EXECUTIVE OFFICE OF THE PRESIDENT
OFFICE OF SCIENCE AND TECHNOLOGY POLICY
WASHINGTON. DC 20600
DRAFT
Isolation of Radioactive Hastes in Geologic Repositories:
Status of Scientific and Technological Knowledge
A Working Paper for the Interagency Review Group on
Nuclear Waste Management
prepared by the
Subgroup for Alternative Technology Strategies
July 3, 1978
DRAFT
-------
DRAFT
Table of Contents
Page
1. Introduction 3
2. Perspectives 11
3. The Operational Period 16
3A. Flooding 16
3B. Gaseous Effluents 16
3C. Water 17
3D. Retr ievability 18
3E. Rock Mechanics Considerations 19
4. Long-Term Isolation 22
4A. Waste Forms and Waste Rock Interactions 22
4B. Properties of the Host Rock 27
4C. Hock Mechanics 29
4D. Hydrogeologic Transport of Radionuclides 32
5. Long-Term Risk Assessment 38
6. Site Selection and Characterization 47
7. Candidate Media 53
7A. Rock Salt 53
7B. Anhydrite 62
7C. Granite and Other Crystalline Rocks - 64
7D. Shale and Related Rock Types 66
7E. Flood Basalt 69
7F. Tuffs 70
7G. Unsaturated Rocks 73
8. References 75
DMFT
-------
DRAFT
1. INTRODUCTION •
The purpose of this paper is to describe the issues pertaining to the
scientific and technological knowledge relevant to the construction, operation,
and ultimate sealing and long-term safety of a mined geologic repository
for radioactive wastes. (Other disposal techniques will be examined in other
papers of the Subgroup for Alternative Technology Strategies of the Interagency
/
Nuclear Waste Management Task Force.) This paper identifies areas where we have
confidence in our current state of knowledge, and also areas where uncertainties
and lack of knowledge still exist. Included to the extent possible will be an
evaluation of the significance of such knowledge gaps and identification of
of areas where work is currently underway or needed to resolve uncertainties.
Radioactive wastes considered in this paper are generated as a result
of processes-within a fission reactor, whether it be a civilian nuclear power
plant, a research or test reactor, or a plutonium production reactor. Two
major classes* of radioactive nuclides comprise such waste:
(1) Fission products: These isotopes are the fragments produced when a
heavy nucleus is split. The bulk of the radioactivity is associated
with radionuclides that change to stable elements over
a period of several hundred years. The fission products produce
intense radiation and are a major source of the heat generated during
the first few hundred years of the disposal period. A few fission
products have half lives of millions of years, but these constitute
a minute fraction of the initial amount of radioactivity.
(2) Actinide elements: These consist of uranium which has not undergone
fission, transuranic elements formed by neutron capture, and the
* In addition, carbon-14 and other radioactive isotopes created by neutron
activation of materials may be present in the wastes.
DRAFT
-------
DMT
4
decay products of these two types of elements. Actinides generally
have considerably longer half-lives .than fission products and an
equivalent mass generates considerably less heat per unit time.
However, integrated over time the actinides can contribute signifi-
cantly to total decay heat (about half of the total heat in high-
level wastes and more than 90 percent of the total heat in
spent fuel).
These two classes of radioactive nuclides occur in several distinct
combinations that are candidates for disposal by emplacement in mined geologic
repositories. One combination is found in spent reactor fuel, comprised mostly
of uranium oxide with zircalloy cladding but also, with a burnup of about
30,000 Mw(th) days/metric ton, containing waste in. the form
of about 1 percent transuranic elements and about 3 percent fission products (1).
Under current U.S. policy to defer indefinitely commercial reprocessing for recovery
of uranium and plutonium, the possibility exists that large quantities of spent
fuel will at some time be disposed of as waste (2). Assuming a nuclear power economy
of 380 GWe in the year 2000, the spent fuel discharged through that time will con-
tain approximately 98,000 metric tons of uranium and other actinides (3).
A second combination of radionuclides is found in the high level waste
streams from the chemical reprocessing of spent fuel. Reprocessing has
occurred in government facilities, and a large quantity (70,000,000 gallons)
of reprocessing waste, the radioactive part of which is composed of roughly
98 percent fission products and 2 percent actinides, has been accumulated (3).
A relatively small amount of high level commercial waste (612,000 gallons)
from the now closed commercial reprocessing plant at West Valley, N.Y. also
awaits treatment and disposal (4). If commercial reprocessing is initiated in
the United States, the major waste product requiring disposal will be high
DM
•i r
-------
DRAFT
level wastes from the nuclear power fuel cycle, rather than spent 5
fuel. The composition, by weight, of high level waste from commercial
power plant fuel (i.e., from once-through PWR fuel having an exposure of
33,000 Mw(th) days/metric ton) would be about 84 percent fission
products and 16 percent actinides (5).
The third combination of waste nuclides of interest to this report
are contained in scrap and trash, which in the course of a wide variety of
nuclear activities, have become contaminated with plutonium and other trans-
uranic nuclides. Large quantities of this so-called TRU waste have been
buried in trenches at the Hanford, Savannah River, and Idaho reservations
and whether this should be exhumed and removed to a deep geologic
repository is under review. TRU wastes with a concentration of
transuranic nuclides above lOnCi/gm have been packaged and stored in
a dry environment above ground since the early 1970s. Fifty thousand
cubic meters of this material have now accumulated in retrievable storage
at the Idaho Falls site and elsewhere (3). Because the long life of the TRU
components requires this material to be isolated from the biosphere, these wastes,
though of low heat generation rate, must be isolated in much the same way as
high level waste.
Each of the three major categories of waste—spent fuel, reprocessing waste
(from Government or from commercial operations) and TRU—have different radio-
nuclide compositions and, therefore, different chemical and radiologic properties.
These variations in properties and their possible interactions with host
rock are discussed in Section 4A.
A comparison of the thermal power and radioactivity of spent fuel (SF)
and high level waste (HLW) that typically would result from the light
water reactor fuel cycle is shown in the following table (11):
DRAFT
-------
DRAFT
Time Since Discharge Thermal Power Radioactivity
of Spent Fuel, Years watts/MTHM (a) Ci/MTHM (a)
SF HIM SF HLW
10
100
1000
10,000
100,000
1,000,000
1200
290
55
14
1.1
0.39
1000
110
3.3
0.47
0.11
0.15
410,000
42,000
1,800
480
58
21
320,000
35,000
130
42 '
21
10
a. MTHM is metric tonsaof heavy metal orginally charged to the reactor.
The composition of the high-level waste assumes recycle of uranium
(not plutonium) to light water reactors (LWRs) and removal during
reprocessing of 100 percent of the tritium and noble gases (He, Kr, and Xe), 99.9
percent of the iodine and bromine, and 99.5 percent of the uranium
and plutonium. The fuel is assumed to have been reprocessed one
year after removal from the reactor. The spent fuel has appreciably
higher thermal power and radioactivity than the high-level waste,
primarily because of the presence of the plutonium and its subsequent
daughter products.
The as-generated TRU waste averages about 9 grams of plutonium per cubic
meter. The volume can be reduced by a factor of about 50 by incineration (and,
therefore, the plutonium concentration increased by the same factor).
The heat generation rate and radioactivity of the contained plutonium as a
function of time since purification (separation from daughter products)
are as follows (11):
DRAFT
-------
DRAFi
Time Since Purification,
Years
1
10
100
1000
10,000
100,000
1,000,000
Thermal Power
watts/MT of Pu
200
230
220
52
14
1
0.2
7
Radioactivity
Ci/MT of Pu
120,000
84,000
7,700
1,600
440
35
9
The above data show that the thermal power and radioactivity
of the THJ wastes change much slower with time than does spent fuel
or high-level wastes. This, of course, is the result of the comparatively
longer half life of transuranic radionuclides.
Nuclear waste management involves the effort to identify and develop
the means to isolate these wastes from the biosphere. The different options
that have been seriously considered for the isolation of radioactive wastes
can be divided into two general categories:
(1) Elimination of portions of the waste from existence on earth,
such as by transmutation and by ejection to space, and
(2) Disposal of the wastes in various geologic media, either in the
continental rock mass or in the seabed.
The approach that has received most attention to date, and which is the
subject of this paper, is disposal in a geologic cavity mined by conventional
technology. Such a mined cavity is commonly referred to as a geologic
repository.
DRAF
-------
DRAFT
8
The safety of a mined geologic repository is generally analyzed
in terms of three time periods:
(1) Operational period: While the repository is open and during
which any emplaced waste can be retrieved by conventional mine
machinery.
(2) Thermal period: For high level waste, the first few hundred years
or for spent fuel the first few thousand years after closure of '
the mine during which time radioactive heat production will be
significant.
(3) Actinide decay period: A period that exceeds several hundred
thousand years, defined by the ultimate decay of the actinide
isotopes.
Relevant issues for the operational period before the repository is
permanently sealed include the possibility of repository flooding, the venting
of radioactive gases and/or flammable gases emanating from the waste or from
waste/container/host rock interactions, the capture of water vapor from air within
the repository, the question of retrievability, and the effects of thermal
stress on the integrity of the repository.
The thermal period is of special interest because it is during this
interval that the thermal effects will be at their maxiumum. As the
near-field temperature of the repository rises and then falls
during this period, the repository contents and surroundings will
be subjected to thermal, mechanical, chemical and, possibly,
thermoelectric effects. Most physical or chemical changes within
the waste form that would affect the relationship between the
waste and its container, or between the host rock and its container,
DRAF
-------
DRAFT
will proceed at the highest rate during this period. The initial
thermal expansion will place the immediate host rock and its surroundings under
stress and could conceivably provide a means for breaching the containment by
fracturing the rock, and then permitting subsequent intrusion of water.
The third and final time period extends from the end of the thermal
period until the risk of the repository has disappeared. While
the stresses generated during the preceding period will have ended,
the long half-lives of the plutonium and other actinides results in a con-
tinuing but gradually diminishing level of risk. It is over this
much longer time period that our predictive capabilities are the least accurate.
The length of the period during which the waste should be considered
hazardous is an issue for which there is a broad range of views. It has been
discussed by several workers (6-10), who variously conclude that it may be from
several hundred to several million years. A comparison of the relative ingestion
hazard of high-level waste with that of uranium ore (6) led some
to suggest that several hundred years of isolation would be sufficient.
At the other extreme of the time scale, it has been suggested
that neptunium-237 and its daughter products might require isolation
from the biosphere for several million years (7).
Isolation for roughly the first 1000 years, in which time most of the
fission products will have decayed to stable isotopes, is a minimum
requirement. Beyond that, one might conclude that the wastes should be
isolated from the biosphere for as long a period as possible,
hopefully for periods longer than several thousand years and, preferably,
for several tens of thousands of years. Complete confinement need
DRAFT
-------
10
not be assured for this long a period if we understand the release and
transport processes adequately to feel confident that total releases will be
small and that transport to the biosphere will be so small as not to
result in an unacceptable risk to humanity.
In the following sections this report considers first a set of pers-
pectives on waste disposal and then discusses in turn the issues important
during the operational period of a repository, factors that affect the
isolation of the waste in a repository, the methodology for and
our ability to assess the long term risks associated with possible
releases from the repository, and questions relating to site selection
and characterization. The report closes with a discussion of the properties
of the various candidate host rocks.
DRAFT
-------
DRAFT
11
2. PERSPECTIVES
The task of accurately predicting the fate of radionuclides emplaced
in a repository over time frames of even several hundreds to many
thousands of years is unprecedented. The principal earth science
disciplines involved in such a task — namely hydrogeology,
geochemistry and rock mechanics — are relatively young sciences with
little experience in making or evaluating predictions that cover time frames
even as short as a few decades. Numerous limitations of our knowledge in these
fields, pertinent to radioactive waste disposal, are identified throughout
this paper. Nevertheless, we believe that' such gaps in our current
knowledge need not rule out successful underground containment
of radionuclides for periods of many thousands of years. The reasons for
this belief follow:
(a) application of the multiple barrier concept to repository
selection and design can compensate for the lack of total
understanding of repository behavior;
(b) wastes can be allowed to cool and/or be fabricated to take
account of the physical properties of the host
media prior to their emplacement in the repository;
(c) the knowledge gaps, enumerated in various sections of this
report, do not necessarily all apply to a specific repository
nor are they of equal importance;
(d) the pertinence of some of the perceived knowledge gaps probably
will be resolved during construction and operation of the
repository and conservative engineering practices can sometimes
be utilized to compensate for remaining uncertainties; and
DRAFT
-------
DRAFT
(e) active R&D programs focused at filling knowledge gaps, 12
currently are underway or planned and our understanding
of the relevant hydrogeology, geochemistry,
rock mechanics, and risk assessment should
increase rapidly over coming years.
These reasons are discussed below in the order listed.
