Characterization of Unsaturated Hydraulic Conductivity at the ...

July 1988

Prepared for the U.S. Department of Energy under Contract DE-AC06-76RLO 1830

Characterization of Unsaturated Hydraulic Conductivity at the Hanford Site

M. 1. Rockhold M. J. Fayer C. W. Gee

Pacific Northwest Laboratory Operated for the U.S. Department of Energy by Battelle Memorial Institute

DISCLAIMER

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor Battelle Memorial Institute, nor any or their employees, makes any warranty, expressed or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily consti- tute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof, or Battelle Memorial Institute. The views and opinionsof authorsexpressed herein do not necessarily stateor reflect thoseof the United States Government or any agency thereof, or Battelle Memorial Institute.

PAC1 FlC NORTHWEST LABORATORY operated by

BA'TTELLE MEMORIAL INSTITUTE for the

UNITED STATES DEPARTMENT OF ENERGY under Contract DE-AC06-76RLO 7830

Printed in the United States of America Available from

National Technical Information Service United States Department of Commerce

5285 Port Royal Road Springfield, Virginia 22161

NTlS Price Codes Microfiche A01

Printed Copy

Price Pages Codes

CHARACTERIZATION OF UNSATURATED HYDRAULIC CONDUCTIVITY AT THE HANFORD SITE

M. 1. Rockhold M. J. Fayer G. W . Gee

July 1988

Prepared f o r t h e U.S. Department o f Energy under Contract DE-AC06-76RLO 1830

P a c i f i c Northwest Laboratory Rich1 and, Washington 99352

EXECUTIVE SUMMARY

The quantification of water movement in the unsaturated zone is an objective of the Hanford Site Performance Assessment Program, sponsored by the U.S. Department of Energy. This program is being conducted by the Pacific Northwest Laboratory (PNL) and West i nghouse Hanford Company. To accomplish this objective, PNL has undertaken a study to evaluate methods for measuring and predicting unsaturated hydraulic conductivities. This report details some recent field measurements and compares predicted and measured values of hydraulic conductivities for three locations at the Hanford Site.

Measurements from small (6-cm-dia.) "point" and large (2-m by 2-m) "plot" areas utilized infiltration and drainage techniques to obtain in situ data for field-saturated and unsaturated hydraulic conductivity. The Guelph permeameter was used for point sampling, and the unsteady drainage-flux method was used on plots for field-saturated and unsaturated hydraulic conductivi ty measurements. Steady-state techniques were used to measure unsaturated hydraulic conductivities in small columns in the laboratory for one of the three soils tested to provide a comparison with data obtained from the field.

The sandy soil at the Buried Waste Test Facility near the 300 North Area buri a1 grounds yielded hydraul ic conductivity values that ranged over five orders of magnitude (6 x 10-3 to 9 x 10-8 cmls) . These values represent a water content range from field saturation to "field capacity" (i .e., well drained), corresponding to values of 0.303 and 0.096 cm31cm3, respectively. The laboratory and field measurements agree within a factor of five for the range of measured values. A power function relationship describes the field- measured hydraulic conductivity data fairly well for typical water contents observed in the field.

The Grass site is approximately 3 km southwest of the Buried Waste Test Facility. The soil profile at the Grass site is layered, with loamy sand overlying sand. Hydraulic head data indicate that lateral spreading of water

in the upper layers of soil occurred during the drainage phase of the experiment at this site. Calculations of hydraulic conductivity by the unsteady

drai nage-f 1 ux method assume one-dimensi onal (vertical) f 1 ow. Therefore, data

from the unsteady drainage-flux experiment at this site were not used to cal cul ate unsaturated hydraul ic conductivities. The experiment was repeated with modifications to ensure one-dimensional (vertical) water movement, and these data are being evaluated.

The McGee Ranch is approximately 37 km northwest of the Buried Waste Test Faci 1 i ty. The unsteady drai nage-f 1 ux method was used to calculate hydraulic conductivity for the fine (loam and silt loam) soil that occurs at this site. The unsaturated hydraulic conductivities at the McGee Ranch range from approximately 9 x 10-4 to 8 x 10-7 cmls over a water content range of 0.40 to 0.16 cm31cm3, i.e., from field saturation to field capacity.

Measured unsaturated hydraulic conductivities and those predicted from particle-size distribution and bulk density data agree within one-half to one and one-half orders of magnitude, depending on soil type. To use a particle- size distribution to estimate water retention characteristics and, subsequently, to predict unsaturated hydraulic conductivities, measurements of water-retention characteristics are necessary to determine a parameter value used in one of the models. Additional soil types would need to be analyzed to determine if a single value of the parameter can be used that will enable adequate prediction of hydraulic conductivities for the soil types of interest on the Hanford Site. If the predictive techniques can be refined and correlated with field measurements, unsaturated hydraul ic conductivities could be adequately estimated without actual field measurements. This information could then be used in the evaluation of potential remediation or disposal sites or for characterizing larger areas for recharge estimation.

No single method for measuring or calculating unsaturated hydraul ic conductivities is appropriate for all Hanford Site soils. Ideally, several methods should be used to take advantage of the strengths of each method, considering the data needs and resources avai 1 able.

ACKNOWLEDGMENTS

We would l i k e t o acknowledge John McBride, Paula H e l l e r , and Randy

Hayden o f t h e P a c i f i c Northwest Laboratory (PNL) f o r per forming 1 aboratory Z analyses used i n t h i s repo r t . We a l so express spec ia l thanks t o Tim Jones

(PNL) f o r rev iewing t h i s document, and t o D r . James B. Sisson (Kansas S ta te

Un ive rs i t y ) f o r h e l p f u l discussions. Thanks are a l so expressed t o Chr is f

Morgan (PNL) f o r word processing assistance.

EXECUTIVE SUMMARY .................................................... ACKNOWLEDGMENTS ...................................................... 1.0 INTRODUCTION .................................................... 2.0 METHODS .........................................................

2.1 STEADY-STATE FLUX CONTROL .................................. 2.2 UNSTEADY DRAINAGE-FLUX ..................................... 2.3 GUELPH PERMEAMETER ......................................... 2.4 PREDICTIVE TECHNIQUES ......................................

3.0 SITE DESCRIPTIONS ............................................... 3.1 BURIED WASTE TEST FACILITY ................................. 3.2 GRASS SITE ................................................. 3.3 McGEE RANCH ................................................

4.0 RESULTS AND DISCUSSION .......................................... 4.1 BURIED WASTE TEST FACILITY .................................

.................... 4.1.1 Steady-State F lux Contro l Method

....................... 4.1.2 Unsteady Drainage-Flux Method

4.1.3 Guelph Permeameter Method ........................... ......................................... 4.1.4 P red i c t i ons

4.2 GRASS SITE ................................................. ....................... 4.2.1 Unsteady Drainage-Fl ux Method

........................... 4.2.2 Guelph Pemeameter Method

......................................... 4.2.3 P red i c t i ons

4.3 McGEE RANCH ................................................ ....................... 4.3.1 U n s t e a d y D r a i n a g e - F l u x M e t h o d

........................... 4.3.2 G u e l p h P e r m e a m e t e r M e t h o d

4.3.3 P r e d i c t i o n s ......................................... 5 . 0 CONCLUSIONS AND RECOMMENDATIONS ................................. 6 . 0 REFERENCES ...................................................... APPENDIX A - WATER RETENTION AND HYDRAULIC CONDUCTIVITY DATA ......... APPENDIX B . P A R T I C L E- S I Z E DATA ......................................

v i i i

FIGURES

Locat ion o f F i e l d S i t e s ........................................ S o i l P r o f i l e s a t t h e F i e l d S i t e s ............................... Layout of t h e Bur ied Waste Test F a c i l i t y Caissons .............. Measurements o f Unsaturated Hydrau l ic Conduct iv i ty o f L-So i l by t h e Steady-State Flux Contro l Method ........................ Water Content P r o f i l e s Observed During t h e Unsteady Drainage- F lux Experiment i n t h e Southeast Caisson ....................... Hydrau l ic Conduct iv i ty as a Funct ion o f Water Content from t h e Unsteady Drainage-Flux Experiment i n t h e Southeast Caisson, from Repacked Columns i n t h e Laboratory, and from t h e Lax So lu t i on (Si sson, Ferguson, van Genuchten 1980) f o r t h e Watson (1967) K(9) Rela t ionsh ip ................................ Depth (z)/Time (t) Versus Water Content from t h e Unsteady Drainage-Flux Experiment i n t h e Southeast Caisson .............. Depth (z)/Time ( t ) Versus Scaled Water Content from t h e Unsteady Drainage-Flux Experiment i n t h e Southeast Caisson ..... Water Content P r o f i l e s Observed During t h e Unsteady Drainage- F lux Experiment i n t h e North Caisson ........................... Hydrau l ic Head P r o f i l e s Observed During t h e Unsteady .................. Drainage-Flux Experiment i n t h e North Caisson

Hydrau l ic Conduct iv i ty as a Funct ion o f Water Content from t h e Unsteady Drainage-Flux Experiment i n t h e North Caisson, from Repacked Columns i n t h e Laboratory, and from t h e Lax So lu t i on f o r t h e Watson K(8) Re la t ionsh ip ............................... Hydrau l ic Conduct iv i ty as a Funct ion o f M a t r i c Head from t h e Unsteady Drainage-Flux Experiment i n t h e North Caisson, from Repacked Columns i n t h e Laboratory, and t h e K(h) Re la t ionsh ip Determined from t h e Average o f 15 Sets o f Guelph Permeameter Measurements a t t h e Bur ied Waste Test F a c i l i t y ................. Water Retent ion Curves F i t t o Data from t h e Unsteady Drainage- F lux Experiment i n t h e North Caisson and t o Water Retent ion Charac te r i s t i cs Predicted by t h e Arya-Pari s (AP) (1981) Model ..........................................................

4.11 Hydrau l ic Conduc t i v i t y as a Funct ion o f Water Content from t h e Unsteady Drainage-Flux Experiment i n t h e North Caisson and Pred ic ted Curves Based on t h e Arya-Paris (AP) (1981) Model Resul ts Shown i n Figure 4.10 ............................. 4.18

4.12 Water Content P ro f i 1 es Observed During t h e Unsteady Drainage- F lux Experiment a t t h e Grass S i t e 4.22 a ..............................

4.13 Water Retent ion Data from t h e Unsteady Drainage-Flux Exper i- ment a t t h e Grass S i t e ........................................ 4.23 !!

4.14 Hydrau l ic Head P r o f i l e s Observed During t h e Unsteady Drainage- F lux Experiment a t t h e Grass S i t e .............................. 4.24 .

4.15 Water Content P r o f i l e s Observed During t h e Unsteady Drainage- Flux Experiment a t t h e McGee Ranch ............................. 4.27

4.16 Hydraul i c Head Pro f i l e s Observed During t h e Unsteady Drainage-Flux Experiment a t t h e McGee Ranch .................... 4.28

4.17 Water Retent ion Data from t h e Unsteady Drainage-Flux Experiment a t t h e McGee Ranch .................................. 4.29

4.18 Hydrau l ic Conduc t i v i t y as a Funct ion o f Water Content f rom t h e Unsteady Drainage-Fl ux Experiment a t t h e McGee Ranch ....... 4.30

4.19 Hydrau l ic Conduc t i v i t y as a Funct ion o f M a t r i c Head f rom t h e Unstead Drainage-Flux Experiment a t t h e McGee Ranch and t h e K(h J Rela t ionsh ips Determined from t h e Average o f 9 Sets .......... o f Guelph Permeameter Measurements a t t h e McGee Ranch 4.32

4.20 Water Retent ion Curves F i t t o Data from t h e Unsteady Drainage-Flux Experiment a t t h e McGee Ranch and t o Water Retent ion C h a r a c t e r i s t i c s Predicted by t h e Arya-Paris (AP) ................................................... (1981) Model 4.34

4.21 Hydrau l ic Conduc t i v i t y as a Funct ion o f Water Content f rom t h e Unsteady Drainage-Flux Experiment a t t h e McGee Ranch and Pred ic ted Curves Based on t h e Arya-Paris (AP) (1981) Model Resul ts Shown i n F igure 4.20 ................................... 4.35

TABLES

Resul ts from t h e Guelph Permeameter f o r t h e Bur ied Waste Test F a c i l i t y ....................................................... Curve- F i t t i ng Resul ts from t h e RETC.F77 Computer Program Based on Data from t h e Bur ied Waste Test F a c i l i t y .................... Resul ts from t h e Guelph Permeameter f o r t h e Grass S i t e .........

........ Resul ts from t h e Guelph Permeameter f o r t h e McGee Ranch

Curve- F i t t i ng Resul ts f rom t h e RETC.F77 Computer Program Based on Data from t h e McGee Ranch ................................... Steady-State F lux Cont ro l Method Resul ts f o r L-Soi 1 ............ Water Content Data from t h e BWTF Southeast Caisson ............. Hydrau l ic Conduc t i v i t y Data from t h e BWTF Southeast Caisson ....

................. Water Content Data from t h e BWTF North Caisson

................... M a t r i c Head Data from t h e BWTF North Caisson

........ Hydrau l ic Conduc t i v i t y Data from t h e BWTF North Caisson

Water Content Data from t h e Grass S i t e ......................... M a t r i c Head Data from t h e Grass S i t e ........................... Water Content Data f rom t h e McGee Ranch ........................ M a t r i c Head Data from t h e McGee Ranch ..........................

............... Hydrau l ic Conduc t i v i t y Data from t h e McGee Ranch

P a r t i c l e - S i z e D i s t r i b u t i o n Data from t h e Bur ied Waste Test F a c i l i t y .................................................. P a r t i c l e - S i z e D i s t r i b u t i o n Data from t h e Grass S i t e ............ P a r t i c l e - S i z e D i s t r i b u t i o n Data from t h e McGee Ranch ...........

1.0 INTRODUCTION

Various disposal systems have been reviewed for the long-term disposal and isolation of hazardous wastes. At the Hanford Site, one concern is that water draining through the unsaturated sediments may carry contaminants to the water table. A fundamental property of the unsaturated sediments that controls the rate at which water transports contaminants is the hydraulic conductivity (U.S. Department of Energy 1987, Appendix M) . For this reason, the Hanford Site Performance Assessment (HSPA) program is evaluating various procedures for measuring and predicting hydraul ic conductivities of soi 1s at the Hanford Site. Although this report uses the term "soils", the methods outlined can be applied to most of the near-surface unsaturated sediments found on the Hanford Site. Note that this report considers the simple case where water is the only wetting fluid. Whenever multiphase flow and physicochemical interactions are significant, these methods must be modified.

The purpose of this report is to present the results of one research project that used three techniques to measure and one technique to predict unsaturated hydraulic conductivities of soils from three locations on the Hanford Site. Objectives of this study were not only to measure and predict unsaturated hydraul ic conductivities by various methods, but a1 so to compare the methods and, if possible, determine which technique(s) provides the most re1 i abl e results.

For each measurement technique used, water flow was measured and the hydraulic conductivity calculated from the appropriate form of Darcy's Law. The technique used in this study for making measurements in the laboratory is a modification of the steady-state flux method of Klute and Dirksen (1986). The modification involved controlling the flux of water into the soil columns with equipment described by Wierenga et a1 . (1986). The techniques used for making measurements at the field sites included the unsteady drainage-flux method (Green, Ahuja, and Chong 1986) and the Guelph permeameter method (Reynolds and El rick 1985) . The steady-state f 1 ux and unsteady drainage-f 1 ux methods have traditional ly been the most accurate techniques. Both tech-

niques are relatively time consuming. Consequently, they may be impractical

for making the large number of measurements needed to characterize areas

(Ni el sen, Biggar, and Erh 1973) . The Guel ph permeameter was used in addition

to the other methods because of its speed, low-water-use requirements, and

portability.

Methods of predicting unsaturated hydraulic conductivity rely on

description of the water retention curve (WRC) rather than measurements of

water flow. Mualem (1986) and van Genuchten (1978) describe many of these

methods. A WRC relates the volumetric water content to the soil water potenti a1 . The WRC can be determined in the 1 aboratory or in the field. Field

measurements o f water retention require more effort than laboratory

measurements, especial ly for relatively dry conditions. An alternative to

measuring water retention characteristics is to predict them from soil tex-

tural and structural properties. This can be done in a variety of ways,

including multSpie regression techniques that relate water contents at

specified soi 1 -water pressures to texture and bulk density (e.g., Hal 1 et a1 . 1977; Gupta and Larson 1979).

