United States
Environmental Protection
Agency
Off ice of
Radiation Programs
Washington DC 20460
EPA 520/4-79-007E
March 1982 •
Radiation
&EPA
Technical Support of
Standards for High-Level
Radioactive Waste
Management
Addendum to
Volumes C and D
-------
TECHNICAL SUPPORT OF STANDARDS FOR
HIGH-LEVEL RADIOACTIVE WASTE MANAGEMENT
ADDENDUM TO VOLUMES C AND D
EPA Contract No. 68-01-4470
Prepared by
Arthur D. Little, Inc.
Cambridge, Massachusetts 02140
March 1982
-------
DISCLAIMER
This report was prepared as an account of work sponsored by the
Environmental Protection Agency of the United States Government
under Contract No. 68-01-4470. Neither the United States nor the
United States Environmental Protection Agency makes any warranty,
express or implied, or assumes any legal liability or responsibility
for the accuracy, completeness, or usefulness of any information,
apparatus, product, or process disclosed, or represents that its use
would not infringe privately owned rights.
11
-------
ACKNOWLEDGMENTS
Many individuals contributed to the work done under the
direction of Arthur D. Little, Inc. for the U.S. Environmental
Protection Agency under Contract No. 68-01-4470. John L. Russell
and Daniel Egan of the Office of Radiation Programs at EPA served as
constant guides in the process of our work. Dr. Bruce S. Old, James
I. Stevens, and David I. Hellstrom of Arthur D. Little, Inc., were
Program Director, Program Manager, and Assistant Program Manager,
respectively, of the overall project. Key individuals involved in
each of the reports prepared under the four tasks were:
TASK A Donald Korn
Arthur D. Little, Inc.
Task Director
Robert McWhorter,
Michael Raudenbush,
and Lester Goldstein
S. M. Stoller Corp.
TASK B Edwin L. Field
Arthur D. Little, Inc.
Task Director
Robert McWhorter and
Michael Raudenbush
S. M. Stoller Corp.
TASK C Dr. P. J. O'Brien
Arthur D. Little, Inc.
Task Director
Dr. Ronald B. Lantz
Intera Environmental
Consultants, Inc.
Dr. John Gormley
D'Appolonia Consulting
Engineers, Inc.
TASK D Dr. Charles R. Hadlock
Arthur D. Little, Inc.
Task Director
Peter D. Mattison and
Dr. Ajit Bhattacharyya
Arthur D. Little, Inc.
111
-------
FOREWORD
A major Federal effort is underway to develop methods for
disposal of high-level radioactive waste in deep geologic
repositories. An important element of this program is the
development and promulgation by the U.S. Environmental Protection
Agency (EPA) of environmental standards for the management of these
wastes.
In anticipation of its efforts to develop these standards, EPA
recognized that it would be necessary to estimate the expected and
potential environmental impacts from potential geologic repositories
using modeling techniques based upon as thorough an understanding as
possible of the uncertainties involved in the quantities and
characteristics of the wastes to be managed, the effectiveness of
engineering controls, and the potential migration and accidental
pathways that might result in radioactive materials entering the
biosphere. Consequently, in March 1977, the EPA contracted with
Arthur D. Little, Inc., for a study to provide a technical support
for its development of environmental regulations for high-level
radioactive wastes. This study was divided into the following four
tasks:
Task A - Source Term Characterization/Definition
Task B - Effectiveness of Engineering Controls
Task C - Assessment of Migration Pathways
Task D - Assessment of Release Mechanisms
Since the completion of these reports several years ago,
research by many organizations has been proceeding at a very rapid
rate, and thus it is of interest to determine to what extent the
contents or conclusions should be updated. In particular, the
purpose of this report is to examine several issues relevant either
to the conclusions of Tasks C and D or to their subsequent use by
EPA and others, and to identify additions or modifications that may
be dictated by such additional data and analysis. The ten major
sections are essentially independent of each other, but are often
dependent on further background information from the Task C and D
reports.
v
-------
TABLE OF CONTENTS
Page
1. Frequency of Fault Movement 1
1.1 Paradox Basin 1
1.2 Permian Basin 3
1.3 Gulf Coast Region 3
1.4 Summary 4
2. Effects of Salt Dissolution on Groundwater Flow 7
3. Future Drilling . . . 19
4. Deep Dissolution Features 22
4.1 Introduction 22
4.2 Breccia Pipes 23
4.3 Breccia Pipe Distribution , 25
4.4 Brine Reservoirs 27
4.5 Brine Pocket Distribution 28
4.6 Detection Methods 29
5. Solution Mining 33
5.1 Introduction 33
5.2 Bedded Salt 33
5.3 Dome Salt 34
6. Thermal Buoyancy Model 44
7. Hydraulic Conductivity of Fault Zones 54
Vll
-------
TABLE OF CONTENTS (continued)
Page
8. Hydrologic Parameters for Host Rocks 57
8.1 Review of Reference Parameters 57
8.2 Assumption of Negligible Permeability for Salt 57
8.3 Vertical Flows in the Presence of Interbeds 63
for Cases of Basalt and Shale
9. Fracture Flow 71
10. Geochemical Retardation 77
10.1 Field Measurements of Retardation Factors 79
10.2 Calculated Retardation Factors Based on 79
Distribution Coefficients
Vlll
-------
- LIST OF TABLES
Table No. Page
1-1 Values for Fault Movement Frequency Used 2
in Task D Report Analyses
1-2 Modified Fault Movement Frequencies for Salt 5
2-1 Effective Vertical Hydraulic Gradient in 8
Permeable Boreholes (Bedded Salt Repository)
2-2 Effective Vertical Hydraulic Gradient in 9
a Fault (Salt Dome Repository)
2-3 Head Differences Between the Underlying and 12
Overlying Aquifers Corresponding to
Hydraulic Gradients Listed in Table 2-1
(Bedded Salt Repository)
2-4 Head Differences Between the Underlying and 13
Overlying Aquifers Corresponding to
Hydraulic Gradients Listed in Table 2-2
(Salt Dome Repository)
2-5 Water Level Differences Between Hypothetical- 16
Wells in the Underlying and Overlying
Aquifers Considering the Effects of Salt
Dissolution on Water Density (Bedded Salt
Repository)
2-6 Water Level Differences Between Hypothetical 17
Wells in the Underlying and Overlying
Aquifers Considering the Effects of Salt
Dissolution on Water Density (Salt Dome
Repository)
3-1 Average Drilling Rates: Comparison of Past, 20
Present, and Estimated Future Values (Holes
Per Square Kilometers Per Year)
5-1 Data on Cavities Shown in Figure 5-2 37
5-2 Results of Regional Characterization Screening 39
5-3 Resource Values of Potentially Suitable Salt Domes 40
6-1 Hydraulic Characteristics of the Generic 47
Granite Stratigraphic Section
-------
LIST OF TABLES (continued)
Table No. Page
6-2 Water Viscosity, Density, Hydraulic Gradient and 52
Minimum Travel Time for Selected Representative
Temperatures
7-1 Hydraulic Conductivity of Faulted Pathways 55
Used in the Task D Report
8-1 Reported Ranges or Values of Porosity 62
of Potential Host Rocks
8-2 Vertical Volumetric Groundwater Flows 64
Through Portion of Salt Bed Containing
Repository, as a Function of Hydraulic
Gradient and Hydraulic Conductivity
8-^3 Equivalent Composite Hydraulic Conductivity 68
of Flow Path Consisting of Higher and
Lower Permeability Zones (cm/sec)
9-1 Effect of Aperture Size on Travel Time 75
10-1 Retardation Factors Based on In-situ 81
Measurements of Radionuclide Migration in
Deep Groundwater
10-2 Representative Retardation Factors 83
-------
LIST OF FIGURES
Page
2-1 Repository in Bedded Salt 10
2-2 Repository in Salt Dome 11
5-1 Potential Demands for Salt Dome Utilization 35
5-2 Cavity Development in a Gulf Coast Salt Dome 36
6-1 Information Flow Chart for Buoyancy Computations 45
6-2 Generic Granite Stratigraphic Section in 46
its Regional Setting
6-3 Temperature Distribution Profile for Generic 50
Granite Repository at a Thermal Loading Rate
of 100 kW/Acre
6-4 Temperature Distribution Profile for Generic 51
Granite Repository with a Thermal Loading
Rate of 200 kW/Acre
8-1 Reported Ranges or Values of Hydraulic 58
Conductivity of Salt (cm/sec)
8-2 Reported Ranges or Values of Hydraulic 59
Conductivity of Granite (cm/sec)
8-3 Reported Ranges or Values of Hydraulic 60
Conductivity of Basalt (cm/sec)
8-4 Reported Ranges or Values of Hydraulic 61
Conductivity of Shale (cm/sec)
8-5 Generic Repository in Basalt 65
10-1 Relationship Between Retardation Factor, R,, 78
and Distribution Coefficient, K,
d
10-2 Illustration of Calculation of R From 80
In-situ Measurement of Travel Times
Between Two Wells
10-3 Ranges of Distribution Coefficients for 82
Various Rock Types
-------
1. FREQUENCY OF FAULT MOVEMENT
The values for frequency of fault movement used in the analyses
presented in the Task D Report were based on an estimate of the time
since the most recent movement along existing faults. The equations
used to calculate first-estimate fault frequency and second-estimate
fault frequency were X = 4/N and X = 10/N, respectively,
where X is the fault failure rate (events/year), N is the time since
the most recent fault movement (years), and the constants are based
on estimates of fault frequency within the area of the repository.
The fault failure rates used in the Task D Report analyses are given
in Table 1-1.
The purpose of this section is to review the previous estimates
in light of more recent data from certain regions with
characteristics similar to those assumed for the generic
repositories in the Task D Report. Only salt formations are
considered, since these are the formations for which new data have
appeared. This review lends support to the first-estimate values
used in the Task D. Report, but it suggests that second-estimate
values should be higher. It should be emphasized that even though
real data are used for these calculations, the resulting numbers
should not be interpreted as applying to the specific sites. It is
outside the scope of this study to estimate fault movement
probabilities for specific regions or sites. Such estimates would
be expected to be based on more extensive data and a detailed
analysis of geologic processes at the site.
1.1 PARADOX BASIN
The 1982 geologic characterization report (Woodward-Clyde,
1982) for the Gibson Dome area of the Paradox Basin indicates that
major faulting occurred during Late-Cretaceous and Cenozoic time
(approximately 65 to 10 million years before present [MYBP]) as a
result of regional tectonic forces. There is also scattered
-------
TABLE 1-1
VALUES FOR FAULT MOVEMENT FREQUENCY
USED IN TASK D REPORT ANALYSES
First Estimate Second Estimate
(events/year) (events/year)
Bedded Salt 2 x 10"8 4 x 10~7
Granite 2 x 10"8 10"5
Basalt 5 x 10~7 10~5
Shale 2 x 10~8 A x 10"7
Dome Salt 3 x 10"7 " 10~5
-------
evidence of Quaternary faulting (less than 2 MYBP), but the general
lack of Quaternary cover makes it difficult to assess and date the
most recent episode of faulting. Using both 65 and 10 MYBP in
first-estimate calculations yields a fault movement frequency of 4 x
—7 —8
10 to 6 x 10 events/year. Using 1 MYBP as a conservative value
i.e., tending to overestimate the risk, for second-estimate
calculations yields a fault movement frequency of 10 events per
year. ...
