ML20198R949

Discusses Compilation of Several Recommended Field Testing Procedures & Other Info on Relative Accuracy of Field & Lab Testing for Hydraulic Conductivity

ML20198R949
Person / Time
Issue date:	09/24/1986
From:	Dale Goode, Weber M, Matt Young NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS)
To:	Ford B NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS)
Shared Package
ML20198R768	List: IR 05000429/2005001 ML20198R949
References
FOIA-99-21 NUDOCS 9901110095
Download: ML20198R949 (20)

Text

.____- _ _ __. _ _ _ - _ _ _ _ _ . _ _ - _ _ . _ _ . . . _ _ _ _ . _ . . _ _ . _

h.e w...a ,

..n , . _ -.-

l -

DANG COVER TESTING .

1 l l NOTE TO: Bill Ford, WMGT g . 24 Sep 86 Fo0M: ode, Mi ghMikeWefe'r,Jo F rstrom, WMGT l~

-s l

SUBJECT:

INFORMATION ON PERFEABILITY TESTING OF SOIL COVEPS AND LINERS l

l in follow up to rur meeting on 22 Sep 86, we have compiled several recomended l 'deld testing procedures ano cther information on the relative accuracy c#

l field and laboratory tests 'or hydraulic conductivity. This informatdcr supperts our position on th'is issue which is that field tests of hydraulic l conductivity should be performed because these tests are cuch more reliable at predicting conservative values of field-scale saturated hydraulic conductivity.

! - These field tests could be performed on.pilet-scale field test fills. As you can see, there apparently is an ASTP stardard technique for the double-ring infiltrometer test.

In addition te this information, the two papers which Steve and Joe passed out yesterday also support our position. The 'irst paper (by Day and Daniel)

I c!carly irdicates the unreliable results often given by lab tests. The fact that the paper identifies'some problems with field tests does not ' imply that the authors would recomend against performing field tests. Cnth field and lab tests are more accurate if proper procedures are used. The second paper (by ,

Mundell and Bailey) also indicates one of the classic problems with lab tests. ,  !

"One sample which exhibited a highar permeability value [almost 10 times higher thantheaverage]wasfoundtocontainacrntinuousverticalsiltseamand judged to be a localized condition . . ." Small localized conditions, which dc '

affect large scale performance, are more likely to be detected by field tests. ,

Furthermore, the permeability values predicted by lab testing were verified by '

taking less disturbed samples and running more lab tests. There is no field scale measure of the permeability of the liner in this study. Therefore, it is r.ot krown if the lab tests agree with the field performance.

cc: JKane SSmykowski l

4

(.

L 9901110095.981217

l. PDR FOIA FONNER99-21 PDR O , , ._ _ __ - - _ . .

y.

NUREG/CR-3038 ,

Tests for Evaluating Sites For Disposal of Low-Level Radioactive Waste 4

Manuscript Completed: July 1982 -

Date Published: December 1982 Prepared by R. J. Lutton, D. K. Butler, R. B. Moede, D. M. Patrick, A. B. Strong, H. M. Taylor, Jr.

Geotechnical Laboratory U.S. Army Engineer Waterways Experiment Station Vicksburg, MS 39180 Prepared for Division of Wacte Management Office of Nuclear Material Safety and Safeguards U.S. Nuclear Regulatory Commission Washington, D.C. 206ti5 NRC FIN B7324 l

-_^ . .. .. .

%Y l'

adequit2 construction qurlity control will be provided to achiev2 the permesbility decretses with de,crezsmg concentration 1his trend son.

j tinues erntd it reaches (21) a "critictl" concentration, where the perme-desired design hydriulic conductivity needed for the project. If close i

! abihty becomes " essentially zero."

quality control is maintained, together with an appropriate predictive

Isboratory testing program (20), the writer believes that laboratory tests The most spectacular changest occur for solutions with Na as the only -

, can be used to successfully predict the conditions necessary to achieve cation, which corresponds to a.SAR of infinity. Usmg this fluid and a an as-built conductivity. Case studies documtriting the performance of Waullema soil, McNeal and Coleman (21) obtained " stable" permeabihty such closely controlled impoundments would be most valuable in fur- of 5.010-' cm s-', a " threshold"i concentration on 800 meq L" (46.75 g I ther substantiating this claim. of Nacl /L). ,

- It is not possible to generalize, because " tap water" thanges with lo-Areasout.-Rtreamces '

cation, and at a famed location il changes with time. Ilut as an example

, of Central Texas " tap water," that on the tampus of Texas A&M Uni-

] 18 Ciroud, J. P., **1mpermeabihty: The Myth and a Rational Approach " Pro- versity has about 0.45 g L-' of dissolved sohds, and typically will contain credings of the laternstaanal Con /cremy on Cannemtrancs, Denver, CO 1984.

200-2.50 ppm of Na,3-5 ppm of Ca, and only traces of Mg. This cation 19 Faure. Y. H., " Design of Drain Beneath Geomembranes: Discharge Estima-tion and Flow Pattems in the Case of taak," Proceedings of tat faternationsi makeyp results in an extremely high SAR, that for practical purposes Caferrwe on Geomem6 rears, Denver, CO,1984. can be taken as inimity. Since the electrolyte concentration of this " tap of a water" (0.45 g L") is less than the " critical" concentration (2.92 g L-').

