EPA-520/3-74-
STORAGE OF LOW-LEVEL RADIOACTIVE WASTES
IN THE GROUND: HYDROGEOLOGIC AND
HYDROCHEMICAL FACTORS
with an Appendix on
The Maxey Flats, Kentucky, Radioactive >vaste
Storage Site: Current Knowledge and Data Needs
for a Quantitative Hydrogeologic Evaluation
U.S. ENVIRON MENTAL PROTECTION AGENCY
Oifict of Radiation Programs
•*"ap» df*** .
$$f
-------
STORAGE OF LOW-LEVEL RADIOACTIVE WASTES IN THE GROUND
HYDROGEOLOGIC AND HYDROCKEMICAL FACTORS
with an Appendix on
THE MAXEY FLATS, KENTUCKY, RADIOACTIVE
WASTE STORAGE SITE: CURRENT KNOWLEDGE
AND DATA NEEDS FOR A QUANTITATIVE
HYDROGEOLOGIC. EVALUATION
By
Stavros S. Papadopulos and Isaac J. Winograd
U.S. Geological Survey
Reston, Virginia
Prepared for the U. S. Environmental Protection Agency
Open-File Report 74-344
1974
-------
FOREWORD
The Office of Radiation Programs carries out a National Program
designed to evaluate the exposure of man to ionizing and nonionizing
radiation and to promote development of controls necessary to protect
the public health and safety and assure environmental quality.
Within the Office cf Radiation Programs,.problem areas have bsan
defined and assigned a priority in order to determine che level of
effort expended in each area. One of these, the waste management
problem area, has been assigned a high priority and requires the
participation and cooperation of several Federal agencies. This
report is directed at a specific Environmental Protection Agency task
of establishing action guidelines based on radiation exposure levels.
Other reports, recommendations, and State assistance will be developed
and executed to fulfill EPA obligations under the interagency agreement,
I encourage users of this report to inform the Office of Radiation
Programs of any omissions or errors. Your additional comments or
requests for further information are also solicited.
•'/n r,
W. D. Rowe, Ph.D.
Deputy Assistant Administrator
for Radiation Programs
111
-------
TABLE OF CONTENTS
Page
ABSTRACT [[[ 1
INTRODUCTION [[[ 3
POTENTIAL RELEASE MECHANISMS AND
HYDROGEOLOGIC SITING CRITERIA .......... .' .................... 5
MATHEMATICAL SIMULATION OF SOLUTE TRANSPORT IN
. HYDROGEOLOGIC SYSTEMS : A BRIEF REVIEW ........... ........... 9
HYDROGEOLOGIC AND HYDROCHEMICAL DATA NEEDS FOR
SITE EVALUATION .............................................. 17
CONCLUDING REMARKS ...................................... ...... 22
APPENDIX- The Maxey Flats, Kentucky, radioactive waste
storage site: Current knowledge and data needs for a
quantitative hydrogeologic evaluation ....................... 25
Current knowledge of site hydrogeology ...................... 25
Inherent obstacles to a quantitative evaluation
of Maxey Flats hydrogeology .............................. 33
Outline of data needs and methodology for an evaluation
of site hydrology ......................................... 35
Definition of ground-water flow system ................... 35
Determination of the magnitude of vertical and
horizontal leakage from the trenches ......... ......... 40
Sorption characteristics of the confining beds ........... 42
-------
LIST OF ILLUSTRATIONS
Page
Figure 1. Sketch map and geologic section,
Maxey Flats, Kentucky 26
VI
-------
CONVERSION FACTORS
Factors for converting English units to metric units are shown to
four significant figures. However, in the text the metric equivalents
are shown only to the number of significant figures consistent with
the values for the English units.
English
inches (in)
feet (ft)
miles (mi)
cubic feet (ft )
gallons (gal)
gallons per minute (gpm)
pounds per square inch
(psi)
Multiply by
25.40
0.3048
1.609
0.02832
3.785
0.06309
0.07031
Metric
millimetres (mm)
metres (m)
kilometres (km)
cubic metres (m )
litres (1)
litres per second (1/s)
kilograms (force)
per square centimetre
(kgf/cm-)
vii
-------
STORAGE OF LOW-LEVEL RADIOACTIVE WASTES IN THE GROUND:
HYDROGEOLOGIC AND HYDRO CHEMICAL FACTORS
ABSTRACT
Hydrogeologic cricaria presented by Cherry and others (1973) are
adopted as a guideline to define the hydrogeologic and hydrochaaicai
data needs for the evaluation of the suitability of proposed or
existing low-level radioactive waste burial sites. Evaluation of the
suitability of a site requires the prediction of flow patterns and of
rates of nuclide transport in the regional hydrogeologic system.
Such predictions can be made through mathematical simulation of flow
and solute transport in porous media. The status of mathematical
simulation techniques, as they apply to radioactive waste burial sites,
is briefly reviewed, and hydrogeologic and hydrochetnicai data needs
are. listed in order of increasing difficulty and cost of acquisition.
Predictive modeling, monitoring, and management of radionuclides
dissolved and transported by ground water, can best be done for sites ia
relatively simple hydrogeologic settings; namely, in unfaulted
relatively flat-lying strata of intermediate permeability such as silt,
siltstone and silty sandstone. In contrast, dense fractured or soluble
media, and poorly permeable porous media.(aquitards) are not suitable
for use as burial sites, first because of media heterogeneity and
difficulties of sampling, and consequently of predictive modeling,
-------
and sacond, because in humid zones burial trenches in aquicards may
overflow. A buffer zone several thousands of feet to perhaps several
miles around existing or proposed sites is a mandatory consequence of
the site selection criteria. As a specific example, the Ma:-;ey Flats,
Kentucky low-level waste disposal site is examined.
-2-
-------
INTRODUCTION
Low and intermediate-level-solid and liquid radioactive wastes
have been released to the 'Subsurface environment for over 2 decades
during which dozens of reports, several symposia, and at least one
short textbook (International Atomic Energy Agency, 1965) have been
published on this practice.
In this paper no claim or attempt is made to summarize the
voluminous literature on radioactive waste storage or disposal into
the ground. First, an excellent summary is available (International
Atomic Energy Agency, 1965). Second, conclusions based on even the
.most detailed of site studies are usually of only limited value for
evaluation of the waste-containing properties of another site with
differing hydrogeologic, pedologic or geochemical setting; that is, the
transfer value of such studies is limited. The objective of this
paper is to present a brief overview of some of the major hydrogeologic
and hydrochemical factors bearing on radioactive waste storage in, and
release from, the ground; special emphasis is placed on a review of the
state of the art of simulation of solute transport, and on complications
of data collection. As a specific example, the Maxey Flats, Kentucky
low-level burial site is examined in the Appendix.
-3-
-------
Some qualifications and definitions are needed at this point.
First, s. thorough evaluation of the potential hazard created by
emplacement of radioactive wastes into the ground would consider
critical nuclides, critical pathways, and critical population groups
(International Atomic Energy Agency, 1971). We consider herein only
those critical pathways suggested by hydrogeoiogic and, to a lesser
degree, geoaorphic factors. It is likely that burial tranches could
be designed which would contain-critical nuclides, for several hundred
years, barring destruction of the site by an earthquake. Existing
practices, however, are to bury the low-level wastes in unlined
trenches or pits, a practice necessitating attention to hydrogeoiogic
pathways of critical nuclide transport. Second, because of the
operational .and economic difficulties of determining a_ priori the
critical nuclides in the effluent from waste dumps, our definition of
low-level waste follows that of the US. Atomic Energy Commission (1967)
namely wastes with about 1 uCi (microcurie) per gallon or cubic foot
of waste. Lastly, although we restrict our discussion to shallow
burial of low-level wastes (as defined) we are aware that, in practice,
low-level wastes may, in fact, consist of any wastes not classified
as high-level and may even contain quantities of the long-lived and
extremely toxic transuranic radioelements.