1.2- PERMIAN BASIN
The area geologic characterization report for the Palo Duro and
Dalhart Basins of Texas (Stone and Webster, 1981) contains a summary
discussion and a table of fault data. Most of the faulting has been
described as Late Mississippian to Early Permian in age and does not
affect the overlying salt strata. Two faults, however, with as much
as 600 feet displacement, are reported to intersect the upper
Permian salt strata. Relative age dating on the basis of
stratigraphic relationships indicates that the most recent movement
along these faults occurred during Late Permian (approximately 250
MYBP) and Triassic (approximately 225 MYBP). Using these data would
—8
result in fault movement frequencies of 1.6 to 1.8 x 10
events/year.
1.3 GULF COAST REGION
The geologic characterization reports for the Mississippi and
Louisiana study areas of the Gulf region (Law Engineering, 1980)
identify several faults that were active during Late Jurassic
(approximately 140 MYBP) and Late Cretaceous (approximately 65 MYBP)
time. The report for the Mississippi study area also states that
interpretation of seismic data indicates that there has been no
post-Eocene (approximately 40 MYBP) fault movement that could affect
the proposed host rock. Investigations are being undertaken,
however, to evaluate what is believed to be relatively shallow
faulting that has occurred in the Quaternary (less than 2 MYBP).
Using the Late Jurassic to Eocene dates for first-estimate
-------
calculations and 1 MYBP for second-estimate calculations yields
fault movement frequencies of 1.0 to 2.8 x 10 events/year to 1 x
10 events/year, respectively.
1.4 SUMMARY
On the basis of the estimates given above, there is no reason
to revise the first-estimate values for faulting from those given in
Task D. It does appear, however, that the second-restimate values
should be increased to 10 events/year. The modified values are
given in Table 1-2. Second-estimate comparisons among media could
be misleading if based on the new salt values and other values from
Table l-il, since more detailed study of potential sites with other
host rock formations might similarly lead to higher values.
-------
TABLE 1-2
MODIFIED FAULT MOVEMENT FREQUENCIES FOR SALT
First Estimate Second Estimate
(events/year) (events/year)
Bedded Salt 2 x 10~8 10"5
Dome Salt 3 x 10~7 10~5
Note: Only the second estimates have been modified from Task D
values.
-------
References
Law Engineering Testing Company. Geologic Area .Characterization.
Volume 1: Introduction, Background and Summary. Gulf Coast Salt
Domes Project. Prepared for Battelle Memorial Institute. August
29, 1980.
Stone and Webster Engineering Corporation. Area Geological
Characterization Report for the Palo Duro and Dalhart Basins, Texas.
Prepared for the Office of Nuclear Waste Isolation. December 1981.
ONWI-292.
Woodward-Clyde Consultants. Geologic Characterization Report for
the Paradox Basin Study Region Utah Study Areas. Prepared for
Battelle Memorial Institute Office of Nuclear Waste Isolation.
January 1982. ONWI 290.
-------
2. EFFECTS OF SALT DISSOLUTION ON GROUNDWATER FLOW
For the evaluation of generic salt repositories, the analyses
of groundwater flow and travel times contained in the Task D Report
assumed that there were no changes in water density caused by salt
dissolution. The initial analyses for the reference bedded salt and
salt dome repositories are reevaluated in this section under the
assumption that a breach through the repository that intersects
overlying and underlying fresh-water aquifers would allow for salt
dissolution that, in turn, would increase the water density. This
density increase can either reduce the upward water flow rate or
even reverse the direction of the flow. Sample calculations are
given to show this effect. Naturally, the importance of this
phenomenon depends in part on the initial degree of salinity of the
water that could flow through a pathway. Many deep groundwaters in
salt regions are already quite saline, thereby diminishing the
effect.
Tables 2-1 and 2-2, which are taken from the Task D Report,
summarize the effective hydraulic gradients used to analyze flow
rates in boreholes and faults in bedded salt and salt dome
repositories, respectively. (These cases are chosen as examples.)
Figures 2-1 and 2-2 are schematics of these reference repositories.
The gradients listed in these tables are equivalent to the head
difference between the overlying and underlying aquifers divided by
the vertical distance between the two aquifers. For the reference
bedded salt and salt dome repositories, the aquifers are separated
by a vertical distance of 200 meters and 330 meters, respectively.
The Task D analyses were based on the conservative assumption that
the head in the lower aquifer was greater than the head in the upper
aquifer and that groundwater flow was from the lower to the upper
aquifer. Tables 2-3 and 2-4 list the relative head differences that
would exist between the two aquifers based on gradients listed in
Tables 2-1 and 2-2.
-------
TABLE 2-1
EFFECTIVE VERTICAL HYDRAULIC GRADIENT IN PERMEABLE BOREHOLES
(BEDDED SALT REPOSITORY)
Hydraulic Gradient, 1
100 Years 1000 Years 10,000 Years
First Estimate 0.13 0.11 0.04
Second Estimate 0.62 0.60 0.53
*
Years after repository closure.
Source: Table D-53, Task D Report.
-------
TABLE 2-2
EFFECTIVE VERTICAL HYDRAULIC GRADIENT IN A FAULT
(SALT DOME REPOSITORY)
100 Years* 1000 Years 10,000 Years
First Estimate 0.13 0.08 0.04
Second Estimate 0.32 0.27 0.23
*
Years after repository closure.
Source: Table D-97, Task D Report,
-------
•Surface -
Deposits''
Surface
jir^-_- Shale r
Salt
Repository
330 Meters
. 360 Meters
410 Meters
460 Meters
510 Meters
560 Meters
590 Meters
Source: Task D Report, Figure D-2.
FIGURE 2-1 REPOSITORY IN BEDDED SALT
10
-------
Surface
Surface
Deposits
a Aquifer
Aquifer
200 Meters
230 Meters
460 Meters
560 Meters
590 Meters
Source: Task D Report, Figure 8.
FIGURE 2-2 REPOSITORY IN SALT DOME
11
-------
TABLE 2-3
HEAD DIFFERENCES BETWEEN THE UNDERLYING AND OVERLYING AQUIFERS
CORRESPONDING TO HYDRAULIC GRADIENTS LISTED IN TABLE 2-1
(BEDDED SALT REPOSITORY)
Water Level Difference, Meters
1000 Years 10,000 Years
First Estimate 22 8
Second Estimate 120 106
*
Note: Freshwater equivalent difference
12
-------
TABLE 2-4
HEAD DIFFERENCES BETWEEN THE UNDERLYING AND OVERLYING
AQUIFERS CORRESPONDING TO HYDRAULIC GRADIENTS LISTED IN TABLE 2-2
(SALT DOME REPOSITORY)
*
Water Level Difference, Meters
1000 Years 10.000 Years
First Estimate 26.4 13.2
Second Estimate 89.1 76
*
Note: Freshwater equivalent difference
13
-------
If a borehole or a fault were to provide a pathway from the
underlying aquifer to the overlying aquifer, then fresh water could
flow through and dissolve part of the salt formation. This process
would change the density of the water as it flowed to the overlying
aquifer.
To evaluate the effect of changes in water density caused by
salt dissolution, it is necessary to evaluate the effect of the
denser water on the hydraulic gradient between the underlying and
overlying aquifer. The density of waters associated with salt
3 3
deposits ranges between 1.027 g/cm and 1.345 g/cm with an average
3
density of 1.15 g/cm (Muller et al., 1981, Table 3). The present
3
analyses assume an average density of 1.15 g/cm .
The change in the height of a water column, owing to a change
in water density, can be calculated by the equation
hs pf
TT —
hf ps
where
h = Height of the column of saline (denser) water
s
h,. = Original height of the column of fresh water
p, = Density of the fresh water
p = Density of the saline water
S
For the present reference calculations, the correction factor for
the change in height of the water column is 0.87, i.e., 1.0/1.15.
This means that if a fresh water column extending from the lower
aquifer to some point above the upper aquifer and with a pressure p
at the bottom were replaced by a column of water also with pressure
p, at the bottom, but with a length L of saline water (p = 1.15)
contained within it, the height of the second column of water would
be 0.13 L (1.00 - 0.87) less than the height of the first column of
water. Using this factor results in a head correction value of 20
14
-------
meters for the bedded salt repository and A3 meters for the salt
dome repository.
Tables 2-5 and 2-6 list the revised water level differences
between the underlying and overlying aquifers attributable to the
effects of increased density of the water between the bottom of the
host salt unit and the overlying aquifer. For most of the
first-estimate calculations, the effect of the increased water
density is to reverse the gradient and to cause water to flow down
from the upper aquifer to the lower aquifer.
15
-------
TABLE 2-5
WATER LEVEL DIFFERENCES BETWEEN HYPOTHETICAL WELLS
IN THE UNDERLYING AND OVERLYING AQUIFERS CONSIDERING
THE EFFECTS OF SALT DISSOLUTION ON WATER DENSITY
(BEDDED SALT REPOSITORY)
Water Level Difference, Meters
1000 Years 10,000 Years
First Estimate 2 -12
Second Estimate 100 86
*
Minus sign (-) means water level in upper aquifer is higher than
in lower aquifer, implying a downward gradient.
16
-------
TABLE 2-6
WATER LEVEL DIFFERENCES BETWEEN HYPOTHETICAL WELLS
IN THE UNDERLYING AND OVERLYING AQUIFERS CONSIDERING
THE EFFECTS OF SALT DISSOLUTION ON WATER DENSITY
(SALT DOME REPOSITORY)
*
Water Level Differences, Meters
1000 Years 10,000 Years
First Estimate -17 -30
Second Estimate 46 33
Minus sign (-) means water level in the upper aquifer is higher
than in the lower aquifer, implying a downward gradient.
17
-------
Reference
Muller, A. B., N. C. Finley and F. J. Pearson, Jr. Geochemical
Parameters Used in the Bedded Salt Reference Repository Risk
Assessment Methodology. Sandia National Laboratories. Prepared for
the U.S. Nuclear Regulatory Commission. September 1981.
SAND-81-0557; NUREG/CR 1996.
13
-------
3. FUTURE DRILLING
This section adds two comments to the discussion of future
drilling scenarios in the Task D Report.
First, the discussion in the Task D Report was not based on a
statistical extrapolation from past or present data, but on
hypothetical scenarios for future activities at the site. The
reason for this approach was the importance of the change in
drilling rates with time. Nevertheless, it may be useful in lending
perspective to the suggested rate estimates to consider past,
present, and estimated future rates on a common scale. Such a
comparison is presented in Table 3-1. In comparing U.S. average
rates with those for sedimentary rocks, it might be desirable to
divide the U.S. average by the fractional area of the sedimentary
basins in which most of the drilling has taken place. Using an
estimate of 10% for this fraction, this adjustment could raise the
rate by one order of magnitude. Similarly, the average rate for
basalt and granite would be lowered slightly.
