TECHNICAL SUPPORT DOCUMENT

         POTENTIAL RECYCLING OF SCRAP METAL
                 FROM NUCLEAR FACILITIES

          PART I:  RADIOLOGICAL ASSESSMENT OF
                   EXPOSED INDIVIDUALS

                    Volume 2: Appendices A-F
                          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

-------
                     APPENDIX A




SCRAP METAL INVENTORIES AT U.S. NUCLEAR POWER PLANTS

-------
                                       Contents
                                                                                   page

A. I  Introduction  	 A-1

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-1 I
       A.3.1.2  Reference PWR			  A-1 I
   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 PlansA-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.x5.l.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	  A1-1
   Reference	  A1-6
                                         A-iv

-------
                                        Tables
                                                                                 page

A-1. Sources of Residual Activities in Reference BWR and PWR 	  A-1 I
A-2. Estimated Activities of Neutron-Activated Reactor Components in a BWR ..........  A-12
A-3. Neutron-Activated Reactor Components in a PWR	 ,x	  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-1 7
A-8. Internal Contamination Levels of Big Point Nuclear Plant at Shutdown . :	  A-18
A-9. Plant Systems Radioactivity Levels at SONGS I  	/	  A-19
A-10. Average Internal Contamination Levels of Reactor Systems at Yankee Rowe	  A-20
A-ll. 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-15. Summary of Contamination Levels in BWR Equipment	  A-25
A-16. Estimated Internal Surface Activities in BWR Systems  	  A-25
A-1 7. 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-3 I
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 M We PWR Facility	  A-4 I
A-30. Breakdown of Materials Used in PWR Plant Structures and Reactor Systems	  A-42
A-3 I. Inventories of Ferrous Metals Used to Construct a 1,000-M We 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 Poof Cooling and Cleanup System	  A-50
A-36. High Pressure Core Spray System  	  A-50
A-37. HVAC Components System	  A-5 I
A-38. Low Pressure Core Spray System	  A-5 I
A-39. Main Steam System  	  A-52
                                         A-v

-------
                                 Tables (continued)
                                                                                page

A-40. Main Steam Leakage Control System  	  A-53
A-4 I. 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	O . . ,	  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-5 I. Radioactive Floor Drain Processing System  	  A-61
A-52. Rad Waste Building Drains System	.x.	  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
                                       /N
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. Residua! 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 PWR Non-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

A I -1. Nuclear Power Reactors Currently  Licensed to Operate 	,^	  A I -2
A I -2. Formerly Licensed Nuclear Power  Reactors 	'.../•	  A I -6
                                        Figures

A-1.  Pressurized Water Reactor	  A-5
A-2.  Boiling Water Reactor  	:".	  A-8
A-3.  Reactor Coolant System in a Four-Loop PWR	.;.  . ,7	  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 200Q).  In the
next three decades, most of the operating licenses of reactors currently in operation—originally
valid for 40 years—will have expired.'                                     /•
                                                                     \
                                                                    \ "   '

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 io prepare and submit
plans and cost estimates for decommissioning their facilities to thelMRC 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
                                          /N
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 staled in Chapter 2. the NKC has issued a rule allou 11152 a licensee to appK tor a 2(>-\ ear reneual of its original
operating license  I o date. ti\e reactors ha\e been granted such license reneuals. a number ot'other reneual
applications are pending, and more applications are anticipated

                                           A-1

-------
     •   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 toxlry 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 GFR Pan 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^lso 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 usfe 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 Rkely  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 I07 dpm/100 cnr at 10 years post-shutdown 10 10s
dpm/100 cnr (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 me 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.
                                         /N
Residual  radioactive contaminants of reactor components/systems and building structures is
generally grouped as:  (I) activation products that are distributed volumetrically, (2) activation
and fission products in the form of corrosion tllms.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,  35 are BWRs manufactured by General Electric and 69 are PWRs manufactured by
Westjnghouse, Combustion  Engineering and Babcock and Wilcox (U.S. NRC 2000).
                                          A-3

-------
Appendix A-1 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 x
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-61 74
(Smith et al.  1996).  The four NUREG reports cited above, along with several other NRC reports
                                        /N
and selected decommissioning plans on tile 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, 201 I.   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-1. 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
                                                   Chemical
                                                   Volume
                                                   & Control
                                                   System
                                                            Feed water
                                                              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  buildin«, 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 (I) 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 pan 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  19 m 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 diesei 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  MVYe) 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 et al. 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 rooting 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
rooting. 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 condensef 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 rooting. This structure is approximately 64,by 49 m in  plan, 32  m in overall
height, and stands as two full floors and one partial  tloor aboye the ground tloor.

