CONF-8409115

Los Alamn- Environmental Reatoration ...,. a .......... ,~. Pl'ftrt:~tt•ina facility

ER Record 1.0.# U011757

PROCEEDINGS OF THE SIXTH ANNUAL PARTICIPANTS' INFORMATION MEETING

DOE LOW-LEVEL WASTE MANAGEMENT PROGRAM

Convened by The DOE Low-Level Waste Management Program

September 11-13 Denver, Colorado

Published December 1984

National Low-Level Radioactive Waste Management Program

Idaho Falls, Idaho 83415

Prepared by EG&G Idaho, Inc. For the U.S. Department of Energy

Under DOE Contract No. DE-AC07-761D0157r

111111111111111111111111111111 3473

•

•

•

TECHNOLOGY DEVELOPMENT FOR THE DESIGN OF SHALLOW LAND BURIAL FACILITIES AT ARID SITES

J. W. Nyhan, w. v. Abeele E. J. Cokal and B. A.Perkins

Los Alamos National Laboratory Environmental Science-Group

Los Alamos. NM 87544

and

L. J. Lane U.S. Department of Agriculture

Southwest Rangeland Watershed Research Center Tucson, AZ 85705

ABSTRACT

The field research program involving technology development for arid shallow land burial sites is described. Field tests of biointrusion barriers at waste disposal sites and in experimental plots at Los Alamos are reported. Results of completed and on-going experiments with migration barriers for water and contaminant movement are presented. An envelope wick experiment for subsurface water management is described and preliminary field data are reported. An integrated field experiment was designed to test individual SLB component tests related to erosion control, biobarriers, and sub­surface capillary and migration barriers, and the progress made in emplacing the experiment is presented. Efforts to utilize the field data col­lected to validate hydrologic models (TRACR3D) important to waste management strategies are also presented.

INTRODUCTION

Reliable and comprehensive experimental data are not available to design a shallow land burial (SLB) facility at an arid site or to predict the future performance of proposed SLB designs. Sufficient i~formation does not exist to allow preparation of manuals to design SLR facilities at arid sites. Particularly important issues are the ability to predict long-term performance and the reliability of those predictions. Moreover, data must be obtained on system interactions when integrated system designs are based on optimal components. Application of component models [e.g., CREAMS (A Field Scale Model forLhemicals, Runoff, and Lrosion From~griculture Management ~stems) for surface and near-surface processes, and unsaturated flow and transport models for subsurface processes need to be evaluated for validity· of predictions in integrated systems environments. These data and evalua-

150

)

)

)

~··

•• ~··

tions are necessary to demonstrate the ability to model, monitor, and predict future performance of SLB facilities. Gaps in knowledge exist on component processes even as our component experiments are producing new information and optimal designs. Addition information is required to control plant and animal intrusion into buried waste of to specify the influence of burrowing animals on the water balance. Additional information is needed to more accurately determine the potential for radionuclide movement within, near, and off-site and to design barriers to control this migration. Addi­tional data and model testing are required to predict long-term SLB perfor­mance using validated mathematical models.

The experiments performed under this task will provide ~xperimental information in several of these areas. Specifically, the experiments per­formed for this task will field test biointrusion barriers, migration bar­riers, and ground and surface water management systems, as well as verifying hydrologic SLB models and performing an integrated field experiment to test former and on-going individual SLB component field tests.

BIOINTRUSION BARRIER TESTING

Questions concerning the performance of different types of biobarriers at various field scales and the effects of layered rock barrier systems on the water balance are being addressed in a field experiment at a low-level waste disposal site (Area G), which will eventually be decommissioned.

Three types of 1-m thick biobarriers are being tested in the Area G experiments in 6 by 21 m plots: (1) 15 em gravel (1-2 em-diameter) on top of 85 em cobble (7.5-12 em-diameter), (2) 1m cobble, and (3) 30 em gravel on top of 70 em cobble. These three treatments are being compared with the con­ventional trench cap treatment containing 1 m of crushed tuff without a biobarrier. Since we were trying to produce treatment effects over a two year period, we used only 15 em of topsoil over these four field treatments to insure that percolation of water through the biobarriers would be detected using neutron moisture gauge techniques. A cesium chloride tracer layer was emplaced immediately beneath each biobarrier (and at a correspond­ing depth in the conventional cap treatment) so that the penetration of roots currently growing through the biobarriers can be detected by collect­ing plant samples and assaying them for their cesium content.

Cesium determinations were performed on native grass vegetation samples collected from all four plots at Area G on 14 sampling dates (June 1982 through November 1983)(1). Elevated levels of cesium were generally not found in these samples, indicating that the biobarriers were performing satisfactorily. However, only the 1 m cobble biobarrier proved to be totally effective in preventing root penetration to the cesium tracer. The gravel­cobble intrusion barriers failed one and three times, for the 15 em gravel­as em cobble and 30 em gravel- 70 em cobble treatments, respectively. The crushed tuff treatment failed four times during this 1 1/2 year period, and demonstrated a significantly higher average cesium concentration in the vegetation from this plot than for similar samples collected on the other three treatments.

