chemical substitution because that is less costly in many cases than alternative means of CFC emissions control. Under any policy strategy, the attempt to control substitute chemicals will make the policy less effective in reducing CFC emissions than the estimates given in this study, which assumes no other changes in regulatory controls for non-CFC chemicals." [Rand, 1980, p. 248]. "The estimates of emissions reductions and compliance costs presume that no regulatory restrictions will be placed on chemicals that might be substituted for CFCs. The effects of this presumption are not trivial." [Rand, 1980, p. 12]. It is clear from the above that any regulatory initiative which necessarily relies on substitution (as does the economic incentives option) must be analyzed carefully opposite the cost and risk from using substitutes and the future availability of the substitutes. We note that the ANPR makes no statements in this regard. The question of risk from regulation and the need for risk-risk comparisons is taken up in Section V. VII-2- ------- Economic Considerations 7. Energy It becomes apparent when considering the important CFC end uses that CFCs gain a significant portion of their current value-in-use due to their energy efficiency. CFC solvent systems are preferred to other systems in certain end uses because the lower boiling point of CFCs means less energy consumption. In the insulation market, foam insulation materials blown with CFCs have the lowest K factor (a measure of relative insulating ability) of any available insulation material, and this is due to the inherent insulating properties of CFCs retained in the foam. CFCs also are the heat transfer medium in heat pumps, considered to have, important potential for greater energy efficiency in residential applications. Three emerging end uses for CFCs (1) use of CFCs as heat transfer fluid in solar energy applications, (2) use of CFCs as the blowing agent in structural foams, and (3) use of CFCs in the beneficiation of coal derive their potential from the nation's desire to reduce its dependence on imported oil. 1) In the solar energy application, CFC systems have been shown to represent the best combinations of energy efficiency, thermal stability and operating reliability among available systems. 2) Structural foams blown with CFCs provide molded parts with high strength with a low plastic resin requirement, resulting in a significant reduction in weight. Uses include automobiles and aircraft and other applications where weight reduction to save energy is an important consideration. o A more detailed discussion of the energy efficiency consequences of CFC regulation appears in Section II-K and Appendix C. VII-29 ------- Economic Considerations 3) In the coal application, a process using CFCs to remove nonburning matrix and sulfur has been commercially demonstrated. Growth of solar energy, structural plastic foams and coal beneficiation would contribute importantly to solving the nation's energy problems. Such growth would obviously be greatly restricted by a CFC production cap. There are other potential energy savings developments only beginning to be explored which would likely require CFCs, for example, waste heat recovery and geothermal applications. These also would be hurt by a res- tricted availability of CFCs. e It is worth noting that certain government actions in other areas relating to energy run totally counter to EPA's regulatory thrust on CFCs. Specifically, in the Department of Energy's recent energy efficiency requirement proposals [DOE, 1980] certain standards provide for an increase in CFC use due to their energy efficiency. 8. Financial Markets Among the several options to control CFC emissions cited by EPA recovery equipment, redesign and replacement of existing production equipment, and alternate product designs most would force CFC end users to make new capital investments. Over time, these regulation - mandated expenditures could become significant in the aggregate which could negatively impact the nation's financial markets. In recent years, these markets have been extremely volatile; interest rates have reached unprecedented levels, particularly in long-term markets. Pressures on financial markets will probably continue largely unabated based on our continued high inflation rate, high level VII-30 ------- Economic Considerations of government spending and capital expenditures by business necessitated by environmental and other regulations. With this scenario in prospect, is it appropriate through such measures as CFC regulation to add to the burden on financial markets? 9. Impact of Uncertainty Perhaps even more important than any of the economic issues addressed thus far, is the impact of uncertainty. In particular, it is the key to reconciling EPA's position that its proposed production cap on CFCs can be accommodated without significant economic penalty and the view of others, shared by us, that such a production cap would create severe economic hardships. We and EPA do not disagree on the direction of future events following imposition of a production cap on CFCs, i.e., higher CFC prices and lower CFC consumption through non-market restraints on demand. The area of disagreement is the magnitude of the impacts once events run their course. EPA would argue that, by merely capping CFC production at current levels, the gap between permissible supply and demand would be sufficiently small initially, and would grow each year in sufficiently modest increments, that readily available options to reduce CFC consumption could cover these gaps at manageable cost, from both the macro-and microeconomic viewpoint. EPA would also argue that this same process could accommodate, not only control of CFC growth, but also subsequent reductions in CFC consumption from base period levels. The precise target EPA has in mind for eventual CFC use reductions has been stated as 50-70% from base period levels [Jellinek, 1980a]. Here is where the uncertainty issue comes to play, specifically in the context of its role in business planning. Business at all levels must manage its affairs with a view toward the long-term because the most critical decisions it makes are investment decisions (those involving capital outlays for plants VII-31 ------- Economic Considerations and equipment) where the wisdom of today's choice cannot properly be judged until many years in the future. Recognizing this, business relies heavily on the process of long-range planning to develop perspectives on the possible direction of future events, the likely impact of these events on the firm and the actions required to either avoid problems or capitalize on opportunities. Essentially, it is a process of adapting to uncertainty; the hoped for end result is a reduction in uncertainty to manageable levels as a basis for prudent investment and other decisions. One characteristic of this process is particularly pertinent to this discussion. Since business is fundamentally risk adverse, a potential future event which would have a significant impact on a firm (particularly if it were negative) would dominate that firm's long-range plans. Moreover, the higher the probability of the future event, the more impact on the firm's strategic thinking. What impact, then, would EPA's proposed CFG production cap have on an affected firm? Two perceptions would quickly be formed in the course of developing plans: In practice, the functioning of the cap mechanism (with its permit system and reliance on bidding wars to determine who gets CFCs, in what quantity and over what time period) would be highly unpredict- able, and There would be a fairly high probability mandated reductions in CFC use over time would be sufficient- ly large to significantly impact all end users, regardless of essentiality and lack of substitutes arguments, but the timing and stepwise incidence of such reductions could not now be predicted. VII-32 ------- Economic Considerations In shorty tremendous uncertainty would be created. How would a firm respond? We believe that the prudent firm, attempting to maintain its health over the long-term, would have no choice but to adopt an "assume the worst" posture. In other words, since (1) the potential long-term impact of the cap is severe opposite the way the firm does business today, and (2) perhaps a ten-year lead time is involved for the firm to attempt to make the kinds of changes necessary to avoid those severe impacts, the firm would be forced (to ensure its long-range viability) to project the longer-term EPA 50-70% use reduction scenario back to today's decision processes. In short, firms would act as though CFCs have been essentially banned, not just capped. Rand expresses general agreement with this view: "In contrast, firms may perceive tax rates or quota levels as highly variable, subject to regulatory whim or political manipulation. If so, firms might be reluctant to undertake long-term investments that would reduce emissions for fear that future regulatory action would make the investment obsolete or reduce its cost- effectiveness." [Rand, 1980, p. 242]. The uncertainty issue cuts deeper, however. For example, how would public policy deal with the following kinds of situations? CFC user firms who have no current alternatives to CFCs would want to take steps to ensure CFC supply over very long periods of time and at levels which permit growth. Since a very large volume of current CFC uses falls in this category, the effect on CFC prices would greatly exceed anything EPA is currently anticipating. CFC user firms who take a gamble on new investments to alter CFC use patterns, and guess wrong as to how VII-33 ------- Economic Considerations regulation finally affects the business, would probably, in most cases, face the survival question. Firms who have capability to do development work that might eventually produce CFC substitutes or alternate downstream products that do not require CFCs would have difficulty justifying the project from the economic viewpoint because of the uncertainties as to payoff. To sum up, EPA's posture that CFC regulation can be approached one step at a time at a measured pace is unworkable in practice. In particular, EPA's first intended regulatory step the production cap will have the capability to create severe and immediate economic consequences, whether or not further CFC use restrictions are implemented. In other words, once the process is started, it may be difficult to alter its course even if this is desired. Major interruption of long-term business plans in mid-course would be simply chaotic from the economic viewpoint. Moreover, the so-called "protected" CFC end uses (those that are essential and without available substitutes), would not be immune from the impact. In fact, it can be argued that because there are no alternative options for these uses, the impact of uncertainty over the future would hit firms in these businesses the hardest. VII-34 ------- Economic Considerations D. CFC REGULATION VIA COMMAND AND CONTROL TECHNIQUES Command and control regulatory approaches could achieve reduction of CFC emissions in several ways: CFC product bans, CFC use bans or limits, Mandated minimum technology standards, and/or Mandated emission reduction techniques. Considering the number of affected CFCs, their wide range of end uses, the several command and control options cited above, and the fact that given options affect different firms in the same industry in different ways, we agree with EPA that regulating CFCs in this manner would be complex. However, as noted in section C, regulation of CFCs by economic incentives options also would be extremely complex and, unlike command and control options, very little is yet known on how the economic incentives options would work in practice or the probable con- sequences. Therefore, we believe that command and control op- tions must be explored in much greater depth as a viable regula- tory strategy. EPA sets forth reasonably well in the ANPR questions that need to be answered to develop command and control regulatory options that reasonably balance economic consequences with CFC emission reduction goals. Unfortunately, few, if any, of these questions have yet been answered to any sort of definitive degree. The applicable studies in hand [Rand, 1980; NAS, 1979d] are seriously deficient in this regard. In-depth study is required to fill the information and analysis gap. The fact that EPA prematurely has identified a preferred regulatory VII-35 ------- Economic Considerations option other than command and control does not excuse EPA from the responsibility to proceed with rigorous analysis of all relevant options, including command and control. In fact, it is arguable whether EPA will have met its legal obligations