-4-
-------
POTENTIAL RELEASE MECHANISMS AND KYDROGEOLOGIC SITING CRITERIA
Potential mechanisms through which critical radio elements in low-
level solid wastes may be released from a burial sice and introduced
into the hydrosphere, atmosphere or biosphere are: a) transport of
dissolved nuclides by water to wells, gaining streams, or springs; b)
transport upward to the soil zone by capillary flow followed by
concentration of the nuclides in plants; and c) exposure and overland
transport, by normal erosion processes (water and wind) , erosion .due to
floods, or erosion following disruption of landscapes by earthquakes.
The suitability of a site for shallow burial of low-level wastes,
therefore, .depends on the extent that its environs are capable of
preventing the occurrence of these release mechanisms. Criteria
for the evaluation of the suitability of a. site for land burial
operations have been presented in reports by Peckham and Belter (1962),
Richardson (1962a, 1962b) , Mawson and Russel (1971), and very recently
by Cherry and others (1973). The report by Cherry and others categorizes
burial sites as a) intermediate-terra sites, suitable for wastes
that decay to safe levels within several decades and for which
protection is mainly provided by engineered structures in which the
wastes are buried, and b) long-term sites, suitable for wastes
with longer life, .and which depend mainly on hydrogeologic conditions
for protection. Hydrogeologic-criteria are presented for both
types of sites. We found the hydrogeologic criteria that they
present to be very comprehensive, and we have adopted them as a
-5-
-------
guideline in defining che type of daca chat are needed for the
evaluation of burial sites. Of course, assessment of a site as suitable
for intermediate or long-term burial applies only to future sites.
Inasmuch as the type of waste buried in existing sites has not been
strictly controlled, these sites require treatment as long-term burial
sites, although all of them may not meet che criteria for such sites.
To provide che background for what follows, the hydrogeoiogic
criteria presented by Cherry and others (1973) are repeated here, with
slight modifications :
"
riteria for Intermediate -Term Burial Sites
(1) the land surface should be devoid of surface water }
except during snowmelt runoff and exceptional periods
of rainfall. In other words the sites should not be
located in [flood plain,]* swamps, bogs, or other types
of very wet [or potentially very wet] terrain.
(2) the burial zone should be separated -from fractured
bedrock by an interval of geologic deposits sufficient
to prevent migration of radionuclides into the fractured
zone-
Except in unusual circumstances the direction and rate of
ground-water flow as well as the retardation effects are ver
-------
(3) the predicted rats of radionuclide transport in the
shallow.... deposits at the site should be 3leu 2r.cu.ch to
provide "any years or decades of delay time before radio-
nuclides would be able to reach public waterways or any
other area which might be considered hazardous in the
biosphere. In other words considerable time would be
available for detection of contamination and for
application of remedial measures if necessary.
(4) the size should have sufficient depth ~c water table to
~ermi~ all burial operations to occur above the wa~er
table, or as an alternative the site should be suitable
for producing an adequate water-table dep~h by flew system
manipulation.
(5) the site should be well suited for effective monitoring
and for containment by flow-system manipulation schemes."
"Criteria for Long-Term Burial Sites
(1) the land should be generally devoid of surface water ar.d
be relatively stable geomorphically. In other words
erosion and weathering should not be proceeding at a rate
which could significantly affect the position and character
of the land surface during the next few hundred years.
(2) the subsurface flow pattern in the area must be such that
the flow lines from the burial zone do not lead to areas
considered to be particularly undesirable, such as
fractured bedrock, public waterways used by man, aquifers
used for water supply, etc.
(3) the predicted residence time of.radionuclides within an
acceptable part of the subsurf ace-flow system must be of
the order of several hundred years. Tne hudrogeologic
conditions must be simple enough for reliable residence-
time yredictior.s to be made, (underlining not in original
quote)
-7-
-------
(4) the natural water table should be bel-~J the burial zone
by at least several meters and the hydrogeologic setting
should be such that large water-table fluctuations are
very unlikely. This condition would provide additional
assurance that leaching of radionuclides would not occur
quickly in the event of corrosion of the waste containers
or in the event that low-level wastes are out direczly
As these criteria indicate, for a complete evaluation cf the
suitability of a site for land burial of wastes, it is necessary to
predict the flow patterns and the rate of transport of radionuclides
in the regional hydrogeologic system. Such predictions require che
simulation of the hydrogeologic system by a mathematical model that
describes the simultaneous transport of radioactive solutes and water
in the system. A brief review of the state of the art in solute
transport simulation is presented in the next section.
-8-
-------
MATHEMATICAL SIMULATION OF SOLUTE TRANSPORT
IN HYDROGEOLOGIC SYSTEMS: A 3RIE? REVIEW
The processes chat control the transport of solutes in a hydrc-
geologic system are: (a) convection by the moving fluid, (b) hydro-
dynamic dispersion, which combines the effects of mechanical dispersion
and molecular diffusion, and (c) chemical reactions which may taka plac =
between various solutes, between solutes and the solid matrix of the
system, and within the solute, namely radioactive decay. Mathematical
simulation of solute transport in a hydrogeologic system, therefore,
requires the simultaneous solution of the differential equations that
describe the movement of the fluid in the system and the transport
of each of the solutes in the fluid by one or more of the processes
stated above. The simulation can be carried out, provided that the
boundary conditions and the parameters defining the system and the
processes that take place within the system are known and provided that
the resulting equations are solvable by known mathematical techniques.
Simulation of fluid flow through porous media has received
considerable attention in the past, and both analog and digital methods
have been developed and extensively applied to analyze actual field
problems. However, because of the difficulties in solving the'solute
transport equation and because of a lack of understanding of all the
types and rates of chemical reactions that may take place as a solute
-9-
-------
a
9
10-
12
13
20 —
moves through a porous medium, until recently little progress had
been, made in simulating solute transport. Only relatively simple
.problems could be handled by analytical techniques, and the finite
difference methods, which were so successful in simulating fluid tlow,
did not always provide satisfactory solutions to the solute transport
equation. Development of more accurate numerical methods, however,
such as the method of characteristics (Gardner and others, 1S64)
and the Galerkin method (Price and others, 1968; Aziz and others,
!1968), has recently made possible the solution of the solute transport
equation and its application to actual field problems under certain
conditions. Progress has also been made in the area of chemical
reactions, and the transport of solutes with equilibrium controlled
ion exchange type reactions has been simulated. Research is continuing
in this area, and it is expected, that solute transport with other
types of reactions will be possible to simulate in the near future.
A few selected examples of recent articles on solute transport that
are indicative of the present state of the art are presented below.
: In unsaturated systems, Bresler and Hanks (1969) considered the
one-dimensional (vertical) transient convective transport of a
:nonreacting solute. Later, Bresler (1973) extended the solution
' to this problem to include ionic diffusion and mechanical dispersion
(hydrodynamic dispersion) by using a high-order finite difference
; method.
_•,/-, I', i. GOVERNMENT HRINTtNG OFFICE: 19«» 0- Illl'l
•^ H7-IC
-------
3
11
In saturated systems, assuming two-dimensional transient flow,
2 ipinder and Cooper (1970) and, independently, Reddell and Sunada (1970) j
' used the method of characteristics to study salt-water intrusion in .
an aquifer. Bredehoeft and Finder (1973) used the same method to
•:~ simulate aquifer contamination in Brunswick, Georgia, and Finder
3 (1973) used the Galerkin method to simulate aquifer contamination in
Long Island, New York. In these problems the solute was assumed to be
.conservative (nonreacting), and transport by convection and hydro-
3
.dynamic dispersion was considered.