The water content of the crushed tuff located 30 em beneath the four biobarriers is shown as a function of time in Fig. 1, along with daily pre­cipitation and snow depth (1). The most obvious characteristic of this data is the seasonal changes in water content, regardless of the overlying trench cap treatment. The greatest increases in water content occurred with winter

151

•

•

•

snowmelt, with smaller increases occurring when rain was the source of pre­cipitation during seasons when evapotranspiration was maximized.

The three biobarrier treatments involving gravel and cobble demon­strated higher water content values (30 em beneath the biobarrier) with time than similar values for the crushed tuff treatment (Fig. 1). This observa­tion reflects the loss of water storage capacity in the 1-m thick layer of gravel and/or cobble that would otherwise be present in the corresponding 1-m thick layer of crushed tuff in this experiment. This loss of water hold­ing capacity can obviously result in significantly greater percolation of precipitation, and, concurrently, more rapid changes in soil water content (Fig. 1).

MIGRATION BARRIER TESTING

Water can infiltrate into a SLB facility and interact with the buried waste by a number of pathways. This interaction can result in the mobiliza­tion and subsequent transport of contaminants out of the burial facility and into the environment surrounding the facility. There does not exist adequate data from carefully-instrumented large scale field experiments on the move­ment of water and contaminants under unsaturated conditions to enable a site operator to define and engineer suitable barriers to prevent the migration of radionuclides out of a SLB facility. Also, the mechanisms of leaching and transport under unsaturated conditions are not well known. Before the need for a migration barrier can be determined and the specification of a migra­tion barrier can be made, the data and information mentioned above must be obtained .

Migration barriers are used in shallow land burial facilities to slow or stop the movement of water and contaminants and are discussed in a recent Los Alamos report as a single component embedded in a complex environmental system (2). Analytical solutions to solute transport equations are used to approximate the behavior of migration barriers and to derive design criteria for control of subsurface water and contaminant migration. Various types of migration barriers are compared and design recommendations are made for shallow land burial trench caps and liners. Needed improvements and sug­gested field experiments for future designs of migration barriers are then discussed relative to the management of low-level radioactive wastes.

Although not much is known about gravel-cobble biointrusion barriers functioning as a subsurface wick or capillary barrier system in a trench cap, field research at Los Alamos has shown that gravel-cobble biobarriers can have an effect on the subsurface migration of water and radionuclides (3). This field research was performed in two 6.1 m-deep caissons with a diameter of 3.05 m, one of which contained a 1 m-thick gravel-cobble biobar­rier (Fig. 2) In the caisson without a barrier, soil water movement at the zone containing Cs, Sr, and Co tracers did not occur as sudden breakthrough surges, as it did in the caisson with the biobarrier. For the same water inputs, the biobarrier treatment exhibited greater migration of tracers than did the tuff treatment, as is typically shown for the Cs data (Fig. 3). This effect was attributed to greater nonuniformity in water content, more water infiltration, and the "pulse" type of change in water content at or near the tracer layer in the biobarrier treatment compared with the tuff treatment (as also shown in Fig. 1) .

152

)

)

' 20

~ 18

16 8 r:t: 14 ~~ ~- 12

..:I u--~ 10

~~ e gz o-> 6 r&l CJ 4 i ~ 2

0

LOS ALAMOS LOW-LEVEL WASTE DISP~AL SITE (AREA G)

SAWPU NG DEPTH: TREATWENT: (30 em BELOW BlOBA.RRIER) 15 em TOPSOIL

100 em BIOBARRID

/\./.:":·.... .··· ....... .. .. . ' \.·. . .... . :,, A • '•, / ''•/' ~··,

!., v -~ .............. /. .. :.--<-. ' . / .. - ... "'- , ...... ,, •... -. !. . .\. lf' '\ ,..... ·~-·

. ···.·: .. ····· . t· I ::{~!.-.... ·•· .. -···· ................ ········· ·····.·. ,.· ·· .... ·· ··--~:~.~~ .... j ' , ..... \ ; "-" v--.~'->.r, ..,...:_, __.....,;,·, v .

GRAVEL/COBBLE ( 15 cm/86 em)

COBBLE (100 em)

1982 1983

NO BIOBA.RRIER

GRAVEL/COBBLE (30 cm/70 am)

1984

5 r-~~~~--~~~-r~--~~~~~~~

4.5 ~ SNOW DEPI'H -E

4 I PRECIPITATION

() - 3.5

~ 3 -E-~ 2.5

0:: 2 -~ 1.5 c:t:: 0.. 1

0.5

0 ~~~~~~~~~~~~~~~~~~~~

Fig. 1. Precipitation and distribution of water 30 em beneath the biobarrier treatments at Area G vs time.