' 20 Mundell, Clay Bamer lohn A., to Layer and Bade Limit , B., "The ercolation Design Through and Testindrvers, landi:t " Hydrsedic Compacted if it is used as the permeating fluid on a soil sarnple containing some Barners in Sed and Rect, ASTM STP V4, A. L Johnson, et al., Eds., American smectites, which are very frequent and abundant m Central Texas sods, Sooety for Testmg and Materials, P:iinndelphia, FA,1905, pp. 246-262. the laboratory results can grossly underestimate the " stable" perme-ability of the soil in question.

! he author has apparently not considered the importance of the pres-ence of Na in the permeating fluid. liowever, as previously reasoned,

, Discusalon t>y Misuet ricornell-Derder,' M. ASCE (l this omission can account for the significant difference in reported i

permeabilities. Therefore, there is no need to resort to the hypothesized The author is unable to explain the difference between the perme- On these grounds, there is no basis for 3 abihtas obtained from laboratory and field tests..He uses this lack of an presence the recommendation of a of microcrack thicker fabric.'kiners. But more im g explanation to postulate the presence of a rrucrocrack fabric. Without an trates the need to perform the laboratory tests with a prepared 'Nd of

{ independent confirmation of their presence, but because of their exis- known chemical composition add not " tap water."

l tence, the author concludes that clay liners ough: to be thicker than 2 To avoid theinterfereth of Na in the laboratory determination - per-j! -

It (0.61 m). hbui% it'Is'neces' arf'os use i permeating fluid with a low SAR value ne writer's opinion is that the observed difference in permeability can such he the solution of . halt of a divalent cation in distilled water. As ij be atinbuted to inadequate testmg procedures. De author uses the loosely the slathor l's aware (10), the standard fluid used by sod physicists in p defined terms " tap water" and ** freshwater" foe the'lluids used in the permhability tests is a 0.01 N solution of calcium sulfate.

, laboratory tests and to fill the pond respectively. Whde these terms ap- Tli permeability measured with this standard fluid is the " stable"

e pear to indicate that the electrolyte concentration was low in both fluids, permeability, which is the maximum to be expected in the field. If the they do not give any indication concerning the presence of sodium (Na) fluid retained by the linet;is relatively salt free, the actual held perme-i in either fluid. abilitf could only be reasonably. estimated if the laboratory permeant is

he presence of Na ions in the permeating solution has been shown adentsoil to the solution flowing' through the hner in the field. Central

(21) to decrease dramatically the permeability of soils with even minor Texas soils frequently contain gypsum and carbonate nodules which

-: ; amounts of smectites. His elfect is particularly apparent at low salt con- despite the fact that they are only slightly solubic, can modify noticeably g centrations. De two characteristics of the permeating fluid that deter- the concentration and the SAR of a relatively salt. free wates flowing j mine the entent of this effect are the electrolyte concentration and the through the liner. If a reliable estimate of the actual held permeabihty y sodium absorption ratio "SAR," which is the ratio of the Na concentra- s needed, the changes imposed by these soluble salts on the pore so-19 tion over the square root of one half of the sum of the Ca and Mg con- lution should be considered.

Fra r es of permeating solutions of decreasing electrolyte concen- heimo Mgrammets

-] tration, but constant SAR, it is possible to distinguish three distinct elec-

~

21. McNeal. B. I~ and Coleman. fd T.. 'Tilest of Solution Compoution on So,i trolyte concentration regions from their effect on the resultant perme-flydraulse Conductivity," Sod Scente Sx erry of Amerna. Pr,=c..t,,en Vol 3Q abihty. At high. concentrations, the permeability is " stable" (22), i.e., 1966, pp. 308-312 4

independent of concentration. There is a " threshold" below which the 22. Quirk, J. P., and Schoticid. R. K "t he i licie of t ic. ,olyse Conum, mon

'Research Anst , Texaa A4 M Univ., Co!!'5e Station, TX 77840. on Sod Permeability," Jos.rsud of Sod &wenc. Vol 6,1955. pp 163. t /et 3

1465

rei,nf41tratiso capacity Infiltration Capacity Field seasurement Infiltration following rate can standard be measured in the field in accordance with the method:

ASTM D3385-75, " Standard Test Method:for Infiltration Rate of Soils in Field Using Double-Ring Infiltrometers," Annual Book of ASTM Standards: Part 19 +: Natural Building Stones; Soil and Rock, 6 pp.

.v . ~

. , c : : ::.i . -

v. 4 When test results are plotted as infiltration rate against elapsed time from the beginning,of the test, a curve commonly called the infiltration capacity is obtained. The ultimate infiltration rate after a more or less constant rate is achieved is of special importance as reflective of long-term capacit, for infiltration.

The infiltration rate may also be escinated on the basis of previous experience with similar soils in the vicinity of the site. Ideally this estimation may amount to application of ' previous test results obtained for agricultural or other purposes, but the condition of the soil, the soil moisture, and the vegetation all must be" integrated into the comparison.- - + 'i- a ,e ~8 *r~*-

, ri .. . ,, i -

ar Indirect seasurement or estimation '

Inflitration rates appropriate for longer periods can be obtained by use of site-specific stream gaging data or from the curve number method of estimating runoff (see Runoff). The runoff anopnt is subtracted from the site-specific precipitation amount for the appropriate time period to obtain the corresponding infiltration amount. The period under con-sideration is always of great importance since infiltration from storms is. invariably less than that from equivalent cumulative rainfall spread over a long interval. '

\ $

, A-46 e

~ ~ ~ ~ - -'

.. i l .