were it to continue to advance a regulatory option without adequately evaluating competing options (See Section III - Legal Considerations). VII-36 ------- Economic Considerations E. INADEQUACY OF RAND REPORT TO SUPPORT A REGULATORY DECISION9 1. Introduction On balance, the Rand [1980] Report is a useful start on the difficult task of assessing the implications of some of the key potential regulatory options which could be applied to CFCs. However, there are a number of serious omissions and deficiencies which must be addressed prior to making a regulatory decision based on the report findings. The study is flawed opposite the use to which it is being put by EPA because it: (1) is based on an obsolete data base, (2) does not adequately examine all CFCs being proposed for regulation by EPA, (3) does not adequately assess all uses of CFCs which would be impacted by EPA's proposals, and (4) only compares regulatory options under artificial time and technical feasibility study parameters. However, most importantly, the study is an empirical comparison of regulatory options under study bounds selected by the authors, not an economic impact study of the consequences of these options were they to be applied to the real world of CFCs (all CFC products, all CFC uses). The fault is not Rand's but rather EPA's for attempting to justify a regulatory decision on the basis of a report not designed for this purpose. A couple of examples are illus- trative: 9 A more detailed commentary on economic incentives options as treated by Rand (and by EPA in the ANPR) appears as Appendix I. A companion product is Du Pont's March, 1980 submission to EPA [Du Pont, 1980c] which critiqued the Rand Draft Report made available to industry for comment. We note that only a few of our comments and none of the major concerns raised in the critique have been addressed in the final report now being used by EPA. Therefore, we attach our full critique as part of this submission. VII-37 ------- Economic Considerations a) EPA [Jellinek, 1980c] states its preference for economic incentives options on the basis that these will be less burdensome than traditional command and control options for a given level of emisssions reduction. EPA cites [Inside EPA, 1930] for this purpose Rand's estimate of real resource costs for a no-growth-in-emissions regulatory policy of $270-600 million for a ten year period. EPA has used these numbers out of context. Rand's estimates were based on only a partial control of CFC production. EPA has used the numbers in the context of total CFC production. Rand examined the cost of control opposite the artificially selected base which covered only 310 million pounds of CFC use [Rand, 1980, p. 35]. The 1900 level which EPA proposes to control is in excess of 300 million pounds (See Appendix J). Further, Rand's base case did not include CFC-22 as evidenced by the following quote: "Therefore, we have not treated CFC-22 as a principle ozone hazard, and home and supermarket air-conditining systems are not included in the list of analyzed products." [Rand, 1980, p. 3], Yet, the proposed controls in the ANPR include CFC-22. Further, Rand only estimated the cost between 1930 and 1990. In reality, costs would continue to mount for every year after regulation. Another point is that the $270-600 million estimate is only for compliance cost. As we shall see, economic incentive options have associated with them VII-3- ------- Economic Considerations costs called transfer payments, which are in economic incentives options have associated with them costs called transfer payments, which are in addition to, and well in excess of, the compliance costs. b. EPA's statements on the cost efficiency of economic incentives options generally have focused on compliance cost and glossed over the much larger economic impact of transfer payments [Inside EPA, 1980], although in the ANPR it was acknowledged that "The Rand Corporation estimates that total transfer payments will eventually exceed $2 billion." In reality, the Rand Report concluded that for its "benchmark" emissions reduction case (a reduction in emissions by 1990 of only about 12 percent from the no regulation case [Rand, 1980, p. 21S] transfer payments would run $1.5-1.7 billion [Rand, 1980, p. 281]. However, as stated above, this was based on a study base of 310 million pounds in 1976, not the 800 million actual pounds used in 1980; and CFC-22 was not included. A more important point, however, is that EPA's ANPR proposal is not for this benchmark level of emis- sions reduction but rather for a ceiling at current levels. Rand terms this zero growth case as "strin- gent" and concludes that transfer payments just for the period 1980-1990 would be up to $6.2 billion [Rand, 1980, p. 230], again based on only the par- tial piece of the CFC market examined. Factoring all CFCs and uses in would put this number at approximately $16 billion. The statement by EPA in the ANPR that transfer payments "...will eventually ------- Economic Considerations exceed $2 billion" grossly understates the economic consequences of the proposed regulation and, at a minimum, shows a poor understanding of the basis for the numbers developed in the Rand Report. The underlying reality is that due to the selected bounds of the study performed by Rand (bounds which limited its analysis to less than half of the current actual CFC market) any citation of cost projections from the report must be strictly qualified. The Rand study was an empirical exercise, not an economic impact study of EPA's proposed production cap. 2. Data Base The Rand study had three objectives: a. To update and extend data on CFC use and emissions. b. To analyze the economic properties of various tech- nologies and procedures by which industry might be able to reduce CFC emissions, and c. To assess the economic implications of alternative regulatory strategies. Each objective depended upon information generated in the previous objectives and was subject to the inherent quali- fications and limitations of each previous step. This point cannot be stressed too strongly. Readers may focus on the findings or conclusions from the assessment of the various regulatory strategies while neglecting the proper qualifications and reservations necessitated by the limitations in the under- lying technical and economic data bases. VII-40 ------- Economic Considerations Therefore, the logical starting point for a critique of the report would be evaluation of the data bases (for accuracy and completeness), followed by an effort to improve these where possible. Unfortunately, this task is made very difficult and in some cases impossible due to lack of publication of the final data bases employed. It is unfortunate the final report is not organized as described by EPA at the inception of the study. As an example, in an interview with the EPA project officer for the Rand study it was stated: "The idea here is to produce a consistent set of data which everyone views as reasonable." "...the other idea is to allow anyone to use precisely the same data we used to reach conclusions." [Mader, 1978]. Rand did produce draft data books which were sent to industry for comment. Reviews were made and comments were returned to Rand and to EPA. But final data books reflecting the draft comments have not been finalized and published for "anyone to use". Fur- ther, the draft data books dealt almost exclusively with emissions profiles and the technological aspects of CFC use, not with the economics, e.g., demand functions. Consequently, our critique of the Rand Report is necessarily limited. In those areas where the data appears in the body of the report, it has been reviewed and comments appear in our critique [Du Pont, 1980c] (attached). In those areas where the raw data are aggregated or not reproduced in the report, our comments center on the analytical assumptions made and the limitations on the use of the findings. One of the key limitations in not having ready access to the full data base is that in a number of cases where we disagree with Rand's conclusions, it is not clear whether the source of VII-41 ------- Economic Considerations disagreement is the basic data or the reasoning leading to the conclusions. It is hoped that the final data on use and emissions as well as the details of the economic analysis, will be made available in the future for full examination and commentf as originally planned. The authors identify major data deficiencies and in a number of instances recommend the need and direction for further work. As a consequence of these data deficiencies, the Rand researchers had to make a number of assumptions concerning use and emissions' data, and the technological feasibility and economics of alternatives and emission control. Therefore, we provide comments on these assumptions, and discuss the limitations which must be assigned to the assessments performed in the report as a consequence of the assumptions employed. As Rand itself states: "The methods and assumptions [used in the impact analyses] deserve close scrutiny because they influence the outcome of analyses presented in later sections [of the Rand Report]." [Rand, 1980, p. 22]. In many instances, not only is the data not available but the specific logic or reasoning leading to the conclusions is missing. It often is not clear to what extent conclusions are reached from hard facts and clear understanding of the uses and users, versus speculation or estimation. The development of the CFC use and price combinations is particularly bothersome. Rand states that the demand schedules represent their "point estimates" of use versus price but neither the data base nor the estimation process is reported. VII-42 ------- Economic Considerations 3. Study Assumptions a. No Regulatory Restriction of Alternatives In the report introduction, Rand states: "The estimates of emissions reductions and compli- ance costs presume that there will be no regulatory restrictions on chemicals that might be substituted for CFCs. The effects of this presumption are not trivial. [Rand, 1980,, p. 12]. Not only are the effects not trivial, but this assumption is unrealistic. Many of the potential substitute chemicals already are under regulatory control due to concern over their photochemical reactivity (e.g., perchloroethylene, trichloro- ethylene and pentane); some are already under regulation for flammability (e.g., pentane and other hydrocarbons); a number of the key substitutes are burdened with unresolved toxicity questions (e.g., methylene chloride and trichloroethylene); several are likely to be impacted by water quality controls; and some are also implicated in stratospheric ozone depletion (methyl chloroform). A move to CFC substitutes, which already are regulated or may be regulated in the future, would increase the cost of regulation beyond Rand's estimates, because these estimates were developed under the assumption such substitution would be a regulatory "free ride" (See section C-5). b. Discount Rates Rand assumes future compliance costs of CFC regulations should be discounted back to 1980 at 11 percent per year. It was stated that this rate was used by Rand for consistency with cost- benefit analyses being conducted for EPA by the University of Maryland [Bailey, 1980, p.38; Bailey, 1979]. We believe that if future compliance costs are discounted back to current dollars at a certain rate, then the quantifiable future benefits of any / VII-43 ------- Economic Considerations proposed regulation should be at the same rate. We are not prepared to argue for or against any specific discount rate, but feel strongly that the one used, implicity or explicitly, should apply equally to both sides of the cost-benefit equation. c. Time Delay of Emissions Rand makes the assumption that: "The ultimate effect on the ozone layer is essentially the same for a given cumulative emis- sions level, regardless of whether the emissions occur in a brief burst or over a period as long as a decade." [Rand, 1980, p. 32]. The statement is true for the ultimate potential effect but needs to be put into perspective. The key to this assumption is the parameter of "a period as long as a decade." A decade seems to have been chosen to fit the arbitrarily selected boundary for the comparative effects analyses (1930 to 1990) . But the banking characteristics of some of the uses, especially insulating foam, result in the emissions occurring over 50 to 75 or more years, not over a decade. In those instances, assuming other emissions reductions were mandated, the effect would be to draw out the emissions profile.over sufficient time that the ultimate poten- tial depletion would not be as great, due to the self-healing nature of the ozone layer. d. Transfer Payaments Not Inflationary