Expert system analysis of non-fuel assembly hardware and spent fuel disassembly hardware

MISSING IMAGE

Material Information

Title:
Expert system analysis of non-fuel assembly hardware and spent fuel disassembly hardware its generation and recommended disposal
Physical Description:
xi, 326 leaves : ill. ; 28 cm.
Language:
English
Creator:
Williamson, Douglas Alan, 1964-
Publication Date:

Subjects

Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1991.
Bibliography:
Includes bibliographical references (leaves 223-231).
Statement of Responsibility:
by Douglas Alan Williamson.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001722549
notis - AJD5028
oclc - 25764297
System ID:
AA00003314:00001

Full Text











EXPERT SYSTEM ANALYSIS OF NON-FUEL ASSEMBLY
HARDWARE AND SPENT FUEL DISASSEMBLY HARDWARE:
ITS GENERATION AND RECOMMENDED DISPOSAL

















By

DOUGLAS ALAN WILLIAMSON


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1991














ACKNOWLEDGEMENTS

The author would like to express his sincere

appreciation to all the people whose time and efforts

contributed to the completion of this dissertation. Special

thanks are given to the dissertation committee members,

Prof. James Tulenko, Dr. William Vernetson, Prof. Glen

Schoessow, Dr. Douglas Dankel, and Dr. Gene Dunnam, for

their continuous guidance and support. The author would

also like to thank his parents, without whose support this

work would not have been possible. And finally, the author

would like to thank his wife, Katherine, for her endless

patience and encouragement.

This research was performed under appointment to the

Radioactive Waste Management Fellowship program administered

by Oak Ridge Associated Universities for the U.S. Department

of Energy.















TABLE OF CONTENTS

ACKNOWLEDGEMENTS . . .. ii

LIST OF TABLES ... . .. .

LIST OF FIGURES . ... .vii

LIST OF ABBREVIATIONS. . ... ix

ABSTRACT .. .. .. ... ...... .. x

CHAPTERS

1 INTRODUCTION . .. .. 1

Purpose . . 3
Classes of Radioactive Waste . 6
Mill Tailings . .. 6
Transuranics . . 7
Spent Nuclear Fuel . 11
High-Level Waste . .. 14
Low-Level Waste . 22
Notes . . 31

2 DOMESTIC HARDWARE WASTE HANDLING . 35

Characteristics Data Base . 56
Utility Survey . . 66
Other Sources .. ............. 73
Domestic Hardware Analysis .. .... 77
Hardware Waste Classification . 79
Notes . . 113

3 FOREIGN HARDWARE WASTE HANDLING . 117

Hardware Waste Programs .. 118
France . . .... 119
United Kingdom ... .. 122
West Germany . . 126
Japan . . 129
Sweden . . 130
Hardware Handling Summary . .... 134
Notes . . 136

4 HARDWARE WASTE EXPERT SYSTEM . .. 141

Expert Systems . . 143


iii










Expert System Definitions
Expert System Components .
Knowledge Representation .
Reasoning Schemes .
Expert System Shells .
Exsys . .
RuleMaster 2 .
Exsys Professional .
Other Shells .
Prototype: Version 1 .
Conversion: Version 2 .
Refinement: Version 3 .
Program Results .
Program Results Verification
Hardware Waste Quantities .
Notes . .

5 CONCLUSIONS . .

Waste Classification .
Waste Quantities .
Hardware Disposal .
Further Work .
Notes . .

REFERENCES . .


. . .
. . .
. . .
. . .
. . .


. . .

. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .


APPENDICES
A ESTIMATE OF TOTAL NON-FUEL ASSEMBLY HARDWARE
INVENTORIES AS OF 1990 . .

B ESTIMATE OF TOTAL NON-FUEL ASSEMBLY HARDWARE
INVENTORIES OUT TO 2010 . .

BIOGRAPHICAL SKETCH . .. .


144
146
150
152
155
157
159
161
163
166
168
175
183
193
198
203

206

207
210
213
218
222

223


232



279

326


. . .
. . .
. .
. .
.. .














LIST OF TABLES


Table 1. TRU waste inventories at the eight TRU-
generating DOE sites as of Dec. 31, 1989. 8

Table 2. A comparison of the volumes and activity levels
of the five major radioactive waste classifications as
of Dec. 31, 1989. . 12

Table 3. Critical isotopes listed in 10 CFR 61 for LLW
classification with their half-lives. 23

Table 4. A chronology of some of the important dates
concerning radioactive waste disposal. .. 30

Table 5. Typical generation of storage space by rod
consolidation . . 45

Table 6. Summary of the contents of the five data bases
of the Characteristics Data Base. ... 57

Table 7. NFA hardware information included within the
Characteristics Data Base. . 62

Table 8. A summary of the estimated classification of
NFA and SFD hardware components based on the
component's materials of construction. 79

Table 9. A summary of the physical characteristics of
the three primary reactor materials studied in this
analysis. . ... .. 83

Table 10. Physical and radiological characteristics of
the critical isotopes and their parent materials 85

Table 11. The maximum permissible initial concentrations
of niobium in NFA hardware as a function of hardware
lifetime and materials of construction. 91

Table 12. The maximum permissible initial concentrations
of nickel in NFA hardware as a function of hardware
lifetime and materials of construction. .. 93

Table 13. The relative volumes of the materials in a
Combustion Engineering Control Rod Assembly and the
calculation of the component's waste classification. 98








Table 14. The relative volumes of the materials in a
Westinghouse Control Rod Assembly and the calculation
of the component's waste classification. ... 105

Table 15. The relative volumes of the materials in a
Palisades Control Blade and the calculation of the
component's waste classification. . 106

Table 16. A summary of the final waste classifications
for NFA and SFD hardware based on material of
construction and burnup. . 112

Table 17. A breakdown of the number of NFA hardware
records associated with the reactor records within the
Characteristics Data Base (CDB). . 170

Table 18. Breakdown of reactors, by vendor and reactor
type, for which estimates were performed by the HWES. 185

Table 19. The total NFA hardware inventories predicted
through the year 1990. . 186

Table 20. The total NFA hardware inventories predicted
through the year 2010. . 191

Table 21. A comparison of the predicted hardware
discharges versus the actual discharges at two
Combustion Engineering reactors on the same site. 194

Table 22. A comparison of hardware values provided by
the Characteristics Data Base to actual values used by
a nuclear utility at two reactor sites. 195

Table 23. A comparison of the predicted hardware
discharges versus the actual discharges at two Babcock
& Wilcox reactors of the same age. ..... 197

Table 24. A comparison of the predicted hardware
discharges versus the actual discharges at three
Combustion Engineering reactors on the same site. 197

Table 25. Total estimated quantities of NFA hardware
discharged as of the year 1990. . 199

Table 26. A summary of the classification results broken
down by NFA hardware type and materials of
construction. . .. 209

Table 27. Summary of NFA hardware waste quantities and
waste classifications. The hardware types are listed
in descending order of importance from a volumetric
standpoint. .............. .212














LIST OF FIGURES


Figure 1. Exploded view of a Combustion Engineering
Upper End Fitting on a PWR fuel assembly. 47

Figure 2. BWR Non-Fuel Assembly Hardware. 49

Figure 3. PWR Control Rod Assemblies. The left diagram
illustrates a Westinghouse Control Rod Assembly while
the right diagram shows three different configurations
used for Combustion Engineering Control Element
Assemblies. . . 50

Figure 4. Westinghouse Thimble Plug Assembly. 51

Figure 5. Illustration of the assumed flux shape within
the reactor and the size of the four irradiation
zones. . . .. 86

Figure 6. A comparison of the allowable initial niobium
concentrations for various irradiation times in
zircaloy, stainless steel, and inconel. ... 90

Figure 7. A comparison of the allowable initial nickel
concentrations for various irradiation times in zircaloy,
stainless steel, and inconel. . 92

Figure 8. Niobium concentrations in inconel which will
cause Nb-94 to exceed its Class C limit as a function
of time. . . 94

Figure 9. Nickel concentrations in inconel which will
cause Ni-59 and Ni-63 to exceed their respective Class
C limits as a function of time. .. 95

Figure 10. Niobium concentrations in stainless steel
which will cause Nb-94 to exceed its Class C limit as
a function of time. . . 99

Figure 11. Allowable initial niobium concentrations in
stainless steel as a function of irradiation time for
three different vertical positions in the core. .100

Figure 12. Nickel concentrations in stainless steel
which will cause Ni-59 and Ni-63 to exceed their
respective Class C limits as a function of time. .101


vii








Figure 13. Nickel concentrations in zircaloy which will
cause Ni-59 and Ni-63 to exceed their respective Class
C limits as a function of time. . ... 108

Figure 14. Niobium concentrations in inconel which will
cause Nb-94 to exceed its Class C limit as a function
of time. . . 109

Figure 15. An example of frames and inheritance. 151

Figure 16. Schematic representation of the HWES Version
2. . . .. 172

Figure 17. Schematic representation of the HWES Version
3. . . . 177


viii














LIST OF ACRONYMS


AEC Atomic Energy Commission
AGR Advanced Gas Reactor
AVM Advanced Vitrification Method
BPRA Burnable Poison Rod Assembly
BRC Below Regulatory Concern
BWR Boiling Water Reactor
CDB Characteristics Data Base
CFR Code of Federal Regulations
DOE Department of Energy
EFPD Effective Full Power Day
EFPY Effective Full Power Year
EPA Environmental Protection Agency
EPRI Electric Power Research Institute
FIS Federal Interim Storage
FWMS Federal Waste Management System
GTCC Greater-Than-Class-C
HANF Hanford Reservation
HLW High-Level Waste
HWES Hardware Waste Expert System
ILW Intermediate-Level Waste
INEL Idaho National Engineering Laboratory
LLW Low-Level Waste
LLWAA Low-Level Waste Ammendments Act
LLWPA Low-Level Waste Policy Act
LMFBR Liquid Metal Fast Breeder Reactor
LWR Light Water Reactor
MRS Monitored Retrievable Storage
MTIHM Metric Tons Initial Heavy Metal
MWD Megawatt-Days
NARUC National Association of Regulatory Utility
Commissioners
NFA Non-Fuel Assembly
NRC Nuclear Regulatory Commission
NWF Nuclear Waste Fund
NWPA Nuclear Waste Policy Act
NWPAA Nuclear Waste Policy Ammendment Act
OCRWM Office of Civilian Radioactive Waste Management
ONRL Oak Ridge National Laboratory
PNL Pacific Northwest Laboratory
PWR Pressurized Water Reactor
RG&E Rochester Gas & Electric
SFD Spent Fuel Disassembly
SNF Spent Nuclear Fuel
TRU Transuranic
WVDP West Valley Demonstration Project
WIPP Waste Isolation Pilot Project














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

EXPERT SYSTEM ANALYSIS OF NON-FUEL ASSEMBLY
HARDWARE AND SPENT FUEL DISASSEMBLY HARDWARE:
ITS GENERATION AND RECOMMENDED DISPOSAL

By

Douglas Alan Williamson

December 1991

Chairman: Prof. James S. Tulenko
Major Department: Nuclear Engineering Sciences

Almost all of the effort being expended on radioactive

waste disposal in the United States is being focused on the

disposal of Spent Nuclear Fuel, with little consideration

for other areas that will have to be disposed of in the same

facilities. One area of radioactive waste that has not been

addressed adequately because it is considered a secondary

part of the waste issue is the disposal of the various Non-

Fuel Bearing Components of the reactor core. These hardware

components fall somewhat arbitrarily into two categories:

Non-Fuel Assembly (NFA) hardware and Spent Fuel Disassembly

(SFD) hardware.

This work provides a detailed examination of the

generation and disposal of NFA hardware and SFD hardware by

the nuclear utilities of the United States as it relates to

the Civilian Radioactive Waste Management Program. All

available sources of data on NFA and SFD hardware are

x








analyzed with particular emphasis given to the

Characteristics Data Base developed by Oak Ridge National

Laboratory and the characterization work performed by

Pacific Northwest Laboratories and Rochester Gas & Electric.

An Expert System developed as a portion of this work is used

to assist in the prediction of quantities of NFA hardware

and SFD hardware that will be generated by the United

States' utilities. Finally, the hardware waste management

practices of the United Kingdom, France, Germany, Sweden,

and Japan are studied for possible application to the

disposal of domestic hardware wastes.

As a result of this work, a general classification

scheme for NFA and SFD hardware was developed. Only NFA and

SFD hardware constructed of zircaloy and experiencing a

burnup of less than 70,000 MWD/MTIHM and PWR control rods

constructed of stainless steel are considered Low-Level

Waste. All other hardware is classified as Greater-Than-

Class-C waste.

Extensive hardware quantity predictions performed by

the Expert System are also presented. Good approximations

resulted for Combustion Engineering and Babcock & Wilcox

reactors, and values for General Electric and Westinghouse

reactors were derived from these results. Methods for

packaging the hardware for disposal are also presented.














CHAPTER 1
INTRODUCTION

One of the most pressing concerns facing the American

nuclear industry is the disposal of the radioactive waste

generated by the nation's commercial nuclear power reactors.

As of December 31, 1990, 111 commercial nuclear power plants

operating nationwide1 were generating 21% of the nation's

commercial electricity.2 The commercial power industry

also generates waste in the course of operations, waste

which requires special handling as a result of the

radioactivity it contains. Most scientists within the field

agree that disposing of this waste in a manner which is safe

and effective is technically possible with today's

technology.3 However, ensuring that such disposal will

indeed be permanent as well as publicly acceptable has

proven to be a formidable task, a task that has come to

require the involvement of the federal government as well as

the private sector.

The origins of commercial radioactive waste (radwaste)

can be found in the year 1957 when the first commercial

nuclear power station began operation. Defense-related

radwaste made its first appearance even earlier during the

Manhattan Project of World War II. In these early days of

nuclear energy, formalized methods of waste disposal and

waste classification had not been developed, so the methods

1






2

used for the management of radwaste were inconsistent.

Since that time, several plans for the management and

disposal of the various classes of radioactive waste have

been developed. However, the implementation of these plans

has met with varying degrees of success.

Management concepts for radioactive waste were

discussed as early as 1955 in Geneva at the first conference

on Peaceful Uses of Atomic Energy. Papers on various

aspects of radwaste disposal both on land and at sea were

presented by scientists from the United States, the United

Kingdom, and Canada.4 In 1972, Congress made its first

attempt at creating a national radwaste program through the

Atomic Energy Commission (AEC).5 Since that time, several

changes have been made in the management and regulation of

radwaste, including the development of a detailed, if rather

complicated, system for classifying radioactive wastes which

will be discussed later in this chapter. Progress has been

made toward the disposal of radwaste, but the progress is

far from uniform and has generally been contested every step

of the way. The most progress has been made in the disposal

of Low-Level Waste which is generally considered resolved,

while the least progress has been made in the disposal of

High-Level Waste which is little better off now than it was

in 1957. In point of fact, all aspects of nuclear energy

and particularly radioactive waste management have become so

controversial and contentious as to require acts of Congress

to make any progress at all.