The utility of a design and site selection philosophy
that relies upon a series of backup systems to provide redundant
protection in the event of the failure of a key component — the so-called
multiple barrier concept — is a widely accepted approach in engineering
practice, particularly in areas with limited empirical information. For
example, a geologic environment combining long groundwater flow paths,
slow groundwater velocity, rocks with high sorptive capacity along the
flow path, and tectonic stability is an example of the multiple natural
barriers that are able to contribute to radionuclide retention.
Multiple barriers may also be "engineered." Examples include
the inertness of the waste form, corrosion resistance
of the canister, and the character of the material packed around
the canisters. Although efficacy of these man-made multiple barriers will
undoubtedly be shorter than those provided by nature; they could still be
useful in some geologic environments, particularly in the first, and,
probably, the second time period.
To date emphasis has been placed predominantly on
the importance and suitability of a particular rock type
to contain radioactive wastes and relatively less on the
system comprising the entire hydrogeologic and geochemical environment.
DRAFT
-------
DRAFT
A change in emphasis appears necessary, both for implementation 13
of the multiple barrier concept within the repository and for optimally
conducting a search for appropriate repository sites.
Heat emitted by the wastes during the first several hundred to few thousand
years is a major cause of containment uncertainty during this time period.
Adjusting the waste content of high level waste canisters, altering
the repository loading factors, and cooling the wastes for several decades
prior to emplacement in a repository can reduce the uncertainty
of the response of the waste, the geologic media, and
groundwater to this heat. The Swedish proposal for placement
of reprocessed wastes in granite, for example, calls for lower waste
concentration in high-level waste canisters than has been considered in the
United States and the cooling of high-level, reprocessed wastes
for 40 years prior to emplacement. Cooling, spacing,
and dilution of the wastes would also permit consideration
of geologic media (for example, zeolitized tuff or alluvium)
that have low heat conductivity but excellent sorptive
properties for radionuclides. The advantages of some surface
cooling of the wastes are clear and current plans involve some
cooling period before emplacement in a'repository. Just how long the wastes
should be cooled for optimal operation of the total waste management
system is a question that requires further study.
In addition to cooling, fabrication of the wastes to make them
compatible with the physical-chemical properties of an intended host medium
is still another way to reduce some of the uncertainties (outlined
in other sections of this report) related to the integrity of the
DRAF
-------
DRA-i
waste form. •»•!••• 14
The knowledge gaps enumerated throughout this paper do not
necessarily apply at every potential repository site, nor are those gaps
applicable at a given site of equal importance. For example, the potential
for radionuclide transport upward via the shafts, fractures induced by the
shaft construction, or exploratory drill holes may not constitute
a major concern in a terrain in which natural groundwater movement is
predominantly downward. Similarly, uncertainties about waste form
solubility are of less importance in an hydrogeologic and geochemical
environment that provides long groundwater flow paths, slow water
velocity, and strata of high sorptive capacity.
Some of the matters perceived to be potential problems may be
resolved by knowledge gained empirically during construction
and initial testing of a repository. Science traditionally advances
by a steady interplay of theory, experimentation, methodology
development, and demonstrations (empiricism). Although theory clearly
dictates what data and problems are pertinent, theories are themselves
adjusted to fit anomalous facts, and facts themselves are, sometimes,
fleeting. It may be necessary to run a variety of in-situ
tests or, perhaps, even to construct and operate a repository
for a decade or longer to resolve the question of the
importance of currently perceived gaps in our knowledge,
as well as to test new methodologies being developed for
site characterization. Problems remaining after such a testing period
will, hopefully, lend themselves to resolution by conservative
engineering design.
DRAFT
-------
DRAM
Although knowledge requirements are addressed throughout this paper, 15
a brief general comment is appropriate here. Some current in-situ tests of
rock properties are not conducted at actual or potential repository sites.
Such studies provide useful generic data on rock properties and are
important for instrument development and data interpretation on method-
ology. However, the variability of geologic data and the systems philosophy
dictate that generic characterization cannot substitute for detailed
studies at actual repository sites. Such site-specific work, including
in-situ heater tests (initially conducted with electrical heaters),
will necessarily take considerable time (at least a year)
and money to construct, and many months to a few years
to operate and interpret. Accordingly, in the future, greater program
emphasis should be placed on conducting such tests in areas
that can be identified as candidate sites.
DRAFT
-------
DRAFT
3. THE OPERATIONAL PERIOD 16
During the period when the repository is open several considerations
are important; these include the safety and cost of operation, the
monitoring of repository phenomena, and the prevention of occurrences that
could jeopardize long term isolation.
3A. Flooding
A sudden flooding of the repository workings could cause serious immed-
iate problems. The potential for flooding depends on the local hydrology
and design features of the facility, and can be minimized by careful site
selection, by conservative engineering and by monitoring during the drilling
and tunneling operations. Sufficient hydrologic and mining knowledge
i
exists to reduce this problem to a very low level of risk.
3B. Gaseous Effluents
Any toxic or explosive gas (e.g., hydrogen, methane, and hydrogen sulfide
generated by chemical reactions within the repository could be
controlled by the mine ventilation system and should not become a safety hazard (12)
The release of radioactive gases (e.g. krypton-85, tritium, and iodine-129, which
are present as a relatively low fraction of the total radioactivity of spent
fuel and an even lower fraction of the radioactivity of high-level and TRU wastes)
to ventilation corriders probably can be controlled during operation of a
repository by engineered barriers to flow and diffusion.
The extent to which this might be necessary would depend on
the amount of radioactive gas that is likely to be released from
the repository and on the level of emissions permitted by environmental
regulations (13,14). The former depends on the waste form and
integrity of the waste container and on possible interactions among the
DRAF
-------
DRAFT
waste, its container, and the host rock. Relatively little attention 17
has been given to repository conditions. Corrosive interaction
between the spent fuel and the host rock could be very important.
High level waste would release relatively little radioactive
gas (essentially only that from spontaneous fission of the actinides) because
the original gaseous fission products would have been removed
and captured during reprocessing. The gaseous wastes that are captured
and concentrated during reprocessing could be encapsulated, normally converted to
stable solid forms, and, also, be isolated in a geologic repository.
3C. Water
For some host rocks, particularly salt, minimizing the amount of free
water within the repository could be very important during the first period
and could even provide the limiting factor for long term confinement.
Depending on the amount of naturally contained water in the formation
and the limits of free water required to assure safety, it might
be necessary to dehumidify air circulating through the repository
for ventilation. Dehumidification can readily be done by standard
industrial techniques.
DRAFT
-------
DRAFT
3D. Retrievability
Maintaining retrievability involves tradeoffs for a repository designed
for disposal rather than for storage of radioactive wastes. On the one hand,
there might be safety advantages in sealing the storage chambers as soon as
they are full. On the other hand, to hedge against the possibility of design,
engineering or construction errors or other unanticipated difficulties, there
might be reason to maintain fairly siinple retrievability for a period of time.
In salt and possibly in shale, the repository openings will have a natural
tendency to close because of the plastic flow that results from the increase in
temperature and from the weight of the overburden. Maintaining retrievability
beyond a few years or decades will involve some additional cost. It is not yet
clear how to optimize the tradeoff between the early backfilling and sealing
of a storage room and letting the rooms remain "open" for ease of retriev-
ability. An analysis of this issue is required.
A repository could also be used to store spent fuel in a
retrievable mode in order to maintain the option of reprocessing at a later
time. If maintaining retrievability for a few decades would lead to addi-
tional problems with respect to flooding, water absorption, radioactive gas
effluents, or other factors affecting safe operation of the repository, the
cost and feasibility of handling these problems will need to be compared with
•
the costs and risks of other options including wet and dry surface storage.
The cost and risks of transportation must, of course, be included
in such analyses. Although the 1978 DOE Task Force (3) recommended against
operating a repository for economic retrievability, this option requires more
analysis to provide perspectives on cost and safety.
DRAFT
-------
DRAFT
19
3E. Rock Mechanics Considerations
From a structural viewpoint, a geologic repository for radioactive waste
is not conceptually different from previously mined cavities. Underground
facilities have been built and operated for many centuries. Some of the
mined cavities have remained open to this day, despite the fact that
they were not constructed for long-term stability, while others have collapsed
or filled with water. An important observation with respect to these cavities,
however, is that, except for the opening, no significant change has been made on
the geologic formations in which the cavities exist. Within a waste repository
a thermal pulse will be generated by the radioactive decay of the
emplaced wastes. Because of the potential for this thermal pulse to alter
the structure of the surrounding rocks, knowledge of the response of the host
rock and its surroundings to the thermal loading is of major importance.
The potential far-field and long-term effects and their implications
will be discussed in Section 4C on rock mechanics under long-term
isolation. Here, only short term effects are considered.
Rock mechanics as applied to repository design is well advanced for repos-
itories in salt (15-17). The major mechanisms for mechanical deformation
of salt have been determined and the laws governing these effects
have been identified. The major design factor for which rock mechanics
must provide analysis is plastic deformation that affects the repository
openings through creep mechanisms. The analytic models needed
to predict the near-field response of the rock mass have been
developed with data obtained from laboratory experiments and from
in-situ heater testing conducted in Project Salt Vault (18) in Kansas.
The validity of applying these models, developed for bedded salt conditions,
DRAFT
-------
DMfT
to done salt will be determined through in-situ heater test experiments 20
being conducted at Avery island (15).
Uncertainty remains, however, about the specific data required for a
rock mechanics analysis. For example, experiments with single crystals of salt
at high temperature and pressure could be utilized to establish a rock mechanics
model. But, single crystals of salt constitute a material of high purity
with specific slip systems available for mechanical deformation and a unique
temperature-deformation relationship. By comparison, rock salt from a geologic
formation is an impure polycrystalline material. Its strength and deform-
ational characteristics are influenced by the presence of impurities, solid
solution, and grain boundaries. The strength is greater than that
of single crystals and the time-dependent behavior -is different.
Also, rock salt is more resistant to the effects of an increase in
temperature than single crystals of salt. Nonetheless, by employing
appropriate engineering conservatism, current knowledge with respect to
rock mechanics is adequate to successfully design repositories in salt.
Numerical analysis of rock mechanics questions in non-salt formations
is also being actively pursued (15, 19-24). The most important problem
that must be solved for hard rock formations is that of identifying
and experimentally confirming the potential failure modes. In-situ
experiments underway in granite in Sweden and at the Nevada Test Site
and those planned for basalt at the Hanford site in 1979, are intended to estab-
lish the mechanisms of deformation in these rocks. The major design consideration
in hard rocks is brittle fracture. The approximation of continuum
mechanics that has been applied effectively to salt may not provide
a good aproximation for a repository developed in a jointed and
fractured rock mass.
DIM/7
-------
DRAFT
21
It is of major importance to distinguish between individual rock
specimen and rock mass properties. The appropriate means of
correlating these two types of samples is still undergoing technical
and engineering debate. Most data currently available are based
upon laboratory experiments and measurement on individual rock specimens.
These data are not completely adequate for engineering design because
they represent the individual specimens and not the rock mass; the
latter commonly contain fractures and heterogeneities that could
significantly alter the response of a structure built in such a
geologic environment.
The program of the Department of Energy in rock mechanics emphasizes those
items related to site-specific design and development of instrumentation
that will be reliable over the time required to conduct in-situ
tests. Experiments will be developed to measure in-situ properties of the
rock mass for non-salt rocks including mechanical properties and
the time and temperature .dependent relationships. The primary
objective of the program is to provide data that can be used to
validate the analytical models which describe the response of the
"rock mass" in the near field. As recommended by the NAS Panel on Rock
Mechanics(25), experiments capable of inducing large scale failures
are being planned and will be undertaken in 1979. The specific
goal of these experiments will be to provide the conceptual basis
for defining permissible stresses, pillar stability, deformation
and failure modes in the near field. In addition, two major experiments
with radioactive waste are planned for Hanford (26) and the Nevada Test
Site (27) to evaluate both thermal and radiologic effects.
DRAFT
-------
DMF.
22
4. LONG TERM ISOLATION.
As discussed above, the ability of a repository to assure long term
isolation of nuclear waste depends on the effectiveness of the various
natural and engineered barriers. These can be aggregated into three
categories: (1) barriers against disintegration of the waste form and
container, (2) barriers provided by the host rock and the surrounding
rock mass, and (3) barriers provided by the length and character of the
hydrologic flow path from the repository to the biosphere. These
will be discussed in this section in turn.
4A. Waste Form and Waste-Rock Interactions.
Chemical and radiological interactions between the host rock
witii its contained water and both the waste form and the container
could lead to disintegration of the packaging and partial dissolution
of the waste. The extent to which this will happen depends on
the nature of the container and the waste form, the chemical properties
of the host rock, the ground water, and the duration of exposure.