Rand argues that tax or permit payments to government by industry are simply transfers of wealth within the economy and, as such, are unlike regulatory compliance costs which reflect increased use of real resources to avoid CFC emissions. And since such payments do not use up real resources and are eventually returned (in some unspecified fashion) to the economy, they do not directly contribute to inflation. We suggest this is VII-44 ------- Economic Considerations an overly fine distinction. This point was discussed in section C-3 so will not be repeated here. e. Scope of Economic Incentives Regulatory Options Rand's quantitative analyses of the economic incentives policy designs assume all applications of CFCs would be subject to the policy. Justifications offered for this feature are that: i) overall economic efficiency would be promoted by encouraging pursuit of the least costly combination of emissions-reducing activities, and ii) it would be less costly to administer. Yet on a more detailed examination, such a policy would have built-in inefficiencies to the extent that certain uses of CFCs (uses where no alternatives are available and little or no emission reduction is possible in the foreseeable future) would be increased in cost but without any meaningful associated CFC reduction. This is not a small concern. Note the findings of Rand: "The non-aerosol CFC applications that do not appear to have technical options currently -- of which foam insulation and refrigeration products are the largest account for the largest fraction of projected CFC use over the next decade." (emphasis added) [Rand, 1980, p. 236] . Thus, under EPA's proposal, the largest fraction of projected CFC use has no technical options, yet would be required to pay the higher prices generated by the production cap. This would be a clear case of regulation with no benefit, but high cost. f. Legal Concerns It is assumed by Rand that the potential for collusion among firms or predatory behavior in the permit marketplace would not be a concern because neither activity would reduce the emissions-reducing potential of a permit policy. Nevertheless, VII-4! ------- Economic Considerations this must be examined carefully because of the potential for anticompetitive effects from a marketable permit system such as presented by Rand and proposed by EPA. The procompeti tive policies of the antitrust laws are important national objectives that cannot reasonably be ignored by EPA. The potential impact of the proposed permit systems could be contrary to the policy of the antitrust laws. (For detailed comments on legal points, see Section III). 4. Limitations on Use of Report Findings a. Introduction The following quotations from the Rand Report taken together illustrate why we have very serious reservations over use of the report findings by EPA to support a decision that economic incentives options are the preferred regulatory option: "First, in several of the applications where CFC demand is inelastic by assumption (e.g., LFF, sterilants, mobile air-conditioning) , options for reducing emissions have been identified but are not reflected in the CFC demand schedules solely because of the lack of cost data." (emphasis added) [Rand, 1930, p. 227]. "Of course, without more data about CFC demand schedules in these applications, we cannot predict the precise magnitude of transfer payments per application."(emphasis added). [Rand, 1980, p. 230], "Although the response of these product areas to economic incentives obviously cannot be predicted precisely with available information, these appli- cations well might contribute to emissions reductions under an economic incentive policy, especially one as stringent as the zero-growth scenario." (emphasis added) [Rand, 1980, p. 227], "Without detailed information on individual plants around the country, it is impossible to predict where plant closures caused by transfer payments might occur." (emphasis added) [Rand, 1980, p. 235], VII-46 ------- Economic Considerations "Transfer payments can cause short-run economic dislocations, both because the transfers might not reenter the economy instantaneously and because some human and physical capital is firm- or industry- specific and fixed in the short-run. Like the short-run phenomenon of rents, temporary dislo- cations due to transfer payments are omitted from this quantitative analysis of policy effects in this study. (emphasis added) [Rand, 1980, p. 27]. "This design of a compensation approach is far beyond the scope of this study." (emphasis added) [Rand, 19SO, p. 236]. "Solvent substitution is not, of course, a panacea. Other solvents appear to impose their own health and environmental hazards, and require some increase in energy utilization. The potential effectiveness and the desirability of using policy to induce substitution among solvents cannot be determined in this study, and remain important issues for further investigation by EPA. (emphasis added) [Rand, 19GO, p. 37]. Overall, it is clear that although the Rand study may have been a good start on the assessment of the practicability and cost of imposing economic incentives options, much additional work needs to b,e done. Detailed discussion follows on some of the major limitations of the report'findings: b. Analytical Conclusions Must be Extrapolated With Care The Rand comparative regulatory option analysis is based upon the selection of a set of benchmark controls. Care must be taken in extrapolating the conclusions reached from this bounded analysis to the real world of CFC uses and regulatory policy choices. In the comparative analyses, Rand first selects a set of "benchmark controls" (command and control steps) that they feel VII-47 ------- Economic Considerations could attain substantial emissions reductions by 1990, using available technology, without undue economic impact, and which could be readily enforced. Economic incentives regulatory options are then set to attain an identical emissions reduction goal and analyzed against the benchmark for efficiency and economic impact. Rand concludes from their analyses that: "If substantial emissions reductions beyond the limited capabilities of mandatory controls are required, the relevant policy choice appears to be between outright bans on CFC use and economic incentives." [Rand, 1980, p. vii] . The validity of this finding is questionable the weakness lying in the extrapolation of results from a tightly bounded comparative analysis to the "real world" of CFC uses and policy choices. Although Rand's findings flow logically from the analyses and seem valid in terms of comparing the regulatory options within the parameters of the analyses, they would not necessarily be valid were a different artificial or arbitrary set of analytical parameters employed. What we are left with then is a useful ordering and comparison of the pluses and minuses of the options, within limiting boundaries, but not a study which fully compares the potential impact of the regulatory options were they to be imposed in the real world of CFC uses. The scope of the selected benchmark controls necessarily bound the comparative analyses. Therefore, it is crucial that the set of controls selected for the benchmark regulatory option be realistic, accurate and complete. This does not seem to be the case from our perspective. Under the mix of benchmark controls selected by Rand, total emission reduction potential from a baseline case of no regulation is only about 12 percent. Preliminary work at Du Pont indicates this may understate, by at least a factor of three, the ultimate potential for emissions VII-43 ------- Economic Considerations reduction from a mix of technically feasible and enforceable command and control regulations. Should this preliminary estimate be confirmed, Rand's conclusion that the policy choice is limited to outright bans and economic incentives in not valid. Any regulatory recommendation made by EPA should be based on what actually is possible or achieveable under various options, not upon an arbitrarily bounded benchmark set of controls. Consequently, a much broader comparative analysis would seem to be required. . c. Regulatory Cost Remains Uncertain The cost advantage of incentives policies over direct controls could be greater or smaller than estimated by Rand. To determine the magnitude of costs imposed on firms by CFC regulation, it is necessary to obtain: i) information on the size of the market for goods whose production involves the use of CFCs, ii) the relationship of the cost of using CFCs to the finished product price, and iii) the production costs when alternative technologies that rely less on CFCs are employed. Due to the difficulty of obtaining this information, Rand relied on engineering estimates to develop the scope of the cost of possible substitutes for CFCs. The lack of good data and the consequent use of this estimating method introduces considerable uncertainty in estimating the costs of CFC regulation. d. Criteria for Benchmark Controls Too Restrictive In its selection of the benchmark control option, Rand groups categories of potential command and control steps based on the implications of using them. Then by imposing a number of restrictions or boundaries, the benchmark set of controls is selected. The selection requirements used are immediacy of VII-49 ------- Economic Considerations emission reduction, adequacy of technological and economic information/ and enforceability. We feel the parameters of thg requirements employed are arbitrary and artificial, leading to an overly restrictive set of possible mandatory controls. Each requirement is examined below: i. Enforceability - Giving undue importance to the question of perfect enforceability may well deprive the rulemaker and the impacted industries of imminently workable options options which could contribute significantly toward the attainment of an emissions reduction goal. ii. Adequacy of Information - Rand excludes certain potential control options from the benchmark set due to the need for further assessment of their technical feasibility and cost. This most often occurs with the options requiring the use of fluorocarbon alternatives. We agree that all of the necessary answers are not yet available but this does not seem to us to be a valid reason for discounting the potential of these technical options. It is understandable that adequate information could not be assembled on all promising control options given the time and dollar constraints under which Rand had to operate. Certain of these options were sufficiently new that adequate data simply did not exist. However, it does not seen defensible to select regulatory options based on the present Rand findings, without an effort being made to fill in the informational deficiencies identified by Rand. These options have VII-50 ------- Economic Considerations not been rejected as unworkable, only as lacking sufficient information upon which to evaluate them. An effort should be made to obtain the required information. iii. Immediacy of Emission Reduction - Another fault we find with Rand's screening process is the criterion that any control step selected for the benchmark option must produce "immediate" emissions reduction, i.e., the majority of the emission reduction gains must begin to occur before 1990. The year 1990 apparently is selected as it is the limit of Rand's data base emissions profiles and economic projections. The difficulty of obtaining credible projections well into the future is understood. However, it does not seem unreasonable to demand the long-term economic projections be performed. If one is to address ozone depletion effects 30 to 100 years in the future, one should be equipped and willing to address regulatory options, their efficiencies and costs within the same time span not to limit the analyses to 10 years. Similarly, in dealing with a long-term potential problem, it makes little sense to limit the evaluation of solutions to those effective only in the next 10 years. Due to the relatively long lead times required to implement many of the more promising control strategies identified, the real effectiveness of these measures often will not occur until after 1990. Arbitrary elimination of longer-term control steps from the analyses results in a severe understatement of the potential for emission reduction command and control steps. There is another problem with the requirement that Vll-Iil ------- Economic Considerations solutions produce immediate payback. This is the questionable regulatory efficiency (in pounds of emissions reduced per dollar control expenditure) under these restrictions. Were regulatory decisions to be based only upon emission reduction achievable by 1990, to the exclusion of potential ultimate emission