3

Purpose

Almost all of the effort being expended on radioactive

waste disposal in the United States is being focused on the

disposal of Spent Nuclear Fuel, leaving very little

consideration for other areas. One area of radioactive

waste that has been considered a secondary part of the waste

issue is the disposal of the various Non-Fuel Bearing

Components of the reactor core, metal hardware that is

activated but not fuel bearing. These hardware components

fall somewhat arbitrarily into two categories: Non-Fuel

Assembly (NFA) hardware and Spent Fuel Disassembly (SFD)

hardware. Neither hardware category is comparable to the

weight burden of Spent Nuclear Fuel, but when the waste is

measured by volume, these hardware wastes can be essentially

equal to Spent Nuclear Fuel. Additionally, there is even

more diversity among Non-Fuel Assembly hardware and Spent

Fuel Disassembly hardware then there is among fuel

assemblies.

These Non-Fuel Bearing Components are somewhat

controversial with regard to classification, which can

affect jurisdiction. First, while these wastes are

certainly either Class C or Greater-than-Class-C (GTCC) Low-

Level Waste (LLW), which classification is actually correct

is in dispute. As the methods of disposing of these two

classes of waste are significantly different, correct

classification is important. Furthermore, proper waste

classification would clarify the jurisdictional






4

responsibility as GTCC LLW is the responsibility of the

federal government. In such a case, these wastes would fall

under the jurisdiction of the Office of Civilian Radioactive

Waste Management (OCRWM), so it is important that the waste

streams be studied, classified, and quantified. No

comprehensive study of these components has yet been

published.

The purpose of this work is twofold: 1) to examine in

depth the generation and disposal of Non-Fuel Assembly

hardware and Spent Fuel Disassembly hardware by the nuclear

utilities of the United States as it relates to the Civilian

Radioactive Waste Management Program and 2) to build an

Expert System which will assist in the prediction of

quantities of Non-Fuel Assembly hardware and Spent Fuel

Disassembly hardware that will be generated by the United

States' utilities and will also estimate the waste

classification of these wastes. Several data gathering

methods are used to provide the most comprehensive and

balanced approach to the problem. The first source is a

comprehensive examination of all available literature on the

subject. The second is a survey of domestic nuclear

utilities for general utility experiences. Direct contact

with select utilities for specific information and direct

questioning is also utilized, as well as contact with other

relevant agencies, corporations, and laboratories. Finally,

direct contact with European organizations which have

routinely handled similar waste forms is used to provide a






5

basis for comparison and a view toward actual waste handling

experience. The information gathered from these sources is

coupled with original analysis to classify and quantify the

NFA and SFD hardware waste streams, as well as to make

general recommendations on the hardware's packaging and

disposal.

Details of the work performed for this dissertation are

organized in the following manner. The remainder of Chapter

1 provides detailed background information on radioactive

waste classification and an overview of the current status

of radwaste disposal programs in the United States. Chapter

2 presents greater detail on NFA hardware and SFD hardware.

The sources and general characteristics of these hardware

types are discussed, and all domestic data sources on NFA

and SFD hardware are analyzed. Hardware classifications are

then presented based on original analysis of these and other

data. In Chapter 3, the practices and experiences of five

foreign nations are examined with emphasis placed on how

these experiences can benefit the American waste management

programs. Chapter 4 begins with general information on

Expert Systems and Expert System shells, and then proceeds

to discuss the development of the Hardware Waste Expert

System (HWES). The results of the HWES are presented at the

end of this chapter. Finally, the conclusions reached as a

result of this work are presented in Chapter 5. Areas of

research which could benefit from additional study are also

identified.






6

Classes of Radioactive Waste

Radioactive waste classification is primarily based on

the concentration of radionuclides in the waste and

secondarily on the process which generated the waste. The

waste classification system has evolved from its early

general categories to the more specific classifications of

today. The modern system for the classification of

radioactive wastes in the United States involves five major

categories of waste. These categories are Low-Level Waste

(LLW), High-Level Waste (HLW), Spent Nuclear Fuel (SNF),

Transuranic (TRU) waste, and mill tailings.



Mill Tailings

Proceeding in reverse order, mill tailings are the

byproducts of uranium mining, extraction, and milling. The

average yield of uranium from the uranium ore is about 0.1%

to 0.5% and is composed of approximately 0.72% 5U and

99.27% 238U.6 The natural uranium is then sent through a

milling, extraction, and purification procedure to eliminate

impurities. Mill tailings are the waste product from this

process, which take the form of a slurry of sand and clay

particles. The large piles of mill tailings which are

currently the subject of the DOE Environmental Restoration

Program resulted when these slurries were pumped into

impoundment ponds to dry.7

The mill tailings retain more than 75% of the

radioactivity in the natural ore, but due to the large






7

quantities of non-radioactive materials in the ore, the

volumes of mill tailings are large while the radioactivity

concentrations are low.8 At the end of 1989, the total

volume of mill tailings at active mill tailing sites, i.e.

those sites not undergoing environmental restoration, was

117.6 X 106 cubic meters. Estimates of the total

radioactivity in mill tailings are not available.9

The mill tailings retain a small percentage of uranium,

but the primary isotopes of concern from a radiation

protection standpoint are 226Ra and 222Rn.1 Increased levels

of radon are detectable within roughly one kilometer of a

mill tailings site. Significant radon release problems have

occurred in a few instances where the mill tailings were

erroneously used as fill for urban developments.11 Current

management procedures for mill tailings place the dried

tailings in piles within enclosures and provide for minimum

controls to prevent the waste from entering the water table

or from becoming airborne.12 Unlike other radwastes, mill

tailings are disposed of at the site where they are

generated.



Transuranics

Transuranic (TRU) waste is material contaminated with

more than 100 nCi/g of alpha-emitting radionuclides which

have a half-life of at least 20 years and atomic number 92

or greater13 and includes such isotopes as s5U, 23U, and

9Pu. TRU waste is generally composed of trash such as






8

rubber gloves, filters, and rags from the reprocessing and

refinement of nuclear fuels. This waste is further divided

into two sub-classifications based on the degree of

contamination. For "Contact-handled" TRU, the shielding

provided by the waste package itself is sufficient for safe

handling. "Remote-handled" TRU, however, exhibits a dose

rate of more than 200 mrem/hr of alpha, beta, and/or gamma

emitters, so additional shielding and remote handling are

required. Fortunately, remote-handled TRU waste represents

only about 2.4% of the total TRU waste inventory as

summarized in Table 1.14

Since commercial reprocessing is currently at a

standstill, virtually all TRU waste is generated during the

extraction of plutonium from production fuels for the

Department of Defense. The current TRU inventories are

located at eight DOE sites: the Savannah River Site; Oak

Ridge National Laboratory; Los Alamos National Laboratory;

the Rocky Flats Plant; Sandia National Laboratory,

Albuquerque; the Nevada Test Site; Idaho National




Table 1. TRU waste inventories at the eight TRU-generating
DOE sites as of Dec. 31, 1989.
Volume Radioactivity Heat Load

Buried TRU 190,837.0 cu. m 211,900 Ci 3,300 W

Stored TRU
Contact 60,057.0 cu. m 1,179,600 Ci 27,200 W
Remote 1,501.9 cu. m 2,486,000 Ci 8,400 W
Subtotal 61,558.9 cu. m 3,665,600 Ci 35,600 W

Total TRU 252,395.9 cu. m 3,877,500 Ci 38,900 W






9

Engineering Laboratory (INEL); and the Hanford Site (HANF).

The vast majority of these wastes are at HANF and INEL.

Table 1 shows the total TRU inventory as of December 31,

1989 which is distributed among these sites. The table

shows the TRU inventories broken down by classification

(remote-handled vs contact-handled) and the waste's current

situation (buried vs stored).15 The inventories listed as

"buried" were emplaced prior to 1970, when land burial was

considered a safe disposal option for these wastes. In

1970, however, the AEC ruled that additional safeguards

should be used to separate TRU wastes from the environment,

so all TRU wastes since that time have been placed into

retrievable-storage pending a final disposal site. When

such a disposal site is available, the previously buried TRU

waste will be exhumed, surveyed, and when necessary, moved

to the new site.16

The Waste Isolation Pilot Project (WIPP) is being

constructed at Carlsbad, New Mexico for the disposal of TRU

wastes from defense activities. The facility is being

constructed at a depth of approximately 2150 feet in a bed

of rock salt, one of the earliest media considered for the

disposal of radioactive wastes. The WIPP facility has been

designed for a nominal operating life of 25 years and has a

final design capacity of roughly 178,000 cubic meters of

contact-handled TRU and 5100 cubic meters of remote-handled

TRU waste.17 In addition to the disposal of TRU waste, the

facility has also been designed "for the purpose of






10

providing a research and development facility to demonstrate

the safe disposal of radioactive wastes resulting from the

defense activities and programs of the United States."18

As such, the first five years of WIPP operation are

designated as a test phase during which contact-handled TRU

waste and pilot project amounts of HLW from defense

activities will be emplaced in the facility for in-situ

testing. If the results of the test phase are favorable,

additional quantities of TRU, including remote-handled TRU,

will be permanently emplaced in WIPP up to the design

capacity. The HLW, however, will have to be removed and is

eventually destined for the Federal HLW Repository, the same

repository that is being pursued for the disposal of

commercial spent fuel.

The WIPP facility was originally scheduled to be

operational in 1988. The site was essentially completed on

schedule, but exhibited some technical uncertainties such as

a water seepage condition which caused concern among

governmental officials. Accordingly, when Congress failed

to approve a land swap between the Department of the

Interior and the Department of Energy (DOE), opening of the

site was postponed indefinitely to allow time for the land

swap to be approved and to give the DOE time to address

these concerns.19 In June of 1990, Secretary of Energy

James Watkins announced that the DOE expects to begin the

test phase of operations at WIPP as soon as certain,

unspecified prerequisites outside of the Department's








control are met.20 One major prerequisite was met on

November 1, 1990 when the Environmental Protection Agency

(EPA) approved the DOE TRU waste emplacement test

program.21 The granting of the administrative land

withdrawal on January 22, 1991 by the Interior Department's

Bureau of Land Management cleared another major obstacle.

The DOE now hopes to resolve any remaining issues in time to

begin the movement of TRU waste drums from INEL to WIPP in

July 1991.22



Spent Nuclear Fuel

The third primary category of radwaste is Spent Nuclear

Fuel (SNF). The Nuclear Waste Policy Act of 1982 (NWPA)

defines Spent Nuclear Fuel as "fuel that has been withdrawn

from a nuclear reactor following irradiation, the

constituent elements of which have not been separated by

reprocessing."1 In the ideal fuel cycle, SNF is

reprocessed and the useful fissile isotopes are extracted

for fabrication into new fuel elements. In recent years,

however, some countries, including the United States, have

opted for direct disposal of SNF, thus eliminating the

reprocessing stage. As such, SNF will be treated as HLW

until such a time as reprocessing is again initiated for

commercial fuels.

In terms of quantities, SNF can be somewhat misleading

because of the number of ways in which these quantities can

be measured. SNF is the only category of radwaste to






12

contain large quantities of uranium and plutonium. As a

result, quantities of spent fuel are usually measured in

terms of metric tons of initial heavy metal (MTIHM). In

these terms, the total SNF inventory as of December 31, 1989

was 19,641 MTIHM. However, for purposes of comparison with

other waste categories, this measurement is inadequate.

Accordingly, the SNF inventory can also be expressed

volumetrically as 7920 m3 or in terms of radioactivity as

2.07x1010 Curies.24 Table 2 shows a comparison of the five

major categories of radioactive waste. While SNF represents

the smallest volume of radwaste, it also contains the

highest level of radioactivity (2.07 x 1010 curies) and the

highest volumetric activity (2.61 x 106 curies/cubic

meter).25 SNF also has a greater waste heat output than

any other radwaste category. Hence, SNF is clearly the most

concentrated form of radioactive waste.

The high radiation levels exhibited by spent nuclear

fuel, coupled with the significant waste heat load it

generates, requires special consideration. As a direct




Table 2. A comparison of the volumes and activity levels of
the five major radioactive waste classifications as of
Dec. 31, 1989.
Total Volumetric Total
Volume Total Activity Heat Load
(cu. m) Curies (Ci/cu. m) (Watts)

LLW 3.91E+06 1.91E+07 4.88E+00 4.32E+04
HLW 3.81E+05 1.11E+09 2.91E+03 3.16E+06
SNF 7.92E+03 2.07E+10 2.61E+06 7.83E+07
TRU 2.52E+05 3.67E+06 1.46E+01 3.89E+04
Tailings 1.18E+08 N/A N/A N/A






13

result of these factors, after being discharged from the

reactor core, spent fuel is stored in spent fuel pools at

the reactor site for a period of time to allow the shorter-

lived fission products to decay, thus reducing the heat load

and radiation level of the assemblies. In the early days of

nuclear power, this cooling period was expected to range

from three months to a year, after which the spent fuel

would be removed for reprocessing.26 In practice, however,

commercial fuel reprocessing is currently unavailable, so

the cooling period has been extended to five years or more.

After such a cooling period, the SNF can be relocated to

higher density storage racks, or even to dry storage

modules. Pressurized Water Reactor (PWR) spent fuel pools

are filled with water containing a boron concentration of

approximately 2000 parts per million (ppm) whereas Boiling

Water Reactor (BWR) spent fuel pools contain only

demineralized water without any boron additive. The water

serves several safety purposes, the most important of which

is the maintenance of subcritical configuration for a keff of

approximately 0.95 or less. Water also provides radiation

shielding, and decay heat removal capacity while still

allowing visual inspection of the fuel in the fuel pool.

The typical spent fuel pool capacity is 4-5 full reactor

cores or about 850 fuel assemblies.27 When the spent fuel

pool reaches capacity, as will soon happen at several of the

nation's nuclear power plants, several options are avail-

able. These options are discussed in detail in Chapter 2.






14

Since SNF is not currently reprocessed in the United

States, the fuel has been accumulating at the nation's

reactor sites and a method of disposing of this waste form

has been needed. The DOE's current plans as mandated by

federal law (see "High-Level Waste," this chapter) call for

the development of a Federal HLW Repository as a permanent

disposal facility for SNF and HLW. The key point of these

plans is to solve the waste management issue now and not to

postpone it so as to place the burden on the shoulders of

future generations. Furthermore, since the NWPA

specifically limits the first repository to a final capacity

of 70,000 metric tons of heavy metal,2 and since SNF is

expected to represent the majority of the waste to be

emplaced in the repository, its quantities and character-

istics are considered extremely important to the Federal

Waste Management System and to OCRWM.