Lai and Jurinak (1971) used finite-difference approximations to
solve the one-dimensional solute transport equation for a.single
12 component, under steady fluid flow conditions, including the effects •
i
of ion exchange controlled by local equilibrium. Rubin and James
" (1973) also assumed one-dimensional steady water flow and local
13~ chemical equilibrium and applied the Galerkin method to cases involving
16 .(1) homogeneous or layered systems, (2) exchange reactions with
constant or concentration dependent selectivity coefficients, (3)
•binary or multi-component exchange, and (4) systems in which one of the
19 ;exchanging ions is also involved in a precipitation-dissolution
"'" reaction. Robertson and Barraclough (1973) used the method of
i: :characteristics model developed by Bredehoeft and Finder (1973) and
!considered radioactive decay and instantaneous-equilibrium, linear-
iisotherm type reversible sorption to simulate the two-dimensional
I transport of radioactive wastes in the Snake Plain aquifer at the
25 —
I
_ 1 1 — f. S. GOVERNMENT PRINTING OFFICE : I?" O • 'Iliri
23
24
• ICO
-------
9 . i iO ;
5-
10-
Natior.al Reactor Tasting Station, Idaho. j
Therefore, with the present state of the art of numerical simulation
of solute transport, the zovement of radionuclides leached from
shallow land burial sites could be simulated by assuming that: (a) '
the flow is one-dimensional (vertical) in the unsaturated zone and tvc-
dirsensional (horizontal) in the saturated zone; (b) isothermal
conditions prevail; (c) the solute transfer processes that take place
-are convective transport, hydrodynamic dispersion, radioactive decay, ,
and instantaneous-equilibrium controlled reversible sorption; and ;
(d) density variations due to changes in solute concentration are ;
'negligible. Under these assumptions the equations applicable to
transient transport of radionuclides, formulated on the basis of the ;
references cited above, are given below. j
In the unsaturated zone: i
15-
and
23
w(z,t)c; - k.
_1 9_ u- *• COVERMMSNT PSINTINC OFFICS : : J1» 0 •
-------
vhere
;
3 = volumetric water content of the nediuzi, dimensicniess;
1
t = time, T;
a = - K(9) $L = vertical flux of solution, LT ";
- . * 2 O Z
T 2 = vertical space coordinate, L;
K(6) = hvdraulic conductivity of aediur: (a function of 5)
LT-1;
h = hydraulic head, L;
y = ..£ W.6(2-2.) = rate of source terms, T ;
i: m = number of sources;
53 W. = flux due- to source j located at 2. , LT ;
M 0 = Dirac delta function;
Li C. = mass of dissolved form of nuclide i per unit volume
1 of solution, ML~3;
15— _
C. = mass of sorbed form of nuclide i per unit mass of
1S the porous medium, dimensionless;
p = bulk density of the medium, ML ;
13 • D = d q .+ D, (9)= effective coefficient of vertical
2 2 2 d 2-1
hydrodynamic dispersion, L T ;
d = vertical dispersivity coefficient, L;
71 j — ' Z
D,(9) = molecular diffusion coefficient (a function of 3)
j;
C! = mass of nuclide i per.unit volume of solucion in
r- i Che source fluid, ML~ ;
2~ \ i = ia 2/(tO-= radioactive decay rate of nuclide i,T
_-- i (t,j.= half life of nuclide i,T.
I'. S. aOVEHKMSMT PHI^TINC OFFICE : I»i? O - HUM
-------
In the saturated zone:
TV, q
3 IT ---£--3?
ar.d
>3C_. 3C.
~ (bsC.-l-bO C.) =1- (D ~ ~ D -r^
at i a i ax xxj x xv iv
¥ (Dyy "57 * V^} ' I* (qxci} - I? (qyci}
- bsyaC. -- + Q(x,y,t) C'. - k beC.
J. O L i 1
where S = storage coefficient of medium, dimensionless ; ;
x,y * space coordinates along principal axes of transtnissivity,,
L; " ;
A t ->t^
q ,q = - T -T— , - T -r— = fluxes of solution in the x and y i
'^ ' yy3y
direction, respectively, L T ;
T ,T = K b,K b = transmissivities of the medium in the
xx yy xx yy
2 -1
x and y direction, respectively, L T ;
K ,K * hydraulic conductivities of the medium in the x and
XX W 1
y* y direction, respectively, LT~X;
b = saturated thickness of the medium (equal to h in
unconfined systems), L;
m _]_
Q =.Z1 Q.6 (x-x ) 5(y-y.) rate of source terms, LT ;
I'. S. COVERNMiiNT HHINTWC OFFICE : I"' O - jili:!
M,.,oo
-------
-• 1
2 • Q. = flux due to source j, located at x y.,L"T ^;
3 £ = porosity of medium, dimensicnless;
D D ,D ,D = components of the effective hydrodynamic
xx' yy xy yx
5- dispersion coefficient for the entire saturated
thickness of the medium (see below); L T"^;
Y = unit weight of solution, ML T ";
3 a = compressibility of the medium, M LT~.
9
All other symbols are as defined previously. Following Scheidegger
10- .
(1961) and Finder (1973), the hydrodynamic dispersion coefficients can
1!
be expressed as
12
q 2 q.2
13 D = DT -S- + D_ -L- 4- D T
xx L q2 T q2 d
q 2 q 2
VDt^- +DTT- + V
xy rx = L T 2 i
Q •
'where D » d q = effective longitudinal dispersion coefficient,L T ;'
•
;Q_. DT = d_q = effective transverse dispersion coefficient, L T ; ;
?i : o . 3 effective molecular diffusion coefficient,L T ;
d . ;
-'2 q = (q2 -t- q2) = magnitude of &ux, L T~ ; ;
x y i
dT,dT = dispersivity coefficients, L; j
4*~ L T I
-- - = cortuosity,' dimensionless.
U. S. COVSaSMENT VHtNTlNG OFFICE : l«" O • JIllT!
i«? • \Q
-------
For equilibrium reactions with linear adsorption isotherms, the
concentration C. of the sorbed-form appearing in the solute transport
equations can be expressed in terms of the concentration C. of the
dissolved form through use of the distribution coefficient K,
(Thompkins and Mayer, 1947; Kaufman, 1963), which is defined as
K. = C./C., dimension lAf1.
a 11
The distribution coefficient K, is a lumped parameter which depends
on many variables, including moisture content in the case of unsaturatad
flow. This parameter will be further discussed in the section that
follows.
-16-
-------
HYDROGEOLOGIC AND HYDROCHE-1ICAL
DATA NEEDS FOR SITE EVALUATION
The types of hydrogeologic and hydrocheaical data needed co
deter-ine whether proposed or existing sites meet the criteria seated
earlier are listed below in•approximate order of increasing difficulty
and (or) cost of acquisition:
1. .Depth to water table, including perched water tables, if
present.
2. Distance to nearest points of ground water, spring water or
surface water usage (Includes well and spring inventory).
3. Ratio of pan evaporation to precipitation minus runoff (by
month for period of at least 2 years).
4. Water table contour map.
5. Magnitude of annual water table fluctuation.
6. Stratigraphy and structure to base of shallowest confined
aquifer.
7. Baseflow data on perennial streams traversing or adjacent to
storage site.
8. Chemistry of water in aquifers and aquitards and of leachate
• from the waste trenches.
9. Laboratory measurements of hydraulic conductivity, effective
porosity, and mineralogy of core and grab samples (from trenches)
of each lithology in unsaturated and saturated (to base of
-17-
-------
shallowest confined aquifer) zone—hydraulic coneuccivi^y to
be measured at different water contents and suctions.
10. Neutron moisture meter measurements of moisture content of
unsaturated zone. Measurements co be made in especially-
constructed holes; at lease 2 years' record needed.
11. In situ measurements of soil moisture tension in upper 15-30
feet (4.5-9,0 IT.) of unsaturatsd zone; at lease 2 years'
record necessary.
12. Three-dimensional distribution of head in all saturated
hydrostratigraphic units to base of shallowest confined aquifer,
13. Pumping,, bailing, or slug tests to determine transmissi'/ity
and storage coefficients.
14. Definition of recharge and discharge areas for unconfined and
shallowest confined aquifers.
15. Field measurements of dispersivity coefficients.
16. Laboratory and field determination of the distribution
coefficient (K,) for movement of critical nuclides through all
hydrostratigraphic units.
