TECHNICAL SUPPORT DOCUMENT POTENTIAL RECYCLING OF SCRAP METAL FROM NUCLEAR FACILITIES PART I: RADIOLOGICAL ASSESSMENT OF EXPOSED INDIVIDUALS Volume 1 Prepared by R. Anigstein, W. C. Thurber, J. J. Mauro, S. F. Marschke, and U. H. Behling S. Cohen & Associates 6858 Old Dominion Drive McLean, Virginia 22101 Under Contract No. 1W-2603-LTNX Prepared for U.S. Environmental Protection Agency Office of Radiation and Indoor Air 401 M Street S.W. Washington, D.C. 20460 Deborah Kopsick John Mackinney Project Officers September, 2001 ------- Note: EPA no longer updates this information, but it may be useful as a resource or reference. Contents page List of Tables vii List of Figures viii List of Appendices ix Executive Summary xi Preface xv 1 Introduction 1-1 1.1 Purpose and Scope 1-1 1.2 Organization of the Report 1-2 Reference 1-4 2 Overview of Scrap Metal Operations 2-1 2.1 Characteristics of Scrap Sources 2-1 2.2 Industry Perspectives 2-3 2.3 Principal Scrap Metal Operations Considered 2-4 2.4 Current Recycle Practice of Nuclear Facilities 2-5 References 2-6 3 Screening Procedures to Define the Scope of the Analysis 3-1 3.1 Objectives 3-2 3.1.1 Characterization of the Potential Sources of Scrap Metal 3-2 3.1.2 Normalized Dose and Lifetime Risk of Cancer to the RME Individual 3-2 3.2 Sources of Scrap Metal: Administrative Categories 3-4 3.2.1 Department of Energy 3-6 3.2.2 Nuclear Regulatory Commission 3-6 3.2.3 Department of Defense 3-7 3.2.4 State or Superfund Authority 3-8 3.3 Types of Scrap Metal Considered 3-8 3.4 Radionuclides Selected for Consideration 3-10 3.5 Exposure Scenarios and Biological Endpoints 3-10 3.5.1 Multiple Pathways 3-11 3.5.2 Personal Devices 3-12 3.5.3 Other Pathways and Scenarios 3-12 3.5.4 Direct Disposal of Scrap Following Clearance 3-12 3.6 Summary of the Screening Process 3-13 3.6.1 Sources of Scrap Metal 3-13 3.6.2 Types of Scrap Metal from Nuclear Facilities 3-13 3.6.3 Scenarios, Pathways, Modeling Assumptions, and Biological Endpoints . . . 3-13 References 3-15 -in- ------- Note: EPA no longer updates this information, but it may be useful as a resource or reference. Contents (continued) page 4 Quantities and Characteristics of Potential Sources of Scrap Metal from DOE Facilities and Commercial Nuclear Power Plants 4-1 4.1 Existing and Future Scrap Metal Quantities Available from DOE 4-1 4.1.1 Background Information 4-1 4.1.2 Existing Scrap Inventories at DOE 4-8 4.1.3 Summary of Existing Scrap Inventories at DOE Sites 4-12 4.1.4 Scrap Metal Inventory by Metal Type 4-13 4.1.5 Scrap Metal from Future Decommissioning 4-15 4.1.6 Summary and Conclusions Regarding DOE Scrap Metal Inventories 4-18 4.2 Scrap Metal from the Commercial Nuclear Power Industry 4-19 4.2.1 Estimates of Contaminated Steel from Commercial Nuclear Power Plants . . 4-21 4.2.2 Contaminated Metal Inventories Other Than Steel 4-22 4.2.3 Timetable for the Availability of Scrap Metal from Decommissioning 4-23 4.3 Recent Recycling Activities (1995 - 1998) 4-22 4.3.1 DOE Materials 4-24 4.3.2 Activities of Members of the Association of Radioactive Metal Recyclers .. 4-27 References 4-29 5 Description of Unrestricted Recycling of Carbon Steel 5-1 5.1 Recycling Scrap Steel—an Overview 5-1 5.2 Reference Facility 5-3 5.3 Exposure Pathways 5-4 5.3.1 External Exposure 5-4 5.3.2 Internal Exposure 5-4 5.4 List of Operations and Exposure Scenarios 5-4 5.4.1 Dilution Factors 5-7 5.4.2 Scrap Transport 5-9 5.4.3 Scrap Processing Operations 5-9 5.4.4 Steel Mill 5-9 5.4.5 Processing EAF Dust 5-12 5.4.6 Use of Steel Mill Products 5-12 References 5-14 6 Radiological Assessment of the Recycling of Carbon Steel 6-1 6.1 Radioactive Contaminants 6-2 6.2 Specific Activities of Various Media 6-5 -IV- ------- Note: EPA no longer updates this information, but it may be useful as a resource or reference. Contents (continued) page 6.3 Exposure Pathways 6-9 6.3.1 External Exposures to Direct Penetrating Radiation 6-9 6.3.2 Inhalation of Contaminated Dust 6-13 6.3.3 Incidental Ingestion 6-17 6.3.4 Radioactive Decay 6-17 6.4 Unique Scenarios 6-19 6.4.1 Ground Water Contaminated by Leachate from Slag Storage Piles 6-19 6.4.2 Ingestion of Food Prepared in Contaminated Cookware 6-35 6.4.3 Impact of Fugitive Airborne Emissions from the Furnace on Nearby Residents 6-35 References 6-39 7 Results and Discussion of Carbon Steel Radiological Assessment 7-1 7.1 Normalized Doses and Risks to the RME Individual 7-1 7.2 Maximum Exposure Scenarios 7-1 7.2.1 Slag Pile Worker 7-2 7.2.2 Scrap Yard Worker 7-2 7.2.3 Lathe Operator 7-4 7.2.4 Sailor Sleeping next to Steel Hull-plate 7-5 7.2.5 Truck Driver: Baghouse Dust 7-5 7.2.6 EAF Furnace Operator 7-5 7.3 Evaluation of the Results of the Radiological Assessment 7-5 7.3.1 Dilution of Potentially Contaminated Steel Scrap 7-6 7.3.2 Exposure Pathways 7-6 7.3.3 Mass Fractions and Partitioning of Contaminants 7-9 7.3.4 Scenario Selection 7-9 7.3.5 Implementation of Clearance Criteria 7-10 References 7-12 8 Radiological Assessment of Recycling Aluminum 8-1 8.1 Distribution of Contaminants 8-1 8.1.1 Material Balance 8-1 8.1.2 Elemental Partitioning 8-2 8.2 List of Operations and Exposure Scenarios 8-4 8.2.1 Dilution 8-4 8.2.2 Scrap Transport 8-6 8.2.3 Secondary Smelter Operations 8-6 8.2.4 Industrial Uses of Mill Products: Aluminum Fabrication 8-8 8.2.5 Use of Finished Products 8-9 8.2.6 Off-Site Individuals Exposed to Smelter By-Products 8-10 -v- ------- Note: EPA no longer updates this information, but it may be useful as a resource or reference. Contents (continued) page 8.3 Results 8-12 8.3.1 Shredder Operator 8-14 8.3.2 Scrap Transport Worker 8-14 8.3.3 Disposal of Dross in an Industrial Landfill 8-15 8.4 Evaluation of the Results 8-15 8.4.1 Dilution of Potentially Contaminated Scrap 8-15 8.4.2 Exposure Pathways 8-15 8.4.3 Airborne Effluent Releases 8-15 8.4.4 Ingrowth of Radioactive Progenies 8-16 References 8-18 9 Radiological Assessment of Recycling Copper 9-1 9.1 Recycling Copper Scrap—an Overview 9-1 9.2 Distribution of Contaminants 9-2 9.2.1 Material Balance 9-2 9.2.2 Elemental Partitioning 9-3 9.3 List of Operations and Exposure Scenarios 9-3 9.3.1 Dilution 9-4 9.3.2 Scrap Transport 9-6 9.3.3 Secondary Smelter Operations 9-7 9.3.4 Electrorefming 9-8 9.3.5 Use of Finished Products 9-9 9.3.6 Impact of Airborne Effluent Emissions on Nearby Residents 9-10 9.4 Results 9-11 9.4.1 Slag Worker 9-11 9.4.2 Airborne Effluent Emissions 9-14 9.4.3 Tank House Operator 9-14 9.5 Evaluation of the Results 9-14 9.5.1 Airborne Effluent Releases 9-14 9.5.2 Other Scenarios 9-14 References 9-17 -VI- ------- Note: EPA no longer updates this information, but it may be useful as a resource or reference. Tables page S-l. Maximum Exposure Scenarios and Normalized Impacts on the RME Individual from One Year of Exposure to Recycling of Carbon Steel, Aluminum, and Copper x 3-1. Inventory of Sites That Are Known to Be Radioactively Contaminated 3-5 4-1. Groupings of DOE Materials in Inventory 4-9 4-2. Existing Scrap Metal Inventories at DOE Sites 4-11 4-3. Estimates of Existing DOE Inventories of Contaminated Scrap Metal 4-13 4-4. DOE Scrap Metal Inventory 4-14 4-5. Existing and Future Contaminated Scrap Metal at DOE Facilities 4-20 4-6. Comparison of Estimates of Ferrous Metal and Nickel Inventories 4-20 4-7. Residually Radioactive Steel from Nuclear Power Plants 4-22 4-8. Contaminated Metal Other than Steel Potentially Suitable for Clearance 4-24 4-9. Anticipated Releases of Scrap Metals from Nuclear Power Plants 4-25 5-1. Exposure Scenarios and Parameters for Radiological Assessments of Individuals 5-6 6-1. Implicit Progenies of Nuclides Selected for Analysis 6-3 6-2. Nuclides Included in Various Combinations and Decay Series 6-5 6-3. Partition Ratios (PR), Concentration Factors (CF), and Distribution Factors (DF) 6-8 6-4. Lung Clearance Types and Ingestion fx Values for Use with ICRP 68 6-16 6-5. Vadose Zone Parameter Values for Site Types A, B, and C 6-21 6-6. Potential Contaminants of Groundwater 6-22 6-7. Composition of Slag Used in Leaching Test 6-23 6-8. Leaching Parameters Values 6-25 6-9. Diffusion Coefficients for EAF Slag Monolithic Samples 6-26 6-10. Fraction of Various Toxic Elements Leached from Slags Using EPA TCLP Protocol 6-27 6-11. Aquifer Parameter Values for Site Types A, B, and C 6-31 6-12. Soil-Water Distribution Coefficients (Kds) for Site Types A, B, and C 6-32 6-13. Locations and Results of CAP-88 Analyses 6-36 6-14. Calculation of Normalized Doses and Risks from Airborne Effluent Emissions 6-38 7-1. Maximum Exposure Scenarios and Normalized Impacts on the RME Individual from One Year of Exposure 7-3 -Vll- ------- Note: EPA no longer updates this information, but it may be useful as a resource or reference. Tables (continued) page 8-1. Partition Ratios (PR) and Concentration Factors (CF) in Aluminum Smelting 8-3 8-2. Exposure Scenarios and Parameters for Radiological Assessments of Aluminum Recycling 8-5 8-3. Normalized Impacts from One Year of Exposure to Fugitive Airborne Emissions ... 8-11 8-4. Maximum Exposure Scenarios and Normalized Impacts on the RME Individual from One Year of Exposure 8-13 9-1. Partition Ratios (PR) and Concentration Factors (CF) 9-5 9-2. Exposure Scenarios and Parameters for Radiological Assessments of Copper Recycling 9-6 9-3. Normalized Impacts from One Year of Exposure to Fugitive Airborne Emissions .... 9-11 9-4. Maximum Exposure Scenarios and Normalized Impacts on the RME Individual from One Year of Exposure 9-12 Figures 4-1. Nuclear Weapons Complex 4-2 5-1. Operations Analyzed in the Carbon Steel Recycle Analysis 5-2 6-1. Transport of Slag Leachate to Domestic Well 6-31 8-1. Simplified Material Flow for Secondary Aluminum Smelter 8-2 9-1. Simplified Mass Flow: Annual Throughput of Secondary Copper Smelter 9-2 9-2. Simplified Material Balance for Electrorefining of Copper Produced from Scrap 9-3 -Vlll- ------- Note: EPA no longer updates this information, but it may be useful as a resource or reference. Volume 2: Appendices A -F A. Scrap Metal Inventories at U. S. Nuclear Power Plants B. Aluminum Recycling C. Copper Recycling D. Selection of Radionuclides for Radiological Assessment E. Distribution of Contaminants During Melting of Carbon Steel F. Distribution of Contaminants During Melting of Cast Iron Volume 3: Appendices G - L G. Dilution of Residually Radioactive Scrap Steel H. Detailed Scenario Descriptions for Carbon Steel I. Leaching of Radionuclides from Slags J. Radiological Impacts on Individuals—by Scenario K. Radiological Impacts on Individuals—by Pathway L. Radiological Impacts of the Disposal of Residually Contaminated Materials in Industrial Landfills -IX- ------- Note: EPA no longer updates this information, but it may be useful as a resource or reference. EXECUTIVE SUMMARY Introduction Large quantities of radioactively contaminated scrap metal are generated during the decommissioning of nuclear facilities and, to a lesser extent, during the normal operation of these facilities. To evaluate the radiological impacts of releasing residually contaminated metals to the environment, the U.S. Environmental Protection Agency (EPA) performed exhaustive analyses of the release and recycling of carbon steel, aluminum, and copper scrap. The aim of the analyses was to calculate the annual dose and the lifetime risk of cancer to the reasonably maximally exposed (RME) individual, normalized to the specific activity of a given radioactive contaminant in the scrap, from one year of exposure. These results, presented as a set of tables that list the normalized doses and risks to the RME individual from each of 44 radionuclides and nuclide combinations that are potential contaminants of the three metals, can be used to assess the potential health effects of releasing scrap with a given level of contamination. Description of Actual Work The first step was constructing a series of exposure scenarios corresponding to the entire life cycle of each metal, comprising the transportation of the scrap; cutting and sorting at a scrap processing or recycling facility; melt-refining at a steel mill, secondary smelter facility, or an integrated copper production facility; fabrication of commercial products; and the use of such products. Also included were exposures to the primary byproducts of the furnace—slag (dross in the case of aluminum) and offgas. In the case of steel and aluminum, most of the offgas, which comprises both volatile and particulate matter, is captured by the emission control system and routed to the baghouse, where the fumes are cooled and filtered. Airborne effluent emissions include uncondensed gases and particulate matter that escape the collection and filtration system. The RME individual is the person who, due to his occupation, location or living habits, would receive the maximum likely exposure from a given radionuclide. To identify this individual, the doses from one year's exposure to each scenario were calculated for all three metals. The person with the highest dose became the RME individual for a given radionuclide. The exposure pathways fall into two general groups: external exposure to direct penetrating radiation and internal exposure from inhaled or ingested radionuclides. The internal exposure -XI- ------- Note: EPA no longer updates this information, but it may be useful as a resource or reference. pathways consist of inhalation of radioactively contaminated dust; incidental ingestion of dust or other loose, finely divided material; and ingestion of contaminated food or water. The 44 individual radionuclides and nuclide combinations studied in this analysis are those most likely to be present in contaminated scrap that may be a candidate for recycling. A literature search as well as thermodynamic calculations were used to develop partition ratios and vaporization fractions of the corresponding elements during the melt-refining of carbon steel, aluminum, and copper. Results Table S-l summarizes the results of the analyses. The maximum normalized doses from one year of exposure span the range of approximately 3 x 10"3 to 700 |lSv/a per Bq/g, reflecting the wide range of chemical and radiological properties of these nuclides. In 29 of the 44 cases, the normalized doses from the maximum exposure scenario for copper scrap are higher than the maximum doses from carbon steel or aluminum. In the majority of cases, the RME individual is a worker directly involved in handling or processing the scrap metal or its refinery byproducts. In several other cases, it is a person who is exposed to finished metal products as a result of his occupation. In three other cases, it is an individual who resides near a recycling or disposal facility and is exposed to airborne effluents or contaminated drinking water. These results allow EPA and other interested parties to evaluate the potential radiological impacts of recycling scrap metals with known levels of residual contamination. Table S-l. Maximum Exposure Scenarios and Normalized Impacts on the RME Individual from One Year of Exposure to Recycling of Carbon Steel, Aluminum, and Copper Nuclide C-14 Mn-54 Fe-55 Co-60 Ni-59 Ni-63 Zn-65 Sr-90+D Maximum Scenario Dross in landfill Lathe operator Slag worker Sailor exposed to hull plate Slag worker Slag worker Truck driver: baghouse dust Slag leachate in groundwater Metal Al steel Cu steel Cu Cu steel steel Dose mrem per pCi/g 3.4e-04 1.06-01 4.1e-05 4.7e-01 9.56-06 2.6e-05 7.16-02 1 .6e-02 jjSv per Bq/g 9.2e-02 2.76+01 1.1e-02 1.36+02 2.6e-03 7.16-03 1.9e+01 4.26+00 Lifetime Risk of Cancer3 per: pCi/g 1.6e-10 7.76-08 1.1e-11 3.56-07 6.4e-12 2.06-11 5.46-08 7.76-09 Bq/g 4.4e-08 2.16-05 2.9e-09 9.56-05 1 .7e-09 5.56-09 1 .5e-05 2.16-06 a Maximum risk—may correspond to a different scenario -Xll- ------- Note: EPA no longer updates this information, but it may be useful as a resource or reference. Table S-l (continued) Nuclide Nb-94 Mo-93 Tc-99 Ru-106+D Ag-110m+D Sb-125+D 1-129 Cs-134 Cs-137+D Ce-144+D Pm-147 Eu-152 Pb-210+D Ra-226+D Ra-228+D Ac-227+D Th-228+D Th-229+D Th-230 Th-232 Pa-231 U-234 U-235+D U-238+D Np-237+D Pu-238 Pu-239 Pu-240 Pu-241+D Pu-242 Am-241 Cm-244 U-Natural U-Separated U-Depleted Th-Series Maximum Scenario Slag pile worker Slag worker Slag worker Lathe operator Lathe operator Sailor on naval support vessel Airborne effluent emissions Truck driver: baghouse dust Truck driver: baghouse dust Slag pile worker Slag worker Slag pile worker EAF furnace operator Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Metal steel Cu Cu steel steel steel steel steel steel steel Cu steel steel Cu Cu Cu Cu Cu Cu Cu Cu Cu Cu Cu Cu Cu Cu Cu Cu Cu Cu Cu Cu Cu Cu Cu Dose mrem per pCi/g 2.3e-01 3.3e-04 1 .8e-04 2.66-02 3.2e-01 6.26-02 3.3e-01 1.86-01 6.6e-02 8.36-03 1 .6e-04 1.76-01 5.6e-01 S.Oe-01 2.4e-01 2.56+00 1.4e+00 2.36+00 3.8e-01 6.66-01 9.8e-01 2.46-01 2.4e-01 2.16-01 6.3e-01 4.36-01 4.3e-01 4.36-01 4.6e-03 4.06-01 1.1e+00 7.26-01 1.6e+00 4.76-01 2.4e-01 2.36+00 jjSv per Bq/g 6.36+01 8.8e-02 5.06-02 7.0e+00 8.56+01 1.7e+01 8.96+01 5.0e+01 1.86+01 2.3e+00 4.26-02 4.6e+01 1.56+02 8.2e+01 6.56+01 6.8e+02 3.76+02 6.2e+02 1.06+02 1.8e+02 2.76+02 6.6e+01 6.46+01 5.7e+01 1.76+02 1 .2e+02 1 .2e+02 1 .2e+02 1.26+00 1.1e+02 3.16+02 1.9e+02 4.46+02 1.3e+02 6.46+01 6.1e+02 Lifetime Risk of Cancer3 per: pCi/g 1 .8e-07 3.36-11 5.9e-11 2.06-08 2.4e-07 4.76-08 1 .5e-07 1 .4e-07 5.0e-08 6.56-09 9.4e-11 1 .3e-07 1 .6e-07 2.16-07 1 .2e-07 1 .2e-07 8.7e-07 4.76-07 4.4e-08 8.26-08 5.3e-08 1.16-07 1.1e-07 9.86-08 2.9e-07 7.46-08 6.8e-08 6.86-08 4.1e-10 6.56-08 3.0e-07 1 .9e-07 5.2e-07 2.16-07 1.1e-07 1 .Oe-06 Bq/g 4.86-05 8.8e-09 1 .6e-08 5.3e-06 6.56-05 1 .3e-05 4.06-05 3.8e-05 1 .4e-05 1 .8e-06 2.56-08 3.5e-05 4.36-05 5.8e-05 3.16-05 3.4e-05 2.36-04 1 .3e-04 1 .2e-05 2.2e-05 1 .4e-05 2.9e-05 3.16-05 2.7e-05 7.86-05 2.0e-05 1 .8e-05 1 .8e-05 1.16-07 1 .7e-05 8.26-05 5.2e-05 1 .4e-04 5.7e-05 S.Oe-05 2.8e-04 Maximum risk—may correspond to a different scenario -Xlll- ------- Note: EPA no longer updates this information, but it may be useful as a resource or reference. PREFACE In March, 1997, S. Cohen and Asociates, under contract to the Office of Radiation and Indoor Air of the U.S. Environmental Protection Agency (EPA), produced a draft report entitled "Technical Support Document: Evaluation of the Potential for Recycling of Scrap Metals from Nuclear Facilities".1 The purpose of that report was to evaluate the potential public health impacts associated with the free release and recycling of scrap metal from nuclear facilities as an alternative to disposal at a licensed low level radioactive waste disposal facility. The report was also intended to be part of the technical basis for determining the need for regulatory action to ensure that recycle of scrap metal from nuclear facilities does not endanger public health and safety. The report was widely distributed by EPA to the U.S. Department of Energy, the U.S. Nuclear Regulatory Commission, representatives of U.S. metal recycling and steel manufacturing industries, the International Atomic Energy Agency, the European Commission, and other stakeholder groups for review. Several meetings were held with these organization to exchange information and receive comments on the Agency's draft report. In addition, a Task Group appointed by the National Council on Radiation Protection and Measurement performed a critical review of the Draft TSD. The Draft TSD has been revised to address many of the questions and concerns raised during the review process and to incorporate a great deal of new information acquired since that report was issued. The present report, which constitutes Part I of the revised TSD, contains an expanded and revised assessment of the potential impacts of the free release of scrap metal from nuclear facilities on exposed individuals. This document was reprinted in July, 1997 with a revised cover page. The text was unchanged. -XV- ------- Chapter 1 INTRODUCTION 1.1 PURPOSE AND SCOPE The cleanup of sites in the U.S. that are contaminated with radioactive material and the decommissioning of nuclear facilities are expected to generate large amounts of scrap metal. In fact, some sites controlled by the U.S. Department of Energy (DOE) have already accumulated significant inventories of scrap metal that are currently in storage awaiting final disposition. The Office of Radiation and Indoor Air of the U.S. Environmental Protection Agency is evaluating a broad range of technical and regulatory issues associated with the disposition of scrap metal from nuclear facilities. The Agency is examining alternatives to disposing of the scrap at a licensed low level radioactive waste disposal facility or otherwise maintaining the material under regulatory control. This report is not a Regulatory Impact Analysis nor an Environmental Impacts Statement supporting an Agency rulemaking. It is intended solely as part of a technical information document for use by the Agency as part of the basis for decision-making with respect to the free release of metal from nuclear facilities. A separate document, "Radiation Protection Standards for Scrap Metal: Preliminary Cost-Benefit Analysis" (IEC 1997), describes a preliminary analysis of the potential costs and benefits of recycling scrap metal from nuclear facilities. The purpose of the present analysis is to assess the radiological impacts of the free release of scrap metal from nuclear facilities on reasonably maximally exposed (RME) individuals. (Throughout this report, the terms "residually radioactive scrap metal," "residually contaminated scrap metal," "scrap metal from nuclear facilities," or simply "scrap metal" are used to refer to any metal that has the potential for free release. "Free release," in turn, refers to the clearance of the material from the regulatory control of DOE, the U.S. Nuclear Regulatory Commission (NRC) or the Department of Defense.) These radiological impacts are stated as doses or risks from one year of exposure, normalized to unit specific activities1 (i.e., 1 pCi/g or 1 Bq/g in scrap) Throughout this report, the terms "radionuclide concentration" and "specific activity" may appear to be used interchangeably. Strictly speaking, concentration refers to a given physical quantity, such as mass, per unit volume or unit mass of the matrix. The concentration of uranium in soil, for example, might be expressed in micrograms of uranium per gram of material. Specific activity is always expressed in units of activity per unit mass, such as pCi/g or Bq/g. For a given radionuclide, of course, the specific activity is proportional to its concentration. Since radionuclides are usually 1-1 ------- of each separate radionuclide or combination of nuclides that is a potential contaminant of scrap metal. The relationship between the concentration of a radionuclide in scrap metal and the potential radiological impacts on RME individuals are intended to help EPA establish clearance levels for the free release of scrap metal from nuclear facilities that are in the Agency's acceptable risk range, should the Agency decide to issue rules or guidance which establishes such levels. The present report does not address the issues related to the implementation of any such rules. Thus, it would be incorrect to predict the numerical values of any future clearance criteria solely on the basis of the present analysis. The analysis addresses metal that is suspected to be lightly or moderately contaminated as a result of radioactive deposition or neutron activation. Scrap metal that has never been exposed to possible radioactive contamination is not considered in the evaluation. Conversely, metal that is so heavily contaminated that it can only be disposed of as a radioactive waste is also excluded from this evaluation. 1.2 ORGANIZATION OF THE REPORT The report comprises three volumes. The first volume consists of nine chapters. Chapter 2 provides an overview of scrap metal operations in the United States and the characteristics of scrap metal from nuclear facilities. Chapter 3 describes the screening procedures used to define the scope of the analyses and discusses the limitations of these analyses. Chapter 4 describes the principal sources of scrap metal, which include the DOE complex and the commercial nuclear power industry. Chapters 5 through 9 present a detailed analysis of the doses and risks to RME individuals from 44 radionuclides or nuclide combinations that may be present in three metals: carbon steel, aluminum and copper. Chapter 5 describes the exposure scenarios used to assess the potential radiological impacts of the free release and recycle of carbon steel scrap on individual members of the public. (The term "members of the public" includes all individuals except radiation workers, whose exposures are governed by existing NRC and DOE regulations.) Chapter 6 presents the methodology and models used to perform these radiological assessments. Chapter 7 discusses the key results of detected and assayed in terms of their activities, not in terms of their masses, specific activity is a more useful concept 1-2 ------- the assessments of carbon steel scrap. Chapter 8 describes the exposure scenarios and methods used to assess the radiological impacts of the free release and recycle of aluminum scrap and discusses key results of these assessments. Chapter 9 presents a similar discussion of the free release and recycle of copper scrap. Volume 2 consists of six appendices. Appendix A presents a discussion of the scrap metal that would be generated by the decomissioning of commercial nuclear power plants in the United States This appendix includes a detailed analysis of carbon and stainless steel that would be available for potential release—the metals that constitute well over 90 percent of the metal inventory used to construct a nuclear power plant—as well as discussions of nine other metals. Appendix B presents a detailed discussion of aluminum recycling, beginning with an analysis of the potential availability of aluminum scrap from nuclear facilities, continuing with an overview of aluminum recycling, and concluding with a series of possible scenarios for assessing the radiological impacts on potentially exposed individuals. Appendix C presents a similar discussion of copper scrap. Appendix D presents a review of published reports, data bases, and computer codes which formed the basis for selecting the radionuclides addressed in the analysis. Appendix E presents the empirical and scientific basis for determining how various trace elements and their compounds are redistributed among the various phases during the melt- refining of carbon steel. Appendix F presents a similar discussion related to the production of cast iron. Volume 3 contains six more appendices. Appendix G discusses the geographical and temporal distribution of anticipated future releases of carbon and stainless steel from commercial nuclear power plants, and the possible sites for the melt-refining of these materials and the processing of the baghouse dust, a byproduct of melt-refining. Appendix H presents a detailed discussion of the data sources and parameters used to construct the exposure scenarios described in Chapter 5. Appendix I discusses the empirical data on the leaching of contaminants from steel slags and presents a model for estimating the leach rate of trace elements. Appendices J and K present the detailed results of the radiological assessments of individuals potentially exposed to each radionuclide by exposure scenario and pathway. Appendix L presents an analysis of the radiological impacts on the RME individual if steel scrap that is free-released from a nuclear facility were buried in a RCRA Subtitle D solid waste landfill, and an assessment of the burial of aluminum dross in a similar landfill. 1-3 ------- REFERENCE Industrial Economics, Inc. (IEC). 1997. "Radiation Protection Standards for Scrap Metal: Preliminary Cost-Benefit Analysis." Prepared for U.S. Environmental Protection Agency, Office of Radiation and Indoor Air, Washington, DC. 1-4 ------- Chapter 2 OVERVIEW OF SCRAP METAL OPERATIONS This chapter provides an overview of the types and quantities of scrap metal potentially available for clearance, the operations of the secondary metal industry in the United States, and current practices for recycling scrap metal from nuclear facilities. It provides a summary description of the world of scrap metal generation and utilization; any scrap metal released from nuclear facilities for recycling becomes a part of this process. It also offers a perspective on the data, modeling, parameter characterization, and associated technical issues considered in the present analysis. This chapter also presents updated summary estimates of the availability of scrap metal from nuclear facilities. 2.1 CHARACTERISTICS OF SCRAP METAL SOURCES Scrap metal released from nuclear facilities could be supplied to the secondary metal industry in the United States. The DOE complex and decommissioned nuclear power plants will generate scrap metals of many types. These include carbon steel, stainless steel, galvanized iron, copper, Inconel, lead, bronze, aluminum, brass, nickel, and precious metals, such as gold and silver. The impacts of recycling scrap metal from nuclear facilities will depend on the quantities and timing of the releases and on the characteristics of the secondary metal industry for the particular metals of concern. To support the analysis, assessments were performed of the available quantities of scrap metals from the DOE complex and from the commercial nuclear power plants, the radioactive contaminants, and the time when the metals would be released. The future releases of scrap metal from nuclear facilities are highly uncertain. In terms of the DOE complex, uncertainty exists as to when and how decommissioning of these facilities and equipment might occur, and when and if any residually radioactive metals will be released for unrestricted use. Uncertainty also exists about the shutdown schedule of commercial nuclear power reactors and the associated decommissioning activities. Operating licenses issued by the NRC are generally valid for 40 years. In 1996, the NRC issued a rule allowing a licensee to apply for a 20-year renewal of its original operating license. To date, license renewals have been issued for two sites, with a total of five reactors (U.S. NRC 2000). A number of other renewal applications are pending, and more such applications are anticipated (U.S. NRC 2001). 2-1 ------- Furthermore, some plants might delay releasing scrap metal for as long as 60 years after shutdown. On the other hand, some plants are being shut down and decommissioned before their present operating licenses expire. The DOE currently has a moratorium on the free release of volumetrically contaminated metals and has suspended the unrestricted release for recycling of scrap metal from radiological areas within DOE facilities. Given these uncertainties, it is possible that the generation of scrap metal from existing DOE and commercial nuclear facilities will span the present century. A comprehensive discussion of iron and steel scrap which would be generated by the decommissioning of commercial nuclear power reactors can be found in Appendix A of the present report; a summary is presented in Section 4.2. Numerous snapshot assessments of scrap metal inventories at DOE sites have been performed over the last decade. These assessments are discussed in Section 4.1. A quantitative assessment of the availability of aluminum scrap from nuclear facilities is presented in Appendix B. A similar discussion of copper recycling can be found in Appendix C. A principal source of information on contaminated metals at DOE facilities is "A Report of The Materials in Inventory Initiative" (U.S. DOE 1996). This report presents a snapshot of scrap metal inventories taken in the summer of 1995. The inventories were characterized as clean (no radioactive contamination), contaminated (known radioactive contamination), and "unspecified" (potentially contaminated). These inventories are subject to large and rapid changes as a result of ongoing operations such as the sale of clean scrap, disposal of radioactive wastes, and batch production of scrap from decommissioning of structures. The report does not include projections of scrap generation and availability since plans for future cleanup and decommissioning activities are highly uncertain. The characterization of scrap metal inventories at DOE sites is further complicated by the use of different definitions of "scrap" at various DOE sites. For example, inventory estimates reported in U.S. DOE 1996 include approximately 1.1 x 10s metric tons (t)1 of carbon steel scrap associated with the uranium enrichment facilities at the Y-12, K-25, Paducah, and Portsmouth Throughout this report, metal stockpiles, capacities of metal recycling facilities, and other parameters characterizing the nuclear or metal refining industries will generally be cited in metric tons (tonnes) or, if English units were cited in the source documents, in short tons. The word "ton" will always mean short ton (1 ton = 0.9072 tonne). When practicable, the metric equivalent will also be listed. In describing scientific analyses, however, the International System of Units (SI) will be employed. 2-2 ------- facilities. Other DOE reports note that these same sites contain on the order of 6.4 x 1051 of carbon steel not yet declared scrap and therefore not included in current scrap metal inventories. These reports do not indicate when this metal might be declared scrap and made available for recycle, although some schedules for decommissioning have been proposed. Recent developments in DOE policy concerning re-industrialization suggest that facilities at these sites might be converted to new uses, in which case potential scrap metal generation would be limited. 2.2 INDUSTRY PERSPECTIVES To gain perspective on the significance of recycling scrap metal from nuclear facilities to the secondary metal industry, the annual industry throughput of iron and steel scrap is compared with estimates of the potential inventories of metals available for recycling from DOE sites and commercial reactors. According to data compiled by the U.S. Geological Survey, a total of 6 x io71 of carbon steel scrap were consumed in the United States in 1996 (Fenton 1997). Table A-81 in the present report indicates that the total amount of carbon steel that could be generated from the decommissioning of all commercial nuclear reactors is about 3.5 x IO61. An estimated 4.7 x 10s t of this total would be residually radioactive carbon steel scrap that could become available for release and recycle over a 50-year period. Table 2-9 of U.S. DOE 1996 indicates that the total inventory of contaminated and "unspecified" carbon steel at DOE sites amounted to 1.05 x IO51, while the HAZWRAP report (Parsons 1995) provides supplemental information which brings the estimate of the DOE inventory to about 1.5 x io51 (see Table 4-3 of the present report). As noted above, the enrichment facilities contain another 6.4 x 10s t of carbon steel which may be declared to be scrap. The inventory of potentially available carbon steel scrap from other DOE sites is small in comparison to the enrichment facilities. In summary, the upper bound of carbon steel scrap generated by DOE facilities and commercial nuclear power plants and available for free release and recycle is on the order of a million metric tons. This scrap would be released to the secondary metal industry over a period of decades. Therefore, the total quantity of carbon steel scrap from nuclear facilities entering the recycling stream is small in comparison to the annual consumption of 6 x IO71. Quantities of other types of scrap metal potentially available from nuclear facilities, such as aluminum and copper, are also small in comparison with the annual consumption. 2-3 ------- 2.3 PRINCIPAL SCRAP METAL OPERATIONS CONSIDERED There are over 200 electric arc furnaces (EAFs) used for iron and steel production in the United States; theoretically, any of these could receive and process iron and steel scrap released from nuclear facilities. However, in practice, steel mills maintain close relationships with nearby scrap metal dealers in order to minimize transportation costs. In turn, the scrap dealers receive their materials from nearby sources, again to minimize transportation costs. Careful segregation of the scrap by specific alloy and physical form is highly desirable to the scrap metal consumer to facilitate optimum melting practices. It is therefore quite likely that nuclear facilities, such as DOE sites and decommissioned commercial nuclear power plants, will send their scrap metal to nearby scrap dealers who in turn will respond to orders from the mills they serve. As shown in Appendix G of the present report, each of the DOE sites that could be a significant source of scrap metal is located near steel mills with electric arc furnaces and, by implication, near scrap processors who could receive and process the recycled scrap. Steel mills with EAFs are also located in the vicinities of most nuclear power plants. Because of the variations in furnace capacities, the characteristics of the working relationships between scrap dealers and mills, and the make-up of individual charges to a furnace, the dilution of residually radioactive scrap metal with uncontaminated metal is highly variable. For example, it is theoretically possible, though highly unlikely, that a single furnace charge could be made up entirely of contaminated scrap. It is far more likely that each charge to a furnace would be a mixture of contaminated and "clean" scrap. Furthermore, most of the scrap metal generated at a nuclear facility will never have been exposed to radioactive contamination. This clean metal would thus serve to dilute the residually contaminated scrap, which would presumably be released at the same time. The levels of residual radioactive materials in the intermediate or finished products will also depend on partitioning that occurs during furnace operations. As a result of chemical phase equilibria, radioactive and other species will be distributed among the metal-melt, slag, and vapor phases associated with melting operations. Partitioning during melting of carbon steel is discussed in Appendix E; partitioning during melting of cast iron is discussed in Appendix F. Partitioning during the smelting of aluminum and copper scrap is discussed in Appendices B and C, respectively. 2-4 ------- 2.4 CURRENT RECYCLE PRACTICE OF NUCLEAR FACILITIES Recycle of scrap metal from nuclear facilities is currently practiced on a limited and directed basis. For example, specialty metals, with low levels of contamination, that are generated at one DOE site are given new use at another site and scrap is converted into containers used for radioactive waste disposal. Despite these initiatives, the general rule of thumb for management of scrap for both the DOE complex and facilities licensed by the NRC is to dispose of residually radioactive scrap metal in a licensed, low-level waste disposal facility. 2-5 ------- REFERENCES Fenton, M. 1997. "Iron and Steel Scrap." (17 June 2001) Parsons Engineering Services, Inc., RMI Environmental Services, and U.S. Steel Facilities Redeployment Group. 1995. "U.S. Department of Energy, Scrap Metal Inventory Report for the Office of Technology Development, Office of Environmental Management," DOE/HWP- 167. Prepared for Hazardous Waste Remedial Actions Program, Environmental Management and Enrichment Facilities, Oak Ridge, TN. U.S. Department of Energy (U.S. DOE), Office of Environmental Management. 1996. "Taking Stock: A Look at the Opportunities and Challenges Posed by Inventories from the Cold War Era," DOE/EM-0275. Vol. 1, "A Report of The Materials in Inventory Initiative." U.S. Nuclear Regulatory Commission (U.S. NRC). 2000. "Information Digest, 2000 Edition," NUREG-1350. Vol. 12. U.S. NRC, Washington, DC. U.S. Nuclear Regulatory Commission (U.S. NRC). 2001. "NRC License Renewal." (3 July 2001) 2-6 ------- Chapter 3 SCREENING PROCEDURES TO DEFINE THE SCOPE OF THE ANALYSIS Evaluating the potential for the free release of scrap metal from nuclear facilities gives rise to a number of questions. What kinds and amounts of scrap metal will be available? What residual radioactive contamination can be anticipated? How will the scrap be handled? Which individuals would be exposed to the scrap and the products and byproducts of recycling? In order to answer these questions, it was necessary to define the objectives of the analysis and to develop a method of screening potential issues and data to determine those which were most significant to the analysis. The objectives of the analysis included characterizing the different types of existing and future sources of scrap metal from nuclear facilities and the levels of residual radioactive contamination in or on the cleared materials. The analysis also defined the relationship between the specific activity of the various radionuclides in the scrap metal and the radiological impacts on individuals from its free release. This chapter presents the screening methods that were used to define the parameters used in the analysis, including the sources of scrap metal, the types of metal that would be available, the radionuclides of concern, the exposure scenarios and pathways, and the types of potentially adverse biological effects that may result. This chapter is divided into six main sections. Section 3.1 provides background information on the specific areas of inquiry and analysis contained in the present report. Section 3.2 describes the existing and potential sources of scrap metal from nuclear facilities and the methods used to select the sources used in the analysis. Section 3.3 describes the different types of scrap metal potentially available for recycle and the methods used to select the specific metals addressed in the present study. Section 3.4 describes the methods used to select the radionuclides of concern to the analysis. Section 3.5 presents the screening methods used to select the specific exposure scenarios, pathways, and biological endpoints addressed in this report. Finally, Section 3.6 summarizes the results of the screening process. 3-1 ------- 3.1 OBJECTIVES 3.1.1 Characterization of the Potential Sources of Scrap Metal The construction of realistic exposure scenarios requires data on the future availability of scrap metal, the expected levels of radioactive contamination, and the temporal and geographic distribution of anticipated releases. 3.1.2 Normalized Dose and Lifetime Risk of Cancer to the RME Individual A key aim of the analysis is the calculation of the normalized dose and the lifetime risk of cancer to the reasonably maximally exposed (RME) individual from one year of exposure to each radionuclide that is a potential contaminant of the cleared materials. The term "normalized dose" refers to the high-end effective (committed) dose equivalent (EDE) that may be received by an individual in any one year, expressed in mrem/y EDE per pCi/g (or |lSv/a per Bq/g) of a specific radionuclide in scrap metal. Similarly, the normalized risk is the lifetime risk of cancer (excluding non-fatal skin cancers) resulting from one year of exposure. The concept of the RME was adopted from the EPA Superfund program. U.S. EPA 1989 states that: The reasonable maximum exposure [RME] is defined here as the highest exposure that is reasonably expected to occur at a site.... The intent of the RME is to estimate a conservative exposure case (i.e., well above the average) that is still within the range of possible exposures. Additional guidance was provided by Habicht (1992): Information about individual exposure and risk is important to communicating the results of a risk assessment. Individual risk descriptors are intended to address questions dealing with risks borne by individuals within a population. These questions can take the form of: • Who are the people at the highest risk? • What risk levels are they subjected to? • What are they doing, where do they live, etc., that might be putting them at higher risk? • What is the average risk for individuals in the population? The high-end of the risk distribution is, conceptually, above the 90th percentile of the actual (either measured or estimated ) distribution. The conceptual range is not meant to precisely 3-2 ------- define the limits of this descriptor, but should be used by the assessor as a target range for characterizing "high-end risk." As used in the present report, the RME is that individual, within the group of people that have the greatest potential for exposure to residually contaminated scrap metal, who would receive the high-end exposure. This group of people, which can be referred to as the critical or limiting population group for a given radionuclide, have job responsibilities or living habits that result in an elevated potential for exposures as compared to other groups. Within the group, there is variability among the members with regard to their individual potential for exposure. The RME is that individual within the group that has a relatively high potential (e.g., 90th percentile) for exposure. Therefore, it is unlikely that many individuals within or outside the group could receive exposures significantly greater than those of the RME individual; most individuals that may be exposed are likely to receive exposures that are substantially lower. The concept of the RME individual is similar to that of the critical group1. The critical group concept was first introduced by the International Commission on Radiation Protection (ICRP) in order to account for the variation in dose in a given population which may arise due to differences in age, size, metabolism, living habits and environment. The critical group is defined by the ICRP as a relatively homogeneous group of people whose location and lifestyle are such that they represent those individuals expected to receive the highest doses as a result of radioactive releases (ICRP 1977, ICRP 1985). As part of the critical group definition, the ICRP specifies the following additional criteria: • Size. The critical group should be small in number and typically include a few to a few tens of people. • Homogeneity among members of the critical group. There should be a relatively small difference between those receiving the highest and the lowest doses. It is recommended that the range between the low and high doses not differ by more than a factor often or a factor of about three on either side of the critical group average. • Modeling assumptions. In modeling the exposures of the critical group, the ICRP recommends that dose estimates be based on cautious but reasonable assumptions. The National Research Council (1995) recommended using a critical group in developing a standard; the U.S. NRC employs this concept in its rulemaking. 3-3 ------- The principal difference between the concept of the RME individual and the average member of the critical group is that the RME individual is conceptually the 90th percentile of the group that itself has a potential for high exposure, while the average member of the critical group receives the mean exposure of such a group. In addition, the critical group could refer to a specific age group, while the RME individual is an adult: the objective of the assessment is to limit the lifetime risk to an individual and not the dose or risk from a given year of exposure. In practice, the differences between the two concepts are not significant. 3.2 SOURCES OF SCRAP METAL: ADMINISTRATIVE CATEGORIES The potential sources of scrap metal can be categorized by function (e.g., reactors, research laboratories) and by the administrative authority responsible for their management and disposition (e.g., DOE, NRC). The process used to screen potential sources of scrap metal was to: (1) review available data within each category of administrative authority, and (2) assess the degree to which the various functional categories were represented. This approach was found to be the most practical method for acquiring scrap metal data because the needed information was more readily accessible by administrative authority. The principal administrative authorities responsible for controlling the release of scrap metal from nuclear facilities are: • Department of Energy • Nuclear Regulatory Commission • Department of Defense (DOD)—many DOD facilities are licensed by NRC • State or Superfund Authority Table 3-1 presents an overview of the various administrative categories of sites containing or contaminated with radioactive materials. A review of available data characterizing contaminated structures within these administrative categories is provided in the Addendum to the Technical Support Document for soil cleanup levels (U.S. EPA 1996, p. E4-7). There are a total of about 30,000 structures in the major jurisdictional sectors, of which about 8,000 are contaminated. These structures, along with scrap metals already in storage at many DOE sites, represent the major potential sources of scrap metal at nuclear facilities in the United States. 3-4 ------- Table 3-1. Inventory of Sites That Are Known to Be Radioactively Contaminated AGENCY Federal DOE DOD SITE TYPE Major DOE Facilities FUSRAPb UMTRAPC Other DOE Sites Major DOD Facilities Sites with Burial Areas Sites with Accident Contamination Sites with DUe Contamination Other DOD Sites Other Federal Sites NRC/Agreement State Licensees Non-federal NPL Sites Nuclear Power Plants Test and Research Reactors Other Fuel Cycle Facilities Rare Earth Extraction Facilities Byproduct Material Facilities Municipal Landfills Radium Sites Other Sites Other State Sites NUMBER 13a 21 2 27 ld 85 1 15 57 2f 121s 37 65 22 4401h 3 7 11 no reliable data Note: Data is for purposes of illustration and is subject to change as facilities are shut down or converted to other uses, and new facilities are opened. Sources: SCA 1997 and others a See Section 4.1.1 Formerly Utilized Sites Remedial Action Program Uranium Mill Tailings Remedial Action Program d Aberdeen Proving Ground—licensed by NRC e Depleted uranium f Watertown Arsenal (GSA), Fremont National Forest (USDA) g Includes 104 currently licensed reactors and 17 formerly licensed reactors in SAFSTOR or scheduled for DECON (See Section A.5.2.2) h NRC licensees only 3-5 ------- 3.2.1 Department of Energy DOE is responsible for cleaning up more than 130 contaminated facilities in over 30 states and territories (U.S. DOE 1995c, p. iii). These include approximately 45 national laboratories and former nuclear weapons production and testing facilities, where environmental restoration and waste management activities are taking place. Many of these are large sites with facilities that have been used for multiple activities related to nuclear weapons research, production and testing over the years and have many areas of contamination. DOE's Environmental Restoration and Waste Management (EM) program is responsible for characterizing, decontaminating and decommissioning these facilities, and restoring the environment at these sites. Information on the status of these programs is provided in many DOE core documents (U.S. DOE 1995a, 1995b, 1995c). In addition, DOE's "Materials in Inventory Report" (U.S. DOE 1996) has estimated the current and projected inventory of potential scrap metal at many of its facilities. These data are discussed in Section 4.2 of the present report. Based on this understanding of the potential quantities of DOE scrap metal, the DOE sites and facilities listed in Table 4-1 are included in the scope of the present analysis. 3.2.2 Nuclear Regulatory Commission The NRC and its Agreement States have licensed about 22,000 facilities for the production and handling of radioactive materials (U.S. EPA 1993). About one-third of these are NRC licensees, while the remainder are licensed by Agreement States under Section 274 of the Atomic Energy Act. Licensees include universities, medical institutions, radioactive source manufacturers, and companies that use radioisotopes for industrial purposes. About 50% of NRC's 7,500 licensees use either sealed radioactive sources or only small amounts of short-lived radionuclides. Activities at these facilities are not likely to result in residual radioactive contamination that will need to be cleaned up and disposed of because: (1) the radionuclides remain encased and cause little (if any) contamination and/or (2) because the radionuclides rapidly decay to non-radioactive nuclides. A small number of licensees (e.g., radioactive source manufacturers, radiopharmaceutical producers, and radioactive ore processors) conduct operations that could result in substantial radioactive contamination in portions of the facility. In addition, about 250 facilities associated with the nuclear fuel cycle2 maintain large inventories of radioactive These include nuclear power plants, non-power (research and test) reactors, fuel fabrication plants, uranium hexafluoride production plants, uranium mill facilities, and independent spent fuel storage installations. 3-6 ------- materials; many of these facilities will need to be cleaned up before their licenses can be terminated. The only sources of scrap metal in this administrative category that are explicitly addressed in the current assessment are commercial nuclear power plants. The other potential sources are not significant, due to the relatively small volume of scrap metal generated and/or the short half-lives of the radionuclides involved. 3.2.3 Department of Defense DOD's Installation Restoration Program (IRP) encompasses over 17,500 potential hazardous waste sites located at 1,877 installations (Baca 1992). DOD sites vary widely in function and size. They include hospitals, laboratories, proving grounds, bombing and gunnery practice ranges, missile launch sites, weapons manufacturing and storage facilities, and nuclear reactors. Only a few of these are currently known to have radioactive contamination; however, these sites have not been fully characterized. Consequently, it is not possible to reliably estimate the number of radioactively contaminated sites. DOD sites may contain small enclosed radiation sources, such as radium and tritium instruments. They may also contain larger sources, such as research reactors, and dispersed sources, such as laboratory waste storage areas and test ranges. Due to the relatively limited potential for scrap metal and the limited availability of data characterizing the scrap metal, these potential DOD sources of scrap metal are not addressed in the present analysis. Naval Nuclear Propulsion Program The U.S. Navy maintains a fleet of over 80 nuclear-powered ships, primarily aircraft carriers and submarines. Submarines have one nuclear reactor and the carriers typically have two. The nuclear propulsion systems on these ships, including the reactors, steam generators, transport piping, pumps and auxiliary systems, are contaminated with much the same mix of radionuclides as any nuclear power plant. When these ships are decommissioned, the nuclear fuel is removed and shipped to INEEL. The reactor compartments, which include most of the nuclear power plant components, are then removed from the ships and transported to the DOE's Hanford Reservation for land burial. The remainder of the dismantled ships may be stored afloat or conceivably sold for scrap. However, there is little or no available data on the residual 3-7 ------- radioactive contamination of this potential source of scrap metal. It was therefore not possible to address it in the present analyses. 3.2.4 State or Superfund Authority This administrative category includes sites that are not licensed by NRC or Agreement States but are under State or Superfund authority. (Sites that are under Superfund authority are those that are on the National Priority List [NPL] and are being cleaned up by the Federal government.) This category includes about 1,000 particle accelerator sites that generally contain only small amounts of short-lived radionuclides after shutdown. Other sites included in this category contain long-lived, naturally-occurring radionuclides which vary in form from small packaged radiation sources to large areas of mostly low-level dispersed contamination, including mining wastes and materials, tailings from ore processing, and residues from academic or commercial research. The scrap metal generated at sites in this administrative category is primarily contaminated with naturally-occurring radioactive material (NORM)—most of this metal is from the oil and gas industry. Metal contaminated with NORM raises issues which are quite different from those posed by metal from nuclear facilities. The quantities of metal are much greater and the contamination is markedly different; furthermore, these materials are not currently under the regulatory control of federal agencies. The disposition of this material represents an economic, regulatory, and administrative framework that is markedly different from that of nuclear facilities. The Agency has therefore determined that issues related to the recycling and reuse of NORM are best addressed separately; this source of scrap metal has thus been excluded from the current assessment. 3.3 TYPES OF SCRAP METAL CONSIDERED Not all scrap metal that fell within the administrative and functional categories discussed above was included in the present study. As stated in Chapter 1, the analysis was limited to contaminated scrap metal that can be potentially decontaminated to meet foreseeable clearance standards. Some metal is so radioactive that it could not be decontaminated by any practical methods and was therefore not included in the present assessment. Examples include the 3-8 ------- canyons at fuel reprocessing facilities and nuclear reactor internals. On the other hand, scrap metal that has never been exposed to possible radioactive contamination is also excluded.3 Tables 4-5, 4-7, and 4-8 list a number of different metals that could potentially be cleared from nuclear facilities. The five most abundant metals—listed in descending order—are: • Carbon steel • Stainless steel • Copper and brass • Nickel • Aluminum These metals and their various alloys comprise about 99% of the total inventories. Carbon steel represents over 80% of the total metal inventory and is therefore the obvious first choice for the assessment. Stainless steel is second in abundance. However, as discussed in Appendix A, the stainless steel scrap generated by the decommissioning of nuclear power reactors, the principal source of this metal, would come from many of the same systems and components as the carbon steel scrap. Hence, the patterns of contamination in the cleared scrap would be similar for the two metals. Likewise, the exposure scenarios that would be used to model the release and recycle of stainless steel would be in many ways similar to those used to assess carbon steel. As discussed in Appendix E, the partitioning of trace contaminants among the various phases during the melt-refining of stainless steel is comparable to that observed in carbon steel. Given the smaller quantities of stainless steel scrap that would be generated, and the high cost of this metal, which places restrictions on its use as compared to the much cheaper carbon steel, it is not likely that the radiological impacts of this metal would be greater than those of carbon steel. A similar argument can be made about nickel, which is an ingredient of most commonly used stainless steels, and which has chemical properties similar to those of iron. Aluminum and copper, however, are markedly different from steel in many respects. Their main sources are the DOE facilities—relatively little copper and very little aluminum would be generated by nuclear power plants. These metals have physicochemical properties that are quite different from those of iron. As a result, the recycling of aluminum and copper scrap uses Such metals are considered when estimating the dilution of residually contaminated cleared materials with "clean" metal in the various commercial processes discussed later in this report. 3-9 ------- processes that are different than those used in the melt-refining of steel. Furthermore, the properties of these metals, as well as their relatively high cost in comparison to steel dictate very different commercial uses. Separate exposure assessments were therefore performed for these two metals. 3.4 RADIONUCLIDES SELECTED FOR CONSIDERATION The following criteria were used to select the radionuclides addressed by the exposure assessments: • Half-life greater than six months. Major nuclear facilities are unlikely to begin clearing scrap metal sooner than about five years after they have ceased to operate. Thus, selecting a six-month half-life as the cutoff would allow for a decay of at least ten half- lives, during which time the activity of the shorter-lived radionuclides would be reduced by a factor of at least one thousand. Although it is theoretically possible that materials very heavily contaminated by only such short-lived nuclides could still pose a health risk, in reality this situation would never occur: such nuclides would always be commingled with longer-lived nuclides which would be the dominant contaminants at the time of release. As it happens, almost all the radionuclides that are potential contaminants of scrap metal and are affected by this cutoff have half-lives of less than 90 days. Hence, a half-life of six months is a convenient selection criterion. • Likely contaminant of scrap metal. The radionuclides that are likely contaminants of scrap metal were identified by a review of the available literature describing the radionuclides associated with the nuclear fuel cycle. A discussion of this selection process is presented in Appendix D. 3.5 EXPOSURE SCENARIOS AND BIOLOGICAL ENDPOINTS A screening process was used to identify the individuals that would have the highest potential exposures to the various radionuclides that are likely contaminants of steel, aluminum and copper scrap cleared from nuclear facilities. Chapter 5 discusses the process used to select scenarios for the radiological assessment of iron and steel scrap cleared from nuclear facilities, and includes a detailed description of the 19 scenarios which were selected. The exposure scenarios for the assessments of aluminum and copper scrap were selected on the basis of separate investigations of these metals, but were guided by the experience in the analysis of steel scrap. The studies of the recycling of these two metals are described in Appendices B and C, respectively. Detailed descriptions of the exposure scenarios are presented in Chapters 8 and 9. 3-10 ------- The values of the exposure parameters used in the analyses were based on empirical data, whenever such data were available. In cases where the data spanned a range of values, conservative, upper-end values were selected so as to give reasonable assurance that the assessment would not understate the reasonable maximum exposure. When such data were unavailable or uncertain, the assigned values were based on reasonably conservative estimates. The biological endpoints of potential concern could include carcinogenic, genetic, and teratogenic effects. However, as discussed on page 3-2, the present analysis addresses dose and lifetime risk of cancer (excluding non-fatal skin cancer). This methodology is consistent with the approach typically taken by EPA in developing its radiation protection standards. The Agency does not quantify the potential for non-carcinogenic health effects because they are far less likely to occur than carcinogenic effects at the dose levels potentially associated with recycling scrap metal (U.S. EPA 1989). This approach is supported by current international radiation protection guidance (UNSCEAR 1993).4 The objective of the screening process was to limit the individuals, scenarios, and biological endpoints to a manageable number without excluding any that could result in significantly greater impacts than those presented in this report. 3.5.1 Multiple Pathways Consideration was given to the possibility that some individuals could be exposed in multiple, unrelated scenarios. For example, is it reasonable to assume that the lathe operator—who receives one of the high-end exposures from y-emitting radionuclides that partition to cast iron—is also the driver of a car made with residually contaminated scrap? For this to happen, not only would his lathe have to be made from a single heat that contained the highest likely fraction of residually contaminated scrap, but so would the engine block of his car. Each of these two circumstances have a small but finite probability of occurring. Given the number of lathes manufactured and the number of cars produced, there is a good chance that at least one such lathe and one such car would be built. The probability that both such items would be used by a single individual is vanishingly small. UNSCEAR 1993 cites a risk coefficient of 5 x 10"4 per rem for lifetime fatal cancer risk in a nominal population of all ages. The risk coefficient cited for genetic effects is 1.2 x 10"4 for a reproductive population for all generations after exposure. For clinically important disorders for the first generation of offspring of exposed parents, the genetic risk coefficient is cited as 2 x 10"5 to 4 x 10"5 per rem for the reproductive part of the population. 3-11 ------- A second example involves an individual who lives near a steel mill. In any one year, there is a possibility that some steel mill will process a significant quantity of scrap cleared from a nuclear facility. There might be a few people who live downwind from and near the mill, and who farm, grow their own food, keep a dairy cow and raise their own meat animals. The probability that one of these few people would also be exposed to one of those few metal products (lathe, automobile, etc.) made from the maximally contaminated steel is likewise extremely small. Similar arguments can be made regarding other combinations of unrelated scenarios. Because of the small probability that they would simultaneously affect the same individual, such combined scenarios are not further addressed in this analysis. Multiple sources of exposure are considered, however, for industrial workers whose various duties could lead to radiation exposures from different individual sources with a common origin: the residually radioactive material being processed. Such scenarios are described in Chapters 5, 8, and 9 and in Appendix H. 3.5.2 Personal Devices Additional studies were performed of the potential impact of the use of recycled scrap metal in several representative personal devices: a baby stroller, a prosthetic hip replacement, dental braces, an aluminum beverage container and a coin made of a copper alloy. In all cases, the radiological impacts were less than those from the scenarios and pathways included in the detailed analysis. 3.5.3 Other Pathways and Scenarios Further scoping analyses were performed to ensure that important scenarios and pathways were not overlooked in the analysis. For example, we examined the potential exposure from food grown in soil that uses slag as a soil conditioner (liming agent). A conservative, upper-bound calculation revealed that, though this is a realistic pathway, the normalized dose from any nuclide would be at least one order of magnitude less than the dose from the maximum exposure scenario for that nuclide. The analysis of the slag agricultural pathway is presented in Appendix H-l. 3.5.4 Direct Disposal of Scrap Following Clearance Scrap metal could be cleared and then disposed directly in a municipal or industrial landfill. This scenario, which is not part of the suite of exposure scenarios used to determine the normalized impacts on the RME individual, is discussed in Appendix L. 3-12 ------- 3.6 SUMMARY OF THE SCREENING PROCESS This section summarizes the results of the screening process and the resulting scope of the analyses. 3.6.1 Sources of Scrap Metal Out of the various functional categories and four major administrative categories that represent the sources of scrap metal from nuclear facilities, the analysis explicitly addresses four functional categories, two administrative categories, and a number of specific sites that contain or are contaminated with radioactive materials. The four functional categories include DOE enrichment facilities, fuel fabrication and weapons assembly plants, reprocessing and extraction facilities, and nuclear power reactors. The two administrative categories are DOE and the NRC. The specific sites include the 17 major DOE facilities listed in Table 4-2. In addition, the analysis addresses all currently operating nuclear reactors as well as shutdown reactors slated for decommissioning. 3.6.2 Types of Scrap Metal from Nuclear Facilities Of the numerous metals and metal alloys included in the potential scrap metal inventories of nuclear facilities, exposure assessments were performed on carbon steel (which is representative of ferrous metals), copper and aluminum. 3.6.3 Scenarios. Pathways. Modeling Assumptions, and Biological Endpoints Out of the virtually unlimited number of possible ferrous metal exposure scenarios, 19 recycling scenarios (plus the agricultural slag scenario), and one landfill disposal scenario were selected for analysis. More limited assessments were performed for personal devices, as discussed on page 3-12. Twelve exposure scenarios were addressed in the analysis of the recycling of aluminum, plus the dross disposal scenario which is discussed in Appendix L. The analysis of copper scrap utilized six scenarios. The pathways selected for analysis include external exposure, inhalation of dust, inadvertent ingestion of particulate matter, and ingestion of food and water. Parameters were selected to represent realistic values for high-end individuals that may be exposed as a result of the free release of the metals and their subsequent recycling or disposal. Of the range of biological endpoints that could be of concern (i.e., dose, risk of cancer, 3-13 ------- hereditary and teratogenic effects), hereditary and teratogenic effects were not explicitly addressed. 3-14 ------- REFERENCES Baca, T. E. 1992. "DOD Environmental Requirements and Priorities." Federal Facilities Environmental Journal. Habicht, F. H. 1992. "Guidance on Risk Characterization for Risk Managers and Risk Assessors." Memo from F. Henry Habicht, Deputy Administrator, U.S. Environmental Protection Agency, to Assistant Administrators and Regional Administrators (26 February 1992). International Commission on Radiological Protection (ICRP). 1977. "Recommendations of ICRP," ICRP Publication 26. Annals of the ICRP, vol. 1, no. 3. Pergamon Press. International Commission on Radiological Protection (ICRP). 1985. "Principles of Monitoring for the Radiation Protection of the Population," ICRP Publication 43. Annals of the ICRP, vol. 15, no. 1. Pergamon Press. National Research Council, Committee on Technical Bases for Yucca Mountain Standards. 1995. "Technical Basis for Yucca Mountain Standards." National Academy Press, Washington, DC. S. Cohen & Associates (SCA). 1995. "Analysis of the Potential Recycling of Department of Energy Radioactive Scrap Metal." 4 vols. Prepared for U.S. Environmental Protection Agency, Office of Radiation and Indoor Air, Washington, DC. S. Cohen & Associates (SCA). 1997. "Radiation Site Cleanup Regulations: Technical Support Document for the Development of Radionuclide Cleanup Levels for Soil." Vol. 1. Prepared for U.S. Environmental Protection Agency, Office of Radiation and Indoor Air, Washington, DC. United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR). 1993. Sources and Effects of Ionizing Radiation. United Nations, New York. U.S. Department of Energy (U.S. DOE). 1995a. "Draft Waste Management Programmatic Environmental Impact Statement for Managing Treatment, Storage, and Disposal of Radioactive and Hazardous Waste," DOE/EIS-0200-D. U.S. Department of Energy (U.S. DOE). 1995b. "Integrated Data Base - 1994: U.S. Spent Fuel and Radioactive Waste Inventories, Projections, and Characteristics" rev. 10, DOE/RW- 0006. U.S. Department of Energy (U.S. DOE). 1995c. "Estimating the Cold War Mortgage: The 1995 Baseline Environmental Management Report," DOE/EM-0230. 3-15 ------- U.S. Department of Energy (U.S. DOE), Office of Environmental Management. 1996. "Taking Stock: A Look at the Opportunities and Challenges Posed by Inventories from the Cold War Era," DOE/EM-0275. Vol. 1, "A Report of The Materials in Inventory Initiative." U.S. Environmental Protection Agency (U.S. EPA), Office of Radiation Programs. 1989. Environmental Impact Statement: NESHAPS for Radionuclides, Background Information Document," EPA/520/1-89-005-1. Vol. 1, "Risk Assessments Methodology." U.S. EPA, Washington, DC. U.S. Environmental Protection Agency (U.S. EPA). 1993. "Issues Paper on Radiation Site Cleanup Regulations," EPA 402-R-93-084. U.S. Environmental Protection Agency (U.S. EPA), Office of Radiation and Indoor Air. 1996. "Radiation Site Cleanup Regulations: Technical Support Document for the Development of Radionuclide Cleanup Levels for Soil." Addendum, EPA 402-R-96-01 ID. U.S. EPA, Washington, DC. 3-16 ------- Chapter 4 QUANTITIES AND CHARACTERISTICS OF POTENTIAL SOURCES OF SCRAP METAL FROM DOE FACILITIES AND COMMERCIAL NUCLEAR POWER PLANTS This chapter provides quantitative data on the amount of scrap metal potentially available for unrestricted release from nuclear facilities controlled by the Department of Energy (DOE) and from the decommissioning of nuclear power plants. Scrap metal quantities for DOE sources and the means by which the data were developed are discussed in Section 4.1. A comprehensive discussion of scrap metal sources that will be generated from the decommissioning of commercial nuclear power reactors is provided in Appendix A—a summary discussion of these data is presented in Section 4.2. Section 4.3 provides a brief summary of recent recycling activities involving scrap metal from commercial and government-owned facilities.1 4.1 EXISTING AND FUTURE SCRAP METAL QUANTITIES AVAILABLE FROM DOE 4.1.1 Background Information The historic role of DOE was to design, test, manufacture, and maintain nuclear weapons. This effort started with the Manhattan Project and the development of the first nuclear weapons that were employed in World War II. Shortly after World War II, deteriorating relations between the United States and the Soviet Union led to a massive nuclear arms race. In the United States, the nuclear arms race resulted in the development of a vast research, production, and testing network of Federal facilities that came to be known as the "nuclear weapons complex." During half a century of operations, the complex manufactured tens of thousands of nuclear warheads and test-detonated more than one thousand. At its peak, this complex comprised 16 major facilities, each with its own mission (Figure 4-1). Weapons production stopped in the late 1980's, initially to correct environmental and safety problems. Subsequently, most of the nuclear weapons activity has been suspended indefinitely. The information on DOE facilities is primarily based on data collected through 1998. As mentioned in Section 2.1, DOE currently has a moratorium on the free release of volumetrically contaminated metals and has suspended the unrestricted release for recycling of scrap metal from radiological areas within DOE facilities. 4-1 ------- The U. S. Nuclear Weapons Complex to Burlinc|lon Assembly r-l.ini ernnld PI i Uranium and Machining i' ;iticlilh P Ut.iniiiiii ! mi Muni! Kansas City PI Eledront. Mechanical, ftasilc Component Rocky FLils Plant Warhead Ir tigers Lawrence Livermore Laboratory Weapons Researcli and Cesign WoMon Spring Hi muni: l-Hiii-1 : and ! !• I ii Foundry H! ,iii-" II,- Fuel fiTarga Fabrlcallon. ^lV iTillltr" 'li:|fll'ltfcl11' ChemicalSeparallon: 1 Trllinrn Hlgl> Explosives Fabrlcallon Final and WaitB Isolation* Pilot PI. .m Bikini and Eniwetok Atolls Nuclear Weapons Production Nonnuciear Componenls Fuel and Plutonium Uranium Target Production Foundry Fabrication Reactors Reprocessing to Separate Pluto it mm Former industrial sites contaminated with ~ radioactivJtyr some but not all of which contributed to nuclear weapons production. Uranium Is mined, milled. and refined irorn ore NnillHlll!" !'!'•• •••• -1 III! • i1.-: miKiK'ii I it/I irv err tried. and depleted uranium Uranium gas InlO ! IP- : ,1 Uranium metal is formed into HI- i and largel reactors Uranium target •:• !ii-in . are irradiated ID Is used ID exlracl plutonium from Uranium and pluionlum are runner processed Tar warhead (riggers Iil'i'i'^ neulron generalcrs. and other elecirlcal and mecnanlcal tornponenis are assembled into cornplele warheads Figure 4-1. Nuclear Weapons Complex (Source: U.S. DOE 1995a) ------- DOE is now in the process of deciding what should be done with facilities, structures and materials that in many instances are radioactively contaminated. Among the materials that pose significant disposition problems are large quantities of metals that have become radioactively contaminated in various phases of extracting, testing, and producing materials for nuclear weapons. Radionuclides Associated with Nuclear Weapons The principal fissile components of nuclear weapons are highly enriched uranium and plutonium. Early nuclear weapons were designed to use either highly enriched uranium or plutonium that, when compressed into a critical mass, would sustain a nuclear chain reaction and result in a nuclear explosion. As designs for nuclear weapons improved, a new generation of thermonuclear weapons evolved that require both plutonium and highly enriched uranium. Thermonuclear weapons also require a third ingredient: tritium, a radioisotope of hydrogen that boosts the explosive power of the nuclear weapon commonly referred to as the hydrogen bomb. The processes by which these three components are produced are the source of radioactive contamination of scrap metals at DOE facilities. Enriched Uranium. About 99.3% of naturally occurring uranium atoms consists of U-238, almost all of the remaining 0.7% being U-235. However, U-235 is the only naturally abundant uranium isotope that can undergo the sustained fission required for the detonation of nuclear weapons. To make uranium highly enriched in U-235, DOE facilities at the Oak Ridge Reservation in Tennessee initially used two elaborate processes to extract U-235 from natural uranium: (1) electromagnetic separation and (2) gaseous diffusion. However, most of the enrichment was done by the gaseous diffusion process which used uranium hexafluoride (UF6) gas as the vehicle for enrichment. Additional diffusion plants were subsequently built at Paducah, Ky. and Portsmouth, Ohio. The enriched UF6 gas must be converted into a metal before it is used in nuclear weapons production. At the Fernald uranium foundry in Ohio, the UF6 was chemically converted into uranium metal. Enriched uranium metal was: (1) used as fissionable material in nuclear weapons and (2) fabricated into nuclear fuel for DOE reactors used to produce plutonium. Between 1944 and 1988, DOE operated 14 plutonium-production reactors at the Hanford and the Savannah River Sites, producing about 100 tons of plutonium. Pu-239 that is required for 4-3 ------- nuclear weapons is produced by the neutron irradiation of depleted uranium metal targets2. Additional weapons-grade plutonium is recovered from the spent fuel of the production reactors. Unfortunately, both sources also contain hundreds of different radionuclides that must be chemically separated from the fissionable material. Scientists developed elaborate physical structures and chemical processes to accomplish this separation in a manner that is consistent with the safety of workers and the public. A total of eight chemical separation plants, called "canyons," were operated for the DOE. These plants employed the PUREX process for the separation and recovery of plutonium and uranium. In total, the eight chemical separation plants generated more than 100 million gallons of radioactive wastes that are currently contained and stored at DOE facilities. Sources of Data Used to Quantify and Characterize DOE Scrap A thorough search for available reports and study data that might contain useful information regarding scrap metal inventories and a characterization of those inventories identified only very limited sources. This was not unexpected when viewed in context of the highly secretive, classified nature of past nuclear weapons activities, the relatively short time since the end of the Cold War, and the yet-undecided future for many DOE facilities. For these reasons, DOE has only in recent years begun to evaluate existing and future material inventories and their management. Some of DOE's earliest attempts to assess material inventories were based on the most cursory of data; data that were further compromised by an uncertain and continuously revised projection of future needs. Earlier reports are, therefore, of limited value and data reported therein have been revised and updated to reflect the most current information, facility status, and future needs. The following reports are among the most informative regarding existing and future scrap metal inventories: 1 "A Report of the Materials in Inventory Initiative. Taking Stock: A Look at the Opportunities and Challenges Posed by Inventories from the Cold War Era" (U.S. DOE 1996). (This report is commonly referred to as the "1996 MIN Report" or simply the "MIN Report") Depleted uranium metal targets are prepared by converting the UF6 gas that is left after the lighter isotopes— U-234 and U-235—have been extracted by the gaseous diffusion process. 4-4 ------- 2 "U.S. Department of Energy Scrap Metal Inventory Report for the Office of Technology Development, Office of Environmental Management," (Parsons et al. 1995). (This report is commonly referred to as the "HAZWRAP Report.") 3 "Scrap Metal Inventories at U.S. Nuclear Facilities Potentially Suitable for Recycling" (SCA 1995a). 4 "Gaseous Diffusion Facilities Decontamination and Decommissioning Estimate Report" (Person et al. 1995). Collectively, these four documents identified 13 DOE facilities as principal sources of scrap metal. A brief description of each of the thirteen sites is presented below (U.S. DOE 1996). • Fernald Environmental Management Project. Located on 1,050 acres in the southwest corner of Ohio, the Fernald Environmental Management Project (FEMP), formerly known as the Feed Materials Production Center, was constructed in the early 1950's to convert uranium ore to uranium metal targets. Uranium targets were subsequently shipped to DOE production reactors, where targets were irradiated for the production of plutonium used in nuclear weapons. Over a 36-year period, this facility produced over 225 million kilograms of purified uranium. Production of uranium targets ceased in 1989. Principal radioactive contaminants include the uranium isotopes and their radioactive progenies and Tc-99. • Hanford. The Hanford reservation encompasses about 560 square miles within the Columbia River Basin in southeastern Washington and borders the Tri-Cities area of Richland, Pasco, and Kennewick to the south. Nuclear materials were produced at Hanford since the early 1940's. Activities once included plutonium production and separation, advanced reactor design and testing, basic scientific research, and renewable energy technologies development. • Idaho National Engineering and Environmental Laboratory. The Idaho National Engineering and Environmental Laboratory (INEEL) encompasses an area of approximately 890 square miles in southeastern Idaho on the edge of the Eastern Snake River Plain. INEEL is a multipurpose laboratory supporting the engineering and operations efforts of DOE and other Federal agencies in the areas of nuclear safety, reactor development, reactor operations and training, waste management and technology development, nuclear fuel reprocessing, and energy technology/conversion programs. Over 50 nuclear reactors, most of them small test reactors, have existed at INEEL. Some of these reactors and their associated support buildings have been decommissioned and demolished. Others are slated for decommissioning. • Los Alamos National Laboratory. Los Alamos National Laboratory (LANL) occupies about 43 square miles approximately 25 miles northwest of Santa Fe, N.M. LANL was 4-5 ------- established in 1943 with the specific responsibility of developing the world's first nuclear weapon. The Laboratory's original mission rapidly broadened to include research programs in nuclear physics, hydrodynamics, conventional explosives, chemistry, metallurgy, radiochemistry, and relevant life sciences. In addition to research, a second important mission of the Laboratory between 1945 and 1978 was to process plutonium metal and alloys from nitrate solution feedstock provided by other DOE production facilities. Other operations included reprocessing of nuclear fuel, processing of polonium and actinium, and producing nuclear weapons components. Although the Laboratory has retained many of the original research programs dealing with national defense, its current mission has been expanded to include research in emerging technologies pertaining to biomedicine, nuclear systems for outer space, materials sciences, computational sciences, and environmental management. Nevada Test Site. The Nevada Test Site (NTS) is 65 miles northwest of Las Vegas and occupies 1,350 square miles, making it the largest facility in the DOE complex. NTS has been the primary site for atmospheric and underground nuclear weapons testing by DOE, with more than 300 nuclear tests conducted above and below ground. This includes tests at NTS and at seven other locations outside Nevada. All nuclear weapons tests at NTS had been conducted underground since 1963. The United States has observed a moratorium on all nuclear tests since 1992. Oak Ridge National Laboratory. Founded in 1942, the Oak Ridge National Laboratory (ORNL) occupies about 2,900 acres within the Oak Ridge Reservation, which lies south and west of Oak Ridge, Tenn. The Laboratory's original mission was to produce and chemically isolate the first gram quantities of plutonium for use in nuclear weapons. With time, the scope of ORNL greatly expanded to include production of other radionuclides, fundamental research in a variety of scientific disciplines, research pertaining to hazardous and radioactive materials, environmental studies, radioactive waste management and disposal, and a wide range of educational programs. Y-12 Plant. Built in 1943 as part of the Manhattan Project, the Oak Ridge Y-12 Plant occupies approximately 811 acres within the Oak Ridge Reservation. This facility consists of some 250 buildings that house about seven million square feet of laboratory, machining, and research and development areas. The initial mission of the Y-12 Plant, which began operation in November of 1943, was the separation and enrichment of U- 235 from natural uranium by an electromagnetic separation process. When gaseous diffusion technology became the accepted process for uranium enrichment, the magnetic separators were taken out of service in 1946. Since that time, the Y-12 Plant's mission has shifted to the disassembly of returned weapons components, quality evaluation for the existing stockpile of nuclear weapons, and research in engineering designs associated with the production and fabrication of nuclear weapons components. 4-6 ------- ETTP. The East Tennessee Technology Park (formerly known as the Oak Ridge K-25 Site) occupies about 1,500 acres within the Oak Ridge Reservation. The K-25 Gaseous Diffusion Plant was built as part of the Manhattan Project to supply highly enriched uranium for nuclear weapons production. Construction of the primary K-25 building started in 1943 and the plant was fully operable by August 1945, with additional buildings involved in the enrichment process brought on stream by 1956. Beginning in 1964, exclusive production of highly enriched uranium for weapons was gradually replaced with the production of commercial-grade, low-enrichment uranium for the emerging nuclear power industry. Because of the declining demand for enriched uranium, the K-25 Plant was placed on standby in 1985 and was permanently shut down in 1987. Paducah. Located just outside Paducah, Ky., the Paducah Gaseous Diffusion Plant site occupies approximately 750 acres of Federally-owned land. The plant was constructed in the early 1950s for the purpose of enriching uranium by the gaseous diffusion process. Since 1991, both Paducah and Portsmouth have produced only low-enriched uranium for use as fuel in commercial nuclear power plants. In 1993, production operations at both gaseous diffusion plants were assumed by the United States Enrichment Corporation (USEC), a government corporation formed under the Energy Policy Act of 1992. Portsmouth. The Portsmouth Gaseous Diffusion Plant, located in Piketon, Ohio (approximately 22 miles north of Portsmouth and 75 miles south of Columbus) is situated on 3,714 acres of federally-owned land. In spite of the then-existing gaseous diffusion programs at the K-25 facility and at Paducah, the Portsmouth facility was built to meet the demand for highly enriched uranium created by the emergence of nuclear submarine reactors and for low-enriched uranium for projected commercial nuclear power reactors. In June 2000, USEC announced plans to shut down enrichment operations at the Portsmouth plant in June 2001 (Bechtel Jacobs 2001). Rocky Flats. The Rocky Flats Environmental Technology Site (RFETS) covers 11 square miles located approximately 16 miles northwest of Denver. Its primary mission was to produce nuclear weapon components, which involved plutonium handling and fabrication. Currently, activities at RFETS include cleaning up contamination and waste from its past activities and converting its facilities to alternative uses. Savannah River Site. The Savannah River Site (SRS) is located in west-central South Carolina and has an area of approximately 310 square miles; its production facilities occupy less than 10% of the total area. SRS was established by the Atomic Energy Commission in 1950 for the purpose of producing Pu-239 and tritium for nuclear weapons. SRS also produced other special radionuclides (Cf-252, Pu-238, and Am-241) to support research in nuclear medicine, space exploration, and commercial applications. To produce these nuclides, metal targets were irradiated in the five production reactors. The radionuclides were recovered from irradiated targets at chemical separation facilities 4-7 ------- also located at SRS. Current operation of chemical process facilities is limited to the recycling of tritium and the extraction of Pu-238 for use in space exploration. • Weldon Spring. The Weldon Spring Site consists of 229 acres, approximately 20 miles west of St. Louis, and comprises the Weldon Spring Chemical Plant and the Weldon Spring Quarry. It was part of a site used by the U.S. Army as an ordnance works in the 1940s. In the 1950s and 1960s, the Atomic Energy Commission processed uranium ore in the Chemical Plant. The plant was subsequently deactivated and no further activities were carried out at the site since remediation began in 1985. Relevant data contained in these four documents are briefly summarized below. Estimates of scrap metal quantities and limited qualitative data are defined in terms of (1) existing scrap metal inventories and (2) projected scrap metal inventories associated with future decommissioning of DOE facilities. Because significant gaps in quantitative information remain, an attempt was made to supplement reported data by direct contact with DOE personnel. Individuals contacted included key members of the administrative staffs at DOE Headquarters and DOE Regional Offices, as well as personnel in DOE field offices. Field personnel included individuals with responsibilities related to scrap metal decontamination, segregation, storage, environmental monitoring, and salvage and recycling operations. In most instances, direct contacts yielded only subjective information that explained the approach and general methods used to arrive at the reported quantities of scrap metal. 4.1.2 Existing Scrap Inventories at DOE Data Reported in 1996 MIN Report DOE's first major undertaking to evaluate its materials management practices dates back to January 1990 with the establishment of the Mixed Waste and Materials Management Workgroup. To support the Workgroup effort, an attempt was made to define and inventory Materials Not Classified As Waste (MNCAW) and resulted in the 1994 MIN Report (formerly known as the MNCAW Report). This and other reports have been combined and collated with new data and analysis to provide information presented in the 1996 MIN Report (U.S. DOE 1996). DOE defines "materials in inventory" as materials that are not currently in use (i.e., have not been used during the past year and are not expected to be used within the coming year) and that have 4-8 ------- not been set aside for national defense purposes. The Department identified ten categories with significant quantities of materials. The ten categories are divided into two subcategories: nuclear materials and non-nuclear materials. Scrap metal and equipment are listed in the non- nuclear materials subcategory (Table 4-1). Table 4-1. Groupings of DOE Materials in Inventory Nuclear Materials Spent Nuclear Fuel Plutonium and Other NMMSS- Tracked Materials Natural and Enriched Uranium Depleted Uranium Lithium Non-Nuclear Materials Sodium Lead Chemicals Weapons Components Scrap Metal and Equipment Scrap metal consists of worn or superfluous metal parts or pieces including, but not limited to, structural steel and other metals from decommissioned buildings, scrap metals accumulated from facility maintenance and renovation in the past, and scrap stored in scrap yards and lay-down yards. Scrap metal includes metals that are clean and metals radioactively contaminated or activated and/or contaminated with hazardous substances. Equipment considered in the MIN Report is defined as all equipment used for construction, production, or manufacturing, and all associated spare parts and hand tools. To estimate scrap material inventories, the Department recruited personnel from each DOE Operations and Field Office and from designated Headquarters Offices. The MIN Scrap Metal and Equipment Team sought information by means of site-specific surveys and, whenever possible, extracted information contained in various DOE databases. MIN data collection was, therefore, constrained by the need to use existing data sources; the project team was neither authorized nor allocated resources to conduct new studies or to develop new information. The report acknowledges its limitations and states: ". . . Because of limited data, this report does not attempt to capture the exact amount of each material in inventory. Rather, it attempts to capture the general magnitude of the inventory of each material [emphasis added]." Despite its acknowledged limitations, the 1996 MIN report is regarded as the principal data source for scrap metal estimates for most DOE facilities. Table 4-2 summarizes these data, as 4-9 ------- well as presenting estimates for several facilities which were not listed in the MIN Report but represent interpolated values. Interpolation was needed because only a few DOE sites provided complete quantitative estimates that defined existing scrap metal inventories as clean or radioactively contaminated. Many facilities either provided only a partial breakdown or no breakdown with regard to quantities of contaminated versus uncontaminated scrap metals. In fact, the largest percentage of DOE scrap metal (-80%) reported in the 1996 MIN Report is designated as "unspecified" with regard to radioactive contamination. For scrap metal inventories designated as "unspecified," it was assumed that 88% of scrap metal was contaminated and 12% was clean and not considered contaminated. The basis for this assumption is Table 1-6, page 16 of Volume 2 of the 1996 MIN Report. Table 4-2 shows that about 90% of the contaminated scrap in existing stockpiles is currently located at five sites. In descending order, they are: Paducah, K-25, SRS, Y-12, and Portsmouth. Information on contaminants identified at each site is also included in Table 4-2. Data Extracted from the HAZWRAP Report (Parsons et al 1995) In 1994, Martin Marietta Energy Systems, Inc., in support of the DOE's Hazardous Waste Remedial Actions Program (HAZWRAP), conducted a study that assessed scrap metal inventories and their economic values for 11 DOE facilities. Collection of information on amounts and locations of scrap metal within the DOE complex was pursued through three independent but complementary methods. A preliminary questionnaire was forwarded to key site personnel, which requested generic demographic data pertaining to scrap metal management along with a "DOE Scrap Metal Data Sheet." Key information sought by the questionnaire included (1) type of material (e.g., steel, aluminum, copper, etc.); (2) "radioactivity", and (3) quantity. A second source of information for developing estimates in the HAZWRAP Report came from a thorough review of published reports and DOE databases. A total of 28 documents were identified as pertinent. 4-10 ------- Table 4-2. Existing Scrap Metal Inventories at DOE Sites DOE Site FEMP Hanford INEEL LANL NTS ORNL Y-12 K-25 Paducah Portsmouth RFETSd SRS Weldon Spring6 Fermilab/ANL-W/BNL Pantex/WIPP Ashtabula SLAC Total Reported Quantitya (tons) 5,115 416C 2,300 0 331 1,411 11,332 36,699 60,473 11,143 — 15,533 — 7,448 393 70 17 152,681 Contaminated Fraction13 0.895 1 0.348 — 0 0.88 0.88 0.88 0.88 0.88 — 0.934 — 0.995 0 0 0 Mass (t) 4,161 378 727 0 0 1,129 9,066 29,359 48,378 8,914 — 13,189 — 6,722 0 0 0 122,023 Identified Contaminants not specified not specified fission products on stainless steel, not specified on carbon steel Cs-137, Sr-90, Co-60 not specified U+ progeny, Tc-99, Pu-239 (trace), Np-237 (trace) same as K-25 same as K-25 H-3, Co, Eu, Cs-137, Am-241, Sb-125 not specified activation products at Fermi Lab Source: U.S. DOE 1995b aU.S. DOE 1995b, Table 1-1 b U.S. DOE 1995b, Table 1-3, or, if not specified, 88% is assumed contaminated per Table 1-6, Note 2 c Hanford scrap is not included in U.S. DOE 1995b, Table 1-1, but is noted as contaminated "mixed" scrap on p. A8-3. d No available data e ROD calls for on-site burial (U.S. DOE 1995b, p. 10) ------- Lastly, the HAZWRAP Project Team visited the sites and met with personnel to examine storage areas and document the locations and amounts of stored scrap metal. Confirmatory estimates of stored scrap metal quantities were based on physical measurements of size and storage density of piles. Scrap metal estimates reported in the HAZWRAP Report were either used directly or updated for the 1996 MIN Report. As indicated in Table 4-2, scrap metal data for LANL, RFETS, and the Weldon Spring facilities were not fully discussed in the 1996 MIN Report. A brief description of the management and current inventories of scrap metals at these three sites, as reported in the HAZWRAP Report, is presented below. Los Alamos National Laboratory. LANL has an active scrap metal recycling program. Existing scrap metal inventories are stored at several locations in small piles, the largest of which is about 1,800 t. The total quantity of contaminated scrap metal at LANL is estimated to be 3,099 t. Rocky Flats. At RFETS, contaminated scrap metal is stored in metal drums and boxes that were inventoried in the Site Waste Management database. Because the material quantities could not be determined using the methods described previously, information from the Waste Management Program was used to quantify amounts and metal types of scrap inventories. The total amount of contaminated scrap metal was estimated to be 24,543 t. Weldon Spring. At the Weldon Spring Site, scrap metal is located in two storage areas. Contaminated scrap metal removed in the past from process piping associated with the Quarry and the Chemical Plant is stored in the Temporary Storage Area and in an eight-acre laydown area called the Material Storage Area. A total of 27,8391 of contaminated scrap metal was estimated to be stockpiled. Since, according to U.S. DOE 1995b, p. 10, the Record of Decision for Weldon Spring specifies on-site burial of the waste, this scrap metal is not included in the inventory presented here. 4.1.3 Summary of Existing Scrap Inventories at DOE Sites Table 4-3 summarizes the current best estimates of contaminated scrap metal quantities stored at DOE facilities. Most of these estimates were derived from data presented in U.S. DOE 1996. The remaining values were derived from information presented by Parsons et al. (1995). 4-12 ------- Based on these data, it is estimated that existing inventories of scrap metal comprise about 150,000 t. Table 4-3. Estimates of Existing DOE Inventories of Contaminated Scrap Metal (t) DOE Site FEMP Hanford INEEL LANL NTS ORNL Y-12 K-25 Paducah Portsmouth Rocky Flats SRS Other Subtotal TOTAL Existing Scrap Metal Quantities MIN Report 4,161 378 727 0 0 1,129 9,066 29,359 48,378 8,914 Not Reported 13,189 6,722 122,023 HAZWRAP Report 3,099 24,543 27,642 149,665 4.1.4 Scrap Metal Inventory by Metal Type Data collected in support of the HAZWRAP Report provided information regarding the composition of scrap metal inventories. Quantity estimates were provided for seven forms of scrap metal classified as: (1) carbon steel, (2) stainless steel, (3) copper and brass, (4) nickel, (5) aluminum, (6) tin and iron, and (7) miscellaneous, which included lead, monel, and cast iron. These data were reviewed and updated by the MIN Scrap Metal and Equipment Team. Table 4-4 summarizes data reported in the 1996 MIN Report by metal type. Inspection of Table 4-4 shows that 3,503 t of scrap metal were found to be free of radioactive contamination. Moreover, an estimated 110,042 t, or about 79.5% of existing scrap, had not been assessed for radioactive contamination and were classified as "unspecified." 4-13 ------- Table 4-4. DOE Scrap Metal Inventory (t) Metal Carbon Steel Nickel Stainless Steel Aluminum Copper and Brass Tin and Iron Miscellaneous Total Percent Clean 1,008 0 1,435 27 24 227 782 3,503 2.5 Contaminated 11,437 0 5,392 14 1,483 0 6,537 24,863 18.0 Unspecified 94,472 8,817 959 5,637 147 0 10 110,042 79.5 Total 106,917 8,817 7,786 5,678 1,654 227 7,329 138,408 100.0 % 77.2 6.4 5.6 4.1 1.2 0.2 5.3 100.0 Contaminated Assumed 86,820 8,817 757 1,925 145 0 9 98,473 71.1 Total 98,257 8,817 6,149 1,939 1,628 0 6,546 123,336 89.1 Scaled3 119,232 10,699 7,462 2,353 1,975 0 7,943 149,665 Source: 1996 MIN Report a Scaling factor =1.213 In order to characterize the "unspecified" scrap and adjust the totals in Table 4-4 to be consistent with those in Table 4-3, the following procedure was used. The total quantity of contaminated scrap was estimated by applying the following formula to each of the metals in Table 4-4: Assumed Contaminated = Known Contaminated Known Contaminated + Clean Unspecified (In the absence of more information, all of the nickel was conservatively assumed to be contaminated.) Using this procedure, 98,474 t of "unspecified" scrap in Table 4-4 were reclassified as "assumed contaminated." The "assumed contaminated" quantities were added to the contaminated quantities of each metal in Table 4-4 to obtain the total amount of contaminated scrap listed in column 8. However, the total contaminated scrap for all metals resulting from this calculation, 123,336 t, is less than the 149,665 t of contaminated scrap for all sites shown in Table 4-3. To account for this discrepancy, each value in column 8 was scaled upward by a factor of 1.213 (149,665 + 123,336 = 1.213). The adjusted inventories are shown in the last column. It should be noted that carbon steel comprises about 80% of the total DOE inventory of contaminated scrap metal. 4-14 ------- 4.1.5 Scrap Metal from Future Decommissioning During peak periods of activity, the nuclear weapons complex included more than 120 million square feet of building structures (U.S. DOE 1995a). These buildings include 14 large production reactors with extensive support structures, research reactors and their associated support structures, eight chemical processing plants with vast quantities of metal piping, tanks, valves, motors, ductwork, and structural components, and an array of buildings used for storage, milling, manufacturing, testing, assembly, and administrative activities. With the end of the Cold War Era and the reduced demand for additional nuclear weapons, many of these structures will be decommissioned over the next several decades. As of June 1995, DOE's Office of Environmental Restoration Decommissioning Inventory slated 865 structures for future decommissioning (U.S. DOE, Office of Environmental Restoration Decommissioning Inventory, June 1995). Several facilities are still awaiting final notification of deactivation and are not yet designated for decommissioning. As a result, assessments aimed at estimating future scrap generation at some DOE sites have not been conducted for these facilities. Site-Specific Estimates For those DOE sites that are slated for partial or total decommissioning, scrap quantities are at best preliminary estimates that are based on limited and incomplete data. Projected scrap estimates associated with future decommissioning activities were derived from three reports that include the following sites: • 1995 SC&A Report (SCA 1995a): FEMP, Hanford, LANL, Rocky Flats • 1996 MIN Report: INEEL, SRS • 1995 ORNL Report (Person et al., 1995): K-25, Paducah, Portsmouth Combined scrap quantities from future decommissioning activities at these sites are estimated to be 925,0001. Scrap sources and site-specific estimates for the nine sites are briefly summarized below. Hanford. To date, only modest attempts have been made to assess future scrap quantities pertaining to decommissioning activities. However, quantities are expected to be substantial. As 4-15 ------- of June 1995, 250 buildings at Hanford had been slated for decommissioning. Massive amounts of structural steel scrap will be produced during the decommissioning of these buildings. Also included are other structures such as exhaust stacks, storage tanks, and river outfall structures as well as carbon steel and stainless steel pressure vessels from the Clinch River Breeder Reactor program. Approximately 91,798 t of scrap are likely to be generated during decommissioning activities. The vast majority of scrap is expected to be carbon steel with significant amounts of stainless steel, lead, and aluminum. The total scrap includes about 1001 of graphite, which is not included in the present analysis. Idaho National Engineering and Environmental Laboratory. Over the past 50 years, more than 50 nuclear reactors (mostly small test reactors) have operated at INEEL. While some of these reactors and their support buildings have already undergone decommissioning, others are targeted for future decommissioning. Many published DOE documents that cite scrap estimates were assessed in SCA 1995a and in the 1996 MIN Report. Future decommissioning activities at INEEL are estimated to generate 33,785 t of surface-contaminated scrap metal. At this facility, carbon steel (55.7%) and stainless steel (44.0%) constitute nearly all the projected contaminated scrap metal. There are also 337,6441 of uncontaminated, non-activated carbon steel at the site and 4721 of activated steel (U.S. DOE 1995b, p. A3-2). In the present analysis, it was assumed that activated steel would not be a candidate for unrestricted release. Los Alamos National Laboratory. LANL's Metal Inventory Report (LANL 1996) not only assessed existing scrap metal inventories but identified future scrap metal quantities associated with decommissioning activities, as well as for scheduled "upgrade" projects. In combination, decommissioning and upgrade activities are estimated to generate a total of 2,686 t of scrap. Fernald. The FEMP production area includes 20 process facilities and supporting structures that are obsolete and beyond their design life. In total, 128 buildings and 72 miscellaneous structures have been designated for decontamination and decommissioning. The dismantling of buildings, process equipment, and structures is estimated to generate 135,623 t of scrap. Savannah River Site. SRS includes five heavy water production reactors that were used in the production of tritium and other weapon materials. All reactors have been shut down and, at present, there are no scheduled restart dates. Scrap associated with the decommissioning of the 4-16 ------- five production reactors and support structures/systems is estimated to be 3,463 t with nearly equal contributions from carbon steel and stainless steel. The fate of the two SRS chemical separation plants and the many facilities that support them remains undetermined. The decommissioning of these facilities would undoubtedly add substantial (but to date undefined) quantities of scrap. Rocky Flats. A literature search in support of SCA 1995a revealed the existence of only one study that estimated future scrap quantities for Rocky Flats. A study by the Manufacturing Sciences Corporation (Floyd 1994) stated that the decommissioning of Rocky Flats is expected to generate about 1,003 t of scrap metal from four buildings that were to be cleaned up by the National Conversion Pilot Project and an additional 25,300 t from the other buildings and site structures. Most scrap is likely to be contaminated with depleted uranium, enriched uranium, and/or plutonium. Oak Ridge, K-25 Facility. The K-25 facility is the first of three DOE gaseous diffusion plants that are slated for decommissioning. Decommissioning of the K-25 site is estimated to take a total of eleven years: two years of planning and nine years of decontamination and decommissioning. Decommissioning activities are currently projected to be completed in the year 2006 (Person et al. 1995, Fig. 2). Decommissioning will include removal of large quantities of metals associated with process equipment, piping, and structural components. Principal contaminants include uranium isotopes and their radioactive progenies, Tc-99, and trace quantities of Np-237 and Pu-239. A total quantity of 406,3721 of recyclable metal was listed by Person et al. (1995) but the report did not specify the fractions of uncontaminated and contaminated scrap metal. Subsequently, personal communications with Gary Person (1996) yielded the following estimates: of the total future inventory of 406,273 t of scrap metal, 193,666 t are estimated to be free of contamination and about 212,7061 are likely to be residually contaminated scrap that is considered suitable for unrestricted release. Portsmouth. Decommissioning of the Portsmouth gaseous diffusion facility is scheduled to begin in FY 2007 (following completion of decontamination and decommissioning activities at the K-25 facility), with a completion date in FY 2015 (Person et al. 1995, Fig. 2). The decontamination and decommissioning of the three gaseous diffusion plants are purposely 4-17 ------- scheduled in series in order to (1) learn from experience, (2) minimize annual expenditures, and (3) provide a steady stream of metal for recycle. The availability of 312,085 t of total scrap metal was reported by Person et al. (1995). Of this quantity, 189,0721 were estimated to be contaminated metal that, after decontamination, could be suitable for unrestricted release. Paducah. The Paducah Gaseous Diffusion Plant will be the third such facility to be decommissioned. Decommissioning is currently projected to start in 2015 and end in 2023 (Person et al. 1995, Fig. 2). The first major phase will be the removal and decontamination of major components (i.e., motors, cell housing, compressors, converters, piping and valves, electrical equipment, and HVAC systems) from the process buildings. Person (1996) said that of the total projected scrap metal inventory of 331,365 t (Person et al., 1995) about 230,886 t are estimated to be scrap that is considered suitable for unrestricted release. 4.1.6 Summary and Conclusions Regarding DOE Scrap Metal Inventories At its peak, the nuclear weapons complex consisted of 16 major facilities that included buildings with a combined area of more than 120 million square feet. These buildings contain large quantities of equipment, structural steel, and other metal components. Over a 50-year period, some of these buildings, their ancillary facilities, and the equipment they housed have been renovated, replaced, and/or demolished. Currently, about 150,000 t of residually contaminated scrap metal that is considered suitable for unrestricted release is stored at various facilities. Estimates of existing scrap metal quantities are mostly based on site-specific reviews of historical inventory data and physical surveys of scrap piles; these estimates can therefore be viewed with reasonable confidence. Future scrap quantities are closely linked to projected decommissioning activities at DOE sites that make up the nuclear weapons complex. At some sites, virtually all structures and their contents will be dismantled and removed; at other sites decommissioning may be limited, and the DOE will continue selected operations considered crucial to national security or important to research. To date, decisions and commitments for decommissioning are not only incomplete but, in instances where such decisions have been made, they remain both tentative and subject to change in scope and schedule. Consequently, estimates of future scrap quantities are uncertain. 4-18 ------- In the present report, future scrap estimates were based on currently scheduled decommissioning activities at nine facilities: FEMP, Hanford, INEEL, LANL, SRS, Paducah, Portsmouth, Y-12, and K-25. Decommissioning of these facilities is estimated to yield more than 925,000 t of contaminated scrap metal that is derived from dismantling large production reactors, research reactors, chemical processing plants, and a vast array of associated support facilities and structures. With effective decontamination, this scrap metal is potentially available for unrestricted release. Table 4-5 provides summary estimates that represent existing scrap inventories and future scrap associated with decommissioning activities. It is important to remember that the information in this table is for contaminated scrap only. Of approximately one million tonnes of scrap, about 85% is carbon steel, while copper, nickel, aluminum, and stainless steel constitute virtually all of the remainder. It is possible that these values may underestimate the total scrap metal quantities because data pertaining to future decommissioning activities are incomplete. In the fall of 2000, DOE made a data call requesting information from field locations on current scrap metal inventories and projected scrap metal generation from decommissioning activities through 2035. The data call was designed to support a feasibility study on a dedicated steel mill to process DOE scrap into containers for DOE use (Geiger 2001). As a consequence, materials not suitable for steelmaking because of economic or radiological reasons were eliminated from the database. The data call was confined to carbon steel, iron, stainless steel, and nickel (a key alloying element in stainless steel). Table 4-6 presents a comparison of information from the 2000 data call with corresponding data from Table 4-5. While there is some shift between carbon steel and stainless steel, the amounts of ferrous metals from the two analyses are remarkably similar. 4.2 SCRAP METAL FROM THE COMMERCIAL NUCLEAR POWER INDUSTRY At the end of 1997, the U.S. commercial nuclear power industry included 104 operating reactors and 27 reactors3 formerly licensed to operate (see Appendix Al). Over the next two to three decades, most of the reactors currently in operation will have reached the expiration date of their initial 40-year operating licenses. However, as stated in Chapter 2, NRC has granted 20-year Only 17 of these reactors are anticipated to release significant quantities of scrap metal (see Section A.5.2.2). 