~.

l l 1

l l

i nat'onal hanc coo < of recommendec metnocs for water-cata acquisition I

i l

l I

4 9 A UNITED STATES CONTalBUTION TO INTERNRTIONAL HYDAOL.OGKR. PROGRAM l Prepared under the sponsorship of the Office of Water Data Coordination Geological Survey U.S. Department of the Interior

• Reston, Virginia l- 1977 f

i

=. .4 .

~

.. .,. 6.D.4. LITERATURE CITED Peters. D.B., 1963. Water avadability, vs Black. C.A.,

i ed.. Methods of soil analysts. Agronomy monograph no. 9 pt.1: Madmos. Wisc.. Am. Soc. Aaronomy, p. 279 283

{ Veihmeyer. F.J.. and Hendrickson. A.H.,

. 1931 The moisture equivalent as a measure of the field capacity of sous: Sou Sci.. v. 32, no. 3. p. Isl.193.

i Youas. K.K., and Dinos. J.D.,

1966 Overestimation of water content at field capacity from sieved sample l data: Sod Sci.. v.101, no. 2. p.104107.

h

! 6.E. SOIL-WATER MOVEMENT i

The entry of water into soil, movement to plant roots, flow to drains, seepage, and evaporation j

are a few of the processes in which the rate of water movement plays an important role. Soil water i

responds to differences in potential and moves from areas of high potentialinto areas oflow poten-i tial. Movement due to temperature and osmotic gradients does occur, but it is often minor. The rate i

of water movement is determined by the potential gradient and the hydraulic conductivity.

} With all studies ofinfiltration and permeability (hydraulic conductivity), at least three rules are j

important: (1) Carefully select representative sites or soils; (2) use usacting and well-designed tech- '

i niques; and (3) repeat the tests as required by the significance level of your design. Variations are normal and the results should be averaged and the variance should be determined. Land use

i. (grassland, forestland, pastureland, plowed ground, and so forth) may affect water. movement in .

j upper soil horizons much more than does soil type.

I

} 6.E.1. INFETRADON

) Inrdtration refers to the movement of water into a soil as contrasted to the mov' e ment of water

!' through a soil. Because the infiltration rate is influenced by the water content and surface condition of the soil, correct use of these factors is important when interpreting the results. To date, no single

(\

i measuring technique that will work under all conditions has been developed. However, two general i a

methods, flooding and rainfall simulators, are widely used.

l

, 6.Lt.a. FED ogsNG 1

The double-ring inflitrometer is probably the most widely used instrun:ent for measuring in-filtration. Infiltration can also be determined by flooding the soil and measuring the rate of water j intake. A large plot bounded by a wall of soil or some impermeable material to contain the water i may be used. Recome sources for this method are Bertrand (1965, p. 202-207) and Johnson

! (1970, p.187-191).

- . r,w , 8 E1 b Ranxm2. seus Obtaining a satisfactory measurement of infiltration with this method requires that the ar-tificial rain closely simulate natural rainfall and that the plot areas be large enough to represent the given soil. Infutration equals the difference between the amount of water applied and the amount of runoff. The reco==* add source for this method is Bertrand (1965, p.198-201).

6.L2. HYDRAULIC CONDUCnVrrY I This section presents recommended methods for measunns hydraulic conductivity in both the lajboratory and in the field.

6g 1/82 ( i l

i t

I f

Either constant head or falling-head methods can be used in the laboratory to measure . I hydtculic c:nductivity under saturated conditions. The constant head system is best suited to -

h samples with conductivities greater than approximately 0.01 cm per minute (relatively pervious I soils), whereas the falling-head system is best suited to samples with conductivities lower than 0.01  !

cm per minute (relatively impervious soils). Laboratory measurements of saturated hydraulic con- ,

ductivity are carried out using either disturbed or undisturbed soil sr.mples that are held in metal or plastic cylinders. Recommended sources for these methods are Klute (1%5a, p. 210-221) and American Society for Testing and Materials (1975, p. 298-304).

Comment: For undisturbed cores, the major disadvantages of these methods are the small sam-pie size (which necessitates the use of a large number of samples) and the possibility of leakage along the interface of the soil core and the sample retainer. 1 6.E.2.b. SAM!aAHD. MtB 1

Five techniques can be used in the field to measure hydraulic conductivity under saturated con- l I

ditions. Two of the' techniques (auger-hole and piezometet methods) measure hydraulic conductivity below the water table,t'and three (double-tube, air. catty, and'shalloiw-well pump-lp methods)

~ -

measure hydraulic conductivity above the water table.  :

1 6.E.2.b.t. AUGEa.HOLa hWTHoo The auger hole method is based on the measurement of flow into an une===d cavity. The  !

hydraulic conductivity e=leula'ed from the results of this test is an average value of the horizontal i conductivity of primarily the layers below.the water table penetrated by the hole. In stratified soil the results are dominated by more permeable horizons; hence, the method is.of most value in unstratified soil / The recommended source for this method is Boersma (1965a, p. 223 229).

Comment: The suger-hole method is difficult to use in rocky soil or in coarse gravel, in soils -

with very high permeability rates, and under conditions in which the water table is at or above the ground surface. '

s s.n.2.b.2. mzonamm urmoo -

The pWa-e method is based on the measurement of flow into an uncased cavity at the 12wer end of a cased hole. Because the vertical dimension of the unlined cavity is small, the method is well suited for measuring the hydr' a ulic conductivity of individual layers of soil. In this method the length of the cavity is generally.several times its diameter, and the horizontal component of con-ductivity is measured.p.. wider.the hole and the shorter the length of the cavity left unlined, the more nearly the"measuruinient haramma the vertical conductivity. The tube method developed by Frevert s' a d WT1948),"s modification of the riezometer methad, is designed to measure ver-tical conductivitjphs.'recommeaMsource' for the piezometer method is Boersma (1965a, p.