Research work has been underway for more than two decades to evaluate
container materials and to develop processes for producing stable
solid forms for high-level and transuranic wastes (28). The development
and assessment of container systems for spent fuel has begun
only within the last year. Substantial further progress on
containers and waste forms awaits performance criteria and the
results of ongoing experimentation.
DRAFT
-------
DRAFT
The current reference form for high level waste assumes that the 23
waste is homogeneously distributed in a borosilicate glass which is
then placed in a container of mild steel (for disposal in salt) or
stainless steel (for disposal in most other formations). One problem with
borosilicate glass is that some highly soluble phases are produced as
the glass devitrifies. There is also recent evidence (29) that contact
with pressurized water at the 200-400 degrees C temperatures that could
occur near the waste in the early years of storage rapidly alters the
•
borosilicate glass and promotes interactions between waste and both
shale and basalt rocks. Examples of interactions are the formation of
a uranyl silicate directly from hydrothermal alteration of the glass and
the reaction of cesium and sodium from the waste with aluminosilicates
in basalt and shale to form pollucite, a stable cesium sodium
aluminosilicate mineral.
Investigation is underway on potentially more stable waste
forms, such as synthetic inorganic polycrystalline materials (e.g.,
ceramics, oxides and silicate minerals) that might prove to be
«
more resistant to radiation and to chemical attack in a given
geologic environment (30). Several years of research and development
will be required to advance oxide and silicate crystalline forms
as a competitor to borosilicate glass. Even more complex waste
forms such as metallic matrices, cermets, and microspheres of
waste coated with impermeable pyrolytic carbon (31-33) are
under study. More corrosion resistant container materials, such as Hastelloy C
and alumina, are also being considered for waste emplacement in salt (34).
The efficacy of coating of a metallic container with flame sprayed oxides
(e.g. alumina) or silicate glazes has yet to be evaluated. A total of
DRAFT
-------
DRAFT
24
three to five years probably will be required to develop an adequate
understanding of chemical reactions between candidate waste forms, con-
tainer materials and the chosen geologic formation.
The current reference concept for storage of spent fuel assumes
that one or more fuel assemblies will be placed within a container of
mild or stainless steel that is filled with helium and sealed by
welding. Substantial (at least several years) work remains to be done
to experimentally determine allowable temperatures to avoid excessive
failure of the fuel rods, to avoid criticality problems if fuel material
is loose in the container, and to measure interactions of the container,
the zircalloy cladding, and the fuel pellets with potential
host rocks. More needs to be known about the chemical forms
of the fission products and actinides in the spent fuel
pellets and cladding or about the resistance of these forms
to leaching or interaction with repository environments. Research
work on these problems is underway by Lawrence Livermore Laboratory,
Rockwell Hanford Co., and the Battelle Pacific Northwest Laboratory.
The age and concentration of high-level wastes and the age and
exposure (burnup) of spent fuel determines their rate of heat generation
and their surface temperature. These parameters, in turn, affect the severity
of the thermal pulse to the host rock and the nature of the geochemical
interactions between the waste and the host rock. The duration of the
pre-disposal cooling time is, therefore, an adjustable parameter that can be
lengthened in order to reduce the rate and severity of some impacts
of the waste on its surroundings. The spacing between canisters and the
concentration of wastes within each container can also be used to control
the magnitude of the thermal pulse.
DRAFT
-------
DRAFT
The effects of gases that are generated by the wastes over the long 25
term must also be considered. The primary means for minimizing release of
radioactive gases (e.g. krypton) to the atmosphere would be to provide
engineered barriers (such as the plugged shafts and boreholes) to flow
and diffusion. Gases resulting from corrosion and waste-rock interactions
(e.g. hydrogen, hydrogen chloride, and chlorine) will, most likely, be
accommodated by diffusion through grain boundaries, flow through fractures,
or dissolution in ground water.
Currently, transuranium wastes (containing contaminated trash
such as paper and neoprene gloves) are normally enclosed in
mechanically sealed steel containers. A problem with this type of waste
is that it generates combustible gases (e.g. hydrogen and methane) by
radiolysis and pyrolysis during storage. These gases escape gradually
from the unwelded containers relieving pressure during storage, and
could be troublesome within a repository. The organic materials in these
wastes might also react with the actinides to form compounds that are less
likely to be sorbed by the surrounding rocks. For these reasons it may
become necessary to incinerate these wastes and to fix the ashes in an
inert solid material such as glass, concrete, or a synthetic
mineral. The choice of waste container and waste form could
be made and the related R&D completed within several years after the
incineration question (currently under review by the D.O.E.) is resolved.
The current status of knowledge and research and development
programs on wasted-rock interactions was the subject of a recent
conference (35). The chemical interactions between waste and host rock,
DRAFT
-------
DRAFT
26
the effects of temperature and the presence of a fluid phase, are
best understood for salt (18). Even so, additional work
is needed on the complex chemical properties of brines and
bitterns containing many ionic species. This is underway
at the USGS and the Oak Ridge National Laboratory (15).
Because of salt's highly corrosive nature, currently planned waste
containers would seem to be breached and substantially corroded
by all but the very driest salt within months to years.
Current work may be able to extend the period of integrity
to decades but waste packaging in the case of salt is now
generally assumed to provide only for safety during transportation
and for ease of retrievability over a relatively short time span.
Less is known about the chemical properties of other candidate
host rocks at the relevant temperatures and pressures and little is
known about the interaction of waste with these rocks. However, there
are indications to suggest that at least in some other media (granite and
basalt are examples) the container and associated overpack material
might be expected to maintain their integrity for much longer times.
Investigations are now underway on basalt at Rockwell-Hanford and
Pennsylvania State University, on granite at Lawrence Berkeley and Lawrence
Livermore Laboratories, on shale at Sandia Laboratory, Georgia Tech and
Pennsylvania State University and on tuff and alluvium at Sandia (15,35).
The interactions with waste of several of these media are also being
conducted in other countries (35).
DRAFT
-------
DRAFT
4B. Properties of the Host Rock. Uflfll • 27
The host rock provides the first natural barrier to waste migration
and strongly influences the detailed design of the engineered repository
within it. The properties of the host rock, therefore, are among the
important determinants of the total geologic environment of the repository.
Several media have been studied and are under study to determine in a
generic sense whether their properties are suitable for a repository
host rock. These candidate media are discussed in Section 7 of this paper.
At this point it is necessary only to identify some general properties that
would seem desirable for a host rock.
Media that exhibit creep or plastic flow might be capable of sealing
repository workings by flow without fracture propagation, or might self-heal
in the event of fault-induced or subsidence-induced fracturing. Creep
is observed in rock salt and to a lesser extent in some shales which,
for example, can flow without fracturing at depths that are reasonable
for repository construction using conventional mining techniques.
Another desirable characteristic of a host medium is that there
should be minimal chemical reactions between the waste and the host rock
to form compounds with high solubility or low melting points. Chemical
reactions between rock and waste can change the source term for chemical
transport because they change the leaching rate, alter the compositional
and pressure gradients, increase the quantity of fluid available, or
change the ionic strength of these fluids. An increase in the ionic
strength decreases the sorptive capacity of the host rock for radionuclides
DRAFT
-------
DRAFT
28
and could increase the corrosion rate of containers. Chemical
reactions that produce liquids or gases throughout the rock mass may
also induce changes of acidity, density, or viscosity of the liquids or
gases already present and thus reduce rock strength. Release of large
quantities of gases or liquids throughout the rock mass could produce
local pressures capable of hydraulic fracturing.
Media with high thermal diffusivity and conductivity and low thermal
expansion over the range of repository tenperatures are desirable. Such
media .could keep waste temperatures lower and undergo less mechanical
deformation during heating and cooling.
Of course, the medium must be considered within the total system
including waste form, mechanics of overlying rock strata, and hydrogeologic
flow system. As already indicated, focusing attention on the medium to
the exclusion of the other components of the waste isolaton system is
inappropriate and misleading. A host rock need not exhibit all these
properties nor necessarily have optimum values for any such
properties that it does exhibit. The properties required of, or desirable
for, any particular candidate rock mass depends on the other system
characteristics for the particular site. Such a systems view leads to
the possibility that a wider range of rock type might be of interest than has
traditionally been considered. Besides salt, granite, basalt and shale, all
of which are receiving considerable attention, anhydrite, tuffs and
unsaturated alluvium might warrant consideration in some environments.
DRAFT
-------
DRAFT
4C. BOCK MECHANICS 29
As pointed out by the National Academy of Sciences (25) and others (36),
designing a waste repository constitutes a major challenge for rock
mechanics both because of the thermal pulse and time frame during which
repository behavior might influence radionuclide migration. The mechanical
response of rock has relevance to problems of long term waste isolation
to the extent that it would affect the transport characteristics
of the geologic media in the vicinity of the repository site.
Important parameters of this response are rock strength (ability
to sustain differential stress) and ductility (ability to flow
without fracture). Both strength and ductility are influenced
by the presence of solutions and by increased temperature; in general,
the strength is decreased and ductility is increased. Thermal
expansion of rocks caused by increased temperature is a mechanical
response of particular concern in the case of a radioactive waste
repository. Aspects of rock mechanics relevant to problems of
operating the repository have been addressed previously in the Section 3E of
"The Operational Period". Considered here are those aspects concerned with
near-field and far-field effects that may influence the security of waste
isolation in the post-operational periods (37,38).
The principal mechanical effect from the waste will be the development
in time of thermal stresses in the rock mass around the* repository. Heating
will generate compressive stresses in the immediate vicinity of the' repository,
with associated tensile stresses beyond the compression zone. The magnitude
and extent of these stresses will depend on the layout and rate of heat gener-
ation of the canisters and will change with time. The specific effects of
these stresses will depend on the relative magnitude of the pre-existing
DRAFT
-------
DRAF
tectonic overburden stresses, and on the physical properties and other 30
physical characteristics of the repository rock and associated geologic
units. It is possible, for example, that the tensile stresses could
reduce tectonic stresses sufficiently to produce slip on pre-existing
fractures or joints in the rock mass. The tectonic stress in
the region should be determined experimentally to allow the probability
of such events to assessed. It is also possible that reduction
of the net compression within the region of tensile stress would
also increase the permeability, but new tension fractures are unlikely
because the region will still be in a state of compression.
Fracturing can also result from increased fluid pressures induced by
heating of the fluid or caused by pore volume reduction where thermal expansion
of the rock begins to close fluid-filled voids. The compressional strength of
all common rocks increases almost linearly"with increasing effective confining
pressure (and, therefore with increasing depth). Increased pore
fluid pressure decreases the effective confining pressure and therefore decreases
the strength of the rock, although any existing differential stress in the rock
is not reduced. Because all permanent deformation other than compaction results
from differential stress, increased pore pressure frequently results in fracture
of the rock. This would, of course, contribute to increased permeability..
The extent of these and related effects must be evaluated by means of the
science of rock mechanics, and will involve acquisition of data from both lab-
oratory and in-situ mechanical testing in order to develop analytical models
of the repository site. It should be emphasized, however, that models validated
by such data for short-term (operational period) and near-field mechanical effects
cannot be validated in the same manner for long-term and far-field effects
DM/T
-------
DRAFT
because of the great length of time required for measurable effects to be 31
observed at large distances away from the repository. Confidence
in the ability of such models to predict far-field deformations
over long periods of time will necessarily have to be based on
the accuracy of short-term predictions and increased understanding
of long term processes. Monitoring, for some period yet to be
determined, will be useful to assure the accuracy (and safety)
of the predictive models, and to provide an early warning system
should the models prove to be unreliable.
As yet, little effort appears to have been devoted to modeling of the region
away from the immediate vicinity of the repository — emphasis has been on near-
field deformations — but, in principle, no reason exists to prevent this. Improved
constitutive stress-strain relationships and better understanding of the
mechanical response of layered media are needed for a refined analysis, but
bounding limits could probably be determined with existing knowledge. The
results would indicate the likely locations and magnitudes of displacements
in the media, and those regions in which the rock strength might be exceeded.
Far more difficult, and beyond the capability of the models, would
be an estimation of the fracture density and other characteristics important
to assessing changes in transport properties of the rocks. Such information
would have to be obtained from combined analysis of data from modeling, exper-
imental, and theoretical studies.
Because of the potentially serious consequences of thermal expansion, the
different thermal loadings produced by reprocessed waste and fuel rods, respect-
ively, are a matter for careful analysis. The capability currently exists to
do this, and appropriate loadings can be determined once the relevant properties
of the media have been specified.
LMFT
-------
DRAFT
32
4D. Hydrogeologic Transport of Radionuclides
After a repository is fully loaded and sealed the most likely
mechanism for the release of radionuclides to the biosphere is by their
dissolution and transport in groundwater. Repositories located
below the water table are expected to fill with groundwater.
