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)