AD-Ai62 528 INSTALLATION RESTORATION GENERAL ENVIRONMENTALTECHNOLOGY DEVELOPMENT TAS (U) NESTON (ROY F) INC NESTCHESTER PA W P LAMBERT ET AL MAR 85 AMXTH-TE-TR-85M4

UNCLASSIFIED DAAKli-82-C-867 F/G 13/2 UL

IEIEEIIEEEIIElllllE-EllIIElllEllllEEElhE

* -. '__.° '

LL. 13. M

11111112._'1""11= n I2.

-~11111L2 LA 14 .6

MICROCOPY RESOLUTION TEST CHART

NATIONAL BUREAU OF STANDARDS- 1963-A.-B -

'.- .2

Contract No. DAAKII-82-C-0017Task Order 8

-. d

P N INSTALLATION RESTORATION GENERAL ENVIRONMENTALLf'

TECHNOLOGY DEVELOPMENT

Task 8. Bench-Scale Investigation ofrI Low Temperature Thermal Removal of TCE from Soil

FINAL REPORT

Walter L. Lambert, Ph.D. D TICLawrence J. Bove r!.. -CTE6 David Burkitt V

Scott Birk DECDavid Jo RussellPeter J. Marks

*0 -

ROY F. WESTON, INC.West Chester, Pennsylvania 19380

1 March 1985

Distribution Unlimited. Approved for Public Release.

Prepared for:

U.S. ARMY TOXIC AND HAZARDOUS MATERIALS AGENCY

Aberdeen Proving Ground (Edgewood Area), Maryland 21010-5401

I 5389A

I OTiC FiLL{ COP

?AV . .

'IL

The views, opinions, and/or findings contained in this report-.3 are those of the authors and should not be construed as an of-ficial Department of the Army position, policy, or decisionunless so designated by other documentation.

53.1I

=X J rU 1 VNT L .,-u

UNCLASSIFIEDSECURITY CLASSIFICATION OF THIS PAGE (When Date ffntered)

READ INSTRUCTIONSREPORT DOCUMENTATION PAGE BEFORE COMPLETING FORM1. REPORT NUMBER 12. GOVT ACCESSION NO. 3. CI ENT'S CATALOG NUMBER

4. TITLE (and Subdd,) S. TYPE OF REPORT & PERIOD COVERED

ench-Scale Investigation of Low Temperature Thermal FINAL REPORTREmoval of TCE from Soil June 1984 - February 198,

6. PERFORMING ORG. REPORT NUMBER

7. AUTHOR(,) S. CONTRACT OR GRANT NUMBER(&)

W.P. Lambert, L.J. Bove, D. Burkitt, S. Birk, andD. Russell, and P.J. Marks DAAK11-82-C-0017

9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT. PROJECT. TASKROY F. WESTON, INC. AREA & WORK UNIT NUMBERS

Weston WayWest Chester, PA 19380

11. CONTROLLING OFFICE NAME AND ADDRESS 12. RE 9T OATESU.S. Army Toxic & Hazardous Materials Agency Feb'ary 1985Aberdeen Proving Ground (Edgewood Area) 3 s. NUMBER OF PAGESWayne E. Sisk, Project Officer PP010-5401

14. MONITORING AGENCY NAME S AOORSS(I dfetmnt f Cm&*ulind Office) IS. SECURITY CLASS. (of tde rapact)

IS. DECASSIFICATON/ DOWNGRADINGSCHEDULE

I. DISTRIBUTION STATEMENT (of dde Repma)

Distribution Unlimited. Approved for Public Release.

17. DISTRIBUTION STATEMENT (of Ih* a6ba mntd in leek 2, if dffet um A e po)

IL SUPPLEMENTARY NOTES

19. KEMY WORCS (Contine On re"ere aide Ui fedosoa Mot Idontit7 br 6109k aMike)

Soil ContaminationTrichloroethylene (TCE)Air StrippingSoil Decontamination

- 2L A@S"ACr (C. -N W1onm b we [ md I rbl k nmmber)

Bench-scale studies were performed to demonstrate the viability ofthe concept of low temperature thermal volatilization of TCE from

* soils. A factorial experimental design using multi-way Analysis ofVariance (ANOVA) for data analysis was employed. Factors weretype of soil (3 levels), TCE concentration (2 levels), operatingtemperature (3 levels), and moisture content (2 levels). Duplicateswere run.

D0 , 173 Rm o Or I NOV s is oUNOLETa

S-cumrY CLASSIFPCATfOlf OF THIS PAGE (MbOM Dtw*e

UNCLASSIFIEDSECURITY CLAIiPICATON OP TH4S PAOIRlhm DOS a lM

The concept was demonstrated. -TCE can be stripped fromcontaminated soils under the conditions of this study. Complete-removal of TCE is dependent on operating conditions.

Statistically significant factors influencing the rate ofvolatilization were operating temperature, soil moisture (dominantfactor), and an undefined interaction between them. Kinetics were~proportional to both temperature and soil moisture. .-

arying amounts of TCE were not stripped from the tested oil atall experimental conditions at temperatures less than 1261C. Thecontent of these residuals was verified by independent anatysis tobe TCE. Only TCE was displayed as output traces by theexperimental apparatus.

All of the experimental factors and each of their binary combinationwere statistically significant to the magnitude of TCE residualconcentrations. Operating temperature was the dominant factor. I..Engineering data implications based on experimental results are: [

(a) Operating temperatures for piloj;-scale equipment should '

be greater than or equal to 120TC for complete removalf TCE in less than 20 minutes .

h-)b 'Pilot-scale equipment should be of sufficientflexibility to permit verification of the results ofthis study and experimentation to optimize operatingparameters.

- I-

., UNCLASSIFIED

SECURITY CLASSIFICAION OF THIS PAGE(Whon DA NteWeQ

"-i

N CONTENTS

Page

Paragraph 1 EXECUTIVE SUMMARY ....................... 11.1 Volatilization kinetics ............... 11.2 TCE residuals ... .. .. .. .. . ... . .. . ... . .. 1

.N1.3 Engineering implications .............. 22 PURPOSE AND OBJECTIVES .......... *........ 42 .1 Reference . .. .... .. . .. .. .. .. .. .. .. .. . .. 4

2.3 objectives .. *.......................**. 4

2.4 Criteria for positive test of concept .43 PARAMETERS AND TEST CONDITIONS .......... 63.1 Test parameters .. .. .. . ... .. . ... .. .. . .. 63.1.1 Definitions ........................... 63.1.2 Relevant test parameters .............. 63.2 Test apparatus ........................ 93.2.1 Background . .. .. .. .. .. .. .. .. .. .. .. .. . .. 93.2.2 Test apparatus . .. .. .. .. .. .. .. .. .. .. . .. 103.3 Initial experimental design ........... 153.4 Shakedown protocol .................... 17

3.4.2 Exjections ................... .......... 17

3.4.3 Conclusions .. ... ..... ... ... .. . ... ..... 173.5 Experimental design ................... 243.6 Experimental procedure ................ 243.6.1 Soil handling .. .. .. .. .. .. .. .. .. .. .. . .. 243.6.2 Preparation of test cells ............. 243.6.3 Experimental apparatus preparation .... 273.6.4 Execution of experimental design ...... 273.6.5 Temperature profile ................... 283.6.6 TCE verification test ................. 283.6.7 Soil characterization ................. 28

4.1 Typical data output ................... 30

4.1.2 Standard .................... ..... ..... 30

rW 4.2 Experimental results .................. 354.2.1 Experimental moisture levels .......... 35

%I4.2.2 TCE removal rate at operatingtemperature -kinetics .............. 35

4.2.3 TCE residual percentage ............... 354.2.4 Temperature profiles .................. 354.2.5 TCE verification test ................. 354.2.6 Soil characterization ................. 35