High-Level Waste

Also destined for the repository is the next category

of radwaste, High-Level Waste. High-Level Waste is a waste

product of the reprocessing of spent nuclear fuel. Since

reprocessing is designed to extract the plutonium and the

remaining usable uranium from spent fuel for fabrication

into new fuel elements, HLW contains very little of these

elements and instead consists mainly of fission products and

other TRU elements.28 Freshly generated HLW exhibits high

levels of radioactivity and generates considerable decay






15

heat. The projected characteristics of HLW forms (canisters

of vitrified HLW) which will be generated at the Savannah

River Site, Hanford, Idaho National Engineering Laboratory,

and the West Valley Demonstration Project (WVDP) range from

a low of 125,200 curies and 382 watts per canister at the

WVDP to a high of 416,000 curies and 1158 watts per canister

at Hanford.2 However, since the process which produces

these wastes removes a significant portion of the longer

lived actinides and dilutes the remaining radioactivity, the

volumetric activity of HLW is three orders of magnitude

lower than the activity of SNF as shown by Table 2. While

the activity and heat load of both SNF and the HLW is

expected to drop by a factor of 100 after only 200 years,

the HLW will always require less shielding and cooling than

the SNF. Some countries like the United Kingdom and France

plan to store their HLW for 50 years or more prior to

disposal to take advantage of this decay.

The HLW is initially a liquid or sludge and, in the

past, has been stored in this form, but over the long term,

these storage facilities are prone to developing leaks which

necessitate expensive remedial cleanup activities.

Accordingly, methods for immobilizing these wastes in solid

blocks were found. In the United States, several studies

were conducted into various ceramic waste forms, but the

most promising waste form was developed in France and is

produced by the Advanced Vitrification Method (AVM).M In

the AVM process, the liquid waste is run through a heated






16

cylinder which dries the waste. The resulting solid

calcinate is mixed with borosilicate glass frit in a ratio

of approximately 38% calcinate and 62% glass frit. The mix

is then melted and poured into molds, where it solidifies

into glass blocks. The AVM process has been in operation at

the Marcoule plant in France since 1978 without any

unforseen difficulties, and is also being installed in the

newer reprocessing facilities at La Hague.31 The vitrified

waste block is the reference waste form for HLW destined for

the Federal HLW Repository.32

As shown in Table 2, HLW represents the second largest

volume of radwaste in the United States, exclusive of

uranium mill tailings. A small portion of this inventory is

commercial HLW from the commercial fuel reprocessing

performed in the 1970's at the West Valley plant in West

Valley, New York. During the plant's brief operation, 650

MTU were reprocessed which produced roughly 2000 cubic

meters of HLW.3 The vast majority of HLW in the United

States, however, is generated by the production of nuclear

materials for national defense activities. The combined

volumes of commercial HLW and SNF represented only 2.6% of

the total 381,000 m3 HLW volume extant in 1989 with the

remainder being defense HLW. Conversely, the l.11x109

curies in current HLW inventories represent only 5.4% of the

total curie content of SNF." Without commercial fuel

reprocessing to produce HLW, defense HLW is expected to

remain the largest portion of the HLW inventory. However,






17

the vitrification programs under development at the three

DOE facilities and the WVDP should help to reduce the total

HLW volume by as much as a factor of ten35 which will alter

the proportions accordingly.

The development of a facility for the ultimate disposal

of HLW has always been the responsibility of the federal

government. The AEC began investigating the possibility of

burying HLW in bedded salt deposits as early as 1957, after

such an approach was suggested by the National Academy of

Sciences and the National Research Council.3 The project

was abandoned in 1971 by order of Congress and no further

significant developments occurred until the passage of the

Nuclear Waste Policy Act in 1982. The NWPA instructed the

Department of Energy to develop one or more repositories for

the final disposal of the nation's SNF and HLW. The three

key provisions of the act are the establishment of 1)

guidelines and milestones, 2) the Office of Civilian

Radioactive Waste Management, and 3) the Nuclear Waste Fund.

The guidelines and milestones provided the program with a

structure within which to work, thus providing much needed

direction to the effort as well as a means of measuring the

project's progress. The Office of Civilian Radioactive

Waste Management was established as a branch of DOE whose

only responsibility was the development and operation of a

federal repository. Finally, the monies paid into the

Nuclear Waste Fund by the utilities provided a means of

paying for the repository, and all the associated work which






18

did not require federal expenditures and simultaneously

committed the utilities to using the facility once

established. The Nuclear Waste Fund is discussed in greater

detail later in this section.

OCRWM was duly established and began work on the

repository. In December 1984, as per the NWPA, the DOE

nominated nine candidate sites for consideration. After

completing the Environmental Assessments for these sites in

April 1986, three of the sites were recommended to the

President for site characterization. These sites were

located in Deaf Smith County, Texas; Yucca Mountain, Nevada;

and the Hanford Reservation, Washington. The next stage

laid out by the NWPA involved site characterization of all

three sites, and a continued search for sites east of the

Mississippi River as candidates for a second repository.

However, all "efforts to survey potential sites in the

eastern part of the United States were hotly contested .

[so] citing lack of need, the DOE suspended its site-

selection program for the second repository."37

By 1987, site characterization was proceeding at only

one site (Yucca Mountain), and the entire program was faced

by so much opposition that Congress was required to

intervene with the Nuclear Waste Policy Amendment Act of

1987 (NWPAA). Based on the concept that "the U.S. problems

are political and not technical, i.e. that the storage of

HLW is no big technical problem,"37 the NWPAA sought to

streamline the selection process. The most important






19

features of the NWPAA are that 1) the site characterization

was officially narrowed to only the Yucca Mountain site, 2)

the second repository is canceled until a report on the need

for such a facility is submitted between the years 2001 and

2010, and 3) financial incentives are now included in the

site selection process to encourage state cooperation.37

Once the site characterization has been completed, then

OCRWM must justify the site, and the site must be reviewed

and approved by the Nuclear Regulatory Commission (NRC).

The review procedure is expected to be long and time-

consuming, thus contributing significantly to the delay time

until the repository can begin operation." This delay,

however, may actually prove to be relatively short in

comparison to the other delays the program has already

suffered.

The NWPA set a target date of January 31, 1998 for the

first repository to begin receiving spent fuel from the

utilities. By 1987, the project had fallen sufficiently

behind that DOE proposed postponing the initial operational

date of the repository to the year 2003.3 This target

date was subsequently confirmed by the enactment of the

NWPAA. Unfortunately, due mainly to the obstacles presented

by the state of Nevada, the project is once again stalled

and the year 2010 is currently considered the earliest date

for first operation." To date, Nevada has refused to

issue the necessary environmental permits needed by the DOE

to perform site characterization studies of Yucca Mountain.






20

On January 25, 1990, DOE filed suit with the U.S. District

Court alleging that Nevada "has prevented the federal agency

from carrying out necessary site investigation work by

unlawfully refusing to act.""4 This suit was subsequently

decided in favor of the DOE, and Nevada's appeal was turned

down by both the U.S. Court of Appeals in September 1990 and

by the U.S. Supreme Court on March 4, 1991. Before

beginning work, however, DOE must now await the outcome of

another lawsuit which is designed to force the issuance of

the needed permits. As this case and its subsequent appeals

are not likely to be resolved until late in 1992, DOE has

also asked Congress for the authority to conduct the site

investigation without regard for the opposition presented by

Nevada.42 If granted, then the DOE can proceed with the

characterization work without further delay. In any case,

no definitive schedule for the repository's development will

be possible until the legal battle is resolved, and even

then, other delays are extremely likely.

A further consideration when discussing the Federal

Waste Management System (FWMS) is the Nuclear Waste Fund

(NWF), which was created by the NWPA and is "funded with an

assessment of one mill per kilowatt-hour on all electricity

generated at commercial nuclear plants."3T Approximately $4

billion has been paid into the NWF by the utilities from the

program's inception through the end of fiscal year 1989.

The NWF's balance at the end of the fiscal year was $2.2

billion.3 The approximate unaudited balance at the end of






21

the 1990 fiscal year was $2.6 billion." Since the NWF's

inception, the General Accounting Office (GAO) and the DOE

have been in dispute as to whether or not these assessments

will be sufficient, with the GAO urging for an increase in

the assessment.45 However, as of November 1990, the DOE

did not feel it necessary to increase the waste fee.6 In

any case, the utilities are already concerned about getting

their money's worth from the waste fund. The same contract

which requires the utilities to pay the one mill per

kilowatt-hour fee also stipulated that the DOE must begin

accepting SNF no later than January 31, 1998.47 Due to

innumerable programmatic delays and continuous opposition to

its efforts, the DOE will be unable to fulfill its end of

the bargain unless legislative changes are made. The only

likely method foreseen at this time by which the DOE can

meet the 1998 deadline is the development of a Monitored

Retrievable Storage (MRS) facility which is not coupled to

the development of the repository. As things stand now, the

utilities must continue to hold the SNF after having

collectively paid out some $4 billion without any

demonstrable gain.

The utilities have also shown concern over the manner

in which the waste fee is calculated. Since the inception

of the NWF, the utilities have won two judgements concerning

the calculation of the waste fee.4 In December 1985, the

court ruled that the waste fees should be based on net

electricity production as opposed to gross electricity






22

production, taking into account power used on the plant

site.9 In March 1989, the court further ruled that

transmission losses should also be excluded from the

calculation.50 The fees are now based solely upon the net

electricity sold to consumers. Before the NWPAA was

enacted, "the electric utility industry was prepared to stop

its contributions to the Nuclear Waste Fund if a

moratorium [on HLW disposal] had been legislated."37 Now

with the HLW program once again stalled in legal

deliberations, Congress may be required to act once again to

expedite the development of the HLW repository.



Low-Level Waste

The final major category of radwaste is also the

broadest in definition. Low-Level Waste is a catchall

category for otherwise uncategorized radioactive wastes, or

more specifically, "Low-level waste [is] radioactive

waste not classified as high-level radioactive waste,

transuranic waste, spent nuclear fuel, or byproduct

material."51 Within the LLW classification, there are four

further subdivisions -- Class A, Class B, Class C, and

Greater-Than-Class-C -- which require progressively more

stringent precautions for disposal. Class A, B, and C

wastes are currently disposed of by shallow land burial.

Class A waste is regarded as the least hazardous and thus

requires the least precautions while Class C wastes are

subject to more restrictive guidelines for disposal.






23

Greater-Than-Class-C waste, on the other hand, is generally

regarded as not suitable for shallow land burial and thus

cannot be treated like other Low-Level Wastes.

LLW classifications are determined by measuring the

concentrations of certain radionuclides within the waste

package which are specified in 10CFR61 as follows:

First, consideration must be given to the
concentration of long-lived radionuclides (and
their shorter-lived precursors) whose potential
hazard will persist long after such precautions as
institutional controls, improved waste form, and
deeper disposal have ceased to be effective .
Second, consideration must be given to the
concentration of shorter-lived radionuclides for
which requirements on institutional controls,
waste form, and disposal methods are effective.51

The actual isotopes of concern are listed in two groups in

10CFR61 and are reproduced here in Table 3. The isotopes in

Group 1 are the long-lived radionuclides referred to above

while the Group 2 isotopes are the shorter-lived radio-

nuclides. Each of these nuclides is listed within one of

two tables given in 10CFR61 along with its allowable

concentration limits.52 By determining the ratio of the

concentration of each nuclide in the waste form to the value




Table 3. Critical isotopes listed in 10 CFR 61 for LLW
classification with their half-lives.
Group 1 (long lived) Group 2 (short lived)

C-14 (5730 y) H-3 (12.26 y)
Ni-59 (80,000 y) Co-60 (5.24 y)
Nb-94 (20,000 y) Ni-63 (92 y)
Tc-99 (210,000 y) Sr-90 (28.8 y)
1-129 (1,600,000 y) Cs-137 (30 y)
Pu-241 (13 y)
Cm-242 (163 d)






24

given for that nuclide in the tables, the LLW classification

of the waste can be determined. If the ratio is greater

than one, the waste is classified as GTCC LLW. Addition-

ally, even if the individual nuclide concentrations do not

exceed the limits, the waste may still be GTCC if the

combination of the ratios of the nuclides is too great. In

this case, the individual ratios calculated earlier for the

individual isotope concentrations are summed together, and

if the resulting sum is greater than one, then the

combination of nuclides is considered to be too great for

disposal as LLW, so the waste is classified as GTCC. In

both cases, the waste is then required to meet all the

handling and disposal criteria for GTCC wastes. Otherwise,

depending on the actual concentrations, the waste will be

either Class A, B, or C and may be disposed of by shallow

land burial. In the specific case of Non-Fuel Assembly and

Spent Fuel Disassembly hardware, since these materials are

activated metals, only four of the isotopes listed in

Table 3 are measured (14C, "Ni, 3Ni, and 4Nb), but the

procedure is otherwise as described above.

The disposal of LLW is the most advanced radwaste

program in the United States as such disposal has been

routinely conducted since 1962 when the first commercial LLW

disposal site opened at Beatty, Nevada. The number of sites

increased to six in 1971 before a reduction to three in

1979. These three remaining sites are located in Barnwell,

South Carolina; Richland, Washington; and Beatty, Nevada.






25

In 1979 and 1980, the states which hosted these sites became

concerned about the inequity of the LLW disposal situation

and made efforts to close their sites to LLW from other

states. These actions in turn prompted Congress to pass the

Low-Level Waste Policy Act (LLWPA) of 1980 and later, the

Low Level Waste Amendments Act (LLWAA) of 1985.53 Under

the provisions of these Acts, each state is responsible for

the permanent disposal of its own LLW. The Acts encouraged

the states to form "compacts" in order to jointly develop a

common site for use of all the member states. Several

milestones were provided within the Acts to ensure that the

states did in fact develop such sites. Failure to meet

these deadlines is penalized by escalating surcharges on

disposal at existing LLW sites with this cost being born by

the waste generators. The final deadline was originally set

for December 31, 1986 by the LLWPA, but was subsequently

postponed to December 31, 1992 by the LLWAA when it became

apparent that the procedure for forming and approving

regional compacts was proceeding much slower than

anticipated. After this final deadline, the LLW from

generators in any state that does not have an operational

LLW disposal site, or that does not belong to a compact that

does, may be excluded from all other disposal sites.

However, in such a case, the generators are entitled to

demand that the state take title to the LLW which they are

thus unable to dispose of.5"55 How this action will be

accomplished if the option is exercised is still an






26

unresolved issue. As 1992 approaches, most compacts are

still developing their sites and do not expect the sites to

be operational before 1995. So, whereas progress is being

made toward the eventual establishment of compact LLW sites,

it is unlikely that the sites will be operational early

enough to meet the deadline set by the LLWAA. The compacts

are currently examining options for interim measures until

their sites are operational.

Whereas the techniques for the disposal of LLW is

clearly established, the disposal of GTCC LLW is not so

clearly defined. Of the twelve isotopes listed in Table 3,

ten are used to judge between Class C and GTCC wastes. The

isotopes 3H and "Co have no Class C limit and therefore can

be present in any concentration without making the waste

GTCC. Furthermore, when the waste being classified consists

of activated metals, only four isotopes are of importance.

These isotopes are Carbon-14 (14C), Nickel-59 (5Ni), Nickel-

63 (6Ni), and Niobium-94 (9Nb). As has been previously

mentioned, when the isotope concentrations exceed the

allowable limits, the waste is not considered suitable for

shallow land burial, but exactly what manner of disposal is

acceptable was not specified in 10CFR61.

The LLWAA "charged the DOE with the safe disposal .

of GTCC waste from any generator,"56 but the DOE was not

prepared to make any plans for its disposal at that time.