Second, the drilling rates given in the Task D models are based
on the assumption of loss of human knowledge and government control
of the site after 100 years. This conservative assumption was a
guideline for the study. If control of the site is lost but
knowledge of the hazard remains, then future drilling scenarios
might be somewhat different. Although all discussion of these
matters must be speculative, some estimates are appropriate since
this is an important concern in siting a repository. The present
authors' best estimate is the following: Knowledge of the presence
of the repository will do little to deter future drilling, whether
or not the government retains control of the site. In fact, if
there are signs of potential resources, it is likely that there will
be effective pressure to drill for resources on the site even before
the hundred-year control period ends. Drillers will pass through
the repository level with greater than average care, however, and
will monitor for hazardous material. Any associated risk to the
drillers becomes their own responsibility, or that of the government
19
-------
TABLE 3-1
AVERAGE DRILLING RATES: COMPARISON OF PAST, PRESENT, AND
ESTIMATED FUTURE VALUES (HOLES PER SQUARE KILOMETER PER YEAR)
Past Rate,
Rock Type First Estimate
Past Rate,
Second Estimate
Past Rate
U.S. Average,
Present Rate,
......., , , , , ,,» Future Rate,
All Formations^ ' All Formations (C) First Estimate^ ' Second Estimate
.U.S., Average, , , Future Rate,
Bedded
Salt
Granite
Basalt
Shale
Dome Salt
1.3 x
1.3 x
?..! x
1.3 x
8.3 x
1C'4
10~5
10-*
io-4
io-3
2.5
1.3
5.4
2.5
8.3
x
X
X
X
X
io-2
io-4
io-4
io-2
io-2
1.7 x
1.7 x
1.7 x
1.7 x
1.7 x
io-3
io-3
10"3
io-3
io-3
4.0 x
4.0 x
4.0 x
4.0 x
4.0 x
io-3
io-3
io-3
io-3
io-3
2.5 x
3.1 x
1.3 x
2.5 x
2.5 x
io-3
io-4
io-3
io-3
io-3
6.3
2.5
6.3
6.3
6.3
x
x
X
X
X
io-3
io-3
io-3
io-3
io-3
(a) Based on existing borehole density, as given in Task D Report, averaged over 30 years. The choice of 30
years is somewhat arbitrary, intended to cover period of heaviest drilling. First and second estimates are
as defined in the Task D Report.
(b) Based on 100 years, roughly age of petroleum industry in United States.
(c) Estimated present drilling rate, based on data for recent years.
(d) Based on Task D models, omitting period of above-average drilling rate.
-------
that permits the operation. Some long-term disruption of the
repository may also occur through the establishment of an
imperfectly sealed borehole pathway, as discussed in the Task D
Report. (For example, it may have a hydraulic conductivity of 10
cm/sec initially or after some degradation.)
21
-------
4. DEEP DISSOLUTION FEATURES
4.1 INTRODUCTION
Dissolution features have been noted in many salt deposits that
are under investigation as potential sites for the disposal of
radioactive waste. The possible vulnerability of a repository to
dissolution of the protective salt requires careful evaluation of
the origin, prevalence, and nature of such features. This section
augments the information provided in the Task D Report, and while it
does not lead to any revision in the numerical estimates given
there, it does highlight the large uncertainties and degree of
variability.
"Dissolution feature" describes structures and petrologic
changes that result from dissolution of salt and other evaporites,
or of less soluble rocks, such as carbonates. In fact, some of these
"dissolution features" may not be wholly caused by dissolution, but
since dissolution plays an important role in their development,
finer distinctions are not useful.
Localized salt dissolution is distinct from 'regional
dissolution, the latter appearing as the progressive and calculable
thinning of evaporite deposits at their margins. Localized
dissolution is not necessarily accompanied by general salt loss over
a large area, but is more erratic, less predictable, and often less
easily discerned.
"Deep," as opposed to "shallow" or "surficial," denotes
features extending to, or originating near, the bottom of a salt
deposit. It is a relative term in that no specific depth is implied.
Although many dissolution features are probably found in most
evaporite regions, two types of localized features—breccia pipes
and brine pockets— are of sufficient scale and complexity to be of
particular concern for a repository. While there is some
controversy over the proper names for these features, "breccia pipe"
and "brine reservoir" are used here to denote, respectively, a
columnar structure filled with fragmented rock and a void or
23
-------
permeable zone in a rock formation that yields brine (usually under
pressure) when tapped.
4.2 BRECCIA PIPES
Breccia pipes are the more often reported features, since they
are often visible at the surface as minor hills or depressions with
a core of brecciated rock usually surrounded by upwarped country
(surrounding) rock. Whether a hill or depression remains appears to
depend upon whether dissolution has lowered the level of the sur-
rounding rock, leaving the insoluble core of the pipe behind. These
structures vary somewhat in shape, but most are reported to be
roughly circular and from a few meters to hundreds of meters in
diameter. In many cases they extend from the bottom of a salt or
other soluble stratum to the surface. They are considered collapse
structures because the material filling the chimney is brecciated
rock typically from overlying formations.
The size of the breccia clasts and their displacement from
their original stratigraphic position vary widely, as do their
physical properties. Some breccias are reported to be completely
cemented with carbonate or evaporite minerals; others are
essentially loose and uncemented. It is difficult or impossible to
determine the time required for cementation for particular cases.
The processes by which breccia pipes are formed in salt are not
completely understood. Three theories are currently prevalent, with
all three assuming the existence of some communication between the
evaporite deposit and an underlying aquifer. Two theories focus on
dissolution and removal of salt to provide void space into which
overlying rock may settle. Stanton (1966) hypothesizes that water in
contact with salt dissolves and softens the evaporite matrix,
permitting the less soluble fractions to detach and float free into
what amounts to a slurry of unbound insoluble fragments. He
concludes that the void space produced by this process is never very
large and that, when overlying rock collapses into the void, the
distance of fall is quite small. If the evaporites are under
24
-------
sufficient stress to flow plastically, evaporite flow would close a
cavity as rapidly as it is formed, and brecciation would occur only
in overlying strata.
Another theory, based to some extent on laboratory experiments,
involves the dissolution of a relatively large cavern through the
connective circulation of water under artesian pressure (Anderson,
1978). Under the proper circumstances, fresh (unsaturated) water
from a deep aquifer contacts the bottom of a salt deposit through
fractures or a permeable zone in an underlying aquiclude. The water
becomes saturated with the overlying salt and the dense brine thus
formed sinks through the fresh water, bringing more fresh water in
contact with the salt. In this way, a convective cell is established
and dissolution propagates upward at a rate determined by the
relative densities of the fluids. Anderson suggests that a cavity
can be formed by this mechanism until it grows too large to support
the overburden, whereupon roof-fall occurs and fills the cavity.
The third theory, proposed by Kopf and discussed by Zand
(1981), suggests that breccia pipe formation could also result from
tectonic and hydrologic interaction. In a manner analogous to the
formation of pingos in a permafrost region, Kopf's mechanism
involves extreme hydrostatic pressures, induced by tectonism and
exceeding the lithostatic stress, that rupture the overburden. Local
dissolution could introduce a weak spot and provide an initial
pathway for pressurized water to escape. Credence to this theory is
supplied by the observations of Christiansen et al. (1981) in their
investigation of a blowout structure that penetrates Saskatchewan
evaporite deposits and forms Howe Lake. They believe that this
structure was formed by the overpressurization of an underlying
aquifer during glaciation. . '
Which mechanism is most instrumental in the formation of a
breccia pipe depends largely on local conditions. The first two
mechanisms require hydraulic communication between an underlying
aquifer and the salt deposit to initiate breccia pipe growth. The
third hypothesis does not require communication but does depend upon
25
-------
the presence of an aquifer below the salt and the (geologically)
rapid overpressurization of that, aquifer.
4.3 BRECCIA PIPE DISTRIBUTION
Breccia pipes or similar collapse structures have been
identified in all of the bedded salt regions under consideration for
siting waste repositories, although not all collapse structures have
been correlated with deep dissolution. There are, as would be
expected, important differences in the location and distribution of
breccia pipes among the various, salt regions. It is therefore
appropriate to discuss each separately.
In the Palo Duro Basin of the Texas Panhandle and neighboring
Oklahoma, numerous pipes have been identified and investigated.
While there is no complete inventory of these pipes, Gustavson
reports at least 78 and alludes to more (Gustavson et al., 1980).
In several cases, the pipes are associated with surface drainage.
During the construction of Sanford Dam on the Canadian river, 27
pipes were uncovered and more were mapped along the river. Because
exposures are best where erosion or regional salt dissolution
provide topographic relief, most of the visible pipes have been
found along or near valleys or escarpments. It is not yet clear
whether pipes are equally distributed throughout the basin, however.
The pipes uncovered during construction of Sanford Dam had not been
previously recognized; but there has been no similar construction
away from the river course that might reveal other undiscovered
pipes.
Interpreting the age or rate of development of the Palo Duro
features is difficult. Some pipes are overlain with undisturbed
Quaternary deposits, while others are filled. In some, the breccia
has been cemented with carbonates, but it remains uncemented in
others. Since most of the chimneys start in Permian salts and
extend through Triassic or younger sediments, it is possible to
argue that collapse took place well after deposition. Whether there
was a single episode of formation or a steady evolution, or whether
26
-------
it is continuing today, has not been determined. This is partly due
to the lack of a continuous geologic record for the region. Local
dissolution is largely controlled by the quantity and quality of
water in aquifers underlying the salt deposits. Present conditions
suggest that deep dissolution, if it is taking place, is progressing
at a very slow rate and that there is little evidence for active
dissolution except in association with surface drainage (Stone and
Webster, 1981).
Delaware Basin salt deposits are similarly disrupted by a
number of breccia pipes, although again a complete inventory is not
available. In the immediate (10 km) vicinity of the proposed Waste
Isolation Pilot Plant (WIPP) site, at least four surface features
have been recognized as possible pipes. As in other regions, many
collapse structures may not have reached the surface. Structure
contours of the top of the Salado formation show three times as many
locations where deep dissolution could account for unusual
structures (Sandia, 1978), although there is no evidence to suspect
that all or any are indeed breccia pipes.
Initially, there was thought to be a correlation between the
reef formation encircling WIPP and the dissolution features. More
recent evidence indicates, however, that dissolution is more
scattered across the bottom of the Salado salt deposits (Anderson,
1981 and 1982).
Numerous collapse structures, including breccia pipes, have
been reported in the Paradox Basin (Sugiura & Kitcho, 1981). About
40 structures are mapped in an area of about 52,500 square miles.
Others may exist below the surface. The structures described range
in size from 10 to 150 meters in diameter; breccia within the pipes
may range from well-cemented clasts to uncemented sands and silts.
The age of these collapse structures is difficult to deduce but, as
some are overlain by Quaternary strata, it is likely that they are
as recent as Tertiary, possibly formed contemporaneously with
regional salt anticlines in the Early Tertiary. There also seems to
be a relationship between collapse and faulting, which may be
27
-------
attributed to groundwater communicating with the salt deposits
through faults.
Several locations in the Mackinac Straits region of the
Michigan Basin have been identified where breccia structures could
be attributed to collapse of overlying rocks into dissolution
cavities in salt (Landes, 1945). More locations have been found
during recent years. As is the case in most other salt provinces,
the numbers of occurrences of pipes cannot be easily established.
It is particularly difficult in the Michigan Basin because recent
glacial action has obliterated most topographic expressions of
breccia pipes, and vegetation and other ground features make
exploration difficult.
4.4 BRINE RESERVOIRS
Permeable and porous zones within some deposits may contain
trapped water with high concentrations of dissolved material, both
solids and gases. These fluid bodies are generally isolated from
aquifers and are located in discrete strata or areas within a larger
formation. Under certain conditions, brine reservoirs may be eco-
nomically important and are tapped for their salt content.