reduction, there is little question that monies would not be spent efficiently. Increasingly large expenditures would be required to achieve incrementally small improvements by 1990, when less expenditure on other projects could produce much larger gains (although the effects would not be felt until after 1990). The goal should be to reduce total emissions over time relative to the perceived environmental risk over time not to maximize short-term solutions to a long-term potential problem. e. Designs of Economic Incentives Options Are Too General The designs of the incentives options analyzed appear more conceptual than workable. Although great care is taken by Rand to describe and bound the benchmark control option, the designs of the incentives options are much less specific. Many potential problems associated with design, implementation, administration and impacts of the options are identified, and in some cases potential answers or solutions are offered and evaluated. But a reader is not able to discern the exact structure of an option, how it would work or what the effects would be. Rand's approach is closer to a sensitivity analysis of the various facets of incentives policies than an analysis of how these options would be applied to CFCs. (For detailed discussion, see Appendix I). VII-52 ------- Economic Considerations f. Transfer Payment Concerns Not Resolved Rand concludes that the question of transfer payments may be the single most complex aspect of economic incentives approaches. A number of political and economic issues which should be addressed prior to implementation of these options are identified. Although a number of potential solutions are discussed, none are convincing. Uncoinpensated transfer payments acknowledged as a real cost of doing business result in greater regulatory costs than do mandatory controls. Compensa- tion schemes are recommended as a means of reducing the potentially large cost resulting from incentives options. However, neither the potential benefits nor the beneficiaries of compensation are clearly defined, and the real cost of some of the compensation schemes may be underestimated. The lack of resolution of this issue raises serious questions about the utility of the report conclusions on economic incentives options and EPA's stated preference for them. We note that EPA does not offer any solutions to these problems in the ANPR, even though many of these concerns were raised by us in our critique [^Djj Pont, 1980c] of the Rand Draft Report. We cite below representative statements in the Rand Report on the impact and importance of this aspect of the economic incentives options. Taken together, they indicate a problem so large that we believe it is mandatory that workable solutions be presented before EPA proceeds with a rule proposal which is based on incentives options involving transfer payments. "For the firms that pay them, transfer payments are an expense that will be reflected in higher prices to the consumer and a greater risk of plant closures and worker unemployment. Because the transfers are not a real resource cost, the negative effects on firms that pay them will be offset by benefits to the ultimate transfer recipients. Nevertheless, transfers are a policy concern because wealth redistribution and its effects on certain consumer VII-53 ------- Economic Considerations prices and on plant closures are politically sensitive issues." [Rand, 1980, p. 18]. "Economic incentives would impose lower costs on the economy as a whole, but could seriously injure CFC user industries unless wealth transfers are 'compensated'." [Rand, 1980, p. vii]. "An uncompensated incentives policy design would generate large transfer payments, ranging from a discounted cumulative total of 1.5 billion dollars for the benchmark-equivalent cost-minimizing design to 6.2 billion dollars for the zero-growth design based on cautious assumptions about the CFC demand curves." [Rand, 1980, p. 230]. "Uncompensated economic incentive designs will result in higher prices for final products made with CFCs than will compensated designs or mandatory controls. Under an uncompensated policy, firms and their customers bear the full burden of the transfer payments and total costs of production are higher. Although prices elsewhere in the economy should fall commensurately, in a trillion-dollar economy it cannot be predicted in which individual industries this effect will be noticeable. In short, the burden of transfer payments will be readily apparent, while the benefits might not be." [Rand, 1980, p. 234]. "However, uncompensated policies would generate transfer payments many times as high as compliance costs, and CFC-using firms would face total regulatory costs much higher under such policies than under mandatory controls. Consequently, there is greater risk of plant closures under uncompensated economic incentives policies." [Rand, 1980, p. 29]. "Devaluation of fixed capital occurs under any form of regulation, whether mandatory controls or economic incentives. However, the magnitude of uncompensated transfer payments implies that the wealth loss from capital devaluation in regulated industries is much greater under uncompensated economic policies than under other policy approaches. For these reasons, most firms would understandably prefer mandatory controls to uncompensated economic incentives." [Rand, 1980, p. 235] . VII-54 ------- Economic Considerations "Despite the advantages of economic incentives for reducing the real resource costs of regulation and achieving substantial emissions reductions, the adverse impa.cts on user industries from an uncompensated incentives policy may not be acceptable. If this is the case and substantial emissions reductions are required to prevent serious environmental damage, the achievement of regulatory goals may rest on the ability to design a compensated policy that does not distort incentives for low-cost emissions reductions." [Rand, 1930, p. 239]. g. Inadequate Attention is Given to Market Structure Effects of Regulatory Option Design The Rand Report portrays taxation and marketable permits as capable of producing broadly equivalent regulatory outcomes, with the potential differences that do exist possibly creating a slight preference for a permit system. But we believe a permit system could lead to increased concentration of either producing or using firms. And, as the efficiency of a permit system is dependent on markets being purely competitive and not becoming concentrated as a result of the permit system, failure to meet these conditions could create two problems. First, the compliance cost of attaining a designated reduction of CFC would not necessarily be minimized under a permit system. Second, if firms were to hoard permits, a permit system may not regulate emissions to their optimal levels, i.e., a permit .system may unavoidably lead to a greater reduction in the use of CFCs than would be socially desirable. (For detailed discussion, see Appendix I) . 5. Needed Further Work a. Consider Mixed Regulatory Options Rand only treats potential regulatory options in the pure form: that is, combinations of strategies for different VII-55 ------- Economic Considerations end-uses or over time are not analyzed. A regulatory mix should be considered. This could be done by analyzing options available for each use on a cost-effectiveness basis. Such analyses should include regulatory administrative costs as well as direct and indirect compliance costs. Cost-effectiveness analyses also should be used to evaluate stepwise introduction of regulatory options over time in a manner consistent with the anticipated increase in knowledge of the benefits of regulation (directly related to knowledge of the degree of ozone depletion and its effects). b. Risk Trade-offs Needed Many of the proposed and potential emission reduction processes or alternative products which would be utilized in response to incentive options have some degree of worker, con- sumer or environmental risk associated with them. Obviously these risks need to be weighed opposite the risk from continued CFC use. Option impact studies need to consider these risks in predictions of the likely response of industry to given incentive levels. If the perceived risk from alternatives is high (and thij» is not factored into the impact analyses), industry may not take the anticipated steps to reduce CFC use at a given incentive level thereby markedly increasing the forecast regulatory cost and decreasing the forecast benefits. c. Alternative Approach to Policy Evaluations The main problem with the utility of Rand's findings is that they are all based upon comparisons with an artifically bounded base case of benchmark controls. The result is that neither the incentive options nor the benchmark set of command and control rules adequately encompasses the real world situation in terms of total achievable emissions reductions from all uses^ and the associated costs. Consequently, the analyses do not VII-56 ------- Economic Considerations logically support a regulatory decision which would have to address the effectiveness and costs of a regulatory recommen- dation upon the total CFC market. Unfortunately, no matter what set of benchmark controls were selected for comparative regulatory analyses, separate groups might critic! ?,'* the selection as overly or insufficiently restrictive, not technically feasible or not enforceable. We suggest a better approach would be for EPA to provide a series of emission reduction goals, either in total pounds or as percentile reductions from base line (no regulation) projections. Analyses could then be performed to spell out specifically how these goals could be attained under the different regulatory options or option combinations under consideration, in what time frame and at what cost. In addition, we recommend that policy options be evaluated opposite their potential impact on environmental risk, not upon pounds of emissions. To facilitate such an evaluation, we urge EPA to adopt ranking of compounds by their "ozone depletion potential" (See Appendix G) . Having clone that, then Rand's analyses and a.ll ensuing analyses must be done on the basis of relative risk, not straightforward pounds of emissions. d. Broaden Analyses Barring such a revised methodology, an effort needs to be undertaken to substantially expand the regulatory effects analyses started by Rand. i. Time Frame - Limiting the comparative analyses to those options which would achieve significant emissions reduction by 1990 is extremely short-sighted. The limitation biases any regulatory selection made on the basis of these short-term VII-57 ------- Economic Considerations solutions. We suspect that a bias to short-term solutions will occur at the expense of steps which could show greater emissions reductions at a lower cost/pound over the long-term. The potential pro- blem is a cumulative long-term (30-100 year) problem. We fail to see any justification for restricting potential solutions to the problem to only those which would show "immediate" results. ii. Technical Assessments - Rand identifies a number of areas requiring further emissions or technical data collection and assessment. Due to lack of technical data on the potential for emission reduction and the use of alternatives, a number of promising options are eliminated from the set of benchmark mandatory controls. We agree that all of the necessary assessments are not yet available but this does not seem to be a valid reason for discounting the potential for so many promising technical options. i ii. Design Rand notes: "The study identified compensation techniques that can substantially mitigate the transfers of wealth under an incentives policy. Such techniques promise to be difficult to design and implement." [Rand, 1980, p. vi.] . "Designing a compensation scheme that does not distort the policy's incentives is not a simple matter operationally." [Rand, 1980, p. 253]. Clearly an expanded analysis is needed on option design implications, especially in the area of the economic and political implications of uncompensated versus compensated transfer payments - both the costs and benefits. ' VII-58 ------- Economic Considerations Other design questions needing work are: Regulatory control point - production, use or emissions? The potential changes in market structure under a marketable permit system. Impact of uncertainty on choice of optimal design. iv. Option Implementation and Administration Rand notes: "For economic incentives policies, the distributive consequences depend critically on how the policy is implemented." [Rand, 1980, p. 229]. "The magnitude of transfer payments depends on how an economic incentives policy is implemented." [Rand, 1980, p. 230]. "An incentives