In this study, prediction of water retention characteristics is based on

an empirical model by Arya and Paris (1981) which also uses particle-size

distribution and bulk density data. This type of analysis is potentially

attractive for use at the Hanford Site because particle-size distribution

data have already been collected from numerous test and observation wells

(i .e., the Westinghouse Hanford Company grain-size data base).

This report provides hydraulic conductivity data for three test loca-

tions at the Hanford Site (see Figure 1.1) : l) the Buried Waste Test Facil-

ity (BWTF) , described by Phil 1 ips et a1 . (1979) ; 2) the Grass site, described by Gee and Ki rkham (1984) ; and 3) the McGee Ranch, described by Last et a1 . (1987). The methods used for measurements, predictions, and data interpreta-

tions of unsaturated hydraulic conductivity are presented in the sections

that follow. Physical property data from the three test locations are

provided in the appendixes.

HANFORD SITE

Ponland

BURIED WASTE TEST FACILITY (BWTF)

(REF) PSB612-7

0 8 KILOMETERS

w rn 0 5 MILES

FIGURE 1.1. Locat ion o f F i e l d S i t e s

2.0 METHODS

Various methods have been developed for measuring the hydraulic conductivity of soils in the laboratory and field (Klute and Dirksen 1986).

t Field measurements are general ly considered to be more representative of actual soil properties and conditions than laboratory measurements, but require more effort. Good agreement between field and laboratory data is often difficult to obtain because the natural soil heterogeneity of in situ soi 1 s is usual ly not represented in 1 aboratory samples . Val id correlations are also made difficult by problems encountered in field studies, such as incomplete saturation, hysteresis effects, and preferential flow.

Methods that predict hydraulic conductivity based on particle-size distribution and bulk density data are generally easier to use than field or laboratory methods, but yield results with more uncertainties than those determined experimentally. These methods are usually based on simplifying assumptions and typically require a considerable amount of field or laboratory data for initial parameter estimation, and in order to make defensible predictions.

A problem affecting all methods is the attainment of a high degree of accuracy. In addition, no technique is completely reliable or adequately deals with all problems of measurement scale, spatial variability, and sample representativeness for all conditions. Consequently, the approach used in this study wi 1 1 be to compare the results of several techniques, and to use the results of the steady-state flux and unsteady drainage-flux methods as standards of relative accuracy.

To understand how various measurement techniques work, it is important to understand the processes controlling water flow in soil. Water moves in

, an unsaturated soil as liquid and vapor. Under isothermal conditions, water generally moves from regions of higher to lower potential energy. This . - potential energy, H, can be expressed as

A

where h p = pressure potential hS = solute potential hm = matric potential h Z = gravitational potential .

Pressure potential represents external forces, such as water ponded on the surface of a f i e l d plot during the in f i l t r a t ion phase of an unsteady drainage-flux method experiment. Solute potential represents the a t t r ac t ive forces of water t o higher solute concentration or osmotic forces. Matric potential represents the capillary and adsorptive forces which a t t r a c t and bind water t o the so i l matrix. Gravitational potential i s the energy associated with the location of water in the Earth's gravitational f i e l d , measured with respect t o some reference point such as the so i l surface. In most cases, pressure and solute potential are considered negligible. Conse- quently, the to t a l potent ial , in the context of t h i s report , i s the sum of

the matric potent ial , h(cm) , and the gravitational potential (or ver t ical distance from the so i l surface) , z(cm). The sum of matric and gravitational potenti a1 s , when expressed on an equivalent height-of -water basis , i s known as the hydraulic head. The to ta l hydraulic head, as used in t h i s report , consists of the matric head and the gravitational head.

The f lux (q) of water through so i l i s proportional t o the hydraulic head gradient (dH/dz) . For saturated s o i l s , the flux can be determined with the Darcy flow equation

where KS i s the saturated hydraulic conductivity ( i .e . , the proportionality factor) . For unsaturated s o i l s , the hydraul i c conductivity i s nonl inearly re1 ated t o the matric head or water content. Equation (2.2) i s usual ly modified t o be

where K i s defined as the flux of water per unit gradient of hydraulic head and B i s the volumetric water content or volume of water per unit bul k volume of s o i l . To describe t ransient , vertical flow, Equation (2.3) must be com- bined with the equation of continuity

where t i s time and z i s depth. This combination i s commonly known as the Richards equation (Richards 1931) .

2.1. STEADY-STATE FLUX CONTROL

Laboratory determi nations of hydraul i c conductivities by the steady- s t a t e flux control method were made using the general method described by Kl ute and Dirksen (1986). The method was modified by control 1 ing the flux of water into the so i l columns with equipment described by Wierenga e t a l . (1986). An acryl ic cylinder of known volume was packed with so i l t o a pre- scribed bulk density. The lower end of the cylinder was covered with a porous s ta in less s teel plate (bubbling pressure = 245 cm H20) within an acryl ic end cap. The end cap had a f i t t i n g t o a1 low connection t o a vacuum chamber. Rubber O-ring seals within the cap ensured an a i r t igh t seal between the cyl inder and the cap. The upper end of the cylinder was covered by an acryl ic cap with a f i t t i n g tha t allowed connection t o a syringe pump and solution reservoir. The top end cap was f i t t e d loosely on the cylinder, so tha t the a i r above the so i l was a t atmospheric pressure. The acryl ic cylinder had two tensiometer ports, a t 5 cm and 25 cm above the s ta in less s teel plate.

The syringe pump was adjusted t o pulse a small volume of water a t regu- l a r intervals t o establish steady-state flow conditions through the so i l column. The pulse volume was minimized and pulse frequency maximized t o the

extent possible. The starting point was a flux equal to the saturated hydraulic conductivity, with a unit gradient or hydraulic head difference equal to

the distance between tensiometers (20 cm) . To establ i sh unsaturated condi - tions within the column, vacuum was applied to the vacuum chamber and the

7 bottom of the column. The syringe pump was adjusted to reduce the flux of water into the top of the column, so that the fluxes entering and exiting the cylinder were equal. This steady-state condition was determined by moni- k

toring the tensiometers with a TENSIMETER pressure transducer (Soi 1 Measure- ment Systems, 1906 South Espina, Las Cruces, NM 88001). When the readings of . both tensiometers were equal, steady hydraulic flow and a uniform volumetric water content were assumed to exist (i .e., unit gradient conditions). For these unit gradient conditions, Equation (2.3) reduces to q = -K(8) and the conductivity is equal to the input flux. The water content associated with the input flux (i .e., hydraulic conductivity) was determined by weighing the entire soil column. The reference weight for the soil column was the weight at approximately 100% saturation. As a datum check at the end of each experiment, the soil was removed from the column and oven dried to calculate a gravimetric water content. Applying higher suctions to the bottom of the column and reducing the input flux appropriately a1 lowed measurement of unsaturated hydraulic conductivities over the range of 0 to -196 cm of matric head.

This method was only used for determining unsaturated hydraulic conductivities of L-soil (97% sand, 2% silt, 1% clay), which is the laboratory designation for soil collected from the BWTF site in 1978 (Phillips et al. 1979; Cass, Campbell, and Jones 1981). Two repetitions with L-soi 1 were conducted at each of two bulk densities, 1.6 and 1.7 g/cm3. We assumed that these 1 aboratory samples are texturally equivalent to samples subsequently collected from this site.

2.2 UNSTEADY DRAINAGE-FLUX

The unsteady drai nage-f 1 ux method is based on Darci an analysis of trans- ient in situ soil-water content and hydraulic head profiles during vertical drainage from field plots. The method, as used in this study, consisted of

ponding water on the surface of a plot until the profile was wetted beyond the maximum depth of interest. The soi 1 surface was then covered with clear plastic and a thin (approximately 3-cm-thick) layer of soil to prevent evaporation and to minimize thermal effects. Isothermal conditions were assumed to exist in the profile during drainage. Water contents and hydraulic heads were then monitored as the water in the profile redistributed and drained.

Ponding was facilitated by using existing caisson walls (e.g., at the BWTF site), using planking installed in narrow trenches around which soil was thoroughly compacted (e.g., at the Grass site), or by berming soil around the plot (e.g., at the McGee Ranch site). Water was supplied from an observation well via an electric pump at the BWTF site, and by hauling water by truck to the other two sites. Water contents were monitored with a model 503DR Hydro- probe (Campbell Pacific Nuclear Corp., 2830 Howe Rd., Martinez, CA 94553) inserted into steel or aluminum access tubes installed vertically in each plot. Matric heads were measured with tensiometers and a TENSIMETER pressure transducer.

Tensiometer and neutron probe readings were taken every 10 to 15 min during the initial drainage and redistribution phase of each experiment, and less frequently as time passed. The tensiometers were placed at 15- to 30-cm-depth increments, down to 180 cm at the BWTF and Grass sites and to 120 cm at the McGee Ranch site. All tensiometer measurements were referenced to the soil surface. Neutron probe readings were taken at depths corresponding to tensiometer placement, with the exception of the BWTF southeast caisson study, where no tensiometers were installed. Tensiometers were not installed in the southeast caisson because the caisson was not large enough to place them far enough away from the neutron access tube so that probe readings would not be affected by the water in the tensiometers. Volumetric water content was determined from neutron probe count readings by field cali- brations at each site.

The unsteady drainage-flux method was first used for field measurements by Richards, Gardner, and Ogata (1956). Further developments in the method

were made by Nielsen et. a1 (1964); Rose, Stern, and Drummond (1965); and

Watson (1966). The actual computations of hydraulic conductivity used in

this study are based on the time-averaging method used by Rose, Stern, and

Drummond (1965), and the instantaneous profile method (after Watson 1966).

To obtain the value of K at depth, L, Equation (2.5) can be integrated ' b

with respect to z, from the soil surface (z = 0) to the maximum depth of

interest (z = -L), by the following equation

Because there is no flow across the plastic-covered soil surface, the second

term on the right-hand side of Equation (2.6) effectively becomes zero. Sub-

stituting (h + z) for H and rearranging Equation (2.6) yields

The values on the right-hand side of Equation (2.7) are evaluated to deter-

mine K(0) at selected times for each depth of measurement.

Using a time-averaging approach, the integral, ae/at dz, of Equa- l tion (2.7) can be estimated by trapezoidal approximation for each depth

interval, as described by Green, Ahuja, and Chong (1986). The water content

from the surface (z = 0) to the first depth of measurement is taken as that

measured at the first depth. For example, for data points at 30-cm-depth

intervals and at depth, zl,

where Bi is the soil-water content measured at the it h point in the profile, measured from the top of the profile, and n is the number of data points down

to depth, zl. The total head gradients are then approximated by

where all variables have been defined previously. Alternatively, head gradients can be determined by curve-fitting techniques as outlined by Green, Ahuja, and Chong (1986). Fluxes are calculated at each depth and measurement time to be equal to the volume change in water stored between measurement depths during a given time interval, as determined from the previously described trapezoidal integration procedure. Time-averaged gradients and water contents are then calculated, and hydraulic conductivity values corresponding to the time-averaged water content are determined by dividing the calculated fluxes by the time-averaged gradients.

Using an instantaneous profile approach, volumetric water content is plotted versus time for each depth of measurement, and curves are fit to these data. The slopes of these curves (-aelat) are then measured at selected times and multiplied by their respective depth increments to obtain the per-layer rate of water content change. The flux through the bottom of each layer is then calculated by accumulating the water content increments of a1 1 layers overlying that depth [i .e., q = (a8/Ot)/dz]. Matric head values are plotted versus time, and the depth of each tensiometer is added to each matric head value to obtain total hydraulic head profiles. Then, the hydraulic conductivity is calculated by dividing the flux values by their corresponding hydraul ic head gradient values.

The time-averaging and instantaneous profile procedures should yield similar results, especially with data from soi 1 profiles that are relatively uniform by depth. Differences between the results obtained by the two procedures are caused by the different approximations of the differential and integral quantities.

Black, Gardner, and Thurtel 1 (1969) studied drainage 1 osses from lysimeters and noted that the "unit gradient" condition was often valid. Davidson et al. (1969) rewrote Equation (2.5) in unit gradient form such that

Using Equation (2.7) t o estimate hydraul i c conducti,vi t i e s requires knowledge of the r a t e of change in water content and the hydraulic head gradient. The unit gradient method modifies t h i s data requirement by assuming tha t the head

gradient i s uniformly equal t o 1. This condition ar i ses when the water conten t i s nearly uniform with depth, and resu l t s in BhIBz RJ 0 and BHIBz w 1.

Sisson, Ferguson, and van Genuchten (1980) solved Equation (2.10) by using a solution scheme proposed by Lax (1972). This solution can be used in t

two ways. F i r s t , i f so i l hydraulic properties are known, the solution describes the water content prof i le between the so i l surface and the advanc- ing drainage front . Second, i f water content i s measured during drainage, the solution can be used t o estimate so i l hydraulic properties. Both appli- cations are limited by the val idi ty of the unit gradient assumption.

Sisson (1987) extended the concept of a unit gradient t o a "fixed gradient", where BHIBz may not be identical t o 1 , b u t i s a function of depth, and i s invariant with time. Scaling theory i s incorporated into the assumption of a fixed gradient t o define new water content and space variables. The fixed gradient thereby becomes a unit gradient, when written in terms of the scaled variables. This extension a1 lows the fixed-gradient problem t o be solved using unit gradient solutions.

The fixed gradient analyses used in t h i s study assume a power function re1 ationship between hydraulic conductivity and water content. This re1 a-

tionship i s the Watson (1967) model

where K f S i s the field-saturated hydraulic conductivity, Om i s the maximum water content obtained during ponding, and P i s an unknown parameter. When the ponding phase of an unsteady drainage-flux method experiment has ended,

the f inal r a t e of in f i l t r a t ion i s used t o estimate Kfs, and Om i s approximated by averaging the water contents a t each depth t o the deepest depth of in te res t . Multiple regression i s then performed on log ( z l t ) versus log 6 t o determine the slopes and intercepts of least-squares f i t s of s t ra ight l ines

t o the data. The depth, z , i s measured from the so i l surface, and the time,

t, i s t h e t ime a t which neutron probe measurements are taken a f t e r ponded water disappears from t h e sur face o f t h e p l o t . These slopes and i n t e r c e p t s

are then used t o sca le t h e data and t o determine t h e p parameter i n t h e

Watson (1967) model.

2.3 GUELPH PERMEAMETER

The Guelph permeameter method (Reynolds and E l r i c k 1985) measures t h e

steady- state r a t e o f water i n t a k e from a c y l i n d r i c a l auger ho le i n which a

constant depth o f water i s maintained. The a i r - i n l e t tube o f t h e Guelph

permeameter i s used t o e s t a b l i s h and main ta in a constant head l e v e l , H, w h i l e

t h e corresponding discharge r a t e , Q, i s measured as t h e r a t e o f discharge

from t h e permeameter water rese rvo i r . Th is method simultaneously measures i n

s i t u f i e l d - s a t u r a t e d hyd rau l i c conduc t i v i t y , Kfs, and m a t r i c f l u x p o t e n t i a l ,

i n t h e unsaturated zone. The Guelph permeameter used i n t h i s study was

obtained from S o i l Moisture Equipment Corp., P.O. Box 30025, Santa Barbara,

CA 93105.

The m a t r i c f l u x p o t e n t i a l i s def ined by Gardner (1958) as

where K(h) i s t h e hydraul i c conduct iv i t y- mat r i c head re1 a t ionsh ip . Calcu-

l a t i o n s us ing t h e Guelph permeameter method assume t h e exponent ia l K(h)

re1 a t i onshi p o f Gardner (1958)

K = Kfs exp (ah); h i J h 0 (2.13)

where a i s t h e s lope o f t h e curve ln (K) versus h, and h i i s t h e i n i t i a l

m a t r i c head i n t h e s o i l . S u b s t i t u t i n g Equation (2.13) i n t o (2.12) and i n t e -

g r a t i n g produces

which simplifies t o

f o r many s o i l s a t "f ie ld capacity" or d r i e r conditions (Scotter, Clothier, and Harper 1982). Field capacity i s not a quantitatively defined water conten t . However, i t can be qual i ta t ively defined as the water content of a <

re la t ively uniform, deep so i l tha t has drained fo r 2 t o 3 days a f t e r thorough wetting. This i s generally considered t o be a water content reached under conditions of no evaporation or water uptake by plants.