The DOE position maintained that it would require several

years to set up a GTCC program and then an additional 8 to






27

10 years before a permanent disposal site could be

constructed. The NRC, on the other hand, has taken steps

toward the immediate resolution of the GTCC problem. The

NRC, to whom OCRWM is answerable under the provisions of the

NWPA, has issued a final ruling which stipulates that

"radwaste that exceeds the upper bounds for Class C low-

level waste must be disposed of in a geological repository

suitable for high-level waste disposal,"57,58 unless

another suitable facility is designed by the DOE and

approved by the NRC. Since considerable quantities of GTCC

waste are expected from reactor decommissioning activities,

the development of a separate facility is a serious

possibility, but no firm plans to develop such a facility

have been announced.

A significant portion of potential GTCC wastes fall

into two categories, Non-Fuel Assembly hardware and Spent

Fuel Disassembly hardware, two waste streams which will be

further discussed in Chapter 2. Both the NFA hardware and

SFD hardware waste streams are directly related to the spent

fuel waste stream, both are subjected to a similar

irradiation history, and both are constructed of similar

materials. The activated metals which comprise these waste

streams will be classified as either Class C or GTCC LLW.

The approximate radioisotopic contents of many of these

components have been calculated by the ORIGEN2 code and

included in the Characteristics Data Base59 (CDB), so an

estimate of the classification of these waste components can






28
be made using the CDB. A preliminary analysis of this data

performed by the author in preparation for this work

indicated that most, if not all, of these hardware elements

will be GTCC waste.

In most of the hardware elements delineated in the CDB,

the single worst element, i.e. the element which has most

often exceeded its Class C limit, is 9Nb. Niobium has a

half life of 20,000 years, so materials containing 9Nb

remain radioactive for a long period of time. The decay of

niobium has a high decay energy (2.1 MeV), 75% of which is

emitted as gamma radiation. Niobium also has a high degree

of solubility in water which causes niobium to pose a

potential leaching problem. For these reasons, niobium has

been given a very low Class C limit of only 0.2 curies per

cubic meter. In certain materials like both Zircaloy-2 and

4, niobium is only present at impurity levels, but for the

irradiation levels to which the hardware components are

exposed, "trace quantities may be sufficient for the

irradiated material to exceed Class C limits."59 In some

other materials, there is an even greater concentration of

niobium, because

there has also been a rapid increase in recent
years in the use of niobium in the steel industry.
Small amounts of niobium markedly increase the
yield strength of mild steel plates and prevent
weld decay and intergranular corrosion in
stainless steels; In a similar way, the
addition of niobium can increase the high-
temperature strength of high-strength heat-
resisting steels and superalloys such as are used
in gas turbines and similar environments.6






29

Thus, the quantities of niobium used in the various reactor

metals is very important to the eventual determination of

the activated metals' waste classification.

Of the other three critical isotopes, only 5Ni

possesses a half life of a magnitude similar to that of

9Nb, namely 80,000 years. However, the decay of 59Ni is

less energetic (1.07 MeV) and releases no gamma radiation,

so it was given a higher Class C limit of 220 curies per

cubic meter. On the other hand, 6Ni is a comparatively

short-lived nuclide with a half life of 92 years. Because

it is expected to decay away in a relatively short time

span, it has a correspondingly high Class C limit of 7000

curies per cubic meter, the highest allowed concentration of

any critical isotope in an activated metal. Finally, 14C

falls in between these two extremes with a half life of 5730

years, and a Class C limit of 80 curies per cubic meter.

All three of these nuclides are found in most reactor

materials such as inconel and stainless steel. All three

also often exceed their respective Class C limits, but, in

most cases, not before niobium has already done so.

The foregoing discussion serves to illustrate the

complexity of the radioactive waste management issue.

Whereas some aspects of radwaste disposal are technically

challenging, the most difficult problems for radwaste

programs to overcome are political in nature. Some of the

most important milestones for radwaste disposal are

summarized in Table 4, the clear majority of which are














Table 4. A chronology of some of the important dates
concerning radioactive waste disposal.

1955 First conference on the Peaceful Uses of
Atomic Energy.

1957 First commercial nuclear power reactor.

1962 First commercial LLW site at Beatty, NV.

1972 First attempt by Congress to establish a
national radioactive waste program.

1979 Washington and Nevada temporarily close LLW
sites; South Carolina reduces allowable
disposal at Barnwell.

1980 Low-Level Waste Policy Act (LLWPA) enacted.

1983 Nuclear Waste Policy Act (NWPA) enacted.

1985 Low-Level Waste Policy Amendments Act
(LLWAA) enacted.

1987 Nuclear Waste Policy Amendments Act (NWPAA)
enacted.

1991 Potential start-up of test phase at WIPP.

1993 Deadline for the development of Compact LLW
sites as set by the LLWAA.

1995 Expected start-up date for most Compact LLW
sites.

1998 Contractual deadline for the Department of
Energy to begin accepting SNF from the
utilities as set by the NWPA.

2003 First operational date of the Federal HLW
Repository as set by the NWPAA.






31

political events or the outcome of political events. If

programs like WIPP and the Compact LLW disposal sites are

successful, then progress in the disposal of HLW and SNF may

become politically possible.

Before proceeding, mention should be made of two

further waste categories. The first is mixed wastes, or

wastes which are considered to represent both a radioactive

and nonradioactive toxicity hazard. Due to the nature of

these wastes, they fall under the jurisdiction of both the

NRC and the EPA. The resulting regulatory situation is

complicated, often contradictory, and beyond the scope of

this work. The second category is Below Regulatory Concern

(BRC) radioactive waste. This is waste which contains

radioactivity at such low levels as not to be considered a

significant hazard. The NRC has recently drawn considerable

criticism for issuing a BRC policy which would classify some

Low-Level Wastes as BRC wastes.61 Discussion of this waste

category is also beyond the scope of this work.

Notes


1"World List of Nuclear Power Plants," Nuclear News
34, no. 2 (February 1991): 53-72.

2John Greenwald, "Time to Choose," Time 137, no. 17
(April 29, 1991): 61.

3Y. S. Tang and James H. Saling, Radioactive Waste
Management (New York, NY: Hemisphere Publishing Corporation,
1990), 1.

4Peaceful Uses of Atomic Energy: Proceedings of the
International Conference in Geneva. Switzerland. Auqust 8-
20. 1955, vol. 9, Reactor Technology and Chemical Processing
by the United Nations (New York, NY: United Nations, 1956),
3-32, 669-726.








5Tang, 15.

6Tang, 2.

7Integrated Data Base for 1990: U.S. Spent Fuel and
Radioactive Waste Inventories. Projections, and
Characteristics, DOE/RW-0006, Rev. 6, (Oak Ridge, TN: Oak
Ridge National Laboratory, October 1990), 125.

8Tang, 227.

9Integrated Data Base, 129.

1Integrated Data Base, 3.

"Tang, 278.

12Tang, 2-12.

13Integrated Data Base, 2.

'4Integrated Data Base, 75.

'5Integrated Data Base, 81.

16Integrated Data Base, 75-9.

1Tang, 187-92.

'896th U.S. Congress, Department of Energy National
Security and Military Application of Nuclear Energy
Authorization Act of 1980, Public Law 96-164, 1980.

19"Late News in Brief," Nuclear News 31, no. 13
(October 1988): 17-8, 107-8.
20"DOE Endorses WIPP Tests," Nuclear News 33, no. 10
(August 1990): 95.
21"EPA Announces Approval of Disposal Test at WIPP,"
Nuclear News 33, no. 15 (December 1990): 77-8.
22"Land Withdrawal Okayed, Giving DOE Access to WIPP,"
Nuclear News 34, no. 3 (March 1991): 78-80.

297th U.S. Congress, Nuclear Waste Policy Act of
1982, Public Law 97-425, 1983.

4Integrated Data Base, 81.

25Tang, 13.

26Kenneth C. Lish, Nuclear Power Plant Systems and
Equipment (New York, NY: Industrial Press Inc., 1972), 117.








27Tang, 54.

28Tang, 6.

"Characteristics of Spent Fuel. High-Level Waste, and
Other Radioactive Wastes Which May Require Long-Term
Isolation, DOE/RW-0184 (Oak Ridge, TN: Oak Ridge National
Laboratory, December 1987), 1:3.1-11.

30Tang, 100.

31H. Bastien Thiry, J. P. Laurent, and J. L. Ricaud,
"French Experience and Projects for the Treatment and
Packaging of Radioactive Waste From Reprocessing
Facilities," in Radioactive Waste Management: Proceedings of
an International Conference on Radioactive Waste Management
Held in Seattle. 16-20 May 1983, by the International Atomic
Energy Agency (Vienna: International Atomic Energy Agency,
1984), 221-7.

32Tang, 106-22.

3Tang, 103.

4Integrated Data Base, 5, 47.

35Integrated Data Base, 47-9.

3Tang, 93.

37John Graham, "Amending NWPA to Prevent a
Moratorium," Nuclear News 31, no. 3 (March 1988): 42-7.


""Late News in Brief," Nuclear News 32, no. 1
(January 1989): 17-8, 131-2.

39John Graham, "The DOE HLW Program: Rough Sledding
Ahead," Nuclear News 30, no. 3 (March 1987): 44-6.

0U.S. Department of Energy, Report to Congress on
Reassessment of the Civilian Radioactive Waste Management
Program DOE/RW-0247 (Washington, DC: U.S. Department of
Energy, November 1989), 8-11.
41"DOE Sues Nevada Over Yucca Mountain Stalling,"
Nuclear News 33, no. 3 (March 1990): 73.
42"Supreme Court Lets Stand Voiding of Nevada 'Veto',"
Nuclear News 34, no. 5 (April 1991): 63.

43U. S. Department of Energy, Nuclear Waste Fund Fee
Adequacy: An Assessment DOE/RW-291P, (Washington, D.C.: U.S.
Department of Energy, November 1990), 2.






34

""DOE Issues Assessment of Nuclear Waste Fund Fee
Adequacy," OCRWM Bulletin DOE/RW-0301P (December 1990 /
January 1991): 1.

45"GAO Cites Cost Hikes, Delays in DOE Program,"
Nuclear News 30, no. 14 (November 1987): 94-5.

U.S. DOE, Nuclear Waste Fund Fee Adequacy, 11-4.

47U.S. Code of Federal Regulations, Title 10, Section
961, Standard Contract for Disposal of Spent Nuclear Fuel
and/or HiQh-Level Radioactive Waste.

4U.S. DOE, Nuclear Waste Fund Fee Adequacy, 9.

4"Utilities Win Waste Fee Suit," Nuclear News 29, no.
1 (January 1986): 95.

50"Court Rules That DOE Overcharged for NWF," Nuclear
News 32, no. 7 (May 1989): 88.

51U.S. Code of Federal Regulations, Title 10, Section
61, Licensing Requirements for Land Disposal of Radwaste.

5210CFR61, 61.55.

3Michael E. Burns, ed., Low-Level Radioactive Waste
Regulation (Chelsea, MI: Lewis Publishers, Inc., 1989), 39-
46.

5496th U.S. Congress, Low-Level Radioactive Waste
Policy Act, Public Law 96-573, 1980.

5599th U.S. Congress, Low-Level Radioactive Waste
Amendments Act, Public Law 99-240, 1986.

56"Nuclear News Briefs," Nuclear News 30, no. 6 (April
1987): 13-4, 115-6.

57"Nuclear News Briefs," Nuclear News 31, no. 8 (June
1988): 21-2, 147-8.

58"Greater-than-Class-C Included by NRC Rule," Nuclear
News 32, no. 9 (July 1989): 84.

5Characteristics of Radioactive Wastes, DOE/RW-0184,
1:2.7-5.

"F. Fairbrother, The Chemistry of Niobium and
Tantalum (London: Elsevier Publishing Company, 1967), 4.
61"NRC Approves BRC Policy; NUMARC Defers Action,"
Nuclear News 33, no. 10 (August 1990): 93-4.














CHAPTER 2
DOMESTIC HARDWARE WASTE HANDLING

Due to the unavailability of commercial fuel

reprocessing and the lack of an acceptable facility for SNF

disposal, i.e. the federal repository, SNF is continuing to

accumulate at the nation's nuclear reactor sites. As a

result, many utilities are currently faced with the problem

of storing far greater quantities of spent fuel than their

spent fuel pools are capable of accommodating.' When most

of these plants began construction,

spent-fuel storage was not considered to be a
problem [as] it was assumed that all spent
fuel would be reprocessed. Government decisions
in the 1970's and the economics of the 1980's,
however, served to eliminate the reprocessing
option in the United States, thus putting the
burden of spent fuel storage onto utilities that
were not prepared for it. The spent-fuel pools of
many older units simply could not accommodate all
spent fuel [generated] over the life of the
plant.2

A 1988 study published by NARUC (the National Association of

Regulatory Utility Commissioners) found that 55 reactor

sites were expected to reach their authorized capacity by

1998. The report further stated that if all 55 of these

reactors made full use of reracking and rod consolidation,

29 units would still exhaust their pool capacity by 1998.1

Congress acknowledged this difficulty in the NWPA by

including provisions whereby






36

the DOE may provide Federal Interim Storage (FIS)
for spent fuel up to a total of 1900 metric tons
of heavy metal [about 19 cores] for nuclear power
plants that have done everything possible to
provide storage for their spent fuel and through
no fault of their own cannot do so and would have
to shut down if additional storage were not
provided.3

Nevertheless, the initial efforts were required of the

utilities who were thus forced to find methods of increasing

the available spent fuel storage space. The option for FIS

expired in January 1990 without being utilized. The amount

of space required at any given plant will depend on the age

of the plant and on when the repository (or an MRS facility)

is actually able to receive spent fuel.

Nuclear reactors maintain a "full-core reserve" in

their spent fuel pools. This is not considered a safety

matter, but instead allows added operational flexibility,

such as being able to unload the entire core during pressure

vessel inspections.4 The obvious drawback to maintaining a

full core reserve is the reduction of available storage

space in the spent fuel pool. To relieve this storage space

problem, the utilities have examined several alternatives

for both long-term and short-term solutions. The short-term

solutions are spent fuel pool reracking, intra-utility

transshipment, and fuel rod consolidation, while long-term

solutions include the construction of additional spent fuel

storage pools, and modular dry storage. Reracking has been

and will continue to be used extensively to provide extra

time to explore other, long-term alternatives. Transship-

ment is also commonly used where possible, but is only






37

available to utilities which have multiple reactors or an

alternate storage pool.

Reracking involves the replacement of the existing

spent fuel storage racks with new racks to decrease dead

space and increase storage capacity. The original spent

fuel rack designs were conservatively designed with regard

to criticality concerns at the direction of the NRC, but 30

years of commercial reactor experience, the use of poisons

within the rack structure, and improved calculational tech-

niques have allowed for new racks to be installed while

still meeting all safety and seismic requirements. By

taking into account the burnup of the fuel and by inserting

borated steel plates between storage cells whenever

necessary, the number of fuel assemblies that can be stored

in the pool can be increased by a factor of four or five.4

For example, reracking performed by the Commonwealth Edison

Company resulted in an increase in "spent-fuel storage at

LaSalle-2 from 1080 to 4078 assemblies."5 Reracking

requires NRC review and approval, but the only real limiting

concern is the ability of the fuel pool to handle the

additional floor loading.