5389A

CONTENTS

Paragraph 4.3 Data analysis ......................... 43

4.3.2 Volatilization kinetics ............... 434.3.3 Residual TCE, percentages ............. 464.3.4 Residual TCE, concentrations .......... 515 DISCUSSION ............... .............. 575.1 Volatilization kinetics ............... 575.2 TCE residuals, percentage ............. 585.3 TCE residuals, concentration .......... 605.4 Carrier gas flow rate ................. 606 CONCLUSIONS AND RECOMMENDATIONS ......... 616.1 Volatilization kinetics ............... 616.2 TCE residuals, percentage ............. 616.3 TCE residuals, concentration .......... 616.4 Engineering implications .............. 616.4.1 Assumptions .. ... .......... .. ... ...... . 616.4.2 Implications for ambient temperature

volatilization ...................... 62 :,,

6.4.3 Implications for thermally inducedvolatilization . ..................... 62

6.4.4 options for pilot-scale testing ofthermally induced volatilization .... 62

APPENDIX A -Test plan for a bench-scaleinvestigation of low temper-ature thermal removal of TCEfrom soil .................. A-i

V7W

iv5389A

Ir-0

TABLES

Page

TABLE 1. Analysis of experimental variables ............. 72. Summary ofshakedown testing ................... 183. Characterization parameters for each soil type .29

4. TCE measured by Hall detector duringexperiment 2112 ... .. ............ *........... * . 32

5. Experimental moisture levels for each soiltested ................. *...................... 36

6. Soil characterization .......................... 427. Analysis of variance for volatilization kinetics

at constant moisture (moisture Level 1)conditions ....... e... ...* *** .*...****** 44

8. Analysis of variance for volatilization kineticsat constant TCE concentrations .......... 47

9. Analysis of variance for percentage residual TCE,constant moisture conditions ........... 49

10. Analysis of variance for percentage residual TCE,constant TCE concentration ...... ....... 52

11. Analysis of variance for TCE residual concentra-tion, constant moisture conditions ...... 55

12. Analysis of variance for TCE residual concentra-tion, constant TCE concentration ....... 56

Accesion ForNTIS CRA&IDTIC TABUnannouinced ElJustification

By

I N W C TEDAvailability cdc-s

Dist Avai d ci or

v5389A

FIGURES

Page

K. FIGURE 1. Schematic of the apparatus to measure thevolatilization of TCE from soil .............. 11

2. Experimental apparatus ......................... 123. Experimental test cell ......................... 134. Hall furnace ... ..... *..... .. ... .. .. ... ... .... . .. 145. Initial experimental design for investigating

low temperature thermal removal of trichloro-ethylene from soils ........................ 16

6. Revised experimental design for investigatinglow temperature thermal removal of trichloro-ethylene from soils .......................... 25

7. Preparation of test cell ....................... 268. Typical curve produced during the execution of

9. Standard curve for 1 uL spike of TCE performedon 10 November 1984 .......................... 34

10. Time (minutes) for 90 percent of TCE removedat the operational temperature to be volati-

11. Residual percentages of TCE to total TCE volati-1lized ........................................ 38

12. Temperature versus time profile for soils at4 ambient moisture levels for 270C experimen-

ta i n . . . . .. . .... . . .. . . . . . . 3

-4-a13. Temperature versus time profile for soils atambient moisture levels for 900C experimen-

1.Temperature versus time profile for soils ata:4- ambient moisture levels for 1200C expe-

15. Operating temperature versus average time for90 percent Al at constant soil moisture levels 45

16. Average time (minutes) for 90 percent removal(of Al) of TCE at a fixed TCE concentration atspecified operating temperatures and soil

* moisture levels prior to ramping operatingtemperature to 175 0C .. 0 ...... .* ........ s. 48

17. Percent residual TCE at a fixed moisture levelcompared to soil type and operating temperature 50

18. Percent residual TCE at fixed TCE concentration(1 uL) compared to operating temperature and

a:' soil moisture level independent of soil

19. Estimated residual concentrations (ppm) priorP to ramping process ........... ............... . 54

vi15389A

1. EXECUTIVE SUMMARY

1.1 Volatilization kinetics. Volatilization kinetics, asrelated to this study, is defined as the rate of volatilizationof TCE from the three selected soils. Volatilization kineticswas dependent upon a subset of experimental factors tested.Statistically significant factors were:

(a) Operating temperature.(b) Soil moisture (dominant factor).(c) Interaction between (a) and (b). -

Relationships indicated by the experimental data were: F

(a) Kinetics proportional to operating temperature.(b) Kinetics proportional to soil moisture.(c) Kinetics proportional to a combination of temperature

and moisture.

1.2 TCE residuals. A residual did exist under most of theexperimental conditions of this study. Evaluation of residualswas made on the basis of residual proportion to total TCE vola-tilized so that direct comparison of data could be made withoutreference to specific soil properties (e.g., sample weight andsoil density). Statistically significant factors were:

(a) Operating temperature (greatest stripping of TCE at op-erating temperature of 1200C).

(b) Soil type.(c) Soil moisture.(d) Interaction between temperature and soil type.(e) Interaction between temperature and soil moisture.

Operating temperature was the dominant factor.. -

Relationships indicated by statistical analysis of the datawere:

(a) Residual fraction inversely proportional to tempera-ture.

(b) Residual fraction inversely proportional to soil mois-ture. prmtr hc lcdtetresisi h

(c) Residual fraction inversely proportional to one or more

following order of increasing residual, other condi-tions remaining constant: Letterkenny, Pennsylvania>Twin Cities, Minnesota >Sharpe, California.

538 9A

(d) Residual fraction inversely proportional to an unde-fined interaction between soil type and operatingtemperature.

(e) Residual fraction inversely proportional to an unde-'~ 1 fined interaction between soil moisture and operating

temperature.

TCE concentration was not statistically significant to the be-havior of residual fractions.

1.3 Engineering implications. The obvious and consistentpatterns in temperature dependency of both kinetics and resid-uals implies that significant residuals should remain from at-tempts to air strip TCE from soils at ambient temperatures andtreatment times less than 20 minutes. The kinetics factor mustbe taken into account. Pilot studies may reveal that ambienttemperatures merely prolong the stripping process and the endresult is the same regardless of temperature. This has to beverified. Data from this study imply that ambient strippingwill leave an undesirable residual.

Temperatures above ambient are requirkd to improve strippingkinetics and to minimize residuals. The operating temperatureshould reach 110 to 1200 C for complete stripping of TCE. Co-volatilization of water vapor with the TCE may assist in corn-pletely scrubbing the TCE residual. This has to be verified.

Machinery for stripping TCE from soil may be of two types.The first is a single chamber device operating at the ppropri-ate temperature throughout. The second is a two-stage devicehaving a low temperature first stage and a high temperaturesecond. A comparative evaluation is required to be able to rec-ommend which generic type may be most applicable.

Equipment most readily adaptable to pilot testing of ther-mally induced volatilization comes from the field of commercialdrying.

The protocol for pilot testing should be designed to verifythe conclusions of this study and to provide answers to engi-neering questions of optimization.

Machinery employed for pilot testing should have sufficientk . controls to permit parametric testing of both design and oper-ational parameters. Compromises will be required to avoid thenecessity of custom construction. Parameters include feed rate,TCE concentration, operating temperature, soil moisture, feedpreparation, agitation rate, and air recycle rate.

25389A

Z:. '.', * -. - * - -

Options for equipment include actual pilot-scale machineryand small-scale commercial machinery. A brief survey of theequipment market revealed that both options are conceptuallyviable. Actual selection will depend on availability, configur-ation, cost, and dependency on proprietary designs.

pN

35389A

. ~ ~ ~~~ .° ..