229 233). " . . J. " , .

Comment: The awa=-'- method is difficult to use in rocky soils. Even when the tube can be installed in these soils, it is difficult to do so without leaving channela along the outside of the tube.

Also, it is difficult to establish cavities of the correct dimensions. The diameter of the cavity is very important when calculating the hydraulic conductivity, making a stable cavity mandatory for '

reproducible results.

1/82 6-9 l

i - . . ,

6.a.2.b.3. DOUnt.a.TUaE METHOD The double-tube method uses an auger hole above the water table. The hole is excavated to the j

j depth at which a measurement of hydraulic conductivity is desired. This method can measure hydraulic conductivity of a well<iefined sample area in the absence of a water table. Results from the double-tube =*thad in the field compare favorably with hydraulic conductivity values obtained

(

i. in the laboratory from soil samples taken at the bottom of the auger hole after completion of the i field tests. The recommended source for this method is Boersma (1%5b, p. 234 242).

! Comment: The double tube method measures hydraulic conductivity in an orientation between l vertical and horizontal. It is time consuming, requiring a day to characterize a volume about the size i of a 4 inch core.

I j 6.a.2.b.4. Am.aNTRY MamOD

)

l With the air-entry method, hydraulic conductivity is calculated from Darcy's equation using in- i j filtration rates measured under high-speed conditions. The recommended source for this method is i

Bouwer and Jackson (1974, p. 631-633). -

Comment: lliis method essentially gives the value of hydraulic conductivity. at. the air-entry value of matric suction.21:ia.value is approximately equal to half of the saturated conductivity (Bouwer,1966). .

6.E.2.6.s. sHAROw.wE1.L PUMP.IN MuTHOD Hydraulic conductivity of soil in which no water table is present can be determined in place by

, measuring the rate of flow of water from a cased or uncased auger hole when a coastant height of water is maintained in the hole. This method is known as the shallow well pump in method or the -

dry-auger-hole method. 'Ihe shallow well pump.in method permits the measurement of an average

hydraulic conductivity for the full depth of the hole being tested..The final value, however, reflects l primarily the conductivity of the more permeable layers. The maandad socce for this method

! is Boersma (1965b, p. 242-244).

4

. Comment: Limitations of tiu shallow-well pump in method are that large quantities of water are needed, considerable equipment is required, and the procedum is time consuming. Values of l hydraulic conductivity obtained with the shallow well pump-in method an usually lower than values  ;

i obtained with the auger-hole test. - l

l i

i s.a.s.a. unaATvuAtun synaAUUC CONDUCT!vrrY 4

Hydraulic' conductivity declines many fold for most soil materials as tension increases from saturation to'0.,1,bar. Measuring unsaturated hydraulic conductivity requires information on both ,

the tension and thalate of water movement. The relattaashira and some msehada are described by Klute (1965b)[M"and Jackson (1974), and Bouma and others (1974). The crust method for measuring unsaturiated hydraulic conductivity in the field, described by Bouma and others (1974), is less time consuming and is less difficult than measuring saturated hydraulic conductivity in the field.

6.L3. UTERATURE CNED s

American somery for Tamins and Materials.1975, saansard D24M es, Peneenbility of sranular soils (commans head), Annaal book of ASTM mandards, pt.19. PN Am. sec. Testias Mesmals, p. 29s404.

6-10 1/82

(

. ~ . . ...

. - . - . - - . . - - . . _ . . - - ~. - ... -_- - . ~. - - _ . - .

Bertrand. A. R.,1965, aats tf water intake in the field, i.1 Black. C. A., ed.. Methods of soil analyus, Aartnomy monograph no. 9. pt. -

1: Mediros. Wisc., Am. Soc. Astosomy p. 197 209.

Boersas, t,yckle,1965s. Field measuransat of hydraube ooeductivity below a water table, in Black. C. A., ed., Methods of soU analysis.

Aeroneasy moneeraph so. 9, pt.1: Madison, wisc., Am. Soc. Agronomy, p. 222233.

1965b, Fleid meansement of hydradac conductivity above a water table. m Black. C. A., ed., Methods of soil analysis.

Agonomy sesesymph'an. 9. pt.1: Madison. Wisc., Am Soc. Agronomy, p. 234 232.

Bouma Johana Baker, F.O., and Veneman P. t M. 1974. Measurement of water novament la soil pedoes above the water table: lo.

formenoa Cire. No. 27: Mediaan Wisc. Univ. Wisc. Geol. and Nat. History Survey,114 p.

Bowser. Hennen.19s6, Rapid fleid measuremsat of air. entry value and hydraulic coeductivity of sod as significaot parameters in flow sysseum analyss: water assources assearch, v. 2. no. 4. p. 229 23s.

Bouwer Hennes, and Jacksoa. R. D.,1974. Determinans soil properties, he Von Schilfgaarde. Jao, ed., Drainage for agriculture, Agroecesy monograph no.17: Madison. Wisc., Am. Soc. Agronomy, p. 631633.

Frevert. a. K.. and Kirkham. Doa.194s. A fleid method for measurias the pennenbility of sod below a water table: Proc. Mishway Rs.

Board, v. 28, p. 433 442.