The second short-term solution, transshipment, is the

cheapest option for increasing a reactor's spent fuel

storage space and, when available, is usually the first

option to be exploited. Transshipment is the transfer of

spent fuel from one reactor's spent fuel pool to another

reactor's spent fuel pool. When a utility has new reactors






38

coming on-line and older reactors already on-line, the space

in the new reactor's pool can be used by both reactors to

buy time. Additionally, in some cases where a new reactor

is being built by a utility which has other nuclear units,

the spent fuel pool of the new reactor is being

intentionally oversized to allow room for fuel from the

older reactor to be stored there. When used together,

transshipment and fuel pool reracking can supply the needed

storage space for several year's worth of spent fuel.

However, the storage space available through these methods

is strictly limited and is, in most cases, insufficient to

accommodate all the spent fuel generated before the Federal

Repository goes on-line.

After transshipment and reracking possibilities have

been exhausted, three other options still remain. One such

option is to build an additional spent fuel storage pool or

to expand an existing one, but this option is more expensive

than both modular dry storage and rod consolidation. Also,

due to a spent fuel pool's reliance on active cooling

systems, the spent fuel pool requires additional maintenance

and safety systems. Due to these considerations, no new

pools have been built to date.

Thus, a second, more flexible option is modular dry

cask storage. When used,

dry storage employs passive cooling of spent
nuclear fuel in large metal casks, vaults,
drywells, or concrete silos. Following storage in
pools for several years, several intact or
consolidated spent nuclear.fuel assemblies can be
loaded into the dry storage modules for extended






39

storage above ground at nuclear power plant
sites.6

This method has the advantage of being modular; additional

storage space can be added one unit (be it cask, silo, or

whatever) at a time as additional storage capacity is

required. However, each additional unit is an additional

capital outlay and, for some reactors, a large number of

such casks will be required. Dry storage removes "cooled"

fuel from the spent fuel pool providing space for hotter,

fresh fuel assemblies. The storage modules, whether casks,

vaults, or silos, will need to be stored within the

exclusion area around the reactor. Only the three San

Onofre reactors are expected to have trouble accommodating

dry storage modules. These reactors are situated on federal

land, the lease for which prohibits the storage of SNF at

any location other than within the spent fuel pool.

However, in order to prevent the premature shutdown of these

plants, the government may lift this restriction to permit

dry storage before the storage capacity of the fuel pools is

exhausted in 1996.2

The passive cooling requirement for dry storage means

that no active or moving components, such as pumps or fans,

are required to cool the assemblies within these modules.

The possibility of the storage casks also doubling as

shipping containers is also being considered. If possible,

the casks would then serve double duty, and would reduce the

amount of fuel handling that will be required. Several

demonstration projects for dry cask storage were performed






40

in the mid-1980s in cooperation with the DOE and the

Electric Power Research Institute (EPRI).' In July 1986,

Virginia Power became the first utility to be licensed for

dry cask storage.8 Castor V metal storage casks which are

manufactured by General Nuclear Systems, Inc. and hold 21

intact PWR fuel assemblies are in use at the Surry Nuclear

Power Stations,9 as are Nuclear Assurance Corporation NAC-

128 S/T casks and Westinghouse MC-10 casks. Including the

two Surry reactors, fifteen reactors operated by eight

utilities had already opted for modular dry storage by

January 1991.10 In July 1990, the NRC issued new

regulations which permit dry storage to be conducted in

certain approved casks under the general plant license.

This eliminates the need for a special license and licensing

procedure for such storage in the approved casks," which

should make dry storage an even more attractive interim

storage option for many utilities.

The third and final option which has been investigated

is spent fuel rod consolidation which, like transshipment

and reracking, relies on making more efficient use of

existing storage capacity. This option has the advantage

that

rod consolidation technology is an extension of
previous experience with the reconstitution of
[failed] fuel assemblies in storage pools, and is
an alternative for [increased storage in] pools
that have sufficient structural strength to safely
support the added weight.6

Hence, rod consolidation is not so much a new idea as it is

an extension of an old idea. In execution,






41

rod consolidation is a process that involves
dismantling the fuel assembly and rearranging the
spent-fuel rods into a close-packed geometry in a
storage canister. As a storage technology, rod
consolidation has the potential to double the
existing water basin storage capacity.6

Cooperative studies in rod consolidation have also been

conducted and "target consolidation ratios of 2:1 or better

have been demonstrated,"2 excluding SFD hardware.

The process of rod consolidation involves dismantling

the fuel rod assemblies and placing the fuel rods into

canisters designed for this purpose. The consolidation

equipment is designed to extract the fuel rods from the

assembly and then place the rods into a consolidation

canister, changing them to a triangular pitch in the

process. The triangular pitch allows more efficient packing

of the fuel rods, which is necessary to achieve the 2:1

consolidation ratio. The majority of the designs extract

one fuel rod and handle one assembly at a time, but there

are a few designs which operate differently. U.S. Tool and

Die, Inc. has a design, in which "rods are pulled by rows

from the fuel assembly and guided down directly into the

canister through a stationary funnel."13 The rods are

placed in the triangular array inside the canister.

Westinghouse has a design which transfers all the fuel rods

from an assembly simultaneously. The rods are moved to a

transition canister, mechanically rearranged into a

triangular array, and then placed in the storage

canister.14 The Fuel Master system, designed jointly by

Babcock & Wilcox and Numatec, pulls only one rod from an






42

assembly at a time, but works on two assemblies

simultaneously.15 The major differences between these

systems are the speed with which they work and the amount of

automation used in the process.

As a storage technology, rod consolidation has both

advantages and disadvantages, just like dry cask storage.

On the positive side, consolidation is an extension of

existing technology/methodology, and should therefore be

easier to implement than dry storage. Consolidation has the

potential to double the storage capacity of the reactor's

existing spent fuel pool, thus minimizing the capital outlay

required of the utility. There is only the one time cost of

consolidation equipment, which should theoretically cost

less than a number of casks of equal capacity. As mentioned

earlier, consolidation also doubles the fuel pool floor

loading which could be a limiting condition for some

reactors. However, rod consolidation techniques can also be

used in combination with the dry cask storage techniques (if

the particular cask is designed to accommodate consolidated

fuel, specifically the additional heat load and increased

radiation dose rate associated with the consolidated fuel)

in which case only half the number of casks will be required

for any particular storage requirement. In early 1988, in a

test at INEL jointly sponsored by DOE and EPRI, a

Transnuclear TN-24P metal storage cask was loaded with 24

consolidated fuel canisters. The cask performed well in

terms of both heat transfer and radiation shielding. Of






43

particular note was that the gamma dose rate for the

consolidated fuel was significantly lower than the gamma

rate for intact fuel because the primary gamma sources, i.e.

the SFD hardware elements, had been removed.16 A 1988

survey of dry storage cask designs further indicated that

the majority of cask designs could accommodate consolidated

fuel canisters.17 If a similar performance can be achieved

when rod consolidation is used with shipping casks, only

half the number of shipping casks and half the number of

shipments would be required. This, of course, excludes the

SFD hardware generated by consolidation which may be shipped

to a separate facility for disposal. Rod consolidation thus

has the potential for use within the FWMS, regardless of its

use or lack of use by the utilities. Consolidation

performed at an MRS facility would result in a reduction in

the number of spent fuel shipments required, while

consolidation performed at the repository would reduce the

number of waste packages required.

On the negative side, some reactors may not be able to

use rod consolidation, because the structural base of the

spent fuel pool is not designed to carry the added weight of

the consolidated spent fuel. However, they might be able to

use consolidation in conjunction with appropriately designed

dry storage casks, as previously mentioned. Rod

consolidation has also been shown to have adverse effects on

reactor operations in two primary ways. First, assemblies

tend to develop a film of crudd" while in the reactor core






44

which, when the fuel rods are pulled from the assembly, is

broken loose. This debris reduces the visibility in the

spent fuel pool and could result in higher exposure rates to

the operators.18 Secondly, consolidation activities are

extremely time-consuming and the man-hours required to

perform consolidation operations would place a severe

financial burden on the plant, as well as interfere with

other required work around the reactor site.19

Consolidation also increases the risk of damaging or

breaking fuel pins when they are withdrawn from the

assembly. These damaged pins would then require special

treatment and pose an additional storage problem. Only a

few of the thousands of fuel pins consolidated during the

course of the various demonstration programs were damaged,

in spite of several of the programs purposely consolidating

warped or deformed rods. This demonstrates that rod

breakage is not a major factor; nevertheless, the concern

cannot be entirely dismissed.

A further difficulty revealed by the demonstration

programs is that the rod consolidation process is not as

efficient as was originally anticipated. Whereas the

desired 2:1 consolidation ratio has been successfully

achieved for the spent fuel rods, the consolidation process

creates another waste, the fuel skeleton, the disposal of

which has not been adequately demonstrated. If the design

compaction goal for these skeletons (10:1) could be

achieved, consolidation would result in a 40% increase in






45

storage space as shown in Table 5. However, the consoli-

dation demonstration programs have had a difficult time

reliably achieving this ratio. Accordingly, current

consolidation technology produces less than a 40% increase

in available space and, at this point, becomes economically

questionable. Two alternative methods for handling this

hardware are being studied. One option is to store the

compacted hardware in canisters above the spent fuel rack.

The most important consideration for this method is

maintaining an optimum depth within the pool to limit

personnel exposure. The second option is immediate disposal

of the SFD hardware at LLW sites. However, this option is

contingent upon the waste classifying as LLW, an issue which

is currently being studied and whose evaluation is one major

purpose of this work. Finally, the continuing delays in the

repository scheduling have further hurt rod consolidation in

that consolidation alone is no longer sufficient to provide

adequate additional storage space at most reactor sites.

Accordingly, most utilities have decided to pursue dry cask

storage initially and to reconsider the use of consolida-



Table 5. Typical generation of storage space by rod
consolidation (in terms of fuel assembly spaces).

Assemblies to be Consolidated: 10
Consolidated Canisters (2:1) : 5
Compacted Skeletons (10:1) : 1
Empty Spaces Created : 4

Increase in Storage Space:
(4/10) X 100% = 40%






46

tion, possibly in conjunction with dry cask storage,

sometime in the future.18'19

As of April 1991, no utility has announced its

intention to pursue a full-scale rod consolidation campaign.

Those utilities which had announced such plans have since

canceled them for the foreseeable future. The most notable

example of this is Northern States Power which had intended

to conduct a 1000 assembly campaign at their Prairie Island

Nuclear Generating Plant,18 but have since switched to using

dry storage in Transnuclear TN-40 metal casks.10 Neverthe-

less, as rod consolidation may be used in the future by

either utilities or the FWMS, an analysis of the resulting

waste stream, SFD hardware, is in order.

When all of the fuel rods have been removed from the

fuel assembly, the empty assembly skeleton remains behind.

The components which comprise the skeleton include, but are

not limited to, guide tubes, grid spacers, and top and

bottom nozzles and are known collectively as Spent Fuel

Disassembly hardware. Figure 1 shows the upper end fitting

for a Combustion Engineering PWR fuel assembly. The diagram

provides a detailed illustration of just a few of the SFD

hardware components associated with a fuel assembly.

Additional spacer grids, typically a total of eight to ten,

are spaced evenly over the length of the assembly. The

bottom of the assembly also has a lower end fitting similar

in size and material to the upper end fitting. These

components were previously expected to be disposed of as






END FITTING


Exploded


View


---oc king Post


Holddown
-- Plate

Holddown
Spring

Flow Plate


CEA Guide
STube


Spacer
SGrid


Figure 1. Exploded view of a Combustion Engineering Upper
End Fitting on a PWR fuel assembly.






48

part of the spent fuel assembly, but when rod consolidation

is used, these components are no longer associated with the

fuel rods and must be handled separately. The empty

assembly skeleton as a whole will most likely classify as

GTCC LLW as will many of the individual components which

make it up. However, due mainly to differing materials of

fabrication, some of these components may not be GTCC, so

with proper hardware sorting, the quantities of GTCC LLW,

i.e. waste that must go to the Federal Repository, could be

minimized. Proper classification of these hardware elements

is one of the goals of this dissertation and is discussed in

greater detail later in the "Domestic Hardware Analysis"

section of this chapter.

Another frequently overlooked waste stream which has an

impact on both the spent fuel storage problem at utility

reactor sites and on the repository planning is Non-Fuel

Assembly hardware. This waste stream is just one portion of

what utilities consider operational waste, the waste

generated through the daily operations of a nuclear plant.

NFA hardware includes all the hardware components that are

related to the reactor core and/or assemblies, but are not

considered to be or to contain fuel. These include, but are

not limited to, burnable poison rods, control rods, neutron

sources, and incore detectors. Figure 2, Figure 3, and

Figure 4 illustrate some of the types of NFA hardware which

must be considered. These components have operational

lifetimes associated with them and must be replaced








CHANNEL


SFabtener
Assembly
pacer ton
Button


CRUCIFORM





B Roller
: Bearing


Figure 2. BWR Non-Fuel Assembly Hardware.


periodically. In most cases, the discarded hardware is
stored in the spent fuel pool pending disposal, either loose
or within void spaces in spent fuel assemblies, i.e. in
empty guide tubes. Control rods in particular are often
stored within the assemblies and many utilities expect them
to be disposed of in this manner. Two problems exist with
this approach. First, the dimensions used in current
shipping cask designs are insufficient to allow for anything














Sp
B


CONTROL ELEMENT ASSEMBLIES


CONTROL ROD CLUSTER



iSpring


Sdys


i T I A


0 E.t

4 Elemnf


\m I "


Retainer


8 Elenat


YY Y Y YU


12 Elent


Figure 3. PWR Control Rod Assemblies. The left diagram
illustrates a Westinghouse Control Rod Assembly while
the right diagram shows three different configurations
used for Combustion Engineering Control Element
Assemblies.


larger than the spent fuel assemblies. Control rod
assemblies, when inserted into fuel assemblies, increase the
assembly length by several inches and would prevent the
shipping cask from closing. To a lesser extent, spent fuel
consolidation also poses a problem for hardware storage.
Since consolidation removes the space available for storage


Sot


03


1.



































Figure 4. Westinghouse Thimble Plug Assembly.


within the assemblies, alternative storage space must be

provided, usually at the expense of spent fuel pool space.

Regardless of the status and extent of consolidation,

NFA hardware must be taken into account as a separate waste
stream needing final disposal. Under the terms of the
"Standard Contract for Disposal of Spent Nuclear Fuel and/or
High-Level Radioactive Waste" entered into by the DOE and
the individual nuclear utilities, the DOE is obligated to


Thimble Plug






52

accept for disposal all non-fuel components. Specifically,

the contract states that

non-fuel components including, but not limited to,
control spiders, burnable poison rod assemblies,
control rod elements, thimble plugs, fission
chambers, and primary and secondary neutron
sources, that are contained within the fuel
assembly, or BWR channels that are an integral
part of the fuel assembly, which do not require
special handling, may be included as part of the
spent nuclear fuel delivered for disposal pursuant
to this contract.2

The contract then proceeds to state that other components

which do not meet these guidelines will be accepted as non-

standard fuel. The waste classification of this hardware is

unclear, but is certainly either Class C LLW or GTCC waste.