2. PURPOSE AND OBJECTIVES

2.1 Reference. The primary reference for this report is,"Test Plan for a Bench-Scale Investigation of Low Temperature -3Thermal Removal of TCE from Soil," prepared by Roy F. Weston,Inc. for the U.S. Army Toxic and Hazardous Materials Agency,January 1984. Full text is presented as Appendix A.

2.2 Purpose. Low temperature stripping of TCE from aqueoussolutions has been demonstrated to be an economical and practi-cal process. However, information concerning low temperatureremoval from a soil medium is limited. The purpose of this in-vestigation was to determine the factors that would affect re-moval efficiency.

This experimentation was the first component of a phaseddevelopmental scheme for promising soil decontamination technol-ogies. Results of this testing will be applied to pilot-scaleinvestigations for verification of the concept and for evalu-ation of engineering design and performance parameters.

2.3 Objectives. The primary objective of the investigationwas to decide if the concept of low temperature thermal removalof TCE merits pilot-scale testing.

. Secondary objectives included the following:

"- " (a) Identification of process sensitive parameters, includ-ing an analysis of sensitivity.

(b) Indications of optimum ranges of operational parame-ters.

(c) Indications of the type of pilot- and full-scale equip-ment that may be most applicable.

2.4 Criteria for positive test of concept. A positive testof concept is volatilization (stripping) of TCE from soils tonondetectable levels of the instrumentation employed for thisstudy.

4-4

5389A

• ".". "' ,? ? '/ "-".. '.- .' . '. <-. . . ' " i. - . -. -.* *-. . * . .: .... ? . " * . ..

Sn J w-t u. L y .. . -. : ,: L , v -r'r r w - r : -; V 1. ry .- . .. --- , . -. -, -. - . l

This criterion is based on anticipated requirements forvolume production through commercial equipment. The availableequipment itself demands an operating temperature in the rangeof ambient to 4000F. Economical throughput requires a soilresidence time measured in minutes. The absence of a univer-sally acceptable residual level of TCE in soils, and the ex-ploratory testing for potential engineering performance (massbalance on TCE around the test system) require that all detect-able levels of TCE be removed for positive test of concept.

55389A

_T 3. PARAMETERS AND TEST CONDITIONS

3.1 Test parameters.

* 3.1.1 Definitions.

(a) Parameter. This is a measurable property or character-istic that may be quantified as part of bench-scaletesting. The word is also used for constants, coeffi-cients, and exponents that describe statistical popu-

* lations. Both uses of the word will be applied inthis test plan.

(b) Experimental variable. This is a parameter that is un-der investigation. Experimental variables are eitherdependent or independent. The latter are controlled,and their values are predetermined. The former areuncontrolled, and their values are monitored andmeasured.

3.1.2 Relevant test parameters. Thirty-seven parameterswere identified as having relevance to this study. Of those, 18were soil characteristics that were fixed by the source and typeof soils to be used. The remainder were associated with eitherexperimental or full-scale operation of potentially adaptablecommercial equipment. Each of the parameters was considered tobe a candidate experimental variable. An analysis of each wasconducted, and the results are shown in Table 1.

All of the parameters in the list were either controllableor not. Some had relevance to bench testing, and all had rele-vance to full-scale operations. All were measurable. If used asan experimental variable, some were judged to be dependent inthat they were judged to be products of a commercial operation.others were judged to be independent because they were eitherfixed by materials to be used or could be controlled independ-ently of one another.

Experimental variables were identified as parameters meetingthe following set of criteria:

(a) Controllable.(b) Relevant to bench-scale investigations.(c) Measurable. x(d) Not fixed by preservation method, source of material,

or condition of supply, e.g., TCE-free purge gas sup-plied in that condlition in cylinders.

65389A

4,4

a 14 40

40

-c 0

0' z~

U) C

4' 0

C4> CO0

1

l's) -

C c

'D 0 Nrz~W. 1~

4'1z1 1 44

0'00

410

E- VC .

woo OS.'300

0 -

au

c' C C0 0

* 0a-4

04 41-

o, II

I CC RC

*4 u ' 0

0~.4C-

0w4V

0 0 0o

0.0 0

40I C

01 CCI

0 U) 0 V V 'Ara C4 41 4' 4) C 0

-4 .C 0 0A 0.0 v 0 0 18o c4

t0 c 0 0 41 A

u z >.

40

M 1" 9600

w v4:0w.413

41w

C 0 m

ol x

VI 0 41 0 0 4 '0 416 )41 2 41 .C 1 1 41 >I0 w

0 -44 41 0 1 1

H 00 CW ,0410 0 4100 0 iwtrI 4

04 4 I i0 11 I; 04 1.,014 04 014 0 0 a 44t

06 41 w 41 141-4 41 A1 41i aN 491 40 44114 4 400 41 41 14 0 0 44 41

gj w00 w. -... M. 3. .. >.- 340.

0'00 0 0

0

4.-I.

* -

Independent experimental variables for consideration in thisstudy were the following:

(a) Moisture content, soil, initial.(b) Temperature, oven.(c) TCE concentration, soil, initial.(d) Flow rate, purge gas.(e) Soil characteristics (18 characteristics).

Dependent experimental variables were the following:

(a) TCE concentration, purge gas, outflow (measured), con- '1

tinuous.(b) TCE concentration, soil, interim (computed).(c) TCE concentration, soil, final (measured).

., Experimental parameters for monitoring were the following:

(a) Temperature, soil sample, initial.*(b) Temperature, soil sample, interim.* (c) Temperature, soil sample, final.

(d) Temperature, purge gas, inflow.(e) TCE concentration, purge gas, inflow.(f) Weight, soil sample, dry.(g) Time.

3.2 Test apparatus.

3.2.1 Background. Thought was given to the types of com-mercial-scale equipment that might be used for volatilization ofTCE from soils. A brief review of chemical and metallurgicalprocessing equipment resulted in industrial dryers being the

.* prime candidate for full-scale operations. A brief survey ofdryer manufacturers and vendors was completed. It was evidentfrom discussions with vendors that a bench-scale study using a -

general apparatus could not simulate the engineering performanceof a full-scale unit. Therefore, the thrust of the bench studywas proof of concept.

Considerations for selection, design, configuration, andoperation of bench-scale apparatus included the following:

(a) A soil sample size in the 1 to 10 gram range.lei (b) A flow-through purge gas.

(C) Temperature control to within 1-degree Celsius.(d) Analytical equipment to determine quantities of TCE.

5389A

" -* -4 ***.* .''' . . ' ' " " " " "" ''" ' ' '" " " '' ' " . .. **.'" '.*" " . --" -. 4. '. " ,

3.2.2 Test apparatus. Based on the considerationsdiscussed above, it was concluded that all of the requirementsfor control, sample size, and detection could be satisfied if

%7 the tests were run using the assembly shown in Figure 1.

Figures 2, 3, and 4 are detailed photographs of variouscomponents of the test apparatus. Major subsystems of a Hewlett-Packard Model 5880A gas chromatograph (GC) were adapted for useas a simulation system for through-circulation volatilizationof TCE from a variety of soils.

The GC oven (Figure 1, item 9) provided a controlled, iso-thermal environment. It also provided ramped temperatures forTCE residuals volatilization, and for drying soil samples atthe end of a run.

Purge gas was dry helium (Figure 1, item 3) which will notinterfere with operation of the Hall cell. In a commercial dry-

N er, it would be most economical to operate with ambient air thatwould have a variable relative humidity. A dry purge gas maxi-mizes the rate of water evaporation, which may cause the slowestrate of TCE volatilization.

Four to six gram samples of spiked soil were held in glass-wool stoppered tubing (Figure 1, item 6) at oven temperature.The sample tube was fitted on either end with swageloc fittingsfor mating to the purge gas tubing. Inside the sample tube, andimbedded in the soil sample, was a thermocouple (Figure 1, item5) for recording the temperature of the soil sample during theexperiment.

Off-gas carrying TCE and moisture was valved (Figure 1, item10) to a Hall furnace (Figure 1, item 13) and a Hall detector(Figure 1, item 14) for quantification of TCE. The heated valveassembly (Figure 1, items 10, 11, and 12) provided a series ofshort duration samples. The series of TCE peaks expected to begenerated during a stripping simulation run provided a TCE massevolution rate (i.e., flux). Integration of the area under theset of curves quantified the total mass of TCE evolved.

Temperatures in the oven and of the soil sample were moni-tored by thermocouples connected to the GC's real-time clockwith printer.

* ~. 10

5389A

~~~*- T- I I F- W- LZ~ -7 v-. fr~ N'T F 7- 77U-7,~TJ* WL T, '[7 TR -7 w~ .JJ R F~' -%. F

00

C.,.L

cm (A

.9-.

CL( U.(

0 w : f. Id 4C* ~ *~~ OA4

:An ~

Forced airoven

(H.P. 5880 GO oven)

Hall conductivity -.

-~ cell

Hall* detector

printer

Figure 2. Experimental appartus.

12

Test cellconnected in

*:. the GCoven4bL

Test cel

Figre3. xprimntlTest cell .

M1

Thrmcopl

*-ri! RO 1. 74.* * *.

6-port,3-cross valve

Figure 4. Hall furnace.

14

a3.3 Initial experimental design. A full factorial designwith three replicates per test case was originally planned prior

* to shakedown and operation of the test apparatus. The number ofruns was:

(2 levels of soil moisture) x (3 drying temperatures) x (3TCE concentrations) x (3 soil types) x (3 replicates) =162runs.

Identification of each run is provided in Figure 5. Each run*is identified by a four-digit number. Each digit denotes one of

four experimental variables (factors), and the value of thedigit denotes the level (value) of that factor.

The values shown in Figure 5 for each level of the experi-mental factors were tentative and were based on best estimates

K' at the time of the initial test plan preparation. The final ex-perimental design was determined during the shakedown phase ofthe project.

The selected sites for sources of test soils were as follows:

(a) Location 1: Sharpe Army Depot, California (SH).(b) Location 2: Twin Cities Army Ammunition Plant, Minne-

sota (TC).(c) Location 3: Letterkenny Army Depot, Pennsylvania (LK).

These selections were made based on known contamination on-site and anticipated differences of the soil structure at eachsite. Uncontaminated soils were collected for use in thesetests.

5389A

I Co CEE

r C . X

r 0- IV

9. IMZ - I.. C

-C

Its. ~ 6 c

z A 2oc C..g

.11 't E4 *--

C.) L

U 0 2 0 CL c_ CLl

ela ri,>O aI-U

1I~c -

EU -ouJ C CY C c>~~~~ m )C

m NY N

c .) Cy C. ) C .) c . C .) C . C .) C

55~C '.. NN N C. C',-o - N N1

S.,

4 -CS) C. C4) C.) C ) C.y en C)E4 *c.. -w Ny N

cmE cN ~ ~ N ~ nN> - NN4 N

* C.) C.) .) C. C.) C.) C) C.

N N ~ N N N C. N - N N .)MN' NA

.

C.) - C.) C) C.

L -d ~ a -0 - -d L C.)l a odCc.)lLOd wf

P" M-O o os O is 1

110) C.) L. C.)9 110 ) C.) a .)106 C ) C.)G

-01AUO0 N0 N ~ BIU O N0 U0N149NIN0N N C301

-4 ~,16

3.4 Shakedown protocol.

*3.4.1 Objectives. The shakedown protocol consisted of aM series of experiments performed to acquire data which wouldidentify operational shortcomings of the experimental apparatusat test conditions specified in the original test plan. Listedbelow were the objectives of the shakedown protocol:

(a) Verify soil handling procedures which included moistureand solvent addition.

(b) Verify integrated experimental apparatus operability atexperimental conditions.

(c) Finalize all experimental operating parameters.(d) Revise initial experimental design to reflect results

of shakedown.

*-3.4.2 Execution. Table 2 summarizes the experiments per-formed during the shakedown protocol. It is important to notethat the protocol was continually revised as new informationwas generated.

3.4.3 Conclusions. Following are the major conclusions ofthe shakedown procedure:

(a) The moisture contents originally specified resulted ina sludgelike mud which would result in poor operation

o f the test apparatus.(b) Soils must not be air-dried before moisture is added.(c) Moisture should be added to soil at ambient moisture

content and mixed thoroughly by hand to avoid unevendistribution.

(d) Experimental apparatus successfully detected and quan-tified TCE when the solvent was spiked into an emptyP test cell.

(e) Purge gas flow-rate was determined to be 20 cubic cen-timeters per minute.

(f) The third experimental temperature was determined to be1200C.

(g) Recovery of TCE was best quantified when solvent wasdirectly spiked into a test cell f illed with soil

* and held overnight.(h) TCE spike volumes were limited by the spiking apparatus

to no less than 0.5 uL.* **(i) TCE recovery and spike reproducability were best ef-

fected when solvent was directly spiked into a testcell and held for at least 16 hours before testing.

(j) Soil moisture adversely affected the detection of TCEby the Hall cell at specified experimental condi-tions. This resulted in TCE recoveries of over 100percent (based on the Hall cell readings). This re-sulted in the modification of the test plan to in-clude dried soils ranging from 0.5 to 1 percent mois-ture by weight.

-* 175389A

.0 004 :3 a L

-. I-I4L 01 -I V )I) ~ ~ 3

w10I44 uI)0 to4000I u4

4)~~ WILO 04)4 V.3 0.4 a 10.40

.1144) 0 W.40 w0)).L .L.In)

0 0-I04 0 m) c 14 0))4 - 4

0 EL .0 3 w I4) - N 0.4-. 0 0 o)0 0 m >4) .11 w)L ") LI..e .ILIIL