4-19 ------- extensions of the operating licenses of five reactors; a number of other renewal applications are pending, and more applications are anticipated. A great deal of data has been compiled by the NRC and the individual utilities regarding the decommissioning of these facilities and the quantities and characteristics of the scrap metal that would be generated in the process. Appendix A presents a detailed summary of the relevant information; an abbreviated version is provided in this section. Table 4-5. Existing and Future Contaminated Scrap Metal at DOE Facilities (t) Site Name FEMP Hanford INEEL LANL ORNL Y-12 K-25 Paducah Portsmouth RFETS SRS Other Total Percent Scrap Metal Database 139,780 90,724 34,511 5,785 1,129 9,066 242,065 279,264 197,986 50,846 16,651 215 1,068,022 100.00 Metal Aluminum — 684 40 18 34 7,988 21,161 6,130 — 14 1 36,070 3.38 Carbon Steel 101,740 87,020 19,018 5,568 992 8,392 232,955 212,921 191,412 33,666 10,213 — 903,897 84.63 Stainless Steel — 787 15,449 177 117 602 753 190 18 2,454 6,413 — 26,960 2.52 Copper/ Brass 38,040 5 44 — 2 38 304 198 408 14,726 11 214 53,990 5.06 Nickel — 24 — — — — — 44,794 — — — — 44,818 4.20 Monel — — — — — — 65 — 18 — — — 83 0.01 Lead — 291 — — — — — — — — — — 291 0.03 Other/ misc. — 1,913 — — — — — — — — — — 1,913 0.18 Note: Restricted to metal whose disposition may be affected by a future release standard. Table 4-6. Comparison of Estimates of Ferrous Metal and Nickel Inventories (1000 t) Material Carbon Steel & Iron Stainless Steel Nickel Total 2000 Data Call3 792 158 34 984 Pre-2000 Estimates'3 904 27 45 976 Difference 14.1% -82.9% 32.4% -0.8% aGeiger2001 b Table 4-5 4-20 ------- A key factor affecting the quantity of scrap metal and associated contamination levels is the basic design of the reactor. The two types of reactors operating in the United States are the pressurized water reactor (PWR) and the boiling water reactor (BWR). Of the 104 reactors operating in the United States, 35 are BWRs manufactured by General Electric and 69 are PWRs manufactured by Westinghouse, Combustion Engineering, and Babcock and Wilcox. Between 1976 and 1980, two studies were carried out for the NRC by the Pacific Northwest National Laboratory (PNNL) that examined the technology, safety, and costs of decommissioning large reference nuclear power reactor plants. Those studies, by Smith et al. (1978) and Oak et al. (1980), for a reference PWR and reference BWR, respectively, reflected the industrial and regulatory situation of the time. To support the final Decommissioning Rule issued in 1988, the earlier PNNL studies have been updated by Konzek et al. (1995) and Smith et al. (1994). These four reports, along with several other NRC reports and selected decommissioning plans on file with the Commission, represent the primary source of information used to characterize Reference PWR and BWR facilities and to derive estimates of scrap metal inventories for the industry as a whole. 4.2.1 Estimates of Contaminated Steel from Commercial Nuclear Power Plants Table 4-7 presents summary data on contaminated steel potentially available for clearance. Estimates for the Reference BWR and PWR were derived by summing component mass values cited in Tables A-32/64 and Tables A-65/79, respectively. Estimates for the entire commercial nuclear industry were derived by taking Reference BWR and Reference PWR values and applying plant-specific scaling factors for each operating and formerly licensed reactor (except for those which are in an ENTOMB status or for which DECON is in progress or completed). The row marked "Total" lists the total quantities of steel used to construct each plant. "Releasable" refers to all contaminated steel that is a candidate for release, excluding only steel that is neutron-activated. (This includes metal that would require very aggressive decontamination methods to achieve any foreseeable clearance criteria.) Approximately 600,0001 of contaminated steel may become available over time for unrestricted release. About 80% of the contaminated steel is carbon steel, with stainless steel representing the balance. The data on contaminated equipment in nuclear power plants is usually presented in terms of areal (surface) activity concentrations. However, as will be discussed in the following chapters, the risk assessments of the recycling of scrap metals are based on specific activities. Converting the areal activities to specific activities involves the average mass thickness of the metal, which is given by the following equation: 4-21 ------- Average Mass Thickness = Mass Area Table 4-7. Residually Radioactive Steel from Nuclear Power Plants (t) Reactor Type Rebar All Other Total Releasable3 Low" Medium0 High" PWR All Steel 2.50e+06 2.98e+05 7.56e+04 4.12e+04 1.81e+05 Carbon Steel 9.35e+05 1.426+06 2.36e+06 2.386+05 6.05e+04 3.296+04 1.456+05 Stainless* 1.506+05 5.96e+04 1.516+04 8.236+03 3.626+04 BWR All Steel 1.24e+06 2.896+05 9.87e+04 1.35e+05 5.586+04 Carbon Steel 6.16e+05 5.486+05 1.16e+06 2.316+05 7.906+04 1.086+05 4.46e+04 Stainless 7.196+04 5.78e+04 1.97e+04 2.696+04 1.126+04 Total Industry All Steel 3.74e+06 5.876+05 1.74e+05 1.766+05 2.376+05 Carbon Steel 1.55e+06 1.976+06 3.52e+06 4.696+05 1.39e+05 1.41e+05 1.89e+05 Stainless 2.22e+05 1.17e+05 3.496+04 3.526+04 4.736+04 * Although data for stainless steel and carbon steel are presented as independent quantities, a significant fraction of stainless steel is unlikely to be segregated as such for the purpose of clearance. Contaminated steel that can be potentially decontaminated Low-level contamination: <105 dpm/100 cm2 Medium-level contamination: 105 — 107 dpm/100 cm2 High-level contamination: >107 dpm/100 cm 2 The total surface area of all potentially contaminated, recyclable carbon steel scrap was determined by taking the sum of the areas of all the inner surfaces of the contaminated components of the Reference BWR and using a scaling factor (based on the reactor's power rating) to determine the area of each actual BWR. A similar procedure was used to determine the contaminated surface areas of PWRs. The average mass thickness—the sum of the areas of all the components in all the commercial power reactors in the United States, divided by the total mass of the contaminated, recyclable carbon steel scrap that could be obtained from these reactors—is 4.79 g/cm2 (see Section A.5.2.3). Assuming a density of 7.86 g/cm3 (the density of plain carbon steel [AISI-SAE 1020]), this corresponds to an average thickness of about 0.61 cm (0.24 inches). 4.2.2 Contaminated Metal Inventories Other Than Steel There are significant quantities of metals and metal alloys other than steel that may be suitable for recycling, including: (1) galvanized iron, (2) copper, (3) Inconel, (4) lead, (5) bronze, (6) aluminum, (7) brass, (8) nickel, and (9) silver. However, there exist no credible data in the open 4-22 ------- literature regarding the estimated fractions of these metal inventories that are likely to be contaminated or the extent of their contamination. In the absence of reported data, a reasonable approach is to assume that the contaminated fraction of each of these metals is the same as the contaminated fraction of carbon steel for the Reference BWR and Reference PWR. Justification for this modeling approach is based on the fact that most of these metals exist as sub-components of larger items consisting primarily of carbon steel. From data cited in Appendix A, the ratio of contaminated carbon steel suitable for recycling to that of total plant inventory corresponds to 20% and 10% for the Reference BWR and the Reference PWR, respectively. Applying these values to other metals yields the quantities of recyclable, contaminated metal listed in Table 4-8. 4.2.3 Timetable for the Availability of Scrap Metal from Decommissioning The currently operating nuclear power plants are assumed to have an operating life of 40 years, plus any renewals that have been approved by the NRC. It was assumed for the purpose of this analysis that releases of scrap metal would take place ten years following reactor shutdown. Thus, for an operating reactor, the earliest date for releasing scrap metal is assumed to be 50 years after startup. As noted previously, there are also 27 reactors which were formerly licensed to operate. Some of these have been placed in an ENTOMB status, some have been or are currently being decommissioned under the DECON option, some have elected DECON but have not commenced decommissioning, and some are in a SAFSTOR status. Only reactors which are slated for DECON or which are in a SAFSTOR status are included in this analysis (see Appendix Al). It is assumed that reactors in SAFSTOR would retain that status for 50 years, with releases of scrap metal taking place ten years later. Table 4-9 summarizes the potential availability of scrap metal, starting with the year 2006, and lists all years during which releases are anticipated. The actual release dates of scrap metal may be later than those listed. First, as mentioned on page 4-19, a number of reactors may receive 20-year extensions to their operating licenses, thereby delaying the projected date of decommissioining. Second, many, if not most, facilities are likely to elect the SAFSTOR decommissioning alternative, thereby delaying releases for up to 50 years. 4.3 RECENT RECYCLING ACTIVITIES (1995 - 1998) This section briefly summarizes recent scrap metal recycling activities involving scrap from both commercial and government sources. The objective is to provide illustrative information rather than an exhaustive analysis. It should be emphasized that several of the activities described 4-23 ------- below involved recycle and reuse within the DOE complex rather than free release into normal commercial channels for scrap metal processing. Table 4-8. Contaminated Metal Other than Steel Potentially Suitable for Clearance (t) Metal Galvanized Iron Copper Inconel Lead Bronze Aluminum Brass Nickel Silver Reference Facility BWR 260 138 24 9.2 5.0 3.6 2.0 0.2 <0.2 PWR 130 69 12 4.6 2.5 1.8 1.0 0.1 <0.1 Industry All BWRs 8,904 4,726 822 315 171 123 68 7 <7 All PWRs 9,354 4,965 863 331 180 130 72 7 <7 Total 18,258 9,691 1,685 646 351 253 140 14 <14 4.3.1 DOE Materials National Center of Excellence for Metals Recycling (CEMR) The National Center of Excellence for Metals Recycling was established by DOE at Oak Ridge, Tennessee in October 1997 (AMM 1998). Recent activities of the Center for Excellence are listed below (Bishop 1999).4 Weldon Spring Site Remedial Action Project. Two hundred eighteen tonnes of suspected radioactive scrap metals were recycled with a cost avoidance to DOE of $336,000 in FY1998. ETTP Recycle of Metal Pallets. The East Tennessee Technology Park (ETTP) surveyed and sold 1200 pallets through a public offering. The total mass of the 1200 pallets was 2441. The associated cost avoidance to DOE in FY1998 was estimated at$912,638. See Note 1. 4-24 ------- Table 4-9. Anticipated Releases of Scrap Metals from Nuclear Power Plants (t) Year 2006 2007 2016 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2043 2044 2045 2046 2047 2049 2052 2056 2057 2058 Total3 § o> .§£ ioW o 4,906 1,169 5,683 1 1 ,522 9,111 8,372 26,266 31,573 52,479 6,252 24,978 9,844 1 1 ,922 10,202 13,527 32,775 20,675 34,307 27,206 46,335 17,730 6,229 13,847 3,634 9,556 5,896 3,564 2,947 917 2,928 1,809 3,255 3,255 4,820 469,490 tO tO — CD CD c £ 'eo W CO 1,227 292 1,421 2,881 2,278 2,093 6,568 7,894 13,122 1,563 6,245 2,461 2,981 2,551 3,382 8,195 5,170 8,578 6,802 11,585 4,433 1,558 3,462 909 2,389 1,474 891 737 229 732 452 81,414 814 1,205 117,389 •a CD N C c o >- CO O 195 45 217 444 355 324 1,012 1,232 2,023 248 973 390 465 405 537 1,268 800 1,340 1,062 1,797 704 244 539 144 380 234 142 117 35 116 72 129 129 184 18,304 s_ CD a. a. o O 103 24 115 235 189 172 537 654 1,074 132 517 207 247 215 285 673 425 711 564 954 374 129 286 77 201 124 75 62 19 62 38 69 69 98 9,715 Inconel 18 4 20 41 33 30 93 114 187 23 90 36 43 37 50 117 74 124 98 166 65 23 50 13 35 22 13 11 3.2 11 6.6 12 12 17 1,690 •a CO CD 6.9 1.6 7.7 16 13 11 36 44 72 8.8 34 14 16 14 19 45 28 47 38 64 25 8.6 19 5.1 13 8.3 5.0 4.1 1.2 4.1 2.5 4.6 4.6 6.5 648 CD N O m 3.7 0.86 4.2 8.5 6.8 6.2 19 24 39 4.8 19 7.5 8.9 7.8 10 24 15 26 20 35 14 4.7 10 2.8 7.3 4.5 2.7 2.3 0.67 2.2 1.4 2.5 2.5 3.5 352 Aluminum 2.7 0.62 3.0 6.1 4.9 4.5 14 17 28 3.4 13 5.4 6.4 5.6 7.4 18 11 19 15 25 9.8 3.4 7.5 2.0 5.3 3.2 2.0 1.6 0.49 1.6 0.99 1.8 1.8 2.6 253 to to 2. m 1.5 0.34 1.7 3.4 2.7 2.5 7.8 9.5 16 1.9 7.5 3.0 3.6 3.1 4.1 9.8 6.2 10 8.2 14 5.4 1.9 4.1 1.1 2.9 1.8 1.1 0.90 0.27 0.89 0.55 0.99 1.0 1.4 141 CD _*: o Z 0.15 0.034 0.17 0.34 0.27 0.25 0.78 0.95 1.6 0.19 0.75 0.30 0.36 0.31 0.41 0.98 0.62 1.0 0.82 1.4 0.54 0.19 0.41 0.11 0.29 0.18 0.11 0.090 0.027 0.089 0.055 0.10 0.10 0.14 14 Note: Adapted from Table A-84 a Totals may differ from sum of listed amounts due to roundoff. 4-25 ------- ORNL Tower Shielding Facility Clean Material Recycle Clean material sold for recycle/ reuse included 30 tons of aluminum, 50 tons of steel, 5 tons of graphite, 40 tons of lead, 85 tons of miscellaneous metal, and 305 tons of concrete. Approximately 30 tons of concrete and 3 tons of activated stainless steel were transferred to the High Flux Isotope Reactor facility for reuse. Total DOE project waste avoidance was 497.21, with a cost avoidance of $2,766,000 in FY1998. Sale of LLW Drums. In FY1998, DOE processed the LLW contained in a number of drums. Since the empty drums were contaminated, they where sold to a commercial vendor for like use (i.e. supercompaction of LLW). The total project waste avoidance to DOE was 54 t and the cumulative cost avoidance to DOE and industry was$178,000. B-25 Boxes. In FY1998, 35 boxes have been shipped to ETTP for reuse on a re-industrialization project. This represents a waste and cost avoidance of 13 t and $10,500 to DOE. ETTP Three Building D&D and Recycling Project BNFL Inc. was awarded a$238 million fixed price contract on 25 August 1997 to deliver vacant and decontaminated buildings (K-29, K-31, and K-33) to DOE/ORO. The $238 million contract cost included a credit back to DOE of$55,569,748 for the recyclable material. This amounts to quarterly cost savings of $2,646,178 over 21 quarters for the materials recycled or reused. The recycling activities began in the fourth quarter of CY1998 and were scheduled to continue throughout the duration of the contract (but see Note 1). The scheduled end date is 31 December 2003. The following materials were recycled in the fourth calendar year quarter of 1998 for a cost savings for this quarter of$2,646,178: • Lube Oil, Hazardous, 83 t • Transformers, MLLW, 119 t • Scrap Metal, LLW, 395 t Approximately 117,162 t of material were to be recycled from the three buildings, including 70,232 t from K-33, 12,138 t from K-29, and 34,792 t from K-31. ETTP K-31 & K-33 Switchyard. DOE has elected to fund Option I under the BNFL ETTP Three-Building D&D and Recycle Project. The equipment removal activities also included the disposition of the equipment as salvage/recycle materials and the disposal of all waste. The 4-26 ------- switchyard materials and equipment are non-radioactive. The estimated total mass of all equipment and materials awaiting disposition is 3,673 t. The dismantlement work began July 14, 1998. Total project savings are estimated at \$1,103,833. As of December 1998, 1,049 t of clean scrap metal from the ETTP Switch Yard had been recycled. SEG Bear Creek Facility. In 1996 INEEL shipped about 46,000 Ib (-211) of radiologically contaminated scrap to SEG for melting and beneficial reuse (INEEL 1997). The INEEL material was scheduled to be remelted into shielding blocks for use at LANL. The slag was to be returned to INEEL for disposal. In April 1997, GTS Duratek acquired SEG and announced in June of that year that staff reductions would be made (GTS 1997b). They noted that the flow of contaminated material to the Metal Melt Facility was neither sufficient nor steady enough to maintain continuous operations. GTS Duratek notes that the SEG facility (a 20-ton, 7,200 kW electric induction furnace) is the largest low-level radioactive metal furnace in the United States and the only one capable of making 10-ton shield blocks for DOE. Since 1992, SEG has converted over 60 million pounds of metal into shield blocks, each weighing 1-10 tons, for use at DOE laboratories (GTS 1997a). Other Activities. Approximately 26,000 Ib (-12 t) of slightly contaminated lead from INEEL was mixed with other metal provided by Lockheed Martin Energy Systems and used to manufacture ten lead-lined shielded storage containers at Manufacturing Sciences Corporation in Oak Ridge, Tenn. The storage containers are being used at the INEEL Radioactive Waste Management Complex to store remote-handled TRU waste (INEEL 1998). 4.3.2 Activities of Members of the Association of Radioactive Metal Recyclers (ARMR) ARMR member companies are responsible for the great majority (over 80%) of residually radioactive scrap metals in the United States that are either recycled or reused, in accordance with established NRC/DOE/State guidelines. Activities of the ARMR between 1995 and 1998 are summarized below (Loiselle 1999). 1995 About 15,0001 of RSM were surveyed and then either free-released or melted into shield blocks. The split was approximately one-half for each path (release or melt). Approximately 6,000 t of this metal originated in commercial nuclear utilities, 4-27 ------- another 6,0001 from the DOD. (The latter metal made into shield blocks). The remainder was from DOE. 1996 About 13,0001 were surveyed and then either free released, melted into shield blocks, or used to fabricate boxes and drums for restricted uses. Approximately 6,000 t, which came from the DOD, were made into shield blocks, 700 t from DOE were converted into restricted use boxes and drums, and most of the remainder was from the nuclear utilities. 1997 About 9,000 t from nuclear utilities were surveyed and free released. During this year, DOE did not release any metals to ARMR members, and no metal melting was required. 1998 About 20,0001 were surveyed and 17,000 t were free-released. The remaining 3,000 t were DOD metals that were melted into shield blocks. Approximately 10,000 t of the 17,0001 of the free-released scrap metal came out of the BNFL Three Building Project. The remainder was from nuclear utilities. 4-28 ------- REFERENCES AMM. 1998. American Metal Market., p. 6, December, 1998. Bechtel Jacobs Company. 2001. "Portsmouth Gaseous Diffusion Plant. (27 June 2001) Bishop, L., (U.S. Department of Energy). 1999. Private communication. Floyd, D. R. 1994. "National Conversion Pilot Project, Stage I, Preliminary Market Analysis Report." Rev. 1. Manufacturing Sciences Corporation, prepared for U.S. Department of Energy, Rocky Flats Office, Golden, CO. Geiger, G. 2001. Presentation to The National Academies/National Research Council Committee on Alternatives for Controlling the Release of Solid Materials from Nuclear Regulatory Commission-Licensed Facilities March 26-28, 2001, Washington, DC. GTSDuratek. 1997a. InSite. Vol. 15, 1st Quarter. GTS Duratek. GTS Duratek. 1997b. "GTS Duratek Announces a Consolidation of Operations at the SEG Bear Creek Facility." Press release, 30 June 1997. Idaho National Engineering and Environmental Laboratory (INEEL). 1997. "Citizen's Guide - A Supplement to the INEEL Reporter." (17 December 1998). Konzek, G. J., et al. 1995. "Revised Analyses of Decommissioning for the Reference Pressurized Water Reactor Power Station," NUREG/CR-5884, PNL-8742. Vol. 1, "Main Report." Pacific Northwest Laboratory prepared for the U.S. Nuclear Regulatory Commission, Washington, DC. Loiselle, V., (Chairman, Association of Radioactive Metal Recyclers). 1999. Private communication. Los Alamos National Laboratory (LANL). 1996. "Los Alamos National Laboratory (LANL) Metal Inventory." Oak, H. D., et al. 1980. "Technology, Safety and Costs of Decommissioning a Reference Boiling Water Reactor Power Station," NUREG/CR-0672. Vol. 2, "Appendices." Pacific Northwest Laboratory, prepared for the U.S. Nuclear Regulatory Commission, Washington, DC. 4-29 ------- Parsons Engineering Services, Inc., RMI Environmental Services, and U.S. Steel Facilities Redeployment Group. 1995. "U.S. Department of Energy, Scrap Metal Inventory Report for the Office of Technology Development, Office of Environmental Management," DOE/HWP- 167. Prepared for Hazardous Waste Remedial Actions Program, Environmental Management and Enrichment Facilities, Oak Ridge, TN. Person, G. A., et al. 1995. "Gaseous Diffusion Facilities Decontamination and Decommissioning Estimate Report," rev. 2, ES/ER/TM-171. Environmental Restoration Division, Oak Ridge National Laboratory, Oak Ridge, TN. Person, G. A., (Lockheed Martin Energy Systems, Inc.). 1996. Personal communication. S. Cohen & Associates (SCA). 1995a. "Scrap Metal Inventories at U.S. Nuclear Facilities Potentially Suitable for Recycling." Prepared for U.S. Environmental Protection Agency, Office of Radiation and Indoor Air, Washington, DC. S. Cohen & Associates (SCA). 1995b. "Analysis of the Potential Recycling of Department of Energy Radioactive Scrap Metal." Prepared for U.S. Environmental Protection Agency, Office of Radiation and Indoor Air, Washington, DC. Smith, R.I., G. J. Konzek, and W. E. Kennedy, Jr. 1978. "Technology, Safety and Costs of Decommissioning a Reference Pressurized Water Reactor Power Station," NUREG/CR- 0130. Vol. 1. Pacific Northwest Laboratory, prepared for the U.S. Nuclear Regulatory Commission, Washington, DC. Smith, R. I, et al. 1996. "Revised Analyses of Decommissioning for the Reference Boiling Water Reactor Power Station," NUREG/CR-6174, PNL-9975. Vol. 2. Pacific Northwest Laboratory, prepared for the U.S. Nuclear Regulatory Commission, Washington, DC. U.S. Department of Energy (U.S. DOE), Office of Environmental Management. 1995a. "Closing the Circle on the Splitting of the Atom." U.S. DOE, Washington, DC. U.S. Department of Energy (U.S. DOE), Office of Environmental Management. 1995b. "Scrap Metal and Equipment: Materials in Inventory." U.S. Department of Energy (U.S. DOE), Office of Environmental Management. 1996. "Taking Stock: A Look at the Opportunities and Challenges Posed by Inventories from the Cold War Era," DOE/EM-0275. Vol. 1, "A Report of The Materials in Inventory Initiative." 4-30 ------- Chapter 5 DESCRIPTION OF UNRESTRICTED RECYCLING OF CARBON STEEL Chapter 2 presented an overview of scrap metal operations in the United States The present chapter examines the recycling of carbon steel scrap in greater detail, with particular emphasis on those operations with the greatest potential for the radiation exposures of individuals. 5.1 RECYCLING SCRAP STEEL—AN OVERVIEW Figure 5-1 presents a simplified schematic diagram of some of the steps that would be involved in recycling carbon steel scrap into consumer or industrial products; this diagram is intended to present an outline of one possible set of recycling scenarios, which are discussed in this chapter. To preserve clarity, some of the scenarios addressed by the radiological assessment are not illustrated. For the sake of completeness, the following discussion makes note of additional or alternative steps in the recycling process which are not shown in the diagram nor addressed in the analysis. All references to these steps are enclosed by square brackets. The process starts with radioactively contaminated steel scrap that is already stored in scrap piles at various DOE and perhaps at NRC-licensed facilities, or that will be generated in the course of the decommissioning of such facilities. [Smaller amounts of scrap are also generated during the normal operations of these facilities.] After initial decontamination to meet ALARA requirements, the scrap is surveyed to determine if it is a reasonable candidate for clearance. Scrap that does not satisfy a putative clearance criterion and that cannot be economically decontaminated to achieve such a criterion is disposed of as low-level radioactive waste. The remaining scrap is decontaminated as required and cleared for release; it is then loaded onto trucks [or rail cars] to be transported off site. As indicated on Figure 5-1, during these operations the material is under regulatory control. The tasks are performed by radiation workers, who are subject to DOE- or NRC-regulated exposure limits and ALARA procedures. Therefore, these operations are not addressed by the present analysis. 5-1 ------- Scrap Piles Survey Regulatory Control Dispose as LLRW Furnace Charge Scrap to Furnace Steel Dust Interim Products Continuous Caster Sheet Metal Unloading jttii T Loading Scrap Processor Release to Atmopshere I HTMR Processing Slag Processor I Road Building Groundwater User Products Kitchen Range Figure 5-1. Operations Analyzed in the Carbon Steel Recycle Analysis (Operations indicated by dark shaded boxes in the diagram are not modeled) 5-2 ------- The scrap is transported to a processor where it is unloaded, sorted and possibly cut up or compacted. [Alternatively, it may be processed in a scrap staging area at or near the facility where the scrap was generated.] The processed scrap is transported to a steel mill where it may be [unloaded to a scrap pile or] sent directly to the furnace. In either case, it is loaded into a charging bucket and charged to an electric arc furnace (EAF), where it is melted. Certain constituents of the furnace charge are either vaporized or entrained in the offgas as particulate matter. Most of these emissions are captured by the emission control system. Once inside the system, the fumes are routed to the baghouse, where they are cooled and filtered. The filters, which are in the form of long bags, are periodically emptied by remotely operated mechanical means. The dust is transferred to a tanker truck and shipped off site. After the scrap is melted, first the slag and then the molten steel are poured into separate ladles. The molten steel is transferred from the ladle to a tundish from which it is fed to a continuous caster, where it is made into slabs. These may be sold as such or made into interim mill products, such as coils of sheet metal. The sheet metal may be made into consumer products, such as a kitchen range. The slag is transported to a slag pile at the steel mill, where it is stored prior to shipment to a slag processing facility. The slag processor sells the slag for various uses, such as ballast for road- building or aggregate which is mixed with cement and used for paving. While the slag is stored at the mill, various components could leach out and percolate through the soil to an underlying aquifer, possibly contaminating an underground source of drinking water. This list of scenarios, which follows the scrap from the generating facility to a specific consumer product, is only one of an endless set of possible variations in the process of recycling steel scrap. Not all possible scenarios could be analyzed, and, as stated earlier, not all scenarios that were analyzed are shown in Figure 5-1. 5.2 REFERENCE FACILITY In the United States, most steel scrap is melted in either an EAF or a basic oxygen furnace (EOF). The charge for an EAF usually consists entirely of scrap, while scrap makes up less than 30% of the feedstock of a BOF, the rest being the pig iron output of a blast furnace. A steel mill equipped with EAFs was therefore selected as the reference mill for the present study. 5-3 ------- The reference steel mill for the present analysis was based partly on the Calumet Steel Co. facility in Chicago Heights, 111., which is described in greater detail in Appendix G. The mill is equipped with two EAFs, each of which has a 12.5-foot diameter shell and produces a 32-ton (29 t) average heat, with a nominal capacity of 75,000 tons (-68,000 t) per year. Other parameters used in the analysis are based on data pertaining to other facilities, on engineering judgement and on analytical assumptions. Thus, the reference steel mill is a hypothetical construct. This analysis should not be construed to predict that radioactively contaminated steel scrap will, in fact, be processed at any specific facility. 5.3 EXPOSURE PATHWAYS The exposure pathways considered in the present analysis fall into two general groups: external exposure to direct penetrating radiation and internal exposure from inhaled or ingested radionuclides. 5.3.1 External Exposure The external exposures are evaluated by use of the MicroShield computer code, which is described in more detail in Section 6.3.1, or by dose coefficients adapted from Federal Guidance Report (FGR) No. 12 (Eckerman and Ryman 1993). 5.3.2 Internal Exposure The internal exposure pathways consist of the inhalation of radioactively contaminated dust, the inadvertent ingestion of contaminated dust, soot or other loose, finely divided material, and the ingestion of contaminated food or water. The following sections describe the geometries and the materials used to model the external exposure from each task, as well as the assumptions regarding the inhalation and ingestion pathways. A detailed discussion of the last two pathways appears in Sections 6.3.2 and 6.3.3. 5.4 LIST OF OPERATIONS AND EXPOSURE SCENARIOS Table 5-1 lists the operations and exposure parameters employed in the assessment of radiological impacts of recycling residually radioactive carbon steel scrap on exposed individuals. These operations and the parameters used to model the corresponding exposure 5-4 ------- scenarios are partially based on an earlier EPA-sponsored study of the recycling of DOE scrap metal (SCA 1995). That study included over 60 exposure scenarios, which were based on studies done by the International Atomic Energy Agency (IAEA 1991) and for the NRC (O'Donnell et al. 1978), as well as on visits to steel mills and scrap processors and private communications with staff members of these facilities. The present analysis incorporates those operations shown to have the maximum potential impacts on the exposed individuals. This study included additional visits to steel mills and scrap processors, further communications with steel industry personnel, and additional research into scrap metal recycling practices. In addition to reviewing published and draft reports, valuable information and insights were obtained by close collaboration and consultation with other organizations which were also investigating the radiological consequences of the clearance of metals and other materials, including NRC, DOE, the European Commission, and the IAEA. As seen in Table 5-1, the study addresses the radiation exposures from several representative finished products which might be made from recycled steel scrap1. These products were selected on the basis of their wide use and their potential radiological impacts on individuals—they are comparable to the finished products in the earlier studies. For many radionuclides, the impacts on end users would be dominated by exposure to external radiation. Therefore, the highest impacts would be produced by massive products that are in close proximity to the exposed individuals for the longest times. Cooking utensils were included to assess radiation exposures from consumption of food potentially contaminated by radionuclides leached from the metal during cooking. Three of these products are made from cast iron, which is produced by a different process than is used to make carbon steel. Since the radiological impacts of iron founding are not included in the present study, these products are not represented in Figure 5-1. However, the contaminant distributions characteristic of cast iron are utilized in the impact assessment of these products (see Section 6.2 and Appendix F). 5-5 ------- Table 5-1. Exposure Scenarios and Parameters for Radiological Assessments of Individuals Description SCRAP TRANSPORT: Truck Driver SCRAP PROCESSING: Cutter STEEL MILL Furnace Operations Crane Operator EAF furnace operator Airborne effluent emissions Interim Products Operator of continuous caster Baghouse Baghouse maintenance Truck driver: baghouse dust Slag Pile Slag pile worker Slag leachate in groundwater PROCESSING EAF DUST s_ O B £ c O '-4— ' _2 Q 0.2 0.055 0.055 0.005 Exposure Pathways External Exposure Time (hr/y) 1000 1750 8 c 2 t/> b 8ft N/Ab Medium scrap scrap Internal Time (hr/y) Medium Dust load (mg/m3) RFa N/A 1500 10 0.5 1750 1750 10m 4-30 ftd scrap 1750 1750 dustc 1.3 2.2 0.58 N/A 1750 2-1 5 ft steel 1750 dust 2.0 0.58 1450 40 250 1000 10m e e 3.5 m scrap dust 1450 40 250 dust 2.3 50 1.2 0.58 0.0076' 0.58 N/A 1000 N/Ab slag 1750 slag 2.6 0.51 N/A 1000 N/Ab dust 1750 dust 10 0.5 INDUSTRIAL USE OF MILL PRODUCTS Using slag in road construction Assembling automobile engines Manufacturing industrial lathes END USERS Using kitchen range Sailor on naval support vessel Taxi driver Lathe operator Cooking in cast iron pan 0.055 0.5 140 1750 1750 1 m 20-70 cm 20-70 cm slag cast Fe 1750 slag 2.6 0.51 N/A 1750 cast Fe 2.7 0.5 525 2000 3300 1750 263 2ft 1.5ft 2ft 20-70 cm 2ft steel cast Fe N/A N/Ag cast Fe N/A Respirable fraction Exposure assessment uses FGR 12 dose coefficients—see discussion in Section 6.3.1 Dust = baghouse dust Range of distances—see discussion in Section 6.3.1 Special model—see discussion in Appendix H Includes respiratory protection factor of 100 s Exposure from ingestion of contaminated food 5-6 ------- 5.4.1 Dilution Factors Potentially contaminated scrap would in most cases be diluted with scrap that had never been exposed to radioactive contaminants. The ratio of the potentially contaminated scrap to the total amount of metal—termed the dilution factor2—is listed for each of the four major groups of operations shown in bold-faced upper-case type in Table 5-1. A detailed discussion of the dilution of scrap steel is presented in Appendix G. The section summarizes the application of these concepts to the scenarios presented in Table 5-1. Scrap Transport A truck driver could spend an entire year transporting the recyclable scrap metal from one large boiling-water reactor (BWR) power plant. However, as noted in Appendix G, only about 20% of such scrap would be potentially contaminated. Therefore, the average specific activity of any given radionuclide in the scrap being transported during the course of the year would be only 20% of its average specific activity in the potentially contaminated scrap. The dilution factor for the scrap transport operation is thus 0.2. Scrap Processing All the recyclable scrap metal from a BWR could plausibly be sent to a single scrap processor in a single year, resulting in the same dilution factor (0.2) as for the scrap transport operation, assuming that the processor accepts no scrap from other sources during that year. However, as discussed in Appendix G, the postulated decommissioning of the largest operating BWR power plant would only yield a total of about 38,0001 of carbon steel scrap—both clean and contaminated. The massive mountains of scrap which characterize this scenario, described later in this chapter, were observed at a facility that processed 50,000 tons (-45,0001) per month—the exposure potential would most likely be much less at the smaller processor. If the approximately 7,500 t of potentially contaminated BWR scrap were sent to the larger facility, the dilution factor would be -0.014. Although the actual size of the facility that would yield the reasonable maximum exposure is unknown, we can make a reasonable estimate of the dilution factor by taking the geometric mean of these two extreme values. The calculated value (-0.053) was rounded up to 0.055, which is the calculated dilution factor for the steel mill (see below). The term "dilution factor" is potentially confusing: the greater the dilution, the smaller the dilution factor. Thus, a dilution factor of one means that there is no dilution, a dilution factor of zero corresponds to infinite dilution. The 1997 draft TSD used the term differently. 5-7 ------- Steel Mill The two EAFs at the reference steel mill have a combined nominal capacity of 150,000 tons (136,000 t) per year. The 7,500 t of potentially contaminated carbon steel scrap from the postulated decommissioning of the BWR power plant would equal 5.5% of the mill's annual capacity. The dilution factor for the steel mill operations is therefore 0.055. Processing EAF Dust All of the contaminated baghouse dust is assumed to be processed at a high temperature pyrometallurgical metals recovery (HTMR) plant owned and operated by the Horsehead Resource Development Company (HRDC). The dilution factor for this scenario was calculated by comparing the processing capacities of the three HRDC facilities with the anticipated year-by- year releases from nuclear power plants in their respective service areas3. The largest dilution factor—leading to the greatest radiological impacts (see Note 2)—would occur at the HRDC facility in Rockwood, Tenn. in 2024, when 24,000 t of potentially contaminated carbon steel scrap would be released in the area comprising the Southeastern United States (NRC Region II) plus the states of Arkansas, Louisiana and Texas. As described in Appendix G, the melt- refining of this scrap would generate 380 t of baghouse dust. If all this dust were processed in the Rockwood facility, it would consume about 0.5% of the plant's annual capacity, yielding a dilution factor of 0.005 for this scenario. The processing of residually radioactive EAF dust at this facility is part of a hypothetical scenario constructed for the present analysis. The analysis should not be construed to predict that radioactively contaminated dust will, in fact, be processed at any specific facility. Industrial Use of Mill Products It was assumed that the three industrial operations using mill products modeled in this analysis obtained all their materials from the reference mill. Thus, the materials are assigned the same dilution factor as the steel mill operations. End Users Any one item could be made from a single heat which could contain a higher-than-average fraction of residually contaminated scrap. However, as discussed in Appendix G, because each heat is made up of scrap from a number of different sources, the probability that all of the scrap For the sake of simplicity, the carbon steel scrap from each nuclear power plant was assigned to the nearest HRDC facility, as determined by measuring point-to-point distances. 5-8 ------- in any one heat would come from a nuclear facility is vanishingly small. The analysis presented in Section G.4.3 showed that the maximum likely fraction of contaminated scrap in any single heat during the course of one year is 50%; therefore, the dilution factor for the finished product in the end-user scenarios is O.5.4 5.4.2 Scrap Transport The scrap transport worker is a truck driver who spends eight hours per day in the cab of a truck, carrying 20-t loads of scrap metal to the scrap processor and returning with an empty truck (or carrying other cargo). His only exposure would be to external radiation from the load of contaminated scrap. 5.4.3 Scrap Processing Operations Assessments were performed on a worker who spends six hours per day cutting the scrap, but spends a total of seven hours in canyons surrounded by scrap. He would also inhale and ingest dust which is assumed to have the same specific activity as the scrap. 5.4.4 Steel Mill Most mills that process scrap metal receive the scrap via truck or rail. Upon arrival at the mill, the scrap is unloaded, charged into an EAF and melted.5 Furnace Operations The scrap is unloaded by means of a large electromagnet and dumped into charging buckets that move the scrap to the furnace. The exposures of two workers performing representative tasks involved with furnace operations are assessed in the present analysis. One is the crane operator who transfers the charging bucket—he would be exposed to external radiation from the scrap in the bucket. The other is the furnace operator, who would be exposed to radiation from the scrap in the furnace while it is melting. They both would be exposed to fugitive furnace emissions which escape capture by the emission control system. This represents the 90th percentile value—there is at most a 10% probability that any of the 2,000 heats melted during one year would have a greater fraction of contaminated scrap. As noted in Section 5.1, the alternate scenario, in which scrap is first unloaded to a scrap pile at the mill, is not included in the analysis. 5-9 ------- Interim Products Once the steel melts, it is poured onto a continuous caster. Torches cut the solidified steel into slabs as the metal runs down a set of rollers. Cooled slabs are stored, reheated and formed into products such as coils of sheet metal, or are sold in their raw form. The operator of the continuous caster would be exposed to external radiation from the molten steel in the tundish as well as from the continuous caster. He would also be exposed to fugitive furnace emissions. Baghouse The baghouse contains rows of filters, suspended from the ceiling, that trap the particulate emissions from the melt-refining process. These bag-like filters are shaken at frequent intervals; the dust settles into collecting hoppers and is fed by a screw mechanism into a tanker trailer. Once filled to capacity, the trailer is transported away from the steel mill to a processing facility for recovery of commercially valuable components, primarily metals such as zinc, cadmium and lead, and for ultimate disposal. Steel mill workers are occasionally assigned to spend a day repairing or changing the baghouse filters. Such a worker typically spends four to six hours6 in the midst of the suspended filters in the dust-laden atmosphere of the baghouse, wearing a respirator equipped with a full facepiece. At a typical facility, this procedure is carried out an average of seven times per year. The analysis assumes that the same worker is assigned to this task every time. While performing such maintenance, the worker would be exposed to external radiation from the residual dust that is retained in the filters after they are emptied, as well as to the dust that has settled on the floor of the baghouse. In addition, one worker typically spends about one hour per day monitoring the control mechanisms and performing maintenance that does not require entering the modules containing the filters. It is conservatively assumed that the same worker who maintains the filters would be assigned to this duty on days he was not inside the modules. The rest of the time, he would be assigned a variety of tasks in the steel mill. His external exposure rate during that time is Rest periods necessitated by work in a confined area and the need to don and remove protective clothing restrict the amount of time the worker can spend on this task. 5-10 ------- assumed to be the same as that of the crane operator.7 His internal exposure rate is assumed to be the average of workers inside a typical mill (see Appendix H). The driver of the tanker truck transporting the dust off site would be exposed to external radiation from the dust in the truck. Since the reference facility operating at full capacity produces about 2,250 tons of dust per year, a truck carrying 25 tons of dust would make 90 trips per year. Since many EAF mills are more than a one-day drive away from the nearest processing facility, transporting the dust could occupy at least one driver full-time. He is assumed to return with an empty trailer. Airborne Effluent Emissions Not all the fugitive furnace emissions are trapped in the baghouse. As noted on p. 5-9, some bypass the collection system, while others pass through the baghouse filters. In particular, radionuclides which form gases or volatile vapors would not be trapped by the filters. A family of subsistence farmers, who are postulated to live one km from the steel mill and who obtain a portion of their produce, meat and milk from their farm, would be exposed to these airborne effluent emissions. In most cases that are significant to the present analysis, the consumption of home-grown foods would constitute the primary exposure pathway. Slag Disposal After the completion of the melt cycle, the EAF is tilted and the slag is poured into a ladle, which is moved by overhead crane to a slag yard outside the building. A worker at a typical facility spends about half his time on a platform on the edge of the slag yard and would be exposed to external radiation from the slag. Since the rest of his time is in the vicinity of the slag, he would be exposed to slag dust during the course of the day. Since the slag pile is exposed to the elements, soluble components of the slag leach out of the matrix and percolate through the soil until they reach an underlying aquifer. (This process takes a number of years—see Section 6.4.1.) A nearby resident who gets his drinking water from a well that is downgradient from the slag pile might, at some time in the future, be exposed to contaminated groundwater. This worker was selected as having the median exposure rate to Co-60, one of the significant radionuclides in the present analysis. 5-11 ------- 5.4.5 Processing EAF Dust A worker at one of HRDC's HTMR plants who operates a front-end loader at the edge of a large pile of dust spends about four hours a day transporting dust from the piles to the conveyors. During this time he would be exposed to external radiation from the pile of dust. However, because the dust is assumed to contaminate the air and settle on accessible surfaces throughout the facility, he would be exposed to the inhalation and inadvertent ingestion of the dust throughout his workday. 5.4.6 Use of Steel Mill Products All products of the steel mill have industrial uses. The present analysis deals with two of these products—finished steel and slag—in addition to the reprocessing of baghouse dust, discussed in the preceding section. Slag As shown in Appendix I, slag is primarily used in road building, as fill or for soil conditioning. A worker employed in road construction would be exposed to external radiation from the slag in the roadbed as well as that in the cement pavement—he would also be exposed to contaminated slag dust. Steel Steel is used to make a virtually endless variety of finished products. The analysis considers the five categories of products which are listed below, along with an example of each category. These products also represent small, medium and large objects, as indicated below. • Large home appliance (medium-sized object): double oven • Automotive component (medium-sized object): engine block • Large industrial equipment (large object): 8-ton metal-working lathe • Cooking utensil (small object): frying pan • Shipbuilding (large extended object): hull plate on naval support vessel Only the oven and hull plate are made from carbon steel, however. The other three are made primarily of cast iron, which is produced by a different process. The radiation exposures of workers producing and assembling two of these products—engine blocks and industrial 5-12 ------- lathes—are assessed in the present analysis. In each case, the workers would be exposed to external radiation from the iron, which is assumed to have the same dilution factor as the steel. The grinding operations on the lathe bed would also expose the lathe maker to the inhalation and ingestion of iron dust. End Users of Finished Products The final group of exposed individuals are people who use the products listed in the previous section. One user of each of the five products, who was judged to be the RME individual for that product, is included in the analysis. A consumer would be exposed to external radiation from the steel in the kitchen range. A taxicab driver would be exposed to external radiation from the engine block, while a lathe operator would be exposed to radiation from the cast iron lathe bed. Another consumer cooking food in a cast iron frying pan would be exposed to external radiation from the cast iron, in addition to eating food which would be contaminated with residual radioactivity that has leached from the pan. Finally, a sailor on a naval support vessel would be exposed to external radiation from a hull plate next to his sleeping quarters. 5-13 ------- REFERENCES Eckerman, K. F., and J. C. Ryman. 1993. "External Exposure to Radionuclides in Air, Water, and Soil," Federal Guidance Report No. 12, EPA 402-R-93-081. U.S. Environmental Protection Agency, Washington, DC. International Atomic Energy Agency (IAEA). 1991. "Exemption Principles Applied to the Recycling and Reuse of Materials from Nuclear Facilities." Draft (unpublished). MicroShield, Ver. 4.2. Grove Engineering, Inc., Rockville, MD. O'Donnell, F. R., et al. 1978. "Potential Radiation Dose to Man from Recycle of Metals Reclaimed from a Decommissioned Nuclear Power Plant, " NUREG/CR-0134. Oak Ridge National Laboratory, Oak Ridge, TN. S. Cohen & Associates (SCA). 1995. "Analysis of the Potential Recycling of Department of Energy Radioactive Scrap Metal." 4 vols. Prepared for U.S. Environmental Protection Agency, Office of Radiation and Indoor Air, Washington, DC. 5-14 ------- Chapter 6 RADIOLOGICAL ASSESSMENT OF THE RECYCLING OF CARBON STEEL Chapter 5 presented the scenarios and modeling parameters used to assess the radiation exposures of individuals that result from recycling potentially contaminated steel scrap. Chapter 6 discusses how these scenarios are used to perform radiological assessments of these individuals1. For the sake of clarity in the presentation, the radioactively contaminated scrap dilution factor presented in Section 5.4.1 is not explicitly included in the present chapter. This factor is, however, incorporated in all the assessments. The dilution factor will be explicitly addressed in the discussion of results in Chapter 7. The concept of the RME individual is central to the assessment—it is discussed here in more detail. For a single exposure scenario and a given radionuclide, such as scrap contaminated with a strong y-emitting nuclide (e.g., cobalt-60 or cesium-137), the choice of the RME individual is relatively straightforward: it is the individual worker who spends the most time nearest to the scrap. For the entire life cycle of a given batch of scrap metal—from the time it leaves the custody of a licensed facility, is transported to a steel mill, is turned into sheet metal, is used to fabricate a kitchen range and finally is delivered to a home2—there may be several exposed individuals. Which is the RME individual is not obvious a priori. To determine who receives the highest exposure, the annual doses to the exposed individuals at each stage of production, transportation, distribution and storage, including the use of the finished product, are compared. The person with the highest dose rate would become the RME individual for a given radionuclide. A number of computer codes dealing with recycling and pathways analysis were reviewed for use in this study but none were found suitable. Initially, a series of computer spreadsheets was developed to perform the calculations described in this chapter. As the analysis progressed, the need for a single integrated computer program became evident. Such a program was developed for this analysis. The program was written in the Fortran 90 computer language and runs on an IBM-compatible personal computer. Although directed to steel, much of the discussion also applies to assessments of aluminum and copper, which are discussed in Chapters 8 and 9, respectively. This is but one example of the sequence of scenarios—the complete set of exposure scenarios was presented in Chapter 5. 6-1 ------- 6.1 RADIOACTIVE CONTAMINANTS The 40 individual radionuclides studied in this analysis were selected on the basis of a review of nine published reports which cast light on the nuclides most likely to be present in potentially contaminated steel scrap that may be a candidate for recycling. The selection process was briefly discussed in Chapter 3—a more detailed discussion is presented in Appendix D. Since a period of years is assumed to elapse between the time the metal was contaminated and the time it would be recycled, short-lived nuclides (i.e. those with half-lives of less than six months) would have decayed to insignificant levels and were therefore omitted from the present analysis. By the same token, short-lived progeny of long-lived parents are assumed to be in secular equilibrium with their parents at the time of recycling. All references to such parent nuclides in this report include the designation "+D" to indicate that the contributions of these implicit progeny are included in the calculated annual doses and risks, which are normalized to unit specific activity of the parent. The implicit progenies of all nuclides selected for the present analysis are listed in Table 6-1. The generation number indicates whether the progeny nuclide is first generation (1), second generation (2), etc. The analysis also addressed steel scrap contaminated with unique combinations of radionuclides, including long-lived members of natural decay series in secular equilibrium with their parents. These include: (1) "U-separated"—the three naturally occurring uranium isotopes (in secular equilibrium with their short-lived progenies but separated from their long-lived progenies) in the ratios of their natural abundances; (2) "U-depleted"—the same isotopes in ratios characteristic of depleted uranium;3 (3) "U-natural"—natural uranium in secular equilibrium with the entire U-238 and U-235 radioactive decay series, and (4) "Th-series"—Th-232 in secular equilibrium with its entire decay series. The calculated radiological impacts of the mixtures of uranium isotopes, as well as the impacts of the uranium series, are normalized to unit activities of U-238, while the impacts of the thorium series are normalized to unit activities of Th-232. The nuclides included in each of these groupings are listed in Table 6-2. Depleted uranium is a byproduct of the uranium enrichment process and contains reduced activities of U-234 and U-235. 6-2 ------- Table 6-1. Implicit Progenies of Nuclides Selected for Analysis Parent Nuclide Sr-90 Ru-106 Ag-llOm Sb-125 Cs-137 Ce-144 Pb-210 Ra-226 Ra-228 Ac-227 Th-228 Half-Life (y) 28.8 1.02 0.684 2.77 30.1 0.781 22.3 1600 5.75 21.8 1.91 y Radiation P P P,Y,e- P,Y,e- P P,Y,e- P,Y,e- a,Y,e' P a,P,Y,e- a,Y,e' Progeny Generation 1 1 1 1 1 1 1 1 2 1 2 3 4 5 1 1 1 2 3 4 5 6 7 7 1 2 3 4 5 6 6 Nuclide Y-90 Rh-106 Ag-110 Te-125m Ba-137m Pr-144m Pr-144 Bi-210 Po-210 Rn-222 Po-218 Pb-214 Bi-214 Po-214 Ac-228 Fr-223 Th-227 Ra-223 Rn-219 Po-215 Pb-211 Bi-211 Tl-207 Po-211 Ra-224 Rn-220 Po-216 Pb-212 Bi-212 Tl-208 Po-212 Branching Ratio 100 100 1.4 22.8 94.6 1.43 98.6 100 100 100 100 99.98 100 99.97 100 1.38 98.6 100 100 100 100 100 99.7 0.27 100 100 100 100 100 35.9 64.1 Half-Life 2.67 d 29.9s 24.6s 58 d 2.55m 7.2m 17.3m 5.01 d 138 d 3.82d 3.04m 26.9m 19.7m 164 |ls 6.13 h 22.0m 18. 7 d 11.4d 3.96s 1.78ms 36.1 m 2.14m 4.77m 0.516s 3.66d 55.6s 0.145 s 10.6 h 1.01 h 3.05m 298ns Radiation P P,Y P,Y Y,e Y,e Y,e P,Y P a a,Y a P,Y,e- P,Y,e- a,Y P,Y,e- P,Y,e- a,Y,e a,Y,e a,Y,e a,Y P,Y,e- a,Y,e- P,Y a,Y a,Y,e- a,Y a P,Y,e- a,P,Y,e P,Y,e- a Note: Only progenies with half-lives of six months or less are included in the implicit progeny of "+D" nuclides. 6-3 ------- Table 6-1 (continued) Parent Nuclide Th-229 U-235 U-238 Np-237 Pu-241 Half-Life (y) 7340 y 7.04e+08 y 4.47e+09 y 2.14e+06y 14.4 y Radiation a,Y,e' a,Y,e' a,e" a,Y,e' P Progeny Generation 1 2 3 4 5 6 6 7 1 1 2 3 1 1 Nuclide Ra-225 Ac-225 Fr-221 At-217 Bi-213 Tl-209 Po-213 Pb-209 Th-231 Th-234 Pa-234m Pa-234 Pa-233 Am-241 Branching Ratio 100 100 100 100 100 2.16 97.8 100 100 100 100 0.16 100 100 Half-Life 14.9 d 10.0 d 4.8m 32.3 ms 45.6m 2.16m 4.20 |ls 3.25 h 1.003 d 24.1 d 1.17m 6.69 h 27.0 d 432 y Radiation P,Y,e- a,Y,e a,Y,e a,Y a,P,Y,e P,Y,e- a P P,Y,e- P,Y,e- P,Y,e- P,Y,e- P,Y,e- a,Y,e Note: Only progenies with half-lives of six months or less are included in the implicit progeny of "+D" nuclides. Because of the variability of contamination patterns and storage conditions, it cannot be assumed that radon isotopes would escape from the surface of the metal. The contamination might have been painted over, for instance, or trapped inside a steel component that was crushed as part of a volume reduction process. Therefore, the assessment of Ra-226+D assumes that Rn-222 and its short-lived daughter products would remain in the scrap in complete secular equilibrium with the radium, while that of natural uranium series assumes that both Rn-222 and Rn-219, as well as their entire progenies, would be in secular equilibrium with U-238 and U-235, respectively. Similarly, the assessments of the Th-232 series and of Th-228+D assume thatRn-220 would remain in the scrap. Except for the "+D" nuclides with their short-lived progeny and the natural uranium and thorium series, no ingrowth of progenies was modeled in the radiological assessment of carbon steel scrap. Exposures of scrap processors, steel mill workers, and handlers and users of slag are assumed to occur within a few months of the release of the scrap for recycling, too short a time for any significant ingrowth of long-lived progeny. Since the finished products that were among 6-4 ------- the subjects of the analysis have useful lives of several years, such ingrowth could potentially occur. However, as will be seen in the discussion of such exposure scenarios later in this chapter, this ingrowth would have no significant impact for the nuclides and the materials considered in the analysis. A similar conclusion applies to isotopes of elements that are released to the atmosphere or accumulate in the baghouse dust, since, as shown in Section 6.2.1, these radionuclides have no radioactive progenies. Table 6-2. Nuclides Included in Various Combinations and Decay Series Series U-Natural U- Separated U-Depleted Th-Series Nuclide U-238+D U-234 Th-230 Ra-226+D Pb-210+D U-235+D Pa-231 Ac-227+D U-238+D U-234 U-235+D U-238+D U-234 U-235+D Th-232 Ra-228+D Th-228+D Activity Fraction 1 1 1 1 1 0.047 0.047 0.047 1 1 0.047 1 0.09123 0.01594 1 1 1 6.2 SPECIFIC ACTIVITIES OF VARIOUS MEDIA When steel scrap is charged to an EAF, chemical agents (fluxes) are added to control the chemical properties of the molten metal. The interactions among the flux, the refractory brick which lines the furnace, and the molten metal affect the final composition of the melt and hence the distribution of radionuclides among the furnace products: the melt, the slag, and the offgas. The melt is cast into various shapes that become the primary output of the mill. Slag is material 6-5 ------- not remaining in the metal and includes the chemical agents, some of the liner material, and small amounts of the base metal, much of which is recovered and charged to the furnace for a subsequent melt. Offgas consists of the fumes and aerosols evolved during melting which are captured by the facility's emission control system. After cooling, most of the offgas (except gases and vapors) is collected in the baghouse in the form of dust. To perform an exposure assessment of a given radionuclide in the scrap, it is necessary to determine how that contaminant is distributed in the various media following the melting of the scrap. The concentration of radionuclide /' in medium m is calculated as follows: (6-1) Cim = specific activity of radionuclide / in medium m Cif = specific activity of radionuclide/' in the furnace charge Ms = mass of scrap in furnace charge Pim = partition ratio (or distribution factor4) of radionuclide /' in medium m Mm = mass of medium m produced from that charge A literature search as well as thermodynamic calculations were used to develop the partition ratios and distribution fractions for EAF melting of carbon steel used in the present analysis, which are listed in Table 6-3. Ranges of partition ratios and distribution factors reflect variability in melting practices. A detailed report of this study appears in Appendix E. A similar Ms study was performed for cast iron production; it is reported in Appendix F. The ratio - in Mm Equation 6-1 can be replaced by the reciprocal of fm, the mass of each medium as a fraction of the mass of the furnace charge. Based largely on the discussion in Section E.7 of Appendix E and a comparable discussion in Appendix F, the following mass fraction values were adopted for the present analysis: Strictly speaking, partitioning occurs between two liquid phases, such as the molten steel and the slag. The term "distribution factor," as used in this report, refers to the fraction of the activity that is found in the baghouse dust. 6-6 ------- • Purchased scrap (scrap from sources outside the mill) ............ 0.95 • Home scrap (metal recovered from by-products of previous melts) . . 0.05 • Finished steel ............................................ 0.9 • Steel slag ............................................... 0.117 • Cast iron slag ............................................ 0.065 • Baghouse dust ........................................... 0.015 • Melt [[[ 0.97 6.2.1 Partition Ratios and Concentration Factors The calculation of the partition ratios is complicated by the fact that the baghouse dust includes contributions from the melt — atomic and molecular species that are vaporized in the furnace and particulate matter entrained in the gas stream — as well as from the slag. As discussed in Section E.7, one-third of the dust is formed from slag. Thus, the mass fraction from the slag is 0.005 (0.015 x Vs = 0.005). Slag formation can thus be visualized as a two-step process. In the first step, slag forms during the melting of steel scrap, and those radionuclides that are predicted to partition to the slag pass into this phase. From a nominal charge of 100 t of total scrap, a total of 12.2 t of slag is initially formed. Next, 500 kg of the slag, along with any contaminants in that material, is entrained in the gas stream and forms part of the dust, resulting in a net 1 1 .7 t of slag. Thus, about 5% of the activity that initially partitions to the slag passes into the dust. This is why the actinides and other elements which, on theoretical grounds, are expected to partition entirely to slag, are assigned slag partition ratios of 95%, the remaining 5% accumulating in the dust. Earlier analyses (SCA 1995) showed that the release of filtered parti culates to the atmosphere does not constitute a significant pathway in the radiological assessment of exposed individuals; therefore, such releases are not included in the release fractions listed in Table 6-3. C. The concentration factor (CFim) for radionuclide / in medium m is defined as , Cis where Cis is the specific activity in scrap. In deriving the concentration factors, the home scrap ------- Table 6-3. Partition Ratios (PR), Concentration Factors (CF), and Distribution Factors (DF) Element Ac Ag Am C Ce Cm Co Cs Eu Fe I Mn Mo Nb Ni Np Pa Pb Pm Po Pu Ra Ru Sb Sr Tc Th U Zn Furnace Charge CFa 0.95 1 0.95 1 0.95 0.95 1 0.95 0.95 1 0.95 0.98 1 0.95 1 0.95 0.95 0.95 0.95 0.95 0.95 0.95 1 1 0.95 1 0.95 0.95 0.96 Steel PR (%) 0 99- 75 100 - 27 99 97 24- 65 99 99 99 99- 80 99 20-0 CF 0 1.02 0 1.03 0 0 1.02 0 0 1.00 0 0.66 1.02 0 1.02 0 0 0 0 0 0 0 1.02 1.02 0 1.02 0 0 0.20 Cast Iron CF 0 1.01 0 1.01 0 0 1.01 0 0 1 0 0.98 1.01 0 1.01 0 0 0 0 0 0 0 1.01 1.01 0 1.01 0 0 0.02 Slag PR (%) 95 0 95 95 95 0- 5 95 2 72- 32 95 95 95 95 95 95 95 95 95 CF 7.71 0 7.71 0 7.71 7.71 0 0.41 7.71 0.17 0 6.03 0 7.71 0 7.71 7.71 0 7.71 0 7.71 7.71 0 0 7.71 0 7.71 7.71 0 Baghouse DFb (%) 5 1 - 25 5 5 5 1 100 - 95 5 1 4-3 1 5 1 5 5 100 5 100 5 5 1 1 -20 5 1 5 5 80 - 100 CF 3.17 16.7 3.17 0 3.17 3.17 0.67 63.3 3.17 0.67 0 2.61 0.67 3.17 0.67 3.17 3.17 63.3 3.17 63.3 3.17 3.17 0.67 13.3 3.17 0.67 3.17 3.17 63.3 Release Fraction (%) 0-73 100 Note: data are relevant to EAF operations, except cast iron concentration factors, which apply to iron foundries. Scrap metal charged to the furnace, which consists of 95% imported scrap and 5% recirculating home scrap. b See Note 4 6-8 ------- In all cases where a range of partition ratios is listed for a given chemical element in a given medium, the high end of the range is used to calculate the corresponding concentration factor. The range results from the variability of melting practices and other factors. Consequently, a given individual may be exposed to radionuclide concentrations corresponding to the high end of the range in one medium, while a different individual could be exposed to the high end of the range for a different medium. In only one scenario in the present analysis—the operator of the continuous caster—is the same individual exposed to radioactive materials in two different media (other than scrap). As shown in Table 5-1, this individual would be exposed to external radiation from the steel while inhaling and ingesting the furnace emissions (i.e., baghouse dust). The radiological impacts on this individual of those nuclides that have a range of partition ratios in both the steel and the dust—isotopes of silver, manganese, antimony and zinc—are overstated. Since, as will be seen in Chapter 7, this individual does not become the RME for these nuclides, this approach has a minor impact on the analyses. In all other cases, however, this method yields a reasonable maximum exposure assessment. 6.3 EXPOSURE PATHWAYS 6.3.1 External Exposures to Direct Penetrating Radiation Table 5-1 shows that the external exposure pathway is included in every scenario except the consumption of groundwater contaminated by leachate from a slag pile. Except for the assessment of exposure to airborne effluent emissions, which are discussed later in this chapter, external exposures were evaluated either by using the MicroShield computer code or by employing the external exposure dose coefficients in Federal Guidance Report (FOR) No. 12 (Eckerman and Ryman 1993). Use of MicroShield Computer Code MicroShield is an industry-standard computer program used to perform y-ray shielding calculations for radioactive sources. The program computes the exposure rate from a uniform distribution of one or more radionuclides within a specified matrix, such as a solid cylinder of iron, with additional shielding material between the source and the receptor point. The code includes attenuation and buildup factors for nine metallic elements as well as for air, concrete, and water. In addition, it is possible for the user to create custom materials by specifying the densities and elemental compositions of the new material. However, the present analysis uses iron to represent the various steel alloys in both the source material being processed and in the components of the furnace that act as radiation shields. Since carbon steel contains over 98% 6-9 ------- iron, it is preferable to model it as pure iron, since the buildup factors for iron are based on actual measurements. Results obtained with MicroShield are generally in good agreement with those performed by photon transport codes employing discrete ordinate or Monte Carlo methods. At photon energies below about 100 keV, the MicroShield results begin to diverge from those calculated by the more exact methods. This limitation, however, is not of concern in the present analysis. The primary dose contribution from most of the y-emitting nuclides is from the high-energy photons. From nuclides that emit only low-energy photons, the dose is dominated by internal exposure. For such nuclides, the external exposure pathway is potentially significant only in scenarios where there are no internal exposure pathways. The only such scenarios in the present study are those involving exposure to finished products, which of necessity address only those nuclides which partition to the metal. The only nuclide in the present study that partitions to the steel and emits photons in the 10 - 100 keV range5 is Mo-93. Because of its long half-life and very low penetrating radiation (the principal photons have an energy of 16.6 keV), the highest exposure would be of the sailor on the naval vessel. The external exposure rate from Mo-93 in the hull plate was estimated by multiplying the exposure rate of Co-60—a strong y emitter—by the ratio of the FOR 12 dose coefficient of Mo-93 to that of Co-60 for a soil layer with a similar mass thickness. Further details are presented in Appendix H. MicroShield utilizes dose coefficients listed in ICRP Publication 51 (ICRP 1987) to calculate the effective dose equivalent for each of five exposure geometries6. For most exposure scenarios, the present analysis assumes that the radiation is incident in the anteroposterior direction, which corresponds to the exposed individual's facing the radiation source. This is a realistic assumption in most cases—it also results in the highest dose. The resulting external exposure factors, expressed in millirem per hour, are utilized in the assessments of the external exposure pathways. An illustrated description of the source and receptor configuration for each scenario analyzed with MicroShield are presented in Appendix H. Photons with energies below 10 keV have very limited penetrating ability and are typically ignored in calculating whole-body doses, viz. FOR 12, ICRP 1987. Both ICRP 51 and FGR 12 use the tissue weighting factors recommended in ICRP Publication 26, rather than those recommended in ICRP Publication 60. The impact of these differences on doses from external exposure is discussed in Chapter II of FGR 12. 6-10 ------- The external exposure factors are used to calculate the normalized doses and risks from external exposure. The source-to-receptor distance and the duration of exposure for each scenario are listed in Table 5-1. Additional details are presented in Appendix H. The annual dose to the maximally exposed individual from external exposure to a given nuclide in a given scenario is calculated by multiplying the appropriate exposure factor by the exposure duration and by the specific activity in the source medium, normalized to a unit specific activity in the scrap. The concentration factors for the various nuclides in the different media are listed in Table 6-3. The dose calculations are shown in Equation 6-2, below. DimxC*) = CimFimx(x)te (6-2) Dimx(x) = dose from one year of external exposure to radionuclide / in medium m at distance x (|lSv/a per Bq/g in scrap) x = distance from source to receptor (m) te = annual exposure duration (hr/y) Fimx(x) = external exposure factor at distance x from radionuclide / in medium TO in a given source configuration (|lSvg/Bq-hr) External Exposure over a Varying Distance In several scenarios, such as the EAF furnace operator, the distance between the source of the external radiation and the exposed individual varies over time—i.e., the operator is at different locations during the course of the day. Although the minimum and maximum distances of a given individual have been observed or can be inferred, the time spent at various locations within this range is difficult to ascertain. The analysis therefore makes the simplifying assumption that the individual spends an equal amount of time at each distance. This is equivalent to assuming that he moves uniformly back and forth, like a sentry walking his post between two points. To determine the integrated exposure during this time, it is necessary to derive the exposure rate at some arbitrary distance from the source, given the exposure rates at two fixed distances. The first step is to calculate the distance and strength of a fictitious equivalent point source that would produce the same exposure rates at the same locations as those calculated for the real source. Applying the inverse square law, we obtain: 6-11 ------- RW - R(x) = exposure rate at distance x from real source A0 = strength of equivalent point source x0 = distance of equivalent point source from real source To evaluate the constants A0 and x0, we substitute the calculated values of R(x) at two known distances: Ri = r^y (6-4) » _ Ao Solving Equations 6-4, we obtain: A = R (\ - \ \2 Ao Ki \xi xo/ v T?'/2 v T?'/2 (6-5) Xj K.J ~ X2 K-2 v" J/ Xo = Vl T? '/2 j - K2 Next, we find the mean value of R(x) over the interval x3 < x < x4 dx "° J (x^ R = 5J (6-6) - Xo) Equation (6-6) is used to evaluate the factor Fimx(x) in Equation 6-2 over the range [x3, x4]. 6-12 ------- Use of FGR 12 Dose Coefficients MicroShield is a useful tool for determining dose rates from relatively compact sources. In some scenarios, however, the external radiation comes from a planar source whose lateral dimensions are large in comparison to the source-to-receptor distance, and which is optically dense—i.e., it has a mass thickness several times greater than the mean free path of the most penetrating radiation of any of the nuclides in the analysis. In those cases, the dose coefficients for soil contaminated to an infinite thickness listed in FGR 12 provide a convenient method of analysis. These factors were applied to the slag yard worker standing at the edge of the slag. Since he is only exposed to one-half of an infinite plane, he would only get one-half the dose predicted by FGR 12. Since the average atomic number of slag is somewhat higher than that of soil, the analysis would tend to overstate the doses. For the nuclides with the most energetic y-rays, for which external exposure is a major pathway, the interaction of the radiation with the source material is primarily by Compton scattering, which is relatively insensitive to the atomic number. The FGR 12 dose coefficients were also used to evaluate the external exposure of the scrap cutter. Since he spends time in canyons surrounded by walls of scrap, it is reasonable to model the sources as two vertical half-planes beginning at the ground surface. The two half-planes together are equivalent to a single infinite plane. Again, the scrap has a higher atomic number than the average for soil, yielding a somewhat conservative but not excessively overstated assessment. 6.3.2 Inhalation of Contaminated Dust During certain of the operations listed in Table 5-1, some of the radioactively contaminated material is assumed to be dispersed in the ambient atmosphere in the form of dust. The radiation exposure of an individual inhaling this dust will depend on his breathing rate, the dust loading of the ambient air, the respirable fraction (i.e., the mass fraction of particles with AMAD < 10 |lm)7, the exposure duration, and on whether or not he uses some form of respiratory protection. The radiological impacts are modeled by the following equations: AMAD is the acronym for Activity Median Aerodynamic Diameter," [which] is the diameter of a unit density sphere with the same terminal settling velocity in air as that of an aerosol particle whose activity is the median of the entire aerosol." (Eckerman et al. 1988). 6-13 ------- = BCimfffrFihteXd (6-7) Dimh = 50-year dose commitment from inhalation of radionuclide /' in medium m during one year (|lSv/a EDE per Bq/g in scrap) B = breathing rate = 1.2(m3/hr) ff = respiratory protection factor (filter factor) fr = respirable fraction Fih = dose conversion factor for inhalation of radionuclide/' (|lSv/Bq) (ICRP 1994) %d = concentration of dust in air (dust loading, g/m3). R^ = excess lifetime risk of radiogenic cancer from inhalation of radionuclide /' in medium m during one year (y"1 per Bq/g in scrap) Gih = risk factor for inhalation of radionuclide i (Bq'1) (U.S. EPA 1994) The dust loading for each exposure scenario is listed in Table 5-1. The derivation of these values is discussed in Appendix H. The analysis assumes that all of the airborne dust emanates from the contaminated material being processed and that the specific activity of a given radionuclide in the dust is the same as that of the material. This assumption is realistic for operations such as handling of baghouse dust or slag, or the use of a cutting torch on scrap. In these cases, the dust results from the operations and would contain the same radionuclides as the material in process. Studies show that the specific activity in the dust may be either greater or less than the radionuclide concentrations in the source of the dust due to enhancement and discrimination processes. For example, the particles that become airborne are usually less than 50 |im in diameter (Peterson 1983). If most of the activity in the source material is found in particles larger than 50 |im, then the specific activities in the dust are likely to be lower than those in the source. Conversely, if the activity is primarily on particles smaller than 50 |lm, the specific activity in the dust can be greater than that in the source. The assumption that the specific activity of a given radionuclide in the dust is the same as that of the source material is a 6-14 ------- reasonable approximation in most cases. Many other physical and chemical properties besides particle size can also produce enhancement or discrimination effects. A discussion of this subject is provided in Envirosphere 1984. The inhalation DCFs are taken from ICRP Publication 68 (ICRP 1994) (commonly referred to as "ICRP 68"),8 while the inhalation risk factors (i.e., slope factors) are from U.S. EPA 1994. ICRP 68 lists DCFs for both 1 |lm and 5 |lm AMAD particles. Since the particle size distributions vary, the higher of the two DCFs was used for each radionuclide for the appropriate Lung Clearance Type. Accordingly, the dose and risk factors used in this study are conservative, high- end values.9 The DCFs for inhalation, as well as the ones for ingestion, discussed in the following section, also depend on the chemical form of the nuclide in question. The nuclides that reach the groundwater will be in their most soluble form—less soluble nuclides would be retarded in their migration to the aquifer (see Section 6.4.1)10. In the case of baghouse dust, slag dust, or vapors from molten metal, the analysis of radionuclide distributions during the melting of carbon steel, presented in Appendix E, indicates that in almost all cases the nuclides will be present as oxides. Consequently, the dose conversion factors corresponding to the respective Lung Clearance Type and fj value, as listed in Tables E. 1 and F. 1 of ICRP 68, were adopted for the analysis. Since the chemical forms of the radionuclides on the scrap metal is unknown, it is assumed that each nuclide has the form corresponding to the highest DCF listed in ICRP 68. For the ingestion of radionuclides left in the melt—the frying pan scenario in the present analysis—the nuclides are assumed to be in the elemental form. The chemical form of each element with radioactive isotopes that may be found in radioactively contaminated carbon steel is listed in Table 6-4, along with the appropriate Lung Clearance Type and fj value. ICRP 68 provides DCFs for intakes of radionuclides by workers. These values are appropriate for the present study, which addresses adult members of the population (see Section 3.1.2). An additional contribution to the dose from dust inhalation is from large particles that are inhaled, refluxed from the air passages and then swallowed. In all scenarios where inhalation exposure is modeled, inadvertent ingestion of deposited particulate matter is also assumed. Since the inadvertent ingestion rate is typically several times larger than the inhalation rate of the same material, the ingestion of these large inhaled particles would not make a significant contribution to the total dose. Entries in the groundwater column ("GW") appear for those elements whose isotopes were addressed in the analysis of groundwater contaminated by leachate from the slag pile, as indicated by checkmarks in Table 6-6. 6-15 ------- Table 6-4. Lung Clearance Types and Ingestion f\ Values for Use with ICRP 68 Element Ac Am Ag Ba Bi C Ce Cm Co Cs Eu Fe 1 Mn Mo Nb Ni Np Pa Pb Pm Po Pr Pu Ra Rh Ru Sb Srb Tc Te Th U Y Zn Metal Melt Form Ag C Co Fe Mn Mo Ni Rh Ru Sb Tc all Type F M F F F F F F F F S I 0.05 1 0.05 0.1 0.1 0.8 0.05 0.05 0.05 0.01 0.8 0.5 Dust and Slag Form Ac2O3 all Ag all Bi203 C02a Ce203 all CoO all all FeO all MnO MoO3 Nb205 NiO all Pa02 all Pm2O3 Po02 Pr203 Pu203 all oxide RuO4 Sb2O3 SrO Tc02 Te02 ThO2 UO2 Y203 all Type S M F F M S M S F M M F M S S M M S F S M S S M S S M F M M S S S S f. 5e-04 5e-04 0.05 0.1 0.05 1 5e-04 5e-04 0.05 1 5e-04 0.1 1 0.1 0.8 0.01 0.05 5e-04 5e-04 0.2 5e-04 0.1 5e-04 1e-05 0.2 0.05 0.05 0.1 0.3 0.8 3e-01 2e-04 2e-03 1e-04 0.5 GW* I 0.01 5e-04 5e-04 0.3 0.02 * Ground-water pathway: only elements that are checkmarked in Table 6-6 are addressed in slag leachate analysis. Carbon in slag would most likely be found as a carbonate; however, ICRP 68 only lists inhalation DCFs for CO2, CO, and organic compounds. All forms except SrTiO3, an unlikely contaminant of potentially contaminated steel scrap 6-16 ------- 6.3.3 Incidental Ingestion Individuals working in a dusty, sooty environment are likely to inadvertently ingest some of the contaminated material, which is generically referred to in this report as soot. The radiological impacts of such incidental ingestion will depend on the soot ingestion rate of the exposed individual and on the duration of exposure. The impacts are modeled by the following equations: img im ig s e (6-8) Rimg ~ Cim Gig Js te Djmg = 50-year dose commitment from ingestion of radionuclide /' in medium m during one year (|lSv/a EDE per Bq/g in scrap) Fig = DCF for ingestion of radionuclide i ((|lSv/Bq) (ICRP 1994)) Is = soot ingestion rate (g/hr) Rjmg = excess lifetime risk of radiogenic cancer from ingestion of radionuclide / in medium m during one year (y"1 per Bq/g in scrap) Gig = risk factor for ingestion of radionuclide i (Bq'1) (U.S. EPA 1994) The EPA's Exposure Factors Handbook (U.S. EPA 1997) presents a detailed discussion of soil and soot ingestion, primarily by children. However, data are also provided for inadvertent soil and soot ingestion rates by adults working in a dusty environment. For adults, the daily soil ingestion rates range from 0.56 mg/day for indoor work to 480 mg/d for outdoor work. Table 4-15 of U.S. EPA 1997 lists soil ingestion rates for assessment purposes—included is a rate of 20 mg/hr by an adult during gardening. Given the nature of the operations at scrap yards and steel mills, this seems to represent a reasonable maximum exposure and was therefore adopted for the present analysis. The one exception is the lathe manufacturing operation, where it is likely that only part of the "dirt" in the area would come from the cast iron that is being ground. Some would come from the grinder itself, for instance. Consequently, an ingestion rate of 10 mg/hr was adopted for the radioactively contaminated component of the soot in that operation. 6.3.4 Radioactive Decay Equations 6-2, 6-7 and 6-8 present methods of calculating dose rates from all scenarios in which the source strength is essentially constant during the course of one year. These are situations in 6-17 ------- which the source is replaced at frequent intervals. For scenarios in which the source is not replaced—the end user of finished products—radioactive decay over the course of a year must be taken into account. In such cases, the exposure is integrated over a period of one year, resulting in the following expression: cOOt, 1 -A,,t I - e ' (6-9) A; = radioactive decay constant of nuclide /' (y"1) t = integration time The risk from external exposure is calculated by multiplying the corresponding dose by the risk factor for doses of low-LET radiation to the whole body: excess lifetime risk of radiogenic cancer from one year of external exposure to radionuclide / in medium m at distance x (y"1 per Bq/g in scrap) risk factor for external exposure 7.6 x 1Q-8 jiSv'1 (U.S. EPA 1994) In addition to decay, ingrowth of progeny was also considered. Eleven of the elements listed in Table 6-3 significantly partition to steel or iron. Of the radionuclides included in the analysis that partition to steel or iron, only Mo-93 has a long-lived progeny: Nb-93m, which has a half- life of 16.1 y. (As stated earlier, the dose and risk factors of all progenies with half-lives of less than six months are included in those of the parent.) The RME individual for Mo-93 is the sailor sleeping next to a hull plate. A naval support vessel has an average life of about 30 years. Even in the last year of anticipated use, the Nb-93m activity in the plate would be less than 75% that of Mo-93, which has a half-life of 3,500 y and would thus not have decayed significantly. The external dose from the Nb-93m would be approximately 13% that from Mo-93. Likewise, in the cookware scenario described in Section 6.4.2, the ingestion dose from Nb-93m during the last year of the 15-year lifetime of the utensil would be about 2% that of Mo-93. Given the other 6-18 ------- uncertainties in the analysis, omitting the Nb-93m contribution to the total Mo-93 dose in the finished product scenarios does not have a significant impact. 6.4 UNIQUE SCENARIOS Three unique scenarios that require special models are described in this section. The exposure assessment of two of the scenarios—the consumption of groundwater contaminated by leachate from slag and the consumption of food cooked in cast iron cookware made from potentially contaminated scrap—required the development of special sub-models. The third analysis, the assessment of the impact of fugitive airborne emissions from the furnace on nearby residents, utilized EPA's CAP-88 model. A discussion of the anticipated impacts of disposing the cleared scrap in an industrial landfill—utilizing the RESRAD code—is presented in Appendix L. 6.4.1 Groundwater Contaminated by Leachate from Slag Storage Piles As discussed in Section 5.4.4, an individual residing near the slag storage yard who gets his drinking water from a well that is downgradient from the slag could be exposed to contaminated groundwater. During storage at the steel mill site, slag will be subjected to weathering and certain components may be leached from the slag and ultimately contaminate the local groundwater. A study, based on a search of available literature, was performed to enable the calculation of leach rates of various radionuclides. Details of this study are presented in Appendix I. In addition, EPA sponsored an experimental study at the Brookhaven National Laboratory to determine the leach rates of constituents of various steel and iron slags. Some results of this study are presented in Appendix 1-1. Some of the information obtained from both studies is presented in this section, followed by the development of a model of the leaching of radionuclides which partition to the slag. The other scenarios discussed in this chapter would produce maximum impacts shortly after the release of the cleared material (except the naval vessel scenario, where there is a lapse of 18 months from the time the steel is delivered until the ship is commissioned). Because of the time required for transport of radionuclides to an underground source of drinking water, the maximum impact could occur many years later, depending on the radionuclide in question and on the assumed hydrogeology of the site. The present analysis limited the period of assessment to 1,000 years from the time of release. 6-19 Continue ------- Back Elemental Selection Criteria Because of the scarcity of data, it was desirable to narrow the scope of the analysis to those elements that have radioactive isotopes which, if leached from the slag, would have a significant radiological impact via the groundwater exposure pathway. The selection criteria include: • Potential contamination of steel scrap by one or more isotopes of the given element • Significant partitioning of the given element to the slag (i.e., concentration factor > 0.1) • Travel time to the aquifer relative to half-life of longest-lived isotope included in study • Travel time relative to the 1,000-year period of impact assessment Travel time. The travel time through the vadose (unsaturated) zone of the /'-th element was determined by Equation E.21 of the RESRAD user's manual (Yu et al. 1993): Az R pe Rs At = : (6-11) Az = thickness of vadose zone Rd = retardation factor for /'-th element pe = effective porosity of vadose zone I = infiltration rate pb = bulk soil density of vadose zone Kd = soil-water distribution coefficient for /'-th element (cm3/g) pt = total porosity of vadose zone Kv = saturated hydraulic conductivity of vadose zone b = soil-specific exponential parameter The hydrogeology of the slag yard was based on one of the three generalized reference sites presented in the technical support document for the development of radionuclide cleanup levels 6-20 ------- for soil (SCA 1997, Section 4.4.1). These in turn were based on EPA's DRASTIC standardized system for evaluating groundwater pollution potential of various hydrogeologic settings. The parameters of the vadose zone of the three reference sites are shown in Table 6-5, below. Table 6-5. Vadose Zone Parameter Values for Site Types A, B, and C Site Type A B C Soil Type Clay Loam Loam/Sandy Loam/Clay Sand/Gravel Az(m) 50 10 3 Kv (m/y) 77.3 1090 5550 ph 1.28 1.36 1.52 b 8.52 5.39 4.05 Pt 0.476 0.435 0.395 Pfi 0.15 0.22 0.23 I (m/y) 0.05 0.13 0.40 The travel time calculated for each element is listed in Table 6-6. The calculations utilized the parameters for site type C, which result in the shortest times. Any element with a travel time greater than 1,000 years, or longer than 20 half-lives of its longest-lived isotope11 (among the radionuclides included in the present analysis), is marked with an "X" in Table 6-6, indicating that it is eliminated from consideration. It is not likely that any significant activity of any isotope of any such element would reach the aquifer during the 1,000- year assessment period under any reasonable environmental conditions. Slag Cement Leaching Studies The American Nuclear Society has developed and formalized detailed procedures for measuring the teachability of solidified low-level radioactive wastes (ANS 1986). This procedure involves testing of controlled-geometry specimens in demineralized water at 17.5°C to 27.5°C to determine releases over successive intervals of time. Mass transport is assumed to be controlled by a diffusion process. When the fraction leached from a uniform sample is less than 20% of the initial activity, the leaching behavior can be approximated by that of a semi-infinite medium where the "effective diffusivity" is given by the following equation: D. = TiT AioAntS (6-11) D; = effective diffusivity of nuclide /' (cm2/d) 11 Twenty half-lives was selected as the cutoff criterion since the activity will decay to 10"6 of its initial value during this time. 6-21 ------- T = mean time of the leaching interval n (d) „ \ 2 n-l ain = activity of nuclide /' released during time interval n V = sample volume (cm3) A;0 = initial activity of nuclide /' in sample Ant = duration of n-th leaching interval (d) S = surface area of sample (cm2) Table 6-6. Potential Contaminants of Groundwater Element Ac Am Ce Cm Cs Eu Fe Mn Nb Np Pa Pb Pm Pu Ra Sr Th U Slag CF 7.71 7.71 7.71 7.71 0.41 7.71 0.19 6.03 7.71 7.71 7.71 (progeny) 7.71 7.71 7.71 7.71 7.71 7.71 V (y) 21.8 432.7 0.8 18.11 30.17 13.6 2.7 0.9 20,300 2.14e+06 32,760 22.26 2.6 24,131 1600 28.6 1.40e+10 4.47e+09 Kd 240 1900 500 4000 270 240 170 50 110 5 110 270 240 550 500 15 3200 15 Atv;(y) 1,594 12,613 3,320 26,553 1,793 1,594 1,129 333 731 34 731 1,793 1,594 3,652 3,320 100 21,243 100 Potential X X X X X X X X • • • X X X • X • Comments will decay will decay will decay will decay will decay will decay will decay will decay will decay At;» l,000y At;» l,000y At;» l,000y Half-life of longest-lived isotope 6-22 ------- -, is greater than 0.2, Equation 6-11 must be When the cumulative fraction leached, corrected for specimen geometry. Using a model and procedures similar to those described in ANS 1986, Japanese investigators have determined the fractional leaching of Sr-90, Co-60, Cs-137, and H-3 from cement/slag composites in deionized water and synthetic sea water (Matsuzuru and Ito 1977; Matsuzuru et al. 1977, 1979). The duration of the leaching tests was about 100 days. The radionuclides were incorporated into the cement via a sodium sulfate solution. The composition of the slag is listed in Table 6-7. Table 6-7. Composition of Slag Used in Leaching Test Component SiO2 A1A Fe203 CaO MgO Insoluble Residue Ignition Loss Composition (wt %) 28.7 11.5 2.3 50.9 3.2 0.8 0.6 Leaching data were analyzed using a plane source diffusion model to derive the expression f. = 2S v N D:t 71 (6-12) f; = fraction of nuclide /' leached in t days. Equation 6-12 can be rewritten as f = 2S mit' D.. (6-13) 6-23 ------- where the expression in the square brackets is represented by m;, the slope of the line obtained by plotting f; vs. t'/2. Once m; is determined, Equations 6-13 can be solved for D;: (6-14) Since the actual leaching process involves an initial rapid leaching rate of a few days' duration (~ 7 d for Sr-90 and ~ 2 d for Co-60), followed by a longer-term linear relation between f; and t'/2, the experimental data are fitted to an equation of the form fi = mjt'7' + cti (6-15) Equation 6-12 can also be used to determine the value of f; for various geometries, as follows: Ail V V2 (6-16) where subscripts 1 and 2 refer to geometries 1 and 2, respectively. Because of certain limitations and problems such as the initial leach rate, Matsuzuru and Ito (1977) defined L, the leaching coefficient, with the same mathematical form as D in Equation 6-14. Values of L for Sr-90 leached from slag cements ranged from 1.2 x 10"7 to 1.7 x 10"7 cm2/day for both deionized water and synthetic sea water at 25°C. Using average values of LSr for samples cured seven days prior to testing in deionized water, and assuming a right circular cylinder, h = 2r, V = 70 cm3, we have derived values for mSr and aSr in Equation 6-15, which are listed in Table 6-8. The teachability of Cs-137 was reported to be about ten times that of Sr-90; it was therefore assumed that mCs =10 mSr and aCs = 10 aSr. Equation 6-16 was then used to derive values that describe leaching from slag particles that are also right circular cylinders, but only 1 cm in diameter—a more typical size for EAF slags. The chromium data was based on particles that passed through a 9.5 mm mesh. The calculated value cited in the section "Other slag leaching studies," below, therefore constitutes an upper bound to mCr for a 1-cm right circular cylinder, listed in Table 6-8. 6-24 ------- Table 6-8. Leaching Parameters Values Element Sr Cs Crb m (d-'70 r = 2.233 cma 5.8e-04 5.8e-03 c r = 0.5 cm 2.59e-03 2.59e-02 6.9e-06 a r = 2.233 cm 4.97e-03 4.97e-02 0 r = 0.5 cm 2.22e-02 0.222 0 a Corresponds to 70 cm3 right circular cylinder ( h = 2r) Cr is used as a surrogate for Nb, Np, Pa and U—see discussion below. c Not applicable, see text. The strontium data were replaced by the data obtained from the Brookhaven National Laboratory (BNL), described below. Data from Brookhaven National Laboratory Results from leaching experiments on EAF slags performed at BNL indicate that the leaching of strontium, the only element checkmarked in Table 6-6 that was measured in the leachate, was governed by diffusion (Fuhrmann 1997). The diffusion coefficients determined from tests on three monolithic samples of EAF slag are listed in Table 6-9. To calculate the value of mSr for a monolithic cylinder, we first invert Equation 6-14: mi 2S V \ D, _ 12 7t r \ Di 7t S = 27u(r2 + hr) = 67tr2 V = Tur3 (h = 2r) msr = 1.54x lQ-2d-1/2 1.51 x 10"11 cm2/s (mean of values in Table 6-9) 1.30x 10-6cm2/d = 0.5 cm 6-25 ------- Since Fuhrmann did not report any initial releases that were not diffusion-controlled, CCSr is set equal to zero. These data were used to model the Sr-90 leaching from the slag in the present analysis. Other Slag Leaching Studies This section describes earlier leaching studies done on pure slags rather than slag/cement composites. Table 6-9. Diffusion Coefficients for EAF Slag Monolithic Samples Slag Sample AS-1 AS-2 AS-3 Diffusion Coefficient (cm2/s) 1.4e-ll 2.5e-ll 6.2e-12 Source: Fuhrmann 1997 Australian researchers at CSIRO incorporated the toxic elements As, Sb, Cd, Zn and Cr into slags of various types by melting them at 1300°C, and subsequently leached the slags according to the EPA TCLP protocol (Jahanshahi et al. 1994). In the TCLP test, a sample of at least 100 g, which has a minimum surface area of 3.1 cm2/g or passes through a 9.5 mm sieve, is treated with about 2,000 g of extractant for 18 ± 2 hr at 22 ± 3°C using rotary agitation. The extractant has a pH of either 4.93 or 2.88, depending on the basicity of the sample (40 CFR 261, Appendix II, Method 1311). The pH is achieved by use of acetic acid that is buffered with sodium acetate for the higher pH (55 FR 11798). Slag samples were prepared by both slow cooling and quenching. Examination of the slag samples with an optical microscope showed that interconnected pores were present in the slow cooled and most of the quenched samples. Slow-cooled slag samples were crushed to either a "coarse" size (100% minus 10 mm)12 or a "fine" size (100% minus 1 mm) for the leaching tests. In generalizing on the results of the TCLP tests, the researchers observed that • As and Sb leached more readily than Cd, Cr or Zn • Fine particles generally leached more readily than coarse particles 100% minus 10 mm" means that 100% of the particles passed through a screen with a 10 mm mesh. 6-26 ------- • Slow cooled samples showed similar behavior to quenched samples In the present analysis, the fraction leached was estimated on the basis of the information presented by Jahanshahi et al. (1994), making the following assumptions: • Slag compositions from Table III of Jahanshahi et al. 1994 • Sample size = 100 g • Extractant volume = 2 L The results are presented in Table 6-10. The compositions of three of the slags (CaFel, CaFeSil, and FeSil) are markedly dissimilar to those expected from EAF melting of carbon steel. The other three slags, while not identical to EAF slags, are useful for developing preliminary modeling parameters. Unfortunately, of the five elements studied, only Cr is expected to be found in the slag in any significant quantity. However, in the absence of element-specific leaching data, Cr can be considered as a surrogate for the stable oxides expected in slags. Assuming that the fraction leached is proportional to t'/2, this fraction can be expressed by the second line of Equation 6-13, where the upper limit of mCr is about 6.9 x 10"6 d"'/2, based on Cr in the BF2 slags and an 18-hour leach test. Table 6-10. Fraction of Various Toxic Elements Leached from Slags Using EPA TCLP Protocol Slag CaFel CaFeSil CaFeSi2 FeSil BF1 BF2 Fraction Leached As 3.48e-03 3.53e-03 5.09e-04 1.54e-04 1.68e-04 9.80e-04 Sb 4.21e-05 2.68e-04 2.37e-04 1.10e-04 1.03e-04 4.29e-04 Cd 3.10e-04 2.40e-04 6.80e-05 1.15e-04 1.10e-04 1.20e-03 Cr O.OOe+00 O.OOe+00 5.63e-07 4.82e-07 O.OOe+00 6.00e-06 Zn 3.00e-05 2.70e-05 2.30e-05 2.30e-05 1.34e-04 1.23e-03 Leach Rate The leach rate from slag is calculated by Equation 6-15, using the BNL data for strontium and the parameter values for 1-cm diameter particles listed in Table 6-8 for other elements. 6-27 ------- The reference steel mill is assumed to produce 150,000 tons (-136,0001) of steel per year and 17,6001 of slag as a by-product. The slag is assumed to be continuously dumped onto a 1-meter high pile and to be removed at the same rate, with a 1-year inventory always remaining in place. The new slag is mixed uniformly with the old slag—the slag that is removed is thus a representative sample of this mixture. To model the age-dependent leach rate, we must first determine the age distribution of the individual particles in the pile. If there is a constant number, N, of particles in the pile, the number of particles added or removed during time dt is dn = Ar N dt Ar = removal rate constant 1 365 = 2.74 x 1Q-3 d'1 Assume that vo particles are added to the pile at some initial time (t = 0). After time t, the number of these particles left in the pile is given by v(t) = By definition, this is the number of particles older than t. The number of particles with ages between t and t + dt is obtained by differentiating the above expression with respect to t and changing the sign: dv = Arvoe r dt Since this expression is independent of the initial time, it can be generalized to all the particles in the pile: dn . -MjA — = *re dt (6-1?) dn = number of particles in pile with ages between t and t+dt N = total number of particles in pile The time-dependent leach rate is derived by differentiating f; in Equation 6-15 with respect to time: 6-28 ------- df. = — dt i /i Multiplying the above expression by the age distribution function of Equation 6-17 yields the leach rate of particles with ages between t and t + dt. Integrating that expression with respect to time yields nt m. A / i (0< t £ T) (6-18) fj (t) = leach rate of nuclide /' in slag pile at time t t = time since the start of recycling operations T = variable of integration T = period during which mill is recycling residually radioactive steel scrap The integral expression does not have an analytical solution but must be evaluated numerically. Equation 6-18 applies to the period (assumed to last one year) during which the mill is recycling residually radioactive steel scrap. After one year, no new contaminated particles are being added to the pile. The general relationship remains the same, but the leaching is now from particles with ages between t and t - T. m A f/(t) = -^-^ T-1/2e~V dt (t>T) (6-19) 2 J t-T The elapsed time since the mill began recycling the potentially contaminated scrap is represented by t, which is thus the age of the oldest residually radioactive particle in the pile, while t - T is the time since the mill stopped recycling that scrap, and thus the age of the newest particle. The concentration in the pore water percolating through the soil (in the absence of radioactive decay) is given by: C. D f/(t) p C. (t) = -^ g (6-20) 6-29 ------- Cip(t) = concentration of nuclide /' in pore water at time t (Bq/mL) Cig = initial specific activity of nuclide /' in slag (Bq/g) D = depth of slag layer = 1m pg = specific gravity of slag = 2 Transport to well. The RESRAD manual (Yu et al. 1993) presents a simple model that describes the transport of the radionuclides through the soil to the aquifer and thence to a drinking-water well downgradient from the source. The transport of the activities from the surface to the aquifer was described on pages 6-20 et seq. Once the contaminated water reaches the aquifer, it moves along a trajectory which is described by a vector sum of the vertical flow rate (the infiltration rate) and the horizontal flow rate. Figure 6-1 shows how the contaminated water moving downward through a volume element of width dx, at a distance x from the well, is deflected in the horizontal direction until it intercepts the well. The travel time in the aquifer is represented by .. Pe/Rd=X Ji °i / Pt ta (x) = travel time of nuclide / through a distance x (y) p'e = effective porosity of aquifer Rd = retardation factor of nuclide /' in aquifer x = distance from source element to well (m) Ka = saturated hydraulic conductivity of aquifer J = hydraulic gradient p'b = bulk density of aquifer (g/cm3) Kd = distribution coefficient of nuclide/'in aquifer (cm3/g) p't = total porosity of aquifer 6-30 ------- Slag Pile dx Inflow from infiltration Water Table Well Aquifer flow ~ ~ ~ - - - _ jnfiltrated water Water from aquifer^--^ Figure 6-1. Transport of Slag Leachate to Domestic Well The aquifer parameters were taken from the three generalized reference sites described on page 6-20. These parameters are shown in Table 6-11. Table 6-11. Aquifer Parameter Values for Site Types A, B, and C Site Type A B C Soil Type Shale/ Metamorphic/Igneous Sandstone/Limestone Sand & Gravel/Basal t/Karst Limestone Ka (m/a) 500 1000 2000 Pt 0.38 0.32 0.39 Pe 0.26 0.27 0.30 J 0.05 0.13 0.40 P'b (g/cm3) 1.64 1.80 1.62 The Kd values used for all the elements in the slag leaching analysis (those checkmarked in Table 6-6) for the three reference sites are listed in Table 6-12. Dilution of the pore water is modeled by the following equation, derived by differentiating the first of Equations E.27 in the RESRAD manual: 6-31 ------- df = dx (6-22) df = incremental dilution factor (concentration in pore water + concentration in well) of element dx dw = screened depth of well = 3 m Table 6-12. Soil-Water Distribution Coefficients (Kds) for Site Types A, B, and C Element Sr Nb Pa Np U Vadose Zone A 110 110 2700 5 1600 B 20 110 1800 5 15 C 15 110 550 5 35 Aquifer A 110 110 50 5 35 B 1.4 110 50 5 2.9 C 23.6 110 50 5 35 The contribution of element dx to the concentration of radionuclide /' in the well is given by the following expression: dwJKa ig D f/(T) (6-23) dC (t) = increment of concentration of radionuclide /' in well at time t T = t - ta;(x) - Atv; The first of Equations 6-23 follows directly from the definition of df given above, the second was derived by substituting the expressions in Equations 6-20 and 6-22 for Cip(t) and df, respectively, 6-32 ------- while the third was derived by differentiating the first line of Equation 6-21, solving for dx, and substituting. The concentration of radionuclide /' in the well at time T is obtained by integrating the last line of Equation 6-23 and introducing radioactive decay, which has been ignored up to now for the sake of clarity: A. t * ~ Atvi Cw.(t) = CigDpg6 I f/(T) dT (AtV;< t < Atv;+ ta;(/)) (6-26) ta (/) = travel time in aquifer of nuclide / through length / / = length of slag pile parallel to aquifer flow = 94m A = area of slag pile = V/D = 8,845 m2 V = slag volume = M/pg = 8,845 m3 M = one year's slag production = 136,000 t steel/a x 0.13 = 17,690 t For t > ta (/) + Atv , the lower limit of integration is changed as follows: t - Atv. f/(T)dT (t>Atv+ tai(/)) (6-27) t-ta:(D Exposure Assessment To calculate the maximum dose and risk from a given nuclide to an individual drinking the contaminated water in any one-year period, it was necessary to find the average concentration during the peak year. The concentration was calculated over successive one-year periods, beginning with t = Atv and proceeding in increments of 0.01 y, until the year of peak concentration was found. The dose or risk to the maximally exposed individual was determined by multiplying this peak one-year-average concentration by the appropriate dose or risk factor and the by drinking water consumption rate: 6-33 ------- ig ig w;\ max/ ig Rig = Cwi(tmax) Giglw Cw (tmaxj = average concentration of radionuclide /' in well during peak year Iw = annual consumption of water = 7.3xl05 mL/y Buildup of radioactive progeny The long travel time of some radionuclides necessitates the consideration of ingrowth of their long-lived radioactive progeny. Of the five elements listed in Table 6-6 which have radioisotopes capable of reaching the aquifer, only three — neptunium, protactinium and uranium — have isotopes which in turn have radioactive progenies with half-lives greater than six months. Neptunium. Table 6-6 shows that it would take neptunium leached from the slag 34 years to reach the aquifer. No significant ingrowth of the long-lived progeny of Np-237, the only neptunium isotope included in the present analysis, would occur during that time. Protactinium. Pa-23 1 is the only nuclide which would have significant ingrowth of long-lived daughter products — Ac-227, which has a half-life of 22.8 years, and its short-lived progeny — during its travel time of 73 1 years. Actinium has a higher Kd than protactinium and would thus travel more slowly. Still, its short half-life in comparison with the travel time of the parent indicates that significant daughter product activity would be found in the aquifer along with the parent. An upper bound of the radiological impact of Ac-227 was calculated by assuming that actinium has the same Kd as protactinium, so that the two nuclides would be in secular equilibrium in the aquifer. Uranium. Table 6-6 shows that it would take uranium leached from the slag 100 years to reach the aquifer. No appreciable ingrowth of the long-lived progeny from any of the three uranium isotopes included in the steel scrap recycling analysis would occur during that time. 6-34 ------- 6.4.2 Ingestion of Food Prepared in Contaminated Cookware One of the finished products examined in this study is cast iron cookware made from potentially contaminated scrap metal. Radionuclides may leach from such cookware into the food and subsequently be ingested. The metal content of food cooked in cast iron cookware can be inferred directly from data presented by Reilly (1985), who lists the concentrations of iron in beef and cabbage cooked in cast iron and copper utensils. Since copper utensils contain little or no iron, the average difference in the iron content of the foods cooked in the two types of vessels enables us to estimate the amount of iron leached from the cast iron one. Using cabbage as a surrogate for all vegetables, we derived a weighted average of 13.5 ± 4.7 mg/kg, based on the relative consumption of beef and vegetables shown below. The equations used to calculate the dose and risk for this exposure pathway are: Rig = CimCmfIfGig Qnf = concentration of iron in food = 1.35xlO-5g/g If = amount of food consumed annually = Ib + Iv = 1.45 x I05g/a Ib = beef consumption rate = 7.5 x I04g/a Iv = vegetable consumption rate = 7.0 x I04g/a 6.4.3 Impact of Fugitive Airborne Emissions from the Furnace on Nearby Residents The impact of fugitive airborne emissions from the furnace on nearby residents was modeled by means of EPA's Clean Air Assessment Package, using the computer code CAP88-PC. To calculate the effects of airborne effluent emissions on the RME individual, the map showing locations of EAF facilities and commercial nuclear power plants and shutdown dates was used to identify seven EAF facilities which could receive the decommissioning scrap from two or more nuclear plants in a single year. The meteorological data accompanying CAP88-PC were used to 6-35 ------- locate the meteorological station nearest to each of these seven facilities13. CAP-88 analyses for releases of C-14 and 1-129 were performed using each of the seven meteorological data sets, as shown in Table 6-13—the highest individual doses from each of the two nuclides from the seven runs were used in the TSD analysis. These analyses, performed for an earlier assessment, assumed annual releases of 11 mCi of C-14 and 15 mCi of 1-129, respectively. The RME individual was assumed to reside 1 km from the emission point; default CAP-88 values for the rural residential scenario were used for all other parameters. Table 6-13. Locations and Results of CAP-88 Analyses Location CAP88 ID PVD0560 HAR0631 ILG1058 TYS1328 MLI0269 LAX0304 ORD0452 State Rhode Island Pennsylvania Delaware Indiana Illinois California Illinois Nuclide 1-129 Dose mrem/y 3.90e-01 6.94e-01 4.48e-01 4.86e-01 3.97e-01 7.91e-01 4.16e-01 X/Q sec/m3 7.25e-07 1.58e-06 9.01e-07 1.02e-06 7.73e-07 1.84e-06 8.12e-07 Rank 7 2 4 3 6 1 5 C-14 Dose mrem/y 3.01e-04 8.66e-04 4.81e-04 8.28e-04 5.72e-04 4.86e-04 4.84e-04 X/Q sec/m3 2.65e-06 9.37e-06 4.67e-06 8.36e-06 5.64e-06 4.65e-06 4.55e-06 Rank 7 1 6 2 3 4 5 The fractions of foods produced on the individual's own land was based on a 1965-66 U.S. Department of Agriculture survey, cited by U.S. EPA (1989, Section A.I). More recent information indicates that there are now very few farms which produce the variety of food needed to supply a family. If another conservative assumption—that the RME individual lived in the worst of the seven locations—were added, the compounded conservatism would exceed the RME guidelines discussed in Section 3.1.2. It was therefore assumed that the individual lives in the median location—rank 4 in Table 6-13—rather than in the worst location—rank 1. These median doses must then be corrected for the releases of C-14 and 1-129 from the reference facility, as postulated by the present analysis. In addition, the doses must be adjusted to reflect the use of ICRP 68 DCFs, rather than the DCFs employed by CAP-88, which in most cases are 13 Although the NRC's schedule of termination of operating licenses has been revised since these seven facilities were selected, their locations, in different geographical regions, constitute a representative range of meteorological conditions. 6-36 ------- identical to ones in FGR 11 (Eckerman et al. 1987). Since the doses from these two radionuclides in this scenario are primarily delivered via the food ingestion pathway, the adjustments were based on the ratios of the DCFs for this pathway from ICRP 68 to those from FGR 11, as shown in the following equation. 'ia p, ig Dia = 50-year dose commitment from airborne effluent releases of nuclide /' D'ia = dose from airborne effluent releases of nuclide /' from previous analysis Fig = DCF for ingestion of nuclide /(ICRP 68) Qi = activity of nuclide/' released in one year = friMc fri = release fraction of nuclide/' (see Table 6-3) Mc = mass of cleared material processed in one year = 7.455 Gg (see Section G.4.2) Fj = DCF for ingestion of nuclide /' (Eckerman et al. 1988) QJ = released activity of nuclide /'in previous analysis The doses, along with the calculated cancer risks, are shown in Table 6-14. Because CAP-88 calculates the risk of cancer fatality, not the risk of cancer incidence, it was necessary to calculate the cancer risks by estimating the intake of each radionuclide and then applying the appropriate slope factor. n _ ia ig Kia -- p - ig Ria = lifetime risk of cancer incidence from airborne effluent releases of nuclide /' Gig = risk factor for ingestion nuclide /' (cancer morbidity per unit intake) (U.S. EPA 1994) 6-37 ------- This calculation is based on the assumption that all the dose is delivered via the ingestion pathway. Since this is the major pathway for these two nuclides, the above equation provides a good estimate of the cancer risk. Table 6-14. Calculation of Normalized Doses and Risks from Airborne Effluent Emissions Nuclide C-14 1-129 CAP-88 Q* (mCi/y) 11 15 Dose (mrem/y) 4.86e-04 4.48e-01 DCF-Ingestion (Sv/Bq) FOR 11 5.64e-10 7.46e-08 ICRP68 5.8e-10 l.le-07 Adjusted Q* (Bq/a) 5.44e+09 7.46e+09 Dose (|iSv/a) 6.68e-02 8.88e+01 Cancer Risk G* (Bq1) 2.79e-ll 4.98e-09 Risk (Bq1) 3.21e-09 4.02e-06 Activity released in airborne effluent emissions 6-38 ------- REFERENCES American Nuclear Society (ANS). 1986. "Measurement of the Leachability of Solidified Low- Level Radioactive Wastes by a Short-Term Test Procedure," ANSI/ANS-16. Eckerman, K. F., A. B. Wolbarst and A. C. B. Richardson. 1988. "Limiting Values of Radionuclide Intake and Air Concentration and Dose Conversion Factors for Inhalation, Submersion, and Ingestion," Federal Guidance Report No. 11, EPA-520/1 -88-020. U.S. Environmental Protection Agency, Washington, DC. Eckerman, K. F., and J. C. Ryman. 1993. "External Exposure to Radionuclides in Air, Water, and Soil," Federal Guidance Report No. 12, EPA 402-R-93-081. U.S. Environmental Protection Agency, Washington, DC. Envirosphere Company, 1984. "Algorithm for Calculating an Availability Factor for the Inhalation of Radioactive and Chemical Materials," EGG-2279. EG&G, Idaho. Fuhrmann, M. 1997. Private communication (see Appendix I-1 of present report.) International Commission on Radiation Protection (ICRP). 1987. "Data for Use in Protection Against External Radiation," ICRP Publication 51. Annals of the ICRP, vol. 17, no. 2/3. Pergamon Press, Oxford. International Commission on Radiological Protection (ICRP). 1994. "Dose Coefficients for Intakes of Radionuclides by Workers," ICRP Publication 68. Annals of the ICRP, vol. 24, no. 4. Pergamon Press, Oxford. Jahanshahi, S., et al. 1994. "The Safe Disposal of Toxic Elements in Slags." In Pyrometallurgy for Complex Materials and Wastes 105-119. Matsuzuru, H. and A. Ito. 1977. "Leaching Behavior of Strontium-90 in Cement Composites." Annals of Nuclear Energy 4:465-470. Pergamon Press, Oxford. Matsuzuru, H. et al. 1977. "Leaching Behavior of Co 60 in Cement Composites." Atomkernenergie (ATKE) 29 (4): 287-289. Matsuzuru, H. et al. 1979. "Leaching Behavior of Tritium From A Hardened Cement Paste." Annals of Nuclear Energy 6:417-423. Pergamon Press, Oxford. MicroShield, Ver. 4.2. Grove Engineering, Inc., Rockville, MD. 6-39 ------- Peterson, H. T. 1983. "Terrestrial and Aquatic Food Chain Pathways." In Radiological Assessment: A Textbook on Environmental Dose Analysis, NUREG/CR-3332, ORNL-5968, eds. J. E. Till and H. R. Meyer. U.S. Nuclear Regulatory Commission, Washington, DC. Reilly, C. 1985. "The Dietary Significance of Adventitious Iron, Zinc, Copper, and Lead in Domestically Prepared Food." Food Additives and Contaminants 2:209-215. S. Cohen & Associates (SCA). 1995. "Analysis of the Potential Recycling of Department of Energy Radioactive Scrap Metal." 4 vols. Prepared for U.S. Environmental Protection Agency, Office of Radiation and Indoor Air, Washington, DC. S. Cohen & Associates (SCA). 1997. "Radiation Site Cleanup Regulations: Technical Support Document for the Development of Radionuclide Cleanup Levels for Soil." Vol. 1. Prepared for U.S. Environmental Protection Agency, Office of Radiation and Indoor Air. U.S. Environmental Protection Agency (U.S. EPA), Office of Radiation Programs. 