. - Steady-state recharge depends on K f S and #rn. The steady-state recharge

r a t e , Q , i s given by

where the f i r s t , second, and th i rd terms on the right-hand s ide of the equation represent the pressure, gravity, and cap i l l a r i ty components, respec- t ive ly . Equation (2.16) i s an approximate analytical solution based on saturated-unsaturated flow theory (Reynolds and El rick 1985), where H i s the head level in the well, a i s the well radius, and C i s the shape of the saturated so i l "bulb" surrounding the well hole. The value of C i s primarily a function of H/a in saturated s o i l s , b u t a lso depends on so i l s t ruc ture , tex- tu re , and i n i t i a l matric head in unsaturated so i l s . Values of C were obtained from standard C-curves in the operating instructions fo r the Guelph permeameter. These standard curves were developed from numerical simulations of steady, saturated-unsaturated flow around we1 1 s in coarse sand, Guel ph

loam, and unstructured clay.

The field-saturated hydraul i c conductivity and matric flux potential in

t h i s study were calculated from steady-state recharge ra tes by a simultaneous equation approach, referred t o as the Richards analysis (GP-R) by Reynolds, Elrick, and Clothier (1985) using Equation (2.16). The GP-R analysis requires two o r more constant head level discharge measurements. Therefore,

when steady-state flow is reached at one head level, the air-inlet tube is simply raised to a different height, and the steady-state recharge at that head level is measured.

* 2.4 PREDICTIVE TECHNIQUES

The RETC.F77 computer program (van Genuchten 1985) was used to fit a )1 mathematical function to the measured and predicted water retention data, and

to predict unsaturated hydraulic conductivities. This program uses non- . linear, least-squares curve fitting to fit a soil WRC of the form

where Or = residual soil water content B s = saturated soil water content h = matric head

a, m, and n = curve-fitting parameters.

Mualem (1976) developed a general model to predict the hydraul ic conductivity from the soil WRC. This model has the form

where

Se = (9 - Br)l(BS - Or), and t. is a parameter.

b Assuming that m = 1 - l/n, van Genuchten (1978) derived a closed-form solu- . - tion to Equation (2.17). This solution is

or in terms of matric head

where Kr (or re1 ative hydraul ic conductivity) is the hydraul ic conductivity divided by the saturated hydraulic conductivity.

One method for predicting the WRC is the physicoempirical model by Arya and Paris (1981) . This is essential ly a capi 11 ary pore model that first translates the particle-size distribution into a pore-size distribution. Cumulative pore volumes, corresponding to increasing pore radii, are divided by the sample bulk volume to give volumetric water content. The pore radii are converted to equivalent matric head values by using the equation of capi 11 ari ty

hi = 27 cosa / pwgri

where hi = soil matric head corresponding to the ith pore increment 7 = surface tension of water a = contact angle of water with soil particles pw = density of water g = gravitational acceleration ri = radius of the ith pore.

In this study, the surface tension was taken as that of pure water at 25OC (71.97 dynes/cm), and the contact angle was assumed to be zero.

To compute the pore volumes and radii, the particle-size distribution is divided into segments. The solid mass in each segment is assumed to form a matrix with a bulk density equal to that of a natural structure sample. For a unit of sample mass, an equivalent pore volume is computed from

and the corresponding pore radius from

where V V ~ = pore volume Wi = solid mass pp = particle density e = void ratio ri = mean pore radius Ri = mean particle radius ni = number of particles a = an empirical constant.

The formulation for the pore radius assumes spherical particles and cylin- dri cal pores.

During the auguring of some of the well holes used for the Guelph permeameter measurements, known volumes of soi 1 were removed from each auger hole at the depth at which permeameter measurements were taken. These samples were sealed in plastic bags to maintain original water content and then oven dried in the laboratory for soil bulk density measurements. A brass cylinder sampler was also used to collect bulk density samples from the Grass site and McGee Ranch unsteady drai nage-f 1 ux experiment plots.

Bulk density samples were also used for determining particle-size distribution by a sieve analysis and hydrometer method (Gee and Bauder 1986). These particle-size distribution and bulk density data were then used to predict water retention characteristics by using the model of Arya and Paris (1981). Predicted water retention characteristics were then fit with the RETC.F77 computer program, and hydraulic conductivities were calculated with

a the program using Mualem's (1976) hydraulic conductivity model. In general, this and other models work best when the predicted hydraulic conductivity values are scaled to one or more measured values. The most common approach

b is to scale the predicted values using the measured saturated hydraulic

conductivity as a matching point between curves.

3.0 SITE DESCRIPTIONS

Three l o c a t i o n s a t t h e Hanford S i t e were se lec ted f o r unsaturated

hydraul i c conduct i v i t y measurements : t h e BWTF, Grass, and McGee Ranch s i t e s .

The l o c a t i o n s o f these s i t e s a re shown i n F igure 1.1. These s i t e s represent

t h r e e d i s t i n c t s o i l p r o f i l e s as shown i n F igure 3.1.

The BWTF and Grass s i t e s are research s i t e s f rom which da ta are being

co l l e c t e d f o r va l i d a t i o n s tud ies o f t h e UNSAT-H unsaturated f l o w code (Fayer,

Gee, and Jones 1986). S o i l f rom t h e McGee Ranch s i t e i s c u r r e n t l y be ing

t e s t e d as t h e sur face cover f o r t h e Hanford S i t e P r o t e c t i v e B a r r i e r s (Kirkham

and Gee 1987). The i n f l u e n c e o f t e x t u r e , b u l k dens i ty , and l a y e r i n g on t h e

h y d r a u l i c p r o p e r t i e s o f s o i l s from these t h r e e l o c a t i o n s i s o f i n t e r e s t f o r

b a r r i e r system design and development, as w e l l as f o r model v a l i d a t i o n .

3.1 BURIED WASTE TEST FACILITY

The BWTF i s l oca ted adjacent t o t h e 300 Nor th Area b u r i a l grounds (see

F igure 1.1). The f a c i 1 i t y cons i s t s ' o f an a r ray o f seven corrugated,

BWTF Grass Site McGee Ranch

1 . . - . . . -- Loam . --

8

8 . 8 @

- - ' 8 v

. FIGURE 3.1. S o i l P r o f i l e s a t t h e F i e l d S i t e s

galvanized-steel caissons of two different diameters, bolted together in the arrangement shown in Figure 3.2, and two weighing lysimeters (not shown).

All seven caissons are 7.6 m long. The three large caissons are 2.7-m dia.

and the four small caissons are 0.6-m dia. These caissons are filled with a

relatively uniform material, consisting of approximately 97% sand, 2% silt, !

and 1% clay (L-soi 1). This soi 1 consists of the same material that was exca-

vated for the facility, but with particles greater than 1.27-cm dia. screened T

out. This facility was originally designed for field water balance and radionuclide transport studies. Construction and originhl instrumentation . specifications are described by Phi 1 1 ips et a1 . (1979).

Samples of L-soil were collected in 1978 during the construction of the BWTF. During the summer of 1986, laboratory measurements of hydraulic conductivity were made on these samples using the steady-state flux control method, described in Section 2.1.

\ Access Port 0.6 m 4

A = North Caisson

B = Southeast Caisson

FIGURE 3.2. Layout of the Buried Waste Test Faci 1 ity Caissons

Two unsteady drainage-flux method s tudies were conducted a t t he BWTF in October 1986. These s tudies were in t he large north caisson and t h e small southeast caisson (A and B, respectively, in Figure 3.2). The upper 20 and 10 cm of f i 11 materi a1 were removed from caissons A and B, respect ively , t o expose t h e tops of t h e caissons. These exposed ends of t he caissons acted as enclosures f o r ponding water during i n f i 1 t r a t i on . In t he southeast caisson,

an addit ional 60 cm of so i l was excavated t o remove a previously emplaced p l a s t i c 1 iner . The removed soi 1 was packed back i n to t he caisson a f t e r removing t h e l i ne r .

Guelph permeameter measurements were taken i n t he area around t he ca i s- sons in September and October 1986, and within t he north caisson in July 1987.

3.2 GRASS SITE

The Grass s i t e i s located approximately 3 km southwest of t he BWTF. I t i s s i tua ted in a broad, shallow topographic depression approximately 900 m wide and several hundred meters long in a northeast-southwest d i rect ion. Ongoing water balance and t ranspira t ion s tudies a r e being conducted a t t h i s 1 ocation (Gee and Ki rkham 1984).

The s o i l a t t he Grass s i t e i s 3.5 m th ick and i s well drained. The upper-most 0.6 m of t he so i l p ro f i l e contains approximately 74% sand, 21% s i l t , and 5% clay, and i s c l a s s i f i ed as a sandy loam t o loamy sand [border- l i n e , but previously c l a s s i f i ed as a loamy sand by Gee and Kirkham (1984)l. From 0.6 t o 3.5 m, t h e so i l consis ts of approximately 91% sand, 6% s i l t , and 3% c lay , and i s c l a s s i f i ed as a sand. A gravel layer t h a t l i e s below the 3.5-m depth i s estimated t o be several meters th ick , based on excavations a t adjacent s i t e s .

This s i t e i s instrumented with 25 neutron-probe access tubes arrayed in a 5 by 5 gr id with a 6-m spacing between tubes. The unsteady drainage-flux experiment conducted a t t h i s s i t e in July 1987 was a repeat of a previous

study (Gee and Kirkham 1984), using t he same p lo t (2 m by 2 m) and neutron- probe access tube (No. 25). The 1984 study was repeated i n an attempt t o

investigate a wider range of water content and t o measure hydraulic head tha t was not measured successfully in the f i r s t study.

Guelph permeameter measurements were made a t depths of 20- and 60-cm fo r various locations around the grid of neutron-probe access tubes in September and October 1986. Additional measurements were made in July and August 1987,

both around and within the unsteady drainage-fl ux experiment plot .

3 . 3 McGEE RANCH

The McGee Ranch i s approximately 37 km northwest of the BWTF. This s i t e has been characterized for near-surface so i l texture and other physical properties (Last e t a l . 1987). The so i l texture a t t h i s s i t e ranges from s i l t loam t o sandy loam. The average part ic le- size dis t r ibut ion of so i l samples collected from the McGee Ranch during t h i s study i s 36% sand, 49% s i l t , and 15% clay, which c l a s s i f i e s the so i l as a loam. The ground surface a t the McGee Ranch slopes 3% t o 5% t o the south.

An unsteady drainage-flux experiment was conducted a t t h i s s i t e in July 1987. The location of the 2-m by 2-m study plot was between the north-south McGee Ranch road and a borrow p i t from which f ine so i l s were taken f o r the Field Lysimeter Test Faci 1 i t y . Several thin (4-cm) cal iche 1 ayers were encountered during ins ta l la t ion of tensiometers a t depths of approximately 35, 80, and 100 cm.

Guelph permeameter measurements were taken in July and August 1987, a t various locations around the borrow p i t a t the McGee Ranch and within the unsteady drainage-flux experiment plot a t t h i s s i t e .

4.0 RESULTS AND DISCUSSION

The following sections describe the results of the methods used at each site. Collected data are reported in tabular and graphic form in the following sections, and in tabular form in the appendixes.

4.1 BURIED WASTE TEST FACILITY

Laboratory measurements of unsaturated hydraul ic conductivity were made on L-soil collected from the BWTF using the steady-state flux control method. Two unsteady drainage-flux method experiments were conducted in the southeast and north caissons (see Figure 3.2). Guel ph permeameter measurements were made in the area immediately surrounding the BWTF site and within the north caisson. Soil samples were collected from the permeameter auger holes and were used for particle-size analysis and subsequent water retention char- acteristic predictions using the Arya and Paris (1981) model. Predicted water retention curves (WRCs) were then used to predict hydraulic conductivi-

ties with the RETC. F77 computer program, using Mualem's (1976) predictive conductivity model.

4.1.1 Steady-State Flux Control Method

Hydraul ic conductivity data from two rep1 ications and two bul k densities for L-soil are displayed on Figure 4.1. The actual 8, h, and K values are listed in Appendix A, Table A.1. The bulk density variation had little discernible effect on the measured hydraulic conductivity values. Each replicated test required one week to pack and saturate the samples and approximately six weeks to collect six to seven data points. Obtaining data points for the lower water contents (achieved with a low flux rate) took the majority of the six-week period, because the time necessary to achieve steady-state flow was 1 onger (because of the low flux rates) . 4.1.2 Unsteadv Drainaae-Flux Method

Water content data for the southeast caisson drainage study are plotted on Figure 4.2 and 1 isted in Appendix A, Table A.2. Because there was no

collection of matric head data during this experiment, we assumed that a unit gradient condition existed. Hydraulic conductivities were then calculated

- - - - - - - O045 - - - CB - - - no - - 0 - - P

0 - - m - - - - - 0 - - 0 m - - - - - - 6 - - -

- - - - - - - - 6 Bulk Density (g/crn3) - - 8 1.6 - - - - - 8 0 1.7 - - -

I I I 1 1 08 0 0.1 0.2 ' 0.3 0.4

FIGURE 4.1 Measurements of Unsaturated Hydraulic Conductivity of L-Soil by the Steady-State Flux Control Method

0.5

using the instantaneous profile method (Watson 1966) and the Lax solution

method (Si sson , Ferguson , and van Genuchten 1980) .

Water Content (cm3/cm3)

Hydraulic conductivities determined using the instantaneous profile

method are plotted on Figure 4.3 and are listed in Appendix A, Table A.3.

The hydraulic conductivity data for soil depths below 90 cm are grouped rela-

tively close together. The hydraulic conductivity data for the three depths

above 90 cm, however, show more variance with respect to water content. We

-

- Time (s)

0 484 - 0 2,280

n 3,480 A 7,080

- 14,300

81,600 4 1,050,oOo -

-

-

-

-

I I I I


FIGURE 4.2. Water Content Profiles Observed During the Unsteady Drainage-Flux Experiment in the Southeast Caisson

believe th is difference resulted from the upper 60 cm of soil being disturbed (to remove a previously emplaced plastic 1 iner) and repacked just prior t o beginning the experiment. The effect of th i s disturbance was t o

create a zone with a lower bulk density than a t the lower depths ( i .e., a 1 ayeri ng effect) .

lo-'

10-2

10-3

A

Q) \

6 - 2 104 .- > .- C 0 =J 0 C 0 0 .- - 3 !! 0 r I

1 o=' Unsteady Drainage-Flux (Field)

Steady-State Flux Control (Laboratory)

1 0-7 - Lax Solution for Watson Model

10" 0 0.1 0.2 0.3 0.4 0.5


FIGURE 4.3. Hydraulic Conductivity as a Function of Water Content from t h e Unsteady Drainage-Flux Experiment i n t h e Southeast Caisson, from Repacked Columns in t he Laboratory, and from t h e Lax Solution (Si sson, Ferguson, van Genuchten 1980) f o r t h e Watson (1967) K(9) Relationship

Also p lo t ted on Figure 4.3 a r e t h e laboratory data from Figure 4.1. The f i e l d data from the upper three depths in the caisson agree w i t h the 1 aboratory data f a i r l y we1 1 , suggesting s imi la r bulk densi ty and packing cha rac t e r i s t i c s between t h e L-soil in t he laboratory and t h e upper t h r ee depths in t h e caisson. Between water contents of approximately 0.12 and 0.25 cm3/cm3, hydraulic conduct ivi t ies from t h e lower depths a r e higher than

the laboratory values by as much as a factor of five. For water contents between 0.10 and 0.12 cm3/cm3, the f i el d-measured conductivities match the laboratory conductivities more closely.

There are several possible explanations for the differences between the laboratory and the field conductivities. As mentioned previously, the variation in packing density expected for field conditions (compared to the re1 ative uniformity of packing within a 1 aboratory column) could have contributed to the differences. Figure 4.1, however, indicates that a bulk density variation of 0.1 g/cm3 had no discernible effect on the laboratory- measured conductivity values. Another explanation is that the neutron probe was not adequately calibrated for the caissons. The neutron probe that was used is undergoing recalibration, but a preliminary analysis of the new cali- bration curve indicates that water contents will not change by much more than 0.01 cm3/cm3, and that calculated conductivities will not change by more than about 5%. A third possibility is that, early in the experiment, a significant amount of entrapped air may have been present (the caisson side ports were sealed and the bottom was partially sealed). The entrapped air would have affected the hydraulic head gradients. Unfortunately, we have no measure of hydraulic head gradients during the experiment and have re1 ied on the assumption of a unit gradient.