o C e.X 0 w).I.O LL li 0 c V 01m4)~ ~~~~~ 4)')O 41 .0)004L 000 w I

- 0 L~C. 0 c I4L0) rI c:I0- ~ ~ ~ ~ ~ 4 "44.4 w)-(0 CCL 0.IOIL4

4)) 0.0..U4 (j1 .eJ4 0)0U))I) I 0 zII(0 4L Oc 04 00 "~01 m)L 0)

1'D m 0 WU'4 C.0000 ) DI e

.)1) QL 2 -3 )4 .00 0 )t14

0 0 00440)~~ 44

.14) 4

.4. co I In 3t0 M1

-- 4E-4 ' ~ 44401L

r aI 0'0. 00.a0 04)L4 m e

LI, aIl~ * -) 4)

4) 0.1L.4 n4

a) 40L I00 - ) L

074j 0 0)L11144

0L

0 0) -.1 W 4))

r4 0 I) 0 4 L~ LI4)ILI

4) 0.OaL4 41 eX ) 4 0.m

01- >U). 'A 0) a)

.. 0J 4 41 014)0LI w'

0- 0 oVgoI aV

) 110 a 12114 20 b4V 0 0-0

V1v0 aLI: . 4) * C a4IE

4) W)4 '0d 0 .e )0L z ) c S .3t 0 0 4) c .I39 " 0 -0 '000 .

x X C. M.L4 doL .14 4) -. 14 I

C2 0 to4-L 4) 30 0,1 E11 0t0oLI .44

LI )1.4)1 LI 0 4)I2II40EL 0 UU03X

In. )0. w4 a. 4)30 *Oc 010 U). I.

c . 4)0 '2 dl0N 010ILL 04 4

E-4 0 j~ C6I 0 4)n c) VLO ) ) 0 4 3 L .

01 0'0 to 4) 3 00 01 0 j0I0 C V 3.4 00 V 4.I.

0: 01 4)IL4, 2104 4. 1 0 L. >L4 )

H -.11~~~~~~~ =..4LO . .~Q *0 . 400-~) .~4.JI 4L

4)d3-0 . 4) > X)40000)4. .))0-0

L 0.0)))0 L 1 d-4 3 0 - w0)V) 0a 0

0 o.24 LII 1 2a .Z IO ! 4)4)00 C..4L )I0 ')

4) 0 0 ...... x . ) ) 0.0 L 4 0 -0 3 O 4 0l :5 xCL. E0 0 - *.. w CL 00 30IM0..0X 0 0 m m m .0 l

0LL00 04 0.I0L-.4 4

O in0)044L

~~4) 0 '00 0 0)'-

0 :5CL A ) LIw

4) 04041- )I W

0~~~ 0 I04w.40 0I1*- 0 %.

04) a 4 0

>) 04) 0 O44 L 0LI4

4) 0 0 0 0L 0 00' 0 0

0 04 1 04 a1.. 0

0. 04)3;-

.0 4 0 0 LI1

a ~ ~ ~ 4 cz 414 L. c.0404

9.., 0 04)~4-4.03 0-02I01 0-04

41-

18

-~ ~~ 'A -*1

40 0 , 0

440 414m1 040 41o

m W -n ul 45 z1 W 4a4) W. M 413 W 41 z 4 4

4l.44 4 1 404

(a U4 ~ 44

to C3 .04 m 00* 44I44

-6 bc u114. 00' 4 1W41 1"1>04

m1 MW1V 00- U' 41 4,4.404 11 4

0) 041 40 .0 C1. 0441 4111Uc43

41 4 1 41 to1 4 4

m a .IW a14 11 a 0

0 Ai.V V30 VWOW

w1 0 c 41 c M0a004cC 0 0. 01 014 3 Mt cw0

W O~ 41- 411 o 41

414 A1. *V 5to4 41 -14 4 04 . Iz w. 1 w 00 ) 02041 414 ) 0

41 2 41 2 040 WOV 0j.ck j41 c QI j ( 0 0.WIA1 00 3110 04 044 c5 0.

0 V44 011. 341 41- " 1 41 c0.. 44 WV > () .a, I 0V- 41

0 6 wA A w 0 414 41 0.4 41067 4144 u4 140 41 4 .0Aj C

a 041 41 o.41 oc40 M04 *~>4 c 0Q 4 1 4

W14 44141 444 41 E41 4 A, 41

040 w1 w0 1. 141 0. I' 0 41 00 4

0 41 41 -'

cu 'a 0 v * 0 . 0 41

41 0 a m-c - 0V

41 41W41) 414141 .00

rz~~ ~ 00 *WO 4u4304 0.14 w'. 0 040 4N 3 4

c >. .3 " 004 41 -1144 V41 411oa0 -4 3 " 3t A 414-.

041 ~ ~ 1 v14 0 04 41V144 414141041(A la 1 4

'A V0 .1 1 4 0 1 - 41) I 4141 0 .. 04 0 '

0411 41141441 014 tr1.4 W 044 weW 100.W 414 4 414 & 401 0.4

* ~0 0O 41 W 0. ~011 49 341 go 4 40 4uz 04 .- 041 VA c.4 -o0)141 0011 144

41 -001 40 4114E 1 41..I 41A 40j~~~~0 A,1144 41410J114 41 m 4140 0 1 V 0 41 0

W1 4140 . 41.A .0V ' Q1 41M4 01 41 I41 u 11

0 W 41040 W m......4 &.c 41 . .v1"v ow 04114 . I i1VV0. I1 .44 .4-4014 4 0 1W Ai 4 u 1. 000 41 'A 41

in11401 w30.0'4V4 1 4

.41 W 41410 4041.I >14to a.41. 014.I41114 0 41..01 A1144 w1 041.041~ ~ c1 411 4 1-~~ 04 0-0 4. 0 fV. 4.

0 441I *V .41 441 >, dca 4114 .W 4

o 0 'A4001

C-4 41.2V .0.4W10 ..

a 414

0 a 4 W4 o 41 -. .4j~0 0 41 1

41 .4 41 VW 0 0a N

0 - : - 041 0 01 410 - 04 41M1 1

4* W- -0 CW U2A 0 0014 a o 4 4 0.41 WV xV W 0

-1 41Aj 0 4-0,1 u0 441 04 41 M-V 4

a 04U 4aa 0 m0 W 4W 1 4 1 4E-4 OW 00 V4 41 41 gM0

4 a V10 0. "40 444 I44 Itr 2-; 0 0 0 w; o 0.0 0.>0 0 1M 041

'a x4 - 0 a0 41 41 ' 0 0 * 4W 1 'C 41 0 w0o0i40 30 04 041 3.1 0441 0

0,

a. 0- 0> 41t.- 41 4 0 410 >

>4 .40 -c 1041 0. .=. 0 I 410 4101

I 41. 410 04m 04 04 041

W1 40441 41 .. 4W 1.10IWW V a .0 41 41 4W 410a 440 1

'o 'W4 Z' I 4 41 'A44 a.44 .44

VV- -4 410 a1W o 0 AWU 41W

.10 c4 004 0414 a44 04141 U

4111I4,-1 .4 Z14 4114 41411.0IO M0 al 411401c

V041 '00 '0 41

lw19

W* -4a04 4))

~ 4) .~ 42) M(D .0 W.4 >wU)0

W . '00 0420 2 4)4) 04 40)0 m 2

4) 4) a.0 42 00.404)

'W 0.4 w)4) 03.4 w4C V4 a,2

0 4)4 0 0.44)) C W .4)(aa 'A0 010). C0 M )04 C*4 3K Mv I0>'0 4m42 z)"' 4 4 04)00. . 0.

>r (424 4).. 0'4 4 0 42440

0D IV 0 12 w 0 4)42E 4.

m )0" .4 4 )-0~ -0.0 0.-w C -4 0 v )44'4 0f '

.40 :1. 4) 4a'-) 0.00w4 .4 2 -)'M 4 * 00""fM20) 04) w Iz f W 4 ).-> a )0

XI.40~~~ 042 404".C42 ))

') 0 4) 0 44)~t 0>4 4

0) 4 4) 4 =4'

C 0 CL4 m> 42

E-4 0. v2 ' 24) 4) .")44 0.4

0 4 ? 4) 0

0 kw0 04)>4) W0 4)

- 4 004 04) 2z40 4) 4)404 'm4)lIw

02 0, A9)4'. zC 10 I i23 0 4 %)4) 4) M49 ct )0ow M- 0'" 0" >C0 . "

z . t a= 0.0 770'04 -4A42a ' w 4 o 0 -> 0) ,

0 - a0 j V c 4 4 1) 0 . o4 42.4

0 4244. a- ; n V4W)C J= t waA a4 w e1 0 0 CL 0

.0'' C C 0'0.w j.04 M "0 4) -I - .o r4 I 0 0) N

O-4.4. 0.4 U - w I'l .)4 '04 "K -j96Cw440 4242 003 w4)4 a)0 '04 CCCaa 2 -0a0 0

04)4 4)4A.'4'040.44)> w200 at 'A-)!. w)

".m 0 4))fl 4 4 Q z )4 4) 0 c IJ .4 4)44 =.') 0 M 0 0

04)mu" A 04 aU.a4oE # C 4CL 0 42 w Q042 0,0 mM wt0-CC4)1 0 - :0 ' 0G 4). 0'n .42X0 C 0to .O0O 4 . 44A .4 O)4.'s cam 0 K C' *UI)0)4 0.4)

'1'.~~~~ ~~~~~~ >)0.4 4)* 4'04U 00 )0)2-

K~ ~~~~~~~~~~ m 4-44- w N0 0)- 0'4w4)4. f4

0 4)0 w0 ow caE4 M440 44~0.)C

V 0 .4-4to 0 3U434*p ~ 4 x. 40I4 w )

'M 4440" 04) 0>) 0)~.4 )w )3 4444 4 20 0034) 414 0424 N. K 4)' 4 '3 -'0

o ))4. 4).4444 w02)44 w4'.'- 4j- 4 4 - '.00. 4)002 '4).44 Co .41 w 4 0. -4 "0).4 In.

.0 CC 0c44 J)-

00 N w 4a0:. ce 0.n c a t

- .0)0 4 4).0&C204

>4).J > .a W UWoz '0 4J

(1 11 C -)

C'UC~ 0U.' oC4) ) 0)

M U4 G .M3 0m - -4C. 'U a4.~0f W41M 0U1.C4CJ -- UV0 W V 4 C64)

'.4 M j 0'l 4)cOMC4)' M V .C- 2, Ij0m

.4 0~)I a W .C'UbAc 41- c4 Oc' q. W). M c

0 0UC ' - 'U a4) 4.- 0 u 0 0 0-'4wAj 'U 1U 0-4 . 0 )- 1 ) 0 02 u U m I4 Z '