17. Rates of denudation and (or) slope retreat.
These data are necessary for a complete definition of flow and
nuclide transport through both the unsaturated and saturated zones. To
obtain such information, exploration costs could total several tens to
several hundreds of thousands of dollars, dependent upon depth to water
-------
cable and complexity of the flow system-. However, nor all Che outlined
information is likely co be needed at all sites. For example, in the
arid or semi-arid zones, daca types 1 and 3 may strongly indicate the
absence of any nee flux of water in the unsaturatsd zone below a depth
of a few feet (Wir.ograd, 1974) and, hence, the probable suitability cf
the sice for storage of solidified wastes, at least in the shore term,
say less than 100 years. Should some liquid wastes also be placed in
the. same trench, data of the type listed in items 1, 9, 10 and 11
.might permit an estimate of the volume of liquid waste that, the un-
saturated zone night accept before flow reaches the water table. Even
in sub-humid terrane, the presence of interbedded fine and well-sorted
coarse-grained sediments within the unsaturated zone might, when
coupled with a moderate depth to water table (say 30 feet or 9 m),
preclude natural recharge during all but the wettest of years
(Winograd, 1974). As another example, field measurements of hydraulic
conductivity and dispersivity at a storage site excavated in a thick
aquitard, say a 60-foot (18 m) thick glacial till, is impractical, as
might be the attempt to measure the three-dimensional distribution of
heads. In such a medium, perhaps only a long-tenn monitoring may
permit determination of velocity, flow direction, and dispersion.
To determine the extent to which sorption processes retard the
movement of a given critical nuclide, the four principal types of
geochemical information necessary are: a) the mineralogy (by size
fraction including colloidal materials) of all hydrostratigraphic units
-19-
-------
traversed by the water, including unics in Che ur.saturated zone; b)
chemiscry of water in aquifers and aquitards; c) chemistry of
representative samples of leachate from the waste field; and d)
laboratory experiments to determine the distribution coefficients
(K ) for the critical nuclides in representative samples of each
hydrostratigraphic unit.
In some of the qualitative literature on waste disposal a
panacealike aura surrounds the term ion-exchanga. In such literature
it is implied that when all else fails, ion-exchange processes will
prevent movement of contaminants to points of water use. A few caveats
are in order. First, there are several sorption processes, of which
ion-exchange is but one. Some processes will capture and "fix" or hold
a given nuclide "irreversibly" within the crystal lattice of a clay
mineral; but other processes provide only a temporary home for
undesirable, elements, which after a while reappear in solution
(Tamura, 1972). For this reason the parameter K,, or distribution
coefficient, is a markedly superior means with which to evaluate
nuclide transport through a given medium. This coefficient is the ratio
of activity, concentration or mass of a sorbed nuclide per unit mass of
solid to the activity or concentration of dissolved nuclide per unit
volume of water (Thomp'tcins and Mayer, 1947; Kaufman, 1963). Explicit
in the definition is the statement that sorption of a nuclide from waste-
bearing water is incomplete; some fraction will always reside in the
-20-
-------
water. Second, as is ion-exchange, K, is a lumped parameter — one cha:
2
is a function of many variables, including pH, radox pocential,
T
solucs chemistry, concentration of tha nuclides, chemistry of leachace,
•mineralogy (including solid solution), and, in the unsaturated ;one,
moisture content. For a given mineralogy it is also a function of
surface area cf the porous medium. Hence, predictions of nuciids
travel based on laboratory measurements of X, are valid only to the
u
degree to which field conditions are duplicated. A major change in
9
leachate chemistry may, in some cases, result in an order of magnitude
10-
decrease in the K, determined in the laboratory (Tamura, 1972, fig. 1).
< i
Likewise, extrapolation of K, values from one area to another is
12 d
valid only if the.-natural water chemistry and fine grained alunino-
12
silicate and (or) oxyhydroxide mineralogy in both areas is similar.
L-i
Third, sorption is-an important retardant of nuclide movement,
principally in rocks with intergranular or intercrystalline porosity,—
rocks with considerable surface area. Sorption is considerably less
effective in fractured or cavernous media due to minimal surface
13 ; ,
'area and maximum velocity (that is, minimal contact time with minerals '
19 :
on fracture or cavern walls). Yet, even'granular media, such as till,
- contain open fractures (Cherry, and others, 1973; Williams and
21 :
Farvolden, 1967);- where such fractures are present, the predictive
•2 '
value of K, data determined from core samples may be marginal. :
23 . '
-91 -
^-L ;'. •••• COVEKMiENT i• H[NT:.\O OFFICE : !>'•! <
-------
CONCLUDING REMARKS
In our brief overview we have considered the hydrogeologic, and
co a lesser excenc the hydrochemical, aspects of subsurface movement
of radionuclides dissolved from low-level wastes buried at shallow
depths. The reader is reminded that a complete safety analysis of a
low-level radioactive burial site would also have to explicitly con-
sider the following matters: a) introduction of radionuclides co
the atmosphere and surface water through long-term erosion and cata-
strophic erosion due to floods and earthquakes; b) uptake of radio-
nuclides from the soil zone by plants; c) identification of critical
nuclides within, and of the critical population.group in the vicinity
of, proposed burial sites (International Atomic Energy Agency, 1971);
d) long term monitoring of the. site to prevent vandalism and
blundering by unaware descendants; and e) methods of trench construc-
tion and waste emplacement designed to reduce or exclude entry cf
water into the trenches.
Within the hydrogeologic domain of concern here, it is emphasized
that predictive modeling, monitoring, and management of dissolved
nuclides are most effective if wastes are emplaced in relatively simple
hydrogeologic settings; namely, in unfaulted, relatively flat-lying
strata of intermediate permeability (for example, silt, silts tone, and
silty sandstone) in a region of low relief. In such a setting, a
modest amount of background hydrogeologic and hydrochemical data of
the types outlined in the previous section of this report may permit
modeling of the rates and directions of movement of selected critical
-22-
-------
radionuclides prior to a commitment Co use of the site. In contrast.
fractured or soluble media and aquitards utilized as burial sites are
difficult to monitor, sample, and model. Moreo%rer, in. humid zor.es,
sites in aquitards may overflow. That is, due to the considerably
higher permeability 'of the trenches than the surrounding strata, the
trenches act as infiltration galleries, fill with' precipitation, and
may spill over through seeps at the trench-fill contacts. The hydro-
geologic suitability of existing sites in aquitards can perhaps bes:
be studied, ex post facto, by long term (one to several decades)
monitoring of nuclide movement through sampling of water and cores
from carefully designed and constructed test wells. Such long tarm
records of nuclide distribution may then be used to empirically predict
future rates and directions of nuclide travel.
A buffer zone several thousands of feet to several miles around new
and existing sites is a mandatory consequence of the site selection
criteria outlined in the section, "Potential release mechanisms and
hydrogeologic siting criteria." The minimum width of the buffer zone
would be governed at each prospective site by the calculated length
of ground-water flow path needed to permit decay to safe levels of the
identified critical nuclides. A buffer zone 20-50 percent greatar
than that calculated is. probably advisable, so that: a) a safety factor
is provided in the event that unknown aquifer heterogeneities result
in longitudinal dispersion considerably greater than .calculated; and
-23-
-------
b) construction of additional disposal sites adjacent co the original
site is possible should the predictions ot solute movement prove
conservative.
Finally, the commonly ;nade distinction between storage versus
disposal of radioactive wastes appears questionable for low-level
wastes. This distinction, though applicable' to problems concer-in;;
the handling of the relatively small volume of high-level rcdioacti"-:
wastes (see for example, Winograd, 1974), does not appear realistic
for the relatively large volumes of low-level waste production projected
for the years 1976-80, 1981-90, and 1991-2000, namely 3.6 X 10°,
/• s t*
14.5 X 10 , and 79 X 10° ft per year (O'Connell and Ho Iconic, 1974).
It appears improbable on economic, logistical, or radiological safety
grounds, that the contents of an existing low-level waste site would
be exhumed even in the event of extensive leakage of contaminated
ground water from the site. Hence, in reality, the term "waste disposal"
may be more correct for shallow burial of low-level wastes than the
term "waste storage". Granting that such emplacement is to all intents
and purposes permanent (in a human time frame) it becomes mandatory
that: a) intensive hydrogeologic and hydrochemical studies precede
choice of all new sites and that they be completed at any existing, but
heretofore un-studied, sites; and b) that intensive radiochetnical
monitoring of ground and surface waters be done during and after the
lifetimes of all operational sites.