The rate of groundwater inflow will depend on the host
rock permeability, the depth of the repository beneath
the water table, the design of the shafts and boreholes,
the effectiveness of shaft and borehole sealing techniques
and, for salt, the rate of repository self-sealing as
a result of plastic flow. After the repository and associated
~.iafts are saturated, ground water flow through and in the
immediate vicinity of the repository, will be influenced by a complex
^upling of: (a) the thermal pulse; (b) the mechanical response of the host
reck to repository construction and to the thermal pulse; and (c) the
natural hydraulic gradients prevalent prior to repository construction. At
a distance of several hundreds to several thousand feet from the
repository ground water flow should be dominated by natural hydraulic
gradients, gradients which have vertical as well as horizontal components.
The ability of the ground water to dissolve the wastes
and transport the radionuclides from the repository depends
on the following factors: (1) solubility of the waste form and of
its container at repository temperatures; (2) the sorptive properties of the
host rock at ambient and repository temperatures, and the sorptive
DRAFT
-------
DRAF
properties of all other media along the groundwater flow path; (3) the chemical 33
properties of the groundwater including its acidity, oxidation potential,
ionic strength, ccmplexing agents present, and chemical changes associated
with emplaced waste; and (4) the rate of flow. Each of those factors is
a function of still other parameters. For example, the flow
rate is a function of permeability, porosity, and hydraulic
gradient; and the rate of dissolving the waste is also
•i function of groundwater chemistry. The sorptive properties of rocks
ni turn are a complex function of: (1) the type of minerals
comprising or lining the interstices and/or fractures through which
;-oundwater moves; (2) the surface area of the sorptive minerals;
3) temperature; (4) acidity and other chemical properties of the
,;oundwater; (5) the changing chemical composition
of the groundwater as it moves along the flow path; and (6) the changing
sorptive properties of the mineral surfaces due to breakdown of emplaced
waste.
Once radionuclides are transferred to the ground water, e.g., as
dissolved, complexed, or .colloidal material, factors such as radioactive
decay, dispersion, or the mixing of aquifer water with
that from an adjacent stratum operate to determine the concentrations
of radionuclides as they are transported towards points of potential
human access. Briefly, prediction of the transport of
radionuclides from a repository requires detailed knowledge
of: the physical chemistry of the waste form; transient repository
temperatures; and the three-dimensional distribution of the
aquifer porosity, permeability, dispersivity, hydraulic gradient,
sorptive characteristics, and water chemistry. These types of hydrogeologic
3RAFT
-------
DRAFT
and geochemical information are currently not fully available even 34
for the best known aquifers and may require considerable effort
to obtain at a repository site owing to the need to minimize
disruption of the repository area by drilling.
Despite the anticipated data limitations, valuable estimates of the
transport of radiocuclides by groundwater have been made by
Burkholder et. al. (39), by Marsily et. al. (40) and others (41)
using groundwater flow and mass transport models. By varying
the boundary conditions and such generic input data as
permeability, porosity, and sorption properties of geologic media
for specific radionuclides, those models yield a measure of the
relative importance of flow path length, flow velocity, and aquifer
sorption characteristics as barriers to radionuclide transport.
The models suggest that if a repository is sited where flow paths are
sufficiently long, such as may be found in regional aquifer systems,
then retention of radionuclides in the lithosphere for
periods of thousands of years may be considered likely. Site specific
analyses would require a firm knowledge of the vertical
components of groundwater flow, components which could transport
radionuclides to relatively shallow aquifers from which well water might
be withdrawn at some time in the future (see Section 5).
Such knowledge can be obtained by standard hydrogeologic
techniques for rocks of moderate to high permeability but will require
special techniques for less permeable rocks.
Not all of the types of data outlined above are needed in equal detail.
For example, if the flow path from a repository to a point of groundwater
DRAFT
-------
DRAFT
35
discharge is a few kilometers or less, and if flow is chiefly
via fractures, the solubility of the waste form and canister and the
sorptive characteristics of the aquifer(s) would be the only major barriers to
radionuclide transport. On the other hand, in regions where flow paths are
on the order of 50 kilometers or more, and when flow is chiefly through rock
interstices, the solubility of the waste might be of lesser importance.
Similarly, despite the acknowledged difficulties of characterizing, and hence
of modeling, the flow of groundwater through fractured media,
such characterization may not be necessary in all hydrogeologic environments.
For example, characterization of flow of water through fractured shale may
not be required if the ground waters are datable (say, by carbon-14) and are
older than 30,000 years in the vicinity of the repository (Methods for dating
groundwaters older than several tens of thousands of years are presently being
investigated by several universities through the sponsorship of the Office
of Waste Isolation and DOE).
It is almost axiomatic to state that groundwater flow and mass transport
models must be employed to help identify the factors affecting the radio-
nuclide transport that are critical at each proposed site for given time
frames. Current modeling efforts now recognize that:
(1) results should state error limits and should show possible ranges
of the effects; and (2) input data and boundary conditions
must reflect appropriate hydrogeologic and geochemical information.
Although a considerable body of data exists on the sorptive characteristics
of various media for selected radionuclides, these extensive measurements
have not been made under conditions that are characteristic of
the temperature, pressure, or chemical complexity of a
selected repository and of the deep aquifer environment; including
DRAFT
-------
DRAh
36
modifications resulting from the chemical breakdown of emplaced
canisters of waste. Hence the utility of existing data is limited.
Similarly, boundary conditions described as "conservative"
by a modeler may prove highly permissive to another scientist.
A range of boundary conditions must be investigated in
generic modeling, both to improve our understanding of the importance
of repository design parameters and to increase the likelihood
that generic modeling experience will be applicable to proposed sites.
The analyses of radionuclide migration cited on the preceeding pages
generally assume that groundwater flow will be controlled
only by the natural hydraulic gradients. However, as stated in the opening
paragraph of this section, ground water flow in the immediate vicinity
of the repository will also be influenced by the thermal pulse and the
mechanical response of the host rocks to repository construction as well
as to the thermal pulse. Detailed studies of such flow are also important.
For example, water-flow within the repository itself may
be of major importance if geochemical redistribution of the waste and
container materials might result in new chemical forms
with changed sorption characteristics, zones of high concentration
of certain elements, and altered time release rates of
radioactivity. Additionally, it has been postulated (42)
that in low permeability media such as shale or salt, inflow of water
into a repository might result in hydr©fracturing of the repository rocks due
due to thermal expansion of the water. Whether this is a realistic scenario
and the extent to which the postulated hydrofracturing might extend from
the repository into overlying aquifers is under appraisal (42) using ground .
iAFT
-------
DRAF
37
water flow models developed for analyses of geothermal reservoirs.
As a third example, in areas with natural vertical components of ground
water flow, shafts and (repository-related) drill holes constitute potential
short circuits of natural ground water flow; they merit serious study.*
* Despite efforts to seal shafts and boreholes, and the fractures induced
during their emplacement, they remain potentially important pathways for
radionuclide release from a repository. This important matter is under study
by OWI. Consensus does not now exist that borehole sealing materials and tech-
nology, currently anticipated to be special concretes, will provide an adequate
seal even over time frames of decades. A field demonstration is needed on bore-
hole and shaft sealing and on the interaction of the sealant with the host rock;
care must be taken to assure that placement of the sealant does not itself
induce microfracturing.
[RAFT
-------
38
UKAri
5. LONG-TERM RISK ASSESSMENT
DRAFT
Risk assessment of a nuclear repository involves two types of
predictions: probabilities of events or chains of events occurring that
lead to the release of radionuclides to the biosphere, and consequences
of these releases. "Risk" is the probability-consequence pair and
risk assessment must include an estimate of both. Within 3 to 5 years
a risk methodology should be available that will permit estimates of
repository risks and that can provide a partial basis for judging the
acceptability of options for radioactive waste disposal. Given the nature
of the uncertainties associated with repository storage, however, complete
risk assessment may never be possible. But risk assessment methodology
can bound the uncertainties, structure what is known and to some extent
identify what is unknown, identify important mechanisms for release, and
quantitatively estimate how uncertainties combine to yield risks. A risk
assessment should not be intentionally conservative. Rather, it should
attempt to be realistic and, to the extent possible, should indicate the
distribution of the risk that derives from uncertainty in the input parameters
and in the model itself.
The inability to do a precise risk assessment should not dissuade one
from attempting it. The insights and information gained from attempting a
complete and comprehensive risk assessment could be more valuable than the
resulting estimates of risk. For example, the results of an attempted
assessment should enable one to identify those characteristics
and properties of the system under study that can contribute significantly
to the public risk. This information would permit one to focus
better on further research and investigation that could significantly
improve the safety of the system.
DRAFT
-------
DRAFT
39
imni i
5A. Methodology of Risk Assessment
Among the most difficult aspects of risk assessment is ascertaining that
all dominant sources of risk or pathways of radioactive wastes to the
human environment have been identified and that these are accommodated
within the existing structure of models developed for the assessment.
These models are necessarily complex and it is difficult to validate
them, inasmuch as very limited data exist for radionuclide transport
in natural systems or for the effects of disruptive events on
such systems. Confidence in the models can be developed by comparing
their output with parts of natural systems for which data exist
(e.g., groundwater flow and discharge) and by comparing the output
from different models that have independently analyzed a hypothetical
but realistic situation.
Uncertainties associated with the results of risk assessment are of
four types. First is the uncertainty caused by lack of data. Presumably
this type of uncertainty can be reduced but not totally eliminated
by continued research and analysis. Detailed site exploration will, of
course, provide data for use in the site-specific risk assessment.
Second, uncertainty results from lack of experience. This type of uncertainty
involves identification of mechanisms and scenarios; it is only partially
reducible by continued research. Third, uncertainty exists due to natural
variations in physical properties. In principle, this uncertainty is reducible
t
with extensive field exploration and analysis, at particular sites, but practical
limits are imposed by cost, time, and the need to protect the integrity of the
site. Fourth, uncertainty results from an inability to predict the future of
long-term geological and climatic processes, and of social evolution. These
uncertainties are predominantly not reducible, but can be estimated by
DRAFT
-------
DRAFT
40
careful selection of appropriate scenarios. The uncertainties from the
sources mentioned above are neither additive nor of equal significance;
sensitivity analyses can be used to determine their relative importance and
contribution to the overall uncertainty of risk estimates.
Possibly the most important aspect of risk assessment of radioactive waste
disposal is to understand how the uncertainties mentioned above enter into the
calculation of risk. Detailed modeling of the processes by which waste could
escape from a the repository, move to the biosphere and, ultimately,
to humans is required if the effects of these uncertainties, both on the
likelihood of release and on the consequences, are to be evaluated.
An important limitation to the ability to assess risk lies in associating
probabilities with geological processes, both for long-term events and for
those of immediate engineering interest. Development of probabilistic
models in geology is recent, but advancement of the techniques
is proceeding rapidly. For the present, the inability to model
geological and hydrological processes is a serious limitation
to risk assessment. However, continued work over the next five
years may yield sufficient progress that at least short and immediate
term predictions can be better made. Comprehensive probabilistic
predictions for the long-term, and even detailed probabilistic predictions for
short and intermediate terms seem farther off. It may be that "conditional"
risks will have to be calculated — i.e., risks that would exist if the assump-
tion made in the analysis about future events (climatic factors, land use, etc.),
do hold.
DRAFT
-------
DRAT
41
Major current work that could be classified as risk assessment of nuclear
wastes in deep geologic formations is being performed by Sandia Laboratories
and Lawrence Livermore Laboratories for the Nuclear Regulatory Commission, by
Arthur D. Little, Inc. for the Environmental Protection Agency, and by Battelle
Pacific Northwest Laboratory for the Department of Energy. These analyses are
for generic (i.e., non-site specific) repositories, and generally are for bedded
salt. Generic analysis is not applicable to particular sites, but does permit
development and exercise of a methodology that can be applied at the appropriate
time to specific sites selected for study. Several years of activity have led to
a consensus that the three most important modes of potential release are:
transport by groundwater, catastrophic release to air (e.g., as
a result of meteorite impact), and removal by man (e.g., from
mineral exploration). Release scenarios that involve migration
in groundwater flow have received the most attention to date
because dissolution and transport in groundwater is believed to be the
most likely mechanism for transfer of radionuclides back to the human
environment.
DRAFT
-------
42
5B. Local Release from the Repository and Movement Through the Subsurface.
It is expected that only areas with a long history of geologic stability
will be considered as potential candidates for waste repositories. Certainly,
past stability is a reasonable requirement for repository site selection.
However, a history of geologic stability is not sufficient to
assure future stability for the area with or without the development
of a repository. The development of a waste repository could
constitute a significant alteration of the local stress field and no amount
of back-filling or sealing of shafts will restore the area com-
pletely to its original conditions. Once a repository is filled and sealed,
the area containing it might evolve differently than if the repository had not
been placed there. Release modes that need to be considered therefore include
those that are self-induced as well as those which could occur independently
of the presence of the repository.