Complete saturation of a soil profile is very difficult, if not impossible, to obtain in a field experiment. All pores are not interconnected or

open, and air may become trapped in some of the open pore spaces, effectively preventing water from filling them. If an unlimited water supply were available, and water could be ponded on the plot or the plot irrigated for an extended period of time, much of the entrapped air would dissolve. Unfor- tunately, such conditions are not possible for most field studies of this type. Therefore, curves fit to field-measured water retention data from most unsteady drai nage-f 1 ux method experiments do not represent true desorption curves, but are actually intermediate scanning curves representing the effects of hysteresis (the nonuniqueness of the water content-matric head relationship). In a typical laboratory setup, columns of soil are saturated from the bottom, or under a vacuum, so that air is driven out the top of the

col umn as the soi 1 becomes saturated. Therefore, 1 aboratory WRCs general ly represent t rue desorption curves. These differences are part of the reason why laboratory and field-measured retention and hydraulic conductivity data typically are not i n complete agreement.

The second method for analyzing the southeast caisson data i s based on the Lax solution (Sisson, Ferguson, and van Genuchten 1980). Mu1 t i p l e regression of log (z / t ) , which equals log (dK/d0) , versus log 8 by the method %

of dummy variables, was performed t o determine the slopes and intercepts of these l ines fo r parameter estimation in the Watson (1967) model. The depth, z(cm), i s measured from the soi l surface, and the time, t (days) , i s measured from when water f i r s t disappeared from the soi 1 surface ( i .e., time zero).

The l ines shown on Figure 4.4 are least-squares f i t s t o data from each depth. Eleven regression l ines are portrayed on t h i s figure (one fo r each depth), but some of them fa1 1 on top of each other. Although the sand in the caisson i s re1 at ively uniform with respect t o part ic le- size dis tr ibut ion, the regression l ines representing data from the upper three depths are separated from the other regression lines. I t i s apparent tha t the disturbed so i l was packed t o a lower bulk density than the r e s t of the caisson s o i l , and tha t the hydraulic properties of the upper 70 cm were thus al tered, as indicated by the separation between regression l ines. This same conclusion was reached a f t e r reviewing the instantaneous profi le calculations and i s consistent with

data from 1 ayered soi 1 profi les (Sisson 1987).

In Figure 4.5, the water content from each depth was adjusted by the amount, (6/Om) x 10 Bk/BO, and replotted as a single curve with the average intercept of the curves shown on Figure 4.4. The Om value i s the maximum water content reached a t each depth. The regression coeff icients , Bk and Bo, are the intercepts and slopes, respectively, of least-squares f i t s of s t ra ight l ines t o data from each depth. Adjusting or scaling the data as shown on Figure 4.5 shows tha t a large portion of the variance observed in measured K(0) values can be removed by adjusting or scaling specif ic water contents by a fixed amount tha t depends on spatial position [see Sisson (1987) fo r fixed gradient model detai ls] . Scaling of the water content data in t h i s way also enables out l ie rs in the data s e t t o be readily ident if ied.


FIGURE 4.4. Depth (z)/Time (t) Versus Water Content from the Unsteady Drainage-Flux Experiment i n the Southeast Caisson. Curves represent Lax so l u t i on (Si sson, Ferguson, van Genuchten 1980) t o Watson (1967) K(6) r e l a t i onsh ip , w i t h i nd i v i dua l curves f i t t e d t o data from each depth.

The i n f i l t r a t i o n r a t e a t t he end o f the 2-h ponding per iod was

0.0063 cm/s. This value was used as an est imate o f KfS. The volumetr ic

water content o f a l l depths was averaged t o obta in an est imate o f OS = 0.262

cm3Icm3. Subs t i t u t i ng these OS and KfS values i n t o t he Watson (1967) equa-

t i o n resu l ted i n the f o l l ow ing K(6) re la t ionsh ip :

Depth (cm)


FIGURE 4.5. Depth (z)/Time ( t ) Versus Scaled Water Content from the Unsteady Drainage-Flux Experiment in the Southeast Caisson. Curve shown'has the same slope and the average intercept of the curves shown in Figure 4.4.

where 8.59 i s the slope of the log (z / t ) versus log 8 regression l ine , plus 1, a f t e r scaling the data. Taking the derivative of the Watson (1967)

equation re su l t s in the following equation:

where 1//3-1 i s t he slope of t he regression l ine . Therefore, 1 must be added t o t he slope before subs t i tu t ing back in to t he or iginal equation f o r 1/P. The so l id l i n e shown on Figure 4.3 shows t he K(0) re la t ionship (Watson 1967) determined from t h i s analysis .

The unsteady drainage-flux method was a l so used t o determine hydraulic conduct ivi t ies in t he north caisson a t the BWTF s i t e . Water content p rof i l es f o r several times during t he north caisson drainage study a re plot ted on Fig- ure 4.6 and l i s t e d in Appendix A, Table A.4. The maximum water content during ponding was approximately 0.30 cm3/cm3 f o r a l l depths, o r 75% sa tur- ation f o r a t o t a l porosity calculated t o be 0.397 assuming bulk and p a r t i c l e dens i t i es of 1.7 and 2.82 g/cm3, respectively. These dens i t i es were determined from laboratory analysis of L-soil. This maximum value of water content i s approximately 25% higher than the maximum value f o r t he southeast caisson data f o r depths below 60 cm. The f a c t t h a t both experiments resul ted in a water content s i gn i f i c an t l y l e s s than t h e t o t a l porosity suggests t ha t entrapped a i r was present. The presence of entrapped a i r would prevent t he attainment of complete sa turat ion (Klute 1986). The di f ference i n maximum water content between the two caissons may r e f l e c t t he f a c t t h a t not a1 1 of the north caisson surface was ponded. The e l e c t r i c pump t h a t supplied water f o r ponding on t he surface of t he caissons did not have a high enough flow

r a t e t o pond water over the e n t i r e surface of t he north caisson. Therefore, water was only ponded on a wedge-shaped section of the north caisson, representing approximately one-third of i t s t o t a l area. By not ponding water over t he e n t i r e caisson surface , a i r could escape more ea s i l y from the north caisson than from the southeast caisson. I f t h i s were t r u e , t h i s mechanism may be par t ly responsible f o r the higher average water content (0.305 cm3/cm3) in t h e north caisson, than in t he southeast caisson (0.262 cm3/cm3) .

Another observation based on Figure 4.6 i s t he rap id i ty w i t h which the p r o f i l e drained. More than half of a l l t he water t h a t eventually drained, drained during t he f i r s t hour. From t h i s observation, we conclude t h a t

i

-

-

-

-

-

- Time (s)

480 0 840 - A 1,380 V 2,040 A 2,880 - V 14,300

89,300 0 188,000 - a 1,150,000

I I I I


FIGURE 4.6. Water Content Profiles Observed During the Unsteady Drainage- Flux Experiment in the North Caisson

during the early drainage phase, many measurements are needed to clearly delineate the shape of the dO/dt curve. Also, the rapid rate of drainage

creates a problem, in that a finite amount of time is needed to obtain a

water content measurement at each depth, and the total time necessary to scan

all depths is significant. To simplify calculations, the recorded time of

measurement was taken as the time at the beginning of the first reading. In

re t rospec t , especia l ly f o r the ear ly drainage times, i t may have been more appropriate t o correct f o r the in te rva l s of time needed t o lower t he neutron probe and t o obtain readings a t each depth.

The matric head data in Appendix A, Table A.5 were used t o ca lcu la te hydraulic head values f o r the BWTF north caisson experiment. These head values were used t o construct t he head prof i l es shown on Figure 4.7.

a Although the head prof i l es indicate uni t gradient conditions throughout most of t he drainage phase of the experiment, there a re times when t he gradient

- - near t he surface i s l e s s than unity. Therefore, hydraulic conductivity ca l -

cula t ions using t he north caisson data were made with t he actual gradient - . measurements ( i .e . , a uni t gradient was not assumed). Hydraul i c conduc-

t i v i t i e s were calculated by the instantaneous p r o f i l e method f o r each measurement time and a re l i s t e d in Appendix A, Table A.6.

Figure 4.8 contains a p lo t of hydraulic conductivity versus water content f o r a l l depths of t he BWTF north caisson, and a plot of t he laboratory data from Figure 4.1. The r e su l t s a r e s imi la r t o t he r e su l t s from the southeast caisson with respect t o t h e i r general re la t ionship t o water content. In f a c t , t he data indicate t h a t the north caisson, l i k e t he south- ea s t caisson, has hydraulic conductivit ies t h a t a r e higher than t he laboratory data a t water contents exceeding 0.12 cm3Icm3. Higher conduct ivi t ies a t lower water content in the caissons suggest t h a t flow through macropores may have had a much grea te r e f f ec t on water content changes in t he caissons than in the laboratory columns a t water contents exceeding 0.12 cm31cm3. I f t h i s were t r u e , i t i s probably t he r e su l t of differences in packing density between t he caissons and t he laboratory columns.

Analysis of t he north caisson data by t he Lax solution method (Sisson, Ferguson, and van Genuchten 1980) and fixed gradient analysis (Sisson 1987)

r. resul ted in t he following K(8) re la t ionship f o r t he Watson (1967) model :

The value of 0.025 cmls represents t he i n f i l t r a t i o n r a t e a t t he end of the 1.5-h ponding period. The water content value of 0.305 cm3Icm3 i s t he average of t he water content values f o r a l l depths a t t he end of t he

Time (s)

0 1,380 A 6,480 7' 20,300

89,300

Total Head (cm)

. FIGURE 4.7. Hydraulic Head Profiles Observed During the Unsteady Drainage-Flux Experiment in the North Caisson

infiltration and start of drainage. The slope of the log (z/t)- versus log 6

regression line was 7.08. Scaling the data had very little effect on

regression parameters because of the uniformity of the profile. The solid

1 ine on Figure 4.8 resulted from substituting values of 8 into Equation (4.3) and plotting the resulting K(6) values.

The Kfs of 0.025 cm/s from the north caisson is four times larger than r)

the KfS of 0.0063 cm/s at the southeast caisson. The potential for three-

dimensional flow resulting from not ponding water over the entire surface of 4

the caisson could explain the higher KfS value obtained in the north caisson.

Once the profile is wetted and ponding has ceased however, flow is essen-

tially vertical in both caissons. The higher Kfs in the north caisson could

also be a result of the higher degree of saturation.

E- m - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - -

Unsteady Drainage-Flux (Field)

Steady-State Flux Control (Laboratory) - Lax Solution for Watson Model

I

Water Content (crn3Icm3)

FIGURE 4.8. Hydraulic Conductivity as a Function of Water Content from the Unsteady Drainage-Fl ux Experiment in the North Caisson, from Repacked Columns in the Laboratory, and from the Lax Solution for the Watson K(8) Relationship

The rate of infiltration of the ponded water after approximately 2 h of ponding may not be truly representative of the field-saturated hydraulic

conductivity. The infiltration rate was not measured as a function of time;

consequently, the actual steady-state infiltration rate normally ascribed to

Kfs may not have been reached. Better estimates of the KfS value used in the

Watson (1967) model could probably be obtained by the Guelph permeameter or

other methods.

The Watson (1967) model curves show higher hydraulic conductivities than are indicated by laboratory data and lower conductivities than are indicated by most of the field data for the southeast and north caissons at water contents between 0.10 and 0.30 cm3Icm3. At lower water contents, the curves show higher conductivities than are indicated by measured data. Overall, 4

this Watson (1967) model K(8) re1 ationship provides a fairly good description of the measured data from the BWTF site.

6

4.1.3 Guelph Permeameter Method

The Guelph permeameter method measures KfS rather than the actual saturated conductivity, KS. Field-saturated hydraulic conductivities are general ly lower than actual saturated conductivities, because the presence of entrapped air reduces the pore space available for flow as previously described. Studies by Stephens et a1 . (1983) and Stephens, Lambert, and Watson (1984) suggest that reasonably accurate estimates of KS can usually be obtained by simply doubling the KfS measurement obtained from the Guelph permeameter method. The arithmetic mean value of KfS for 15 sets of measurements by the Guelph permeameter at the BWTF site is 0.0045 cmls. The arithmetic mean of the four laboratory measurements of Ks (Appendix A, Table A.l) is 0.0084 cmls. Hence, for the BWTF soi 1 , the Stephens et a1 . (1983) and Stephens, Lambert , and Watson (1984) approximations appear to be val id.

Table 4.1 shows the results of the Guelph permeameter analyses from 15 sets of measurements taken around the BWTF site and within the north caisson at the BWTF site. Plotted on Figure 4.9 is the exponential K(h) relationship determined from the average of these 15 measurements. This relationship is

K = 0.0045 exp [O .0573 (h)] (4 4)

where 0.0045 is the field-saturated hydraul ic conductivity (cmls) , 0.0573 is the slope of the lognormal K versus h line, and h is the matric head. Included on Figure 4.9 are the laboratory and field measurements of unsaturated hydraul ic conductivity (see Appendix A) . Examination of Figure 4.9 raises the question of whether or not the exponential K(h) relationship assumed in the Guelph permeameter analysis adequately describes the K(h)

TABLE 4.1. Resul ts from t h e Guelph Permeameter f o r t h e Bur ied Bur ied Waste Test F a c i l i t y

Locat i on

Outside Caissons, 30-cm depth

Average

Wi th in North Caisson

Average

Overa l l Average

A = 30-cm depth. B = 60-cm depth.

re1 a t i onsh ip o f t h i s so i 1. Th is exponential re1 a t i onsh ip matches t h e 1 abora-

t o r y data w i t h i n approximately 1 order o f magnitude over t h e range o f m a t r i c

heads shown. The f i e l d data show more o f a s t r a i g h t - 1 i n e K(h) r e l a t i o n s h i p

than t h e l abo ra to ry data, b u t t h e slope o f t h e l i n e constructed from t h e

Guelph permeameter data does n o t match t h e t rend o f t h e f i e l d data from t h e

unsteady dra inage- f lux method experiment.

4.1.4 P red ic t i ons

F igure 4.10 shows field-measured water r e t e n t i o n data from t h e unsteady

dra inage- f lux experiment i n t h e no r th caisson. The s o l i d l i n e was f i t t o t h e

data by t h e RETC.F77 computer program w i t h t h e Mualem-based (1976) r e s t r i c -

t i o n , m = 1- l /n . Also shown on Figure 4.10 are RETC.F77 curve f i t s t o water

Unsteady Drainage-Flux (Field)

10" Steady-State Flux Control (Laboratory)

- Guelph K (h) Relationship

1 o4 A

cn \

k - .C 10-5 > .- C 0 3 u c 8 .s! - 10" 3 2 B I

1 o-'

- - loS - - - - - - 1 o - ~

0 -50 - 100 - 150 -200 - 250 -300

Matric Head (cm)

FIGURE 4.9. Hydraulic Conductivity as a Function of Matric Head from the Unsteady Drai nage-Fl ux Experiment in the North Caisson, from Repacked Columns in the Laboratory, and the K(h) Relationship Determined from the Average of 15 Sets of Guelph Permeameter Measurements at the Buried Waste Test Facility

retenti on values predicted by the Arya and Paris (1981) model. These water

retention predictions are based on a composite particle-size distribution of

samples BWTF-18A and -18B collected within the north caisson at depths of

30 and 60 cm, respectively. A bulk density of 1.7 g/cm3 and a particle

density of 2.82 g/cm3 were used in the model to calculate a saturated volu-

metric water content of 0.397 cm3/cm3. The dashed line is a curve fitted to

FIGURE 4.10. Water Retention Curves Fit t o Data from the Unsteady Drainage-Flux Experiment in the North Caisson and to Water Retention Characteristics Predicted by the Arya-Pari s (AP) (1981) Model . Predicted values were generated from a composite parti cl e-size distribution of Samples 18A and 18B with the AP model "a" = 1.38 and 1.18

b water retention predictions with the "a" term in the Arya and Paris (1981)

model set a t 1.38. This value was the best-fit value of the "a" parameter

determined by Arya and Paris (1981) for the range of soils in their study.

The dashed-dotted line is a curve fitted to water retention predictions with

the "a" term set at 1.18. This value of "a" was determined by visual fit to the measured data.