153

••

•

•

R -f-GRAVEL{ CO&Bl

~~:ffi:~~lffi~ -TRACERS .-CRUSI£D

TlEF

COWPACTEO_i:=:-­CRUSHED _

TUFF

Cs,Sr,Co

NEUTRON WOISTURE

PROBE ACCESS TUBES

SAND liiUif[L

DIWII

c OWPACTED CRUSHED

TUrF

Fig. 2. Biointrusion barrier/transport experiment.

-gu -)( E f 20 E Q. Q. -,6 c

..0

8 12 -0

Ill 8 c:: 0 ·--~ 4 -c:: II CJ

······· ... ······ ...

· .... ~ ....

--- CRUSHED TUff TREA Tt.4£NT

·············•· B I OBARR I ER TREA Tt.4ENT

·· .... ···· ...

··. ··. ·· ...

...... '··· ...... ··--·-···-·-·-··

§ 0~------,-------,-------~-------r-------r-------, 300 (.) 0 50 100 150 200 250

DIs t once from the tracer I aye r (em)

Fig. 3 • Vertical distribution of a cobalt tracer in caissons.

154

)

)

' )

'

'

Thus, a gravel or cobble layer incorporated into the cap may result in problems if the soil in the region adjacent to the gravel becomes saturated. This may be solved by perforated pipes and pumps emplaced strategically beneath the trench cap. Another solution would be to make certain the layer above the gravel-cobble layer performed as a satisfactory wick layer. This would mean that a gravel-cobble layer in the trench cap would have a signi­ficant slope (the slope in the caisson experiment shown in Fig. 2 is equal to zero) and slope length, and that the thickness of the overlying capillary layer would also be adequate in this configuration to effectively transmit subsurface water.

GROUND AND SURFACE WATER MANAGEMENT SYSTEMS

The purpose of the Ground and Surface Water Management System Testing subtask is to field test systems that can be used to control the movement of water on top of or around SLB trenchs. These systems will incure the execu­tion of the performance objectives for low-level radioactive waste disposal sites set up by the Nuclear Regulatory Commission (NRC) Part 61 and in simi­lar DOE requirements.

Small-scale modelin~ has demonstrated it is possible, by using capil­lary barriers, to mainta1n dry structure in porous media. The percolating liquid will only penetrate the coarser material after the overlying finer material is near saturation, provided the soil moisture characteristic curves of the two materials reach a certain degree of dissimilarity. Conse­quently, the structure, which is enclosed in the coarser material, remains dry. As long as the pressure at the coarse/fine interface remains negative, water infiltrating into the finer layer will not cross the interface but will flow within the finer layer; percolation occurs where the saturated water front reaches the edge of the coarse layer. The limiting granulometric differences, beyond which this phenomenom ceases to exist, have been deter­mined. This concept has sometimes been referred to as the 11 Wick effect ... It was found that under unsaturated conditions,a gravel lens caused lateral flow in a finer-textured overlying material. The lateral distance over which the water can be transported is limited and will be influenced by the slope of the interface.

At equal matric potentials, a fine-grained medium will contain more water than a coarse-grained medium. During drainage, an initially saturated soil will drain its largest pores first, whereas in wetting a dry soil, the smallest pores will fill first as the matric potential is allowed to increase (getting less negative), the water content increases as progres­sively larger pores are filled. This will be accomplished by an increase in hydraulic conductivity and lateral diversion of water already promoted by the presence of a sloping interface. The coarser medium, on the other hand, with its predominantly large pores, will remain relatively dry with a low hydraulic conductivity until the matric potential reaches zero and near zero at the interface {Figs. 4 and 5). The system will fail when or before the matric potential at the interface ceases to be negative. Point readings at the interface are provided through tensiometers. Proximity of the failure point to zero will be function of abruptness and magnitude of particle size difference at the interface.

155

•

•

1-z LoJ 1-z 0 u 0:: LoJ 1-

~ u 0:: 1-L.o.J

F'INE s•NO

~ CO.RSE GUVEL -J 0

> ~~---------------------------~ATRIX POTENTIAL

Fig. 4. Hypothetical wetting curve for a fine sand and a coarse gravel.

fiNE S•NO

VOLUMETRIC WATER CONTENT -

Fig. 5. Hypothetical K(8) curves for a fine sand and a coarse gravel •

10 ... 1-..-___.IL...____.J,_--L _ __.L.. _ _.J,.._......J

0.0 -4.0 8.0 12.0 16.0 20.0 2o4.0

PERCENT BY DRY WEIGHT 0[ AQWIXTUR£

Fig. 6. Effect of adding bentonite to Los Alamos tuff •

156

)

J

Our wick system designs indicate that an effective wick system is one where the suction at the wick interface is about four kPa tension or less before failure. To obtain such a low suction between our backfill and sand, the addition of a small amount of bentonite to the backfill materials at our land disposal site is necessary (Fig. 6).