-24-
-------
APPENDIX
THE MAXEY FLATS, KENTUCKY, RADIOACTIVE
WASTE STORAGE SITE: CURRENT KNOWLEDGE
AND DATA NEEDS FOR A QUANTITATIVE
HYDROGEOLOGIC EVALUATION
Current Knowledge of Site Hydrogeology
Maxey Flats * a low-level radioactive wasce storage sice
administered and monitored by the.State of.Kentucky and operated by
Nuclear Engineering Co. (NECO) is located atop a dissected plateau in
Fleming County near Morehead, Kentucky. The site, approximately
2,500 feet (760 .m ) long and 1,200 feet (370 m) wide, is bounded by
scarps and steep.slopes on the east, south and west (fig. 1). The
upland is about 250-350 feet (76-110 m) above the surrounding valleys
(locally called hollows), which contain small perennial creeks.
Average annual precipitation is 46 inches (1,170 mm). Storage of
wastes began in May 1963, and about 2.5 million cubic feec (70,000 m )
of radioactive wastes have been buried in trenches through 1971. An
inventory of type and quantity of wastes buried is available in a report
by Clark (1973).
The stratigraphy of the site and surrounding regions is wail known
and clearly depicted on a 1:24,000 geologic map (McDowell, Peck and
"Mytton, 1971). The site is underlain by gently dipping (25 feet per
-25-
-------
33°35'
12^J. i/J
" /2J
•^J I 1 *3 si C '
333I5'
•^1000
Noncy Member of the Sorden Formarion
Formers Member sf (he Sordsn Formation
and Sunbury Shale
Bedford Shale
Ohio Shale
Crab Orchard Formation
ond older rocks
2000 «000 Feet
I i 1 Vertical Eiaggerotion 4X
1000 Metres
Figure 1. Sketch map and geologic section, Maxey Flats, Kentucky
(base map, USGS 1:24,000 Plummers Landing quadrangle;
geology simplified after McDowell and others, 1971).
-26-
-------
alia of 4.7 m/km co the east-southeast) Silurian, Devonian, and
Mississ.ippian rocks consisting chiefly of fissile carbonaceous shale,
clay-shale, and some siltstone and sandstone. These rocks comprise -he
following formations, listed in order of. decreasing age: Crab Orchard
Formation (upper part, 30-130 ft or 24-40 a thick); Ohio Shale (150-220
ft or 46-57 tn chick); Bedford Shale (10-40 fc or 3-12 -a chick); Sunbury
Shale (15-20 ft or 5-6 a thick); and the Farmers (33-95 ft or 10-29 m
thick) -and Nancy (155-195 ft or 47-59 m thick) Members of the Borden
'Formation. At Maxey Flats only the lower 40 feet (12 m) of the Nancy
Member is present. The trenches are entirely within the Nancy Member.
The hydrogeology of the site is poorly understood. A summary of
several reconnaissances (Hopkins, written commun. 1962; Walker, 1962;
and Whitman, written commun. 1971) and observations made during a 2-day
traverse of the site by I.J. Winograd follows.
All the cited stratigraphic units are aquitards, and only one of
these has been hydraulically tested at the storage site. The Ohio Shale
is probably the most transmissive of the strata beneath the site
by virtue of its highly fissile nature, general absence of interbedded
clay-shales (some which are slightly to very plastic when wet), and
great thickness (150-200 ft or 46 to 67 m). In areas adjacent to the
site, this formation reportedly yields 100 to 500 gallons (400-2000
litres) per day to wells, particularly those dug or drilled in valleys
(Hall.and Palmquist, 1960); water quality is marginal. The Crab
-27-
-------
Orchard Formation, in contrast,, consists predominantly of. expansive
clay-shale, reported to be very plastic when wet, and to flow and
slump into excavations (McDowell, Peck and Mytton, 1971; Dobrovclny
and Morris, 1965); open fractures are therefore unlikely in this
formation within the zone of saturation, and it may be the least
permeable of the aquitards beneath the site. The Crab Orchard Formation
apparently controls the occurrence of the numerous springs and seeps
reportedly emerging from the base of the overlying Ohio Shale
(Dobrovolny and Morris, 1965); such control, if documented by a
detailed inventory of springs at and around the site, suggests that
the Crab Orchard, by virtue of its stratigraphic position, tnay be
considered the "hydraulic basement" for significant movement of ground
water in the region.
The transmissivity of the remaining four formations or members
is probably intermediate between that of the Crab Orchard and the
Ohio Shale. Sandstone and siltstone near the base of the Nancy Member
and in the Farmers Member of the Borden Formation contain open joints
and bedding planes in outcrop and in exposures. But these competent
strata are interbedded with clay-shales, which tnay retard and control
the vertical movement of water beneath Maxey Flats. Open bedding planes
observable in the Farmers Member along the eastern scarp of Maxey Flats
appear to be due to removal, by mechanical weathering, of the soft
interbedded clay-shales, which are in place a few yards in from the
outcrop face.
-28-
-------
The highly fissile nature of the Ohio and Sunbury Shales and the
high degree of induration of silts tones and sandstones of the Nancy
and Fanners Members of the Borden Formation suggest that flow through
these strata occurs principally via secondary openings, namely joints
and planes of fissility. Whether water flow in the clay-shales of
the Nancy and Farmers Mashers and the Bedford Shale also occurs
principally through fractures rather than intersticas is an open
question; some of these clay shales are plastic when wet (particularly
those in the Bedford Shale) and may contain no open fractures when
saturated.
A well inventory and test drilling in 1962 (Hopkins, written
comraun. 1962; Walker, 1962) suggest that many shallow dug wells
tapping the basal Nancy Member on the Flats are actually cisterns
periodically recharged with water from the soil zone. Pressure
injection tests were conducted by Walker (1962) on several 40 co 50
foot (12-15 m) test holes at the NECO site. Assuming that no leakage
occurred around the packers, the results of these tests indicate that
injection rates ranged from 0.00002 to 0.005 gallons per minute per
foot of injected interval per pound per square inch of injection
pressure (0.00006-0.015 [(l/s)/m]/(kgf/cm )); the median value for
10 tests was 0.0002 gallons per minute per foot per pound per square
7
inch .(0-0006[(1/s)/m]/(kgf/cm )). After drilling, but prior to the
pressure injection testing in 1962, water was not detected in any of
the test holes. The reported absence of a measurable water level in
the test holes may support the cistern concept or may simply reflect
- 29 -
-------
inadequate time for recovery of water level after completion of the
drilling. In aquitards such recovery may take hours to days
(W-inograd, 1970) even in holes drilled by the air-rotary or cable
tool methods. The drill holes contained water when measured by
H. Hopkins in October 1963.
Eased on water level data from shallow dug and drilled wells,
from two deep wells, and upon hydrogeologic inference, the following
model of ground-water flow may be postulated for the Maxey Flats
area. Ground water in the soil zone is perched above the poorly
permeable Nancy Member of the Borden Formation (Hopkins, written
commun. 1962). It is this water that supplies shallow dug wells on
Maxey Flats. This perched water table (altitude 1,035-1,055 ft, or
315-322 m) slopes southeastward, paralleling the regional dip of the
21
Nancy Member (Hopkins, written commun. 1962)— . Water levels in the
two "deep" wells (reportedly drilled to 110 and 165 feet or 34 and 50 tr
but only open to 59 and 64 feet or 18 and 20 m, at the time of
measurement in 1962) were as much as 30 feet (9 m) lower than the
perched water table in 1962. The lower levels may represent a lower
water table, as suggested by Hopkins (written commun. 1962), or be
evidence of decreasing heads with depth, namely suggestive of vertical
ground-water flow within the Nancy Member.
— In October 1963 Hopkins also measured water levels in 1 of the
8 test holes drilled at the site. Water level altitudes in these
40-50 foot (12-15 m) holes range from 1,025-1,040 feet (312-317 m)
with depths to water of 6-20 feet (2-6 m). Interpretation of these
water levels will require detailed study of waste burial operations
during the months preceding the measurements.