1989. Environmental Impact Statement: NESHAPS for Radionuclides, Background Information Document," EPA/520/1-89-005-1. Vol. 1, "Risk Assessments Methodology." U.S. EPA, Washington, DC. U.S. Environmental Protection Agency (U.S. EPA). 1994. "Estimating Radiogenic Cancer Risk," EPA 402-R-93-076. U.S. EPA, Washington, DC. U.S. Environmental Protection Agency (U.S. EPA), National Center for Environmental Assessment, Office of Research and Development. 1997. "Exposure Factors Handbook," EPA/600/P-95/002Fa. Vol. 1, "General Factors." U.S. EPA, Washington, DC. Yu, C., et al. 1993. "Manual for Implementing Residual Radioactive Material Guidelines Using RESRAD," ANL/EAD/LD-2. Argonne National Laboratory, Argonne, IL. 6-40 ------- Chapter 7 RESULTS AND DISCUSSION OF CARBON STEEL RADIOLOGICAL ASSESSMENT This chapter presents a summary of the radiological impacts of recycling carbon steel scrap from nuclear facilities on the RME individuals, as well as a brief discussion of the results of these analyses. The dose and risk from each radionuclide, each pathway, and every exposure scenario are tabulated in Appendix J. The same results are tabulated more concisely by exposure pathway in Appendix K. 7.1 NORMALIZED DOSES AND RISKS TO THE RME INDIVIDUAL The doses and risks from one year of exposure, normalized to unit specific activities in the scrap, are presented in Appendix K. Table K-l lists the normalized doses to the maximally exposed individual worker from the radionuclides likely to be found on potentially contaminated steel scrap, calculated for each scenario described in Chapter 5. Tables K-2 to K-4 list the contributions to the dose for each exposure pathway: external radiation, inhalation and ingestion. The corresponding lifetime risks of cancer resulting from one year's exposure to the same operations are listed in Tables K-5 to K-8. 7.2 MAXIMUM EXPOSURE SCENARIOS Table 7-1 lists the scenario which would result in the maximum dose to the RME individual from one year's exposure to each radionuclide in the present analysis, as well as the radiation dose and lifetime risk of cancer from one year of exposure to that individual. Several observations can be made about the data in Table 7-1: • The normalized doses vary by more than six orders of magnitude, from a low of 2.1 x 10'4 |lSv per Bq/g from Ni-59 to a high of 330 |lSv from Ac-227+D. Seven of the 44 nuclides and nuclide combinations studied would produce maximum doses greater than 100 |iSv, while 25 others would be in the range of 10 to 100 |lSv. • Individuals exposed as a result of their occupations are the RME individuals for almost all nuclides. The three exceptions are: 1. Sr-90, which, due to its high teachability and low Kd, will readily leach from the slag pile into the groundwater. For this nuclide, the RME individual would be a 7-1 ------- person living near the slag storage yard whose drinking water comes from a potentially contaminated well. 2. C-14, which is potentially volatile (as CO2) and would thus be released to the atmosphere and be incorporated in food crops and animal fodder grown in the vicinity of the steel mill. The RME individual would be a person living near the steel mill who gets a large portion of his food from these crops and farm animals. 3. 1-129, which is volatile and would also be released to the atmosphere and contaminate food crops and animal fodder grown in the vicinity of the steel mill. The RME individual would again be a person living near the steel mill who gets a large portion of his food from these crops and farm animals. • Eight scenarios account for the reasonably maximum doses from all 44 nuclides and nuclide combinations. The groundwater potentially contaminated by slag leachate and the airborne effluent emissions scenarios are discussed above; the remaining six scenarios are discussed in the following sections. 7.2.1 Slag Pile Worker The slag pile worker would receive the highest doses from isotopes of many of the elements that concentrate in slag: Nb-94, Ce-144+D, Pm-147, Eu-152, isotopes of radium, and all the actinides except Ac-227, Th-230, and the plutonium isotopes. For the strong Y-emitters—Nb-94, Ce-144+D, Eu-152, and the two radium isotopes—the primary pathway is external exposure. This is a result of the worker's spending four hours per day exposed to slag in the slag yard—a massive source in close proximity. For the remaining nuclides, the primary pathway is inhalation of slag dust. 7.2.2 Scrap Yard Worker The worker cutting scrap in the scrap yard would receive the highest doses from many of the nuclides that do not concentrate in the slag and are not strong Y-emitters: Fe-55, isotopes of nickel, Tc-99, Ac-227, Th-230, and the plutonium isotopes. The primary pathway is dust inhalation. The worker's use of a cutting torch causes the metal to volatilize, potentially enhancing the radionuclide concentrations in the ambient air. 7-2 ------- Table 7-1. Maximum Exposure Scenarios and Normalized Impacts on the RME Individual from One Year of Exposure Nuclide C-14 Mn-54 Fe-55 Co-60 Ni-59 Ni-63 Zn-65 Sr-90+D Nb-94 Mo-93 Tc-99 Ru-106+D Ag-llOm+D Sb-125+D 1-129 Cs-134 Cs-137+D Ce-144+D Pm-147 Eu-152 Pb-210+D Ra-226+D Ra-228+D Ac-227+D Th-228+D Th-229+D Th-230 Th-232 Pa-231 Maximum Scenario Airborne effluent emissions Lathe operator Scrap yard worker Sailor exposed to hull plate Scrap yard worker Scrap yard worker Truck driver: baghouse dust Slag leachate in groundwater Slag pile worker Sailor exposed to hull plate Scrap yard worker Lathe operator Lathe operator Sailor exposed to hull plate Airborne effluent emissions Truck driver: baghouse dust Truck driver: baghouse dust Slag pile worker Slag pile worker Slag pile worker EAF furnace operator Slag pile worker Slag pile worker Scrap yard worker Slag pile worker Slag pile worker Scrap yard worker Slag pile worker Scrap yard worker Dose mrem EDE per pCi/g 2.5e-04 l.Oe-01 3.7e-06 4.7e-01 7.9e-07 1.9e-06 7.1e-02 1.6e-02 2.3e-01 4.3e-05 1.3e-05 2.6e-02 3.2e-01 6.2e-02 3.3e-01 1.8e-01 6.6e-02 8.3e-03 3.5e-05 1.7e-01 5.6e-01 3.0e-01 1.9e-01 1.2e+00 4.3e-01 4.0e-01 7.5e-02 l.Oe-01 2.5e-01 |lSv per Bq/g 6.7e-02 2.7e+01 l.Oe-03 1.3e+02 2.1e-04 5.1 e-04 1.9e+01 4.2e+00 6.3e+01 1.2e-02 3.6e-03 7.0e+00 8.5e+01 1.7e+01 8.9e+01 5.0e+01 1.8e+01 2.3e+00 9.6e-03 4.6e+01 1.5e+02 8.1e+01 5.2e+01 3.3e+02 1.2e+02 1.1 e+02 2.0e+01 2.8e+01 6.6e+01 Lifetime Risk of Cancer* per pCi/g 1.2e-10 7.7e-08 8.6e-13 3.5e-07 5.0e-13 1.4e-12 5.4e-08 7.7e-09 1.8e-07 3.3e-ll 4.8e-12 2.0e-08 2.4e-07 4.7e-08 1.5e-07 1.4e-07 5.0e-08 6.5e-09 3.06-11 1.3e-07 1.6e-07 2.1e-07 1.2e-07 5.8e-08 3.0e-07 9.9e-08 8.6e-09 1.3e-08 1.4e-08 per Bq/g 3.2e-09 2.1e-06 2.3e-ll 9.5e-06 1.4e-ll 3.8e-ll 1.5e-06 2.1e-07 4.8e-06 8.8e-10 1.3e-10 5.3e-07 6.5e-06 1.3e-06 4.0e-06 3.8e-06 1.4e-06 1.8e-07 8.2e-10 3.5e-06 4.3e-06 5.8e-06 3.1e-06 1.6e-06 8.2e-06 2.7e-06 2.3e-07 3.4e-07 3.7e-07 Maximum risk—may correspond to a different scenario 7-3 ------- Table 7-1 (continued) Nuclide U-234 U-235+D U-238+D Np-237+D Pu-238 Pu-239 Pu-240 Pu-241+D Pu-242 Am-241 Cm-244 U-Natural U-Separated U-Depleted Th-Series Maximum Scenario Slag pile worker Slag pile worker Slag pile worker Slag pile worker Scrap yard worker Scrap yard worker Scrap yard worker Scrap yard worker Scrap yard worker Slag pile worker Slag pile worker EAF furnace operator Slag pile worker Slag pile worker Slag pile worker Dose mrem EDE per pCi/g 3.7e-02 5.2e-02 3.6e-02 1.2e-01 8.0e-02 8.8e-02 8.8e-02 1.6e-03 8.2e-02 1.8e-01 1.2e-01 6.4e-01 7.6e-02 4.0e-02 7.3e-01 |lSv per Bq/g l.Oe+01 1.4e+01 9.7e+00 3.3e+01 2.2e+01 2.4e+01 2.4e+01 4.3e-01 2.2e+01 4.9e+01 3.1e+01 1.7e+02 2.0e+01 l.le+01 2.0e+02 Lifetime Risk of Cancer* per pCi/g 1.6e-08 2.9e-08 1.8e-08 6.5e-08 1.4e-08 1.4e-08 1.4e-08 1.5e-10 1.4e-08 5.1 e-08 3.2e-08 2.6e-07 3.5e-08 2.0e-08 4.3e-07 per Bq/g 4.4e-07 7.9e-07 4.8e-07 1.8e-06 3.8e-07 3.9e-07 3.9e-07 4.0e-09 3.7e-07 1.4e-06 8.5e-07 7.0e-06 9.6e-07 5.3e-07 1.2e-05 Maximum risk—may correspond to a different scenario 7.2.3 Lathe Operator The lathe operator would receive the highest doses from three relatively short-lived nuclides that partition strongly to cast iron: Mn-54, Ru-106+D, and Ag-1 lOm+D. His only potential exposure would be to external radiation from the cast iron in the lathe, since he would be exposed to negligible (if any) amounts of ingestible material or respirable particulates from the metal in this machine. The reason this individual would receive higher external exposures from these nuclides than, say, the scrap yard worker is due to the residually radioactive scrap dilution factors. As was discussed in Chapter 5, only 5.5% of the scrap yard's annual throughput would consist of potentially contaminated scarp. Thus, the scrap worker's exposure is reduced due to the 94.5% uncontaminated scrap in his surroundings. The lathe, however, was assumed to come from a single furnace heat that contained 50% potentially contaminated scrap. Although the radiation source is less massive, this is more than compensated by the nine-fold higher concentration. 7-4 ------- 7.2.4 Sailor Sleeping next to Steel Hull-plate The sailor on a naval support vessel would receive the highest doses from three relatively long- lived photon-emitters that partition strongly to steel: Co-60, Mo-93, and Sb-125. Because of the 18-month delay between the fabrication of the steel hull-plate in the EAF and the vessel's entering normal service (see discussion in Appendix H-2), relatively short-lived nuclides (i.e., those with half-lives of less than 18 months) will have undergone substantial decay prior to the sailor's occupying his berth next to the hull plate. The sailor's exposure from the longer-lived nuclides is greater than the lathe operator's because the large area of the hull plate more than compensates for its smaller mass and correspondingly smaller total activity. 7.2.5 Truck Driver: Baghouse Dust The truck driver hauling baghouse dust would receive the highest dose from the three strong y- emitters that concentrate in the dust: Zn-65 and the two cesium isotopes. Although the scrap yard worker is exposed to a much greater mass of material and spends more time in proximity to the source, the high concentration of these nuclides in the dust results in higher exposures of the driver. 7.2.6 EAF Furnace Operator The EAF furnace operator would receive the highest doses from Pb-210+D because of his internal exposure to potentially contaminated dust and soot. Lead is volatile at steel-melting temperatures; however, the lead vapors condense to an aerosol dust in the cooler air outside the furnace. This dust is inhaled by the steel mill workers; when it settles and forms soot it is also inadvertently ingested. According to measured data on dust loadings at various work stations, the furnace operator would have the highest intake of Pb-210 of the workers modeled in the present analysis. Since Pb-210 is a p-emitter with only one low-intensity, low-energy y-ray, external exposure would not be a significant pathway. 7.3 EVALUATION OF THE RESULTS OF THE RADIOLOGICAL ASSESSMENT The analysis was designed to produce a conservative but reasonable assessment of the potential doses and accompanying risks to individuals resulting from the recycling of potentially contaminated steel scrap. This assessment required the authors to make many assumptions 7-5 ------- regarding the scenarios and the physical processes involved. Several of the assumptions that had a significant effect on the results are discussed in the following sections. 7.3.1 Dilution of Potentially Contaminated Steel Scrap1 Perhaps the most critical assumptions relate to the dilution of the potentially contaminated steel scrap by uncontaminated scrap during and after recycling. Relatively little potentially contaminated steel scrap is being currently released for unrestricted recycling. Once large-scale decommissioning of nuclear facilities takes place, it is difficult to predict how much scrap will in fact be released for recycling, over what period, and with what geographic distribution. The present analysis made a conservative assumption regarding the maximum likely fraction of contaminated scrap in the process materials. Insufficient data is available to determine the probability that all the decommissioning scrap from a given nuclear power plant (or the equivalent amount from more than one plant) would be sent to the same scrap processor in one year. The assumption regarding the end users' being exposed to products containing 50% contaminated steel scrap is conservative but reasonable. Although the average heat in the reference mill over one year was assumed to contain 5.5% contaminated scrap, a statistical analysis showed that there was at least a 10% probability that at least one heat during the year would contain 50% contaminated steel scrap. The use of such statistics is consistent with EPA's philosophy regarding the reasonably maximum exposure as the 90th percentile of the distribution—i.e., exposure conditions that would be exceeded only 10% of the time (see Section 3.1.2). 7.3.2 Exposure Pathways A number of assumptions were made in modeling the exposure pathways for each scenario. These will be discussed separately for each pathway. External Exposure Federal Guidance Report (FOR) No. 12 (Eckerman and Ryman 1993) provides a highly accurate means of assessing the external exposure from an idealized source geometry, if the receptor is a person standing on the source and the source has the same elemental composition as the soil used See Appendix G for a comprehensive discussion of this topic. 7-6 ------- in the FGR 12 dose calculations. FGR 12 gives a reasonable approximation to the three scenarios—the slag storage yard, the road built with slag, and the scrap yard—to which it was applied. In all three cases, the roughness of the surface would tend to reduce the actual exposures from the FGR 12 predictions, as would the higher effective atomic number of the source material.2 The external exposures in the remaining scenarios were modeled using MicroShield 4.2. MicroShield is an industry-standard shielding code that produces reliable results for nuclides with principal y-ray energies greater than 100 keV. With the exception of Mo-93, which is discussed in Section 6.3.1, all the nuclides in the present study which emit photons only in the 10 - 100 keV range would have their principal impacts via the internal exposure pathways. This assumption was confirmed by comparing the dose from the maximum exposure scenario for every such nuclide to the external exposure dose calculated from FGR 12. In every such case, the FGR 12 dose from soil of infinite depth was a small fraction of the calculated dose. Since none of the scenarios have an external exposure geometry that could produce doses higher than FGR 12, any underestimation of the external exposures from low-energy photons would have a negligible effect on the doses listed in Table 7-1. Inhalation The major parameters that affect the dose via the inhalation pathway are the atmospheric concentration (dust loading) and composition of the dust. The dust loading was, in most cases, based on measured values for similar operations. Thus, the dust in the areas occupied by the steel mill crane operator, the furnace operator, and the operator of the continuous caster were based on reported measurements for such workers at an actual EAF steel-making facility, albeit one that primarily produced stainless steel. Since only analyses for toxic trace constituents in the dust were reported, it was not possible to ascertain the origin of the dust, which would have enabled a determination of its hypothetical radioactive contamination. It was therefore assumed that all of the dust came from the furnace emissions—i.e., that it had the same composition as baghouse dust. Since the furnaces are the primary source of airborne emissions in a steel mill, this is not an unreasonable assumption. A discussion of the anticipated effect of atomic number on calculations using the FGR 12 dose coefficients can be found in Section H.2.1. 7-7 ------- The dust loading in the scrap yard was more difficult to determine, since scrap processors do not routinely monitor dust levels. Newton et al. (1987) reported that cutting metal with an oxy- acetylene torch produces aerosol concentrations of 15 mg/m3, and that these particles were primarily of respirable size. Thus, a concentration of 15 mg/m3 and a respirable fraction equal to 1 would be an upper limit for the scrap cutting operation. However, the scrap cutter works outdoors, so the aerosols from his torch would have more of a chance to disperse. It was thus assumed that the total dust is equal to the ACGIH Threshold Limit Value (TLV) of 10 mg/m3 for nuisance dust, and that only 50% of this is respirable. This yields a concentration of 5 mg/m3 of respirable dust, which is also the OSHA PEL. This is a key assumption, inasmuch as dust inhalation by the scrap cutter is the major contributor to the maximum annual dose from several radionuclides. Another key assumption in the scrap cutting scenario is that the dust has the same specific activity as the scrap. An argument could be made that much of the radioactive contamination will be on the surface and that it is these surface layers that are the primary sources of the dust. To counter that argument, one observes that the scrap would have undergone surface decontamination prior to being released, so that loose surface activity would have been removed. The cutting operation was assumed to be the major source of the dust. The aerosols are produced by the melting and volatilization of the steel; their composition can therefore be assumed to be the same as the overall composition of the scrap. Ingestion Ingestion of radionuclides is a major contributor to the maximum doses of two of the RME individuals. The inadvertent ingestion rate of soot by the EAF furnace operator (the RME individual exposed to Pb-210+D), is based on the EPA's "Exposure Factors Handbook" (U.S. EPA 1997) and represents a reasonable maximum value. The soot was assumed to have the same composition as baghouse dust, since, like the dust in the air, its primary source is the fugitive emissions from the furnace. (See discussion of the inhalation pathway, above). Ingestion is the only pathway for the RME individual exposed to Sr-90: the person whose drinking water well may become contaminated by leachate from the slag pile. This analysis includes a number of conservative assumptions, which are discussed in this section. The first assumption regards the teachability of strontium from slag. The leach rate was calculated using diffusion coefficients which were in turn calculated using data from EPA- 7-8 ------- sponsored experimental research conducted by the Brookhaven National Laboratory. One source of uncertainly is the variability of the experimental data—the highest of the three reported values was four times greater than the lowest. Another is the assumed size of slag particles. The analysis assumed an average particle size of 1 cm. If the slag consisted primarily of finer particles, the leach rate would be correspondingly larger. Another assumption concerned the Kd of the soil layer under the slag. For all elements considered in this study, the lowest reasonable Kds for a given soil type, which had been identified in SCA 1997, were used to model the transport of the radionuclides through the soil. The metallic elements (including strontium) are more mobile in an acidic environment. Slag, however, is basic. Thus, the leachate would be basic, causing elements like strontium to be retarded in the soil. This would prolong its migration time, leading to more radioactive decay of Sr-90 and perhaps even preventing any significant amount from reaching the aquifer. In such a case, of course, another scenario would produce the maximum dose. 7.3.3 Mass Fractions and Partitioning of Contaminants The mass fractions of metal and non-metallic components of the steel or cast iron, slag, dust, and home scrap were determined from a definitive study of the literature and consultations with other research workers and technical experts. The data on partitioning of contaminants among these various media was less definitive. Nevertheless, a major and largely successful effort was made to combine the observed partitioning with thermodynamic calculations to produce a set of reasonable and defensible concentration factors. The only conscious conservatism that was introduced into this phase of the analysis was the simultaneous use of high-end partition fractions for two or more media, which, as was discussed at the end of Section 6.2, would overstate the exposure of the operator of the continuous caster to four of the radionuclides studied. Since this individual did not prove to be the RME individual for any nuclide, this assumption has no effect on the maximum doses listed in Table 7-1. 7.3.4 Scenario Selection The scenarios used in the present analysis were selected from a much longer list which had been examined in an earlier analysis of recycling residually radioactive steel scrap (SCA 1995). Scenarios in the previous analysis which were redundant or which had no potential for producing the maximum doses from any nuclide were dropped from the present analysis. Given the 7-9 ------- conservative assumptions used, it is improbable that any plausible scenario would produce greater impacts than those studied. One scenario that was not part of the assessment was the use of slag as a soil conditioning agent. A scoping analysis was performed to determine if this could be a significant exposure pathway for any of the radionuclides considered in the present analysis. A brief description of this assessment is presented below. A more detailed discussion can be found in Appendix H-2. Because of its high lime (CaO) content, slag is sometimes used to raise the pH of acidic soils. According to a vendor of gardening supplies, even highly acidic soils do not require more than about 100 Ib of liming agent per 1,000 ft2 (-500 g/m2). The liming agent supplied by this vendor contained 48% lime, which is comparable to the CaO content of steel slags listed in Appendix I. The doses from the consumption of agricultural products grown in this soil were calculated, assuming that slag were applied to agricultural soil in the same quantity, and that it were mixed into the top six inches (15 cm) of soil (the assumed plow depth). The normalized annual dose to the maximally exposed individual via the food ingestion pathway was calculated for each radionuclide that partitions to the slag by using the dose factors calculated for the agricultural produce pathway for a generic site with radioactively contaminated soil, listed in Table 3-1 of SCA 1997. The doses calculated for this scenario for each radionuclide that would partition to the slag were one to three orders of magnitude less than the doses listed in Table 7-1. 7.3.5 Implementation of Clearance Criteria The normalized dose from each radionuclide listed in Table 7-1 was calculated on the assumption that all of the residually radioactive carbon steel scrap released from a nuclear facility would be uniformly contaminated by that nuclide alone. Were that the case, it would be appropriate to calculate a clearance level in terms of the specific activity of that nuclide which would meet a stipulated release criterion. Say the normalized dose from nuclide /' is equal to D; and the Agency were to stipulate that the dose to the RME individual should not exceed a value Dj.3 Then the maximum allowable concentration of nuclide /' could be calculated by simply dividing Dj by D;: At this time, the Agency has made no decision regarding dose limits or clearance criteria for any material released from any nuclear facility. 7-10 ------- C'y = clearance level of radionuclide /' corresponding to Dj, assuming uniform specific activity (Bq/g) Dj = hypothetical dose limit (|lSv/a) D; = normalized dose from radionuclide/(Table 7-1) (|lSv/a per Bq/g) In reality, the specific activities of residually contaminated steel would span a range of values, with the lower end of the range being essentially zero—no activity detectable above background. In most of the exposure scenarios presented in Table 7-1, the dose to the exposed individual is based on the average concentration in the waste stream over a one-year period. If most of the material that could be cleared at level C'y were well below this level (i.e., if the average specific activity were well below the maximum), the clearance level could be set considerably higher without exceeding the (hypothetical) stipulated dose limit to that individual. Analogous considerations would apply to the end-user scenarios (i.e., the sailor sleeping next to steel hull- plate or the lathe operator). Furthermore, in most cases the residually radioactive metals would be contaminated with a suite of radionuclides that do not necessarily have the same maximum exposure scenarios—a notable example are the gaseous diffusion plants, where the primary contaminants are uranium and Tc-99. In such cases, the sum-of-fractions rule which is applied by the NRC to gaseous and liquid effluents need not be used in implementing clearance levels, since the same RME individual would not be exposed to all the nuclides. The above brief discussion illustrates some of the issues involved in using the normalized doses listed in Table 7-1 to implement a dose- or risk-based clearance criterion. 7-11 ------- REFERENCES Eckerman, K. F., and J. C. Ryman. 1993. "External Exposure to Radionuclides in Air, Water, and Soil," Federal Guidance Report No. 12, EPA 402-R-93-081. U.S. Environmental Protection Agency, Washington, DC. Newton, G. J., et al. 1987. "Collection and Characterization of Aerosols from Metal Cutting Techniques Typically Used in Decommissioning Nuclear Facilities." American Industrial Hygiene J. 48:922-932. S. Cohen & Associates (SCA). 1995. "Analysis of the Potential Recycling of Department of Energy Radioactive Scrap Metal." 4 vols. Prepared for U.S. Environmental Protection Agency, Office of Radiation and Indoor Air, Washington, DC. S. Cohen & Associates (SCA). 1997. "Radiation Site Cleanup Regulations: Technical Support Document for the Development of Radionuclide Cleanup Levels for Soil." Vol. 1. Prepared for U.S. Environmental Protection Agency, Office of Radiation and Indoor Air, Washington, DC. U.S. Environmental Protection Agency (U.S. EPA), National Center for Environmental Assessment, Office of Research and Development. 1997. "Exposure Factors Handbook," EPA/600/P-95/002Fa. Vol. 1, "General Factors." U.S. EPA, Washington, DC. 7-12 ------- Chapter 8 RADIOLOGICAL ASSESSMENT OF RECYCLING ALUMINUM Detailed descriptions of the practices of recycling aluminum are presented in Appendix B. The present chapter recapitulates those aspects of these practices which are relevant to the assessment of the radiation exposures of individuals. The exposure pathways are the same as those discussed in Section 5.3. Differences in exposure parameters applicable to the aluminum assessment are described in the following sections. Figure 8-1 presents a simplified diagram depicting the mass flow during aluminum recycling. As shown in the figure, the aluminum scrap from normal commercial sources as well as from a nuclear facility is sent to a reverberatory furnace to be smelted. The furnace produces aluminum alloys, as well as the smelting by-products: dross and offgas. The dross, primarily consisting of metallic oxides and halide salts, is analogous to the slag produced during the melt-refining of carbon steel, while the offgas contains volatile products and aerosols emitted by the furnace. 8.1 DISTRIBUTION OF CONTAMINANTS 8.1.1 Material Balance The following mass fractions were adopted for the present analysis1 (see Section B.6.1): Furnace Charge: • Aluminum scrap 0.735 • Heel from previous melt . 0.25 • Silicon 0.015 Output: • Aluminum casting alloy . . 0.943 (0.25 left in furnace) • Baghouse dust 0.00225 (0.0015 metal) • Dross 0.112 (0.056 A12O3, 0.045 halide salts, 0.0112 Al metal) Most of these values are lower than those cited in Section B.6.1. The latter value refers to the scrap + silicon charged to the furnace, while the present values refer to the total metal in the furnace, which includes the heel from the previous melt. 8-1 ------- flux Silicon scrap processing contaminated Al scrap clean Al scrap 22% direct automotive 44% automotive related 8% small engine smelting (reverberatory furnace) offgas emissions control system house dust landfill Figure 8-1. Simplified Material Flow for Secondary Aluminum Smelter The sum of the output fractions is greater than 1 due to the addition of flux to the furnace charge, as well as oxidation of the aluminum in the dross. 8.1.2 Contaminant Partitioning Table 8-1 lists the partition ratios of the various elements, taken from Table B-13. The concentration factors are calculated according to the methodology presented in Section 6.2, using the mass fractions in Section 8.1.1. The calculation of the concentration factor in the furnace charge assumes that the scrap consists of 98.5% old scrap, and 1.5% scrap recovered from the dross produced in previous melts. This recovered scrap would have the same specific activities as the finished metal. 8-2 ------- Table 8-1. Partition Ratios (PR) and Concentration Factors (CF) in Aluminum Smelting Element Ac Ag Am C Ce Cm Co Cs Eu Fe I Mn Mo Nb Ni Np Pa Pb Pm Pu Ra Ru Sb Sr Tc Th U Zn Furnace Charge CFa 0.85 0.99 0.85 0.75 0.85 0.85 0.98 0.72 0.85 0.98 0.72 0.98 0.99 0.98 0.98 0.85 0.98 0.99 0.85 0.85 0.85 0.99 0.99 0.75 0.99 0.85 0.85 0.98 Metal PR (%) 1 -50 100 1 -50 1 -10 1 -50 1 -50 90-99 0 1 -50 90-99 0 90-99 100 90-99 90-99 1 -50 1 -99 100 1 -50 1 -50 1 -50 100 100 1 -10 100 1 -50 1 -50 90-99 CF 0.44 1.03 0.44 0.08 0.44 0.44 1.01 0.00 0.44 1.01 0.00 1.01 1.03 1.01 1.01 0.44 1.01 1.03 0.44 0.44 0.44 1.03 1.03 0.08 1.03 0.44 0.44 1.01 Dross PR (%) 50- 99 50- 99 90- 99 50- 99 50- 99 1 - 10 100 50- 99 1 - 10 50- 100 1 - 10 1 - 10 1 - 10 50- 99 99- 1 50- 99 50- 99 50- 99 90- 99 50- 99 50- 99 1 - 10 CF 7.07 0 7.07 7.07 7.07 7.07 0.95 7.11 7.07 0.95 7.11 0.95 0.00 0.95 0.95 7.07 7.07 0 7.07 7.07 7.07 0 0 7.07 0.00 7.07 7.07 0.95 Release Fraction13 8.88e-05 2.05e-04 8.88e-05 1.56e-05 8.88e-05 8.88e-05 2.03e-04 O.OOe+00 8.88e-05 2.03e-04 0.0-0.5 2.03e-04 2.05e-04 2.03e-04 2.03e-04 8.88e-05 2.03e-04 2.05e-04 8.88e-05 8.88e-05 8.88e-05 2.05e-04 2.05e-04 1.56e-05 2.05e-04 8.88e-05 8.88e-05 2.03e-04 Refers to the scrap aluminum (98.5% old scrap and 1 Atmospheric releases, assuming compliance with the .5% scrap recovered from dross) + silicon charged to the furnace draft EPA emission standard of 0.4 Ib per ton of furnace charge 8-3 ------- 8.2 LIST OF OPERATIONS AND EXPOSURE SCENARIOS Table 8-2 lists the operations and exposure parameters employed in the radiological assessment. The descriptive title of each individual exposure scenario—used to assess the exposure of a given individual—is italicized. (Sub-scenarios listed beneath the individual scenario refer to different activities performed by the same individual.) These operations are described in the following sections. The scenarios were selected from a larger list of possible scenarios discussed in Appendix B. The aim of the selection was to ensure that the reasonable maximum exposures from each radionuclide would be evaluated. Scenarios were omitted if the radiation sources, exposure pathways, and exposure durations were such that the radiation exposures would be bounded by the scenarios already selected. Further details are found in Appendix B. 8.2.1 Dilution Scrap Transport It is assumed that aluminum scrap from a nuclear facility would be transported to a smelter in a dedicated truck. Therefore, the dilution factor for this operation is equal to 1. Smelter Operations Unlike carbon steel, movement of aluminum scrap is not geographically constrained by haulage costs. The reference smelter for the aluminum recycling assessment is based on the Wabash Alloys facility in Dickson, Tenn., which has an annual capacity of about 75,000 tons (68,000 metric tons [t]). As discussed in Section B.I.I, the largest anticipated release of aluminum scrap from a single facility in any one year is the 2,527 t of aluminum from Paducah available each year from 2016 to 2022. Assuming all of this scrap is processed at the reference smelter, the dilution factor is 0.037. The same factor is applied to the industrial use scenario. Dross Disposal The dilution factor for the dross being transported for disposal is the same as for the smelter operations. In the realistic landfill burial scenario described in Appendix L, the dross from the reference smelter is commingled with dross from other smelters, resulting in an overall dilution factor of 1.3 x 10"3. The derivation of this factor is discussed in Appendix L. 8-4 ------- Table 8-2 Exposure Scenarios and Parameters for Radiological Assessments of Aluminum Recycling Description SCRAP TRANSPORT. Truck driver SECONDARY SMELTER Scrap Operations Scrap handler Shredder operator Furnace Operations Furnace operator: near furnace misc. duties Airborne effluent emissions Handling Ingots Skimmer and stacker skim ingot stack ingots operate fork lift Dross Disposal Dross transport: truck driver Dross buried in landfill Dilution factor 1. 0.037 0.0013 Exposure Pathways External Exposure Time (hr/y) 1000 § CO % b 8ft Medium scrap Internal Time (hr/y) Medium Dust load (mg/m3) RFa N/A 875 875 1750 b 10ft 3-10ftd scrap 1750 1750 scrap 0.85 10 0.6 0.5 1500 250 6ft 25ft scrap 1750 dustc 0.57 0.6 N/A 250 750 750 1.5ft 40 in 15 in 1.5- 6ft 3ft metal dross metal metal metal 1750 dust 0.85 0.6 1000 8ft dross N/A N/A INDUSTRIAL USE OF MILL PRODUCTS Aluminum fabricator END USERS Taxi driver: engine block Truck driver: fuel tank Cooking in aluminum pan 0.037 0.5 1300 1 ft metal 875 metal 7.66 1.0 3300 3000 263 0.8 m 2.5ft 2ft metal N/A N/Ae metal N/A Respirable fraction 1 Exposure assessment uses FOR 12 dose coefficients—see discussion in Section 6.3.1 Dust = baghouse dust Range of distances—see discussion in Section 6.3.1 Exposure from ingestion of contaminated food 8-5 ------- End Users As was the case with carbon steel scrap, it is highly unlikely that all of the aluminum scrap in a single furnace heat will be from a nuclear facility. In addition to dilution caused by charging the furnaces with different batches of scrap, about 25% of the charge consists of molten aluminum from the previous heat. Thus, a dilution factor of 0.5—the value used for finished products made of carbon steel—is reasonably conservative for the aluminum assessment. 8.2.2 Scrap Transport The scrap transport worker is a truck driver who spends eight hours per day in the cab of a truck, carrying 20-t loads of scrap metal to the scrap processor and returning with an empty truck (or carrying other cargo).2 His only exposure would be to external radiation from the load of contaminated scrap. The MicroShield computer program was used to calculate normalized dose rates to the driver from external exposure. The scrap was assumed to fill a trailer that was 48 ft long, 8 ft wide and 9l/2 ft high—typical dimensions for a large cargo trailer. The driver was assumed to sit 8 ft in front of the load, the doses being calculated for the posteroanterior (PA) exposure geometry, which corresponds to the driver having his back to the load. The attenuation of the walls of the trailer and cab and of the driver's seat were neglected, leading to a slightly conservative assessment. 8.2.3 Secondary Smelter Operations Aluminum recycling operations at a secondary smelter may be divided into four categories: (1) scrap handling and processing, (2) furnace operations, (3) handling and processing of ingots, and (4) disposal of dross. Two scrap workers—a scrap handler and a shredder operator—were included in the exposure assessments, as were a furnace operator, a worker who performs different tasks involved in the processing and handling of ingots, and a truck driver who transports the dross for disposal. All these workers would be exposed to direct radiation from the residual radioactivity of the metal. All but the driver would also inhale potentially contaminated dust in the ambient air and ingest deposited particulate matter. According to Section B.6.2, transporting the 2,527 t/a of scrap from generated at Paducah to Dickson, Term. would require 126 trips. The analysis of this scenario, however, considers the possibility that the scrap may be smelted at a more distant facility, which would require the driver of a dedicated truck to spend an entire year transporting this material. 8-6 ------- Scrap Handler The scrap handler moves scrap from the stockpiles to the shredder or the furnace using a front- end loader with a 5-yd3 bucket—the bucket would be loaded one-half of the time. The sources of his external exposure would be the load in the bucket and the scrap piles. The contents of the bucket were modeled as a 3-m-long cylinder, 1.3 m in diameter, with a bulk density of 1.08 g/cm3. Because of the size and distribution of the scrap piles, this source was modeled as one-half of an infinite plane, using the dose coefficients for exposure to soil contaminated to an infinite depth, as listed in Federal Guidance Report (FGR) No. 12 (Eckerman and Ryman 1993), but dividing each value by 2. The dust loading is the average of the eight measured values listed in Table B-7. The respirable fraction—the mass fraction with aerodynamic diameters < 10 |lm—was taken from the size distribution of particles in uncontrolled particulate emissions from refining operations in secondary aluminum smelters employing reverberatory furnaces, as listed in U.S. EPA 1995. Shredder Operator The shredder operator stands near the scrap conveyor, which transports a stream of scrap, 3 ft wide by 1A ft deep with a bulk density of 50%. A dust loading of 12.17 mg/m3 had been measured at a scrap conveyer (see Section B.4.3). The analysis assumes that the operator's non- radiological inhalation exposure would comply with recommendations of the American Council of Governmental and Industrial Hygienists (ACGIH 1996),3 so that the total dust loading would not exceed a time-weighted average (TWA) of 10 mg/m3 and the respirable dust concentration would not exceed 5 mg/m3. Furnace Operator The furnace operator spends most of his time tending the furnace at an average distance of 6 ft from the charge and about one hour per day performing other duties which place him 25 ft from the furnace. The ACGIH total dust loading is lower than the OSHA PEL of 15 mg/m3 for nuisance dust; however, the TWA concentration of respirable dust is the same as the OSHA PEL. 8-7 ------- Skimmer and Stacker During the pouring of the melt from the furnace, a worker is assigned to skimming dross from the ingot surface. Because this is a part-time activity, the same person is assumed to work on the crew stacking ingots onto pallets. These crew members divide their time between manually stacking the ingots and transporting the pallets with a forklift. During the dross skimming operation, the worker's external exposure is from an ingot which measures 4 x 4 x 22.5 inches and weighs 35 Ib, and from a waste container 40 inches away that, on average, is half-full of dross. The dross is modeled as a rectangular solid measuring 20 x 78 x 41 in. During stacking, he is exposed to one ingot carried close to his body (assumed to be 15 inches from the center of the body) and to the pile of 27 ingots—one-half of a fully stacked pallet—at a distance that varies between 1.5 and 6 ft. While operating the fork lift, he is exposed to a fully stacked pallet of 54 ingots. Dross Transport The dross transport worker is a truck driver who spends eight hours per day in the cab of a truck, carrying 20-ton loads of dross for processing or disposal and returning with an empty truck (or carrying other cargo). His only exposure would be to external radiation from the load of contaminated dross. This assessment used the same exposure geometry as that of the carbon steel scrap truck, which is described in Section H.I.I. 8.2.4 Industrial Uses of Mill Products: Aluminum Fabrication The aluminum fabrication worker performs gas metal arc welding on a wrought aluminum base. His exposure pathways consist of external exposure to the base metal and the inhalation of fumes from welding. When not wearing his helmet, he would be exposed to the inadvertent ingestion of deposited particulates. The source of external exposure is modeled as a 4.7-ft-square, /^-inch- thick sheet of aluminum. The exposure rates for all y-emitting nuclides except Mo-93 were calculated by means of the MicroShield computer program, as described in Section 6.3.1. The exposure rate from Mo-93 was calculated in an manner analogous to that used to calculate the exposure rate from the hull plate, also described in Section 6.3.1. The fume concentration inside the helmet was assumed to be the highest of the values listed in Table B-16. ------- 8.2.5 Use of Finished Products As is the case with steel, aluminum is also is used to make a virtually endless variety of finished products. The analysis considers three users of finished products: the driver of a truck with an aluminum fuel tank, the driver of a taxi with an aluminum engine block, and a person who cooks in an aluminum utensil. Taxi Driver The maximally exposed taxi driver is assumed to be an owner/operator who drives a taxi with an aluminum engine. According to the information presented in Section B.6.2, the largest aluminum engine block weighs about 80 Ib (36 kg). MicroShield was used to calculate normalized external dose rates to this driver. The dimensions of the block, as well as other exposure parameters, are assumed to be the same as for the corresponding scenario in the carbon steel analysis (see Sections H.I 1.1 and H.I3.2). The dose rates from Th-232 and Ra-228+D were adjusted to account for the ingrowth of progeny during the useful life of this product, as discussed in Section 8.4.4. Driver of Truck with Aluminum Fuel Tank The only exposure pathway of the driver of a truck with an aluminum fuel tank is to direct radiation from the tank, which is located under the cab of the truck. The tank is made of H-inch thick aluminum and is 1 ft high and 3.7 ft square, with a capacity of 100 gal. On average, the tank would be half-full of fuel. The drivers sits 21/2 feet above the tank, and is shielded by an additional Vi inch of aluminum in the floor of the cab. Cooking Utensil A consumer cooking food in an aluminum frying pan may be exposed to direct radiation from the metal in addition to eating food which may be contaminated with residual radioactivity that has leached from the pan. Blumenthal (1990) wrote that"... a person using uncoated aluminum pans for all cooking and food storage every day would take in an estimated 3.5 milligrams of aluminum daily." This intake rate serves as a conservative upper bound for the leaching of aluminum from the residually contaminated frying pan. The external exposures calculated for the cast iron pan serve as a conservative upper bound for the present analysis. This is because the greater mass of the iron pan would contain a higher total activity of a given radionuclide, which more than compensates for the slightly higher self-absorption of iron vs. aluminum. Since the external exposure from this small object, which is used for a relatively few hours per year, is 8-9 ------- not a significant dose pathway, the use of the more conservative results has little impact on the analysis. 8.2.6 Off-Site Individuals Exposed to Smelter By-Products Additional exposure assessments were performed on two off-site individuals. One is a nearby resident who is exposed to the unfiltered airborne effluents from the smelter. The other resides near an industrial landfill used to dispose of the dross. Impact of Fugitive Airborne Emissions on Nearby Residents The assessment of nearby residents exposed to fugitive airborne emissions of C-14 and 1-129 from the furnace utilized the results of the analysis of this pathway described in Chapter 6. The analysis of other radionuclides was based on a previous assessment of the recycling of carbon steel scrap (SCA 1995), which explicitly evaluated this pathway for 24 of the nuclides in the present analysis. The impacts were adjusted for the annual releases of each nuclide, as follows: D. -^ Q; Dia = 50-year dose commitment from airborne effluent releases of nuclide /' D'ia = dose commitment from airborne effluent releases of nuclide /' from previous analysis Q; = activity of nuclide /' released from smelter in one year = friMc fri = release fraction of nuclide /' (see Table 8-1) Mc = mass of cleared aluminum scrap processed in one year = 2.527 Gg (2,527 t) Q; = released activity of nuclide / in previous analysis The doses from C-14 and 1-129 were adjusted for the dose conversion factors from ICRP Publication 68 (ICRP 1994), as discussed in Section 6.4.3. The assessment of impacts from the airborne emissions scenario is presented in Table 8-3. 8-10 ------- Table 8-3. Normalized Impacts from One Year of Exposure to Fugitive Airborne Emissions Nuclide C-14 Mn-54 Co-60 Ni-59 Ni-63 Sr-90+D Tc-99 Ru-106+D 1-129 Cs-134 Cs-137+D Pb-210+D Ra-226+D Ra-228+D Ac-227+D Th-228+D Th-229+D Th-230 Th-232 Pa-231 U-234 U-235+D U-238+D Np-237+D Pu-238 Pu-23 9/240 Pu-241+D Pu-242 Am-241 Reference Analysis'1 |lCi/y 11,000 13.1 13.1 — 13.1 13.1 13.1 13.1 15,000 1,310 1,310 1,310 13.1 13.1 13.1 13.1 13.1 13.1 13.1 13.1 13.1 13.1 13.1 13.1 — 13.1 13.1 — 13.1 Dose (mrem/y) 4.86e-04 7.46e-05 8.80e-04 — 4.16e-07 9.01e-05 8.41e-07 3.77e-05 4.48e-01 3.56e-02 4.08e-02 1.60e-01 4.86e-03 7.67e-04 4.01e-01 1.44e+00 1.29e-01 1.94e-02 9.77e-02 7.65e-02 7.87e-03 7.39e-03 7.04e-03 3.23e-02 — 2.56e-02 4.93e-04 — 2.65e-02 Risk 2.13e-10 4.306-11 4.80e-10 — l.OOe-13 9.20e-12 5.00e-13 2.106-11 2.52e-07 2.07e-08 2.25e-08 2.60e-08 3.21e-09 3.00e-10 4.85e-09 5.41e-07 5.08e-09 1.03e-09 1. 15e-09 1.46e-09 8.40e-10 8.10e-10 7.50e-10 2.08e-09 — 1.66e-09 1.706-11 — 2.31e-09 Aluminum Release Fraction13 1.56e-05 2.03e-04 2.03e-04 2.03e-04 2.03e-04 1.56e-05 2.05e-04 2.05e-04 5.00e-01 O.OOe+00 O.OOe+00 2.05e-04 8.88e-05 8.88e-05 8.88e-05 8.88e-05 8.88e-05 8.88e-05 8.88e-05 2.03e-04 8.88e-05 8.88e-05 8.88e-05 8.88e-05 8.88e-05 8.88e-05 8.88e-05 8.88e-05 8.88e-05 |lCi/y 0.0394 0.513 0.513 0.513 0.513 0.039 0.518 0.518 1,264 0.000 0.000 0.518 0.224 0.224 0.224 0.224 0.224 0.224 0.224 0.513 0.224 0.224 0.224 0.224 0.224 0.224 0.224 0.224 0.224 Dosec 1.8e-09 2.9e-06 3.4e-05 5.9e-09 1.6e-08 2.7e-07 3.3e-08 1.5e-06 5.6e-02 O.Oe+00 O.Oe+00 6.3e-05 8.3e-05 1.3e-05 6.9e-03 2.5e-02 2.2e-03 3.3e-04 1.7e-03 3.0e-03 1.3e-04 1.3e-04 1.2e-04 5.5e-04 4.0e-04 4.4e-04 8.4e-06 4.2e-04 4.5e-04 Riskd l.le-15 1.7e-12 1.9e-ll 1.3e-15 3.9e-15 2.8e-14 2.0e-14 8.3e-13 2.5e-08 O.Oe+00 O.Oe+00 l.Oe-11 5.5e-ll 5.1e-12 8.3e-ll 9.3e-09 8.7e-ll 1.8e-ll 2.06-11 5.7e-ll 1.4e-ll 1.4e-ll 1.3e-ll 3.6e-ll 2.8e-ll 2.8e-ll 2.9e-13 2.7e-ll 4.0e-ll a C-14 and 1-129 results from carbon steel analysis b Table 8-1 0 mrem EDE per pCi/g in scrap d Lifetime risk of cancer per pCi/g in scrap in present report; results for other nuclides from SCA 1995 8-11 ------- Fourteen of the nuclides considered in the present analysis were not included in either of the analyses cited above. In some of the omitted cases, however, a different isotope of the same element can serve as a surrogate for the missing radionuclide. The impacts of Ni-59 can be estimated from the results for Ni-63. Both isotopes emit p-rays or Auger electrons, and neither emits penetrating photons. The report of the earlier analysis (SCA 1995) shows that the principal impact of atmospheric emissions of Ni-63 was via the inhalation pathway. The Ni-59 dose was therefore estimated from the ratio of the dose conversion factors (DCFs) for inhalation of the two isotopes, as listed in Federal Guidance Report No. 11 (Eckerman et al. 1988), the source of the DCFs in the earlier analysis. The risk was estimated from the ratio of the corresponding slope factors (U.S. EPA 1994). Pu-238, Pu-239, Pu-240 and Pu-242 are all CC-emitters with no short-lived progenies. Pu-239 and Pu-240 have almost identical DCFs for both the ingestion and inhalation pathways—they are therefore listed on the same line in the table. The doses and risks from Pu-238 and Pu-242 via this pathway were estimated on the basis of the corresponding values for Pu-239 by a method analogous to the one used for Ni-59. Disposal of Dross in an Industrial Landfill Dross is commonly buried in a RCRA Subtitle D solid waste landfill. It is possible that a nearby resident would be exposed by drinking groundwater contaminated by leachate from the landfill. This could only occur after the closure of the landfill and loss of institutional control, after which the cap is assumed to degrade and fail. The details of the analysis are presented in Appendix L. 8.3 RESULTS The results of the aluminum recycling assessment are shown in Table 8-4. Several observations can be made about these data: • The normalized doses from all but one of the radionuclides (C-14) from the recycling of aluminum are lower than from the recycling of carbon steel. • Workers are the RME individuals for all nuclides except C-14,1-129, and Np-237. • Three scenarios account for the reasonably maximum doses from all 44 nuclides and nuclide combinations. These are discussed in the following sections. 8-12 ------- Table 8-4. Maximum Exposure Scenarios and Normalized Impacts on the RME Individual from One Year of Exposure Nuclide C-14 Mn-54 Fe-55 Co-60 Ni-59 Ni-63 Zn-65 Sr-90+D Nb-94 Mo-93 Tc-99 Ru-106+D Ag-llOm+D Sb-125+D 1-129 Cs-134 Cs-137+D Ce-144+D Pm-147 Eu-152 Pb-210+D Ra-226+D Ra-228+D Ac-227+D Th-228+D Th-229+D Th-230 Th-232 Pa-231 Maximum Scenario Dross in landfill Scrap truck driver Scrap shredder Scrap truck driver Scrap shredder Scrap shredder Scrap truck driver Scrap shredder Scrap truck driver Scrap shredder Scrap shredder Scrap truck driver Scrap truck driver Scrap truck driver Dross in landfill Scrap truck driver Scrap truck driver Scrap truck driver Scrap shredder Scrap truck driver Scrap shredder Scrap truck driver Scrap truck driver Scrap shredder Scrap truck driver Scrap shredder Scrap shredder Scrap shredder Scrap shredder Dose mrem EDE per pCi/g 3.4e-04 6.7e-02 2.9e-06 2.0e-01 6.2e-07 1.5e-06 4.6e-02 3.7e-04 1.3e-01 2.6e-05 9.3e-06 1.7e-02 2.2e-01 3.3e-02 6.5e-02 1.3e-01 4.5e-02 3.5e-03 8.0e-06 8.9e-02 l.Oe-02 1.4e-01 7.3e-02 9.4e-01 1.2e-01 1.7e-01 5.9e-02 6.1e-02 1.9e-01 |iSv per Bq/g 9.2e-02 1.8e+01 7.8e-04 5.4e+01 1.7e-04 4.0e-04 1.3e+01 9.9e-02 3.4e+01 7.1e-03 2.5e-03 4.5e+00 5.9e+01 9.0e+00 1.7e+01 3.4e+01 1.2e+01 9.5e-01 2.2e-03 2.4e+01 2.8e+00 3.7e+01 2.0e+01 2.5e+02 3.1e+01 4.5e+01 1.6e+01 1.7e+01 5.1e+01 Lifetime Risk of Cancera per pCi/g 1.6e-10 5.1 e-08 6.7e-13 1.5e-07 4.0e-13 l.le-12 3.5e-08 9.9e-ll 9.6e-08 2.06-11 2.9e-12 1.3e-08 1.7e-07 2.5e-08 2.9e-08 9.5e-08 3.4e-08 2.7e-09 4.7e-12 6.7e-08 2.8e-09 l.Oe-07 5.6e-08 3.2e-08 8.8e-08 3.3e-08 6.8e-09 1.1 e-08 9.6e-09 per Bq/g 4.4e-09 1.4e-06 1.8e-ll 4.1e-06 l.le-11 3.06-11 9.5e-07 2.7e-09 2.6e-06 5.4e-10 7.9e-ll 3.4e-07 4.5e-06 6.8e-07 7.9e-07 2.6e-06 9.3e-07 7.2e-08 1.3e-10 1.8e-06 7.6e-08 2.8e-06 1.5e-06 8.5e-07 2.4e-06 8.8e-07 1.8e-07 3.0e-07 2.6e-07 Maximum risk—may correspond to a different scenario 8-13 ------- Table 8-4. (continued) Nuclide U-234 U-235+D U-238+D Np-237+D Pu-238 Pu-239 Pu-240 Pu-241+D Pu-242 Am-241 Cm-244 U-Natural U-Separated U-Depleted Th-Series Maximum Scenario Scrap shredder Scrap shredder Scrap shredder Dross in landfill Scrap shredder Scrap shredder Scrap shredder Scrap shredder Scrap shredder Scrap shredder Scrap shredder Scrap shredder Scrap shredder Scrap shredder Scrap truck driver Dose mrem EDE per pCi/g 1.3e-02 l.le-02 l.le-02 6.5e-02 6.3e-02 6.9e-02 6.9e-02 1.2e-03 6.4e-02 5.7e-02 3.7e-02 1.5e-01 2.4e-02 1.2e-02 1.9e-01 |iSv per Bq/g 3.4e+00 3.1e+00 2.9e+00 1.8e+01 1.7e+01 1.9e+01 1.9e+01 3.4e-01 1.7e+01 1.5e+01 9.9e+00 4.1e+01 6.4e+00 3.3e+00 S.le+01 Lifetime Risk of Cancera per pCi/g 5.5e-09 8.3e-09 4.9e-09 4.7e-08 l.le-08 l.le-08 l.le-08 1.2e-10 l.le-08 1.5e-08 9.7e-09 l.le-07 l.le-08 5.5e-09 1.4e-07 per Bq/g 1.5e-07 2.2e-07 1.3e-07 1.3e-06 3.0e-07 3.0e-07 3.0e-07 3.1e-09 2.9e-07 4.2e-07 2.6e-07 2.9e-06 2.9e-07 1.5e-07 3.9e-06 Maximum risk—may correspond to a different scenario 8.3.1 Shredder Operator The operator of the scrap-shredding machine is the RME individual for Fe-55, the two nickel isotopes, Sr-90+D, Mo-93, Tc-99, Pm-147, Pb-210+D, all but two of the actinides, and the three combinations of uranium isotopes and their progenies. These nuclides deliver most of their dose via the internal exposure pathways. Since the shredder operator is exposed to high concentrations of contaminated dust in the ambient air, he would have the greatest potential exposures via this pathway. 8.3.2 Scrap Transport Worker The driver of the truck transporting cleared scrap to the scrap processor is the RME individual for 14 of the radionuclides that are strong y-emitters as well as for the thorium radioactive decay series. The large mass of metal at a relatively short distance results in the highest external exposures, which are the principal exposure pathways for this group of nuclides. 8-14 ------- 8.3.3 Disposal of Dross in an Industrial Landfill Burial of the dross in a landfill leads to the highest exposures from C-14,1-129, and Np-237. These are long-lived isotopes of elements that have low Kds and hence would reach the aquifer within the 1,000 year time frame of the assessment. All three nuclides partition to the dross and deliver their doses via the internal exposure pathways. 8.4 EVALUATION OF THE RESULTS Many of the observations in Section 7.3 regarding the radiological assessment of steel scrap are applicable to the aluminum analysis. The relevant issues from the steel analysis, as well as questions that are unique to aluminum, are discussed in the following paragraphs. 8.4.1 Dilution of Potentially Contaminated Scrap The assumption that all the aluminum scrap released from Paducah would be sent to a single facility is not unreasonable. The secondary smelter in question is relatively near Paducah; however, the intrinsic value of aluminum is such that transportation costs are not the determining factor in selecting a recycling facility. 8.4.2 Exposure Pathways External Exposure The comments about the external exposure calculations in Section 7.3.2 are applicable here and need not be repeated. Inhalation In the scrap shredder scenario, where the inhalation of dust and/or fumes is a major pathway, the aerosol concentration is based on an actual measured value. The assumption that all of the dust comes from the metal is reasonable for this scenario. 8.4.3 Airborne Effluent Releases The evaluation of airborne effluent releases presented in SCA 1995, the basis for the assessment of the radionuclides listed in Table 8-3 (except C-14 and 1-129) used much more conservative parameters than those used with CAP-88 in the present analysis. The earlier study assumed that all the food consumed by the RME individual was home-grown, that the radionuclide 8-15 ------- concentrations in the soil reflected buildup over a ten-year period of continuous emissions, and that the individual resided between 100 m and 500 m from the facility, rather than the 1-km distance assumed in the present analysis. Additional differences are caused by the use of FOR 11 dose factors in the earlier analysis4. The calculated doses for this pathway are one to five orders of magnitude smaller than the maximum doses from the same radionuclides as listed in Table 8-4, with the exception of 1-129, which employed the more realistic CAP-88 assessment5. Consequently, this scenario would not deliver the maximum dose from any of the nuclides listed in Table 8-3, regardless of which model were employed. Ten nuclides were omitted from the airborne effluent release analyses. Six of these nuclides—Zn-65, Nb-94, Ag-llOm+D, Sb-125+D, Ce-144+D, and Eu-152—are strong y- emitters which deliver their doses primarily via external exposure, the RME individual being either the scrap truck driver or the taxi driver. The normalized doses to the scrap truck driver from Co-60 and Ru-106+D, two strong y-emitters which were included, are four orders of magnitude greater than the corresponding doses from the airborne emissions pathway. Since the release fractions of these nuclides are similar to or greater than those of the six omitted y- emitters, there is no reason to believe that the doses from this pathway would be significant for the latter six nuclides. The four other omitted nuclides—Fe-55, Mo-93, Pm-147, and Cm-244—deliver their doses via internal exposure, the RME individual being the scrap shredder. Since the release fractions of these nuclides are equal to or smaller than those of other nuclides which also decay by a- or P- emission or by electron capture, there is again no reason to believe that the airborne release scenario would produce significant doses from these nuclides. 8.4.4 Ingrowth of Radioactive Progenies As was noted in Section 6.3.4, there was no need to consider the ingrowth of radioactive progenies in the assessment of manufactured products made from cleared carbon steel scrap, As discussed in Section 8.2.6, the atmospheric releases of C-14 and 1-129 were assessed using the CAP-88 code and were corrected for the ICRP 68 DCFs. One other exception is Th-228+D. The dose from atmospheric releases, calculated by the SCA 1995 model, is about one-fifth the maximum dose from this nuclide in the present analysis. However, if the atmospheric assessment were corrected for the ICRP 68 DCF for inhalation, the principal pathway for this nuclide in this scenario, the dose would be approximately one-tenth the maximum dose. 8-16 ------- inasmuch as radionuclides which partitioned to the finished steel would not display significant ingrowth during the useful life of such products. This is not the case for aluminum recycling. As shown in Table 8-1, the partitioning of most contaminants is much less clear-cut than in the case of steel. Of the two manufactured product scenarios included in the aluminum analysis, only the taxi driver would receive doses which are comparable (i.e., within an order of magnitude) to the maximum dose from any radionuclide. (The doses from frying pan scenario are significantly smaller than the maximum dose from any radionuclide.) Since all of the dose in this scenario is through external exposure, only those y-emitting progenies that would have significant ingrowth during the 7.3-year useful life of this product needed to be considered. The only such nuclides that needed to be addressed were Ra-228+D and Th-228+D, which form the Th-232 decay chain. The ingrowth of this progeny, modeled by means of the Bateman equations, was explicitly incorporated into the assessments of Th-232 and Ra-228+D in this scenario. 8-17 ------- REFERENCES American Conference of Governmental Industrial Hygienists (ACGIH). 1996. "1996 TLVs and BEIs: Threshold Limit Values for Chemical Substances and Physical Agents, Biological Exposure Indices." ACGIH, Cincinnati, OH. Blumenthal, D. 1990. "Is That Newfangled Cookware Safe?" FDA Consumer 24 (8): 12. Cassidy, P. (Municipal and Industrial Solid Waste Division, Office of Solid Waste, Environmental Protection Agency). 1998. Private communication. Eckerman, K. F., A. B. Wolbarst and A. C. B. Richardson. 1988. "Limiting Values of Radionuclide Intake and Air Concentration and Dose Conversion Factors for Inhalation, Submersion, and Ingestion," Federal Guidance Report No. 11, EPA-520/1 -88-020. U.S. Environmental Protection Agency, Washington, DC. Eckerman, K. F., and J. C. Ryman. 1993. "External Exposure to Radionuclides in Air, Water, and Soil," Federal Guidance Report No. 12, EPA 402-R-93-081. U.S. Environmental Protection Agency, Washington, DC. Huddleston, G. (Wabash Alloys). 2000. Private communication. International Commission on Radiological Protection (ICRP). 1994. "Dose Coefficients for Intakes of Radionuclides by Workers," ICRP Publication 68. Annals of the ICRP, vol. 24, no. 4. Pergamon Press, Oxford. National Council on Radiation Protection and Measurements (NCRP). 1989. "Screening Techniques for Determining Compliance With Environmental Standards—Releases of Radionuclides to the Atmosphere," NCRP Commentary No. 3 (revised). NCRP, Bethesda, MD. S. Cohen & Associates (SCA). 1995. "Analysis of the Potential Recycling of Department of Energy Radioactive Scrap Metal." 4 vols. Prepared for U.S. Environmental Protection Agency, Office of Radiation and Indoor Air, Washington, DC. U.S. Environmental Protection Agency (U.S. EPA). 1994. "Estimating Radiogenic Cancer Risk," EPA 402-R-93-076. U.S. EPA, Washington, DC. U.S. Environmental Protection Agency (U.S. EPA), Office of Air Quality Planning and Standards. 1995. "Compilation of Air Pollutant Emission Factors," AP-42, 5th ed. Vol.1, "Stationary Point and Area Sources." U.S. EPA, Research Triangle Park, NC. 8-18 ------- Chapter 9 RADIOLOGICAL ASSESSMENT OF RECYCLING COPPER Detailed descriptions of recycling practices of copper are presented in Appendix C. The present chapter presents a brief summary of the copper recycling process and recapitulates those aspects of the process which are relevant to the radiological assessment. The exposure pathways are the same as those discussed in Section 5.3. Differences in exposure parameters applicable to the copper assessment are described in the following sections. 9.1 RECYCLING COPPER SCRAP—AN OVERVIEW Copper scrap can enter copper refining and processing operations in a variety of ways, depending on factors such as quality of the scrap and its alloy content. For example, some copper scrap may be refined at primary copper smelters and some at secondary smelters. Copper alloy scrap may be remelted at brass mills, ingot makers, or foundries. Scrap copper released from nuclear installations is likely to be carefully sorted, high-quality material. As such, it would most likely be introduced into the secondary refining process at the fire refining stage where it would be used to produce anodes for electrorefining or finished mill products such as sheet and tubing. Expected partitioning of elements during fire refining is summarized in Table 9-1. While additional partitioning occurs during electrorefining, the result of that process is to further reduce the impurities in the metal. Therefore, it is unlikely that worker exposures to cathode copper will exceed those of workers involved in handling the fire- refined product. The possible exceptions are handling of anode slimes and electrolyte bleed streams from the electrolysis cells. Figure 9-1 presents a simplified diagram depicting the mass flow during the fire refining of copper scrap. As depicted in the figure, the copper scrap from various sources is sent to a reverberatory furnace to be smelted. The furnace produces copper as well as the smelting byproducts—slag and dust—together with airborne effluents (not shown). The mass balance is not exact due to rounoff of the values and to losses during the refining process. Figure 9-2 presents a simplified flow diagram of the electrorefining of scrap copper; the reverberatory furnace depicted in the previous figure constitutes one stage of this process. 9-1 ------- 45,500 tons scrap charcoal and slag formers 200-ton reverberatory furnace 110 tons dust (75% Cu) .. 910 tons slag (40% Cu) - 45,000 tons copper air green logs Figure 9-1. Simplified Mass Flow: Annual Throughput of Secondary Copper Smelter (values are rounded) 9.2 DISTRIBUTION OF CONTAMINANTS 9.2.1 Material Balance Fire Refining Based on the discussion in Section C.5.1.6, the following mass fractions were adopted for the analysis of fire refining of copper scrap: Furnace Charge: • Copper scrap 0.75 • Heel from previous melt . . . 0.25 Output: • Copper 0.99 (25% left in furnace) • Baghouse dust1 0.0018 (75% Cu) • Slag2 0.015 (40% Cu) This value is lower than that cited in Section C. 5.1.3. The latter value refers to the scrap charged to the furnace, while the present value refers to the total metal in the furnace, which includes the heel from the previous melt. This value is lower than that cited in Section C.5.1.2 for reasons cited in Note 1. 9-2 ------- 138,000 tons anode scrap 102,000 tons blister copper Anode Scrap Meltina (shaft fee.) Fire Refining (reverb. fee.) 24,000 tons No. 2 scrap acid Electro- Refining Electrolyte Clean Up bleed 1300 tons nickel sulfate 450,000 tons copper -»• 3200 tons anode slimes slag 190,000 tons purchased anodes Figure 9-2. Simplified Material Balance for Electrorefining of Copper Produced from Scrap The sum of the output fractions is greater than unity due to the addition of charcoal, slag formers and other reductants, as well as oxidation of the copper in the slag. Electrorefining As shown in Figure 9-2, an integrated refinery with an annual production of 450,000 tons (408,000 t) of electrorefmed copper generates 3,200 tons (2,9001) of anode slimes. A mass fraction of 0.0071 (3,200 H- 450,000 = 0.0071) was therefore adopted for the analysis of this byproduct. 9.2.2 Elemental Partitioning Table 9-1 lists the partition ratios and release fractions of the various trace elements during fire refining and electrorefming, taken from Tables C-18 and C-20, respectively. The concentration factors are calculated according to the methodology presented in Section 6.2, using the mass fractions listed in Section 9.2.1. 9.3 LIST OF OPERATIONS AND EXPOSURE SCENARIOS Section C.5 presents a list of possible exposure scenarios for the radiological assessment of copper recycling. The basis of selecting scenarios for the present analysis was to ensure that the reasonable maximum exposure from each radionuclide would be evaluated. Scenarios were 9-3 ------- omitted if the radiation sources, exposure pathways, and exposure durations were such that the radiation exposures would be bounded by the scenarios already selected. Table 9-2 lists the operations and exposure parameters employed in the present assessment, which are discussed in the following sections. The descriptive title of each individual exposure scenario—used to assess the exposure of a given individual—is italicized. (Sub-scenarios listed beneath the individual scenario list different activities performed by the same individual.) Further details are found in Appendix C. 9.3.1 Dilution Scrap Transport It is assumed that copper scrap from a nuclear facility would be transported to a smelter in a dedicated truck. Therefore, the dilution factor for this operation is equal to 1. Smelter Operations Like aluminum, movement of scrap copper is not geographically constrained by haulage costs. There were six secondary copper smelters in the United States in 1995 (U.S. EPA 1995); such smelting is one of several avenues for the recycling of copper scrap. The present assessment assumes that copper scrap released during the decomissioning of three gaseous diffusion plants is fire refined in a 200-ton reverberatory furnace, which has an annual capacity of about 45,500 tons (41,300 t). As discussed in Section C. 1.1, the largest amount of copper scrap generated by the decommissioning of a single facility in any one year is the 2,080 t of copper from the K-25 plant in Oak Ridge, anticipated each year from 2003 to 20053. Assuming all of this scrap would be processed at the reference smelter, the dilution factor would be 0.050. Electrorefining In an alternative scenario, the entire output of the reverberatory furnace is electrolytically refined. In this process, the 2,0801 of residually radioactive copper scrap is blended into a total of 450,000 tons (408,000 t) of metal from various sources, resulting in a dilution factor of 5.1 x 1Q-3. DOE is not expected to begin releasing scrap metal before 2003; consequently, it was assumed that newly generated scrap would be sent to the reference facility, while any previously stockpiled copper would have a different disposition 9-4 ------- Table 9-1. Partition Ratios (PR) and Concentration Factors (CF) for Fire Refining and Electrorefming of Copper Scrap Element Ac, Ce, Eu, Nb, Pa, Pm, Ra, Th, U Ag Am, Cm, Np, Pu C,I Co, Ni, Tc Cs Fe, Mn Mo, Sr Pb Ru Sb Zn Furnace Charge CFa 0.76 0.90 0.75 0.75 0.78 0.75 0.76 0.75 0.81 1.00 0.81 0.80 Metal PR (%) 0.1 -2 30-59 0.1 -1 0 5-10 0 2-5 0 22 100 8-25 10-20 CF 0.015 0.53 7.6e-03 0 0.078 0 0.039 0 0.18 1.01 0.21 0.16 Slag PR (%) 98 - 99.9 41 - 70 99 - 99.9 0 90- 95 10- 20 95- 98 100 73 - 78 0 75- 92 80- 90 CF 50 42 50 0 49 10 50 50 42 0 50 48 Dust PR (%) 0 0 0 0 0 80- 90 0 0 0- 5 0 0- 5 0- 5 CF 0 0 0 0 0 372 0 0 22 0 22 22 Release Fraction13 0 0 0 1 0 0.9 0 0 0.054 0 0.054 0.053 Anode Slimes PRC (%) 0 96 0 0 0 0 36 50 99.7 65- 70 0 0 CF 0 72 0 0 0 0 2.0 0 25 99 0 0 Refers to the copper charged to the furnace, which consists of 75% scrap and 25% heel from the previous melt. Fraction of furnace charge released to atmosphere, assuming no effluent emission controls With respect to metal ------- Table 9-2 Exposure Scenarios and Parameters for Radiological Assessments of Copper Recycling Description SCRAP TRANSPORT. Truck driver SECONDARY SMELTER Scrap handler Furnace Operations Slag worker: Slag handling & metal recovery General duties Airborne effluent emissions Electrorefining Tank house operator END USERS Cooking in copper pan Dilution factor 1. 0.05 0.0051 0.5 Exposure Pathways External Exposure Time (hr/y) 1000 Distance 8ft Medium scrap Internal Time (hr/y) Medium Dust load (mg/m3) N/A RFa 1000 b scrap 1750 scrap 2.3 0.6 500 b slag NA 500 1250 slag dust 10 2.3 N/A 0.5 0.6 1750 3ft slimes N/A 263 2ft N/AC metal N/A Respirable fraction Exposure assessment uses FOR 12 dose coefficients—see discussion in Section 6.3.1 Exposure from ingestion of contaminated food End Users As was the case with carbon steel and aluminum scrap, it is highly unlikely that all of the copper scrap in a single furnace heat would be from a nuclear facility. In addition to dilution caused by charging the furnaces with different batches of scrap, 25% of the charge consists of molten copper from the previous heat. Thus, a dilution factor of 0.5—the value used for finished products made of steel or aluminum—is reasonably conservative for the copper assessment. 9.3.2 Scrap Transport The scrap transport worker is a truck driver who spends eight hours per day in the cab of a truck, carrying 20-t loads of scrap metal to the scrap processor and returning with an empty truck (or carrying other cargo). His only exposure would be to external radiation from the load of contaminated scrap. 9-6 ------- The MicroShield exposure calculations used for the scrap steel transport scenario, with the dilution factor cited in Section 9.3.1, serve as a reasonable estimate for the present analysis. Although solid copper is denser than steel (8.96 vs. 7.86 g/cm3), the assumed bulk density of the scrap steel, 1.57 g/cm3, is already conservative. (Scrap shipped for recycling, as opposed to burial, is not highly compressed.) The higher atomic number of copper would result in slightly higher self-shielding, leading to a slightly more conservative assessment. 9.3.3 Secondary Smelter Operations For the purposes of the present analysis, copper recycling operations at a secondary smelter were divided into two categories: scrap handling and furnace operations. Two workers—a scrap handler and a worker recovering bulk copper from slag—were included in the exposure assessments. Both these workers would be exposed to external radiation from the residual radioactivity of the materials. They would also inhale potentially contaminated dust in the ambient air and ingest deposited particulate matter. Additional exposure assessments were performed on a nearby resident who is exposed to the unfiltered airborne effluents from the smelter. These analyses are discussed in Section 9.3.6. Scrap Handler The scrap handler prepares scrap for charging into the furnace. His external exposure is to the 200 tons of scrap which are stockpiled in the area. The scrap pile may be envisioned as a large rectangular block, with the worker positioned at one of the vertices. Such a source constitutes one-fourth of a semi-infinite slab. The external exposure rates were estimated by taking one- fourth of the dose coefficients for soil contaminated to an infinite depth, as listed in Federal Guidance Report No. 12 (FGR 12) (Eckerman and Ryman 1993). Because the scrap handler will move around the area in the course of his work, it is assumed that he is in the immediate vicinity of the pile for only four hours per day. The dust loading and respirable fraction are taken from the results of measurement at the smelting furnace listed in Table C-23. 9-7 ------- Slag Worker After the slag is skimmed from the melt, it is transported to another area of the smelter. There, a worker breaks up the slag with a pneumatic hammer and culls copper by hand from the slag. This operation is estimated to take about two hours per day. If the three tons of slag that are processed each day were spread on the floor in a 5-cm-thick layer4—assuming the crushed slag has a bulk density similar to soil, 1.6 g/cm3—it would occupy an area of 34 m2. The external exposure from this source was calculated by applying dose coefficients for soil contaminated to a depth of 5 cm which are listed in FGR 12. A correction was made for the finite area of the source by applying the area factors listed in the RESRAD manual (Yu et al. 1993, Table A.2). Plotting these values on log-log paper yields a straight line (except the first of the listed values, which appears to be a misprint); an area of 34 m2 corresponds to a correction factor of 0.43. This factor was divided by 2 to correct for the worker's moving over the area rather than staying in the center. The breaking and sorting of the slag in an indoor location generates considerable dust. The value selected for this portion of the slag worker's exposure is based on ACGIH time-weighted average limits for a 40-hour work week (ACGIH 1996) and is consistent with OSHA PELs (see discussion of shredder operator in Section 8.2.3). Since the exposure time is only a fraction of a 40-hour week, the dust concentration could be higher without violating these prescribed limits. However, it is unlikely such a high level would persist for any extended period. When not handling slag, the worker is assigned other duties in the mill which do not expose him to direct radiation from residually radioactive materials. However, he does inhale the dust in the ambient air, which consists of fugitive emissions from the smelter, and ingests deposited particulate matter from the same source. 9.3.4 Electrorefming After fire refining in a reverberatory furnace, the metal may be further processed by electrorefining. The duties of the tank house operator in the electrorefining facility include collecting anode slimes and packing them into 55-gallon (208 L) drums. Since the electrolytic tanks are encased in concrete that would shield the tank's contents, the primary source of Such a thin layer would be needed to allow visual identification for copper nuggets included in the slag. 9-8 ------- exposure would be the material in the drums. The assessment assumes that the operator's average source of external exposure is one full drum. This source was modeled as a right circular cylinder, two feet in diameter and 35 inches high, encased in a 0.05-inch-thick iron shield. Given the wet contents of the drums and the fact that no volatile radioactive material would be present, no significant inhalation exposure is anticipated. Because of the corrosive nature of the material, any significant inadvertent ingestion is also unlikely. Only silver, ruthenium and lead exhibit significant concentrations in the anode slimes, as shown in Table 9-1. Of the three isotopes of these elements included in the present analysis, Ru-106+D and Ag-110m are strong y-emitters—the external exposure in this scenario would have a significant effect on the radiological assessment of the RME individual. Pb-210+D would have a lower concentration in the slimes; furthermore, it is a weak photon emitters—external exposure, the only environmental pathway in this scenario, would not be a significant pathway for this nuclide and its progeny. Thus, only Ru-106+D and Ag-110m were addressed by the assessment of the tank house operator. 9.3.5 Use of Finished Products A consumer cooking food in a copper pan would be exposed to external radiation from the metal in addition to eating food which may be contaminated with residual radioactivity that has leached from the pan along with the copper matrix. The copper content of food cooked in copper cookware was inferred directly from data presented by Reilly (1985), who listed the concentrations of copper in beef and cabbage cooked in cast iron and copper utensils. Since iron utensils contain little or no copper, the average difference in the copper content of the foods cooked in the two types of vessels enables an estimate of the amount of leached copper, which is equal to 4 mg/kg of beef and 1.3 mg/kg of cabbage. The annual intake of 390 mg of copper was derived by using cabbage as a surrogate for all vegetables and using the annual consumption of beef and vegetables listed in Section 6.4.2. It is assumed that the ingested metal has the same concentration of radionuclides as the metal in the pan. The external exposures calculated for the cast iron pan serve as an estimate of the exposures in the present scenario. Since, as discussed in Section 8.2.5, the external exposure to this small 9-9 ------- object for a relatively few hours per year is a small contributor to the dose, the minor differences between an iron and a copper pan are not significant to the present analysis. 9.3.6 Impact of Airborne Effluent Emissions on Nearby Residents The assessment of airborne effluent emissions from the furnace on nearby residents was performed in a manner similar to that described in Section 8.2.6. However, unlike the case with aluminum smelters, there are no current or pending NESHAPS standards for secondary copper smelters. Consequently, unfiltered particulate emissions may be vented directly to the atmosphere, resulting in high release fractions of some trace elements, as shown in Table 9-1. Six of the elements listed in Table 9-1 have significant atmospheric release fractions. Of the seven radioisotopes of these elements included in the present analysis, two—C-14 and 1-129—were addressed by the CAP-88 analysis performed for the carbon steel recycling assessment. The results of that assessment are directly transferable to the copper analysis, as was done for the aluminum studies described in Section 8.2.6. Because the large release fractions could lead to significant doses, the remaining five nuclides were explicitly modeled, using CAP88-PC. Of the seven locations described in Section 6.4.3, Moline, 111. was found to produce the median doses from these five nuclides, which would be released as particulates. The doses from the ingestion and inhalation pathways were adjusted for the DCFs in ICRP Publication 68 (ICRP 1994), as described in Section 6.4.3. The risks from each radionuclide were calculated separately for each pathway (see Section 6.4.3) and then summed. The results, adjusted for the releases from the copper smelter, are presented in Table 9-3. Column 2 of Table 9-3 lists the release rate of each radionuclide that was used as input to the CAP88-PC code. Columns 3 and 4 list the dose and risk corresponding to this release. In the case of C-14 and 1-129, the dose was taken directly from the output of CAP88-PC, as listed in Table 6-14. For the other nuclides, the dose contribution from each pathway was corrected for ICRP 68 DCF. Column 5 lists the airborne release fraction for each radionuclide, as presented in Table 9-1. Column 6 lists the activity released from the smelter during the refining of 2,080 t of copper scrap cleared from the K-25 facility (see page 9-4), normalized to a specific activity of 1 pCi/g of each nuclide in the cleared scrap. Columns 7 and 8 list the dose and risk corresponding to the normalized releases from the smelter. The doses from C-14 and 1-129 are corrected for the ICRP 68 DCF—that correction was applied in Column 3 to the doses from the other five nuclides. 9-10 ------- Table 9-3. Normalized Impacts from One Year of Exposure to Fugitive Airborne Emissions Nuclide C-14 Zn-65 1-129 Sb-125+D Cs-134 Cs-137+D Pb-210+D CAP-88 Analysis mCi/y 11 1,000 15 1,000 1,000 1,000 1,000 Dose3 4.86e-04 1.12e+00 4.48e-01 8.11e-01 3.57e+00 7.54e+00 4.62e+01 Risk 2.40e-10 7.90e-07 2.99e-07 6.05e-07 2.58e-06 5.60e-06 1.28e-05 Secondary Copper Smelter R. F.b 1.0 0.053 1.0 0.054 0.9 0.9 0.054 mCi/y 2.08 0.11 2.08 0.11 1.87 1.87 0.11 Dosec 9.45e-05 1.23e-04 9.16e-02 9.11e-05 6.68e-03 1.41e-02 5.19e-03 Riskd 4.546-11 8.716-11 4.15e-08 6.806-11 4.83e-09 1. 05e-08 1.44e-09 >uu i H-.uzcTui | i.zoe- C-14 and 1-129 doses as reported for carbon steel analysis ii b From Table 9-1 corrected for ICRP 68 DCFs corrected for ICRP 68 DCFs I'luiii i auic 7-1 mrem EDE per pCi/g in scrap, i/un^u^u n Lifetime risk of cancer per pCi/g in scrap 9.4 RESULTS The results of the copper recycling analysis are shown in Table 9-4. Three key observations can be made about these data. • Normalized doses from the recycling of copper are generally higher than from the recycling of carbon steel. • The slag worker is the RME individual for all nuclides except C-14, Ru-106+D, and 1-129. • Three scenarios account for the reasonably maximum doses from all 44 nuclides and nuclide combinations. 9.4.1 Slag Worker The worker who recovers copper from the slag is the RME individual for all but three radionuclides. All the strong y-emitters except Ru-106+D partition strongly to the slag. The principal pathway for these nuclides is external exposure. In the cases of &- and p-emitters that are not strong y-emitters, the principal pathway is usually inhalation, except for a few cases, such as Pb-210+D, in which ingestion dominates. 9-11 ------- Table 9-4. Maximum Exposure Scenarios and Normalized Impacts on the RME Individual from One Year of Exposure Nuclide C-14 Mn-54 Fe-55 Co-60 Ni-59 Ni-63 Zn-65 Sr-90+D Nb-94 Mo-93 Tc-99 Ru-106+D Ag-llOm+D Sb-125+D 1-129 Cs-134 Cs-137+D Ce-144+D Pm-147 Eu-152 Pb-210+D Ra-226+D Ra-228+D Ac-227+D Th-228+D Th-229+D Th-230 Th-232 Pa-231 Maximum Scenario Airborne effluent releases Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Tank house operator Slag worker Slag worker Airborne effluent releases Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Dose mrem EDE per pCi/g 9.45e-05 8.80e-02 4.14e-05 2.56e-01 9.45e-06 2.61e-05 5.93e-02 4.30e-03 1.69e-01 3.26e-04 1.84e-04 9.38e-03 2.44e-01 4.52e-02 9.16e-02 6.82e-02 3.61e-02 7.71e-03 1.57e-04 1.21e-01 3.05e-01 3.03e-01 2.40e-01 2.50e+00 1.36e+00 2.30e+00 3.78e-01 6.64e-01 9.82e-01 |lSv per Bq/g 2.55e-02 2.38e+01 1.12e-02 6.92e+01 2.55e-03 7.05e-03 1.60e+01 1.16e+00 4.57e+01 8.81e-02 4.97e-02 2.54e+00 6.59e+01 1.22e+01 2.48e+01 1.84e+01 9.76e+00 2.08e+00 4.24e-02 3.27e+01 8.24e+01 8.19e+01 6.49e+01 6.76e+02 3.68e+02 6.22e+02 1.02e+02 1.79e+02 2.65e+02 Lifetime Risk of Cancera per pCi/g 4.54e-ll 6.69e-08 1.06e-ll 1.95e-07 6.41e-12 2.03e-ll 4.51e-08 1.91e-09 1.29e-07 1.40e-ll 5.93e-ll 7.13e-09 1.86e-07 3.43e-08 4.15e-08 4.90e-08 2.51e-08 6.02e-09 9.36e-ll 9.16e-08 8.39e-08 1.69e-07 9.14e-08 1.24e-07 8.69e-07 4.72e-07 4.35e-08 8.16e-08 5.26e-08 per Bq/g 1.23e-09 1.81e-06 2.86e-10 5.27e-06 1.73e-10 5.49e-10 1.22e-06 5.16e-08 3.49e-06 3.78e-10 1.60e-09 1.93e-07 5.03e-06 9.27e-07 1.12e-06 1.32e-06 6.78e-07 1.63e-07 2.53e-09 2.48e-06 2.27e-06 4.57e-06 2.47e-06 3.35e-06 2.35e-05 1.28e-05 1.18e-06 2.21e-06 1.42e-06 Maximum risk—may correspond to a different scenario 9-12 ------- Table 9-4 (continued) Nuclide U-234 U-235+D U-238+D Np-237+D Pu-238 Pu-239 Pu-240 Pu-241+D Pu-242 Am-241 Cm-244 U-Natural U-Separated U-Depleted Th-Series Maximum Scenario Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Slag worker Dose mrem EDE per pCi/g 2.43e-01 2.37e-01 2.12e-01 6.29e-01 4.26e-01 4.26e-01 4.26e-01 4.55e-03 3.98e-01 1.13e+00 7.20e-01 1.62e+00 4.66e-01 2.38e-01 2.27e+00 |lSv per Bq/g 6.57e+01 6.41e+01 5.73e+01 1.70e+02 1. 15e+02 1. 15e+02 1. 15e+02 1.23e+00 1. 08e+02 3.05e+02 1.95e+02 4.38e+02 1.26e+02 6.43e+01 6.14e+02 Lifetime Risk of Cancera per pCi/g 1.08e-07 1.13e-07 9.84e-08 2.89e-07 7.37e-08 6.83e-08 6.83e-08 4.09e-10 6.46e-08 3.04e-07 1.92e-07 5.16e-07 2.11e-07 1.10e-07 1.04e-06 per Bq/g 2.92e-06 3.05e-06 2.66e-06 7.81e-06 1.99e-06 1.85e-06 1.85e-06 l.lle-08 1.75e-06 8.22e-06 5.19e-06 1.39e-05 5.70e-06 2.97e-06 2.81e-05 Maximum risk—may correspond to a different scenario As can be inferred from Table 9-1,37 of the 40 individual nuclides addressed in the present analysis exhibit significant partitioning to the slag. Because the mass fraction of the slag is only 1.5%, the concentration factors of these nuclides range from 10 to 50. Because the slag worker is in close contact with the slag for about two hours per day, during which time he is exposed both to the direct radiation from the slag spread on the floor and to a high concentration of dust in the ambient air, his doses are higher than those of comparable workers at a steel mill. This is because the latter work in an outdoor location with a lower dust loading; furthermore, the specific activities of radionuclides in copper slag, normalized to unit specific activity in scrap, are at least six times greater than those of the same nuclides in steel slag. 9-13 ------- 9.4.2 Airborne Effluent Emissions The nearby resident exposed to unfiltered airborne effluent emissions is the RME individual in the case of two radionuclides: C-14 and 1-129. Carbon and iodine both vaporize in the smelter and are completely released to the atmosphere. 9.4.3 Tank House Operator The tank house operator is the RME individual in the case of Ru-106+D. Ruthenium is the only element which partitions entirely to the metal during smelting and then accumulates in the anode slimes during the subsequent electrorefining. The principal pathway for this strong y-emitter is external exposure, which is the only pathway in this scenario. 9.5 EVALUATION OF THE RESULTS Many of the observations in Section 7.3 regarding the radiological assessment of scrap steel are applicable to the copper studies. The issues that are unique to copper are discussed in the following paragraphs. 9.5.1 Airborne Effluent Releases The airborne effluent release scenario addressed the seven radionuclides which would be released to the atmosphere during the smelting of copper scrap. These were assessed using the EPA's CAP88-PC code, which uses generally conservative assumptions. The actual normalized doses and risks would most likely be smaller than those calculated in the present analysis. 9.5.2 Other Scenarios Several scenarios described in Section C.5.2 were omitted from the detailed exposure assessments. These scenarios were subjected to scoping analyses which demonstrated that they would not lead to the maximum exposures from any of the nuclides in the present analysis. Baghouse Dust Agglomeration Operator The baghouse dust agglomeration operator would be exposed to a large mass of wetted dust. The only element that primarily accumulates in the dust is cesium; consequently, the nuclides of concern are Cs-134 and Cs-137. Since this scenario takes place at a large integrated facility with an annual production of 51,100 tons (46,400 t) of dust, the cesium in the 2,080 t of contaminated 9-14 ------- scrap, 90% of which is assumed to accumulate in the dust, would have a dilution of 0.04 (2080 x 0.9 + 46400 ~ 0.04). Thus, the cesium is more dilute than in the scrap handler scenario, where the dilution factor is 0.051. Furthermore, the dust is contained in a concrete bunker, which would provide substantial shielding of the direct radiation. Consequently, the worker in this scenario would receive a smaller dose from the cesium isotopes than would the scrap handler. Furnace Operator The primary external exposure of the furnace operator would be during the two hours per day that he spends raking the dross from the melt, which has a mass of 200 tons. During this time, he reaches into the furnace through a 3 x 2 ft door. Only the portion of his body that is exposed to this opening would be irradiated, resulting in a reduced dose. Furthermore, most of the source material would not be in the line-of-sight of this opening. In contrast, the scrap handler would be exposed to the same activity in 200 tons of scrap, without the benefit of shielding, while the slag worker would be exposed to that portion of the activity that partitions to the slag. The internal exposure of the furnace operator would be comparable to that of the other two workers. Casting Machine Operator The casting machine operator would receive external exposure from cast copper logs which weigh up to 5 t. His exposure would not be greater than that of the scrap handler, who is exposed to a greater mass of metal, especially since all contaminants except ruthenium are wholly or partly removed from the cast metal. Exposure to Electrolyte Bleed Streams As shown in Table C-21, a number of elements would be expected to partition to the electrolyte bleed during the electrorefining of copper. However, with the exception of ruthenium, all of these elements would have been removed from the copper during the prior fire refining step, so that their final concentrations in the bleed stream would be lower than their concentrations in slag. The concentration of ruthenium in the bleed stream would be similar to its concentration in the anode slimes. Consequently, the dose from Ru-106+D to a worker who came in contact with this material would be similar to the dose from this radionuclide to the tank house operator. Furthermore, it should be noted that Ru-106—a fission product with a one-year half-life—would not play a significant role in the radiological assessment of the recycling of copper scrap, for the following reasons. First, there would be little or no Ru-106 in the scrap cleared during the decomissioning gaseous diffusion plants, the major source of copper from nuclear facilities. In 9-15 ------- the case of nuclear power plants, this relatively short-lived nuclide would have largely decayed away prior to the commencement of decommissioning activities. External Exposure to Finished Products Unlike steel and, to a lesser extent, aluminum, there are no commonplace scenarios in which an individual other than one engaged in recycling is exposed to a large mass of copper or a copper alloy. Section C.5.3 lists the copper content of motor vehicles and a few common home appliances. None of these contain massive amounts of copper; furthermore, their copper content would be derived from more than one source. Consequently, the probability that a worker or a consumer would be exposed to direct radiation from a significant mass of copper from a single melt, which could contain a large fraction of residual radioactive metal, is remote. In addition, the only radionuclide that strongly partitions to the metal in the fire refining process is Ru-106, which, as noted above, does not play a significant role in the radiological assessment of copper recycling. Uses of Copper Slag As noted in Section C.5.1.2, slag generated by the fire refining of copper is used in the manufacture of abrasives, shingles, road surface bedding, mineral wool, and cement/concrete materials. An individual residing in a home with a roof covered with shingles made from copper slag could be exposed to direct radiation from radionuclides that partition to the slag. It is highly improbable that all the shingles would be made from slag generated during a single melt. Consequently, the dilution factor would not be substantially greater than that assumed for the secondary smelter. The shingles are substantially thinner than the 5-cm-thick layer of slag on the floor of the smelter in the slag worker scenario; furthermore, the slag is used as the backing of the shingle, which contains other ingredients. Finally, the distance between the resident of the home and the roof, as well as the shielding by intervening construction materials, would further reduce the radiation exposure. Consequently, it is unlikely that such a scenario would lead to higher doses than those to the slag worker. 9-16 ------- REFERENCES American Conference of Governmental Industrial Hygienists (ACGIH). 1996. "1996 TLVs and BEIs: Threshold Limit Values for Chemical Substances and Physical Agents, Biological Exposure Indices." ACGIH, Cincinnati, OH. Eckerman, K. F., and J. C. Ryman. 1993. "External Exposure to Radionuclides in Air, Water, and Soil," Federal Guidance Report No. 12, EPA 402-R-93-081. U.S. Environmental Protection Agency, Washington, DC. U.S. Environmental Protection Agency (U.S. EPA), Office of Air Quality Planning and Standards. 1995. "Compilation of Air Pollutant Emission Factors, AP-42, 5th ed. Vol.1, "Stationary Point and Area Sources." U.S. Environmental Protection Agency, Research Triangle Park, NC. Reilly, C. 1985. "The Dietary Significance of Adventitious Iron, Zinc, Copper, and Lead in Domestically Prepared Food." Food Additives and Contaminants 2:209-215. Yu, C., et al. 1993. "Manual for Implementing Residual Radioactive Material Guidelines Using RESRAD," ANL/EAD/LD-2. Argonne National Laboratory, Argonne, IL. 9-17 ------- APPENDIX A SCRAP METAL INVENTORIES AT U.S. NUCLEAR POWER PLANTS ------- Contents page A.I Introduction A-l A.2 Characteristics of Reference Reactor Facilities A-3 A.2.1 Reference PWR Design and Building Structures A-4 A.2.1.1 Reactor Building A-6 A.2.1.2 Fuel Building A-6 A.2.1.3 Auxiliary Building A-7 A.2.1.4 Control and Turbine Buildings A-7 A.2.2 Reference BWR Design and Building Structures A-7 A.2.2.1 Reactor Building A-8 A.2.2.2 Turbine Building A-9 A.2.2.3 Radwaste and Control Building A-9 A.3 Residual Activities in Reference Reactor Facilities A-9 A.3.1 Neutron-Activated Reactor Components and Structural Materials A-10 A.3.1.1 Reference BWR A-ll A.3.1.2 Reference PWR A-ll A. 3.2 Internal Surface Contamination of Equipment and Piping A-14 A.3.2.1 Measurements of Internal Surface Contamination at Six Nuclear Power Plants A-14 A.3.2.2 Internal Surface Contamination Levels Reported in Decommissioning Plans\-17 A.3.2.3 Levels of Internal Surface Contamination Derived for Reference BWR . . A-20 A.3.2.4 Levels of Internal Surface Contamination for Reference PWR A-23 A.3.3 Contamination of External Surfaces of Equipment and Structural Components A-28 A.3.3.1 Data for Reference Facilities A-32 A.3.3.2 Surface Contamination Levels Reported by Facilities Preparing for Decommissioning A-35 A.4 Baseline Metal Inventories A-37 A.4.1 Reference PWR A-37 A.4.2 Reference BWR A-38 A.5 Metal Inventories with the Potential for Clearance A-43 A.5.1 Contaminated Steel Components with the Potential for Clearance A-47 A.5.1.1 Reference BWR A-47 A.5.1.2 Reference PWR A-68 A.5.1.3 Summary of Steel Inventories of the Reference Reactors A-76 A. 5.2 Applicability of Reference Reactor Data to the Nuclear Industry A-78 A.5.2.1 Scaling Factors A-78 A.5.2.2 U.S. Nuclear Power Industry A-79 A-iii ------- Contents (continued) page A.5.2.3 Estimating the Metal Inventories of U.S. Nuclear Power Plants A-80 A.5.3 Metal Inventories Other Than Steel A-82 A.5.4 Timetable for the Release of Scrap Metals from Nuclear Power Plants A-83 References A-85 Appendix A-l: U.S. Commercial Nuclear Power Reactors Al-1 Reference Al-6 A-iv ------- Tables page A-l. Sources of Residual Activities in Reference BWR and PWR A-l 1 A-2. Estimated Activities of Neutron-Activated Reactor Components in a BWR A-12 A-3. Neutron-Activated Reactor Components in a PWR A-13 A-4. Activation Levels at Trojan Nuclear Plant One Year after Shutdown A-13 A-5. Residual Activities and Operating Parameters of Six Nuclear Power Plants* A-15 A-6. Relative Activities of Long-Lived Radionuclides at Six Nuclear Power Plants* A-16 A-7. Distribution of Activities in Major Systems of Three PWRs (%) A-17 A-8. Internal Contamination Levels of Big Point Nuclear Plant at Shutdown A-18 A-9. Plant Systems Radioactivity Levels at SONGS 1 A-19 A-10. Average Internal Contamination Levels of Reactor Systems at Yankee Rowe A-20 A-l 1. Activated Corrosion Products in the Reference BWR A-21 A-12. Distribution of Activated Corrosion Products on Internal Surfaces of Reference BWR A-22 A-13. Contact Dose Rate and Internal Surface Activity of BWR Piping A-23 A-14. Estimates of Internal Contamination for Reference BWR Piping A-24 A-l 5. Summary of Contamination Levels in BWR Equipment A-25 A-16. Estimated Internal Surface Activities in BWR Systems A-25 A-17. Internal Surface Contamination in the Reference PWR Primary System A-28 A-18. Activated Corrosion Products on the Interiors of PWR Systems A-28 A-19. Non-RCS Contaminated PWR Piping A-29 A-20. Radionuclides in Primary Coolant in the Reference PWR A-30 A-21. Radionuclide Concentrations in Reactor Coolant of Reference BWR A-31 A-22. Surface Contamination Levels for Reference BWR at Shutdown A-32 A-23. Estimated External Structural Contamination in the Reference BWR A-33 A-24. External Surface Activity Concentrations at Six Nuclear Generating Stations A-35 A-25. Radionuclide Inventories on External Surfaces at Trojan Nuclear Plant A-36 A-26. Contamination of Floor Surfaces at Trojan Nuclear Plant Prior to Decommissioning A-36 A-27. Radiation Survey Data for Humboldt Bay Refueling Building A-39 A-28. Radiation Survey Data for Humboldt Bay Power Building A-40 A-29. Inventory of Materials in a 1971-Vintage 1,000 MWe PWR Facility A-41 A-30. Breakdown of Materials Used in PWR Plant Structures and Reactor Systems A-42 A-31. Inventories of Ferrous Metals Used to Construct a 1,000-MWe BWR Facility .... A-43 A-32. Containment Instrument Air System A-48 A-33. Control Rod Drive System A-48 A-34. Equipment Drain Processing System A-49 A-35. Fuel Pool Cooling and Cleanup System A-50 A-36. High Pressure Core Spray System A-50 A-37. HVAC Components System A-51 A-38. Low Pressure Core Spray System A-51 A-39. Main Steam System A-52 A-v ------- Tables (continued) page A-40. Main Steam Leakage Control System A-53 A-41. Miscellaneous Items from Partial System A-54 A-42. Reactor Building, Closed Cooling Water System A-55 A-43. Reactor Building Equipment and Floor Drains System A-55 A-44. Reactor Core Isolation Cooling System A-56 A-45. Reactor Coolant Cleanup System A-56 A-46. Residual Heat Removal System A-57 A-47. Miscellaneous Drains System A-58 A-48. Chemical Waste Processing System A-59 A-49. Condensate Demineralizers System A-60 A-50. HVAC Components System A-60 A-51. Radioactive Floor Drain Processing System A-61 A-52. Rad Waste Building Drains System A-61 A-53. Standby Gas Treatment System A-62 A-54. Feed and Condensate System A-62 A-55. Extraction Steam System A-63 A-56. Heater Vents and Drains System A-63 A-57. HVAC Components System A-64 A-58. Offgas (Augmented) System A-64 A-59. Recirculation System A-65 A-60. Turbine Building Drains System A-65 A-61. Reactor Building A-66 A-62. Primary Containment A-66 A-63. Turbine Building A-67 A-64. Radwaste and Control Buildings A-67 A-65. External Surface Structures Equipment System A-68 A-66. Internally Contaminated Primary System Components System A-69 A-67. Component Cooling Water System A-69 A-68. Containment Spray System A-70 A-69. Clean Radioactive Waste Treatment System A-70 A-70. Control Rod Drive System A-71 A-71. Electrical Components and Annunciators System A-71 A-72. Chemical and Volume Control System A-72 A-73. Dirty Radioactive Waste Treatment System A-73 A-74. Radioactive Gaseous Waste System A-73 A-75. Residual Heat Removal System A-74 A-76. Safety Injection System A-74 A-77. Spent Fuel System A-75 A-78. Structural Steel Components A-75 A-79. Reference PWRNon-RCS Stainless Steel Piping A-76 A-80. Summary of Reference PWR and BWR Steel Inventories A-77 A-vi ------- Tables (continued) page A-81. Steel Inventories of U.S. Nuclear Power Facilities A-80 A-82. Average Mass Thickness of Carbon Steel Inventories A-82 A-83. Inventories of Metals Other Than Steel A-82 A-84. Anticipated Releases of Scrap Metals from Nuclear Power Plants A-84 Al-1. Nuclear Power Reactors Currently Licensed to Operate Al-2 Al-2. Formerly Licensed Nuclear Power Reactors Al-6 Figures A-l. Pressurized Water Reactor A-5 A-2. Boiling Water Reactor A-8 A-3. Reactor Coolant System in a Four-Loop PWR A-27 A-vii ------- SCRAP METAL INVENTORIES AT U.S. NUCLEAR POWER PLANTS A.I INTRODUCTION At the end of 1999 the U.S. commercial nuclear power industry was represented by 104 operating reactors and 27 nuclear power reactors formerly licensed to operate (U.S. NRC 2000). In the next three decades, most of the operating licenses of reactors currently in operation—originally valid for 40 years—will have expired.1 With the publication of the NRC's Decommissioning Rule in June 1988 (U.S. NRC 1988), owners and/or operators of licensed nuclear power plants are required to prepare and submit plans and cost estimates for decommissioning their facilities to the NRC for review. Decommissioning, as defined in the rule, means to remove nuclear facilities safely from service and to reduce radioactive contamination to a level that permits release of the property for unrestricted use and termination of the license. The decommissioning rule applies to the site, buildings, and contents and equipment. Currently, several utilities have submitted a decommissioning plan to the NRC for review. Historically, the NRC has defined three classifications for decommissioning of nuclear facilities: • DECON is defined by the NRC as "the alternative in which the equipment, structures, and portions of a facility and site containing radioactive contaminants are removed or decontaminated to a level that permits the property to be released for unrestricted use shortly after cessation of operations." • SAFSTOR is defined as "the alternative in which the nuclear facility is placed and maintained in a condition that allows the nuclear facility to be safely stored and subsequently decontaminated (deferred dismantlement) to levels that permit release for unrestricted use." The SAFSTOR decommissioning alternative provides a condition that ensures public health and safety from residual radioactive contamination remaining at the site, without the need for extensive modification to the facility. Systems not required to be operational for fuel storage, maintenance and surveillance purposes during the dormancy period are to be drained, de-energized and secured. As stated in Chapter 2, the NRC has issued a rule allowing a licensee to apply for a 20-year renewal of its original operating license. To date, five reactors have been granted such license renewals; a number of other renewal applications are pending, and more applications are anticipated. A-l ------- • ENTOMB is defined as "the alternative in which radioactive contaminants are encased in a structurally long-lived material, such as concrete; the entombed structure is appropriately maintained and continued surveillance is carried out until the radioactive material decays to a level permitting unrestricted release of the property." Over the years, the basic concept of the three alternatives has remained unchanged. However, because of the accumulated inventory of spent nuclear fuel (SNF) in the reactor storage pool and the requirement for about seven years of pool storage for the SNF before transfer to dry storage, the timing and steps in the process for each alternative have had to be adjusted to reflect present conditions. For the DECON alternative, it is assumed that the owner has a strong incentive to decontaminate and dismantle the retired reactor facility as promptly as possible, thus necessitating transfer of the stored SNF from the pool to a dry storage facility on the reactor site. While continued storage of SNF in the pool is acceptable, the 10 CFR Part 50 license could not be terminated until the pool had been emptied, and only limited amounts of decontamination and dismantlement of the facility would be required. This option also assumes that an acceptable dry transfer system will be available to remove the SNF from the dry storage facility and to place it into licensed transport casks when the time comes for DOE to accept the SNF for disposal at a high level waste repository. In addition, the amended regulation stipulates that alternatives, which significantly delay completion of decommissioning, such as use of a storage period, will be acceptable if sufficient benefit results. The Commission indicated that a storage period of up to 50 years and a total of 60 years between shutdown and decommissioning is a reasonable option for decommissioning a light water reactor. In selecting 60 years as an acceptable period of time for decommissioning of a nuclear power reactor, the Commission considered the amount of radioactive decay likely to occur during an approximately 50-year storage period and the time required to dismantle the facility. In summary, the reactor facility will need to adequately cool the high-burnup assemblies from the final fuel core in the pool for up to seven years and must fulfill the regulatory requirements that critical support systems be maintained in operable conditions. Therefore, the time between shutdown, decontamination and the earliest date of dismantling efforts that would generate scrap metal is likely to be about 10 years. This interval may extend up to 60 years under the SAFSTOR decommissioning alternative. A longer time interval has the obvious benefit of greatly reducing radionuclide inventories through radioactive decay. However, a simple inverse A-2 ------- correlation between reduced levels of contamination and increased quantities of scrap metal with a potential for clearance cannot be inferred. It is likely that for most scrap metal, the longer decay time may merely affect the choice of decontamination method and/or decontamination effort required to meet a desired standard. For example, a storage period that reduces beta/gamma surface contamination of 107 dpm/100 cm2 at 10 years post-shutdown to 105 dpm/100 cm2 (i.e., a 100-fold reduction) would still require substantial decontamination in order to meet current standards defined by NRC Regulatory Guide 1.86 (U.S. AEC 1974). However, since the reduced activity would most likely be dominated by Cs-137, the method and level of effort required for successful decontamination would be different than that employed at an earlier time. The potential for clearance of scrap metal is, therefore, dictated by the cost-effectiveness with which materials can be decontaminated to acceptable levels. Estimates of scrap metal quantities must consider starting levels of contamination and whether the contamination is surficial or volumetrically distributed. Residual radioactive contaminants of reactor components/systems and building structures is generally grouped as: (1) activation products that are distributed volumetrically, (2) activation and fission products in the form of corrosion films deposited on internal surfaces, and (3) contamination of external surfaces that result from the deposition of liquid and airborne radioactive materials associated with steam, reactor coolant and radioactive waste streams. Most of the scrap metal generated by the complete dismantling of a nuclear power plant is not expected to be radioactive. The non-radioactive scrap includes the large quantities of structural metals and support systems that have not been exposed to radioactivity during reactor operations. Conversely, some metal components will undoubtedly be so contaminated as to render them unsuitable for clearance. A.2 CHARACTERISTICS OF REFERENCE REACTOR FACILITIES A crucial factor affecting the quantity of metal and associated contamination levels is the basic design of the reactor. Each of the nuclear power reactors currently operating in the U.S. is either a pressurized water reactor (PWR) or a boiling water reactor (BWR). Of the 104 operating reactors, 3 5 are BWRs manufactured by General Electric and 69 are PWRs manufactured by Westinghouse, Combustion Engineering and Babcock and Wilcox (U.S. NRC 2000). A-3 ------- Appendix A-l provides a complete listing of U.S. nuclear power reactors along with demographic data that includes projected year of shutdown. In the 1976-1980 time frame, two studies were carried out for the NRC by the Pacific Northwest Laboratory (PNL) that examined the technology, safety and costs of decommissioning large reference nuclear power plants. Those studies—"Technology, Safety and Costs of Decommissioning a Reference Pressurized Water Reactor Power Station," NUREG/CR-0130 (Smith et al. 1978) and "Technology, Safety and Costs of Decommissioning a Reference Boiling Water Reactor Power Station," NUREG/CR-0672 (Oak et al. 1980)—reflected the industrial and regulatory situation of the time. To support the final Decommissioning Rule issued in 1988, the earlier PNL studies were updated with the issuance of "Revised Analyses of Decommissioning for the Reference Pressurized Water Reactor Station," NUREG/CR-5884 (Konzek et al. 1995) and "Revised Analyses of Decommissioning for the Reference Boiling Water Reactor Power Station," NUREG/CR-6174 (Smith et al. 1996). The four NUREG reports cited above, along with several other NRC reports and selected decommissioning plans on file with the Commission, represent the primary source of information used to characterize Reference PWR and BWR facilities and to derive estimates of scrap metal inventories for the industry at large. A.2.1 Reference PWR Design and Building Structures The Reference PWR facility is the 3,500 MWt (1,175 MWe) Trojan Nuclear Plant (TNP) at Rainier, Oregon, operated by the Portland General Electric Company (PGE). Designed by Westinghouse, this reactor is considered a typical PWR that has been cited as the Reference PWR (Smith et al. 1978; Konzek et al. 1995). The NRC granted the operating license for the TNP on November 21, 1975, and the plant formally began commercial operation on March 20, 1976. TNP's operating license was scheduled to expire on February 8, 2011. However, on November 9, 1992, the TNP was shut down when a leak in the "B" steam generator was detected and the licensee notified the NRC of its decision to permanently cease operations in January 1993. Following the transfer of spent fuel from the reactor vessel to the spent fuel pool in May of 1993, TNP's operating license was reduced to a possession only license. TNP's 17-year operating period encompassed 14 fuel cycles and approximately 3,300 effective full-power days. In the decommissioning plan A-4 ------- submitted by PGE, the licensee has proposed the DECON approach with a five-year delay period prior to decontamination and dismantlement (Portland General Electric 1996). In a PWR, the primary coolant is heated by the nuclear fuel core but is prevented from boiling by a pressurizer, which maintains a pressure of about 2,000 psi. The principal systems and components of the nuclear steam supply system are illustrated in Figure A-l. Components of interest are the reactor vessel, which contains the fuel and coolant, and the reactor coolant system (RCS). The reactor vessel also contains internal support structures (not shown) that constrain the fuel assemblies, direct coolant flow, guide in-core instrumentation and provide some neutron shielding. The RCS consists of four loops for transferring heat from the reactor's primary coolant to the secondary coolant system. Each loop consists of a steam generator, a reactor coolant pump and connecting piping. Steam generated from secondary feedwater is passed through the turbine, condensed back to water by the condenser and recycled. Containment Boundary I Steam Jet Air Ejector Feedwater Pump Cooling Water Secondary Makeup Water Primary Makeup Water Denotes Reactor Water System or Radioactive Water Figure A-l. Pressurized Water Reactor (Dyer 1994) Also included in the primary loop is a small side-stream of water that is directed to the chemical volume and control system (CVCS). The CVCS provides chemical and radioactive cleanup of the primary coolant through demineralizers and evaporators. The primary coolant is reduced in A-5 ------- both pressure and temperature by the CVCS before being processed; therefore, the CVCS is often referred to as the letdown system. The water processed through the CVCS is returned to the primary loops by the charging pumps. Note that the primary coolant processed through the CVCS is brought through the containment boundary or out of the containment building, but the primary coolant providing the heat transfer to the steam generators does not pass through the containment boundary. As shown in Figure A-l, highly contaminated components of a PWR are those associated with the primary coolant system. Low-level contamination of the secondary loop is a result of steam generator tube leakage in which limited quantities of primary coolant are introduced into the recirculating steam/water. Other major contaminated systems of PWRs not shown in Figure A-l include the radioactive waste handling system and the spent fuel storage system. The principal structures requiring decontamination for license termination at the Reference PWR are the (1) reactor building, (2) fuel building and (3) auxiliary building. In addition to housing major plant systems, all three buildings contain contaminated systems and substantial quantities of contaminated structural metals that are candidates for clearance. A.2.1.1 Reactor Building The reactor building houses the nuclear steam supply system. Since its primary purpose is to provide a leak-tight enclosure under normal as well as accident conditions, it is frequently referred to as the containment building. Major interior structures include the biological shield, pressurizer cubicles and a steel-lined refueling cavity. Supports for equipment, operating decks, access stairways, grates and platforms are also part of the containment structure internals. The reactor building is in the shape of a right circular cylinder, approximately 64 m tall and 22.5 m in diameter. It has a hemispherical dome, a flat base slab with a central cavity and an instrumentation tunnel. A.2.1.2 Fuel Building The fuel building—approximately 27 m tall, 54 m long, and 19m wide—is a steel and reinforced concrete structure with four floors. This building contains the spent-fuel storage pool and its cooling system, much of the CVCS, and the solid radioactive waste handling equipment. Major steel structural components include fuel storage racks and liner, support structures for fuel A-6 ------- handling, and components, ducts and piping associated with air conditioning, heating, cooling and ventilation. A.2.1.3 Auxiliary Building The auxiliary building—approximately 30 m tall, 35 m long and 19 m wide—is a steel and reinforced concrete structure with two floors below grade and four floors above grade. Principal systems contained in the auxiliary building include the liquid radioactive waste treatment systems, filter and ion exchanger vaults, waste gas treatment system, and the ventilation equipment for the containment, fuel and auxiliary buildings. A.2.1.4 Control and Turbine Buildings Other major building structures with substantial metal inventories include the control building and the turbine building. The principal contents of the control building are the reactor control room, and process and personnel facilities. The principal systems contained in the turbine building are the turbine generator, condensers, associated power production equipment, steam generator auxiliary pumps, and emergency diesel generator units. Barring major system failures (e.g., steam generator failure) most scrap metal derived from these systems can be assumed to be free of contamination and can, therefore, be excluded from the inventories of scrap metal which are candidates for clearance. A.2.2 Reference BWR Design and Building Structures The 3,320 MWt (1,155 MWe) Washington Public Power Supply System (WPPSS) Nuclear Project No. 2 located near Richland, Wash., is the basis for the Reference BWR facility (Oak et al. 1980; Smith etal. 1996). The design of a BWR (see Figure A-2) is simpler than a PWR inasmuch as the reactor coolant water is maintained near atmospheric pressure and boiled to generate steam. This allows the coolant to directly drive the turbine. Thereafter, the steam is cooled in the condenser and returned to the reactor vessel to repeat the cycle. In a BWR, the contaminated reactor coolant comes in contact with most major reactor components, including the reactor vessel and piping, steam turbine, steam condenser, feedwater system, reactor coolant cleanup system and steam jet A-7 ------- Reactor Vessel Steam Jet Air Ejector V Reactor Water Clean-up System Containment Boundary Reactor Pump Cooling Water Denotes Reactor Water System or Radioactive Water Figure A-2. Boiling Water Reactor (Dyer 1994) air ejector system. As with the PWR, other major contaminated systems include the radioactive waste treatment system and spent fuel storage system. The principal buildings requiring decontamination and dismantlement in order to obtain license termination at the reference BWR power station are the reactor building, the turbine generator building, and the radwaste and control building. These three buildings contain essentially all of the activated or radioactively contaminated material and equipment within the plant. A.2.2.1 Reactor Building The reactor building contains the nuclear steam supply system and its supporting systems. It is constructed of reinforced concrete capped by metal siding and roofing supported by structural steel. The building surrounds the primary containment vessel, which is a free-standing steel pressure vessel. The exterior dimensions of the Reactor Building are approximately 42 m by 53 m in plan, 70 m above grade and 10.6 m below grade to the bottom of the foundation. A-8 ------- A.2.2.2 Turbine Building The turbine building, which contains the power conversion system equipment and supporting systems, is constructed of reinformed concrete capped by steel-supported metal siding and roofing. This structure is approximately 60 m by 90 m in plan and 42.5 m high. A.2.2.3 Radwaste and Control Building The radwaste and control building houses, among other systems: the condenser off-gas treatment system, the radioactive liquid and solid waste systems, the condensate demineralizer system, the reactor coolant cleanup demineralizer system and the fuel-pool cooling and cleanup demineralizer system. The building is constructed of reinforced concrete, structural steel, and metal siding and roofing. This structure is approximately 64 by 49 m in plan, 32 m in overall height, and stands as two full floors and one partial floor above the ground floor. A.3 RESIDUAL ACTIVITIES IN REFERENCE REACTOR FACILITIES Significant levels of contamination remain in a nuclear power station following reactor shutdown, even after all spent nuclear fuel has been removed. Neutron-activated structural materials in and around the reactor pressure vessel contain most of the residual activity in a relatively immobile condition. Other sources of radioactive contamination comprise activated corrosion products and fission products leaked from failed fuel, which are transported throughout the station by the reactor coolant streams. The origin and mobility of radioactive contaminants following reactor shutdown leads to grouping of residual activities into five categories of different binding matrices. These categories include: 1. Activated Stainless Steel. Reactor internals, composed of Type 304 stainless steel, become activated by neutrons from the core. Radionuclides have very high specific activities and are immobilized inside the corrosion-resistant metal. 