Our field experiment was desi9ned to test the performance of a wick system emplaced in a 6.1-m deep ca1sson with a diameter of 3.05 m for a wick system of known thickness (2m), slope (10%), and slope length (2m), and for one combination of porous materials [a bentonite-tuff mix having a 0.02 bentonite ratio by mass overlying Ottawa sand] subjected to a known addition rate of water (Fig. 7). Using the soil water and tension data collected over time at several positions within the wick layer, we will then be able to field validate a hydrologic model describing the two-dimensional, unsaturated flow of water thorugh a multilayered system. This data set will then be available to field-validate several hydrologic models, which can subsequently be used for the construction of future wick designs.

Although this experiment was just emplaced in June 1984, preliminary measurements of soil water content and tension collected adjacent to the interface between the bentonite-tuff and Ottawa sand layers are presented for several sampling times (Figs. 8 and 9). The average volumetric water content of the bentonite-tuff mixture gradually with sampling depths from 2.5 to 3.7 m (Fig. 8), as the wetting front progressed toward the underlying layer of Ottawa sand (Fig. 7). The tensiometric values observed in the bentonite-tuff mixture (Fig. 9), are seldomly showing signs of positive pressure, even after 98 days of submission to a 1 m-thick overlying water layer. The volumetric water content curves seem to approach an asymptotic value of about 22%. These values are probably more representative of some­thing like field capacity moisture content rather than saturation water con­tent.

The tensiometric values seem to indicate (Fig. 9) that the wick system is actively slowing the flow of water into the lower sand layer (positive tensions were observed along the sand/bentonite-tuff interface at the 4.02 to 4.18 m sampling depths).

Additional data collection activities will continue in this field experiment after saturated flow conditions are met. After this, as the bentonite-tuff mixture dries, our plans are to collect additional tension and soil water content data for subsequent model verification and validation activities.

Model Verification

To predict how wei I a disposal site may be able to contain toxic materials, simulations of the movement of these waste components are necessary. The TRACR3D computer code has been developed at Los Alamos to simulate transport of solutes through unsaturated as wei I as saturated soils and rock. The model computes water and/or air flow under soi I moisture conditions ranging from fully saturated to completely dry. Material properties can vary spatially. A variety of boundary and source/ sink conditions Is possible. It also simulates transport of sorbable species. Transport mechanisms include advection, diffusion, and disper­sion. Sorption can be either equil lbrium or kinetic and saturable or not. Decay chains and leachin9 sources are allowed. The TRACR3D model has

157

_, <.1'1 ():)

• 20

18

16

,... s 14 c 0 • • 'i 12 u -0

• 10 0 --0 Jj 8 E 0 .. -• 6 u c 1111 .. ii 4

2

0

'--"

Subturface Wick Experl•ent bide view)

I ,/Neutron probe •~ lube

o- -TenlliOmeter -::Ao

1_ ___ 0 ----_T~ft-~lo;-- -o J --- o·· -Mixture -o

o -o o--- ---- --u

0

8

5

4

3

,... E ...., c 0 --'i u -0

E 0 --0 Jj

• 0 2 .!::

1

0

• u c 1111 -• s

• Subsurface Wick Experiment (plan views)

10 1 Section --------...._ r 3

-... -........

8

6

4

2

Q) 0

M

g 10 l Sec lion !J BB rn ..... 0 8

6

4

2

0

2

l

0

3

2

1

0

•

-e ~

Q) 0 a «< +)

en . .... 0

Fig. 7. Modified envelope wick design.

"-' \...,

~

-J

(J"'

\.0

f'A. '

21

Modified Wlck Experiment

- 11 • • 2.50m

.... • • • 2.80m c 22 • • • 3.10m CD .... ,. .. " 3.40m c 0 20 • • • 3.70m u L. Q) 18 -c ~ 16 u ·-L.

-; 14

E :J - 12 0 >

10 0 14 28 42 56 70 84 98 112

Time (days)

Fig. A. Distribution of volumetric water content for five sampling depths vs time in the envelope wick experiment.

,..,

ENVELOPE WICK EXPERIMENT: •• SAWPLE LOCATIONS ABOVE TUFF-cLAY/SAND INTERFACE

~ 2.5m ._...._ 2.8m - ..__ 3.1 m f

.! ,...__ 3.4 m

E -- 3.7 m -z 0 -til z r&l E-r&l ~ < ~ r&l

?<

-76

82 83 85 96 &7 100

TIME {days)

75 SAMPLE LOCATIONS AT TUFF-CLAY/SAND INTERfACE

• 50 ) -.. ~ g

.&J E 25 -z 0 Vl 0 z YJ 1-

YJ B---8--tl 4.02 (m)

" -25 __, 4.06 (m)

""' ~ __........ 4.10 (m) YJ

~ ,.__ 4.14 (m) -50 • • • 4.18 (m)

-75

82 u 84 95 96 97 98 99 100

TIME (days)

Fig. 9. Temporal distribution of soil water tension at various sampling depths above and at the surface of the tuff-clay/sand interface of the enve-

• lope wick experiment •

,)

160

been used to study migration of species from underground nuclear explo­sions, high-level underground waste storage, and lo·.-1-level radioactive shallow waste burial.