Resolution of this matter is another goal of this

dissertation and is discussed throughout the remainder of

this work.

In order to plan and implement the disposal of the NFA

and SFD hardware waste streams properly, several pieces of

information are necessary, including the hardware's physical

and radiological characteristics. The hardware's length,

width, and susceptibility to crushing determine the volume

that this waste will occupy. The weight of the components

is important for transportation planning and determining

crane loads at the disposal site. The initial materials of

construction, and the isotopic composition of those

materials, can be used to predict the radioisotope

concentrations after irradiation. However, the initial

concentrations of some isotopes exhibit considerable

variations and are often not measured at all, particularly






53

in the case of trace elements like niobium, thus making it

difficult to predict the future radiological characteristics

of these isotopes. Alternatively, in the absence of

detailed isotopic compositions, direct or indirect

measurement of the radioisotopes of interest can be

performed after irradiation. Whatever the process used,

these radioisotope concentrations are necessary for the

accurate determination of the hardware waste classification

according to 10CFR61. The curie content of the hardware,

either by direct measurement or by calculation, is also

important for shielding purposes at all stages of disposal.

The most important datum, however, the quantity of

individual hardware items to be disposed of, is also the

most difficult to determine.

To predict the number of NFA hardware elements for

disposal accurately requires information that is not easily

located in the open literature. First, the types and number

of NFA hardware components that are used at each reactor

must be determined. These values vary widely from one

reactor to another due to reactor design changes and

differing utility operational practices, as well as

differing reactor vendors. NFA hardware, like reactor fuel

assemblies, has also evolved over the life of the plant, so

that the initial configuration may bear little resemblance

to the current configuration. Hardware lifetimes represent

another important facet of the problem. Hardware vendors

can provide design lifetimes for all of the hardware that






54

they manufacture, but the utility of this information is

questionable or at least in need of verification. In actual

practice, manufacturers' proposed lifetimes are frequently

not met either through premature component failure or

through preventive replacement, i.e. purposely replacing the

component before its design lifetime elapses. Additionally,

the changing needs of the reactor may cause the use of some

components to be discontinued which may a) cause some of

these components to be discharged earlier than expected and

b) result in fewer total components discharged than

expected. For example, at the Prairie Island Nuclear

Generating Plant near Minneapolis-St. Paul, Minnesota, "all

use of BPRA [burnable poison rod assemblies], thimble plugs,

and source assembly inserts has been discontinued."18 Many

hardware lifetimes are also given in terms of Effective Full

Power Days, and therefore depend on the operational history

of the reactor. Finally, to complicate the issue further,

many reactor sites have shipped some or all of their NFA

hardware components directly to LLW sites for disposal.

Whereas this policy helps to relieve the anticipated burden

on the FWMS, it makes the determination of how much will

require disposal in the future that much more difficult.

On a single reactor basis, quantities of SFD hardware

are usually somewhat easier to predict. The quantity of SFD

hardware is directly related to the number of assemblies

which are consolidated, which is dependent upon both the

utility and the specific reactor site. The number of






55

assemblies that can be consolidated is strictly limited to

the number of assemblies stored and generated on-site,

barring transshipment from other reactors. Each assembly

consolidated will produce one assembly skeleton of SFD

waste. However, even when given the quantity of assemblies

consolidated, SFD hardware quantities could still show a

considerable variance based on the fuel assembly type.

Seventy-nine fuel assembly types have been identified by the

CDB,21 and each type has its own specific hardware

quantities. There is a significant difference between the

SFD hardware weights of a BWR assembly and a PWR assembly

and, in general, between any two PWR or BWR assemblies.

Typical total weights of all SFD hardware components within

a PWR assembly range from 24.6 kilograms to 44.0 kilograms

per assembly, while BWR assembly hardware weights range from

8.0 kilograms to 16.8 kilograms per assembly.22 As can be

seen from these figures, BWR assemblies have roughly one-

third the SFD hardware, by weight, of PWR assemblies, which

is as expected due to the smaller size of BWR assemblies.

Since no utilities are currently pursuing fuel rod

consolidation, the only SFD hardware currently of concern is

that generated by the consolidation demonstrations.

However, if more utilities decide to utilize consolidation,

these quantities could become almost as difficult as NFA

hardware quantities to predict.

As a portion of the effort directed toward the

development of the repository, waste characterization has






56

become a major research area. The majority of the

characterization effort has been directed toward HLW and

SNF, so there is incomplete information on NFA and SFD

hardware. Even reprocessing studies which produce both SFD

hardware and cladding hull wastes generally only include a

brief mention of this hardware waste, when it is mentioned

at all. There has been an increase in interest in this area

recently, however, so information has become more available.

What follows is a review and analysis of that information.


Characteristics Data Base

The first significant source of data to be developed

which dealt with NFA hardware and SFD hardware was the

Characteristics Data Base (CDB). The CDB was developed in

1987 by Oak Ridge National Laboratory (ORNL) to serve as a

standard data source for information on all forms of waste

which may be disposed of in the federal repository.3 In

the summer of 1988, the author performed an analysis of the

CDB for his practicum research at ORNL. A classification

scheme for LWR fuel assemblies resulted from this work.2

The CDB is composed of five user-oriented, menu-driven

data bases which can be used on any IBM PC or compatible

computer. The data bases contain information on the

physical and radiological characteristics of the radioactive

wastes which are expected to be emplaced in the Federal HLW

Repository for final disposal. The five data bases which

comprise the CDB are 1) the LWR Assemblies Data Base, 2) the








LWR Radiological Data Base, 3) the LWR Quantities Data Base,

4) the LWR NFA Hardware Data Base, and 5) the High-Level

Waste Data Base. General descriptions of the contents of

these data bases are provided in Table 6. Only two of these

data bases are directly applicable to this work, the LWR

Assemblies Data Base and the LWR NFA Hardware Data Base.



Table 6. Summary of the contents of the five data bases of
the Characteristics Data Base.


Data Base


LWR Assemblies








LWR Radiological




LWR Quantities




LWR NFA Hardware





High-Level Waste


Contents


Physical descriptions of intact fuel
assemblies (dimensions, weight, number
of fuel rods, materials, etc) for 58
distinct assembly types at time of
publication; Radiological properties
(curie content, isotopic content, and
heat generation rate) and limited
physical properties (weight and
location) of SFD hardware elements.

Radiological properties (composition,
curie content, and heat rate) of
intact fuel assemblies as a function
of assembly type, burnup, and decay
time.

Historical inventories and projected
quantities of discharged LWR fuel
assemblies based on Energy Information
Administration data and predictions.

Physical descriptions (dimensions,
weight, materials, etc) and
radiological properties (curie
content, heat rate, composition) of
NFA hardware by vendor, hardware type,
and reactor.

Physical descriptions (chemical and
isotopic compositions, canister type,
and age) of current HLW inventories
for both commercial and defense
wastes; General characteristics of
immobilized waste forms based on base-
line solidification programs.






58

The data bases were originally developed using dBase

III+, then the menus were written and compiled separately

using a dBase compiler called Clipper. When using any of

the five data bases, it is the compiled version that is

being used. Use of the menu-driven files eliminates the

need for any knowledge of the dBase program, but as a direct

consequence, limits the amount of data which can be

extracted from the individual dBase data files which

comprise the menu-driven data bases. Accordingly, for the

purposes of this work, the individual data files were

accessed directly to take maximum advantage of the available

resources.

Several diverse sources of data were used to develop

the data bases of the CDB, but only two types of sources

were important to the LWR Assemblies Data Base and the LWR

NFA Hardware Data Base. Both of these data bases derived

their physical properties data from materials provided by

the original fuel vendors and their radiological information

from the ORIGEN2 computer code. The information on the

physical properties of the assemblies and the related

hardware was sought from the vendors of the original

components. Individual contracts were entered into between

ORNL and each supplier including GA Technologies, Babcock &

Wilcox, Combustion Engineering, Westinghouse, and Advanced

Nuclear Fuels Corporation (formerly Exxon) under the terms

of which the corporations provided all available information

on all the fuel and reactor hardware that they had ever






59

manufactured and/or sold.25 The quality and quantity of

information provided by each corporation varied considerably

resulting in very inconsistent records within the CDB. In

addition, General Electric (GE) declined to enter into a

contract, so very little information on GE fuels or hardware

is included in the CDB. However, the CDB represents one of

the only sources of information for this data and, in spite

of its deficiencies, is by far the largest compilation of

this nature made to date.

Both the LWR Assemblies Data Base and the LWR NFA

Hardware Data Base also include an extensive file containing

radiological information which can be applied to the

hardware elements. The radiological data was provided by

the ORIGEN2 computer code, ORIGEN2 being an acronym for Oak

Ridge Isotope Generation and Decay code Version 2.2 A

detailed model of the specific reactor conditions is

required for each calculation, and as the code was

originally developed to predict the isotopic content of fuel

assemblies after irradiation, the most accurate models

available are for the core region. Therefore, the

information for hardware in the fueled region is more

accurate than that for hardware in the top, bottom, or gas

plenum regions. At the time of publication of the CDB,

efforts to improve the modeling of the regions immediately

outside the active core region were ongoing.27

Furthermore,

the flux decreases significantly in the two zones
adjacent to the core zone [while] the effective






60

cross sections outside the core zone increase up
to 570%, depending on the element (Co, Ni, Nb, or
N), the zone, and the reactor type. This increase
is presumably due to resonance and a higher
fraction of thermalized neutrons outside the core
zone.28

These changes give rise to scaling factors used to correct

the values calculated for these regions. These scaling

factors are discussed in greater detail within the context

of the current Pacific Northwest Laboratory research (see

the "Other Sources" section of this chapter). With due

consideration for the foregoing, the data from ORIGEN2 is

used within the data bases to help estimate the waste

classifications of the assorted hardware components.

Details of the hardware information contained within

the CDB can be found within the data bases and, to a lesser

extent, within the hardcopy report which is associated with

it. There are 78 assembly types, 39 SFD hardware types, and

95 NFA hardware types listed within the CDB. In the case of

SFD hardware, the hardcopy report includes general

information on typical hardware elements, but the data base

does not contain more precise information on the physical

properties of the hardware. Where information on the SFD

hardware is available, it is restricted to the hardware's

general type, weight, material of construction, and vertical

location within the core. For example, the 39 hardware

types specified in the LWR Assemblies Data Base do not truly

represent 39 different hardware items. Five of these

records represent grid spacers with the only difference

between these records being their vertical location on the






61

assembly. Overall, the level of detail available on SFD

hardware is very low, even in comparison to the information

available on NFA hardware.

The SFD hardware information is a portion of the LWR

Assemblies Data Base, which consists of 19 dBase data files.

Most of these files, however, contain information on the LWR

fuel assemblies, so only five of them were used for this

work. Details of the five files are as follows:

1) HARDWARE.DBF: Contains 310 records linking SFD
hardware elements to individual assembly
types.

2) MATERIAL.DBF: Contains 354 records listing the
materials) of construction and weight of
each hardware element.

3) MATNAME.DBF: Contains 28 records which convert
the material code given in the MATERIAL.DBF
file into a material name and density.

4) SFDNAME.DBF: Contains 40 records which convert
the SFD code given in the HARDWARE.DBF file
to hardware names and also to the irradiation
zone where this hardware resides.

5) INDUCED.DBF: Contains 26,979 records which
provide the radiation characteristics for all
of the SFD hardware elements.

The INDUCED.DBF file is particularly large due to the number

of cases it contains. In order to facilitate transferring

this file from one computer to another, a smaller version of

this data base containing only the 1474 records needed by

the Hardware Waste Expert System (HWES) was developed.

For NFA hardware elements, the 95 types identified in

the CDB are broken down as follows: 7 BWR fuel channels, 23

burnable poisons, 26 incore sources, 9 incore instrumen-

tation thimbles, 28 control assemblies, and 5 guide






62

tube/orifice plugs. Table 7 shows what NFA hardware

information is included for each of the reactor vendors.

Additionally, hardware that is listed as "included" was

presumably complete when the data base was created in 1987,

but as no updates have been completed since then, additional

hardware types may have been subsequently introduced.

The NFA hardware information is contained in the LWR

NFA Hardware Data Base, in a group of 27 dBase data files.

Whereas all of the information contained in these files

relates to NFA hardware, not all of the information was

relevant to this dissertation, so only 13 of them were used.

Details of the 13 files are as follows:

1) CHANNELS.DBF: Contains 7 records providing
specific information on BWR fuel channels.





Table 7. NFA hardware information included within the
Characteristics Data Base.
Included Not Included

Babcock & Guide Tube Plugs Incore Instrumentation
Wilcox Control Assemblies
Neutron Sources
Burnable Poisons

Combustion Incore Instr. Guide Tube Plugs
Engineering Control Assemblies
Neutron Sources

General BWR Fuel Channels Incore Instrumentation
Electric Control Blades
Burnable Poison Curtains
Neutron Sources

Westinghouse Guide Tube Plugs Incore Instrumentation
Control Assemblies
Neutron Sources
Burnable Poisons








2) CLIMIT.DBF: Contains 8 records listing the
10CFR61 Class C limits for isotope
concentrations in activated metals.

3) COMPOSED.DBF: Contains 444 records which
indicate the materials of which the various
NFA hardware elements are composed.

4) CONTROL.DBF: Contains 25 records which provide
specific information on control assemblies.

5) INSTRMNT.DBF: Contains 9 records which provide
specific information on incore
instrumentation.

6) MATNAMES.DBF: Contains 46 records for the
conversion of the material codes provided in
the 6 hardware data files into specific
material names and densities.

7) PARTNAME.DBF: Contains 96 records which convert
the part codes given in the 6 hardware data
files into specific hardware names and
weights.

8) PLUGS.DBF: Contains 5 records listing specific
information on guide tube plugs.

9) POISONS.DBF: Contains 23 records with specific
information on burnable poisons.

10) PWRPLANT.DBF: Contains 1092 records which link
the various hardware types to the individual
reactors.

11) RCTNAMES.DBF: Contains 126 records listing all
of name of the reactors which are represented
within the CDB.

12) SOURCE.DBF: Contains 26 records listing
specific information on incore sources.

13) RADDATA.DBF: Contains 26,086 records which are
used to provide all the radiation
characteristics for the various hardware
elements.

The RADDATA.DBF file, like the INDUCED.DBF, is extremely

large, so a reduced version of the file was developed which

contains only those cases necessary to the functioning of


the HWES.






64

The CDB serves to illustrate both the large quantity of

necessary data for complete hardware characterization and

the still incomplete nature of that data. Information on GE

hardware is virtually non-existent while information from

the other vendors does not include all hardware categories.

Also, the information on the hardware included in the data

base has several gaps. Many records do not include complete

weights, dimensions, and materials of composition, all of

which are important for waste classification and packaging

purposes. More importantly, however, the most commonly

absent data items for all hardware entries are the hardware

lifetime, the reactors at which the hardware is used, and

the number of components used at each of these reactors. Of

the 95 hardware records, 59 do not have lifetimes listed,

nor can lifetimes be readily assumed for these components.