2. Activated Carbon Steel. Reactor pressure vessels are made of SA533 carbon steel that becomes activated by neutron bombardment. The specific activities are considerably lower than in the stainless steel internals, and the binding matrix is much less corrosion resistant. 3. Activated Structural Steel, Steel Rebar and Concrete In the reactor cavity, these components become activated by neutrons escaping from the reactor vessel. Significant A-9 ------- activation occurs along approximately 15 feet of the reactor cavity vertically centered on the reactor core and to a depth of about 16 inches in the concrete. 4. Contaminated Internal Surfaces of Piping and Equipment. Activated corrosion and fission products travel through the radioactive liquid systems in the plant. A portion forms a hard metallic oxide scale on the inside surfaces of pipes and equipment. 5. Contaminated External Surfaces. External surfaces may become contaminated over the lifetime of the plant, primarily from leaks, spills and airborne migration of radionuclides contained in the reactor coolant water (RCW). The specific activity of RCW is low, but the contamination is easily mobilized and may be widespread. All of the neutron-activated metals/materials are contained in the reactor pressure vessel, vessel internals, and structural components inside and within the concrete biological shield. Total quantities and the relative radionuclide composition of the residual activity are not only affected by reactor design (BWR vs. PWR) but are also strongly influenced by numerous other factors including (1) fuel integrity, (2) rated generating capacity and total years of operation, (3) composition of metal alloys in reactor components and the RCS, (4) coolant chemistry and water control measures, and (5) the performance and/or failures of critical systems and their maintenance over the initial 40-year span of the operating license (see footnote on page A-l). Table A-l provides summary estimates of typical residual activities for each of the five major source categories. Inspection of the data reveals that the volumetrically activated stainless steel represents the overwhelming majority of the residual activities. Much smaller activities are found in volumetrically activated carbon steel and internal and external surface contamination consisting of activation and fission products. A more detailed discussion of residual activity by source category is given below. A.3.1 Neutron-Activated Reactor Components and Structural Materials Contamination of reactor components and structural materials by neutron activation is the result of normal reactor operation. The interaction of neutrons with constituents of stainless steel, carbon steel and concrete in and around the reactor vessel results in high in-situ activities. The radionuclide inventories include significant activities of Cr-51, Mn-54, Fe-55, Fe-59, Co-58, Ni- 59 and Ni-63. The specific activities of various radionuclides in materials exposed to a neutron flux is highly variable and depends upon (1) the concentration of the parent nuclide and its A-10 ------- neutron cross-section, (2) the radioactive half-life of the radionuclide, (3) the neutron flux intensity at the given location, and (4) the duration of neutron exposure. Table A-l. Sources of Residual Activities in Reference BWR and PWR Source Activated Stainless Steel Activated Carbon Steel Activated Structural Components, Rebar, Metal Plates, I-Beams Internal Surface Contamination of Piping and Equipment External Contamination of Equipment, Floors, Walls, Other Surfaces Residual Activity (Ci) BWRa 6.6e+06 2.9e+03 1.2e+03 8.5e+03 l.le+02 PWRb 4.8e+06 2.4e+03 1.2e+03 4.8e+03 l.le+02 c aOaketal. 1980 b Smith etal. 1978 c Implied value (U.S. NRC 1994) A.3.1.1 Reference BWR The average activity concentrations and estimated total activities for Reference BWR structural components with significant amounts of neutron activation are listed in Table A-2. The Reference BWR reactor vessel is fabricated of SA533 carbon steel about 171 mm thick and is clad internally with 3 mm of Type 304 stainless steel. The total mass of the empty vessel is about 750 metric tons (t). The major internal components include the fuel core support structure; steam separators and dryers; coolant recirculation jet pumps; control rod guide tubes; distribution piping for feedwater, core sprays and liquid control; in-core instrumentation, and miscellaneous other components. Collectively, these internals, made of stainless steel, represent about 250 t. A.3.1.2 Reference PWR The right circular cylinder of the Reference PWR is constructed of carbon steel about 216 mm in thickness and is clad on the inside with stainless steel or Inconel having a thickness of about 4 mm. The approximate dimensions of the vessel are 12.6 m high and 4.6 m in outer diameter. The vessel weighs about 400 t. A-ll ------- Table A-2. Estimated Activities of Neutron-Activated Reactor Components in a BWR Component (number) Core Shroud (1) Jet Pump Assembly (10) Reactor Vessel (1) Cladding Shell Wall Steam Separator Assembly (1) Shroud Head Plant Steam Separator Risers Top Fuel Guide (1) Orificed Fuel Support (193) Core Support Plate (1) Incore Instrument Strings (55) Control Rod (185) Control Rod Guide Tube (185) Total Average Activity Concentration (Ci/m3) 1.68e+06 2.62e+04 1.07e+03 1.12e+02 1.03e+04 2.53e+03 9.71e+04 l.Ole+03 2.56e+02 7.67e+05 5.11e+05 2.16e+02 Total Activity (Ci) 6.30e+06 2.00e+03 2.16e+03 9.60e+03 3.01e+04 7.01e+02 6.50e+02 1.10e+04 1.78e+05 9.47e+01 6.55e+06 Source: Oak et al. 1980 The vessel's internal structures support and constrain the fuel assemblies, direct coolant flow, guide in-core instrumentation and provide some neutron shielding. The principal components are: the lower core support assembly, which includes the core barrel and shroud, with neutron shield pads, and the lower core plate and supporting structure; and the upper core support and in- core instrumentation support assemblies. These structures are made of 304 stainless steel and have a total mass of about 1901. Based on 40 years of facility operation and assuming 30 effective full-power years (EFPY) of reactor operation, the total activity contained in the activated vessel and internals is estimated to be 4.8 million curies (see Table A-3). Extra-vessel materials subject to significant neutron activation (=10 curies) includes the reactor cavity steel liner and a limited quantity of reinforcement steel (rebar). Additionally, the concrete bioshield contains an estimated total inventory of about 1,200 curies. A-12 ------- Table A-3. Neutron-Activated Reactor Components in a PWR Component Shroud Lower 4.7 m of core barrel Thermal shield Vessel inner cladding Lower 5.02 m of vessel wall Upper grid plate Lower grid plate Total Average Activity Concentration (Ci/m3) 2.97e+06 3.07e+05 1.45e+05 7.73e+03 9.04e+02 4.20e+04 1.12e+06 Total Activity (Ci) 3.43e+06 6.52e+05 1.46e+05 1.50e+03 1.76e+04 2.43e+04 5.53e+05 4.82e+06 Source: Smith et al. 1978 The projected estimates of Table A-3 for the Reference PWR (i.e., Trojan Nuclear Plant) made in 1978 can be compared to the more current estimates contained in that plant's decommissioning plan (submitted to the NRC in 1996). Table A-4 identifies revised calculated inventories of activation products for 1993, or one year after shutdown. The recalculated value of about 4.2 million curies is about 13% lower than the original estimate of 4.8 million curies and principally reflects the difference between 17 years of actual plant operation and the initial projection of 40 years. Table A-4. Activation Levels at Trojan Nuclear Plant One Year after Shutdown System Reactor Vessel Reactor Vessel Internals Vessel Clad and Insulation Bioshield Wall Total Activity (Ci) 6.20e+03 4.16e+06 2.37e+04 8.30e+02 4.19e+06 The considerably higher activities calculated for a Reference BWR primarily reflect the larger size and mass of the vessel and its internals. A-13 ------- For both PWR and BWR plants, the range of activity concentrations among individual reactor components at time of shutdown is likely to vary over several orders of magnitude. Nevertheless, even those components with the lowest activity concentrations would still have residual activities far in excess of any conceivable levels that would permit clearance. (Note: at a specific gravity of 7.86, a cubic meter of steel containing one curie has a specific activity of 0.13 |lCi/g.) Furthermore, these components also exhibit high levels of interior surface contamination. While surface contamination is potentially removable, the volumetrically distributed activation products are not. For this reason, the reactor vessel and all internal components identified in Tables A-2 and A-3 must be excluded from plant material inventories which are potential candidates for clearance. Excluded for similar reasons are certain metal components used for structural support and reinforcement (i.e., rebar, I-beams, and floor and reactor cavity liner plates) that exhibit significant levels of activation products. Scrap metal that can potentially be cleared can therefore originate only in reactor systems and structural components where contamination is limited to interior and exterior surfaces. A.3.2 Internal Surface Contamination of Equipment and Piping Activated corrosion products from structural materials in contact with the reactor coolant and fission products from leaking fuel contribute to the radioactive contamination of reactor coolant streams during plant operation. Although most of these contaminants are removed through filtration and demineralization by the CVCS, a small portion remains in the coolant. With time, some of the contaminants, principally the neutron-activated, insoluble corrosion products, tend to deposit on inner surfaces of equipment and piping systems. The resulting metal oxide layer consists primarily of iron, chromium and nickel with smaller, but radiologically significant, quantities of cobalt, manganese and zinc. This section characterizes the mixture of internal surface contaminants and their relative distribution within major components associated with BWR and PWR power plants. A.3.2.1 Measurements of Internal Surface Contamination at Six Nuclear Power Plants In a 1986 PNL study, six nuclear power plants—three PWRs and three BWRs—were assessed for residual inventories and distributions of long-lived radionuclides following plant shutdown (Abel et al. 1986). Residual concentrations in the various plant systems decreased in the A-14 ------- following order: (1) primary coolant loop, (2) radwaste handling system, and (3) secondary coolant loop in PWRs and condensate system in BWRs. Table A-5 lists total estimated activities at the six plants, as well as the electrical ratings and the approximate number of operational years of the plants at the time of the assessments. The operational periods ranged from 8.3 years for Turkey Point Unit 3 to slightly over 18 years for Dresden Unit 1. Table A-5. Residual Activities and Operating Parameters of Six Nuclear Power Plants* Stations Humboldt Bay Dresden- 1 Monticello Indian Point- 1 Turkey Point-3 Rancho Seco Total Inventory (Ci) 600 2,350 514 1,050 2,580 4,470 Period of Operation (y) 13 18.3 10 11 8.3 8.8 Power Rating (MWe) 63 210 550 170 660 935 Reactor Type BWR BWR BWR PWR PWR PWR Source: Abel et al. 1986 * Total inventory includes radionuclides with half-lives greater than 245 days (i.e., Zn-65); inventories in activated metal components of the reactor pressure vessel and internals and activated concrete are excluded. The relative radionuclide composition of internally contaminated surfaces at the six plants also showed considerable variation (see Table A-6). Fluctuations in compositions were due to numerous factors including: (1) the elapsed time since reactor shutdown; (2) rated generating capacity; (3) materials of construction of the operating systems; (4) reactor type (PWR or BWR); (5) coolant chemistry and corrosion control; (6) fuel integrity during operations; and (7) episodic equipment failure and leakage of contaminated liquids. Inventories include only the radioactive contamination of corrosion film and crud2 on surfaces of the various plant systems, and do not include the highly activated components of the pressure vessel. The most abundant radionuclides in samples two to three months old included Mn-54, Fe-55, Co-58, Co-60 and Ni-63. Zinc-65 was present in relatively high concentrations in BWR corrosion film samples. However, Fe-55, and Co-57+Co-60 were the most abundant radionuclides at all stations except Monticello. These radionuclides constituted over 95% of the A colloquial term for corrosion and wear products (rust particles, etc.) that become radioactive (i.e., activated) when exposed to radiation. The term is actually an acronym for Chalk River Unidentified Deposits, the Canadian plant at which the activated deposits were first discovered. A-15 ------- estimated inventories at Humboldt Bay and Turkey Point. At Indian Point-1, Dresden-1, Turkey Point-3 and Rancho Seco, they accounted for 82, 74, 98 and 70%, respectively, of the total estimated inventory. Although Fe-55 and Co-60 accounted for the majority of the inventory (greater than 60% at five of the six stations), the relationship between the two radionuclides was quite variable. The transuranic nuclides (Pu-238, Pu-239, Pu-240, Am-241, Cm-242 and Cm- 244) constituted varying percentages of the total inventory, ranging from 0.001% at Rancho Seco to 0.1% at Dresden-1. Table A-6. Relative Activities of Long-Lived Radionuclides at Six Nuclear Power Plants* Radionuclide Mn-54 Fe-55 Co-57 Co-60 Ni-59 Ni-63 Zn-65 Sr-90 Nb-94 Tc-99 Ag-llOm 1-129 Cs-137 Ce-144 TRU" Total (Ci) Relative Activity, Decay-Corrected to Shutdown Date (%)T BWRs Humboldt Bay 3 90 — 6 — 0.2 — 4e-03 < 4e-03 3e-04 — < 3e-06 0.5 — 5e-03 596 Dresden- 1 0.9 28 — 46 0.09 5 19 7e-03 < 3e-03 4e-05 — < le-05 0.04 1 0.1 2,350 Monticello 1 1 — 11 — 0.04 84 2e-03 <0.1 8e-05 — < le-06 2 — 8e-03 448 PWRs Indian Point- 1 4 67 — 15 0.02 2 11 7e-04 8e-04 8e-05 — 2e-05 0.5 — 2e-03 1,070 Turkey Point-3 0.4 31 43 24 4e-03 0.1 1 8e-04 < 4e-03 8e-03 — < 3e-03 — 0.2 6e-03 2,580 Rancho Seco 4 28 24 18 0.1 19 0.09 <0.01 < 4e-03 < 5e-03 4 < le-05 0.4 <0.04 le-03 4,460 Source: Abel et al. 1986 * Excludes activated metal components of the reactor pressure vessel and internals and activated concrete. ' Relative activity of each nuclide as a percentage of total activity at each power plant ** Transuranic alpha-emitting radionuclides with half-lives greater than 5 years, including Pu-238, Pu-239, Pu-240, Am-241, Am-243 and Cm-244. A-16 ------- Secondary coolant loops in PWRs and condensate systems in BWRs contained much lower activity concentrations than observed in primary loop or feedwater samples. Typically, concentrations were two or more orders of magnitude lower in secondary system samples. As expected, the steam generators contained the single largest repository of internally deposited radionuclides at the PWR stations examined (see Table A-7). The percentages of the total residual radionuclide inventories in the steam generators were 77, 89 and 94% for Indian Point-1, Turkey Point-3 and Rancho Seco, respectively. The other repository of significance in a PWR is the radwaste system, which typically contained 5 to 10% of the total residual inventory. Table A-7. Distribution of Activities in Major Systems of Three PWRs (%) System Steam Generators Pressurizer RCS Piping Piping (Except RCS) Secondary Systems Radwaste Turkey Point-2 89 0.5 0.9 <0.01 0.1 9.2 Indian Point- 1 77 0.5 2.6 14 0.2 7 Rancho Seco 94 0.33 0.71 <0.01 0.05 5 Average 86.7 0.4 1.4 4.7 0.1 7.1 Source: Abel et al. 1986 A.3.2.2 Internal Surface Contamination Levels Reported in Decommissioning Plans A small number of commercial nuclear power facilities, which have experienced a premature shutdown or have projected shutdown within the next few years, have submitted a decommissioning plan to the NRC for review. Summarized below are system-specific internal contamination levels reported for one BWR and two PWRs. Big Rock Point Nuclear Plant The Big Rock Point Nuclear Plant is a small (67 MWe) BWR designed by the General Electric Company and constructed by Bechtel Power Corporation. Owned and operated by Consumers Power Company, the plant started commercial operation in March 1963 and was shut down in August 1997. Table A-8 presents summary data of systems internally contaminated (Consumers Power 1995). A-17 ------- Table A-8. Internal Contamination Levels of Big Point Nuclear Plant at Shutdown System Liquid Rad Waste Tanks Nuclear Steam Supply RDS Main Steam System Fuel Pool Liquid Radwaste System Condensate System Resin Transfer System Off-gas System Control Rod Drive Rad Waste Storage Fuel Handling Equip Heating & Cooling System Surface Contamination Level (dpm/100 cm2) 3e+10 9e+09 3e+09 4e+08 4e+08 4e+08 5e+07 3e+07 3e+07 6e+06 9e+05 7e+05 3e+05 San Onofre Nuclear Generation Station Unit 1 (SONGS 1) SONGS 1 is a 436-MWe PWR that started operation in 1968. As a result of an agreement with the California Public Utility Commission, operation of SONGS 1 was permanently discontinued on November 30, 1992 at the end of fuel cycle #11. A preliminary decommissioning plan, submitted to the NRC on December 1, 1992, proposed to maintain SONGS 1 in safe storage until the permanent shutdown of SONGS 2 and 3. SONGS 2 and 3 are licensed to operate until 2013. In support of the SONGS 1 decommissioning plan, scoping surveys and analyses were performed that supplemented an existing radiological data base (Southern California Edison 1994). The containment building, fuel storage building and radwaste/auxiliary building were identified as the principal structures containing significant levels of radioactivity within plant systems. Systems were grouped by contamination levels defined as (1) highly contaminated, (2) medium-level contaminated and (3) low-level contaminated. Based on total radionuclide inventories and surface areas, an average contamination level for each of the three groupings was derived (see Table A-9). A-18 ------- Table A-9. Plant Systems Radioactivity Levels at SONGS 1 Plant Systems High-Level Contaminated Systems: IDS Letdown PAS Post Accident Sampling System PZR Pressurizer Relief RCS Reactor Coolant RHR Residual Heat Removal RSS Reactor Sampling SFP Spent Fuel Pool Cooling VCC Volume Control Medium-Level Contaminated Systems: BAS Boric Acid CWL Containment Water Level RCP RCP Seal Water RLC Radwaste Collection RMS Radiation Monitoring RWG Radwaste Gas RWL Radwaste Liquid CRS (Containment Spray) Recirculation SIS Safety Injection Low-Level Contaminated Systems: AFW Auxiliary Feedwater CCW Component Cooling CND Condensate SHA Sphere Hydrazine Addition CSS Condensate Sampling CVD Condensate Vents & Drains CVI Cryogenics CWS Circulating Water FES Flash Evaporator FPS Fire Protection FSS Feed Sampling FWH Feedwater Heaters FWS Feedwater MSS Main Steam MVS Miscellaneous Ventilation PSC Turbine Sample Cooling SOW Service Water SWC Salt Water Cooling TCW Turbine Cooling Total Area (cm2) 1 .26e+08 1 .25e+08 3.18e+08 Surface Contamination Level (dpm/100 cm2) 3.6e+09 1 .9e+06 8.36+03 Total Activity (Ci) 2.08e+03 1.086+01 1.21e-02 A-19 Continue ------- Back Yankee Rowe Yankee Rowe is a 167-MWe PWR with a startup date of August 19, 1960. It started commercial operation in July, 1961 and was shutdown in October, 1991 following 21 fuel cycles and 8,052 EFPD. In the 1993 decommissioning plan submitted to the NRC, systems with significant internal surface contamination were identified, as shown in Table A-10 (Yankee Atomic 1995). Table A-10. Average Internal Contamination Levels of Reactor Systems at Yankee Rowe System Main Coolant Spent Fuel Cooling Waste Disposal Primary Plant Vent & Drain Charging & Volume Control Shutdown Cooling Fuel Handling Letdown/Purification Primary Plant Sampling Safety Injection Safe Shutdown Vol. Control Heating & Cooling Vol. Control Vent. & Purge Post Accident H2 Control Chemical Shutdown Surface Contamination Level (dpm/100 cm2) 7.1e+09 3.3e+08 1.2e+07 1.2e+07 1.2e+07 1.2e+07 1.7e+06 1.4e+06 1.4e+06 1.4e+05 1.4e+05 1.2e+04 1.2e+04 1.2e+04 l.le+04 The data on facilities that have submitted decommissioning plans have limited applicability to a generic analysis because of: (1) their limited years of operation, (2) abnormal events and operating conditions that prompted premature shutdown and/or, (3) size and design of the facilities. A.3.2.3 Levels of Internal Surface Contamination Derived for Reference BWR Internal surface contamination levels in BWR systems and piping reflect the radionuclide concentrations in the reactor coolant, steam and condensate. Summary estimates of activities in A-20 ------- corrosion films deposited on internal surfaces of equipment and piping are cited by Oak et al. (1980) for a Reference BWR. The radionuclide composition of corrosion films is shown in Table A-l 1. About 86% of the estimated inventory at shutdown was due to two nuclides, Co-60 and Mn-54 (Co-60 constituted nearly half of the total inventory). It should be noted that internal surface deposited nuclides generally do not include large amounts of fission products. Although fission products do exist in the reactor coolant, they are generally soluble and remain in solution rather than plate out along with neutron-activated corrosion products. The buildup of coolant contaminants is controlled by the CVCS system, which continuously removes both insoluble (particulate) and soluble contaminants. Table A-l 1. Activated Corrosion Products in the Reference BWR Nuclide Cr-51 Mn-54 Fe-59 Co-58 Co-60 Zn-65 Zr-95 Nb-95 Ru-103 Ru-106 Cs-134 Cs-137 Ce-141 Ce-144 Total Half-Life 27.7 d 312.1 d 44.5 d 70.88 d 5.271 y 244.26 d 64.02 d 34.97 d 39.27 d 373.6 d 2.065 y 30.07 y 32.5 d 284.9 d Relative Activity at Various Times After Shutdown* 0 2.1e-02 3.9e-01 2.5e-02 9.3e-03 4.7e-01 6.1e-03 4.0e-03 4.0e-03 2.3e-03 2.8e-03 1.9e-02 3.4e-02 3.0e-03 8.1e-03 1.0 lOy — 1.2e-04 — — 1.3e-01 1.9e-07 — — — 3.2e-06 — 2.7e-02 — l.le-06 1.5e-01 30 y — — — — 9.1e-03 — — — — — — 1.7e-02 — — 2.6e-02 50 y — — — — 6.6e-04 — — — — — — l.le-02 — — l.le-02 Activities of individual nuclides, normalized to the total activity at shutdown The total radionuclide inventory has been estimated at 8,500 curies, with 6,300 curies associated with internal equipment surfaces and the remaining 2,200 curies associated with internal piping surfaces (see Table A-12). A-21 ------- Table A-12. Distribution of Activated Corrosion Products on Internal Surfaces of Reference BWR Location Piping Equipment: Reactor Building Turbine Building Radwaste & Control Total Surface Area (m2) 3.4e+04 8.6e+03 2.0e+05 1.4e+03 2.4e+05 Areal Activity Concentration (Ci/m2) 6.5e-02 2.2e-01 6.0e-03 2.3e+00 2.6e+00 Total Surface Activity (Ci) 2.2e+03 1.9e+03 1.2e+03 3.2e+03 8.5e+03 Source: Oak et al. 1980, vol. 1, Table 7.4-10 For the residual inventory of 6,300 curies on equipment, an estimated 30% was associated with equipment in the reactor building, about 19% was associated with the condenser and feed-water heaters located in the turbine building, and about 51% involved internal deposition on equipment in the radwaste and control building. Of the 2,200 curies present in piping, approximately 56% were estimated to be associated with the reactor coolant piping and 44% with condensate piping. Presented below is a more thorough analysis of piping data. Contaminated Piping Internal surface contamination levels of BWR piping can be most useful when grouped according to direct or indirect contact with reactor coolant, steam/air and condensate. Deposition levels for reactor coolant and condensate were based on empirical dose rate measurements that were correlated to contamination levels for a specific pipe size and schedule. A summary of measured dose rate data and derived deposition levels is shown in Table A-13. Table A-14 provides a detailed accounting of radionuclide inventories derived for various size piping made of aluminum, carbon steel, and stainless steel in contact with reactor coolant, steam/ air, or condensate. A-22 ------- Table A-13. Contact Dose Rate and Internal Surface Activity of BWR Piping Medium in Pipes Reactor Coolant Steam/ Air Condensate Nominal O.D. (mm) 610 914 610 Wall Thickness (mm) 59.5 20.4 26.0 Contact Dose Rate (mR/hr) 700 70 50 Areal Activity Concentration (Ci/m2) 1.1 0.005 0.05 Contaminated Equipment Contamination on internal surfaces of BWR equipment in contact with reactor coolant was estimated from measurements taken on the heat exchanger in the reactor coolant cleanup system. In general, equipment in contact with steam or condensate was assumed to reach the same levels as previously cited for BWR piping. Exceptions were the lower values assigned to steam surfaces for the turbine and feedwater heaters. Table A-15 provides estimates of contamination levels assigned to BWR equipment. Table A-16 identifies the major system components and radionuclides inventories based on location and contact with reactor coolant, steam, condensate and radwaste. A.3.2.4 Levels of Internal Surface Contamination for Reference PWR Radioactive contamination levels associated with internal surfaces of piping and equipment for a Reference PWR have been estimated by Smith et al. (1978). At time of shutdown, the fractional contributions of various radionuclides deposited on internal surfaces of the primary loop of a PWR are shown in Table A-17. Estimates of internal surface activity concentrations for major systems and components were based on models which correlated external dose rate measurements with internal contamination analyses, taking into account source geometry and shielding factors (see Table A-18). Empirical dose rate measurements showed that reactor vessel and steam generator internal surfaces in contact with primary coolant, on average, would yield contamination levels of about 0.23 Ci/m2 at time of shutdown. A-23 ------- Table A-14. Estimates of Internal Contamination for Reference BWR Piping Pipe Material/ Contact Medium Outer Diameter (mm) 60 L (m) A (m2) Act. (Ci) 152 L (m) A (m2) Act. (Ci) 356 L (m) A (m2) Act. (Ci) 533 L (m) A (m2) Act. (Ci) 660 L (m) A (m2) Act. (Ci) 914 L (m) A (m2) Act. (Ci) Total L (m) A (m2) Act. (Ci) Aluminum Steam/Air Condensate 4,300 — 81 — 0.4 — 1,400 14 640 6.7 3.2 0.3 130 — 140 0.7 — — — — — — — — — 5,830 14 861 7 4 0.3 Carbon Steel Rx coolant Steam/Air Condensate 380 1,200 7,400 71 220 1,400 78 1.1 7.0 1,500 1,800 8,300 700 880 3,900 770 4.4 200 61 5,600 5,100 68 6,300 5,700 75 32 280 55 1,200 2,800 92 2,000 4,600 100 10 230 — 950 370 — 200 770 — 9.8 38 — 440 210 — 1,300 610 — 6.3 31 1,996 11,190 24,180 931 10,900 16,980 1,023 64 786 Stainless Steel Rx coolant Steam/Air Condensate Total 8 280 7,000 20,568 1.5 53 1,300 3,127 1.6 0.3 66 154 34 — 1,600 14,648 16 — 780 6,923 18 — 39 1,035 61 — 220 11,172 68 240 12,516 75 12 475 55 — 4,110 92 — 6,784 100 — 440 — — 1,320 — — 970 — — 48 — — 650 — ~ 1,910 — — 37 158 280 8,820 52,468 178 53 2,320 32,229 195 0 117 2,189 to Note: Average contamination level = 68 mCi/m2 (1.5 x 109 dpm/100 cm2) ------- Table A-15. Summary of Contamination Levels in BWR Equipment Equipment Category Reactor Coolant Equipment Steam Equipment Turbine Condensate Equipment Main Condenser Feedwater Heaters Concentrated Waste Tanks/Equipment Areal Activity Concentration (Ci/m2) 3.6e-01 5.0e-03 5.0e-04 5.0e-02 5.0e-03 5.0e-03 5.0e+00 The total surface activity on the reactor vessel and its internal components, which have a total surface area of 570 m2, was estimated to be about 130 Ci. The surface activity on the four steam generators, which have a total mass of 1,2511 and a combined surface area of about 19,000 m2, was estimated to be approximately 4,400 Ci, which represents 90% of the total deposited activity. The areal concentration of activated corrosion products in the 89-metric ton pressurizer was assumed to be about 0.04 Ci/m2. Since the internal surface area is about 87 m2, the total deposited activity was estimated to be about 4 Ci. Table A-16. Estimated Internal Surface Activities in BWR Systems Building/System Reactor Building Fuel Pool Heat Exchangers Skimmer Surge Tanks Fuel Pool, RxWall, Dryer & Sep. Pool RBCC Water Heat Exchangers RMCU Regenerative Heat Exchangers RWCU Nonregenerative Heat Exchangers RHR Heat Exchangers Reactor Vessel Total Total Internal Area (m2) 8.0e+02 1 .Oe+02 1.4e+03 1.8e+03 2.5e+02 1 .76+02 1.5e+03 2.66+03 8.6e+03 Areal Activity Concentration (Ci/m2) 5.06-02 5.0e-02 5.06-02 5.0e-02 3.66-01 3.6e-01 3.66-01 3.6e-01 Total Activity (Ci) 4.06+01 5.0e+01 7.06+01 9.0e+01 9.06+01 6.0e+01 5.46+02 9.4e+02 1.96+03 A-25 ------- Table A-16 (continued) Building/System Turbine Generator Building Main Condenser Steam Jet Air Ejector Condenser Gland Seal Steam Condenser Condensate Storage Tanks Low-Pressure Feedwater Heaters Evaporator Drain Tanks Reheater Drain Tanks Moisture Separator Drain Tank Main Turbine Steam Evaporator Turbine Bypass Valve Assembly Moisture Separator Reheaters Seal Water Liquid Tank Pumped Drain Tank High-Pressure Feedwater Heaters Total Radwaste and Control Building Condensate Phase Separator Tanks Condensate Backwash Receiver Tank Waste Collector Tank Waste Surge Tank Waste Sample Tanks Floor Drain Collector Tank Waste Sludge Phase Separator Tank Floor Drain Sample Tank Chemical Waste Tanks Distillate Tanks Detergent Drain Tank Decontamination Solution Cone. Waste Tk. Spent Resin Tank Cleanup Phase Separator Tanks Decontamination Solution Concentrator Total Total Internal Area (m2) 7.9e+04 1.6e+03 3.5e+02 1 .6e+03 7.5e+04 1.06+01 8.4e+02 S.Oe+01 2.6e+03 2.06+03 1.5e+01 1.86+04 1.2e+01 2.76+01 1 .7e+04 2.06+05 1 .8e+02 8.56+01 1 .Oe+02 1.96+02 1 .6e+02 1.16+02 6.1e+01 7.86+01 1.5e+02 1.56+02 3.2e+01 2.36+01 1.3e+01 6.86+01 1.9e+01 1.46+03 Areal Activity Concentration (Ci/m2) 5.0e-03 5.06-02 5.0e-02 5.06-02 5.0e-03 5.06-02 5.0e-02 5.06-03 5.0e-04 5.06-03 5.0e-03 5.06-03 5.0e-02 5.06-02 5.0e-03 5.06+00 5.0e+00 5.06-02 5.0e+00 5.06-02 5.0e-02 5.06+00 5.0e-02 5.06-02 5.0e-02 5.06-02 5.0e+00 5.06+00 5.0e+00 5.06+00 Total Activity (Ci) 3.9e+02 8.06+01 1.7e+01 8.06+01 3.7e+02 5.06-01 4.2e+01 1.56-01 1.3e+00 1.06+01 7.5e-01 9.06+01 6.06-01 1.4e+00 8.56+01 1.2e+03 9.06+02 4.2e+02 5.06+00 9.5e+02 8.06+00 5.5e+00 S.Oe+02 3.9e+00 7.56+00 7.5e+00 1.66+01 1 .2e+02 6.56+01 3.4e+02 9.56+01 3.2e+03 Source: Oak et al. 1980, vol. 2, Table E.2-7 RCS piping includes those sections of piping interconnecting the reactor vessel, steam generators, reactor coolant pumps and various other components, as shown in Figure A-3. RCS A-26 ------- Aux Spray From CVCS L i Shield \ ; X Pressuriz Reactor Coolant Pump »f^ Loop 2 Steam __ Generator ^ ' -A /^ n> |( « . - . , ™__ j-L Shield J_ k V- Prpssuri7pr * : i i [ k. Dnlinf T-inlr \I P er Heater Controller Reactor j—0.699 m I.D. Pipe ^uu,«,,,i , »„,„ y \ / ¥ \ / Steam \ \ / Generator \ / Nfil Loop 3 0.736 m I.D. Pip^ / /\ 0.787 m I.D. Pipe Steam Generator Tube Safety Injection Safety Injection Safety Injection Safety Injection Shell Steam Generator Reactor Coolant Pump Reactor Coolant Pump Reactor Vessel From CVCS Normal Charging From CVCS ALT Charging Figure A-3. Reactor Coolant System in a Four-Loop PWR (Abel et al. 1996) piping primarily involves large diameter, thick-walled pipes. The inside diameter typically ranges from 699 mm to 787 mm, with a corresponding wall thickness of between 59 and 66 mm. From dose rate measurements—about 600 mR/hr—the internal surface activity concentration on RCS piping was estimated at 0.86 Ci/m2. The total activity on the RCS piping, which has an internal surface area of about 190 m2 and a mass of 100 t, is estimated to be 160 Ci. The average activity concentration on the inner surfaces of non-RCS or auxiliary system piping is estimated to be about 0.06 Ci/m2, based on external dose rate measurements. This value, together with the pipe specifications listed in Table A-19, yields a total surface activity of about 71 Ci on the inner surfaces of all non-RCS PWR piping. A-27 ------- Table A-17. Internal Surface Contamination in the Reference PWR Primary System Radionuclide Cr-51 Mn-54 Fe-59 Co-58 Co-60 Zr-95 Nb-95 Ru-103 Cs-137 Ce-141 Total Half- Life 27.7 d 312.1 d 2.73 y 70.88 d 5.271 y 64.02 d 34.97 d 39.27 d 30.07 y 32.5 d Areal Activity Concentration (|lCi/m2) 5.30e+03 8.00e+03 1.80e+03 l.OOe+05 7.10e+04 8.80e+03 1.20e+04 5.90e+03 2.60e+02 1.50e+04 2.30e+05 Relative Activity at Various Times After Shutdown* 0 2.40e-02 3.60e-02 8.20e-03 4.60e-01 3.20e-01 5.60e-02 5.60e-02 2.60e-02 1.20e-03 6.60e-02 1.0 lOy — l.le-05 — — 8.6e-02 — — — 9.5e-04 — 8.7e-02 30 y — — — — 6.2e-03 — — — 6.0e-04 — 6.8e-03 50 y — — — — 4.5e-04 — — — 3.8e-04 — 8.3e-04 Source: Smith et al. 1978, vol. 1 Activities of individual nuclides, normalized to the total activity at shutdown Table A-18. Activated Corrosion Products on the Interiors of PWR Systems Systems Reactor Vessel and Internals Steam Generators Pressurizer Piping (Except RCS) RCS Piping Total Surface Area (m2) 5.7e+02 1.9e+04 8.7e+01 l.le+03 1.9e+02 2.1e+04 Areal Activity Concentration (Ci/m2) 0.23 0.23 0.05 0.05 0.84 Total Activity (Ci) 130a 4,400 4 60 160 4,800 Source: Smith et al. 1978, vol. 2, Table C.4-5 Excluding volumetrically distributed activation products A.3.3 Contamination of External Surfaces of Equipment and Structural Components External surfaces of system components as well as floors, walls and structural components become contaminated over the operating lifetime of a nuclear power plant from leaks or spills of radioactive materials originating from the reactor coolant. While most liquid contamination A-28 ------- remains localized in the vicinity of the leak or spill, some contamination may experience limited transfer through physical contact. More widespread contamination of external surfaces occurs when contaminants become airborne and passively settle out. Airborne contaminants are also the principal source of contamination of ducts, fans, filters and other equipment that are part of the heating and ventilation and air conditioning systems (HVAC). Table A-19. Non-RCS Contaminated PWR Piping Nominal Pipe Size (in.) '/2 3/4 1 l'/2 2 3 4 6 8 10 12 14 Total Schedule 80 160 40 80 160 40 80 160 40 80 160 40 80 160 160 160 160 160 140 140 140 ID. (in.) 0.546 0.464 0.824 0.742 0.612 1.049 0.957 0.815 1.610 1.500 1.338 2.067 1.939 1.687 2.624 3.438 5.187 6.813 8.500 10.126 11.188 Length (m) 120 120 240 360 570 60 180 420 120 330 540 300 480 1,050 140 180 300 140 365 90 100 6,205 Mass (kg) 198 238 205 400 1,675 152 590 1,800 493 1,811 3,967 1,655 3,642 11,850 2,985 6,128 20,972 15,924 29,750 18,370 25,475 148,280 Inside Area (m2) 5.2 4.4 15.8 21.3 27.8 5.0 13.7 27.3 15.4 39.5 57.7 49.5 74.3 141.3 29.3 49.4 124.2 76.1 247.6 72.7 89.3 1186.9 Total Activity (Ci) 0.3 0.3 0.9 1.3 1.7 0.3 0.8 1.6 0.9 2.4 3.5 3.0 4.5 8.5 1.8 3.0 7.5 4.6 14.9 4.4 5.4 71.2 Radionuclides typically found in the primary coolant and their relative abundance in a PWR and BWR are given in Table A-20 and Table A-21, respectively. A-29 ------- Table A-20. Radionuclides in Primary Coolant in the Reference PWR Radionuclide Cr-51 Mn-54 Fe-55 Fe-59 Co-58 Co-60 Sr-89 Sr-90+D Zr-95 Nb-95 Te-129m 1-131 Cs-134 Cs-136 Cs-137 Total Half-Life 27.7 d 312.1 d 2.73 y 44.5 d 70.88 d 5.271 y 50.52 d 28.78 y 64.02 d 34.97 d 33.6 d 8.04 d 2.065 y 13.16d 30.07 y Relative Activity at Various Times After Shutdown* 0 6.9e-04 1.4e-03 2.2e-02 8.7e-04 7.5e-03 7.5e-02 1.2e-03 6.9e-04 2.5e-04 2.5e-04 3.1e-04 1.4e-02 1.2e-01 l.le-03 7.5e-01 1.0 lOy — 4.2e-07 1.7e-03 — — 2.0e-02 — 5.4e-04 — — — — 4.2e-03 — 6.0e-01 0.62 30 y — — l.le-05 — — 1.5e-03 — 3.4e-04 — — — — 5.1 e-06 — 3.8e-01 0.38 50 y — — 6.7e-08 — — l.Oe-04 — 2.1e-04 — — — — 6.2e-09 — 2.4e-01 0.24 Source: Smith et al. 1978, vol. 1 * Activities of individual nuclides, normalized to the total activity at shutdown The amount of external surface contamination following 40 years of operation is likely to vary significantly among nuclear power plants and is influenced by fuel integrity, primary coolant chemistry, operational factors and reactor performance. A key operational factor is the effort expended to clean up spills and to decontaminate accessible areas on an ongoing basis. Although all nuclear utilities conduct routine radiological surveys that assess fixed and removable surface contamination, only limited data have been published in the open literature from which average contamination estimates can be derived. In this section, estimates of external surface contamination are provided that reflect (1) modeled data, (2) data published in the open literature and (3) data from individual utilities that have submitted a decommissioning plan. A-30 ------- Table A-21. Radionuclide Concentrations in Reactor Coolant of Reference BWR Radionuclide P-32 Cr-51 Mn-54 Fe-55 Fe-59 Co-58 Co-60 Ni-63 Zn-65 Sr-89 Sr-90 +D Y-91 Zr-95 Ru-103 Ru-106 Ag-llOm Te-129m 1-131 Cs-134 Cs-136 Cs-137 Ba-140 +D Ce-141 Ce-144 Pr-143 Nd-147 Total Half-Life (days) 14.28 d 27.7 d 312.1 d 2.73 y 44. 5 d 70.88 d 5.271 y 100.1 y 244.26 d 50.52d 28.78y 58. 5 d 64.02 d 39.27 d 373.6 d 249.8 d 33.6 d 8.04d 2.065 y 13.16d 30.07 y 12.75 d 32.5 d 284.9 d 13.57d 10.98 d Specific Activity Gio/g) 2e-04 5e-03 6e-05 le-03 3e-05 2e-04 4e-04 le-06 2e-04 le-04 6e-06 4e-05 7e-06 2e-05 3e-06 le-06 4e-05 5e-03 3e-05 2e-05 7e-05 4e-04 3e-05 3e-06 4e-05 3e-06 1.3e-02 Relative Activity at Various Times After Shutdown* 0 1. le-03 5.3e-02 7.2e-04 3.7e-01 5.3e-04 5.6e-03 2.9e-01 3.4e-03 1.8e-02 2.0e-03 1.5e-02 8. le-04 1.6e-04 2.9e-04 3.9e-04 8.8e-06 4.9e-04 1.5e-02 8.8e-03 l.Oe-04 1.8e-01 2.0e-03 3.4e-04 2.9e-04 2.0e-04 1.2e-05 lOy — — 2.2e-07 2.9e-02 — — 7.8e-02 3.2e-03 5.7e-07 — 1.2e-02 — — — — 3.5e-10 — — 3. le-04 — 1.4e-01 — — 4.0e-08 — — 2.7e-01 30 y — — — 1.8e-04 — — 5.6e-03 2.8e-03 — — 7.3e-03 — — — — — — — 3.7e-07 — 9.0e-02 — — — — — l.le-01 50 y — — — 1. le-06 — — 4.0e-04 2.4e-03 — — 4.5e-03 — — — — — — — 4.5e-10 — 5.7e-02 — — — — — 6.4e-02 Activities of individual nuclides, normalized to the total activity at shutdown A-31 ------- A.3.3.1 Data for Reference Facilities Oak et al. (1980) have modeled the surface contamination on structures of the Reference BWR. The model was based on an assumed release rate of one liter of primary coolant per day for 40 years. Levels of deposited contaminants on external surfaces were correlated to ambient dose rates by means of the computer code ISOSHLD and divided into two discrete categories. The first category is low-level contamination, defined by dose rates of 10 mR/hr in air at 1 meter from the surface. The second category was defined as higher contamination with dose rates of 100 mR/hr in air at 1 meter from the surface. Based on the radionuclide composition of Reference BWR coolant, these two contamination levels were estimated to correspond to areal activity concentrations of 2.5 x 10"3 Ci/m2 and 2.5 x 10"2 Ci/m2, respectively. Table A-22 summarizes the distribution of external surface contaminants at shutdown. The total deposited activity on structural surfaces in the Reference BWR was estimated to be 114 curies. Table A-22. Surface Contamination Levels for Reference BWR at Shutdown Building Reactor Building Contamination Level la Contamination Level 2b Turbine Generator Bldg. Contamination Level la Contamination Level 2b Radwaste & Control Bldg. Contamination Level la Contamination Level 2b Total Surface Area (m2) 5145 2403 2742 1817 1767 50 1953 579 1374 8915 Deposited Activity (Ci) 74 5.7 68.3 4.4 3.2 1.2 35.8 1.4 34.4 114.2 Surface Contamination Level at Shutdown (dpm/100 cm2) 3.19e+08 5.27e+07 5.53e+08 5.38e+07 4.02e+07 5.33e+08 4.07e+08 5.37e+07 5.56e+08 2.84e+08 Source: Oak et al. 1980, vol. 2, Table E.2-10 a Contamination Level 1 corresponds to 2.5 x Contamination Level 2 corresponds to 2.5 * 10'3 Ci/m2. lO'2 Ci/m2. Table A-23 provides a more detailed breakdown of contamination levels by identifying major equipment/systems that are located within each of the aforementioned facility buildings. A-32 ------- Table A-23. Estimated External Structural Contamination in the Reference BWR Building/Associated Equipment/System/Structure Reactor Building Containment Atmosphere Control Condensate (Nuclear Steam) Control Rod Drive Equipment Drain (Radioactive) Floor Drain (Radioactive) Fuel Pool Cooling & Cleanup Fuel Pool Cooling & Cleanup High-Pressure Core Spray Low-Pressure Core Spray Main Steam Miscellaneous Wastes (Radioactive) Reactor Building Closed Cooling Reactor Core Isolation Cooling Reactor Water Cleanup Reactor Water Cleanup Residual Heat Removal Standby Gas Treatment Traversing Incore Probe Primary Containment Total Turbine Generator Building Air Removal Condensate (Nuclear Steam) Condenser Off Gas Treatment Equipment Drain (Radioactive) Floor Drain (Radioactive) Heater Drain Main Steam Miscellaneous Drain & Vent Reactor Feedwater Miscellaneous Wastes (Radioactive) Total Contaminated Area (m2) 1.6e+01 3.3e+01 1.8e+02 1.8e+01 7.4e+01 1.2e+03 2.8e+02 l.le+02 1.4e+01 3.0e+02 8.3e+01 1.2e+01 1.5e+01 1.5e+02 1.7e+02 1.7e+02 4.0e+01 8.0e+01 2.2e+03 3.9e+01 6.6e+02 1.8e+02 2.5e+01 2.5e+01 9.1e+01 1.7e+02 1.9e+01 6.9e+02 9.0e+00 Contamination Level 1 1 1 2 2 1 2 1 1 1 1 1 1 1 2 1 1 1 2 1 1 1 2 2 1 1 1 1 1 Deposited Activity (Ci) 4.0e-02 8.2e-02 4.5e-01 4.5e-01 1.8e+00 3.0e+00 7.0e+00 2.7e-01 3.5e-02 7.5e-01 2.1e-01 3.0e-02 3.8e-02 3.8e-01 4.2e+00 4.2e-01 l.Oe-01 2.0e-01 5.5e+01 7.4e+01 9.7e-02 1.6e-01 4.5e-01 6.2e-01 6.2e-01 2.3e-01 4.2e-01 4.7e-02 1.7e+00 2.2e-02 4.4e+00 A-33 ------- Table A-23 (continued) Building/Associated Equipment/System/Structure Radwaste and Control Building Condensate Filter Demineralizer Condenser Off Gas Treatment Equipment Drain (Radioactive) Equipment Drain (Radioactive) Floor Drain (Radioactive) Floor Drain (Radioactive) Floor Pool Cooling & Cleanup Miscellaneous Wastes (Radioactive) Miscellaneous Wastes (Radioactive) Process Waste (Radioactive) Process Waste (Radioactive) Reactor Water Cleanup Total Contaminated Area (m2) 3.6e+02 3.2e+02 4.3e+01 1.8e+02 1.2e+01 1.9e+02 5.4e+01 2.4e+01 1.9e+02 1.8e+02 2.7e+02 1.3e+02 Contamination Level 2 1 1 2 1 2 2 1 2 1 2 2 Deposited Activity (Ci) 9.0e+00 8.0e-01 l.le-01 4.5e+00 3.0e-02 4.8e+00 1.4e+00 6.0e-02 4.8e+00 4.5e-01 6.7e+00 3.2e+00 3.6e+01 Source: Oak et al. 1980 Note: Estimated total deposited radioactivity on contaminated external surfaces = 1.14 x 102 Ci Model Estimates Versus Empirical Data External surface contamination corresponding to Level 1 (2.5 x 10"3 Ci/m2 or 5.2 x 107 dpm/100 cm2) and Level 2 (2.5 x 10"2 Ci/m2 or 5.5 x 108 dpm/100 cm2) is not uncommon and has been observed in most reactor facilities. Table A-24 presents study data that focused on the most highly contaminated surfaces at six nuclear power plants (Abel et al. 1986). Contamination levels corresponding to modeled values (i.e., Level 1 and Level 2), however, were restricted to small areas that had experienced spills, leaks, or intense maintenance, such as the reactor sump area, RCS coolant pumps and radwaste system components. The study data also showed that when surfaces were coated with sealant or epoxy paint, nearly all contamination resided on or within the surficial coating and was readily removable. In summary, the modeled external surface contamination levels cited by Oak et al. (1980) for the Reference BWR appear excessive in terms of their projected surface areas and total plant inventory. The primary model parameter regarding the release of one liter of primary coolant per day that is allowed to buildup over a forty-year period of plant operation is not only without A-34 ------- technical basis but ignores the ongoing decontamination efforts that exist at all nuclear facilities. For these reasons, the modeled data cited by Oak et al. (1980) are not considered suitable for characterizing the contaminated material inventories of BWR power plants. Table A-24. External Surface Activity Concentrations at Six Nuclear Generating Stations Radionuclide Co-60 Ni-59 Ni-63 Sr-90 Tc-99 Cs-137 Eu-152 Eu-154 Eu-155 Pu-238 Pu-239, 240 Am-241 Cm-244 Areal Activity Concentration Range (pCi/cm2) 590 - 460,000 30 - 2,400 3,100-6,400 1.6-480 0.27-2.4 550- 2.0 E6 9-3,100 90- 1,500 10-500 0.025 - 48 0.089-21 0.10-30 0.013-0.026 Average (dpm/100 cm2) 2.4e+07 1.9e+05 l.Oe+06 3.7e+04 3.5e+02 8.1e+07 2.2e+05 1.5e+05 1.3e+04 3.1e+03 1.7e+03 1.9e+03 3.5e+00 N* 5 3 2 4 O 6 O 3 2 4 4 4 O Number of reactor units included in calculation A.3.3.2 Surface Contamination Levels Reported by Facilities Preparing for Decommissioning PWR By coincidence (as was previously noted), the Trojan Nuclear Plant (TNP), which was used as the Reference PWR facility by Smith et al. (1978), has been permanently shutdown and has submitted a decommissioning plan. External surface contamination inventories at this facility are summarized in TNP's decommissioning plan and have been reproduced in Table A-25. Estimates were based on historical survey data and recent structural surveys performed in support of the radiological site characterization required by the decommissioning plan. Combined radionuclide inventories in the containment building, auxiliary building, fuel building and the main steam support structure are estimated to be 30 mCi. Note that this value is about A-35 ------- three orders of magnitude lower than the estimate for the Reference BWR modeled by Oak et al. (1980), presented in Table A-23. Table A-25. Radionuclide Inventories on External Surfaces at Trojan Nuclear Plant Structure Containment Building Auxiliary Building Fuel Building Main Steam Support Structure Turbine Building Total Total Activity (mCi) 24 2 1 1 2 30 More detailed data relating to contamination of external surfaces at TNP were recently cited in a draft report issued by the NRC (1994). The survey data primarily measured removable floor contamination levels obtained by smears. However, such measurements may reasonably be assumed to also represent metal surfaces of reactor systems and structural components. A summary of removable external surface contamination levels at TNP are given in Table A-26. Table A-26. Contamination of Floor Surfaces at Trojan Nuclear Plant Prior to Decommissioning Building Containment Auxiliary (6 levels) Fuel Building (5 levels) Turbine Building Control Building Total Area (m2) 1,900 4,000 5,000 5,700* 700* Contaminated Fraction (%) 100 1 -5 1 -5 «1 «1 Area (m2) 1,900 40 - 200 50-250 ~ 0 ~ 0 Removable Surface Contamination (dpm/100 cm2) 1,100-55,000 < 1,100-7,900 < 1,100-5,000 < 1,000 < 1,000 Source: NRC 1994 * per level The auxiliary and fuel buildings also exhibited some areas of floor contamination, but not to the extent of that observed in the reactor containment building. Based on survey reports, about 1% to 5% of the floor area (representing about 40 m2 to 200 m2) in the auxiliary building has radioactive contamination levels in the range of 1,100 to 7,900 dpm/100 cm2. The fuel handling A-36 ------- building also has a small area of contaminated floor, ranging from 50 m2 to 250 m2, with contamination levels ranging of about 1,100 to 5,000 dpm per 100 cm2. Other buildings, including the turbine building and the control building, did not have measurable, removable contamination on any surfaces. It is important to note, however, that the quantitative estimates in Table A-26 reflect contamination that is removable (i.e., by wiping a 100 cm2 area with a dry filter paper). Reasonable estimates of total surficial contamination levels (i.e., fixed and removable) may be obtained by multiplying values in Column 5 of Table A-26 by a factor whose value may range from 5 to 10. BWR Values similar to those reported in the TNP's decommissioning plan have also been reported in the decommissioning plan submitted for Humboldt Bay Unit 3 (Pacific Gas and Electric 1994). Excerpts of survey measurements (as they appear in the decommissioning plan) are shown in Tables A-27 and A-28. Horizontal surfaces (i.e., floors) exhibited contamination levels that, on average, were about one order of magnitude higher than vertical surfaces (i.e., walls) with values ranging from below detection limits up to several million dpm per 100 cm2 for certain floor areas (e.g., under the reactor vessel). When relatively small areas of high contamination are excluded, average external surface contamination was generally between 5,000 and 100,000 dpm/100 cm2. From the above-cited data, it is concluded that, within the common variability of contamination levels in nuclear plants, the survey data reported in decommissioning plans for the Trojan and Humboldt Bay facilities provide a reasonable basis for estimating surface contamination levels at other PWR and BWR power plants, respectively. A.4 BASELINE METAL INVENTORIES A.4.1 Reference PWR The total amounts of metals contained in significant quantities in a typical 1,000 MWe PWR power plant have been quantified in a 1974 study of material resource use and recovery in nuclear power plants (Bryan and Dudley 1974). Material estimates were made using various methods that included: (1) amounts of raw materials purchased for construction (e.g., reinforcing steel and structural steel required for construction), (2) weights of materials contained in A-37 ------- equipment and machinery based on manufacturers' specifications and technical journals (e.g., determination of carbon steel, stainless steel, copper and other metals in electric motors); and (3) the U.S. Atomic Energy Commission facility accounting system, which identified individual items. Summary estimates of composite materials used to construct a 1971-vintage 1,000 MWe PWR power plant are given in Table A-29. Carbon steel is the predominant metal used in the construction of a nuclear power plant. It is used in piping and system components when the need for corrosion resistant stainless steel is not of significant importance. A large percentage is also used in structural components that include rebar, I-beams, plates, grates and staircases. A breakdown of material quantities used in reactor plant structures and plant systems is provided in Table A-30. Structural components comprise 16,519 t out of a total of 32,7311 of carbon steel, with the remainder used in plant equipment. Of the more than 16,000 t of carbon steel employed in plant equipment/systems, 10,958 t are contained in turbine plant equipment. Barring significant leakage in steam generators, equipment in this grouping as well as electric plant equipment, equipment identified as "miscellaneous," and "structures" are not likely to be exposed to radioactive contamination and are therefore not likely to contribute significant quantities of residually contaminated scrap metal. The primary sources of contaminated scrap metal in a PWR are underlined in Table A-30 and involve all items associated with reactor plant equipment with additional quantities contributed by "Fuel Storage," certain structural components, HVAC systems and other items that are identified in detail in Section A. 5. Table A-30 also shows that the use of corrosion resistant stainless steel is almost totally confined to reactor plant and turbine plant systems. Of the total 2,080 t of stainless steel, essentially all of the 1,154.6 t associated with reactor plant systems and the 21.11 that line the fuel pool can be assumed to be contaminated. A.4.2 Reference BWR Inventories for a 1,000-MWe BWR reference plant have been estimated by adjusting Bryan and Dudley's 1974 Reference PWR plant data taking into account the characteristics of a BWR (Oak etal. 1980). A-38 ------- Table A-27. Radiation Survey Data for Humboldt Bay Refueling Buildinga Location +12 ft elevation Access Shaft -2 ft elevation -14 ft elevation -24 ft elevation -34 ft elevation -44 ft elevation -54 ft elevation -66 ft elevation Cleanup: HX Room -2 ft elevation Cleanup: Demin Room -2 ft elevation Shutdown: HX Room -14 ft elevation West Wing -66 ft elevation Under Reactor -66 ft elevation New Fuel Vault +0 ft elevation TBDT Area -14 ft elevation floor wall floor wall floor wall floor wall floor wall floor wall floor wall floor wall floor wall floor wall floor wall floor wall floor wall floor wall floor wall Dose Rateb (mR/h) Gamma 10 7g 2g 1g 1g 7g 18 12 65 6 55 110 23 5 23 Beta <1 h 0 h h 1.5 1.1 0 0 1.5 1.1 7.5 21 47 35 Contamination Levels (uCi/100cm2) Contact0 Alpha f f f f f f f f f f f f f f f f f f f f f f f f 1 .7e-03 f 3.4e-04 f f f Beta- Gamma 3.6e-02 9.8e-03 1 .6e-02 2.1e-03 4.26-03 2.4e-03 3.16-03 1.0e-03 2.16-03 f 8.3e-02 1 .Oe-02 1.2e-01 2.16-02 1.4e-01 6.46-02 1.0e-01 4.26-02 2.1e-01 2.16-03 f 2.1e-02 f 9.66-02 2.0e+00 3.26-02 2.3e+00 f 1.66-01 3.4e+00 Smearable Alpha 3.96-06 2.2e-06 7.16-06 f 4.7e-06 2.36-06 1 .4e-05 f 1 .2e-05 f 4.5e-06 f 4.56-06 f 2.3e-06 f 2.16-05 f 1 .Oe-04 2.06-06 3.7e-06 2.86-07 1 .2e-05 5.66-07 9.0e-04 6.56-05 1 .9e-05 1.16-06 4.2e-06 1.16-06 Beta- Gamma 1.16-03 3.3e-04 1.56-03 2.7e-05 2.36-03 7.6e-04 2.46-03 f 3.0e-03 f 1.36-03 2.7e-05 1 .2e-03 f 6.1e-04 f 9.46-03 1.9e-05 4.26-02 3.5e-04 2.86-03 2.0e-05 2.76-03 f 3.3e-01 4.46-03 5.4e-03 6.36-04 9.6e-04 9.16-03 Average values of PG&E survey conducted May 1984 unless otherwise specified. Ion chamber Minimum sensitivity: alpha: 1 x 10"4 uCi/100cm2 beta: Cutie Pie 5 x 10'3 ^Ci/lOOcm2 HP-210 2 x lO'6 j^Ci/lOOcm2 d Based on Cs-137 e Based on Sr-90 (10%), Co-60 (45%) and Cs-137 (45%) Not detected s Previous survey Data not recorded A-39 ------- Table A-28. Radiation Survey Data for Humboldt Bay Power Buildinga Location Condenser/ Demineralizer Cubicle Condenser/ Demineralizer Regeneration Room Condenser/ Demineralizer Operations Area Condenser Pump Room Air Ejector Room Condenser Area Pipe Tunnel Feed Pump Room Seal Oil Room Turbine Enclosure +27 ft elevation Turbine Washdown Area +27 ft elevation Hot Lab Laundry/Demin Area +27 ft elevation Laundry/Hot Lab +34 ft elevation floor wall floor wall floor wall floor wall floor wall floor wall floor wall floor wall floor wall floor wall floor floor floor floor Dose Rateb (mR/h) Gamma 11 14 14g 13g 55 19 15 <1" 0.005g <1fl <19 <1fl <1fl h Beta 0 1.5 h h 56 <1 1.5 h h h h h h h Contamination Levels (uCi/100cm2) Contact0 Alpha f f 2.6e-04 LOe-03 f f f f f f f f f f f h f h f f f f f f Beta- Gamma 3.2e-02 3.2e-02 3.56-02 7.1e-02 3.56-03 8.8e-03 f f 5.66+00 f 6.0e-03 f 4.76-03 f 5.2e-04 h f h 3.16-03 4.2e-03 1 .Oe-03 1 .2e-02 2.66-03 1 .Oe-03 Smearable Alpha 8.56-06 f 1.1e-05 1.16-05 1 .4e-06 f 2.06-06 f 1 .7e-06 h 5.76-07 h 1.1e-06 1 .4e-07 f h f h 8.5e-07 2.86-07 1 .7e-06 f 4.36-07 f Beta- Gamma 1 .4e-03 9.7e-05 2.76-03 1 .5e-03 1 .5e-04 6.1e-05 5.06-04 2.8e-05 7.86-02 1 .5e-03 5.76-04 h 2.9e-04 2.16-05 8.4e-05 h 2.16-05 h 1 .2e-04 f 6.16-05 7.3e-05 7.76-05 2.0e-04 Average values of PG&E survey conducted May 1984 unless otherwise specified Ion chamber Minimum sensitivity: alpha: 1E-4 uCi/100cm2 beta: Cutie Pie 5E-3 uCi/100cm2 HP-210 2E-6 uCi/100cm2 d Based on Cs-137 e Based on Sr-90 (10%), Co-60 (45%) and Cs-137 (45%) Not detected g Previous survey Data not recorded A-40 ------- Table A-29. Inventory of Materials in a 1971-Vintage 1,000 MWe PWR Facility Metal Carbon Steel Rebar All Other Stainless Steel Galvanized Iron Copper Inconel Lead Bronze Aluminum Brass Nickel Silver Total Mass (t) 3.3e+04 1.3e+04 2.0e+04 2.1e+03 1.3e+03 6.9e+02 1.2e+02 46 25 18 10 1.0 < 1.0 Source: Bryan and Dudley 1974 With regard to the steel inventories, there are two significant differences between a PWR and BWR. A BWR has less heat-transfer piping and lacks a steam generator, but has more extra- vessel primary components, including a pressure suppression chamber. A second difference is the estimated quantity of rebar used for concrete reinforcement. Of the 32,700 tons of carbon steel in the Reference 1,000 MWe PWR, Bryan and Dudley estimated that about 13,300 tons is rebar; for the 1,000 MWe Reference BWR, the total mass of rebar was estimated at 18,300 tons (Oaketal. 1980). Although the amount of steel required to construct a BWR is only slightly greater than for a PWR, a greater fraction of the steel (and other metals) is contaminated. This is because primary- to-secondary leakage causes radioactive contamination of the BWR steam flow, which in turn contaminates turbine plant equipment; in a PWR, such equipment is usually uncontaminated. Table A-31 identifies material estimates for a 1,000-MWe BWR plant. Material estimates for metals other than carbon and stainless steel for the 1,000-MWe Reference BWR are assumed to be identical to those of the 1,000-MWe Reference PWR. A-41 ------- Table A-30. Breakdown of Materials Used in PWR Plant Structures and Reactor Systems (t) System Structures/Site Site Improvements Reactor Building Turbine Building Intake/Discharge Reactor Auxiliaries* Fuel Storage Miscellaneous Bldgs. Reactor Plant Equipment Reactor Eguipment Main Heat Trans. System Safeguards Cool. System Radwaste System Fuel Handling System Other Reactor Eguipment Instrumentation & Control Turbine Plant Equipment Turbine-Generator Heat Rejection Systems Condensing Systems Feed-Heating System Other Eguipment Instrumentation & Control Electric Plant Equipment Switchgear Station Service Eguip. Switchboards Protective Eguipment Structures & Enclosure Power & Control Wiring Miscellaneous Equipment Transportation & Lifting Air & Water Service Sys. Communications Eguip. Furnishings & Fixtures Entire Plant Carbon Steel 16519.3 1692.9 7264.2 3641.2 333.7 1358.7 364.6 1864 3444.9 430.0 1686.5 274.2 35.2 82.0 823.5 113.5 10958.3 4138.6 2501.1 1359.8 1367.7 1541.3 49.8 965.5 30.4 654.1 87.0 5.9 112.5 75.6 843.2 529.3 232.5 4.7 76.7 32731.2 Stainless Steel 28.6 0.0 5.7 0.0 0.0 0.0 21.1 1.8 1154.6 275.1 202.5 199.1 31.9 67.0 230.3 148.7 883.2 129.9 9.1 392.3 221.2 89.4 41.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 13.7 0.0 6.0 0.0 7.7 2080.1 Galvanized Iron 814.2 17.9 301.2 196.4 3.6 109.8 43.4 141.9 5.5 0.0 1.6 1.1 0.8 0.3 1.7 0.0 4.7 0.5 2.2 0.6 0.5 0.9 0.0 431 1.4 8.5 0.0 0.0 421.1 0.0 2 0.0 0.0 0.6 1.4 1257.4 Copper 33.1 1.5 9.3 1.6 0.2 0.8 0.3 19.4 50.4 6.8 9.8 2.9 0.2 0.2 1.5 29.0 51.4 35.2 3.0 1.3 1.2 0.7 10.0 556.5 2.8 19.0 13.5 39.0 0.0 482.2 2.6 0.5 1.1 1.0 0.0 694 Inconel 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 124.1 124.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0 0.0 0.0 0.0 0.0 124.1 Lead 33.1 0.7 0.0 0.0 0.0 0.0 0.0 32.4 4.5 0.0 0.0 0.0 0.0 0.0 4.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 6.8 0.0 6.8 0.0 0.0 0.0 0.0 2 0.0 0.0 0.0 2.0 46.4 Bronze 0.2 0.0 0.0 0.1 0.0 0.0 0.0 0.1 0.5 0.0 0.0 0.1 0.0 0.0 0.4 0.0 21.5 19.7 0.7 0.3 0.3 0.5 0.0 2.5 0.7 0.7 0.1 0.5 0.0 0.5 0.4 0.0 0.0 0.0 0.4 25.1 Aluminum 1.2 0.1 0.1 0.8 0.0 0.0 0.1 0.1 5.2 0.0 0.0 0.0 0.0 0.0 0.0 5.2 1.2 0.0 0.0 0.0 0.0 0.0 1.2 4.1 0.0 0.0 4.1 0.0 0.0 0.0 6.5 0.0 0.0 0.4 6.1 18.2 Brass 2.9 0.0 0.3 1.4 0.0 0.2 0.1 0.9 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 6.9 0.0 0.4 1.5 3.9 1.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.3 0.0 0.3 0.0 0.0 10.1 Nickel 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.6 0.0 0.0 0.0 0.0 0.0 0.6 0 0.0 0.0 0.0 0.0 0.7 Silver 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.4 0.3 0.1 0.0 0.0 0.0 0.0 0 0.0 0.0 0.0 0.0 0.5 to Source: Bryan and Dudley 1974 * Underlined text identifies equipment/systems with significant amounts of radioactive contamination. ------- Table A-31. Inventories of Ferrous Metals Used to Construct a 1,000-MWe BWR Facility Metal Carbon Steel Rebar All Other Stainless Steel Total Mass (t) 3.4e+04 1.8e+04 1.6e+04 2.1e+03 Source: Oak et al. 1980 A.5 METAL INVENTORIES WITH THE POTENTIAL FOR CLEARANCE From data presented in previous sections, two important conclusions can be stated: (1) only a fraction of metal inventories is likely to be significantly contaminated and (2) not all contaminated metal inventories are candidates for clearance. The potential for clearance is largely determined by the practicality and efficacy with which contaminated scrap can be decontaminated to an acceptable level. The choice of available decontamination methods needed to make scrap metal candidates for clearance is largely dependent on the initial level of contamination, the type of surface, physical accessibility to the surface, the radionuclides involved and their chemical states, and the size and configuration of the metal object. Several techniques are currently used in decontamination efforts at nuclear facilities. Their applicability, however, is not without restrictions and for nearly all approaches, there are numerous factors that affect their efficacy. Examples include the choice of cleaner/solvent/ surfactant for hand wiping, the selection of chemical solvents for the dissolution and removal of radioactive corrosion films or base metal, or the innovative use of dry-ice (CO2) pellets for abrasive blasting. These techniques and their general applicability and limitations are briefly summarized below. Hand Wiping Rags moistened with water or a solvent such as acetone can be an effective decontamination process. Wiping can be used extensively and effectively on smaller items with low-to-medium external contamination levels and easily accessible internal contamination. This method may not work well if the item is rusty or pitted. It requires access to all surfaces to be cleaned, is a relatively slow procedure, and its hands-on nature can lead to high personnel exposure. On the A-43 ------- positive side, wiping can provide a high decontamination factor (DF), generates easily handled decontamination wastes (contaminated rags), requires no special equipment, and can be used selectively on portions of the component. Steam Cleaning This may be performed either remotely in a spray booth or directly by decontamination personnel using some type of hand-held wand arrangement. In the former case, only minimal internal decontamination is possible; however, reasonable external cleaning can be accomplished quickly while minimizing external exposure to direct radiation. Containment of the generated wastes and protection of personnel from radioactive contamination may be difficult, however. Abrasive Blasting This is a highly effective procedure even for surfaces that are rusty or pitted. As with hand-held steam cleaning, this method suffers from internal accessibility problems. It also generates large amounts of solid wastes and, being a dry process, produces significant quantities of airborne radioactive dust. Abrasive blasting may be used if its high effectiveness can be justified after taking into account the radiation exposures, generation of radioactive waste and limited accessibility to internal surfaces. Some of the aforementioned disadvantages are obviated when dry ice pellets are used. Hydrolasing The use of high pressure water jets for decontamination falls somewhere between steam cleaning and abrasive blasting in effectiveness. Less effective than abrasive blasting, it has the advantage of producing liquid wastes (that can be processed) rather than solid wastes. As an external cleaning technique, it has the advantage of reducing the generation airborne radioactive dusts, although this is offset by the potential of spreading contamination by splashing. The use of hydrolasing is generally limited to cases where access to internal surfaces is not required. Ultrasonic Cleaning Since this is an immersion process that is limited to smaller items, it is generally unsuitable for large-scale decontamination. Although ultrasonic cleaning can be especially effective in removing contamination from crevices, it is doubtful that clearance levels can be reached consistently with this technique to make it a viable option. A-44 Continue ------- Back Electropolishing This is an electrochemical process where the object to be decontaminated serves as the anode in an electrolytic cell and radioactive contamination on the item is removed by anodic dissolution of the surface material. Although it is a relatively new process and has not yet been used for a full scale decontamination operation, it nevertheless requires consideration as a technique on the basis of its superior effectiveness in cleaning almost any metallic surface to a completely contamination-free state. On the other hand, this process has several limitations including the size of contaminated objects, the cost of the electrolytes and special equipment, the consumption of considerable power and the production of highly radioactive solutions. Chemical Decontamination Chemical flushing is recommended for remote decontamination of intact piping systems and their components. This technique uses concentrated or dilute solvents in contact with the contaminated item to dissolve either the contamination film covering the base metal or the base metal itself. Dissolution of the film is intended to be nondestructive to the base metal and is generally used for operating facilities. Dissolution of the base metal, however, can be considered in a decommissioning program where reuse of the item will not occur. Based on starting levels of contamination and required decontamination efforts, scrap metal inventories at nuclear power plants can be grouped into four categories. A description of each of these categories appears below, followed by a list of examples of major components that, under normal operating conditions, are most likely to be grouped in that category. 1. Low-level, surface-contaminated. This category is likely to comprise components that may be removed from buildings with significant residual radionuclide inventories but involve systems that are completely isolated from primary coolant, coolant waste streams and other media with substantial levels of radioactivity. A sizeable fraction of scrap metal within this category will exhibit contamination that is limited to external surfaces and not exceed 10s dpm/100 cm2. Decontamination strategies are most likely to be routine with 100% success at achieving foreseeable clearance levels. a. Structural metals in the turbine building, auxiliary building and support buildings b. Control and instrumentation cables, cable trays c. Mechanical systems/piping not associated with primary coolant and radwastes 2. Medium-level, surface-contaminated. Metal components in direct contact with media that are less contaminated than the primary coolant and liquid radwastes may have A-45 ------- internal and/or external surface contamination levels between 10s and 107 dpm/100 cm2. Scrap metal in this category requires substantial decontamination efforts with less than 100% success in achieving unrestricted release. Examples include: a. Containment spray recirculation systems b. Most auxiliary support systems c. BWR steam lines d. BWR turbines e. BWR condenser f Containment building crane, refueling equipment, etc. g. Reactor building structural steel h. Fuel storage pool liner and water cleanup system 3. High-level surface-contaminated. Scrap metal in this category will be represented by systems internally exposed to and contaminated by primary coolant and liquid radwastes leading to contamination levels in excess 107 dpm/100 cm2. Variable fractions of such metals are likely to be decontaminated to a level that permits unrestricted release. a. PWR primary recirculation piping b. PWR primary pumps and valves c. Liquid radwaste systems/tanks d. PWR steam generators e. Primary coolant cleanup system f. PWR pressurizer g. Coolant letdown and cleanup h. Spent fuel pool cooling 4. Volumetrically contaminated. Components proximal to the reactor core may contain volumetrically distributed activation products that range from nominal to extremely high levels. (Some of these components may also have high levels of surface contamination.) Removal of volumetrically distributed contaminants by standard processes is not achievable. a. Reactor vessel b. Reactor vessel top head A-46 ------- c. Reactor vessel internals d. Control rod drive lines e. Reactor building components proximal to pressure vessel (< 10%) f. Rebar (~ 1 % of plant total) A.5.1 Contaminated Steel Components with the Potential for Clearance The steel components and systems of the Reference BWR and PWR which are candidates for clearance are described in the following sections. (As discussed above, metals with significant levels of volumetrically distributed activation products would not be considered for clearance.) These tables in each of these sections list the system components and their corresponding masses. The materials composing the individual components have not been adequately defined. While a considerable number of components could be identified to consist exclusively of carbon steel or stainless steel, large quantities of steel exist as thick-walled carbon steel that is clad with thin- walled stainless steel (e.g., large piping, valves, vessels, tanks). When stainless steel provides corrosion resistant cladding, it is in effect physically inseparable from its large carbon steel component. In other instances, a given item will consist of many independent parts, each having a different composition. For example, a recirculation pump may have a carbon steel casing and base with stainless steel shaft, impellers and other internals. Although potentially separable, segregation of such individual components is labor intensive and may be precluded by considerations of worker exposures (and ALARA) and/or economic factors. A prudent approach may, therefore, be to assume that all steel scrap containing nickel be categorized as "stainless steel" (even if the nickel content is well below that of standard stainless steel alloys) because it is easier to upgrade scrap by adding nickel and other alloying material than it is to remove nickel for the production of mild steel or carbon steel. A.5.1.1 Reference BWR For the Reference BWR, a total of 29 contaminated systems are identified that are grouped by location (i.e., reactor building, radwaste building and turbine building). The systems in each building are listed in alphabetical order in Tables A-32 to A-60, together with the system-average level of contamination as previously defined on page A-45. Piping inventories for the Reference BWR have been quantified and segregated by plant location in Tables A-61 to A-64. A-47 ------- In total, it is estimated that about 8,4001 of contaminated steel from the Reference BWR are candidates for clearance. Based on material composition data cited by Oak et al. (1980), it is further estimated that of this total, stainless steel comprises nearly 1,7001. Stainless steel that is physically bound to carbon steel, however, may not be readily segregated. Reference BWR Steel Inventories in the Reactor Building Table A-32. Containment Instrument Air System Number 22 1 1 222 Total Component Instrument Air Accumulators 6" Check Valve 6" Valve Valves (3/4 - 2" dia.) Mass (kg) Each 129 68 82 NA Total 2,838 68 82 4,008 6,996 Note: System average contamination level = low Table A-33. Control Rod Drive System Number 460 225 185 370 210 2 2 2 2 2,660 Total Component CRD Blade CRD Mechanism Direction Control Set Scram Valve Scram Accumulator CRD Pump Scram Discharge Volume Pump Suction Filter CRD Drive Water Filter Valves (% - 4" dia.) & Components Mass (kg) Each 182 218 36 32 64 1,816 908 182 45 NA Total 83,720 49,050 6,660 11,840 13,440 3,632 1,816 364 90 48,830 219,442 Note: System average contamination level = 80% low; 20% medium A-48 ------- Reference BWR Steel Inventories in the Reactor Building (continued) Table A-34. Equipment Drain Processing System Number 1 1 1 1 1 1 1 1 1 2 2 199 Total Component Waste Demineralizer Waste Collector Filter Waste Filter Hold Pump Waste Collector Tank & Educator Waste Collector Pump Spent Resin Tank Spent Resin Pump Waste Surge Tank & Educator Waste Surge Pump Waste Sample Tank & Educator Waste Sample Pump Valves (1 - 8" dia.) Mass (kg) Each 907 1,812 318 10,229 284 657 102 18,282 284 6,960 230 NA Total 907 1,812 318 10,229 284 657 102 18,282 284 13,920 462 5,374 52,631 Note: System average contamination level = medium A-49 ------- Reference BWR Steel Inventories in the Reactor Building (continued) Table A-35. Fuel Pool Cooling and Cleanup System Number 15 1 2 2 2 2 1 2 1 1 2 1 1 165 Total Component Spent Fuel Racks Fuel Pool Liner FPCC Pumps FPCC Demineralizer Skimmer Surge Tank FPCC Heat Exchanger Supp. Pool Cleanup Pump Resin Tank Agitator Fuel Pool Precoat Pump (Precoat) Dust Evacuator FPCC Hold Pump FPCC Precoat Tank FPCC Resin Tank Valves (1 - 10" dia.) & Components Mass (kg) Each 18,424 32,000 527 1,566 5,354 2,038 527 36 284 104 195 227 227 NA Total 276,360 32,000 1,054 3,132 10,708 4,076 527 72 284 104 390 227 227 8,038 337,199 Note: System average contamination level = high Table A-36. High Pressure Core Spray System Number 2 1 1 61 Total Component 24" Suction Strainer 12 X 24" Pump 1 X 2" Pump Valves (24 - 3/4" dia.) Mass (kg) Each 172 27,410 82 NA Total 344 27,410 82 18,459 46,295 Note: System average contamination level = medium A-50 ------- Reference BWR Steel Inventories in the Reactor Building (continued) Table A-37. HVAC Components System Number 7 5 17 NA Total Component Containment Recirculation Fans Containment Fan Coil Units Emergency Fan Foil Units Ducts (750 linear meters) Mass (kg) Each 636 1,500 955 NA Total 4,452 7,500 16,235 29,975 58,162 Note: System average contamination level = low Table A-38. Low Pressure Core Spray System Number 2 1 1 1 1 45 Total Component 24" Suction Strainer Vent Strainer 14 X 24" Pump Pump Pit 1 X 2" Pump Valves (3/4 - 24" dia.) Mass (kg) Each 172 43 9,625 182 82 NA Total 344 43 9,625 182 82 10,523 20,799 Note: System average contamination level = medium A-51 ------- Reference BWR Steel Inventories in the Reactor Building (continued) Table A-39. Main Steam System Number 1 2 2 2 1 2 1 1 2 2 4 2 2 4 18 36 18 1 6 6 8 1 4 2 1 2 2 2 2 8 951 Total Component HP Turbine LP Turbine RFW Turbine Steam Chest Gland Steam Condenser Ejector Condenser Moisture Separator Bypass Valve Assy. Moisture Separator Reheater Steam Evaporator 2" Strainer 4" Strainer 12 Stop Check 30"FlowRestrictor 8" AO SRV 10" Vacuum Breakers 24 x 12" Quenchers 72" MOV Stop Valves Interceptor Valves 30" MSIV 24" MOV 24" Relief Valve 20" Relief Valve 16" MOV 16" Check Valve 14" Check Valve 14" MOV 12" MOV 28" HOV Governor Valves Valves (1 - 10" dia.) Mass (kg) Each 194,169 371,130 18,160 55,565 1,861 1,816 908 5,266 208,386 13,472 43 100 894 1,362 921 408 749 51,900 18,160 4,540 636 3,223 4,190 3,496 1,920 1,534 1,008 1,253 1,135 3,632 NA Total 194,169 742,260 36,320 111,130 1,816 3,632 908 5,266 416,772 26,944 172 200 1,788 5,448 16,578 14,724 13,482 51,900 108,960 27,240 5,088 3,223 16,760 6,992 1,920 3,068 2,016 2,506 2,270 29,056 69,592 1,922,200 Note: System average contamination level = 60% medium; 40% low A-52 ------- Reference BWR Steel Inventories in the Reactor Building (continued) Table A-40. Main Steam Leakage Control System Number 8 28 2 14 4 4 20 2 2 4 Total Component 1/2" Valve 3/4" Valve 1 " Flow Element 1" Valve 1" Check Valve l-1/^" Flow Element I-1// MOV I-1// Check Valve MSLCFan(3") MSLC Heater Mass (kg) Each 11 14 17 23 17 21 23 21 204 57 Total 88 392 34 322 68 84 460 42 408 227 2,125 Note: System average contamination level = low A-53 ------- Reference BWR Steel Inventories in the Reactor Building (continued) Table A-41. Miscellaneous Items from Partial System Number 5 2 5 5 5 1 set 2 1 1 2 1 1 1 1 1 1 2 1 185 1 1 20 9 4 1 1 Total Component TIP Drive Unit TIP Indexing Unit TIP Ball Valve Explosive Shear Valve TIP Shield Pig TIP Tubing Hogger (mechanical vacuum pump) Refueling Bridge Reactor Service Platform Refueling Mast CRD Removal Turntable CRD Removal Trolley Incore Instrument Grapple Fuel Support Piece Grapple Control Blade Grapple Spent Fuel Pool Work Table Fuel Prep Machine Channel Measurement Machine Blade Guide In Core Instrument Strongback Manipulators, crows feet, etc. In-vessel Manipulator Poles Drywell Recirculation Fan Stud Tensioner RPV Head Strongback Dryer/Separator Strongback Mass (kg) Each 361 9 23 23 154 295 3,171 24,918 5,210 295 2,492 173 36 41 59 445 381 422 73 100 136 14 254 1,044 2,134 60 Total 1,805 72 115 115 770 295 6,342 24,918 5,210 590 2,492 173 36 41 59 445 762 422 13,505 100 136 280 2,286 4,176 2,134 60 67,339 Note: System average contamination level = 55% low; 45% medium A-54 ------- Reference BWR Steel Inventories in the Reactor Building (continued) Table A-42. Reactor Building, Closed Cooling Water System Number 3 2 1 5 1 3 7 6 4 1 218 Total Component RBCCW Heat Exchanger RBCCW Pump RBCCW Surge Tank Drywell Cooler & Fans 14" MOV 12" Valve 10" MOV 10" Valve 10" Check Valve 10" Flow Element Valves (3/4 - 8" dia.) Mass (kg) Each 7,460 1,597 531 745 449 331 250 250 168 16 NA Total 22,380 3,194 531 3,725 449 993 1,750 1,500 672 672 6455 42,321 Note: System average contamination level = low Table A-43. Reactor Building Equipment and Floor Drains System Number 4 O 1 1 97 Total Component Drain Sump Pump Drain Sump Pump Equipment Drain Heat Exchanger Drywell Equipment Drain HX Valves (3/4 - 6" dia.) Mass (kg) Each 523 650 680 680 NA Total 2,908 1,950 680 680 3,725 9,943 Note: System average contamination level = medium A-55 ------- Reference BWR Steel Inventories in the Reactor Building (continued) Table A-44. Reactor Core Isolation Cooling System Number 1 1 1 1 1 1 1 2 4 1 1 284 Total Component Pelton Wheel Turbine/Pump Barometric Condenser Condenser Pump Water Leg Pump Vacuum Pump Vacuum Tank Steam Condensate Drip Pot 8" Suction Strainers 3/4" Steam Trap 10" Exhaust Drip Chamber Turbine Exhaust Sparer Valves (3/4 - 10" dia.) Mass (kg) Each 6,260 553 679 216 453 407 109 66 25 309 241 NA Total 6,260 553 679 216 453 407 109 112 100 309 241 12,115 21,554 Note: System average contamination level = medium Table A-45. Reactor Coolant Cleanup System Number 2 2 1 1 1 2 O 2 1 2 1 259 Total Component RWCU Pump Clean Up Hold Pump Clean Up Precoat Pump Sludge Discharge Pump Decant Pump Non-regenerative HX Regenerative HX Filter Demineralizer Batch Tank Phase Separator Tank Precoat Agitator Valves 0/2 - 6" dia.) Mass (kg) Each 590 534 454 284 102 4,086 4,131 3,178 227 2,043 27 NA Total 1,180 1,068 454 284 102 8,172 12,394 6,356 227 4,086 27 13,170 47,520 Note: System average contamination level = high A-56 ------- Reference BWR Steel Inventories in the Reactor Building (continued) Table A-46. Residual Heat Removal System Number 3 1 1 1 1 6 2 3 2 1 11 8 5 3 2 4 4 2 3 2 3 3 3 1 2 1 324 Total Component RHRPump Water Leg Pump Drywell Upper Spray Ring Header Drywell Lower Spray Ring Header Wetwell Spray Ring Header Suppression Pool Suction Strainers RHR Heat Exchanger 24" MOV 20" MOV 20" Valve 18" MOV 18" Valve 18" Check 18" Flow Element 18" Restricting Orifice 16" MOV 14" MOV 14" Valve 14" Air Operated Check 14" Restricting Orifice 12" MOV 12" Valve 12" Air Operated Check 12" Restricting Orifice 10" Valve 10" Check Valve Valves (3/4- 3" dia.) Mass (kg) Each 7,792 397 8,562 13,063 5,347 195 29,190 7,150 4,086 4,086 4,603 4,603 2,762 2,762 2,762 2,724 1,544 1,544 971 944 1,017 1,017 581 549 731 399 NA Total 23,376 397 8,562 13,063 5,347 1,171 58,380 21,450 8,172 4,086 50,633 36,828 13,810 8,286 5,524 10,896 6,176 3,088 2,913 1,888 3,051 3,051 1,743 549 1,462 399 12,100 306,401 Note: System average contamination level = low A-57 ------- Reference BWR Steel Inventories in the Reactor Building (continued) Table A-47. Miscellaneous Drains System Number 1 1 174 Total Component Misc. Drain Tank #1 Misc. Drain Tank #2 with Pumps Valves (!"- 6" dia.) Mass (kg) Each 487 654 NA Total 487 654 6,509 7,650 Note: System average contamination level = medium A-58 ------- Reference BWR Steel Inventories in the Radwaste Building Table A-48. Chemical Waste Processing System Number 2 2 2 2 2 1 2 2 2 2 2 1 2 2 2 2 2 1 2 2 2 293 Total Component Chemical Waste Tank Detergent Drain Tank Detergent Drain Pump Concentrator Feed Pump Chemical Waste Pump Detergent Drain Filter Chemical Addition Pump Tank Agitators Chemical Addition Pump Distillate Tank Distillate Tank Pump Distillate Polishing Demineralizer Decon Solution Concentrator Decon Sol. Concentrator Tank Decon Cone. Recycle Pump Decon Concentrator Condenser Decon Concentrator Pre Heater Decon Concentrator Waste Pump Chemical Waste Stream Mixer Condensate Receiver Tank Condensate Receiver Tank Pump Valves (!"- 8" dia.) Mass (kg) Each 5,024 1,834 175 254 478 1,133 257 36 175 5,024 230 454 3,405 711 843 2,305 3,143 254 111 950 102 NA Total 10,048 3,668 350 508 956 1,133 454 72 350 10,048 460 454 6,810 1,422 1,686 4,610 6,286 508 222 1,900 204 7,654 59,803 Note: System average contamination level = medium A-59 ------- Reference BWR Steel Inventories in the Radwaste Building (continued) Table A-49. Condensate Demineralizers System Number 6 6 6 1 1 1 1 2 363 Total Component Filter Demineralizers Resin Trap (with Basket) Demin Hold Pump Condensate Backwash Receiving Tank Sludge Disc Mixing Pump Condensate Decant Pump Condensate Backwash Transfer Pump Condensate Phase Separator Tank Valves & Components (1 - 36") Mass (kg) Each 5,300 953 159 6,912 420 420 420 3,178 NA Total 31,800 5,718 954 6,912 420 420 420 6,356 36,783 89,783 Note: System average contamination level = medium Table A-50. HVAC Components System Number 11 O NA Total Component Radwaste Air Handlers Filter Units and Fans Ducts (1,980 linear meters) Mass (kg) Each 1,327 11,123 NA Total 14,597 33,369 54,785 102,751 Note: System average contamination level = low A-60 ------- Reference BWR Steel Inventories in the Radwaste Building (continued) Table A-51. Radioactive Floor Drain Processing System Number 1 1 1 1 1 1 1 1 1 1 1 171 Total Component Floor Drain Demineralizer Floor Drain Sample Tank Floor Drain Sample Pump Floor Drain Filter Aid Pump Floor Drain Filter Hold Pump Floor Drain Filter Floor Drain Collector Pump Floor Drain Collector Tank Waste Decant Pump Waste Sludge Discharge Mixing Pump Waste Sludge Phase Sep. Tank Valves 0/2 - 8" dia.) Mass (kg) Each 907 6,960 230 118 317 1,812 284 10,229 102 288 5,490 NA Total 907 6,960 230 118 317 1,812 284 10,229 102 288 5,490 4,500 31,237 Note: System average contamination level = medium Table A-52. Rad Waste Building Drains System Number 1 2 3 38 Total Component Chemical Drain Sump Pump EDR Sump Pump FDR Sump Pump Valves & Components (% - 3" dia.) Mass (kg) Each 666 585 483 NA Total 666 1,170 1,449 612 3,897 Note: System average contamination level = high A-61 ------- Reference BWR Steel Inventories in the Radwaste Building (continued) Table A-53. Standby Gas Treatment System Number 42 2 14 2 2 8 4 Total Component 2" Check Valve 18" Valves 18" Damper, MOV 1 8" Damper, AOV SGT Filter Unit 3/4" Valve Blower Mass (kg) Each 25 2,225 563 563 8,898 14 2,043 Total 1,050 4,450 7,882 1,126 17,796 112 8,172 40,588 Note: System average contamination level = medium Reference BWR Steel Inventories in the Turbine Building Table A-54. Feed and Condensate System Number 2 3 3 1 2 1 2 2 3 3 3 3 2 2 2 407 Total Component Turbine and Feed Pump Condensate Booster Pump Condensate Pump Gland Exhaust Condenser Air Ejector Condenser & Ejectors Off Gas Condenser #6 Feedwater Heater #5 Feedwater Heater #4 Feedwater Heater #3 Feedwater Heater #2 Feedwater Heater #1 Feedwater Heater Condensate Storage Tanks Seal Steam Evaporator Seal Steam Evaporator Slowdown Cooler Valves 0/2 - 24" dia.) Mass (kg) Each 54,821 12,006 21,883 4,032 6,614 897 73,394 68,863 35,338 50,288 51,194 62,974 50,475 13,451 213 NA Total 109,642 36,018 65,649 4,032 13,228 897 146,788 137,726 106,014 150,864 153,582 188,922 100,950 26,902 426 350,478 1,592,118 Note: System average contamination level = medium A-62 ------- Reference BWR Steel Inventories in the Turbine Building (continued) Table A-55. Extraction Steam System Number 6 6 10 10 5 5 2 2 6 4 4 10 12 85 Total Component 24" MOV 24" Stop Check 20" MOV 20" Stop Check 18" MOV 18" Stop Check 16" MOV 16" Stop Check 8" AOV 6" MOV 4" AOV 2" AOV 2" Restricting Orifice Inst. Root (typ. 3/4" globe) Mass (kg) Each 3,223 2,583 2,633 2,107 2,225 1,780 1,920 1,536 511 267 122 34 25 15 Total 19,338 15,498 26,330 21,070 11,125 8,900 3,840 3,072 3,066 1,068 488 340 300 1,275 115,710 Note: System average contamination level = medium Table A-56. Heater Vents and Drains System Number 2 2 2 4 4 841 Total Component Steam Evaporator Drain Tank Heater Drain Tank Moisture Separator Drain Tank Reheater Drain Tank Reheater Drain Tank Valves & Components (l-l/2 - 20" dia.) Mass (kg) Each 898 6,274 1,715 1,134 6,274 NA Total 1,796 12,548 3,430 4,536 25,096 151,369 198,775 Note: System average contamination level = medium A-63 ------- Reference BWR Steel Inventories in the Turbine Building (continued) Table A-57. HVAC Components System Number 4 1 10 NA Total Component Exhaust Air Units Standby Gas Treatment Air Handlers & Filter Units Ducts (1,000 linear meters) Mass (kg) Each 4,900 8,853 829 NA Total 19,600 8,853 8,290 48,503 85,246 Note: System average contamination level = low Table A-58. Offgas (Augmented) System Number 2 2 1 1 2 2 2 8 2 2 2 4 2 2 2 9 18 175 Total Component Catalytic Recombiner Vessel Preheater Heat Exchanger Offgas Condenser Water Separator Lab Vacuum Pump Lab Vacuum Pump Water Separator Charcoal Ads. Vessel Cooler Condenser Pre-filter Vessel After-filter Vessel Desiccant Dryers Dryer Heater Dryer Chiller Regenenerator Blower 6" Air Operated Valve 6" Valve Valves (3/4 - 4" dia.) Mass (kg) Each 453 538 897 271 45 45 1,359 4,077 906 1,133 1,133 622 3,625 2,265 636 82 82 NA Total 906 1,076 897 271 90 90 1,718 32,615 1,812 2,266 2,266 2,488 7,250 4,530 1,272 738 1,476 2,722 64,483 Note: System average contamination level = medium A-64 ------- Reference BWR Steel Inventories in the Turbine Building (continued) Table A-59. Recirculation System Number 2 2 4 258 Total Component Recirculation Pump with Motor 24" HOV 24" MOV Valves (3/4 - 2" dia.) Mass (kg) Each 43,617 4,767 4,767 NA Total 87,234 9,534 19,068 4,700 120,536 Note: System average contamination level = low Table A-60. Turbine Building Drains System Number 4 4 25 Total Component Equipment Drain Sump Pump Floor Drain Sump Pump Small Valves (2 -3" dia.) Mass (kg) Each 586 484 NA Total 2,344 1,936 450 4,730 Note: System average contamination level = medium A-65 Continue ------- Back Reference BWR Piping Inventories Table A-61. Reactor Building Piping Material Outside Diameter (mm) <60 73 - 254 305 - 406 457-610 660 - 762 914- 1,829 Total Carbon Steel Length (m) Mass (kg) 2,323 8,479 3,922 110,368 505 61,897 952 127,160 55 14,850 — — 322,754 Stainless Steel Length (m) Mass (kg) Total Mass (kg) 6,169 18,674 500 4,551 54 2,143 — — — — — — 25,368 348,122 Oi Note: average contamination level: medium Table A-62. Primary Containment Piping Material Outside Diameter (mm) <60 73 - 254 305 - 406 457-610 660 - 762 914- 1,829 Total Carbon Steel Length (m) Mass (kg) 263 1,366 1,084 63,181 211 29,760 1,239 554,877 374 145,312 559 234,882 1,029,378 Stainless Steel Length (m) Mass (kg) Total Mass (kg) 3,850 10,603 110 3,411 64 8,789 55 21,440 — — — — 44,243 1,073,621 Note: average contamination level: high ------- Reference BWR Piping Inventories (continued) Table A-63. Turbine Building Piping Material Outside Diameter (mm) <60 73 - 254 305 - 406 457-610 660 - 762 914- 1,829 Total Carbon Steel Length (m) Mass (kg) 3,336 14,153 2,632 115,525 1,647 176,600 1,832 386,321 465 240,698 559 234,882 1,168,179 Stainless Steel Length (m) Mass (kg) Total Mass (kg) — — 38 1,474 103 6,421 — — — — — — 7,895 1,176,074 Note: average contamination level: low Table A-64. Radwaste and Control Buildings Piping Material Outside Diameter (mm) <60 73 - 254 305 - 406 457-610 660 - 762 914- 1,829 Total Carbon Steel Length (m) Mass (kg) 3,087 10,267 3,337 75,778 338 29,221 12 4,584 — — 99 29,410 149,260 Stainless Steel Length (m) Mass (kg) Total Mass (kg) 1,150 4,747 1,026 10,164 55 1,756 — — — — — — 16,667 165,927 Note: average contamination level: high ------- A.5.1.2 Reference PWR Tables A-65 to A-79 list major contaminated PWR components by function and location. The total inventory of contaminated steel (excluding the reactor pressure vessel and its internals) is estimated at about 4,100 t. It should be pointed out, however, that about 2,000 t comprise primary system components that include steam generators, pressurizer, reactor coolant piping, etc. (see Table A-66). The long-term buildup of activated corrosion products and fission products on internal surfaces among these components is projected to be high. Even with intense and aggressive decontamination efforts, the free release of these components may not be technically achievable. The balance of about 2,1001 includes 11 internally contaminated reactor support systems and piping that are associated with the Auxiliary Building/Fuel Storage facility and a variety of structural components where contamination is limited to external surfaces. It is estimated that nearly 20% of all of this metal is stainless steel. Reference PWR Steel Inventories in the Reactor Building Table A-65. External Surface Structures Equipment System Component Refueling Cavity Liner Base Liner Reactor Cavity Liner Floor and Cavity Liner Plates CRD Missile Shield Stairways/Gratings Miscellaneous Equipment Total Mass (kg) 17,000 54,000 14,500 139,000 11,000 45,000 13,600 294,100 Note: System average contamination level = 70% low; 30% medium A-68 ------- Reference PWR Steel Inventories in the Reactor Building (continued) Table A-66. Internally Contaminated Primary System Components System Number 4 4 1 NA 1 4 1 2 1 1 Component Steam Generator Rx Coolant Pumps Pressurizer Containment Spray Piping Pressurizer Relief Tank Safety Injection System Accumulator Reactor Cavity Drain Pump Containment Sump Pump Excess Letdown Heat Exchanger Regenerative Heat Exchanger Mass (kg) Each 312,000 85,350 88,530 12,338 34,700 363 635 726 2,994 Total 1,248,000 341,400 88,530 90,800 12,338 138,800 363 1,270 726 2,994 Reactor Coolant Piping Size (mm) Total 686 - 787 ID 51 -356OD Length (m) 81 677 100,698 11,793 2,037,712 Note: System average contamination level = high Reference PWR Steel Inventories in the Auxiliary and Fuel Storage Buildings Table A-67. Component Cooling Water System Number 2 2 1 1 9 169 Total Component CCW Heat Exchanger CCW Pump CCW Surge Tank Chem. Addition Tank Sample Heat Exchanger Valves (3/4 - 24" dia.) Mass (kg) Each 31,780 6,810 908 477 3,178 Total 63,560 13,620 1,816 954 28,602 104,700 213,252 Note: System average contamination level = low A-69 ------- Reference PWR Steel Inventories in the Auxiliary and Fuel Storage Buildings (continued) Table A-68. Containment Spray System Number 2 2 1 6 6 46 Total Component Pump Pump Tank Small Electrical Equipment Large Electrical Equipment Valves (3/4-l 8" dia.) Mass (kg) Each 3,087 45 2,490 34 68 NA Total 6,174 90 2,490 204 408 37,875 47,241 Note: System average contamination level = medium Table A-69. Clean Radioactive Waste Treatment System Number 1 2 1 1 2 2 2 2 1 2 1 2 1 1 1 1 83 Total Component Rx Coolant Drain Tank Rx Coolant Drain Pump Rx Coolant Drain Filter Spent Resin Storage Tank Clean Waste Receiving Tank Clean Waste Receiving Pump Treated Waste Monitor Tank Treated Waste Monitor Pump Aux. Building Drain Tank Aux. Building Drain Pump Chem. Waste Drain Tank Chem. Waste Drain Pump Waste Cone. Hold Tank Waste Cone. Hold Pump Clean Waste Filter Clean Radwaste Evaporator Valves (2 -3" dia.) Mass (kg) Each 758 227 159 3,087 4,975 227 5,085 104 949 590 2,452 91 949 104 30 18,160 NA Total 758 454 159 3,087 9,950 454 10,170 208 949 1,180 2,452 182 949 104 30 18,160 3,935 53,181 Note: System average contamination level = medium A-70 ------- Reference PWR Steel Inventories in the Auxiliary and Fuel Storage Buildings (continued) Table A-70. Control Rod Drive System Number 4 4 1 Total Component Small Electric Equipment Large Electric Equipment Large Mech. Equipment Mass (kg) Each 34 68 68 Total 136 272 68 476 Note: System average contamination level = low Table A-71. Electrical Components and Annunciators System Number 2 2 1 1 1 7 7 1 12 2 22 Total Component 125 VDC Power (Small) 125 VDC Power (Medium) 125 VDC Power (Large) 4. 16 kV AC & Aux. (Small) 4. 16 kV AC & Aux. (Large) 480 kV AC Ld Cntr (Small) 480 kV AC Ld Cntr (Large) 480 kV AC MCC 480 kV AC MCC Annunciators (Elec. Port.) Annunciators (Mech. Port.) Mass (kg) Each 68 227 2,270 227 9,080 227 908 227 9,080 34 34 Total 136 454 2,270 227 9,080 1,589 6,356 227 108,960 68 748 130,115 Note: System average contamination level = low A-71 ------- Reference PWR Steel Inventories in the Auxiliary and Fuel Storage Buildings (continued) Table A-72. Chemical and Volume Control System Number 3 1 1 1 2 1 3 2 2 1 1 1 2 1 2 1 2 1 3 2 2 1 1 2 2 1 1 378 Total Component Regenerative Heat Exchanger Seal Water Heat Exchanger Letdown Heat Exchanger Excess Letdown Heat Exchanger Centrifugal Charge Pump Volume Control Tank Holdup Tank Monitor Tank Boric Acid Tank Batch Tank Resin Fill Tank Reciprocal Charge Pump Boric Acid Pump Reactor Coolant Filter Mixed Bed Demineralizer Cation Ion Exchange Seal Injection Filter Concentrate Hold Tank Evaporator Feed Ion Exchange Evaporator Condensate Ion Exchange Condensate Filter Concentrates Filter Cone. Hold Tank Transfer Pump Gas Stripper Feed Pump Boric Acid Evaporator Skid Assembly Ion Exchange Filter Recirculation Pump Valves (3/4 - 6" dia.) Mass (kg) Each 2,724 772 863 726 7,759 2,202 13,620 9,080 9,080 658 118 8,036 281 91 477 477 749 1,589 477 477 18 18 91 227 9,489 68 288 NA Total 8,172 111 863 726 15,518 2,202 40,860 18,160 18,160 658 118 8,036 562 91 954 477 1,498 1,589 1,431 954 18 18 182 454 18,978 68 288 17,481 159,288 Note: System average contamination level = high A-72 ------- Reference PWR Steel Inventories in the Auxiliary and Fuel Storage Buildings (continued) Table A-73. Dirty Radioactive Waste Treatment System Number 1 2 1 2 1 2 2 46 Total Component Rx Cavity Drain Pump Rx Cont. Sump Pump Dirty Waste Monitor Tank Dirty Waste Monitor Tank Pump Dirty Waste Drain Tank Dirty Waste Drain Tank Pump Aux Building Sump Pump Valves (2 -3" dia.) Mass (kg) Each 363 681 2,633 91 2,969 181 590 NA Total 363 1,362 2,633 182 2,969 362 1,180 2,280 11,331 Note: System-average contamination level = medium Table A-74. Radioactive Gaseous Waste System Number 1 4 2 2 2 1 2 4 2 1 83 Total Component Surge Tank Decay Tank Gas Compressor Moisture Separator HEPA Prefilter Exhaust Fan Br. Seal Water Heat Exchanger Large Electrical Equipment Large Mechanical Equipment HVAC Equipment Valves (3/4 - 4" dia.) Mass (kg) Each 404 4,900 3,632 45 91 45 3,496 68 2,270 68 NA Total 404 19,600 7,264 90 182 45 6,992 272 4,540 68 4,607 44,064 Note: System-average contamination level = medium A-73 ------- Reference PWR Steel Inventories in the Auxiliary and Fuel Storage Buildings (continued) Table A-75. Residual Heat Removal System Number 2 2 12 11 1 42 Total Component Pump Heat Exchanger Unit Small Electrical Equipment Large Electrical Equipment Small Mechanical Equipment Valves (3/8 - 14" dia.) Mass (kg) Each 3,087 10,487 34 68 34 NA Total 6,174 20,974 408 748 34 49,032 77,370 Note: System-average contamination level = high Table A-76. Safety Injection System Number 4 1 2 1 1 10 10 1 89 Total Component Accumulator Tank B. Injection Tank Safety Injection Pump Refueling Water Tank Primary Water Storage Tank Small Electrical Equipment Large Electrical Equipment Small Mechanical Equipment Valves (3/4 - 10" dia.) Mass (kg) Each 34,731 12,939 3,904 80,721 45,037 34 68 34 NA Total 138,924 12,939 7,808 80,721 45,037 340 680 34 12,114 298,597 Note: System-average contamination level = medium A-74 ------- Reference PWR Steel Inventories in the Auxiliary and Fuel Storage Buildings (continued) Table A-77. Spent Fuel System Number 1 2 1 2 1 1 2 53 1 Total Component Pump Pump Pump Filter Filter Demineralizer Heat Exchanger Valves (3/4 - 10" dia.) Fuel Pool Liner Fuel Storage Racks Fuel Handling System Overhead Crane Mass (kg) Each 454 409 318 163 68 998 2,769 NA 37,000 Total 454 918 318 326 68 998 5,538 14,117 37,000 49,079 18,470 113,000 240,286 Note: System-average contamination level = high Table A-78. Structural Steel Components Number NA NA NA NA NA NA Total Component Wall Support Roof Support Stairs/Grates/Tracks/Hand-rails I-beams HVAC Ducts HVAC Components Mass (kg) Each NA NA NA NA NA NA Total 24,200 16,300 33,200 207,000 26,550 76,500 383,750 Note: System-average contamination level = low A-75 ------- Table A-79. Reference PWRNon-RCS Stainless Steel Piping Nominal ID. (in.) '/2 3/4 1 l'/2 2 3 4 6 8 10 12 14 Total Schedule 80 160 40 80 160 40 80 160 40 80 160 40 80 160 160 160 160 160 140 140 140 ID. (in.) 0.546 0.464 0.824 0.742 0.612 1.049 0.957 0.815 1.610 1.500 1.338 2.067 1.939 1.687 2.624 3.438 5.187 6.813 8.500 10.126 11.188 Length (m) 122 122 122 183 580 61 61 427 122 335 549 305 488 1,067 140 183 311 143 192 88 100 Inside Area (m2) 5.315 4.517 8.022 10.84 28.32 5.106 4.658 27.77 15.67 40.10 58.62 50.31 75.51 143.6 29.31 50.2 128.7 77.7 130 71.1 89.3 1055 Mass (kg) 198 238 205 400 1,671 152 590 1,803 493 1,810 3,967 1,655 3,642 11,840 2,985 6,128 20,972 15,923 29,750 18,370 24,474 147,266 Source: Smith et al. 1978, vol. 2, Table C.4-4 Notes: Includes piping for the following systems: residual heat removal, chem and volume control, emergency core cooling, containment spray, auxiliary feedwater, spent fuel pool cooling, condensate facility, station service, component cooling, service cooling, makeup water system. Contamination levels vary over several orders of magnitude from near background levels to 107 dpm/100 cm2. About 80% is assumed to be low-level contaminated with the remaining 20% medium-level. A. 5.1.3 Summary of Steel Inventories of the Reference Reactors Table A-80 presents a summary of steel inventories of the reference reactors—the rebar data is copied from Tables A-29 and A-31. Estimates of the contaminated steel inventories (comprising A-76 ------- both carbon and stainless steels) of the Reference BWR and PWR were derived by summing the masses of the components listed in Tables A-32 to A-64 and A-65 to A-79, respectively. Estimates of the stainless steel portions of these steel inventories were developed from information provided by Bryan and Dudley (1974), Oak et al. (1980) and Smith et al. (1978). These data were used to construct Table A-30, which presents a breakdown of the stainless steel used to construct a Reference PWR—the radioactively contaminated components are underlined. This table shows that 1,155 t of stainless steel in reactor plant equipment and 211 in spent fuel storage were contaminated, for a total of about 1,176 t, as listed in Table A-80. Included in this total, however, is about 348 t that is neutron activated at levels that would preclude the metal being cleared. Consequently, the releasable stainless steel inventory is about 828 t. Subtracting this from the total mass of 4,138 t of contaminated steel— the sum of the components listed in Tables A-65 to A-79—results in 3,310 t of contaminated, releasable carbon steel. The carbon and stainless steel inventories for the two metals with the three levels of contamination—shown in Table A-80—were derived, assuming that the low-, medium- and high-level contaminated components all contain the same proportions of carbon and stainless steel.3 Table A-80. Summary of Reference PWR and BWR Steel Inventories (t) Rebar All Other Total Potentially Releasablea Low-level13 Medium-level0 High-level11 Total Contaminated Volumetric PWR All Steel 34,811 4,138 1,051 572 2,515 Carbon Steel 13,000 19,731 32,731 3,310 841 458 2,012 864 Stainless 2,080 828 210 114 503 1,176 348 BWR All Steel 36,100 8,442 2,882 3,932 1,628 Carbon Steel 18,000 16,000 34,000 6,753 2,306 3,145 1,302 Stainless 2,100 1,689 576 786 326 Contaminated steel that can be potentially decontaminated to meet foreseeable clearance standards b <105dpm/100cm2 c 105— 107dpm/100cm2 d >107dpm/100cm2 The values displayed in this and other tables in this appendix are rounded; consequently, there may appear to be slight disparities in the totals shown. A-77 ------- The row marked "Total" lists the total quantities of steel used to construct each plant. "Releasable" refers to all contaminated steel that is a candidate for release, excluding only steel that is neutron-activated. (This includes metal that would require very aggressive decontamination methods to achieve any foreseeable clearance criteria.) The total mass of releasable, contaminated steel from the Reference BWR—the sum of the components listed in Tables A-32 to A-64—is 8,442 t. The carbon and stainless steel inventories for the BWR shown in Table A-80 were estimated assuming the same ratio of carbon steel to stainless as in the PWR4. A.5.2 Applicability of Reference Reactor Data to the Nuclear Industry The material inventories cited by Bryan and Dudley (1974) can be applied to other U.S. nuclear power plants; however, these inventories must be adjusted for the characteristics of individual plants, and the limitations inherent in this procedure must be acknowledged. The current U.S. nuclear power plant inventory comprises not only different designs but also varied power ratings. Nuclear power plant designs reflect standards for plant safety and the protection of the environment that have evolved over four decades. For example, Bryan and Dudley's reference plant used run-of-river cooling, which is not applicable to more recent nuclear facilities that employ cooling towers of various designs, holding ponds, sprays, etc. Significant quantities of materials are involved in some of these alternative cooling systems. Additionally, the 1979 accident at Three Mile Island mandated revised safety standards, which have added to the material inventory of more recent nuclear plants. Material inventories that reflect evolving changes in plant design have not been adequately addressed in the open literature, however. It is therefore not feasible to address such design changes in the present analysis. Instead, the material inventories of individual facilities will be based on those of the reference facilities, adjusted only for the individual reactor's power rating. A.5.2.1 Scaling Factors It is reasonable to assume a correlation between a plant's power rating and its material inventory. By this means, data collected for Reference PWRs and BWRs can be utilized to estimate inventories for the industry at large. In reports prepared for the DOE, Argonne National Laboratory (ANL) employed a scaling method based on the mass of PWR and BWR pressure A materials inventory for the stainless steel in the Reference BWR, such as the one for the PWR shown in Table A-30, could not be constructed from the available data. A-78 ------- vessels (Nuclear Engineering International 1991, 1992, 1993). ANL assumed that all metal inventories for both PWRs and BWRs can be calculated from those at the corresponding reference plant based on the design power, as follows: M = mass of metal (e.g., carbon steel) in actual reactor Mr = mass of same metal in reference reactor P = power rating of actual reactor (MWe) Pr = power rating of reference reactor ( v\~ The quantity, — 3, is referred to as the scaling factor. A.5.2.2 U.S. Nuclear Power Industry Table Al-1 in Appendix A-l lists the 104 nuclear power reactors currently licensed to operate by the NRC. The table also lists the scaling factors for PWRs and BWRs in separate columns. Scaling was based on the net maximum dependable capacity reported by the NRC (U.S. NRC 2000). It is recognized that this capacity may vary with time and a more constant metric would be the licensed thermal capacity of each reactor. However, since the inventory of materials listed in Table A-29 is for a 1000 MWe PWR, scaling was based on electrical rather than thermal output. Given the other uncertainties inherent in the scaling process, this choice should not significantly affect inventory estimates. In addition to the operating reactors, there are 27 nuclear power reactors which were formerly licensed to operate. (Of these, six were not light water reactors.) Only reactors which are in SAFSTOR or scheduled for DECON are included in this scrap metal analysis. Reactors where DECON is in progress or has been completed are excluded, as are reactors which are in an ENTOMB status. Thus, from the total population of formerly licensed nuclear power reactors, eight PWRs are included together with six BWRs and three other reactors (which are treated as BWRs5). Table Al-2 lists these 17 reactors, along with the scaling factors and dates when scrap metal releases might be expected. These reactors include Fermi-1, CVTR and Peach Bottom-1. Since these are all small plants (less than 200 MWt), treating them as BWRs will have little impact on the total scrap metal inventories. A-79 ------- A.5.2.3 Estimating the Metal Inventories of U.S. Nuclear Power Plants The following relationship was used to estimate metal inventories of U.S. nuclear power plants: 77 M = m S S j = i m 44 ; j = i (A-l) total inventory of metal category /' (e.g., contaminated stainless steel) from all nuclear power plants inventory of metal category /' in Reference PWR scaling factor for actual PWRy (see Tables Al-1 and A1-2) inventory of metal category /' in Reference B WR scaling factor for actual BWRy (see Tables Al-1 and Al-2) The results are shown in Table A-81. Approximately 587,000 t of contaminated steel may, over time, become candidates for clearance. About 80% of the contaminated steel is carbon steel with stainless steel representing the balance. The terms "Total" and "Releasable" were explained in connection with Table A-80. Table A-81. Steel Inventories of U.S. Nuclear Power Facilities (t) Reactor Type — Sum of Scaling Factors Rebar All Other Total Releasable3 Low" Medium0 High" PWR — 71.954 All Steel 2.50e+06 2.98e+05 7.56e+04 4.12e+04 1.81e+05 Carbon Steel 9.35e+05 1.426+06 2.36e+06 2.386+05 6.05e+04 3.296+04 1.45e+05 Stainless 1.50e+05 5.966+04 1.51e+04 8.236+03 3.62e+04 BWR — 34.249 All Steel 1.24e+06 2.896+05 9.87e+04 1.35e+05 5.586+04 Carbon Steel 6.16e+05 5.486+05 1.16e+06 2.316+05 7.906+04 1.086+05 4.46e+04 Stainless 7.196+04 5.78e+04 1.97e+04 2.696+04 1.126+04 Total Industry All Steel 3.74e+06 5.876+05 1.74e+05 1.766+05 2.376+05 Carbon Steel 1.55e+06 1.976+06 3.52e+06 4.696+05 1.39e+05 1.41e+05 1.89e+05 Stainless 2.22e+05 1.17e+05 3.496+04 3.526+04 4.736+04 a Contaminated steel that can be potentially decontaminated to meet foreseeable clearance standards Low-level contamination: <10 dpm/100 cm c Medium-level contamination: 10 —10 dpm/100 cm High-level contamination: >10 dpm/100 cm The radioactive contaminants of most of the metal components that are candidates for clearance will be found on the surface. Therefore, in the preceding sections of this appendix, contamination levels have been cited as areal activity concentrations, in units of dpm/100 cm2 or A-80 ------- Ci/m2. However, in the exposure scenarios discussed in Chapters 5 and 6, the radiation sources are modeled as bulk material. Thus, whether the source is a pile of assorted scrap, or the residually radioactive metal products and non-metallic byproducts of the steel refining process, contamination expressed as mass activity concentrations (i.e., specific activities), in units such as pCi/g, is a more meaningful quantity. Specific activities can be derived from areal activity concentrations by the following relationship: Sjj = specific activity of nuclide /'in component y'(pCi/g) Qj = areal activity concentration of nuclide /' in component y' (pCi/cm2 = 108 Ci/m2) Oj = mass thickness of component y' (g/cm2) aj nij = mass of component y' (g) aj = area of contaminated surface of component y' (cm2) Since the present radiological assessment addresses the clearance and subsequent recycle of large quantities of cleared metals rather than individual components, it is useful to calculate the average mass thickness of all carbon steel that will be potentially cleared from U.S. nuclear power facilities. This quantity can be expressed as follows: 44 Ap S •» + A — = - j = i Mc *p P; a = area of component / of Reference PWR ab = area of component /' of Reference BWR Mc = mass of all carbon steel potentially cleared from U.S. nuclear power facilities, given by Eq. A-l The areas of the individual PWR components were based on data presented by Smith et al. (1978), while the corresponding BWR data was presented by Oak et al. (1980). A-81 ------- Table A-82. Average Mass Thickness of Carbon Steel Inventories Reactor Type PWR BWR Total Sum of Scaling Factors 71.954 34.249 Reference Reactor Mass (g) 3.31e+09 6.75e+09 Area (cm2) 2.19e+08 2.40e+09 Total Mass (g) 2.38e+ll 2.31e+ll 4.69e+ll Mass Thickness (g/cm2) Total Area (cm2) 1.58e+10 8.22e+10 9.80e+10 4.79 A.5.3 Metal Inventories Other Than Steel Although steel is clearly the predominant metal used in the construction and components of a nuclear power plant, there are also significant quantities of other metals. Tables A-29 lists the total inventories of nine metals for the Reference PWR. (In the absence of other data, the same total inventories were adopted for the Reference BWR.) There are no available data on the radiological contamination of these metals. However, most of these metals are in components that are made primarily of carbon steel. It is therefore assumed that these metals have contamination profiles similar to those of the carbon steel components of the Reference PWR and the Reference BWR, respectively. Table A-83. Inventories of Metals Other Than Steel (t) Metal Galvanized Iron Copper Inconel Lead Bronze Aluminum Brass Nickel Silver Total Inventory — Industry 138,064 73,280 12,744 4,885 2,655 1,912 1,062 106 <106 Contaminated — Subject to Clearance* Reference Facility PWR 131 70 12 4.7 2.5 1.8 1.0 0.1 <0.1 BWR 258 137 24 9.1 5.0 3.6 2.0 0.2 <0.2 Nuclear Power Industry All PWRs 9,460 5,021 873 335 182 131 73 7.3 <7.0 All BWRs 8,844 4,694 816 313 170 122 68 6.8 <6.7 Total 18,304 9,715 1690 648 352 253 141 14 <14 Contaminated metals that can be potentially decontaminated to meet foreseeable clearance standards A-82 ------- A.5.4 Timetable for the Release of Scrap Metals from Nuclear Power Plants The projected year of shutdown for each of 104 operating units is listed in Table Al-1. For the purpose of the present analysis, it was assumed that any scrap metal would be released ten years after reactor shutdown.6 As described in Section A.5.2.2, Table Al-2 lists the 17 shut-down commercial nuclear power reactors included in the present analysis, along with the dates when significant scrap metal releases might be expected. Table A-84 summarizes the availability of scrap for each year during which one or more plants would begin releasing scrap metal. The amount of each metal released during that year is calculated by a formula similar to Eq. A-l: Mik = mPi .^ sp. + mb; £ sb. Mik = total inventory of metal /' from all nuclear plants dismantled in year k nkp = number of PWRs dismantled in year k nkb = number of BWRs dismantled in year k Columns 2 and 3 list the sum of the scaling factors of the PWR and BWR plants, respectively, that are expected to begin major decommissioning activities in the year listed in Column 1. The remaining columns list the mass of each metal that would be released that year, assuming that all metal from a given plant would be released in one year. It is recognized that, in fact, the releases from each plant would span a period of several years, and that there would be considerable overlap in the releases from various plants that shut down within a few years of each other. Nevertheless, this table presents an overview of the anticipated rate of release in future years. The actual release dates of scrap metal may be later than those listed. First, as mentioned in Note 1, a number of reactors may receive 20-year extensions to their operating licenses, thereby delaying the projected date of decommissioning. Some facilities are likely to elect the SAFSTOR decommissioning alternative, thereby delaying releases for up to 50 years. In the case of reactors for which the SAFSTOR decommissioning alternative was selected, clearance is asumed to occur 60 years after shutdown (see Appendix A-l). A-83 ------- Table A-84. Anticipated Releases of Scrap Metals from Nuclear Power Plants (t) Year 2006 2007 2016 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2043 2044 2045 2046 2047 2049 2052 2056 2057 2058 Total Z scaling factors3 PWR 1.48 0 0 0.6 1.39 0.81 1.65 5.12 3.38 1.89 3.71 2.82 1.83 3.08 4.09 3.06 1.97 5.8 4.4 5.23 5.36 1.16 1.99 1.1 2.89 1.78 1.08 0.89 0 0.88 0.55 0.98 0.98 0 72 BWR 0 0.17 0.84 1.41 0.67 0.84 3.08 2.16 6.11 0 1.88 0.1 0.87 0 0 3.35 2.09 2.24 1.87 4.3 0 0.35 1.08 0 0 0 0 0 0.14 0 0 0 0 0.71 34.2 c O CD .Q CD 8" 4,906 1,169 5,683 1 1 ,522 9,111 8,372 26,266 31,573 52,479 6,252 24,978 9,844 1 1 ,922 10,202 13,527 32,775 20,675 34,307 27,206 46,335 17,730 6,229 13,847 3,634 9,556 5,896 3,564 2,947 917 2,928 1,809 3,255 3,255 4,820 469,490 CO CO — CD CD <= £ ro W CO 1,227 292 1,421 2,881 2,278 2,093 6,568 7,894 13,122 1,563 6,245 2,461 2,981 2,551 3,382 8,195 5,170 8,578 6,802 11,585 4,433 1,558 3,462 909 2,389 1,474 891 737 229 732 452 814 814 1,205 117,389 Galvanized Iron 195 45 217 444 355 324 1,012 1,232 2,023 248 973 390 465 405 537 1,268 800 1,340 1,062 1,797 704 244 539 144 380 234 142 117 35 116 72 129 129 184 18,304 s_ CD Q. Q. O 0 103 24 115 235 189 172 537 654 1,074 132 517 207 247 215 285 673 425 711 564 954 374 129 286 77 201 124 75 62 19 62 38 69 69 98 9,715 Inconel 18 4 20 41 33 30 93 114 187 23 90 36 43 37 50 117 74 124 98 166 65 23 50 13 35 22 13 11 3.2 11 6.6 12 12 17 1,690 •a CO CD 6.9 1.6 7.7 16 13 11 36 44 72 8.8 34 14 16 14 19 45 28 47 38 64 25 8.6 19 5.1 13 8.3 5.0 4.1 1.2 4.1 2.5 4.6 4.6 6.5 648 CD N O m 3.7 0.86 4.2 8.5 6.8 6.2 19 24 39 4.8 19 7.5 8.9 7.8 10 24 15 26 20 35 14 4.7 10 2.8 7.3 4.5 2.7 2.3 0.67 2.2 1.4 2.5 2.5 3.5 352 E ^ c E ^ < 2.7 0.62 3.0 6.1 4.9 4.5 14 17 28 3.4 13 5.4 6.4 5.6 7.4 18 11 19 15 25 9.8 3.4 7.5 2.0 5.3 3.2 2.0 1.6 0.49 1.6 0.99 1.8 1.8 2.6 253 CO CO 2. CO 1.5 0.34 1.7 3.4 2.7 2.5 7.8 9.5 16 1.9 7.5 3.0 3.6 3.1 4.1 9.8 6.2 10 8.2 14 5.4 1.9 4.1 1.1 2.9 1.8 1.1 0.90 0.27 0.89 0.55 0.99 1.0 1.4 141 CD _*: o '•z. 0.15 0.034 0.17 0.34 0.27 0.25 0.78 0.95 1.6 0.19 0.75 0.30 0.36 0.31 0.41 0.98 0.62 1.0 0.82 1.4 0.54 0.19 0.41 0.11 0.29 0.18 0.11 0.090 0.027 0.089 0.055 0.10 0.10 0.14 14 Values displayed are rounded; however, full precision was used in calculation A-84 ------- REFERENCES Abel, K. H., et al. 1986. "Residual Radionuclide Contamination Within and Around Commercial Nuclear Power Plants," NUREG/CR-4289. Pacific Northwest Laboratory, prepared for the U.S. Nuclear Regulatory Commission, Washington, DC. Bryan, R. H., and I. T. Dudley. 