policy might seriously disrupt the CFC-using industries, depending on the magnitude of transfer payments; compensated economic incentives could mitigate transfer, but may be quite difficult to implement." [Rand, 1980, p. 255]. In addition, further attention needs to be given to the following aspects of incentives options: The public sector expense associated with the development, implementation and enforcement of incentives options. Enforcement issues, especially if incentives options become complex through inclusion of exemptions. VII-59 ------- Economic Considerations How to establish regulatory predictability. Thought needs to be given to establishing a readily understood decision formula for control adjustment (e.g., level of taxes) and how to communicate regulatory intent in advance (e.g., the future permit level, where, how and when permits would be available, etc.), so that industry may make plans for emission reduction. v. Legal Issues - Work is needed in a number of areas. These are discussed in detail in Section III. e. Expand and Add Detail to Economic Incentives Option Structures In Rand's treatment of economic incentives regulatory options, the designs and implementations suggested are too general to allow detailed impact analyses and position-taking by industry. One can discuss the pluses and minuses of the concepts but has difficulty in arriving at a defensible endorsement of the options because important structural detail is lacking. One could accept the principle but find the formulation and applica- tion to be totally unsatisfactory. Consequently, there remains a great need for EPA or its contractors to develop a detailed explanation of exactly what these options would look like, how they would be implemented and administered, and how payments would be handled. Having surveyed the potential problems and knowledge deficiencies, solutions now must be proposed. We note that this comment was first made to EPA by us in March, 1980 [Du Pont, 1930c]. Yet, the October, 1980 ANPR shows no indication that any of this needed work has yet been per- formed. VII-60 ------- Economic Considerations F. MISCELLANEOUS ANPR POINTS HAVING ECONOMIC IMPLICATIONS Comments below relate to specific topics, points, arguments, etc., contained in the ANPR where economic implica- tions are pertinent. Those already covered in other sections will only be highlighted here. 1. Choosing a Regulatory Strategy EPA's thought process in approaching CFC regulation, as reflected in the ANPR, appears to have been to: (1) choose a CFC emisssions target based principally on its potential impact in persuading the rest of the world to take action, and then, (2) set about to determine how best to achieve this emissions target considering the trade-offs (including economic) involved. Since evolving public policy increasingly is placing as much emphasis on "cost" as "benefit" in the cost/benefit regulatory equation, it would seem EPA's process of choosing an initial CFC emissions target should have included economic considerations at a much earlier stage. Accordingly, it would be appropriate for EPA to rethink its overall regulatory strategy based on this approach. 2. Cost/Benefit Analysis Recent Administrations, the Congress and the regulatory agencies themselves have strongly embraced the need for careful cost/benefit analyses of proposed regulations to ensure that such regulations will be in the broadest public interest. For any regulatory proposal, this requires that the decision process address two separate but related questions: Does the proposed regulation achieve the stated regulatory objective at the lowest cost? VII-S1 ------- Economic Considerations In structuring the regulatory objective (specifi- cally, in determining the degree of control), are expected benefits and costs equated at the margin? (In other words, does regulation stop where the incremental cost begins to exceed the incremental benefit?) Let us adddress these questions in the context of the proposal to further regulate CFCs. Least Costly Regulatory Approach EPA's stated regulatory objective is to hold CFC emissions from U.S. sources to current levels. EPA further states it prefers the CFC production cap route to accomplish this objective. But is this the least costly regulatory option? In our view, it is clear, on the basis of information and analyses now available, that there is no justification to conclude that a production cap is the most cost effective means of further regulating CFCs. The ANPR does not state a conclusion on this point, nor does it even offer a comparison of the relative costs of achieving the given regulatory objective in alternate ways. Equating Marginal Benefits and Marginal Costs According to EPA, the benefit of the proposed regulatory step is not CFC emissions reductions per se but rather the potential to catalyze international action leading toward CFC emissions reduction. So, the real benefift of the proposed production cap would be a function of how much better a catalyst it is than some alternate action VII-62 ------- Economic Considerations or set of actions, with the latter not necessarily being regulatory actions. But, what about the cost of this "catalyst"? Our analysis above attempted to deal with this question and, although the results cannot be precisely quantified at this stage, clearly the cost of a CFC production cap would be very high in economic terms. In this light, it is impossible to reconcile EPA's proposed CFC production cap with conventional cost/benefit methodology. The "benefit" is highly subjective and theoretical, if it exists at all; but, the cost is extremely high. In view of the foregoing, we feel there is strong justification to insist that EPA perform much more substantive and rigorous cost/benefit analyses before proceeding with further CFC regulation. 3. EPA's Long-Term Regulatory Intent EPA states that the only acceptable long-term strategy for CFCs is "substantial emissions reductions." However, no consideration has been given to the economic impact of this extreme control option. EPA relies heavily on the Rand Report but its most extreme analysis was an option in which emissions were held constant, notJ^substantially reduced from current levels. 4. Product/End Use Bans EPA places bans low on the priority list of regulatory options, in part because of the potential for overcontrol of some uses and the undercontrol of others. However, if EPA followed the analytical approach it outlines to deal with product/end use ban questions, the likelihood of overcontrolling or undercon- VII-63 ------- Economic Considerations trolling a particular end use through selective bans would be greatly reduced. Moreover, under the production cap proposal, a reverse kind of situation is just as likely to eventuate truly essential end uses, with no substitutes, might suffer severe economic hardship while less essential end uses might not. Determining factors would be the financial strength of the ,-tffected firms and the availability of suitable substitutes, but not the question of product essentiality or equity. 5. Economic Incentives or Disincentives? The economic options presented in the ANPR are not incentive options. True incentive approaches would include items such as tax credits to encourage investment in CFC recycle equipment. 6. Tax or Surcharge on CFC Use The Rand Report devoted considerable space to taxes as an economic control option. A tax may avoid some of the uncertainty associated with an allocation or auction system. EPA may have prematurely seized on a permit system as the preferred economic approach. Given the magnitude of what is at stake, we believe all possible options should be explored. 7. Base Year Choice of the base year for a proposed production cap should recognize evolutions in the marketplace, swings in the business cycle and .other structural factors. Using a formula approach which combines several representative years might accomplish this. One concept we feel EPA should explore is that of choosing a base year which is out into the future, say 1985. VII-64 ------- Economic Considerations The production ceiling would be based on an estimate, absent regulations, of how large the market would be at that time if current demand trends continue. Since CFC use is only growing moderately in the U.S., incremental contribution to world CFC emissions would be small in relation to total emissions. The benefit of this approach would be that it would reduce the uncertainty facing business and allow firms time to adjust to CFC use restrictions at a more measured pace, thus reducing potential economic and societal impacts. 8. Term of Permits Rand notes: "The authorization interval and the mix of maturity dates for outstanding permits should be chosen according to two basic principles. First, the authorization interval should be long enough to allow firms to buy and sell permits as needed to insure that demand and supply are equalized. Second, the interval should be long enough and the mix of maturity dates should overlap enough so that there are not major swings in the permit price from one issue to the next because of short-term fluctuations in demand." [Rand, 1980, p. 240]. Yet, EPA has not addressed at all in the ANPR the issue of the term for permits. We believe that orderly business functioning would require that CFC permits, however obtained granted or auc- tioned (to end users or producers) have a long life ideally at least ten years so that investment decisions can be handled on a rational basis. Otherwise, uncertainty would breed chaos and, ultimately, significant economic dislocation. However, if the duration of permits exceeds more than a couple of years, legal implications appear to begin to override. Moreover, from the economic viewpoint, inefficiencies due to lack of competition could develop. VII-65 ------- Economic Considerations 9. Direct Allocation of Permits to Manufacturers This approach may be undesirable from an economic viewpoint because some producers may not need to actively compete with each other for business, and therefore, may have little incentive to operate efficiently. This is so because the price mechanism would not function in its typical manner to equilibrate supply and demand. Moreover, this would be the case whether the "transfer payments" flow to producers or the government. Shortages could develop under this approach, and price could not be relied upon to remove them. What happens to the users who are shorted? Will the government then mandate, in addition to the production cap, formal allocation schemes (especially for those uses deemed to be "essential") in an attempt to treat everyone equitably? How would such a program be administered without a significant cost and without becoming itself a major source of inequity? (For further discussion, see Section Ill-Legal Issues). 10. Direct Allocation of Permits to Users This approach would be unworkable in practice 'because of the large number of firms involved, the many and varied CFC end users and the dynamic technological and market situation which operates here. Witness the tremendous administrative problems, inequities, etc., with previous gasoline allocation schemes which, in fact, are probably simpler to design and implement because of the standardized product and single end use characteristic. As in the case of permit allocation to producers, this approach could create economic inefficiences at the producer VII-66 ------- Economic Considerations level. This is so because under a system of allocation of per- mits to users, some producers may not have an incentive to be efficient; and others, reacting to the potential for extreme year to year variation in production demand on them by the users, could conclude that the associated uncertainties were suffi- ciently great to undermine the incentive to remain in the CFC production business. 