Any mathematical model, to be useful, must be verified and validated. Verification means comparison to analytic solutions. Val !dation means comparison of the model with experimental data. Examples of both for TRACR3D are discussed by Travis (3).

Laboratory experiments are usually wei I control led but do not allow alI length and time scales to be tested. Field experiments allow a wider range of length and time scales but are frequently subject to un­certainty regarding the spatial distribution of material properties. The Los Alamos caisson experiments allow testing of larger length and time sea I es than I aboratory experiments a I I o~o; but without the uncertain­ties of large scale field experiments. To validate a code, both caisson and field experiments are useful.

In FY-1984, validation of the TRACR3D code in a one-dimensional form was obtained for flow of soil water in three experiments (4). In the first experiment, a pulse of water entered a crushed tuff soil and initially moved under conditions of saturated flow quickly followed by unsaturated flow. In the second experiment, steady state unsaturated flow took place. In the final experiment, two slugs of water entered crushed tuff under field condi­tions.

Experiment I

The first field experiment was performed in a 3-m diameter, 6-m deep caisson, which was fl I led with crushed tuff (5,6). Tensiometers and access tubes for neutron moisture gauge determinations of volumetric water content were emplaced at 75-cm Increments within the caisson and measurements collected frequently for six weeks. A 13-cm pulse of water was added to the surface of the caisson, and the top of the caisson was sealed to prevent evaporation of water. Figure 10 shows the resulting changes In soil water content with time.

The change In volumetric moisture pulse decreased In ampl !tude rather quickly as It moved downward, and, below 400 em, no change In volumetric moisture could be detected experimentally (fig. 10).

The data indicate that, although movement of soi I water could not be detected below approximately 400 em at any time during the experiment, drainage from the caisson would require that soil water was moving below this horizon with flux in equal to flux out In those regions in which no change In soil moisture was detected.

The data also Indicate that the Introduction of water at the surface was felt very quickly at the bottom of the caisson, with drainage increasing to day 12-13, after which the drainage rate remained the same through day 22, when the experiment was terminated.

The code model (4) predicts that, for the initial given conditions. once a 13-cm slug of water Is added, there wi II be saturated flow near the

1 61

•

•

•

-SURF' ACE

30 29 ~

28 -27 -

26 -

25 -

2 .. -

23 -

22 -

21 -

20 -

19 -18 r-

17 ~

16 0

0

• c

•

0

'

I

• c

•

0

v 0

•

100

• 0

0

v

I

I

l

0

• v

•

I I

0

I 0

I I

200 300 .. oo DEPTH FROM SURF ACE (em)

I

I

• 1/21 0 1/22 • 1/23 a 1/24 • 1/27 0 10/1 9 11/1

J

500

-------------600

Fig. 10. Water pulse moisture as a function of depth and time.

sol I surface just after pending. At a few centimeters below the surface~

this flow becomes unsaturated~ with the volumetric moisture pulse having a smal I width at 3.2 hours. Very quickly alI flow is unsaturated and the

pulse width continues to broaden~ with a decrease in amp! itude as the pulse

moves downward. By 400 em~ the effect of the pulse becomes very smal I in

terms of changes In volumetric moisture and below this~ the changes are too

smal I to detect experimentally. At the bottom of the caisson at the tuff/sand and sand/gravel Interfaces~ moisture levels at each interface are

predicted to increase until approximately day 12.

Very close agreement was observed between the TRACR3D code predictions

and the experimental data from the field experiment. When both types of

data were compared in terms of saturation ratios~ they generally differed by

a saturation ratio of only 0.02 (4).

Experiment 2

A second caisson was fi lied and instrumented similar to the caisson In

the first experiment. Using a drip irrigation system, water was uniformly

applied across the surface of the caisson at a uniform flow rate of 200 rnl/min unti I equi I ibrium conditions (same flow in and out) were achieved.

The experimental and the code data both showed the crushed tuff in

the caisson to have saturation ratios ranging from 0.84 to 0.86, (4) again

showing close agreement as In the first experiment .

162

.)

)

j

"II{ ~~-·

Experiment 3

The purpose of the final experiment was to compare ~he TRACR3D code simulation with a field experiment In which moisture levels were initially low and data were taken over almost a year. The field experiment was much closer to actual conditions In a burial site because the material properties of the fil I had not been determined (as thei were In the caisson experiments). In the field Instal latlon excavated Into Bandelier tuff, 220-cm of crushed tuff is overlain by 61-cm of 7.6- to 17.8-cm-diameter cobble, covered by 30-cm of 1.9- to 2.5-cm-diameter gravel, and final Jy, at the surface, covered with 3D-em of topsoi 1.

After emplacement of the experime~t, soi I moisture as a function of depth was measured at regular intervals over a period of almost a year <Fig. II).