1974. "Estimated Quantities of Materials Contained in a 1000- MW(e) PWR Power Plant," ORNL-TM-4515. Oak Ridge National Laboratory, prepared for the U.S. Atomic Energy Commission. Consumers Power Company. 1995. "Decommissioning Cost Study for the Big Rock Point Nuclear Plant." Dyer, N. C. 1994. "Radionuclides in United States Commercial Nuclear Power Reactors," WINCO-1191, UC-510, ed. T. E. Bechtold. Westinghouse Idaho Nuclear Company, Inc., prepared for the Department of Energy, Idaho Operations Office. Konzek, G. J., et al. 1995. "Revised Analyses of Decommissioning for the Reference Pressurized Water Reactor Power Station," NUREG/CR-5884, PNL-8742. 2 vols. Prepared by Pacific Northwest Laboratory for the U.S. Nuclear Regulatory Commission, Washington, DC. Nuclear Engineering International. 1991. World Nuclear Industry Handbook 1991. Nuclear Engineering International, Surrey, U.K. Nuclear Engineering International. 1992. World Nuclear Industry Handbook 1992. Nuclear Engineering International, Surrey, U.K. Nuclear Engineering International. 1993. World Nuclear Industry Handbook 1993. Nuclear Engineering International, Surrey, U.K. Oak, H. D., et al. 1980. "Technology, Safety and Costs of Decommissioning a Reference Boiling Water Reactor Power Station," NUREG/CR-0672. 2 vols. Pacific Northwest Laboratory, prepared for the U.S. Nuclear Regulatory Commission, Washington, DC. Pacific Gas and Electric Company. 1994. "SAFSTOR Decommissioning Plan for the Humboldt Bay Power Plant, Unit 3" Portland General Electric. 1996. "Trojan Nuclear Plant Decommissioning Plan," PGE-1061. A-85 ------- Smith, R.I., G. J. Konzek, and W. E. Kennedy, Jr. 1978. "Technology, Safety and Costs of Decommissioning a Reference Pressurized Water Reactor Power Station," NUREG/CR- 0130. 2 vols. Pacific Northwest Laboratory, prepared for the U.S. Nuclear Regulatory Commission, Washington, DC. Smith, R. I, et al. 1996. "Revised Analyses of Decommissioning for the Reference Boiling Water Reactor Power Station," NUREG/CR-6174, PNL-9975. 2 vols. Pacific Northwest Laboratory, prepared for the U.S. Nuclear Regulatory Commission, Washington, DC. Southern California Edison Company. 1994. "San Onofre Nuclear Generating Station, Unit 1, Decommissioning Plan." U.S. Atomic Energy Commission (U.S. AEC). 1974. Regulatory Guide 1.86: "Termination of Operating Licenses for Nuclear Reactors." U. S. AEC, Washington, DC. U.S. Nuclear Regulatory Commission (U.S. NRC). 1988. "General Requirements for Decommissioning Nuclear Facilities," Federal Register, Vol. 53, No. 123, June 27, 1988. U.S. Nuclear Regulatory Commission (U.S. NRC). 1994. "Generic Environmental Impact Statement in Support of Rulemaking on Radiological Criteria for Decommissioning of NRC- Licensed Nuclear Facilities," NUREG-1496 (Draft). U.S. NRC, Washington, DC. U.S. Nuclear Regulatory Commission (U.S. NRC). 2000. "Information Digest, 2000 Edition," NUREG-1350, Volume 12. U.S. NRC, Washington, DC. Yankee Atomic Electric Company. 1995. "Yankee Nuclear Power Station Decommissioning Plan." A-86 ^ Continue ------- Back APPENDIX A-l U.S. COMMERCIAL NUCLEAR POWER REACTORS ------- U.S. COMMERCIAL NUCLEAR POWER REACTORS Table Al-1 presents a list of the 104 commercial nuclear power reactors in the U.S. currently licensed to operate by the NRC. The reactor type (BWR or PWR) is listed, along with its electrical generating capacity, and its scaling factor, which is described in Section A.5.2.1. The scaling factors for PWRs and BWRs are listed in separate columns to enable the sum of these factors for each type of reactor to be calculated separately; however, the factors for individual PWRs and BWRs are calculated by the same formula. The year of projected shutdown is based on the expiration date of the current operating license, including, in three cases, credit for construction recapture. Construction recapture is defined as "[t]he maximum number of years that could be added to the license expiration date to recover the period from the construction permit to the date when the operating license was granted. A licensee is required to submit an application for such a change." (U.S. NRC 2000) Al-1 ------- Table Al-1. Nuclear Power Reactors Currently Licensed to Operate Electric Utility Arizona Public Service Arizona Public Service Arizona Public Service Baltimore Gas & Electric Baltimore Gas & Electric Boston Edison Carolina Power & Light Carolina Power & Light Carolina Power & Light Carolina Power & Light Centerior Energy Cleveland Electric Commonwealth Edison Commonwealth Edison Commonwealth Edison Commonwealth Edison Commonwealth Edison Commonwealth Edison Commonwealth Edison Commonwealth Edison Commonwealth Edison Commonwealth Edison Consolidated Edison Consumers Energy Detroit Edison Duke Power Duke Power Duke Power Duke Power Duke Power Duke Power Duke Power Duquesne Light Duquesne Light Entergy Operations, Inc. Entergy Operations, Inc. Entergy Operations, Inc. Reactor Palo Verde 1 Palo Verde 2 Palo Verde 3 Calvert Cliffs 1 Calvert Cliffs 2 Pilgrim 1 Brunswick 1 Brunswick 2 H. B. Robinson 2 Shearon Harris 1 Davis-Besse Perry 1 Braidwood 1 Braidwood 2 Byron 1 Byron 2 Dresden 2 Dresden 3 LaSalle 1 LaSalle 2 Quad Cities 1 Quad Cities 2 Indian Point 2 Palisades 1 Fermi 2 Catawba 1 Catawba 2 McGuire 1 McGuire 2 Oconee 1 Oconee 2 Oconee 3 Beaver Valley 1 Beaver Valley 2 Arkansas Nuclear 1 Arkansas Nuclear 2 Grand Gulf 1 Type PWR PWR PWR PWR PWR BWR BWR BWR PWR PWR PWR BWR PWR PWR PWR PWR BWR BWR BWR BWR BWR BWR PWR PWR BWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR BWR Power Rating (MWe)a 1,227 1,227 1,230 835 840 670 767 754 683 860 873 1,160 1,100 1,100 1,105 1,105 772 773 1,036 1,036 769 769 951 730 876 1,129 1,129 1,129 1,129 846 846 846 810 820 836 858 1,179 Scaling Factor13 PWR 1.146 1.146 1.148 0.887 0.890 — — — 0.776 0.904 0.913 — 1.066 1.066 1.069 1.069 — — — — — — 0.967 0.811 — 1.084 1.084 1.084 1.084 0.895 0.895 0.895 0.869 0.876 0.887 0.903 — BWR — — — — — 0.766 0.838 0.828 — — — 1.104 — — — — 0.842 0.842 1.024 1.024 0.839 0.839 — — 0.916 — — — — — — — — — — — 1.116 Year of Projected Shutdown 2024 2025 2027 2034 2036 2012 2016 2014 2010 2026 2017 2026 2026 2027 2024 2026 2006 2011 2022 2023 2012 2012 2013 2011C 2025 2024 2026 2021 2023 2033 2033 2034 2016 2027 2014 2018 2022 Source: U.S. NRC 2000 a Net maximum dependable capacity Scaling factor = (power rating/1000) (see text) c Year assuming construction recapture Al-2 ------- Table Al-1 (continued) Electric Utility Entergy Operations, Inc. Entergy Operations, Inc. Florida Power Corp. Florida Power & Light Florida Power & Light Florida Power & Light Florida Power & Light GPU Nuclear GPU Nuclear Illinois Power Indiana/Michigan Power Indiana/Michigan Power IES Utilities Nebraska Public Power New York Power Authority New York Power Authority Niagara Mohawk Niagara Mohawk North Atlantic Energy Northeast Nuclear Energy Northeast Nuclear Energy Northern States Power Northern States Power Northern States Power Omaha Public Power Pacific Gas & Electric Pacific Gas & Electric PECO Energy PECO Energy Pennsylvania Power Pennsylvania Power Philadelphia Electric Philadelphia Electric Public Service E & G Public Service E & G Public Service E & G Rochester Gas & Electric South Carolina E & G Reactor River Bend 1 Waterford 3 Crystal River 3 St. Lucie 1 St. Lucie 2 Turkey Point 3 Turkey Point 4 Oyster Creek Three Mile Island 1 Clinton D. C. Cook 1 D. C. Cook 2 Duane Arnold Cooper James A. Fitzpatrick Indian Point 3 Nine Mile Point 1 Nine Mile Point 2 Seabrook 1 Millstone 2 Millstone 3 Monticello Prairie Island 1 Prairie Island 2 Fort Calhoun Diablo Canyon 1 Diablo Canyon 2 Peach Bottom 2 Peach Bottom 3 Susquehanna 1 Susquehanna 2 Limerick 1 Limerick 2 Hope Creek 1 Salem 1 Salem 2 Ginna 3 Summer Type BWR PWR PWR PWR PWR PWR PWR BWR PWR BWR PWR PWR BWR BWR BWR PWR BWR BWR PWR PWR PWR BWR PWR PWR PWR PWR PWR BWR BWR BWR BWR BWR BWR BWR PWR PWR PWR PWR Power Rating (MWe)a 936 1,104 818 839 839 693 693 619 786 930 1,000 1,060 520 764 762 965 565 1,105 1,158 871 1,137 544 513 512 478 1,073 1,087 1,093 1,093 1,090 1,094 1,105 1,115 1,031 1,115 1,115 470 945 Scaling Factor13 PWR — 1.068 0.875 0.890 0.890 0.783 0.783 — 0.852 — 1.000 1.040 — — — 0.977 — — 1.103 0.912 1.089 — 0.641 0.640 0.611 1.048 1.057 — — — — — — — 1.075 1.075 0.605 0.963 BWR 0.957 — — — — — — 0.726 — 0.953 — — 0.647 0.836 0.834 — 0.683 1.069 — — — 0.666 — — — — — 1.061 1.061 1.059 1.062 1.069 1.075 1.021 — — — — Year of Projected Shutdown 2025 2024 2016 2016 2023 2012 2013 2009 2014 2026 2014 2017 2014 2014 2014 2015 2009 2026 2026 2015 2025 2010 2013 2014 2013 2021 2025 2013 2014 2022 2024 2024 2029 2026 2016 2020 2009 2022 a Net maximum dependable capacity Scaling factor = (power rating/1000) 3 (see text) Al-3 ------- Table Al-1 (continued) Electric Utility Southern California Edison Southern California Edison Southern Nuclear Southern Nuclear Southern Nuclear Southern Nuclear Southern Nuclear Southern Nuclear STP Nuclear STP Nuclear Tennessee Valley Authority Tennessee Valley Authority Tennessee Valley Authority Tennessee Valley Authority Tennessee Valley Authority Tennessee Valley Authority Texas Utilities Electric Texas Utilities Electric Union Electric Vermont Yankee Nuclear Virginia Electric & Power Virginia Electric & Power Virginia Electric & Power Virginia Electric & Power Washington Public Power Wisconsin Electric Power Wisconsin Electric Power Wisconsin Public Service Wolf Creek Nuclear Reactor San Onofre 2 San Onofre 3 Edwin 1. Hatch 1 Edwin I. Hatch 2 Joseph M. Farley 1 Joseph M. Farley 2 Vogtle 1 Vogtle 2 South Texas 1 South Texas 2 Browns Ferry 1 Browns Ferry 2 Browns Ferry 3 Sequoya 1 Sequoya 2 Watts Bar 1 Comanche Peak 1 Comanche Peak 2 Callaway Vermont Yankee North Anna 1 North Anna 2 Surry 1 Surry2 Washington Nuclear 2 Point Beach 1 Point Beach 2 Kewaunee Wolf Creek 1 Type PWR PWR BWR BWR PWR PWR PWR PWR PWR PWR BWR BWR BWR PWR PWR PWR PWR PWR PWR BWR PWR PWR PWR PWR BWR PWR PWR PWR PWR Power Rating (MWe)a 1,070 1,080 805 809 812 822 1,162 1,162 1,251 1,251 1 ,065d 1,065 1,065 1,117 1,117 1,117 1,150 1,150 1,171 510 893 897 801 801 1,107 485 485 511 1,163 Total Scaling Factor13 PWR 1.046 1.053 — — 0.870 0.878 1.105 1.105 1.161 1.161 — — — 1.077 1.077 1.077 1.098 1.098 1.111 — 0.927 0.930 0.862 0.862 — 0.617 0.617 0.639 1.106 65.866 BWR — — 0.865 0.868 — — — — — — 1.043 1.043 1.043 — — — — — — 0.638 — — — — 1.070 — — — — 32.327 Year of Projected Shutdown 2022= 2022= 2014 2018 2017 2021 2027 2029 2027 2028 2013 2014 2016 2020 2021 2035 2030 2033 2024 2012 2018 2020 2012 2013 2023 2010 2013 2013 2025 Net maximum dependable capacity Scaling factor = (power rating/1000) 3 (see text) Assuming construction recapture Based on design characteristics—reactor has no fuel loaded and requires NRC approval to restart. Al-4 ------- Table Al-2 lists the commercial nuclear power reactors that were formerly licensed but have been shut down. As was stated in Section A.5.2.2, the list excludes reactors whose owners have chosen the ENTOMB decommissioning alternative, and those with the DECON alternative that have begun or already completed decommissioning. It is unlikely that reactors in these categories would be clearing scrap metal in the foreseeable future. As before, scaling factors for PWR and BWR plants are listed in separate columns. For the purpose of the present analysis, the three non-light water reactors are treated as if they were BWRs. The last column lists the date that significant quantities of scrap metal would be released from these reactors. For reactors in SAFSTOR, this is assumed to be 60 years after the shutdown date, while for those with the DECON alternative it is ten years after shutdown. Al-5 ------- Table Al-2. Formerly Licensed Nuclear Power Reactors Reactor Big Rock Point CVTR Dresden 1 Fermi 1 GEVBWR Haddam Neck Humboldt Bay Indian Point 1 La Crosse Maine Yankee Millstone 1 Peach Bottom 1 Rancho Seco San Onofre 1 Three Mile Island 2 Zion 1 Zion 2 Type BWR PTHW3 BWR SCFe BWR PWR BWR PWR BWR PWR BWR HTGRe PWR PWR PWR PWR PWR Power Rating (MWe)a 72 20 210 60 15 548 60 185 50 732 603 34 832 404 831 975 975 shut down reactors (see note) Total including currently licensed reactors Scaling Factor13 PWR — — — — — 0.670 — 0.325 — 0.812 — — 0.885 0.547 0.884 0.983 0.983 6.088 71.954 BWR 0.173 0.074 0.353 0.153 0.061 — 0.153 — 0.136 — 0.714 0.105 — — — — — 1.922 34.249 Alternative0 DECON SAFSTOR SAFSTOR SAFSTOR SAFSTOR DECON SAFSTOR SAFSTOR SAFSTOR DECON SAFSTOR SAFSTOR SAFSTOR' SAFSTOR y SAFSTOR SAFSTOR Year Shutdown 1997 1967 1978 1972 1963 1996 1976 1974 1987 1996 1998 1974 1989 1992 1979 1997 1996 Released 2007 2027 2038 2032 2023 2006 2036 2034 2047 2006 2058 2034 2049 2052 2039 2057 2056 Source: U.S. NRC 2000 Note: excludes reactors at which DECON has started or been completed and those in ENTOMB status Licensed thermal capacity x 0.3 Scaling factor = (power rating/1000)% (see text) c Selected decommissioning alternative Year that significant quantities of scrap metal will be released—10 years after shutdown for the DECON alternative, 60 years for SAFSTOR e Metals inventory and contamination levels assumed same as for BWR Dismantlement of radioactive secondary piping and components is ongoing g In monitored storage until TMI-1 is shut down, then both will be decommissioned REFERENCE U.S. Nuclear Regulatory Commission (U.S. NRC). 2000. "Information Digest, 2000 Edition," NUREG-1350, Volume 12. U.S. NRC, Washington, DC. Al-6 ------- APPENDIX B ALUMINUM RECYCLING ------- Contents page B. 1 Inventory B-l B. 1.1 Scrap Metal Inventory B-l B.I.2 Radionuclide Inventory B-5 B.2 Recycling of Aluminum Scrap B-6 B.2.1 Secondary Aluminum B-6 B.2.2 Composition of Scrap Aluminum B-6 B.3 Structure of the Scrap Industry B-7 B.4 Secondary Aluminum Industry B-9 B.4.1 Scrap Handling and Preparation B-10 B.4.2 Melting Practice B-13 B.4.3 Dust Handling B-15 B.4.4 Partitioning of Contaminants B-16 B.4.4.1 Thermochemical Considerations B-16 B.4.4.2 Observed Partitioning B-20 B.4.4.3 Baghouse Dust B-24 B.4.4.4 Proposed Partitioning B-25 B.4.5 Dross Processing B-28 B.4.6 Handling Baghouse Dust B-31 B.4.7 Product Shipments B-32 B.5 Product Markets B-32 B.6 Basis for Exposure Scenarios B-35 B.6.1 Exposure Parameters B-35 B.6.2 Workers in the Secondary Aluminum Industry B-38 B.6.3 Users of End-Products B-40 References B-44 Appendix B-l. Description of Selected Secondary Smelters Appendix B-2. Secondary Aluminum Smelter Operations at Arkansas Aluminum Alloys Inc. B2.1 Facility Description B2-1 B2.2 Process Description B2-1 Reference B2-3 B-iii ------- Tables page B-l. Aluminum Scrap Potentially Available from Nuclear Facilities B-l B-2. Availability of Potentially Contaminated Aluminum from Nuclear Facilities B-3 B-3. Current Inventory of Potentially Contaminated Aluminum Scrap at DOE Facilities . . . B-4 B-4. U.S. Consumption of Aluminum Scrap by Primary Producers, Foundries, Independent Mill Fabricators and Others in 1995 B-8 B-5. U.S. Consumption of Purchased Old and New Scrap by Secondary Smelters in 1995 . . B-9 B-6. TCLP Values for Dust Samples and Spent Refractory B-l6 B-7. Secondary Aluminum Smelter Dust Levels B-17 B-8. Standard Free Energy of Formation (AF°) for Various Metal Chlorides at 1,000 K . . . B-19 B-9. Selected Metal Chlorides with Boiling Points Below 1000 K B-20 B-10. Partitioning of Uranium in Aluminum Melts in Zirconia Crucibles at 1573 K B-22 B-ll. Cation Impurities in 3XX Aluminum Residue-Oxide Samples B-24 B-12. Composition of Particulate Matter From Secondary Aluminum Smelter B-25 B-13. Proposed Partitioning of Selected Elements During Secondary Aluminum Smelting . B-27 B-14. Production of Secondary Aluminum Alloys by Independent U.S. Smelters in 1995 . . B-33 B-l5. Representative Applications for Aluminum Casting Alloys B-34 B-l6. Concentrations in Ambient Air Inside and Outside the Welder's Helmet During Aluminum Welding and Cutting B-42 B-17. Dust Levels During Plasma Arc Cutting of Wrought Metal 2090 B-43 Bl-1. Description of Selected Secondary Smelters Bl-1 Figures B-l. Typical Secondary Aluminum Smelter Flow Diagram (after Viland 1990) B-ll B-2. Handling of Scrap Turnings from Forged Aluminum Auto Wheels at IMCO's Uhrichville OH plant B-12 B-3. Scrap Shredder at Secondary Aluminum Smelter B-12 B-4. Aluminum Liquid Metal Transporter B-14 B-5. Proposed Salt Cake Recycling Process B-30 B-6. Simplified Material Balance for Secondary Aluminum Smelter B-37 B-iv ------- ALUMINUM RECYCLING This appendix provides information on the recycling of aluminum and the use of its products, byproducts, and wastes. B.I INVENTORY Based on the review provided in this section, the total quantity of aluminum scrap metal, both clean and potentially contaminated, attributable to the nuclear industry, is listed in Table B-l. Table B-l. Aluminum Scrap Potentially Available from Nuclear Facilities (t) Commercial Nuclear Power Plants Total 1,900 Contaminated 253 DOE Facilities Contaminated 36,070 Total Contaminated 36,323 A more recent DOE summary states that the total aluminum available as radioactive scrap metal from DOE and NRC-licensed facilities (other than nuclear power plants) is 30,000 tons1 (Adams 1998). Presumably this is contaminated and suspect contaminated material. The DOE estimate is in reasonable agreement with the quantities tabulated above. B.I.I Scrap Metal Inventory Chapter 4 of the present report summarizes information on the potential quantities of aluminum scrap available for recycle from DOE and commercial facilities. However, there is no available information as to the portion of the aluminum that may be contaminated and the radionuclide composition of the contamination. Most of the aluminum from commercial nuclear power plants is expected to be in gratings, switch gear, and component housings. It is proposed in Section 4.2.2 that, for the purpose of the present analysis, a reasonable approach is to assume that the contaminated fraction of aluminum among total nuclear power plant scrap metal inventories This appendix includes numerous references with widely varying units of measurement. The authors of this appendix have generally chosen not to convert the units to a consistent system but rather have chosen to quote information from the various sources in the original units. When the cited information is distilled into scenarios for modeling doses and risks, consistent units are used. B-l ------- parallels the contaminated fraction of carbon steel for the Reference BWR and the Reference PWR. According to Table A-83, the total of amount of aluminum in all commercial nuclear power reactors is about 1,900 metric tons (t). Only a fraction of this inventory is expected to be significantly contaminated and not all of the contaminated inventory may be potentially suitable for recycling. Assuming that all metals have the same contamination profiles as steel, it is estimated that 20% of the aluminum in the Reference BWR and 10% in the Reference PWR is contaminated but potentially recyclable2. Applying these factors to the entire U.S. commercial nuclear power industry yields 122 t from all BWRs and 131 t from all PWRs, for a total of 253 t. For currently operating reactors, it is assumed that the scrap will be available ten years after the expiration of the current operating license. The methodology for assessing formerly licensed reactors is presented in Appendix A-l. Using this decommissioning schedule, annual availability of scrap can be established as shown in Table B-2 (based on Appendix A, Table A-84). It can be seen that 28 t of aluminum would be released from commercial nuclear power plants in the peak year: 2024. Based on a survey of DOE data, it is estimated that 2,353 t of contaminated, potentially releasable aluminum were in inventory at the end of 1996 (see Table 4-4) and that 33,717 t of contaminated aluminum will be generated from future decommissioning activities, resulting in a total of 36,070 t of contaminated aluminum (see Table 4-5)3. Approximately 98% of this aluminum scrap is expected to come from dismantling the gaseous diffusion plants (GDP) at K-25 (Oak Ridge, Tenn.), Portsmouth, Ohio, and Paducah, Ky. Decommissioning schedules for the diffusion plants are assumed to be as follows (see Section 4.1.5): • K-25 1998 to 2006 • Portsmouth 2007 to 2015 Garbay and Chapuis (1991) concluded that a PWR contained 20 to 100 t of aluminum, mostly as electrical cable. The authors assumed that about 25% was contaminated and selected 20 t as the value for modeling exposures. They further assumed that two PWRs would be decommissioned each year, resulting in 40 t of contaminated aluminum available for recycle annually. This value appears to be conservative (i.e., high) since Compere et al. (1996) note that only 20,100 t of radioactive aluminum/copper will be available from the three diffusion plants while Table 4-5 lists a total of 35,300 t. B-2 ------- • Paducah 2015 to 2023 For the purposes of analyzing the DOE facilities, it was assumed that no scrap metal is generated in the first year (of a nine-year decommissioning period), 9% is generated in the final year, and 13% is generated in each of years 2 through 8. Table B-2. Availability of Potentially Contaminated Aluminum from Nuclear Facilities (t) Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 Total DOE Facilities 7237 979 979 679 — 780 780 780 780 780 780 780 540 2,636 2,636 2,636 2,636 2,636 2,636 2,636 1,746 — — — 36,075 Commercial Nuclear Power Plants — — — 2.7 0.6 — — — — — — — — 3.0 — — 6.1 4.9 4.5 14 17 28 3.4 13 Year 2027 2028 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2043 2044 2045 2046 2047 2049 2052 2056 2057 2058 Commercial Nuclear Power Plants 5.4 6.4 5.6 7.4 18 11 19 15 25 9.8 3.4 7.5 2.0 5.3 3.2 2.0 1.6 0.5 1.6 1.0 1.8 1.8 2.6 253 Note: Values may differ to roundoff error B-3 ------- Although dismantlement of the K-25 facilities is in progress, DOE is not currently releasing any scrap metals generated in the process. In January 2000, the Secretary of Energy issued a moratorium on the Department's release of volumetrically contaminated metals "pending a decision by the ... NRC... whether to establish national standards.... On July 13, 2000, the Secretary of Energy issued a memorandum ... [which] suspended the unrestricted release for recycling of scrap metal from radiological areas within DOE facilities. This suspension will remain in effect until improvements in DOE release criteria and information management have been developed and implemented" (Michaels 2000). Based on these DOE policy decisions, it is assumed in this report that releases of scrap metal from DOE facilities will not begin until 2003. Some information as to the breakdown by location of aluminum scrap in the DOE inventory can be found in U.S. DOE 1996, vol. 2. These data are reproduced in Table B-3. Since most of this material is not specified to be clean or contaminated in the source document, the same methodology used in Chapter 4 is applied here. Table 4-4 indicates that 271 are "clean," 141 contaminated, and 5,6371 "unspecified." It was therefore assumed that 34.1% (14 ^ [14 + 27] = 0.341) of the "unspecified material" at each site was contaminated while the rest was clean. Furthermore, the quantity reported for each site was multiplied by a scaling factor of 1.213 to ensure that the total of all the sites conform to the totals in Table 4-4. Table B-3. Current Inventory of Potentially Contaminated Aluminum Scrap at DOE Facilities (t) Site K-25 ORNL Y-12 Paducah Portsmouth Total Clean — — — — — 27 Contaminated — — — — — 14 Unspecified 1,100 20 38 4,165 314 5,637 Total 1,100 20 38 4,165 314 5,678 Contaminated Assumed 376 7 13 1,422 107 1,925 Total 376 7 13 1,422 107 1,939 Scaled3 456 8 16 1,725 130 2,352 Source: U.S. DOE 1996, vol. 2, Appendix A6, Table 2-1 Note: Values may differ to roundoff error B-4 ------- The contaminated aluminum scrap from future decommissioning activities at facilities other than the diffusion plants—766 t—is assumed to be released uniformly over the period 2016 to 2022. The availability of potentially recyclable aluminum scrap from DOE facilities is summarized in Table B-2. Clearly, any aluminum scrap recycling scenarios will be dominated by scrap from DOE facilities rather than from nuclear power plants. The maximum amount of scrap available in any year is 7237 t, which is the expected inventory by the year 2003. The largest source of this material is the K-25 plant. B.I.2 Radionuclide Inventory As noted above, about 98% of the aluminum scrap from the DOE complex will be generated from the decommissioning of the gaseous diffusion plants at Portsmouth, Paducah, and Oak Ridge. The radioactive contamination of these materials is attributed to a limited suite of radionuclides. The predominant contaminants are isotopes of uranium and their radioactive progenies. Smaller amounts of Tc-99 and trace quantities of Pu-239 and Np-237 may also be present. Indicated contamination levels for aluminum scrap metal items in inventory at the diffusion plants are as follows (U.S. DOE 1986): • U/U-235 <500 ppm • Tc-99 <10 ppm • Np-237 <0.05 ppb • Pu-239 <0.05 ppb • Th <500 ppm It has been estimated that the following radionuclide inventories were fed to the Paducah GDP (National Research Council 1996, Appendix E): • U-236 900 Ci •Tc-99 11,200 Ci •Np-237 13 Ci • Pu-239 20 Ci • Th-230+D 140 Ci •Pa-231+D 16 Ci B-5 ------- Much of this activity was removed during the cascade upgrade and improvement programs. Recent studies have shown that, for the cast aluminum compressor blades used in the diffusion plants, much of the contamination is internal, caused by UF6 entering surface-connected voids (Compere et al. 1996). The UF6 hydrolyzes to UO2F2 (National Research Council 1996). B.2 RECYCLING OF ALUMINUM SCRAP B.2.1 Secondary Aluminum Secondary aluminum, or the aluminum recovered from scrap, has become an important component of the supply/demand relationship in the United States. The industry's recycling operations, commonly referred to as the "secondary aluminum industry," use purchased scrap as "raw" material. Purchased aluminum scrap is classified as "new" (manufacturing) scrap and "old" scrap (discarded aluminum products). In 1996, metal recovered from both new and old scrap reached an historic high of approximately 3.3 million tons, according to data derived by the U.S. Geological Survey from its "Aluminum Scrap" survey of 90 U.S. companies and/or plants (Plunkert 1997a). Fifty-three percent of this recovered metal came from new scrap and 47% from old scrap. The predominant type of purchased scrap was aluminum used beverage container (UBC) scrap, accounting for more than one-half of the old scrap consumed. Aluminum recovered from scrap has increased tenfold since 1950. The recovery of aluminum from old scrap has shown an even more rapid expansion over the same period of time. Increased costs for energy and growing concerns over waste management have provided the impetus for increased recycling rates. Improvements in recycling technologies and changes in the end-use consumption patterns have also contributed to the increase in aluminum scrap recovery. B.2.2 Composition of Scrap Aluminum Aluminum scrap enters the supply stream of the secondary aluminum industry through two major, broadly classified sources: (1) new scrap, generated by the fabrication of aluminum products, and (2) old scrap, which becomes available when consumer products have reached the end of their economic life and have been discarded. New scrap includes solids, such as new casting scrap, clippings or cuttings of new sheet, rod, wire and cable, borings and turnings from B-6 ------- machining operations; residues (e.g. drosses, skimmings, spillings, and sweepings); and surplus products (mill products and castings). Old scrap includes products such as automobiles, aluminum windows/doors/siding, used beverage cans, and cooking utensils. Obsolete industrial products, such as transmission cables, aircraft, and other similar items; outdated inventory materials; production overruns; out-of-specification products; etc., are also classified as old scrap. Aluminum alloys are divided into two distinct categories according to how they are formed: cast alloys and wrought alloys. Controlling the composition of aluminum recovered from scrap is essential to producing marketable secondary alloys. Cast alloys are those specially formulated to flow into a sand or permanent mold, to be die cast, or to be cast by any other process into the final form for end use. Wrought alloys are alloys that have been mechanically worked after casting. The "wrought" category is broad, since aluminum can be formed by virtually every known process. Wrought forms include sheet and plate, foil, extrusions, bar and rod, wire, forgings, and tubing. The application or end product use of the aluminum determines which of these two major alloy categories is employed for the product. Application requirements determine the specific alloying elements and proportions of each element present in the product. The mix of alloys recovered in aluminum scrap at a given time varies depending on (1) patterns of use and discard of these products, (2) the collection systems that act to intercept the discarded waste materials, (3) the separation efficiency with regard to control of scrap shape and size, and (4) degree of processing required to remove certain contaminants. New industrial scrap, assuming proper segregation and identification, can be melted with minimal corrective additions. The processing of post consumer scrap, on the other hand, is much more difficult to predict because the scrap has a variable composition. B.3 STRUCTURE OF THE SCRAP INDUSTRY Aluminum scrap is handled by both major segments of the aluminum industry: (1) the primary producers (integrated aluminum companies), and (2) independent secondary producers. The primary producers recover aluminum from bauxite ore via an electrolytic process in cells or "pots." Such large pot-line plants are devoted to the production of ingots alloyed to particular B-7 ------- specifications necessary for fabrication of various products. The primary aluminum production plants do not recycle any outside material; however, an integrated aluminum company will utilize scrap aluminum feed in other facilities, separate from the primary pot-line plant. In general, the primary producers practice recycle, mostly for UBC's, in large reverberatory smelters. They also recycle "new" scrap from their customers in very large smelters, and return the particular product to their customers. Such plants are not suitable for a feed scrap stream having many different alloy compositions since, if the smelter produced an "off-spec" material, the rework of very large smelter volumes makes such an event very costly. Primary producers consumed 2,180,000 t of old and new scrap in 1995, as summarized in Table B-4. Table B-4. U.S. Consumption of Aluminum Scrap by Primary Producers, Foundries, Independent Mill Fabricators and Others in 1995 (t) NEW SCRAP Solids Borings and turnings Dross and skimmings Other3 Total New Scrap 783,000 31,600 15,900 198,000 1,028,500 OLD SCRAP Castings, sheet, clippings Aluminum-copper radiators Aluminum cans Other" Total Old Scrap Sweated Pig Grand Total 329,000 2,710 799,000 14,200 1,144,910 10,300 2,183,710 Includes foil, can stock clippings and other miscellaneous. Includes municipal waste and fragmented auto shredder scrap. In 1996, about 15.5% of all scrap processed by the primary and secondary smelters (567,000 t) was handled under tolling arrangements where the smelter remelts the scrap and returns it to the supplier (Plunkert 1997a). B-8 ------- A great variety of feed compositions are now handled by the independent secondary producers and it can be expected that recycle of decontaminated material, being diverse in alloy composition, will go to these producers, with their smaller smelters and experience with varying feeds. B.4 SECONDARY ALUMINUM INDUSTRY The secondary aluminum industry comprises those firms which melt aluminum scrap and manufacture various mill products which are sold to foundries and fabricators. In 1995, secondary aluminum smelters consumed 1,300,000 t of purchased new and old aluminum scrap and recovered 1,050,000 t of metal containing 978,000 t of aluminum (Plunkert 1996). The sources of this scrap are summarized in Table B-5. Table B-5 U.S. Consumption of Purchased Old and New Scrap by Secondary Smelters in 1995 (t). NEW SCRAP Solids Borings and Turnings Dross and Skimmings Other3 Total New Scrap 177,000 204,000 208,000 207,000 796,000 OLD SCRAP Castings, Sheet, Clippings Aluminum-Copper Radiators Aluminum Cansb Other0 Total Old Scrap Sweated Pig Total Secondary Smelters 324,000 10,200 118,000 44,500 496,700 4,340 1,297,040 a Includes data on foil, can stock clippings, and other miscellaneous. Includes UBCs toll treated for primary producers c Includes municipal waste (includes litter) and fragmented scrap (auto shredder) B-9 ------- According to a recent EPA report, the secondary aluminum industry operates about 68 plants4 and employs about 3,600 (U.S. EPA 1995). Another source states that the North American industry involves 46 companies with 81 smelting operations (Novell! 1997). A major product of the secondary smelters is feed stock for production of aluminum castings. Aluminum casting alloys are tolerant to a variety of alloying elements, so mixed scrap can be used. If the scrap is carefully segregated, wrought alloys with less tolerance to impurities can be produced. It is this segment of the industry which is of primary interest to the present analysis, since it is the segment which processes a wide variety of scrap materials and typically utilizes nearly 100% scrap in the recycle operation. In practice, secondary smelter sourcing, processing, and marketing can be highly complex. Illustrative of this are the operations at IMCO Recycling Inc.—a publicly-owned company broadly involved in aluminum recycling. In 1996, IMCO had available 1,575 million pounds of aluminum recycling capacity at nine facilities and experienced a 92% operating rate. Scrap materials recycled included dross, used beverage cans, post- consumer and commercial scrap, and new scrap from manufacture of cans and other products. About one-half of the material was from the beverage can and packaging industry; the balance was from transportation and construction market sectors. The product mix was 40% for cans and packaging, 27% for construction and 23% for transportation. The balance was supplied to the steel industry and miscellaneous customers. In 1997, IMCO expected that 90% of production would involve tolling arrangements for customer-owned materials while the remainder would be based on buy/sell transaction which involve purchase of scrap aluminum on the open market, and then processing and selling it (IMCO 1997). In contrast, Wabash Alloys, which has five U.S. smelters and one in Canada, purchases all of its scrap from the open market and mainly produces casting alloys which are sold to the automotive industry (Viland 1990). A flow diagram for typical secondary smelter processing is shown in Figure B-l. B.4.1 Scrap Handling and Preparation Scrap is purchased for a given facility from hundreds of brokers and dealers. In contrast to carbon steel, shipping costs are not a major factor in the aluminum scrap market. Imported aluminum scrap is sometimes used by secondary smelters under favorable market conditions. This total probably includes plants dedicated to UBC remelting. B-10 ------- f FINISHED PRODUCT: I 1 AL ALLOY I Figure B-l. Typical Secondary Aluminum Smelter Flow Diagram (after Viland 1990) Scrap is generally shipped to secondary smelters in trucks with 45,000-lb (20 t) capacity. Rail shipment is also used. Scrap yard operations are illustrated in Figure B-2. As indicated in Figure B-l, crushing (or shredding) may be required for size reduction prior to melting. A shredder at a secondary aluminum smelter is shown in Figure B-3. During the sizing operation, discrete iron contaminants are magnetically separated. The scrap may be dried to remove moisture and organic contaminants such as cutting oils and plastics. Rotary kilns with baghouse dust collection systems are often used for this operation. Some smelters have fixed radiation detection systems installed to monitor incoming and outgoing materials for radioactive contamination, some use hand-held detectors, and some do not monitor but rather rely on their suppliers to ensure against inadvertent contamination. Potash (KC1), a fluxing agent, can trigger radiation detection systems due to naturally-occurring K-40. Occasionally, a small scrap dealer may melt some of the scrap into ingot for sale to a larger scrap dealer if the economics are appropriate (i.e., the value of the remelt ingots is greater than the value of the unprocessed scrap plus the cost of melting the scrap into ingots). Such an operation might involve a small gas-fired pot furnace with a fume collection hood which vents to the atmosphere. During operation at such a facility, an americium source was inadvertently melted. The incident was detected when the ingot was delivered to a larger dealer with radiation B-ll ------- Figure B-2. Handling of Scrap Turnings from Forged Aluminum Auto Wheels at IMCO's Uhrichville OH Plant (IMCO 1997) monitoring equipment. App arently, cleanup after the incident was reasonably straightforward in that most of the Am remained with the aluminum and was not spread around the facility (Mobley 1999). A description of the features of several secondary smelters is included in Ap p endix B -1. Ap p endix B -2 provides a detailed description of the secondary smelter operations at Arkansas Aluminum Alloy Inc. in Hot Springs (Kiefer et al. 1995). Figure B-3. Scrap Shredder at Secondary Aluminum Smelter B-12 Continue ------- Back B.4.2 Melting Practice Melting for general scrap recovery is done almost exclusively in gas- or oil-fired reverberatory furnaces, typically of 40,000 to 220,000-lb (18 to 40 t) capacity (Viland 1990). Halide salts (such as mixtures of NaCl, KC1, and NaF) are added to form a cover over the melt and reduce oxidation. For casting alloys, Si (2% to 13%) is added in secondary smelting process to promote casting alloy fluidity. (Silicon also imparts other desirable properties such as wear resistance.) Die casting alloys generally can accept higher limits on Fe, Mn, Cu, Zn, and Cr. For corrosion resistance (e.g., outboard motors), copper limits "are greatly reduced." Permanent mold and sand casting alloys must have reduced Fe levels to improve ductility (Viland 1990). The melting cycle for a typical reverberatory furnace consists of charging scrap into the forewell of the furnace, blending and mixing alloying materials, addition of fluxing salts, magnesium removal, gas removal, skimming off the dross, and pouring. A heel consisting of 20 to 40% of the furnace capacity is generally left in the furnace to shorten the melting cycle (Plunkert 1995). Scrap is charged into the furnace, either with a front-end loader or a belt conveyor, over a 16- to 18-hour period. Magnesium and gas removal require two to four hours and tapping requires an additional three to four hours resulting in a total cycle of about 24 hours. According to Crepeau et al. (1992), dressing fluxes typically constitute about 0.2% to 1% of the metal charged5. Use of NaF in the flux will add traces of Na to the melt; K2TiF6 can be used to add Ti, and KBF4 can be used to add B. A1F3 will tend to remove Ca, Sr, and Mg, while chlorine-releasing compounds promote removal of Mg, Na, and Sr. Phosphorus can be added to the melt via flux containing amorphous phosphorus. Prior to tapping the furnace, the melt is typically treated with chlorine gas to reduce magnesium to acceptable levels6. During this "demagging" process, other metallic impurities which form chlorides more stable than A1C13 are also removed from the melt and transferred to the dross. Hydrogen is also removed but, for that impurity, removal is by solubility in the C12 gas rather than by HC1 formation. It should be noted that this is the amount of flux charged not the amount of dross produced, the latter being much higher. Magnesium is not undesirable in all alloys. Some aluminum alloys contain up to 10% Mg. B-13 ------- Neff notes that alkali and alkaline earth metals such as Li, Na, K, and Ca can be removed from aluminum either by chlorine injection of pot-line vessels or in-line degassers (Neff 1991). Furnace output is typically cast into ingots or sometimes into sows (1,000-lb cast blocks). In North America, about 500 million Ib/year is shipped in liquid form in crucibles via trucks (Viland 1990). Truck shipment of molten aluminum is shown in Figure B-4. Figure B-4. Aluminum Liquid Metal Transporter During the melting cycle, dross is skimmed from the melt surface and collected in containers adjacent to the furnace. Dross is processed to recover the contained aluminum by physical separation using hammer mills or by melting in rotary salt furnaces. Some secondary smelters use rotary furnaces, particularly for the processing of low-grade or light scrap. "For every 1 million pounds of scrap processed, 760,000 pounds of secondary aluminum is produced, and 240,000 pounds of dross residues, and 3,000 pounds of baghouse dusts are generated. The dross residues are not hazardous but contain salts and are generally disposed of in solid waste landfills" (Viland 1990). Salt recovery systems have not been very successful because of the extremely corrosive nature of the salts. Baghouse dusts may contain Cd and Pb above the limits of the EPA Toxicity Characteristics Leaching Procedure (TCLP) test. In many cases, these dusts are disposed of in hazardous waste landfills. B-14 ------- B.4.3 Dust Handling Not all secondary aluminum smelters use baghouse dust collection systems. Those that do may not process all of the furnace offgas through the baghouse. For example, at one smelter, each furnace has a canopy exhaust system which is connected to a baghouse for dust collection. About 40% of the flue gases is also exhausted through the baghouse to maintain the gas temperature above its dew point. Condensation of halides can cause severe corrosion problems in the exhaust system. The balance of the flue gases is exhausted directly through the stack. The baghouse has eight modules. Lime-coated bags are used because of the acidic nature of the offgas. Dust collected from blowdown is accumulated in the baghouse hoppers and transported via screw conveyors to reinforced plastic bags attached to the ends of the enclosed conveyors. The filled plastic bags are temporarily held in a nearby commercial steel dumpster and ultimately taken by the disposal contractor to an approved municipal landfill. A maintenance operator typically spends about one hour per day in the baghouse area. The fabric filter bags are replaced every two years. Although some hazardous volatiles accumulate in the dust, the collected waste at this smelter meets EPA TCLP requirements. (TCLP results are summarized in Table B-6.) Cadmium in the dust may come from paint while multiple sources of lead are possible. Comparison of the crusher fines and the furnace dust data suggests that the furnace dust is enriched in the volatile elements Cd and Hg and depleted in Ba and Cr. Some data on airborne dust concentrations have been obtained from a small aluminum foundry where three electric furnaces were used to melt aluminum under chloride/fluoride fluxes. The molten aluminum was transferred to a ladle and then poured into steel molds (Michaud et al. 1996). Dust samples were collected at fixed sampling locations: between two of the furnaces, near the core maker, next to a mold, and in the middle of the foundry room. The average total dust concentration was 2.5 mg/m3 and the respirable concentration was 1.1 mg/m3. The respirable fraction, as defined by the American Council of Governmental and Industrial Hygienists (ACGIH 1996), has a range of particle aerodynamic diameters (AD) with a median value of 4 |im. The total dust concentration included an average of 0.05 mg/m3 of Al and 0.03 mg/m3 of Mg. Using SIMS and XPS analytical probe techniques, Ca and Si were found to be associated with the coarse fraction (i.e., >4|im AD) and S, Zn and Cl were concentrated in the fine particles. Na, K, Al and C exhibited higher intensities in the fine fraction (i.e., <1 |lm AD) than in the coarse fraction. Fluorine was strongly detected in all size fractions. B-15 ------- Table B-6. TCLP Values for Dust Samples and Spent Refractory (mg/L) Element As Ba Cd Cr Pb Hg Se Ag TCLP Limits 5 100 1 5 5 0.2 1 5 Furnace Dust <0.70 0.42 0.08 <0.010 <0.2 0.003 <0.7 <0.01 Crusher Fines <0.70 0.78 0.023 0.023 <0.2 <0.0004 <0.7 <0.01 Spent Refractorya <0.70 1.2 0.054 0.87 <0.2 <0.0004 <0.7 <0.01 a Solid material, not dust Additional dust sampling results are available from a NIOSH study at the Arkansas Aluminum Alloys Inc. smelter which uses three 220,000-lb (100 t) reverberatory furnaces (Keifer, et al. 1995). Prior to the referenced study, area samples collected in 1992 showed respirable dust concentrations of 2.3 mg/m3 near furnace #2 and 4.4 mg/m3 near furnace #4. Earlier samples taken in 1989 found 12.17 mg/m3 of total dust at the scrap conveyor and 15.38 mg/m3 of total dust at the baghouse. In the referenced 1995 study, NIOSH took samples in a variety of locations that were analyzed for total dust and component metals. Details, including time-weighted average (TWA) concentrations, are presented in Table B-7. No Cr, Pb, nor Ni was detected in the samples collected. In two samples, Cd was reported between the analytical detection limit and the limit of quantification. Although not so stated by the authors, other values in the table, which appear in parentheses, presumably fall within the same range—i.e., measurements were made, but the values are so low as to be suspect. B.4.4 Partitioning of Contaminants B.4.4.1 Thermochemical Considerations This section examines the expected partitioning of contaminants during the melting process. As noted above, the primary radioactive contaminants in DOE aluminum scrap are expected to be U, Tc, Np, Th, and Pu. Some of these elements may be transferred to the dross during the demagging operation, depending on the relative thermodynamic stability of the respective chloride species. B-16 ------- Table B-7. Secondary Aluminum Smelter Dust Levels Activity Sampled Skimming/pouring - Furnace #2 Skimming/pouring - Furnace #2 Furnace #4 operator - South side Furnace #4 operator - North side Furnace #2 operator - South side Furnace #2 operator - North side General area - sweeping/cleaning Pouring area - sweeping/cleaning Sampling Time (min) 366 364 463 480 486 420 118 125 Total Dust (mg/m3) 0.45 0.26 0.64 0.46 0.62 0.55 0.60 3.24 TWAa Concentration (|jg/m3) Al 40 18 57 12 50 37 27 370 Zn 0.1 3.1 2.3 1.9 7.7 Cd (0.2)b (0.3) Mg (2.9) (2.4) 5.5 (2.4) 4.8 8.8 (5.1) 12 Mn 0.16 0.2 0.44 (0.1) 5.2 Fe 6.3 2.5 19 4.0 10 12 8.9 49 Cu 0.8 0.4 1.5 0.6 1.5 1.4 2.0 1.9 Ti 1.6 0.6 0.8 0.18 0.48 0.4 0.9 21 a Time-weighted average b Values in parentheses are assumed to be less than the lower quantification limit Representative values for the free energy of formation for the following reaction at 1000 K (a typical pouring temperature for aluminum) are presented in Table B-8. - M + C12 = - MXC1 y y y Assuming that the above equation represents the governing chemistry, that equilibrium is obtained and that the dilute solutions behave as pure substances, it is assumed that all the elements below A1C13 in Table B-8 will be transferred to the dross and that those above A1C13 will tend to remain with the aluminum. Hydrogen (tritium) should also be substantially but not totally removed from the melt and released to the atmosphere. As noted previously, hydrogen removal is by solution in the chlorine rather than by HC1 formation, which is thermodynamically unfavorable. Thermodynamic equilibria based on pure substances suggest that solute elements with standard free energies of formation of the solute metal chlorides higher (less negative) than B-17 ------- that of A1C13 will remain in the melt. However, there is virtually no information available on activity coefficients for the same substances in dilute solutions. Thus, the thermochemical calculations in Table B-8 provide only rough guidelines as to the expected partitioning during melting. It may be noted from Table B-8 that if protactinium is in the +5 valence state, it would be expected to remain in the melt but if it is in the +3 valence state it would be expected to partition to the dross. However, any pentavalent chloride which forms would be reduced by aluminum, so Pa should partition to the dross. Many chlorides are volatile at low temperatures and this attribute may play a role in the partitioning process. Addition of chlorine to the melt for demagging and hydrogen removal might result in the formation of volatile chlorides. Selected metal chlorides with boiling points below the melting point of aluminum are listed in Table B-9. The gas volumes passing through the liquid metal and the liquid flux can be large and three interactive partitioning mechanisms are possible—between the gas and the metal, between the gas and the flux, and between the flux and the metal. As suggested by Table B-9, many chlorides will have a perceptible vapor pressure at 1000 K and can be transferred from the melt to the gas. Some of these displaced chlorides will terminate in the dross and some in the fume which will either condense on the ducting or in the baghouse. Removal of a portion of the iron and silicon, but not copper, has been observed during the treatment of aluminum melts with C12 in the laboratory. Iron and silicon chlorides condensed on the walls of the system ducting. The partitioning mechanism was not elucidated but may involve small partial pressures of the solute metal chlorides in a volatile aluminum chloride. The gaseous aluminum chloride is dense and is not transported a significant distance in the offgas system. These experiments involved large quantities of flux and highly specialized melting practices not representative of those expected in a secondary aluminum smelter. In a typical smelting operation, impurities such as iron are not preferentially removed. Iron, Sb, Ce, Co, Nb, Sr, Th, and U have no reported solubility in molten aluminum; rather, they form intermetallic compounds which are in equilibrium with pure aluminum (Davis 1993). Thus, volatile chloride formation would require a reaction between chlorine and, say, UA14, rather than between chlorine and uranium dissolved in the aluminum. If a volatile chloride did form with an impurity less stable (per Table B-8) than A1C13, it would most likely be immediately reduced before it could exit the melt. B-18 ------- Table B-8. Standard Free Energy of Formation (AF°) for Various Metal Chlorides at 1,000 K Metal Chloride RuCl3 MoCl6 TcCl3 NbCl5 PbCl4 MC12 AgCl CuCl SbCl3 CoCl2 HC1 FeCl2 SiCl4 ZnCl2 MnCl2 PaCl5 A1C13 UC13 NpCl3 MgCl2 ThCl3 PuCl3 PaCl3 AmCl3 SrCl2 CsCl -AF° (Kcal/g-atom Cl) decomposes at 900 K 3.23 7.37 11.4 18.6 18.8 19.1 20.9 21.2 22.4 23.9 26.6 27.9 32.2 40.1 41.3 45.5 53.5 55.2 57.4 58.9 59.4 63.9 66.6 82.6 83.0 The possibility also exists that some elements expected to be transferred to the dross would also volatilize to some extent and either condense on the ducting or be collected in the baghouse dust. Based on Tables B-8 and B-9, uranium might be expected to exhibit such behavior. B-19 ------- Table B-9. Selected Metal Chlorides with Boiling Points Below 1000 K Metal Chloride A1C13 FeCl3 MoCl6 MnCl3 MnCl4 NbCl5 PaCl5 PbCl4 SbCl3 SiCl4 TcCl5 UC15 UC16 Boiling Point (K) 453 (sublimes) 592 630 900 384 519 659 400 492 330 505 690 550 Source: Glassner 1957 Note: no information available on chlorides of Eu and Pm While the simple free energy calculations presented in Table B-8 suggest that any U, Th, Pu, or Np dissolved in an aluminum melt will be removed by chlorine during the demagging process, the radioactive contaminants may be in the form of oxides. It is not clear whether such oxides will be either reduced by aluminum or converted to the halide form. For example, the thermodynamics are unfavorable for converting UO2 to either a fluoride or chloride at 1,000 K. In addition, the free energy change for the reaction between UO2 and Al to form A12O3 and U is about zero at 1,000 K, suggesting that this reaction is also unlikely to proceed. However, as will be discussed in Section B.4.4.2, formation of uranium-aluminum intermetallics has been observed. B.4.4.2 Observed Partitioning The partitioning of uranium in aluminum melts has been experimentally measured by Copeland and Heestand (1980). In this work, aluminum melts were equilibrated with a slag of unspecified composition containing 0.3 wt% uranium at 973 K and the uranium pickup by the aluminum was measured. Based on this type of laboratory measurement, the partition ratio—defined as the concentration of the uranium in the slag to the concentration in the metal—was determined to be B-20 ------- 190. The experimental results, which suggest that some decontamination of the melt will occur, are in contrast to thermodynamic calculations made by these authors for an oxide system which suggested a value on the order of 10"3 for the partition ratio7. In another set of experiments, these authors prealloyed uranium with aluminum and found that the partition ratio was only 2 to 3, as compared to 190 when uranium-containing slag was equilibrated with the molten aluminum. Copeland and Heestand also examined drip melting, where surface-contaminated aluminum was placed on a metal screen and then heated to above the melting point. The molten aluminum dripped through the screen to a crucible below while the dross remained on the screen. In this experiment, the metal contained 16 ppm U while the dross contained 2,100 ppm U. When the drip melting process was scaled to multi-kilogram size ingots, the separation was less effective, with 4 ppm U in the aluminum and 25 to 75 ppm in the dross. Heshmatpour and Copeland (1981) described additional laboratory measurements of uranium partitioning during aluminum melting. In these experiments, 500 ppm of UO2 was added to aluminum, and the melts were held at 1,573 K under various slags. Experimental results are summarized in Table B-10. While the results generally show some preferential partitioning of uranium to the slag, there are some results which appear anomalous. Sample 5 shows very little decontamination even though companion tests (samples 3 and 4) with slightly different fluxes show much higher partition ratios. The flux compositions used for samples 1 and 18 are significantly different than would be expected in commercial secondary smelting. Except for sample 5, the uranium content of the melt ranged from about 1 to 100 ppm when halide or cryolite-type fluxes were used. It should also be noted that all of these tests were conducted at a substantially higher temperature than used in commercial secondary smelting. It is not clear from this work what effect the higher temperature has on the partition ratios. However, a study by Uda et al. (1986) showed that the residual uranium content in aluminum melts doped with 500 ppm U increased as the melting temperature increased. The melting was conducted under a flux of 14% LiF-76% KCl-10% BaCl2 and the mass of the flux was 10% of This partition ratio is based on the reaction of uranium in the aluminum melt with A12O3 in the slag to produce UO2 in the slag. The calculation assumes that the weight of the slag is 10% that of the melt, that the thermodynamic activity of A12O3 in the slag is 0.1, that the activity of UO2 in the slag is 0.01, and that the Henry's Law constants for U in the aluminum melt and UO2 in the slag are unity. B-21 ------- that of the metal charge. The residual uranium content of alloy 5083, containing 4.45% Mg, increased from about 1 ppm at 800°C to about 10 ppm at 1000°C. For alloy 1050 (99.5%A1), the residual uranium content increased from about 20 ppm to about 70 ppm over the same temperature range. The experimental program showed that the uranium removal increased exponentially with increasing magnesium content in the aluminum. Table B-10. Partitioning of Uranium in Aluminum Melts in Zirconia Crucibles at 1573 K Sample 1 2 3 4 5 6 7 8 9 18 Metal (9) 76 81 81 80 78 50 50 166 503 250 Flux (9) 7.6 8.1 8.1 8.0 7.8 0 0 8.3 25.15 25 U02 (ppm) 500 500 500 500 500 500 500 500 500 500 Uranium (ppm) Metal 1.2 111 0.9 2.4 315 469 430 31.4 81.1 308 Slag 9610 1360 405 570 150 1760 4190 255 Partition Ratio3 801 1.2 45 24 0.05 3 3 0.08 Flux (%) AIF3 AI203 CaF2 100 60 40 20 CaO Fe203 NaF 100 40 60 80 Si02 No flux No flux 35 35 10 10 10 5 50 5 55 55 30 a Amount of contaminant in the slag divided by amount of contaminant in metal The experimental observation that uranium removal from aluminum increases as the temperature decreases is opposite of that which is predicted from the calculated equilibrium constant for the reaction: UO, + -Al = U + -ALO, 2 3 3 2 3 No satisfactory explanation was provided by the authors for the difference between the experimental observations and the thermodynamic calculations. The increased uranium removal associated with higher magnesium content is attributed to the formation of strong intermetallic compounds between Al and Mg which reduce the ability of the aluminum to reduce the UO2. This argument appears specious since all of the aluminum is not tied up as intermetallics. In a subsequent paper, Uda et al. (1987) described the electroslag melting of aluminum alloy 5052 under a flux of 14% LiF, 76% KC1 and 10% BaCl2. The aluminum alloy electrode was contaminated by drying a solution of known uranium concentration on the surface. The amount B-22 ------- of uranium was such that the concentration in the finished ingot would be 500 ppm if none were lost to the slag or elsewhere. The actual uranium concentration in the finished ingot was 3 to 5 ppm. Insufficient information is provided by the authors to calculate a partition ratio. Mautz et al. (1975) described the results of melting some aluminum scrap from the Portsmouth gaseous diffusion plant in a oil-fired reverberatory furnace of unspecified size. Fluxing agents were not used. The aluminum scrap consisted of die-cast, wrought, and cast parts which had extended exposure to UF6. The scrap was chemically decontaminated prior to melting. Sixty- two ingots from die cast scrap contained residual uranium ranging from a minimum of 0 to 100 to a maximum of 1300 to 1400 ppm. (Since bar charts rather than actual data were provided by the authors, only ranges for the minimum and maximum could be determined.) Ingots produced from cast and wrought scrap were generally lower in uranium than ingots produced from die-cast scrap. Some experimental work has shown that UO2 can react with Al in the solid state at temperatures of 873 K to form various intermetallic compounds such as UA12, UA13, and UAl4(Waugh 1959). Reaction between UO2 and Al to form UA1X and A12O3 was 90% to 100% complete in 10 hours. The U-A1 binary phase diagram predicts that the equilibrium phases formed during the solidification of melts containing small quantities of uranium should be UA14 (or U09A14) and aluminum (Davis 1993). If the same reaction occurs in the liquid state, it would tend to promote partitioning of the uranium to the melt (as UA1X) rather than to the slag (as UO2). Heshmatpour et al. (1983) described one experiment where 500 ppm of PuO2 was melted with 100 g of Al at 800°C without any flux. The solidified sample contained 5.4 ppm Pu while the surface Pu concentration was 18,300 ppm. These results suggest that if plutonium is present as the oxide it is likely that most of it will be removed with the dross. As noted under B.4.4.1 above, oxide, as well as chloride, reactions can occur between elements and compounds in the melt and in the slag. Hryn et al. (1995) have measured the cation content of the oxide residue of dross generated by melting series 3XX aluminum casting alloys. (These oxide residues were byproducts of the process of aluminum recovery from the dross.) The results are summarized in Table B-l 1. These measurements indicate that some of the metals which would be predicted to partition to the melt on the basis on Table B-8 are also found in the dross. These include silicon, zinc, copper, manganese, and iron. B-23 ------- Table B-l 1. Cation Impurities in 3XX Aluminum Residue-Oxide Samples Element Mg Si Ca Ti Zn Mn Fe Cu 3XX Residue-Oxide (%) 4.7 5.3 1.4 0.3 0.3 0.14 1.5 0.5 B.4.4.3 BaghouseDust As noted earlier, not all secondary aluminum smelters use baghouse dust collection systems. Some of those that do may collect only a portion of the offgas and pass it through the baghouse. Limited data are available to predict the partitioning of particular elements to the dust. As part of the EPA program to develop an air emissions standard for secondary aluminum smelters, some measurements have been made of the composition of the dusts based on stack samples. During the standards development program, two sets of particulate samples were taken from a furnace at the Alcan Recycling Facility in Berea, Ky. (U.S. EPA 1990). No information was provided on the composition of the metal being melted, so it is not possible to develop a detailed estimate of the how the various elements partition to the dust. However, if one assumes that the material being melted in alloy 3004—the standard material used for the aluminum can bodies (Davis 1993)—some insight into partitioning can be derived. Table B-12 compares the composition of alloy 3004 with the furnace particulate matter. From this table it can be seen that the particulates are enriched in magnesium and iron, depleted in manganese and essentially unchanged in zinc. Small quantities of other elements including Sb, Ba, Co, Pb, and Ni, were also found in the particulate matter. The limited information available does not suggest that particular elements have orders of magnitude concentration increases in the dust. Consequently, it is assumed that the dust has the same composition as the scrap with regard to metallic elements. Any particulates released to the atmosphere are also assumed to have the same metallic composition as the scrap. B-24 ------- Table B-12. Composition of Particulate Matter From Secondary Aluminum Smelter Element Al As Ba Cd Co Cr Fe Hg Mg Mn Ni Pb Sb Se Ti Zn Alloy 3004 (%) a a a a a 0.70 max. a 0.8 to 1.3 1.0 to 1.5 a a a a a 0.25 max Alcan Furnace (Run 1) Ib/hr 3.19e+00 5.07e-04 1.69e-02 2.11e-04 4.22e-04 1.44e-03 6.36e-02 8.45e-05 9.38e-01 5.07e-04 <1.69e-03 1.69e-03 3.38e-03 1.69e-04 <6.76e-02 1.02e-02 % 0.016 0.53 0.0066 0.013 0.045 2.0 0.0026 29 0.016 <0.053 0.052 0.11 0.0052 <2.1 0.32 Alcan Furnace (Run 2) Ib/hr 6.79e-01 <2.10e-04 <7.00e-03 7.00e-05 <2.80e-05 6.30e-04 3.54e-02 7.00e-05 9.38e-01 1.05e-03 <1.40e-03 7.00e-04 2.80e-03 1.40e-04 <5.60e-02 1.89e-03 % <0.032 <1.0 0.010 O.0041 0.093 5.2 0.010 138 0.15 <0.21 0.10 0.41 0.021 <8.2 0.28 a All other elements limited to 0.05% max. and 0.15% total B.4.4.4 Proposed Partitioning Based on the information presented here, coupled with technical judgement, the suggested partitioning ratios for the various elements between melt, dross, baghouse dust, and the atmosphere are summarized in Table B-13. Since the data are limited and conflicting, ranges are proposed in many cases. In the case of the uranium partition ratio, the very low and very high values in Table B-10 were discarded and it was assumed that the partition ratio could vary from 1 to 100. In the absence of other information and based on the assumption of similar chemical and thermodynamic behavior, this same range was assigned to Ac, Am, Ce, Eu, Np, Pa, Pm, Pu, Ra, and Th. The possibility also exists that some uranium which partitions to the dross could volatilize and collect in the baghouse dust. Where no experimental evidence exists to the contrary, partitioning is assumed to follow predictions based on the thermodynamic calculations in Table B-8 (e.g., Cs, and Ag). In some instances the calculations in Table B-8 were tempered B-25 ------- by the observations on oxides in the dross included in Table B-l 1. In applying the data in Table B-l 1, Ni and Co were assumed to be analogous to Fe and Nb to be analogous to Ti. Additional comments on various alloying elements are summarized below (Davis 1993): • silver has substantial solubility in both liquid and solid aluminum • lead has very limited solubility in both liquid aluminum (0.2 at%) and solid aluminum (0.02 at%) but lead is sometimes added to certain alloys to improve machinability • carbon is occasionally found in aluminum as an oxycarbide or a carbide (A14C3), although fluxing operations usually reduce C to the ppm level • antimony is present in trace amounts in primary commercial-grade aluminum and is used as an alloying element in certain aluminum alloys • cobalt has been added to some Al-Si alloys containing iron to improve strength and ductility • cerium has been added to experimental casting alloys to increase fluidity and reduce die sticking • manganese is a common impurity in primary aluminum and is a frequently used alloying additive • strontium is found in trace amounts in (0.01 to 0.1 ppm) in commercial aluminum • molybdenum is a low level impurity in aluminum (0.1 to 1 ppm) and has been added as a grain refiner • nickel has limited solubility in aluminum (0.04%) but nickel has been added to Al-Si alloys to increase hardness and strength at elevated temperatures B-26 ------- Table B-13. Proposed Partitioning of Selected Elements During Secondary Aluminum Smelting Element Ac Ag Am C Ce Co Cs Cu Eu Fe 1 Mn Mo Nb Ni Np Pa Pb Pm Pu Ra Ru Sb Si Sr Tc Th U Zn Partition Ratio (PR) (%) Metal 1/50 100 1/50 1/10 1/50 99/90 99/90 1/50 99/90 99/90 100 99/90 99/90 1/50 1/99 100 1/50 1/50 1/50 100 100 99/90 1/10 100 1/50 1/50 99/90 Dross 99/50 99/50 99/90 99/50 1/10 100 1/10 99/50 1/10 50/100 1/10 1/10 1/10 99/50 99/1 99/50 99/50 99/50 1/10 99/90 99/50 99/50 1/10 Baghouse3 Atmos.b 50/0 Comments 1 ------- that some uranium may concentrate in the dust due to condensation of a uranium chloride volatilizing from the slag, but insufficient information is available to quantify this possibility. B.4.5 Dross Processing Significant concentrations (10% - 80%) of aluminum are found in the dross, necessitating reprocessing of this waste stream for maximum metal recovery. One of two techniques is generally used for dross processing: • physical separation • melting in rotary salt furnaces When physical separation is employed, the dross is passed through hammer mills and across screens. The screen oversize, which is rich in aluminum, is returned to the smelting process while the undersize, containing primarily salt and some oxides, is shipped to a landfill. Some landfills may have leachate liners. Dross processing may be done on site or at a dedicated facility. In some cases, the dross is sold to a processor and the recovered aluminum is repurchased. Rotary furnaces produce larger quantities of salt waste (salt cake) which contains relatively small amounts of aluminum as compared to dross. It has been estimated that recovery of aluminum from skim and dross in rotary furnaces generates about 460,000 t of salt cake annually. The salt cake contains 5 to 7 wt% aluminum, 10 to 50 wt% salts, and 30 to 85 wt% residue oxides. The residue oxide is primarily aluminum oxide with minor amounts of cryolite, magnesium oxide, magnesium aluminate, and other contaminants (Graziano et al. 1996). Most of the salt cake is landfilled. Given long-term concerns about landfill availability, processes are being developed to reduce the quantity of salt cake which must be buried. The Ford Motor Company has initiated a process to handle about 11,000 t of aluminum salt cake annually from their foundry in Essex, Ontario. The salt cake will be shipped by Browning Ferris Industries to a facility in Cleveland for processing by the Aluminum Waste Technology, Inc. Aluminum and salt are recovered from the process and sold to secondary smelters, while aluminum oxide is recovered and sold to the steel industry for topping compounds (Wrigley 1995). Aluminum Waste Technology, Inc., is a wholly owned subsidiary of Alumitech, Inc (which is, in turn, owned by Zemex Corporation). Alumitech, Inc. is also seeking other markets for the metallic oxides recovered from the process, which it describes as non-metallic products (NMP). B-28 ------- To further this product strategy, Alumitech has built a metallurgical plant in Cleveland to prepare NMP feedstock for the production of refractory ceramic fiber (Zemex 1998). Calcium aluminate is also recovered as a separate product for use as a steel slag ingredient. Because of European landfill restrictions, dross from Austria is being shipped to Alumitech for processing ("Aluminum Smelters Export" 1995). Graziano et al. (1996) evaluated the economics of various salt cake recycling options. Their base case design was predicated on combining processes that had been commercialized, licensed, or developed by the industry. The base case process is described as follows (see Figure B-5): In the solids preparation section, the salt cake is dry-crushed, screened, and magnetically separated to recover an aluminum-rich, iron-free product for remelting in a secondary aluminum furnace. We assumed that 70% of the aluminum in the salt cake is recovered in this byproduct stream at 50% purity. The effluent from the solids preparation section is salt cake, depleted in aluminum and crushed to 1-mm size, for feed to leaching. In base case process, crushed salt cake from the solids preparation section is fed to a leaching tank, where the salts are dissolved in water at ambient conditions (25°C, 1 atm) to yield a brine concentration of 22 wt% salts. Insolubles (aluminum oxide) in the leach effluent are separated from the brine and washed with water to remove residual salts. The wet oxide is landfilled or further processed for sale. The clarified brine solution is fed into a forced-circulation evaporator system designed for energy recovery (single effect with vapor recompression or multiple effect). The NaCl and KC1 salts crystallize as the water is evaporated. The slurry effluent from the evaporator is then routed to product recovery.... In the product recovery section the salt solids are separated from the brine solutions with a centrifuge and then dried and stored for sale.... The filtrate from the centrifuge is then recycled back to the evaporator to maximize recovery of salts. The gas treatment section is required to control emissions of toxic and explosive gases generated when salt cake is leached in water. According to European sources, hydrogen, ammonia, methane, phosphine, and hydrogen sulfide are emitted from the leaching action. ...the gas treatment section consists of a thermal oxidizer followed by a chlorine scrubber. The authors modeled a plant which processed 30,0001 of salt cake per year with a 90% on- stream factor. The salt cake was assumed to contain 6 wt% Al, 14 wt% NaCl, 14 wt% KC1, and 66 wt% aluminum oxide. Assuming that a 20% return on investment was needed, the base case plant had a negative net present value, indicating lack of economic viability. They also B-29 ------- Salt Cake Feed Make- Up Water Aluminum By-Product Vent Gases Solids Preparation Leaching and Oxide Removal (25C.1 atm) T Wet Oxide Residue Product Salts Gas Treatment Brine Recycle Evaporation Recycle Water Product Recovery Figure B-5. Proposed Salt Cake Recycling Process (Graziano et al. 1996) considered alternative flow sheets involving high temperature leaching of the salt cake followed by flash crystallization, a solvent/anti-solvent process to replace evaporation, and the use of B-30 ------- electrodialysis to replace evaporation. None of these alternatives was economically viable. The base case process could be made more attractive if the scale of the operation were increased and if the aluminum oxide residue were recovered for sale rather than landfilled. Higher landfill costs also improve process economics. However, producing a marketable product would probably require additional processing to meet specifications for selected applications. Graziano et al. (1996) were aware of only three operations in the United States where salt cake recycling was practiced. These included Aluminum Waste Technology (Cleveland), Reynolds Metals Company (Richmond, Va.), and Insamet (Litchfield Park, Ariz). Salt cake recycling is more prevalent in Europe, driven by landfill restrictions. More recently EVICO Recycling Inc. (1998) has described the process at the Litchfield, Ariz. plant, which is 70% owned by EVICO. The plant recycles aluminum scrap and turnings under tolling arrangements. It also processes concentrates from purchased dross and salt cake in a patented wet milling process. The recovered aluminum is melted and sold on the open market. Aluminum oxide, which is a byproduct of the wet-milling process, is sold for use in making Portland cement. The salt will be recovered from evaporation ponds and some will be used as flux in EVICO's aluminum smelting operations. At its Utah facility, IMCO operates a joint venture with Reilly Industries where salt cake is recycled into aluminum concentrates, aluminum oxide, and brine. The brine is transferred to a solar recovery system operated by Reilly Industries. The recovered salts are used for a variety of purposes including fluxes. While salt cake recycling is not widely practiced, the salt cake may be mechanically treated to remove a portion of the residual aluminum prior to landfilling the treated salt cake. Roth (1996) characterizes "standard existing technology" as involving a primary jaw crusher and a high- speed, horizontal shaft, plate-and-breaker-bar impact mill. This system produces a concentrate containing 60 - 70% aluminum from salt cake initially containing 3-10% aluminum. B.4.6 Handling Baghouse Dust Not all furnaces have baghouse dust collection systems. If such systems are used, baghouse dust is shipped to landfills for disposal or buried in landfills on site. The dust may contain lead and consequently stabilizing agents may be added to insure that the product meets the EPA TCLP requirements. Because of the demagging operations, many trace radionuclides will be converted to chloride salts which are non-volatile and will remain with the dross. As such, the potential for B-31 ------- radionuclides to concentrate in the baghouse dust is markedly lower at an aluminum smelter than at an EAF shop where steel is melted. The EPA has recently proposed, under 40 CFR Part 63, to regulate emissions of hazardous air pollutants from secondary aluminum production. The proposed rule requires that particulate emissions be limited to 0.4 Ib/ton and that HC1 emissions be limited to 0.40 Ib/ton (or be reduced by 90%). The proposed standard is based on achievable emissions limitations when melting dirty charge materials with unlimited fluxing and collecting the emissions in a fabric filter baghouse with continuous lime injection. However, the required limits can be achieved with other means, such as improved work practices, reduced flux usage, process design changes, etc. In the proposed standard, total particulates are measured as a surrogate for hazardous particulates and HC1 is measured as a surrogate for HC1, HF, and C12. B.4.7 Product Shipments As noted above, approximately 230,000 t/y of remelted aluminum is shipped in the molten state. This is roughly 7% of all aluminum alloy shipments (based on a calculated metallic recovery of 3,190 million t in 1995 [Plunkert 1997a]). Hot aluminum is shipped in covered crucibles mounted on flatbed trucks (see Figure B-4). The crucible, which is typically made of 1.9-cm (0.75-inch) steel, is lined with approximately 13 cm (five inches) of refractory and contains 13.6 t of molten aluminum (Viland 1997). Haulage distances range from 35 to 250 miles. Hauling distances are limited to those within a five- to six-hour driving range. B.5 PRODUCT MARKETS According to Viland (1990), markets served by secondary smelters are as follows: • Direct automotive 22% • Automotive related 44% • Small engine 8% • Appliance 7% • Other 19% Another perspective on the output of secondary smelters is presented in Table B-14. B-32 ------- The total in Table B-14 is less than that in Table B-5. One reason for the difference is that Table B-14 does not include toll-processed aluminum beverage can stock. In addition, more estimation is involved in developing Table B-14 (Plunkert 1997b). From this table, it can be seen that most of the secondary smelter output is casting alloys. About 17% of the output is extrusion billets used to produce wrought alloys. These wrought alloys are based on new scrap of known, specific chemistry which can be remelted into compositions suitable for extrusion into various mill products (Plunkert 1999). Table B-14 Production of Secondary Aluminum Alloys by Independent U.S. Smelters in 1995 (t) Secondary Product Die-cast Alloys Sand and Permanent Mold Alloys Wrought Alloys: Extrusion Billets Aluminum-base Hardeners Other a Total Less primary feedstocks (Al, Si, other) Net Metallic Recovery Production 619,600 150,400 163,000 5,400 39,600 978,000 120,000 858,000 Source: Plunkert 1996 Includes other die-cast alloys and other miscellaneous. Additional detail on the wide variety of products produced from various aluminum casting alloys is included in Table B-15. In addition to these applications, the steel industry uses about 450 million pounds (205,000 t) of aluminum each year as a deoxidant, and as an ingredient in slag conditioners and desulphurizers. Aluminum is also added to steels as a grain refiner. As an example of how this market is served, IMCO Recycling Inc. has plants in Elyria and Rock Creek, Ohio which process aluminum scrap. At these plants, presses, mills, and shredders are used for physical processing of dross and scrap. No melting is involved. The recovered aluminum is sold to about 70 customers. The majority of these customers blend the aluminum with other materials such as lime and fluorspar and sell the blended products to the steelmakers. Some of these blended products may be melted and cast at an IMCO facility in Oklahoma (IMCO 1997, IMCO 1998). B-33 ------- Table B-15. Representative Applications for Aluminum Casting Alloys Alloy 100.0 200.0 208.0 222.0 238.0 242.0 A242.0 B295.0 308.0 319.0 332.0 333.0 354.0 355.0 356.0 A356.0 357.0 359.0 360.0 A360.0 380.0 A380.0 384.0 390.0 413.0 A413.0 443.0 514.0 A514.0 518.0 520.0 535.0 A712.0 713.0 850.0 A850.0 Representative Applications Electric rotors larger than 152 mm (6 in.) in diameter Structural members: cylinder heads and pistons; gear, pump, and aerospace housings General-purpose castings; valve bodies, manifolds, and other pressure-tight parts Bushings; meter parts; bearings; bearing caps; automotive pistons; cylinder heads Sole plates for electric hand irons Heavy-duty pistons; air-cooled cylinder heads; aircraft generator housings Diesel and aircraft pistons; air-cooled cylinder heads; aircraft generator housings Gear housings; aircraft fittings; compressor connecting rods; railway car seat frames General-purpose permanent mold castings; ornamental grilles and reflectors Engine crankcases; gasoline and oil tanks; oil pans; typewriter frames; engine parts Automotive and heavy-duty pistons; pulleys; sheaves Gas meter and regulator parts; gear blocks; pistons; general automotive castings Premium-strength castings for the aerospace industry Sand: air compressor pistons; printing press bedplates; water jackets; crankcases. Permanent: impellers; aircraft fittings; timing gears; jet engine compressor cases Sand: flywheel castings; automotive transmission cases; oil pans; pump bodies. Permanent: machine tool parts; aircraft wheels; airframe castings; bridge railings Structural parts requiring high strength; machine parts; truck chassis parts Corrosion-resistant and pressure-tight applications High-strength castings for the aerospace industry Outboard motor parts; instrument cases; cover plates; marine and aircraft castings Cover plates; instrument cases; irrigation system parts; outboard motor parts; hinges Housings for lawn mowers and radio transmitters; air brake castings; gear cases Applications requiring strength at elevated temperature Pistons and other severe service applications; automatic transmissions Internal combustion engine pistons; blocks; manifolds; and cylinder heads Architectural; ornamental; marine; and food and dairy equipment applications Outboard motor pistons; dental equipment; typewriter frames; street lamp housings Cookware; pipe fittings; marine fittings; tire molds; carburetor bodies Fittings for chemical and sewage use; dairy and food handling equipment; tire molds Permanent mold castings of architectural fittings and ornamental hardware Architectural and ornamental castings; conveyor parts; aircraft and marine castings Aircraft fittings; railway passenger car frames; truck and bus frame sections Instrument parts and other applications where dimensional stability is important General-purpose castings that require subsequent brazing Automotive parts; pumps; trailer parts; mining equipment Bushings and journal bearings for railroads Rolling mill bearings and similar applications Compiled from Aluminum Casting Technology. American Foundrymen's Society. 1986. Source: Davis 1993 B-34 ------- B.6 BASIS FOR EXPOSURE SCENARIOS The information collected in the course of the present study of aluminum recycling can be used to construct a set of representative exposure scenarios for the radiological assessment of this process. The present section discusses possible scenarios and suggests one or more values for the exposure parameters. These data form the basis for the radiological assessment which is presented in Chapter 8. B.6.1 Exposure Parameters8 Dilution Unlike carbon steel, movement of aluminum scrap is not geographically constrained by haulage costs. If all the DOE scrap available in 2003—7237 t, as listed in Table B-2—were melted in a single 220,000-pound (100 t) capacity reverberatory furnace with 100% scrap feed, a 25% furnace heel, and 90% on stream time, it would use 29 % of the furnace capacity under optimum operating conditions (7237 H- [100 t/d x 365 d/y x 0.9 x 0.75] ~ 0.29]). Based on the April, 1997 operating rate for a specific smelter, a more realistic operating rate might be 47 million pounds (-21,000 t), in which case the DOE scrap would utilize 34% of the furnace capacity for one year. Since the specific smelter has four furnaces, three of which are typically in operation, the effective dilution in terms of worker exposure would be 0.11, assuming a separate crew for each furnace. But, if all the aluminum were melted in a single dedicated furnace, the dilution would be 0.34. Whether or not all the scrap would be handled in a single furnace would depend on the composition of the scrap, the scrap availability over time, and the product requirements at the particular time the scrap was processed. As noted in Section B.4.2, some small furnaces may have a capacity limited to 40,000 Ib (18 t) per year. It is not known whether a furnace of this size could be the only furnace at a facility or whether the facility would have multiple furnaces. It would require about 1.6 years to process the 7237 t of DOE aluminum through such a furnace. If the scrap consists of a variety of alloys, it is unlikely that it would be processed through a single furnace. A plausible scenario for the limiting case is that all of the 2,527 t of aluminum from Paducah, available each year from 2016 to 2022, would be processed at the Wabash Alloys facility in Data on a typical secondary smelter, presented in this section, is based on information from Graham (1997). B-35 ------- Dickson, Tenn. The capacity of this facility is about 150 million pounds (68,000 t) per year. In such a case, the contaminated scrap would represent about 3.7% total capacity mill of the mill—i.e., the contaminated scrap dilution factor would be 0.037. Dross Production Dross production at a typical secondary smelter with reverberatory furnaces is about 15% of the metal charge and this dross contains 8 to 12% aluminum metal. The balance of the dross is halide salts and oxides. While this is typical for a specific smelter, as noted in Section B.4.4, some dross may contain as much as 80% aluminum. On a national basis, in 1996, U.S. secondary smelters consumed 1.44 million t of scrap with a calculated metallic recovery of 1.1 million t (Plunkert 1997a). This suggests that about 24% of the scrap charge is lost as aluminum and aluminum oxide in the dross. Dust Production Based on the information in Section B.4.2, six pounds of baghouse dust are generated for each ton of scrap melted. In metric units, this corresponds to 3 kg per t, for a ratio of 0.3%. Some Pb and Cd may partition to the baghouse dust. Dust could be buried in a municipal landfill or on site. Material Balance The following simplified material balance was developed for a typical secondary aluminum smelter using reverberatory furnaces to produce casting alloys based on 1,000 kg of metal charged into the furnace: Furnace Charge: • Aluminum scrap 980 kg • Silicon 20 kg • Flux 60 kg Output:2 • Aluminum casting alloy .... 943 kg • Baghouse dust 3 kg, containing 2 kg of metal The output is greater than the furnace charge due to pick up of oxygen in the dross products. B-36 Continue ------- Back flux 4338 MT/y Silicon scrap processing contaminated Al scrap 2527 MT/y (Paducah) clean Al scrap 1446 MT/y DF=0.04 68330 MT/y smelting (reverberatory furnace) 68182 MT/y Al alloy 22% direct automotive 44% automotive related 8% small engine 7% appliance 19% other 10845 MT/y dross dross processing 1084 MT/y Al alloy 5423 MT/y oxides offgas 4338 MT/y salts emissions control system bag- house dust 217 MT/y dust inc. 145 MT/y Al, landfill Figure B-6. Simplified Material Balance for Secondary Aluminum Smelter • Dross 150 kg, containing 60 kg of salts, 15 kg of Al, and 75 ka of oxide 75 kg of oxide This simplified material balance, which is illustrated in Figure B-6, ignores the minor effects of C12 injection and Mg removal. The material flows in Figure B-6 are for a full year. Karvelas et al. (1991) quoted processing results from secondary aluminum smelters in the United States in 1988. For each 1,100 tons of aluminum produced, 114 tons of black dross and 10 tons of baghouse dust were generated. The composition of the black dross was 12% - 20% Al, 20% - 25% NaCl, 20% - 25% KC1, 20% - 50% aluminum oxide, and 2% - 5% other compounds. That study yields results similar to the simplified material balance proposed here. Karvelas et al. reported that 17 tons of aluminum were recovered from every 114 tons of black dross in 1988. B-37 ------- B.6.2 Workers in the Secondary Aluminum Industry Scrap Metal Transporter If the 2,527 t of scrap to be generated at Paducah were transported by a truck with 22-ton (20-t) capacity to a secondary smelter 170 miles (-275 km) away, it would take 126 trips. A driver would be exposed to the residually radioactive scrap for about four hours during each trip. However, since haulage costs are not the deciding factor in selecting the recycling facility, it is plausible for the scrap to be transported a greater distance, in which case a single driver could be occupied full time, hauling the scrap one-half the time and returning with an empty truck (or hauling other cargo). Scrap Handler An operator is assumed to spend eight hours per day moving scrap from the stockpiles to the shredder or the furnace using a front-end loader with a five cubic yard bucket (the bucket would be loaded 50% of the time). In addition to exposure from the load being transported, he would receive additional external radiation exposure from the scrap piles and internal doses from dust inhalation or ingestion. The scrap is stored in piles and stacked bales of shredded metal. Assuming that the desired inventory level is 15 days' supply, a facility with an annual capacity of 68,000 t would typically have at least 3,000 t of inventory on hand. The actual inventory might be larger to accommodate special purchasing situations or seasonal needs. Shredder Operator A typical shredder operator is assumed to spend seven hours per day running a scrap shredder (Figure B-3). The operator is assumed to stand beside the scrap conveyor which transports a stream of scrap 3 ft wide by 0.5 ft deep, with a 50% bulk density. Less than half the scrap is shredded. Furnace Operator The furnace operator is assumed to do a variety of jobs in close proximity to the furnace. For example, he skims dross from the melt surface in the charging well using a mechanized skimmer on an extendable arm located at one side of the well. The operator sits in a booth on the skimming machine about 6 ft from the melt and transfers the dross to a container in front of the charging well. During the course of a week the operator spends 15 hours skimming dross, and 25 hours feeding alloying or fluxing agents into the furnace or performing other furnace-related work. Other work might include manually raking the furnace to remove bulk steel objects which B-38 ------- settle to the bottom. This is done twice per shift and requires 30 to 45 minutes per event (Kiefer etal. 1995). Ingot Stacker Once the ingots are removed from the molds, they may require stacking onto pallets. According to Kiefer et al. (1995), this labor-intensive job requires a crew of four—two stackers and two forklift operators. The stackers pick up ingots from a rotary table and place them on a stacking pallet. It requires about 20 minutes for each stacker to load a 2,000-lb pallet. The forklift operators transport the pallets to a storage area. The stackers and the forklift operators trade jobs frequently during a shift. Dross Hauler Dross containing 10% Al (with Co, Fe, Mn and Tc) and 90% salts and oxides (including elements such as U, Pu, Np and Cs ) might be shipped 400 miles (-645 km) by truck with a 20- ton (18-t) capacity. Approximately 11,000 t of dross—about 600 truck-loads—is produced each year at the reference facility described in Figure B-6 . A one-way trip would take over eight hours; therefore, transporting the dross would be a full-time occupation for four or five drivers. Aluminum Fabricator Plasma arc cutting (PAC), gas metal arc welding (GMAW), and gas tungsten arc welding (GTAW) are processes typically used in fabrication of aluminum structures. An extensive study has been made of the metal fume levels associated with these processes (Grimm and Milito 1991). Tests were conducted using an instrumented mannequin in a special room where the air flow did not exceed 15 ft/min (~ 5 m/min or 7.6 cm/s). The mannequin was instrumented to measure fume concentrations inside and outside a welding helmet. Both a wrought base metal (2090) and a cast base metal (A356) were tested with different weld filler metals (1100, 2319, and 4043). Fume measurements are summarized in Tables B-16 and B-17 and indicate that the maximum fume level observed inside the welder's helmet was 7.66 mg/m3, associated with gas metal arc welding of alloy 2090. It is expected that the welder would be exposed to these fume levels no more than 50% of the time, with the balance of the workday involving setup, workpiece handling, and other operations. B-39 ------- B.6.3 Users of End-Products Automobiles The average amount of aluminum used in North American cars and light trucks is 250 pounds, 65% of which is recycled metal (IMCO 1997, Lichter 1996). The aluminum content in luxury and specialty cars is higher—for example, the Plymouth Prowler uses 963 Ib of aluminum (Drucker Research Company 1998). The use of aluminum in cars is a fast-growing market, having increased 35% over the last five years. If this trend is sustained for another five years, the average recycled aluminum content can be estimated to be 220 pounds (250 x 1.35 x 0.65). Most of the recycled aluminum would likely be associated with under-the-hood components. Another author estimated that by 2010 domestic vehicles would use 283 pounds of aluminum castings ("Automotive Aluminum Recycling" 1994). A recent study by the Drucker Research Company estimated that in 1999, the total aluminum content of passenger cars and light trucks will be 3.815 billion pounds based on 15.362 million units of production (Drucker Research Company 1998). Secondary aluminum made from old and new scrap will account for 63% of the 3.8 billion pounds (primarily as die and permanent mold castings). The total aluminum content per vehicle will average 248 pounds (of which 156 pounds will be secondary aluminum). The largest single component is most likely the engine block. The approximate weight of a four- cylinder block is 40 Ib (18 kg), a V-6 block weighs 55 Ib (25 kg), while a V-8 ranges from 60 to 80 Ib (27 to 36 kg) (Klimish 2001). Home Appliances Sources of exposure include ingestion of food cooked in cast aluminum frying pans10 and external exposure to cast aluminum components in appliances. Aluminum usage in typical home appliances is as follows (Aluminum Association 1985): • room air conditioners 10 Ib • ranges 2 Ib • refrigerators 10 Ib Kitchen cookware is commonly made from wrought aluminum alloys such as 6061 rather than cast alloys. Some cast aluminum (e.g., 383 alloy) might be used for skillets (Graham 1997). B-40 ------- • dishwashers 2 Ib • washers 15 Ib • dryers 4 Ib Truck The tractor of a large truck can contain about 700 Ib of aluminum in the cab shell (including the sleeper compartment) and under the hood. On a long haul the driver is limited by Department of Transportation regulations to a maximum of 15 hours per day of driving and on-duty time, including a maximum often hours of driving. The driver is also limited to 60 hours of on-duty plus driving time in a seven-day period. On-duty time includes such actions as loading and unloading the vehicle. In addition, the driver may spend time resting in the sleeper compartment. However, the cab is made from a large number of aluminum parts and the likelihood of all the parts coming from the same heat of aluminum is nil. The largest aluminum component that is made from one or two pieces of aluminum mill products is assumed to be a 100-gallon fuel tank that is mounted on the left side of the cab behind and below the driver.11 If such a tank were fabricated from A-inch aluminum sheet, it would weigh about 180 Ib. 16 ' & Motor Home The floor of an aluminum motor home contains about 600 Ib of aluminum. As is the case with the truck cab, the motor home will be constructed from a variety of shapes, making it unlikely that all the material would come from a single heat. The Freightliner Cl 12 Tractor with 58-inch raised roof sleeper cab is configured in this way. According to a Freightliner spokesman, tanks weigh about 200 pounds. B-41 ------- Table B-16. Concentrations in Ambient Air Inside and Outside the Welder's Helmet During Aluminum Welding and Cutting Component NO NO2 03 Total fume AI203 SiO2 Fe203 CuO Cr203 MgO MnO NiO TiO2 ZrO2 Li20 Sb BeO Be Total oxides Oxide •*• total fume Units ppm mg/m3 ug/m3 mg/m3 % GMAW* 2090/2319 Inside <0.25 <0.01 0.16 7.66 7.12 — — 0.15 — — — — — — — — — — 7.29 94.9 Outside <0.25 <0.02 0.22 42.9 40.60 — 0.07 1.09 — <0.03 0.07 — 0.07 — 0.06 — — — 42.00 98.6 GMAW 2090/1100 Inside <0.25 <0.01 0.09 5.76 5.71 — 0.131 0.05 0.04 — — — — — — — — — 6.04 106.2 Outside <0.25 0.03 0.14 27.4 25.97 — 0.05 0.05 — — — — — — — — — — 26.08 95.6 GTAW3 2090/2319 Inside <0.25 <0.01 <0.01 0.20 0.05 — — — — — — — — — — — — — 0.06 30.0 Outside <0.25 <0.01 0.08 0.57 0.23 — — — — — — — — — — — — — 0.23 NV GMAWA 356/4043 Inside <0.25 <0.01 0.28 1.14 0.96 0.12 — 0.03 0.03 — — — — — — — <2.91 <1.04 1.10 92.1 Outside <0.25 0.23 5.75 14.5 13.97 0.99 — 0.03 — — — — — — — — 28.40 10.22 15.05 104 GMAWA 356/4043 Inside <0.25 <0.01 0.16 0.73 0.70 — 0.04 — — — — — — — — — <1.87 <0.67 0.75 122 Outside <0.25 <0.01 0.68 4.96 3.48 — 0.04 0.03 — — — — — — — — <3.30 <1.22 3.52 73.6 GMAWA 356/4043 Inside <0.25 <0.01 0.06 0.78 0.36 — — — — — — — — <2.14 <0.77 0.36 53.7 Outside <0.25 <0.01 0.18 2.82 1.59 — <1.87 <0.67 1.60 68.9 td -U to Note: — indicates analyses completed, but values do not exceed lower limit of detection (LOD). (For SiO2, LOD=0.03 mg/m3, for all other oxides, except BeO, LOD=0.02 mg/m3). a Gas Metal Arc Welding b Gas Tungsten Arc Welding ------- Table B-17. Dust Levels During Plasma Arc Cutting of Wrought Metal 2090 (mg/m3) Component Total fume ALA SiO2 Fe203 CuO Cr2O3 MgO MnO MO TiO2 ZrO2 Li2O BeO (|ig/m3) Be (iig/m3) Total oxides (mg/m3) Total oxide/total fume (%) Inside Helmet 3.40 2.65 — — <0.03 — — — — — — 0.16 <1.40 0.50 2.83 71.4 Outside Helmet 3.28 2.25 — — — — — — — — — .14 <1.40 <0.50 2.39 66.5 B-43 ------- REFERENCES Adams, V. 1998. "National Center of Excellence for Metals Recycle." U.S. Department of Energy. Aluminum Association. 1985. "Aluminum Recycling Casebook. " American Conference of Governmental Industrial Hygienists (ACGIH). 1996. "1996 TLVs and BEIs: Threshold Limit Values for Chemical Substances and Physical Agents, Biological Exposure Indices." ACGIH, Cincinnati, OH. "Aluminum Smelters Export Slag, Dross." 1995. American Metal Market. "Automotive Aluminum Recycling in 2010." 1994. Automotive Engineering 102 (8): 17 Compere, A. L., et al. 1996. "Decontamination and Reuse of ORGDP Aluminum Scrap," ORNL/TM-13162. Oak Ridge National Laboratory, Oak Ridge, TN. Copeland, G. L., and R. L. Heestand. 1980. "Volume Reduction of Contaminated Metal Waste." In Waste Management '80: The State of Waste Disposal Technology, Mill Tailings, and Risk Analysis Models. Vol. 2, 425-433. Eds. M. E. Wacks and R. G. Post. Tucson AZ. Crepeau, P. N, M. L. Fenyes and J. L. Jeanneret. 1992. "Solid Fluxing Practices for Aluminum Melting." Modem Casting 82 (7): 28-30. Davis, J. R., ed. 1993. Aluminum and Aluminum Alloys, ASM Speciality Handbook. ASM International. Drucker Research Company, Inc. 1998. "Drucker Research Company 1999 Passenger Car and Light Truck Aluminum Content Report Highlights" (18 November 1998) Garbay, H., and A. M. Chapuis. 1991. "Radiological Impact of Very Slightly Radioactive Copper and Aluminum Recovered from Dismantled Nuclear Facilities," Final Report, EUR- 13160-FR. Commission of the European Communities. Glassner, A. 1957. "The Thermochemical Properties of the Oxides, Fluorides, and Chlorides to 2500 K," ANL-5750. Argonne National Laboratory, Argonne, IL. Graham, R. (Wabash Alloys, Inc., Dickson, TN). 1997. Private communication. Graziano, D., et al., 1996. "The Economics of Salt Cake Recycling." In Light Metals, 1255- 1260. The Minerals, Metals, and Materials Society. B-44 ------- Grimm, R. E., and R. A. Milito. 1991. "Evaluation of Atmosphere at Operator's Position When Gas Metal Arc Welding, Gas Tungsten Arc Welding and Plasma Arc Cutting Selected Aluminum Alloys," Report No. 52-91-01. Alcoa Laboratories. Heshmatpour, B., G. L. Copeland, and R. L. Heestand. 1983. "Decontamination of Transuranic Contaminated Metals by Melt Refining." Nuclear and Chemical Waste Management., 4:129-134. Heshmatpour, B., and G. L. Copeland. 1981. "The Effects of Slag Composition and Process Variables on Decontamination of Metallic Wastes by Melt Refining," ORNL/TM-7501. Oak Ridge National Laboratory, Oak Ridge, TN. Hryn, J. N., et al. 1995. "Products from Salt Cake Residue-Oxide." In Third International Symposium on Recycling of Metals and Engineered Materials, 905-915. Eds. P. B. Queneau and R. D. Pearson. The Minerals, Metals & Materials Society. IMCO Recycling Inc. 1997. "1996 Annual Report." Irving TX. IMCO Recycling Inc. 1998. "1997 Annual Report." Irving TX. Karvelas, D., E. Daniels, B. Jody, and P. Bonsignore. 1991. "An Economic and Technical Assessment of Black-Dross- and Salt-Cake-Recycling Systems for Application in the Secondary Aluminum Industry," ANL/ESD-11. Argonne National Laboratory. Argonne, IL. Kiefer, M., et al., 1995. "Health Hazard Evaluation Report: Arkansas Aluminum Alloys Inc., Hot Springs, Arkansas," HETA 95-0244-2550, NTIS PB96210067. National Institute for Occupational Health and Safety. Klimish, D. (The Aluminum Association, Automotive and Light Truck Group). 2001. Private communication (1 May 2001). Lichter, J., 1996. "Aluminum Applications Expand." Advanced Materials and Processes 150 (4): 19-20. ASM International. Michaels, D. 2000. Statement issued by Assistant Secretary David Michaels in October, 2000. Michaud, D., et al. 1996. "Characterization of Airborne Dust from Two Nonferrous Foundries by Physico-chemical Methods and Multivariate Statistical Analyses." Journal of the Air & Waste Management Association 46: 450-457. Mautz, E. W., et al., 1975. "Uranium Decontamination of Common Metals by Smelting: A Review," NLCO-1113. National Lead Company of Ohio. B-45 ------- Mobley, M. 1999. Private communication (23 February 1999). National Research Council. 1996. "Affordable Cleanup? Opportunities for Cost Reduction in the Decontamination and Decommissioning of the Nation's Uranium Enrichment Facilities." National Academy Press. Neff, David V. 1991. "Scrap Melting and Metallurgical Processes Employed in Aluminum Recycling." In Extraction, Refining, and Fabrication of Light Metals., 18-21. Ottawa. Novell!, Lynn R. 1997. "The Fate of Secondary Aluminum Smelting." Scrap 54 (6): 49-58. Plunkert, P. (Bureau of Mines). 1995. Private communication (20 September 1995). Plunkert, P., 1996. "Aluminum Annual Review: 1995." Bureau of Mines, U.S. Department of Interior. Plunkert, P., 1997a. "Aluminum Annual Review: 1996." U.S. Geological Survey. Plunkert, P. (U.S. Geological Survey). 1997b. Private communication (May 1997). Plunkert, P. (U.S. Geological Survey). 1999. Private communication (22 February 1999). Roth, David J., 1996. "Recovery of Aluminum from Rotary Furnace Salt Cake by Low Impact Rotary Tumbling." Light Metals, 1251-1253. Uda, T., et al. 1986. "A Melt Refining Method for Uranium Contaminated Aluminum." Nuclear Technology 72:178-183. Uda, T., et al. 1987. "Melting of Uranium-Contaminated Metal Cylinders by Electroslag Refining." Nuclear Technology 79:329-337. U.S. Department of Energy (U.S. DOE). 1986. Request for Proposal No. DE-RP05- 86OR21069. U.S. Department of Energy (U.S. DOE), Office of Environmental Management. 1996. "Taking Stock: A Look at the Opportunities and Challenges Posed by Inventories from the Cold War Era," DOE/EM-0275. Vol. 2. U.S. Environmental Protection Agency (U.S. EPA). 1990. Docket A-92-61, II-D-11. Letter from R. Shafer, Alcan Recycling, to S. Dillis, Permit Review Branch, Division of Air Quality, Kentucky Department for Environmental Protection, enclosing report: "Compliance Emission Testing on the Melt Furnace, Decoater, Hold Furnace, Alpur Filter, Dross B-46 ------- Baghouse, Hot and Cold Baghouses at the Alcan Ingot and Recycling in Berea, Kentucky." (4 December 1990) U.S. Environmental Protection Agency (U.S. EPA). 1995. "Profile of the Nonferrous Metals Industry" EPA 310-R-95-010 Viland, J. S. 1990. "A Secondary's View of Recycling." In Second International Symposium: Recycling of Metals and Engineered Materials. The Minerals, Metals & Materials Society. Viland, J. S. (Wabash Alloys). 