While brine reservoirs are distinct to each locale, those found
in the Delaware Basin are typical enough to serve as examples.
Here, as far as is known, the large reservoirs are essentially
restricted to the Anhydrite 2 and Anhydrite 3 zones within the
Castile formation (Powers, 1982). Smaller pockets of brine have
been found by potash miners drilling into clay seams above the
potash ore horizons. The volume of the large reservoirs cannot be
determined with accuracy, since few have been tapped and measured,
but a number have yielded at least several million gallons of brine.
The material contained in these pockets is primarily brine
saturated, or nearly saturated, with sodium and calcium salts and
showing more than trace quantities of lithium and other cations.
Gases—primarily nitrogen, carbon dioxide, and hydrogen
sulfide—under high pressure generally accompany these brines.
28
-------
As previously noted, the reservoirs are generally confined to
specific strata. Seismic studies of the Delaware Basin have revealed
that the reservoirs are also associated with such structural
features as domes, anticlines, and synclines affecting those strata
(Powers, 1982). Although not every structure contains a brine :
reservoir, every known sizable reservoir has been associated with
some discernible structure. The extent and height of these
structures, together with measured quantities of brine from tapped
reservoirs, provide estimates of their area and volume. The only
currently available means of confirming that brine is present is to
drill into the structure. Since drilling is often difficult or
expensive, the extent, nature, and size of brine reservoirs can only
be estimated.
4.5 BRINE POCKET DISTRIBUTION
There is less published information for brine pockets than for
breccia pipes and similar collapse structures, largely because
pockets are usually found through drilling, an expensive and very
site-specific technique. Brine reservoirs in the Delaware Basin have
already been discussed. In the other regions of interest, brines
have occasionally been found, usually during exploration for
petroleum or mineral reserves. Except where brines have been of
commercial interest, or have ceased to flow soon after discovery,
their characteristics have not been well documented.
In the Paradox and Michigan Basins, and around the Salton Sea
area, natural brines have been exploited for their mineral content.
Most of the brine reservoirs in the Paradox Basin are believed to be
small and associated with the evaporites of the Paradox Formation
(Stone and Webster, 1981). Michigan Basin brines have been
productive for years, but most are apparently the result of
dissolution by shallow waters rather than deep dissolution features.
Investigation of the larger brine reservoirs is most active in
the immediate vicinity of the WIPP site in the Delaware Basin.
Preliminary chemical and radiochronology tests have indicated that
29
-------
these features are relatively young, with the water on the order of
1 million years old. The water was apparently meteoric, and not the
result of dehydration of gypsum or other hydrous minerals (Powers,
1982). Investigations are continuing to determine whether the brine
reservoirs represent one stage in the development of breccia pipes,
or are independent features. Also remaining to be established is the
source of the water in the brine and whether the reservoirs that
have been found are still connected to a source. There is some
recent evidence that they are actually interconnected and behave
like an aquifer (Anderson, 1982).
Earlier studies of brines discovered during potash mining in
the Delaware Basin suggest that the recently discovered potash mine
brines are decidedly different (GEI.1978). A survey of six mining
operations revealed that the brine pockets encountered were
generally confined to clay strata near the ore horizons. The volume
of brine usually was no more than 10,000 gallons, seldom required
pumping, and was never sufficient to disrupt mining operations.
4.6 DETECTION METHODS
Locating deep dissolution features by existing techniques
requires both luck and judgment. Standard methods for finding
collapse features usually include topographic examination, assisted
by aerial or satellite photography and field checking. If the
surface expression of a feature warrants further study, seismic
surveys may be performed to determine if the underground form of the
structure is consistent with breccia pipes. Finally, drilling is
necessary to confirm the nature of a structure. Similarly, brine
reservoirs may be found through a combined field and seismic survey.
Confirmation of the existence of deep brine also depends upon
drilling into the structure.
Neither drilling nor seismic surveying can be accomplished
quickly or inexpensively, and most investigators believe that many
buried features remain undiscovered. Furthermore, an extensive
drilling program around a potential repository site may do more to
30
-------
compromise the integrity of the site than the mere presence of a
pipe or brine reservoir.
31
-------
References
Anderson, R. Y. Deep Dissolution of Salt, Northern Delaware Basin,
New Mexico. Report to Sandia Laboratories. 1978.
Deep-seated Salt Dissolution in the Delaware Basin,
Texas and New Mexico. New Mexico Geological Society Special
Publication No. 10, 133-145. 1981.
Upper Castile Brine Aquifer Northern Delaware Basin
New Mexico. Department of Geology, University of New Mexico. April
1982.
Christiansen, E. A., D. J. Gendzwill, and W. A. Meneley. Howe Lake:
A Hydrodynamic Blowout Structure. Preprint, 1981.
Geotechnical Engineers, Inc. (GEI). Final Report on Uncertainties
in the Detection, Measurement, and Analysis of Selected Features
Pertinent to Deep Geologic Repositories. July 10, 1978.
Gustavson, T. C. et al. The University of Texas at Austin. Geology
and Geohydrology of the Palo Duro Basin, Texas Panhandle. Prepared
for the U.S. Department of Energy. 1980.
Landes, K. K. Mackinac Breccia, Chap. 3 p. 121-154. In: Geology of
the Mackinac Straits Region and Sub-surface Geology of Northern
Southern Peninsula. (Landes, K. K., G. M. Ehlers and G. M. Stanley,
eds.). Mich. Geological Society Pub. 44, Geol. Ser. 37, 204 p.,
1945.
Powers, D. W., Sandia National Laboratory, personal communication,
1982.
32
-------
Sandia Laboratories (Powers, D. W. et al., eds.). Geological
Characterization Report, Waste Isolation Pilot Plant (WIPP) Site,
Southeastern New Mexico. Prepared for the U.S. Department of
Energy. August 1978. SAND 78-1596.
Stanton, R. J., Jr. The Solution Brecciation Process. Geological
Society of America Bulletin, No. 77, pp. 843-848, August 1966.
Stone & Webster Engineering Corporation. Area Geological
Characterization Report for the Palo Duro and Dalhart Basins, Texas.
Prepared for the Office of Nuclear Waste Isolation, December 1981.
ONWI-292.
Sugiura, R. and C. A. Kitcho. Collapse Structures in the Paradox
Basin. Rocky Mountain Association of Geologists—1981 Field
Conference.
Zand, S. M. State of New Mexico, Environmental Evaluation Group.
Dissolution of Evaporites and Its Possible Impact on the Integrity
of the Waste Isolation Pilot Plant (WIPP) Repository.
September 15, 1981. EEG-14.
33
-------
5. SOLUTION MINING
5.1 INTRODUCTION
Solution mining was not modeled as a separate failure element
in the Task D Report. The decision not to do so was based on the
belief that a repository would be detected during the preliminary
drilling and testing associated with the development of a solution
mine. Furthermore, since salt deposits are widespread and solution
mining is limited, even the consideration of solution mining at a
repository site was regarded as a highly improbable event. This
section reexamines this approach on the basis of a more extensive
review of available information, including the more recent salt dome
siting studies.
5.2 BEDDED SALT
There are approximately 30 salt solution mines in bedded salt
regions of the United States, with at least one such operation in
each of the Permian, Paradox, Silurian, and Williston Basins (USGS,
1970; United States Department of Interior, 1980). (Another 15 or
so solution mines in the salt domes of the Gulf Region are discussed
in Section 5.3.) All the mines are mining sodium chloride except
for one potash solution mine in the Paradox Basin; a solution mine
for trona (a sodium mineral) is being developed in Wyoming. Adding
the number of solution cavities developed for other purposes, such
as underground storage, in the same regions, it appears that there
are 50 to 100 such operations in the United States. On the other
hand, there are about 1 million drillholes in those regions to a
depth that may be considered relevant to a repository (Task D
Report). Thus, if the repository site becomes available for
exploration or exploitation, and if drilling and solution mining
continue in the future at a ratio similar to that of the past,
drilling would probably lead to discovery of the repository before
solution mining.
35
-------
5.3 DOME SALT
The situation with salt dome deposits is quite different from
that with bedded salt. Whereas there may be only about 50 solution
5 2
mining operations in 8 x 10 km of evaporite basins in the United
States, there may be an equal number within the approximately 250
salt domes in the Gulf Region. Given the relatively small size of
most salt domes (a few square kilometers), this number represents
much greater exploitation of the available resource. Furthermore,
dome salt has certain properties that make it a favorable target for
solutioning. These properties include generally higher purity,
lower likelihood of continuous interbeds and other structures that
can cause leaks from storage cavities, and much greater thickness.
As pointed out in the Task D Report, there are multiple, often
competing, uses for salt domes, as illustrated in Figure 5-1. The
geometry of past solutioning cavities in one salt dome is shown in
Figure 5-2, taken from an environmental impact study for its use as
part of the Strategic Petroleum Reserve (Federal Energy
Administration, 1976). Numerical data characterizing the cavities
shown in Figure 5-2 are given in Table 5-1.
The U.S. Department of Energy and its contractors have recently
carried out studies to identify salt domes with potential for a
nuclear waste repository. For several reasons, such as greater salt
dome stability and distance from the sea, the effort focused on the
salt domes of the Gulf Interior region. Seventeen criteria and
screening specifications were used, but three of the criteria
dominated the screening (Law Engineering, 1980).
1) Dome tops should be at depths less than about 3000 feet.
2) Domes should have areas greater than 1000 acres in lateral
extent, plus a 500-foot buffer zone.
3) Domes should not have been used by industry for production,
storage, or other mineral-related use.
36
-------
CHEMICAL REACTION
CHAMBER
HYDROCARBON
STORAGE
AIR STORAGE
FOR PEAKING
RADIOACTIVE
WASTE ISOLATION
SPACE FOR
UNDERGROUND
STRUCTURES
UNDERGROUND
DEGRADATION
OF WAST$
Source: Martinez and Thorns, 1978.
FIGURE 5-1 POTENTIAL DEMANDS FOR SALT DOME UTILIZATION
37
-------
5OO 1000
LEGEND
0> ABANDONED-NOT USABLE n
6 COMMITTED STORAGE «
O USABLE CAVERNS FEET
« UNDEVELOPED CAVERNS
COMMITTED AVAILABLE STORAGE SPACE
2000
Source: Martinez and Thorns, 1978. Modified from Federal
Energy Administration, 1976.
FIGURE 5-2 CAVITY DEVELOPMENT IN A GULF COAST SALT DOME
38
-------
TABLE 5-1
DATA ON CAVITIES SHOWN IN FIGURE 5-2
Javity
1
2
3
8A
11
13
15
16
17
18
19
20
21
Top of
Cavity
(depth/ft)
988
741
876
1242
1067
1112
2597
2620
2598
3500
2980
4126
New Solution
Bottom
of
Cavity
(depth/ft)
1816
1608
1830
1978
1812
1883
3297
3264
4043
4285
4312
4328
Well
Maximum
Diameter/
Height
(ft)
418/1007
*/1121
295/1130
235/765
400/953
300/918
480/980
*/644
295/1543
480/883
415/1420
355/326
Gross -
Volume x 10
(ft3)
8,102
6,000
4,236
3,155
10,454
5,055
16,618
8,800
12,175
10,000
9,000
6,000
No sonar caliper survey available
Source: Martinez and Thorns, 1978.