Hydraul ic conductivities calculated by the instantaneous profile method for the north caisson data are shown on Figure 4.11. Also shown on Fig- ure 4.11 is the hydraulic conductivity curve based on field-measured water

Water Content ( c d I c m 3 )

FIGURE 4.11. Hydraulic Conductivity as a Function of Water Content from the Unsteady Drainage-Flux Experiment in the North Caisson and Predicted Curves Based on the Arya- Paris (AP) (1981) Model Results Shown in Figure 4.10

retention data with the KS value fixed at 0.0154 cm/s and the Os value internally fitted by the program at 0.309 cm3/cm3. This KS value is two times the arithmetic mean of nine Guelph KfS measurements within the north caisson (see Appendix B, Table B.l). The "Q." parameter used in the Mualem (1976) model was fixed at 0.5, which was the best-fit value of the parameter determined by Mualem (1976) in an analysis of several soils. The restric-

9 tions of m = 1-l/n and Q. = 0.5 were imposed on all of the curves fit to measured data. The fit to the measured hydraulic conductivity data can be

- . improved by a1 lowing the RETC.F77 program to fit values for m and Q. and/or by simultaneously fitting water retention and hydraulic conductivity data. As shown on Figure 4.11, the measured data could apparently be fit better by fixing Ks at a higher value or by allowing the program to fit a Ks value.

The dashed line on Figure 4.11 represents hydraulic conductivities calculated from the water retention values predicted by the Arya and Paris (1981) model, with a = 1.38 and KfS and OS fixed at 0.0154 cmls and 0.397 cm3/cm3, respectively. The calculated Os value of 0.397 was fixed to correspond with the KS value of 0.0154 cm/s in the curve-fitting process. The dashed-dotted 1 ine represents hydraul ic conductivities calculated by the same method with a = 1.18. Predicted and measured conductivities differ from one another by an order of magnitude or less at water contents exceeding 0.10 cm31cm3. At 1 ower water contents, however, differences between measured and predicted values are much greater.

Changing the "a" term in the Arya and Paris (1981) model from 1.38 to 1.18 lowered the predicted matric head values by a factor of 2 to 6 between water contents of 0.40 and 0.025 cm3/cm3. Differences between predicted matric head values at lower and higher water contents were relatively small and almost negligible at saturation and at water contents less than approximately 0.025 cm3/cm3. Changing the "a" parameter had very little effect on

I.

the predicted hydraulic conductivities shown on Figure 4.11. The general , . shapes of the water retention and hydraulic conductivity curves in Figure t 4.11 are very similar.

The hydraulic conductivities based on the Arya and Paris (1981) model water retention predictions agree more closely with the laboratory data than the field data from the north caisson (see Figure 4.8). The calculations of

pore volumes associated with each soil-particle grain-size fraction in the

Arya and Paris model assume that particles in each size fraction are packed in a discrete domain and that, when all domains are considered, the resulting assemblage has a bulk density equal to that measured for a natural-structure

t sample. The model also assumes that the total pore space calculated from the particle and bulk densities is available for filling and is filled at saturation. Therefore, predicted hydraulic conductivities are likely to agree more c

closely with the laboratory data than with the field data. This is a result of the uniform packing of the laboratory columns to the bulk density used for . -

predicting water retention values and the thorough saturation of the laboratory col umns. The RETC. F77 computer program curve-f i tti ng results are shown in Table 4.2. See Section 2.2 for parameter descriptions.

4.2 GRASS SITE

An unsteady drainage-flux experiment was conducted at the Grass site. Guelph permeameter measurements were made around the neutron probe access well grid at the site, and within the unsteady drainage-flux experiment plot.

TABLE 4.2. Curve-Fitting Results from the RETC.F77 Computer Program Based on Data from the Buried Waste Test Facility

Parameters (a) Data Set Or & a - m n - 1 Ks

BWTF-North Caisson Water Retenti on Data 0.09 0.307 0.0931 R 3.6956 0.5" 0.0154*

AP-Predicted Water Retenti on from Samples 18A and 18B (a = 1.38) 0.0095 0.397* 0.0531 R 2.2719 0.5* 0.0154~

AP-Predicted Water Retention from Samples 18A and 18B (a = 1.18) 0.0106 0.397* 0.0972 R 2.5554 0.5* 0.0154*

(a) See Section 2.2 for arameter definitions. AP = Arya and Paris (1981 ! model R = Mualem (1976) based restriction, m = 1-l/n * = Value was fixed

Soi 1 samples, col lected from the auger holes used f o r permeameter measurements, were used f o r par t i c le- s ize analysis .

4.2.1 Unsteady Drainage-Fl ux Method

A t t he Grass s i t e , water content and matric head were measured as functions of depth and time. These data a re in Appendix A, Tables A.7 and A.8, respectively. Figure 4.12 shows the water content p rof i l es as a function of time. The maximum water content reached a t the 15-cm depth was 0.218 cm3Icm3. The maximum water content reached a t the 180-cm depth was 0.142 cm3Icm3. These water content values a re much l e s s than the t o t a l porosity of each so i l layer (approximately 0.5 f o r the upper layer and 0.4 fo r the lower l ayer ) . The tensiometer data l i s t e d in Appendix A, Table A.8 indicate near-saturated flow conditions a t the maximum water content shown. These r e su l t s suggest t h a t entrapped a i r i s preventing complete sa tu ra t ion , a t l e a s t f o r t he upper soi 1 1 ayer. The lower soi 1 l ayer , which i s coarser textured than the upper layer , could not be wetted t o complete sa tu ra t ion , because t he maximum f lux through the upper so i l layer i s not su f f i c i en t t o maintain sa turat ion in the lower layer.

During i n f i 1 t r a t i o n , t he wetting f ront e s sen t i a l l y stops a t the coarse- grained layer un t i l the matric head increases ( t o nearly ze ro) , a t which time the larger pores in the coarser- textured zone begin t o f i l l with water. Lateral flow will occur unt i l t h i s matric potential i s reached. Hence, differences between t he maximum water content reached in t he upper and lower soi 1 layers a t the Grass s i t e can be a t t r ibu ted t o t he e f f ec t of t he soi 1 1 ayeri ng .

According t o Hi 1 l e l (1980) , the advance of a wetting f ron t across a boundary from a fine-grained t o a coarse-grained horizon may not be even and sudden "breakthrough flows" may occur in spec i f ic locat ions , where f inger l ike in t rus ions take place. This unstable flow phenomenon has been t he subject of numerous s tud ies (e.g., Raats 1973; Phil ip 1975; Parlange and Hill 1976; S t a r r , Parlange, and Frink 1986). Preferential flow along the tensiometers i n s t a l l ed a t the Grass s i t e would be somewhat analogous t o the "breakthrough

flows" described by Hi1 l e l (1980). The resul t ing e f f ec t could be sa turated conditions immediately surrounding the tensiometer cups when the rest of the p ro f i l e was ac tua l ly unsaturated. For such condit ions, t he tensiometers

Time (s)

V 16,700 68,900

A 427,000 1,380,000

200 1 I I I I I 0 0.05 0.10 0.15 0.20 0.25


FIGURE 4.12. Water Content Profiles Observed During the Unsteady Drainage-Flux Experiment at the Grass Site

would not accurately measure matric heads in the plot profile, at least

during early drainage measurements.

Figure 4.13 shows the field-measured water retention data for the

Grass site. Because of entrapped or encapsulated air, complete saturation

o f the profile was not attained.


FIGURE 4.13. Water Retention Data from the Unsteady Drainage- Flux Experiment a t the Grass S i t e

Figure 4.14 shows to ta l head plotted against depth fo r various times

p during drainage a t the Grass s i t e . Note that hydraulic head values in the upper 60 cm decreased much more rapidly than in the lower part of the soi 1

b - profi le . This observation suggests that water moved out of the upper so i l , 1 ayer by some process other than drainage (e.g., evaporation, transpira-

t i on, or 1 ateral f 1 ow). Because the plot was covered and the vegetation

surrounding the plot was dormant, la teral flow i s the 1 i kely cause of the

hydraulic head changes. Tensiometers in the upper so i l layer,

Total Head (cm)

FIGURE 4.14. Hydraulic Head Profiles Observed During the Unsteady Drainage-Flux Experiment at the Grass Site

approximately 12 m from the test plot at the Grass site, indicated dry

conditions exceeding the range of tensiometer measurement (c-800 cm) prior

to running the experiment. These adjacent tensiometers were not close

enough to detect lateral flow out of the study plot, but matric head

gradients between the plot and the surrounding dry soil may have been

great enough for water to be drawn laterally out of the upper soil layer.

An important assumption of the unsteady drainage-flux method is that

lateral flow in or out of the test plot profile is negligible. This

assumption is usually justified by ponding water over a large enough area

for a sufficiently long period of time, so that a buffer zone is created

which minimizes the lateral flow component within the test plot during

drainage. This assumption is reasonable for the BWTF drainage

experiments, where caisson walls physical ly res t r ic ted any la te ra l movement of water out of the t e s t plot profi le . However, t h i s assumption does not appear t o be jus t i f ied fo r the drainage experiment a t the Grass

s i t e . Therefore, the instantaneous profi le method was not used t o calcu-

1 a t e hydraul i c conductivities from these data.

The Lax solution method (Sisson, Ferguson, and van Genuchten 1980) i s

based on a unit gradient assumption. From Figure 4.14, i t i s obvious tha t

unit gradient conditions do not ex is t across the en t i r e prof i le , although

they do appear t o ex is t below the 60-cm depth. The fixed gradient analy-

s i s (Sisson 1987) assumes tha t the hydraulic head gradient may vary by

depth, b u t i s invariant with time. However, Figure 4.14 shows tha t the

gradient varies with time above the 60-cm depth. Therefore, neither the

Lax solution nor the fixed gradient analysis was used with the Grass S i t e

data. Efforts are in progress t o repeat t h i s experiment, with modif-

ications t o eliminate la teral movement within the upper 60 cm of the

profi le .


Results fo r the Guelph permeameter analyses from measurements a t the

Grass s i t e are shown in Table 4.3. The mean K f s value of 0.0092 cm/s for

TABLE 4.3. Results from the Guelph Permeameter fo r the Grass S i t e

Locat i on

20-cm depth

Average

60-cm depth

Average

the lower soil layer is approximately 9 times larger than the mean value of 0.001 cm/s for the upper layer. This difference supports the conten- tion that the upper soil layer restricts water infiltration to the lower layer by limiting the flux, such that the lower layer cannot be completely saturated during an infiltration experiment.

4.2.3 Predictions

Particle-size distribution data from the Grass site (see Appendix B, Table B.2) will be used in the Arya and Paris (1981) and Mualem (1976) models to predict hydraulic conductivities after repeating the unsteady drainage-flux experiment at that site. This second set of data will then be available to further assess the predictive capabilities of these models in layered soil profiles.

4.3 McGEE RANCH

An unsteady drainage-flux experiment was conducted at the McGee Ranch. Guelph permeameter measurements were taken around the borrow pit at the site, and within the unsteady drainage-flux test plot. Soil samples, collected from the auger holes used for Guelph permeameter measurements, were used for particle-size analysis and WRC prediction by the Arya and Paris (1981) model. These predictions were then used in the Mual em (1976) model to predict hydraul ic conductivities.

4.3.1 Unsteady Drainage-Flux Method

The water content data for the unsteady drainage-flux experiment at the McGee Ranch are listed in Appendix A, Table A.9. The water content profiles on Figure 4.15 show that water content decreased uniformly with depth during drainage. Data from the 120-cm depth were not analyzed, because steady-state flow had not been reached at that depth, and time constraints and water availability limited additional infiltration. After approximately 15 days of drainage, matric head values had reached -323 to -340 cm for a1 1 depths under consideration (see Appendix A, Table A.lO). Figure 4.16 shows total head versus depth for various times during drainage at this site. The mean head gradient is equal to 0.83. If lateral flow were appreciable at the McGee Ranch site, it would not be as

Time (s)

0

1,500 4,200 9,900

59,700

120 1 I I I I I 0 0.1 0.2 0.3 0.4 0.5

Water Content (crn3/crn3)

FIGURE 4.15. Water Content Profiles Observed During the Unsteady Drainage-Flux Experiment at the McGee Ranch

apparent in the total head data of Figure 4.16 as it was on Figure 4.14, because of the relative uniformity of the soil profile.

Field-measured water retention data for the McGee Ranch site are listed in Appendix A, Tables A.9 and A.lO. These data are plotted on Fig- ure 4.17. Hydraulic conductivities were calculated by a time-averaging approach (Rose, Stern, and Drummond 1965), using actual head gradients rather than an assumed unit gradient. These data are listed in Appen- dix A , Table A.11, and are plotted on Figure 4.18. The close grouping of the data on Figure 4.18 indicate that the upper 1 m of soil at this site is re1 atively uniform with respect to hydraul ic conductivity.

-- 0 Time (s) 0

Total Head (cm)

FIGURE 4.16. Hydraul ic Head P r o f i l e s Observed During the Unsteady Drainage-Flux Exper iment 'at the McGee Ranch

Figure 4.16 ind ica tes t h a t u n i t g rad ient cond i t ions d i d no t e x i s t a t

the McGee Ranch s i t e dur ing the unsteady drainage- f lux experiment. Based

on F igure 4.16 i t appears as though the gradients are r e l a t i v e l y constant

i n t ime. Therefore, t he Lax so l u t i on (Sisson, Ferguson, and van Genuchten

1980) and f i x e d grad ient ana lys is were used t o determine t he parameters i n

t he Watson (1967) model f o r the McGee Ranch data. Scal ing o f t he water

content data had very l i t t l e e f f e c t on t he regression parameters, because

o f t he un i f o rm i t y o f the p r o f i l e . The r e s u l t i n g Watson (1967) model

re1 a t ionsh ip determined from t h i s ana lys is i s

0 .- L C

9 t Depth (cm)

Water Content ( cm3/ cm3)

FIGURE 4.17. Water Retention Data from the Unsteady Drainage- Flux Experiment a t the McGee Ranch

where 0.0017 cmls i s the rate of fa l l of the level of ponded water on the surface of the t e s t plot. The average water content for a l l depths a t time zero was 0.399 cm31cm3, and the slope of the log (z l t ) versus log 8

l ine i s 7.53. This K(B) relationship i s shown as the solid line in Figure 4.18.


Results of analysis of the Guelph permeameter data from around the borrow p i t and within the unsteady drainage-flux plot a t the McGee Ranch

are shown in Table 4.4. Figure 4.19 shows the field-measured K(h) data from the McGee Ranch unsteady drainage-flux experiment. The solid line on Figure 4.19 represents the average K ( h ) re1 ationship determined from

Water Content (crn3/cm3)

FIGURE 4.18. Hydraulic Conductivity as a Function of Water Content from the Unsteady Drai nage-Fl ux Experiment at the McGee Ranch

analysis of nine Guelph permeameter measurements shown in Table 4.4. The arithmetic mean value of Kfs, based on analysis of the 9 Guelph

permeameter measurements, i s 0.0009 cmls. The .KfS val ues calculated from Guelph permeameter data for samples 9 A and 9B, which were measured within the. unsteady drainage-flux study plot, are 0.0005 and 0.0007 cmls, respec-

tively. The K(h) relationships determined from these data are also

TABLE 4.4. Results from the Guelph Permeameter for the McGee Ranch

Location Kfq, cmls #m, cm21s a

20-cm depth

Average 0.0006 0.0058 0.1034

60-cm depth

Average 0.0013 0.0140 0.0929

Overall Average 0.0009 0.0095 0.0947

plotted on Figure 4.19 as the dashed and dashed-dotted 1 ines, representing samples 9A and 9B, respectively. The slope of the line determined from analysis of Guelph permeameter results of sample 9B comes closest to matching the trend of the measured hydraul ic conductivity values. There is less than an order of magnitude difference between values that fall on this 1 ine and the measured values at matric heads of about -60 cm. This difference increases to almost 3 orders of magnitude at a matric head of -300 cm. This relationship suggests that the calculation of the slope of

the line comparing lognormal K to h that was used in the assumed exponential K(h) relationship, may not be appropriate for the soils in this study [see Equation (2.15)]. However, this failure of the K(h) line based on analysis of Guelph permeameter data, to fit the measured K(h) data may be a result of natural soil heterogeneity within the plot and across the McGee Ranch site.

The Guel ph permeameter measures hydraul ic conductivity in the vicin- ity of the auger hole. The neutron probe and tensiometers measure water contents and matric heads over a larger volume of soil such that the measurements and subsequent hydraulic conductivity calculations represent

Matric Head (cm)

FIGURE 4.19. Hydrau l ic Conduc t i v i t y as a Funct ion o f M a t r i c Head f rom t h e Unstead Drainage-Flux Experiment a t t h e McGee Ranch and t h e K(h 3 Rela t ionsh ips Determined from t h e Average o f 9 Sets o f Guelph Permeameter Measurements a t t h e McGee Ranch. Guelph 9 A and 9B represent 2 se ts o f measurements, from t h e 20-cm (9A), and 60-cm (9B) depths, w i t h i n t h e unsteady dra inage- f 1 ux experiment f i e l d p l o t

L

more o f t h e n a t u r a l heterogenei ty . Consequently, d i f f e rences i n h y d r a u l i c

c o n d u c t i v i t y obta ined by t h e Guelph permeameter and unsteady d ra inage- f lux

methods l i k e l y r e s u l t f rom s p a t i a l v a r i a b i l i t y and sca le d i f f e r e n c e s

between the two methods. Differences may also result from different approximations of differential and integral quantities in the two methods.