1997. Private communication (April 1997). Waugh, R. C. 1959. "The Reaction and Growth of Uranium Dioxide-Aluminum Fuel Plates and Compacts." Nuclear Science and Engineer ing Supplement, vol. 2, No. 1. Wrigley, A., 1995. "Ford Begins Aluminum Salt Cake Recycling." American Metal Market 103 (123): 6. Zemex 1998. "Zemex Corporation Completes Acquisition of Aluminum Dross Processor." (5 June 1998). B-47 ------- APPENDIX B-l DESCRIPTION OF SELECTED SECONDARY SMELTERS ------- Table Bl-1. Description of Selected Secondary Smelters Facility Ohio Valley Aluminum, Shelbyville KY Rock Creek Aluminum, Rock Creek OH Alcan Recycling, Shelbyville TN Sceptar Industries, New Johnsonville TN IMCO Recycling, Morgantown KY IMCO Recycling, Uhrichville OH U.S. Reduction, Toledo OH Wabash Alloys, Dixon TN Bag House Type None ?? On shredder, decoater, and furnaces. Furnace bags coated w/ Ca(OH)2 On rotary fur- naces but not on reverberatories Lime-coated bags. One ton of dust per 100 tons of feed. Unknown Lime-coated bags Dust Disposal N/A ?? BFI ships to secured landfill To on-site landfill Both on-site & off- site landfills used. On-site equivalent to Sub-Title C, although not required Off-site Off-site BFI to municipal landfill Pretreatment None Crushing and screening Shredding and decoating Very little pre- processing Shredder Large and small crusher and dryer Shredder Dross Handling Skimmed into containers and sold N/A Sold to Tennessee Processors, Al repurchased Dross is remelted Shipped to independent process Shipped to company plant in Benton, Ark. Radiation Detectors Not used Hand-held Geiger counter Fixed Ludlum detectors Not used Not used Fixed Furnaces Three, 9.5 million Ib/mo total 60 million/y, no melting Two reverberatory, 40 to 50,000 tons/y total Two reverberatory, three rotary 12-14 million Ib/month Rotary furnaces: reverberatory under construction, 220 million Ib/y current capacity 360 million Ib/y Two reverberatory Four reverberatory, 220,000 Ib each, 150 million Ib/y total. td ------- APPENDIX B-2 SECONDARY ALUMINUM SMELTER OPERATIONS AT ARKANSAS ALUMINUM ALLOYS INC. ------- SECONDARY ALUMINUM SMELTER OPERATIONS AT ARKANSAS ALUMINUM ALLOYS INC 12 B2.1 FACILITY DESCRIPTION Arkansas Aluminum Alloys, Inc. (AAAI) is an aluminum recycling facility (secondary aluminum smelter) that has been in business since 1974. AAAI produces aluminum stock with varied elemental composition depending on customer specifications. Approximately 165 employees (administration and production) work at the facility. The facility operates 24 hours per day, 355 days per year, with four rotating work shifts. Employees receive two 10-minute breaks and a 30- minute lunch period per shift. There are three gas-fired reverberatory furnaces at the smelter. However, except for times of extreme production demands, only two furnaces are operated at one time. Office, warehouse, and production space occupies 57,130 square feet, situated on nineteen acres. Smoking is permitted in the manufacturing areas. B2.2 PROCESS DESCRIPTION AAAI receives and processes all types of reclaimable aluminum scrap except cans. Most (98%) of the scrap aluminum is delivered by tractor-trailer truck, weighed, scanned for radioactivity, unloaded, and spread in the storage area. The scrap is then placed on a conveyor where it is visually inspected and manually sorted. Iron, stainless steel, zinc, brass, and other materials are removed at this station. The scrap is then sampled and analyzed and placed in storage bins based on elemental composition. AAAI has an on-site laboratory with a sophisticated elemental analyzer that requires very little sample preparation and provides rapid results. Some of the sorted scrap is shredded and crushed and screened to removed dirt. A magnet is used to separate iron from the aluminum. The shredded scrap is then placed in bins. A gas-fired kiln located at the back of the facility is used to dry machined turnings prior to processing in the melting furnace. There are three 220,000-lb capacity gas-fired furnaces at AAAI. Each furnace is equipped with exhaust ventilation to control flue gas, as well as fume control (canopy hoods). Fume exhaust is conveyed to a roof-mounted baghouse system. Furnace runs last approximately 20 hours, followed by a 41A> hour pour time. The pour temperature of the melt is approximately 1380°F. 12 Source: Keifer et al. 1995 B2-1 ------- About 80,000 Ib of molten aluminum are left in the furnace to prime the next run. To charge the furnace, the furnace operator will open large overhead doors on one side of the furnace and use a front-end loader to place the scrap into wells adjacent to the furnace. After charging, the overhead doors are closed, and the scrap melts and flows into the main furnace body. Samples are periodically taken from the melt with a ladle and analyzed to ensure that the final product meets customer specifications (elements are added if necessary to meet customer requirements). Copper and silicon are the major elements added; this is done by placing into a hopper at the front of the furnace. The majority (over 95%) of AAAI customers purchase the finished aluminum in 30-lb ingots. AAAI will also accommodate those few customers who request 1000- Ib aluminum "sows." Magnesium is a common contaminant that must be scavenged (by demagging) from the melt to reduce the concentration below 0.1%. At AAAI, this is accomplished by injecting chlorine gas into the melt—piped from a 55-ton tank car, through vaporizers, to each furnace—via a graphite pump and carbon tubes. The chlorine combines with the magnesium to form MgCl2, which is then skimmed off the top of the melt. If necessary, A1F3 can be used instead of chlorine for this "demagging" operation. According to AAAI, A1F3 is rarely used. Salt (NaCl), potash, and cryolite are added to every charge as a flux to remove dirt and prevent oxidation of the melt. Iron is considered a major detriment to the product, and every attempt is made to eliminate it during initial inspection and by the use of magnetic separation prior to processing. However, some iron inevitably gets into the furnace, sinks to the bottom, and must be manually removed. Periodically (twice per shift), furnace operators manually drag a large rake along the bottom of the melt to pull the iron out of the furnace. Each raking event takes about 30 to 45 minutes. During pouring, the furnaces drain into an insulated open trough. To start the pour, a furnace plug is removed and the molten metal flows continuously through the trough into V/2 ft long, 30- lb molds (or 100-pound molds if necessary). The 30-lb molds are on a carousel/conveyor system and pouring occurs as the molds move sequentially through a water bath. This area is shielded because of the potential for violent reactions in the event molten aluminum contacts the water. After the molds have passed through the water, two workers stand adjacent to the conveyor line and skim dross from the ingots using hoe-like hand tools. The ingot molds are then elevated on the carousel and rotated to release the ingots onto a conveyor belt. Graphite is used as a mold- release agent. An automated pneumatic hammer is used to remove the ingots from the molds if necessary. B2-2 ------- The ingots are then conveyed to the stacking area where they are dropped onto a rotating table. The surface temperature of the ingots is approximately 230°F when received at the stacking station. Stacking is a 3- or 4-man labor-intensive operation (2 stackers, 2 forklift operators), and workers continuously rotate between stacking and forklift operation. As the ingots are deposited onto the table, the stacker will pick up the ingot and place it in position on a stacking pallet. Stackers are also required to inspect the ingots and recycle those found to be defective. Each stacker will load one 2000-lb stack (approximately 18-20 minutes), and then switch jobs with the forklift operator. The fully stacked pallets are then moved to a cooling room, and finally to the warehouse. AAAI has a fleet of trucks for shipping product to customers. REFERENCE Kiefer, M., et al., 1995. "Health Hazard Evaluation Report: Arkansas Aluminum Alloys Inc., Hot Springs, Arkansas," HETA 95-0244-2550, NTIS PB96210067. National Institute for Occupational Health and Safety. B2-3 ------- APPENDIX C COPPER RECYCLING ------- Contents page C.I Inventory of Potentially Recyclable Copper Scrap C-l C. 1.1 Scrap Metal Inventory C-l C. 1.2 Radionuclide Inventory C-4 C.2 Recycling of Copper Scrap C-6 C.2.1 Types of Copper Scrap C-6 C.2.2 Scrap Handling and Preparation C-8 C.2.3 Copper Refining Operations C-10 C.2.3.1 Copper Smelting Practices C-l 1 C.2.3.2 Copper Converting C-18 C.2.3.3 Fire Refining C-19 C.2.3.4 Electrolytic Refining C-19 C.2.3.5 Melting, Casting, and Use of Cathodes C-23 C.2.3.6 Slag Handling C-23 C.2.3.7 Offgas Handling C-24 C.2.3.8 Illustrative Secondary Smelter C-24 C.2.4 Brass and Bronze Ingot Production C-27 C.2.5 Brass Mills C-27 C.2.6 Aluminum Bronze Foundries C-30 C.3 Markets C-31 C.3.1 Scrap Prices C-31 C.3.2 Scrap Consumption C-32 C.4 Partitioning of Contaminants C-32 C.4.1 Partitioning During Copper Refining C-33 C.4.1.1 Thermochemical Considerations C-33 C.4.1.2 Experimental Partitioning Studies C-33 C.4.1.3 Proposed Partitioning of Contaminants C-38 C.4.2 Partitioning During Brass Smelting C-46 C.5 Exposure Scenarios C-46 C.5.1 Modeling Parameters C-46 C.5.1.1 Dilution of Cleared Scrap C-47 C.5.1.2 Slag Production C-47 C.5.1.3 Baghouse Dusts C-48 C.5.1.4 Electrolyte Bleed C-49 C.5.1.5 Anode Slimes C-49 C.5.1.6 Summary Model for Fire-Refined Products C-50 C.5.1.7 Summary Model for Electrorefming C-51 C-iii ------- Contents (continued) page C.5.2 Worker Exposures C-51 C.5.2.1 Baghouse Dust Agglomeration Operator C-52 C.5.2.2 Furnace Operator C-53 C.5.2.3 Scrap Handler C-53 C.5.2.4 Casting Machine Operator C-53 C.5.2.5 Scrap Metal Transporter C-53 C.5.2.6 Tank House Operator C-54 C.5.3 Non-Industrial Exposures C-54 C.5.3.1 Driver of Motor Vehicle C-54 C.5.3.2 Homemaker C-54 References C-56 Appendix C-l. Partitioning During Fire Refining and Electrorefining of Copper Scrap C-iv ------- Tables page C-l. Current Inventory of Copper Scrap at DOE Facilities C-2 C-2. Availability of Copper from Decommissioning of Nuclear Facilities C-5 C-3. Copper Recovered from Scrap Metal Processed in the United States in 1997 C-10 C-4. Copper Consumption from Copper-Base Scrap in the United States in 1997 C-l 1 C-5. Composition of Process Streams from the Smelting of Copper Scrap in a Cupola Blast Furnace C-13 C-6. Composition of Products Obtained from Treating Copper Blast Furnace Slag in an EAF C-15 C-7. Partitioning During Blast Furnace Smelting of Copper Scrap C-15 C-8. Composition of Converter Products from the Smelting of Copper Scrap C-19 C-9. Composition of Anodes Produced in a 250-t Reverberatory Furnace C-20 C-10. Anode Compositions at Various U.S. Electrolytic Copper Refineries C-22 C-ll. Consumption of Copper-Base Scrap in 1997 C-29 C-12. Standard Free Energies of Formation for Various Oxides at 1,500 K C-35 C-13. Calculated Partition Ratios of Various Contaminants Between Copper and an Oxide Slag at 1,400 K C-36 C-14. Partitioning of Uranium in Laboratory Melts of Copper C-37 C-15. Distribution of Iridium and Ruthenium During Electrorefming of Copper C-37 C-l6. Distribution of Iridium and Ruthenium after Electrolyte Purification C-38 C-17. Observed Partition Fractions in the Melting of Low-grade Copper in a Blast Furnace C-39 C-l 8. Partition Fractions of Impurities in the Melting of Low-grade Copper Scrap in a Blast Furnace C-40 C-19. Partition Fractions of Impurities in the Fire Refining of Copper C-42 C-20. Composition of Anode and Cathode Copper and Anode Slimes at the Southwire Co. C-43 C-21. Partition Fractions of Impurities in the Electrorefming of Copper C-44 C-22. Half-cell Electrode Potentials of Elements less Noble than Copper C-45 C-23. Airborne Dust Concentrations At Primary Copper Smelter C-52 Cl-1. Partitioning During Fire Refining and Electrolysis of Copper Scrap Cl-1 Figures C-l. Simplified flow diagram for copper-base scrap in 1997 C-9 C-2. Process diagram for the flow of copper scrap in primary and secondary copper refining C-12 C-3. Flow Diagram of the Copper Division of Southwire (CDS) C-26 C-4. Proposed Material Balance for Modeling Copper Produced by Fire Refining C-51 C-5. Simplified Material Balance for Electrorefming of Copper Produced from Scrap .... C-52 C-v ------- COPPER RECYCLING This appendix presents background material to support an analysis of exposures expected from the recycling of copper scrap. C. 1 INVENTORY OF POTENTIALLY RECYCLABLE COPPER SCRAP C.I.I Scrap Metal Inventory The Scrap Metal and Equipment Appendix to the 1996 MIN Report (U.S. DOE 1995) identified 1,691 metric tons1 (t) of copper and brass scrap in inventory. This inventory was classified as containing 1,4901 of contaminated metal, 53 t of clean scrap metal, and 148 t of material unspecified as to its state of contamination. (These amounts are slightly higher than the inventory listed in Table 4-4 of the present report)2. A detailed breakdown by location is provided in Table C-l. Based on the ratio of clean to contaminated scrap, 143 t of the unspecified material was categorized in the present study as contaminated, resulting in a total of 1,633 t of potentially contaminated 58 t of clean copper and brass scrap. As discussed in Section 4.1.4, the HAZWRAP Report (Parsons 1995) listed inventories of contaminated scrap metal at LANL and Rocky Flats which were omitted from the MIN Report. It is therefore likely that some unreported copper scrap may be in inventory at these two sites. Obviously, most of the current inventory is at Fernald. DOE has entered into an arrangement with Decon and Recovery Services LLC (DRS) of Oak Ridge, Tenn. to process about 1,200 t of copper scrap (primarily motor windings) from Fernald (Deacon 1999). DRS will mechanically remove the insulation, which is slightly contaminated, leaving behind clean copper that, in the future, could be released for unrestricted sale under the provisions of DOE Order 5400.5.3 This appendix includes numerous references with widely varying units of measurement. The authors of this appendix have generally chosen not to convert the units to a consistent system but rather have chosen to quote information from the various sources in the original units. When the cited information is distilled into scenarios for modeling doses and risks, consistent units are used. These data are slightly higher than those in Summary Table 1.4 of U.S. DOE 1995 because that table did not include all individual sites. As noted in Chapter 2, DOE currently has a moratorium on the free release of volumetrically contaminated metals and has suspended the unrestricted release for recycling of scrap metal from radiological areas within DOE facilities. C-l ------- Table C-l. Current Inventory of Copper Scrap at DOE Facilities (t) Location Fernald ANL-W Hanford BNL FermiLab SRS WIPP NTS SLAC LBL K-25 Y-12 ORNL Portsmouth Paducah Total Clean 6.3 33 2.5 0.23 0.90 4.8 4.8 53 Contaminated 1270 200 9.2 11 1490 Unspecified 42 44 1.8 21 39 148 The principal future sources of DOE copper scrap are the gaseous diffusion plants at Oak Ridge; Paducah, Ky., and Portsmouth, Ohio. It has been estimated that these plants contain 40,200 t of copper scrap (National Research Council 1996)4 with individual facility totals as follows: •K-25 16,000 t • Portsmouth 13,600 t •Paducah 10,6001 The copper is present in the form of wire, tubing, and valves, with the following breakdown reported for the K-25 plant (U.S. DOE 1993): These values were derived from a 1991 study by Ebasco Services, Inc., which estimated that the total radioactive scrap metal arising from decommissioning the three gaseous diffusion plants would be 642,000 t. This estimate did not include carbon steel in the building structures but did include electrical/instrumentation equipment and housings. Person et al. (1995) estimated that 1,047,000 t of scrap metal would be recycled including structural steel. Of this total, 60.3% is estimated to be potentially contaminated and the balance to be clean. Thus, these authors predicted the same total amount of radioactive scrap metal as the earlier Ebasco study; they did not provide a breakdown by metal type. C-2 ------- • copper tubing/valves 0.191 • large copper wire 8.6 t • small copper wire 7.2 t The three plants contain an additional 20,200 t of "aluminum/copper," but the two metals are not separated by type. The above estimates do not include any copper in "miscellaneous electrical/instrumentation and housings" (U.S. DOE 1993). No information is available on copper scrap expected to be generated at other DOE facilities. To develop a recycling schedule for DOE facilities, the procedure described below was used. Existing scrap is assumed to be available for processing in 2003. The existing inventory is adjusted to remove the Fernald motor windings, since this scrap is being handled currently. The decommissioning schedule for the three diffusion plants is as follows (see Section 4.1.5): • K-25 1998-2006 • Portsmouth 2007-2015 •Paducah 2015-2023 It is assumed that no scrap is generated in the first year of a nine-year decommissioning period, 13% is generated in years 2 through 8, and 9% in the final year. Scrap generation based on this schedule is summarized in Table C-2. Table A-29 lists the amounts of copper, brass, and bronze used to construct a 1971-vintage, 1,000 MWe PWR facility. Specific information is not available on the amount or contamination level of radioactively contaminated copper scrap that would be generated during the decommissioning of such a facility. Consequently, it is assumed that the contaminated fraction of copper scrap is the same as contaminated fraction of carbon steel from the Reference BWR and Reference PWR facilities. Extending the data in Table A-29 to the entire U.S. commercial nuclear power industry leads to the conclusion that approximately 73,000 t of copper would be generated by the decomissioning of the facilities listed in Appendix A-l . Only a small portion of this metal is expected to be contaminated. Some of the contaminated inventory may not be suitable for free release. Based on the results for carbon steel presented in Appendix A, it is assumed that 20% of the copper scrap from the Reference BWR would be residually radioactive metal that is potentially C-3 ------- recyclable, while 10% of the copper scrap from the Reference PWR would fall into this category. Applying these factors yields 9,6911 of potentially recyclable contaminated copper, as shown in Table 4-8. As shown in that table, the nuclear power plants also contain small quantities of brass and bronze. These copper alloys were not included in this analysis. Since the annual availability of these alloys should be less than 50 t in toto, sizable dilution with uncontaminated scrap is expected; thus, the omission of these metals should have no significant impact on the radiological assessment. The schedule of anticipated releases of scrap metals from nuclear power plants is presented in Table 4-9. The data for copper are reproduced in Table C-2. From Table C-2, it can be seen that the maximum projected annual amount of DOE and commercial nuclear power plant copper scrap to be available for clearance is 10,833 t in the year 2003. This includes the 1,633-t inventory derived from U.S. DOE 1995 (less 1,200 t of Fernald scrap assumed to have been removed to date), and a stockpile of copper scrap accumulated during five years (1999 - 2003) of decommissioning and dismantlement of the K-25 facility. This projection is based on the assumption that DOE will resume clearing scrap metal for recycle by 2003 (see Section B.I.I). The total of 50,3001 of potentially recyclable scrap in Table C-2 is in good agreement with a more recent DOE estimate of 51,000 t of radioactive copper scrap (Adams 1998). C.I.2 Radionuclide Inventory As indicated in Section C. 1.1, the majority of scrap copper will be generated from the gaseous diffusion plants. The naturally occurring uranium isotopes and their short-lived progenies are the principal source of contamination at the diffusion plants. Other contaminants include Tc-99, U-236, and traces of Pu-239 and Np-237. It has been estimated that the following activities were introduced into the Paducah gaseous diffusion plant, relative to 250 kCi of U-238 (National Research Council 1996): • U-236 900 Ci •Tc-99 11,200 Ci •Np-237 13 Ci • Pu-239 20 Ci • Th-230 (+ progeny) 140 Ci C-4 ------- Pa-231 (+ progeny) 16 Ci Table C-2. Availability of Copper from Decommissioning of Nuclear Facilities (t) Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 Total DOE Facilities 10,833 2,080 2,080 1,440 — 1,770 1,770 1,770 1,770 1,770 1,770 1,770 1,210 1,380 1,380 1,380 1,380 1,380 1,380 1,380 940 — — — 40,633 Commercial Nuclear Power Plants — — — 103 24 — — — — — — — — 115 — — 235 189 172 537 654 1,074 132 517 Year 2027 2028 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2043 2044 2045 2046 2047 2049 2052 2056 2057 2058 Commercial Nuclear Power Plants 207 247 215 285 673 425 711 564 954 374 129 286 77 201 124 75 62 19 62 38 69 69 98 9,715 Much of this contamination was removed during the cascade upgrade and improvement programs of the 1980's (National Research Council 1996). The other significant source of copper scrap is Fernald. Beginning in 1953, the Feed Materials Production Center (now known as the Fernald Environmental Management Project [FEMP]) converted uranium ore to uranium metal C-5 ------- targets for nuclear weapons production. Over a 36-year period, this facility produced over 225,000 t of purified uranium. The principal radioactive contaminants include the uranium isotopes (and their short-lived progenies) and Tc-99. In commercial nuclear power plants, activation of copper should be negligible. Naturally occurring copper consists of two isotopes: Cu-63 (69%) and Cu-65 (31%). In a nuclear power reactor, thermal neutrons create only small amounts of Cu-64 and Cu-66, because the neutron- capture cross-sections of the naturally-occurring copper isotopes are small. These radioisotopes, with respective half-lives of 12.7 hr and 5.1 min, undergo p-decay to the stable isotopes Zn-64 and Zn-66 in. Thus, the major source of radioactive contamination will be surface contamination caused by a broad suite of radionuclides (Epel 1997). C.2 RECYCLING OF COPPER SCRAP Copper scrap can enter copper refining and processing operations in a variety of ways, depending on factors such as the quality of the scrap and its alloy content. For example, some copper scrap may be refined at primary copper smelters and some at secondary smelters. Copper alloy scrap may be remelted at brass mills, ingot makers, or foundries. This section characterizes the manner in which copper and copper alloy scrap are recycled. C.2.1 Types of Copper Scrap The Institute of Scrap Recycling Industries (ISRI) and the National Association of Recycling Industries recognize various major classes of copper scrap (NARI1980, Newell 1982, Riley et al. 1984). The major unalloyed scrap categories are termed No. 1 copper, which must contain more than 99% copper, and No. 2 copper, which must contain a minimum of 94% copper. For copper alloys, ISRI has identified 50 separate scrap classifications. Additional classifications exist for copper containing waste streams, such as skimmings, ashes and residues generated in copper smelting and refining processes. Copper scrap is further categorized as either "old" or "new" scrap. New scrap is generated during fabrication of copper products. For example, copper-containing end-products that are manufactured from intermediates, such as copper sheet, strip, piping, or rod, may have product yields as low as 40%. These new scrap materials generated from borings, turnings, stampings, cuttings, and "off-specification" products are commonly sold back to the mills that produced the original intermediates from which the new scrap was generated. Since both new scrap and C-6 ------- manufactured scrap are recycled within the copper industry, neither is considered to be a new source of copper. Old scrap, which is generated from worn-out, discarded, or obsolete copper products, does constitute a new (i.e., from outside the industry) source of metal for the secondary copper industry. Since World War II, the reservoir of copper products in use has increased dramatically, both in the U.S. and globally. The U.S. scrap inventory increased from 16.2 million tons in 1940 to nearly 70 million tons in 1991 (Bureau of Mines 1993). The availability of copper scrap is linked with the quantity of copper-containing products and their life-cycles. Estimates of life cycles have been made for major products: copper used in electrical plants and machinery averages about 30 years, in non-electrical machinery about 15 years, in housing 40 years or more, and in transportation about 10 years (Carlin et al. 1995). Copper scrap may also be broadly categorized into four main types based on copper content and the manner in which it is treated for copper recovery (as quoted from Davenport 1986): • Low-grade scrap of variable composition (10-95% Cu). This material is smelted in blast or hearth furnaces and then fire and electrolytically refined. It may also be treated in Peirce-Smith converters of primary smelters. • Alloy scrap, the largest component of the scrap recovery system, consists mainly of brasses, bronzes, and cupronickels from new and old scrap. There is no advantage in re- refining these alloys to pure copper, and hence they are remelted in rotary, hearth, or induction furnaces and recast as alloy stock. Some refining is done by air oxidation to remove aluminum, silicon, and iron as slag, but the amount of oxidation must be closely controlled because desirable alloy constituents (Zn in brasses and Sn in bronzes) also tend to oxidize. • Scrap, new or old, which is by and large pure copper but which is contaminated by other metals (e.g. metals used in plating, welding, or joining). This scrap, is melted in the Peirce-Smith converters of primary smelters or the anode furnaces of primary or secondary refineries, where large portions of the impurities (e.g. Al, Fe, Zn, Si, Sn) are removed by air oxidation. The metal is then cast into copper anodes and electro-refined .... It may also be sold as fire-refined copper for alloy making. • Scrap which is of cathode quality and requires only melting and casting. This scrap originates mainly as wastes from manufacturing (e.g. reject rod, bare wire, molds). It is melted and cast as ingot copper or alloyed and cast as brasses or bronzes. C-7 ------- According to the U.S. Geological Survey, in 1997, about 496,000 t of copper were recovered from old scrap and 956,000 t from new scrap. This resulted in 1,450,000 t of copper consumption in the U.S. from scrap (Edelstein 1998). This quantity of copper was contained in 1,750,000 t of scrap metal. Table C-3 summarizes the kinds of scrap involved in copper recycle and the form in which the copper was recovered. It is important to note that alloy scrap will typically be reused in similar alloys. Aluminum scrap containing copper will be used in aluminum alloys; brass scrap will be used in brass, etc. However, pure recycled copper can conceptually be used either as pure copper or as an alloying agent. In 1997, consumers of this scrap included about 35 brass mills, several brass and bronze ingot makers, 15 wire mills, four secondary smelters, seven primary smelters, six fire refineries, eight electrolytic plants, and 600 foundries, chemical plants, and miscellaneous consumers (USGS 1998). The quantities of old and new copper-base scrap used by these consumers are summarized in Table C-4. The total in this table is less than the total in Table C-3 because Table C-4 includes only copper-base but not other copper-containing scrap. A simplified flow diagram for the copper scrap consumption documented in Table C-4 is included as Figure C-l. This figure illustrates the disposition of 1,370,000 t of copper in copper- base scrap. It is apparent from the diagram that the flow paths are numerous and complex. Information presented by Edelstein (1998) indicates that, of the 383,000 t of copper in scrap that is processed by smelters and refiners (i.e. the box on the left of Figure C-l), about 39% is No. 1 wire and heavy scrap. Although Figure C-l indicates that scrap was processed by four secondary smelters in 1997, currently only two secondary smelters are operating (Chemetco in Hartford, 111. and Southwire in Carrollton, Ga). Chemetco produces anodes, which are sent to another processor (Asarco) for electrolytic refining. Southwire does its own electrolytic refining. C.2.2 Scrap Handling and Preparation Copper scrap is collected by a national network of processors and brokers. The scrap is visually inspected and graded. Chemical analyses are performed when necessary. Loose scrap is baled and stored until needed. Alloy scrap is segregated and identified by the alloy and the impurity content of each batch. Scrap of unknown composition may be melted and analyzed to determine its chemistry (CDA 1998a). The major processes involved in secondary copper recovery are scrap metal pretreatment and smelting. Pretreatment prepares the scrap copper for the smelting process. Smelting is a pyrometallurgical process used to separate, reduce, or refine the copper. ------- 28% (383) 7 primary smelters 4 secondary smelters 6 fire refiners 59% (809) (151) ! electrolytic plants (233) 10% (132) brass and bronze ingot makers 35 brass mills 15 wire-rod mills 600 foundries, chemical plants, misc. manufacturers 4% (55.4) Figure C-l. Simplified flow diagram for copper-base scrap in 1997. Units are percent of total copper consumed from copper-base scrap and metric tons (in parentheses). Pretreatment includes cleaning, and concentrating the scrap materials to prepare them for the smelting process. Pretreatment can be accomplished by: (1) concentration, (2) pyrometallurgical, or (3) hydrometallurgical methods. These methods may be used separately or combined. Pretreatment by concentration is performed either manually or mechanically by sorting, stripping, shredding, or magnetic separation. The resulting scrap metal is then sometimes briquetted in a hydraulic press. Pretreatment by the pyrometallurgical method includes sweating, burning of insulation (especially from scrap wire), and drying (burning off oil and volatiles) in rotary kilns. The hydrometallurgical method includes flotation and leaching with chemical recovery. After pretreatment the scrap metal is ready for smelting (U.S. EPA 1995). C-9 ------- Table C-3. Copper Recovered from Scrap Metal Processed in the United States in 1997 Scrap Kind of Scrap Form of Recovery New Scrap Old Scrap Copper-base Aluminum-base Nickel-base Zinc-base Total Copper-base Aluminum-base Nickel-base Zinc-base Total Grand total As unalloyed copper As alloys and compounds At electrolytic plants At other plants Total Brass and bronze Alloy iron and steel Aluminum alloys Other alloys Chemical compounds Total Grand total Amount (t) 909,000 46,800 91 — 955,891 465,000 30,300 28 19 495,347 1,451,238 233,000 161,000 394,000 979,000 743 77,500 113 252 1,057,608 1,451,608 Source: Edelstein 1998 Note: Totals differ due to round-off errors. C.2.3 Copper Refining Operations Copper scrap is utilized by both primary and secondary producers of copper. Locations in the copper refining process where copper scrap may be introduced are summarized in Figure C-2. This diagram does not address the large amount of copper-alloy scrap, which is used by brass mills, ingot makers, and foundries. Based on the data in Table C-4, the figure illustrates the disposition of 63% of old scrap. In this figure, typical secondary copper operations are described by the dashed boxes. C-10 ------- Secondary smelters use several processes that are equivalent to those employed as primary pyrometallurgical processes for mined copper ores. A first stage smelting process is most commonly performed in either a blast furnace, reverberatory furnace, or an electric furnace. This is followed by treatment in a converter furnace and then in an anode furnace. The copper may be further purified by electrolytic refining. Depending on the grade, copper scrap may enter the flow stream at numerous locations. Some slag from the process is sold or landfilled; the remaining slag is recycled back into the smelting furnace because of its copper content. Sulfur dioxide, a by-product gas from primary smelting, can be collected, purified, and made into sulfuric acid for sale or for use in hydro-metallurgical leaching operations. Each of the major processes used in recycling copper scrap is described below. Table C-4. Copper Consumption from Copper-Base Scrap in the United States in 1997 (t) Type of Operation Brass/bronze ingot makers Copper refineries Brass and wire-rod mills Foundries and manufacturers Chemical plants Total From New Scrap 35,200 91,400 771,000 11,200 252 909,052 From Old Scrap 96,500 292,000 32,800 43,900 — 465,200 Total 132,000 383,000 804,000 55,100 252 1,374,352 Note: Totals differ due to round-off errors. C.2.3.1 Copper Smelting Practices Blast Furnace The vertical shaft furnace, also known as the blast furnace or cupola, has the ability to smelt copper-bearing material of an extremely diverse physical and chemical nature. It is the unit that is commonly employed in the pyrometallurgical treatment of low-grade secondary copper material and largely controls the metal losses in the system (Nelmes 1984). Low-grade copper scrap containing skimming, grindings, ashes, iron-containing brasses, and copper residues is typically smelted in a blast furnace, where coke is added as a reductant and limestone is added to assist in forming a calcium-iron-silicate slag. The molten "black copper" product from the blast furnace is transferred via a ladle to a converter for further purification. It is then fire refined and electrorefined. Dusts from the blast furnace are collected in a baghouse. C-ll ------- o float. cone. (20-35% * Cu) 1 I slac oxidized ores and wastes P flash smelter furnace i i T solvent extraction l matte (40-70% Cu) Cu-rich slag 1 Cu-rich strip sol'n electrowinning plant spent electrolyte Cu start! #1 and #2 scrap (limi i converter k ) ted) i blister Cu anode Cu ^ furnace anodes ^ electrolytic (>98.5% f Cu) refm Cu-rich slag off- gases emissions control system blowdown slurry (KD64) 1 ^ r I e "~ plant I I anode elec slimes b i i "! i" . i i i • ' i 1 smelting i 1 • 1 1, I P»i converter 1 1*£ furnace • 1 "I | 1 " 1 r » i i i ' i * i . _ _ j L _____' l_ f low-grade refinery Cu scrap brass acid plant ng Cu #1 scrap V V -'u vertical Ca » shaft (99 9+% (cathode) Cu) furnace trolyte eed 1 1 slabs anode \ in9ot furnace 1 rod.ca|- 1 1 J, #1 and #2 scrap, blister Cu T sulfuric acid Figure C-2. Process diagram for the flow of copper scrap in primary and secondary copper refining. (Dashed boxes represent secondary processor's operations.) ------- The ranges of compositions for blast furnace process streams, as reported by several authors, are summarized in Table C-5. The feed to the cupola described by Opie et al. (1985) contained about 30% copper. The average dust composition from a cupola has also been reported by Garbay and Chapuis (1991): • Cl 3% • Cu 4% • Zn (ZnO) 55% • Sn 4% • Pb 9% The dust composition, which is typical of French smelting practice, is encompassed by the ranges of values in Table C-5. Table C-5. Composition of Process Streams from the Smelting of Copper Scrap in a Cupola Blast Furnace (%) Item Cu Ni Sb Sn Fe Zn Pb SiO2 Cl F CaO A1203 Other Black Copper Kusik and Kenahan 75-88 0.1 - 1.7 1.5 3-7 4- 10 1.5 Nelmes 80 4 4 5 O 4 <1 Opie 65-70 7.5- 12 0.5- 1.5 2-4 5- 10 2-4 2-4 Slag Nelmes 0.9 1.5 0.3 30 3 0.6 27 14 9 15 Opie 1.5-2 1 - 1.5 1 -2 1 -2 30-35 2-4 1.5-3 Dust Kusik and Kenahan 0.1 0.1 5- 15 58-61 2-8 0.1 -0.5 Nelmes 1.5 1 50 15 32.5 Opie 8- 12 0.1 -0.5 0.3 -0.8 1.5-2 20-35 13 - 15 4-7 6- 10 1 -5 Sources: Kusik and Kenahan 1978, Nelmes 1984, and Opie et al. 1985. C-13 ------- During the blast furnace smelting operation, the scrap charge is fed onto a belt conveyer, which in turn discharges into one of two skip hoist buckets (Browne 1990). These buckets are hoisted and alternately dumped into opposite sides of the furnace. Coke is added as a reducing agent along with silica, lime, or iron oxide. Air is injected by means of tuyeres. The copper-bearing material initially enters at the top of the furnace into a zone at 400-600°C. It subsequently descends into the tuyere zone and increases in temperature to about 1,400°C5 (Schwab 1990). According to Nelmes (1984), many secondary copper blast furnaces have an area of about 35 ft2 with the range being from 12 to 140 ft2. Assuming a melting rate of 6 tons/ft2/day, a typical blast furnace would have an output of 210 tons/day. A mixture of molten copper and slag flows down a launder into an oil-fired rocking furnace that can rotate. This furnace is large enough to give the slag sufficient time to separate from the copper. Rotating the furnace in one direction allows the liquid copper to fill a preheated ladle on a rail car below the rocking furnace. Rotation in the opposite direction allows the slag to pour into a granulating trough. Granulation is accomplished by impinging the liquid slag with a high pressure jet of water. The slag and water are collected in a pit that is large enough to remove the slag with a clamshell bucket on a crane. When granulated blast furnace slag is dried, crushed, and screened, it is used to manufacture a variety of commercial products. It is useful for making a variety of abrasives, filler for asphalt shingles, roofing sealers, grit for sand blasting, road surface bedding, and in the manufacturing of mineral wool and light-weight cement aggregates (Nelmes 1984, Schwab 1990, Mackey 1993). The metal content of the slag is typically 1% copper or less (Mackey 1993). Some slag is stored or discarded in piles on site (U.S. EPA 1995). In some cases the slag may be treated for recovery of additional metal values prior to granulation. Opie et al. (1985) describe a processing step in which the blast furnace slag is pyrometallurgically treated in an electric arc furnace with 2% coke added as a reductant. The arc furnace temperature is 100 to 200°C higher than in the blast furnace. A small amount of additional black copper is produced, dust is collected in a separate baghouse, and a slag with reduced metal values is obtained. The composition ranges for these products are presented in Table C-6 and are based on treating the blast furnace slag described by Opie et al. (1985) (see Table C-5). The melting point of pure copper is 1,083°C. C-14 ------- Table C-6. Composition of Products Obtained from Treating Copper Blast Furnace Slag in an EAF Element Cu Ni Sb Sn Fe Zn Pb Black Cu (%) 55-60 5- 10 0.5- 1.5 2-4 5-7 1.5-2.0 1.0- 1.5 Final Slag (%) 0.2-0.5 0.2-0.4 0.1 -0.20 0.05-0.1 30-35 0.5- 1.0 0.5- 1.0 Baghouse Dust (%) 1 -2 0.2-0.3 0.1 -0.2 1.5-3.0 0.5-0.7 45-55 15-20 Source: Opie et al. 1985 For a 100-ton blast furnace charge consisting of copper scrap, coke, and slagging agents, the expected output is 40 tons of black copper, 40 tons of slag, and 5 tons of baghouse dust (Nelmes 1984). Carbon in the charge is converted to CO/CO2, which is exhausted through a stack. The overall elemental partitioning for a copper blast furnace, based on these mass partitioning values and the elemental compositions included in Table C-5, is presented in Table C-7. Table C-7. Partitioning During Blast Furnace Smelting of Copper Scrap (% recovery) Output Metal Dust Slag Cu 98.64 0.25 1.11 Sn 90.4 2.82 6.78 Fe 14.29 — 85.71 Zn 24.49 51.02 24.49 Pb 61.78 28.96 9.26 Ni 63.9 — 36.1 A1203 — — 100 CaO — — 100 SiO2 — — 100 Source: Nelmes 1984 Table C-7 does not include 1.6 tons of "Other" material reporting to the dust and 6.0 tons reporting to the slag. Reverberatory Furnace Reverberatory furnace smelting began in the nineteenth century. It still accounts for a significant fraction of both primary and secondary copper production and recycling of secondary scrap metal. Disadvantages of these furnaces are the long melting cycle times and low fuel efficiencies (Davenport 1986). C-15 ------- In a reverberatory furnace, the scrap copper is charged into one or more piles located behind one another, in front of several high capacity end-wall-fired burners. These high capacity conventional burners typically are fired above the copper scrap and use the reverberatory effect for heat transfer, i.e., re-radiation from the refractory roof and walls to the scrap. During the melting cycle, when the process requirements for energy are high, the surface of scrap exposed to the flame radiation and to radiative heat transfer from the furnace refractory surfaces is small relative to the total surface area of the scrap. This is because the top layers of scrap shade the interior scrap surfaces from the radiation, resulting in low rates of radiative transfer to the entire scrap charge. In addition, convective heat transfer to the interior of the scrap charge is limited by low circulation of gases within the scrap. A typical reverberatory furnace is charged with approximately 250 tons of scrap and about 100 tons of liquid metal in order to maintain a 24-hour operating cycle; the melting portion of the cycle is 8 hours. This represents an average "melt-in" rate of cold scrap of about 31 tons per hour (Wechsler and Gitman 1991). The reverberatory furnace is charged by fork-lift trucks or by charging machines. Impurities are removed during melting by air oxidation and skimming away the resultant slag. The oxygen content of the melt is then reduced to the desired level (e.g., 0.03% to 0.04%) by adding a hydrocarbon source (e.g., natural gas) and the copper is cast into shapes such as cakes, billets, or wire-bar. In some cases melting of copper scrap in a reverberatory furnace may be the only step in the refining process. At Reading Tube Co., for example, No. 1 copper scrap is the sole feed. All of the incoming scrap is visually inspected for known forms of suspect copper. An in-depth visual inspection is made of selected samples from the scrap; chemical analyses are taken from samples to screen for impurities. (The scrap is not monitored for radioactivity.) The scrap is charged into a 200-ton reverberatory furnace,6 melted, and blown with air or oxygen to oxidize impurities. The oxide slag is skimmed from the melt. The melt is covered with charcoal and "poled" to remove oxygen. In the poling process, green hardwood logs are thrust into the molten copper bath, where the hydrocarbons react with the oxygen to form CO/CO2. The molten copper is then laundered. In this process the copper flows under charcoal into a ladle which is covered with a carbon-based product. The laundering removes additional oxygen from the melt. Final deoxidation is promoted by the addition of phosphorus; the melt is cast into billets for subsequent One heat per day is typically produced. The furnace undergoes an annual maintenance shutdown. Reading also operates a shaft furnace, which can produce 100 tons per day. C-16 ------- fabrication into tubing (Reading 1999). The slag is sold to an outside processor for recovery of additional copper values. Offgases from the furnace pass through an after-burner to convert CO to CO2 and to destroy any hydrocarbons; they are then exhausted through a stack. Stack offgas is monitored for total particulates, opacity, and SO2. Electric Arc Furnace The electric arc furnace (EAF) is also used in secondary copper smelting (5/26/99). At Halstead Industries (now part of Mueller Industries, Inc.) in Wynne, Arkansas, bales of copper scrap, cathode sheets, or copper ingots (from Codelco in Chile) are preheated with natural gas to about 1,000°F and charged into a 16,000-volt EAF7. In the EAF, the copper is melted and heated to between 2,200-2,300°F and then poured into a graphite-covered launder at a rate of 640 pounds per minute. Phosphorus pellets are added to the molten copper stream for deoxidation8. The copper flows from the launder to the casting machine, where four logs, each 9 inches in diameter and 25 ft long, are simultaneously cast at a rate of about 8 inches of ingot length per minute. The logs weigh 6,160 Ib each. The launder then swings to a second set of molds while the logs produced from the first set of molds are raised from the casting pit under the molds and transferred with an overhead crane to the billet cutter. At the billet cutter each log is sawed into 14 extrusion billets, each 20.25 inches long and weighing 420 Ib. The EAF is rated at 72 tons and produces 310 to 330 tons per day (Blanton 1999). The charge is 75% to 80% scrap and 20% to 25% cathodes or ingots. Incoming scrap is screened with a Geiger counter for radioactivity. Plant procedures call for an alert at twice background and automatic rejection of the shipment at three times background. In the past four to five years there have been two alarms, both traceable to truck drivers who had been treated with radioisotopes. The furnace is equipped with a baghouse for dust collection. The dust generation rate is about 5 Ib/ton and the dust contains 73% to 76% copper, some zinc, small amounts of iron and tin, and about 0.1% to 0.15% lead. Significant carbon, attributable to melt poling, is also present. Slag is skimmed from the furnace using hand rakes. The slag contains 30% to 50% copper, considerable carbon, Mueller Industries also has smelting facilities in Fulton, Mississippi where, until recently, all melting was done in a shaft furnace. They have now added a Maerz reverberatory furnace at that production location. Q The alloy produced is C12200 or Phosphorus-Deoxidized High Residual Phosphorus Copper, containing 99.9% copper (min.). C-17 ------- calcium from bone ash (a slagging agent), zinc, and iron oxide. Both the baghouse dust and the slag are sold to Chemetco for further processing. A metric for slag generation was not available. C.2.3.2 Copper Converting The product from the smelting furnace may contain significant amounts of Fe, Sn, Pb, Zn, Ni. and S. These elements are removed either by reduction and evaporation or by oxidation. At smelting temperatures, oxides of most metals are more stable than CuO or Cu2O. Thus, from an equilibrium thermodynamics perspective, these metals would be transferred to the slag under oxidizing conditions. Impurity metals with high vapor pressures (e.g., Pb, Cd, Zn) or with high- vapor-pressure oxides (e.g., SnO, Cs2O, P2O3) may volatilize and be collected in the zinc-rich dust. Tin is recovered from baghouse dust and used as tin/lead alloy for solder, and zinc is recovered and converted to ZnO for the pigment industry (Gockman 1992). The conversion process employs either a Peirce-Smith converter or a top blown rotary converter (TBRC). Oxygen-enriched air or pure oxygen is used for the removal of impurities (Davenport 1986; Roscrow 1983). The charge is melted under reducing conditions to avoid premature oxidation of copper. Lead, tin, and zinc are also reduced to metals. Zinc-rich dust is collected in a baghouse. Iron reacts with silica flux to form a silicate slag. The furnace is then run in an oxidizing mode using air or oxygen. The remaining iron, zinc, tin and lead are removed. When processing black copper produced from scrap in a converter, the converter must be "blown hard" to remove nickel, tin, and antimony from the melt. This results in a slag containing over 30% copper. The slag is returned to the blast furnace for copper recovery (Opie et al. 1985). The resultant converter product is blister copper (-96% Cu). A typical furnace can produce from 4,000 to 15,000 tons per year of blister copper (O'Brien 1992). Based on metal content, the baghouse dust may be shipped to zinc smelters or to tin and lead refiners for metal recovery. The composition of the blister copper, the slag, and the baghouse dust from a converter operation based on secondary copper smelting is summarized in Table C-8. C-18 ------- Table C-8. Composition of Converter Products from the Smelting of Copper Scrap (% Element Cu Ni Sb Sn Fe Zn Pb Blister Copper 94-96 0.5- 1.0 0.1 -0.3 0.1 -0.2 0.1 -0.3 0.05-0.1 0.05- 1.0 Slag 30-35 10- 15 0.5- 1.5 2-4 20-25 1.0- 1.5 2.5-4.0 Baghouse Dust 2-3 0.5- 1.0 0.5- 1.5 10-20 0.5- 1.0 25-35 20-25 Source: Opie et al. 1985 C.2.3.3 Fire Refining The blister copper from the converter is then processed in an anode furnace, which is generally some type of reverberatory furnace. Anode production is the last processing step prior to electrolytic refining and is called "fire refining." Sulfur and other readily oxidizable elements are removed by air oxidation. The dissolved oxygen is then removed from the melt by reaction with hydrocarbon gases prior to anode casting. During fire refining, the melt is first saturated with O2 (about 0.8 to 0.9% O) and the oxygen is then decreased to about 0.2%. Oxidized impurities are collected in the slag, which is recycled either on-site or at another refinery. The anodes are then cast in copper molds on a rotating horizontal wheel. Anode thickness is controlled by weighing the copper poured. The anodes contain about 99.5% copper with impurities such as Ag, As, Au, Bi, Fe, Ni, Pb, Sb, Se, and Te (Kusik and Kenahan 1978, Davenport 1986). Garbay and Chapuis (1991) list the composition of fire-refined anodes produced from a French smelting operation in a 250-t reverberatory furnace, as listed in Table C- 9. Schloen (1987) summarized typical anode chemistries at nine U.S. electrolytic copper refineries which were operating at the time. Results of this survey are presented in Table C-10. C.2.3.4 Electrolytic Refining The final stage in copper purification employs an electrolytic refining process that yields copper which may contain less than 40 ppm of metallic impurities (Ramachandran and Wildman 1987). C-19 ------- During electrorefming, copper anodes and pure copper cathode starter sheets are suspended in a CuSO4-H2SO4-H2O electrolyte, through which an electrical current is passed at a potential of about 0.25 Vdc. The electrolytic refining process requires 10 to 14 days to produce a cathode weighing about 150 kg. During electrolysis the copper dissolves from the anode and deposits on the cathode. Impurities such as Au, Ag, and other precious metals, as well as Pb, Se, and Te collect in the anode slimes9. These anode slimes are collected and sent to a precious metals refinery (Davenport 1986). Other elements such as Fe, Ni, and Zn dissolve in the electrolyte10 and are removed from the copper electrolysis cells in a bleed stream. The bleed stream is sent to "liberator" cells, where the solution is again electrolyzed and soluble copper is plated out on insoluble lead anodes. The bleed stream is then treated for NiSO4 recovery by concentrating the solution in evaporator vessels, where NiSO4 crystals precipitate. The remaining liquor is called "black acid." Both the NiSO4 and the black acid are typically salable products (Kusik and Kenahan 1978). Table C-9. Composition of Anodes Produced in a 250-t Reverberatory Furnace (ppm) Ag As Pb Ni 600 1,110 2,200 500 Sn Sb Se Te 400 250 100 100 Bi Fe Zn S 20 50 100 10 Source: Garbay and Chapuis 1991 Note: Balance Cu The processing conducted at the ASARCO's Amarillo copper refinery (Ramachandran and Wildman 1987) is illustrative of electrorefming operations. Blister copper is shipped to the refinery in solid bottom gondola rail cars, which are unloaded either in a storage area or at the Anode Casting Department. Blister copper from the storage area is transferred to the Anode Casting Department via 11-ton fork lifts. Usage of blister copper is 8,500 tons per month (tpm). Number 2 copper scrap is received loose in box cars or trucks. The scrap is sampled and briquetted into bales which measure about 40 x 36 x 17 inches. Scrap usage is up to 6,000 tpm. The blister copper and the scrap are melted in a 350-ton Maerz tilting reverberatory furnace, y According to U.S. patent 4,351,705, a typical slimes composition is 5-10% Cu, 4-8% Ni, 6-8% Sb, 15-25% Sn, 5- 12% Pb, 0-2% Ag, and 4-8% As. According to Davenport (1986), As, Bi, Co, Fe, Ni, and Sb report to the electrolyte. C-20 ------- which operates on a 22-hour cycle. Copper for anodes, each weighing about 765 pounds, is poured into molds in a casting machine. The finished anodes are transferred to the tankhouse with a 20-ton straddle carrier. The refinery also uses a 50 ton per hour shaft furnace to remelt anode scrap from outside sources and reject anodes. Output from the shaft furnace is transferred to a 15-ton holding furnace, which feeds the same casting wheel as used with the reverberatory furnace. Monthly anode production is about 22,000 tons. Typical anode chemistry is: • Cu 98.6 - 99.4% • Ni 0.04 - 0.08% • Sb 0.05 -0.08% • As 0.03 - 0.09% • Se 0.06 - 0.07% The tankhouse contains six independent modules, each with its own rectifier, circulation system, reagent system, and operating crew. Each module contains 400 cells. The annual output of the plant is about 460,000 tons. Additional anodes required to maintain tankhouse operation at capacity are obtained from external sources. A typical analysis of the cathode copper is: • Cu 99.96% •S 6 ppm • Se <1 ppm • Sb 1 ppm • As 1 ppm • Bi 0.2 ppm • Fe 2 ppm • Nickel 2 ppm • Pb <1 ppm • Sn <1 ppm • Zn < 3ppm C-21 ------- Table C-10. Anode Compositions at Various U.S. Electrolytic Copper Refineries Element/ units Cu% Ag ppm Se ppm Te ppm As ppm Sb ppm Bi ppm Pb ppm Ni ppm O, ppm 00 CO CD co ,-; oo <~! CO W ~^ ^ |_ O 03 oo h- co < 0 99.2 120 50 50 50 350 50 1,500 1,500 2,000 o O 00 o •a O _: CL — CD-I 8:° o-1 O ••— ' CO iZ "oo CD CO O UJ 99.6 210 20 130 100 80 20 1,600 500 1,200 o O oo CD ^ co CD" ce: .E i— n Q. CD o — o € O ?• 99.5-99.8 700 25 ~ 200 50 5 80 150 1,000-1,500 Q_ O o O) 'E M — CD ce: •Q CD 0 o CD t~~ ^ C £Z '*••' CD CO ^ CO 99.63 435 490 69 500 210 41 380 510 1,228 ^ c5 -^ c 2 "CD a: c" 5 to o c 0 m U* E a) CD .c ^ CO 99.6 403 500 140 560 50 33 140 220 1,960 6 0 N CD< o i CO CO E^ O) c cc co ^ CO 99.78 175 415 6 24 73 3 63 121 900-1,200 Q. O O O) £— ,E CD a: CD . O) X •a CD Q - o Q- co CD Q. .c CL UJ 99.5 225-300 200-450 25-50 25-50 35-50 5-15 15-150 100-700 1,400-2,800 ^ O CO .^ Q CD a. CJ O E § > H— < £ o ^ fc O CO CO O 99.3 600 ~ 20 500 700 30 1,000 3,000 2,000 CO CD ------- A continual bleed of electrolyte is taken from the electrorefining cells to a separate building containing copper-removal cells. Here the copper is passed through a number of primary liberator cells plumbed in series, where the copper content of the electrolyte is reduced from 40 to 20 g/L. The cathodes from these primary cells are returned to the Anode Casting Department for recasting into new anodes. A portion of the partially purified liquor is returned to the main tankhouse and the balance is sent to secondary recovery cells, where the copper content of the electrolyte is further reduced to about 1 g/L. The cathodes from the secondary cells may be returned to the Anode Casting Department or shipped to a smelter in El Paso, Texas for reprocessing. The treated electrolyte, which contains 15-20 g/L of Ni, is processed through one of two submerged combustion evaporators to produce NiSO4. A single evaporator can produce about 115 tpm of NiSO4 on a dry-weight basis11. The black acid remaining after nickel removal is either returned to the tankhouse for use in acid makeup or is used to leach the slimes. The crude nickel sulfate, which contains about 5% H2SO4 and 3% H2O, is shipped to nickel producers. Slimes are processed at the electrorefinery. C.2.3.5 Melting, Casting, and Use of Cathodes The cathodes are washed, melted, and cast into shapes for fabrication and use. The melting is usually done in a vertical shaft furnace in which stacks of cathodes are charged near the top and melt as they descend, heated by combustion gases. The operation is continuous, and the molten copper may be cast and rolled to form rod for wiremaking, or into slabs and billets for other wrought products. C.2.3.6 Slag Handling The slags from the copper converters and the anode furnaces are rich in copper and are returned to the smelting furnace for recovery of additional copper values. The smelting furnace slag is stored or discarded in slag piles on site. Some slag is sold for railroad ballast and for blasting grit (U.S. EPA 1995). Most of the radioactive contaminants would end up in the slag because they tend to be more easily oxidized than copper. If the plant processes 460,000 tons of copper anodes containing 0.08% Ni and produces 92% NiSO4, the nickel sulfate production would be about 88 tpm if all the nickel forms NiSO4, which in turn contains 38% Ni by weight. C-23 ------- C.2.3.7 Offgas Handling Offgases from the converters at primary producers are collected by a hood system and processed through an emission control system, which typically consists of an electrostatic precipitator (ESP) and a wet scrubber12. The scrubbed gas is processed through an acid plant and converted to sulfuric acid. Since secondary producers do not handle high sulfur matte, they do not have acid plants in their systems. C.2.3.8 Illustrative Secondary Smelter Operations at the Southwire Company in Carrollton, Ga. are briefly described to indicate the complexity and variability of the operations at a large secondary refiner. Examples of the types of scrap handled by Southwire include blister copper, spent and reject anodes, No. 1 copper scrap, No. 2 copper scrap, No. 3 copper scrap, and miscellaneous copper-bearing materials (e.g. bronze, brass, and small motors) (2/24/99). Southwire has a fixed Nal scintillation detector system built by Eberline to monitor incoming trucks for radioactive contamination. The system has alarmed three or four times—once by radon in propane from a Texas salt dome (McKibben 1999). Southwire uses a blast furnace to process low-grade scrap, a top-blown rotary converter to process the blast furnace output into blister copper, a reverberatory furnace to melt No. 2 scrap, and a shaft furnace to melt and refine blister copper and No. 1 scrap and produce anodes. The high copper slags from the other furnaces are returned to the blast furnace for the recovery of additional metal values. The blast furnace slag is granulated, dried, and screened. It is sold to the roofing industry for use in shingles (Gerson 1999). The Southwire flowsheet is shown in Figure C-3 (McDonald 1999). The brick plant in Figure C-3 was scheduled to be replaced by a new central mixing facility (Capp 1997). In the new facility, baghouse dust from the Maerz reverberatory furnace, the anode shaft furnace, the anode holding furnace, and the slag plant are collected in dust-tight tote bins. When the tote bins are full they are transported by fork-lift truck to the central mixing facility. Tote bins are filled approximately once per 12-hour shift from the reverberatory furnace While some sources have suggested that scrubber blowdown at primary copper facilities is RCRA-regulated waste (K064), this is not the case. In a 1990 decision, a federal district court remanded the K064 listing to EPA for reconsideration. No further action has been taken by the Agency. The wastes may be characteristically hazardous due to acidity or metals content. C-24 ------- baghouse, once per shift from the slag plant baghouse, and once every one to three days from the other sources. Dust is transported from the tote bin via an enclosed screw auger to a 200 ft3 storage silo (called a day bin), which holds about a three-day inventory. The dust is then moved by a second enclosed screw auger to an agglomeration unit with a design capacity of 20 tons per hour (tph), where water is added and a paste is produced. This paste is transferred to a wet bin for storage until the product is needed for feed to the blast furnace. When required, the paste is moved with a front-end loader to the blast furnace charge beds, where it is blended with other feed materials. The central mixing facility has an annual design input of about 51,100 tons per year (TPY) of baghouse dust. The facility design calls for limiting emissions through two low stacks (18 and 20 feet above grade) to 1.64 tpy of particulate material with the following indicated contaminants: • As 0.07 tpy • Cr 0.05 tpy • Se 0.05 tpy • Cd 0.004 tpy • Ni 0.004 tpy • Sb 0.000 tpy • Co 0.000 tpy • Mn 0.000 tpy • Be 0.000 tpy These estimates were based on the analysis of baghouse fines. Each furnace has at least one baghouse and some have a backup. Dust from the blast furnace is disposed of in a hazardous waste landfill because of Cd, Pb, and other heavy metals. Dust from the converter is sold to an overseas customer, who recovers metal values such as Pb, Sn, and Zn. Dust from the reverberatory furnace and the shaft furnace is returned to the process as described above. It is difficult to obtain a figure of merit for dust generation because it varies significantly with the type of scrap being processed. For example, a high-brass furnace charge will generate more zinc dust. C-25 ------- COPPER WIRE MANUFACTURING o to TRUCK SCRAP RAILCAR GO oa BtTSTOT SCRAP BALER MA'ERZ" REVERBATORY rURNACE LOW-GRADE MATERIALS FUGITIVE SOURCES sc PR FINES 1 »» RAP ^" E-TREATMENT BRICK PLANT ANODES j 137 FT f'""1 1STACK : IBAGHOUSEJ IBAGHOUSEJ 1 i 1 t I f T f T f 1 SAFflR"'] f ] f""* ; r"* i JBURNER! IBAGHOUSEJ ^AGHOUSE! BAGHOUSE) : I l J t j ' i i I I TANKHOUSE ~f~f~ it : 4 TANK HOUSE BOILERS^ Jl / PLANT] EE FTTATDTT CATHODE PACKAGING 1 BY-PRODUCTS O L OO GO jDEMISTER jH2 RECOVERY I ^ i PROCESS FLOW EMISSION FLOW TO WIRE PLANTS STEAM FLOW ! COPPEF bo co ROD 0 Figure C-3. Flow Diagram of the Copper Division of Southwire (CDS) ------- Anodes are electrolytically refined. The anode slimes are sold to an offshore processor for precious metal recovery. Copper is removed from the electrolyte bleed by electroplating. The solution is then evaporated. Nickel sulfate is crystallized and recovered for sale. Cathodes from the electrorefming operation are melted in a shaft furnace and cast into copper rod. In 1998, the output of the rod-mill shaft furnace was about 342,000 tons (McDonald 1999). Operations at Chemetco, a secondary smelter in Hartford, Illinois are somewhat different. Chemetco has four 70,000 Ib reverberatory furnaces and four top-blown rotary converters to process scrap (Riga 1999). They process scrap ranging from high-grade copper wire to low- grade slags and skims. Slags are sold for railroad ballast, road beds, and asphalt shingles. Anodes are sold to Asarco for electrorefming. C.2.4 Brass and Bronze Ingot Production As shown in Figure C-l, about 10% of copper-base scrap is consumed by brass and bronze ingot makers. At the ingot manufacturer, scrap is melted in a reverberatory furnace. Fluxing agents such as borax and sodium nitrate are added. Alloying agents such as tin may also be included in the furnace charge. Zinc evolved in the melting process is collected in a baghouse. Slag is either returned to a smelter for reprocessing or shipped for disposal (Kusik and Kenahan 1978). Aluminum bronze is melted in gas- or oil-fired crucible furnaces, coreless induction furnaces, or in reverberatory furnaces (for very large castings) (U.K. CDA 1999). The furnace charge typically involves addition of cathode copper, aluminum (either as ingot or a 50% Al-50% Cu master alloy), and iron and nickel (either in elemental form or as a master alloy). Process scrap is generally added when the ingots are remelted to produce the final castings but may be added at the end of the alloying schedule. During melting, most of the copper together with the iron and nickel are introduced into the furnace under a charcoal blanket and the melt is heated to about 1,300°C. The remaining copper is then added, the charcoal is removed and the aluminum is charged. A small amount of cryolite or fluoride flux is then stirred into the melt to clean entrapped metal from the dross before pouring the melt into ingot molds. C.2.5 Brass Mills Brasses are alloys of copper with up to 40% zinc. Other alloying elements such as Al, Fe, Mn, Pb, and Sn may be added at levels of up to a few percent of each metal, depending on the specific C-27 ------- alloy being produced. As shown in Table C-l 1, brass mills are major consumers of yellow and red brass scrap. An example is the Chase Brass and Copper Company, which produces brass rod primarily from scrap. Chase currently has an annual capacity of about 300 million Ib per year and is expanding to 400 million Ib per year. The scrap is melted in four induction furnaces and cast into logs, which are 23 ft long and 10 inches in diameter. About 80% of their scrap requirements are obtained through purchase and tolling arrangements with their customers. In 1997 there was a price differential of 5 cents per pound between the metal selling price to the customer and the metal buying price (i.e., the scrap price) from the customer. The balance of their requirements are purchased from scrap dealers at the free-market price. Chase uses hand- held detectors to check scrap from unknown (i.e., open-market) sources for radioactivity. They have had no instances where any activity has been detected in the scrap. Several million pounds are typically in inventory at the plant site. A baghouse system is used to collect dust from the furnace offgas. Dross is removed from the furnace and run through a vibratory screening system to collect metal for internal recycle. Both the undersize from the dross processing and the baghouse dust are drummed and sold to an off-site reprocessor (Warner 1999, Woodserman 1999). The reprocessor treats these waste streams with mineral acids and then crystallizes various metal salts from the solutions. Typically, the salts are sold to the steel industry for use in fluxes. Chase seldom uses copper scrap in its melting operations. Use of copper in the furnace charge requires a higher melting temperature, which increases zinc losses from the melt. Chase does not have a figure of merit for baghouse dust production. The value is quite variable depending on the alloy being melted, the quantity of scrap in the furnace charge, etc. Olin Brass in East Alton, 111. produces 60 to 70 different copper and brass alloys. Most of the scrap used is either run-around (internal) scrap or customer returns (either direct or handled by a broker). A portable spectrometer may be used to check the chemistry of an incoming truckload of scrap. Occasionally, pure copper is used for selected products. Melting is done in small induction furnaces that feed a large holding furnace. The furnace charge is typically baled scrap. Most Olin alloys are cast by the direct chill method, in which multiple ingots are cast simultaneously. Each rectangular cross-section ingot is about 25-ft long and weighs 18,000 Ib. The ingots are reduced to sheet and strip via a series of hot and cold rolling operations (Olin 1995). Furnace offgas is processed through cyclone separators and a baghouse. During melting, dross formation is not intentionally promoted. However, use of highly reactive alloying additions may enhance dross formation. Dross disposition practices, which are proprietary, are designed to maximize process economics (presumably by using some sort of recycling). The same considerations apply to treatment of baghouse dust (Shooter 1999). C-28 ------- Table C-ll. Consumption of Copper-Base Scrap in 1997 Scrap Type and Processor No. 1 wire and heavy: Smelters, refiners, and ingot makers Brass and wire-rod mills Foundries and misc. manufacturers No. 2 mixed light and heavy: Smelters, refiners, and ingot makers Brass and wire-rod mills Foundries and misc. manufacturers Total unalloyed scrap: Smelters, refiners, and ingot makers Brass and wire-rod mills Foundries, and miscellaneous manufacturers Red brass:3 Smelters, refiners, and ingot makers Brass mills Foundries and miscellaneous manufacturers Leaded yellow brass: Smelters, refiners, and ingot makers Brass mills Foundries and miscellaneous manufacturers Yellow and low brass: all plants Cartridge cases and brass: all plants Auto radiators Smelters, refiners, and ingot makers Foundries and miscellaneous manufacturers Bronzes Smelters, refiners, and ingot makers Brass mills and miscellaneous manufacturers Nickel-copper alloys: all plants Low-grade and residues Smelters, refiners, and miscellaneous manufacturers Consumption (t) 149,000 413,000 35,800 230,000 34,900 2,770 379,000 448,000 38,600 58,300 8,780 10,100 28,800 404,000 1,930 53,900 66,800 72,200 4,470 12,100 14,900 17,800 87,100 Source: Edelstein 1998 Includes composition turnings, silicon bronze, railroad car boxes, cocks, and faucets, gilding metal, and commercial bronze. C-29 ------- Table C-ll (continued) Scrap Type and Processor Other alloy scrapb Smelters, refiners, and ingot makers Brass mills and miscellaneous manufacturers Total alloyed scrap Smelters, refiners, and ingot makers Brass mills Foundries and miscellaneous manufacturers Total Scrap Smelters, refiners, and ingot makers Brass and wire-rod mills Foundries and miscellaneous manufacturers Consumption (t) 38,400 6,570 303,000 558,000 24,100 682,000 1,010,000 62,700 Includes refinery brass, beryllium copper, and aluminum bronze. C.2.6 Aluminum Bronze Foundries Aluminum bronzes may be produced from prealloyed ingots (see Section C.2.4) or from directly alloyed components. In the latter case, the copper is melted together with copper/iron and copper/nickel master alloys at 1,200°C under a charcoal cover (U.K. CDA 1999). The melt is then deoxidized with a copper/manganese alloy and the charcoal cover is removed. The manganese oxide is skimmed off at this point to prevent its subsequent reduction by aluminum. An aluminum/copper master alloy is next added in small increments. The melt is then degassed with nitrogen (which also facilitates mixing) and a small quantity of a fluoride-base flux is added to remove metal from the dross. The bronze is then cast into appropriate molds. Melting of large charges in a reverberatory furnace may require use of a cover flux to reduce oxidation losses. Melt temperature and melting time are kept to a minimum to control hydrogen pickup in the furnace. At 1,200°C, the hydrogen solubility in an aluminum bronze containing 8% Al is about 3.5 cnrVlOO g and this increases to about 5.8 cnrVlOO g at 1,400°C. (The solubility of hydrogen in pure copper at comparable temperatures is more than twice as high.) C-30 Continue ------- Back C.3 MARKETS The leading consumers of refined copper are wire mills, accounting for 75% of the refined copper consumption. Brass mills producing copper and copper alloy semi-fabricated shapes are the other dominant consumers at 23%. The dominant end-users of copper and copper alloys are the construction and electronic products industries, accounting for 65% of copper end-usage. Transportation equipment, such as vehicle radiators, accounts for an end-usage of 11.6%. A passenger car typically contains 50 Ib of copper wire (BHP 1997). Copper and copper alloy powders are used for brake linings and bands, bushings, instruments, and filters in the automotive and aerospace industries, for electrical and electronic applications, for anti-fouling paints and coatings, and for various chemical and medical purposes. Copper chemicals, principally CuSO4, CuO, and Cu2O, are widely used in algaecides, fungicides, wood preservatives, copper plating, pigments, electronic applications, and numerous special applications. End-use markets for brass rod include: • construction and remodeling 48% • industrial equipment and machinery 30% • electrical and electronics 8% • transportation equipment 8% • exports 4% • consumer durables 2% Typical products include plumbing fixtures, industrial valves and fittings, welding and cutting equipment, cable and electronic connectors, gas grill components, brake hose assemblies, and decorative hardware. C.3.1 Scrap Prices Scrap prices are related to the refined copper price, but the price spread must be sufficient to allow for collection, sorting, shipping, chopping, etc. If the price spread is too narrow, the processor cannot charge enough for the end product, which also is determined by the refined copper price, to make a profit. When refined copper prices are high, more copper scrap is offered to processors. If refined copper prices are low, less scrap enters the market. As the gap between C-31 ------- scrap price and refined price narrows, the processing cost may make the scrap uneconomical (Carlinetal. 1995). C.3.2 Scrap Consumption Copper-base scrap consumption in 1997 by type of scrap and by processor is summarized in Table C-l 1 (Edelstein 1998). The total consumption of 1,755,0001 is greater than the total of 1,370,000 t shown in Table C-4 because the latter table is based on the copper content of the scrap while the former is based on the gross weight of the copper-base alloys. Both of these tables are based on copper-base scrap, while Table C-3 includes other alloys where copper is not the primary alloying element. Table C-l 1 emphasizes the diversity of copper scrap uses. Unalloyed scrap is consumed by smelters, refiners, ingot makers, brass mills, wire-rod mills, and foundries. While about 63% of alloy scrap is consumed by brass mills, a significant fraction is also processed by ingot makers, smelters, refiners, and foundries. It is worth noting that environmental restrictions on lead associated with copper pose obstacles to recycling certain copper alloys, particularly some brasses. The addition of up to 8% lead in brass castings and rod improves machinability and casting characteristics. New drinking water standards may require elimination of most of the lead from brass plumbing fixtures (Carlin et al. 1995). As can be seen in Table C-l 1, leaded brass is a major component of copper-base scrap recycling. C.4 PARTITIONING OF CONTAMINANTS This section discusses the manner in which impurities partition during the various metallurgical operations involved in the refining of copper scrap. The main application of copper is as an electrical conductor. As such, extremely high purity levels are required to maintain low electrical resistance. As little as 0.08% iron or 0.05% phosphorus will reduce the conductivity of copper by 33% (CDA 1998b). Typical output from the cathode furnace may be electrolytic tough-pitch copper which contains a minimum 99.90% copper or oxygen-free copper, which contains a minimum of 99.95% copper. Thus, the aim of copper refining is to remove most of the impurities from the metal. The following sections discuss the expected distribution of contaminants in scrap that is introduced into the copper processing cycle (see Figure C-2). The expected partitioning from scrap which is introduced into brass mills, foundries, and the like will be discussed in a later section. C-32 ------- C.4.1 Partitioning During Copper Refining C.4.1.1 Thermochemical Considerations Most impurities in copper scrap introduced into blast furnaces, converters, or anode (fire refining) furnaces will tend to be oxidized during processing and removed with the slag. Theoretically, this will include all oxides whose free energies of formation per gram-atom of oxygen are more negative than that of CuO. The free energy of formation of CuO at 1,500 K (1,227°C) is about -6 Kcal/gram-atom of oxygen (Glassner 1957). Oxides of metals such as Po, Te, and the platinum group (Pt, Pd, Rh, Ir) are less stable than CuO and the respective metals should remain with the copper. Cs2O boils below 1,000 K and would be volatilized. Other species with low boiling points such as Cd, Po, Ra, Se, and Zn may also be partially volatilized (see Table E-3). Relevant free energy data for various oxides are summarized in Table C-12. Of the elements whose oxides are listed in this table, only Ag and Ru are expected to remain in the copper under equilibrium conditions. Copeland et al. (1978) calculated the partition ratios between copper and an oxide slag for several contaminants, based on free-energy data. The authors assumed that: (1) the weight of the slag was 10% of the weight of the metal, (2) the activity of the copper oxide in the slag was 0.1, and (3) the activity of the contaminant oxide in the slag was 0.01. Henry's Law constants for the contaminant and the contaminant oxide were assumed to be unity (i.e., ideal solution behavior). The partition ratio was defined as the weight of the contaminant in the slag divided by the weight of the contaminant in the ingot. Calculated partition ratios at 1,400 K are summarized in Table C-13. These calculations suggest that all the elements listed except cobalt will partition to the slag and that concentrations of most of these contaminants in the copper will be very low. However, blister copper leaving the converter is reported to contain small amounts of impurities such as As, Bi, Fe, Ni, Pb, Sb, Se, Te, and precious metals (Davenport 1986). This emphasizes that predictions based on thermochemical calculations and vapor pressures are only guidelines to impurity behavior during processing. C.4.1.2 Experimental Partitioning Studies Some experimental work has been done to measure partitioning of radionuclides during copper smelting. Heshmatpour et al. (1983) found that plutonium strongly partitioned to the slag, as would be expected from thermodynamic considerations. Three tests were conducted, in which C-33 ------- 500 ppm of PuO2 was melted with 200 grams of copper in recrystallized alumina crucibles at 1,400°C. The slag weight was 10% of the metal weight. Slags included a borosilicate composition (80% SiO2 13% B2O3, 4% Na2O, 2% A12O3, 1% K2O), a blast furnace composition (40% CaO, 30% SiO2, 10% A12O3, 15% Fe2O3, 5% CaF2) and a high silica composition (60% SiO2, 30% CaO, 10% A12O3). The respective partition ratios (defined as the ratio of total Pu in the slag to total Pu in melt) were 3,225, 157, and 107. In each case less than 1 ppm of Pu remained in the copper. In the last two cases, a significant fraction of the input PuO2 was not accounted for, rendering these values suspect. Copeland and Heestand (1980) measured the partition ratio of uranium in copper in a laboratory experiment by equilibrating copper at 1,100°C with a slag containing 0.3 wt% U. The measured partition ratio was 600, which is many orders of magnitude lower than the predicted value (see Table C-13). The final uranium concentration in the copper was 5 ppm. Other experimental details were not provided. A laboratory drip-melting experiment was also described, in which surface contaminated copper was placed on a screen and melted. The molten copper passed through the screen into a crucible below. Assay of the dross and the ingot showed that the former contained 3,400 ppm U, while the latter contained 1.4 ppm U. In a scaled-up experiment, about 40 kg of copper scrap surface contaminated with UO2 was drip melted. The copper ingots contained 0.07 ppm U, while the slag contained 1,250 ppm U, resulting in a partition ratio of 18,000. In subsequent work, Heshmatpour and Copeland (1981) conducted a series of laboratory experiments, in which 500 ppm UO2 was added to small melts of copper produced with various fluxes. The samples were melted in recrystallized alumina or zirconia crucibles and held at about 1,250°C to equilibrate the melt and the slag. The results, which are summarized in Table C-14, show that the partition ratios vary from 49 to 3182. Mautz (1975) and Davis et al. (1957) summarized the results of melting 40 heats (about 100 tons) of uranium-contaminated copper scrap with surface activities up to 150,000 dpm/100 cm2 in an oil-fired reverberatory furnace with a 125-ft stack. Ten samples taken from the copper product showed uranium values ranging from <0.022 ppm to 3.1 ppm. Six slag samples contained 1,440 to 1,730 ppm of U, while two samples contained only 0.43 and 0.47 ppm. No explanation for these low values was provided, although it is possible that the copper melts from which these slag samples were taken were initially very low in U. Uranium contamination of the furnace lining was also detected. Activity in the stack averaged 4 x io~n |iCi/cc. No air activity C-34 ------- was detected outside the furnace in excess of 1.7 x 10"12 |iCi/cm3, which is 10% of the MFC value listed in NBS Handbook 52 for a controlled area. Samples collected to detect fallout showed no measurable uranium contamination of areas inside or outside the furnace building. Table C-12. Standard Free Energies of Formation for Various Oxides at 1,500 K Metal Oxide Ag20 RuO4 CuO Cs2O Cu2O PbO TcO2 Sb2O3 CoO MO FeO ZnO MnO SiO2 PaO2 AmO2 Np02 RaO CeO2 UO2 Pu203 SrO ThO2 -AF° (Kcal/g-atom O) decomposes at 460 K 1.9 5.8 9.4 14.2 19.1 19.9 26.0 26.5 26.5 38.6 39.2 65.7 73.4 89.8 89.8 91.6 94.6 94.6 99.0 99.9 102 113 Source: Copeland et al. 1978 Abe et al. (1985) also conducted laboratory experiments to examine melt refining as a copper decontamination scheme. In these studies, 100 grams of metal and 10 grams of flux were melted in an alumina crucible under argon. Using a 1,550°C melting temperature, a melting time of one hour and a flux consisting of 40% SiO2, 40% CaO, and 20% A12O3, decontamination factors C-35 ------- ranged from 100 for an initial uranium concentration of 10 ppm to 104 for 1,000 ppm. The final uranium concentration in the ingot appeared to be relatively insensitive to the amount of uranium introduced into the melt. This suggests that the uranium content in the melt would not be less than about 0.1 ppm under the conditions of these experiments. However, the minimum observed uranium concentration in the melt-refined ingot—0.083 ppm—is very close to the 0.075 ppm of uranium in the copper feed stock used in this experiment. Table C-13. Calculated Partition Ratios of Various Contaminants Between Copper and an Oxide Slag at 1,400 K Contaminant Th Hf U Np Ti Pu W Tc Co Partition Ratio 1031 1026 1024 1024 1021 1020 108 103 10° Source: Copeland et al. 1978 In another study, Ren et al. (1994) conducted a series of laboratory experiments to optimize the removal of uranium contamination from copper. Samples weighing 100 grams were doped with 238 ppm uranium and melted with various fluxes. The investigation showed that residual uranium in the copper was at a minimum when the basicity of the flux was about 1.1. The highest decontamination factors were obtained when the flux was made from a blast furnace slag with the nominal composition: 38.1% SiO2, 41.4 %CaO, 3.8 %MgO, 2.6% Fe2O3, and 14.1% A12O3. To minimize the residual uranium in the copper, the mass of flux needed to be at least 5% of the metal charge. The researchers also found that over a range of uranium concentrations of 2.4 to 238 ppm, the residual uranium content in the copper ingot was unchanged. This is the opposite of the finding of Abe et al. (1985) discussed in the previous paragraph. The maximum decontamination factor achieved in the laboratory tests was 236. C-36 ------- Table C-14. Partitioning of Uranium in Laboratory Melts of Copper J£ "Hn GO 1 2 3 4 5 6 7 8 9 10 11 12 Metal (g) 100 100 100 100 100 100 100 100 100 250 250 170 Flux (g) 10 10 10 10 10 10 10 10 10 25 25 — U concentration (ppm) Slag 934 341 411 213 265 390 1813 1273 943 1590 1650 — Metal 0.13 0.37 0.11 0.14 0.54 0.45 0.83 0.04 0.25 1.36 0.14 1.96 Partition Ratio3 718 92 374 152 49 87 218 3182 377 117 1179 — Flux Composition A1203 25 20 15 10 10 10 10 10 CaF — — — — — — — — — CaO 25 20 15 30 20 30 10 10 30 CuO — — — 5 5 5 — — — Fe2O3 — — — — — — 5 5 5 SiO2 50 60 70 65 65 55 75 65 55 borosilicate glass 10 5 50 — 5 30 no flux Source: Heshmatpour and Copeland 1981 Mass of uranium in slag divided by mass in metal Vorotnikov et al. (1969) studied the behavior of iridium and ruthenium during the electrorefining of copper. They used copper anodes with 0.4% Ni, to which Ru-106 and Ir-192 were added. The distribution of these radionuclides during electrorefining in laboratory cells at current densities of 175 to 350 A/m2 is summarized in Table C-15. Table C-15. Distribution of Iridium and Ruthenium During Electrorefining of Copper Current Density (A/m2) 175 240 350 Ir(%) Electrolyte 14 15 15.5 Slimes 84 83 81 Cathode none none none Ru (%) Electrolyte 65 67 70 Slimes 29.8 27.4 20.1 Cathode 3.8 3.2 3.0 Source: Vorotnikov et al. 1969 As can be seen, most of the iridium reports to the slimes, while most of the ruthenium reports to the electrolyte. The electrolyte was then decoppered at a current density of 400 A/m2; the C-37 ------- resultant solution was boiled to produce nickel sulfate. Distribution of the iridium and ruthenium after electrolyte purification is shown in Table C-16. Table C-16. Distribution of Iridium and Ruthenium after Electrolyte Purification Product Regenerated Copper Copper Sponge Nickel Sulfate Electrolyte Ir(%) None Undetermined Undetermined 90 Ru (%) 5.0 21.0 12 70 Source: Vorotnikov et al. 1969 Even after purification of the electrolyte, most of the iridium and ruthenium remain in that process stream. C.4.1.3 Proposed Partitioning of Contaminants Blast Furnace Smelting Based on the information presented in Table C-5, expected partition ratios of contaminants during the processing of low-grade copper scrap in a blast furnace were developed using the studies of Opie et al. (1985) and Nelmes (1984). The study of Kusik and Kenahan (1978), also included in Table C-5, was not used to estimate partition ratios since those authors did not include information on slag compositions. The slag resulting from the blast furnace operation characterized by Opie et al. (1985) in Table C-5 is rich in recoverable metals. These authors describe a processing step in which the blast furnace slag is further treated in a EAF, to which 2% coke is added as a reductant (see Section C.2.3.1, Table C-6). The slag from this step is assumed to be granulated and sold. Slags generated from downstream operations are returned to the blast furnace for recovery of additional metal values. By assuming that the metal streams and the dust streams are combined, overall observed partitioning from the blast furnace/EAF processing can be calculated from the Opie study. This additional step was not used in analyzing the Nelmes data. The results of the partitioning studies are summarized in Table C-17. In developing this table, it was assumed that each 100 tons charged to a blast furnace produces 40 tons of black copper, 40 tons of slag, and 5 tons of baghouse dust (Nelmes 1984). To develop the ranges shown in Table C-17, the maximum and minimum values were selected from among the data from the various studies. C-38 ------- U.S. Patent No. 4,351,705 (related to the work of Opie et al. [1985]) provides information on the partitioning of silver. In one example from the patent, 1,455 tons of converter slag containing 17.2 oz/ton Ag were smelted in a blast furnace to produce 420 tons of black copper containing 43.2 oz/ton Ag and an unspecified quantity of blast furnace slag containing 0.81 oz/ton Ag. When the blast furnace slag was cleaned in an arc furnace, the silver content was reduced to 0.5 oz/ton. Based on additional information included in the patent, it can be estimated that approximately 1,170 tons of blast furnace slag were produced. The silver input to the smelting process from the converter slag was 25,000 oz; the silver output was 18,100 oz to the black copper and 950 oz to the blast furnace slag, leaving about 6,000 oz unaccounted for. In order to achieve a material balance, it is assumed here that the unaccounted material is contained in the baghouse dust. Using methodology similar to that for other metals during the slag cleaning process, one can estimate that the 950 oz of silver in the blast furnace slag are distributed as follows: • black copper from EAF 410 oz • slag from EAF 540 oz • baghouse dust from EAF: set to zero (the quantity will be small relative to that collected in the converter baghouse). These calculations provide the basis for the silver partition fractions in Table C-17. Table C-17 Observed Partition Fractions in the Melting of Low-grade Copper Scrap in a Blast Furnace Element Cu Ni Sb Sn Fe Zn Pb Cl F Ag Metal Min. 0.99 0.73 0.80 0.89 0.14 0.24 0.47 0 0 0.74 Max. 0.99 0.97 0.84 0.91 0.24 0.40 0.62 0 0 0.74 Dust Min. 0.0023 0.0020 0.056 0.028 0.00 0.51 0.29 1.0 1.0 0.022 Max. 0.0039 0.0053 0.060 0.066 0.00029 0.52 0.31 1.0 1.0 0.022 Slag Min. 0.0027 0.023 0.10 0.019 0.84 0.080 0.093 0 0 0.24 Max. 0.011 0.27 0.14 0.068 0.86 0.24 0.13 0 0 0.24 C-39 ------- The observed partitioning during the smelting of copper scrap in a blast furnace, as summarized in Table C-17, is combined with chemical analogies for certain elements and thermodynamic predictions from Table C-12 to arrive at the proposed partitioning for the desired suite of elements. This summary is presented in Table C-18. Most of the actinides form very stable oxides and are expected to be removed from the copper and concentrated in the slag. Even if removal is not 100%, as proposed in Table C-18, when the black copper is blown in a converter, the strongly oxidizing conditions can be expected to remove residual quantities of these elements to the converter slag, which is recycled to the blast furnace. Table C-18 Partition Fractions of Impurities in the Melting of Low-grade Copper Scrap in a Blast Furnace Element Ag Am Ce Co Cu Cs Fe Mn Ni Np Pa Pb Pu Ra Ru Sb Si Sr Tc Th U Zn Metal 0.74 0.73/0.97 0.99/0.99 0.14/0.24 0.14/0.24 0.73/0.97 0.47/0.62 0.99/0.99 0.80/0.84 0.73/0.97 0.24/0.40 Slag 0.02 1.0 1.0 0.023/0.27 0.0027/0.011 0.10/0.20 0.84/0.86 0.84/0.86 0.023/0.27 1.0 1.0 0.093/0.13 1.0 1.0 0.0027/0.011 0.10/0.14 some 1.0 0.023/0.27 1.0 1.0 0.080/0.24 Baghouse Dust 0.24 0.0020/0.0053 0.0023/0.0039 0.80/0.90 0.00/0.00029 0.00/0.00029 0.0020/0.0053 0.29/0.31 0.0023/0.0039 0.056/0.060 some 0.0020/0.0053 0.51/0.52 Basis for Estimate Table C-17 Table C-12 Table C-12 Same as Ni, Table C- 13 Table C-17 Table C-12, WCT Table C-17 Same as Fe Table C-17 Table C-12, Table C-13 Table C-12 Table C-17 Table C-12, Table C-13 Table C-12 Same as Cu Table C-17 Table C-5 Table C-12 Same as Ni, Table C-13 Table C-12, Table C-13 Table C-12, Table C-13 Table C-17 WCT = Author judgement C-40 ------- Converting Some information on the composition of the process streams emanating from a copper converter is presented in Table C-8. However, no mass balance information was available to develop estimates of partition ratios. If copper scrap is introduced directly into the converter, it is expected that partitioning will be similar to that in the blast furnace. The strongly oxidizing conditions should insure that any actinides and other strong oxide formers will be oxidized and removed with the slag. If the scrap were introduced at the blast furnace stage, removal of additional Fe, Ni, Sb, Sn, Pb and Zn would be expected, based on the information included in Tables C-5 and C-8, resulting in blister copper with fewer impurities. Fire Refining and Electrolysis Expected partitioning of impurities in fire-refined copper and in electrorefined copper is summarized in Tables C-19 and C-21, respectively. Both fire-refined copper and electrorefined copper are included since both are used to produce end products. For example, fire-refined copper is used to produce sheet and tubing while electrorefined copper is used to produce wire. The elemental partitioning proposed in Table C-19 is appropriate for evaluating scenarios involving production for non-electrical applications where, say, No. 1 scrap is used to make a copper product such as tubing for plumbing applications or sheet for roofing. If the scrap is introduced earlier in the process then, with the exception of silver and ruthenium, which are not easily oxidized, the quantities of radioactive contaminants remaining with the metal should have been reduced during prior processing steps. The values for Ag, Fe, Ni, Pb, Sb, and Zn were developed using the data in Table C-8 for the feed composition and the data of Garbay and Chapuis (1991) is cited in Table C-9 for the chemistry of the fire-refined anodes. While the use of two unrelated data sets is a recognized problem, better data were not uncovered during the current study. This concern is ameliorated, in part, by providing a range for many of the partition factors. As was discussed in Section C.2.3.1, a reverberatory furnace used for fire refining may not be equipped with a baghouse for dust collection. Offgas exiting the furnace after-burner may be exhausted directly through a stack. There are no NESHAPS standards for secondary copper smelters. Brunson and Stone (1975) provide information of the composition of the anode and cathode copper, as well as anode slimes at the Southwire Co. The compositions are listed in Table C-20. C-41 ------- Table C-19. Partition Fractions of Impurities in the Fire Refining of Copper Element Ag Am Co Cs Fe Mn Ni Np Pa Pb Pu Ru Sb Si Sr Tc Th U Zn Metal 0.30/0.59 0.001/0.01 0.05/0.10 0.02/0.05 0.02/0.05 0.05/0.10 0.001/0.01 0.001/0.02 0.22 0.001/0.01 1 0.08/0.25 0.001 0.001/0.02 0.001/0.02 0.10/0.20 Slag 0.41/0.70 0.99/0.999 0.90/0.95 0.10/0.20 0.95/0.98 0.95/0.98 0.0.90/0.95 0.99/0.999 0.98/0.999 0.73/0.78 0.99/0.999 0.75/0.92 1 1 0.999 0.98/0.999 0.98/0.999 0.80/0.90 Offgas 0.80/0.90 0.00/0.05 0.00/0.05 0.00/0.05 Basis for Estimate Table C-8, Table C-12, Garbay and Chapuis 1991 Same as Pu Table C-12, same as Ni Table C-12, WCT Table C-8, Table C-12, Garbay and Chapuis 1991 Table C-12, Same as Fe Table C-8, Table C-12, Garbay and Chapuis 1991 Same as Pu Same as U Table C-8, Table C-12, Garbay and Chapuis 1991, WCT Tables C-12 and C-1 3, Heshmatpour et al. 1 983 Table C-12 Table C-8, Table C-12, Garbay and Chapuis 1991, WCT Table C-1 2 Table C-1 2 Table C-1 2 and C-1 3 Same as U Tables C-12 and C-1 3, Heshmatpour and Copeland 1981 (Table C-1 4) Table C-8, Table C-12, WCT, Garbay and Chapuis 1991 WCT = author judgement Table C-21 presents partition fractions of selected impurities in the electrorefining process, based on the data reported by Brunson and Stone (1975). Cobalt and manganese were assumed to behave like nickel and iron, respectively. Strontium was assumed to behave similarly to calcium. When a contaminant was identified in both the anode slimes and in the cell bleed (i.e., Fe, Sb, and Zn), the unaccounted for material was assumed to accumulate in the nickel sulfate, which is recrystallized from the cell bleed after copper is removed in the liberator cells. Detailed calculations are summarized in Appendix C-1. Ruthenium partitioning is based on data of Vorotnikov et al. (1969). Metal partitioning can also be estimated for a limited suite of elements using the data of Ramachandran and Wildman (1987) presented in Section C.2.3.4. Comparing these data with the values in Table C-21 indicates that the latter values are conservative (i.e., show slightly higher partitioning to the metal) for use in predicting radiation exposures to residual radioactive contaminants in metal. C-42 ------- Table C-20 Composition of Anode and Cathode Copper and Anode Slimes at the Southwire Co. Element Cu O s Pb Ni As Sb Bi Au Ag Se Te Sn Fe Zn Ca Si Typical Anode (%) 99.50 0.10 0.003 0.19 0.10 0.005 0.010 0.0007 0.0012 0.024 0.031 0.0003 0.025 0.025 0.013 — — Typical Cathode 99.99% — — 5 ppm 7ppm 1 ppm 1 ppm 0.1 ppm — 10 ppm 0.5 ppm 1 ppm 1 ppm 6 ppm — — — Anode Slimes (%) 8.77 — — 31.45 0.75 — — 0.55 4.65 — — 9.28 1.20 — 1.10 3.50 Source: Brunson and Stone (1975) Note: Slimes also contain 0.001% Pt and 0.001% Pd. The literature on the electrorefining of copper abounds with consideration of the removal of impurities typically associated with copper, including Ag, As, Bi, Ni, Pb, Sb, Se, and Te. Virtually no information was uncovered in the course of this study on actinides and fission products, which are among the possible contaminants of copper cleared from nuclear facilities. To provide a quantitative perspective on the expected behavior of these contaminants during electrorefining, recourse was taken to some general electrorefining principles. According to Demaeral (1987): During the electrorefining of copper, anode impurities either dissolve in the electrolyte or remain as insoluble compounds in the anode slime. Elements less noble than copper such as zinc, nickel and iron easily dissolve in the electrolyte. Elements more electropositive than copper, e.g. selenium, tellurium, silver, gold, and the platinum group metals and elements which are insoluble in sulphuric acid, such as lead, are concentrated in the anode slime. A C-43 ------- third group of elements, comprising the impurities which have a dissolution potential comparable to copper, such as arsenic, antimony, and bismuth, behave in a different way. Depending on anode composition and other operational parameters they either report to the slime or to the electrolyte with a widely fluctuating distribution pattern. Further, these elements can, depending on the respective concentration in the electrolyte, undergo several side reactions in the bulk of the electrolyte, resulting in a wide range of insoluble compounds and floating slimes. Table C-21. Partition Fractions of Impurities in the Electrorefining of Copper Element Ag Am Ca Co Cs Fe Mn Ni Np Pb Pu Ru Sb Si Sn Sr Tc Th U Zn Metal 0.04 0.01 0.02 0.02 0.01 0.003 0.03/0.04 0.01 0.001 Anode Slimes 0.96 0.5 0.36 0.36 0.997 0.65/0.70 1.0 0.999 0.5 Electrolyte Bleed 1.0 0.5 0.99 1.0 0.62 0.62 0.99 1.0 1.0 0.20/0.30 0.99 0.5 1.0 1.0 1.0 1.0 Electrode potentials for half-cells of various elements less noble than copper are listed in Table C-22. From this tabulation, it can be deduced that all the listed elements should report to the electrolyte and that a fraction should be continuously removed from the electrorefining circuit with the electrolyte bleed. In the absence of modifying information, all the elements less noble than copper are assumed to report 100% to the electrolyte. During treatment of the electrolyte C-44 ------- bleed, it is not known whether many of these elements would concentrate in the black acid or in the crystallized nickel sulphate. Based on its electrode potential, strontium is expected to concentrate in the electrolyte. However, as noted by Brunson and Stone (1975), some calcium (and, by chemical analogy, strontium) is found in the slimes. Since the calcium content of the anodes is not reported by these authors, a partition ratio cannot be calculated. For Table C-21 it was arbitrarily assumed that calcium (and strontium) is distributed equally between the electrolyte and the slimes. Most of the nickel and probably the zinc, iron, cobalt, and manganese would be recovered from the electrolyte bleed as mixed sulfate crystals13. Table C-22. Half-cell Electrode Potentials of Elements less Noble than Copper Reaction Cs Sr Am Pu Th Np U Zn Tc Fe Co Ni Cu = Cs+ + e = Sr2+ +2e- = Am3+ + 3e = Pu3+ +3e- = Th4+ +4e- = Np3+ +3e' = U3+ +3e- = Zn2+ +2e' = Tcx+ +xe' = Fe2+ +2e- = Co2+ +2e- = M2+ +2e- = Cu2+ +2e- Potential (V) -2.92 -2.89 -2.32 -2.07 -1.90 -1.86 -1.80 -0.763 -0.71 -0.44 -0.277 -0.25 0.337 Sources: Lewis and Randall 1961, Snyder et al. 1987. (All values quoted by Snyder et al. (1987), except the one for Tc, were taken from Latimer 1953.) Note: Potentials at 25°C For copper wire and other electrical conductors produced from fire-refined copper, estimating the partition fractions of contaminants in the metal involves combining the factors in Tables C-19 and C-21. Thus, if there were 1 kg of lead in a unit of copper scrap, there would be 220 g of lead in the fire-refined copper and 0.7 g in the electrolytic copper. 13 Dobner (1997) has indicated that the composition of crude nickel sulfate (NiSO4.2H2O) is 27% Ni, 0.7% Zn, 0.3% Fe, 0.18% As, and 0.12% Sb. C-45 ------- C.4.2 Partitioning During Brass Smelting Partitioning of contaminants during brass smelting is expected to be different from that in fire refining of copper. In fire-refining operations, the objective is to remove, by oxidation and slagging, as many impurities as possible. In brass melting, on the other hand, one objective is to minimize losses of alloying elements such as Zn, Fe, Mn, Pb, Al, and Sn. Consequently, from a conservative perspective in assessing radiation exposures to radioactive contaminants in metal, it should be assumed that all the contaminants remain in the metal. C.5 EXPOSURE SCENARIOS C.5.1 Modeling Parameters As discussed in the previous sections, there are numerous options for the introduction of copper scrap into the copper refining process. Worker exposures to the contaminated scrap prior to smelting would be relatively independent of where the scrap is introduced into the secondary recovery process but would vary with the type of scrap. Typical operations may involve sorting, shredding, briquetting, and transportation. Insulation removal is required for the recycling of most copper wire. It is likely that slag generated at any step in the process will be returned to a blast furnace for further processing and only blast furnace (or cleaned blast furnace) slag will exit the process. This slag will be sold or disposed of. The blast furnace operation may be at a different location than the initial secondary smelting operation. In that case, haulage of contaminated slag may be required. Since slag volumes will be smallest when introducing No. 1 copper scrap directly into a fire-refining furnace, the concentrations of any radionuclides that partition to the slag will be greatest for that type of operation. This slag will be diluted when reprocessed in a blast furnace. Scrap copper released from nuclear installations is likely to be carefully sorted high-quality material. As such, it would most likely be introduced into the secondary refining process at the fire refining stage where it would be used to produce anodes for electrorefming or finished mill products such as sheet and tubing. Expected partitioning of contaminants during fire refining is summarized in Table C-19. While additional partitioning occurs during electrorefming, the result of that process is to further reduce the impurities in the metal. Therefore, it is unlikely that electrorefming of cleared scrap would lead to higher radiation exposures than received during the C-46 ------- fire-refining of such scrap. Possible exceptions could be exposures to anode slimes and electrolyte bleed streams from the electrolysis cells. C.5.1.1 Dilution of Cleared Scrap The information presented in Section C.I.I indicates that a maximum of 10,833 t of copper scrap would be cleared in any one year. This represents about 0.8% of the total annual consumption of copper scrap, as listed in Table C-4. Thus, if this scrap were uniformly distributed amongst all consumers, the dilution factor would be 0.008. If all this scrap was processed through a single 200-ton reverberatory furnace, which has an annual capacity of 45,500 tons (-41,300 t) the dilution factor would be 0.26. This calculation assumes that the furnace operates 330 days per year on a 24-hour cycle with 25% of the charge left in the furnace to facilitate the subsequent melting cycle. A more reasonable assumption is that the reference facility—the 200-ton reverberatory furnace cited above—would process the 2,080 t/a of copper scrap generated during the decommissioning of the K-25 Plant at Oak Ridge, while the scrap stockpiled during the years when no scrap was cleared by DOE would have a different disposition. In such a case, the dilution would be 0.05. C.5.1.2 Slag Production Slag production in a reverberatory furnace varies as a function of the percentage of copper in the charge. With increasing copper grade (Biswas and Davenport 1976): • Copper concentration in slag increases • Slag weight decreases • Copper loss decreases High-copper-content scrap metal, ranging from 85-95% copper, loaded in a 350-ton-per-day reverberatory furnace, may generate about 30 tons per day of slag. The slag contains an economically recoverable concentration of copper, which may be recycled to a blast furnace for recovery (Murrah 1997). Slag is used for the manufacture of abrasives, shingles, road surface bedding, mineral wool, and cement/concrete materials (Carey 1997). Slags from a Peirce-Smith converter have an economically viable copper content and may be recycled to a reverberatory or blast furnace to reduce copper loss (Biswas and Davenport 1976). C-47 ------- The process options are myriad; each processor has its own preferred operational cycle. These range from simple remelting and casting, to smelting and recycling the slag, depending upon the available options (Murrah 1997). One producer, who uses a reverberatory furnace to melt high grade copper scrap and cast logs from which extrusion billets are cut, estimates that the slag weight is about 2 to 2.5% of the charge weight (Burg 1999). Based on the available information, it is proposed for modeling purposes that a reverberatory furnace melting and fire refining No. 1 copper scrap generates 0.02 tons of slag per ton of scrap charged. Since many oxidizable impurities concentrate in the slag, a small slag volume will increase concentrations of these elements in the slag. C.5.1.3 Baghouse Dusts In the copper conversion process, baghouse filtration is used at various processing stages to collect zinc, tin and lead dusts. The composition of the dust is a function of the copper charge composition. Thus, dust capture will vary strongly with alloy composition. Assuming a typical converter charge, about 0.25% of the copper in the feed will enter the baghouse collection system as oxide. Dust, depending on the alloy composition of the charge, is sent to lead, zinc, or tin smelters to recover these metals (Edelstein 1997). In a reverberatory furnace, the dust produced may be as much as 1% of the charge. The dust is frequently recycled to the furnace if the copper content is significant. Dust from a Peirce-Smith converter may contain as much as 11% copper; it is almost always recycled to a smelting furnace (Biswas and Davenport 1976). The mass of dust generated by an EAF used for copper smelting is about 0.25% of the mass of scrap metal charged to the furnace. However, as noted previously, some operations do not use a baghouse for dust control, so that the species that accumulate in the offgas, as noted in Table C-19, would be released to the atmosphere. C-48 ------- C.5.1.4 Electrolyte Bleed During the final electrolytic purification of copper, part of the electrolyte is bled off to control impurity build-up in the electrolytic cells. The soluble impurities include As, Bi, Co, Fe, Ni, Sb, and Zn. As noted in Section C.4.1.3, As, Bi, and Sb may report either to the electrolyte or to the anode slimes depending on such factors as anode chemistry and cell operating parameters. Actinide elements are also assumed to report to the electrolyte. Some of these impurities are removed from the bleed stream by evaporation and crystallization and may be contained in products which are sold. Other impurities may remain in the electrolyte and be returned to the electrorefining process or used to leach slimes. The implication is that this added step in the processing of copper creates the potential for a new source of exposure by reconcentrating residual metals. However, most of the residual radioactive contaminants in the cleared copper scrap will have partitioned to the slag or been removed in the offgas well before this stage. The principal exceptions are isotopes of Co, Fe, Ni, Ru, and Zn. If a large electrolytic refinery uses 460,000 tpy of copper anodes containing 0.1% Ni, the nickel content in the feed is 460 tons. According to Table C-21, 99% of Ni is concentrated in the electrolyte bleed stream. If this nickel is crystallized as NiSO4, which is 38% Ni by weight, and if the crude nickel sulfate contains 5% H2SO4 and 3% water, then the annual production of the crude precipitate is about 1,300 tons (460 x 0.99 H- [0.92 x 0.38] ~ 1,300). The concentration of nickel in the crude nickel sulfate is 35% (0.38 x 0.92 = 0.35), or about 350 times that of the nickel in the anodes. By chemical analogy, cobalt should be similarly concentrated. While the behavior of other impurities in the electrolyte bleed is unknown, it is likely that some of these will be crystallized with the nickel sulfate. According to Garbay and Chapuis (1991), a 50,000-t French electrorefining plant produces about 500 t of residual sulfuric acid, about 30 t of arsenical sludge, and about 601 of nickel sulfate. The nickel sulfate production rate quoted by Garbay and Chapuis—1.2 kg/t of Cu—is lower than that described in the previous paragraph—equivalent to 2.9 kg/t of Cu—partly because the nickel content in the French anodes is only 0.05% (see Section C.2.3.3). C.5.1.5 Anode Slimes Brunson and Stone (1975) cite a slimes generation rate of 15 Ib of anode slimes produced per ton of copper refined at the Southwire Co. This rate of slimes production—7.5 kg/t of Cu—is more than an order of magnitude higher than the 600 g/t quoted by Garbay and Chapuis (1991). The C-49 ------- cause of this difference is not known. However, data quoted by Schloen (1987) corresponds to slimes generation rates ranging from 1 to 7.3 kg/t of anodes for nine U.S. electrolytic refineries, suggesting that the higher figure is more typical of U.S. experience. C.5.1.6 Summary Model for Fire-Refined Products Based on the information presented above, the following model is proposed for fire-refined products, such as copper tubing. A 200-ton reverberatory furnace is used to melt No. 1 copper scrap. The furnace operates 12 out of every 14 days, with two days down for routine maintenance. The furnace also is shut down for an additional two weeks per year for major maintenance. The furnace operates on a 24-hour cycle with the following cycle elements : • Charging 4.5 hr • Melting 4.5 hr • Refining and slagging 5.5 hr • Poling 2.5 hr • Casting 7 hr Since about 25% of the melt remains in the furnace as a heel for the subsequent heat, the daily output is 150 tons and the annual output is 45,000 tons. The annual furnace input is 45,500 tons of copper scrap. The furnace produces 910 tons of slag and 110 tons of dust (dust generation of about 5 Ib per ton) annually. The slag contains about 40% copper and the dust contains about 75% copper. The dust is either collected in the baghouse or released to the atmosphere. The slag and the dust (if captured) are sent to an outside processor for recovery of additional metal values. Elemental partitioning is presented in Table C-19. The approximate material balance is illustrated in Figure C-4. The slag from the reverberatory furnace is shipped to an outside processor who treats the material in a 50 tph blast furnace with an annual capacity of 36,000 tons (50 tph x 24 hr/day x 300 days/year = 36,000 tons). Thus, the slag from the reverberatory furnace undergoes a further dilution of 0.025 (910 H- 36,000 ~ 0.025). The blast furnace slag is then sold for industrial applications such as use in abrasives, roofing materials, or road building materials. C-50 ------- 45,500 tons scrap charcoal and slag formers 200-ton reverberatory furnace 110 tons dust (75% Cu) .. 910 tons slag (40% Cu) - 45,000 tons copper air green logs Figure C-4. Proposed Material Balance for Modeling Copper Produced by Fire Refining (values are rounded) C.5.1.7 Summary Model for Electrorefming Based on the previously presented information, the following model is proposed for high conductivity electrical products, such as wire and cable, which require electrorefining after fire refining for further impurity removal. Annual output from the electrolytic refinery is 450,000 tons of copper, 3,200 tons of anode slimes, and 1,300 tons of crude nickel sulfate (Schloen 1987). Sulfuric acid recovered from the electrolyte bleed circuit is assumed to be used for electrolyte makeup; accordingly, it is returned to the process. The nickel sulfate, containing 5% H2SO4 and 3% H2O, is sold to nickel producers for metal recovery. The nickel sulfate also contains contaminants, such as iron and zinc. The annual input to the reverberatory furnace at the electorefinery is assumed to be 24,000 tons of No. 2 copper scrap and 102,000 tons of blister copper from primary producers. The average nickel content of the anodes is 0.1%. An approximate material balance is presented in Figure C-5. Elemental partitioning can be calculated by combining the factors included in Tables C-19 and C-21. C.5.2 Worker Exposures Dust sampling at a primary copper smelter has been reported by Michaud et al. (1996). Samples were taken at a smelting furnace and a converter located in separate buildings. Results are C-51 ------- 138,000 tons anode scrap 102,000 tons blister copper Anode Scrap Melting (shaft fee.) Fire Refining (reverb. fee.) 24,000 tons No. 2 scrap acid Electro- Refining Electrolyte Clean Up bleed 1300 tons nickel sulfate 450,000 tons copper •*• 3200 tons anode slimes slag 190,000 tons purchased anodes Figure C-5. Simplified Material Balance for Electrorefining of Copper Produced from Scrap summarized in Table C-23. Cadmium and nickel were not detected in the dusts. Table C-23. Airborne Dust Concentrations At Primary Copper Smelter (mg/m3) Unit Smelting Furnace Converter Total 2.3 2.1 Respirable 0.6 0.8 Lead 0.21 0.15 Copper 0.10 0.32 Arsenic 0.02 0.02 Source: Michaud et al. 1996 C.5.2.1 Baghouse Dust Agglomeration Operator As noted in Table C-19, cesium is the main contaminant that would distribute to the offgas during fire refining of copper scrap. The exposure scenario developed here is designed to capture worker exposure to this dust and is based primarily on information presented in Section C.2.3.8. Basic assumptions include: • Copper output 342,000 tpy • Baghouse dust from fire-refining furnaces 51,100 tpy • Cesium partitioning to dust 90% C-52 ------- Based on these assumptions, the dust generation rate will be 0.15 tons of dust per ton of copper product (51,100 + 342,000). The cesium reconcentration factor due to preferential partitioning to the dust will be 6:1 (5,000 x 0.9 ^ 750). The operator would be exposed for 7 hours per day, 5 days per week to the mass of wetted dust in a concrete bunker that is about 20 x 30 x 12 ft high. It is assumed that the bunker contains a maximum of three days' output from the agglomerator or 420 tons (20 tph x 7 hr/d x 3 d = 420 tons). If the recycling facility used a reverberatory furnace without a baghouse, then all the cesium would be exhausted up the stack and become airborne. C.5.2.2 Furnace Operator A furnace operator would be part of a crew that spends full time in the vicinity of the reverberatory furnace that holds 200 tons of copper. For about two hours per shift, he would be standing 5 to 10 ft from an open furnace, skimming slag from the furnace with a rake into a metal box about 4 * 4 x 1 ft. Another operator would transport the slag box with a forklift truck about 200 ft to an area on the furnace room floor where the box is dumped. The cooled slag is broken up by an operator with a pneumatic hammer; copper is then culled by hand from the slag. At other times the operator will be shoveling charcoal and slag-forming agents into the furnace or tapping the furnace to allow the molten metal to flow through launders to the holding furnace. C.5.2.3 Scrap Handler The scrap handler would spend full time in the vicinity of the scrap piles preparing the material for charging into the furnace. This might include loading material into a briquetting machine and transporting the briquetted scrap to a staging area with a fork-lift truck. On average, about 200 tons of scrap are stockpiled in the scrap-handling area. C.5.2.4 Casting Machine Operator A casting machine operator would cast the copper into logs and assist in moving the cooled logs from the casting machine cooling pit to the billet-cutting machine. The operator would spent full time working near several copper logs that are about 26 feet long and up to 12 inches in diameter. C-53 ------- C.5.2.5 Scrap Metal Transporter If all the scrap from the largest annual DOE source (i.e. 2,080 t from the K-25 plant in Oak Ridge) were shipped to Southwire in Carrollton, Ga. for recycling, 104 shipments in a 20-t truck would be required. The distance is about 250 miles; the estimated driving time is six hours. Thus the total driver exposure would be about 624 hours. Other situations, which would lead to greater exposures, are possible. To accommodate this possibility, it is conservatively assumed that a truck driver spends full time driving a 20-t truck, with the truck loaded only one-half of the time (i.e., about 1,000 hr/y). C.5.2.6 Tank House Operator A tank house operator in a 450,000 tpy electrolytic refining plant would collect and drum 3,200 tons of anode slimes for transport to a refinery for metals recovery. C.5.3 Non-Industrial Exposures C. 5.3.1 Driver of Motor Vehi cl e The average amount of copper used in automobiles or light trucks is 50 pounds. The radiator contains about 80% of this; the electrical system contains about 20%. These elements are mostly under the hood presenting minimal exposure hazards. The radiator would consist of recycled scrap (CDA 1997). It is likely that the copper would come from several lots of material with differing processing histories. C.5.3.2 Homemaker Home appliances and heating and cooling systems contain copper produced from recycled scrap. Copper usage in home appliances is as follows (CDA 1997): • Central Air Conditioner 50 Ib • Refrigerator 5 Ib • Dishwasher 5 Ib • Washing Machine 4.4 Ib • Dryer 2 Ib • Range 1.3 Ib C-54 ------- • Garbage Disposer 2.3 Ib • Dehumidifier 2.7 Ib • Heat Pump 48 Ib Radiation exposures from any residual radioactive contaminants in these products would be very low relative to those associated with handling copper scrap and finished and semi-finished products made from this metal during the various stages in the copper refining process. This is primarily because of the small quantities of copper in these products, and because the copper would be obtained from many different lots of material, not all of which would be produced from cleared scrap. C-55 ------- REFERENCES Abe, M, T. Uda, and H. Iba. 1985. "A Melt Refining Method for Uranium Contaminated Steels and Copper." In Waste Management '85 3: 375-378. Tucson, AZ. Adams, V. 1998. "National Center of Excellence for Metals Recycle." U.S. Department of Energy. Biswas, A. D., and W. G. Davenport. 1976. Extractive Metallurgy of Copper. Pergamon Press. BHP 1997. "Bringing Copper to Market." On CU, vol. 1, No. 4. BHP Copper. Blanton, C. (Mueller Industries, Inc.). 1999. Private communication (26 May 1999). Brunson, W. W., and D. R. Stone. 1975. "Electrorefming at the Copper Division of Southwire Company, Carrollton, Georgia, U.S.A." Transactions of the Institution of Mining and Metallurgy, Section C, vol. 85, C150-C156. Browne, E. R. 1990. "A Little Copper Goes A Long Way," Scrap Processing and Recycling, 47 (1): 90. Bureau of Mines. 1993. "Recycled Metals in the United States." U.S. Department of Interior, Washington, DC. Burg, G. (Reading Tube Company). 1999. Private communication (May 1999). Capp, J. A. 1997. "Review of Application No. 10037 for New Fines Process for Blast Furnace Operation at Copper Division of Southwire." Memorandum to John R. Yntema, Georgia Department of Natural Resources, Environmental Protection Division, Air Protection Branch. Carey, J. (Chemetco Inc.). 1997. Private communication. Carlin, J. F., Jr., et al. 1995. "Recycling-Nonferrous Metals: Annual Report 1993." Bureau of Mines, U.S. Department of Interior, Washington, DC. Copeland G. L., R. L. Heestand, and R. S. Mateer. 1978. "Volume Reduction of Low-Level Contaminated Metal Waste by Melting: Selection of Method and Conceptual Plan," ORNL/TM-6388. Oak Ridge National Laboratory, Oak Ridge, TN. Copeland, G. L., and R. L. Heestand. 1980. "Volume Reduction of Contaminated Metal Waste." Waste Management '80, vol. 2, 425-433. Tucson, AZ. Copper Development Association (CDA). 1997. Private communication. C-56 ------- Copper Development Association (CDA). 1998a. "Newly Mined Copper: Why Do We Need It?" Copper Development Association (CDA). 1998b. "Recycled Copper." Davenport, W. G. 1986. "Copper Production." In Encyclopedia of Materials Science and Engineering. Vol 2, 841-848. Pergamon Press. Davis, D. M., J. C. Hart, and A. D. Warden. 1957. "Hazard Control in Processing Stainless Steel and Copper Contaminated with Uranium." Ind. Hyg. Quart. 18:235-241. Deacon, K. (U.S. Department of Energy, Oak Ridge Office). 1999. Private communication (February 1999). Demaeral, J. F. ,1987. "The Behavior of Arsenic in the Copper Electrorefining Process." In The Electrorefining and Winning of Copper. Eds. J. E. Hoffman et al. TMS 116th Annual Meeting, February 24-26, 1987, Denver, CO. The Metallurgical Society/AIME, Warrendale, PA. Dobner, R. F. 1997. "Bleed-Off Treatment at HK's Secondary Copper Electrorefinery." In Proceedings of the 1997 TMS Annual Meeting, 413-424. The Metallurgical Society/AIME, Warrendale, PA. Edelstein, D. L. (U.S. Geological Survey). 1997. Private communication (10 June 1997). Edelstein, D. L. 1998. "Copper - 1997 Annual Review: Mineral Industry Surveys." U.S. Geological Survey. Epel, L. (Brookhaven National Laboratory). 1997. Private communication. Garbay, H., and A. M. Chapuis. 1991. "Radiological Impact of Very Slightly Radioactive Copper and Aluminum Recovered from Dismantled Nuclear Facilities," Final Report, EUR- 13160-FR. Commission of the European Communities. George, D. B. 1993. "Oxy/Gas Rotary Furnaces Benefit Metals Industry." Plant Engineering 47 (3): A4. Gerson, L. (Southwire Company). 1999. Private communication (February 1999). Glassner, A. 1957. "The Thermochemical Properties of the Oxides, Fluorides, and Chlorides to 2500 K," ANL-5750. Argonne National Laboratory, Argonne, IL. Gockman, K. 1992. "Recycling of Copper." CIMBulletin 85 (958): 150. C-57 ------- Heshmatpour, B., G. L. Copeland, and R. L. Heestand. 1983. "Decontamination of Transuranic Contaminated Metals by Melt Refining." Nuclear and Chemical Waste Management., 4:129-134. Heshmatpour, B., and G. L. Copeland. 1981. "The Effects of Slag Composition and Process Variables on Decontamination of Metallic Wastes by Melt Refining," ORNL/TM-7501. Oak Ridge National Laboratory, Oak Ridge, TN. Kusik, C. L., and C. B. Kenahan. 1978. "Energy Use Patterns for Metal Recycling," Information Circular 8781. U.S. Bureau of Mines. Latimer, W. M. 1953. Oxidation Potentials. 2d ed. Prentice Hall, New York. Lewis, G. N., and M. Randall. 1961. Thermodynamics. 2d ed. McGraw-Hill Book Company. Mackey, T. 1993. "Outlook for Copper Scrap Recovery." American Metal Market 101 (70): 14. Mautz, E. W., et al., 1975. "Uranium Decontamination of Common Metals by Smelting - A Review," NLCO-1113. National Lead Company of Ohio. McDonald, R. P. 1999. "Southwire in Carrollton, GA," Memorandum to Lou Musgrove, Georgia Department of Natural Resources, Environmental Protection Division, Air Protection Branch (January 7, 1999). McKibben, G. (Southwire Company, Copper Division). 1999. Private communication (March 1999). Michaud, D., et al. 1996. "Characterization of Airborne Dust from Two Nonferrous Foundries by Physico-Chemical Methods and Multivariate Statistical Analyses." Journal of the Air and Waste Management Association 46:450-457. Murrah, C. (Southwire Copper Corp.). 1997. Private communication. National Association of Recycling Industries (NARI). 1980. "Standard Classification for Nonferrous Scrap Metals," NARI Circular NF-80. National Research Council. 1996. "Affordable Cleanup? Opportunities for Cost Reduction in the Decontamination and Decommissioning of the Nation's Uranium Enrichment Facilities." National Academy Press. Nelmes, W. S. 1984. "The Secondary Copper Blast Furnace." Trans., Inst. of Mining & Metallurgy 93 :C180. C-58 ------- Newell, R., et al. 1982. "A Review of Methods for Identifying Scrap Metals," Information Circular 8902. U.S. Bureau of Mines. O'Brien, N. M. 1992. "Processing Secondary Materials in a Top Blown Rotary Converter." In Conference: Copper in the 90's, p. 76. Bombay. Olin Brass Co. 1995. "Olin Brass - A Tradition of Leadership for Today and Tomorrow." Opie, W. R., H. P. Rajcevic, and W. D. James. 1985. "Secondary Copper Smelting." In Conference - Physical Chemistry of Extractive Metallurgy 379-385. The Metallurgical Society/AIME, New York. Parsons Engineering Services, Inc., RMI Environmental Services, and U.S. Steel Facilities Redeployment Group. 1995. "U.S. Department of Energy, Scrap Metal Inventory Report for the Office of Technology Development, Office of Environmental Management," DOE/HWP- 167. Prepared for Hazardous Waste Remedial Actions Program, Environmental Management and Enrichment Facilities, Oak Ridge, TN. Person, G. A., et al. 1995. "Gaseous Diffusion Facilities Decontamination and Decommissioning Estimate Report," rev. 2, ES/ER/TM-171. Environmental Restoration Division, Oak Ridge National Laboratory, Oak Ridge, TN. Ramachandran, V., and V. L. Wildman. 1987. "Current Operations at the Amarillo Copper Refinery." In The Electrorefining and Winning of Copper. Eds. J. E. Hoffman et al. TMS 116th Annual Meeting, February 24-26, 1987, Denver, CO. The Metallurgical Society/AIME, Warrendale, PA. Reading Tube Corp. 1999. (3 March 1999). Ren, Xian Wen, et al. 1994. "Melt Refining of Uranium Contaminated Copper, Nickel, and Mild Steel." In Spectrum '94, Nuclear and Hazardous Waste Management. International Topical Meeting, August 14-18, 1994, Atlanta GA. Riga, M. (Chemetco Inc. ) 1999. Private communication (March 1999). Riley, W. D., R. E. Brown, and D. M. Soboroff. 1984. "Rapid Identification and Sorting of Scrap Metals." Conservation and Recycling 6:181. Roscrow, W. J. 1991. "Furnaces for Non-Ferrous Metal Reclamation." Metallurgia 50 (4): 158. Schloen, J. H. 1987. "Electrolytic Copper Refining: Tank Room Data." In The Electrorefining and Winning of Copper. Eds. J. E. Hoffman et al. TMS 116th Annual Meeting, February 24-26, 1987, Denver, CO. The Metallurgical Society/AIME, Warrendale, PA. C-59 ------- Schwab, M. I, A. W. Spitz and R. A. Spitz. 1990. "Blister Copper Production from Secondary Materials." In Second International Symposium: Recycling of Metals and Engineered Materials, p. 139. Eds. J. H. L. van Linden, D. L. Stewart, Jr. and Y. Sahai. The Minerals, Metals and Materials Society, Warrendale PA. Shooter, D. (Olin Brass). 1999. Private communication (22 March 1999). Snyder, T. S., et al. 1987. "Experimental Results for the Nickel Purification: Phase I of the Oak Ridge Scrap Metal Decontamination Program," Contract No. DE-ACO5-860R-21670. Westinghouse R&D Center. U. K. Copper Development Association (U. K. CD A). 1999. "Melting Practice." . U.S. Department of Energy (U.S. DOE) 1993. "Oak Ridge K-25 Site Technology Logic Diagram." Vol. 1, "Technology Evaluation," Report K-2073. U.S. DOE, Office of Technology Development, Oak Ridge K-25 Site Office. U.S. Department of Energy (U.S. DOE) 1995. "Scrap Metal and Equipment: Materials in Inventory." Appendix to "Taking Stock: A Look at the Opportunities and Challenges Posed by Inventories from the Cold War," DOE/EM-0275. U.S. Department of Energy, Office of Environmental Management. U.S. Environmental Protection Agency (U.S. EPA) 1995. "Profile of the Non-Ferrous Metals Industry," EPA 310-R-95-011. U.S. EPA. U.S. Geological Survey (USGS) 1998. "Recycling Metals: 1997 Annual Mineral Industry Surveys." U.S. Geological Survey. Vorotnikov, N. V., et al. 1969. "Behavior of Iridium and Ruthenium in Electrorefming of Copper." The Soviet Journal of Non-Ferrous Metals 10 (2) 73-74. Warner, J. (Chase Brass and Copper Co.). 1999. Private communication (March 1999). Wechsler, T. E. F., and G. M. Gitman. 1991. "Combustion Enhancement of Copper Scrap Melting and Heating." In Conference EPD Congress 91, 421-436. The Mineral, Metals and Materials Society, Warrendale PA, Woodserman, J. (Chase Brass and Copper Co.). 