11. Government Auction of Permits a. The auction concept is nothing more than a rather complicated way to impose a tax on CFC use. As an alternative, EPA should explore more actively the more direct approach of an excise tax on CFC use, which might be more equitable and simplier to implement and administer. b. As in any free market for a "paper" commodity, speculators would become a major factor. Would EPA permit this? If not, how would auction participation be monitored? c. Future success/failure for users of CFCs will tend to become highly dependent upon how they fare in this auction process, regardless of the relative essentiality of their CFC uses or the strength of their underlying competitive positions. What steps would EPA plan to take to avoid the inequities that might result, or to mitigate against the impact of such inequities? d. How will EPA deal with those situations, which inevitably will come up, where essential end users of CFCs don't get enough to cover their needs and this causes economic or other hardships either to themselves or consumers? A mechanism to handle this would need to be developed. VII-67 ------- Economic Considerations e. The open auction process would inevitably create artificially high prices since there would be a significantly higher perceived demand than actual demand if producers, end users and others all participate. What steps would EPA take to avoid this? f. As time passes, the auction process would create a significant flow of dollars to the government with very significant potential implications from an economic growth and inflation viewpoint. How would these payments be distributed? Who should have jurisdiction in this process due to the potential impact on the economy (Commerce), levels of employment (Labor), financial markets (Treasury), etc.? Rand found the issue of transfer payments to be of paramount importance: "Ultimately, the resolution of the implementation issues raised by transfer payments may be one of the most critical policy choices required by CFC destruction of the ozone layer." [Rand, 1980, p. 239] We concur. VII-68 ------- Economic Considerations G. SUMMARY After examination of EPA's proposed regulatory options, with particular attention to the Agency's stated preference for economic incentives options, we conclude: 1. EPA drastically understates the potential adverse economic consequences of its production cap proposal, particularly the consequences of the large induced uncertainty under such a system. 2. EPA's comments on the potential mechanics of a production cap/production or use permit system demonstrate a lack of appreciation of the complexities involved; as such, EPA grossly underestimates the practical complications of operating such a system. 3. The Rand Report [Rand, 1980] upon which EPA relies heavily to justify its preference, is not adequate from the economic viewpoint as a basis for a regulatory decision. The study, while a useful empirical exercise, was not designed nor performed as an economic impact study of the consequences of economic incentives options. 4. In reaching its preference for a production cap, EPA has not considered in sufficient depth command and control options, true economic incentives approaches such as tax credits for CFC recycle investments, or use of excise taxes as a demand dampening technique. 5. A great deal more in-depth analysis in several areas is required before a defensible conclusion can be reached that economic incentives control options can be accomodated without severe adverse economic consequences. Moreover, the burden of V1I-5S ------- Economic Considerations proof should be on EPA since the Agency's proposals appear to depart from conventional economic wisdom. We find it particularly disturbing that EPA seems already to have made up its mind, without benefit of in-depth analyses, that a CFC production cap is the appropriate regulatory method. In view of the potentially severe economic consequences, EPA's position is not tenable. This becomes quite apparent when we consider that EPA has barely scratched the surface in assessing alternative options. 6. It is possible that EPA advances a production cap as its regulatory preference for reasons other than the potential to meet regulatory goals in the most cost-effective manner. In particular, from a political viewpoint, the cap proposal seems to offer EPA several advantages: Unfocused opposition - the "you don't need to worry; the cap will affect somebody else" argument; A quick regulatory solution, without the necessity to go through the rigorous analyses demanded by command and control; and Operation of the "new toy" theory, which tends to attract supporters to the proposed regulation for all the wrong reasons. 7. In all discussion of the need for CFC regulation, and the pluses and minuses of the regulatory options, (whether economic incentives or command and control), it must be remembered that CFCs have a huge quality- of- life, economic and employment significance to the United States: VII-70 ------- Economic Considerations a. CFCs perform a wide range of jobs considered essential to today's way of life. b. For many current uses there are no safe suitably performiny alternatives available nor are there likely to be in the foreseeable future. c. About $500 million of CFCs are sold annually. d. More than 730,000 jobs are related to CFC use. e. There are approximately 260,000 domestic business locations which use CFCs. f. The annual value of goods and services which depend to some extent upon CFCs exceeds $28 billion. g. The value of installed products which use CFCs is more than $135 billion. Regulatory decisions on CFCs will effect major indus- tries, many workers and the consumer through the wide array of CFC-dependent products. Should regulation ultimately prove to be necessary, great care will have to be taken to insure that the regulatory option and the degree of control selected will be the most cost-effective overall to society. VII-71 ------- VIII. THE SEARCH FOR ALTERNATIVES A. INTRODUCTION B. CRITERIA FOR ALTERNATIVES C. SCOPE OF PROGRAM D. PROGRAM STATUS AND PLANS E. TIMETABLE F. SUMMARY Paqe VIII-1 ------- Search for Alternatives A. INTRODUCTION From early in the controversy of whether continued use of CFCs will lead to a depletion of stratospheric ozone, efforts have been underway to develop acceptable replacement compounds should their use be required. This effort in no way signals an acceptance of the validity of the theory but rather just a prudent and responsible decision to be prepared should the need arise. At Du Pont, this effort to date has cost over $15 million. Unfortunately, all promising compounds identified so far have one or more limitations, such as possible toxicity, no known commercially viable manufacturing process to make them, or, as yet unspecified criteria for environmental acceptability opposite the potential for ozone depletion. Consequently, we feel that if fully satisfactory fluorocarbon alternatives are available at all, they are a minimum of seven but more likely ten years away. Periodically, the status of Du Font's program has been formally reported to EPA [Du Pont 1978; 1980d] and numerous informal updates also have been made. Nevertheless, the Agency routinely has issued statements of optimism on the results of our work [e.g., DeKany, 1980] far in excess of what we believe the facts can support. Consequently, we are including a section in the submission on our alternatives program, its objectives, parameters, status and future. More detailed discussion appears in Appendix B. B. CRITERIA FOR ALTERNATIVES The ozone depletion theory predicts that certain chlorine-containing, volatile compounds are sufficiently stable VIII-2 ------- Search for ATter natiVes (or long-lived) in the lower atmosphere that significant amounts survive to reach the stratosphere, where the chlorine is released by ultraviolet radiation. It is theorized this chlorine may deplete ozone. The amount of predicted ozone depletion is calculated by computer "models" designed to simulate the atmosphere. The potential for this theoretical depletion of ozone can be reduced or avoided in two ways. If a CFC molecule can be made marginally less stable, most will not survive the journey to the stratosphere. However, it must not be so unstable that it contributes to smog, as does, for example, trichloroethylene. Hydrogen-containing fl desirable middle ground. Hydrogen-containing fluorocarbons have a stability in this Alternatively, fluorocarbons which do not contain chlorine could be substituted for CFCs. Fluorine is not involved in the ozone depletion mechanism. Acceptable CFC substitutes also must meet other criteria. An acceptable alternative must provide product performance, low toxicity and safety-in-use. Cost must be compatible with value-in-use, and an economic incentive for manufacture must exist. Lastly, a commercial process for its manufacture must exist or be developed. For purposes of this discussion, compounds containing at least one chlorine and one fluorine will be referred to as chloro- fluorocarbons (CFCs) , whereas compounds containing no chlorine will be referred to as fluorocarbons (FCs). VIII-3 ------- Search for Alternatives C. SCOPE OF PROGRAM As outlined in the previous section the most promising candidates were identified early on as either fluorocarbons containing no chlorine or as chlorofluorocarbons containing hydrogen. This led to an examination of all practical fluorocarbon and chlorofluorocarbon compounds meeting one or the other of these criteria. Only fourteen compounds were found to comprise this category: CFC-21 (CHC12F) CFC-22 (CHC1F2) CFC-31 (CH2C1F) FC-32 (CH2F2) CFC-123 (C2HC12F3) CFC-124 (C2HC1F4) FC-125 (C2HFV CFC-132b (C2H2C12F2) CFC-133a (C2H2C1F3) FC-134a (C2H2F4) CFC-141b (C2H3C12F) CFC-142b (C2H3C1F2) FC-143a (C2H3F3) FC-152a (C2H4F2) These compounds have been or are being evaluated for product performance (as refrigerants, foam blowing agents and solvents), safety (flammability and toxicology) and manufacturing capability. Only CFC-22, FC-134a, CFC-141b, CFC-142b and FC-152a have survived all the tests performed to date. However, the results of all long-term toxicology studies, which would be VIII-4 ------- Search for Alternatives necessary before more broad use of these compounds would be permitted, are not yet available. Only CFC-22 is a major commercial product. CFC-142b and FC-152a are manufactured in very limited quantities. The reasons for discontinuing work on the other candidates are given in Section VIII - F. D. PROGRAM STATUS AND PLANS Du Font's research and development effort on alter- natives is continuing. The results of long-term toxicity studies on CFC-22, FC-152a and CFC-142b are expected over the next two years. Field tests of CFC-22 and CFC-142b in automotive air-conditioning are underway. Limited testing of 'FC-134a in refrigeration and air- conditioning equipment is ongoing. Basic data on equipment modifications necessary in refrigeration and air-conditioning equipment is under development. Methods to determine the long-term insulating performance of alternatives in rigid polyurethane foam are being developed. Although no processes yet exist for the commercial production of FC-134a or CFC-141b, process research is continuing. E. TIMETABLE Seven to ten years may be necessary to reach commercial production for most alternatives, assuming all technical and toxicological programs yield favorable results. Even with existing production processes for CFC-22, CFC-142b and FC-152a, new or expanded facilities would be needed for increased production and for raw materials. The ten-year estimate includes pilot plant construction and operation, long-term chronic VIII-5 ------- Search for Alternatives toxicity testing, development of basic design data, acquisition of a plant site, obtaining production equipment and environmental permits, and plant construction and start-up. An examination of the regulatory protocol and EPA's priority for this issue suggests that regulations for nonpro- pellant uses of CFCs could be promulgated in final form as soon as 1981, and perhaps become effective a year later. The gap between the regulatory timetable and the most optimistic .time- table for the development of alternatives is a major source of concern. VIII-6 ------- F. Surcnary - Chlorof luorocarbon Alternatives Number Formula Boiling Point, °F 48 -41 16 -61 82 12 r55 116 45 -16 CFC-21 CFC-22 CFC-31 FC-32 CFC-123 CFC-124 FC-125 CFC-132b CHC12F CHC1P2 CH2C1F CH2F2 (a) CHC12CF3 CHC1FCF3 CHF2CF3 (a) CH2C1CC1F2 CFC-133a FC-134a CFC-141b CFC-142b FC-143a FC-152a CH2C1CF3 CH2FCF3 (a) CH3CC1F2 CH3CF3 ta) CH3CHF2 90 14 -54 (a) -13 Potential Application Flammable Slewing Agent No Refrigerant No Propellant Yes None (b) Yes Blowing agent, refrigerant No Refrigerant, other No Refrigerant No Cleaning agent - too No aggressive (b) Blowing agent, propellant No Refrigerant, other No Blowing agent Yes Blowing agent, refrigerant Yes Refrigerant Yes Propellant, refrigerant Yes Manufacturing Process Yes Yes NC NC No ft>) No (b) No (b) No NC(US) No Yes (d) Yes NC Yes Toxicology Toxic (b) Weak Mutagen (c) Toxic (b) Low Low Low Not Known Very Incomplete Fjribryotoxic (b) Very Inconplete (testing in progress) Weak Mutagen Weak Mutagen (c) Incomplete Low (c) (a) Contains no chlorine. (b) Work discontinued principally for this reason. (c) Long-term toxicity test in progress. (d) Developmental process only. NC-Not Commercial. NC(US)-Not Commercial in United States, e f rt- sl w ------- IX. CONCLUSIONS AND RECOMMENDATIONS IX-1 Page A. CONCLUSIONS B. RECOMMENDATIONS 13 ------- Conclusions and Recommendations A. CONCLUSIONS1 The discussions and