A rain gauge at a nearby station indicated that between Dece~ber I, 1982 and February 23, 1983, approximately 9-cm of moisture as snow fel I at the site. During February, the snow melted. In March, approximately 2.5-cm of additional precipitation occurred. The field moisture data indicate the fate of this winter precipitation. Because of low evaporation in the winter and the flat terrain, most of the precipitation infiltrated into the 30-cm of topsoil. Because of the soi !/gravel interface, buildup of soi I water occurred in the topsoi I unti I the potential at the interfaces became the same and a slug of moisture entered the cobble and moved downward into the underlying crushed tuff fi I 1. The data indicate that at the top of the crushed tuff two "slugs" of water entered, creating a physical situation in the field simi Jar to the experiment run earlier In the first caisson.

Integration of the "before" and "after" moisture in the crushed tu~f indicates that approximately 8-cm of water entered the tuff in a fairly short Interval before February 23, 1983 and another 2-cm slug of water just before March 2i, 1983. Significant evaporation from the crushed tuff is prevented by the upper cobble/gravel cover. During the rest of the year, despite additional precipitation, moisture losses from evapotranspiration processes prevented bu i I dup and "breakthrough" at the top so i I /grave I inter­face.

Figure I I Indicates the downward movement of the volumetric sci I moisture pulses and the corresponding decrease in amp I itude. The changes In soli moisture appear to be "darrtped" completely by 160-cm below the top of the crushed tuff. The general behavior of the pulses is simi Jar to that observed in the first caisson and modeled In the first TRACR3D simulation. However, because of the Initial low moisture levels and slightly smaller pulse Inputs, the volumetric soi I moisture pulse movement is much slower and complete damping {defined as no detectable change In volumetric moisture) occurs at a higher horizon.

The field and simulation agreement are not as close as in the first caisson pulse experiment. {4) In spite of the unknown material properties of the field fil I and probably greater heterogeneity of the fl I I, the simulation results are In surprisingly good agreement. The "damping" point found in the s.lmulation at 160-cm is the same as was found experimentally <Table I).

163

•

•

•

22 <> ..:+ • o• 0

42 )(() ... 110

&2 ~

82 ,.

102 ,.. -E Cm&.£/TUFT I~ACI

u 122 ·- 0 -w 142 • ,..<> eo 0 0 • ~ 162 • )(i() .. 0 D: ;:) en 182 • 0< «>• .. ~ 0 202 _, ., Ql )8() •

w £1) 222 •c x• ~ ..... 242 •o. INITIAL CONDITIONS e 1/tt Q.. w 0 2/23 c 262 -· • 3/21

a 3/D

282 • • C/U 0 1/17

302 • + •I• )( 11/•

322 • 342 •

0 • 8 12 16 20 24 28 32 36 •o MOISTURE BY VOLUNE

Fig. 11. "Damping" of moisture pulses under field conditions.

INTEGRATED SYSTEM DEMONSTRATION

Research conducted by the Los Alamos Environmental Science Group over the last three years has led to an improved understanding of water dynamics and biological interaction in the SLB site environs. That research has led to the development and testing of improvements in the design of isolated components of an SLB site (7-11). The purpose of this field demonstration is to integrate some of the positive features of the isolated variable experi­ments into a state of the science design for optimal water and biota manage­ment to minimize radionuclide transport at SLB sites.

Specific objectives of this demonstration are to measure the perfor­mance of a conventional and improved trench cover design: (1) on subsurface water dynamics, (2) in limiting biological intrusion, and (3) under both natural and enhanced precipitation regimes to identify the performance lim­its of the two designs. An important adjunct to the project will be to apply and demonstrate offshelf computer-based techniques for in-situ monitoring of SLB facility performance. Time series measurements will be made on the relevant water balance components including precipitation input, soil mois­ture, soil solution, and leachate production including tracer content of the leachate and soil solution. The rate of root intrusion through the cover will also be measured by the tracer content of vegetation.

164

)

)

j

TABLE 1. DEGREE OF SATURATION AS A FUNCTION OF DEPTH AND TIME FOR EXPERIMENTAL (Experiment 3) AND SIMULATION kESULTS.4

Distance Below Initial

Interface Conditions 10 days 10 months

(em) E s E s E s 20 0.22 0.21 0.55 0.49 0.35 0.37 40 0.17 0.17 0.52 0.48 0.35 0.37 60 0.15 0.15 0.42 0.43 0.32 0.37 80 0.16 0.16 0.24 0.16 0.32 0.36

100 0.18 0.18 0.18 0.18 0.29 0.34 120 0.18 0.18 0.18 0.18 0.25 0.32 140 0.18 0.18 0.18 0.18 0.22 0.26 160 0.18 0.18 0.18 0.18 0.20 0.18 180 0.18 0.18 0.20 0.18 0.20 0.19

E - Experimental S - Code simulation Total saturation - 40% by volume Degree of saturation - volumetric moisture/total volumetric moisture at saturation

Plot Construction

Four 3 x 10m plots were emplaced in the last half of FY-1984 to emphasize two major variables, namely, cover design and moisture application rate (Fig. 12). The two trench cover designs include one representing con­ventional trench cover design technology [a topsoil/backfill configuration (Fig. 13) and an improved technology (Fig. 14) employing a topsoil/layered rock system that functions both as a water ar.d biological intrusion barrier when properly configured (7,9). Not only does the layered rock system per­form as a water and biological intrusion barrier, but the material cost is relatively inexpensive, rock does not readily deteriorate in the environ­ment, and it is generally available.