1999. Private communication (April 1999). C-60 ------- APPENDIX C-l PARTITIONING DURING FIRE REFINING AND ELECTROREFINING OF COPPER SCRAP ------- Table Cl-1. Partitioning During Fire Refining and Electrolysis of Copper Scrap Reverb charge Reverb output Electrolytic Cell output Cu Ni Sb Sn Fe Zn Pb Ag Bi As Te Se Ca Si 45500 tons 910 tons in slag 110 tons in dust 910 tons at 40% Cu 110 tons at 75% Cu 45000 tons in anode Cu 44500 tons as cathodes 337.5 tons as slimes 15lb/ton 128.7 tons as nickel sulfate (38%Ni) Anodes (wt. %) 99.5 0.1 0.01 0.025 0.025 0.013 0.19 0.024 0.0007 0.005 0.0003 0.031 Total tons 44775 45 4.5 11.25 11.25 5.85 85.5 10.8 0.315 2.25 0.135 13.95 44965.8 Cathodes (ppm)** 99.99% 7 1 1 6 0 5 10 0.1 1 1 0.5 tons 44495.55 0.31 0.04 0.04 0.27 0.00 0.22 0.45 0.00 0.04 0.04 0.02 44497 Metal Partition 0.0069 0.0099 0.0040 0.0237 0.0000 0.0026 0.0412 0.0141 0.0198 0.3296 0.0016 **unless other units shown Slimes (wt %) 8.77 0 0 9.28 1.2 0 31.45 5.2 0 0.75 0 0 1.1 3.5 Slimes tons 29.60 0.00 0.00 31.32 4.05 0.00 106.14 17.55 0.00 2.53 0.00 0.00 3.71 11.81 194.91 Slimes Partition 0.000 0.000 2.784 0.360 0.000 1.241 1.625 0.000 1.125 0.000 0.000 0.500* 1.000* * assumed Bleed tons 44.69 Bleed Partition 0.99 Material Balance tons unaccounted 0.00 4.46 -20.11 6.93 5.85 -20.87 -7.20 0.31 -0.33 0.09 13.93 -3.71 -11.81 -32.46 140 tons of slimes not accounted for add bal. to bleed add bal. to anodes add bal. to bleed add bal. to bleed subt. bal. fr. slimes subt. bal. fr.slimes add bal. to bleed subt. bal. fr. slimes add bal. to slimes add bal. to slimes add bal. to anodes add bal. to anodes Adjusted Slimes Partition 0.000 0.000 0.999 0.360 0.000 0.997 0.959 0.000 0.980 0.670 0.998 0.500 1.000 Adjusted Bleed Partition 0.993 0.990 0.000 0.616 1.000 0.000 0.000 0.986 0.000 0.000 0.000 0.500 0.000 Adjusted Metal Partition 0.0069 0.0099 0.0014 0.0237 0.0000 0.0026 0.0412 0.0141 0.0198 0.3296 0.0016 0.0000 0.0000 Partition Check 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 o ------- APPENDIX D SELECTION OF RADIONUCLIDES FOR RADIOLOGICAL ASSESSMENT ------- Contents page D.I Sources Used to Make Recommendations D-l D.I.I IAEA-TECDOC-855 D-l D.1.2 NUREG/CR-0134 D-l D.1.3 WINCO-1191 D-2 D.1.4 NUREG/CR-0130 D-2 D.I.5 NUREG/CR-3585 D-4 D.1.6 NUREG/CR-4370 D-4 D.1.7 SAND92-0700 D-5 D.I.8 ORIGEN D-7 D 1.9 SAND91-2795 D-8 D.2 Radionuclides Recommended for Inclusion D-8 D.2.1 Basis for Recommendations D-8 References D-16 Tables D-l. Nuclides from WINCO-1191 D-3 D-2. Nuclides Included in NUREG/CR-0130 D-4 D-3. Nuclides Analyzed in NUREG/CR-4370 D-5 D-4. Nuclides Analyzed by SAND92-0700 for WIPP D-6 D-5. Nuclides from ORIGEN with Normalized Activity-Weighted Dose Factors D-9 D-6. Selection of Nuclides to Be Included in Scrap Recycle Analysis D-12 D-iii ------- SELECTION OF RADIONUCLIDES FOR RADIOLOGICAL ASSESSMENT D. 1 SOURCES USED TO MAKE RECOMMENDATIONS The following sources were reviewed and used to arrive at the recommendations as to which long-lived (i.e., half-lives greater than six months) radionuclides should be included in the present analysis. The nuclides selected from each source and considered as candidates for the analysis are listed in Table D-6. Each source is referred to by a mnemonic or a short title, which in most cases is the document number. D.I.I IAEA-TECDOC-855 Table I of "Clearance Levels for Radionuclides in Solid Materials: Application of Exemption Principles" (IAEA 1996) presents clearance levels—expressed in units of Bq/g—for the unconditional release of material with radioactive contamination. To determine these levels, the IAEA reviewed a large number of documents. The following four documents are relevant to the release of metals (including steel, aluminum, and copper): "Principles for the Exemption of Radiation Sources and Practices from Regulatory Control," Safety Series No. 89 (IAEA 1988); "Radiological Protection Criteria for the Recycling of Materials from Dismantling of Nuclear Installations," Radiation Protection No. 43 (CEC 1988); "Basis for Criteria for Exemption of Decommissioning Waste" (Elert et al. 1992); and "Radiological Impacts of Very Slightly Radioactive Copper and Aluminium Recovered from Dismantled Nuclear Facilities" (Garbay and Chapuis 1991). The radionuclides that were included in the radiological assessments of clearance (along with their respective release limits) in each of these four documents are listed in Table 1.3 of IAEA 1996. Only those nuclides that are associated with clearance of metals are considered as candidates for the present analysis. D.1.2 NUREG/CR-0134 In "Potential Radiation Dose to Man from Recycle of Metals Reclaimed from a Decommissioned Nuclear Power Plant," NUREG/CR-0134 (O'Donnell et al. 1978), the authors present individual and population dose factors resulting from scrap metal recycle for 27 radionuclides. These nuclides "... include fission and activation products (except gaseous species) that may be encountered during decommissioning, and that have radioactive half-lives longer than about 40 days, 239Pu and 241Am (to characterize transuranic contaminants), and 234U, 235U, and 238U." D-l ------- D.1.3 WINCO-1191 The radionuclides reported in "Radionuclides in the United States Commercial Nuclear Power Reactors," WINCO-1191 (Dyer 1994) were taken from a study of pipe samples and pipe surface contamination from pressurized and boiling water reactors; they are listed in Table D-l. The samples were from 11 pressurized water reactors (PWRs) and "over" eight boiling water reactors (BWRs). The data were based on surface samples taken from the inside of stainless steel piping, a main coolant system check valve, and from fuel element hardware. The study also includes an analysis of the Shippingport reactor material samples. Radionuclides that are found exclusively in the coolant or within the fuel cladding are not considered to be candidates for inclusion in the present analysis. The study notes that between 86% and 99% of the activities from the pipe walls and pipe surfaces are the activation products Fe-55, Co-60, and Ni-63. The author goes on to note that the distribution of radionuclides in reactor component appears to be the same whether the activities are on surfaces or are within the metal. D.1.4 NUREG/CR-0130 Appendix J of "Technology, Safety and Costs of Decommissioning a Reference Pressurized Water Reactor Power Station," NUREG/CR-0130 (Smith et al. 1978) presents five sets of "reference radionuclide inventories" that were used to characterize a PWR at the time of its decommissioning. Four of the reference inventories are associated with contaminated metal components, and are listed in Table D-2, while the fifth set is for contaminated concrete, and is not relevant to the present study. The metals removed during PWR decommissioning which are contaminated with either activated corrosion products or surface contamination would be candidates for recycling. The authors include the "stainless and carbon steel activation products" classes of radionuclides, which are the contaminants on the reactor vessel and its internals. In a PWR at the time of decommissioning, this metal would be too highly activated to be a candidate for recycling. However, stainless and carbon steel can become activated by other means, or a reactor may have operated for only a short time (e.g., Shoreham), therefore, the radionuclides in these two sets are candidates for inclusion in the present analysis. D-2 ------- Table D-1. Nuclides from WINCO-1191 Nuclide C-14a Mn-54a Fe-55a Co-57b Ni-59a Co-60a Ni-63a Zn-65b Nb-93ma Nb-94a Ag-110mb Mo-93c Sb-125c I-129a Ce-144+Db Pu-238a Pu-239/240a Cm-244a Half-Life (y) 5.73e+03 8.55e-01 2.73e+00 7.44e-01 7.60e+04 5.27e+00 l.OOe+02 6.69e-01 1.46e+01 2.03e+04 6.84e-01 3.50e+03 2.76e+00 1.57e+07 7.81e-01 8.77e+01 2.41e4/6.56e3 1.81e+01 Surface Activity at Shutdown (|lCi/cm2) < 5.9e-08 6.9e-03 2.7 1.78e-05 6.80e-03 2.0 1.55 1.68e-06 1.2e-02 8.4e-05 1.3e-04 1.8e-08d 1.0e-05d <1.6e-08 2.49E-6 1.2e-07 4.7e-08 2.6e-08 a Sample taken from Shippingport B-loop Primary Coolant Check Valve. Total activity in sample: 6.27 |_lCi/cm2. b Sample taken from Ranch Seco Nuclear Power Plant. Total activity in sample: 0.252 |_lCi/cm2. c Sample taken from Shippingport reactor internals. Total activity in sample: 3.85E-3 |_lCi/g. d Specific activity (|_lCi/g) Konzek et al. (1995) revised the PWR decommissioning analysis originally presented by Smith et al. (1978) to reflect current regulations, practices and costs. The authors did not re-analyze the radiological source terms presented in Appendix C by Smith et al. (1978), although they did use "as built" drawings, rather than design drawings, for estimating the volume of waste material and equipment (Bierschbach 1996). This could change the radionuclide inventories but would not result in any major changes to the expected radionuclide distributions in PWR components at the time of decommissioning. D-3 ------- Table D-2. Nuclides Included in NUREG/CR-0130 Nuclide Mn-54 Fe-55 Co-60 Ni-59 Ni-63 Zn-65 Sr-90 Mo-93 Nb-94 Ru-106 Cs-134 Cs-137 Stainless Steel AP a /b / / / / / — / / — — — Carbon Steel AP / / / / / — — / — — — — Activated Corrosion Products / — / — — — — — — / — / Surface Contamination / / / — — — / — — — / / AP = activation product A check mark (V") indicates that the radionuclide is included in the NUREG/CR-0130 reference inventory. D.I.5 NUREG/CR-3585 In "De Minimis Impacts Analysis Methodology," NUREG/CR-3585, (Oztunali and Roles 1984), the authors present an analysis of the impacts of clearance of metals. Any metal which met the de minimis activity level would have been considered to be a candidate for clearance, since it would no longer have been under regulatory control. D.1.6 NUREG/CR-4370 "Update of Part 61 Impacts Analysis Methodology," NUREG/CR-4370 (Oztunali and Roles 1986) was reviewed as a source of information concerning the radiological profile of scrap which would be disposed of as low-level waste—cleared scrap would have a similar profile. The report analyzed 53 radionuclides, increased from the 23 analyzed in the original Part 61 analysis methodology. Table D-3 list these 53 nuclides. Oztunali and Roles (1986) identified 148 waste streams, for which they developed radionuclide characterizations. Only three of the 148 streams are directly applicable to the recycling of scrap: D-4 ------- 1. The nuclear power plant decommissioning contaminated metals 2. The West Valley Demonstration Project equipment and hardware 3. Non-compressible trash Table D-3. Nuclides Analyzed in NUREG/CR-4370 Nuclide H-3 C-14 Na-22 Cl-36 Fe-55 Co-60 Ni-59 Ni-63 Sr-90 Nb-94 Tc-99 Ru-106 Ag-108m Cd-109 Sn-126 Sb-125 1-129 Cs-134 Notes a, b, c a, b, c NI -- a, c a, c a, c a, b, c a, b, c a, c a, b, c b NI NI b b a, b, c b Nuclide Cs-135 Cs-137 Eu-152 Eu-154 Pb-210 Ac-227 Th-228 Th-229 Rn-222 Ra-226 Ra-228 Th-230 Th-232 Pa-231 U-232 U-233 U-234 U-235 Notes a, b, c a, b, c b b NI HLW -- NI NI -- NI HLW NI HLW HLW -- c a, c Nuclide U-236 U-238 Np-237 Pu-236 Pu-238 Pu-239 Pu-240 Pu-241 Pu-242 Pu-244 Am-241 Am-243 Cm-242 Cm-243 Cm-244 Cm-248 Cf-252 Notes c a, c a, b, c c a, b, c a, b, c a, c a, b, c a, b, c NI a, b, c a, b, c b, c a, b, c a, b, c HLW HLW a Associated with the nuclear-power-plant-decommissioning contaminated metals waste streams b Associated with the West Valley Demonstration Project equipment and hardware waste streams c Associated with non-compressible trash waste streams NI Nuclide was not included in the characterization of any of the waste streams in NUREG/CR-4370, may be included as a decay product of another nuclide which is included in the waste stream characterization. HLW Nuclide was only included in the spent fuel reprocessing high-level liquid waste stream. D.1.7 SAND92-0700 In volume 3 of the "Preliminary Performance Assessment for the Waste Isolation Pilot Plant," SAND92-0700/3, Peterson (1992) estimates the radionuclide inventories in DOE-generated D-5 ------- transuranic (TRU) waste that would be disposed of at the Waste Isolation Pilot Project (WIPP). Because the radionuclides present in TRU waste are a likely source of the contamination of metals present at DOE facilities, Peterson's memo is included in the present review. The memo classified TRU waste as to whether it can be contact handled (CH) or whether remote handling (RH) is required. Both types of TRU waste are considered for the scrap recycle analysis—Table D-4 indicates the type of TRU waste in which the radionuclide may be found. Table D-4. Nuclides Analyzed by SAND92-0700 for WIPP Nuclide Mn-54 Co-60 Ni-63 Sr-90 Tc-99 Ru-106 Sb-125 Cs-134 Cs-137 Ce-144 Pm-147 Eu-152 Eu-154 Eu-155 Half-Life (y) 8.56e-01 5.27e+00 l.OOe+02 2.91e+01 2.13e+05 l.Ole+00 2.77e+00 2.06e+00 3.00e+01 7.78e-01 2.62e+00 1.33e+01 8.80e+00 4.96e+00 RHa / / / / / / / / / / / / / / CHb — — — / — / — — / / / — — — Nuclide Th-232 U-233 U-235 U-236 U-238 Np-237 Pu-238 Pu-239 Pu-240 Pu-241 Pu-242 Am-241 Cm-244 Cf-252 Half-Life (y) 1.41e+10 1.59e+05 7.05e+08 2.34e+07 4.47e+09 2.14e+07 8.77e+01 2.41e+04 6.56e+03 1.44e+01 3.75e+05 4.33e+02 1.81e+01 2.64e+00 RHa / / / / / / / / / / / / / / CHb / / / — / / / / / / / / / / Waste requires remote handling due to high external exposure rate Waste can be handled by direct contact D-6 ------- D.I. 8 ORIGEN The Oak Ridge Isotope Generation and depletion code (ORIGEN) (Croff 1980) includes a radionuclide library with approximately 1,700 entries collected into three groups: activation products, transuranics, and fission products. Included are 1,040 individual nuclides (a given nuclide can appear in more than one group), 127 of which have half-lives greater than six months. To determine which of these 127 radionuclides should be included in the present analysis, an ORIGEN analysis was performed to calculate the activity in spent fuel at the time of discharge from the reactor. An initial enrichment of 3.04% U-235 was assumed, with a burnup of 44,340 MW-days per metric ton of initial heavy metal (MWD/MTIHM), and the characteristics of PWR fuel with impurities. For the purpose of this selection process, it was assumed that the specific activity of a given nuclide in scrap metal from a nuclear facility would be proportional to its activity in the spent fuel inventory. Furthermore, it was assumed that the dose to an exposed individual from a given nuclide, via one of the three pathways (inhalation, ingestion and external exposure) considered in the radiological assessments presented in the main body of this report, would be proportional to the dose conversion factor (DCF) for that pathway. (The DCFs are listed in Federal Guidance Reports (FGR) No. 1 1 [Eckerman et al. 1988] for internal exposure and No. 12 [Eckerman and Ryman 1993] for external exposure.)1 We therefore assigned a "significance," which we define as the product of the activity in spent fuel and the DCF, to each of the 127 nuclides. For each pathway, we found the nuclide with the highest significance. We then calculated the ratio of the significance of each nuclide for each pathway to the significance of the maximum nuclide — the one with the highest significance where: Ry = significance ratio for radionuclide /' and pathway y The scoping analysis described in this section was performed in support of the 1997 Draft "Technical Support Document: Evaluation of the Potential for Recycling of Scrap Metals from Nuclear Facilities." This scoping analysis was but one of nine criteria used in the radionuclide selection process, and contributed at most 2 points out of a possible score of 30. Although the radiological assessments presented in the main body of the present report utilized the revised internal exposure DCFs from ICRP Publication 68 (ICRP 1994), it is unlikely that the selected radionuclides would change if the more current DCFs were used in the selection process. D-7 ------- Aj = spent fuel activity for radionuclide /' Fy = dose conversion factor for radionuclide /' in pathway j (FOR 1 1 for internal, FGR 12 infinite soil coefficients for external) Am = spent fuel activity for radionuclide with the maximum significance for pathway j mjj = DCF for the radionuclide with the maximum significance for pathway y The results of this scoping analysis are listed in Table D-5. D 1.9 SAND9 1-2795 The "Yucca Mountain Site Characterization Project, TSPA 1991: An Initial Total-System Performance Assessment for Yucca Mountain, SAND91-2795 (Barnard et al. 1992) presents an analysis of the impacts from the disposal of spent fuel. Because the radionuclides present in spent fuel are a likely source for the contamination of metals present in nuclear power plants and other tail-end fuel cycle facilities, this report was included in the present review. D.2 RADIONUCLIDES RECOMMENDED FOR INCLUSION Table D-6 lists all radionuclides with half-lives greater than six months which were included in the present review. A check mark (------- Table D-5. Nuclides from ORIGEN with Normalized Activity-Weighted Dose Factors Nuclide H-3 Be-10 C-14 Na-22 Si-32 Cl-36 Ar-39 Ar-42 K-40 Ca-41 V-49 V-50 Mn-54 Fe-55 Co-60 Ni-59 Ni-63 Zn-65 Se-79 Kr-81 Kr-85 Rb-87 Sr-90 Zr-93 Nb-91 Nb-93m Nb-94 Mo-93 Tc-97 Tc-98 Tc-99 Ru-106 Eu-154 0° Soil O.OOe+00 2.96e-15 3.95e-12 O.OOe+00 2.09e-16 1.38e-ll 3.33e-14 Inhalation 3.04e-08 1.16e-12 7.27e-10 O.OOe+00 2.16e-14 1.50e-10 O.OOe+00 Ingestion 2.31e-06 1. 15e-12 5.51e-08 O.OOe+00 1.74e-14 1.57e-09 O.OOe+00 Not in FOR 11 or 12 2.73e-15 O.OOe+00 O.OOe+00 3.85e-17 1.49e-13 O.OOe+00 4.39e-15 1.076-11 O.OOe+00 Not in FOR 11 or 12 4.64e-06 O.OOe+00 8.77e-04 O.OOe+00 O.OOe+00 2.44e-04 3.75e-12 1.05e-14 6.17e-05 1.37e-15 8.11e-04 O.OOe+00 7.15e-09 1.63e-08 1.40e-05 1.666-11 6.27e-09 1.59e-06 2.35e-09 O.OOe+00 O.OOe+00 3.73e-14 5.09e-02 3.24e-07 2.24e-07 2.79e-07 1.31e-04 9.796-11 4.36e-08 8.55e-05 1.58e-07 O.OOe+00 O.OOe+00 4.31e-12 4.53e-01 1.27e-07 Not in FOR 11 or 12 6.54e-12 8.24e-10 2.15e-13 O.OOe+00 3.486-11 7.79e-10 4.30e-01 5.38e-02 2.18e-09 4.186-11 1.236-11 O.OOe+00 1.10e-13 6.13e-08 l.SSe-Ol 2.37e-03 2.95e-09 5.476-11 4.426-11 O.OOe+00 1.78e-12 8.16e-07 8.20e-01 6.01e-03 Nuclide Rh-102 Pd-107 Ag-108m Ag-llOm Cd-109 Cd-113m In-115 Sn-119m Sn-121m Sn-126 Sb-125 Te-123 1-129 Cs-134 Cs-135 Cs-137 Ba-133 La-137 La-138 Ce-142 Ce-144 Nd-144 Pm-145 Pm-147 Pm-146 Sm-145 Sm-146 Sm-147 Sm-148 Sm-149 Sm-151 Eu-152 U-233 oo Soil 1.16e-05 O.OOe+00 6.40e-08 6.04e-02 9.07e-09 2.45e-08 2.30e-21 4.15e-07 2.57e-10 5.11e-06 1.94e-02 1.20e-20 2.15e-10 l.OOe+00 6.88e-12 l.Sle-01 1.75e-36 O.OOe+00 7.05e-15 Inhalation 1.27e-07 1.09e-09 2.23e-09 3.35e-04 8.36e-08 6.86e-05 2.53e-17 1.02e-06 1.72e-09 5.19e-08 1.31e-04 2.28e-20 3.42e-09 5.79e-03 9.70e-10 2.01e-03 8.16e-39 O.OOe+00 1.44e-15 Ingestion 8.42e-07 9.72e-10 4.54e-09 3.42e-03 7.29e-07 5.48e-04 8.10e-17 1.73e-05 2.47e-08 8.19e-07 2.61e-03 6.86e-19 4.13e-07 6.96e-01 1.14e-07 2.38e-01 2.70e-37 O.OOe+00 4.69e-16 Not in FOR 11 or 12 1.71e-01 2.33e-01 l.OOe+00 Not in FOR 11 or 12 O.OOe+00 2.27e-06 8.39e-06 O.OOe+00 O.OOe+00 O.OOe+00 O.OOe+00 2.11e-03 3.31e-07 O.OOe+00 1.10e-ll 4.466-11 O.OOe+00 4.28e-03 6.28e-07 O.OOe+00 2.05e-12 8.40e-12 Not in FOR 11 or 12 Not in FOR 11 or 12 1.72e-10 1.68e-05 7.03e-15 6.23e-06 6.26e-07 8.08e-10 6.13e-06 1.39e-06 1.31e-10 D-9 ------- Table D-5 (continued) Nuclide Eu-155 Eu-150 Gd-152 Gd-153 Tb-157 Ho- 163 Ho- 166m Tm-171 Lu-176 Hf-182 Ta-180 Re- 187 Os-194 IT- 192m Pt-190 Pt-193 Tl-204 Pb-204 Pb-205 Pb-210 Bi-208 Bi-210m Ra-226 Ra-228 Ac-227 Th-228 Th-229 Th-230 Th-232 Pa-231 U-232 0° Soil 8.27e-04 7.03e-ll O.OOe+00 6.08e-06 O.OOe+00 Inhalation 2.23e-04 2.56e-12 2.76e-17 7.01e-07 O.OOe+00 Ingestion 6.24e-04 4.62e-12 1.38e-18 2.62e-06 O.OOe+00 Not in FOR 11 or 12 3.24e-08 2.01e-12 4.83e-33 O.OOe+00 O.OOe+00 O.OOe+00 5.32e-17 1.84e-14 2.88e-09 1.956-11 1.50e-33 O.OOe+00 O.OOe+00 1.81e-19 7.74e-17 1.68e-15 2.28e-09 6.966-11 1.26e-33 O.OOe+00 O.OOe+00 2.41e-18 1.41e-16 2.25e-15 Not in FOR 11 or 12 1.73e-19 O.OOe+00 8.25e-18 O.OOe+00 3.27e-16 O.OOe+00 Not in FOR 11 or 12 6.92e-21 1.39e-17 4.56e-18 6.26e-14 1.44e-16 1.49e-12 Not in FOR 11 or 12 1.31e-14 7.07e-14 5.70e-18 3.12e-13 1.26e-08 1.46e-13 9.83e-15 4.79e-21 1.06e-12 5.60e-12 8.51e-14 6.43e-14 5.73e-18 1.24e-09 5.06e-07 2.35e-10 3.14e-09 1.79e-14 8.47e-09 4.85e-06 8.16e-14 7.53e-13 1.24e-16 2.06e-10 8.98e-08 3.326-11 4.01e-10 2.26e-15 5.30e-09 7.31e-07 Nuclide U-234 U-235 U-236 U-238 Np-235 Np-236 Np-237 Pu-236 Pu-238 Pu-239 Pu-240 Pu-241 Pu-242 Pu-244 Am-241 Am-242m Am-243 Cm-243 Cm-244 Cm-245 Cm-246 Cm-247 Cm-248 Cm-250 Bk-249 Cf-249 Cf-250 Cf-251 Cf-252 Es-254 oo Soil 1.32e-10 2.89e-09 2.166-11 1. 80e-08 1.426-11 1.71e-12 1.76e-07 1.10e-10 2.51e-07 3.99e-08 3.36e-08 1.35e-06 1.74e-10 1.38e-12 2.83e-06 2.73e-07 1.68e-05 1.14e-05 4.28e-07 1.22e-07 1.246-11 8.17e-13 1.31e-16 4.83e-19 4.75e-14 4.21e-12 1.27e-14 3.71e-13 3.21e-14 l.lle-12 Inhalation 5.17e-05 5.57e-07 1. 50e-05 1.67e-05 2.166-11 4.58e-10 1.03e-04 8.24e-05 7.71e-01 6.88e-02 1.17e-01 7.00e-01 6.63e-04 3.28e-10 3.41e-02 2.09e-03 9.82e-03 7.12e-03 l.OOe+00 1.93e-04 5.71e-05 2.16e-10 2.92e-09 2.83e-15 1.16e-08 1.56e-09 3.34e-08 4.91e-10 3.39e-08 9.71e-12 Ingestion 8.40e-06 9.20e-08 2.43e-06 2.87e-06 9.596-11 2.89e-10 6.40e-05 5.04e-05 4.77e-01 4.30e-02 7.30e-02 4.41e-01 4.12e-04 2.05e-10 2.12e-02 1.30e-03 6.14e-03 4.42e-03 6.17e-01 1.20e-04 3.55e-05 1.35e-10 1.82e-09 1.77e-15 7.59e-09 9.69e-10 2.06e-08 3.07e-10 1.78e-08 5.64e-12 D-10 ------- For each radionuclide identified in one or more of the sources reviewed, a score was calculated by summing the weighting factors for each source in which the radionuclide appeared. These scores are shown in the second column from the right (headed "score") in Table D-6. Those radionuclides with a score of 10 or greater are recommended for inclusion in the scrap recycle analysis, as indicated by a check mark in the last column of Table D-6. Members of the thorium and uranium radioactive decay series have been recommended for inclusion even if they have scores below 10, to enable the radiological assessment of the entire series in secular equilibrium. D-ll ------- Table D-6. Selection of Nuclides to Be Included in Scrap Recycle Analysis Nuclide H-3 C-14 Na-22 Cl-36 Mn-54 Fe-55 Co-57 Co-60 Ni-59 Ni-63 Zn-65 Se-79 Rb-86 Sr-90 Zr-93 Nb-93m Nb-94 Mo-93 Tc-99 Ru-106 Pd-107 Source (weighting factor) NUREG/ CR-0134 (5) — / / — / / — / / / / — — / — — — — / / — IAEA 1996 (6) — — — — / / — / — / / — — / — — / — / / — WINCO 1191 (4) — / — — / / / / / / / — — — — / / / — — — NUREG/ CR-0130 (4) — — — — / / — / / / / — — / — — / / — / — NUREG/ CR-3585 (3) / / / / / / / / / / / — / / — — / — / / — NUREG/ CR-4370 (2) / / — — — / — / / / — — — / — — / — / / — SAND 92 -0700 (2) — — — — / — — / — / — — — / — — — — / / — ORIGEN (2) — — — — — — — / — — / — — / — — — — — / — SAND 91-2795 (2) — / — / — — — — / / — / — / / — / / / — / 2 0 o GO 5 16 8 5 24 24 7 28 20 28 24 2 3 26 2 4 21 10 20 24 2 Include | — / — — / / — / / / / — — / — — / / / / — ------- Table D-6 (continued) Nuclide Ag-108m Ag-llOm Cd-109 Cd-113m Sn-121 Sn-126 Sb-125 1-129 Cs-134 Cs-135 Cs-137 Ce-144 Pm-147 Sm-151 Eu-152 Eu-154 Eu-155 Pb-210 Ra-226 Ra-228 Ac-227 Source (weighting factor) NUREG/ CR-0134 (5) — — — — — — — — / — / / — — — — — — — — — IAEA 1996 (6) — / — — — — — — / — / / / — / — — — — — — WINCO 1191 (4) — / — — — — / / — — — / — — — — — — — — — NUREG/ CR-0130 (4) — — — — — — — — / — / — — — — — — — — — — NUREG/ CR-3585 (3) / / / — — / / / / / / / — — / / — / / / / NUREG/ CR-4370 (2) — — — — — / / / / / / — — — / / — — — — — SAND 92 -0700 (2) — — — — — — / — / — / / / — / / / — — — — ORIGEN (2) — / — / — — / — / — / / / — — / / — — — — SAND 91-2795 (2) / — — — / / — / — / / — — / — — — / / — / CL> O O GO 5 15 3 2 2 7 13 11 24 7 26 22 10 2 13 9 4 5 5 3 5 Include | — / — — — — / / / — / / / — / — — / / / / ------- Table D-6 (continued) Nuclide Th-228 Th-229 Th-230 Th-232 Pa-231 U-232 U-233 U-234 U-235 U-236 U-238 Np-237 Pu-236 Pu-238 Pu-239 Pu-240 Pu-241 Pu-242 Pu-244 Am-241 Am-242 Source (weighting factor) NUREG/ CR-0134 (5) — — — — — — — / / — / — — — / — — — — / — IAEA 1996 (6) — — — — — — — / / — / / — — / / / — — / — WINCO 1191 (4) — — — — — — — — — — — — — / / / — — — — — NUREG/ CR-0130 (4) — — — — — — — — — — — — — — — — — — — — — NUREG/ CR-3585 (3) / / / / / / / / / / / / / / / / / / / / — NUREG/ CR-4370 (2) — — — — — — — / / / / / / / / / / / — / — SAND 92 -0700 (2) — — — / — — / — / / / / — / / / / / — / — ORIGEN (2) — — — — — — — — — — — / — / / / / / — / — SAND 91-2795 (2) — / / — / / / / / / / / — / / / / / — / / CL> O O GO O 5 5 5 5 5 7 18 20 9 20 17 5 15 26 21 17 11 3 22 2 Include | / / / / / — — / / — / / — / / / / / — / — ------- Table D-6 (continued) Nuclide Am-242m Am-243 Cm-242 Cm-243 Cm-244 Cm-245 Cm-246 Cm-248 Cf-252 Source (weighting factor) NUREG/ CR-0134 (5) — — — — — — — — — IAEA 1996 (6) — — — — / — — — — WINCO 1191 (4) — — — — / — — — — NUREG/ CR-0130 (4) — — — — — — — — — NUREG/ CR-3585 (3) — / — / / — — / / NUREG/ CR-4370 (2) — / / / / — — — — SAND 92 -0700 (2) — — — — / — — — / ORIGEN (2) / / — / / / — — — SAND 91-2795 (2) — / — / / / / — — CL> O O GO 2 9 2 9 21 4 2 3 5 Include | — — — — / — — — — ------- REFERENCES Barnard, R. W., et al. 1992. "Yucca Mountain Site Characterization Project, TSPA 1991: An Initial Total-System Performance Assessment for Yucca Mountain," SAND91-2795. Sandia National Laboratories, Albuquerque, NM. Bierschbach, M. C., (Pacific Northwest Laboratory). 1996. Private communication. Commission of the European Communities (CEC). 1988. "Radiological Protection Criteria for the Recycling of Materials from Dismantling of Nuclear Installations," Radiation Protection No. 43. Croff, A. 1980. "A User's Manual for the ORIGEN2 Computer Code," ORNL/TM-7175. Oak Ridge National Laboratory, Oak Ridge, TN. Dyer, N. C. 1994. "Radionuclides in United States Commercial Nuclear Power Reactors," WINCO-1191, UC-510, ed. T. E. Bechtold. Westinghouse Idaho Nuclear Company, Inc., prepared for the Department of Energy, Idaho Operations Office. Eckerman, K. F., A. B. Wolbarst and A. C. B. Richardson. 1988. "Limiting Values of Radionuclide Intake and Air Concentration and Dose Conversion Factors for Inhalation, Submersion, and Ingestion," Federal Guidance Report No. 11, EPA-520/1 -88-020. U.S. Environmental Protection Agency, Washington, DC. Eckerman, K. F., and J. C. Ryman. 1993. "External Exposure to Radionuclides in Air, Water, and Soil," Federal Guidance Report No. 12, EPA 402-R-93-081. U.S. Environmental Protection Agency, Washington, DC. Elert, M., et al. 1992. "Basis for Criteria for Exemption of Decommissioning Waste," Rep. Kemakta Ar 91-26. Kemakta Konsult AB. Garbay, H., and A. M. Chapuis. 1991. "Radiological Impacts of Very Slightly Radioactive Copper and Aluminium Recovered from Dismantled Nuclear Facilities," Rep. EUR-13160- FR. Commission of the European Communities. International Atomic Energy Agency (IAEA). 1988. "Principles for the Exemption of Radiation Sources and Practices from Regulatory Control," Safety Series No. 89. IAEA, Vienna. International Atomic Energy Agency (IAEA). 1996. "Clearance Levels for Radionuclides in Solid Materials: Application of Exemption Principles," Interim Report for Comment, IAEA- TECDOC-855. IAEA, Vienna. D-16 ------- International Commission on Radiological Protection (ICRP). 1994. "Dose Coefficients for Intakes of Radionuclides by Workers," ICRP Publication 68. Annals of the ICRP, vol. 24, no. 4. Pergamon Press, Oxford. Konzek, G. J., et al. 1995. "Revised Analyses of Decommissioning for the Reference Pressurized Water Reactor Power Station," NUREG/CR-5884, PNL-8742. Vol. 1, "Main Report." Pacific Northwest Laboratory prepared for the U.S. Nuclear Regulatory Commission, Washington, DC. O'Donnell, F. R., et al. 1978. "Potential Radiation Dose to Man from Recycle of Metals Reclaimed from a Decommissioned Nuclear Power Plant," NUREG/CR-0134. Oak Ridge National Laboratory, Oak Ridge, TN. Oztunali, O. I, and G. W. Roles. 1984. "De Minimis Waste Impacts Methodology," NUREG/CR-3585. U.S. Nuclear Regulatory Commission, Washington DC. Oztunali, O. I, and G. W. Roles, 1986. "Update of Part 61 Impacts Analysis Methodology NUREG/CR-4370. U.S. Nuclear Regulatory Commission, Washington DC. Peterson, A. C., 1992. "Preliminary Contact Handled (CH) Radionuclide and Nonradionuclide Inventories and Remote Handled Radionuclide Inventory for Use in 1992 Performance Assessment." Memorandum in "Preliminary Performance Assessment for the Waste Isolation Pilot Plant." Vol. 3, "Model Parameters," SAND92-0700/3, p. A-135. Sandia WIPP Project Office, Sandia National Laboratories, Albuquerque, NM. Smith, R.I., G. J. Konzek, and W. E. Kennedy, Jr. 1978. "Technology, Safety and Costs of Decommissioning a Reference Pressurized Water Reactor Power Station," NUREG/CR- 0130. 2 vols. Pacific Northwest Laboratory, prepared for the U.S. Nuclear Regulatory Commission, Washington, DC. D-17 ------- APPENDIX E DISTRIBUTION OF CONTAMINANTS DURING MELTING OF CARBON STEEL ------- Contents page E.I Introduction E-l E.2 Thermodynamic Calculation of Partition Ratios E-l E.3 Correlation with Other Forms of Partition Ratio E-7 E.4 Estimates of the Partitioning of Other Contaminants E-8 E.5 Observed Partitioning E-l 1 E.5.1 Americium E-12 E.5.2 Antimony E-14 E.5.3 Carbon E-16 E.5.4 Cerium E-16 E.5.5 Cesium E-16 E.5.6 Chlorine E-18 E.5.7 Chromium E-18 E.5.8 Cobalt E-19 E.5.9 Europium E-20 E.5.10 Hydrogen E-21 E.5.11 Iridium E-22 E.5.12 Iron E-22 E.5.13 Lead E-23 E.5.14 Manganese E-23 E.5.15 Molybdenum E-25 E.5.16 Nickel E-25 E.5.17 Niobium E-26 E.5.18 Phosphorus E-27 E.5.19 Potassium and Sodium E-27 E.5.20 Plutonium E-27 E.5.21 Radium E-28 E.5.22 Silver E-28 E.5.23 Strontium E-29 E.5.24 Sulfur E-29 E.5.25 Thorium E-30 E.5.26 Uranium E-30 E.5.27 Zinc E-31 E.5.28 Zirconium E-32 E.6 Inferred Partitioning E-32 E.6.1 Curium E-32 E.6.2 Promethium E-32 E.7 Summary E-32 References E-38 Appendix E-l: Extended Abstracts of Selected References El-1 E-iii ------- Contents (continued) page Appendix E-2: Composition of Baghouse Dust E2-1 References E2-3 Tables E-l. Partition Ratios at 1,873 K for Various Elements Dissolved in Iron and Slag E-5 E-2. Standard Free Energy of Reaction of Various Contaminants with FeO at 1,873 K E-9 E-3. Normal Boiling Point of Selected Potential Contaminants E-l 1 E-4. Selected References on the Distribution of Potential Contaminants During SteelmakingE-13 E-5. Distribution of Antimony Between Slag and Metal E-14 E-6. Distribution of Cs-134 Following Steel Melting E-17 E-7. Hydrogen and Oxygen Concentrations in Liquid Iron E-22 E-8. Proposed Distribution of Potential Contaminants During Carbon Steelmaking E-35 El-1. Distribution of Radionuclides in Tracer Tests at WERF El-3 El-2. Specific Activities of Ingots and Slags El-4 El-3. Distribution of Radionuclides Following Laboratory Melts El-9 E2-1. Composition of Baghouse Dust E2-2 E-iv ------- DISTRIBUTION OF CONTAMINANTS DURING MELTING OF CARBON STEEL E.I INTRODUCTION During the melting of potentially contaminated steel, the contaminants may be distributed among the metal product, the home scrap, the slag, the furnace lining, and the offgas collection system. In addition, some contaminants could pass through the furnace system and be vented to the atmosphere. In order to estimate the radiological impacts of recycling potentially contaminated scrap steel, it is essential to understand how the contaminants are distributed within the furnace system. For example, a gaseous chemical element (e.g., radon) will be exhausted directly from the furnace system into the atmosphere while a relatively non-volatile element (e.g., manganese) can be distributed among all the other possible media. This distribution of potential contaminants is a complex process that can be influenced by numerous chemical and physical factors, including composition of the steel bath, chemistry of the slag, vapor pressure of the particular element of interest, solubility of the element in molten iron, density of the oxide(s), steel melting temperature, and melting practice (e.g., furnace type and size, melting time, method of carbon adjustment, and method of alloy additions). This appendix discusses the distribution of various elements with particular reference to electric arc furnace (EAF) steelmaking. The next three sections consider the calculation of partition ratios for elements between metal and slag based on thermodynamic considerations1. Section E.5 presents laboratory and production measurements of the distribution of various elements among slag, metal, and the offgas collection system. Section E.6 proposes distributions for those elements where theoretical or practical information is lacking and Section E.7 provides recommendations for the assumed distribution of each element of interest. E.2 THERMODYNAMIC CALCULATION OF PARTITION RATIOS Partitioning of a solute element between a melt and its slag under equilibrium conditions can be calculated from thermodynamic principles if appropriate data are available. Consider a divalent Reference to a given element does not necessarily imply that it is in the elemental form. For instance, a metallic element might be found in the elemental state in the melt while its oxide is found in the slag. E-l ------- solute element M, such as cobalt, dissolved in molten iron, which reacts with FeO in the slag according to the following equation: M + FeO(slag) = MO(sla (E-l) where M is the symbol for solute dissolved in liquid iron. Equation E-l can be written as the difference between the following equations: M + l/£>2 = MO and Fe + Y2O = FeO (E-2) (E-3) The Gibb's free energy for Equation E-l, energies of Equations E-2 and E-3, viz.: can be expressed as the difference in the free = AF° - AF° Thermodynamic data for Equation E-2 are normally tabulated assuming that the standard state for M is the pure liquid or solid, but it is often desirable to convert from the pure elemental standard state to a hypothetical standard state where M is in a dilute solution. In steelmaking, 1 wt% M in solution in iron is commonly used for this new standard state2 as defined by the transformation: M (E-4) The free energy change for M from the pure state to M in the dilute state is (Darken and Gurry 1953): AF° = RTln Y°M: 100M Fe M Concentrations are expressed here as wt% instead of mass % since wt% is commonly used in the steelmaking literature. The terms are synonymous. E-2 ------- T = absolute temperature in kelvin (K) R = universal gas constant = 1.987cal/mole-K Y°M = Henry's Law activity3 coefficient (based on atom fraction) of M at infinite dilution in iron MFe = atomic weight of iron = 55.85 MM = atomic weight of M Equation E-2 can also be written as the difference of Equation E-5 (below) and Equation E-4. M (pure) (E-5) Therefore, AF°2 = AF°5 - AF°4 and the Gibb's free energy change for Equation E-l can be written as AF° = AF° - AF° - AF° = AF°MO-AF°Fe0- RTln Y°M, Fe 100M M where AF°f is the free energy of formation of the particular oxide. (E-6) At equilibrium AF° = - RTlnK = - RT In where a is the activity of each species in Equation E-l and Kx is the equilibrium constant. In the steel bath, aFe can be assumed to be 1, while aFe0 = YFeoNFe0- To estimate NFe0 (the mole fraction of FeO in the slag), the nominal composition of the slag was assumed to be 50 wt% CaO, 30 wt% SiO2, and 20 wt% FeO. Thus, NFe0 = 0.167. Various investigators have described the activity of FeO in ternary mixtures of CaO, FeO, and SiO2 (Philbrook and Bever 1951, Ansara In Sections E. 1, E.2, and E.3, activity refers to thermodynamic activity, not radioactivity. E-3 ------- and Mills 1984). For the slag composition assumed here, based on the ternary diagram by Ansara and Mills (1984), when NFe0 is 0.2, aFe0 is about 0.4 (i.e., Ypeo i§ about 2). Consequently, aFe0 = 0.333. For the dilute standard state, aM is equal to wt% M and, for dilute solutions of MO in the slag, one can assume that aMO = NM0. It follows that (E-8) v ' wt%M RT NMO where - is one form of the partition ratio for M between the melt and the slag. wt% M For metal oxides other than those formed from divalent cations, the different stoichiometries must be accommodated in Equations E-6, E-7, and E-8. Using values of y° for various solute elements in iron at 1,873 K tabulated by Sigworth and Elliott (1974)4 and free energy of formation data for oxides tabulated by Glassner (1957), partition ratios between melt and slag were calculated for the present analysis and are presented in Table E-l. Values in the last column of Table E-l will be described in Section E.3. When the partition ratio is large, the solute element is strongly concentrated in the slag under equilibrium conditions. This is true for Al, Ce, Nb, Ti, U, and Zr, which all have partition ratios (as defined here) of 80,000 or greater. Similarly, when the partition ratio is small, the solute element is concentrated in the molten iron. Examples of this are Ag, Co, Cr, Cu, Ni, Pb, Sn, Mo, and W, which all have partition ratios of 0.008 or less. Mn, Si, and V, with partition ratios ranging from about 3 to 40, are expected to be more evenly distributed between melt and slag. Silver will not react with FeO in the slag, so on the basis of slag/metal equilibria, this element should remain in the melt. However, silver has a relatively high vapor pressure at steelmaking temperatures (i.e., 10"2 atm at 1,816 K), so some would tend to be removed at a rate dependent on the rate of transfer of silver vapor through the slag. The value of y° for cerium is from Ansara and Mills 1984. A compendium of values for y° similar to that by Sigworth and Elliot 1974 has been prepared by the Japan Society for the Promotion of Science (1988). Some differences exist between values in Sigworth and Elliot 1974 and JSPS 1988, particularly for W, Co, Pb, and Ti. JSPS 1988 proposes a value of y° for Ce(1) of 0.332. This difference in y° values does not affect the conclusions about cerium partitioning. E-4 ------- Table E-l. Partition Ratios at 1,873 K for Various Elements Dissolved in Iron and Slag M A§(i) ^0) Ca(g) Ce(1) Cod) Cr(i> Cua) Mn(1) Mo(s) Nt>(.) Nio) Pba) sia) Sn Ti(S) ua) vw w(s) Zr(s) Oxide Ag20 A1203 CaO CeO2 CoO Cr203 Cu2O MnO MoO3 Nb2O5 MO PbO SiO2 SnO2 TiO2 UO2 V205 WO3 ZrO2 v° y M 200 0.029d 2240 0.026 1.07 1.14 8.6 1.3e 1.86 1.4 0.66 1400 0.0013 2.8 0.038 0.027 0.1 1.2 0.037 AF°f,MO (kcal/mole)a +20.6 -257 -104 -176 -18.2 -80.0 -11.0 -58.0 -89.1 -275 -19.0 -15.5 -129 -47.6 -147 -180 -206 -96.2 -178 Partition Ratio (NMO/wt%M) 3.89e-04b'c 1.32e+05b 1.53e+09 4.33e+07 4.79e-05 1.21e-04b 1.99e-03b 2.74e+00 1.23e-05 8.12e+04b 3.72e-05 8.55e-03 3.76e+01 6.07e-06 7.72e+04 8.87e+07 7.68e+00b 2.77e-05 1.59e+08 (mass in slag/ mass in metal) l.le+10 l.le+09 5.0e-04 2.7e+01 2.1e-04 3.9e-04 3.2e-01 1.9e+02 1.3e-04 6.6e+05 3.8e+09 9.1e-04 2.6e+09 a AF°fFe0 = -34.0 kcal/mole b PR = N'/2/wt% M c Ag will not react with FeO, Ag2O unstable at 1,873K d According to Ansara and Mills (1984), Y°AI = °-005 e According to Ansara and Mills (1984), yV =1-48 It is instructive to examine the impact of assuming a dilute solution in iron rather than the pure element as the standard state for the solute. For those elements that tend to partition strongly to the melt (Co, Cr, Cu, Mo, Ni, Sn, and W), change of standard state from the pure metal to the dilute solution increases partitioning to the melt by factors of about 10 to 300. Lead is an exception, presumably due to its strong deviation from ideal solution behavior. Similarly, use of a dilute solution as the standard state decreases partitioning to the slag for the strong oxide formers such as Al, Ce, Nb, Ti, U, and Zr by factors of about 100 to 16,000. The exception is E-5 ------- calcium with strong positive deviation from ideality. These observations emphasize the importance of using a dilute solution as the standard state when adequate data are available. As noted previously, the calculations in Table E-l assumed, for simplicity, that the activity of MO in the slag was equal to the mole fraction (i.e., YMO = !)• This may not be a good assumption. If, for example, YMO = 0.01, NMO would increase 100-fold. Work by Ostrovski (1994) on the partitioning of tungsten in steel melted in a 25-t EAF illustrates the impact of melting practice and slag chemistry on the activity of WO3 in the slag. When the steel was melted under strongly oxidizing conditions utilizing a 30-minute oxygen blow, the activity coefficient was found to be a function of the ratio %CaO:%SiO2 in the slag and varied from about 10"2 to about 10"4 as the CaO:SiO2 ratio increased from 1:1 to 4:1. Typical measured values of log-^— — — — were between 1 and 2, where (% W) and [% W] are the tungsten contents of [wt% W] the slag and the metal, respectively5. A good fit between experimental and calculated partition ratios was obtained using the following equations: logYwo =-2.076 -0.592 & rw°3 (%Si02) and , (%W) 3054 . _, , , , . M 108 = — - 4-56 - log Y- + 3 log + log[MW03(%eO+ nCaO+ nSiO2 + nWO3)] where n is the number of moles per 100 grams of the various slag components. With this melting practice, approximately 94% of the tungsten in the feed was transferred to the slag, 4% remained in the melt, and the balance was lost. This emphasizes that special melting practices can produce substantially different results from the predictions in Table E-l. The thermodynamic treatment used to derive the partition ratios in Table E-l assumes that the melt is a binary system of iron and solute M, while in practice the melt will actually be a multi- component solution. In recent years, a considerable amount of work has been done to develop, both theoretically and experimentally, a solution model which considers interactions between The convention of using (x) and [y] to signify concentrations or components in the slag and the metal, respectively, is commonly used in the technical literature and will generally be used in this appendix. E-6 ------- solute elements (Engh 1992, Sigworth and Elliot 1974, Ansara and Mills 1984). The activity of element /' in dilute solution can be expressed as: a; = f; (wt% i) where f; is the Henry's Law activity coefficient (for concentrations expressed in wt%). The first order interaction coefficients e;j are defined by the equation log f; = £ 6ij (% j) (Higher order terms are possible but are not considered here.) Using, for illustrative purposes, a low alloy 4140 steel with the nominal composition 0.4% C, 0.04% S, 0.9% Cr, and 0.1% Co, and the interaction coefficients for cobalt with these elements in liquid iron from Engh 1992, fCo was calculated to be 0.975. For this example, the impact of the binary interactions on cobalt activity in iron is quite small. Unfortunately, interaction coefficients for many of the elements of interest in the melting of potentially contaminated scrap metals are not available to refine the calculations summarized in Table E-l. E.3 CORRELATION WITH OTHER FORMS OF PARTITION RATIO In the literature, the partition ratio (PR) may be expressed in a variety of ways. For example, in Chapter 9 of SCA 1995, partition ratios are expressed as "mass in slag/mass in steel." It is of interest to compare this formulation with the definition in column 5 of Table E-l (i.e., NMO/wt% M). The SCA 1995 PR may be expanded as: (wt%M)m PR = —* (E-9) [wt% M] ms l ' mg = mass of slag ms = mass of steel and, if one assumes that the relevant reaction is that in Equation E-2, one can write: (wt%MO)m M PR = [wt%M]msMMO E-7 ------- where MM and MMO are the atomic weight of M and the molecular weight of MO, respectively. Equation E-10 is based on the premise that the reaction involves a divalent solute metal. It is equally true for all oxides where the ratio of the anion to the cation is an integer. For simplicity, if one assumes that the slag consists of two oxide components MO and RO and that wt% MO is « wt% RO, then one can write that (wt% MO)/M MO or that 100 N M (wt% MO) = M0 M0 (E-12) AA ^- * 1V1RO which can be substituted into Equation E-10 to give ^m MM PR = M0 g M (E-13) [wt%M]msMRO l > Equation E-13 relates the partition ratio as defined in SCA 1995 to that in Table E-l. Column 6 of Table E-l converts the partition ratios in column 5 to the formulation in SCA 1995 (i.e., mass in slag/mass in metal), using the assumptions and simplifications described above, and further assuming that the ratio, mass of slag : mass of metal is 1:10 and RO is CaO. This conversion is only done for those oxides where the anion/cation ratio is an integer. E.4 ESTIMATES OF THE PARTITIONING OF OTHER CONTAMINANTS Values of the Henry's Law activity coefficient (Y°M) are n°t available for many solute elements of interest in recycling potentially contaminated steel scrap. However, an indication of partitioning between the melt and the slag can be obtained by calculating the Gibb's free energy for the reaction (E-14) E-8 ------- where M is the pure component rather than the solute dissolved in the melt and FeO and MxOy are slag components. Values of the standard free energy change for Equation E-14 are summarized in Table E-2 for all instances where the reaction occurs in the direction written. Table E-2. Standard Free Energy of Reaction of Various Contaminants with FeO at 1,873 K Element Ac(1) Am(1) Ba(1) Bi(g) Cd(g) Cs(1) Ir(S) K(g) Na(g) NP(D Paa) p°(g) p% Ra(g) Re(S) Ru(s) sb(g) Se(g) Sma) Sr(g) Tc(S) Th(S) Ya) Zn(g) Oxide Ac2O3 Am2O3 BaO Bi203 CdO Cs2O IrO2 K2O Na2O Np02 PaO2 PoO2 PuO3 RaO ReO2 RuO4 Sb2O3 SeO2 Sm2O3 SrO TcO2 ThO2 Y203 ZnO AF° (kcal) -120 -103 -57.1 -100 -94.7 -103 -47.7 -102 -58.6 -142 -101 Comments Ac should partition to slag Am should partition to slag Ba should partition to slag Bi will not react with FeO, some may vaporize from melt CdO unstable at 1873 K, Cd should vaporize from the melt Cs2O unstable at 1873 K, Cs should vaporize from melt, some Cs may react with slag components IrO2 unstable above ~ 1 100 K, Ir should remain in melt K2O less stable than FeO, other K compounds stable in slag Na2O less stable than FeO, other Na compounds stable in slag Np should partition to slag Pa should partition to slag PoO2 unstable above ~ 1300 K, Po assumed to vaporize from melt Pu should partition to slaga Ra should partition to slag Re will not react with FeO, Re should remain in melt RuO4 unstable above =1700 K, Ru should remain in melt Sb will not react with FeO, some may vaporize from melt Se will not react with FeO, some may vaporize from melt Sm should partition to slag Sr should partition to slag, but low boiling point could cause some vaporization Tc will not react with FeO, should remain in melt Th should partition to slag Y should partition to slag Zn will not react with FeO, Zn should vaporize from melt The reaction between Pu and FeO to form PuO2 is slightly more forward thermodynamically than the reaction to form PiuO, E-9 ------- Table E-2 shows that Ac, Am, Ba, Np, Pa, Pu, Ra, Sm, Sr, Th, and Y all will react with FeO to form their respective oxides as indicated by the calculated free energies. Thus, these elements should be preferentially distributed to the slag. By chemical analogy to similar species in Table E-l, one can estimate that the partition ratios (NMO/wt% M) should be on the order of 104 or greater6. The solute elements Bi, Cd, Cs, Ir, K, Na, Re, Ru, Sb, Se, Tc, and Zn do not react with FeO either because the oxides are unstable or because Equation E-14 is thermodynamically unfavorable. Of these elements, Ir, Re, Ru, and Tc are expected to remain in the melt. As indicated in Table E-3, the solute elements Bi, Cd, Cs, Po, Sb, Se, and Zn have low boiling points and would be expected to vaporize from the melt to some degree at typical steelmaking temperatures of 1,823 K to 1,923 K. For example, cesium would tend to be removed at a rate dependent on the rate of transfer of vapor through the slag unless some stable compound such as Cs2SiO3 forms in the slag. Should Cs2O form during the melting process before a continuous slag had formed, it would be volatilized since the boiling point of the oxide is about 915 K. The boiling point of metallic cesium is in the same temperature range. Even though an element may have a low boiling point, it cannot be assumed, a priori, that the element will completely vaporize from the melt. Some may remain in the melt and some may be contained in the slag. For example, elements such as Ca, Mg, K, and Na are found as oxides and silicates in steel slags (Harvey 1990). Pehlke (1973) has shown that, for a solute M dissolved in a solvent (liquid Fe), the following equation applies: PM = vapor pressure of Mover melt PM° = vapor pressure of pure M YM = activity coefficient of M in melt The free energies in Table E-2 were recalculated assuming that y° in Equation E-6 was unity, and partition ratios were then calculated using Equation E-8. All partition ratios calculated in this manner for elements expected to partition to the slag were greater than 104 except Ba (6,300) and Ra (320). If all these calculated partition ratios were reduced by a factor of 103 to adjust for the fact that values of y° are expected to be less than unity, estimated partition ratios are greater than 103 for all slag formers except Ba (6.3), Ra (0.321), and Sr (15). These three elements are in Group II of the periodic table and have electronic structures and chemical properties similar to calcium. As discussed previously in Section E.2, calcium has a value of y° = 2,240. By analogy, one would expect that the partition ratios of Ba, Ra, and Sr would actually be higher than calculated with y° = 1 • For example, if YRa° = 2,000, the partition ratio for radium, as defined by Equation E-8, would be 6 x 105. E-10 ------- = mole fraction of M in melt Table E-3. Normal Boiling Point of Selected Potential Contaminants Contaminant Bi Cd Cs Pb Po2 Ra S2 Se2 Sb2 Zn Normal Boiling Point (K) 1900 1038 963 2010 1300 1410 1890 1000 1890 1180 Source: Darken and Gurry 1953 Thus, as the temperature of the melt increases, the quantity of the volatile element M in the melt decreases by an amount determined by the temperature dependency of PM°. Based on vapor pressure data for Pb, Sb, and Bi by Brandes and Brooks (1992) and Zn from Perrot et al. (1992), one can estimate that increasing the temperature of the iron bath from 1,873 K to 1,923 K will reduce the amount of Pb, Sb, and Bi by about 25% while that of Zn will be reduced by about 18% (assuming that YM i§ independent of temperature over the same range and PM is constant). Actually, YM ls an increasing function of temperature for antimony (Nassaralla and Turkdogan 1993) and a decreasing function for zinc (Perrot et al. 1992). E.5 OBSERVED PARTITIONING This section discusses available experimental and production information on the distribution of possible contaminant elements among melt, slag, and the offgas collection system in steelmaking. Several of the key references are abstracted in Appendix E-l, which describes test conditions and relevant results from selected publications. Since many of the references cited in this section discuss the distribution of multiple elements in a single test, it would be cumbersome to repeat all the experimental details here for each element. Table E-4 summarizes the references by contaminant element. Substantial additional information on these and other references are presented by Worchester et al. (1993). Some additional perspective concerning the E-ll ------- concentrations of impurities and alloying elements can be obtained by examining the composition of a typical low carbon steel (i.e SAE 1020) as shown below: •C 0.18-0.23% •Mn 0.60-0.90% •P < 0.04% • S < 0.05% Thus, the steel melting process must control carbon and manganese within specified ranges and insure that the maximum concentrations of sulfur and phosphorus are not exceeded. The furnace charge, the melting conditions, and the slagging practice must all be carefully managed to achieve the desired steel chemistry. E.5.1 Americium Based on the thermodynamic equilibria, americium would be expected to partition strongly to the slag. Gomer of British Steel reported that, when melting reactor heat exchanger tubing contaminated with Am-241 in a 5-t EAF, traces of Am-241 were found in the slag. No other Am-241 was detected (Pflugard et al. 1985). In laboratory steel melting experiments in a 5-kg furnace, the Am-241 distribution was 1% in the ingot, 110%7 in the slag, and 0.05% in the aerosol offgas filter, resulting in a partition ratio between slag and metal of about 100 (Schuster and Haas 1990, Schuster et al. 1988). Americium is chemically similar to uranium which partitions strongly to the slag (Harvey 1990). On the basis of the available information, americium is expected to partition to the slag as predicted by the thermodynamic calculations. However, one caveat is offered by Harvey (1990). Since the density of the AmO2 is high (11.68 g/cm3), transfer of americium to the slag may be retarded by gravity. In small-scale laboratory experiments using mild steel (see Section E.5.20 for details), americium was observed to partition to the slag (Gerding et al. 1997). Ratios of the concentration of americium in slag to the concentration of americium in metal generally exceed 1000:1. Because of differences in detection efficiencies, more radioactivity is sometimes detected in the products than was measured in the furnace charge. E-12 ------- Table E-4 Selected References on the Distribution of Potential Contaminants During Steelmaking Element Ag Am C Ce Co Cr Cs Eu Fe H Ir Mn Mo Nb Ni P Pb Pu Ra S Sb Sr Th U Zn Zr References Sappok et al. 1990, Harvey 1990, Menon et al. 1990 Pflugard et al. 1985, Schuster and Haas 1990, Schuster et al. 1988 Schuster and Haas 1990, Stubbles 1984b Sappok et al. 1990, Harvey 1990 Nakamura and Fujiki 1993, Pflugard et al. 1985, Sappok et al. 1990, Larsen et al. 1985a, Schuster and Haas 1990, Harvey 1990, Schuster et al. 1988, Menon et al. 1990 Stubbles 1984a Nakamura and Fujiki 1993, Larsen et al. 1985a, Larsen et al. 1985b, Pflugard et al. 1985, Sappok et al. 1990, Harvey 1990, Menon et al. 1990 Sappok et al. 1990, Larsen et al. 1985a, Harvey 1990 Schuster and Haas 1990, Schuster et al. 1988 Stubbles 1984b Larsen et al. 1985b Nakamura and Fujiki 1993, Sappok et al. 1990, Stubbles 1984a, Meraikib 1993, Harvey 1990, Menon et al. 1990 Stubbles 1984a, Chen et al. 1993 Stubbles 1984a, Harvey 1990 Harvey 1990, Stubbles 1984a, Schuster and Haas 1990 Stubbles 1984b Stubbles 1984a Gerding et al. 1997, Harvey 1990 Starkey etal. 1961 Stubbles 1984b Harvey 1990, Menon et al. 1990, Stubbles 1984a, Kalcioglu and Lynch 1991, Nassaralla and Turkdogan 1993 Nakamura and Fujiki 1993, Larsen et al. 1985b, Schuster and Haas 1990 Harvey 1990 Harvey 1990, Larsen et al. 1985a, Schuster and Haas 1990, Heshmatpour and Copeland 1981, Abe et al. 1985 Harvey 1990, Nakamura and Fujiki 1993, Sappok etal. 1990, Stubbles 1984a, Menon etal. 1990 Stubbles 1984a E-13 Continue ------- Back E.5.2 Antimony As described previously, antimony will not react with FeO in the slag and therefore is expected to remain in the melt. However, as noted in Table E-3, the normal boiling point of antimony (1890 K) is at steelmaking temperatures and at least some vaporization would be expected. Contrary to this prediction, Harvey (1990) reports "...that when antimony is added to steel it is recovered with high yield.". This view is supported by Philbrook and Bever (1951), who observed that antimony is probably almost completely in solution in steel. On the other hand, Stubbles (1984a) indicates that antimony is volatilized from scrap during EAF melting. In no case is adequate background information provided to support the statements8. Kalcioglu and Lynch (1991) found that antimony could be removed from carbon-saturated iron (typical of blast furnace operations) if temperatures exceeded 1,823 K and the slag basicity, B (CaO) + (MgO) (Si02) + (A1203) was greater than 1. Using very small samples consisting of 2 g of slag and 3 g of steel, about 45% to 51% of the antimony was vaporized at 1,823 K when the slag basicity was unity. The distribution of antimony between slag and metal is presented in Table E-5. Table E-5. Distribution of Antimony Between Slag and Metal [wt%Sb]a 0.40 0.46 0.51 T b J-sb 0.55 0.59 0.67 [wt%Sb] = concentration in metal b Lsb = (wt%Sb)/[wt%Sb] (wt%Sb) = concentration in slag When the slag basicity was 0.818, values of Lsb ranged from 0.09 to 0.13, and when the basicity was 0.666, Lsb ranged from 0.05 to 0.08 at 1,823 K. The reaction which caused the marked In a recent telephone conversation, Dr. J. R. Stubble, currently Manager of Technology at Charter Steel Company, advised that his conclusions in Stubbles 1984a were based on the high vapor pressure of antimony rather than experimental steel melting evidence. He would not argue against Harvey's conclusions (Stubbles 1996). E-14 ------- increase in antimony partitioning to the slag when the basicity was increased to 1 was not identified. In a proposed follow-on study to the work of Kalcioglu and Lynch, Zhong (1994) suggested that the reaction 2Sb +3(FeO) +(O2 ) = 2(Sb(V) +3Fe(/) has an estimated value for AF° of-4 kcal. While not strongly favoring partition to the slag, the reaction can proceed as written particularly since aFeOand aO2-tend to be high in basic slags. Using data presented by Zhong, the partition ratio for the above reaction can be roughly estimated to be 0.006—a value similar to those for copper and lead in Table E-l9. The calculation supports the conclusion that antimony will not partition to the slag to a significant degree. This conclusion is reinforced by the work of Nassaralla and Turkdogan (1993) who stated that "....most of the antimony will remain in the metal phase. However, it should be possible to remove some antimony from the hot metal by intermixing it with lime-rich flux under highly reducing conditions." Using values of y°sb developed by these investigators, one can calculate a partition ratio for antimony of 8 x 10"6 at 1,873 K. Based on calculated partition ratios (above and in Table E-l), vapor pressures of the pure metals (Table E-3), and vapor pressures of the metal oxides10, one would expect that antimony and lead would behave similarly. It is therefore unclear why antimony tends to remain in the melt and lead is primarily collected in the bag house. This may be a manifestation of significantly higher activity of lead as compared to antimony in molten iron. Menon et al. (1990) measured the distribution of Sb-125 from two heats of stainless steel. Activities of 4.3 x io5 Bq were detected in the melt and 1.7 x io3 Bq in the baghouse dust. No activity was reported in the slag. This calculation uses a value for Y°Sb measured in carbon-saturated iron. According to Perry and Green (1984), the vapor pressures of PbO and Sb4O6 are one atmosphere at 1,745 K and 1,698 K, respectively. E-15 ------- E.5.3 Carbon Carbon is a carefully controlled element in steelmaking. Excess carbon is often added to the melt and then reduced to its final level by oxygen decarburization. This process promotes slag/metal reactions and assists in removing hydrogen from the melt (Stubbles 1984b). CO produced by the decarburization reaction combines with atmospheric oxygen in the offgas to form CO2, which is exhausted from the system (Philbrook and Bever 1951). If, for example, 5 kg/t of charge carbon are added to a melt that nominally contains 2.5 kg of carbon per tonne of scrap and the objective is to produce steel with a final carbon content of 0.2% (i.e., an SAE 1020 steel), 0.55 wt% C must be removed. Thus, about 73% of the carbon would be exhausted from the system and the balance would remain in the melt. The distribution of carbon between the melt and the offgas is dependent upon the carbon content of the scrap charge, the melting practice (i.e., use of charge carbon), and the desired carbon content of the finished steel. E.5.4 Cerium Based on thermodynamic calculations, cerium should strongly partition to the slag as CeO2 or Ce2O3. Sappok et al. (1990) have described experience in induction melting of contaminated steel from nuclear installations. All Ce-144 contamination was found in the slag, although details of the melting and slagging practice were not discussed. Cerium is sometimes added to steel to react with oxygen and sulfur. Since CeO2 has a density of 6.9 g/cm3, which is similar to that of molten steel, Harvey (1990) suggests that the density of the oxide retards transfer to the slag and, consequently, some CeO2 may remain as non-metallic inclusions in the steel. According to JSPS (1988), Ce2O3 rather than CeO2 is the stable oxide during steelmaking. In addition, JSPS recommends a value of 0.322 for y° in dilute iron solutions. These differing assumptions do not alter the conclusion—developed from the calculations in Section E.2—that cerium strongly partitions to the slag. Using the data recommended by JSPS, the partition ratio M 1/2 ,, . ^MO • -, -, , -, „« for cerium, , is 1.15 x 108. wt%M E.5.5 Cesium Based on free energy and vapor pressure considerations, cesium would be expected to volatilize from the melt. Furthermore, cesium has no solubility in liquid iron. According to ASM 1993: E-16 ------- From the scant data reported here and by analogy with other iron-alkali metal binary phase diagrams, it is evident that Cs-Fe is virtually completely immiscible in the solid and liquid phases. A number of investigators have reported measurements on the experimental distribution of cesium during steel melting. Sappok et al. (1990) observed that during air induction melting of about 2,000 tons of steel, no Cs-134/137 remained in the melt. Cesium was found both in the slag and in the dust collection system but the distribution was not quantified. At the Japanese Atomic Energy Research Institute (JAERI), Nakamura and Fujiki (1993) obtained similar results from air induction melting of both ASTM-A33511 and SUS 304 steels. The Cs-137 was about equally distributed between the slag and the dust collection system, but only about 77% of the amount charged was recovered. At the Idaho National Engineering Laboratory (INEL), Larsen et al. (1985a) found cesium both in the slag and in the baghouse dust when melting contaminated scrap from the Special Power Excursion Reactor Test (SPERT) III. In tracer tests, Larsen et al. (1985b) found that 5% to 10% of the cesium remained in Type 304L stainless steel ingots. Gomer described results of three 5-t EAF and one 500-kg induction furnace melts in which the chemical form of cesium addition and the slag chemistry were varied (Gomer and Lambley 1985, Pflugard et al. 1985). The distribution of this nuclide, based on the fraction of Cs-134 recovered, is summarized in Table E-6. Table E-6. Distribution of Cs-134 Following Steel Melting Furnace Type EAF Induction EAF EAF Cs Addition CsCl CsOH CsOH Cs2SO4 Cs Distribution (%) Steel 0 0 0 0 Slag 0 100 7 66 Off Gas 100 0 93 34 Cs Recovery (%) 100 91 50 64 11 This ASTM specification covers various seamless ferritic alloy steel pipes for high temperature service. E-17 ------- In the melt where the cesium was added as CsCl, the chloride, which is volatile below the steel melting temperature, was not collected in the slag because the slag had not formed before the CsCl had completely evaporated. In the induction furnace test, CsOH was added to the liquid steel under a quiescent acid slag. In the related arc furnace test with CsOH, the slag was not sufficiently acid to promote extensive formation of cesium silicate, which would be retained in the slag. In the arc furnace melt with the Cs2SO4 addition, this compound was apparently incorporated into the slag to a significant extent. Harvey (1990) concluded that the hot, basic slags typical of EAF melting were not conducive to cesium retention in the slag. A comparison of three arc furnace melts with varying slag compositions showed the following amounts of cesium retention in the slag 16 minutes after cesium was added to the melt: •SiO2:CaO= 3.1:1 50% recovery • SiO2:CaO= 1.3:1 < 4% recovery •SiO2:CaO= 0.41:1 0 recovery In these tests, no cesium remained in the melt. Menon et al. (1990) recounted that no cesium was found in the ingots or the slag after melting 332 metric tons (t) of carbon steel in an induction furnace, but that substantial Cs-137 (21,000 Bq/kg) was collected in the ventilation filters. During production of two heats of stainless steel, no cesium was found in the ingots; 32% was in the slag; and 68% in the baghouse dust (Menon et al. 1990). E.5.6 Chlorine The disposition of chlorine depends on its form at the time of introduction into the EAF furnace. Any chlorine gas would be desorbed from the scrap metal surface and vented to the atmosphere. If the contaminant exists as a metal chloride, it is likely to be distributed between the slag and the baghouse dust. Cl" has been reported in baghouse dust (McKenzie-Carter et al. 1985). E.5.7 Chromium From a theoretical viewpoint, chromium would be expected to remain primarily in the melt. However, Stubbles (1984a) suggests that chromium recovery in the melt during EAF steelmaking E-18 ------- is only 30% to 50%. Stubbles' observation is not consistent with the calculations in Table E-l, which show chromium remaining primarily in the melt. Xiao and Holappa (1993) have studied the behavior of chromium oxides in various slags at temperatures between 1,773 K and 1,873 K. They reported that chromium in the slag was mainly (i.e., 88% to 100%) Cr2+ when the mol% CrOx in the slag was 10% or less and the NCa0:Nsi02 ratio was unity. The calculations in Table E-l assumed Cr+3 to be the predominant species. Using free energy data presented by these authors for the reaction Cr(s) + V2O2 = CrO(1) (AF° = -79,880 + 15.25 T cal) and other relevant data from Table E-l, the partition ratio involving CrO rather than Cr2O3 is calculated to be 0.42. This suggests that a significant portion of the chromium will partition to the slag if Cr+2 is the principal cation in the slag. E.5.8 Cobalt Free energy calculations indicate that cobalt should remain primarily in the melt. Nakamura and Fujiki (1993) found this to be the case in 500-kg air induction melts of carbon steel and stainless steel where Co-60 was detected only in the ingots. During the melting of six heats of contaminated carbon steel scrap at INEL, some (unquantifiable) Co-60 activity was detected in the dust collection system and some in the slag (Larsen et al. 1985a). In subsequent tracer tests with three heats of Type 304L stainless steel, between 96% and 97% of the Co-60 was recovered in the ingots (Larsen et al. 1985b). Sappok et al. (1990) noted that, during the induction melting of steel, Co-60 was mostly found in the melt although unquantifiable amounts were detected in the slag and in the dust collection system. In an earlier paper, Sappok cited the Co-60 distribution from nine melts totaling 24 t as 97% in the steel, 1.5% in the slag, and 1.5% in the cyclone and baghouse (Pflugard et al. 1985). Schuster and Haas (1990) measured the Co-60 distribution in laboratory melts of St37-2 steel and reported 108% in the ingot, 0.2% in the slag, and 0.2% in the aerosol filter. According to Harvey (1990), " ...cobalt-60 will almost certainly be retained entirely in the steel in uniform dilution in both electric arc and induction furnaces." In support of this conclusion, Harvey described two steel melts in a 5-t EAF. In one test, highly reducing conditions were employed (high carbon and ferrosilicon) while, in the other, the conditions were oxidizing E-19 ------- (oxygen blow). In neither case was any measurable cobalt activity found in the slag. The amount of Co-60 found in the melt was in good agreement with the amount predicted from the furnace charge. No Co-60 was found in the furnace dust although some was expected based on transfer of slag and oxidized steel particles to the gas cleaning system. Harvey concluded that the low level of radioactivity in the furnace charge (ca. 0.23 Bq/g) coupled with dilution from dust already trapped in the filters resulted in quantities of Co-60 in the offgas below the limits of detection. Menon et al. (1990) commented on the air induction melting of 33.6 t of carbon steel. No Co-60 was detected in the slag, but a small quantity (1,300 Bq/kg) was detected in the baghouse dust. The amount remaining in the ingots was not quoted. In two heats of stainless steel weighing a total of 5 t, 26 MBq of Co-58/Co-60 were measured in the ingots, 40 kBq in the slag, and 78 kBq in the baghouse dust. E.5.9 Europium Based on its chemical similarity to other rare-earth elements such as samarium, cerium, and lanthanum, europium is expected to partition to the slag. During induction melting of steel scrap from nuclear installations, Sappok et al. (1990) reported that all the Eu-154 was in the slag. Larsen found some europium in the slag and some in the baghouse dust during induction melting of scrap from the SPERT III reactor. The europium content was below the limits of detection in the feed material, so presumably some unquantified concentrating effects occurred in the slag and the offgas dust (Larsen et al. 1985a). Eu-152 concentrations in the baghouse dust were very low—on the order of 0.8 pCi/g. Harvey (1990) described production of an experimental 3.5-t melt of steel in an arc furnace to study europium partitioning. During the melting operation, oxygen was blown into the melt to remove 0.2% C (typical of normal steelmaking practice). The radioactivity of the metal was too low to be measured and no europium was found in the dust from the fume extraction system. Europium activity was detected only in the slag. Even though there was some concern expressed that, because of the similar densities of steel and Eu2O3 (7.9 g/cm3 and 7.4 g/cm3, respectively), the Eu2O3 would not readily float to the metal/slag interface, the experimental results suggest this was not an issue. With regard to the fact that no europium was found in the fume collection system, Harvey (1990) observed: It is inevitable, however, because of the nature of the process, that some slag is ejected into the atmosphere of the arc furnace and is then entrained in the offgas and is collected in the gas cleaning filters. Hence any radioactive component present in the slag will be present to E-20 ------- some extent in the offgas. The fact that it is not detected on this occasion reflects the small amount of radioactivity used, and the mixing and dilution of dust which occurs in the gas cleaning plant. E.5.10 Hydrogen Hydrogen is an undesirable impurity in steel, causing embrittlement. Thus steelmaking practice seeks to keep the contaminant at very low levels. As noted in Section E.5.3, removal of charge carbon by blowing oxygen through the melt reduces the hydrogen as well. Stubbles (1984b) described tests on the rate of hydrogen removal as a function of time and carbon reduction rate. For steel with an initial hydrogen content of 9 ppm, the hydrogen level was reduced to 1 ppm after 15 minutes when the rate of carbon removal was 1% per hour and to 5 ppm over the same interval when the carbon removal rate was 0.1% per hour. Stubbles' work is consistent with results reported by Deo and Boom (1993) who showed that the rate of hydrogen removal was directly related to the rate of carbon removal. They also described the work of Kreutzner (1972) who investigated the solubility of hydrogen in steel at 1,873 K and 1,973 K. From a graphical presentation of Kreutzner's work, one can estimate that the solubility of hydrogen in steel at 1,873 K can be expressed as [H] = 27 PH* where [H] is the hydrogen solubility in ppm and ?„ is the hydrogen partial pressure in atmospheres. Thus, when PH is 0.01 atm, the eqiulibrium hydrogen concentration is 2.7 ppm. Since the most likely source of hydrogen is from water in the charge components or the furnace atmosphere, the following reaction should also be considered (Philbrook and Bever 1951): H20(g) = 2H + O At 1,873 K, the equilibrium hydrogen concentration is %H = 1.35-10"3 ao, E-21 ------- where a0 is the activity of oxygen in the melt. One can see from this equation that the %H increases as a0 decreases. Table E-7 lists the co oxygen concentrations when PH Q is 0.003 atm. increases as a0 decreases. Table E-7 lists the concentrations of H for various assumed dissolved Table E-7. Hydrogen and Oxygen Concentrations in Liquid Iron (PH Q= 0.003 atm) Concentration (%) O 0.1 0.01 0.001 H 2.5e-04 8e-04 2.5e-03 If the oxygen content of the bath is low, the steel can absorb more hydrogen from water vapor than from pure hydrogen at 1 atm. Hydrogen or water vapor in materials added to the bath after carbon removal or to the furnace ladle will tend to be retained in the product steel (Philbrook and Bever 1951). E.5.11 Iridium Iridium would be expected to remain in the melt during steelmaking. Iridium and iron are completely miscible in the liquid phase (ASM 1993). INEL conducted one induction melting test at the Waste Experimental Reduction Facility (WERF) where Ir-192 was added to Type 304L stainless steel to produce about 500 Ib of product. About 60% of the charged iridium was recovered in the ingot but only small quantities were detected in the slag. Although the material balance was poor, there is no basis to conclude that iridium does not primarily remain in the melt (Larsenetal. 1985b). E.5.12 Iron Iron oxide is a major slag component. According to a 1991 survey by the National Slag Association, the average FeO content of steel slags is 25% (NSA 1994). If one assumes that the ratio of slag mass to steel mass is 0.1, then about 2% of the iron in the charge would be distributed to the slag. Schuster et al. reported some laboratory tests where Fe-55 was added to small melts of steel conducted under an Ar + 10% H2 atmosphere and reducing conditions (Schuster and Haas 1990, Schuster et al. 1988). No Fe-55 was found in the slag or the aerosol E-22 ------- filter. However, these results have little relevance to expected partitioning under actual steelmaking conditions. E.5.13 Lead As shown in Table E-l, lead should remain with the melt rather than with the slag. At 1,873 K, lead has limited solubility in molten iron — about 0.064 to 0.084 wt% (ASM 1993). Although the boiling point of lead (2,010 K) is above normal steelmaking temperatures, lead has a significant vapor pressure (ca. 0.4 atm) at 1,873 K. In addition, any PbO which forms during initial heating of the furnace charge could volatilize before the steel begins to melt since PbO is a stable gas at steelmaking temperatures (Glassner 1957, Kellog 1966). Consequently, much of the lead should be transferred from the melt either as lead vapor or as gaseous PbO and be collected in the offgas system. Stubbles (1984a) reports that, when leaded scrap is added to liquid steel, the lead boils off like zinc and is collected with the fume. If lead in the form of batteries or babbitts is added to the furnace charge, the lead will quickly melt and sink to the bottom of the furnace where it may penetrate the refractory lining. E.5.14 Manganese Manganese is a common element in steelmaking. As discussed above, a typical carbon steel contains 0.6 to 0.9% Mn. Calculations in Section E.2 show that manganese should be more concentrated in the slag than in the metal. For EAF melting, Stubbles states that about 25% of the manganese is recovered in the steel. This establishes the partition ratio based on the mass of manganese in slag to the mass of manganese in steel at 3:1. Meraikib (1993) complied information on manganese distribution between slag and molten iron based on a large number of heats in a 70-ton EAF. He showed that the ratio of the concentration of manganese in the slag to manganese in the metal, r^, is given by the following equation: (Mn) = - 0.0629 B - 7.3952 (Mn) = concentration of Mn in slag (wt%) E-23 ------- [Mn] = concentration of Mn in melt (wt%) a[0] = activity of oxygen in melt f[Mn] = activity coefficient for [Mn] All other terms have been defined previously. For the range of manganese concentrations (0.06 to 1.0 wt%) and the range of temperatures (1,823 K to 1,943 K) studied, f[Mn] is essentially unity (i.e., 0.9503). If one assumes that B = 2 and a[0] = 0.004, then the variation of r^ with temperature can be calculated as follows: 1,843 K ^ = 6.3 1,943 K ^ = 2.9 indicating that the ratio of the concentrations manganese in slag and in metal can vary by a more than factor of two for a 100 K change in melt temperature. Based on the work of Meraikib, the partitioning of manganese between slag and metal (assuming a slag:metal ratio of 1:10) is an order of magnitude lower than observed by Stubbles and about two orders of magnitude lower than estimated from thermodynamic principles in Section E.2. This suggests that the oxygen activity in the steel in equilibrium with the slags used in Meraikib's work is lower than implied in the free energy calculations in Section E.2 Nakamura and Fujiki (1993) conducted four 500-kg air induction melting tests (two with ASTM-A335 steel and two with SUS 304 stainless steel) to which 24 MBq of Mn-54 were added. In two tests with SUS 304 and one test with ASTM-A335, about 90% of the activity was contained in the ingot, while in the other ASTM-A335 ingot only 50% of the Mn-54 was recovered. For the one ASTM-A335 ingot where the slag concentration was also reported, the distribution based on input radioactivity was: • ingot 91% • slag 8% • unaccounted 2% Sappok et al. (1990) described experience in melting about 2,0001 ofons contaminated steel in a 20-ton induction furnace. The melting process generated only a small amount of slag (i.e., about 1.2%). During a 200-t melting campaign, no Mn-54 was found in the melt. Up to 21.9% of the E-24 ------- total slag activity was attributed to Mn-54 and up to 2.1% of the total activity in the dust collection system was from this nuclide. Harvey (1990) notes that manganese tends to be more concentrated in the slag when melting under oxidizing conditions although the reverse result can be obtained when the furnace conditions are reducing. Manganese is relatively volatile having a vapor pressure of 0.08 atm at 1,900 K. In two stainless steel heats melted at Studsvik, the combined manganese distribution was (Menon etal. 1990): • ingot 44 kBq • slag 3.6 kBq • baghousedust 0.36 kBq E.5.15 Molybdenum As described previously in Section E.2, molybdenum should remain primarily in the melt. Stubbles (1984a) supports this view, indicating that 100% of molybdenum is recovered in the steel during EAF melting. Studies by Chen et al. (1993) on the reduction kinetics of MoO3 in slag also buttress this conclusion. In 1-kg-scale laboratory tests, Chen found that the reduction of MoO3 in slag over an iron-carbon melt was completed in about five minutes. E.5.16 Nickel Nickel is chemically similar to cobalt and should remain in the melt during steelmaking. Stubbles states that nickel recovery during arc melting is 100% (Stubbles 1984a). According to Harvey, it is common practice to add NiO to a steel melt and quantitatively recover the nickel. He further notes: "Nickel cannot be volatilized from molten steel, and there do not appear to be any slags which will absorb nickel selectively." (Harvey 1990). Schuster described the distribution of Ni-63 in laboratory melts of 3 to 5 kg under inert gas (Schuster and Haas 1990). About 82% of the nickel was recovered in the ingot, 0.04% in the slag and 0.06% in the aerosol filter, with the remainder unaccounted for. E-25 ------- E.5.17 Niobium On the basis of the thermodynamic calculations in Section E.2, niobium should partition primarily to the slag. According to Stubbles (1984a), the recovery of niobium from scrap in the ingot is zero during EAF melting, which is consistent with the theoretical calculations. Harvey (1990) notes that niobium can be retained in the steel under reducing conditions, but under oxidizing conditions will clearly be transferred to the slag according to the reaction: 2Nb + 6O + Fe = FeONb2O5 The equilibrium constant for this reaction is : K, = 1 2 6 aNb indicating that the equilibrium is very sensitive to the activity of the oxygen in the steel. At 1,873 K, Kj = 2.4 x 1010. Wenhua et al. (1990) studied the kinetics of Nb2O5 reduction in slag by silicon dissolved in iron according to the reaction: 5 Si + 2(Nb2O5) = 4Nb + 5(SiO2) The reaction was assumed to be divided into five steps: 1. Nb2O5 diffuses through slag towards reaction interface 2. Si diffuses through molten iron towards reaction interface 3. Reaction occurs at interface 4. Reaction product niobium diffuses from interface into molten iron 5. Reaction product SiO2 diffuses from interface into slag Using a slag with a CaO:SiO2 (basicity) ratio of about 2:1 and a ferrosilicon reductant (ca 0.42% Si), niobium was rapidly transferred from the slag to the melt, reaching a value of 1.5% after 10 minutes. Wenhua found that the rate controlling step was the diffusion of niobium in liquid iron. E-26 ------- E.5.18 Phosphorus Phosphorus is an undesirable impurity in steel which is typically removed by oxidation. The transfer of phosphorus from the metal to the slag can be represented by the following simplified reaction (Stubbles 1984b): 2P + 5O = (P2O5) The amount removed from the melt will depend on the phosphorus content of the scrap charge and the desired phosphorus content of the melt. Phosphorus removal is facilitated during EAF melting by increasing the basicity and oxidation level of the slag. By injecting 35 kg of powered lime per tonne into the melt together with oxygen, the phosphorus content can be reduced to about 10% of its initial value. E.5.19 Potassium and Sodium Since K2O is less stable than FeO, potassium should be removed from the melt because of its low boiling point. However, various potassium compounds such as silicates and phosphates are present in slags (Harvey 1990). The same considerations apply to sodium. Na2O has also been collected in EAF baghouse dust (Brough and Carter 1972). Given the fact that Na2O in the slag can be reduced by carbon in the melt (Murayama and Wada 1984), that observation is not surprising. The appropriate chemical equation is: Na20(1) + C = 2Na(g) + CO(g) AF° for this reaction at 1,873 K is -48 kcal/mole. Removal of Na2O from the slag would be enhanced by higher carbon levels in the melt. Presumably, any sodium from this reaction would be vaporized and subsequently condensed in the baghouse as Na2O. E.5.20 Plutonium Thermodynamic predictions suggest that plutonium will partition strongly to the slag. Harvey assumed, based on the chemical similarity of plutonium with thorium and uranium, that the plutonium will form a stable oxide and be absorbed in the slag (Harvey 1990). However, he notes that because of its high specific gravity (11.5), transfer of PuO2 to the slag could be slow and some could possibility fall to the base of the furnace and not reach the slag. E-27 ------- Gerding et al. (1997) conducted small-scale (i.e., 10 g and 200 g) tests with plutonium oxide and mild steel in an electric resistance furnace. The melts were held in contact with various slags for one to two hours at 1,773 K under helium at about 0.5 atm. Slag:steel weight ratios ranged from 0.05 to 0.20. The studies showed that the plutonium partitioned to the slag and the partition coefficients (concentration in slag + concentration in metal) were 2 x 106 to 8 x 106. Decontamination efficiency was about the same at 400 and 14,000 ppm Pu, and differences in composition among the various silicate slags were not significant to the partitioning. E.5.21 Radium Radium forms a stable oxide in the presence of FeO and thus would be expected to be found mainly in the slag. Starkey et al. (1961) described results from the arc furnace melting of eight heats of steel contaminated with radium. The average concentration of the radium in the steel was <9 x 10"13 g Ra/g steel and in the slag was 1.47 x 10"9 g Ra/g slag. Slag/metal mass ratios were not reported, but assuming the mass slag/mass metal is 0.1, then the partitioning ratio (mass Ra in slag/mass Ra in metal) is >160. E.5.22 Silver As noted in Section E.2, silver will not react with FeO because Ag2O is unstable at steelmaking temperatures. Silver has no solubility in liquid iron and thus the two metals will coexist as immiscible liquids (ASM 1993). Since silver has a significant vapor pressure (ca. 10"2 atm at 1,816 K), some volatilization might be expected. Sappok et al. (1990) reported that induction melting of steel contaminated with silver resulted in the silver being primarily distributed to the metal, but some was detected both in the slag and in the offgas dust. However, the distribution was not quantified. Harvey (1990) concluded, based on the instability of Ag2O and the expected similarity to the behavior of copper in steel, that silver "would be expected to remain in the melt under all normal steelmaking conditions." Ag-110m activity was measured for two heats of stainless steel at Studsvik (Menon et al. 1990). The Ag-110m activity was distributed as follows: •ingot 290 kBq • slag 1.3 kBq • baghousedust 93 kBq E-28 ------- E.5.23 Strontium Strontium is predicted to partition to the slag. Nakamura and Fujiki (1993) studied the partitioning of Sr-85 during the air induction melting of ASTM-A335 steel in a 500-kg furnace with a slag basicity of 1. All of the Sr-85 was found in the slag (recovery was 75%). Larsen et al.(1985b) described the melting of three heats of Type 304L stainless weighing 500 to 700 Ib each in an air induction furnace. The amount of strontium remaining in the ingots was 1% in two cases and zero in the third. Sr-85 was found in the slag and the baghouse dust but no mass balance was provided. Slagging practice was not documented other than to state that a small amount of a "slag coagulant" was added to aid in slag removal. Schuster and Haas melted St37-2 steel in a 5-kg laboratory furnace using a carborundum crucible. Lime, silica, and alumina were added as slag formers. The melt was allowed to solidify in situ. About 80% of the Sr-85 was found on the ingot surface, 6.3% in the slag, 0.5% in the ingot, and 0.02% in the aerosol filter. The material on the ingot surface would most likely have been found in the slag under more realistic production conditions. Strontium can also react with sulfur and the resultant SrS should partition to the slag (Bronson and St. Pierre 1985). E.5.24 Sulfur Sulfur is a generally undesirable element except in certain steels where higher sulfur levels are desired for free machining applications. As indicated at the beginning of this section, the maximum sulfur content of a typical low carbon steel is 0.05%. Sulfur is difficult to remove from the melt. One mechanism for sulfur removal is reaction with lime in the slag to form calcium sulfide according to the reaction: CaO + S = CaS + O This reaction is facilitated by constant removal of high basicity slag and agitation. According to Stubbles, the concentration ratio — rarely exceeds 8 in EAF melting of steel (Stubbles 1984b). Although sulfur has a very low boiling point (see Table E-3), the compounds it forms within the slag (e.g., CaS) are very stable at steelmaking temperatures. E-29 ------- Engh (1992) described the partitioning of sulfur between slag and metal as a function of slag acidity and FeO content of the slag. Assuming that the slag contained 25% FeO and 20% acid components (SiO2, P2O5, B2O3, and TiO2), the ratio — would range between about 16 and 26. [S] E.5.25 Thorium Based on the stability of ThO2, thorium should partition to the melt. Harvey (1990) notes that the stability of ThO2 has been exploited by using the material in steel melting crucibles. However, because of their high specific gravity (9.86), ThO2 particles may settle in the melt and not reach the slag. E.5.26 Uranium Free energy calculations suggest that uranium should partition to the slag. Heshmatpour and Copeland (1981) conducted a number of small-scale partitioning experiments where 500 to 1,000 ppm of UO2 was added to 50 to 500 g of mild steel and melted in either an induction furnace or a resistance furnace. Slag and crucible composition were varied as well. With the use of highly fluid basic slags and induction melting, partition ratios (mass in slag:mass in metal) from 1.2:1 to >371:1 were obtained. Larsen et al. (1985a) reported that, although uranium was not detected in the feed stock, it was sometimes found in the slag and in the baghouse dust. Schuster and Haas (1990) determined in small laboratory melts that when slag formers were added, the uranium content was reduced from 330 |lg U/g Fe to 5 |lg U/g Fe. Harvey (1990) commented that British Steel had occasionally used uranium as a trace element in steelmaking. Based on their experience, the uranium was absorbed in the slag in spite of the fact that UO2, which has a density (10.9 g/cm3) significantly higher than that of iron, could conceivably settle in the melt. Abe et al. (1985) studied uranium decontamination of mild steel using small (100 g) melts in a laboratory furnace. Melting was done in an argon atmosphere at a pressure of 200 torrs in alumina crucibles with 10 wt% flux added to the charge. The uranium decontamination factor was found to be a function of the initial contamination level, varying from about 200 to about 5,000 as the uranium concentration increased from 10 to 1,000 ppm. Optimum decontamination occurred when the slag basicity was 1.5 with a CaO-Al2O3-SiO2 slag. Decontamination was further enhanced by additions of CaF2 or NiO to the slag. E-30 ------- E.5.27 Zinc Zinc is not expected to react with the slag constituents and, because of its low boiling point, some fraction should evaporate from the melt. In fact, dust from steelmaking operations is an important secondary source of zinc. In 1990, about 100,000 tonnes of zinc were recovered from baghouse dust in Europe (Perrot et al. 1992). Hino et al. (1994) studied the evaporation of zinc from liquid iron at 1,873 K and found that the evaporation rate was first order with respect to the zinc content of the melt. The mass transfer coefficient in the liquid phase was estimated to be 0.032 cm/s. Nakamura and Fujiki (1993) observed that, when induction melting both ASTM-A335 and SUS 304 steels, about 60% to 80% of added Zn-65 remained in the ingot. In one test with ASTM- A335 steel, 90.7% of the added zinc was recovered. Of the total amount recovered, about 14% was found in the offgas and 1% in the slag, with the balance remaining in the ingot. Sappok et al. (1990) reported that, in some instances, zinc was found only in the offgas collection system and, in another melting campaign, some zinc was found in the ingot and the slag as well as in the offgas system. The causes of these differences are not apparent. On the other hand, Stubbles states that zinc is volatilized during EAF melting (Stubbles 1984a). Harvey (1990) supports the view of Stubbles noting that zinc is volatilized during melting and collected as ZnO in the baghouse filters. "The volatilization is very efficient, and the residual content of zinc in the steel is likely to be below 0.001%, whereas the zinc oxide content of the dust is often more than 10%." Perrot et al. (1992) note that in spite of its low boiling point and expected ease of evaporation, zinc removal from liquid steel is far from complete. Industrial experience indicates that the zinc content is often above 0.1 wt.% in liquid cast iron at 1,573 -1,673 K but is somewhat lower in liquid steel at 1,773- 1,873 K. At 1,773 K, assuming that the zinc vapor pressure over the melt is 0.01 atmosphere, the calculated solubility of zinc in iron is about 72 ppm. The solubility of zinc in liquid iron is decreased by other solute elements with ion interaction coefficients greater than zero (e.g., Al and Si) and decreased by solutes with coefficients less than zero (e.g., manganese and nickel). E-31 ------- Richards and Thorne (1961) studied the activity of ZnO in slags with various CaO:SiO2 ratios, over the temperature range 1,373 to 1,523 K, based on the assumption that the following slag/metal reaction controlled the equilibrium: (ZnO) + Fe(s) = (FeO) + Zn(g) The parentheses indicate slag components, as usual. Further assuming that the gas phase contained 3 vol% Zn, they calculated that, at 1,473 K, the amount of zinc in the slag could be represented by the expression: 0.022 (wt%FeO) (yFe0) (wt%Zn) = - - fe° where all components of the equation involve the slag phase. For a fixed FeO concentration, the amount of zinc in the slag decreased with increasing temperature and increasing ratios of CaO:SiO2. For example, at 1,473 K, when the CaO:SiO2 ratio was 0.3:1, the slag contained 1.2 wt% Zn and, when the CaO:SiO2 ratio was 1.2:1, the zinc content of the slag had dropped to 0.8 wt%. If one extrapolates these results to 1,873 K, the amount of zinc in the slag would be only about 0.009%. Menon et al. (1990) found that, during the melting of two stainless steel heats, the Zn-65 was about equally distributed between the melt and the baghouse dust. From the available information it appears that, when the scrap metal charge has a reasonably high zinc content, significant amounts of zinc will be volatilized but, when the zinc levels in the charge are low, vaporization will be more difficult. Virtually no zinc should remain in the slag. E.5.28 Zirconium Based on free energy considerations, zirconium would be expected to partition to the slag. Stubbles' information for EAF steel melting supports this hypothesis (Stubbles 1984a). E-32 ------- E.6 INFERRED PARTITIONING No theoretical or experimental evidence exists for the partitioning of several elements that may be contaminants in steel. This section proposes the distribution of these nuclides based on chemical and/or physical behavior. E.6.1 Curium Curium should behave like other elements in the actinide series such as americium and partition to the slag. E.6.2 Promethium Promethium should behave like other rare-earth elements such as europium and samarium and partition to the slag. E.7 SUMMARY In summarizing the distribution of the various potential contaminants that might be introduced into the steel melting process, one must define certain process parameters including: • ratio of mass of steel produced to total mass of scrap charged to furnace (Rj) • ratio of mass of slag to mass of steel produced (R2) • ratio of mass of baghouse dust to mass of steel produced (R3) • fraction of baghouse dust from slag (%S1) • fraction of baghouse dust from steel (%St) The following values were adopted for each of these process parameters: •R12 0.9 Pulliam (1996) stated that Bayou Steel typically produces 0.882 ton of steel billets per ton of scrap charged. When averaged over the total U.S. production, the process efficiency is much higher. According to the U.S. Geological Survey for the year 1994, the amount of recirculating home scrap was 132,300 tons, while 39.5 million tons of EAF steel were produced. Thus, the annual average ratio of home scrap to steel produced was 0.3% ( Fenton 1995). (Throughout this appendix, capacities of metal recycling facilities, and other parameters characterizing the metal refining industries will generally be cited in metric tons [tonnes] or, if English units were cited in the source documents, in short tons. The word "ton" will always mean short ton ] 1 ton = 0.9072 tonne]. ) E-33 ------- •R213 0.13 •R314 15kg/tof steel melted (16.5 to 18 kg per tonne of carbon steel produced in EAF) (A.D. Little 1993) •%S115 33.3 •%St 66.7 The Rj value is based on the following assumptions: • 5% of metal in each heat becomes home scrap, which is returned to the furnace in a later heat • 1.5% of metal is lost to baghouse dust • 2% of metal is lost to slag • 1.5% is unaccounted for Based on these process parameters and the information presented previously, the assumed distribution of the various elements in summarized in Table E-8. Since the amount of baghouse dust contributed by the melt is 5 kg/t, if a potential radioactive contaminant tended to concentrate in the melt, the dust would contain 1% of the activity in the melt. Similarly, since the amount of baghouse dust contributed by the slag is 5 kg/t of metal, and since the mass of the slag is — the mass of the melt, if such a contaminant tends to concentrate in the slag, 5% of the slag activity would be transported to the baghouse. For simplicity, the baghouse efficiency is assumed to be 100% in evaluating partition ratios. Where varying results are presented by different investigators, emphasis was placed on results which represented EAF melting of carbon steel with basic slags. According to R. West of International Mill Services, a major slag marketer, between 0.12 and 0.14 tons of slag are generated per ton of steel produced (West 1996). Since this appears to be a more realistic figure than the 10% cited in Stubbles 1984a, the average of 0.13 was adopted for the present analysis. Additional information on baghouse dust is included in Appendix E-2. Based on the baghouse dust composition reported by SAIC (McKenzie-Carter et al. 1985), adjusted for the ZnO content, and assuming that all the Fe2O3 and one-half the MnO and SiO2 are from the melt, the %S1 is 33%. E-34 ------- Table E-8. Proposed Distribution of Potential Contaminants During Carbon Steelmaking Element Ac Ag Am Ba Bi C Ca Cd Ce Cl Cm Co Cr Cs Cu Eu Fe H I Ir K Mn Mo Na Nb Ni Np P Pa Pb Pm Distribution (%) Melt 99/75 100/27 99 99/40 99 97 10 99 24/65 99 99 9 Slag 95 95 95 95 95 50 95 0/57 0/5 95 2 50 72/32 50 95 95 87 95 95 Baghouse 5 1/25 5 5 100 5 100 5 50 5 1 1/3 100/95 1 5 1 1 50 4/3 1 50 5 1 5 4 5 100 5 Atmosphere 0/73 90 100 Comments Assumed same as Pb Depends on melting practice Some Cl in baghouse dust (McKenzie- Carteretal. 1985) Longest-lived isotope: t,/2 = 27.7 d Longest-lived isotope: t,/2 = 2.58 d Needs further analysis Needs further analysis Needs further analysis Longest-lived isotope: t,/2 = 25.3d E-35 ------- Table E-8 (continued) Element Po Pu Ra Re Rn Ru S Sb Se Sm Sr Tc Th U Y Zn Zr Distribution (%) Melt 99 99 19 99/80 19 99 20/0 Slag 95 95 77 77 95 95 95 95 95 95 Baghouse 100 5 5 1 1 4 1/20 4 5 5 1 5 5 5 80/100 5 Atmosphere 100 Comments Slag % is max. expected. Melt % may be higher. (Maximum t,/2 = 87.2 d.) Conflicting reports on Sb distribution Assumed to behave like S Zn difficult to remove from melt at low concentrations Additional factors which may alter the results presented in Table E-8 are presented below. • In some cases, results are quoted for stainless steels rather than carbon steels. The thermodynamic activity of solutes in the highly alloyed steel melt should be different from that in plain carbon steels and the slag chemistry will be significantly altered. • Perspective on kinetically driven processes may be altered by the scale of the melting operation. • Melt temperatures and holding times in the molten state may be quite different in cited experiments as compared to commercial practice. This can significantly impact conclusions, especially with regard to volatile elements. The mass concentrations of potential contaminants in free-released steel scrap would be quite low. Consequently, some of the partition predictions made here may be overridden by other factors. For example, if evaporation kinetics of volatile elements control the release, small quantities E-36 ------- of zinc may remain in the steel. For strong oxide formers which should partition to the slag, transfer may be impeded due to the high density of many of the actinide and rare- earth oxides. The experimental evidence of this possibility is mixed. For example, EuxOy seems to be removed from the melt during normal EAF melting, but CeO2 may not be completely removed. One investigator reported that the uranium decontamination factor in mild steel increased with increasing contaminant levels (Abe et al. 1985). In addition, the expected partitioning may be altered significantly if the melting practice is changed. Examples presented in this appendix include the removal of niobium from the slag to the melt and movement of tungsten in the opposite direction. The information in Table E-8 does not explicitly consider home scrap or contaminated furnace refractories. Home scrap (i.e., the scrap from the melting process that is recirculated into future furnace charges) should have the same contaminant distribution as the melt from which it was produced. The contamination of furnace refractories was not studied in the present analysis. However, it should be noted that residuals remaining in the furnace from a melt are frequently recovered in the next one to two melts. For example, when melting a low alloy steel containing, say, 1% Cr, the following heat or two will contain more chromium than would be expected if the only source were the furnace charge for the ensuing heats (Stubbles 1996). E-37 ------- REFERENCES Abe, M, T. Uda, and H. Iba. 1985. "A Melt Refining Method for Uranium Contaminated Steels and Copper." In Waste Management '85 3:375-378. Tucson, AZ. A. D. Little, Inc. 1993. "Electric Arc Furnace Dust - 1993 Overview," CMP Report No. 93-1. EPRI Center for Materials Production. Ansara, I, and K. C. Mills. 1984. "Thermochemical Data for Steelmaking." Ironmaking and Steelmaking 11 (2): 67-73. ASM International. 1993. Phase Diagrams of Binary Iron Alloys. Brandes, E. A., and G. B. Brooks, eds. 1992. Smithells Metals Reference Book. Butterworth- Heinemann Ltd. Bronson, A., and G. R. St. Pierre. 1985. "Electric Furnace Slags." Chap. 22 in Electric Furnace Steeling Making 321-335. Iron and Steel Society. Brough, J. R., and W. A. Carter. 1972. "Air Pollution Control of an Electric Arc furnace Steel Making Shop." J. Air Pollution Control Association vol. 22, no. (3). Chen W., et al. 1993. "Reduction Kinetics of Molybdenum in Slag." Steel Research 63 (10): 495-500. Darken, L. S., and R. W. Gurry. 1953. Physical Chemistry of Metals. McGraw-Hill Book Company. Deo, B., and R. Boom. 1993. Fundamentals of Steelmaking Metallurgy. Prentice Hall International. Engh, T. A. 1992. Principles of Metal Refining. Oxford University Press. Fenton, M. D. (U.S. Geological Survey). 1995. Private communication (25 June 1995). Gerding, T. J., et al. 1997. "Salvage of Plutonium- and Americium-Contaminated Metals." In AIChE Symposium Series 75 (191): 118-127. Glassner, A. 1957. "The Thermochemical Properties of Oxides, Fluorides, and Chlorides to 2500°K," ANL-5750. Argonne National Laboratory. E-38 ------- Gomer, C. R., and J. T. Lambley. 1985. "Melting of Contaminated Steel Scrap Arising in the Dismantling of Nuclear Power Plants," Contract No. DED-002-UK, Final Report. British Steel Corporation, for the Commission of the European Communities. Harvey, D. S. 1990. "Research into the Melting/Refining of Contaminated Steel Scrap Arising in the Dismantling of Nuclear Installations," EUR-12605. Commission of the European Communities. Heshmatpour, B., and G. L. Copeland. 1981. "The Effects of Slag Composition and Process Variables on Decontamination of Metallic Wastes by Melt Refining," ORNL/TM-7501. Oak Ridge National Laboratory. Hino, M., et al. 1994. "Evaporation Rate of Zinc in Liquid Iron." ISIJ Int. 34 (6): 491-497. Japan Society for the Promotion of Science (JSPS). 1988. SteelmakingData Sourcebook. Gordon and Breach Science Publishers. Kalcioglu, A. F., and D. C. Lynch. 1991. "Distribution of Antimony Between Carbon-Saturated Iron and Synthetic Slags." Metallurgical Transactions, 136-139. Kellog, H. H. 1966. "Vaporization Chemistry in Extraction Metallurgy." Trans. Met. Soc. AIME 236:602-615. Kreutzner, H. W. 1972. Stahl undRisen 92:716-724. Larsen, M. M., et al. 1985a. "Sizing and Melting Development Activities Using Contaminated Metal at the Waste Experimental Reduction Facility," EGG-2411. EG&G Idaho, Inc. Larsen, M. M., et al. 1985b. "Spiked Melt Tests at the Waste Experimental Reduction Facility," PG-Am-85-005. Idaho National Engineering Laboratory, EG&G Idaho, Inc. McKenzie-Carter, M. A., et al. 1995. "Dose Evaluation of the Disposal of Electric Arc Furnace Dust Contaminated by an Accidental Melting of a Cs-137 Source," Draft Final, SAIC- 95/2467&01. Prepared by Science Applications International Corporation for the U.S. Nuclear Regulatory Commission. Menon, S., G. Hernborg, and L. Andersson. 1990. "Melting of Low-Level Contaminated Steels." In Decommissioning of Nuclear Installations. Elsevier Applied Science. Meraikib, M. 1993. "Manganese Distribution Between a Slag and a Bath of Molten Sponge Iron and Scrap." ISIJInternational^ (3): 352-360. E-39 ------- Murayama, T., and H. Wada. 1984. "Desulfurization and Dephosphorization Reactions of Molten Iron by Soda Ash Treatment." In Proceedings of Second Extractive and Process Metallurgy Fall Meeting, 135-152. The Metallurgical Society, Lake Tahoe, NV. Nakamura, H., and K. Fujiki. 1993. "Radioactive Metal Melting Test at Japan Atomic Energy Research Institute." Nassaralla, C. L., and E. T. Turkdogan. 1993. "Thermodynamic Activity of Antimony at Dilute Solutions in Carbon-Saturated Liquid." Metallurgical Transactions B, 248:963-975. National Slag Association (NSA). 1994. "Steel Slag: A Material of Unusual Ability, Durability and Tenacity," NSA File: 94/pub/steelslag.bro. Ostrovski, O. 1994. "Remelting of Scrap Containing Tungsten and Nickel in the Electric Arc Furnace." Steel Research 65 (10): 429-432. Phelke, R. D. 1973. Unit Processes in Extractive Metallurgy. American Elsevier Publishing Co. Perrot, P., et al. 1992. "Zinc Recycling in Galvanized Sheet." In The Recycling of Metals (Proc. Conf). Dusseldorf-Neuss Germany. Perry, R. H., and D. W. Green. 1984. Perry's Chemical Engineers'Handbook. 6th ed. McGraw-Hill Book Co., Inc. Pflugard, K., C. R. Gomer, and M. Sappok. 1985. "Treatment of Steel Waste Coming From Decomissioning of Nuclear Installations by Melting." In Proceedings of the International Nuclear Reactor Decommissioning Planning Conference., NUREG/CP-0068, 349-371. Bethesda, MD. Philbrook, W. O., and M. B Bever, eds. 1951. Basic Open Hearth Steelmaking. American Institute of Mining and Metallurgical Engineers. Pulliam, A. (Bayou Steel). 1996. Private communication (25 June 1996). Richards, A. W., and D. F. J. Thorne. 1961. "The Activities of Zinc Oxide and Ferrous Oxide in Liquid Silicate Slags." In Physical Chemistry of Process Metallurgy, Part 1:277-291. AIME Interscience, New York. Sappok, M., et al. 1990. "Melting of Radioactive Metal Scrap from Nuclear Installations." In Decommissioning of Nuclear Installations, 482-493. Elsevier Applied Science. E-40 ------- S. Cohen and Associates (SCA). 1995. "Analysis of Potential Recycling of Department of Energy Radioactive Scrap Metal." Prepared for U.S. Environmental Protection Agency, Office of Radiation and Indoor Air, Washington, DC. Schuster, E., et al. 1988. "Laboratory Scale Melt Experiments with 241Am, 55Fe, and 60Co Traced Austenitic Steel Scrap." In Waste Management '88, vol. II, 859-864. Schuster, E., and E. W. Haas. 1990. "Behavior of Difficult to Measure Radionuclides in the Melting of Steel." In Decommissioning of Nuclear Installations. Elsevier Applied Science. Sigworth, G. K., and J. F. Elliott. 1974. "The Thermodynamics of Liquid Dilute Iron Alloys." Metal Science 8: 298-310. Starkey, R. H., et al. 1961. "Health Aspects of the Commercial Melting of Radium Contaminated Ferrous Metal Scrap." Industrial Hygiene Journal 489-493. Stubbles, J. R. 1984a. "Tonnage Maximization of Electric Arc Furnace Steel Production: The Role of Chemistry in Optimizing Electric Furnace Productivity - Part V." Iron and SteelmakingU (6): 50-51. Stubbles, J. R. 1984b. "Tonnage Maximization of Electric Arc Furnace Steel Production: The Role of Chemistry in Optimizing Electric Furnace Productivity - Part VII." Iron and Steelmaking 11 (8): 46-49. Stubbles, J. R. (Manager of Technology, Charter Steel Company). 1996. Private communication (1 July 1996). Wenhua, W., C. Weiqing, and Z. Rongzhang. 1990. "The Kinetics of the Reduction of Niobium Oxide from Slag by Silicon Dissolved in Molten Iron." 10th International Conference on Vacuum Metallurgy, 1:138-149. West, R. (International Mill Services). 1996. Private communication (25 June 1996). Worchester, S. A., et al. 1993. "Decontamination of Metals by Melt Refining/Slagging -An Annotated Bibliography," WINCO-1138. Idaho National Engineering Laboratory. Xiao, Y., and L. Holappa. 1993. "Determination of Activities in Slags Containing Chromium Oxides." ISIJInternational?,?, (1): 66-74. Zhong, X. 1994. "Study of Thermochemical Nature of Antimony in Slag and Molten Iron." Thesis proposal prepared under supervision of Prof. David C. Lynch, Dept. of Materials Science and Engineering, University of Arizona, Tucson AZ. E-41 ------- APPENDIX E-l EXTENDED ABSTRACTS OF SELECTED REFERENCES ------- Chen W. et al. 1993. "Reduction Kinetics of Molybdenum in Slag." Steel Research 63 (10): 495-500. Reduction of molybdenum oxide in slag over an iron-carbon melt is completed in 5 min in 1-kg lab melts. The reaction may be: (Mo03) + 3[C] = [Mo] + 3COgas AF° = 82.35 - 0.2370T [kJ] or a two-step process (MoO3) + 3Fe = [Mo] + 3FeO AF° = -213.6 + 0.0386T[kJ] and (FeO) + [C] = Fe + COgas AF° = 98.65 - 0.0919T [kJ] At 1,440 to 1,500°C the reaction rate is controlled by molybdenum diffusion in slag and, from 1,500 to 1,590°C, the reaction rate is controlled by molybdenum diffusion in the melt. El-1 ------- Gomer, C. R., and J. T. Lambley. 1985. "Melting of Contaminated Steel Scrap Arising in the Dismantling of Nuclear Power Plants," Contract No. DED-002-UK, Final Report. British Steel Corporation, for Commission of the European Communities. This paper discusses the same tests but in somewhat greater detail than Pflugard et al. (1985). The EAF slag is about 5% to 10% of the metal cast weight and involves chiefly additions of carbon, lime and ferrosilicon plus eroded refractories and general oxidation products. Melts were about 2.5 t each. In the arc furnace melt with a CsCl addition, cesium was added with melt charge and, since CsCl is volatile below steelmaking temperature, the CsCl volatilized before any could be incorporated into non-reactive basic slag. In an induction furnace test, CsOHwas added into liquid steel pool with complete cover of relatively cool, quiescent acid slag. In an arc furnace test with CsOH, cesium was added to the molten pool but slag conditions are not described nor is the hold time after addition stated. However, Gomer stated that, although the slag was made as acidic as the furnace liner could withstand, it still did not contain enough silica to fix the cesium as cesium silicate. The limited cesium recovery of only 50% was attributed to cesium condensation on cooler duct walls upstream of sampling point. In an arc furnace test with Cs2SO4, cesium was added as in the previous arc furnace test with CsOH. The higher cesium recovery in the slag is attributed to incorporation of Cs2SO4 into the slag. Larsen, M. M., et al. 1985a. "Sizing and Melting Development Activities Using Contaminated Metal at the Waste Experimental Reduction Facility," EGG-2411. EG&G Idaho, Inc. This report describes melting of contaminated carbon steel from the SPERT III reactor in a 1,500-lb coreless induction furnace at the Waste Experimental Reduction Facility (WERF). Six heats were thoroughly sampled. All showed only Co-60 in feed stock. However, due to concentrating effects, Eu, Cs, and occasionally U were found in the slag, while the baghouse dust contained Co, Cs, Eu, and U, and spark arrestor dust contained Co and Eu. This occurred even though, except for Co-60, all these nuclides were not seen in the feed at the limits of detection. Molten metal samples either contained Co-60 or emitted no detectable radiation. Detectable quantities of Co-60 were seen in slag and baghouse and spark arrester dust. Of 35,900 Ci of Co-60 charged into six melts, 1,361 Ci were recovered in the baghouse and spark arrestor dust (3.8%). El-2 ------- Larsen, M. M., et al. 1985b. "Spiked Melt Tests at the Waste Experimental Reduction Facility," PG-Am-85-005. Idaho National Engineering Laboratory, EG&G Idaho, Inc. Tracer tests were conducted at WERF in a 1,500-lb induction furnace using Type 304L stainless steel. Three heats, weighing 474 to 689 pounds each, were made. All were doped with Co-60, Cs-137 and Sr-85, while Ir-192 was added to only one. Melt temperatures were not specified; slag chemistry was not specified but apparently no slag formers were added16. A small amount of slag "coagulant" was added to aid in slag removal. Tracers were added to the initial furnace charge. The fraction of each radionuclide partitioning to the metal was determined on the basis of melt samples, as listed in Table El-1. Subsequent analysis of the ingots suggested that these analyses were biased low because of the large sample sizes taken from the melts which caused self- shielding. Averaged results from ingot tests (percent of activity in ingot), also listed in Table El-1, are believed to be more reliable. The last column lists the fraction of the charge recovered in the ingot in each test. Table El-1. Distribution of Radionuclides in Tracer Tests at WERF (% Test No. 1 2 3 Co-60 melt 87 73 77 ingot 96 96 97 Sr-85 melt 1.7 2.3 2.3 ingot 1 0 1 Cs-137 melt 1.3 1.8 1.8 ingot 10 8 5 Ir-192 melt — — 57 ingot — — 60 Ingot fraction 93 98.4 95.4 Some problems were encountered with entrained metal in the slag samples. Poor results were obtained on activity measurements of slag and baghouse dust; consequently, no activity balance was calculated. 16 A subsequent publication reported that the composition of the slag was 72% Si02, 13% A12O3, 4.5% Na2O, 5.0% K2O and 0.7% CaO (Worchester et al. 1993). El-3 ------- Menon, S., G. Hernborg, and L. Andersson. 1990. "Melting of Low-Level Contaminated Steels." In Decommissioning of Nuclear Installations. Elsevier Applied Science. Studsvik AB in Sweden has a 3-t induction melting furnace where low-level radioactive scrap is remelted. Based on the melting of 33.611 of carbon steel, the weight of ingots was 32.27 t, the weight of slag was 1.321 and the weight of dust was 0.019 t. No Cs-137 was measured in the ingots and the activity levels in the slag were also below the measurement threshold for the detection equipment. Dust contained the following nuclides: •Co-60 1,300 Bq/kg •Zn-65 14,400 Bq/kg • Cs-137 21,800 Bq/kg Menon et al. also reported on the results of two stainless steel melts weighing a total of 5,409 kg. The weight of slag in melt 92 was 1.1% of the total and in melt 93 it was 0.5%. The weight of dust from the combined melts was 2.49 kg. Activity measurements are listed in Table El-2. Table El-2. Specific Activities of Ingots and Slags (Bq/kg) Melt No. 92 93 Material ingot slag ingot slag Baghouse dust Co-58/Co-60 1350 720 3440 207 264/31,200 Mn-54 8.2 73 10 146 Cs-134/Cs-137 2320 1493 1,125/134,650 Ag-llOm 54 30 37,450 Sb-125 29 50 670 Zn-65 34 52,250 El-4 ------- Meraikib, M. 1993. "Manganese Distribution Between a Slag and a Bath of Molten Sponge Iron and Scrap." ISIJInternational?,?, (3): 352-360. The manganese distribution ratio is given by the expression: - CMh) ~ [Mn] ' 27005 _ .„ - 7'2324 for a temperature range of 1,550 to 1,670°C. This equation is based on 80 metal samples from melts in a 70-ton EAF, and reflects Meraikib's finding a limited influence of slag basicity on the manganese distribution ratio. A different expression, explicitly including the influence of basicity, was presented in Section E.5.14. Extensive thermodynamic calculations are included. El-5 ------- Nakamura, H., and K. Fujiki. 1993. "Radioactive Metal Melting Test at Japan Atomic Energy Research Institute." Air melting was accomplished in a high frequency (1,000 Hz) induction furnace of 500 kg capacity. Researchers studied the effects of melting temperature, slag basicity and type of steel (ASTM-A335 and SUS 304) on partitioning using radioactive tracers: Mn-54, Co-60, Sr-85, Zn-65 and Cs-137. The slag basicity (CaO/SiO2) was 1 for A335 and 3 for SUS 304. Five radioactive tracer heats (three ASTM-A335 and two SUS 304) and six JPDR decommissioning heats were produced. The average material balance was 99.5%, with the maximum difference being 3%. Material distribution was: 95% ingot, 2-3% slag, 0.1% dust, 1-2% other (metal on tundish and metal splash). The melt temperature was 1,873 K. Results from one of the three A335 tracer tests are as follows: • Mn-54: recovery 98%, about 7% of which was in slag, balance in ingot (approximate Mn content of other three ingots was 90%) • Co-60: 99.5% recovery, all in ingot • Zn-65: 90.7% recovery, about 14% of which was in exhaust gas, 1% in slag and balance in ingot • Sr-85: 72.7% recovery, 100% in slag • Cs-137: 77% recovery, 50% of which was in slag and 50% in exhaust gas The other four tracer tests showed similar tendencies. The melt was held at temperature for about 20 minutes after tracers were added before casting the ingot. Tracers were not present in initial melt charge, but rather were added after melting was completed and the desired temperature of 1,873 K was reached. Exhaust gas analyses were based on sampling about 0.04% of total exhausted volume. El-6 ------- Ostrovski, O. 1994. "Remelting of Scrap Containing Tungsten and Nickel in the Electric Arc Furnace." Steel Research 65 (10): 429-432. This paper discusses partitioning of tungsten between slag and melt during melting of tungsten-bearing steel scrap in a 25-t EAF with slags of varying basicity. Melting under strongly oxidizing conditions (30 min. oxygen blow) and high CaO/SiO2 ratio resulted in 94% of the tungsten in slag, 4% in metal and 2% lost. Thermodynamic equations for calculating the partition ratio are provided. Pflugard, K., C. R. Gomer, and M. Sappok. 