39
-------
Most domes failed to meet one or more of these criteria, as shown in
Table 5-2, which indicates some of the reasons for failure. Eleven
domes were identified; the number was later reduced to seven. These
seven domes were then studied in greater detail. Among the results
were a set of rankings on the basis of resource value for mineral
development or other purposes, as shown in Table 5-3 (Law
Engineering, 1980). The potential for the production of brine or
salt is very high, as is the potential for use for various types of
underground storage. The cited study points out that there are many
other domes with equally favorable characteristics for brining or
salt mining, and the use of one of these domes would not represent
the loss of a unique resource. Nevertheless, given the basic
assumptions of this study (Task D Report, p. 208), it cannot be
assumed that knowledge of the existence of the repository or control
of the site will deter site use after 100 years. It appears
obvious, therefore, that under these assumptions there is a very
high probability that solution mining, for any of several purposes,
will take place in the future. Furthermore, it is quite possible
that the repository would go unnoticed during part or all of this
operation.
This conclusion arises from the premises for the Task D study.
Stated differently, it does not say that solution mining will
probably occur in a salt dome repository, but only that prevention
of such mining is dependent on human controls indefinitely into the
future.
40
-------
TABLE 5-2
RESULTS OF REGIONAL CHARACTERIZATION SCREENING
NOItTtlLAST AND SOUTH TbXAS IMS INS
El tfrtnat ion
Lt-ss than 1000 acres 500 *
PALElil ItJL
1) rook 6
Grand Sail
But Ier
Wnf tehuui>e
KLtCtll
Ht . Uylvan
Ca t Tyler
OA W(JGD
Ha nvsvi 1 le
be hel
Boggy Creek
Day
Kittrcll
La Rue
Concord
Hoca
Uilvorth Ranch
Pcscadito
HOHTI1 I/XMS1ANA
KAYUURNS
Wii.nf teld
CeJ.ii Creek
VACIILHIL
Gibbland
Prot ho
Prices
Arcadia
Bist inuau
Chestnut
Hi lam
Che s t e r
Sikes
Pack ton
Castor Cict-k
Mclr.tofth
fclCllTUN •
Crouville
Tatua
LAHPTON
Petal
Gilbert
Haielhurht
Arm
HcLaurln
Kichuond
bruinaburg
Byrd
Leedo
half |
>
J
,1
t
•>
3
}
>
Ier
let
00
iff
ier
ler
ier
ier
ier
•
ier
ier
ier
ici
a deep
1
a deep
n dec
M dee
ft dec
B dee
B due
• d««-
n dee
n dee
a deu
f*
f
f
1
fj
J
B deep
B dee
a dee
a dec
h dee
3
>
B deep
• deep
a deep
B depD
_
hap 1000 «crea 4500* barrier
• Chan 915 * < r
leeo
han 1000 aicres +f>00* barrier
Source: Law Engineering Testing Company, 1980
-------
TABLE 5-3
RESOURCE VALUES OF POTENTIALLY SUITABLE SALT DOMES
POTENTIAL
RESOURCE
OR USE
ABSORBENT MINERALS
AGRICULTURAL LIME
AGRICULTURAL PRODUCTS
AND FARMING
AMBER
ASPHALT 1C ROCK
BAUXITE AND BAUXITIC CLAY
BENTONITF. AND OTHER
VOLCANIC MATERIALS
BITTERNS
BLEACHING CLAYS
BRINES
BUILDING STONE
CARBON DIOXIDE
CEMENT MATERIALS
CLAYS AND CLAY MATERIALS
CHALK, MARL & LIMESTONE
CHERT & TRIPOLI
COAL (SEE LIGNITE)
CONSTRUCTION MATERIAL
FOTODRY SAND
— ~i
KEECHI <
0-1
1-<1
3-6
0
3-4
<1
1-2
0-<1
0-1
9-10
0-<1
1-2
1-2
1
1-2
0
0
1-3
0
rotencia.
DAKWOOD
0-1
0-<1
3-6
0
0-<1
<1
1-2
0-<1
0-1
8-10
0-<1
1-2
0-1
1
0-1
0
0
1-3
0
L Kar.K on acaie 01
RAYBURNS VACKERIE
0-1 0-1
1 0-<1
3-7 3-7
0 0
0-<1 0-<1
<1 ' <1
1-2 1-2
0-<1 0-<1
0-1 0-1
9-10 8-10
0-<1 0-<1
1-2 1-2
1-2 0-1
1 1
1-2 0-1
0 0
0 0
1-3 1-3
0 0
i-iu inig
CYPRESS
CREEK
0-1
0-<1
3-7
0
0-<1
0-<1
1
0-<1
0-1
7-10
0-«1
1-3
0-1
1
0
0
0
1-3
0
LAMPTON
0-1
0-<1
3-7
0
0-<1
0-<1
1
0-<1
0-1
6-10
0-<1
1-3
0-1
1
0
0
0
1-3
0
RICHTON
0-1
0-<1
3-7
0
0-<1
0-«1
1
0-<1
0-1
8-10
0-<1
1-3
0-1
1
0
0
0
1-3
0
42
-------
TABLE 5-3 (continued)
POTENTIAL
RESOURCE
OR USE
FULLER'S EARTH
GEOTHERJ1AL & GEOPRESSURED
MATERIALS
GLASS SAND
GLAUCONITE
GYPsra
HEAVY MINERALS
HYDROGEN SULFIDE
IRON ORE
KAOLIN £ KAOLINITIC CLAY
LIGHTWEIGHT AGGREGATE
LIGNITE
MANGANESE
MEDICINE SALTS
MINERAL FILLERS
OIL ANP NATURAL GAS
PEAT
PIGMENTING MATERIALS
AND OCHRE
POZZOLANIC MATERIALS
RECREATION & GENERAL USE
REFRACTORY MATERIALS
ROAD MATERIALS
KEECHT
0
3-4
1-2
1-2
0-<1
il
il
<1
1
il
4-6
0
0-<1
il
1-2
0
il
»l
5-10
il
1-4
OAKWOOD
0
3-4
1-2
1-2
0-<1 .
il
il
<1
1
il
4-6
0
0-«1
il
2+
0
il
>1
5-10
il
3-4
RAYBURTJS
0
3-4
1-2
1-2
0-<1
i3
il
1-2
<1
il
3
0
0-<1
il
il
0
il
>1
5-10
il
1-4
VACHKRIE
0
3-4
1-2
1-2
0-<1
+3
il
1-2
<1
il
4-6
0
0-<1
il
1-2
0
il
>1
5-10
il
3-4
CYPRESS
CREEK
0
3-4
1-2
0-1
0-<1
i3
3-4
0-<1
0
il
0-<1
0
0-<1
il
2+
0-1
il
>1
5-10
il
3-4
LAMPTON
0
3-4
1-2
0-1
0-<1
i3
3-4
0-<1
. 0
il
1-2
0
0-<1
ii
il
0
il
>1
5-10
il
3-4
RICHTON
0
3-4
1-2
0-1
0-<1
i3
3-4
0-<1
0
il
0-<1
0
0-<1
il
il
0
il.
>1
5-10
il
3-4
ROCK PHOSPHATES
43
-------
TABLE 5-3 (continued)
POTENTIAL
RESOURCE
OR USE
SALTS
SAND & GRAVEL
SHELLS
SULFUR
UNDERGROUND STORAGE
USUAL GASES
(SEE C02 & H,S)
URANIUM
KEECT.I
9-10
2-3
0
0-1
'7-9
0-<1
<1
OAKWOOD
9-10
3-4
0
0-1
7-9
0-<1
<1
RAYBURNS VACHERIE
9-10 8-10
:-3 3-4
0 0
0-1 0-1
7-9 7-9
0-<1 0-<1
<1 <1
CYPRESS
CREEK
8-10
3-4
0
0-1
7-9
0-<1
<1
LAMPTON
8-10
3-4
0
0-1
7-9
0-<1
<1
RICHTON
8-10
3-4
0
0-1
7-9
0-<1
<1
*
Best-guess estimate rankings of potential mineral resources and uses,
based on available data and subject to modification with changes in
data or knowledge.
1 Non-prospective, non-exploitable, or non-usable
0-1 Negligible to speculative
1-3 Poor to fair
4-6 Fair to good
7-9 Good to excellent
10 Productive/in use or capable of production/use
Source: Law Engineering Testing Company, 1980.
44
-------
References
Federal Energy Administration. Strategic Petroleum Reserve - Final
Environmental Impact Statement for Bayou Choctaw Salt Dome. 1976.
FES 76-5; FEA/S-76/501.
Law Engineering Testing Company. Geologic Area Characterization.
Volume 1: Introduction, Background, and Summary. Gulf Coast Salt
Domes Project. Prepared for Battelle Memorial Institute. August
29, 1980.
Martinez, J. D. and R. L. Thorns. A Systems Concept of Space
Utilization in Gulf Coast Salt Domes. In: Storage in Excavated Rock
Caverns. (Magnus Berman, ed.), Rochester 77, ^. Pergamon Press,
New York, 1978.
U.S. Department of the Interior, Geological Survey (USGS). The
National Atlas of the U.S.A. 1970.
U.S. Department of the Interior, Bureau of Mines. Mineral Facts and
Problems. 1980.
45
-------
6. THERMAL BUOYANCY MODEL
The Task D Report describes a simplified method of analyzing
the effect of heat generated by radioactive wastes on vertical
groundwater flow (Appendix D-VI). This method requires the
following data:
• Temperature distribution in the overlying rock at the time
of interest.
• Water density as a function of temperature.
• Water viscosity as a function of temperature.
• Hydraulic conductivity of the water column pathway.
• Effective porosity of the water column pathway.
This information is entered into the computational framework
illustrated in Figure 6-1. The output from the computations are
volumetric flow rates and travel times for water from the repository
to an overlying aquifer when driven by the thermally induced
buoyancy effect.
To evaluate the reasonableness of this approach, the simplified
model was applied to the generic granite repository described in
Dames & Moore (1978). Comparison of the results of the Task D
approach with those of detailed numerical simulations gives further
evidence that application of the simplified model is adequate for
generic repository evaluations. Figure 6-2, which is taken from the
Dames & Moore report, is a schematic of the reference granite
repository. The repository is located at a depth of 2000 feet (610
meters) below the land surface in unweathered granite. The granite
stratlgraphic section was described as having four zones with
different hydraulic characteristics, as shown in Table 6-1.
47
-------
S1
S2
S3
S4
S6
S7
S9
S10
Pick time t.
Obtain temperature
distribution from
other documents.
Compute corresponding
density distribution.
Compute mass of
vertical column
of water.
S5
Compute mass of column
of water at normal tem-
perature.
Determine vertical force
on column and corres-
ponding pressure.
Convert to equivalent
hydraulic gradient.
S8
Compute water
viscosity for ambient
conditions.
Determine flow rates and
interstitial velocity by
Darcy's law.
Determine travel time
f'om repository to aquifer.
Source: Task D Report.
