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

-------