1985. "Treatment of Steel Waste Coming From Decomissioning of Nuclear Installations by Melting." In Proceedings of the International Nuclear Reactor Decommissioning Planning Conference, NUREG/CP-0068, 349-371. Bethesda, MD. Sappok described nine melts totaling 24 t (plus starting blocks, i.e., furnace heel) in 10-t and 20-t induction furnaces. Mass balance: 28,000 kg steel, 800 kg slag, 20 kg furnace lining, and 64 kg cyclone and baghouse dust. Co-60 and Cs-137 distributions were: Co-60: 97% in steel, 1.5% in slag, 1.5% in cyclone and baghouse Cs-137: 90% in slag, 1% in furnace lining, (5% in baghouse tubes and dust). Activities accounted for: Co-60-96%; Cs-137-73%. No discussion of slagging practices or melting practices and temperatures was included. Gomer used a 500 kg high frequency induction furnace, a 5-t EAF and a 3-t EOF (no results reported). Non-quantitative tests from two 5-t arc furnace melts showed that all the Co-60 was reported in the melt; quantities in slag and fume were below detection limits. Traces of Am-241 were found in slag when melting contaminated heat exchanger tubing in the arc furnace. The results of three quantitative tests of cesium in 5-t EAF's and one in a 500 kg induction furnace are listed in Table E-6 of the present report. Gomer notes that cesium stays in slag in an induction furnace and can be made to stay largely in slag in an arc furnace but conditions "may not be fully practical in production furnaces." No information on melting and slagging practice is included. El-7 ------- Sappok, M., et al. 1990. "Melting of Radioactive Metal Scrap from Nuclear Installations." In Decommissioning of Nuclear Installations, 482-493. Elsevier Applied Science. Melting to date has totaled 2,000 tons of steel (steel presumed from Pflugard et al., but not so stated in report) in a 20-ton induction furnace. (A new dedicated facility with a 3.2-ton medium frequency induction furnace had recently been completed but no radioactive scrap had yet been melted in the new equipment). When melting zinc-plated metal, zinc is "found in the filter dust." Typical mass balance: 98.6% metal, 1.2% slag and 0.2% filter dust. For the melting period May 17, 1985: Ce-144 all in slag, Zn-65 all in offgas, Mn-54 distributed between slag and offgas, Cs-134/137 distributed between slag and offgas, Co-60 mostly in melt but some in slag and some in offgas (Co-60 is only the radionuclide detected in the melt). For the melting period September 27-28, 1985: Mn-54 distributed between slag and offgas; Zn-65 all in offgas; Eu-154 all in slag; Ag-110m distributed among metal, slag and offgas; Cs-134/137 distributed between slag and offgas; Co-60 distributed among melt, slag and offgas, but mostly in the melt. For the melting period January 1, 1986 - March 14, 1986 (200 t): Cs-134/137 distributed between slag and offgas; Mn-54 distributed between slag and offgas; Zn-65 distributed among slag, metal and offgas; Ag-110m distributed among slag, metal and offgas, but mostly in metal; Co-60 distributed among slag, metal and offgas, but retained mostly in metal. No discussion of slagging or melting practice was included. El-8 ------- Schuster, E., and E. W. Haas. 1990. "Behavior of Difficult to Measure Radionuclides in the Melting of Steel." In Decommissioning of Nuclear Installations. Elsevier Applied Science. Laboratory melts were made using a Nernst-Tammann high-temperature furnace with temperatures to 1,700°C and a 3- to 5-kg melt size. Melt additions included: (1) electro- deposited Co-60, Fe-55 and Am-241 on steel disks, (2) carbonate or hydroxide precipitates or elemental carbon on SiO2 filters, (3) direct insertion of uranium and UO2. The melts were allowed to solidify in the carborundum tube crucible. About 60% to 80% of the slag was recovered when melting St37-2 steel under Ar + 10% H2. The results are presented in Table El-3. Table El-3. Distribution of Radionuclides Following Laboratory Melts Sample Location Ingot Slag Aerosol Filter Percentage of Nuclide in Each Medium Co-60 108 0.2 0.2 Fe-55 70 n.d. n.d. Ni-63 ~ 82 0.04 0.06 C-14 91 0.4 < 0.001 In a test for strontium distribution where slag-forming oxides CaO, SiO2 and A12O3 were added, the Sr-85 distribution was: surface layer of ingot—ca. 80%, slag—6.3%, ingot—0.5%, aerosol filter— 0.02%. In a test with Am-241, the isotope distribution was: ingot—1%, slag—110% and aerosol filter—0.05%. In tests with UO2, when slag formers were added, the uranium concentration in the ingot was reduced from 330 to 5 ppm. Starkey, R. H., et al. 1961. "Health Aspects of the Commercial Melting of Radium Contaminated Ferrous Metal Scrap." Industrial Hygiene Journal 489-493. Melting of 40 tons of radium-contaminated steel scrap blended with 20 tons of uranium- contaminated steel scrap in an EAF is discussed. Based on eight heats, the average concentration of radium in steel ingots was <9 x 10"11 g of Ra per g of steel, and the radium content of slag was 1.47 x 10"9 g Ra per g of slag. No information on melting and slagging conditions was provided. El-9 ------- Stubbles, J. R. 1984a. "Tonnage Maximization of Electric Arc Furnace Steel Production: The Role of Chemistry in Optimizing Electric Furnace Productivity - Part V." Iron and SteelmakingU (6): 50-51. Stubbles notes that recovery (from scrap) of Cb, B, Ti, Zr, V, Al, and Si in steel is zero and recovery of Mo, Ni, Sn, and Cu is 100%. Pb, Zn, and Sb are volatilized. Cr and Mn are distributed between slag and metal based on the degree of slag oxidation (the "FeO" level). Chromium recovery ranges from about 30% to 50% and manganese recovery from about 10% to 25%. No supporting information is provided for these recovery values. According to Stubbles, lead from babbitts, batteries, etc. melts and quickly sinks to the furnace bottom, often penetrating the refractory lining. However, when leaded scrap is added to liquid steel, the lead will go into solution and boil off like zinc, exiting with the fume. Stubbles, J. R. 1984b. "Tonnage Maximization of Electric Arc Furnace Steel Production: The Role of Chemistry in Optimizing Electric Furnace Productivity - Part VII." Iron and Steelmaking 11 (8): 46-49. Stubbles cites the following charge to produce one ton of liquid steel: metals 2,100 Ib flux 40 Ib gunning material (high MgO) 10 Ib charge carbon 10 Ib In this example, the initial slag volume is 100 Ib per ton (see Note 12 on p. E-33). Most input sulfur remains in metal and is extremely difficult to transfer to slag. The theoretical sulfur distribution -^ rarely exceeds 8 in EAF's. Working down sulfur during melting requires [S] constant removal of high basicity slag plus agitation. One reason for adding excess carbon above the desired final level is to use decarb oxygen from a lance to promote slag/metal reactions and help boil out hydrogen. Hydrogen levels on the order of 1 ppm can be obtained after a 15-minute carbon boil where the rate of carbon removal is 1%/hr. If the carbon removal rate is 0.1%/hr, the comparable hydrogen level is about 5 ppm (based on an initial level of 9 ppm). El-10 ------- APPENDIX E-2 COMPOSITION OF BAGHOUSE DUST ------- COMPOSITION OF BAGHOUSE DUST Various studies have reported measurements of the composition of baghouse dust. Results of measurements reviewed in this study are reported here. Babcock and Wilcox Company (Kaercher and Sensenbough 1974) provided the baghouse dust composition at its No. 3 EAF melt shop at Koppel, Pa. The melt shop included one 50-ton, one 75-ton and three 100-ton furnaces used for the production of carbon, alloy and stainless steels. The dust composition (in wt%) was: Fe2O3 52.7 CaO 13.6 A12O3 0.9 SiO2 0.9 MgO 12.6 Mn2O3 0.6 ZnO 6.3 MO 0.1 Cr2O3 0.6 CuO 0.1 Loss on ignition at 1100°C 6.8 Balance 4.6 The average dust collection was 12 Ib per ton of steel melted. More recently, dust collection has been increasing, reaching a level of 26 Ib per ton of carbon steel melting capacity in 1985 and 30 Ib per ton of carbon steel melting capacity in 1992 (A. D. Little 1993). Arthur D. Little (ADL) (1993) prepared a survey on EAF dust generation for the Electric Power Research Institute in 1993 based on 52 shops which melted carbon steel. ADL estimated that about 600,000 tons of dust were generated in 1992 from U.S. carbon steel operations. The dust composition (in wt%) was: E2-1 ------- Fe 28.5 Zn 19. Cd <0.01 Pb 2.1 CrO 0.39 CaO + MgO 10.7 The high levels of zinc in the dust are the result of large amounts of galvanized steel in the furnace charge. According to ADL, the disposition of the baghouse dust in 1992 was: • Disposal to landfill 1.2% • Shipped to fertilizer 2.3% • Shipped to zinc recovery 86.5% • Miscellaneous, delisted 0.1% Lehigh University (1982) conducted a study on EAF dust for the Department of Commerce in 1982. Dust composition from stainless steel and carbon steel melts is shown in Table E2-1. Table E2-1. Composition of Baghouse Dust (wt%) Component Stainless Steel Dust Carbon Steel Dust Fe Zn Cd Pb Cr CaO 31.7 1.0 0.16 1.1 10.2 3.1 35.1 15.4 0.028 1.5 0.38 4.8 McKenzie-Carter et al. (1995) described the composition of EAF dust taken from an earlier work by Brough and Carter (1972). The dust composition (in wt%) as quoted by Brough and Carter and interpreted by McKenzie-Carter et al. is: E2-2 ------- Fe2O3 52.5 ZnO 16.3 CaO 14.4 MnO 4.4 SiO2 2.6 MgO 1.9 Na2O 1.5 C12 1.2 Other 5.2 Based on the original source, C12 should be Cl and 4.4% of "Other" is ignition loss. The dust was a by-product of melting low alloy carbon steels. REFERENCES A. D. Little, Inc. 1993. "Electric Arc Furnace Dust - 1993 Overview," CMP Report No. 93-1. EPRI Center for Materials Production. Brough, J. R., and W. A. Carter. 1972. "Air Pollution Control of an Electric Arc furnace Steel Making Shop." J. Air Pollution Control Association, vol. 22, no. 3. Kaercher, L. T., and J. D. Sensenbough. 1974. "Air Pollution Control for an Electric Furnace Melt Shop." Iron and Steel Engineer 51 (5): 47-51. Lehigh University. 1982. "Characterization, Recovery, and Recycling of Electric Arc Furnace Dust." Sponsored by U.S. Department of Commerce. McKenzie-Carter, M. A., et al. 1995. "Dose Evaluation of the Disposal of Electric Arc Furnace Dust Contaminated by an Accidental Melting of a Cs-137 Source," Draft Final, SAIC- 95/2467&01. Prepared by Science Applications International Corporation for the U.S. Nuclear Regulatory Commission. E2-3 ------- APPENDIX F DISTRIBUTION OF CONTAMINANTS DURING MELTING OF CAST IRON ------- Contents page F.I Background F-l F.2 Material Balance F-5 F.2.1 Cupola Furnaces F-5 F.2.2 Electric Arc Furnaces F-6 F.2.3 Chemistry Adjustments F-6 F.3 Partitioning Based on Reduction of FeO in Slag F-7 F.4 Adjustments to Henry's Law for Dilute Solutions F-7 F.5 Observed Partitioning During Metal Melting F-9 F.5.1 General Observations F-9 F.5.2 Antimony F-ll F.5.3 Carbon F-12 F.5.4 Cerium F-12 F.5.5 Cesium F-13 F.5.6 Iron F-13 F.5.7 Lead F-13 F.5.8 Manganese F-13 F.5.9 Niobium F-14 F.5.10 Zinc F-15 F.6 Partitioning Summary F-l7 F.6.1 Elements Which Partition to the Melt F-17 F.6.2 Elements Which Partition to Slag F-18 References F-l9 ------- Tables page F-l. Chemical Composition of Ferrous Castings F-3 F-2. Amounts of Byproducts from Various Foundries F-6 F-3. Standard Free Energy of Reaction of Various Contaminants with FeO at 1,573 K F-8 F-4. Partition Ratios of Two Elements at Typical Iron- and Steel-Making Temperatures F-8 F-5. Partition Ratios at 1,573 K for Various Elements Dissolved in Iron and Slag F-9 F-6. Distribution of Foundries in Bureau of Mines Tramp Element Study F-10 F-7. Lead Levels at Two Different Types of Foundries F-10 F-8. Average Concentrations of Tramp Elements in Cast Iron F-ll F-9. Distribution of Antimony Between Slag and Metal F-ll F-10. Partition Ratios of Manganese at Different Partial Pressures of CO F-14 F-ll. Proposed Partitioning of Metals Which Remain in the Melt F-18 Figure F-l. Flow Diagram of a Typical Cast Iron Foundry F-2 IV ------- DISTRIBUTION OF CONTAMINANTS DURING MELTING OF CAST IRON This appendix discusses the expected partitioning of contaminants during the production of cast iron. The approach taken here is to use the information developed for partitioning during the melting of carbon steel in electric arc furnaces (EAFs) presented in Appendix E, and by analogy predict the expected behavior of selected trace elements during the production of cast iron. To the extent possible, the deductive process takes into account differences in melting and slagging practice. This discussion should be viewed as a supplement to the information developed in Appendix E. Many of the same references are used as information sources and the detailed thermodynamic discussion is not repeated here. In order to assess radiation exposures to products made of potentially contaminated cast iron, it is necessary to estimate the partitioning to cast iron of the elements listed in Table 6-3. The present discussion of partitioning during the production of cast iron therefore includes these elements. F.I BACKGROUND Cast iron is an alloy of iron and carbon (ca. 2 to 4.4 wt%) which also typically contains silicon, manganese, sulfur, and phosphorous. The high carbon content of the alloy results in a hard, brittle product which is not amenable to metalworking (as is steel); hence the alloy is cast into the desired end-use form. As noted by the United States Steel Corporation, now USX, (U.S. Steel 1951): Castings are of innumerable kinds and uses, roughly grouped as chilled-iron castings, gray- iron castings, alloyed-iron casting, and malleable castings. In general, castings are made by mixing and melting together different grades of pig iron; different grades of pig iron and foundry scrap; different grades of pig iron, foundry scrap, and steel scrap; different grades of pig iron, foundry scrap, steel scrap and ferroalloys, and other metals. Representative chemical compositions of cast iron are presented in Table F-l. Cast iron is usually melted in a cupola furnace, an EAF, an electric induction furnace, or an air (reverberatory) furnace. A flow diagram for a typical iron foundry is shown in Figure F-l. The cupola is similar to a small blast furnace where the iron ore in the charge is replaced by pig iron and steel scrap. As described in U.S. Steel 1951: F-l ------- FURNACI CHARGE PREPARATION MEITING AND CASTING to CUPOLA EAF INDUCTION REVERBERATORY GATE AND RISER KNOCKOFF CLEANING AND FINUHINO CORI AND MOLD PREPARATION •PATTERNS Figure F-l. Flow Diagram of a Typical Cast Iron Foundry (from U.S. EPA 1995) ------- The charge is composed of coke, steel scrap, and pig iron in alternate layers of metal and coke. Sufficient limestone is added to flux the ash from the coke and form the slag. The ratio of coke to metallics varies depending on the melting point of the metallic charge. Ordinarily, the coke will be about 8 to 10% of the weight of the metallic charge. It is kept as low as possible for the sake of economy and to exclude sulfur and some phosphorus absorption by the metal. During melting, the coke burns as air is introduced at a 10 to 20 ounce (-0.4 - 0.8 kPa) pressure through the furnace tuyeres. During melting some of the manganese combines with the sulfur forming MnS which goes into the slag. Some manganese and silicon are oxidized by the air blast; the loss is proportional to the amount initially present. Carbon may be increased or reduced depending on the initial amount present in the metallic charge. It may be increased by absorption from the coke or oxidized by the blast. Phosphorus is little affected but sulfur is absorbed from the coke. Prior to casting, the slag is removed from the slag-off hole which is located just below the tuyeres. The molten metal is then tapped through a hole located at the bottom level of the furnace. The depth between these two tapping holes and the inside diameter of the furnace governs the capacity of the cupola (U.S. Steel 1951). Table F-l. Chemical Composition of Ferrous Castings (wt%) Element C Mn P Si S Gray Iron 2.0-4.0 0.40- 1.0 0.05- 1.0 1.0-3.0 0.05-0.25 Malleable Iron (as white iron) 1.8-3.6 0.25-0.80 0.06-0.18 0.5- 1.9 0.06 - 0.20 Ductile Iron 3.0-4.0 0.5-0.8 <0.15 1.4-2.0 <0.12 Steel Scrap3 0.18-0.23 0.60-0.90 < 0.40 — < 0.05 Source: U.S. EPA 1995 Nominal composition of a low carbon steel (e.g., SAE 1020) The melting temperatures used in producing cast irons are lower than those used in steel making. The melting point of pure iron is 1,538°C (1,711 K), while steel making temperatures are typically about 1,600°C (1,873 K). Furthermore, carbon depresses the melting point of iron: the melting point of an iron alloy containing 3.56% C and 2.40% Si is 1,250°C (1,523 K), while one containing 4.40% C and 0.6% Si has a melting point of 1,088°C (1,361 K) (U.S. Steel 1951). Fluxing agents added to the furnace charge to promote slag formation include carbonates (e.g., limestone and dolomite), fluorides (e.g., fluorspar), and carbides (e.g., calcium carbide) (U.S. F-3 ------- EPA 1995). Obviously, the furnace environment during the production of cast iron is more highly reducing than that in typical steel melting. Emissions from the cast iron melting furnaces include particulate matter, CO, SO2, and small quantities of chlorides and fluorides. These emissions are from incomplete combustion of carbon additives, oxidation of sulfur in coke (for cupola melting), flux additions, and dirt and scale in the scrap charge (U.S. EPA 1995). Melting of ductile iron requires the addition of inoculants such as magnesium in the final stages of melting. The magnesium addition to the molten bath results in a violent reaction and the production of MgO particulates and metallic fumes. Most of these emissions are captured by the emission control system and routed to the baghouse, where the fumes are cooled and filtered. Cupolas are also equipped with an afterburner in the furnace stack to oxidize the carbon monoxide and burn any organics. In 1998, U.S. shipments of iron and steel castings were (Fenton 1999): • Ductile iron castings 4,070,000 t • Gray iron castings 5,460,000 t • Malleable iron castings 292,0001 • Steel castings 1,200,0001 • Steel investment castings 83,000 t • Total 11,100,000 t Scrap consumption by manufacturers of steel castings and by iron foundries and miscellaneous users in that year is summarized below (Fenton 1998 ): • Electric arc furnace 7,600,000 t • Cupola furnace 7,500,000t • Air furnaces and other 3,000 t • Total 15,100,000 t Of this total, 5,800,000 t was home scrap. In addition, 1,200,000 metric tons (t) of pig iron and 12,000 t of direct-reduced iron were consumed by the iron and steel foundries. The total metal consumption in 1998 was F-4 ------- 16,300,000 t, which is about 47% greater than cast iron and steel shipments. This difference may be due to generation of home scrap. From a recycling perspective, a significant observation is that cast iron contains more than 90% scrap metal. In 1989, about half of all iron castings were used by automotive and truck manufacturing companies and half of all ductile iron castings were used in pressure pipe and fittings (U.S. EPA 1995). F.2 MATERIAL BALANCE Using the results of several studies, EPA (1995) has compiled emission factors for uncontrolled emissions from two types of gray iron foundries: • Cupola furnace 13.8 Ib/ton1 metal • Electric arc furnace 12.0 Ib/ton metal F.2.1 Cupola Furnaces Based on a 1980 EPA-sponsored environmental assessment of the iron casting industry, Baldwin (1980) reported that a typical cupola producing a medium-strength cast iron from a cold charge would utilize the following materials (as a percentage of iron input): • Scrap steel 48% • Foundry returns (i.e., foundry home scrap) 52% • Ferrosilicon 1.1% • Ferromanganese 0.2% • Coke 14% • Limestone 3% • Melting loss 2% Throughout this appendix, capacities of metal recycling facilities, and other parameters characterizing the metal refining industries will generally be cited in metric tons (tonnes) or, if English units were cited in the source documents, in short tons. The word "ton" will always mean short ton (1 ton = 0.9072 tonne). When practicable, the metric equivalent will also be listed. F-5 ------- Baldwin also documented the quantities of material produced for three foundries: a malleable iron foundry using a induction furnace, a ductile iron foundry using a cupola, and a gray and ductile iron foundry using a cupola for primary melting which duplexes into induction furnaces. The amounts of byproducts are listed in Table F-2. Table F-2. Amounts of Byproducts from Various Foundries Byproduct Slag Dust Collector Discharge Amount Generated (Ib per ton of metal melted) Malleable Iron 34.5 7.19 Ductile Iron 173 Gray and Ductile Iron 130 78.6 F.2.2 Electric Arc Furnaces According to a study conducted for EPA, a typical charge for an electric arc furnace (EAF) includes (Jeffery 1986): • 50% 60% scrap iron • 37% 45% scrap steel • 0.5% 1.1% silicon • 1.3% 1.7% carbon raisers2 Arc furnaces for cast iron melting range from 500-pound to 65-ton capacity, 25 tons being a common size (Baldwin 1980). According to Jeffery (1986), 94% to 98% of the EAF charge is recovered as iron. F.2.3 Chemistry Adjustments As noted in Section F.2.1 and F.2.2, the furnace charge typically contains about 45% steel scrap. If this scrap were similar to that listed in the last column of Table F-l, then, to achieve the cast iron chemistries indicated in that table, it would be necessary to add carbon, phosphorous, sulfur, silicon, and possibly manganese to the furnace charge. Carbon raisers are additives introduced into the bath to increase the carbon content of the cast iron, if required. F-6 ------- Production of ductile iron requires making additions to the melt which alter the shape of the graphite particles in the cast iron from flakes to a spheroidal form. Typically, the melt is inoculated with magnesium just before pouring to produce the ductile iron. Much of the magnesium boils off in the process. Sometimes barium, calcium, cerium, neodymium, praseodymium, strontium, and zirconium are also added as inoculants (Baldwin 1980). To reduce the costs of adding magnesium in larger ductile iron production operations, the melt is desulfurized before magnesium is added. This is frequently done by adding CaC2 (Baldwin 1980). F.3 PARTITIONING BASED ON REDUCTION OF FeO IN SLAG As discussed in Section E.4 of Appendix E, an indication of contaminant partitioning between the melt and the slag can be obtained by calculating the free energy change for the reaction (F-l) where M is the pure component rather than the solute dissolved in the melt and FeO and MxOy are slag components. The standard free energies of reaction of various contaminants with FeO at 1,873 K, a typical temperature for the production of carbon steel in an EAF, were presented in Table E-2. Recalculation of these values for a temperature of 1,573 K, which is typical for cast iron production, indicates no substantive changes from the previous conclusions regarding which elements are expected to concentrate in the slag and which are expected to concentrate in the melt. The assumed 300 K temperature difference between steel melting and cast iron melting produces small changes in the free energies based on Equation F-l, but no significant shifts in the expected equilibria. The free energies of reaction at 1,573 K are listed in Table F-3. F.4 ADJUSTMENTS TO HENRY'S LAW FOR DILUTE SOLUTIONS Partition ratios presented in Table E-l for carbon steel were also recalculated for a furnace temperature of 1,573 K. While slight changes in partitioning ratios were obtained at the lower temperature, no significant shifts in equilibria resulted. An example is the comparable partition ratios for cobalt and uranium, which are shown in Table F-4. Calculations of partition ratios at 1,573 K are summarized in Table F-5. Values of y° were calculated using temperature-dependent values of the free energy change for transference of the F-7 ------- pure substance to a dilute solution in liquid iron. All values were obtained from Sigworth and Elliot (1974) except cerium, which was taken from JSPS 1988. Table F-3. Standard Free Energy of Reaction of Various Contaminants with FeO at 1,573 K Element Acm Amm Barn Cs(1) Npm Pam Pum Ra(2, Ru(s) Sbto Sr(g) T<™ Thrs, Ym Zn^ Oxide Ac2O3 Am2O3 BaO Cs2O Np02 PaO2 Pu203 RaO RuO4 Sb2O3 SrO TcO2 ThO2 Y203 ZnO AF° (kcal) -121 -105 -59.6 -104 -100 -89.1 -55.0 -65.8 -147 -104 Comments Ac should partition to slag Am should partition to slag Ba should partition to slag Cs2O unstable at 1,573 K, Cs should vaporize from melt, some Cs may react with slag components Np should partition to slag Pa should partition to slag Pu should partition to slag Ra should partition to slag Ru should remain in melt Sb will not react with FeO, some may vaporize from melt Sr should partition to slag, but low boiling point could cause some vaporization Tc will not react with FeO, should remain in melt Th should partition to slag Y should partition to slag Zn will not react with FeO, Zn should vaporize from melt Table F-4. Partition Ratios of Two Elements at Typical Iron- and Steel-Making Temperatures Element Co U Partition Ratio (NMO/wt%M) 1,573 K l.Oe-4 1.4e+8 1,873 K 4.8e-5 8.9e+7 F-8 ------- Table F-5. Partition Ratios at 1,573 K for Various Elements Dissolved in Iron and Slag M Agm Alm Cara, Ce(1) Com Crw Cum Mnrn Mo(s, Nbw Nim Pbffi Sim Snrn Ti^ UjD v(s) w^ Zr™ Oxide Ag20 A1203 CaO Ce2O3 CoO Cr203 Cu2O MnO MoO3 Nb2O5 MO PbO SiO2 SnO2 TiO2 UO2 V205 WO3 ZrO2 v° y M 546 0.013 1330 0.26 1.08 1.45 12.9 1.36 2.60 1.79 0.51 11900 2.7e-4 3.44 0.035 0.014 0.078 1.73 0.029 AT<° Hr f,MO (kcal/mole)a +16.5 -280 -118 -302 -25.0 -111 -14.0 -64.3 -95.3 -298 -25.1 -17.8 -143 -61.7 -159 -193 -228 -110 -191 Partition Ratio (NMO/wt%M) 1.06e-03b'c 2.63e+05b 1.15e+10 1.79e+07b l.OOe-04 1.86e-03b 2.56e-03b 5.24e+00 3.49e-06 1.22e+05b 4.98e-05 4.56e-02 4.00e+01 3.70e-05 2.22e+05 1.44e+08 9.93e+00b 6.56e-05 4.52e+08 a AF°fFe0 = -38.1 kcal/mole b PR = N'/2/wt% M c Ag will not react with FeO, Ag2O unstable at 1,573K F. 5 OB SERVED PARTITIONING DURING METAL MELTING F.5.1 General Observations Because of concerns that tramp elements might be accumulating in cast irons from contaminants in steel scrap and affecting casting behavior, the U.S. Bureau of Mines conducted an extensive study over a period of more than five years to evaluate the impurities in cast iron (Natziger et al. 1990). While this study does not specifically address partitioning, the results can provide confirmation of inferred partitioning. Samples were obtained from 28 ductile iron foundries and F-9 ------- 52 gray iron foundries at various times over the course of the study. The distribution of foundries by geographical location, furnace type and product is shown in Table F-6. Table F-6. Distribution of Foundries in Bureau of Mines Tramp Element Study Zone Northeast Great Lakes Southeast Upper Midwest West Ductile Iron Furnace Type Cupola 1 5 1 4 1 Electric 0 0 1 1 0 Induction 2 2 3 O 4 Size3 A 1 1 3 0 5 B 1 2 1 8 0 C 1 4 1 0 0 Gray Iron Furnace Type Cupola 6 12 4 11 3 Electric 0 0 0 1 1 Induction 2 2 3 4 O Size3 A 3 4 3 0 5 B 5 7 2 12 1 C 0 3 2 4 1 Source: Natziger et al. 1990 a A: < 1,000 tons per month; B: 1,000 to 8,000 tons per month; C: >8,000 tons per month With limited exceptions, cerium, niobium, lead, and antimony were not found at the limits of detection (wt%) listed below for the 23 calendar quarters over which sampling was conducted: Ce .... Nb .... Pb .... Sb .... 0.02 - 0.1 0.01 - 0.05 0.005 - 0.2 0.02 - 0.1 Lead levels above the lower detection limit were observed in four quarters, as shown in Table F-7. Table F-7. Lead Levels at Two Different Types of Foundries Quarter 1 2 O 20 Pb Above Detection Limits (wt%) Ductile Iron 0.005-0.007 < 0.005-0.008 Gray Iron < 0.005-0.007 < 0.005-0.010 < 0.005-0.006 < 0.005-0.007 Source: Natziger et al. 1990 F-10 ------- Average analyses for other elements of interest are included in Table F-8. Table F-8. Average Concentrations of Tramp Elements in Cast Iron (wt%) Zone Northeast Great Lakes Southeast Upper Midwest West Ductile Iron Co 0.008 0.007 0.009 0.008 0.012 Mn 0.378 0.405 0.453 0.409 0.415 Mo 0.020 0.022 0.017 0.024 0.025 Ni 0.067 0.117 0.171 0.257 0.186 Zn 0.003 0.003 0.004 0.002 0.005 Gray Iron Co 0.009 0.010 0.010 0.009 0.009 Mn 0.726 0.703 0.675 0.701 0.670 Mo 0.025 0.051 0.030 0.040 0.040 Ni 0.073 0.192 0.142 0.107 0.086 Zn 0.002 0.002 0.003 0.002 0.002 Source: Natziger et al. 1990 F.5.2 Antimony Thermodynamic calculations based on Equation F-l indicate that antimony will not partition to the slag. Experimental work by Kalcioglu and Lynch (1991) showed that when antimony is added to carbon-saturated iron at 1,723 K and allowed to react with an acidic slag (basicity ratio = 0.666), the resulting partition ratios were those listed in Table F-9. Table F-9. Distribution of Antimony Between Slag and Metal [wt%Sb]a 0.45 0.87 1.03 1.06 L b ^Sb 0.067 0.022 0.020 0.018 [wt%Sb] = concentration in metal Lsb = (wt%Sb)/[wt%Sb] (wt%Sb) = concentration in slag Based on these values for Lsb and an assumed slag-to-metal mass ratio of 0.05, the quantities of antimony in the slag are insignificant (i.e., <1%). Antimony recoveries ranged from 47% to 71% for these four tests, the losses being presumably due to vaporization. Nassaralla and Turkdogan (1993) cite the following equation for the activity of antimony in carbon-saturated iron: F-ll ------- logY°= - sb This yields a value for y° of 6.2 at 1,573 K, which, when combined into the Henry's Law (N )* relationship, indicates that the partition ratio, —^?—, is 2.6 x 10"5, supporting the view that [wt% Sb] antimony partitions strongly to the melt. Although, as noted in Section F.5.1, no antimony was found in cast iron samples at the lower limit of detection (0.02 - 0.1 wt%), this does not necessarily vitiate the thermodynamic partitioning argument. Antimony may not be present in the feed materials at the detection limit. Although some antimony may vaporize from the melt, insufficient evidence is available to quantify this possibility. To avoid possibly underestimating exposures to cast iron products potentially contaminated with antimony, antimony is assumed to remain in the melt. F.5.3 Carbon As was noted in Sections F.2.1 - F.2.3, carbon is added to the furnace charge to achieve the levels desired in the finished product (e.g. 1.8% to 4.0% C). During the melting process, some of the carbon in the scrap steel may be oxidized and removed from the system as CO; however, there is a net addition of carbon to the melt, rather than a net removal. Since it is impossible to predict how much carbon is removed from the scrap steel and later replaced with carbon from other charge materials, it is conservatively assumed that all the carbon in the scrap remains in the cast iron. F.5.4 Cerium Cerium is sometimes used as an inoculant in ductile irons (Baldwin 1980); consequently, small amounts must remain in the melt, in spite of the fact that thermodynamic calculations suggest that cerium partitions strongly to the slag. In addition, as noted in Section F.5.1, cerium was not found in cast iron at the limits of detection in samples from 28 ductile iron foundries. Given this conflicting information, the most likely situation is that minute amounts of cerium will remain in the cast iron. However, no evidence has been uncovered which suggest that the amount of cerium remaining in the melt is greater than 0.5% of the total.3 Partition ratios in the present analysis are calculated to the nearest 1%. Thus, any partition ratio less than 0.5% is assigned a value of zero. F-12 ------- F.5.5 Cesium Cesium is expected to partition to the slag and to accumulate in the baghouse dust. None is expected to remain in the melt (Harvey 1990). F.5.6 Iron Some iron is expected to be oxidized and to transfer to the slag. However, no detailed composition data have been located in this study to permit quantification of this expected partitioning. Therefore, it is conservatively assumed that no iron partitions to the slag. F.5.7 Lead Based on thermodynamic equilibrium calculations, lead is expected to remain in the melt. However, lead has very limited solubility in liquid iron. Furthermore, it has a vapor pressure of 0.01 atm at 1,408 K (Darken and Gurry 1953) and 0.05 atm at 1,462 K (Perry and Green 1984). At the limits of detection, lead is seldom found in cast iron (see Section F.5.1). Lead has been detected in leachates from baghouse dust collected by cupola emission control systems. Leachate levels based on the EP toxicity test ranged from about 10 to about 220 mg/L (Kunes et al. 1990). Since it is not possible to quantitatively relate these leachate results to contaminant levels in the dust, one can only reach the qualitative conclusion that some lead vaporizes from the cast iron melt and is collected in the baghouse. The combined evidence indicates that, for the purposes of the present analysis, lead can be assumed to completely vaporize from the melt. F.5.8 Manganese Based on thermodynamic calculations which assume that Y°Mn = 2.6, the partition ratio of manganese between slag and iron is calculated to be about 5 at 1,573 K (see Table F-5), which suggests that significant amounts of manganese will be present in both the slag and the melt. Meraikib (1993) determined that during steelmaking, the distribution of manganese between the slag and the melt could be described by the equation F-13 ------- (Mn) [Mn] a[0]f[Mn]eXP 27530 - 0.0629 B - 7.3952 (Mn) [Mn] a[Q] f[Mn] T B concentration of manganese in slag (wt%) concentration of manganese in melt (wt%) activity of oxygen in melt activity coefficient for [Mn] absolute temperature (K) slag basicity Although there are risks in extrapolating this equation to cast iron melting, the calculation was undertaken in the absence of better information. Partition ratios at two different partial pressures of CO were estimated, assuming T = 1,573 K, B = 0.63, f[Mn] = 0.95, and 130 Ib of slag generated per ton of metal melted. These values are listed in Table F-10. Table F-10. Partition Ratios of Manganese at Different Partial Pressures of CO "co (atm) 1 0.1 "HMn 0.45 0.045 Partition Ratio (see text) (mass in slag/mass in metal) 0.03 0.003 Note: The oxygen activity is calculated using free energy values for C and O dissolved in iron (JSPS 1988) and the CO free energy of formation given by Glassner (1957). The calculated values are in close agreement with information presented by Engh (1993, p. 67). F.5.9 Niobium On the basis of thermodynamic calculations, niobium is expected to partition primarily to the slag. However, according to Harvey (1990), niobium can be retained in steel under reducing conditions. The expected reaction is 2Nb + 6O + Fe = FeO'Nb.O, F-14 ------- where the elements on the left side of the equation are melt constituents and the compound on the right is a slag constituent. The equilibrium constant for the reaction is = a a2 a6 a£e aNb aO (T = 1,873K) Assuming that l-Nb,Ot = 1, values of aNb corresponding to two assumed values of a0 were calculated, as listed below: ao 1 0.01 aNb 6.5e-6 6.5 The value of K1573, the equilibrium constant at 1,573 K, is not available; however, based on the values of the free energies of formation of Nb2O5 at 1,573 K and 1,873 K, it is expected that K1573 > K1873 Thus, a highly reducing environment (a0 < 1) would be required to retain niobium in the melt at the lower temperature. As noted in Section F.5.1, niobium is not detected in cast iron at the detection limit, which indicates that either there are no significant quantities of niobium in steel scrap or the typical melting conditions are not sufficiently reducing to cause niobium to be retained in the melt. F.5.10 Zinc Under steelmaking conditions, zinc is expected, from a free energy perspective, not to partition to the slag and, because of its high vapor pressure, to vaporize from the melt to a large extent. Cast iron melting temperatures, though lower, are still well above the normal boiling point of zinc (1,180 K). Based on information presented by Perrot et al. (1992), the solubility of zinc at 1,573 K is expected to be about 140 ppm when the partial pressure of zinc is 10"2 atm. Silicon in the cast iron will tend to increase the zinc solubility while manganese will have the opposite effect. As noted in Section F.5.1, from 20 to 50 ppm of zinc are typically found in cast iron, which suggests that it is unrealistic to assume that 100% of the zinc volatilizes and collects in the baghouse. F-15 ------- Assume, for example, that a furnace charge contains 45% steel scrap and 55% cast iron scrap, and that both the cast iron scrap and the product contains 30 ppm Zn, as listed in Table F-8. If the steel scrap contains less than 0.67 wt% Zn, then 1% or more of the zinc would remain in the melt (see Note 3) (Koros 1994). According to Koros (1994), typical galvanized scrap contains about 2% Zn. The same author reported that, in 1992, 35% of the scrap classified as No. 1 bundles and busheling is galvanized steel. Other grades of scrap likely to contain significant quantities of galvanized steel include shredded scrap and No. 2 bundles (Fenton 1996). For 1993, No. 1 bundles, No. 1 busheling, shredded, and No. 2 bundles accounted for 46% of the carbon steel scrap used in iron foundries (Bureau of Mines 1995). Using the above information, it can be estimated that about 2% of the zinc will remain in the cast iron and the balance will be transferred to the baghouse dust, based on the following calculation: pZn pZn UFe rFe fFe' r Zn f s fg' f g r Zn TFe UFe' + TFe Ts Tg' Ug PFB" = partition fraction of zinc in cast iron = 0.0205 CFzen = mass fraction of zinc in cast iron product = 3 x 1Q-5 f FFee = mass ratio of cast iron scrap : cast iron product = 0.55 CFB" = mass fraction of zinc in cast iron scrap = 3 x 1Q-5 f Fe = mass ratio of steel scrap : cast iron product = 0.45 fs9 = fraction of galvanized-steel-bearing scrap sources in steel scrap = 0.46 Ffl' = 0.35 fs9 = fraction of galvanized steel in galvanized-steel-bearing scrap sources = mass fraction of zinc in galvanized steel = 0.02 F-16 ------- F.6 PARTITIONING SUMMARY F.6.1 Elements Which Partition to the Melt It is assumed that 1% of the total melt will be transported from the furnace and collected in the baghouse. This is approximately the geometric mean of the values for two types of foundries listed in Table F-2 and is consistent with the values cited in U.S. EPA 1995 (see Section F.2). Based on thermodynamic equilibria, the following elements are expected to partition 99% to the melt and 1% to the baghouse dust: cobalt, molybdenum, nickel, ruthenium, and technetium. Free energy calculations also suggest that silver partitions to the melt but, for EAF melting of carbon steel, this information was tempered by the facts that silver has a significant vapor pressure at steelmaking temperatures (10"2 atm at 1,816 K) and some work on stainless steel melting done at Studsvik (Menon et al. 1990) had shown silver in the baghouse dust. However, the vapor pressure of silver is at least an order of magnitude lower at temperatures used in cast iron melting (e.g., 10"3 atm at 1,607 K)(Darken and Gurry 1953). Consequently, in cast iron, silver is assumed to partition 99% to the melt and 1% to the baghouse dust. Although there is reason to suspect that some niobium might be found in the melt under highly reducing conditions, no evidence was uncovered to support that supposition. For reasons discussed in Section F.3.3 above, carbon and antimony are expected to remain in the melt except for small quantities contained in dust transferred to the baghouse (i.e., 1%). Manganese is predicted to remain primarily in the melt. It is expected that no more than about 2% of the manganese will partition to the slag. Most of the zinc is expected to volatilize and be collected in the baghouse. Only about 2% is assumed to remain in the melt. Table F-l 1 lists the partition ratios of elements which are expected to show significant (i.e., at least 1%) partitioning to the melt. F-17 ------- F.6.2 Elements Which Partition to Slag For those elements which are strong oxide formers and are expected to partition to the slag, the assumption is made here that 5% of the slag will be transported to the baghouse as dust. This is the same assumption as made for melting carbon steel in electric arc furnaces. Based on this assumption, thermodynamic equilibrium calculations at 1,573 K and chemical analogies, the following elements are expected to partition 95% to the slag and 5% to the baghouse dust: Ac, Am, Ce, Cm4, Eu4, Nb, Np, Pa, Pm4, Pu, Ra, Sr, Th, and U. Table F-l 1. Proposed Partitioning of Metals Which Remain in the Melt Element Ag C Co Fe Mn Mo Ni Ru Sb Tc Zn Distribution (%) Melt 99 99 99 99 97 99 99 99 99 99 2 Slag 2 Baghouse 1 1 1 1 1 1 1 1 1 1 98 Since thermodynamic data were not available for these elements, partitioning was assumed to be analogous to similar elements in the rare-earth and actinide series in the periodic table. F-18 ------- REFERENCES Baldwin, V. H. 1980. "Environmental Assessment of Iron Casting," EPA-600/2-80-021. Research Triangle Institute. Bureau of Mines, U.S. Department of Interior. 1995. "Recycling Iron and Steel Scrap." Darken, L. S., and R. W. Gurry. 1953. Physical Chemistry of Metals. McGraw-Hill Book Company. Engh, T. A. 1992. Principles of Metal Refining. Oxford University Press. Fenton, M. D., (Iron and Steel Specialist, U.S. Geologic Survey). 1996. Private communication (3 September 1996). Fenton, M. D. 1998. "Iron and Steel Scrap." InMinerals Yearbook., U.S. Geological Survey. Fenton, M. D. 1999. "Iron and Steel Scrap." In Minerals Yearbook, U.S. Geological Survey. Glassner, A. 1957. "The Thermochemical Properties of Oxides, Fluorides, and Chlorides to 2500°K," ANL-5750. Argonne National Laboratory. Harvey, D. S. 1990. "Research into the Melting/Refining of Contaminated Steel Scrap Arising in the Dismantling of Nuclear Installations," EUR 12605 EN. Commission of the European Communities. Japan Society for the Promotion of Science (JSPS). 1988. SteelmakingData Sourcebook. Gordon and Breach Science Publishers. Jeffery, J., et al. 1986. "Gray Iron Foundry Industry Particulate Emissions: Source Category Report," EPA/600/7-86/054. GCA/Technology Division, Inc. Kalcioglu, A. F., and D. C. Lynch. 1991. "Distribution of Antimony Between Carbon-Saturated Iron and Synthetic Slags." Metallurgical Transactions, 136-139. Koros, P. J. 1994. "Recycling Galvanized Steel Scrap." In Proceedings of the CMP Electric Arc Furnace Dust Treatment Symposium IV, CMP Report No. 94-2. Prepared for the EPRI Center for Materials Production. Kunes, T. P., et al. 1990. "A Review of Treatment and Disposal Technology Applied in the USA for the Management of Melting Furnace Emission Control Wastes." In Conference: Progress in Melting of Cast Irons. Warwick, U.K. F-19 ------- Menon, S., G. Hernborg, and L. Andersson. 1990. "Melting of Low-Level Contaminated Steels." In Decommissioning of Nuclear Installations. Elsevier Applied Science. Meraikib, M. 1993. "Manganese Distribution Between a Slag and a Bath of Molten Sponge Iron and Scrap." ISIJInternational^ (3): 352-360. Natziger, R. H., et al. 1990. "Trends in Iron Casting Compositions as Related to Ferrous Scrap Quality and Other Variables: 1981-86," Bulletin 693. U.S. Bureau of Mines. Nassaralla, C. L., and E. T. Turkdogan. 1993. "Thermodynamic Activity of Antimony at Dilute Solutions in Carbon-Saturated Liquid." Metallurgical Transactions B, 24B: 963-975. Perry, R. H., and D. W. Green. 1984. Perry's Chemical Engineers'Handbook. 6th Ed. McGraw-Hill Book Co., Inc. Perrot, P., et al. 1992. "Zinc Recycling in Galvanized Sheet." In The Recycling of Metals (Proc. Conf.) Dusseldorf-Neuss, Germany. Sigworth, G. K., and J. F. Elliott. 1974. "The Thermodynamics of Liquid Dilute Iron Alloys." Metal Science 8: 298-310. United States Steel Company (U.S. Steel). 1951. The Making, Shaping, and Treating of Steel. 6th ed. Pittsburgh. U.S. Environmental Protection Agency, Office of Air Quality (U.S. EPA). 1995. "Compilation of Air Pollutant Emission Factors," AP-42. 5th ed. Vol. 1, "Stationary Point and Area Sources." U.S. EPA, Research Triangle Park, NC. F-20 ------- APPENDIX G DILUTION OF RESIDUALLY RADIOACTIVE SCRAP STEEL ------- Contents page G.I Introduction G-l G.2 Average Case G-2 G.3 Reasonable Maximum Exposure Case G-2 G.4 Adopted Approach to Dilution G-8 G.4.1 Scrap Transport Scenarios G-8 G.4.2 Recycle Scenarios G-8 G.4.3 Finished Product Scenarios G-9 G.4.4 Processing of Baghouse Dust G-10 References G-13 Table G-l. Mass of Residually Contaminated Carbon Steel Scrap Released in Rockwood HRDC Service Area G-12 Figures G-l. Electric Arc Furnace Shops in NRC Region I (Northeast) G-4 G-2. Electric Arc Furnace Shops and Nuclear Facilities in NRC Region II (Southeast) .... G-5 G-3. Electric Arc Furnace Shops and Nuclear Facilities in NRC Region III (North Central) G-6 G-4. Electric Arc Furnace Shops and Nuclear Facilities in NRC Region IV (West) G-7 G-iii ------- DILUTION OF RESIDUALLY RADIOACTIVE SCRAP STEEL This appendix describes the development of the dilution factors discussed in Section 5.4.1 and used to assess the radiation exposures of individuals that result from recycling potentially contaminated steel scrap. G.I INTRODUCTION Chapter 5 discusses the operations and scenarios used to assess the radiation exposures of the RME individual resulting from the recycling of potentially contaminated steel scrap. Each operation exposes the individual to materials or products generated during a certain stage of the recycling process. It is unlikely that for an entire year,1 any steel mill would be exclusively supplied with scrap resulting from the dismantling of potentially contaminated components. To determine the largest fraction of steel scrap that would be potentially contaminated, the anticipated release of scrap steel by various generator sites nationwide was matched to the scrap processing capacities of nearby steel mills. This appendix presents a discussion of that analysis. "Electric Arc Furnace Roundup" (1996) listed 213 furnaces with a combined nominal capacity of 57,850,000 tons per year.2 The largest furnace in this survey was a 370-ton furnace with a nominal capacity of 950,000 tons per year; the smallest was a 10-ton furnace with an annual capacity of 4,000 tons. The average annual capacity of all the furnaces in the survey was 272,000 tons. EAF steel production in 1995 was 40,619,000 tons (AISI 1995), which suggests that the industry was running at about 70% of capacity during that year. One important factor in developing worker exposure scenarios is the number of furnaces at a site. If there are multiple furnaces at a site, the worker exposure may be related to the total steel tonnage produced at the site rather than the tonnage produced by a single furnace. Recognizing the importance of these and other factors, one can make some estimates as to how operating conditions may alter worker exposure from melting residually radioactive steel scrap. First, an average exposure case will be considered, followed by a reasonable maximum exposure case. In The potential radiological impacts on the RME individual are assessed for the year of peak exposure. Statistical data on U.S. steel production and the steel inventories of nuclear power plants are normally presented in English units (1 ton = 907.2 kg). To present the data as published and to avoid tedious repetition, these values are not generally converted to metric units in this appendix. G-l ------- addition to determining the dilution factors for the steel mill scenario, the discussion will also cover the dilution of potentially contaminated scrap in the truck carrying steel to the scrap processor, the maximum likely dilution factor for any one furnace charge, and the dilution factor of contaminated EAF dust at a high temperature pyrometallurgical metals recovery plant. G.2 AVERAGE CASE According to Table A-81, the total inventory of carbon steel in U.S. commercial nuclear power plants—the 104 currently licensed reactors and the 17 reactors which are permanently shut down and in SAFSTOR or scheduled for DECON—is estimated to be about 3.5 million metric tons (t). As shown in Table A-84, the release of scrap metal from these facilities is expected to begin in 2006, with the bulk of the metal being released during a 40-year period starting in 2019. An average of 89,0001 of carbon steel scrap would be generated each year during this period. If all of this steel were shipped as scrap to a single "average" EAF, it would represent about 35% of the annual capacity of that furnace alone. If it were evenly distributed among all the furnaces in the United States, this scrap would represent 0.16% of total EAF capacity. G.3 REASONABLE MAXIMUM EXPOSURE CASE The NRC has divided the 48 contiguous states into four administrative regions, which are depicted in Figures G-l to G-4. Superimposed on these maps are the locations of steel mills employing EAFs, as well as the locations of nuclear power plants and major DOE facilities that constitute present and future sources of potentially contaminated scrap metal. These maps show that both EAFs and nuclear facilities are broadly distributed across the country. A cursory examination reveals that, with two exceptions, each state that is host to a nuclear facility also has one or more EAF shops or is adjacent to a state that has such shops3. Since transportation costs would be a major factor in determining which EAF shop receives the scrap from a given nuclear facility, the geographical distribution of nuclear facilities and scrap melters should lead to the scrap being distributed among many EAFs. However, the simultaneous shutdown of two or more nuclear power plants in the same vicinity could lead to the release of a relative large amount of scrap at a single location for a brief period of time. A few hypothetical examples of such releases, and their consequences, are discussed in this section. The exceptions are Maine and New Hampshire. The nuclear plants in these states are nevertheless closer to the nearest EAFs than are some of the nuclear facilities in the West. The scales of the maps, which are different for the Northeast and Western regions, may give a different visual impression. G-2 ------- To develop the reasonable maximum exposure case, it was assumed that scrap steel tends to move the shortest possible distance to minimize transportation costs. For example, when the five nuclear power plants in southern California (San Onofre 1, 2, and 3, and Diablo Canyon 1 and 2) are dismantled, it was assumed that the carbon steel scrap would be shipped to TAMCO, near Riverside, Calif., for melting. Based on the time table developed in Section A.5.4, scrap from these five plants would be released between 2031 and 2052. Two of these reactors, San Onofre 2 and 3, are scheduled to shut down in 2022. Although decommissioning of a nuclear power plant can take several years (Smith et al. 1978), for the purpose of a conservative analysis, it was assumed that all the recyclable scrap metal would be released in a single year. According to Table A-29, the decomissioning of the Reference 1,000 MWe (1 GWe) PWR would generate up to 33,000 t of carbon steel scrap. Applying the scaling factors that reflect the power ratings of these reactors (see Section A.5.2.1) and converting to English units, it was found that up to 76,000 tons would be available in 2032 from these two units. This is about 19% of the 400,000- ton nominal annual capacity of TAMCOfor that year alone. By the same logic, the other three units, each scheduled to be shut down in a different year, would use less than 10% of TAMCO's capacity in any one year. Not all the carbon steel scrap generated by the decommissioning of a commercial nuclear power plant would consist of the potentially contaminated, recyclable metal that is the subject of this analysis. Some of the scrap generated during decommissioning would never have been exposed to radioactive contamination (and would therefore be outside the scope of the analysis), while other metal would have neutron activation products throughout its volume or would be so heavily contaminated that it would not be a candidate for clearance. Table A-80 indicates that a maximum of 3,3111 of carbon steel from the Reference PWR and 6,754 t of carbon steel from the Reference 1 GWe BWR would be residually radioactive metal potentially suitable for clearance. Again applying the appropriate scaling factors and converting to English units, it was found that only about 7,700 t of potentially contaminated scrap from San Onofre 2 and 3 would be available for clearance. Such scrap would constitute about 1.9% of TAMCO's nominal annual capacity. In this hypothetical scenario, any stainless steel available for recycle would have to be shipped elsewhere, since TAMCO is a carbon steel shop. G-3 ------- Figure G-l. Electric Arc Furnace Shops in NRC Region I (Northeast) / KEY I OPERATING REACTOR 1 OR MORE REACTORS HAVE BEEN SHUTDOWN AT THIS SITE ELECTRIC ARC FURNACE SHOP [# of Furnaces] (Total Annual Capacity) (xl.OOO Tons) F i tzpatri ck / Nine Mi Ie Point 1 & 2 Crucible Motenols Corp. \ Vermooit Yankee / m (75) Allegheny Ludlum Corp.N Special Materials Div. 1 G i nna J. >^ Armco Advanced Materials Corp. [3] (840) Al Tech Specialty / Steel Corp. [2] (125) Auburn Steel Co. (4001 Bethlehem Steel Corp.X BethForge Div. [5] (1300) North Star Steel Co. Susquehonna \ l nd; an foi nt 1>^1 &V * «ll« Milton Div. \ ' - ~ N trolloy [3](I50) \^ Standard Steel Corp. Luk?"= ' Burham Plant f I2J l85' [3] (140.5) ^ FlrstMlSS Steel Inc. (1100) Three Mile Island 1 & 2 Roriton Steel Corp. 1£70) New Jersey Steel Corp. (500) L i mer i ck 1 & 2 Washington Steel Corp [21 (279) Jessop Steel Corp [33 nr" Allegheny Ludlum Corp. Special Materials Div. [3](B4B) Edqewater Steel Corp. Armco Inc. ,60) Baltimore Specialty Steel Corp. Hope Creek 1 Salem 1 & 2 Yankee-Rows 1 Haddam Neck Carpenter Technology Corp. IB) (6001 Peach Bottom 1 , 2 & 3 Co I vert CI iffs 1 & 2 Bethlehem Steel Corp. 1 Rail Products & Pipe Div. 151(16501 2 Bar, Rod, and Wire Div. [2] (H001 ------- NS Group Inc Newport Steel Corp. [31 (550) Florida Steel Corp. Knoxville Steel Mill [21 (170) NS Group Inc Kentucky Electric Steel Corp. [2] (2S0) x:\ Steel of West Virginia Inc. [2] 1400) Tennessee Valley [3] (630) North Anna 1 & 2 Hoeganoes Corp. 1175) Florida Steel Corp. Tennessee Steel Mill (4501 Bellefonte 1 & 2 - Browns Ferry 1 .2 x'Roanoke Electric Steel Corp-___- [2D C5500) Summer 1 i ^ hlondes bleel Lorp. » F\ Jacksonville Steel Mi Crystal River 3 Nucor Corp. igti [51 (505) Nucor Corp. Dorlinqton Plont Farley 1 & 2 Atlantic Steel Co. [2] I1000I Grand Gulf Birmingham Steel Corp Mississippi Steel Div. Lucie 1 &. 2 •ft MAJOR DOE FACILITIES I OPERATING REACTOR 41 ELECTRIC ARC FURNACE SHOP [* of Furnaces] (Total Annual Capacity) (xl.OOO Tons) Figure G-2. Electric Arc Furnace Shops and Nuclear Facilities in NRC Region II (Southeast) Turkey Point 3 & ------- KEY MAJOR DOE FACILITY OPERATING REACTOR 1 OR MORE REACTORS HAVE BEEN SHUTDOWN AT THIS SITE ELECTRIC ARC FURNACE SHOP [w of Furnaces] (Total Annual Capacity) 1x1.000 Tons Fermi 2 Big Rock Point Mont ice I Io A North Star Steel Corp. Minnesota Dlv. 1310) Prairie Is land 1 & 2 Kewaunee J Lacrosse Point&Beach Quad Cities 1 & 2 Maynard Steel o, Casting Co. *^ [43 141.4) _ Charter Steel Co. 12001 Duone Arnold X -Byron 21 Zion 1 6® National Steel Corp. [231720) McLouth Steel [231560) Rouge Steel Co. [21 (150) MACSTEEL Michigan Dlv. [23 (4101 :ook 1 & 2 g •PoIisades North Star ^teel Co.^Sa , , g 1 & 2 Dresden 2'& 3 (^ Harrison Steel Braidwood 1 & 2 Birmingham Steel Corp. Illinois Steel Divf. A 123IS00) Davis-Besse 1 Slater Steels Corp. [23 I60I Marion Steel Co. [231380) \ Keokuk SteelNB ff V Costing Inc. I V \ Arnco Inc. S^ Cal ' aw°y Kansas City Works [2] 1850) s ST,one oieei ,_. .£• _, Wire Inc. 133 116.5) [23 11300) 2 C 1 i nton 1 _aclede Steel Co. [21 1100) Vy~*A ^f X i ®nour if a 11 > ten "P A " C01"P- 1 ® C2] 120) 1 f NUCOP Corp. ? Crowfordsville v Plant 1 [21 11000) , — /" ( J> -^f^ North Star Steel Co. Michigan Dlv. 1480) Sandusky International [2! 18) Perry 1 & 2 - Champion Steel Co. 127) Copperweld Steel Co. [43(480) xNorth Star Steel Co. North Star Steel Ohio [21 1440) LTV Steel Co. [2] (7121 [23(201 I Worthington Industries Inc. [21 (2351 V FernaId Timken Co. Conton Plont [4] 15200) Faircrest Plont £700) Republic Engineered Steels Inc. CGJ 12580) Portsmouth 1 Charter Electric Melting Co. (120) 2 Finkl. A. & Sons [2] (^0) 3 U.S. Steel Corp. [3] n000) 4 Island Steel Bar Co. C23 1540) 5 Calumet Steel Co. C2) I15BJ 6 Thomas Steel Corp. [211400] 7 Northwestern Steel & Wire Co. E31 (1550) • The capacity of the thud furnace at this facility is Figure G-3. Electric Arc Furnace Shops and Nuclear Facilities in NRC Region III (North Central) G-6 ------- Birmingham Steel Corp. Salmon Bay Steel Dw. [2] I44 Trojan Birmingham Steel Corp. Salmon Bay Steel Div. C2J 1449) ESCO Corp. 13] (114) Oregon Steel / Mills Inc. Humboldt Bay O Diablo Canyon • San Onofre 1 . 2 & Washi ngton Nuclear 2 / I daho F a fTlfTj'Qt i ona I / Engineerina' Lab S\ Cooper Atchison Costing Corp. Nevada/Test Site 1 Texes Steel Co. [2] t40> 2 Hensley, G.H. [2] [ill 3 Chaparral Steel Co. [2] I15BBI Palo Verde 1,\2 5. 3 Comanche Peak 1 A 2 Wolf Creek Arkansas Steel Associates MACSTEEL Arkansas Div. IS30> / Nucor-Yomo-to Steel Co. Nucor Corp. Hickman Plant Arkansas/NucI ear 1 & 2 Lone Star Steel Inc. [2J(400) n Le Tout-neau Co. C21(80) MAJOR DOE FACILITY OPERATING REACTOR 1 OR MORE REACTORS HAVE BEEN SHUTDOWN AT THIS SITE ELECTRIC ARC FURNACE SHOP [ft of Furnaces] (Total Annual Capacity) Bend 1 Water-ford 3 CMC Steel Grpuo .ructural Metals Inc. 1650) Texas Foundries [2] (40) Figure G-4. Electric Arc Furnace Shops and Nuclear Facilities in NRC Region IV (West) Nucor Corp. Jewett Plont ------- The three peak years for reactor shutdowns are expected to be 2013, 2014, and 2026. Nine reactor operating licenses are due to expire that first year4. Again assuming a ten-year delay between shutdown and release of scrap metal, up to 260,000 tons of carbon steel would be released in 2023. Two of these plants—Kewaunee and Point Beach 2—although belonging to different utilities, are near one another. The total amount of carbon steel scrap from these plants—about 46,000 tons—is still less than the amount from San Onofre 2 and 3. Of the remaining seven plants, each is located in a different state. Eleven plants are anticipated to shut down in 2014, resulting in the release of up to 350,000 tons of carbon steel in 2024. Only two of these facilities—Three Mile Island 1 and Peach Bottom 3—are located in the same state. These plants are owned by different utilities; although they are only about 40 miles apart, the profusion of EAF shops in the area makes it unlikely that all the carbon steel scrap from both plants would be recycled in the same facility during the same year. The remaining nine nuclear plants are each located in a different state. Nine plants are anticipated to shut down in 2026, resulting in the release of up to 350,000 tons of carbon steel in 2036. Two of these plants, Braidwood 1 and Byron 2, are both owned by Commonwealth Edison and are less that 100 miles apart. Up to 77,000 of scrap is projected to be released from these plants in 2036, about the same as the amount from San Onofre 2 and 3. Thus, little or no new geographical concentration is projected in any of these three years. G.4 ADOPTED APPROACH TO DILUTION G.4.1 Scrap Transport Scenarios Once the scrap metal is cleared, there would be little reason to segregate residually contaminated metal from scrap that has never been exposed to radioactive contamination. Given this assumption, the highest fraction of contaminated scrap would be generated during the decomissioning of a BWR—as stated above, out a total of 34,000 t of carbon steel in a 1.0 GWe BWR Reference reactor facility, 6,753 t, or about 20%, would be residually radioactive metal that could potentially be cleared. G.4.2 Recycle Scenarios The largest total amount of carbon steel scrap—as well as the largest amount of residually radioactive scrap that could potentially be cleared—from any single commercial facility is 4 See Table Al-1 in Appendix A-l. G-8 ------- anticipated to be from the decommissioning of Perry 1 in northeastern Ohio in 2036. The total amount of carbon steel scrap in this 1,160-MWe reactor is calculated to be 37,540 t, of which 7,455 t would be potentially contaminated. As shown in Figures G-l and G-3, there are a number of EAF facilities in western Pennsylvania and northeastern Ohio which are relatively near to this reactor site. The annual capacity of the EAF shops in northeastern Ohio alone varies from a few thousand tons to over one million tons. Since it is difficult to predict which of these shops are likely to receive this scrap, it was assumed that the scrap would be recycled at the reference facility described in Chapter 5. Since this 150,000-ton-per-year EAF shop, with two furnaces, has a smaller annual production than the 272,000-ton-per-furnace national average, such an assumption is reasonably conservative. Factors which could further reduce the quantity of scrap from nuclear facilities melted in a given shop include: • incompatibility of scrap with product specifications • incompatibility of large, single-source commitments with other purchasing arrangements • reluctance to handle such scrap irrespective of actual risks • scrap buy-back arrangements with customers • release of scrap from the decommissioning of a reactor over a period of several years One factor which could possibly increase the use of such scrap by a given recycling facility is the possibility that its price would have to be heavily discounted in comparison to comparable non- nuclear scrap, and that some marginal melt shops might seize the opportunity to purchase cheap scrap for a quick profit. G.4.3 Finished Product Scenarios If each EAF charge consisted of scrap from a single source, it would be quite likely—indeed, inevitable—that some of the 2,000 heats produced during the one year that the reference facility is processing decomissioning scrap would be composed entirely of residually contaminated steel. In reality, that is never the case. According to Tom Danjczek (1999), President of the Steel Manufacturer's Association and a former EAF supervisor, a single charge would contain steel from 5 to 20 sources. Using the geometric mean of this range—10 sources per furnace charge—a computerized Monte-Carlo simulation was performed to determine the maximum likely fraction of contaminated scrap in any single charge. In this simulation, the first 7.5 tons of G-9 ------- the first 75-ton charge was randomly selected from the annual supply of 150,000 tons of scrap, comprising 7,500 tons of contaminated scrap and 142,500 tons of clean scrap. Whichever source was utilized was decreased by 7.5 tons and the process was repeated until the 75-ton furnace was fully charged. The next charge was then made up in the same manner, utilizing the now- decreased scrap supply; the process was continued until the entire supply was exhausted. The simulation was repeated 1,000 times. The highest fraction of contaminated scrap in any heat in 1,000 simulations of 2,000 heats each was 0.6. The 90th percentile fraction of contaminated scrap was equal to 0.5—this was the highest fraction in any of the 2,000 heats that was exceeded in fewer than 10% (100 out of 1,000) of the simulations. (In fact, the 95th percentile fraction was also 0.5.) Consistent with EPA's definition of reasonable maximum exposure, the 90th percentile value—a dilution factor of 0.5—was adopted for the exposure assessment of the finished product scenarios. G.4.4 Processing of Baghouse Dust Most of the EAF dust generated in the United States between 1992 and 1995 was shipped to high temperature pyrometallurgical metals recovery plants owned and operated by the Horsehead Resource Development Company (HRDC). HRDC operates three regional Wealz kiln plants, located in Palmerton, Penn.; Chicago; and Rockwood, Tenn., that have a cumulative annual capacity of about 450,000 tons per year (Bossley 1994, Schmitt 1996). Based on information in U.S. EPA 1994, HRDC was assumed to have a total of six Wealz kilns, three of which are in Palmerton, two in Chicago and the remaining one in Rockwood. Apportioning the processing capacity equally among the six kilns, the annual capacity of the Palmerton facility was assumed to be 225,000 tons; Chicago: 150,000 tons; and Rockwood: 75,000 tons. All baghouse dust generated by the melt-refining of carbon steel scrap released during the decomissioning of a nuclear power plant was assumed to be processed at the HRDC facility nearest to that plant. The maximum amount of potentially contaminated scrap released during any one year in each of the three HRDC facilities' assumed service areas was compared to the processing capacity of that facility. As might be anticipated, the highest concentration of contaminated dust would occur at the Rockwood facility, which has the smallest processing capacity. This facility's assumed service area encompasses all of NRC Region II except eastern Virginia, as well as the states of Arkansas, Louisiana and Texas. Table G-l lists the nuclear power plants in this area, along with the amount of potentially contaminated carbon steel scrap that would be generated and the anticipated year of release. In 2024, the peak year for releases in G-10 ------- this area, about 21,000 1 of contaminated carbon steel scrap would be generated by the decomissioning of four nuclear power plants. As discussed in Section 6.2, the amount of baghouse dust generated by the melting of the potentially contaminated steel scrap is calculated as follows: M f , M . = f s Md = Mass of baghouse dust generated by the melting of contaminated steel scrap = 333 t= 368 tons Ms = Mass of potentially contaminated steel scrap released = 21,1211 fd = mass of baghouse dust as a fraction of metal charged to furnace = 0.015 fs = mass of scrap imported to the facility as a fraction of metal charged to furnace = 0.95 The dilution factor at Rockwood would therefore be approximately 0.005 (368 + 75,000 ~ 0.005). G-ll ------- Table G-l. Mass of Residually Contaminated Carbon Steel Scrap Released in Rockwood HRDC Service Area Reactor Name Arkansas Nuclear 1 Arkansas Nuclear 2 Shearon Harris 1 H. B. Robinson 2 Catawba 1 Catawba 2 McGuire 1 McGuire 2 Oconee 1 Oconee 2 Oconee 3 Crystal River 3 St. Lucie 1 St. Lucie 2 Turkey Point 3 Turkey Point 4 Vogtle 1 Vogtle 2 South Texas 1 South Texas 2 Waterford 3 Summer Sequoyah 1 Sequoyah 2 Watts Bar 1 Comanche Peak 1 Comanche Peak 2 Brunswick 1 Brunswick 2 Hatch 1 Hatch 2 River Bend 1 Grand Gulf 1 Browns Ferry 1 Browns Ferry 2 Browns Ferry 3 Total Reactor Type PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR PWR BWR BWR BWR BWR BWR BWR BWR BWR BWR Power Rating (MWe) 836 858 860 683 1,129 1,129 1,129 1,129 846 846 846 820 839 839 666 666 1,105 1,103 1,250 1,250 1,075 885 1,122 1,122 1,170 1,150 1,150 767 754 744 762 936 1,143 1,065 1,065 1,065 Scaling Factor1" 0.887 0.903 0.904 0.775 1.084 1.084 1.084 1.084 0.894 0.894 0.894 0.876 0.89 0.89 0.763 0.763 1.069 1.068 1.16 1.16 1.049 0.922 1.08 1.08 1.11 1.098 1.098 0.838 0.828 0.821 0.834 0.957 1.093 1.043 1.043 1.043 Mass (t) 2938 2989 2994 2568 3590 3590 3590 3590 2962 2962 2962 2901 2945 2945 2525 2525 3539 3535 3842 3842 3475 3052 3575 3575 3677 3634 3634 5659 5595 5545 5634 6463 7384 7044 7044 7044 145,365 Year* + 10 2024 2028 2036 2020 2034 2036 2031 2033 2043 2043 2044 2026 2026 2033 2022 2023 2037 2039 2037 2038 2034 2032 2030 2031 2045 2040 2043 2026 2024 2024 2028 2035 2032 2023 2024 2026 Mass Released by Year (t)+ 2023 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2525 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 7044 0 0 9,568 2024 2938 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5595 5545 0 0 0 0 7044 0 21,121 2026 0 0 0 0 0 0 0 0 0 0 0 2901 2945 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5659 0 0 0 0 0 0 0 7044 18,548 2028 0 2989 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5634 0 0 0 0 0 8624 2031 0 0 0 0 0 0 3590 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3575 0 0 0 0 0 0 0 0 0 0 0 0 7165 2032 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3052 0 0 0 0 0 0 0 0 0 0 7384 0 0 0 10,436 2033 0 0 0 0 0 0 0 3590 0 0 0 0 0 2945 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5535 2034 0 0 0 0 3590 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3475 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 7065 2036 0 0 2994 0 0 3590 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5584 2037 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3539 0 3842 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 7381 2043 0 0 0 0 0 0 0 0 2962 2962 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3634 0 0 0 0 0 0 0 0 0 9558 Year of shutdown + 10 (see Table Al-1) See Section A.5.2.1 Tabulation is for years during which two or more reactors are scheduled for decommissioning G-12 ------- REFERENCES American Iron and Steel Institute (AISI). 1995. "Pig Iron and Raw Steel Production," Report AIS7 (preliminary). Bossley, J. J. 1994. "Proceedings of the CMP Electric Arc Furnace Dust Treatment Symposium IV," CMP Report No. 94-2. EPRI Center for Materials Production. Danjczek, T. A., (President, Steel Manufacturer's Association). 1999. Private communication. "Electric Arc Furnace Roundup." 1991. Iron and SteelMaker, May, 1996. Schmitt, R. 1996. "Proceedings of the CMP Electric Arc Furnace Dust Treatment Symposium V," CMP Report No. 96-3. EPRI Center for Materials Production. Smith, R.I., G. J. Konzek, and W. E. Kennedy, Jr. 1978. "Technology, Safety and Costs of Decommissioning a Reference Pressurized Water Reactor Power Station," NUREG/CR- 0130. 2 vols. Pacific Northwest Laboratory, prepared for the U.S. Nuclear Regulatory Commission, Washington, DC. U.S. Environmental Protection Agency (U.S. EPA). 1994. "Report to Congress on Metal Recovery, Environmental Regulation & Hazardous Waste," EPA 530-R-93-018. U.S. EPA, Washington, DC. U.S. Nuclear Regulatory Commission (U.S. NRC). 2000. "Information Digest, 2000 Edition," NUREG-1350, Volume 12. U.S. NRC, Washington, DC. G-13 ------- APPENDIX H DETAILED SCENARIO DESCRIPTIONS FOR CARBON STEEL ------- Contents page H.I Driver Transporting Carbon Steel Scrap—SCRPDRVR H-l H. 1.1 External Exposure H-2 H.2 Scrap Yard Worker Processing Scrap—SCRAPCUT H-2 H.2.1 External Exposure H-2 H.2.2 Inhalation of Gaseous or Suspended Radionuclides H-5 H.3 Crane Operator Moving Scrap by Charging Bucket—OP-CRAKE H-5 H.3.1 External Exposure H-5 H.3.2 Inhalation of Fugitive Furnace Emissions H-7 H.4 EAF Furnace Operator—FURNACE H-8 H.4.1 External Exposure H-8 H.4.2 Inhalation of Fugitive Furnace Emissions H-9 H.5 Operator of Continuous Caster—OPCASTER H-9 H.5.1 External Exposure H-10 H.5.2 Inhalation of Fugitive Furnace Emissions H-ll H.6 Baghouse Maintenance Worker—BAGHOUSE H-ll H.6.1 External Exposure H-ll H.6.2 Inhalation of Potentially Contaminated Dust H-15 H.7 Driver Transporting Baghouse Dust—DUSTDRIV H-l6 H.7.1 External Exposure H-16 H.8 Slag Pile Worker—SLAGPILE H-16 H.8.1 External Exposure H-16 H.8.2 Inhalation of Slag Dust H-17 H.9 Construction Worker Using Slag in Road-building—SLAGROAD H-l8 H.9.1 External Exposure H-19 H.9.2 Inhalation of Slag Dust H-20 H. 10 Worker Processing Baghouse Dust at HTMR Facility—DUSTPROC H-20 H. 11 Worker Assembling Automobile Engines—ENGNWRKR H-20 H. 11.1 External Exposure H-21 H-iii ------- Contents page H.I2 Worker Manufacturing Large Industrial Lathes—LATHEMFG H-21 H. 12.1 External Exposure H-21 H.12.2 Inhalation of Contaminated Dust H-22 H. 13 End-user Scenarios H-23 H.I3.1 Consumer Cooking on Large Double Oven—COOKRNGE H-24 H.I 3.2 Driver of Taxi Exposed to Cast Iron Engine Block—TAXIDRVR H-24 H.I3.3 Production Worker Using Large Industrial Lathe—OP-LATHE H-25 H.I3.4 Consumer Cooking in Cast Iron Frying Pan—FEFRYPAN H-25 H.I3.5 Sailor Sleeping next to Hull Plate Made from Contaminated Scrap —HULLPLAT H-26 References H-28 Appendix H-l: Exposure from the Use of Slag in Agriculture Appendix H-2: U.S. Naval Ship Construction H2.1 Time from Steel Production to Operational Status H2-1 H2.2 Size of Plate Steel H2-2 H2.3 Location of the Steel Plate H2-3 References H2-4 H-iv ------- Tables page H-l. Composition of Baghouse Dust H-13 Hl-1. Normalized Annual Doses via Agricultural Slag Pathway HI-2 Figures H-l. Truck Driver MicroShield Geometry H-3 H-2. Closeup View of Pile of Steel Scrap at Scrap Yard H-3 H-3. Conveyer Belt Depositing Light-Gauge Shredded Scrap onto Scrap Pile H-4 H-4. Scrap in Steel Mill H-6 H-5. Crane Operator MicroShield Geometry H-7 H-6. Furnace Operator MicroShield Geometry H-8 H-7. Worker Handling Steel Slabs Produced by a Continuous Caster H-9 H-8. Continuous Caster Operator MicroShield Geometry - Steel Slab H-10 H-9. Continuous Caster Operator MicroShield Geometry - Tundish H-ll H-10. Plan Drawing of Baghouse — Dimensions Are Typical of All Modules H-12 H-ll. Baghouse with Tank Trailer H-13 H-12. Hoppers and Transfer Chutes Under Floor of Baghouse H-14 H-13. The Heil Co., Super Jet Model 1040 Dry Bulk Trailer H-15 H-14. Baghouse Dust Truck Driver MicroShield Geometry H-16 H-15. Slag Used in Road Base Construction MicroShield Geometry H-20 H-16. Auto Engine Assembly Microshield Geometry H-21 H-17. Example of an Industrial Lathe H-22 H-l8. Lathe Manufacture or Operation—MicroShield Geometry H-23 H-19. Range User MicroShield Geometry H-24 H-20. Frying Pan User MicroShield Geometry H-25 H-v ------- DETAILED SCENARIO DESCRIPTIONS FOR CARBON STEEL This appendix presents detailed discussions of some of the assumptions and parameters used in the analysis of the exposure scenarios for the recycling of carbon steel, which are presented in Table 5-1. As shown in Table 5-1, the annual exposure duration of most industrial workers is 1,750 hours, which is based on the observation that workers typically spend seven hours of a nominal eight-hour day in close proximity to the potential source of radiation exposure. Exceptions to this assumption are discussed in the following sections. The external exposure rates calculated by the MicroShield computer program can be converted to effective dose equivalents for photons incident on an anthropomorphic phantom in one of five geometries: • anteroposterior (AP), • posteroanterior (PA), • lateral (LAT), • rotational (ROT), and • isotropic (ISO) Since the AP geometry results in the highest doses and since it is reasonable to believe that workers would spend most of their time facing their work, which is the source of the external exposure, the AP orientation was assumed unless otherwise stated. The scenarios are described in the order in which they are listed in Table 5-1, along with the mnemonic by which they are identified in the summary tables of results which appear in Appendix K. H. 1 DRIVER TRANSPORTING CARBON STEEL SCRAP—SCRPDRVR The truck driver transporting scrap would be exposed to direct penetrating radiation from x- and y-emitting radionuclides in the load of potentially contaminated scrap. He is assumed to spend his full time (40 hours per week, 50 weeks per year) in the cab of a truck. During one-half of this time, the truck carries potentially contaminated scrap; the remainder of the time, the driver would be driving the empty truck back for another load or transporting other cargo. Since he does not come in intimate contact with the material, he would not receive any significant internal exposure. H-l ------- H. 1.1 External Exposure The MicroShield computer program was used to calculate normalized dose rates from external exposure in this scenario. A load of scrap was assumed to weigh 20 tons1 and to have an average bulk density of 1.57 g/cm3, one-fifth the density of steel. The load was modeled as a semi- cylinder—the MicroShield cylinder geometry was used and the results divided by two. Assuming an aspect ratio of cylinder length to diameter of 5:1, the load was calculated to be approximately 30 ft long and 6 ft wide. The driver was assumed to be located in the cab, 8 ft in front of the load. The MicroShield geometry is illustrated in Figure H-l. H.2 SCRAP YARD WORKER PROCESSING SCRAP—SCRAPCUT The RME individual at the scrap processing facility would be the worker who sections oversized pieces of scrap with a cutting torch. This scrap cutter would be exposed to direct, penetrating radiation from x- and y -emitting radionuclides in the potentially contaminated scrap, to inhalation of radionuclides that would be volatilized along with the steel during the cutting process, and to inadvertent ingestion of such nuclides in the particulate matter that is generated from the scrap. Figure H-2 shows a closeup view of a large pile of scrap steel at a scrap yard. Figure H-3 shows a conveyer belt depositing light-gauge shredded scrap onto a large pile. H.2.1 External Exposure As observed by a member of the project team during a visit to a large scrap yard, workers spend time in narrow passages —resembling canyons—between mountainous piles of scrap. Since each wall of the canyon constitutes a half-plane, the two walls together can be conservatively modeled as an infinite plane. The doses from external exposures to such an infinite plane can best be calculated by use of the dose coefficients for exposure to soil contaminated to an infinite depth, which are listed in Table III. 7 of Federal Guidance Report (FGR) No. 12 (Eckerman and Ryman 1993). Data on U.S. industrial practices are normally presented in English units. To present the data as published and to avoid tedious repetition, these values are not generally converted to metric units in this appendix. H-2 ------- Case Title: Truck with 20 tons of scrap - driver - divide by 2 uase X V Z H R Air Gap i itie fQQt 0 37 0 29 2 7 . irucK inches 0 7 0 7 11 12 Ri^c .0 .3 .0 .3 .5 .0 . Ui Side Uieu - Cylinder Uolune - End Shields Figure H-1. Truck Driver Micro Shield Geometry Figure H-2. Closeup View of Pile of Steel Scrap at Scrap Yard. H-3 ------- Figure H-3. Conveyer Belt Depositing Light-Gauge Shredded Scrap onto Scrap Pile This approach yields a conservative but reasonable estimate of the effective dose equivalent (EDE) in the cases of interest. Based on the mass fraction of each element modeled by Eckerman and Ryman (1993, Table II.3), the average atomic number of soil was calculated to be 7.13. Since the atomic number of iron, the chief constituent of steel scrap, is 26, and since the mass attenuation coefficient of energetic photons (x-rays and y-rays) tends to increase with the atomic number of the absorber, it might at first appear that using soil as a surrogate for steel would understate the absorption and thus significantly overstate the external exposure. However, this did not prove to be a significant factor in the present analysis. As shown on pp. J-2 and J-3, there are 14 radionuclides for which external exposure constitutes the dominant pathway. All the principal y-rays2 of these nuclides have energies greater than 400 keV. The mass attenuation coefficient of nitrogen (Z = 7) at this energy is .0295 cm2/g, while Ranked according to total energy (Ey x I = A) of each y-ray from each nuclide. H-4 ------- that of iron is 0.0306 cm2/g, a difference of less than 4%. Given the other uncertainties in the analysis, such a difference is insignificant. H.2.2 Inhalation of Gaseous or Suspended Radionuclides According to a scrap yard superintendent (Schiffman 1996), a scrap cutter spends up to six hours a day actually cutting scrap—the rest of his time is spent going from one yard location to another or waiting for the scrap to be brought to his location. Since the suspended and vaporized contaminants would be produced by the cutting process, the duration of the cutter's exposure via the inhalation and inadvertent ingestion pathways would be up to 1,500 hours per year. The concentration of dust and vapor in the ambient air is based on an experiment conducted at the Idaho National Engineering Laboratory (Newton et al. 1987). Cutting stainless steel pipe with an oxy-acetylene torch in a ventilated enclosure produced average concentrations of respirable particles (0.1 - 10.3 |lm AMAD) of 15 mg/m3. Such a high concentration is unlikely in the worker's breathing zone in an outdoor location. Furthermore, it would be in violation of OSHA PELs, which restrict average total dust loading to 15 mg/m3 and the concentration of respirable particles to 5 mg/m3. (Both values refer to time-weighted average exposures during a 40-hour work week.) However, since the experiment does indicate the potential for the cutting process to generate high dust concentrations, the average concentration of respirable dust was assumed to be equal to the OSHA PEL of 5 mg/m3. H.3 CRANE OPERATOR MOVING SCRAP BY CHARGING BUCKET—OP-CRANE Figure H-4 presents a view of the inside of an EAF shop, with scrap pans on the floor, a load of metal being hoisted by a crane in the right foreground, and a furnace in the background. The diffusion of the light from the overhead lamps shows the high dust loading of the ambient air. H.3.1 External Exposure MicroShield was used to calculate normalized external dose rates to the crane operator. The primary source of external exposure was assumed to be the charging bucket, which was modeled H-5 ------- Figure H-4. Scrap in Steel Mill H-6 ------- as a rectangular solid, 30 ft wide, 12 ft high, and 10 ft long3. Although the bucket has steel walls that are approximately one inch thick, the attenuation of radiation by this additional shielding was conservatively neglected. The crane operator was assumed to be 10 meters from the bucket. Advantage was taken of the symmetry of the source to make the best use of the computation time: instead of modeling the entire source volume with the dose point along the central axis, a source with one-half the width and one-half the height, with the dose point along one edge, was modeled. The calculated results were then multiplied by four to account for the three missing but identical quadrants. The model geometry is depicted in Figure H-5. — L.ase iiTie. ^narging HucKei; Lf. w x jo L Top Uieu X V z L U H _/" ./I Air Gap Side Uieu ) ,_-. v X J.O fQCt 42 0 0 10 15 6 32 -| — J.^1 Ol inches 9.7 0.0 0.0 0.0 0.0 0.0 9.7 size nuii; oy i ^* » I ** ' **^ Figure H-5. Crane Operator MicroShield Geometry H.3.2 Inhalation of Fugitive Furnace Emissions The crane operator inhales air containing fugitive furnace emissions. The average dust loading, 1.3 mg/m3, is based on a report of the measured dust concentration at a crane operator's work station at an operating steel mill. The respirable fraction of fugitive furnace emissions in this and other scenarios is taken from U.S. EPA 1995. Whenever possible, the designation of rectangular dimensions as length, width and height conforms to the convention of the MicroShield program, which always labels the dimension along the X-axis (i.e., towards the dose point) as length. In some cases, as when this dimension is very much smaller than the others, calling it length would be contrary to the conventional understanding of the term. H-7 ------- H.4 EAF FURNACE OPERATOR—FURNACE H.4.1 External Exposure MicroShield was used to calculate normalized external dose rates to the furnace operator. The electric arc furnace (EAF) of the reference steel mill was based partly on the Calumet Steel Co. facility in Chicago Heights, 111., as described in "Electric Arc Furnace Roundup" (1991). This article lists a shell diameter of 12.5 ft. Other dimensions were based on the professional experience and judgement of the project team. The furnace was assumed to have a 2-inch thick steel outer shell and a 6-inch thick inner shell of refractory brick, which was modeled as concrete, one of the built-in MicroShield materials. The source was assumed to be a load of potentially contaminated scrap steel (modeled as iron) which, prior to melting, has an average bulk density 2 g/cm3. Advantage was taken of the symmetry of the source to reduce the computation time: instead of modeling the entire source volume with the dose point in the plane bisecting the cylinder, a cylinder of one-half the height, with the dose point in the plane of the base, was specified. The calculated dose rates were doubled to account for the missing but identical half of the cylinder. Observations of a furnace operator indicated that his distance from the furnace ranged from 4 to 30 ft. Dose rates were calculated at these two distances; the average normalized dose rates, assuming the worker's distance from the furnace varied uniformly over this range, were determined using Equations 6-5 and 6-6. The MicroShield geometry for the nearer distance is shown in Figure H-6. Case Title: EAF - during nelt Side Uieu - Cylinder Uolune - Side Shields X V z H R T1 T2 - Air Gap V feet 1O O O 2 5 O O 3 inches 3.O .O .0 10. 5 7.O 6.O ' 2.O 12.O Figure H-6. Furnace Operator MicroShield Geometry H-8 ------- H.4.2 Inhalation of Fugitive Furnace Emissions The furnace operator would also be exposed to air containing fugitive furnace emissions. The average dust loading, 2.2 mg/m3, is based on a report of the measured dust concentration at a furnace operator's work station at an operating steel mill. H.5 OPERATOR OF CONTINUOUS CASTER—OPCASTER Figure H-7 shows a worker handling steel slabs produced by a continuous caster, which are ready to be shipped out of the plant. The light-colored materials hanging above the worker and the slab are heat-retaining curtains. Figure H-7. Worker Handling Steel Slabs Produced by a Continuous Caster. H-9 Continue ------- Back H.5.1 External Exposure MicroShield was used to calculate normalized external dose rates to the operator of the continuous caster. There are two potential sources of external exposure in this scenario: the bloom—a long steel slab that is produced by the caster—and the molten steel in the tundish that feeds the caster. The dimensions of the bloom—20 ft wide, 3 ft high, and 1 ft long—are based on conversations with James Yusko of the Pennsylvania Department of Environmental Resources and on information gathered by the project team while touring three steelmaking facilities. As discussed in Section H.3.1, the source was represented by one quadrant and the results quadrupled. The model geometry is depicted in Figure H-8. = Case Title: Steel slab - 20' x 3' x 1' - Side Uieu - Rectangular Uolune moaei Lf*\ 01 siao — nun; X V z L w H - Air Gap uy feet 4 O O 1 1O 1 3 i inches 3.4 .O .O .O .O 6.O 3.4 Figure H-8. Continuous Caster Operator MicroShield Geometry - Steel Slab The tundish was modeled as a rectangular solid, 5 ft, 2 inches long; 4 ft, 10 inches wide; and 5 ft, 2 inches high, with a 4-inch-thick inner wall of refractory brick and a 1-inch-thick steel outer wall. As in the case of the furnace, concrete, one of the built-in MicroShield materials, was used as a surrogate for the refractory bricks. As before, the source was represented by one quadrant and the results quadrupled. The model geometry is shown in Figure H-9. Observations of a caster operator indicated that his distance from both the bloom and the tundish ranged from 2 to 15 ft. The average dose rates to this worker were calculated assuming his position varied uniformly over this range, as discussed in Section H.4.1.4 The model geometries shown in Figures H-7 and H-8 are for an intermediate distance. H-10 ------- ------- volume. The composition of the dust, shown in Table H-l, is modeled after that found at a specific steel mill.5 The worker is assumed to be in the central module, marked "0" in the drawing, facing in the direction indicated by the arrow. The modules are separated by H-inch-thick steel walls. The contribution of each module to the external exposure rate was calculated separately, using the dose conversion factors for AP, PA, or LAT geometries, depending on whether the module is in front of, behind, or alongside the worker. Module 0 was modeled with the dust and the Nomex divided into two sources of equal size, with a 12-inch-wide space in the middle for the worker. The exposures from modules 0, a - d, and /' were modeled assuming the worker was in the center of module 0. However, the contributions from modules e - h were calculated assuming the worker was at the wall separating module 0 from module /'. The attenuation due to this wall was modeled assuming the radiation was normally incident on the wall, which results in less attenuation and therefore produces a somewhat more conservative result. 0 - a h 9 h i 15'2" -« 13'5" Figure H-10. Plan Drawing of Baghouse — Dimensions Are Typical of All Modules Exterior Maintenance During the time the baghouse worker is performing outside maintenance and is monitoring the control panels, his external exposure would be from two sources: the residual dust in the baghouse and the tank trailer that is normally parked under the baghouse. Figure H-l 1 shows such a baghouse and tank trailer, while Figure H-l2 shows the machinery external to the baghouse in greater detail. This composition is somewhat different than that listed in Appendix E-2, and is more representative of stainless steel rather than carbon steel melt shops. For the radionuclides of interest, however, the exact composition has a negligible effect on the external dose rates. H-12 ------- Figure H-l 1 . Baghouse with Tank Trailer Table H-l . Composition of Baghouse Dust Compound FeA CaO Cr203 MO ZnO PbO Composition (% weight) 54.5 24.7 10.9 5.9 3.0 1.0 Exposure to Residual Dust in Baghouse. Each filter was modeled as a rectangular solid source, 120 ft, 9 inches long; 30 ft, 4 inches wide; and 30 ft high, elevated 23 ft above ground level. In addition to the residual dust on the filters, an equal amount of dust is assumed to have settled and collected on the floor of each module, which consists of a %-inch-thick steel plate. H-13 ------- Figure H-12. Hoppers and Transfer Chutes Under Floor of Baghouse. This dust would thus form a layer 120 ft, 9 inches long; 30 ft, 4 inches wide; and weighing 12,960 Ib (720 Ib/module x 18 modules = 1,2960 Ib). Since the worker moves around under the baghouse, his exposure was calculated along a line from the center to one corner, using Equations 6-5 and 6-6. The dose point is one meter above ground. Exposure to Tank Trailer. A tank trailer used to collect and transport baghouse dust is normally parked under one side of the baghouse. A description of the trailer was provided by Fellows (1993). The trailer is approximately 29 ft long and 9l/2 ft in diameter—an engineering drawing is shown in Figure H-13. The trailer was modeled as a semi-cylinder with a horizontal axis. Since the trailer arrives empty and leaves when it is full, it was modeled as being half-full on average. The mid-line of the load is 8 ft, 8 inches above ground. The worker's position is assumed to vary uniformly over a range of 1 to 6 m from the truck. The dust has an average bulk density of 57.5 lb/ft3 (0.92 g/cm3). The walls of the tank are aluminum or steel sheet metal, which would not significantly attenuate the radiation from y-emitting radionuclides in the dust. The shielding due to the walls is therefore neglected. H-14 ------- NOTICE : T*NK «BOHT INCLUOtS fUM. HUBS & CENTMFUSE DRUMS WIM 22,5 « 8.25 ALUM. DISC WHEELS * 11R22.S OOOOIfEAR UNISTEtL S1H H-PIT WES, 674.75" (S6'-Z75") 0-A.L 495.25" (4V-3.2S") O.A.L TOAILER 236.5" 4S8.75" (38-2.75") INTERNAL BRIDGE 620,75 (51 -8.75 ) EXTERNAL BRIDGE Figure H-13. The Heil Co., Super Jet Model 1040 Dry Bulk Trailer Steel Mill Duties Not Involving the Baghouse Except on the days that he performs interior maintenance and during the one hour per day he spends on exterior maintenance, the baghouse worker performs other duties inside the mill. Since no particular mill worker is assigned to baghouse maintenance, the baghouse worker, during the time spent away from the baghouse, is assumed to have the same exposure rate as the crane operator, one of the three mill workers modeled. H.6.2 Inhalation of Potentially Contaminated Dust While inside the baghouse, the worker is exposed to dust concentrations estimated to be 40 mg/m3, with a respirable fraction of 0.76 (U.S. EPA 1995). He wears a respirator equipped with a full facepiece, which has a rated filter efficiency of 99% (10 CFR 20, Appendix A). While monitoring the controls and performing maintenance outside the baghouse, he is exposed to an atmospheric dust loading of 1.2 mg/m3, which is the reported dust concentration for a baghouse maintenance worker at an operating steel mill. While he performs duties away from the baghouse, the dust loading at his work station is assumed to be the average of the reported concentrations at nine other work stations at an operating steel mill. H-15 ------- H.7 DRIVER TRANSPORTING BAGHOUSE DUST—DUSTDRIV Since the truck driver transporting baghouse dust does not come in direct contact with the dust, his only significant exposure would be to direct penetrating radiation from the potentially contaminated dust inside the trailer. H.7.1 External Exposure MicroShield was used to calculate normalized external dose rates to this worker. The load was modeled as described in Section H.6.1, above. The position of the driver in the cab was scaled from the engineering drawing shown in Figure H-13 and determined to be 11 ft 4l/2 inches in front of the load. The model geometry is shown in Figure H-14. X Y Z H R Air Gap feet O 40 O 29 4 11 = ^ase inches .O 7.6 .O 3.3 9.1 4.4 Case Title: Bag-House Dust: Cab of truck Side Uieu - Cylinder Uolune - End Shields Figure H-14. Baghouse Dust Truck Driver MicroShield Geometry H. 8 SLAG PILE WORKER—SLAGPILE H.8.1 External Exposure The external exposure to the slag pile worker was assessed using the FGR 12 dose coefficients, as discussed in Sections 6.3.1 and H.2.1. Since the worker was assumed to stand at the edge of the slag pile, his rate of exposure would be one-half of what it would be in the center. H-16 ------- H.8.2 Inhalation of Slag Dust The atmospheric dust concentration was estimated on the basis of actual field measurements performed as part of an EPA-sponsored study of fugitive emissions from slag loading operations (Bohn et al. 1978). In order to determine the emissions due to the loading operation, the investigators placed air samplers upwind from the emission source to determine the background concentration—i.e., dust concentrations in the air that are not attributable to the activity being monitored. Six background dust concentration measurements were performed at a slag plant attached to a steel mill. The readings ranged from 0.5 to 3.2 mg/m3, with an average of 2.6 mg/m3. These measurements were made using a high-volume air sampler which is not sensitive to particles larger than about 30 |im. For the purpose of the exposure assessment, it is necessary to derive the concentration of respirable particles (AMAD < 10 |lm). Although Bohn et al. (1978) do not present such data directly, the report shows that the ratio of particles with mass median diameters < 5 |lm to particles < 30 |lm varies from 0.27 to 0.31, with an average value of 0.29. U.S. EPA 1995 presents a more detailed distribution of aerodynamic diameters for fugitive emissions from aggregate piles; these data were combined with the data reported by Bohn et al. to calculate the respirable fraction of slag dust as follows: F ( F p. _ rio.E r5, r!0,B p p r5,E V r30,B> FIO,B = respirable fraction of fugitive dust, based on Bohn 78 = 0.51 FIO,E = respirable fraction of fugitive dust, reported in EPA 95 = 0.35 F5E = fraction of particles, AD < 5 |lm, reported in EPA 95 = 0.20 F5B —'— = average ratio of F5 to F30 reported in Bohn 78 = 0.29 H-17 ------- H.9 CONSTRUCTION WORKER USING SLAG IN ROAD-BUILDING—SLAGROAD The exposure time of the road construction worker depends on the fraction of potentially contaminated slag that is used in road construction. This, in turn, depends on the rate of road construction and the rate of slag production at the reference steel mill. R.S. Means 1997, a standard reference for contractors, states that a road construction crew laying down a 300-mm (~ 1-foot) -deep pavement base of 40-mm crushed stone has a production rate of 1,505 m2 per day. A crew laying down 100-mm (~4-inch) -thick asphaltic concrete has a rate of 3,462 m2 per day. Assuming that the same crew lays down both the pavement base and concrete, the area of road produced in a day can be determined as follows: A = R,,x = Rc(l - x) (H-l) A = Rate of road production (m2/d) Rb = Production rate of road base = 1,505 m2/d Rc = Production rate of concrete pavement = 3,462 m2/d x = fraction of day spent laying down road base Solving the second equation for x, we find Substituting this expression in the first of Equations H-l, we obtain R»,R, A = ^ c R* + RC = 1,049 m2/d The quantity of slag used per day can now be readily determined: M = A(dcfc+ H-18 ------- M = rate of slag utilization = 797.2 t/d = 878.8 tons/day dc = thickness of concrete = 0.1 m fc = fraction of slag in asphaltic concrete = 0.8 db = thickness of road base = 0.3m p = bulk density of slag = 2 g/cm3 Since the reference steel mill has a melting capacity of 150,000 tons of steel per year, and since the mass fraction of slag, as listed in Section 6.2, is 0.117, the production rate of slag is 17,550 tons per year, or enough for about 20 days of road construction. Assuming an exposure duration of 7 hours per day, the road workers would be exposed for 140 hours per year. H.9.1 External Exposure MicroShield was used to calculate normalized external dose rates to a worker using slag in road construction. This worker is assumed to be exposed to two primary sources: slag used in the road base and slag used as an aggregate in the concrete paving. The road was modeled as a rectangular solid source 4,000 m long (infinitely long), 36 ft wide, and 6 inches thick, with a 1- foot-thick concrete cover6. The worker was assumed to be standing in the center of the road, the dose point being one meter above the surface. As discussed in Section H.3.1, the source was represented by one quadrant and the results quadrupled. The model geometry of the road base is depicted in Figure H-15. Because of its thickness, density, and area, the exposure rate from the concrete would not differ significantly from that of soil contaminated to an infinite depth. The external exposure from the concrete was therefore assessed using the FOR 12 dose coefficients, as discussed in Sections 6.3.1 and H.2.1. The calculated dose rates were multiplied by fc, the fraction of slag in asphaltic concrete. These dimensions are taken from SCA 1993. H-19 ------- Case Title: Road Bed - ••X- Side View Top Uieu Rectangular Uolune X V z L w H T1 Air Gap feet 4 O O O 6561 18 1 3 inches 9.4 .O .O 6.O 8.2 .O .O 3.4 Figure H-15. Slag Used in Road Base Construction Micro Shield Geometry H.9.2 Inhalation of Slag Dust The road construction workers were assumed to be exposed to the same dust concentrations as the slag pile workers, as described in Section H.8.2. H. 10 WORKER PROCESSING BAGHOUSE DUST AT HTMR FACILITY—DUSTPROC The external exposure of the EAF dust processing worker was computed using FGR 12 dose coefficients for external exposure to soil contaminated to an infinite depth—the same as for the slag yard worker described in Section H.8.1. As was the case for the slag pile, the pile of dust was assumed to be a half-infinite plane: the FGR 12 dose rate was divided by 2. In the absence of specific data, it was assumed that the concentration of respirable dust in the ambient air was equal to the OSHA limit of 5 mg/m3. H. 11 WORKER ASSEMBLING AUTOMOBILE ENGINES—ENGNWRKR Because of his close proximity to a large mass of potentially contaminated metal, a worker assembling V-8 engine blocks was selected as the maximally exposed automobile worker. Since there is little opportunity for particulate matter to evolve from this operation, the only significant exposure pathway of this worker would be direct penetrating radiation from the cast iron block. H-20 ------- H. 11.1 External Exposure MicroShield was used to calculate normalized external dose rates to an automobile engine assembler. The weight and dimensions of a typical V-8 engine were obtained from ADK, the engine rebuilder that formerly supplied engines to Sears, Roebuck and Co. The shipping weight of the engine is 350 Ib; the crate itself weighs about 5 Ib and has overall dimensions of 2 x 2 x 21A ft. Assuming that the crate is one-half inch thick, the engine dimensions would be 23 x 23 x 29 inches. The weight was divided by the volume to obtain an average density of 0.632 g/cm3. Since the worker would be moving back and forth while performing this task, dose rates were calculated at distances of 20 cm and 70 cm from the source. The average dose rates between these two distances were calculated using Equations 6-5 and 6-6. The model geometry for an intermediate distance is depicted in Figure H-16. Case Title: Car engine Side Uieu — Rectangular Uolune X V z L w H Air Gap feet 2 1 O 1 1 2 1 inches 11. 0 2.5 11.5 11.0 11. 0 5.O .0 Figure H-16. Auto Engine Assembly Microshield Geometry H. 12 WORKER MANUFACTURING LARGE INDUSTRIAL LATHES—LATHEMFG A large industrial lathe is illustrated in Figure H-17. H.12.1 External Exposure MicroShield was used to calculate normalized external dose rates to a worker manufacturing large industrial lathes. A large lathe observed in a commercial machine shop weighed 8 tons. The lathe bed, which would comprise most of this mass, was 3 ft wide and 1 ft thick. Assuming the bed contained all of the mass, it was calculated to be approximately 11 ft long and was modeled as a H-21 ------- A Figure H-17. Example of an Industrial Lathe rectangular solid. As discussed in Section H.3.1, the source was represented by one quadrant and the results quadrupled. Since the worker would be moving back and forth while performing this task, dose rates were calculated at distances of 20 cm and 70 cm from the source. The average dose rates between these two distances were calculated using Equations 6-5 and 6-6. The model geometry for the 20-cm distance is depicted in Figure H-18. H.12.2 Inhalation of Contaminated Dust The grinding of the lathe bed could produce airborne dust. Newton et al. (1987) report that cutting metal with a side-arm grinder in a ventilated enclosure produced dust concentrations averaging 2.7 mg/m3. This value was adopted for assessing the inhalation exposure of the lathe manufacturing worker. H-22 ------- X Y Z L W H Air Gap feet 3 O O 3 O 5 O Top Uieu inches 7.9 .O .O .O 6.O 6.O 7.9 " 1 Side Uieu V -JP V h^ rt • ' Figure H-18. Lathe Manufacture or Operation—Micro Shield Geometry H.13 END-USER SCENARIOS The scenarios describing the exposures of the end users of finished products have several features in common. First, the RME individual is assumed to use a product made from a furnace charge containing the maximum likely fraction of potentially contaminated scrap metal, as described in Section GAS. While it is implausible that a lathe fabricator, for instance, would be exposed during an entire year to cast iron that was made from a maximally contaminated furnace charge, it is reasonable to believe that at least one lathe made from such metal could be produced. Since the lathe operator could be assigned to the same machine for one year, he would be exposed to such a source during this time. The same is true for the other products, all of which have useful lives of more than one year. The second distinguishing feature of the end-user scenarios is that, since the user would have the same product for at least a year, the radionuclides would be decaying during this time. Consequently, Equation 6-9 in Section 6.3.4, which explicitly accounts for radioactive decay, is used to calculate the dose during that year. Finally, since no significant erosion of the metal in the finished product is expected in normal use, there are no significant internal exposure pathways, except for the potential contaminants leached from the cast iron frying pan. H-23 ------- H.13.1 Consumer Cooking on Large Double Oven—COOKRNGE MicroShield was used to calculate normalized external dose rates to a user of a large kitchen range, modeled after a Sears Kenmore 30-inch double oven, model No. 78509. It is 66 inches high, 29 inches wide, 28 inches deep, and weighs 284 Ib; the average density of 0.1417 g/cm3 was calculated by dividing the weight by the volume. The dose point is 2 ft in front of the source. The model geometry is depicted in Figure H-19. ------- Fromberg estimated that the owner/driver might work 6 days per week, for a total of 68 to 72 hrs per week. In the present analysis, it was assumed that he works 12 hrs/day but spends 1 hr/day on rest and meal stops. His annual exposure is therefore 11 hrs/day x 6 days/week x 50 weeks/y = 3,300 hrs/y. H.13.3 Production Worker Using Large Industrial Lathe—OP-LATHE The normalized external dose rates to the operator of a large industrial lathe are calculated using the same geometry as described in Section H. 12 for the lathe manufacturing worker. H. 13.4 Consumer Cooking in Cast Iron Frying Pan—FEFRYPAN The MicroShield program was used to calculate normalized external dose rate to a person cooking with a cast iron frying pan. The pan was modeled as a flat disc, 11.8 inches in diameter and weighing about 6 Ib. The dose point is 2 ft from the edge of the pan. The model geometry is depicted in Figure H-20. X Y Z H R Air Gap feet 2.O O.O O.O 0.0 O.O 2.O I — L ^^^i inches 5.9 6.O .O .2 5.9 .O ____ ^-"s. c • I 1 i V 1 1 . T + -* X^_. ^ Figure H-20. Frying Pan User MicroShield Geometry H-25 ------- H.13.5 Sailor Sleeping next to Hull Plate Made from Contaminated Scrap—HULLPLAT8 A typical hull plate is estimated to be 20 ft wide, 10 ft high, and % inch thick, and to weigh about V/2 tons. Such a plate could be made from a single heat in an EAF. Enlisted men's bunks are found immediately next to the hull. A sailor sleeping in a bunk next to the center of the hull plate is assumed to spend 250 days per year aboard ship. Of this time, he would spend 8 hr a day sleeping or relaxing on his bunk, for a total exposure time of 2,000 hours per year9. His average distance from the hull plate is assumed to be about one-half the width of the bunk, or about 18 inches. (Because the lateral dimensions of the plate are large with respect to the distance, a small variation in the distance will have little effect on the external exposure rate.) Since the sailor's orientation to the plate would vary, the ROT geometry was assumed. A minimum of 18 months is estimated to elapse from the time the steel plate is fabricated to the time the sailor begins to occupy his berth. The initial activities in the steel were therefore reduced by radioactive decay during this period. As mentioned in Section 6.3.1, the external exposure to Mo-93 in the hull plate was calculated in a different manner. This exposure rate was estimated by multiplying the exposure rate of Co-60—a strong y emitter—by the ratio of the FGR 12 dose coefficient of Mo-93 to that of Co-60 for a soil layer with a similar mass thickness10. The mass thickness of the %-inch thick steel plate is equal to 7.5 g/cm2 (0.375 inch x 2.54 cm/inch x 7.86 g/cm3 = 7.5 g/cm2). A 5-cm thick layer of soil, as modeled in FGR 12, has a mass thickness of 8 g/cm2 (5 cm x 1.6 g/cm3 = 8 g/cm2), not a significant difference. The exposure rate from Mo-93 in the steel plate was calculated as follows: E F TJ _ ^Co-60 r5,Mb-93 ^^00-93 p, F 5, Co-60 EMo-93 = normalized dose rate from Mo-93 in steel plate The information in this section is based on a report by Cdr. J. Harrop, USN (ret.). A more complete version of this report appears in Appendix H-2. Some sailors might spend more time on board ship and more of their off-duty hours on or in the vicinity of their bunks. Given the other conservative assumptions in this scenario, a less conservative (i.e., shorter) exposure duration was assumed. See discussion of the applicability of FGR 12 dose coefficients to contaminated steel in Section H.2.1. H-26 ------- ECo-6o = normalized dose rate from Co-60 in steel plate, as calculated by Micro Shield, using conversion coefficients for ROT geometry FS, Mo-93 = dose coefficients for exposure to soil contaminated with Mo-93 to a depth of 5 cm (Eckerman and Ryman 1993) FS, co-60 = dose coefficients for exposure to soil contaminated with Co-60 to a depth of 5 cm (Eckerman and Ryman 1993) H-27 ------- REFERENCES Bohn, R., T. Cuscino, and C. Cowherd. 1978. "Fugitive Emissions from Integrated Iron and Steel Plants," EPA-600/2-78-050. U.S. Environmental Protection Agency, Office of Research and Development, Washington, DC. Eckerman, K. F., and J. C. Ryman. 1993. "External Exposure to Radionuclides in Air, Water, and Soil," Federal Guidance Report No. 12, EPA 402-R-93-081. U.S. Environmental Protection Agency, Washington, DC. "Electric Arc Furnace Roundup - USA." 1991. Iron and Steel Maker, May, 1991. Fellows, D. (The Heil Company, Midwest Region). 1993. Private communication. Fromberg, A. (Asst. Commissioner for Public Affairs, City of New York Taxi & Limousine Commission). 1998. Private communication (19 March 1998). Newton, G. J., et al. 1987. "Collection and Characterization of Aerosols from Metal Cutting Techniques Typically Used in Decommissioning Nuclear Facilities." American Industrial Hygiene J. 48:922-932. R. S. Means Company. 1997. Means Heavy Construction Cost Data, Metric Version. R. S. Means Company. S. Cohen & Associates (SCA) and Rogers & Associates Engineering. 1993. "Diffuse NORM Waste: Waste Characterization and Preliminary Risk Assessment." Prepared for U.S. Environmental Protection Agency. Schiffman, W. (Supervisor, Tube City, Inc.). 1996. Private communication. U.S. Environmental Protection Agency (U.S. EPA), Office of Air Quality Planning and Standards. 1995. "Compilation of Air Pollutant Emission Factors," AP-42, 5th ed. Vol. 1, "Stationary Point and Area Sources." U.S. EPA, Research Triangle Park, NC. H-28 ------- APPENDIX H-l EXPOSURE FROM THE USE OF SLAG IN AGRICULTURE ------- EXPOSURE FROM THE USE OF SLAG IN AGRICULTURE This appendix presents a scoping calculation used to assess the potential dose to a subsistence farmer using slag as an agricultural conditioner. Because of its high lime content (up to 50%), slag can be used as a soil conditioner. In common practice, 50 to 100 Ib of lime is applied to 1000 ft2 of soil for pH adjustment. Assuming a plow depth of 15 cm and a soil density of 1.6 g/cm3, the normalized dose to the RME individual via this pathway can be approximated as follows: Dia = normalized dose from radionuclide /' via the slag agricultural pathway (mrem/y per pCi/g in scrap) cig = concentration factor of radionuclide /'in slag (see Table 6-3) Dis = normalized dose from radionuclide /' via the soil agricultural pathway (mrem/y per pCi/g in soil— U.S. EPA 1994, Table 3-1) fc = dilution factor for potentially contaminated scrap = 0.055 fgs = fraction of slag in soil (by weight) mg = mass of slag = 100 Ib = 4.54xlQ4g A = 1000ft2 = 9.29 x 105 cm2 ps = soil density = 1.6 g/cm3 ds = plow depth of soil layer = 15 cm The results for radionuclides that concentrate in the slag are presented in the following table. The column headings correspond to the terms defined above. Hl-1 ------- Table Hl-1. Normalized Annual Doses via Agricultural Slag Pathway (mrem/y per pCi/g) Radionuclide Sr-90 Nb-94 Ce-144 Pm-147 Eu-152 Ra-226 Ra-228 Th-228 Th-229 Th-230 Th-232 Pa-231 U-234 U-235 U-238 Np-237 Pu-239 Am-241 Cm-244 Dis 5 Dia 4.31e-03 negligible 0.012 3.e-04 1.04e-05 O.OOe+00 negligible 4.35 1.6 0.07 0.014 1.5 2.1 5.1 0.16 0.12 0.16 9.6 0.7 0.08 0.21 3.75e-03 1.38e-03 6.04e-05 1.21e-05 1.29e-03 1.81e-03 4.40e-03 1.38e-04 1.04e-04 1.38e-04 8.28e-03 6.04e-04 6.90e-05 1.81e-04 Table 7-1 2.10e-02 2.35e-01 8.59e-03 6.25e-05 1.70e-01 3.00e-01 1.72e-01 6.61e-01 2.13e+00 3.13e-01 1.38e+00 1.16e+00 1.55e-01 1.62e-01 1.43e-01 7.23e-01 3.61e-01 5.73e-01 3.19e-01 These results show that the reasonable maximum dose via the agricultural slag pathway is a small fraction of the dose to the RME individual for each of the nuclides listed. REFERENCE U.S. Environmental Protection Agency (U.S. EPA), Office of Radiation and Indoor Air. 1994. "Radiation Site Cleanup Regulations: Technical Support Document for the Development of Radionuclide Cleanup Levels for Soil," Review Draft, EPA 402-R-96-011 A. U.S. EPA, Washington, DC 20460. Hl-2 ------- APPENDIX H-2 U.S. NAVAL SHIP CONSTRUCTION ------- U.S. NAVAL SHIP CONSTRUCTION1 This report addresses the issue of how U.S. Navy ship construction might incorporate residually radioactive steel scrap. In particular, the report addresses: • where in the ship's structure might the steel be used; • what size plate steel is used; and • what the time frame is from ordering steel from a mill until the ship using the steel is operational. The discussion that follows only addresses U.S. Navy non-combatant vessels, such as oilers and supply ships. Nothing in this report should be construed to apply to warship construction. In a U.S. Navy ship, the RME individual may be a sailor who lives (berths) on the ship, and berths in a compartment having a bulkhead or hull plating manufactured from steel that contains residual radioactive contamination. Such a sailor may be assigned to a ship for three years, and may spend as many as 6,000 hours (250 days x 24 hours/day) onboard the ship. Of the 6,000 hours, as many as 2,000 may be spent in the berthing compartment.2 U.S. Navy ships may be built with steel plate and structural members produced from steel mills with EAFs. H2.1 TIME FROM STEEL PRODUCTION TO OPERATIONAL STATUS Shipyards generally procure steel plate as needed, and do not stockpile large quantities. Thus, the storage time in a yard is short, on the order of days. A non-combatant may take 18 months from keel laying to delivery. Steel plate is used almost immediately, but since the ship is built from the keel up, a few months may pass before the steel plate to be used in a berthing compartment (which typically is above the waterline) arrives at the shipyard (Mancini 1999). Adapted from a report prepared by Cdr. John Harrop, USN (ret.). See Note 9 on page H-26. H2-1 ------- Once delivered, the crew will man the ship, and begin living on the ship. However, a shakedown period of several months is typically required before the ship is in a full operational status. Thus, the steel plate used in the berthing compartment will be formed (poured and rolled) about 18 months before the sailor begins berthing in the berthing compartment. H2.2 SIZE OF PLATE STEEL The size of the plate steel delivered for ship construction depends on the size of the steel mill equipment and the size of equipment at the shipyard. All new ship construction is completed at commercial shipyards. These shipyards have the capacity to handle anything made by an EAF mill. The mill will pour an ingot, or use a continuous cast system. Plates are produced by rolling in a rolling mill. A typical size plate delivered to a shipyard is 10 ft by 20 ft. The shipyard will size the plate upon delivery, according to specific construction needs. Plates may be delivered in various thicknesses (Mancini 1999). Hull (or shell) plates on the ship may be the full 10 by 20 ft. Hull plate thickness varies, and is determined by a standard formula (American Bureau of Shipping 1997): D. Typical parameter values, used for the present analysis, are listed below t = minimum thickness = 8.4 mm = 0.33 inch t0 = 2.5 mm s = spacing of transverse frames or longitudinals = 3 ft = 0.914m s0 = 0.645 m L = vessel length = 565 ft = 172.2 m (Jane's Fighting Ships 1988) L0 = 15.2m d = molded draft = 15 ft = 4.57 m (Jane's Fighting Ships 1988) H2-2 ------- Ds = molded depth = 40 ft = 12.2 m (Jane's Fighting Ships 1988) Note that this represents the minimum required thicknesses. A more typical thickness is % inch (Mancini 1999). H2.3 LOCATION OF THE STEEL PLATE Steel plate may be used for hull (shell) plating, or to form bulkheads. Steel plate of the dimensions listed above may be used as hull plating near midships berthing compartments. Smaller size (length and width) plating may be used near the bow and the stern. H2-3 ------- REFERENCES Mancini, A. (Captain, U.S. Naval Sea Systems Command). 1999. Private communication (19 May 1999). American Bureau of Shipping. 1997. "Rules for Building and Classing Steel Vessels 1997." Part 3, "Hull Construction and Equipment," p. 15-1. American Bureau of Shipping, New York, Jane's Fighting Ships. 1988. Jane's Information Group, Inc. H2-4 ------- APPENDIX I LEACHING OF RADIONUCLIDES FROM SLAGS ------- Contents page I.I Steel Slags—Background Data 1-1 1.2 Slag Cement Leaching Studies 1-2 1.2.1 Strontium-90 1-4 1.2.2 Cobalt-60 1-5 1.2.3 Tritium 1-5 1.3 Slag Leaching Studies 1-5 1.4 Possible Modeling Approach 1-15 References 1-17 Appendix 1-1. Results of Leach Rate Study Performed by Brookhaven National Laboratory Tables 1-1. Fraction of Various Toxic Elements Leached from Slags Using EPA TCLP Protocol ... 1-7 1-2. Constituents Leached from Slag 2 1-8 1-3. Blast Furnace Slag Solubility Data 1-12 1-4. Constituents of Concern in Steel Furnace Slag Leachates 1-13 1-5. Nominal Compositions (wt%) of Slag Mixtures Studied by de Villiers (1995) 1-14 1-6. Variation in the Concentration of Elements Leached from Slags 1 and 3 in SPLP Solutions 1-14 1-7. Comparison of Corps of Engineers and de Villiers Leaching Data 1-15 Figures 1-1. Weathering of Slag 2: Ca and K 1-9 1-2. Weathering of Slag 2: Ba, Cr, and Mn 1-10 I-iii ------- LEACHING OF RADIONUCLIDES FROM SLAGS As described in Appendix E, a number of radionuclides are expected to partition strongly to the slag during the electric arc furnace (EAF) melting of contaminated carbon steel scrap. Typically, this slag is stored at the steel mill for as long as several months before disposal. During storage and use (or disposal)