Analyses of local releases have considered such factors as
waste-rock interactions and thermomechanical behavior in the near field. Those
processes initiated by the presence of the repository constitute one category
of disruptive events. In addition, external events independent of the presence
of the repository—such as seismic activity, magma intrusion, or meteorite
impact—may enhance processes that directly or indirectly lead to release of
the waste.
One assumption commonly made in assessments is that immediately
upon closure of the repository, or at some postulated time after
closure, radionuclides in the wastes are completely dissolved
in circulating groundwater. The tine required for their movement
to the environment is then estimated by assuming that the radionuclides
move through overlying formations along specified paths to the surface. The
BRA
-------
DRAFT
43
pathways may be arbitrarily defined, or determined by sophisticated ground
water transport modeling. Such an assumption is conservative,
because it overestimates the rate of release from the source.
To the extent that radionuclides are prevented from going into
solution, the release rates are proportionately reduced.
If waste is locally released from the repository, it must move through sur-
rounding formations before entering the human environment. This
movement will be substantially controlled by the hydrogeologic
regime, inasmuch as the groundwater flow field largely defines
the potential for movement of dissolved radionuclides. The flow
field and potential pathways for dissolved waste to reach the surface
are difficult to predict for times in the future, but appropriate bounds can be
placed on them. For the near term (e.g., first few hundred years) the pathways
depend on the present distribution of permeabilities (including
fractures), the hydraulic gradients, the thermal field, geological
inhomogeneities, unplugged boreholes, and other phenomena either
resulting from the creation of the repository or exogenous
to it. In the distant future (e.g., beyond a few hundred years) the pathways
depend also on climatologic and geologic changes.
Analyses now in progress treat hypothetical generic profiles, modeled as Darcy
flow regimes. Fracturing and faulting are modeled as short circuits in this
flow. Parametric uncertainty is treated through sensitivity studies, but at
present not by assigning probability distributions to the input. The results of
these analyses are rates of release to the surface for given source terms
and flow conditions as a function of time. Future risk assessment will
have to address probabilistic description of flow conditions and changes in those
ORAf
-------
BRtot
44
conditions, and the possibility of non-Darcy flow regimes caused by convection
currents or other processes.
5C. Movement of Radionuclides Through the Environment and Uptake by Humans.
If radionuclides become dissolved in circulating ground water, they could
enter the human environment in a variety of ways, such as discharge of ground
water to surface water (e.g., springs, rivers, etc.) or direct withdrawal of
contaminated well water by humans, or indirectly (i.e., through food).
The subsequent movement of radionuclides through the environment,
their accumulation in soil, air, and water and their eventual
uptake by humans has received considerable attention by the IAEA (62).
Analysis of environmental impact and human uptake of radionuclides could add
as much, or even more, uncertainty to the evaluation of risk as that associated with
release of wastes from the repository and their movement through the subsurface.
Surface features, particularly those which may be influenced by humans such as river
water flow rates, creation of lakes and dams, and changes in water runoff patterns,
can be expected to change more rapidly than subsurface features, and are consider-
ably less predictable.
Factors directly related to future human actions play a prominent role
in determining the effects on humans of released nuclear wastes. One of the
more important uncertainties is population density in areas that could
potentially become contaminated. Population density coulc'. vary
from zero to far higher than in the now most densely populated
cities. Furthermore, factors related to water usage are important.
If agricultural areas should become contaminated, irrigation from
a contaminated river or contaminated ground water would significantly increase
DRAFT
-------
DRAFT
45
the contamination of food crops and thus increase the human uptake of radio-
nuclides through food chains. Nevertheless, as with certain geologic and
climatic factors, bounds can be placed on the ranges of demographic
and other human factors. Because answers to questions regarding human actions
are unknowable, consistent assumptions must be made so that comparison of
relative risk is possible.
5D. Accuracy, Gaps of Knowledge, and Future Development of Risk Assessment.
The accuracy of risk assessment is not limited by the methodologies of risk
analysis but by the uncertainties associated with geologic
and hydrogeologic processes and future human activity. The construction of
scenarios leading to the release of waste from a repository is generally
agreed to be an important issue in evaluating those risks.
Important knowledge gaps for risk assessment include:
0 The ability to incorporate the variability of physical and
chemical properties in geologic and ground water flow modeling.
0 Identification of important mechanisms of waste-host rock
interaction and quantification of levels of uncertainty in
those mechanisms.
0 The ability to characterize the natural variability
of repository sites by means of exploration programs.
0 The ability to predict long-term geological, hydrological,
and social changes.
I RAFT
-------
DRAFT
46
In these four areas, improvements in knowledge does not necessarily mean
reduction of uncertainty. Uncertainties associated with the first three
of these areas are potentially reducible to numbers. Continued or
accelerated work should significantly improve our abilities to deal with the
problems, particularly in better understood waste emplacement media.
Uncertainties associated with the fourth area (long-term predictions)
may not be quantifiable, but further work directed at developing
a way of handling long-term predictions may provide satisfactory
alternatives to explicit quantification.
DRAF
-------
DRAFT
6. SITE SELECTION AND SITE CHARACTERIZATION 47
The establishment of a site for geologic disposal of nuclear wastes
must necessarily proceed in a certain orderly sequence. First, the
properties of a site which make it desirable for such disposal must
be specified, at least broadly. Second, some candidate regions which
may have such properties must be selected. Third, these regions must be
studied in some greater detail to determine some candidate sites of lesser
geographical extent for even more detailed investigation. The process
to this stage can be broadly denoted as "site selection", and the sequence
outlined above depends strictly on scientific or technical
information.
The site selection process can also take on a different character.
Sites can be selected for detailed investigation based on availabiligy, ease of
access, or simply ownership by appropriate bodies (e.g., DOE). These
sites may then be studied in more detail to determine their suitability
for a repository. In either case (selection by technical/scientific
culling or selection according to other factors) the investigations
of the site which lead to its designation as a repository include much
the same studies and factors.
In the discussion that follows, a sequence of purely technical and
scientific decisions is assumed. All of the information required for
the decision is needed no matter which of the selection processes is
followed.
DRAFT
-------
DRAFT
48
The identification of candidate sites by scientific culling should
ideally begin with regional studies with a broad program that would
include geologic, hydrologic, geochemical, and geophysical studies
needed to confirm, or disconfirm, the presence of multiple natural barriers
within the geology and regional ground water flow systems. Involved
here would be geologic mapping, test drilling, and regional geophysics.
Areas of perhaps 2,000 to 10,000 square miles would be sought
that have favorable hydrologic and geologic characteristics. A
subsequent more detailed exploration program within such large
areas would identify 40 to 400 square mile blocks that appear
particularly favorable as potential repository sites. Finally,
candidate sites would be selected with areal extents of perhaps
10 square miles.
At each step in this process, appropriate technical criteria will be
required to guide the work and facilitate judgments of suitability.
Criteria are under development by the Nuclear Regulatory Commission (44),
by the Department of Energy (46), and others (43, 45, 47). Such
criteria are based on the notion that we must limit, to the extent possible,
the likelihood of radioactive wastes being released from the repository to
the human environment, and the consequences of the release. Therefore,
the criteria should address systematically the significant features of
the repository and its environment that bear on its abiUty to isolate
the wastes for adequate periods of time.
Although satisfaction of appropriate technical criteria is essential
at each stage of the work, other factors also are relevant to the site
DRAFT
-------
ORAF
49
selection process, and could dominate. Among these are ease and
cost of access, distance from other societal activities, and societal
acceptance of the location as a candidate repository site. Large areas
of the country may be ruled out by these latter nongeologic
factors, totally independent of their possible attractiveness from
a geologic perspective. Other areas may be highly favored by the non-
geologic factors. Indeed, the practical aspects of gaining access to
land for reconnaissance and exploration may impose severe restructions on
the areas available for consideration at least over the near term
If (as seems likely now) the criteria for suitability of a site
cannot be specified in great detail because of the complexity of the
geologic settings, it is possible that the initial selection will need
be done mainly on the non-technical factors. Whether a site is selected
this way or by a strictly scientific search and culling it will need be
examined in detail and compared against the underlying radiological
safety criteria.
Once a potential site is identified, an intensive exploration program
will be required to characterize it. The predictions and hypotheses made
during the preceding more general reconnaissance program can be verified
or disproved. Detailed geophysical studies and some coring will be
required as part of a lengthy program. Among the most important data
to be gathered are details of the local ground water hydrology with
respect to the regional flow system, the sorptive characteristics of the
rocks along the flow paths; the geochemical and thermal properties,
DRAF
-------
DRAFT
50
the fracture patterns and stress field of the candidate host rock and its surrounding
rock structures; and the resource potential of the site, if any. The
latter point is important both because the presence of resources in the area
might provide incentive for human penetration of the repository at some
future time after memory of the repository's existence has disappeared and
because competition with possible near term resource exploration claims might
argue against the site's use for a repository. A site-specific safety
assessment, of course, will also be required to compare the consequences
predicted over a range of scenarios with programmatic and regulatory
standards.
The comparison of the information gained by the site characterization with
the criteria for site suitability is a task of some magnitude and
some importance. Only with an ongoing and intense program for the
evaluation of the forthcoming information can the data be shown to
be satisfying the needs of the criteria and the decision process.
Each site, of course, will have a unique set of features and characteristics.
Only the ability of the total systejn to guarantee containment is of interest,
however. Considerable flexibility is therefore required in specifying
and applying site suitability criteria. But caution must be exercised
to prevent the detailed criteria from being adjusted simply to
accommodate the actual features of a particular site under study.
This care can best be taken in'the formulation of the criteria.
Although, in principle, the data required for site characterization can
be acquired by employing standard geophysical, geochemical and hydrological
exploration techniques, in practice important limitations exist .
DRAFT
-------
DRAFT
51
Geophysical techniques are limited in their ability to obtain information
at depth and to detect and identify small anomalies in the rock
structure. The frequency and hydraulic interconnection of fractures
are particularly difficult to determine either by geophysics or by
permeability tests from a limited number of boreholes. The ability to
extrapolate from the rock immediately surrounding such holes to the large
scale characteristics of a site is limited. As already discussed, more
boreholes could be drilled. Under some circumstances this may be undesirable.
It is prudent, therefore, to consider the trade-off between
the need to obtain sufficient information to characterize a site
and the need to maintain its natural integrity. Uncertainties will
always remain and we must be careful not to demand more data that can
be obtained, or than is really needed. The degree of detailed information
required of a particular site can, to some extent, be determined by
conducting sensitivity analysis in the site specific safety assessment
work. Fortunately, current techniques do seem to permit the acquisition
of sufficient data to characterize a site satisfactorily from an
overall systems perspective. New geophysical techniques including
high-resolution seismic and acoustic, short-pulse radar and continuous
wave interferometry methods now under development will contribute to
greater understanding of a potential site.
DRAF.
-------
DRAFT
52
NOt everything will be known, however, before a shaft is sunk and
horizontal tunneling can be done. Important data will be continuously
collected during the site characterization study during the various
phases of the repository construction such as shaft sinking, drift excavation,
and during operation of the repository.
It is important to point out that to date most modeling and
on-location experimentation and measurements have been done for the purpose
of generally characterizing media. Such work is very useful but must be
considered primarily as having generic value by developing methodology
development and heightening awareness of important questions that must
ultimately be applied to specific sites. The importance of examining
actual potential sites cannot be overestimated. Moreover, given
the long time required to do the necessary work, there would seem
to be advantages in conducting detailed site characterization
studies at a number of sites simultaneously.
DRAFT
-------
[RAFT
53
I I till •
7. CANDIDATE MEDIA
Information available and work underway on candidate host
rocks will be summarized in this section. There is, of course, a
range in the physical, chemical, and mechanical properties of
any rock type both at any given site and at different sites.
The rather general tone of the subsequent discussion of geologic
media is not intended to obscure the fact that these variations
occur and must be taken into account and that generic data about
a media, though a necessary first step, are no substitute for
the required detailed site investigations.
7A. Rock Salt.
Rock salt, of either the bedded or dome variety, has received the
bulk of the attention over the past decades.
The original NAS-NRC Committee (48) recommended salt because
of its thermal and physical properties, and its very existence
for 200 million years or more which demonstrated its isolation
from aquifers and the stability of the geologic formations in
which the salt is located. Another desirable feature of many bedded
salt basins, a result of their evaporite origin and subsequent
tectonic history, is their relatively simple structure and predictable
stratigraphy. It is often possible to establish geologic structure
and predict lithology of the formations over a wide area with
relative ease. Because of the early start on investigations of
salt considerable data are available on the properties of salt
and its response to thermal loading (18). Extensive laboratory
testing is underway at Sandia (49) and by OWI (15) and in-situ heater
DRAF.
-------
54
experiments with control of fluid pressures are being
conducted at Asse, Germany (50) and Avery Island, LA. (15).