Johnson A. I.1910 sussened mathed of test for infutrados rate in field usias double.rtag infiltrosasters, u Spedal procedures for testias soil and rock for ensineertas purposes. AS% spetaal techancel pubication 479: Pldladsiplus. Am. Soc. Testas Matenais, p.137191.

luuta. Arnold.196Sa, laboratory neumrement of hydradac condamivity of saturased soil, h Black. C. A., ed., Methods of soG analyss. Agromossy monceraph so. 9. st.1: Madason. Wisc., Am. Soc. Agromaeny, p. 210 221.

1965b, if .,,,,, measurement of hydraulic conductivity in unsanarated sog. m Black. C. A., ed., Methods of soil analysis.

Apomony moneyept me. 9, pc.1: Masinos, wisc., Am. sec. Ayamony, p. 2ss.asi.

6.F. QUALITY CONTROL

. m. . . .

6.F.1. DISCUSSION ....

Quahty control in water-data acquisition may be accomplished in three ways: (1) improving techniques and procedures to minimiw potential sources of experunental errors; (2) sampling ade-quate representatives of a system; and (3) making a sober and realistic appraisal as to what con-stitutes normal field variability and, thus, what range of data values may be considered acceptable.

TAsul 6-2.-Coef5cient of Variation (C.), description of data, and source for specified values of g water content (#)

CeeSdest of Water Coesent, # verlados, C, Descripties Soares for spectand (cat */ca') (pareset) of data valess of weser costant At saturation 8 56 between cores Mason and others,1957 4 11 within series Rogowski,1972 7-11 within series Mason and others,1957 3-11 within series Cassel and Bauer,1975

20 between soils Rogowski,1972

Field soil 8 5-23 by weight Reynolds,1970 i 3-17 by volume Reynolds,1970

{ 9 21 - bare soil Reynolds,1970 1 10 33 vegetated Reynolds,1970

! Moisture characteristic 10 23 150. hectare Beid Nielson and others,1973

$ At one-third bar ' 10 within plots Ike and Outter,1968

-16 19 within series Ike and Outter,1968 At 15 bars 14 16 within plots 1ke and Outter,1968 20 28 within series 1ke and Outter,1968 7 35 within series Rogowsid,1972 25 43 between soils Rogowski,1972 11-16 within series Cassel and Bauer,1975 13-55 with depth Cassel and Bauer,1975 $

i c=I~i ead from values siven for tr.dk density assumms i.65 as particle density.

8 Calculated from values given in tables 3 and d of Reynolds (1970).

7 1/82 6-11

l

~ .- .

1

~ . ' *..~

i

}

1 SOIL PROPERTIES, CLASSIFICATION, AND  ;

HYDRAULIC CONDUCTIVITY TESTING i

4

! i

}

Draft Technical Resource Document

• for Public Comment l (SW-925) i i I

A P

re i

MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY OFFICE OF RESEARCH AND DEVELOPMENT U.S. ENVIRONMENTAL PROTECTION AGENCY CINCINNATI, OHIO 45268 OFFICE OF SOLID WASTE AND EMERGENCY RESPONSE U.S. ENVIRONMENTAL PROTECTION AGENCY WASHINGTON, D.C. 20460 thrch 1984

~~

5 SECTION 7

SUMMARY

t The conclusions that can be drawn f rom this study are: (1) The professions of geology, hydrology, area of soil testing the for hydraulic c science as all these disciplines have made attempts to measuresoil engi the rate of liquid movement thru soil materials; (2) A high determination have been developed for agricultural g or en purposes design other thanwaste of hazardous the application disposal sites; to the feasibility and/or (3) conditions due to small size ofmethods suffer from potential field misrepres samples when transported or remixed; samples (4) and/or disruption of testing techniques textured has generally been limited to more coExperience soils rather with field than fine-grained soils arse-appropriate possible at for hazardous waste disposal sites; that are more testing (5) It is not results caused by the variation inherentthis time to disce tosting method compared to the variation of the in the soil ties of the soil itself; and (6) spatial proper- ,

of an applicable saturated - unsaturated transport m prodiction waste or estimation disposal site. of behavior of a proposed hazardous Important testing which are methods shown are summarized in the Soilo sTesting Matrix Meth dcon in Tables 7.1 and 7.2.

l information tion'of for laboratory and field methods forTable 7.1 summarizes saturated hydraulic conductivity whilethe determina-Table 7.2 is dirGeted at unsaturated hydraulic conductivity methods tion methods, and diffusivity methods. , calcula-121 i

\

, _- .. - - -. - -- . . . - .. . . .. ..-.... - ~ - . . -..,.~ . ..- - _

! ' f .. , * .' . 4 - .

s. .

I TABLE 7-1 SOIL TESTING METHODS MATRIX / SATURATED HYDRAULIC CCNDUCTIVITY l

\

l t

1 i

METHOD APPLICATION PRECISION AND ACCURACY Fair-many samples necessary PRESSURE CELL Land treatment to obtain 956 confidence

, Itmits i O Not available (new method) i 0 COMPACTION MOLDS Liner evaluation A

l e j l -J i Fair-direct measurement of O consolidated sample is much j

g CONSOLIDATION CELL Liner evaluation more precise than K computed q from consolidatien and  !

compresston data 1

>= Good-reproducible results

1. MODIFIED TRIAXIAL l #

i Liner evaluation APPARATUS i

Fair-measure average of vertical and horizontal i PIEZOMETERS Land treatment components of K in all

soil layers below water table =

l .

Good-reproducible results DOUSLE RING INFILTROMETER Land treatment l 3 / PERMEAMETER f-

.I a .