FIGURE 6-1 INFORMATION FLOW CHART FOR BUOYANCY COMPUTATIONS
48
-------
Depth
(ft)
0
Low Ground Relief
Water Table Slope = Regional Dip
Ground Surface
)
)
«\
>•
M
• (0
•TJ
c
\ D
1 O
0)
£
IT}
o
0)
(i
4
inn
1500 -
— W.&0-
[ 720° ]
•^tA^W^
^%'f^1
^v-^k7
&>&£*;
vj.'iVoV*'
'^^'N'sX/x-V
i,^-;,;-;^;
,'M '/ x, ',» '-
-\\'—x,vl-«»
• -V-/ ', iv i^ ,
^^
-V'N-"^"/
v^>^V
' xN 1 - \ X ' 1 s •
:>^;4'^
x^> '_'\/^^l
vl^WrVr^
v.-V'xr^r,1,.-
r-4, -<''¥'"r»; --_J
.'V^»NiV'7V''
r.,-',-'v-/Ji'x'
•^-•K^
rtV^y
* \l / ' 1 / 1 .* x~
•^.Jr _•• i\'m m\,
•W\'v-x/^'>
x'.' 'V'-> ' •./ ^S
•» OXN"X/_X ' 1^
i'-r,';-^V.)
X~/%'N'X«* / ' ,'
-'x;xic.>\<»-v',N
"»* •-'/A-'yX
Nc»af-hfar^^ fli-snit-o K =K =
x y
Partially Weatherec" Granite
Repository
2000 ft depth
Unweathered Granite K =K =
x y
K =K =
x y
Lower Boundary (considered
i i I
20 10 6
Distance in thousands of
5x10 cm/s
K =K = 5xlO~6cm/s
x y .
5xlO~8cm/s
-9
5x10 cm/s
impermeable)
i r i 1
10 20 30
feet
,
(
undary
o /
CO (
o> /
en (
^
(0 ,
x: '
o
w /
S (
(
(
40
Source: Dames & Moore, 1978.
FIGURE 6-2 GENERIC GRANITE STRATIGRAPHIC SECTION IN ITS REGIONAL SETTING
-------
TABLE 6-1
HYDRAULIC CHARACTERISTICS OF THE GENERIC
GRANITE STRATIGRAPHIC SECTION
Isotropic Hydraulic
Conductivity
5 x 10~4
5 x 10'6
5 x 10'8
5 x 10~9
Effective
Porosity
1 x 10~3
1 x 10"4
1 x 10~4
1 x 10~4
cm/s
ID'3
io-4
io-4
io-4
Depth, Meters (ft)
0-61 (0-200)
61-457 (200-1500)
457-1554 (1500-5100)
1554-2195 (5100-7200)
Note: For flow calculations, the three zones that extend from the
land surface to the repository depth (610 meters) can be
represented by an equivalent zone 156 meters thick with a
_o
vertical hydraulic conductivity of 5 x 10 cm/s.
Source: Dames & Moore, 1978.
50
-------
The Task D model uses the following equation to calculate
travel time from the repository to the overlying aquifer:
7 - (6.1)
where
T = travel time from the repository to the overlying
aquifer or target horizon, years.
D = distance from the repository to the overlying
aquifer or target horizon, meters.
K = hydraulic conductivity, meters/year.
i = effective thermal buoyancy gradient, dimensionless.
100 y = viscosity of water at the representative temperature
divided by the viscosity of water at 20°C, dimensionless.
n = effective porosity, dimensionless.
The viscosity term is based on the average temperature of the water
column and the gradient term is related to the average density of
the water column by:
i = 1-p* (6'2)
*
when p is the average density of the water column under
consideration.
To estimate the appropriate viscosity and gradient terms, a
running profile of water density is plotted from the overlying
aquifer to below the repository and the representative water density
is determined when the incremental density addition equals
(approximately) the average density of the column. (Cf. Appendix
D-VI.) The gradient term is calculated by Eq. (6.2), and the
viscosity term is based on the average water temperature as
estimated from the average water density. Substituting values used
51
-------
for the representative generic granite section (D = 156 meters, K =
-2 -4
1.58 x 10 meters/year, n = 10 ) into Eq. (6.1) yields
T = 100 (•£•) years.
Following the method described in the Task D Report, the
estimated travel time for the generic granite repository 1000 years
after repository sealing is determined by plotting temperature
profiles from the calculated temperature distribution data included
in the Dames & Moore report. There is one profile for a 100 kW/acre
loading rate and another for a 200 kW/acre loading rate, included as
Figures 6-3 and 6-4. These profiles are also indicative of density
variations. From Figure 6-3 the average density of the water column
with maximum vertical velocity may be estimated to correspond to an
average temperature of 30 - 50°C. Similarly, from Figure 6-4 an
average temperature of 40 - 60°C may be estimated. Table 6-2
contains the viscosity, density, gradient, and travel times
corresponding to these temperature values.
Table 6-2 shows that using the Task D model, calculated travel
time from the repository to land surface for the generic granite
repository with a 100 kW/acre leading rate is between 45 and 205
years, compared with 100 years using the numerical model (Dames &
Moore, 1978, Table 2-3). For a 200 kW/acre loading rate, the
calculated travel time with the simplified model is between 30 and
80 years, compared with 40 years using the numerical model (Dames &
Moore, 1978, Table 2-3). The close agreement between the calculated
travel times of the simplified model and the numerical model
suggests that for generic repository analyses the simplified model
is adequate.
-4
Using an effective porosity (n) of 10 for the whole section
instead of making an adjustment for the higher effective porosity
_2
(10 ) reported for the weathered granite reduces the calculated
travel time by less than one year.
52
-------
0)
0)
8 1000
<0
•a
c
03
rn 2000
a
0)
Q
3000
t = 853 years
I
I
I
I
1
I
I
I
I
I
I
250
500
8
CO
"i
750
2
v
ca
a
-------
1000
0)
u
ro
**-
L_
3
CO
•o
to
O
» 2000
CO
a
0)
Q
3000
t = 853 years
250
500
o
u
I
.o
-------
TABLE 6-2
WATER VISCOSITY, DENSITY, HYDRAULIC GRADIENT AND
MINIMUM TRAVEL TIME FOR SELECTED REPRESENTATIVE TEMPERATURES
Temperature
°C
Viscosity, y,
Poise
Density, p*
. 3
gm/cm
Gradient, i
Time, T
years
30
40
50
60
.0082
.0065
.0054
.0047
.996
.992
,988
.983
.004
.008
.012
.017
205
80
45
30
55
-------
Reference
Dames & Moore. Technical Support for GEIS: Radioactive Waste
Isolation in Geologic Formations. Volume 21. Groundwater Movement
and Nuclide Transport. Prepared for the U.S. Department of Energy.
1978. Y/OWI/TM-36/21.
56
-------
7. HYDRAULIC CONDUCTIVITY OF FAULT ZONES
In the Task D Report, analyses of groundwater flow along
pathways resulting from fault movement included an estimate of the
hydraulic conductivity of the actual faulted pathway. Table 7-1
lists the estimated hydraulic conductivity values used in the Task D
analyses. Recent hydrologic investigations of the Climax Stock
granite at the Nevada Test Site have provided estimates of hydraulic
conductivity for fractured zones in granitic rocks (Murray, 1981).
Permeability tests of the Climax Stock at the Nevada Test Site
-4 -1
yielded permeability values of 10 to 10 darcies for moderately
to highly fractured rock. These values correspond to a hydraulic
-7 -4
conductivity of 10 to 10 cm/s. Laboratory permeameter tests on
material taken directly from fault shear zones and, therefore,
representative of the hydraulic conductivity of the fault zone
_2
itself, yielded hydraulic conductivity values of 3 x 10 cm/s to 6
_2
x 10 ' cm/s. On the basis of an analysis of the testing procedure,
the investigations concluded that higher laboratory values, i.e., 6
_2
x 10 cm/s, are probably mere representative of field conditions
than are the lower values. These higher values are in close
agreement with those used for granite in the Task D analyses.
57
-------
TABLE 7-1
HYDRAULIC CONDUCTIVITY OF FAULTED PATHWAYS USED IN THE TASK D REPORT
Rock Type Hydraulic Conductivity cm/s
Bedded Salt 10~4
_2
Granite 10
_9
Basalt 10
Shale 10~4
Dome Salt 10"4
58
-------
Reference
Murray, W. A. Lawrence Livermore Laboratory. Geohydrology of the
Climax Stock Granite and Surrounding Rock Formations, NTS. Prepared
for the U.S. Department of Energy. May 1981. UCRL 53138.
59
-------
8. HYDROLOGIC PARAMETERS FOR HOST ROCKS
8.1 REVIEW OF REFERENCE PARAMETERS
The Task C and D Reports each presented hydraulic conductivity
and porosity values for various rock types, based on values reported
in the literature. The purpose of this section is to compare these
values and to supplement them with additional recent data. For
hydraulic conductivity (also loosely referred to as "permeability"
in some sources), this comparison is given in Figures 8-1 through
8-4. For porosity, a comparison is given in Table 8-1. Allowing
for differences in the meaning of terms, e.g., total porosity vs.
effective porosity, the original values are in good agreement and
are confirmed by more recent results.
8.2 ASSUMPTION OF NEGLIGIBLE PERMEABILITY FOR SALT
There is support in the literature for the use of a zero value
to characterize salt hydraulic conductivity for model calculations,
although non-zero values have also been determined in a number of
laboratory and in-situ tests. Higher values may be due in part to
any or all of the following: disturbance of core samples by
reduction of pressure, even if pressure is later reapplied;
measurements with gases or organic liquids that do not lead to
sealing of flow pathways; transmission of fluids into or through
more permeable interbeds (as in measurements of hydrocarbon losses
from storage cavities). Only the last of these is inherent to the
salt deposit itself, and if the emphasis in modeling were on
horizontal flow through the salt formation, then this factor might
need to be considered. Vertical flow, which would lead to more
rapid introduction of radionuclides into an aquifer, appears to be
of greater concern, however, and the horizontal interbeds have
little or no effect in transmitting water in this direction.
Several of the release mechanisms for a salt repository
discussed in the Task D Report led to the calculation of volumetric
flow rates for groundwater moving from the repository to the upper
aquifer. In every case, the pathway for such flows was through some
61
-------
N>
10'1 10'2 10~3 10'4 10'5 10'6 10'7 10'8 10'9 10'10 10'11 10'12 10'13 10'14
Task C (Table C-1)
Task C (Table C-2)
TaskD
GEIS
NBS (Monograph 167)
NUREG/CR-0912
i i i i r
1 I I I I I I
I
"negligible'
Note: For bibliographic citations, see the reference list at the end of Section 8.
FIGURE 8-1 REPORTED RANGES OR VALUES OF HYDRAULIC CONDUCTIVITY OF SALT (cm/sec)
-------
Task C (Table C-1)
Task C (Table C-2)
TaskD
GEIS
NUREG/CR-0912
10"1 10'2 10"3 10'4 10"5 10~6 10"7 10~8 10"9 10'10 10~11 10"12 10"13 10"14
"~1 I I I I I I I I \ I I I |
"very low if no joints or fractures'
Note: For bibliographic citations, see the reference list at the end of Section 8.