-30-
-------
The emergence of springs and seeps from the Ohio Shale around the
periphery of Maxsy Flats and adjacent regions (McDowell, Peck, and
Myccon, 1971) indicate the presence of a water table in this formation,
but its altitude and configuration beneath the Flats or adjacent
areas is unknown. Whether' any perched zone(s) of saturation exists
between the perched water in the soil zone and the postulated water
table in the Ohio Shale is conjectural. Whitman (1971, written
commun.) postulated such a zone within siltstones of the Fanners
Member of the Borden Formation. A perched zone of saturation con-
ceivably might also exist in the Sunbury Shale, above the reportedly
expansive clay-shales of the Bedford Shale. Also unknown is the magni-
tude of interflow, namely ground-water flow within the relatively
permeable soil and colluvium blanketing parts of the slopes
surrounding the Flats.
To a first approximation, the perennial creeks west and northwest
.(Drip Springs Hollow), east and southeast (an unnamed creek), and
south (Rock Lick Creek) of the site (fig. 1) comprise the hydrologic
boundaries of the local ground-water flow system. The first two
creeks reportedly are perennial except for a 2- to 4-week period each
year; water persists in depressions in the creeks during periods of no
flow (Hopkins, written commun. 1962). The near-perennial nature of
these creeks suggests, in this area of probably low infiltration rates
and high runoff, that the creek discharge is primarily baseflow
from the Ohio Shale and (or) the Crab Orchard Formation-and possibly
from the colluvium blanketing parts of the slopes around the site.
-31-
-------
The discharge in Drip Springs Hollow was about 1.7 cubic feet per
second (0.043m /s) on August 1, 1962, and that in the unnamed hollow
east of Maxey Flats, 2.2 cubic feet per second (0.062 a /s) on the
same date; both creeks were dry near their heads (Hopkins, written
comnun. 1962).
In summary, the stratigraphy of Maxey Flats is well known and its
local hydrologic boundaries approximately defined. ?Ic data are
available, however, on the depth to or configuration of the water
table(s), head changes with depth within the zone of saturation,
transmissivities, or porosities of any of the strata older than the
Nancy Member of the Borden Formation; and data for the Nancy Member
are marginal. The absence of substantive hydrogeologic data reflects,
in part, the difficulties of obtaining hydraulic data from wells
penetrating only aquitards and, in part, the absence of wells or test
holes of proper construction tapping units older than the Nancy
Member, either at or in the immediate vicinity of the site.
-32-
-------
Inherent Obstacles to a Quantitative
Evaluation of the Maxey Flats Hydrogeology
An outline of data needs and methodology for a quantitative
evaluation of the hydrogeology of Maxey Flats is presented in the next
section. Here we list several caveats which merit careful considera-
tion prior to a commitment to a costly subsurface exploration program
of the type outlined below. The study of and predictions of nuclide
movement away from potential waste storage sites can be expedited for
sites with relatively simple hydrogeology namely, sites underlain by
non-fractured aquifers of moderate permeability. The strata beneath
Maxey Flats, by contrast, are aquitards whose fractured nature- and
low permeability constitute major obstacles to successful quantitative
definition of the hydrogeology of the region, regardless of funding
levels. The low permeability of these strata constitutes an obstacle
in two ways. First, determination of static water levels of, or the
obtainment of representative samples of formation water from, selected
depth intervals in the a'quitards may take days or possibly weeks.
Second, determination of transmissivities by pumping may not be
feasible, necessitating use of less satisfactory' techniques, such as
slug injection.testing. (See below.) Similarly, the low trans-
missivities may preclude utilization of two-well tracer'tests for
in-situ measurements of aquifer dispersivity and distribution
coefficients.
-33-
-------
The high probability of ground-water flow through fractures,
particularly in the Ohio Shale, the Sunbury Shale, and the Earners
Member of the-Borden Formation, constitutes the second major obstacle.
First, the fractures introduce a measure of heterogeneity that nay
not readily be handled using che assumptions of flow through inter-
granular porous media. Ground-water velocities, computed using such
an assumption, aay be two or more orders of magnitude less than, the
maximum velocity. Second, data from two-well tracer tests are
inherently difficult to interpret in fractured media, and the transfer
value of such data is uncertain.
In view of the above considerations it might be preferable that the
detailed hydrogeologic subsurface exploration program outlined in the
next section is implemented in several phases. If initial phases of
the drilling'and testing program support our reservations on the
possibility of a meaningful quantitative evaluation of the hydrogeology
of the site, the program could be discontinued. In this event,
consideration could be given to preventive measures that would:
a) retard or exclude entry of water into the trenches for decades;
fa) retard movement of radionuclides out of the trenches even in the
presence of ponded water; and c) retard exhumation of the wastes
by erosion for decades. Such measures might, for example, include:
a) placement of a concrete and (or) asphalt cover over all filled
trenches; b) stratification of the wastes, in active and future
trenches, with layers of soil or crushed shale, and c) exclusion
-34-
-------
from future burial of wastes containing the long-lived transuranic
radioelenients, such as plutonium and. arnericium.
Outline of Data Needs and Methodology
for an Evaluation of Site Hydrology
Studies necessary for an evaluation of the hydrogeology of Maxey
'Flats, as it relates to modeling and monitoring of the movement of
dissolved radionuciides from the trenches to off-sice areas, fall
into three general categories: a) definition of. the local and regional
ground-water flow system; b) determination of magnitude of vertical and
lateral leakage of ground water from the trenches; and c) measurement
of the degree of retardation of specific nuclides during 'their
migration through the aquitards underlying the site. The types of data
necessary and methodology commonly used to obtain such data are out-
lined in this section, but such an outline should not be construed as
being fixed. It is expectable that, in practice, the methodology will
have to be tailored in varying degrees to meet the hydrogeologic con-
ditions at the Maxey Flats site.
Definition of Ground-water Flow System
Definition of the local and regional ground-water flow systems,
specifically the rates and directions of ground-water movement beneath
and along the periphery of the site, is a prerequisite for any mass-
transport modeling of nuclide movement. Definition of the flow systems
requires information on the head distribution in and the
-35-
-------
transnissivity and porosity of each water-bearing strat'Lgraphic unit
beneath and in the region surrounding the site.
Head distribution within a water-bearing unit is commonly deter-
mined using: a) piezometers (namely cased and cemented wells open to
the foroiatioa tapped only at the bottom) drilled to different depths;
b) cased and cemented wells of differing depth open to the formation
only opposite the bottom few feet to tens of feet of hole; and c)
uncased holes completely penetrating the aquifer or aquitard, but in
which selected zones are isolated for head measurement using inflatable
packers. Because of the probable sparsity of water-bearing fractures
in the aquitards beneath Maxey Flats, and the probable low perme-
ability of such fractures, method "b" appears well suited for use at
this site.
Several groups or nests of wells will probably be needed to define
the three-dimensional distribution of head. Data obtained from the
first group would determine that needed from the second group, etc.
The first well in a group of six might, for example, be completed to
determine the head in the first detectable water-bearing fracture in
the Nancy or Farmers Member of the Borden Formation; such a well would
be drilled, cased and cemented to the top of the members and then air-
rotary drilled deeper till water entered the bore. An adjacent well
of similar construction might test the water-bearing fractures in
fissile shales of the Sunbury Shale. Successive wells might test the
upper, middle and basal parts of the relatively thick fissile shales
of the Ohio Shale and the clay-shales in the underlying Crab Orchard
.-36-
-------
Formation. Perhaps as many as four such groups or nests of wells
might be needed at the NECO site.
After the head has been determined in each isolated zone, swabbing
to obtain water samples may be feasible in the most permeable fractured
intervals, as well as determination of permeability via slug injection
tests. (See below.)
Determination of the head distribution in the aquitards will
probably be extremely time-consuming, owing to the .low permeability of
these strata and to the influence of the drilling itself (Winograd,
1970). It is, therefore, necessary that modern air-rotary drilling,
cementing, and testing equipment be used and that the entire operation
be under the supervision of an experienced hydrogeologist. Rotary
drilling with water or mud or drilling with cable tools would not be
practical.
Transmissivity may be determined by means of pumping or injection
tests of selected zones. Determination of transmissivity by pump
testing may not be feasible in the aquitards because they may not
yield even a few gallons a minute to a pump for periods in excess of
a few minutes to tens of minutes. Accordingly, a series of slug
injection tests may be necessary; methodology and interpretation of
slug testing are discussed by Cooper and others, 1967; Elankennagel,
1967, 1963; Papadopulos, and others, 1973. As mentioned previously,
slug tests can be made in the nests of wells drilled for head deter-
mination. However, because such holes are designed to sample only a.
fraction of the thickness of each aquitard, tests in these holes, may
-37-
-------
be adequate only for the thinner aquitards. Determination of the
transmissivity of the Ohio Shale, Crab Orchard Formation, and perhaps
the Farmers Member' of the Borden Formation will require several
additional holes penetrating the entire thickness of these aquitards.