Good-reproducible results l O AIR-ENTRY that compare favorably with pgpugAMETER L888 If 8888'#I other methods l  %

i 4 i g Good,large site of sample p-4 more representative of in W CUSE Land treatment situ conditions, can measure vertical and horisontal R separately Good-large sample site and reproducible results, can seasure both saturated CRUST Land treatment and unsaturated K 122 e

i i ,_ _ _ .._

s.

LIMITATIONS OF TEST METHOD STATUS COMMENTS (1) Small sample may =e unrepresentatsve Agricultural Simple and iner-of actual field conditions, (2) several standard pensive equip- ,

days required for fine-textured soils, ment and (3) saturatton of sample not assured (1) Small sample, (2) excessive gradients Experimental Special equipment may cause sidewall flow, (3) interaction developed for use between metal cell and waste, (4) satur- of particular atton of sample not assured, and (5) waste and test will take 1-5 months compacted soil (1) Small sample, (2) falling head pro- Engineering Slight modifica-cedure may require many days to perform standard for tion of common test, and (3) saturatton of sample not consolidation engineering assured data laboratory equipment (1) Small sample, and (2) recommended Engineering Major modification hydraulte gradients in range of 5-20 standard, of common engi-common for clay neering laboratory soils with low equipment ydraulle conductivity *

(1) Errors due to smear zones, (2) re- Standard test Many variations quares presence of water table, and (3) for areas with in equipment and ,

measures both vertical and horizontal shallow depths and procedures hydraulic conductivity to water table (1) Care must be taken during placement ASTM star.dard, Simple and inen-of rings into soil, (2) air trapped be- also commonly pensive equip-low wetting front will effect reeults, used in agri- ment, easy method (3) a few days required for fine-tes- culture and to perform tured soils, and (4) if uncovered, artigation correction for evaporation snould be made (1) Care must be taken during placement Relatively new Moderately inen-of cylinder into soil, (2) will not method, use sn- pensive equipment work wer11 on inattally wet soils, and creasing due to (3) difficult on soils with gravel or ease of proce-stones dure (1) Method will require a few days for Relatively new Very inexpensive clay sotis, (2) sample saturation can- method, use sn- equipment and not be assured, and (3) swelling of creasing due to materials sample may effect measurement ease of proce-dure

, (1) Difficult to assure good contact Relatively new Moderately ines-

! between soil pedestal and ring method, devel- pensive equipment, oped in connec- easy to perform tion with EPA for saturated K sponsered  %

university research i

123  !

l l

,j I'

ig 1

U- -

TABLE 7-2

, t SOIL TESTING METHODS MATRIX / UNSATURATED HYDRAULIC CONDUCTIVITY i

t l

l METHOD APPLICATION PRECIS!ON AND ACCUR ACY w

0- Fair-variability de=

1 creases as length of l 8 8 A I^ column increases

/ COLUMN Unsaturated zone o 1 D

e i =J s j

~

O UNSTEADY STATE rair-many variations of is / INSTANTANEOUS Unsatu:ated zone method, field method  ;

k.

PROFILE more accurate i 4

E f Good-accuracy of, and range q THERMOCOUPLE of suction of psychrometers

1. PSYCHROMETERS Unsaturated zon? makes methoc particularly 2 applicable to compacssd t 3 arid so11s

, Gooa-large sample size

.I and reproducible results, i

• CRUST can measure both saturated

\ Unsaturated zone and unsaturated K O

%l ll > +

4 j.

$. Cood-probably the most

' k accurate field method

( INSTANTANEOUS i

e PMOFILE Uneaturated zone because of the large 3 sample size j 3 -

i

, I' j ,

s Fair-calculated values 6 O never as good as measured b O VARIOUS "'1"

O' E

, J h. PROCEDURES Unsaturated zone t 4 > 188 o.(a 3 3 .'. .

h l l$  ; Fair-disagreement among authors regarding precision 1: ,

h PRESSURE OUTFLOW Unsaturated zone and accuracy o

5 e

j w Fair-because of dependence e ik on the slope of the water g g a HOT AIR Uneaturated tone content curve and determina-tion of water contents y l gravinetrically a 1 -

t i' l 124 i

l-t

. ~.s - w .. . - . . . . _ . .

• s**

I

LIMITATIONS OF TEST METHOD STATUS COMMENTS (1) Metnod wall require longer time for Agricultural clay sotis, (2) small sample, (3) pro- standard Inexpenstve equipment cedure yields K(h), not K(e), (4) K determined from desorption rather than  !

absorption data, and (5) suction lim- I ited to range of tensiometer measure- I ment (1) Method wall require longer time for Agricultural Espenstne and clay soils, (2) small sample, and (3) standard results limited to range of tenstometer potentially dan-measurement gerous equipment, detailed proce4ure (1) Small sample, (2) applicable to clays Experimental Moderagely expen.

with degrees of saturation between 30-904, and sands less than 50%, (3) psy- save equipment, chrometer corrosion in acid soils, and detailed proce-dure *

(4) cannot measure K near saturation (1) Several days required to achieve Relatively steady state flow under crusts of high Moderately ines-new method, pensive equipment, resistance, (2) difficult to ass tr6 good developed in repetitive proce-contact between soil pedestal and ring, conoection dure with crusts and (3) results limited to range of with EPA of different tenstometer measurement sponsored resistance university research (1) Results limited to range of tensio- Agricultural Moderately expen-meter measurement, (2) field plot must standard be level, (3) not applicable in soils sive equipment, easy procedure with high lateral flow, and (4) plots once set up should be larger if surrounding area is strongly evapotranspiring (1) Limited to more coarse-textured soils, Experimental Large variety of (2) matching factors must be determined, methods and (3) matching factors most often deter-mined at or near saturation (1) small sample, and (2) many variations Agricultural Moderately inen-of method using disturbed and undisturbed and ASTM pensive equipment, samples, inflow rather than outflow meas- standard detailed procedure ured, or one large pressure increment used instead of several small ones (1) Small sample, (2) requires moisture Relatively new Inesponsive equip-retention curve to calculate K. (3) not method, use in- ment, easy proca-very reliable near saturation, and (4) creasing due to dure limited to soils with relatively low short time g conductivities in the low tension range period for test 12' M

_ _7 ._ _ .