A.3  RESIDUAL ACTIVITIES IN REFERENCE REACTOR FACILITIES
                                         /N
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:

     I. 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 I 5 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 biolo'gical shield.
                                                             \
Total quantities and the relative radionuclide composition of the residual activity are not only
                                                       N
affected by reactor design (BWR vs. PWR) but are also strongly influenced by numerous other
factors including (I) fuel  integrity, (2) rated generating capacity and  total years of operation, (3)
                                          /N
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-1).
                                        \
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 (I) the concentration of the parent nuclide and its
                                           A-IO

-------
neutron cross-section, (2) the radioactive half-life of the radionuclide, (3) the neutron tlux
intensity at the given location, and (4) the duration of neutron exposure.
            Table A-1.  Sources of Residual Activities in Reference BWR and PWR
Source
Activated Stainless Steel
Activated Carbon Steel
Activated Structural Components, Rebar, Metal Plates, 1-Beams
Internal Surface Contamination of Piping and Equipment
External Contamination of Equipment, Floors, Walls, Other Surfaces
Residual Activity (Ci)
BWRa
6.6e+t)6
2,9e^03
l.2e+03
, 8,Se+03
l.le+02
PWR"
4.8e+06
2.4e+03
1 .2e+03
4.8e+03
l.le+02^
 Oakctal  I WO
 Smith ct al IV7X
" Implied \aluc (I I S NKC W4)
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 feed water, 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 weiuhs  about 400 t.
                                           A-l I

-------
     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 ( 1 85)
Total
Average
Activity Concentration
(Ci/m-')
l.68e+06
2.62e+04

l.07e+03
I.l2e+02

l.03e+04-
2.53e+03
9.7le+04 -x
l.0le+0l •,
2.56e+02
7.67e+tf5
5.116+05
2.l6e+02

Total Activity
(Ci)
6.30e+06.
2.0Qe+03 N
2.16^+03'
\
\ " '
' 9.60e+03
/
3.0le+04
7.0le+02
6.50e+02
I.IOe+04
l.78e+05
9.47e+OI
6.55e+06
         Source  (>;ik et ;il  19X0

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 190 t.

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 ( =  )0 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 4.7m 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/m!)
2.97e+06
3.07e+05
l.45e+05
7.73e+03
9.04e+02
4.20e+04
I.l2e+06

Total
Activity
(Ci)
3.43e+06.
6.52e+X)5
1 .46e+05
l.50e+03
l.76e+04
2.43e+04
5'.53e+05
4.82e+06
              Source Smith ct al IV7X
                                                        \
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.l6e+06
2.37e+04
8.30e+02
4.l9e+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 tar 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 (iCi/g.)  Furthermore, these components also exhibit high levels of interior surface
contamination.  While surface contamination is potentially removable, the volurnetrically
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
                                         /N
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,. I 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 residua! inventories and distributions of long-lived radionuclides following  plant shutdown
(Abe! et al. 1986).  Residual concentrations in  the various plant systems decreased in the

                                          A-14

-------
following order:  (I) 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 I.

     Table A-5. Residual Activities and Operating Parameters of Six Nuclear Power Plants"
Stations
Humboldt Bay
Dresden- 1
Monticello
Indian Point-l
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
1 1
8.3 ,
8.8
Power Rating x
(MWe) "
63
210
, 550
J70
; 660
935
-Reactor Type
BWR
BWR
BWR
PWR
PWR
PWR
Source  Abel et al I Wf,
                                           s\
  I otal nnentoiA  includes radionuclides uith half-lues greater than 2-45 da\ s (i e . /.n-fO). inventories in actuated metal
  components of the reactor pressure \essel and internals arkl actuated 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:  (I) 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 onJy the radioactive contamination of corrosion film and crud: 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 uear products (rust particles, etc ) that become radioactive (i e . actuated)
when exposed to radiation  I lie term is actualh an acroin m for Chalk Kuer I Imdentified I )eposits. the Canadian plant at
•which'the actuated deposits uere first discovered
                                            A-1 5