Thermal experiments are planned for a potash mine near
the proposed WIPP site in FY 1979. Intrinsically
desirable properties of salt include uniformly low
permeability, high thermal conductivity, abundant availability
in thick masses, and plasticity that enables fractures to self heal at
feasible repository depths.
Several questions have been raised concerning specific phenomena that
might affect the ability of salt to afford the permanence of isolation of
waste required in a repository; some and possibly all may be accommodated
by appropriate engineering or operational approaches. All the problems
identified in the subsequent discussion are receiving intensive study
domestically and abroad, and it is expected that within the next 5 years
that the relative importance of these problems will be established and
many will be resolved. The R&D facility associated with WIPP (a proposed
transuranic waste repository) will provide experimental in situ verification
of these results.
One potential problem is associated with the plasticity of
salt and its impact on long-term retrievability. Salt is more plastic
at any given temperature than most other rock types.
Two specific effects must be considered: the first is the closure
of the hole in which the waste is placed and the second is the
closure of the storage rooms. Engineering approaches (such as controlling
the temperatures in the salt, or use of mechanical restraints)
can be employed to mitigate such effects. Each of these approaches
liRAFT
-------
DRAFT
55
along with numerous others, has been considered and specific
practical approaches have been defined to accommodate
the situation. Retrievability in a salt formation could
be maintained, The cost would be determined in part by the
length of time retrievability is required.
Since water can lower the mechanical strength of salt, its presence
and variable concentration could be a problem. The mean water content
in salt is low (usually less than 1 percent), but local variations
over wide ranges occur within salt masses. The water content
tends to be lowest in the salt domes along the Gulf Coast where
the deformation and flow process which formed the domes seem to have kneaded
the water from the salt. Bedded salt strata such as those in New Mexico,
Utah, mid-Continental and Eastern USA are generally more variable than
dome salts in their chemical composition and mineralogy. They commonly
contain complex hydrous halides, sulfates, and larger variations in contents
of free and combined water. More detailed mineralogical characterization
may be required for sites in bedded salt than in domal salt.
The high sensitivity of salt to solution processes requires that
sufficient knowledge of regional and site hydrogeology be obtained
before a repository is operated, and that some assurance about possible
future groundwater flow regimes also be obtained. At present such
understanding depends in part on our limited ability to predict tectonic
stability and future climate, both of which need improvement in levels of
effort and methods of study. However, only limited increased capability
is probable over the next decade, although programs are underway at
several universities, the USGS, and other organizations.
Another potential problem is related to the
DRAF
-------
[RAFT
56
interstratification of the non-halite members and shale beds associated
with bedded salt formations that result in higher average water content,
and a broader range in chemical composition and mineralogy. It has been
postulated that the thin interbeds of shale, anhydrite or dolomite could
increase lateral flow rates. However, it is now clear that these thin
interbeds need not be active conduits for water flow since
the effects of dissolution can be detected and the presence
of flowing water can be measured. (For example, at WIPP such interbeds
are known to exist and the cores taken indicate no such
past flow.) The interbeds might conceivably become a conduit
if they were connected to an aquifer in the future or if
the repository were penetrated by water. In general, however, these
interbedded layers have low permeability. The consequences
of flow through the pores of the interbeds can be estimated
by means of existing calculational techniques and are
likely to be less significant than other plausible
failure paths.
The solubility of rock salt in water is two orders of
magnitude greater than that of any other candidate medium. In the
event that construction, operation, or loss of integrity over geologic
time causes water to flow through a salt repository, dissolution
could cause release of the waste. Although the principal channels
for flow in the salt might be along shale interbeds or fractures that
may have concentrated clay minerals, the sorptive capacity of these
channels will be diminished by the presence of solutions of high ionic
strength. Similarly, the sorptive capacity of other geologic materials
along the flow path will be decreasd by the high salt concentrations.
[RAFT
-------
DRAF
57
Another potential problem is the presence of major heterogenieties
in the salt formation such as structural features that result from flow of
salt or large brine or high pressure gas pockets which
cannot be detected until encountered. (Gas pockets are
not limited to salt formations - they are characteristic
of many geologic media and formations). These defects
may not represent a threat to the long-term integrity of
the repository, but they could represent a threat to the safety of those
who are constructing and operating the facility. For example,
mining into a large brine pocket which is not connected
to any source of water could harm the miners but should not threaten
the capability of the repository to perform its isolation function.
Special geophysical methods are being developed for characterizing geologic
formations which have a high electrical resistivity such as salt.
These techniques have been used successfully in a few
cases but only limited applications have been achieved.
A major national program (at Sandia, LLL, and USGS)
on instrument development is underway to increase the range
(up to a kilometer) and resolution (down to tens of centimeters) of
these methods which offer the potential to locate trapped brine, disturbed
rock structures, variations in water content, etc., in advance of
mining.
The thermal expansion of salt is almost three times that of other
candidate media. Thermal expansion is a significant physical property
in that it is the major design parameter controlling the vertical uplift
resulting from the increase in temperature in the repository volume
and the surrounding geologic media. A major impact of the thermally induced
I RAFT
-------
nun
58
uplift is the flexural deformation of the overlying strata which has protected
the salt formation from the groundwater above. In this area the rock
mechanics analysis is adequate to predict rock stresses and strains for a
particular site. However, the capacity of these stresses to induce fractures
must be determined for each site and coupled with fluid flow phenomena.
With prudent design matrices that limit the temperature increase, fractures
within the overlying strata can be minimized.
Another issue with rock salt is related to the physical-chemical
behavior of the small brine and bittern (brines containing high concentrations
of magnesium and calcium chlorides) inclusions which are present
within and at the boundaries of crystals of bedded salt (51,52). Experiments
(18) show that in salt these brine inclusions will migrate up a thermal
gradient toward the heat source. The pressure in the brine inclusions
increases when they are heated. If the temperature near the waste canisters
is high enough, the fluid pressure in the inclusions may exceed the strength
of the salt and cause it to flow or even to decrepitate, releasing the
fluids and decreasing the thermal conductivity of the salt. It is expected
that these brines will reach the holes in which the waste is placed. The
nature of the effects that will take place when and if the brine arrives
in volume is not determined, but are likely to be deleterious.
Several hypotheses with respect to the potential reactions that may
take place have been proposed. One hypothesis proposes that substantial
quantities of brine will be drawn from the surrounding salt, diminishing
DRA1
-------
DRAFT
59
its moisture content. The brine phase concentration in the salt in a small
volume surrounding the waste will increase causing the mechanical strength
to be reduced. The strength of salt which has been depleted in brine should
increase. The brines are expected to fill the void space around the
waste canister so that corrosive liquids that have low pH values (by
hydrolysis at repository temperatures) will be in contact with the waste
canister. Under these conditions it is expected that these brines, with
high concentration of dissolved salts, will attack waste canisters and
subsequently the waste. However, the rates and extent of such reactions are
still unknown.
Another hypothesis states that elements from the containers and waste
added to the brines will increase the total amount of dissolved salts
and to some, extent will increase the amount of brine present at constant
water content. An alternative hypothesis states that the point must be
considered as to how fast glass or spent fuel will dissolve in a solution
of saturated brine. It does not take into account the fact that, as the
metallic canister corrodes to form sulfides, oxides, chlorides, and hydroxide
corrosion products, the amount of water will decrease by its
reaction to form hydrogen, oxides and hydroxides. The heat generation
rate is maximum when waste is first emplaced, and after
an initial heat-up period, the temperature will decrease,
reducing the potential for bringing water to the wastes. As the temperature
DRAFT
-------
DRAFT
60
decreases with time and as the quantity of water decreases due to corrosion
interactions, dissolved salts may precipitate, minimizing the rate at which
the high level waste could be attacked. Specific steps to reduce the potential
for corrosive brines contacting the waste and canister must be taken. Materials
can be placed in the vicinity of the bore hole specifically selected on the basis
of their ability to react with brines to form solid corrosion products, thereby
reducing the detrimental effects of high brine content on the mechanical
properties of salt.
The chemical reactions which will take place in the vicinity of
the waste, accelerated by elevated temperatures and high pressures, are
complex and far from clear. Complicating the picture further
is the unknown but certain effect of the radiation on these
^
chemical reactions. Gaps in our knowledge here need to be filled by
laboratory data and in-situ data, if possible.
>
Often design features may be devised that can resolve potential
problems. For example, it has been proposed that after a repository
closure brines might attack the seals of bore holes and shafts. If the
shafts and bore holes are located at a distance and down
gradient from the heat source, it is not clear that brines
would collect in these areas and pose any threat to the seal.
In bedded salt, large diameter (the order of 1000 ft)
vertical, and possibly permeable structures, known as breccia pipes, occur.
They are formed by a solution mechanism but the nature of the formation process
DRAFT
-------
DRAFT
61
is not understood. Such pipes exist in many salt basins and
they extend vertically through several of the geologic strata of both saline
and non-saline origin. These pipes, if located near a repository,
and if permeable, could provide a shortened path to the
biosphere. If such a feature is in existence its presence
can be detected with existing geophysical and geologic exploration
and evaluation techniques. The existence of a permeable
breccia pipe in an area would provide the necessary reason not
to construct a repository in the vicinity. The main question concerning
the understanding of the formation of breccia pipes is to assure
that one could not form in the vicinity of a repository after the
repository was closed. If it is shown that breccia pipe formation is
associated with geologic processes which are no longer active in a given
area then they will not represent any threat to a repository in salt.
Sandia has underway an extensive program to provide an answer to
this question and should be able to resolve the question within the next
few years.
In summary, salt is best understood of all the geologic media in
»
many respects and it offers advantages in thermal properties and plasticity.
With conservative engineering salt may be an acceptable repository medium.
However, salt is soluble and it does not provide the sorptive qualities
of other rock types nor is it benign to interactions with the waste and
container. All of these could be troublesome in the event of breachment of
repository integrity.
DRAFT
-------
DRAFT
62
7B. Anhydrite (Calcium Sulfate)
Anhydrite is not yet a candidate repository host rock in this
country, though it has been considered in Switzerland. Anhydrite is
available in thick, homogeneous masses of low permeability in many
areas of the U.S. west of the Mississippi River. Anhydrite has a thermal
conductivity as high as rock saltj but expands only one-third as much
at the same temperatures. Anhydrite is stable to very high temperatures,
and, in general, contains little free water. The solubility
of anhydrite in either weak or strong brines, decreases as temperature
is increased up to 230 degrees C, a unusual phenomenon. This means
that fluids cannot migrate up the thermal gradient into the repository
by solution-dissolution mechanisms. From the few data available (Sandia)
anhydrite seems to have significantly higher sorptive capacity than the
salt strata with which anhydrite commonly is associated. Anhydrite
is more difficult to mine than rock salt. It is also brittle and fractures
rather than flows at reasonably repository temperatures and pressures, and
the' fractures that develop do not self-heal.
#
Anhydrite hydrates to gypsum in the presence of water
of low ionic strength, but not with water of high ionic strength. The
hydration of anhydrite is accompanied by swelling of 62 percent by volume.
Fresh water that reached anhydrite would be partly fixed as
gypsum, and the resultant expansion could possibly prevent further flow or
could even cause fracturing.
In 1979, the USGS proposes to initiate studies of anhydrite to evaluate
its potential as a repository host rock.
DRAF.
-------
DRAFT
63
Anhydrite occurs in the same geohydrologic environment as salt, and
is usually interbedded with salt, shale, and dolomite. These bedded
sections tend to have long flow paths for liquids that escape from
individual beds and may offer a considerable range of sorptive formations
along the flow paths. Though much less soluble than rock salt, anhydrite
is still a soluble rock. The hydrologic regime is characterized by
horizontal flows along bedding planes, but some channeling (cavern formation)
has been observed in anhydrite similar to that of limestone and gypsum.
JMFT
-------
DRAT
64
7C. Granite and other Crystalline Bocks
Granite and other crystalline igneous rocks and metamorphic rocks, such as
gneiss, have been proposed as potential geologic media for a repository
because they occur in very large and relatively homogeneous masses. Crystalline
rocks such as granite are attractive because of their strength and structural
stability. The water content of the rock is low and largely in fractures.
Granite has good sorptive properties and the ionic strength of water in
such formations is usually low, minimizing corrosion rates and the effects on
sorptive characteristics. Because of these favorable natural conditions it
has been estimated that the waste containers in such conditions might maintain
their integrity over many hundreds of years (53).
While the U.S. has placed major emphasis on salt, other countries with
abundant granite such as Sweden (53) Canada (20) and the United Kingdom (54)
have been evaluating it as a repository medium. Secondary
barriers within granite to enhance the retention of radioactive
nuclides within the repository are also being studied.
One technique proposed by the Swedish Power Board is backfilling
the volume around the waste canister with a sorptive material
mixture (quartz and smectite) to buffer the ionic strength
and acidity of the water and minimize corrosive attack
on the canister while still retaining high sorption in
the mineral mixture. While this approach has been considered
a possible approach to the retention of radioactive nuclides in a
repository, the validity of this proposal has not yet been demonstrated.