Information connecting specific hardware components to

specific reactors is particularly difficult to locate,

because the utilities do not typically refer to their

hardware by the same nomenclature used by the vendors or by

the developers of the CDB. In the case of the Westinghouse

information, no data was provided linking specific hardware

types to specific reactors. The creators of the CDB desired

to be conservative with the given data, so no restrictions

were created which did not already exist. Accordingly, any

hardware component which could possibly be employed at a

given reactor is assumed to be used at that reactor. As a

result, each Westinghouse reactor has between 14 and 19






65

hardware records assumed to be associated with it, a number

which is far higher than that for reactors where hardware

was specifically associated. These records are listed as

"assumed" by the CDB to distinguish them from records for

which better information is available, but the numbers are

nonetheless misleading.

In spite of these difficulties, the CDB has proven to

be a very useful data source. The information provided for

Combustion Engineering and Babcock & Wilcox reactors is

substantially complete and can be used to estimate values

for Westinghouse reactors. In terms of hardware volumes,

the major PWR NFA hardware categories, i.e. Burnable Poison

Rod Assemblies and control rods, are included in the data

base. The incore instrumentation and guide tube plugs which

are not described are expected to represent only a small

fraction of the overall hardware waste volume (see "Program

Results," Chapter 4). For BWRs, the only major volume of

hardware not described is the control blades; the CDB does

describe the fuel channels, the NFA hardware waste category

with the largest anticipated volume. Thus, while additional

data is still required, the CDB provides a starting point

for future data gathering efforts. The CDB was used as the

primary data source for this work and the development of the

HWES.






66

Utility Survey

Among all the possible sources of information on NFA

and SFD hardware, the source which is most likely to have

the desired information are the nuclear utilities. All NFA

hardware waste has ultimately been generated by the nuclear

utilities, and most consolidation pilot projects were

carried out in direct cooperation with a nuclear utility.

Accordingly, any existing records which detail how many

control rod assemblies have been discharged from a reactor,

for example, are most likely to be in the possession of the

utilities. Additionally, the utilities should also be able

to specify what lifetime was achieved for the hardware,

where it is being stored, what effect if any, hardware

storage has on spent fuel storage, and a host of other

details. Therefore, a survey of the nuclear utilities was

developed as a portion of this work.

The original survey design had two purposes: 1) to

provide numbers to validate (or correct) the results of the

HWES and 2) to gather additional information about the NFA

hardware elements. At first glance, the first goal appeared

straightforward since the desired information consisted of a

series of numbers which were presumably readily available.

Upon later examination, however, it became apparent that the

desired comparisons were not as easy as originally

anticipated. First, a clear distinction was needed between

how many elements had been discharged and how many were

currently in storage as several utilities have already






67

disposed of various NFA hardware elements at LLW sites.

Second, comparison with the results of the HWES proved to be

difficult at best. The results generated by the Expert

System are based on the information available in the CDB

which frequently turns out to be either too sparse or too

detailed. The problem with the Westinghouse hardware

information discussed in the previous section provides a

good example. Each Westinghouse reactor listed in the CDB

(56 reactors in total) is listed as having between 14 and 19

different NFA components, which are too many entries.

Conversely, since these associations are assumed and not

based on known facts, no values are provided indicating how

many of each type of element are used at a reactor. Here,

information is too sparse. While the general information in

the CDB is sufficient for the HWES, the lack of detailed

information limits the usefulness of the utility survey.

The second goal, to gather additional information,

suffers from another problem in comparing data to CDB

information. Any information gathered must be matched with

an item from the CDB to improve the data base. All hardware

entries within the CDB have specific names as provided by

the original vendors, or where none were provided, by the

CDB developers. However, the test utility responses did not

generally use any names at all and, when names were used,

they did not match the hardware names listed in the CDB. On

the other hand, the CDB contains almost no data on BWR

hardware, so any information that can be uniformly






68

categorized would be an improvement. Additionally,

responses providing actual hardware lifetimes would prove

useful as the lifetimes given in the CDB are theoretical

lifetimes for each hardware type. Information on how the

hardware is currently being stored, how much has been

disposed of, and how it is affecting spent fuel storage

would also be helpful in gauging the extent of the problem

represented by the hardware. Hence, a wide variety of

information was believed to be available from the utilities.

A survey was drafted for distribution to the utilities.

The survey essentially comprised three pages consisting of a

one page table for information about NFA hardware types,

quantities, and lifetimes; one page for general NFA hardware

storage information; and one page for general SFD hardware

information. The survey was completed in late August 1989

and was ready to begin distribution by late September 1989.

Before distribution of the survey could begin, however,

OCRWM indicated that they were also planning to perform a

survey of this nature, and requested to review the author's

survey. The utilities were then consulted by OCRWM and

requested that OCRWM, instead of the author, conduct the

survey; thus, OCRWM requested that the author's survey be

canceled. The author was able to participate in the early

development stages of the OCRWM survey, but in general, the

original intent of the author's survey was lost along with

any control over the survey's execution. The OCRWM survey

was duly distributed to the utilities for completion on a






69

voluntary basis. As of the beginning of September 1990,

approximately 50% of the utilities had responded and a

distillation of that information was made available to the

author.

The original survey had been designed for the general

purpose of improving and expanding on the hardware

information available within the CDB. The OCRWM survey was

designed to gather general NFA hardware information and to

act as a test for another future OCRWM survey, and thus did

not provide the data in a format consistent with the goals

of this work. The problems thus posed to this work were

numerous. Since the data which was made available was only

a distillation of the utilities' responses, the already

limited nature of the data was further emphasized,

particularly with regard to comments made by the utilities.

More importantly, to improve upon the information in the CDB

required that the survey develop direct connections between

reactors and hardware types, and between hardware types and

hardware characteristics. In the OCRWM survey, at the

utilities' request, the data released removed any direct

references to reactor names. As a result, no direct

comparison of the survey data and the CDB data is possible.

The original survey would have established direct

interaction between the author and the utilities which could

then have been exploited to gather additional information,

verify the information already provided, and to clarify any

questions which might arise as a result of the data.






70

Working through the DOE and the complete anonymity required

by the utilities precluded this option. Accordingly, the

DOE survey resulted in 1) highly incomplete data, 2) no

method of verification or clarification, and 3) no practical

improvement in the CDB data.

In spite of these difficulties, some interesting

conclusions can be drawn from the survey data which further

the goals of this work. The responses collected represent

32 Pressurized Water Reactors and 20 Boiling Water Reactors,

for a total of 52 reactors. Of the 32 Pressurized Water

Reactors, 7 indicated that one or more types of NFA hardware

were having detrimental effects on their spent fuel storage

capacity. Of these seven, three went further by stating

that they had plans to dispose of one or more types of NFA

hardware prior to the operation of the federal repository.

Additionally, six PWRs had already sent some NFA hardware to

be buried at LLW sites, while three other reactors had sent

various components to research laboratories for testing. An

unknown quantity of these NFA hardware types has been sent

to LLW sites by these reactors: control rods, Rod Cluster

Control Assemblies (RCCAs), thimble plugs, orifice rods,

Burnable Poison Rod Assemblies (BPRAs), neutron sources,

incore detectors, incore instrumentation, and retainers.

Two reactors indicated that the components which they had

disposed of (BPRAs, thimble plugs, RCCAs, and neutron

sources) had been crushed, sheared, and placed in metal

canisters, one indicated the hardware (BPRAs, orifice rods,






71

retainers, incore instrumentation, and control rods) had

been reduced before shipment, and the other three made no

indication of how the hardware had been packaged. Finally,

three reactors commented that they expected consolidated

fuel and its SFD hardware waste to be removed on the same

schedule as intact fuel. Three reactors expect hardware

waste from fuel reconstitution to be accepted with the fuel,

and two expect BPRAs to be shipped with the fuel.

In contrast to these results, the Boiling Water Reactor

responses show a much stronger bias toward immediate

hardware disposal. Of the 20 reactors, 12 found that

control blades produced adverse effects on spent fuel

storage while 2 reactors said the same about Local Power

Range Monitors (LPRMs), 2 other reactors said the same about

fuel channels, and lastly, 2 other reactors made this

comment about neutron sources. While very few found NFA

hardware to be a generic storage problem, most intend to

dispose of their components on their own instead of waiting

for the federal government to take them. Specifically, 17

intend to dispose of control blades, 18 intend to dispose of

LPRMs, 8 will dispose of fuel channels, 10 will dispose of

neutron sources, and 3 will dispose of poison curtains.

Furthermore, these numbers do not give a clear indication of

the results, because a large number of reactors did not

specify either way. In all of the above cases, the number

with plans to dispose of their hardware are in the clear

majority of those who expressed their intentions. As for






72

hardware which had already been disposed of, eight indicated

that they had done so and the discarded hardware included

the following types: LPRMs, incore instrumentation, control

blades, fuel channels, neutron sources, spring clips, and

poison curtains. Finally, the BWR comments reflected that

one reactor expected consolidated fuel and its SFD hardware

waste to be accepted on the same schedule as intact fuel,

two intend to ship the fuel channels with the fuel, three

reactors store the channels with the fuel, and presumably

intend to ship them this way, and one reactor expects to

have storage difficulties if it is unable to ship its NFA

hardware to LLW disposal.

One of the most important facts learned from this

report is that many utilities have in the past and expect to

continue in the future to dispose of their NFA hardware at

LLW sites. Additionally, the utilities expect (or at least

those involved in consolidation activities expect)

consolidated fuel and its SFD hardware to be accepted like

intact fuel. Some utilities also expect some NFA hardware,

most notably BWR fuel channels, to be shipped integral to

the spent fuel. While these opinions were expressed by only

a small number of the respondents, no dissenting views were

expressed either, so there is no reason to assume that these

viewpoints are not shared by other utilities. All of these

practices could have a significant impact on the FWMS and,

as such, should be studied in greater depth in the near

future, while the repository is still in its early design






73

stages. Only such prompt action will prevent greater

difficulties later on.



Other Sources

Due to the low disposal priority which has been given

to these wastes to date, the number of sources which deal

with NFA and SFD hardware are limited. NFA hardware, for

the most part, has received no attention and is generally

unknown in the literature. Other than the CDB, there are

two exceptions. The first exception involves work conducted

at the Pacific Northwest Laboratories (PNL) in Richland,

Washington. Previous work at PNL included onsite inspection

of nuclear utilities to help determine the types of hardware

generated at each reactor site. Currently, work is underway

which aims to characterize various NFA components, including

a BWR control blade. The project will take samples from

these components in order to determine the exact isotopic

and chemical composition of the hardware materials. The

procedures used here will be similar to those used

previously for SFD hardware (see the later pages of this

section). Results from this work should be available in

late 1991.2

The other reference was produced by E. R. Johnson

Associates, Inc. and attempts to provide a rough

approximation of the total quantities of NFA hardware at

reactor sites. The study also details the impact which

acceptance of NFA hardware will. have on the FWMS. Several






74

key questions are identified which need to be resolved

before a more detailed analysis of these impacts can be

made. This work is the first to attempt to deal with these

points in a comprehensive manner, but by their own

admission, is not accurate enough for final design purposes.

Further, more detailed work will be required after some of

the key issues have been resolved.3

SFD hardware, on the other hand, has received a greater

degree of attention. For one, SFD hardware can be generated

as a result of spent fuel reprocessing and, as such, is

occasionally mentioned, but more commonly is merely

inferred, in conjunction with fuel pin hulls and their

disposal. Such references include no characterization and

little actual data.31"32 The only other source which

frequently mentions SFD hardware is literature dealing with

spent fuel rod consolidation. In these references, SFD

hardware is usually referred to as assembly skeletons and

typically does not include detailed characterization.

However, since the packing, storage, and disposal of SFD

hardware has come to be the pivotal aspect of rod

consolidation, there has been an increase in detail

available in recent reports.

There are currently two other sources of information on

SFD hardware besides the CDB. The first source was produced

by Pacific Northwest Laboratory. Before beginning the NFA

hardware characterization currently underway, PNL performed

similar studies on SFD hardware, the results of which are






75

now available. The lab took a total of 38 samples from 3

fuel assemblies with burnups ranging from 27,500 to 41,800

MWD/MTU. The samples were subjected to radiochemical

analysis to determine the concentrations of the four

controlling isotopes (14C, 59Ni, 63Ni, and 9Nb) for LLW

disposal of activated metals, as well as "Co due to its

importance in determining the heat rate of the hardware.

The samples were also subjected to elemental analysis to

determine the concentrations of the parent materials. The

study determined that, in general, inconel and stainless

steel components were GTCC, while zircaloy components were

close enough to the Class C limit to be questionable. In

particular,

in Zircaloy, elemental analyses reflected niobium
values from below detectable limits up to several
ppm. Curiously, "Nb was detectable in most of
the samples. This is an important result since
most items made of Zircaloy are disposed of as
low-level waste with no consideration of 94Nb
content. Our analyses showed that levels of 9Nb
were a significant fraction of 10 CFR 61 limits
(17%-97%).33

At 17% of the Class C limit the sample is still well below

the 10CFR61 limit. However, at 97% of the Class C limit

uncertainties in isotopic measurements are sufficient to

make disposal as LLW a questionable proposition.

Another result of this project was the development of

new scaling factors for use in ORIGEN2 calculations to

provide more accurate activation values outside of the

active core region. The project used the ORIGEN2 code to

predict the expected activation levels in the hardware






76

samples. The predicted values were compared to the measured

values with the intention of producing new scaling factors.

Comparison with a similar analysis done at Battelle Columbus

Laboratory, however, produced conflicting results.

Therefore, the scaling factors were derived based solely on

the measured sample values.34 The resulting scaling

factors are 10 to 20 times higher than the initial factors

developed in 1978, while ranging from to 5 times the

revised scaling factors of 1987. These figures are

considered approximations based on the available data with

an uncertainty of 50%, so further work will be required to

improve the precision of these results.35

The final noteworthy information source on SFD hardware

is the work performed by Rochester Gas and Electric (RG&E)

in conjunction with their rod consolidation demonstrations.

As an extension of the rod consolidation demonstrations,

they have performed a considerable amount of work on the

classification, treatment, and packaging of SFD hardware.

The first finding of this work is that the desired 10:1

compaction ratio for SFD hardware is not being achieved;

instead only 5 or 6:1 is being reliably achieved.3

Second, the report discusses various alternatives for

storing the hardware until DOE takes title to it.6,37 The

storage methods discussed are similar to those previously

discussed in this work (see the beginning of this chapter).

Of greater interest here, however, are the results of

the classification study. Samples were taken from the SFD






77

hardware generated by the second RG&E consolidation

demonstration at Battelle Columbus Laboratory. These

samples were subjected to radiochemical analysis and the

preliminary results tend to show lower activation levels

than the PNL work. Specifically, early reports indicate

that all inconel components are GTCC waste, as are Stainless

Steel components which have been subjected to a high burnup.