The conventional, or control, trench cover design consists of 15 em of sandy loam topsoil over 75 em of sandy backfill (Fig. 13). The improved trench cover design consists of from 60-75 em of topsoil (see Fig. 14) over a minimum of 25 em of pit-run gravel (<2 em-diameter) and 90 em of river cobble (7-12 em-diameter). The interface between the topsoil and gravel is sloped at 5% [along the 3m plot dimension (see Fig. 12)] and will be separated with a polypropylene geotextile to prevent soil interpenetration into the gravel.

Each plot is lined with an impermeable liner to allow for mass balance calculation of water dynamics. Both the control and improved trench cover plots contain perforated drain pipes (trench drain) to allow for measurement of leachate production and tracer content of leachate. Each plot is fitted

l6'i

•

•

•

e 9

CONTROL DESIGN

'I'Opg:)IL ••••••••••• BACKfiLL

PRECIPITATION REGIME

INPROVED DESIGN

. •..............••

CONTROL DESIGN

TO~IL ••••••••••• BACJ<riU.

._•-3m___.•~ ~ : ~ ..-------. AUTOMATED

DATA CONTROL CENTER

IWPROVED DESIGN

TOP001L .. ............... .

Fig. 12. Plan view of field plot design of integrated systems test .

166

)

,,

t-am

L

.-'L • ~ • .-exm

!_

,-l&:m • ' ?lX:m

t-Zc:m

~

1. 2

1.

I

,... -1

-2

1----

1

1

2

LEGEND =

I PLAN VIEW I 1. 1. 2 DR.A.Hi 2

1. 1. 10m

I SIDE VIEW I ,...

CR~ ,...

1 TOPSOIL 1

2 SI.NDY 2 BACKnU.

1 !ANDY 1 BACKnl..L

l.t.ACHAT£ DRAIN

I END VIEW I TOPSOIL 1

SANDY 2 BACKnU.

SANDY BACKrlU.

e NEUTRON ACC~ TUBE

51

-- TR.AC'!!RS (Ct..Br)

3 I

1 PSYCHROlU1 £RS. POROUS CUPS

2 TENS 1 OW L1 £RS

S FLOW WETER

(MOT TO s;At.l)

Fig. 13. Control trench cover design for integrated system test.

167

•

•

t lSIOcm

l

t ~

~ &

ZXm

' t IIOclll

~ -t

8cm

~

LEGEND =

I P1.A)( V1 n I •• ,. ,. I 2 2 DJUJ" 2

s• •• ,. I TO~IL DRAIN

M~-------- 10m--------• I SIDE van I

r-CR.J3S

r-

r-- 1---

~ TOPS:HL ~ 1-- CRAY£l ~ 1--

S I

S I

/ oruca (() :II ,.. roP9J I L DR.Al N

C:OBBL.t ~ ,_!II 1--

1 BAC'XnU.. 1

~

I END vu:w I

TOPS::>IL T1UC£R 6 I (I) eocm

I I ' ~~ !\,.OPE 2 2 ' CJUYEL

31cm

~

alBBL!

TIUCER (c..Br)

BACJCP"t l.L

e NEUTRON ACX:C3S TUB!

l PYSCHROW.E"I £RS. POROUS CUPS 2 TENS I OW:t:l £RS

S FLOW Wl1£R (MO'r TO S:HI)

Fig. 14. Improved trench cover design for integrated systems test.

,)

)

• ) lfiR

'

with a standard end plate and gutter system to allow for collection and measurement of runoff and erosion, should it occur (11).

Because this project emphasizes water percolation and biological intru­sion (and not runoff and erosion) the surface of the soil in each plot will be treated similarly. The topsoil surface is graded to a 2% slope along the 10m length dimension (see Fig. 12) to minimize runoff is erosion and is vegetated with a mixture of native grasses that are used for revegetating low-level waste sites at Los Alamos.

Precipitation Addition Regime As shown in Fig. 12 we will apply supplemental precipitation to estab­

lish and maintain the vegetative cover and to roughly approximate the long­term average precipitation at Los Alamos (45 em). This average annual value of 45 em (and the approximate monthly distribution) represents a target level. We may have more or slightly less precipitation depending upon the weather observed during the experiment and the amount of supplemental irri­gation required for the vegetative cover. The value of about 90 em per year represents roughly twice the average annual value and will be obtained by adding artificial precipitation. This enhanced level of precipitation will stress the systems to failure levels and thus determine operational limits under actual field conditions, after data have been collected over a suffi­cient time to characterize system performance under more usual climatic con­ditions.

Tracer Methods Two hydrologic tracer ions, bromide and iodide, will be emplaced to

demonstrate movement of water through the various zones of the trench cap. Those ions have been previously shown, in our work and that of others, to be conservative in experiments of this type. Bromide in the chemical form of lithium bromide (LiBr) was added to the topsoil of the improved treatment plots (Fig. 14), to provide confirmation of lateral (due to the wick) and downward migration of water through those plots.