FIGURE 8-2 REPORTED RANGES OR VALUES OF HYDRAULIC CONDUCTIVITY OF GRANITE (cm/sec)
-------
ON
-P-
10'1 10'2 10'3 10'4 10'5 10'6 10'7 10'8 10'9 TO'10 ID'11 10'12 10'13 ID'14
Task C (Table C-1)
Task C (Table C-2)
TaskD
GEIS
NUREG/CR-0912
NWTS-15
I I I I T \ \ Y
I I
1 I I I
"very low if unfractured and not jointed'
i "less" for dense columnar portions
Note: For bibliographic citations, see the reference list at the end of Section 8.
FIGURE 8-3 REPORTED RANGES OR VALUES OF HYDRAULIC CONDUCTIVITY OF BASALT (cm/sec)
-------
10~2 10"3 10"4 10"5 10"6 10~7 10"8
10'9 10'10 10'11 10'12 10'13 10'14
Task C (Table C-1)
Task C (Table C-2)
TaskD
GEIS
NUREG/CR-0912
I I I
I I I I I
not available due to misprint
I I
Note: For bibliographic citations, see the reference list at the end of Section 8.
FIGURE 8-4 REPORTED RANGES OR VALUES OF HYDRAULIC CONDUCTIVITY OF SHALE (cm/sec)
-------
TABLE 8-1
REPORTED RANGES OR VALUES OF POROSITY
OF POTENTIAL HOST ROCKS
Rock
Type
Reference
Porosity
Salt
Task C
Task D
GEIS
NUREG
NBS
< 0.01
none used
0.014 - 0.10
< 0.01
< 0.02
Granite
Basalt
Task C
Task D
GEIS
NUREG
Task C
Task D
GEIS
NUREG
0.06 - 0.09
0.001 - 0.0001
0.005
(fracture porosity)
5 x 10~ - 5 x 10~ (fracture porosity)
0.03 - 0.04
0.001 - 0.0001
0.006
0.20 - 0.001
(fracture porosity)
Shale
Task C
Task D
GEIS
NUREG
0.03 - 0.07
0.001 - 0.0001
0.0 - 0.45
0.01 - 0.40
(fracture porosity)
66
-------
anomalous feature, such as a partially sealed shaft or borehole, a
fault, or a breccia column. (See especially Tables D-2, D-3, D-4,
D-6, D-9, and D-10, of the Task D Report.) For the purpose of
comparison, it is useful to calculate the Darcy flow through the
entire repository itself, assuming some (small) positive hydraulic
conductivity. Table 8-2 presents a set of representative values,
f> 2
based on a cross-sectional area of 8 x 10 m for the flow path,
i.e., the repository area. These flows are very small and indicate
that flow through the bulk rock is generally not important in the
case of salt.
8.3 VERTICAL FLOWS IN THE PRESENCE OF INTERBEDS FOR CASES OF BASALT
AND SHALE
The generic repository setting for basalt was illustrated in
Figure D-5 of the Task D Report and is repeated here as Figure 8-5.
The basalt is shown to have a thickness of 200 meters. In fact, no
single basalt flow in the Columbia River Basalt Sequence has such a
thickness, and the generic basalt should be interpreted as a
sequence of flows along with a number of interbeds. (This
configuration is similar to the case of bedded salt, except that the
interbeds in the basalt can be quite thick and rather permeable.)
The purpose here is to calculate an equivalent hydraulic
conductivity, K, for the composite vertical sequence, based on the
characteristics of more than one type of layer. This is a
straightforward calculation, but it is useful to have available both
the formula and some representative results.
For simplicity, assume that the vertical flow pathway
connecting the upper and lower aquifers through the repository
consists of two basic types of material: dense basalt with hydraulic
conductivity K and more permeable zones (flow tops, interbeds) with
hydraulic conductivity K^. Let L.. be the sum of the thicknesses of
the dense basalt layers and L~ the sum of the thicknesses of the
other layers. Thus, for the generic repository setting,
67
-------
TABLE 8-2
VERTICAL VOLUMETRIC GROUNDWATER FLOWS THROUGH PORTION
OF SALT BED CONTAINING REPOSITORY, AS A FUNCTION OF
HYDRAULIC GRADIENT AND HYDRAULIC CONDUCTIVITY
(Q in m3/yr)
Hydraulic Conductivity, K Hydraulic Gradient,
(cm/sec) 0.001 0.01 0.02
io-14
io-13
io-12
io-n
io-10
5 x
5 x
5 x
5 x
5 x
io-5
io-4
io-3
io-2
io-1
5 x
5 x
5 x
5 x
5
io-4
io-3
io-2
io-1
io-3
io-2
io-1
1
10
i
5 x
5 x
5 x
5
50
0.1
io-3
io-2
io-1
*
Note: Viscosity of water is assumed to be 0.005 poise, corresponding to
an average temperature of 60°C. This would vary with time.
68
-------
-Surface
Surface
• — Deposits
^Aquifer ™;ee;;
460 meters
Hi ^^e/^m^immm^mmmiimm^ If0 JJJ^'S
Source: Task D Report, Figure D-5.
FIGURE 8-5 GENERIC REPOSITORY IN BASALT
69
-------
L2 = L = 200 m (8>1)
From the point of view of vertical flow calculations, the results
are not affected by grouping the dense and permeable zones into two
composite zones. Let A be the cross-sectional area of the vertical
flow path. For the generic repository this would just be 8 x 10 ,
the repository area. Let A be the total head drop over the dense
zones, and A? the drop over the permeable zones. By continuity of
flow,
A. A_
Q = K. ^ A c(y) = K- -^- A c(y) (8.2)
1 Ll L L2
where Darcy's law (with viscosity correction) has been applied to
each zone. The unknowns here are A and A?, the proportional drops
over the two zones, and these are related by the additional equation
A = Ai + A2 (8'3)
It follows from solving the simultaneous equations that
L,
KnA
A
1 h . L2
L2
Q = ; : A c(y)
Jl + Jl
K K (L + L )
1 Z L Z A -A c(y) • (8.4)
L1+K1L2 Ll+ L2
70
-------
Comparing this with Darcy's law shows that the composite system has
an effective conductivity given by
K1K2(L1 + V
K= KV +\ L (8.5)
K2L1 + K1L2
Selected sample values are given in Table 8-3. These values
illustrate that the low-permeability zone controls the flow, with
essentially all the head drop taking place over it, i.e., A = A.
(This could also be deduced directly from the equation for K, or by
considering an electrical resistance analogue.)
Suppose now that the dense zone has an effective porosity n,
and the permeable zone an effective porosity n_, where it is
possible that r\ is much larger than n . It is desired to calculate
the average travel time of fluid as it passes through the flow
pathway. The vertical linear velocity in the less permeable zone is
given by
V (8.6)
X J.
and in the more permeable zone by
v2 = Q/n2A (8.7)
The total travel time T is the sum of the times in the two zones,
given by
T " Tl + T2
Vl V2
L.. n,A L-TI A
X L + L L (8.8)
71
-------
TABLE 8-3 ; <
EQUIVALENT COMPOSITE HYDRAULIC CONDUCTIVITY OF
FLOW PATH CONSISTING OF HIGHER AND LOWER
PERMEABILITY ZONES
(cm/sec)
a) Higher Permeability Zones Equal to 50% Total Thickness
Kl
(Less Permeable
10
10
10
10
Zone)
-9
-8
-7
-6
(More
!0-6
2 x 10"9 2
2 x 10"8 2
1.8 x 10"7 2
10"6 1.8
K2
Permeable
ID'5
x 10"9
x 10"8
x 19"7
x ID'6
Zone)
ID'4
2 x 10"9
2 x 10"8
2 x 10"7
2 x ID'6
io-3
2 x 10"9
2 x 10"8
2 x 10"7
2 x 10"6
b) Higher Permeability Zones Equal to 20% Total Thickness
Kl
(Less Permeable Zone)
10
10
10
10
-9
-8
-7
-6
io-6
1.2 x 10"9
1.2 x 10"8
1.2 x 10"7
io-6
(More
K2
Permeable
Zone)
io-5 io-4
2 x 10
2 x 10
2 x 10
1.2 x 10
-9
-8
-7
-6
2
2
2
2
x 10
x 10
x 10
x 10
-9
-8
-7
-6
io-3
2 x 10
2 x 10
2 x 10
2 x 10
-9
-8
-7
-6
72
-------
Thus if L«ru is much larger than L.TI , which is quite likely, then
the control on travel time is the more permeable and porous zone.
In practice L^ru may be one or two orders of magnitude higher, the
effect being to increase the fluid travel time significantly. Thus
the following conclusion can be drawn:
While the permeable zones have a relatively small effect
in increasing volumetric flow rates, they can have a
large effect in decreasing fluid velocities and hence
increasing travel times.
For reference purposes, observe from Eq. (8.8) that
L n + L n, (L + L,)A
' 2
so that the effective porosity for velocity calculations is
V2 (8.10)
Ll + L2
which equals the average porosity. (Thus, porosity can be averaged
while conductivity cannot.)
The same analysis applies to shale (and even to salt) , but
there are generally few or no significant interbeds in potential
repository formations and so the results are not significant.
73
-------
References
Lawrence Livermore Laboratory. Geoscience Data Base Handbook for
Modeling a Nuclear Waste Repository. Volume I. Prepared for the
U.S. Nuclear Regulatory Commission. January 1981. NUREG/CR-0912;
UCRL-52719.
U.S. Department of Commerce. National Bureau of Standards
(Gevantman, L. H., ed.), Physical Properties Data for Rock Salt.
January 1981. NBS Monograph 167.
U.S. Department of Energy. Final Environmental Impact Statement.
(GEIS). Management of Commercially Generated Radioactive Waste.
October 1980. DOE/EIS-0046F.
U.S. Department of Energy. Proceedings of the 1981 National Waste
Terminal Storage Program Information Meeting. November 1981.
DOE/NWTS-15.
74
-------
9. FRACTURE FLOW
The purpose of this section is to review the so-called "cubic
law" for the hydraulic conductivity of fractured rock (based on
unknown parallel fractures) and to survey various investigations
that have tried to validate or modify it. This law was used in the
Task C Report for certain calculations on flow through fractured
rock. This model does appear to be valid for the generic
calculations carried out.
Snow (1965) modeled flow in a fracture as flow between two
smooth parallel plates. For a single fracture, the discharge is
where
f\
Q = discharge of a fracture (cm /s)
e = fracture opening (cm)
2
g = gravitational acceleration constant (981 cm/s )
2
v = coefficient of kinematic viscosity (0.013 cm /s for
pure water at 10°C)
dh/d£ = head gradient (cm/ cm)
W = width of the fracture (cm).
For a fracture of unit width, Eq. (9.1) becomes
0 = M_ **L (92)
M 12 v
-------
Q - KA (9.3)
2
By equating (9.1) and (9.3), and noting that A = e«l cm , an
expression for the hydraulic conductivity of a single fracture can
be developed:
2
K = &e (Q A1
Kf 12v ^<4J
The velocity in a single fracture is, then
The effective hydraulic conductivity of a unit area of an array
of parallel, evenly spaced smooth fractures of equal opening e, can
be derived from Eq. (9.4), by multiplying K by the equivalent
porosity of a unit area of material. This equivalent porosity can
be written as the product of the area of a fracture per unit width
of material and the number of fractures per unit length of material.
This latter is equivalent to the inverse of the spacing between
fractures . Thus , we have
K
where
b = spacing between fractures (cm) Hook and Bray (1974) .