-------
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-24 I, Cm-242 and Cm-
244) constituted varying percentages of the total  inventory, ranging from 0.00]% atxRancho Seco
to 0.1% at Dresden-1.
                                                                            /•
   Table A-6.  Relative Activities of Lonu-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- 1 1 Om
1-129
Cs-137
Ce-144
TRIT
Total (Ci)
Relative Activity, Decay-Corrected to Shutdown Date (%)'
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 v
< 3e-03
4e-05
— " •
< fe-05
0.04
1
O.I
2,350
Monticello
1
1
—
,-n
—
' 0.04.
84
2e-03
<0.l
8e-05
—
< le-06
2
—
8e-03
448
TnVRs
Indian x
Poiqt-l
• ,4 '
, 67
—
15
0.02
2
1 1
7e-04
8e-04
8e-05
—
2e-05
0.5
—
2e-03
1,070
Turkey
Point-3
0.4
31
43
24
4e-03
O.I
1
8e-04
< 4e-03
8e-03
—
< 3e-03
—
0.2
6e-03
2,580
Rancho Seco
4
28
24
18
O.I
19
0.09

 l-Acludes actuated metal components of the reactor pressure \essel and internals and actuated concrete
 Relative uctn itv  oi'each nuclide as a percentage ot'total actn it\ at each pouer plant
  I ninsuranic alpha-emitting radionuclides v\itli halt-lues greater than 5 \ears. including l'u-2.
-------
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

-------
      Table A-8. Internal Contamination Levels of Bi» 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/IOOcnr)
3e+IO
9e+09
3e+09
4e+08
4e+08
4e+08
5e+07, .
3e+07
3e+07
6e+06
9e+OS
7e+05
3e+05
San Onofre Nuclear Generation Station Unit 1  (SONGS 1)
SONGS I  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 I  was permanently discontinued
on November 30, 1992 at the end of fuej cycle #11.  A preliminary decommissioning plan,
submitted  to the NRC on December I, 1992, proposed to maintain SONGS  I 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 I  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 (I) 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
Plant Systems
High-Level Contaminated Systems:
LDS 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/100cm2)
3.6e+09 ^
/•
1 .9e+06
8.36+03
Total Activity
(Ci)
, 2.08e+03
1.086+01
1.21e-02
                        A-I9
                                                                  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
2.56+02
1 7e+02
1.56+03
2.6e+03
8.66+03
Areal Activity
Concentration
(Ci/m2)

5.0e-02
5.06-02
5.0e-02
5.06-02
3.6e-01
3.66-01
3.6e-01
3.66-01

Total Activity
(Ci)

4.0e+01
5.06+01
7.0e+01
9.06+01
9.0e+01
6.06+01
5.4e+02
9.46+02
1.9e+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 .8e+04
1.2e+01
2.76+01
1 .7e+04
2.06+05

1.8e+02
8.56+01
1.0e+02
1 .9e+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 .4e+03
Area I 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
(|iCi/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
10 y
—
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
10 y
—
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
50.52 d
28.78y
58. 5 d
64.02 d
39.27 d
373.6 d
249.8 d
33.6 d
8.04 d
2.065 y
13.16d
30.07 y
12.75 d
32.5 d
284.9 d
13.57d
10.98 d

Specific
Activity
(na/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

10 y
—
—
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 Building"
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/1 00cm2)
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
1.66-02
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
1.1e-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 Building3
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
<18

-------
       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

-------
                                               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 Releasable3
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
5.96e+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
5.48e+05
1.16e+06
2.316+05
7.906+04
1.086+05
4.46e+04
Stainless


7.19e+04
5.78e+04
1.97e+04
2.696+04
1.126+04
Total Industry
All
Steel


3.74e+06
5.87e+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
I 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
w
us —
CD CD
C £
'ro to
to
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
i_
CD
Q.
Q.
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
T3
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
1
00
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
3
C
E
3
<
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
us
c/5
E
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
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

-------
               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
Surry 2
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)
Tnt^l
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
8 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
Scaleda
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

-------
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 Refractory"
<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
(g)
76
81
81
80
78
50
50
166
503
250
Flux
(g)
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
Ratio8
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
SiO2