m. )4 M m .. a 030 w 41 41.w 0 4)W4 M4)Asa ~ ~ .c4- 0*4 Mt o aG - w' w2 O I MC

lo .. a 0

4) I-a j309 v4. ) 0A

.4w 'UjV >

0 a4M ' . I. - . I4' .On 0 cU . C)1 0-- 0Q 4-41X -000a t

ck 0J4)0 4 1. A.U44 4)W4) 9 J

0' 0 4 0wI 0c 4 0UM4"4 44 010 *-.

VU'4 24'AV U C --

v4 cW4' W ' W 0)). 0~.'06 0))4 'A 0 .4 )I .

0 0O 10. 0.Um 4 a & 4 4) m4)l4 toU c c 0 a 0 0 w " u0 'UJ 40 .'UWo)U> )OM - wz 414M

Q~~ ~ 'U .UMU>).O M 0 16v 4)-4)'U..'UU14IJL U4)M4 m

0 ''4 , 4) 'u 'A U

'r4m c~ Ij 'aV - 'UV lau 04.40w4)0 'A .4 AI A .CU

m E .'j m.4 4d0 U0 r j 4)4 0 ).

0 .4 0. 0 M 1 0)'4 0"4 MOU' ".

.0W0 ' 0 'U 04 0.44. a-4 w112 .40 4.a.- 0.c-E.'~ a4U * 0 0 a a 0 W'0 0J

4'm m" -=.4 0.U...U)r a4 0 AJ 411X 4.(A

m c .0 Q)4 4) z 2 9 4 4? w.4 .f.C -41 Ul- -A ' m 40 0C1)M) aQ.r_ Ia ,0 C1 cI 04 W'U m

.4o 0 CM 0 44 w M' 41'U 01 '~g~42)~

j 0. 03u4) lm444> 6 CC*.- 0 a v) 0 0 414mW

2 v.m c '.4 u4 .4 'w 0 0'eJ O *0M44 '.44 a 'U 0 0V1.M 0I 04) .4 4)UC: v DI3 a M

-IJ. 1 V 4 'U M a .%.2 0 E.* (. 'U W u1 M U 0 ' .'bg a .l C0 4) MIA 9 w0C 1 04'U'4 11 r.. 31O u 0 a,0 c L0'. 4 U 0 'Uw0 4 A'-'Ucflm 0 .4 44 m c

c u34 c00V ) j 2o 4 0,0 M4) w 4 9 - w _

o 4-.w ?10"zw4j 14)01 w 41 C

-40M4 'A 4) 4 aa ja

N 0'4 CU W 1> c0;A) 410- 4) 0'Uj4 0 a .- %4

m .w0 c Wo m4o4:Q x0 4.1 -. 0oD 0

0 0C.0 v.' c0 Av - w jA

de 44 )4.4j0,a -t 4 3 'U

0. >C 044'

4104)0 'W. 440 4U 0 0wW.

4 = c 21

-M~ to. VV)W0

-0 J- - 7 4 4)" 0- 4j 4) 0

0.20 (a 0 2 555 C) l 2 V U1 4) 4r 4, u4 " - C CSal

to -)c 0V) 3M cq to0 '- W4 to M u aa w "aM US M 0. 41 a c" 0V j C a'-

0)0go 4) 4u " 4)5 O'O 2 c V ) = S. Wto M5S0 S ) CEO 4C SW 4) C

C cW4JI wC' w wA C Go53 ~41 to44 4) CC 4

c 40 0 22b W9 . v4UwM 0 '4 "4. a 0 a2(4 C m4 W o40W 4)4 S 04 0 2 4) V40

42". -. S 4 C '2 C.. ) CMC S. .. ' V C0' -MO WS u o C 0 A C. W C U (

S M4) 02M5. u4 0.'4. "5C~~~~~ ~~~ Ca'')) 4C '4' ) 4

U2.C-50M C> C W0 w M4a. C V0024 *CMWC UM4)02w CW a '-' -c3a, 4) >)Q>VAa 'A2:MWSC" u to w-4 cc M ' 03 2 023"3a)a 25 4Jr4 0245 Z2C ' 1D2 C-OIW4)

W4)V4 W>J GCP a3S ) 20 cC

4, 'a 4. t00M300 00 S41 0

coccoa 0 Cc a 4u~) 04 5m 3 0

C C aC0NCW1 0; 0) U coU Ca'4) 0 0O ll at &A> M) %a - 04)

Ch C -0-4 02- 02 54li

41 . . .. SCC2 I w2 Cm 000 C )

4) 0000 0 04) 5

0 0 00 Z C 0. 42 0C~ ~ 'ACW0 4 12 )4W

E- N'0.444 c 02 Wj Wq CL coo. ,aC

a 0- 4) O

VWS 0 c0 NMJ9 4r 2 ,

4 )W4 4) v 4 0 '*' C C

04 2 24) - 0.0 0 0. c M w

Zj ff W Ov wf w at WC 02 W CU C 5 12U IW 00.' 4 4) 0 . z w 4

0 C0 B

0 M 0 V 4)0 4)0 - 3 -. J -'

x ~ ~ ~ MU 4) A 0 'A oCE CE- 3.W Z OW I )0. 0' 00 4) me' 04).,w)O Z

ale 4) I 0 O4c4 'CW MM "0.0 a5 4 MM 0CW S)'A 4

0z 41 U4) S6W * WCC WV 0 C 4 w 0C 'G0 E5 -

" W W0 2C ) Mj 020 M UC1)U M UC C 02w5 40w)0

*G ( .. MuS 4)- 04 5o C4 S0M aV VS4) C

4, r.,I 01 .4SS WS " HUE CO SMM))4 M44 O"4c"' 4) cM 4) V E 4) C 024) 0) 024 aC -. 4 C0 0

L. AJ Z4~a 4) a1 MM.' MW .46 C WSEMC.0 2' Mv44 a 4 c 0044)0000 a CS a).0) CS M0 r.)2

- I 0& w 4) SC -4 505 -4 '1Cw WS 4CCWc -M )to u 4 S 0W M CW 3J WO0' - -0t 0c '4. i

C CCIIW -0 lW' V4*54o00 A4 020 0 -- 4w Q w C i Q Mm w0UJ C >24 CA 54 j 4

'A 0 000=0. 4) U.U OWO C M r')0~~~~~ 0)540 m50- 0.)44' 02' 4" 0 0-U a)'M~ ~~~ N) 04 I" 1W 9 .W .4 ) 0 W4 4 01

-1 0254 0 0 4 t 4C '2) )00 USO w40. ESOMOOAMCJ

02 . '2CM '0%4) -VS r. '.V 'C0 4.'44).U W4) 0k~0 02MWa, C 020M aSS C

02.4~~~~ W CC c)4 4 C2'4 V)2'S4

4) WS MS

w) 4 0 0 4) 40

W W toU'. 'AI

02 0

to .0 45

44 W~'- C 0

CS 2 C2 .

a 1 - l

4)'.)22

c C 41 03c 4,C41 0 X1 -C .(D o

0 m1 0104 w1 V x 'j U4 -41 014 M 0 0 04M4J 04

Wj0 o 0 00 41 0. 4.-0 0.4(

ui 2)1 - 141 W '. 04 4 141 4 414 .41 4 'v 4 11041 (4- D4 4. .000 41 40 03 >c to-4 41'w

3 '..14 4..4>.4) 1 2 0 =-414 1 404 1 Wc 010 10 m, D.C 00 41

c 414 ' XI C1 0 m1( W.C A . 4 1 414 m 41w j l 3 04

0 *04 onw 4 04111411. 0 0 4.'0 .. Ai C

wU ri 4 aU).(a0 441 M W 001 C a-C Ai 2 a aW c 'i a00 0WC 41-w 0

U )C.0 4) V4 411 0-.41.1a- -m.0 -5C O04 W a Q, 3

0 > 4 W 0 w*.. 040. U1 114 0 0W41V 1 410*0 414'4~I04-

0) Ai.41=1c 0 V 414. -- 'aW W 4041

4141a C 0 w 10C Ca 1 44141 A a0- 41 .W41

01 410 WSO 410 41 411 00'W 4

Uto 0 V 11 .4'~41 4 .0 a4 0-. W

4O 4141 0 00 6. '40

c ~ 0. w0414 04

0 0 a 0 i 11C41 W 41.a 4 t.. 0 Q, U

0 4 0SC 02 fA -a W -"-41w 41C"C 4) 2 010 -o4 "40 9 1 W04~O14.

(. N) C 0W- =C.= C 41 0 U 00-1414 a-) 41 4 10-444 U~

4.4 0 a1 C41 C * wCM334v0 W0. bi CO. 41C c .L OC

Ai 414 arJV 41 v N~ ..4.4 44Cl 42 pM 14 40 . 0i w J a 1 4 U0 3j 0 ww140 Sc 4 W 0 .4 10

0- 41 4 0 a .0 91 0 41c04'

E-44'al1'40 W C1W4-4-40'dia M C 04140 41 a .. UC.

0 '4..0 w C14 w 4 '44 414O.OmC n 00'4 0 0 " 14V 0

41 , 0'4 410.08 coo Nw411 09 11 4 s r.0 81-4414~ W 1~1

WW4 wCCI.41 0 -4 0.*44-

0 00 1 &C 4 L. 41 c

3. w $ . W uI44' 1 i0 a 00 W G 40 1 C

C 33. 50"M4t F0 0- 2.4 0. m 4 l

.4 5- cc0415-a 0 4w CW j4 .4J3W01 .04 Id C'4 0 c a 6

0. 4 9 0.4ww >13'a4

0 0 a > . W 0AJW41 A". a D Aiw0 0 G0 >OW40 v4 0(4 OWa mvia .,d 01 00

W 1. 4 C 41C 3 I.2 %a-.3 -0 A .M0t

WV 40

AO 41A

0 10

23

3.5 Experimental design. Information learned during the _q_

shakedown phase was applied against the original experimentaldesign for 162 runs. It was evident that a modified test planwas appropriate. A two-phase design using the same factorialconcepts as the original design was formulated. Phase I was:

(1 level of soil moisture) x (3 operating temperatures) x(2 TCE concentrations) x (3 soil types) x (2 replicates)-36 runs

Phase II was:

(1 level of soil moisture) x (3 operating temperatures) x(1 TCE concentration) x (3 soil types) x (2 duplicates)-18 runs

A summary of the experimental design is illustrated in Fig--_ure 6.

3.6 Experimental procedure. The shakedown protocol pro-* duced data which directly affected the experimental procedure.

The following subsections outline the procedures used to performthis experiment.

3.6.1 Soil handling. Approximately 2.0 kilograms of soilcollected from the designated site at ambient moisture condi-tions was sieved through a 2-millimeter screen. The resultantsoil was mixed and divided into two approximately equal volumes. -

One volume was immediately stored in a 1-liter amber glass con-tainer. The remaining soil was dried overnight at 600C. At600C, soil is dried to less than 2 percent (weight percent-age) of the mixture while minimizing the potential for changingorganic properties. Moisture content was determined using stand-ard tests for both dried and ambient soils before execution ofthe experiment.

3.6.2 Preparation of test cells. Prefabricated test cellswere filled with soil of a specific type and moisture contentusing a metallic laboratory spatula as pictured in Figure 7. Thesoil was packed into the tube using a pipe cleaner.

Cells were spiked with either 1 uL or 5 uL of TCE via hypo-dermic needle injection directly into the test cell at roomtemperature. Once spiked, the cell was immediately sealed andstored at room temperature for at least 16 hours before testing.The spiking technique is pictured in Figure 7.

245389A

E0.

1~ 06_ 0

3l - o M

I In

2, --

00

r u- .- E

o2 0 0= a x

Em N 0

w.2 cm

0 C.

Co

-~ -2 x.

Nm N - c~

.2 EIaS 6 G 0 o a 0ualnue N0 N ~ ale0 331

LISC

- - 25

Filling test cell with soil.

Spiking test cell with TCE.

Figure 7. Preparation of test cell.

26

3.6.3 Experimental apparatus preparation. The experimentalapparatus required daily standardization to determine equipmentoperability. TCE spikes of 1 uL and 5 uL were directly injectedinto an empty test cell. The cell was immediately placed into

A' the GC oven where it was subjected to experimental conditionsat 270C for 15 minutes. The amount recovered was observed onthe Hall cell printout. This test was performed in duplicate.

The resultant Hall cell readings were used as the dailystandard to determine the recovery percentages during the exe-cution of the experiment. Typical standard results and theirapplication during data analysis are explained in Section 4.

In addition, several procedures were performed on an as-needed basis for maintenance purposes. These procedures includedreplacing the Hall reaction chamber, replacing tubing whichfeeds Hall reaction chamber, and replacing the n-propanol sol-vent used as part of the Hall detection cell.

3.6.4 Execution of experimental design. A prepared testcell was connected to the purge gas inlet and outlet lines in-side the GC oven. Once connnected, the test cell was subjectedto the specified experimental temperature for 20 minutes, fol-lowed by temperature ramping to 1750C. The ramping proceedsiuat a rate of 250C per minute. The system operates in theramping mode for 15 minutes.

Effluent purge gas from the test cell was automaticallysampled every 30 seconds via the 6-port, 3-cross valve. Twiceeach minute the valve opened for approximately 1 second tocollect effluent purge gas. The sample was directed to the Hallfurnace which operated at temperatures exceeding 8000 F. Theseconditions converted the chlorines of the TCE to hydrochloric :1.acid (HCI). The effluent of the Hall furnace was fed into a Hall

• " detector cell. This cell used the electrochemical properties ofHCl in conjunction with a standard solvent (in this case n-pro-panol) to measure the amount of HCI in the Hall cell. These

. readings were displayed on a strip chart.

Upon completion of a run, the test cell was removed fromthe GC oven. The cell was emptied of the test soil and cleanedfor re-use using a nylon brush and a water/detergent solution.Once cooled (if necessary) the test apparatus was ready for thenext experimental run.

275389A

* * **'

3.6.5 Temperature profile. At the completion of experimen-tation, additional data were needed which compared oven tempera-ture and test cell temperature versus time. These profiles weregenerated using the experimental apparatus. A test cell wasfilled with soil at its ambient moisture level. Oven temperaturewas programmed to operate at an experimental temperature for 20minutes, followed by temperature ramping to 175 0 C at a rateof 250C per minute. In addition, soil temperature within thecell was recorded every 4 minutes.

This process was repeated for all three soils at ambientmoisture levels at the specified experimental operating tem-per atures. 7-

3.6.6 TCE verification test. During experimentation, itwas observed that the Hall cell would detect what was thoughtto be TCE during the ramping process. It was decided that anadditional test should be performed to verify the presence ofresidual TCE in the effluent gas from the test cell during theramping process.

A test cell was filled with SH soil at the ambient moisturelevel and connected to the experimental apparatus. An experimentwas performed at 270C using the procedures discussed in Sub-section 3.6.4. At 20 minutes temperature was ramped and effluentgas was collected in a gas bag typically used during air sam-pling. The collected gas was then submitted for a GC/MS (massspectroscopy) scan.