4.3 COMPACTION / HYDRAULIC CONDUCTIVITY TESTING ERRORS The normal procedure for determination of the hydraulic conductivity of a compacted soil sample is to compact the soil in a

mold and then to test for hydraulic conCictiv Ly on that sample. The samples so tested are usually cylindrical or disc-  ;

shaped with when trying to estimate diameters between 3 and 15 centimeters. However, laboratory field hydraulic conductivity from compaction and hydraulic conductivity testing, there are many sources of error possible during both laboratory compac-tion procedures conductivity testing.as The welltypes as during laboratory hydraulic and sources of these errors are ,

discussed below.

Effect of Comnaction Water Content It has been clearly established that hydraulic conductivity of saturated samples is relatively high for samples compacted dry of optimum water content while the hydraulic conductiivity is relatively low for samples compacted wet of optimum water con-tent.

Daniel (1981) reported that the hydraulic conductivity of '

soils- comoacted dry of optimum micht tvoically be 10 to 1000 times larger than the hydraulic conductivity of soil comoacted wet of optimum.

For this reason, gross errors in predicting tielc hydraulic occur conductivity from laboratory determinations may if the field compaction water content is not as anticipated.

Maximum Size of Soil Accrecates During. laboratory tests, the soil aggregates from the field sample are usually broken down into smaller chunks than exist in the field. Such disturbance of the natural aggregation of soils will influence hydraulic conductivity. ,

Daniel ' (1981) reported that during testing of the same soil with maxim.us aggregate sises of 3/8 inches, 3/16 inches, and 1/16 inch the hydraulic conductivity of the smallest size class was nearly two orders of magnitude less than the hydraulic conduc- '

tivity of the largest size class. This implies that if aggregate sizes field, are much smaller in the laboratory sample than exist in the the laboratory tests may determine hydraulic conductivities that are much lower than true field values.

Presence of Deleterious Substances

.Similar to the situation with differences in soil aggregate sizes between laboratory specimens and field conditions, the presence or deleterious substances in the field such as roots or rocks or any other material not included in the 3-15 cm $

laboratory specimen may cause substantial discrepancies between the hydraulic conductivity measured in the laboratory and what will actually occur under field conditions.

48

Method ~ of comenction While most laboratory hydraulic conductivity tests on soils are perSrmed on samples prepared with impact compaction using a drop hammer,. such equipment bears no resemblance to any pieces of field compaction-machinery.

Figure '4.10 presents a comparison of field and laboratory compaction on.the same soil. The figure illustrates.the difficu'lty of choosing-a laboratory test that_ reproduces a given field compaction procedure. The laboratory curves generally-  ;

-yield a somewnat lower optimum water content than the actual l field optimum.

is 2e t . .

e,

- I 3 is ,

2 +

n <r -

3.::, .

t - s ,

5 is _

'y j ts s , ,< '

se i  ! t f so to ao as .

WATER CONTENT '(%)

Figure 4.10 Comparison of field and laboratory compaction. (1)

Laboratory static compaction,13.8 MN/m2 (2) Modi- ,

fled AASBO (3) Standard AASBO (4) Laboratory static compaction 1.38 MN/m2 (5) Field compaction, rubber- i i

tired load, 6 coverages (6) . Field compaction, sheeps-foot roller, 6 passes. Notes Static compac-tion f rom top and bottom of soil sample (Lambe and Whitman, 1979).  ;

t Additionally, Mitchell et al. (1965) compared static compac- ,

tion'and kneading compaction and reported that similar hydraulic '

conductivities were found _cn samples compacted dry of optimum while kneading compaction F oduced hydraulic conductivities 3 nearly five times less than stetic compaction when samples were

. compacted wet of optimum.

i 49

,c I[

I b l

s i L- ,

. comnactive Effort Many researcher' slave found that hydra ~ulic conductivity of compacted soils is very sensitive _ to compactive effort. Mitchell et al.

soil (1965) reported that in studies on compacted silty clay that the hydraulic conductivity may decrease by two orders of magnitude, with no change in density or moisture content, simply by changing the compactive effort. Therefore, it is very 5

important to make certain that the compactive effort used in the laboratory is reasonably close to the compactive effort that will be used in the field.

Air in the samnle I. testing compacted samples, it is often assumed that soaking the samples f rom the bottom, with the top open to the atmosphere, will yield saturated samples. However, Jackson (1963) reported that for loam soils, only 79-914 of the total porosity was fillable by water. Because water cannot pass through air pockets, such pockets will effectively reduce the pore space tnat can be occupied by water and thus recuce hydraulic conductivity. This phenomena is one of the main reasons why laboratory hydraulic conductivi";v results are generally lower than hydraulic conductivities unc er actual field s

conditions.

N Excessive Hydraulic Gradients It is virtually impossible to duplicate field hydraulic gradients (usually less than 1) in the laboratory as testing time becomes excessive as well as it is difficult to obtain accurate measurements of the low flows and heads associated with very low hydraulic gradients.