4.3.3 Predictions

Figure 4.20 shows field-measured water retention data from the McGee Ranch. The solid line represents a least-squares fit to the data using the RETC.F77 computer program. The dashed 1 ine curve is fit 'to predicted water retention values based on the particle-size distribution composited from samples MCG-9A and -9B (listed in Appendix B, Table B.3). These samples were collected from the 20- and 60-cm depths within the unsteady drainage-flux study plot at the McGee Ranch. The measured bulk density of 1.54 g/cm3 and particle density of 2.77 g/cm3 were used with the "a" parameter in the Arya and Paris (1981) model set at 1.38 to calculate predicted water retention values. The dashed-dotted line on Figure 4.20 is fit to predicted water retention values based on the same particle-size distribution, with the same bulk and particle densities, but with a =

1.10. This value was determined by visual fit of a curve to the measured data. With a = 1.10, the predicted water retention values agree with the measured data within a factor of 3 of matric head values, between water contents of approximately 0.10 to 0.40 cm3/cm3.

Figure 4.21 shows field-measured hydraul ic conductivity data from the unsteady drainage-flux experiment at the McGee Ranch. The solid curve was fit to f ield-measured water retention data using the RETC. F77 computer program with Mual em's (1976) predictive conductivity model. The Ks value was fixed at 0.0012 cm/s, which is 2 times the average KfS value for samples MCG-9A and -9B, as determined by analysis of the Guelph permeameter data. The curve does not fit the measured data very well. As mentioned previously, a much closer fit to the measured data can be obtained with RETC.F77 if m and Q. are fitted independently and/or if a simultaneous fit to retention and hydraul ic conductivity data is made. The fit could probably also be improved if more data for the drier portion of the range of soil moisture conditions were available. The RETC.F77

program fit the Or value at 0.0 cm3/cm3. This value would be difficult,

if not impossible, to reach under field conditions.

FIGURE 4.20. Water Retention Curves F i t t o Data from the Unsteady Drainage-Flux Experiment a t the McGee Ranch and t o Water Retention Character is t ics Predicted by the Arya-Paris (AP) (1981) Model. Predicted values were generated from a composite pa r t i c l e - s i ze d i s t r i b u t i o n o.f Samples 9A and 9B w i t h the AP model "a" = 1.38 and 1.10

The dashed l i n e shown on Figure 4.21 corresponds t o the dashed l i n e

on Figure 4.20, which i s based on Arya and Par is (1981) model p red ic t ions

w i t h a = 1.38. The dashed-dotted l i n e on Figure 4.21 corresponds t o the

dashed-dotted 1 i n e on Figure 4.20 w i t h a = 1.10. Both o f these 1 ines

match measured hydrau l ic conduc t i v i t y data w i t h i n a f a c t o r o f f i v e up t o a

Water Content (cm31cm3)

FIGURE 4.21. Hydraul ic Conduct iv i ty as a Function o f Water Content from the Unsteady Drainage-Flux Experiment a t the McGee Ranch and Predicted Curves Based on t he Arya- Par is (AP) (1981) Model Results Shown i n Figure 4.20

P

water content o f approximately 0.40 cm3/cm3. The value o f a = 1.10 gives

a much b e t t e r f i t t o measured water r e t en t i on and hydrau l i c conduc t i v i t y # data than a = 1.38. These curves were generated by f i x i n g t h e KS value a t

0.0012 cm/s as was done w i t h the curve f i t t o measured data. The Os value

f o r these curves was f i xed a t a water content value o f 0.444 cm3/cm3

(determined from the p a r t i c l e and bu lk dens i t i es ) , bu t the BS value f o r

t he curve f i t t o measured data was f i t t e d by t he program a t a water

content value of 0.392 cm3/cm3. Therefore, differences between measured

and predicted conduct ivi t ies a t water contents above 0.392 cm3/cm3 do not

necessari ly r e f l e c t on t he predic t ive a b i l i t y of t he Arya and Par is (1981)

model. Table 4.5 shows t he RETC.F77 curve- f i t t ing r e s u l t s f o r McGee Ranch

data.

In t h e sense t h a t t he "a" parameter i s empirically determined, t h e

Arya and Par is (1981) model i s not t r u l y predic t ive . However, Arya and

Par is used an i t e r a t i v e procedure t o determine a b e s t - f i t "a" (1.38) t h a t

minimized t he sum of t h e absolute value of t he log of the measured matric

head values minus t he log of t he calculated matric head values f o r a range

of s o i l s . They then used t h i s value f o r predicting water re tent ion values

TABLE 4.5. Curve-Fitting Results from the RETC.F77 Computer Program Based on Data from the McGee Ranch

Parameters (a) Data Set 8rL a - m n 1 Ks

McGee Ranch Water Retention Data 0.000 0.409 0.0058 R 2.3563 0.5* 0.0012*

Simultaneous F i t t o McGee Ranch Water Retention and Conductivity Data 0.019 0.409 0.0059 R 2.4299 1.897 0.0006

AP-Predicted Water Retention Based on Samples MCG-9A and -9B (a = 1.38) 0.000 0.444" 0.0024 R 1.5420 0.5* 0.0012*

AP-Predicted Water Retention Based on Samples MCG-9A and -9B (a = 1.65) 0.000 0.444" 0.0110 R 1.7619 0.5* 0.0012*

(a) See Section 2.2 f o r arameter def in i t ions . AP = Arya and Par is (1981 ! model

R = Mualem (1976) based r e s t r i c t i o n , m = 1 - l / n * = Value was f ixed

for other soils in their study. They did not use water retention predictions for estimating hydraulic conductivities, however. For the soils studied here, the best-fi t "a" value differs from the value of 1.38 determined by Arya and Paris. This observation raises the question of whether a

+ single value of "a" would be appropriate for predicting water retention characteristics and subsequently predicting hydraul ic conductivity for a1 1

t Hanford Site soils. The question can be answered only by analyzing additional Hanford Site soils.

. Differences between measured and predicted water retention values could be real, thereby suggesting 1 imitations in the Arya and Paris (1981) model,

- . or they could result from errors in the particle-size analysis or bulk density measurements. According to Arya and Paris, uncertainties of t5% in the particle-size analysis and tO.l g/cm3 in the bulk density are not uncommon (e.g., Coel ho 1974; Keisl ing 1974; Alexander 1980). A1 so, an iterative procedure, such as that used by Arya and Paris, could be used to calibrate the model to optimize the fit of predicted values to measured data for the soils in this study. This should help reduce the differences between measured and predicted hydraulic conductivities. Other possible explanations for the variations between measured and predicted conductivities are differences in the field- and laboratory-tested soil materials, within- plot variability, and the initial parameter estimates used in the curve- fitting process.

5.0 CONCLUSIONS AND RECOMMENDATIONS

The most important conclusion, based on the results of this study, is that no single method or measurement technique should be used for generating unsaturated hydraulic conductivity data for the Hanford Site. Each method used in this study produced results sufficiently different from the other methods, that to rely solely on one method would be unwise. The most appropriate method ultimately depends on the specific job or application. Ideally, more than one method should be used to take advantage of the strengths of each method, considering the data needs and resources available.

The laboratory steady-state flux control method provided accurate hydraulic conductivity measurements for repacked columns of L-soil from the Buried Waste Test Faci 1 i ty. These measurements agreed with field measurements within one order of magnitude. Using repacked columns may not yield results that are truly representative of natural conditions at other sites because of the disturbed nature of the samples. Therefore, using this method with undisturbed core samples would be preferable, and tests should be initiated using this method with undisturbed samples from the other field sites. This method is time consuming. It has an advantage, however, over other methods in that samples can be completely saturated so that true desorption curves, rather than intermediate scanning curves, can be measured.

The unsteady drainage-fl ux method provided re1 ati vely accurate hydraulic conductivity measurements at two of the three field sites. At the third site (Grass site) , a textural transition (i .e., 1 ayering) resulted in lateral flow, so that the one-dimensional (vertical) flow assumption used to calculate hydraulic conductivity was not valid. This experiment is being repeated with modifications to ensure one-dimensional flow.

A power function relationship, using parameters estimated by the Lax (1972) solution (Sisson, Ferguson, and van Genuchten 1980), provided reasonable descriptions of the measured hydraul ic conductivity data from the BWTF and McGee Ranch sites. Scaling of water content data with a fixed gradient model (Sisson 1987) appears to be useful as a data reduction technique and for describing some layered soil profiles. The RETC.F77 computer program

(van Genuchten 1985) provides accurate descriptions of measured data, especially when no restrictions are imposed on the curve-fitting parameters.

The Guelph permeameter method provides rapid, re1 atively accurate, field-saturated hydraulic conductivity data. Because of the portability of the apparatus, 1 ow water requirements, and speed with which measurements can be made, this method should be useful for spatial variability studies. The

adequacy of the method for describing the K(h) relationships of soils tested in this study, however, remains questionable. The failure of the K(h) rela-

tionship determined from Guelph permeameter analyses to agree with other measured data may be a result of natural soil heterogeneity and scale differences between methods.

The predictions of hydraulic conductivity based on particle-size distribution and bulk density data were within one-half to one and one-half orders of magnitude of measured values, depending on soil type. This agreement may or may not be considered adequate, depending on the nature of the information needs, but the technique could be useful as a first approximation of hydraulic conductivity and would allow use of the Westinghouse Hanford Company grain-size data base.

The differences in hydraulic conductivities measured by the various techniques in this study illustrate several unresolved problems. One of these is how to reconcile laboratory and field data that have different KS and OS values; this is often attempted by scaling data or by using matching factors. With hysteresis effects resulting from incomplete saturation because of entrapped ai r, f i el d-measured water retention curves wi 11 have different shapes than those measured in the laboratory regardless of matching factors. Consequently, it is not realistic to expect complete agreement between 1 aboratory and f i el d data.

Field data are generally considered to be more representative of natural

conditions and, thus, are preferable to laboratory data. On a large scale, it becomes impractical to try to characterize the variability of soil hydraulic properties with the detailed analyses used in this study. There- fore, geostatistical approaches should be evaluated as a means of using a small set of data to characterize large areas.

6.0 REFERENCES

Alexander, E. B. 1980. "Bulk Dens i t ies o f C a l i f o r n i a S o i l s i n Re la t i on t o Other S o i l Proper t ies." S o i l Sci . Soc. Am. J. 44:689-692.

Arya, L. M., and J. F. Par is . 1981. "A Physicoempir ica l Model t o P r e d i c t t h e Soi 1 Moisture C h a r a c t e r i s t i c from P a r t i c l e - S i z e D i s t r i b u t i o n and Bulk Densi ty Data." S o i l Sc i . Soc. Am. J. 45:1023-1030.

Black, T. A., W. R. Gardner, and G. W. T h u r t e l l . 1969. "The P r e d i c t i o n o f ~ v a p o r a t i o k Drainage, and Soi 1 Water Storage f o r a Bare Soi 1 .I1 Soi 1 Sci . Soc. Am. Proc. 33:655-660.

Cass, A., G. S. Campbell, and T. L. Jones. 1981. Hydrau l ic and Thermal P rope r t i es o f S o i l Samples from t h e Bur ied Waste Test F a c i l i t y . PNL-4015, P a c i f i c Northwest Laboratory, Richland, Washington.

Coelho, M. A. 1974. Spa t i a l V a r i a b i l i t y o f Water Related S o i l Physical Proper t ies . Ph.D. Thesis, U n i v e r s i t y o f Arizona. U n i v e r s i t y M ic ro f i lms , Ann Arbor, Michigan (Diss. Abstr. 75-11061).

Davidson. J. M.. L. R. Stone. D. R. N ie lsen. and M. E. LaRue. 1969. " F i e l d ~easurement and Use o f ~ o i i -water p roper t ies . " Water Resour. Res. 5: 1312-1321.

Fayer, M. J., G. W. Gee, and T. L. Jones. 1986. UNSAT-H Version 1.0: Unsaturated Flow Code Documentation and App l i ca t i ons f o r t h e Hanford S i te . PNL-5899, P a c i f i c Northwest Laboratory, Richland, Washington.

Gardner, W . R. 1958. "Some Steady-State So lu t ions o f t h e Unsaturated Mois ture Flow Equation w i t h App l i ca t i on t o Evaporat ion from a Water Table." Soi 1 Sc i . 85 (4) : 228-232.

Gee, G. W., and J. W. Bauder. 1986. " P a r t i c l e - S i z e Analys is ." I n Methods of S o i l Analys is , Pa r t I. 2nd ed. E d i t o r A. Klu te , pp. 383-411. American Soc ie ty o f Agronomy, Madison, Wisconsin.

Gee, G. W., and R. R. Kirkham. 1984. A r i d S i t e Water Balance: Evapotrans- p i r a t i o n Modeling and Measurements. PNL-5177, P a c i f i c Northwest Laboratory, Richland, Washington.

Green, R. E., L. R. Ahuja, and S. K. Chong. 1986. "Hydrau l ic Conduc t i v i t y , D i f f u s i v i t y , and S o r p t i v i t y o f Unsaturated Soi 1 s: F i e l d Methods. " I n Methods o f S o i l Analys is , Pa r t I. 2nd ed. E d i t o r A. K lu te , pp. 771-798. American Soc ie ty o f Agronomy, Madison, Wisconsin.

Gupta, S. C:, and W. E. Larson. 1979. "Es t imat ing Soi l-Water Retent ion Character1 s t i c s f rom P a r t i c l e - S i z e D i s t r i b u t i o n , Organic Ma t te r Percent, and Bulk Density." Water Resour. Res. 15:1633-1635.

Hall, D. G. M., M. J. Reeve, A. J. Thomasson, and V. F. Wright. 1977. Water Retention, Porosity, and Density of Field Soils. Tech. Monograph No. 9, Rothamsted Experiment Station, Harpenden, England.

Hillel, D . 1980. Fundamentals of Soil Physics. Academic Press, New York.

Keisling, T. C. 1974. Precision with Which Selected Physical Properties of Similar Soils can be Estimated. Ph.D. Thesis, Oklahoma State University., University Microfilms, Ann Arbor, Michigan. (Diss. Abstr. 75-8812).

Kirkham, R. R., and G. W. Gee. 1987. Field Lysimeter Test Facility for Protective Barriers: Experimental Plan. PNL-6351, Pacific Northwest Laboratory, Richland, Washington.

Klute, A. 1986. "Water Retention: Laboratory Methods." In Methods of Soil Analysis, Part I. 2nd ed. Editor A. Klute, pp. 635-662. American Society of Agronomy, Madison, Wisconsin.

Klute, A., and C. Dirksen. 1986. "Hydraulic Conductivity and Diffusivity: Laboratory Methods." In Methods of Soil Analysis, Part I. 2nd ed. Editor A. Klute, pp. 687-734. American Society of Agronomy, Madison, Wisconsin.

Last, G. V., M. A. Glennon, M. A. Young, and G. W. Gee, 1987. Protective Barrier Materials Analysis: Fine Soil Site Characterization. PNL-6314, Pacific Northwest Laboratory, Richland, Washington

Lax, P. D. 1972. "The Formation and Decay of Shock Waves." Am. Math Monthly 79:227-241.

Mualem, Y. 1976. "A New Model for Predicting the Hydraulic Conductivity of Unsaturated Porous Media. " Water Resour. Res. 12(3) :513-522.

Mualem, Y. 1986. "Hydraulic Conductivity of Unsaturated Soils: Prediction and ~ormulas." In Methods of Soil ~nalysis, Part I. 2nd ed. Editor A. Klute, pp. 799-823. American Society o f Agronomy, Madison, Wisconsin.

Nielsen, D. R., J. W. Biggar, and K. T. Erh. 1973. "Spatial Variability of Field-Measured Soil-Water Properties." Hilgardia 42:215-259.

Nielsen, D. R., J. M. Davidson, J. W. Biggar, and R. J. Mi 1 ler. 1964. "Water Movement Through Panoche Clay Loam Soil." Hilqardia 35:491-506.