Zircaloy components were not found to represent a problem

for either storage or disposal.37 However, it is important

to note that the consolidated assemblies from which the

hardware samples were taken had only been subjected to a

burnup of 20,000 to 22,000 MWD/MTU.3 As most assemblies

receive at least 30,000 MWD/MTU and the industry is trending

toward even higher burnups, these measurements are not

necessarily representative of the greater majority of SFD

hardware.

A computer code, called FuelCalc, was also developed as

a portion of this project to provide estimates of the

activations levels and waste classification of SFD hardware

items. However, the program will not be available for

analysis until the final report is released. At that time,

a detailed analysis of both the FuelCalc program and the

sample analysis methodology will be possible.



Domestic Hardware Analysis

An analysis of the preceding data sources confirms that

the disposal of NFA and SFD hardware is not a simple






78

proposition. To begin with, the utility survey conducted by

the DOE indicates that 27% of the responding reactors had

disposed of NFA hardware at LLW sites in the past, and 40%

intended to do so in the future. Since the DOE has

undertaken the responsibility for the disposal of these

hardware wastes at some time in the future,39 for the 40%

of the respondents which indicated plans to prematurely

dispose of this hardware, the value of the storage space

occupied by this hardware presumably exceeds the cost of LLW

disposal. The remaining 60% either have no current need to

dispose of the NFA hardware or have not committed to doing

so in spite of the hardware's detrimental effects upon the

reactor's storage capacity. SFD hardware is not currently

considered to be a concern as a full-scale rod consolidation

campaign has yet to be performed.

A fundamental consideration in this issue is the

10CFR61 waste classification of the NFA hardware. If the

waste classifies as GTCC LLW, the only legal disposal option

foreseen at this time is eventual emplacement in the future

Federal HLW Repository. Waste classifications performed in

the past have yielded results which permit NFA hardware

disposal as LLW.4 However, more recent studies such as

those examined here indicate that some, if not all, of these

hardware elements are not suitable for shallow land burial.

A summary of the findings of the sources examined here are

presented in Table 8. The results are presented based on

the material of construction of.the hardware components.






79

Table 8. A summary of the estimated classification of NFA
and SFD hardware components based on the component's
materials of construction.
Low-Level Greater-Than
Waste Class C Questionable

Stainless
Characteristics None Steel None
Data Base Zircaloy
Inconel

Pacific Northwest None Stainless Zircaloy
Laboratories Steel
Inconel

Rochester Gas & Zircaloy Inconel Stainless
Electric Steel


Components constructed of the materials listed in the "Low-

Level Waste" column are considered suitable for shallow land

burial by the source in question, while those in the

"Greater-Than-Class-C" column are not. For items listed in

the "Questionable" column, the reports are not conclusive.

Components constructed of these materials may or may not be

suitable for shallow land burial depending on the actual

initial concentration of 9Nb in the material and the actual

burnup experienced by the component. As can be seen from

the chart, there are both patterns and inconsistencies to

the results. These inconsistencies are resolved in the next

section of this chapter.



Hardware Waste Classification

As illustrated by the previous section, the 10CFR61

waste classifications produced by previous researchers for

NFA and SFD hardware have shown considerable variation. In






80

all of these sources, however, the waste classification is

primarily determined by the concentrations of 3Ni and 94Nb;

the 14C concentrations measured by these sources have not

been critical to the waste classification while the 59Ni

concentration is usually subordinated to the 63Ni

concentration. Since nickel is relatively abundant in

inconels and stainless steels, the two nickel isotopes are

critical to the classification of hardware components

composed of these materials. Zircaloy, however, does not

contain sufficient concentrations of nickel for these

isotopes to be of concern for waste classifications.

Niobium plays a key role in the classification of zircaloy

components, as well as contributing significantly to the

classification of inconels and stainless steels.

Niobium is considered an impurity in most reactor

materials. The American Society for Testing and Materials

(ASTM) specifications for zircaloy and stainless steel do

not include any guidelines on niobium content in these

materials.41,42 In actual practice, PNL found that niobium

concentrations in zircaloy ranged from approximately 40 ppm

to as high as 200 ppm, with an average concentration of 127

ppm. In the stainless steel samples, the niobium levels

ranged from 7 ppm to 350 ppm, with an average concentration

of 109 ppm.43 Due to the low Class C limit for 4Nb (0.2

iCi/cc), even these "trace quantities may be sufficient for

the irradiated material to exceed the Class C limits.""

In stainless steels, the presence of niobium is usually






81

overshadowed by the "Ni concentration. In zircaloys, which

do not contain significant quantities of nickel, the

concentration of niobium is a critical concern. The

concentration of niobium in inconel ranges from 0.7% to 5.5%

and is sufficient to guarantee a GTCC classification to any

irradiated inconel component.4"5

Any attempt to resolve the discrepancies presented by

the previous researchers must revolve around the actual

initial elemental compositions of the hardware and some

approximation of the irradiation history of the component.

Therefore, the first assumption made in this analysis is

that only stainless steel, inconel, and zircaloy make

significant contributions to the Class C limit calculations

for NFA and SFD hardware. There are two possible

exceptions, but neither should invalidate this assumption.

First, BWR control blades contain satellite rollers which

typically become very highly activated and would bias the

overall waste classification. However, previous waste

disposal campaigns conducted on BWR control blades have

shown that these bearings are easily removed from the

blades, so they will be ignored in this discussion. Second,

some neutron sources used in the reactors contain TRU

materials and thus may classify as TRU waste. However, if

the sources do contain sufficient quantities to classify as

TRU waste, then they are not suitable for disposal as either

LLW or GTCC waste and are thus beyond the scope of this

work. Accordingly, this analysis assumes that source






82

material quantities are not significant to the waste

classification.

The second assumption is that the hardware components

are constructed of the materials, and in the proportions,

listed in the CDB. While most SFD hardware elements are

constructed of only a single material, a given NFA hardware

element can be constructed at several, all of which affect

the waste classification. Since only three materials are

assumed to contribute to the Class C limits, by using the

values presented in the CDB, it is possible to estimate the

contributions made by these materials for each hardware

type. Additionally, since it is allowable to average the

concentrations over the total metal volume of the hardware,

those materials which do not contribute to the classifi-

cation can be used to lessen the total concentrations by

averaging. Thus, the relative proportions of the

contributing and non-contributing materials are assumed to

be as specified in the CDB.

The elemental compositions of the materials, specifi-

cally the stainless steels, inconels, and zircaloys of which

the hardware is constructed, are taken from two sources.

The weight percentages of the major constituents are taken

from the appropriate ASTM material specifications and are

based on these particular grades of the materials: Inconel

625, Stainless Steel 304, and Zircaloy-4. These are the

dominant grades of each material used in NFA and SFD

hardware. As many of the specifications present a range of






83

permissible values, the calculations will usually be based

on the average of the upper and lower limits. When a value

other than the average is used for illustration purposes,

the exception will be noted. For those elements which are

considered impurities in the materials, i.e. for those

elements not listed in the ASTM specifications, the average

concentrations found in the samples measured by PNL are

used. The physical characteristics of these materials as

applied to this work are summarized in Table 9.

The next assumption is that, in most cases, only the

concentrations of "Ni and "Nb need be analyzed to determine

the waste classification. The concentration of 14C is

ignored completely by this analysis as this isotope made no



Table 9. A summary of the physical characteristics of the
three primary reactor materials studied in this
analysis.

Inconel (based on Inconel 625)
Elements Concentrations Variance Basis
Nickel 58 w/o (min) -0.45 w/o spec.
Niobium 3.15-4.15 w/o 0.15 w/o spec.
Density = 8.19 g/cc
Number Density = 8.97x1022 nuclei/cc

Stainless Steel (based on SS 304)
Elements Concentrations Variance Basis
Nickel 8.0-10.5 w/o 0.15 w/o spec.
Niobium 109 ppm (avg) measured
Density = 8.02 g/cc
Number Density = 8.52x1022 nuclei/cc

Zircaloy (based on Zircaloy 4)
Elements Concentrations Variance Basis
Nickel 0.03-0.08 w/o 0.01 w/o spec.
Niobium 127 ppm (avg) measured
Density = 6.56 g/cc
Number Density = 4.29x1022 nuclei/cc






84

significant contributions to any of the previous analyses.

The 59Ni concentration is only of concern if the "Ni and 9Nb

concentrations are not conclusive, in which case the 59Ni

concentration will be estimated for use in a sum-of-the-

fractions analysis as described in the "Low-Level Waste"

section of Chapter 1. The three isotopes are primarily

produced by the following reactions:

Ni (n, ) 59Ni
6Ni (n,1) 6Ni
fNb(n, ) 9Nb

For this analysis, these reactions are assumed to be the

only sources of these isotopes. The physical and

radiological properties of the materials used for this

analysis are presented in Table 10. The cross sections

presented in Table 10 are for monoenergetic neutrons with an

energy of 0.025 eV. Since the average neutron energies in a

reactor are actually higher, the actual cross sections in

the reactor will be lower; thus, these cross sections will

produce conservative results.

Next, to approximate the activation levels achieved in

the components, the radiation environment experienced by the

hardware must be modeled. First, the reactor core is

assumed to be divided into four zones: the top zone, the gas

plenum zone, the incore zone, and the bottom zone. The

incore zone corresponds to the active fuel region of the

core, while the gas plenum zone corresponds to the gas

plenum section of the fuel assemblies. The top and bottom

zones match the top and bottom nozzles of the fuel assembly,






85

Table 10. Physical and radiological characteristics of the
critical isotopes and their parent materials. (a, =
absorbtion cross section, r, = half-life).

Parent Element: Nickel
Density = 8.90 g/cc
Number Density = 9.13x1022 nuclei/cc
Isotopes
59Ni
Abundance of precursor (58Ni) = 67.9%
aa = 4.4 barns
T = 80,000 years
6Ni
Abundance of precursor (6Ni) = 3.66%
Ga = 15 barns
r, = 92 years

Parent Element: Niobium
Density = 8.57 g/cc
Number Density = 5.56x1022 nuclei/cc
Isotopes
9Nb
Abundance of precursor (93Nb) = 100%
a, = 1 barns
'T = 20,000 years




respectively. The relative sizes of these zones are

illustrated in Figure 5. The usage of these zones is

consistent with the work performed by ORNL and PNL.

Second, the flux levels experienced by components in

each of these regions must be approximated. For these

calculations, the incore flux is assumed to be 1013

neutrons/cm2-sec. This value was assumed to be a typical

average flux after analyzing several PWR and BWR Final

Safety Analysis Reports. Then, the scaling factors

developed by PNL are used to approximate the flux in the

other regions. The resulting flux levels are also

illustrated in Figure 5. In all cases, the flux is assumed







Flux (nucl.l/cm2-s.o)


Flux (nuclel/cm2-sec)
1.0E+14




1.0E 13


1.OE*12 aGa Plenum Zone (1i') /
Bottom Zone (6")
Top ZTne (6')|



1.0E*11 -
0 20 40 60 80 100 120 140 160 180
Distance From Top of Assembly (inches)

Figure 5. Illustration of the assumed flux shape within the
reactor and the size of the four irradiation zones.


to be constant across the core and within a given region.

Whereas this is admittedly an oversimplification of the real

flux shape, since this analysis is intended to represent

average components, this approximation provides an average

flux for each region. Furthermore, it is also assumed that

the flux is zero outside the core region and when the

reactor is shutdown. The only components which are affected

by these assumptions are the PWR control rod assemblies, but

the effects of this assumption are negligible as will be

explained later in this section.

Some assumptions are also made concerning concentration

averaging. Concentration averaging is a practice which






87

allows the measured isotopic concentrations to be averaged

over the volume of the waste. In the case of NFA hardware,

the concentrations are averaged over the total metal volume

of the components, thus excluding any void spaces. This

practice produces more uniform waste classifications and

lessens the impact of any "hot spots" on the overall

classification. The current disposal sites also allow

averaging between two or more separate components of the

same type, such as between two control rod assemblies.

However, based on the assumptions made previously, this

analysis is already being performed on an average assembly;

therefore, concentration averaging between more than one

such assembly produces no changes in the classifications.

Accordingly, any concentration averaging performed will be

conducted on an individual component basis.

There are a few final assumptions concerning the

reactor operational history that must be mentioned. First,

one reactor fuel cycle is assumed to be 1.5 years in

duration for both PWR's and BWR's. Second, one assembly

lifetime is assumed to consist of three fuel cycles, or

roughly 4.5 years, again for both PWR's and BWR's. These

values represent an idealized PWR operating cycle with one-

third of the core replaced each fuel cycle. A BWR operating

cycle usually has fuel cycles only one year in length after

which one-fourth or one-fifth of the fuel is replaced; thus,

a total core replacement occurs every four to five years.

For the purposes of this study, this is roughly equivalent






88

to the PWR cycle, and will be approximated as such.

Finally, the reactors are assumed to operate at a capacity

factor of 0.688. This capacity factor was the national mean

in 1989, and thus is taken as a representative value. The

capacity factor is used to convert lifetimes given in units

of EFPD to actual years of operation and vice versa.

In order to make proper comparisons with the previous

research, approximations of the initial parent element

(nickel and niobium) concentrations which will cause the

critical isotopes (59Ni/63Ni and 9Nb, respectively) to exceed

their Class C limits would be useful. Given the assumptions

presented above, it is possible to estimate the activation

levels of an isotope in an irradiated material by using the

following formula:

A=4 aN(1-e-t) decays (1)
cm3 -sec

where:

A = activity (decays/cm3-sec)
0 = neutron flux (neutrons/cm2-sec)
a = absorption cross section (cm2)
N = number density (nuclei/cm3)
A = isotopic decay constant (years"1)
t = time of irradiation (years)

This formula assumes a constant flux and no decay time after

irradiation. Constant flux is one of the assumptions made

earlier, and due to the long half-life of 4Nb (20,000

years), the actual decay time of 5 to 15 years is

negligible. The decay time has a greater significance to

3Ni since its half-life is only 92 years, but since the

values being calculated here are only approximations, the






89

actual decay will still be of little consequence to this

analysis. Also of note is that t represents the duration of

the exposure to the given flux, 0. Accordingly, t is

measured in Effective Full Power Years (EFPY), not actual

years of operation.

If the activity is known, then equation (1) can be

solved in terms of the number density, N, resulting in

equation (2). For this analysis, the 10CFR61 Class C limits


N=A (2)
a ,(1i-e-1t)

are taken to be the maximum permissible activation levels,

so by using the radiological characteristics presented above

(see Table 10), the maximum permissible initial

concentrations of the parent elements can be calculated.

The resulting values are in terms of nuclei/cm3, but by

comparing the results with the characteristics of the parent

materials (zircaloy, stainless steel, and inconel), the

concentrations can be converted to units of parts per

million (ppm) or weight percent (w/o), as appropriate.

Figure 6 shows the initial niobium concentrations in

zircaloy, stainless steel, and inconel which will cause the

material to exceed its Class C limit, as a function of

irradiation time. Due to its higher atomic number, zircaloy

permits the highest concentration of the three materials,

while stainless steel and inconel permit almost identical,

somewhat lower, concentrations. A comparison of these

results with the initial concentrations presented in Table 9