[RAF
-------
DRAFT
65
The groundwater flow paths through crystalline rocks are normally, but not
always, shorter than those in bedded strata such as shale. This path
length depends, of course, on the geohydrologic setting. Crystalline rocks
commonly occur in geohydrological environments that have experienced
complex tectonic histories during which the rocks were metamorphosed and
thermally disrupted several times. Changes in rock properties probably occurred
during these events. Rack properties may have been homogenized by pervasive
events or may be quite variable from place to place and difficult to ascertain
adequately for repository design. Geohydrological characterization of
such crystalline rock terrains will be difficult.
Granites are usually fractured; consequently the permeability of the
rock mass depends upon flow through a network of fractures
rather than flow through pores. The permeability of the "rock mass"
is much less than that of the individual fractures. However, flow through
a fracture depends on the size of the aperture which to a large extent
is controlled by the normal stresses acting across the fracture.
Since these stresses increase with depth, the permeability
of fractured rock usually decreases with depth. Development of a model for
fracture flow is a difficult technical problem that is receiving considerable
attention. The use of granite for a repository, therefore, may
require sites at depths at least as deep as 500 meters below the surface
where the fracture permeability will be sufficiently low
so that, considered from a conservative position, it may not represent a
threat. Another possible approach would be to inject a thin grout into the
fractures to reduce permeability.
WAFT
-------
FT
66
7D. Shale and Related Rock Types
Shale and related rock types, here collectively called shale, have
been identified as geologic media which have a number of attributes which
make them attractive for the isolation of nuclear waste. These properties
include low permeability, plastic flow under lithostatic load, good
sorptive characteristics and low solubility in water. Shale is
abundant in thick masses throughout the Midwestern and Western United
States. However shale is a generic name for geologic materials which
vary widely in character and composition. The suitability of shale
will have to be determined on a case by case basis.
The response to a repository of shale formations will be similar to
that of granite and other hard rock formations. After
the repository is closed, water will fill all voids and
the waste canisters will be in intimate contact with the
water. Repositories will have to be designed with this
characteristic in mind with appropriate provisions made to minimize
the potential for the transport of radionuclides. While the filling of a
repository with water may be cause for concern, several factors
must be considered. If the residence time of the water is long, or the
solubility of the radionuclides in the water is small, and if the sorptive
properties of the shale are high, then the potential for the radionuclides
reentering the biosphere is small. Two reasons why shale is
considered attractive is that it has high sorptive characteristics for
radionuclides and it has low interstitial and generally low fracture permeability,
DRAFT
-------
DRAFT
67
limiting the ability of water to flow through the formations. The
composition of the water in shale varies widely, and in some areas is
quite saline.
While shale offers many positive characteristics, a feature which
may limit the usefulness of some varieties as a repository medium is thermally
induced mineralogical changes which occur at temperatures as low as 100 degrees C
in those varieties such as expandable clays. It may be postulated that the
partial dehydration of some of the constituent clay minerals as well as heating
pore water will increase the fluid pressure on the rock and, thus,
drastically change the stress deformation properties of the rock. The visco-
elastic properties of shale, which are temperature dependent, would also be
affected but little data has been available to estimate the effects (10)
At temperatures above 100 degrees C mineralogical transformations
take place at rates that increase the permeability and porosity of shales
during typical repository containment periods and thus affect the radionuclide
transport rates. Attempts have been made to perform single preliminary field
heating tests in two shaly rocks (Conasauga shale, TN; Eleana argillite, NV) to
determine appropriate experimental methods to measure the temperature sensitivity
of shale and to obtain generic data for the initial and post-heating states of
these rocks. Experimental study of waste-shale-fluid interactions are
underway at the Pennsylvania State University. It is probable that more than
five years of intensive effort will be required to obtain adequate data to
develop generic knowledge of the maximum temperature that shale
repositories can sustain.
The concern of the potential deterioration in shale, associated with
temperatures over 100 degrees C in large volumes of rock, cannot be
evaluated at this time. If one considers profiles for various areas and at
DRAFT
-------
DRAFT
68
various times the volume of shale that would be above 100 degrees C in
temperature, assuming a waste canister temperature of 250 degrees C, is
quite small in comparison to the repository volume. However, the
effects of heat on ground water in the repository after closure, for
example, the potential for hydrofracturing needs careful examination.
The sorptive capacities of shale (55, 56) are generally
high for most radionuclides and they are not significantly
reduced at elevated temperatures above 100 degrees C.
Moreover, the diminution of the sorptive properties will not be a significant
factor if the repository is designed so that only a small fraction of
the shale in the repository will experience temperatures
greater than 100 degrees C. Even with the knowledge
of the sorption coefficient, forecasting the transport of
radionuclides in shale is considered to be a difficult problem. Because of
the low permeability of the shale, transport is expected to occur principally
by fracture flow, and there is not yet an adequate theory for
accurate forecasting of such flow. Although there are several active
research and development programs with respect to this problem, five years
or more may be required to develop an adequate model.
A characteristic of shale which must be viewed as a potential
drawback is the difficulties associated with mining and keeping the
tunnels open. Inhomogenities in shale which significantly affect its
structural characteristics are difficult to identify in advance of mining.
An example of such effects can be found in the Eleana argillite at the
Nevada lest Site (NTS). Based upon core holes there, it is estimated that
DRAFT
-------
ORAFi
69
about 20 percent of the volume of the shale is a highly
plastic material which readily deforms to close unconstrained openings.
A detailed engineering study (57) of the requirements to mine
and maintain the opening would require extensive rock bolting and sets, and
and would add approximately 25 percent to the total cost of mining tunnels at
the NTS.
7E. Flood Basalt
Flood basalt is a candidate medium at the Hanford site, W&, where it
occurs in a thick section near the middle of the extensive
basaltic terrain of the Columbia Plateau. There are other thick basaltic
sections in the Northwestern United States in Idaho and Oregon. These
basaltic terrains are geologically young, and are tectonically active.
Columbia Plateau basalt is commonly intensively jointed in a
columnar (i.e., vertical) fashion but these joints may be filled with
alteration products (predominantly clay minerals), and so the rock mass
has low permeability. Such masses of low permeability rock are being
investigated as potential repository horizons. However, other basaltic
strata in the section are fractured and not sealed by alteration products.
There are also sedimentary interbeds within the section that are water-
bearing. The section at Hanford thus comprises a system of alternating
aquifers and zones that are relatively impermeable to water. Groundwater
flow through heterogeneous fractured media like these are
imperfectly understood. Neither theory nor models exist today for such
complex flow.
DRAF
-------
DRAFT
70
The partially altered fractured basalt and the sedimentary interbeds
between the basalts have moderate to high sorptive capacity. The mineralogy
of the altered fracture zones and sediments is different from that of the
unaltered basalt, and so, sorptive properties of all the minerals in the fresh
and altered basalt as well as the sediments must be determined in order to
model radionuclide transport. An intensive program of laboratory studies is
underway to develop these data before 1980. The water in the horizons where
most flow occurs usually has low ionic strength. Basalt apparently
has the ability to tolerate a high thermal load in a repository, but no
actual field heater tests will be made until 1979.
7F. Tuffs.
Tuff is an extrusive geologic rock produced by volcanic activity; it has
a broad range of physical, chemical, and mechanical properties. Although
tuffs have not been identified previously in the United States as a candidate
geologic medium for waste repositories, they are currently being evaluated
by Sandia and Los Alamos as well as USGS. Japan has identified tuff as a
possible repository medium.
T\*o different forms of tuff are of interest for repository use.
The first is "densely welded" tuff, which can have high density, low
porosity and water content, and a high level of thermal stability. The
compressive strength, thermal conductivity and thermal expansion of densely
welded tuffs are comparable to those of basalt. The second is
zeolitized tuffs which have low density, high porosity, a high water
content and extremely high sorptive properties for radionuclides.
Zeolitized tuffs have moderate compressive strength, and a moderate
-------
DRAFT
71
thermal conductivity.
Dehydration of some zeolites begins about 100 degrees C. It has been
proposed that this fact coupled with the low thermal conductivity
compromise the usefulness of zeolite tuff for a repository. Zeolites heated
at temperatures in excess of a few hundred degrees C could release
fluids which could lead to hydrofracture of the medium or decomposition to
•
new materials with less sorptive capacity. Due to these effects, use
of zeolitized tuffs will impose constraints on thermal loading
of a repository.
Whereas zeolitized tuffs are limited by thermally induced changes,
densely welded tuffs are less limited by such constraints. An important
geologic feature which makes tuffs of great interest is that some large
deposits of welded tuff are underlain by zeolitized tuffs. Preliminary
data (58) show that some welded tuffs can be held at 500 degrees
C without significant change in their mineralogy. Consequently,
it appears that it might be possible to place radioactive waste, without
restrictive limit on temperature, in the welded tuff with its high thermal
stability and maintain the significant benefit from the added protection
of a highly sorptive barrier underlying the repository. A two year work plan
has been proposed at the Nevada Test Site to assess the viability of tuffs
as a geologic medium suitable for a repository and to identify suitable
areas.
Welded and zeolitized tuffs are widespread and occur in thick
sections in the Western United States though they have not yet been
DRAFT
-------
DRAFT
72
sufficiently characterized as to their homogeneity and their hydrologic
characteristics. Most tuffs in the Western United States are relatively
young geologically, and yet they have been broken into blocks tens of
kilometers in size by tectonic forces that were active during and after the
time of their eruption. The hydrogeologic environments in which tuffs
occur are dominated by topographic features that reflect the tectonic
activity. However, a single hydrological system in the Western U.S. can be
large enough to include many such blocks of zeolitized tuffs tens of
kilometers in size as in the case of the Nevada Test Site. Potentially
acceptable sequences of welded and zeolitized tuffs at NTS are located
beneath Yucca Mountain and the Timber Mountain caldera. Specific exploration (27)
is underway to confirm this evaluation.
A potential difficulty with welded tuff is that it may have
significant fracture permeability. For this reason exploration of tuffs
is in part directed toward identifying occurrences of welded tuffs situated
above the water table (58).
BRAFT
-------
DRAFT
73
7G. Unsaturated Rocks
Water tables 200-600 meters deep occur beneath certain valleys
and mesas of the Southwestern United States. Alluvium, tuff, sandstone,
basalt, and many other rock types comprise the strata in the thick
unsaturated zones (that is, the volume of rock between the land surface
and the water table) beneath these valleys and mesas. These unsaturated
zones have been proposed as potential repository sites (59-61).
The major assets of this concept are:
(a) The radioactive wastes will be placed several hundred
meters above, rather than hundreds of meters or more below
the water table;
(b) Under present climatic conditions transport of radionuclides
down to the water table is negligible (due to the near-zero
mean annual ground-water recharge);
(c) Alluvium, tuff, and sandstone—rocks commonly found in
unsaturated zones—have moderate to high sorptive capacity
for radionuclides;
(d) selected thick unsaturated zones occur above regional groundwater
basins with flowpaths on the order of 50-100
kilometers; such long ground-water flowpaths provide an
major barrier to radionuclide transport to the biosphere;
(e) thick unsaturated zones occur within large Federal tracts
of land that have been closed to the public for decades
and which contain few boreholes or abandoned shafts; and
(f) relative ease of repository construction and waste retrieval.
DRA-T
-------
DRAFT
74
Major liabilities of the concept include:
(a) Surface cooling of the wastes for several decades will be required
prior to emplacement in certain media. The rocks
comprising unsaturated zones have thermal conductivities one-half to
1/30 that of salt; hence the thermal loading for a repository
in alluvium, for example, would have to be held to below
100 degrees C. On the other hand, TRU wastes might be emplaced
in alluvium with no requirement for cooling;
(b) In the event of a return of a wetter climate to the Southwest
(such as existed there prior to 10,000 years ago), groundwater
flow vertically through unsaturated zones would undoubtedly
increase somewhat, although inundation of the wastes by a rising
water table is unlikely;
(c) the relatively shallow (150-300 meter) burial of the wastes
might invite penetration by future man; on the other hand,
the very deep water tables would discourage such penetration
say for irrigation waters.
Considerable research, perhaps taking as long as a decade, is
required to quantitatively evaluate this concept. This concept relies on
the hydrologic environment, rather than on a rock type, to provide the
major barrier to radionuclide transport.
DRAFT
-------
DRAFT
75
8. REFERENCES
1. Alexander, C.W., Kee, C.W., Croff, A.G. and Blcmeke, J.0.f "Projections
of Spent Fuel to be Discharged by the U.S. Nuclear Power Industry",
ORNL-TM-6008 (October 1977).
2. U.S. Energy Research and Development Administration, "Alternatives
for Managing Wastes from Reactors and Post-Fission Operations in
the LWR Fuel Cycle", ERDA-76-43, five volumes (May, 1976).