conclusions in the preceding Sections provide ample support for adoption of the "Assessment and Surveillance" regulatory option, as outlined in the Introduc- tion (Section I). Assessment and Surveillance is the only option that assures that the theory will be more thoroughly researched without unreasonable risk developing, while at the same time avoiding the inefficiencies and severe economic impact of unilateral over-regulation. The availability of ozone trend analysis provides an early warning system, and insures that prompt action can be taken if it becomes apparent that a problem is developing. To date, ozone trend analysis has detected ru> depletion of the ozone layer. Specifically, we note that: There are substantial uncertainties surrounding the 1979 ozone depletion calculations. Advances in the science, made since those calculations, reduce the calculated depletion by half or more. In the next few years, it is expected that substantial progress will be made toward resolving the remaining key uncertainties. The research is already underway. Importantly, it can be conducted under the umbrella of ozone trend analysis. Given the availability of ozone trend analysis (a system to survey trends in ozone concentration in the The summary comments which follow are brief because each pre- ceding Section (II-VIII) contains a full summary at the end of the Section. In addition, the Executive Summary provides a high-spot summary of all key points discussed in the full submission. IX-2 ------- Conclusions and Recommendations stratosphere), coupled with periodic scientific reassessment, the risk is minimal in deferring regulation until more accurate and complete scientific information is available. This combination assures EPA that significant potential changes in the ozone can be detected in time to take appropriate action, if it becomes apparent that a problem is developing. Ozone depletion, if it occurs, is truly an international problem. Unilateral action by the U.S. will have no appreciable impact on the environment but will have severe economic impact on U.S. industry and the economy. In light of the failure of the U.S. aerosol ban to stimulate foreign action, it is difficult to see how unilateral non-aerosol regulation will meet with any greater success. Rather, a policy of Assessment and Surveillance is more likely to provide the scientific background necessary to achieve the needed international consensus, and regulatory program, should regulation be determined to be required. In short, we do not believe that the initiation of any further regulation of CFCs at this time is justified by the current body of scientific information. We believe further that were such regulations enacted, it would not accomplish the EPA's stated goals, but it would create severe and unfair burdens for U.S. industry and the U.S. economy. That industry and the regulators of that industry disagree is not unusual. When the disagreement is over the facts of an issue, it can be beneficial to proper decision-making. However, our disagreement with EPA over the CFC/Ozone Depletion Issue has a much broader basis than dispute over the IX-3 ------- Conclusions and Recommendations interpretation or significance of shared facts. In this case, we believe very strongly that the Agency's programs and process have been and continue to be inadequate in the areas of: Adherence to the scientific method Information gathering Interpretation of the data Balancing of risks versus benefits Balancing of risks versus risks Providing for needed research, and Bringing all elements of the issue together to enable reasoned decision-making. Some specific examples of these inadequacies, taken from the history of the CFC/0-, Depletion Issue, and from the ANPR, follow: 1. The Science In the scientific method, scientists collect and then analyze data, postulate a theory to explain the data (generally adding several assumptions to do this) and then, importantly, put the theory to test by performing experiments to prove or disprove key segments of the theory, particularly putting to test the assumptions. Where possible, measurements in the real world are also made to see whether these results fit the matrix of the theory and its predictions. Generally, the theory has to be remodeled several times before reaching a stage which can consistently account for all experimental data and measurements. IX-4 ------- Conclusions and Recommendations At this point, one can say the theory has been proved, at least according to current knowledge. Scientists try not to be "for" or "against" a theory but rather to challenge the theory until it can be demonstrated to their satisfaction that all observations can be accounted for by the theory. Unfortunately, the original CFC/CU Depletion Theory, a good theory based on the available data at the time of its advancement, has not always received this classic scientific treatment i.e., being objectively put to test, questioned, probed, dissected, reassembled, etc., by the qualified scientific community. Perhaps due to the almost "science fictional" flavor of the theory's predictions (skin cancer, crop failure, fish kills, climate changes, etc., due to the release to the atmosphere of odorless, colorless and seemingly benign chemicals used throughout society with great benefit), the scientific process for this issue became distorted or at least partially subjugated to outside influences. What should have been science spilled over into the media and political arenas. Participants in the issue often chose sides (others were forced onto sides) -- one was either "for" or "against" the theory, with those in the "for" camp casting themselves as for the protection of mankind, and casting those in the "against" camp as against human health, etc., for the sake of short-term profit. As a consequence, many chose not to play at all [Margulis, 1980]. Unfortunately, many of those in this latter category were those who were most needed. Others, e.g., Allaby and Lovelock [1980], came to feel that the treatment of this issue became so politicalized that they could not obtain objective hearings. And still others, most of whom were scientists employed by industry, have had to fight an ongoing effort to get involved in the arena at all because once one was classified as "against IX-5 ------- Conclusions and Recommendations the theory" in this case due to place of employment -- presumably all scientific objectivity was lost even though those "for the theory" presumably retained this objectivity. The search should have been and should be for the facts for objective analysis. In contrast, consider the following developments: ** a) The industry research program (CMA/FPP) has invited EPA scientists to its periodic reviews of the science [CMA, 1980c], yet EPA has organized forums, e.g., SRI [1980], in which the organizers excluded industry scientists. b) The U.S. delegations to international conferences to discuss the issue have not included industry scientists or their academic consultants (This is in contrast to the European delegations). In addition, U.S. delegations have consisted almost exclusively of individuals both in and out of government who publicly have stated their conclusions that the theory is valid or proved. Further, U.S. industry scientists have been excluded from report writing workshops of these various international reviews, for example, UNEP [1979] and OECD [1980]. c) EPA's public pronouncements on this issue press releases, letters and information sheets [Jellinek, 1980a; 1980c; EPA, 1980a; 1980c; 1980f; 1980g] have overstated conclusions reached by available studies, ignored conflicting studies and not acknowledged new developments in the science. d) In the face of highly uncertain and conflicting information, EPA has prematurely, and for all practical purposes, conclusively announced that the theory has been sufficiently proved that further regulation is called for, and as justi- fication relies almost exclusively on worst case risk scenarios [Jellinek, 1980c; EPA, 1980a; 1980c; 1980f; 1980g; 1980h]. IX-6 ------- Conclusions and Recommendations e) And last, the Agency has argued that regulators do not have the luxury of following good scientific practices -- that regulators must respond promptly to any theory that suggests present or future harm to the public or the environment. [Jellinek, 1980b]. In short, we submit that EPA has not followed good scientific practices on this issue. By exaggerating conclusions from available reports, by selective use of available informa- tion, by treating uncertain findings as fact, by not giving equal credence to dissenting opinion (even on occasion discouraging the airing of dissenting opinion) and by focusing on and publicizing worst case risk scenarios, the Agency has turned what should have been, and should still be, a scientific issue into a media and political issue. The specifics of the above examples are developed in the body of the submission. The following statements from Dr. Philip Handler, past president of the National Academy of Sciences seem pertinent: "A primary obligation of scientists is to communicate their understandings and the limits of their under- standings. Since these are all probabilistic, it becomes very difficult for the media to deal with them. These past 15 years, those areas of science and technology that seemed to carry some elements of risk have been overemphasized while the magnitude of specific risks has frequently been dealt with somewhat cavalierly." He went on to say: "Scientists have called attention to the carbon dioxide problem, to the effect of freon (sic) on the ozone layer, to the consequences of ionizing radiation, etc. Once scientists have reported what they know, discussion has frequently been taken over by individuals who claim to represent "public science" or "critical science." They seize upon a problem and adopt the philosophy of the Delaney Amendment to the Food and Drug Act, which is that only zero risk is tolerable, with little concern IX-7 ------- Conclusions and Recommendations for the benefits that might be lost. By and large the media have shared the values of these critics and much that appears in the news is offered from that point of view. Once this happens, public discussion becomes polarized, and those who try to state the other side of the case all too easily come to be seen as reckless of the public health." [Handler, 1980]. The ANPR The subject ANPR does nothing to change our views or allay our fears of the Agency's process. a) The presentation on the science fails to mention critiques of the science justification relied upon by EPA, [e.g., Du Pont 1980a; 1980b], fails to mention conflicting reports, [e.g., UK DOE, 1979; EEC, 1980], and ignores information available to EPA, [e.g., Brasseur, 1980; CMA, 1980a] , on recent changes in the science changes which substantially reduce the calculated "predictions" of ozone depletion upon which EPA justifies its need to regulate now. Perhaps even more dis- turbing, the results from analyses of actual ozone measurements provided to the Agency [CMA, 1980a; 1980c] results which indicate depletion is not occurring as predicted are not even mentioned. b) The question of risk continues to receive a black and white' treatment regulate now or wait until its too late to head off major damage -- while not mentioning the middle ground of waiting for several additonal years, under a close monitoring of the situation, in order to develop the science needed to resolve the underlying key uncertainties. Neither did the ANPR make mention of the findings of its own consultant [SRI, 1980] which issued a report based on a workshop of participants invited by EPA in which it was concluded that almost all the key uncertainties could be resolved within roughly 5 years, and that it would be "cost-effective" to do so. IX-8 ------- Conclusions and Recommendations c) Submissions [Masten, 1980; Du Pont, 1980c; 1980e; Block, 1980] have been made to EPA pointing out errors in EPA analyses, for example, permit pound calculations, growth rates, and erroneous inclusion of non-emitting CFCs under the intended regulatory scope. And EPA acknowledged overestimates had been made [Muir, 1980]. Yet, all these errors persist in the ANPR. d) The proposals for unilateral regulation presented in the ANPR make little sense from an effectiveness standpoint. EPA presents these regulations as being necessary, among other reasons, to "stimulate coordinated worldwide [regulatory] cooperation" [EPA, 1980e], yet the ANPR offers no analysis or support for any finding that such a response would be forthcoming. e) The EPA's presentation in the ANPR of economic incentives regulatory options for comment is not adequate in content to support the apparent decision to employ these options. The Agency bases its presentation on work done for it by the Rand Corporation [Rand, 1980] . Industry was asked to critique the final draft version of this study. Du Pont did so [Du Pont, 1980c] , pointing out that the study was a good start but should not be used as a document on which to base a decison because: 1) the data base was obsolete, 2) the scope of the study was artificially constrained in a manner not consistent with real world production and use of CFCs, and 3) the study was not an economic impact study. We also pointed out that we were unable to provide in-depth critique of the economic incentives regulatory options analyzed by Rand because only the concepts were presented no "How to implement?", or "What would they look like?", or "How would they actually work?" This was in March, 1980. Yet, the ANPR presents the same unfleshed-out proposals for comment. Our previous comments and questions remain unanswered and unacknowledged but now we asked to comment again on the same exact concepts. Even more disturbing is the IX-9 ------- Conclusions and Recommendations fact that EPA apparently has used the Rand study to conclude that these regulatory options are preferrable [EPA, 1980e; Jellinek 1980a; 1980c] , without having performed the necessary work (as pointed out in our March, 1980 submission) to allow such a determination. A related point is that industry generally opened its books to EPA's contractor, the Rand Corporation, with the understanding that the final data base, on which the regulatory options would be evaluated by Rand and by EPA, would be made available to industry to enable parallel analysis [Mader, 1978], Yet, to date, this data base has not been released. f) Du Pont has met with EPA [Du Pont, 1980f] and provided submissions [Du Pont, 1978; 1980d] on our efforts to develop alternative fluorocarbon products. We have made it very clear that great difficulties have been encountered. Yet, EPA has informed audiences [DeKany, 1980] that great progress was being made and EPA was optimistic that Du Pont was near a breakthrough. Further, the ANPR makes no mention of the difficulties in, or low probability of, developing safe alternative compounds. g) Du Pont has outlined the risks which could be expected to develop from use of currently available products (which would have to be used were EPA to be successful in restricting the availability of CFCs) [Du Pont, 1978; 1980c], yet no mention of these risks appears in the ANPR. h) EPA was informed of the possible energy penalty from CFC regulation [Battelle, 1980]. Yet no mention of energy consequences appears in the ANPR. i) And last, any hope we had been harboring that our response points could make some difference in the Agency's IX-10 ------- Conclusions and Recommendations program was dashed after viewing the rulemaking timetable in EPA's Chlorofluorocarbon Phase II Development Plan [EPA, 1980, p. 3] a timetable showing: A due date (Nov., 1980) for the initial draft of the proposed rule bef o re comments on the ANPR are received (due by January 5, 1981). . A due date (Jan., 1981) for the final draft of the proposed rule before ANPR comments could be digested. A plan to publish a formal proposed rule (March, 1981) 3 months before EPA will receive a final report from its contractor, just hired in Oct., 1980, to study in more detail various critical aspects of the the economic incentives regulatory options now being favored by EPA for use. Whatever vested interest industry may have in a given decision-making situation, there are major mutual benefits to joint open cooperation with the regulating agency. Without such cooperation at each step in the decision-making process, vital inputs are lost which could minimize the economic cost of regulation, identify adverse consequences of regulatory scenarios and expand the regulatory options under consideration. We reiterate for the record our fervent belief that regulatory decision-making which: fails to measure the validity of the theory against real world measurements, fails to consider the need for, or the consequences of, intermediate delay versus immediate action, IX-11 ------- Conclusions and Recommendations fails to assess fully the consequences of regulatory action, and fails to balance the risk generated by regulation against the risk from no regulation, is seriously flawed, and a luxury which the nation, its busi- nesses, and its citizens cannot afford. We conclude, therefore, that both legal and policy considerations mandate that EPA defer CFC regulation until more accurate and complete scientific information is obtained and other necessary studies are performed. We earnestly request that EPA commit to a solution of the uncertainties in the ozone depletion theory and work towards an international consensus before reaching a decision to engage in further unilateral regulation of CFCs. Specific recommendations for action by EPA follow in the next section. IX-12 ------- Conclusions and Recommendations B. RECOMMENDATIONS The following are Du Font's recommendations for actions to be taken, or at least initiated, by the EPA, which we believe will lead to a proper resolution of the CFC/ozone Controversy. Should this resolution dictate the need for further regulation, such actions as outlined herein will help ensure a balanced and cost-effective regulation. EPA should promptly arrange for an updated assessment of ozone trend analysis by a qualified outside body, such as the NAS. If an outside review f body cannot be employed, a joint industry/govern- ment/academia symposium should be held to review objectively the method opposite the questions: How sensitive is it? What is the confidence range? How and when can it be further improved? A companion recommendation would be for EPA to support the further development of trend analysis. EPA should arrange for an objective thorough review of the science (both the theory itself and effects of ozone depletion) by an international panel of qualified scientists. A joint NAS/UK Royal Society effort would be a logical starting point. Inter- national political organizations such as the Organi- zation for Economic Cooperation and Development (OECD) are not adequate for this assessment due to the limited participation of scientists and the political pressures present in such groups. Even if an international review cannot be promptly arranged, EPA should recontract with NAS for an updated review of the science, followed by a yearly reassessment. The predicted problem is a long-term IX-13 ------- Conclusions and Recommendations problem. The science is changing rapidly. Any regulatory decision based on the science at any point in time must be reassessed as the scientific justification for that decision changes. Between NAS reports, EPA should meet quarterly with the Chemical Manufacturers Association (CMA) Fluoro- carbon Project panel (FPP), and other appropriate advisors, to stay current with the board spectrum of scientific developments. The Agency needs to publish the parameters of its decision-making on the issue: i. What specific level of ozone depletion does EPA consider to pose an unreasonable risk to health and the environment? ii. What will it take to convince EPA there is or is not a serious problem, e.g., a. What sensitivity of ozone trend analysis is accepted (and on what basis)?; What ozone trend analysis results would be viewed as a significant indication of a developing problem? b. What other science developments would be viewed as significant? c. What criteria does EPA use to judge the credibility of sources and reported developments, and which sources meet these criteria? IX-14 ------- Conclusions and Recommendations d. What will EPA do to ensure staying abreast of developments? e. What is the process EPA uses to get developments assessed and to the attention of the regulatory decision-makers? f. What must happen internationally to convince EPA of the need or lack of need for further U.S. regulation? By whom? In what time period? iii. How does the specific proposed regulation result in reduction of risk on this issue and what is the magnitude of this reduction? If in the periodic reviews of the science it is determined the risk has decreased signifi- cantly, what are the parameters of regulatory response? The Agency needs to redefine the problem of ozone depletion generically and then determine and justify whether CFCs should be treated in isolation from other potential depleting compounds and in isolation from potential ozone increasing compounds. The charge to EPA of the 1977 Clean Air Act Amendments is protection of stratospheric ozone, not the regu- lation of CFCs. What is the justification for including CFC-22 under the regulation when methyl chloroform represents a greater total potential problem? Conversely, modelers now include the CCU/ozone augmentation effect. This needs to be factored into EPA's assessment. IX-15 ------- Conclusions and Recommendations On the international level, EPA should abandon its excessively political strategy in favor of an effort to help obtain the needed global scientific assessment and resolution. The Agency should publish its plans for furthering the scientific resolution of this issue. As pointed out in previous sections, EPA must com- plete a significant body of work before it can support the proposed regulatory options. Further assessment and study are needed in the areas of: i. Emission reduction and alternatives. What is achieveable, in what time-frame and at what cost? ii. Impact of economic incentives options. What would be the actual impact on industry and consumers if the options were applied to all CFCs and all CFC uses as proposed? iii. Energy penalty of regulations. iv. Risk from alternatives substituted for CFCs, and a risk-risk assessment of continued CFC use versus use of alternatives. v. A detailed fleshing-out of the incentives options for comment specifying exactly how they would be structured, how they would be implemented and how they would function. We recommend that EPA hold a series of informational exchange meetings around the country to discuss its IX-16 ------- Conclusions and Recommendations proposals, hear concerns and gather information to help its studies of i.-iv. above. EPA needs to employ a more realistic timetable. The current timetable shows a completion date of January, 1981 for the final draft of the proposed rule. The ANPR comment period closes January 5, 1931. Further, we question how EPA can digest and evaluate ANPR comments and submissions in time to publish a formal proposed rule in March, 1981. The proposed timetable appears unrealistic unless EPA has no interest in the ANPR comments and has already made up its mind on how to proceed. Given the magnitude of the issues which remain to be addressed, particularly on the economic incentives options, we fail to see how a reasonable proposal can be finalized in this period. We ask for clarification. Due to the untried nature of the economic incentive options, the numerous questions and concerns which have yet to be answered, and the total lack of experience with these regulatory options in the real world, if EPA elects to proceed with their use, we would strongly urge that a pilot test first be undertaken. The options should first be applied in- to a carefully monitored industry or industry seg- ment, and the impacts thoroughly evaluated, rather than immediately applying this theoretical regula- tory approach to the CFC industry which has such broad and major impact on the total economy. If EPA decides to promulgate a rule, (regardless of which regulatory option is selected) the Agency should issue an annual report detailing: IX-17 ------- Conclusions and Recommendations i. Results of actual ozone measurements. Has depletion been detected? If so, how much and at what rate? ii. Computer calculated or estimated ozone depletion based on best current information. What are the current model calculations? iii. Status of U.S. versus world regulatory t situation. Has the U.S. regulation achieved the Agency's goals? Have other countries followed EPA's lead or is the U.S. example being ignored? iv. The continued nee'd for the regulation as promulgated. v. The economic impact of the promulgated regulation, particularly if new regulatory concerns are involved. This should be compiled by major market segment and business size, as well as a summary report. vi. The identity of substitutes employed in place of CFCs by use category. For all substitutes (and especially new substitutes) safety data, toxicity data, energy efficiency, development cost of replacements and redesign cost for manufacturers should be monitored for a period of 10-20 years to determine the true cost of regulation for guidance in future rulemaking efforts. IX-18 ------- |