Assuming the water table divide in the Ohio Shale parallels the
long axis of the NECO site, that baseflcw data from the creeks
bounding the site (on east and west) are available, and that wacar in
this shale is the principal source of the baseflow, then the gross
transmissivity of this formation might also be determined using the
drain method of Jacob (1943).
As in all fractured media, field determination of porosity will be
extremely difficult. The difficulty will be compounded if the thick
aquitards, such as the Ohio Shale, are too impermeable to pump. If
the fracture porosity cannot be determined by field tests, limits can
probably be placed on this parameter by examination of cores, outcrop
and exposures.
An up-to-date bibliography of the methodology of aquifer testing,
laboratory measurements of porosity and permeability, and geophysical
borehole logging is presented in the report, "Recommended Methods for
Water-Data Acquisition" (U. S. Geol. Survey, 1972, chapter II).
In summary, hydraulic testing methodology is available for deter-
mination of head, transmissivity, and porosity needed for the
evaluation of the local and regional ground-water flow systems
beneath Maxey Flats. Undoubtedly such methodology will require
modification when applied to the aquitards. Use of highly trained
-33-
-------
drilling and hydrogeologic personnel and modern air drilling
technology is probably necessary to assure success of such a study.
In addition to the cited data obtained by drilling, certain other
types of data are necessary to complete a study of sita hydrogeology.
i. A modarr. inventory of wells and springs within a 4-mile
(6 km) radius of the site may prove helpful, first to verify and expand
upon certain conclusions of the 1962 reconnaissances, and second to
determine what geologic controls localize the springs and seeps of the
region.
2. Placement of a recording gage above the mouth of Drip Springs
Hollow and the creek in the unnamed hollow east of Maxey Flats. As
mentioned previously, the baseflow characteristics of these creeks are
potentially valuable, when coupled with head data, for a determination
of the gross transmissivity of.the Ohio Shale. In addition, several
seepage runs made in the Drip Springs Hollow could determine whether
the bulk of baseflow is derived from the Ohio Shale, or the Crab
Orchard Formation.
3. Placement of a recording raingage at the N'ECO site on Maxey
Flats.
4, Determination of the water chemistry of the creeks, particularly
during periods of baseflow, and springs- around the Flats. Selected
+ 7-
chemical constituents, for example, Cl , Na , NO., and SO", might
prove as useful as nuclides in long-term monitoring of waste movement
from the site, hence the need to establish the background chemistry
at these sources. Chemistry of water in the aquitards and in the
-39-
-------
trenches will be determined as part of the test drilling program and
chat program (described below) designed to study leakage from the
trenches.
5. Map extent, thickness, and gross beta-gamma activity of
colluvium on slope of Maxey Flats to determine if this material
might under certain circumstances be capable of conducting ground
water (namely interflow) off site. (See below.)
6. Laboratory measurements of effective intergranular porosity
and permeability and X-ray study of the mineralogy of cores obtained
from the test holes; the mineralogic analysis would be principally
of the clay and colloidal size fractions.
7. Measurement of rates of slope retreat.
Determination of the Magnitude of Vertical and Horizontal Leakage
from the Trenches
Definition of the local and regional ground-water flow systems is
only the first step in an analysis of radionuclide movement from a
burial site. Nuclides dissolved in water in the trenches may leave
the trenches in several ways: a) vertical movement into the underlying
aquitards and thence to the perennial creeks flanking the burial
site; b) lateral movement, via the soil and colluvium, namely interflow
on the slopes flanking the site and thence to the creeks; and c) sur-
face runoff, should both the trenches and soil zone become waterlogged.
Which of these routes is likely to dominate under a worst-case
condition, namely when water is permitted to pond in the trenches
-40-
-------
after abandonment of the site, and to what degree night vertical
Leakage modify natural ground-water gradients necessitate study. It
is recognized that, as a result of the reported ponding of water in
some trenches prior to 1971, the ground-water flow system, as defined
today, may already reflect leakage from the trenches.
The magnitude of expectable vertical infiltration into the Nancy
Member of the Sorden Formation at the bottom of the trenches can be
estimated via a series of infiltration tests using double ring
infiltrotneters. Infiltrometers are placed over visible fracture
traces, or, if none are visible, at random. Methodology of such
tests has been described by Johnson (1963).
Infiltration and laboratory tests of. soil between trenches should
permit a direct comparison of the relative permeability of the soil
and the Nancy Member and an estimate of the magnitude of lateral
interflow through the soil zone. Infiltration tests of the colluviuin
will be difficult due to its occurrence only on slopes; laboratory
analyses of permeability of this bouidery material will be difficult.
The depth of vertical and lateral movement of nuclides beneath
and adjacent to the trenches to date (1973) may be determined either
by: a) water sampling, if a ground-water mound is found beneath the
trenches; b) by means of periodic gamma ray logging of specially
designed holes, if flow is unsaturated; and c) radio-chemical
analyses of cores from the test holes and of samples of soil and
colluvium.
-41-
-------
Sorgtion Characteristics of the Aquitards
Sorpcion properties of porous media are usually investigated in
the laboratory by batch mixing or column studies. (See, for example,
Schroeder and Jennings, 1963; Wahlberg and Dewar, 1965 and Kaufman,
1963.) Field measurements are few. (See, for example, Parsons, 1961;
Robertson and Baraclough, 1973.) If water movement through the acui-
tards beneath Maxey Fiats is principally via joints, as suggested
by the reconnaissances of Walker (1962) and Whitman (written commun.
1971), then determination of sorption by lab methods (designed solely
for granular media) will not yield meaningful results except for soil
and colluvium samples. Similarly, paired well tests will probably
not be applicable unless transmissivities are high enough to permit
pump testing. It appears likely that the sorption characteristics of
the aquitards at Maxey Flats will have to be determined by comparison
of the measured nuclide content of water in trenches with that sampled
from springs and wells in the aquitard, and from soil moisture suction
cups placed in the soil and colluvium. The activity distribution so
sampled will be a net result of sorption and hydrodynamic dispersion
properties of the media traversed.
Monitoring
Calibration of a mass-transport model of the Maxey Flats site
will require several to many years of water level, baseflow, radio-
chemical and chemical data from selected wells, springs, creeks, and
burial trenches at and around the site. An intensive and thorough
-42-
-------
radio-chemical monitoring program has been conducted at the site by
the Kentucky Department of Health since Che start of waste burial
at the site. This program could be expanded to sample water from
selected test wells and the baseflow in the creeks flanking the site,
Periodic measurements of water levels and gamma radiation in the
test holes would have to begin after completion of drilling programs.
-43-
-------
REFERENCES
Aziz, K., Hoist, P. H., and Karra, ?. S., 1968, Natural convection
in porous media: Patrol. Soc. of CIM, paper 6813, 19 p.
Biankennagsl, R. K., 1967, Hydraulic testing techniques of dee? drill
holes at Pahute Mesa, Nevada Test Site: U.S. Geol. Survey,
open-file report, 50 p.
1968, Geophysical logging and hydraulic tasting, Pahuca '.-'es£,
Nevada Tesc Site; Ground Water, v. 6, no. 4.
Bredehoeft, J. D., and Finder, G. F., 1973, Mass transport in flowing
ground-water: Water Resour. Res., vol. 9, no. 1, p. 194-210.
3resler, E., and Hanks, R. J., 1969, Numerical method for estimating
simultaneous flow of water and salt in unsaturated soils: Soil
Sci. Soc. Araer. Proc. vol. 33, p. 827-832.
Bresler, E., 1973, Simultaneous transport of solutes and water under
transient unsaturated flow conditions: Water Resour. Res., vol. 9,
no. 4, p. 975-986.