3. U.S. Department of Energy, Directorate of Energy Research, "Report
of Task Force for Review of Nuclear Waste Management, DRAFT,
DOE/ER-0004/D (February, 1978).
4. U.S. Nuclear Regulatory Commission, "Alternative Processes for
Managing Existing Commercial High-Level Radioactive Wastes", NUREG-
0043 (April, 1976).
5. Blomeke, J.O., and Kee, C.W., "Projections of wastes to be generated",
in Proceedings of the International Symposium on the Management
of Wastes from the LWR Fuel Cycle, Denver, Colo. July 11-16, 1976:
U.S. Energy Research and Development Adm. Rept. CONF-76-0701,
p. 96-117 (1976).
6. Blomeke, J.O., Nichols, J.P., McClain, W.C. "Managing Radioactive
Wastes", Physics Today, Vol. 26, No. 8, pp. 36-42 (August 1973).
7. Gera, Ferruccio, "Geochemical behavior of long-lived radioactive
wastes;" U.S.Energy Research and Devel. Adm., Oak Ridge Natl. Lab.
(Rept.) ORNL-TM-4481, 99 pp. (1975).
8. Hamstra, J., Radioactive hazard measure for buried solid radioactive
waste: Nuclear Safety, v. 16, no. 2, p. 180-189 (1975).
9. Winograd, I.J., "Geologic Containment of High-Level Radioactive
Wastes — What Time Frames are Pertinent and Predictable?" Geological
Society of America, Abstracts with Programs, V.9, No. 7, p. 1231 (1977)
10. Bredehoeft, J.S., et.al., "Geologic Disposal of High-Level Radioactive
Wastes — Earth-Science Perspectives", U.S. Department of the Interior,
Geological Survey Circular 779 (1978).
11. Kisner, R.A., et.al., "Nuclear Waste Projections and Source-Term Data
for FY 1977," Y/CWI/TM-34 (April 26, 1978).
12. Blomeke, J.O., "1972 Preliminary Safety Analysis Report Based on a
Conceptual Design of a Proposed Repository in Kansas", ORNL/TM-5764
(August, 1977).
DRAF
-------
DRAF
76
13. Bishop, W.P., et.al., "Proposed Goals for Radioactive Waste
Management", U.S. Nuclear Regulatory Commission, NUREG-0300
(May, 1978).
14. Arthur D. Little, Inc., "Technical Support for Radiation Standards
for High-Level Radioactive Waste Management", DRAFT REPORTS,
Contract 68-01-4470, Office of Radiation Programs, U.S. Environ-
mental Protection Agency (1978).
15. Zerby, C.D., "National Waste Terminal Storage Program Progress
Report", Y/OWI-9 (April, 1978).
16. Hansen, P.O. and Hellegard, K.D.: "Creep Behavior of Bedded Salt
from Southeastern New Mexico at Elevated Temperature", Technical
Memorandum Report RSI-0062, Prepared for Sandia Laboratories Under
Contract Document Nos. 02-8858 and 05-7466, (November 9, 1977),
33 pp.
17. Callahan, G.D. and Gnirk, P.F.: "Analytical Approximations to the
Thermoelastic Behavior of Repository Configurations", Technical
Memorandum Report RSI-0064, Prepared for the Office of Waste Isolation
Under Subcontract No. 89Y-22303C with Union Carbide Corporation,
Nuclear Division (January 6, 1978), 29 pp.
18. Bradshaw, R.L., and McClain, W.C., (Eds.), "Project Salt Vault: A
Demonstration of the Disposal of High-Activity Solidified Wastes
in Underground Salt Mines", Oak Ridge National Laboratory, ORNL-4555
(April, 1971).
19. Ratigan, J.L. "Groundwater Movements Around a Repository, Phase
2, Rock Mechanics Analysis", KBS Report No. 54:04, Stockholm,
Sweden, (September, 1977).
20. Mantab, M.A. and Ratigan, J.L.: "Topical Report II - Room Stability",
Prepared for the Atomic Energy of Canada Limited (July 1977).
21. Mantab, M.A. McCreath, P.R. "Topical Report 04-Rock Mechanics",
Prepared for the Atomic Energy of Canada Limited, (July 1976).
22. Ratigan, J.L. Chartlwood, R.G., and Skiba, E.: "Topical Report
02-Thermal Loading", Prepared for the Atomic Energy of Canada
Limited (April 1976).
23. Osnes, J.D., Wagner, R.A., and Waltman, H.: "Parametric Thermal-
electric Analysis of High-Level Waste and Spent Fuel Repositories
in Granite and other Non-Salt Rock Types", Technical Memorandum
Report RSI-0066, Prepared for the Office of Waste Isolation (Cur-
rently under Review).
24. St.John, C.M., Computer models and the design of underground radio-
active waste repositories: M.I.T. Summer Session "Current Develop-
ments in Rock Engineering" (1978).
01AFT
-------
DRAFT
77
25. Panel on Rock Mechanics Problems that Limit Energy Resource Recov-
ery and Development, U.S. National Committee for Rock Mechanics,
National Research Council, Limitations of Rock Mechanics in Energy
Resource Recovery and Development, National Academy of Sciences
(1978).
26. Dietz, H.B., Rockwell International, "Waste Isolation Program Presenta-
tion to the Office of Science and Technology Policy," (June 14, 1978).
27. Personal communication, Richard Lynch, Sandia Laboratories
(June 14, 1978).
28. Proceedings from the Symposium on the Management of Radioactive
Wastes from Fuel Reprocessing, OCED/NEA, Paris (1972).
29. McCarthy, G.J., et.al., "Interactions between Nuclear Waste and Sur-
rounding Rock", NATURE, Vol. 273, pp. 216-217 (May 18, 1978).
30. McCarthy, G.J., Nuclear Technology 32, 92-105 (1977).
31. Russin, J.M., et.al., Battelle Pacific Northwest Laboratory,
"Multibarrier Waste Forms, Part I: Development", PNL 2668-1
(to be published).
32. Russin, J.M., et.al., Battelle Pacific Northwest Laboratory,
"Multibarrier Waste Forms, Part II: Characterization and
Evaluation", PNL-266B-1 (to be published).
33. Aaron, W.S., Quinby, T.C., and Kobisk, E.H., "Cermet High Level
Waste Forms", ORNI/TM-6404 (June 1978).
34. Personal Communication: Richard Westerman, Battelle Pacific
Northwest Laboratory (June 1978).
35. Jenks, G.H., "NWTS Conference on Waste-Rock Interactions",
Y/CWI/SUB 77/14268 (July 1977).
36. Ad Hoc Panel of Earth Scientists, The State of Geological Know-
ledge Regarding Potential Transport of High-Level Waste from
Deep Continental Repositories, U.S. Environmental Protection
Agency, EPA/520/4-78-004, (June 1978).
37. Gnirk, P.F., Pariseau, W.G., Russell, J.E., Wawersik, W.R.,
Callahan, G.D., and Hovland, N.: "Analysis and Evaluation of the
Rock Mechanics Aspects of the Proposed Salt Mine Repository",
Summary Progress Report RSI-0005, Prepared for the Oak Ridge National
Laboratory under Subcontract No. 3706 with Union Carbide Corpora-
tion, Nuclear Division (September 28, 1973), 153 pp.
38. Gnirk, P.P., Callahan, G.D., Pariseau, W.G., Van Sambeck, L.L.,
and Wawersik, W.R.: "Analysis and Evaluation of the Rock Mechanics
Aspects of the Proposed Salt-Mine Repository Concept-Ill", Summary
Progress Report RSI-0012, Prepared for Oak Ridge National Labor-
atory under Subcontract No. 3706 with Union Carbide Corp., Nuclear
-------
FT
78
Division (September 28, 1974), 62 pp.
39. Burkholder, H.C.f Stottlemyre, J.A., and Raymond, J.R.:
"Safety assessment and geosphere transport methodology for the
geologic isolation of nuclear waste minerals" in Risk analysis
and geologic modeling in relation to the disposal of radioactive
wastes into geological formations — (Proceedings of workshop
organized jointly by the OECD Nuclear Energy Agency and the
Commission of European Communities, Ispra, Italy, May 23-27, 1977)
pp. 216-229 (1977).
40. Marsily, G. de, Ledoux, E., Barbreau, A., and Margat, J.,
"Nuclear waste disposal: Can the geologist guarantee isolation?" '
Science Magazine, vol. 197, no. 4303, p. 519-527 (1977).
41. Pines, David (Ed.), "Report to the American Physical Society
by the Study Group on Nuclear Fuel Cycles and Waste Management,"
Reviews of Modern Physics, Vol. 50, No. 1, Part II, pp. 5121-
5128 (January 1978).
42. Mercer, J., and Faust C., USGS, oral communication, 1977 Annual
Geological Society of America Meeting, Seattle, WA (Nov. 8, 1977).
43. Arthur D. Little, Inc., Technical Support for Radiation
Standards for High-Level Waste Management-Assessment of
.Geological Site Selection Factors, Draft Report (1977).
44. Office of Nuclear Material Safeguards and Safety, U.S. Nuclear
Regulatory Commission, "Workshop Material for State Review
of USNRC Site Suitability Criteria for High-Level Radioactive
Waste Repositories," NUREG-0326 (September, 1977).
45. Pines, David (Ed.), "Report to the American Physical Society by
the Study Group on Nuclear Fuel Cycles and Waste Management,"
Reviews of Modern Physics, Vol. 50, No. 1, Part II, pp. 5120-
5128 (January 1978)
46. Brunton, G.D., and McClain, W.C., "Geological Criteria for
Radioactive Waste Repositories," Y/OWI/TM-47 (November 28, 1977).
47. International Atomic Energy Agency, Site Selection Factors
for Repositories of Solid High-Level and Alpha-Bearing
Wastes in Geological Formations, Technical Reports Series No. 177
(1977).
48. National Academy of Sciences, Division of Earth Sciences, Committee
on Waste Disposal, "the Disposal of Radioactive Wastes on Land",
NAS-NRC Pub. 519 (1957).
49. Sandia Laboratories, "Waste Isolation Pilot Plant": Program Plan
for FY 1978", Working Draft Report (April 1978).
-------
79
50. Kuhn, Klaus, et.al., "Geologic Isolation of Radioactive Waste in
the Federal Republic of Germany and the Respective Programs of the
Netherlands", Proceedings of the International Conference on
the Management of Wastes from the LWR Fuel Cycle, July 11-16,
Denver, Colorado, pp. 580-600, CONF 76-07-01 (1976).
51. Anthony, T.R. and Cline, H.E., "Thermomigration of liquid droplets
in salt", 4th Symposium on Salt, 1, 313-321. A.H. Cooper, Ed.,
Cleveland, Ohio, Northern Ohio Geological Society (1974).
52. Roedder, E. and Belkin, H.E. (in press), "Application of studies
of fluid inclusions in Permian Salado Salt, N.M.", Problems of
siting a nuclear waste repository, Symposium. Material Research
Society, Nov., 1978.
53. Magnus Bergman (Ed.), Storage in Excavated Rock Caverns, Bookstore
77, Proceedings of the First International Symposium, Stockholm
(5-8 Sept 1977), Vol 2, Pergamon Press (1977).
54. Hill, M.D. and Grimwood, P.O., United Kingdom Radiological Protection
Board, "Preliminary Assessment of the Radiological Protection Aspects
of Disposal of High-Level Waste in Geologic Formations", Report
NRPB-R-69 (Jan. 1978).
55. Tamura, T., "Sorption Phenomena Significant in Radioactive Waste
Disposal" AAPG Memo 18, Underground Waste Management and Environ-
mental Impact Implications, pp. 318-330 (1972).
56. Dosch, R.G., Lynch, A.W., "Interaction of Radionuclides with
Argillite from the Eleana Formation on th Nevada Test Site", Sand
78-0893 (In Preparation).
57. Waltman, D., "Minability Study of the Terminal Waste Storage Pro-
gram Eleana Investigation", Department of Energy Report
NVO-0038-35 (April 1978).
58. Tyler, L. et.al., "Physical, Mechanical, and Chemical Properties
of Tuffs from the Nevada Test Site" (In Preparation).
59. Winograd, I.J., Radioactive waste storage in the arid zone:
EOS (American Geophysical Union Trans.), vol. 55, no. 10,
pp. 84-94, (1974).
60. Angino, E.E., "High-Level and Long Lived Radioactive Waste
Disposal", Science, Vol. 198, pp. 885-890 (1977).
61. Panel on Hanford Wastes, Committee on Radioactive Waste Management,
Radioactive Wastes at the Hanford Reservation; A Technical Review,
pp. 108-110, 115-118, National Academy of Sciences (1978).
62. International Atomic Energy Agency, Transurnaium Nuclides in
the Environment, Vienna (1976).
-StfT
------- |