This expression is used in the Task C Report to describe the
effective hydraulic conductivity of fractured rock. The velocity in
a single fracture can be evaluated from Eq. (9.6) by dividing by the
porosity, e/b, and multiplying by the head gradient, yielding Eq.
(9.5).
The cubic law has formed the basis of theoretical
investigations into fracture flow. Recent work has tended to
confirm its validity, although extensions of the
76
-------
theory to account for more realistic conditions have outstripped
detailed experimental investigations. Louis and Maini (1970), Sharp
et al. (1972), and Gale (1975) emphasized the concept of an
"equivalent" aperture to account for fracture wall roughness and
nonuniform flov7, thus circumventing the difficulties of measuring
roughness and nonuniform flow directly. While the calculation of
effective fracture openings provides useful simplifications for
modeling purposes, it has been noted that the approach produces
values orders of magnitude smaller than actual measured openings
(Gale, 1975). The relation of these differences to the actual
velocities and transit time's of fluid particles in a fracture in
situ has not been thoroughly explored. Witherspoon et al. (1980)
studied the effects of cyclic loading of rough fractures in granite,
basalt, and marble and found that deviations from the ideal parallel
plate model could be accounted for by introducing a roughness
factor, f, as a divisor in the flow equation, Eq. (9.2). Neuzil and
Tracy (1981) extended the cubic law to the case of a fracture
opening that varies continuously perpendicular to the flow but is
uniform in the direction of the flow. They note the lack of
suitable data from experimental work for testing the validity of
current fracture flow theory.
Detailed physical description of fracture spacing and fracture
openings is only one technique for measuring bulk conductivity. A
given bulk conductivity value can result from a range of pairs of
values of fracture opening and fracture spacing. Since the flow
velocity in a fracture depends on the square of the opening, a small
difference in fracture opening can significantly affect the time of
first arrival of a particle of fluid or solute in the fluid.
Furthermore, the parallel plate model analysis leads to a viscous
flow with a maximum velocity 1.5 times the average velocity within a
given fracture.
A range of fracture openings from 0.1 to 1000 ym is
representative of the extremes of available data (Snow, 1968;
77
-------
Witherspoon et al., 1980). Average velocities and times to travel 1
km, calculated for several values in the range, are shown in Table
9-1.
While there are few data available on fracture spacing and
fracture aperture sizes at the depths being considered for
repositories, data presented by Snow (1968) indicate fracture
openings in the range of 50 - 100 ym at depths of 30 - 100 meters.
It would be reasonable to expect smaller average values at greater
depths, and consequently slower transit times.
78
-------
TABLE 9-1
EFFECT OF APERTURE SIZE ON TRAVEL TIME
Average
Fracture Opening (e), ym Velocity (m/y) Time to Travel 1 km (years)
0.1 2.0 x 10~3 5.04 x 105
1.0 2.0 x lO'1 5.04 x 103
10.0 2.0 x 101 5.04 x 101
100.0 2.0 x 10~3 5.04 x 10"1
1000.0 2.0 x 105 5.04 x 1(T3
2
Note: Values here are calculated using v = 0.013 cm /s, and a
hydraulic gradient of 0.01 in Eq. (9.4).
79
-------
References
Gale, J. A Numerical, Field, and Laboratory Study of Flow in Rocks
with Deformable Fractures. University of California, Berkeley,
Ph.D. dissertation. 1975.
Hook, E. and J. Bray. Rock Slope Engineering, Union Brothers, Ltd.,
England, 1974.
Louis, C. and Y. N. T. Maini. Determination of In-situ Hydraulic
Parameters in Jointed Rock. In: Proc. 2nd Congress International
Society for Rock Mechanics. Belgrade. J^, 1970.
Neuzil, C. E. and J. V. Tracy. Flow in Fractures. Water Resources
Res., l]_, 191-199, February 1981.
Sharp, J. C., Y. N. T. Maini, and T. R. Harper. Influence of
Groundwater on the Stability of Rock Masses. Trans. Sec. A., Inst.
Min. Metall. 811(782), A13, 1972.
Snow, D. T. A Parallel Plate Model of Permeable Fractured Media.
University of California, Berkeley, Ph.D. dissertation. 1965.
Rock Fracture Spacings, Openings, and Porosities. J.
Soil Mech. Foundations Div. ASCE 94(5MI), 73. 1968.
Witherspoon, P. A., J. S. Y. Wang, K. Iwai, and J. Gale. Validity
of Cubic Law for Fluid Flow in a Deformable Rock Fracture. Water
Resources Research 16, 1016-1024, December 1980.
80
-------
10. GEOCHEMICAL RETARDATION
The purpose of this section is to update the retardation
factors reported earlier in the Task C Report.
Geochetnical retardation affects radionuclide transport by
causing ions to move at velocities lower than that of the
transporting groundwater. The retardation factor is a measure of
the ratio of the velocity of the flowing groundwater to that of the
radionuclide under consideration and represents the sorptive ,
capacities of the geologic medium. The sorptive capacity is
represented by a distribution coefficient that reflects the
partitioning of the radionuclide between the groundwater and the
solid part of the medium. The retardation factor is usually
calculated on the basis of the physical properties of the medium
together with the distribution coefficient, or it is measured in the
field or laboratory using pulse tracer tests.
The retardation factor (R,) is related to the distribution
d
coefficients by the following relationships:
For porous media, R, = 1 + -r K, (dimensionless) (10.1)
For fractured media, R, = 1 + R-K (dimensionless) (10.2)
d i a
where
3
p = bulk density of the medium (g/cm )
= porosity
2
Rf = surface to volume ratio of the fracture (cm /ml)
K,,K = distribution coefficients for porous and fractured
d a
media, respectively.
If there Is no sorption, the distribution coefficients are zero and
the retardation factors take their minimum value (R, = 1). Figure
10-1 shows the relationship between the retardation factor and the
distribution coefficient for two porous rock media.
81
-------
2
Log Rd
Source: Adapted from NUREG/CR 0912,1981.
FIGURE 10-1 RELATIONSHIP BETWEEN RETARDATION FACTOR, Rd, AND DISTRIBUTION
COEFFICIENT, Kd
82
-------
Measurements of density, porosity, and distribution
coefficients are easily made in the laboratory, but it is not known
how well the laboratory measurements of K, represent field
conditions. Thus, it is uncertain how well the calculated
retardation factors represent field conditions.
10.1 FIELD MEASUREMENTS OF RETARDATION FACTOR
Field measurements of retardation factors are commonly made by
trace injection tests. Radionuclides are injected into groundwater
at one well and measured at a second well some distance away. The
in-situ retardation factors are calculated by dividing the travel
time of each radionuclide by the travel time for tritium (a
nonreactive tracer) as shown in Figure 10-2.
In-situ measurements of retardation factors are generally more
reliable than those based on laboratory measurements because they
are representative of the local geologic environment, taking into
account the heterogeneity of the geologic material. The effects of
such parameters as temperature, oxidation potential, and pH cannot
be determined, however, because these factors are generally constant
in the field environment. Data from in-situ measurement of
radionuclide migration in deep groundwater are limited and are
summarized in Table 10-1.
10.2 CALCULATED RETARDATION FACTORS BASED ON DISTRIBUTION
COEFFICIENTS
Figure 10-3 is an updated version of Figure C-l of the Task C
Report. This figure was modified to include more current
information regarding distribution coefficients for strontium and
cesium. The new information was contained in a report prepared for
the U.S. Nuclear Regulatory Commission by Muller et al. (1981).
Table 10-2 was prepared from information contained in the
Lawrence Livermore report (1981) and summarizes the range of
retardation factors estimated from distribution coefficients
compiled from the literature. The table is subdivided into three
major groups: fission products with no sorption, fission products
83
-------
c
o
c
0>
Spike containing '.-',.
3Hand85Sr For constant flow,
o ic,\ — Sr
'Sr
Time
Source: Adapted from NUREG/CR 0912,1981.
FIGURE 10-2 ILLUSTRATION OF CALCULATION OF Rd FROM IN-SITU MEASUREMENT
OF TRAVEL TIMES BETWEEN TWO WELLS
84
-------
TABLE 10-1
RETARDATION FACTORS BASED ON IN-SITU MEASUREMENTS
OF RADIONUCLIDE MIGRATION IN DEEP GROUNDWATER
Strontium Cesium Technetium Iodine Neodymium
Basalt1 4 NMD* -
Granite2 6 NMD* 1 1 NMD*
Robertson and Barraclough, 1973
2
Landstrom et al., 1979
*
No migration detected
85
-------
Cesium
Mixed-Fission
Products
Strontium
Iodine
Note: Distribution coefficient values for ions shown.
Sources: SAND81 -0557 and BNWL-1 900.
Legend:
Granite
Limestone and Dolomite
Basalt
FIGURE 10-3 RANGES OF DISTRIBUTION COEFFICIENTS FOR VARIOUS ROCK TYPES
86
-------
TABLE 10-2
REPRESENTATIVE RETARDATION FACTORS
Fission products with no sorption
Iodine and technetiura
Fission products with sorption
Retardation Factor
Strontium
Cesium
100 (for dilute groundwater;
> 100 substantially smaller
for brines)
Actinides and daughter products
Thorium
Plutonium
Americium
Neptunium
Radium
100 - 100,000
Based on laboratory measurements of distribution coefficients and
applicable for modeling a generic repository.
Source: Adapted from NUREG/CR-0912.
87
-------
with sorption, and actinides and their daughter products. As there
was a considerable range of distribution coefficients listed for
each of the radionuclides within each group, representative
retardation factors, applicable to generic repository model
analyses, were tabulated.
Experimental evidence has shown that retardation factors are,
in part, functions of rock types, pH, water composition, and flow
rates, but the interrelationships among these parameters and their
effect on retardation are largely unknown. Until laboratory
experiments are correlated with in-situ conditions, considerable
uncertainty may be expected in the results of analyses of
radionuclide transport (Lawrence Livermore Laboratory, 1981).
88
-------
References
Landstrom, 0., C. Klockars, K. Holmberg, and S. Westerberg. In-sltu
Experiments on. Nuclide Migration in Fractured Crystalline Rocks.
Proc. Symposium on Science Underlying Radioactive Waste Management.
(Materials Research Society). 1979.
Lawrence Livermore Laboratory. Geoscience Data Base Handbook for
Modeling a Nuclear Waste Repository. Volume I. Prepared for the
U.S. Nuclear Regulatory Commission. January 1981. NUREG/CR-0912.
Muller, A. B., N. C. Finley and F. J. Pearson, Jr. Geochemical
Parameters Used in the Bedded Salt Reference Repository Risk
Assessment Methodology. Sandia National Laboratories. Prepared for
the U.S. Nuclear Regulatory Commission. September 1981.
SAND-81-0557; NUREG/CR 1996.
Robertson, J. and J. Barraclough. Radioactive and Chemical Waste
Transport in Groundwater at National Reactor Testing Station, Idaho:
20 Years Case History and Digital Model, in Underground Waste
Management and Artificial Recharge. (J. Brownstein, ed.) Am. Assoc.
Petrol. Geol. _!, p. 291-332. Minasha, Wisconsin, George Banta
Co., 1973.
Schneider, K. J. and A. M. Platt (eds.). High-Level Waste
Management Alternatives. BNWL 1900, Battelle Pacific Northwest
Laboratories, Richland, Washington. May 1974.
89
------- |