Cherry, J. A., Grisak, G. E., and Jackson, R. E., 1973, Hydrogeological
factors in shallow subsurface radioactive-waste management in
Canada, in, Proc. Intematl. Conf. on Land for Waste Management,
Ottawa, Canada (Oct. 1-3, 1973).
Clark, D. T., 1973, A history and preliminary inventory report on the
Kentucky radioactive waste disposal site: Radiation Data and
Reports, v. 14, no. 10, p. 573-585.
-44-
-------
Cooper, H. H., Jr., Bredehoeft, J. D., and Papadooulos, 3. S., 1967,
Response of a finite diameter well to an instantaneous charge of
water: Water Resour. Res., v. 2, no. 1, p. 263-269.
Dobrovolny, E., and Morris, R. H., 1965, Map showing foundation and
excavation conditions in the 3urtonville Quadrangle, Kentucky:
U.S. Geol. Survey Map 1-460.
Gardner, A. 0., Jr., Peacernar., D. W. , and Pozzi, A. L. , Jr., 1964,
Numerical calculation of multidimensional miscible displacement
by the method of characteristics: Soc. Petrol. Eng. J., vol. 4,
no. 1, 'p. 26-36.
Hall, F. R., and Palmquist, W. N. , Jr., 1960, Availability of ground
water in Bath, Fleming', and Montgomery County, Kentucky: U.S.
Geol. Survey, Atlas HA-18.
International Atomic Energy Agency, 1965, Radioactive waste disposal
into the ground: IA.EA, Safety Series Rept. No. 15, 111 p., Vienna.
International Atomic Energy Agency, 1971, Disposal of radioactive
wastes into rivers, lakes, and estuaries: IAEA Safety Series
Rept. No. 36, 77 p., Vienna.
Jacob, C. £..,.1943, Correlation of ground water levels and precipita-
tion on Long Island, New York: Am. Geophys. Union Trans.,
pt. 2, p. 564-573 (Jan. 1944).
Johnson, A. I., 1963, A field method (ring infiltrometer) for
measurement of infiltration: U.S. Geol. Survey Water-Supply
Paper 1544-F, p. F1-F27.
-45-
-------
Kaufman, W. J. , 1963, An appraisal of Che distribution coefficient
for estimating underground movement of racioisotopes : Hazleton
Nuclear Science Corp. report to the U.S. AEC (Nevada Operations
Office, Las Vegas, Nev.), 23 p.
Lai, 5. H., and Jurinak, J. J., 1971, Numerical approximation of
cation exchange i^niscible displacement through soil columns:
Soil Sci. Soc. Amer. Proc., vol. 35, p. 394-899.
McDowell, R. C., Peck, J. H., and Mytton, J. W., 1971, Geologic ma? of
the Plummers Landing quadrangle, Fleming and Rowan Counties,
Kentucky: U.S. Geol. Survey, Map GO-964.
Mawson, C. A., and Russel, A. E., 1971, Canadian experience with a
national waste-management facility, in, Management of low-and
intermediate-level radioactive wastes: Intematl. Atomic Energy
Agency, Vienna, p. 183-194.
O'Connell, M. F. and Holcomb, W. F., 1974, A summary of low-level wastes
buried at commercial sites between 1962-1973 with projections to
the year 2000: Radiation Data and Reports (in press).
PapadopuloSj S. S., Bredehoeft, J. D., and Cooper, H. H., Jr., 1973,
On the analysis of "slug test" data: Water Resour. Res., v. 9,
no. 4, p. 1087-1089.
Parsons, P. J., 1961, Movement of radioactive waste through soil.
Part 3. 'Investigating the migration of fission products from
high-ionic liquids deposited in soil: Atomic Energy Canada Limited
Rept. CRER-1018, 46 p.
-46-
-------
Peckhao, A. E., and Belter, W. G., 1962, Consideration for selection
and operation of radioactive waste burial sites, in, Second
ground disposal of radioactive wastes conference, Chalk River,
Canada (Sept. 1961), USAEC Rept. TED-7668 (Book 2), p. 428-436.
Finder, G, F., 1973, A Galerkin finite-element simulation of ground-
water contamination on Long Island, New York: Water Rasour. 3.es. ,
vol. 9, no. 6, p. 1657-1669.
Finder, G. F., and Cooper, H. H., Jr., 1970, A numerical technique
.for calculating the transient position of the saltwater front:
Water Resour. Res., vol. 6, no. 3, p. 875-882.
Price, H. S., Cavendish, J. C., and Varga, R. S., 1968, Numerical
methods of higher order accuracy for diffusion-convection equations:
Soc. Petrol. Eng. J., vol. 8, no. 3, p. 293-303.
Reddell, D. L., and Sunada, D. K., 1970, Numerical simulation of
dispersion in groundwater aquifers: Colo. State Univ., Hydrol.
Pap. 41, 79 p.
Richardson, R. M., I962a, Northeastern burial ground studies, in,
Second ground disposal of radioactive waste conference, Chalk
River, Canada (Sept. 1961) : USAEC Rept. TID-7628 (Book 2), p. 460-461.
1962b, Significance of climate in relation to the disposal of
radioactive waste at shallow depth below ground, in. Proc. on
Retention and Migration of Radioactive Ions through the Soil:
Commissariat a 1'Energie Atomique, Institut National des Sciences
et Techniques Nucleaires, Saclay, France, p. 207-211.
-47-
-------
Robertson, J. B., and Barraclough, J. T., 1973, Radioactive and
chemical-waste transport, in groundwater at National Reactor
Testing Station, Idaho: 20-year case history and digital model, in
Underground Waste Management and Artificial Recharge, J. Braunstain,
editor: American Assoc. Petrol. Geol., vol. I, p. 291-322.
Rubin, J., and James, R.. V., 1973, Dispersion-affected transport of
reacting solutes in saturated porous media: Galerkin method
applied to. equilibrium controlled exchange in unidirectional
steady water flow: Water Resour. Res., vol. 9, no. 5, p. 1332-1356.
Scheidegger, A. E., 1961, General theory of dispersion in porous media,
Jour. Geophys. Res., vol. 66, no. 10, p. 3273-3273.
Schroeder, M. C.,. and Jennings, A. R., 1963, Laboratory studies of the
radioactive contamination of aquifers: Univ. Calif., Lawrence
Radiation Lab., rept. UCRL-13074, 52 p. 2 appendices.
Tamura, T., 1972, Sorption phenomena significant in radioactive-waste
disposal, in, Underground waste management and environmental
implications, T. D. Cook, editor: American Assoc. Petrol. Geol.,
Mem. 18, p. 318-330.
Thompkins, E, R., and Mayer, S. W., 1947, Equilibrium of rare earth
complexes on exchange resins: Am. Chem. Soc. Jour., vol. 69,
p.. 2859.
U. S. Atomic .Energy Commission, 1967, Nuclear terms, a glossary:
U.S.A.E.G. Office of Information Services, World of Atom Series
Booklet, p. 64.
-48-
-------
U.S. Geological Survey, 1972, Recommended methods for water-data
acquisition: Prelim. Repc. of Fed. Interagency Work Group on
Designation of Standards for Water Data Acquisition, Ch. II.
Wahlberg, J. S., and Dewar, R. S., 1965, Comparison of distribution
coefficients for strontium exchange from solutions containing
one and two competing cations: U.S. Geoi. Survey Bull. 1140-D
p. DI-D10.
Walker, I. R., 1962, Geologic and hydrologic evaluation of a proposed
site for burial of solid .radioactive wastes northwest of Morehead,
Fleming County, Kentucky: Unpublished report prepared for State
of Kentucky, Dept. Health, Frankfort, Kentucky, 33 p.
Williams, R. E., and Farvolden, R. N., 1967, The influence of joints
on the movement of groundwater through glacial till: Jour.
Hydrology, v. 5, no. 2, p. 163-170.
Winograd, I. J,, 1974, Radioactive waste storage in the arid zone:
EOS, Trans. Amer. Geophys. Union, v. 55, no. 10, p. 884-894.
1970, Non-instrumental factors affecting measurement of static
water levels in deeply buried aquifers and aquitards, Nevada Test
Site: Ground Water, v. 8, no. 2, p. 19-28.
-49-
------- |