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
Baghouse8





























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

-------
                                  flux
                               4338 MT/y


                                Silicon
           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
                                                                                 4338 MT/y
                                                                                   salts
           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
GTAW"
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
A1203
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.
                                          C-8

-------
          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
O
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



CD
0
3
CD Q
CD UJ
~7 «
1- o
o3 v)
|— CD
< O
99.2
120
50
50
50
350
50
1,500
1,500
2,000

o
O
-a
CL ~
0 •§
o — '
C_3 •*—'
o ^
fc 00
0 CD
O LiJ
99.6
210
20
130
100
80
20
1,600
500
1,200




o
O w
0 g
c
CD 0
o: .E
!_ n
CL 0
o ."t^
O €
O ?•
99.5-99.8
700
25
~
200
50
5
80
150
1,000-1,500


e-
o
O
O3
C
'c
M—
0
o:
o ®
0 o
0 ("~
^ c
c •*-
0 CD
*: m
99.63
435
490
69
500
210
41
380
510
1,228


c5^
w= 3
0
ry ^
CD ^
5 CD
O C
0 =
g^
c o>
0 .E

99.6
403
500
140
560
50
33
140
220
1,960



6
0 N
0
£ o
3 fc
O CD
03 O
99.3
600
~
20
500
700
30
1,000
3,000
2,000





03 m
1-iS
<*.&}-

O t—
DC o.^
<3^ o CD
(/3 o E
< 0 <
99.3
1,200
600
100
400
550
45
500
1,700
1,600
Source: Schloen 1987
O

to
to

-------
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. CD A 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
-2
"cL
C/3
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
A1A
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 Electrorefming 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 Electrorefming 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
(|iCi/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 (/") 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/MTHTM), 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
~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
1.07e-ll
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
1.66e-ll
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
9.79e-ll
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
3.48e-ll
7.79e-10
4.30e-01
5.38e-02
2.18e-09
4.18e-ll
1.23e-ll
O.OOe+00
1.10e-13
6.13e-08
1.88e-01
2.37e-03
2.95e-09
5.47e-ll
4.42e-ll
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
~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
1.81e-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
4.46e-ll
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
Ir-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
~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
1.95e-ll
1.50e-33
O.OOe+00
O.OOe+00
1.81e-19
7.74e-17
1.68e-15
2.28e-09
6.96e-ll
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
3.32e-ll
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
~Soil
1.32e-10
2.89e-09
2.16e-ll
1.80e-08
1.42e-ll
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
1.24e-ll
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
2.16e-ll
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
9.59e-ll
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
o
o
c/a
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)
/
—
—
—
/
/
—
/
—
/
/
—
—
/
—
—
—
/
/
—
/

-------
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)
—
/
/
—
/
/
/
/
/
/
/
/
—
/
/
/
/
/
—
/
/

-------
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)
—
/
—
/
/
/
/
—
—

-------
                                    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)
Cofl)
Cr(i,
Cud)
Mn(1)
Mo(s)
**>(.)
Nio)
Pba)
sia)
Sn
T1(S)
u«
vw
w(i,
Zr(s)
Oxide
Ag20
A1203
CaO
CeO2
CoO
Cr203
Cu2O
MnO
MoO3
Nb2O5
MO
PbO
SiO2
SnO2
TiO2
UO2
V205
WO3
ZrO2
v°
J 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°fiMo
(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%
Rafe)
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 slag3
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
  Pu,O,
                                           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
^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 (%)
Q
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
   SteelmakingU (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 FINISHING
                                                                                               CORE 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(e)
R^
Sbto
Sr(g)
T<™
Thrs,
Ym
Znro,
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
Mofrt
Nbw
Nim
Pbm
Sim
Snrn
Ti^
um
v(s)
w(i)
Zrrirt
Oxide
Ag20
A1203
CaO
Ce2O3
CoO
Cr203
Cu2O
MnO
MoO3
Nb2O5
MO
PbO
SiO2
SnO2
TiO2
UO2
V205
WO3
ZrO2
v°
If 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
AF°
"r 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
-------
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
3
Size3
A
3
4
3
0
5
B
5
7
2
12
1
C
0
3
2
4
1
Source: Natziger et al. 1990
3 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
T 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
'HMH
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
3Nb
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.  TheMaking, 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

-------