Parlange, J. -Y., and D. E. Hill. 1976. "Theoretical Analysis of Wetting Front Instabi 1 ity in Soils." Soi 1 Sci . 122:236-239.

Philip, J. R. 1975. "Stability Analysis of Infiltration." Soil Sci. Soc. Amer. Proc. 39:1042-1049.

P h i l l i p s , S. J., A. C. Campbell, M. D. Campbell, G. W. Gee, H. H. Hoober, and K. 0. Schwartzmil ler. 1979. A F i e l d Test F a c i l i t y f o r Moni to r ing Water/ Radionucl i d e Transport Through P a r t i a l l y Saturated Geologic Media: Desiqn, Construct ion, and Pre l i m i nary Descr ipt ion. PNL-3226, P a c i f i c Northwest Laboratory, Rich1 and, Washington.

Raats, P. A. C. 1973. "Unstable Wett ing Fronts i n Uniform and Nonuniform So i l s . " S o i l Sci. Soc. Amer. Proc. 37:681-685.

Reynolds, W. D., and D. E. El r i c k . 1985. " I n - S i t u Measurement o f F i e l d - Saturated Hydraul ic Conduct iv i ty , S o r p t i v i t y and t h e a-Parameter Using t h e Guel ph Permeameter. " Soi 1 Sci . 140:292-302.

Reynolds, W. D., D. E. E l r i c k , and B. E. C lo th ie r . 1985. "The Constant Head We1 1 Permeameter: E f f e c t o f Unsaturated Flow. " Soi 1 Sci . 139(2) : 172-180.

Richards, L. A. 1931. " C a p i l l a r y Conduction o f L iqu ids i n Porous Mediums." Physics 1:318-333.

Richards, L. A., W. R. Gardner, and G. Ogata. 1956. "Physical Processes Determining Water Loss from So i l . " S o i l Sci. Soc. Am. Proc. 20:310-314.

Rose, C. W., W. R. Stern, and J. E. Drummond. 1965. "Determinat ion o f Hydrau l ic Conduct iv i ty as a Funct ion o f Depth and Water Content f o r S o i l i n Si tu." Aust. J. S o i l Res. 3:l-9.

Scot te r , D. R., B. E. C l o t h i e r , and E. R. Harper. 1982. "Measuring Saturated Hydrau l ic Conduct iv i ty and S o r p t i v i t y Using Twin Rings." Aust. J. S o i l Res. 20(4) :295-304.

Sisson, J. B. 1987. "Drainage from Layered F i e l d So i l s : Fixed Gradient Models. " Water Resour. Res. 23 (11) :2071-2075.

Sisson, J. B., A. H. Ferguson, and M. Method f o r P red ic t i ng Drainage from 44:1147-1152.

Th. van Genuchten. 1980. "Simple F i e l d Plots." S o i l Sci. Soc. Am. J.

S ta r r , J. L., J. -Y. Parlange, and C. R. F r ink . 1986. "Water and Ch lor ide Movement Through a F i e l d So i l . " S o i l Sci. Soc. Am. J. 50:1384-1390.

Stephens, D. B., S. Ty ler , K. Lambert, and S. Yates. 1983. " F i e l d Exper i- ments t o Determine Saturated Hydrau l ic Conduct iv i ty i n t h e Vadose Zone." I n Role o f t h e Unsaturated Zone i n Radioact ive and Hazardous Waste Disposal. E d i t o r J. W. Mercer. Ann Arbor Science, Ann Arbor, Michigan, pp. 113-126.

Stephens, D. B., K. Lambert, and D. Watson. 1984. " In f l uence o f Entrapped A i r on F i e l d Determinat ions o f Hvdraul ic Proper t ies i n t h e Vadose Zone." Proc. Conf. Charac ter iza t ion a n d - ~ o n i t o r i n g ;n t h e Vadose Zone. Nat ional Water We1 1 Associat ion, Worthington, pp. 57-76.

U.S. Department o f Energy. 1987. F i n a l Environmental Impact Statement, Disposal o f Hanford Defense High-Level, Transuranic and Tank Wastes, Hanford S i t e , Richland, Washington. DOEIEIS-0113F (Vol. 3 ) , U.S. Depart- ment o f Energy, Washington, D.C.

van Genuchten, M. Th. 1978. Ca lcu la t i ng t h e Unsaturated Hydrau l ic Conduc- 9 t i v i t y w i t h a New Closed-Form A n a l y t i c a l Model. Report 78-WR-08, Depart- ment o f C i v i l Engineering, Princeton, New Jersey, pp. 63.

van Genuchten, M. Th. 1985. Proqram t o Analyze Observed S o i l Water Tension II,

and Hydraul i c Conduc t i v i t y Data. U.S. Sal i n i t y Laboratory Special Report, Rivers ide, C a l i f o r n i a . 1

Watson, K. K. 1966. "An Instantaneous P r o f i l e Method f o r Determining t h e Hydrau l ic Conduc t i v i t y o f Unsaturated Porous Mater ia ls . " Water Resour. . - Res. 1:577-586. -

Watson, K. K. 1967. "The Measurement o f t h e Hydrau l ic Conduc t i v i t y o f Unsaturated Porous Mate r ia l s U t i l i z i n g a Zone o f Entrapped A i r . " S o i l Sci. Soc. Am. Proc. 32:716-720.

Wierenga, P. J., L. W. Gelhar, C. S. Simmons, G. W. Gee, and T. J. Nicholson. 1986. V a l i d a t i o n o f Stochast ic Flow and Transport Models f o r Unsaturated Soi 1 s: A Comprehensive Fie1 d Study. NUREGICR-4622, U.S. Nuclear Regulatory Commission, Washington D.C.

APPENDIX A

WATER RETENTION AND HYDRAULIC CONDUCTIVITY DATA

APPENDIX A

WATER RETENTION AND HYDRAULIC CONDUCTIVITY DATA

TABLE A.1. Steady-State Flux Control Method Results for L-Soil

Col umn El pb = 1.6 g/cm3 8, cm3/cm3 h , cm K, cm/s

0.435 M 0 7.62E-3

0.308 18 1.74E-3

0.233 2 1 7.89E-4

0.173 2 9 1.01E-4

0.129 4 7 1.32E-5

0.100 83 1.09E-6

0.086 140 7.70E-8

Column E, pb = 1.7 g/cm3 8, cm3/cm3 h, cm K, cm/s 0.400 w 0 7.12E-3

0.307 20 2.57E-3

0.227 23 5.15E-4

0.189 3 5 1.72E-4

0.145 52 2.15E-5

0.125 7 2 5.32E-6

0.100 130 3.36E-7 0.091 175 1.08E-7

0.083 215 5.40E-8

8 = volumetric water content h = matric head K = hydraulic conductivity

pb = bulk density.

Column F, pb = 1.6 g/crn3

0, cm3/cm3 h, cm K, cm/s

0.422 w 0 9.78E-3

0.310 18 3.16E-3

0.250 22 9.90E-4

0.177 2 9 1.01E-4

0.138 4 7 1.36E-5

0.110 8 4 1.59E-6

Column F, pb = 1.7 g/cm3

0, cm3/cm3 h, cm K, cm/s

0.386 w 0 8.91E-3

0.297 20 2.52E-3

0.229 23 5.07E-4

0.186 3 5 1.68E-4

0.155 42 4.23E-5

0.124 76 4.98E-6

0.100 125 3.47E-7 0.092 170 1.12E-7

0.086 200 5.58E-8

TABLE A.2. Water Content Data from BWTF Southeast Caisson

Time, s

8. 88E+88 4.84 E+82 1.88E+83 1.68E.83

2.28E+83 2.88E+83 3.48E+83 4.38E+83

7.88E.83 8.88E.83 1.43E+84 1.79E+04 2.22E+84 2.65E.84 8.16E.84 1.07E+85 1.96E.85

3.67E+85 7.11E+85

1.85E+86

Water Content, cm3/ca3, a t Depth, cm

98 185 128 135 158 - - - - - 8.248 8.256 8.248 8.243 8.246

8.238 8.247 8.242 8.238 0.243

8.225 8.231 8.232 8.231 8.229

8.212 8.219 8.218 8.218 8.224

8.281 8.215 8.289 8.286 8.216 8.189 8.281 8.281 8.283 8.284 0.187 8.193 8.192 8.194 8.197 8.182 8.186 8.188 8.185 8.191

8.167 8.172 8.176 8.176 8.175 8.163 8.168 8.167 8.165 8.168

8.158 8.153 8.163 8.154 8.166 8.147 8.151 8.147 8.151 8.151 8.142 8.141 8.142 8.143 8.146 8.148 8.148 8.143 8.139 8.143

8.121 8.123 8.128 8.122 8.121 8.118 8.118 8.119 8.118 8.128 8.118 8.115 8.111 8.112 8.113 8.187 8.111 8.189 8.111 8.112 8.186 8.187 8.188 8.186 8.189

8.182 8.182 8.184 8.184 8.183

TABLE A.3. Hyd rau l i c Conduc t i v i t y Data from t h e BWTF Southeast Caisson

Hydraulic Conduct iv i ty, cm/s, a t Depth, cm

Time, s

4.84E+82 1.88E+83 1.68E.83 2.28E.83 2.88E+83 3.48E.83 4.38E+83 7.08E.83 8.88E+83

1.43E.84

1.79E+84 2.22E+84

2.65E+84

8.16E+84

1.87E+85 1.96E+85 3.67E.85

7. l lE+85

1.85E+86

TABLE A.4. Water Content Data from t h e BWTF North Caisson

Time, s

0.00E+00 4.80E+02 8.40E+02 1 .38 E+03 2.04E+03 2.88E+03 4.68E+03 6.48E+03 8.28E+03 1.43E+04 2.03 E+04 2.75E+04 8.93 E+04 1.88E+05 2.75E+05 6.23 E+05 1.15E+06

Water Content, cm3/cm3, a t Depth, cm

45 60 90 120 150 180 210 ------- 0.312 0.323 0.297 0.302 0.307 0.300 0.295 0.216 0.223 0.298 0.302 0.303 0.298 0.296 0.195 0.196 0.222 0.252 0.293 0.297 0.295 0.177 0.180 0.199 0.202 0.218 0.251 0.290 0.164 0.167 0.188 0.189 0.195 0.209 0.222 0.155 0.161 0.179 0.179 0.182 0.191 0.201 0.142 0.147 0.163 0.168 0.172 0.179 0.187 0.138 0.137 0.156 0.159 0.162 0.169 0.176 0.132 0.136 0.149 0.152 0.152 0.164 0.170 0.124 0.127 0.138 0.141 0.147 0.151 0.159 0.120 0.123 0.133 0.137 0.138 0.148 0.153 0.110 0.119 0.128 0.129 0.131 0.139 0.146 0.105 0.106 0.116 0.116 0.117 0.124 0.126 0.101 0.102 0.111 0.110 0.113 0.117 0.122 0.100 0.101 0.110 0.113 0.106 0.113 0.118 0.096 0.099 0.107 0.105 0.107 0.109 0.113 0.096 0.093 0.104 0.106 0.105 0.106 0.111

TABLE A.5. M a t r i c Head Data f o r t h e BWTF North Caisson

M a t r i c Head, cm, a t Depth, cm

Time, s - 15 - 30 4 5 - 6 0 - 9 0 - - 120 150 180

TABLE A.6 . Hydrau l ic Conduct iv i ty Data from t h e BWTF North Caisson

Hydraul i c Conductivity, cm/s, a t Depth, cm

Time, s 30 45 60 90 120 150 180 210 240 ---------

TABLE A.7. Water Content Data from t h e Grass S i t e


TABLE A.8. M a t r i c Head Data from t h e Grass S i t e

Time, s

0.00E+00 4.84E+02 1 .08E+03 1.68E+03 2.88E+03 4.08E+03 5.28 E+03 7.08E+03 8.88E+03 1.31E+04 1.67E+04 1.97E+04 6.89E+04 9.93 E+04 1.87E+05 4.27E+05 6.18E+05 7.67 E+05 1.03 E+06 1.38E+06 1.98E+06

M a t r i c Head, cm, a t Depth, cm

15 - 3 0 - 4 5 - 60 - 9 0 - 120 - 150 - 180 -

- 18 - 1 - 1 - 1 - 1 - 1 - 1 - 1 -20 -6 - 3 - 2 - 1 - 1 - 1 - 1 -23 -11 -5 - 4 - 1 -2 - 2 -2 -26 -17 -6 -5 -2 - 3 - 3 - 3 -32 -24 - 11 -8 - 3 -5 -5 - 5 -39 -28 -16 -11 -5 -7 -6 -6 -44 -32 -20 - 13 - 6 - 9 - 7 - 7 -45 -33 -20 -14 -8 -10 -8 -9 -46 -33 -20 -14 -10 -11 - 9 - 10 -52 -38 -24 -16 -14 -14 -12 - 12 -54 -39 -26 -16 -16 -16 -14 -12 -55 -39 -26 -16 -16 -16 -14 -12 -65 -50 -3 1 -22 -20 -20 -17 -17 -70 -54 -34 -23 -20 -20 -17 -17 -87 -70 -51 -35 -21 -23 - 18 -17 -126 -108 -79 -47 -21 -24 - 18 - 18 -148 -129 -100 -57 -24 -26 -19 -19 -166 -150 -106 - 58 -24 -27 -19 -20 -189 -171 -131 -72 -25 -28 -20 -20 -193 -174 -106 -85 -19 -31 -22 -21 -236 -216 -123 -86 -30 -33 - 23 -21

.." TABLE A.9. Water Content Data from t h e McGee Ranch S i t e

Time, s

0.00E+00 5.96E+02 1.50E+03 2.70E+03 4.20 E+03 6.30E+03 9.90E+03 5.97E+04 8.07E+04 3.23 E+05 6.67E+05 1.27E+06


15 30 45 60 90 ----- 0.397 0.413 0.399 0.395 0.391 0.381 0.405 0.397 0.394 0.389 0.374 0.398 0.393 0.387 0.389 0.364 0.395 0.385 0.386 0.386 0.357 0.386 0.386 0.388 0.385 0.342 0.375 0.379 0.381 0.383 0.327 0.365 0.373 0.375 0.381 0.250 0.266 0.302 0.313 0.343 0.237 0.253 0.286 0.298 0.324 0.186 0.197 0.224 0.231 0.226 0.163 0.176 0.196 0.204 0.186 0.153 0.158 0.176 0.183 0.157

TABLE A.9. (contd)

Time-Averaged Water Content, cm3/cm3, a t Depth, cm

Time, s

2.59E+02 1.04E+03 2.07E+03 3.46E+03 5.27 E+03 8.12E+03 3.48E+04 7.02E+04 2.02E+05 4.95E+05 9.70E+05

TABLE A. lO. Ma t r i c Head Data from t h e McGee Ranch S i t e

Ma t r i c Head, cm, a t Depth, cm

Time, s 2.59E+02

Time-Averaged M a t r i c Head Data, cm, a t Depth, cm

TABLE A . l l . Hydraulic Conductivity Data from the McGee Ranch Site

Hydraulic Conductivity, cm/s, at Depth, cm

Time, s 2.59E+02 1.04E+03 2.07 E+03 3.46E+03 5.27E+03 8.12E+03 3 .48 E+04 7.02E+04 2.02E+05 4.95E+05 9.70E+05

APPENDIX B

PARTICLE- SIZE DATA

PARTICLE-SIZE DATA

TABLE B.1. P a r t i c l e - S i z e D i s t r i b u t i o n Data f o r Samples f rom t h e BWTF S i t e . Where d isc repenc ies were found i n t h e p a r t i c l e - s i z e d i s t r i b u t i o n , da ta were i n t e r p o l a t e d t o o b t a i n smooth curves.