The interface between the bottom of the trench cap and the underlying layer of crushed tuff backfill material in all plots was similarly treated with cesium, strontium, and iodide ions (Fig. 14). Cesium is known (our pre­vious work) to be relatively immobile in terms of solution transport, due to strong adsorption on soil/rock particle surfaces, but cesium is readily available to plant roots which may penetrate to the tracer zone. Thus, cesium in above-ground plant tissue samples can be taken as indicative of root penetration of the rock biobarrier or backfill material. Strontium is known (from our caisson work) to be relatively mobile in the subsurface environment, and thus will serve both as an indicator of root penetration and as an indicator of water penetration to the lower zone. The primary indicator of water movement in the lower zone will, however, be the iodide ion (because it is a conservative ion and is weakly adsorbed).

Data Collection and Analysis

The performance of the control and improved trench cover designs will be demonstrated by a suite of chemical and hydrologic parameter measure­ments. These measurements will be conducted primarily by automated, off­shelf, computer-controlled data acquisition system techniques involving a

169

•

•

•

combination of in-situ sampling and analysis, together with laboratory analysis where necessary. A subsidiary goal of the project is the demonstra­tion of appropriate, state-of-the-art monitoring technology for use in con­junction with the SLB technology itself.

The amount (percentage) and availability (tension) of the vadose mois­ture is designed to be determined (Figs. 13 and 14) by a combination of mutually reinforcing techniques. Each plot was equipped with tensiometers, gypsum blocks, thermocouple psychrometers, and access tubes for neutron moisture gauge readings.

The monitoring of ground water chemical parameters such as pH, redox potential (Eh), and dissolved carbon dioxide in the SLB environment may be of critical importance because of their effect on the migration of both radionuclides.

The flow rate of leachate issuing from the test plot drainage outflow pipes will be measured using a combination of both paddle-wheel, electronic flow meters for high-flow situations, and fill-dump, pulse-transmitting gauges for the usual low-flow situations. Flow meter and gauge signals will be transmitted to the DAS and incorporated in the data base.

REFERENCES

1. Hakonson, T. E., "Evaluation of Geologic Materials to Limit Biological Intrusion Into Low-Level Radioactive Waste Disposal Sites," Los Alamos National Laboratory report (in preparation).

2. Lane, L. J., end J. W. Nyhan, "Water and Contaminant Movement: Migration Barriers," Los Alamos National Laboratory report (In preparation).

3. Travis, B. J., "TRACR3D: A Model of Flow and Transport in Porous/ Fractured Media," Los Alamos National Laboratory report LA-9667-MS (May 1984).

4. Perkins, B., end B. Travis, "Soli Water Flow Under Saturated/Unsaturated Conditions--Validation of the TRACR3D Code In Three Experiments," Los Alamos National Laboratory report (in preparation).

5. Abeele, W. V., ''Hydraul lc Testing of Crushed Bandelier Tuff," Los Alamos National Laboratory report LA-10037-MS ( 1984).

6. DePoorter, G. L., W. v. Abeele, and B. W. Burton, "Experiments to Determine the Migration Potential for Water and Contaminants in Shallow Land Burial facilities: Design, Emplacement, and Preliminary Results, Waste Management '82, Vol. I 1," in Proceedings of the Symposium on Waste Management, Tucson, Arizona, March 8-11, 1982, pp. 649-655.

7. Abeele, W. V., end G. L. OePoorter, "Field Scale Determinations of the Saturated and Unsaturated Hydraulic Conductivity of Porous Materials. 1. Crushed Bandelier Tuff," (submitted to Water Resource Research) •

170

)

)

)

...

8. Hakonson, T. E., G. C. White, and E. M. Karlen, "Evaluation of Geologic Materials to Limit Biological lntruslon of Low-Level Waste Site Cover," ANS Topical Meeting, Richland, Washington (1982).

9. Hakonson, T. E., J .. F. Cline, and W. H. Rickard, "Biological Intrusion Barriers for Large Volume Waste Disposal Sites," M. G. Yalcintas, Ed., in Symposium on Low-Level Waste Disposal, Facil tty Design, Construction, and Operating Practices, NUREG/CP-0028, CONF-82091 I, Vol. 3, pp. 289-308 (1982).

10. J. W. Nyhan and L. J. Lane, "Use of a State-of-the-Art Model in Generic Designs of Shallow Land Repositories for Low-Level Wastes," R. G. Post, Ed., In Waste Management '82, Tucson, Arizona, pp. 235-244 (1982).

11. Nyhan, J. W., G. L. DePoorter, B. J. Drennon, J. R. Slmanton, and G. R. Foster, "So II Eros I on and Water Ba I ance of Earth Covers Used In Sha 1 1 o~· Land Burt·al of Radioactive Material," Journal of Environmental Quality (1983).

171