Since Darcy's Law indic4ces a linear relationship between flow rate and hydraulic gradient, many workers have used elevated hydraulic gradients to reduce testing time. However, if hy-draulic gradients become excessive, piping or particle migration may ments.

occur and adversely affect hydraulic conductivity measure-Criteria for selecting an appropriate hydraulic gradient depends on the soil type and the proposed use of the hydraulic conductivity study. In comparative studies where qualitative, rather than quantitative analyses are needed, larger gradients may be used. Wardell and Doynow (1980) used hydraulic gradients of 48 and 67 in a triaxial cell, while Brown and Anderson (1982) utilized hydraulic gradients of 61.1 and 361.6 in a rigid-wall permeameter.

In ooth studies, no piping, particle migration, or non-Darcy behavior was observed.

Bowever, where the objective is to quantitatively estimate 50

field hydraulic conductivity values from laboratory results, ,"

Olson and Daniel (1981) have suggested use of hydraulic gradients as close to those encountered in the field as is economically '

feasible. Likewise, Zimmie et al. (1981) have recommended use of hydraulic gradients of.5-20 (with gradients nearer the lower end of the range to be preferred) for laboratory studies attempting to duplicate field conditions.

Samole Size The measurement of hydraulic conductivity in small cores offers many practical problems as such cores hay not be represen-tative of in situ conditions where root holes, cracks, and fis- i sures are present. Thus, the size of the sample used to test -

hydraulic conductivity is impor tant if such information is used .

to predict field behavior.

Anderson and Bouma (1973) experimented with a series of cores of different lengths to determine the effect on hydraulic conductivity. They found that 17 cm long cores had hydraulic conductivities that were half a magnitude less than 5 cm. long. I Daniel (1981) measured the hydraulic conductivity of a com-pacted clay liner on samples of various sizes in the laboratory with one very much larger sample tested in ghe field. The r e-sults were: 3.8 cm diameter core 1 x 10-' cm sec-1; 6.4 cm .

diameter corg, 8 x 10-9 cm sec-1; an,d 243.8 cm diameter core, 3 x 10-5 cm sec-A. Additionally, the average hydraulic conductivity '

of the liner was bact-calculated f rom measured leakage rates and found to be 1 x 10-3 cm sec-1 Such results demonstrate the  ?

significance of sample size _ in predicting field hydraulic conductivity values.

Table 4-2 is a summary of testing errors possible when testing for the hydraulic conductivity of compacted soils (Dan i el, 19 81) . It also shows estimatas of the possible magni-tude of error associated with each problem and an indication of .

whether the laboratory hydraulic conductivity is likely to be higher or lower than the field value. The estimates of error are based on available data 'and are intended to show trends rather than precise values.

h i I 51

l . .

V** : ,

i TABLE 4-2

SUMMARY

OF SOURCES OF ERROR IN ESTIMATING FIELD HYDRAULIC CONDUCTIVITY OF COMPACTED CLAY LINERS FROM LABORATORY TESTS ,

l Possible Number of Orders of Laboratory K Magnitude of Potential sources of error Too Hich or Low? Error Compaction at a higher water content in laboratory than field Low 1 to 3 l Maximum size of clay chunks smaller i

in laboratory than field Low 1 to 2 Deleterious substances present in the field but not in laboratory  :

samples Low 1 to 3 l Use ot static or impact compaction rather than kneading compaction to prepare laboratory specimens High 0 to i Use'of more compactive effort in the laboratory than in the field Low 1 to 3 l Air in laboratory samples Low 0 to 1 l Use of exc,essive hydraulic gradients Low 0 to 1  !

! Sample size Low 0 to 3 l

l 1 I

,. 4 i

d a

52 l

L I

W ,

W ,

%W j

, Liner .

hwas: M 74>% V

V Piqure 5.2 Water and waste movement through a soil liner.

Because soil liners are constructed from disturbed or admixed materials, thete is no simple and reliable way to test ,

them in situ. Accordingly, hydraulic conductivity must be i ,

performed on compacted laboratory specimens that will be used in ;j l

the field. Therefore, as the facility is constructed, the field.

density should be checked to ascercain that density and associated hydrauli,c conductivities are related to the laboratorv model.

Laboratory methods for determining saturated hydraulic con-ductivity on compacted specimens are: ,

1. Compaction Molds (Section 6, pp. 60)

2. Consolidation Cells (Section 6, pp. 64)

3. Triaxial Apparatus (Section 6, pp. 66)

4. Thermocouple Psychrometers (Section 6, 97)

• 5.3 THE UNSATURATED ZONE Another type of liquid movement that is relevant in all types of land disposal facilities is movement in the unsaturated zone between the root zone or liner and ground water table or bedrock as depicted in Figure 5.3.

As described in Section 3, unsaturated hydraulic i conductivity is more difficult to measure than saturated '

hydraulic conductivity due to the fact that the unsaturated hydraulic conductivity varies with both moisture content and pressure head and therefore must be determined over a range of i

values while saturated hydraulic conductivity will be a constant value.

Also, it can be noted that testing of the unsaturated zone during feasibility and design stages will be of benefit later as for most systems there will be the requirement for monitoring of the unsaturated zone af ter construction of the f acility. Good decisions made during f easibility and design stages for types and 3 locations of unsaturated hydraulic conductivity tests will facilitate the unsaturated zone monitoring requirements.

55

. . _ - - . . ~ .. 1 . . . - ...

_ ______________ ____________- ____ _