P a r t i c l e S i ze Sample I D (microns)

BWTF-15A 2 .OE+3 96% Sand 1 .OE+3

3% S i l t 5 .OE+2 1% Clay 2.5E+2

l . lE+2 7.5E+1 5.3E+1 5.1E+1 3 .OE+l 1.6E+1 9.4E+O 6.6E+O 5.4E+O 4.7E+O 1.4E+O

BWTF-16A 2.OE+3 97% Sand 1 .OE+3

2% S i l t 5 .OE+2 1% Clay 2.5E+2

l . lE+2 7.5E+1 5.3E+1 5.1E+1 3 .OE+l 1.6E+1 9.4E+O 6.6E+O 5.4E+O 4.7E+O 1.4E+O

% Less Than Sample I D

98.0 BWTF-15B 88.6 96% Sand 36.2 3% S i l t

9.5 1% Clay 5.0 4.0 3.5 3.8 3.4 3.1 2.3 1.3 1.3 1.1 1.1

P a r t i c l e S ize (microns)

2 .OE+3 1 .OE+3 5 .OE+2 2.5E+2 l . lE+2 7.5E+1 5.3E+1 5.1E+1 3 .OE+l 1.6E+1 9.5E+O 6.6E+O 5.4E+O 4.7E+O 1.4E+O

BWTF-160 2 .OE+3 96% Sand 1 .OE+3

3% S i l t 5 .OE+2 1% Clay 2.5E+2

l. lE+2 7.5E+1 5.3E+1 5.1E+1 3 .OE+l 1.6E+1 9.5E+O 6.6E+O 5.4E+O 4.7E+O 1.4E+O

% Less Than

TABLE B.1. (contd)

Sample ID

BWTF-17A 98% Sand 1% Silt 1% Clay

3% Silt 1% Clay

2% Silt 2% C 1 ay

Particle Size (microns) % Less Than Sample ID

37.6 2% Silt 12.0 1% Clay

2.8

96.2 BWTF-18B 84.3 95% Sand 32.7 3% Silt

Particle Size (microns)

2 .OE+3 1.OE+3 5.OE+2 2.5E+2 l . lE+2 7.5E+1 5.3E+1 5.1E+1 3 .OE+l 1.6E+1 9.4E+O 6.6E+O 5.4E+O 4.7E+O 1.4E+O

% Less Than

96.8

TABLE B.2. Particle-Size Distribution Data from the Grass Site

Sample ID GS-6A

79% Sand 18% silt

3% Clay

GS-7A 73% Sand 23% Silt

4% Clay

GS-8A 69% Sand 26% Silt

5% Clay

Particle Size (microns)

2.OE+3 1 .OE+3 5 .OE+2 2.5E+2 l . lE+2 7.5E+1 5.3E+1 4.9E+1 2.9E+1 1.6E+1 9.2E+O 6.5E+O 5.3E+O 4.6E+O 1.3E+O

% Less Than

98.7 95.7 44.2 33.2 28.3 24.9 21.2 17.5 12.5 10.0

7.5 6.5 5.0 4.0 3.3

Sample ID

GS-6B 91% Sand

7% Silt 2% Clay

GS-7B 92% Sand

5% Silt 3% Clay

GS-BD m 24% Silt

5% Clay

Particle Size (mi crons)

2 .OE+3 1 .OE+3 5.OE+2 2.5E+2 l . lE+2 7.5E+1 5.3E+1 4.9E+1 2.8E+1 1.6E+1 9.3E+O 6.6E+O 5.4E+O 4.6E+O 1.3E+O

% Less Than

TABLE 6.2. (contd)

P a r t i c l e Size Sample I D (microns)

GS-9A 2 .OE+3 76% Sand 1 .OE+3 20% S i l t 5 .OE+2

4% C l ay 2.5E+2 l . lE+2 7.5E+1 5.3E+1 4.9E+1 2.9E+1 1.6E+1 9.2E+O 6.5E+O 5.3E+O 4.6E+O 1.3E+O

GS-1OA 2 .OE+3 73% Sand 1 .OE+3 22% S i l t 5 .OE+Z

5% Clay 2.5E+2 l . lE+2 7.5E+1 5.3E+1 4.9E+1 2.9E+1 1.6E+1 9.3E+O 6.6E+O 5.4E+O 4.6E+O 1.4E+O

% Less Than Sample I D

98.7 GS-9B 96.6 91% Sand 53.4 6% S i l t 37.1 3% Clay 32.1

P a r t i c l e S ize (mi crons)

2 .OE+3 1 .OE+3 5 .OE+2 2.5E+2 l. lE+2 7.5E+1 5.3E+1 5.OE+1 2.9E+1 1.6E+1 9.3E+O 6.6E+O 5.4E+O 4.6E+O 1.4E+O

% Less Than

99.7 95 .O 40.1 14.1 11.1

9.9 9.1 6.3 5.3 4.8 3.3 2.8 2.5 2.5 2.5

TABLE B.3. Par t ic le-Size Distr ibution Data from t h e McGee Ranchsite

Sample ID

MCG-1A 41% Sand 42% S i l t 17% Clay

MCG-2A 35% Sand 57% S i l t

8% C 1 ay

MCG-3A 44% Sand 37% S i l t 19% Clay

Pa r t i c l e Size (microns)

2 .OE+3 1 .OE+3 5 .OE+2 2.5E+2 l . lE+2 7.5E+1 5.3E+1 4.4E+1 2.7E+1 1.5E+1 8.8E+O 6.3E+O 5.1E+O 4.5E+O 1.3E+O

% Less Than

100.0 99.3 97.3 95.8 86.9 70.4 59.5 60.0 45.0 35.0 28.8 23.3 21.3 20.0 15.8

Pa r t i c l e Size Sample ID (mi crons) % Less Than

MCG-1B 2 .OE+3 53% Sand 1 .OE+3 32% S i l t 5 .OE+2 15% Clay 2.5E+2

l . lE+2 7.5E+1 5.3E+1 4.5E+1 2.7E+1 1.5E+1 9 .OE+O 6.3E+O 5.2E+O 4.5E+O 1.3E+O

MCG-2B 2 .OE+3 7% Sand 1 .OE+3

75% S i l t 5 .OE+2 18% Clay 2.5E+2

l. lE+2 7.5E+1 5.3E+1 4.1E+1 2.5E+1 1.4E+1 8.5E+O 6.1E+O 5.1E+O 4.4E+O 1.3E+O

MCG-3B 2 .OE+3 33% Sand 1 .OE+3 53% S i l t 5.OE+2 14% Clay 2.5E+2

l . lE+2 7.5E+1 5.3E+1 4.4E+1 2.7E+1 1.5E+1 9 .OE+O 6.3E+O 5.2E+O 4.5E+O 1.3E+O

TABLE B.3. (contd)

Sample ID MCG-4A

45% Sand 38% Silt 17% Clay

MCG-9A 35% Sand 54% Silt 11% Clay

MCG-1OA 48% Sand 46% Silt 6% Cl ay

Particle Size (microns) % Less Than Sample ID

100.0 MCG-4B 99.8 28% Sand 96.7 55% Silt 93.5 17% Clay 77.8 63.8 55.0 54.5 41 .O 32.5 25.5 22.5 20.0 19.5 16.3

100.0 MCG-9B 99.8 43% Sand 98.6 51% Silt 96.7 6% C 1 ay 87.8 76.2 63.8 56.1 40.6 31.6 23.1 18.1 15.4 14.8 8.9

99.7 MCG-1OB 98.8 48% Sand 94.9 47% Silt 91.5 5% Cl ay 80.1 66.8 52.4 45.1 28.3

Particle Size (microns) % Less Than

2 .OE+3 100.0 1 .OE+3 100 .O 5.OE+2 97.9 2.5E+2 95.8 l.lE+2 91.2 7.5E+1 81.5 5.3E+1 72.5 4.3E+1 70 .O 2.6E+1 53.3 1.5E+1 39.5 8.8E+O 29.8 6.2E+O 24.0 5.1E+O 21.3 4.5E+O 20.0 1.3E+O 14.8

DISTRIBUTION

No. o f Cop i es

No. o f Cop i es

OFFSITE

10 DOEIOffice o f S c i e n t i f i c and Technical In format ion Center

6 DOE O f f i c e o f C i v i l i a n Radioact ive Waste Management

Fo r res ta l B u i l d i n g Washington, DC 20585 ATTN: L. H. B a r r e t t , RW-33

C. R. Cooley, RW-40 J. R. H i l l e y , RW-30 S. H. Kale, RW-20 D. E. Shelor, RW-32 R. Ste in, RW-23

4 DOE O f f i c e o f Defense Waste & Transportat ion Management

GTN Washington, DC 20545 ATTN: T. C. Chee, DP-123

K. A. Chorey, DP-123 G. H. Daly, DP-124 J. E. L y t l e , DP-124

4 DOE O f f i c e o f Remedial Act ion 81 Waste Technology

GTN Washington, DC 20545 ATTN: J. A. Coleman, NE-24

T. W. McIntosh, NE-24 W. R. Voight, NE-20 H. F. Walter, NE-24

3 DOE Idaho Operations O f f i c e 550 Second S t ree t Idaho Fa1 l s , I D 83401 ATTN: J. P. Hamric

S. T. Hinschberger J. L. L y l e

W. J. Brumley DOE Savannah River Operations

O f f i c e P.O. Box A Aiken, SC 29801

F. T. Fong DOE San Francisco Operations 1333 Broadway Oakland, CA 94612

M. R. Jugan DOE Oak Ridge Operations O f f i c e P.O. Box E Oak Ridge, TN 37830

E. Maestas DOE West Val l e y P ro jec t O f f i c e P.O. Box 191 West Val ley, NY 14171

J. M. McGough DOE Albuquerque Operations

O f f i c e P.O. Box 5400 Albuquerque, NM 87185

C. S. Abrams Argonne Nat ional Laboratory P.O. Box 2528 Idaho Fa1 l s , I D 83401

B a t t e l l e Memorial I n s t i t u t e P ro jec t Management D i v i s i o n 505 King Avenue Columbus, OH 43201 ATTN: W. A. Carbeiner

W. S. Madia Technical L i b r a r y

No. o f Copies

No. o f Cop i es

Paul Bembia U.S. Nuclear Regulatory

Commission Geochemical Sect ion Mai l Stop 62355 Washington, D.C. 20555

J. R. Ber re th Westinghouse Idaho Nuclear

CO., Inc . P.O. Box 4000 Idaho F a l l s , I D 83401

W. Brewer O f f i c e o f H i gh-Level Nuclear

Waste Management Washington S ta te Department

o f Ecology Mai l Stop PV-11 Olympia, WA 98504

G. S. Campbell Agronomy Department Washington S t a t e U n i v e r s i t y Pullman, WA 99164

A. T. C la rk D i v i s i o n o f Fuel Ma te r ia l Safety Nucl ear Regul a t o r y Commi s s i on Washington, DC 20555

G. A. Dinwiddie U. S. Geological Survey 12201 Sunr ise Val l e y Dr i ve Reston, VA 22092

E. I. du Pont de Nemours Company

Savannah R ive r P lant B u i l d i n g 704s A i ken, SC 29808 ATTN: R. G. Baxter

M. D. Boersma J. G. Glasscock J. R. Knight M. J. Plodinec C. T. Randal 1

J. F ischer Low-Level Radioact ive Waste

Program U.S. Geological Survey Water Resources D i v i s i o n 12201 Sunrise Va l l ey Dr i ve Reston, VA 22092

L. Frank Department o f Energy 625 Marion S t ree t , N.E. Salem, OR 97310

P. G. Hagan J o i n t I n t e g r a t i o n O f f i c e Carlmont Execut ive 1 4308 C a r l i s l e N.E. Albuquerque, NM 87107

T. E. Hakonson Los A1 amos Nat iona l Laboratory P.O. Box 1663 Los Alamos, NM 87545

D. H i l l e l Department o f P lan t and S o i l

Science 12A Stockbr i dge Hal 1 U n i v e r s i t y o f Massachusetts Arnherst, MA 01003

E. A. Jennr ich EG&G Idaho P.O. Box 1625 Idaho Fa1 l s , I D 83415

W. A. Ju ry Department o f S o i l s U n i v e r s i t y o f C a l i f o r n i a

a t R ive rs ide Rivers ide, CA 92502

D. A. Knecht Westinghouse Idaho Nuclear

Company P .O. Box 4000 Idaho F a l l s , I D 83403

No. o f Copi es

No. o f Copi es

S . Meyers Envi ronmental P ro tec t i on Agency O f f i c e o f Radiat ion Programs

( ANR-458) 401 M S t ree t S.W. Washington, DC 20460

T. J. Nicholson U.S. Nuclear Regulatory

Commission Div. o f Engineering Safety MS NL-005 Washington, DC 20555

J. W. Nyhan Los A1 amos Nat ional Laboratory P.O. Box 1663 Los Alamos, NM 87545

Oak Ridge Nat ional Laboratory P.O. Box Y Oak Ridge, TN 37830 ATTN: J. 0. Blomeke

W. D. Burch R. T. Jubin L. J. Mezga

D. T. Oakley, MS 619 Los A1 amos Sci en t i f i c Laboratory P.O. Box 1663 Los Alamos, NM 87544

E. OIDonnel 1 Earth Sciences Branch D i v i s i o n o f Health, S i t i n g

and Waste Management Research U. S. Nuclear Regul a t o r y

Commi s s i on Washington, DC 20555

L. D. Ramspott Lawrence L i vermore Nat ional

Laboratory U n i v e r s i t y o f C a l i f o r n i a P.O. Box 808 Livermore, CA 94550

J. Rensel High-Level Waste Management Washington S ta te Department

o f Ecology Mai l Stop Pu I 1 Olympia, WA 98504

E. M. Romney Laboratory o f Biomedical and

Envi ronmental Sciences U n i v e r s i t y o f C a l i f o r n i a

a t Los Angeles Los Angeles, CA 90024

Sandia Laborator ies P.O. Box 5800 Albuquerque, NM 87185 ATTN: R. W. Lynch

Technical L i b r a r y

R. Shaw E l e c t r i c Power Research

I n s t i t u t e 3412 H i l l v i e w Avenue P.O. Box 10412 Pal o A1 t o , CA 94304

R. Stanley Supervisor Hanford Sect ion Washington S ta te Department

o f Ecology Mai l Stop PV- 11 Olympia, WA 98504

M. J. S t e i n d l e r Argonne Nat ional Laboratory 9700 South Cass Avenue Argonne, I L 60439

V. S t e l l o O f f i c e o f t h e Executive

D i r e c t o r f o r Operations Mai l S t a t i o n 6209 Nuclear Regulatory Commission Washington, DC 20555

No. o f Copies

No. o f Cop i es

D. Stephens J. M. Peterson Dept. o f Geosciences G. W. Rosenwald New Mexico I n s t i t u t e o f Min ing M. W. Shupe

and Techno1 ogy Socorro, NM 87801 18 Westinqhouse Hanford Company

S. T y l e r Desert Research I n s t i t u t e P.O. Box 60220 Reno, NV 89506

E. P. Weeks U . S . Geol og i c a l Survey Federal Center Ma i l Stop 413 Denver, CO 80225

West Va l l ey Nuclear Services Company

P.O. Box 191 West Val l e y , NY 14171 ATTN: C. C. Chapman

J. C. Cwynar J. E. Krauss S. J. Marchette J. M. Pope

56 J. L. White, Chairman Energy Research & Development

A u t h o r i t y Empire S ta te Plaza Albany, NY 12223

P. J. Wierenga Dept. o f S o i l Science New Mexico S ta te U n i v e r s i t y Las Cruces, NM 88004

I. J. Winograd U. S. Geological Survey Nat ional Center - Mai l Stop 432 Reston, VA 22092

ONSITE

7 DOE Rich1 and Operations O f f i c e

E. A. Bracken G. J. Bracken R. D. Gerton R. D. I z a t t

M. R. Adams T. B. Bergman L. C. Brown J. W. Cammann R. A. Carlson J. D. Davis C. DeFi gh-Pri ce C. J. Geier D. S. Landeen R. E. Lerch H. E. McGuire K. W. Owens S. J. P h i l l i p s J. F. Relyea W. W. Schulz N. R. Wing D. D. Wodrich R. D. Wojtasek

P a c i f i c Northwest Laboratory

W. W. Ba l l a rd , J r . P. A. Beedlow L. L. Cadwell M. D. Campbell J. W. Cary J. L. Downs D. W. Dragnich M. J. Fayer (5) M. G. Foley H. D. Freeman G. W. Gee S. M. Goodwin M. J. Graham J. M. Hales P. C. Hays P. R. H e l l e r T. L. Jones C. T. K inca id R. R. Kirkham G. V. Last M. W. L i go tke S. 0. L ink

No. of Cop i es

J. F. McBride J. L. McElroy D. H. Mitchell E. M. Murphy R. W. Nelson M. L. Rockhold (10) L. E. Roger 0. Sagar

No. of Copies

C. S. Simmons R. L, Skaggs J. A. Stottlemyre G. P. Streile R. L. Treat R. E. Wildung Pub1 i shing Coordination (2) Technical Report Files (5)

Date post:	24-Jan-2017
Category:	Documents
Upload:	duongdan
View:	225 times
Download:	6 times

Characterization of Unsaturated Hydraulic Conductivity at the ...

Documents