3.6.7 Soil characterization. During shakedown and experi-mental procedures, physical properties of all test soils werecharacterized. Table 3 lists tests performed on all soils.

285389A

_-A

TABLE 3. CHARACTERIZATION PARAMETERS FOR EACH SOIL TYPE

Parameter Method

Soil pH ASA 60-3 (glass electrode pH meter)a

Total organic carbon ASA 90-2 (wet combustion)

Total exchangeable bases ASA 59-2 (residual carbonate method)(base saturation)

Sand Hydrometer

Silt and clay Hydrometer

Particle size distribution Standard sieve analysis

- Kaolinite ASA 49-4 (X-ray diffraction)

Illite ASA 49-4

Vermiculite ASA 49-4

Montmorillonite ASA 49-4

Chlorite ASA 49-4

Interstratified combina- ASA 49-4tions of 2:1 type

* components

Carbonate (CaCO3 ) ASA 91

Aluminum, total ASA 67-2

Cation exchange capacity ASA 57-3 (sodium saturation)

. Exchange acidity ASA 59-2 (residual carbonate)

aBlack, C.A. (Editor-in-chief), D.D. Evans, J.L. White, L.E.Ensminger, F.E. Clark, and R.C. Dinauer. Methods of Soil Analy-sis, Part 1: Physical and Mineralogical Properties, IncludingStatistics of Measurement and Sampling and Part 2: Chemical andMicrobiological Properties. Madison, Wisconsin; American Soci-ety of Agronomy, Inc., 1965.

295389A

- --,4. .

4. RESULTS

4.1 Typical data output.

4.1.1 Experiment. Figure 8 and Table 4 present character-istic output for an experiment. Figure 8 presents a Hall cellTCE trace during periodic sampling of sample tube off-gas. Thenumber at the top of each peak is the time in minutes at whichthe peak was measured. Table 4 presents individual and totalareas under the traces of Figure 8. Areas are proportional toTCE volatilized.

The experimental protocol described in Subsection 3.6 spe-cified that the test operating temperature was to be maintainedfor 20 minutes before the temperature was ramped at a rate of

d 250C per minute to 175 0C. This procedure often resulted intwo sets of peaks, as seen in Figure 8. A certain amount of TCEwas volatilized at experimental operating conditions, but a TCEresidual remained in the soil and required higher temperaturesto volatilize.

The total area under the peaks observed between Time 0 and20 minutes is Area 1 (Al). The total area under peaks observedduring temperature ramping (between 20 minutes and completionof the experiment) is Area 2 (A2). Total area for the experimentis the sum of Al and A2. For the experimental results presentedin Figure 8, Al was 8,114.0 area units and A2 was 3,744.1 areaunits. Total area was 11,858.1.

4.1.2 Standard. In order to determine estimations ofamounts of TCE volatilized, standard curves were producea on adaily basis. Standards were run for 1 uL and 5 uL volumes ofTCE.

-. °

30

3389A

. + I 0 .64 a

1. 14 "

1.64

2.15

1 ,3.15

33.66.4.16 Area (A)4.66'

* .,,

5.67

oupti rmexeiet21 TwnCte o" W C,20Ca777- 23 .. ,

7 e uo24.66

25 25.77 Area 2 (A2)J L . 2 6 . e 2 .9

-7+ 76 .9

a Time in minutes at which peak was measured.

Note: Temperature ramped to 1750C (250C perminute) after 20 minutes of this experimental run. ,t.

.,.-.. Figure 8. Typical curve produced during the execution of an experiment. This ." " output is from experiment 2112 (Twin Cities soil, 1,uL TCE, 270C at"7.5 percent moisture) performed on 10 November 1984. .

OR. . . .T .i -. .: -. i " : : ? - i+ - • - " . -. . . . . .

TABLE 4. TCE MEASURED BY HALL DETECTOR DURINGEXPERIMENT 2112.

RT AREA TYPE WIDTH HEIGHT BASELINE AREA %

0.60 BASELINE @ START RUN = -0.076.00 THRESHOLD @ START RUN = 2r.00 PEAK WIDTH @ START RUN = 0.040.13 7.26 8 2.19 -0.02 0.0610.64 1214.29 BV 6.06 * 336.10 0.91 10.2401.14 2844.75 8 0.06 * 579.86 7.75 17.2441.64 1869.94 BB 8.86 * 511.59 11.25 15.7622.15 1315.41 88 6.86 * 332.38 11.50 11.0932.65 811.94 PB 0.06 * 198.86 11.19 6.8473.15 419.28 BV 6.87 * 95.46 9.53 3.536*.66 292.97 BV 0.08 * 41.12 8.29 1.712

- 4.16 110.98 BV ----- * 19.19 6.92 0.936*1 4.66 68.22 BV ----- * 10.73 5.45 9.575

5.17 32.18 BY 4.61 4.66 0.2715.67 16.67 BB 2.69 4.20 0.68

,. 6. 17 1.60 8B 804 4.34 0.017.1

".," 23.16 4.43 88 1.21 0.59 0.03723.65 31.93 8B 6.93 0.33 0. 26924. 14 107. 34 8V ------* 19.38 1.13 0.905

-.. 24.66 66.66 BP * 11.80 1.72 0.56225.01 371.93 PV 91.71 0 77 3.13625.21 66.22 VYV 13.35 8.92 0.558.25.23 271.17 V' - 186.71 0.98 2. 28725.36 116.99 VV - 44.04 1.85 0.987'25.46 559.77 /7 97.43 1.11 4.72125.79 58.59 '/V 12.18 1.31 0.49425.77 269.49 V -- 117.67 1.36 2.2731-,17 t3 ....26.89 608.583 '7 *-- 37. 49 1.53 5.131 ]26.29 66.06 VV -6.34 1.69 8.55726.32 248.44 VV ------ 72.74 1.78 2.09526.48 312.69 VV 30.62 1.90 2. 63626.57 89.18 VY 24.36 1.97 8.752< 26.91 222.93 P7 .. * 54.61 3.58 1.89

27.17 26.74 BP 7.34 6.66 0.2252.: 17.39 14.14 PV 5.35 5.28 0.11927.52 203.01 /V7 73.44 4.96 1.71227.64 27.98 VP 6.07 4.69 0.236

TOTAL RREA = 11858.10

NOTE: TCE detection is the total area under the integratedcurve in Figure 8.

32'A.-.5389A

t,'..

The daily standard at 1 uL TCE is presented in Figure 9.Total standard area is 20,524 for that day. This indicated thatrecovery (following ramping to 1200 C) for experiment 2112 was11,858.1/20,524.3 x 100 = 57.8 percent, or 42.2 percent of theTCE was lost during spiking and storage of the test cell beforeexperimentation. Examples of other relevant calculations follow:

6 1.466 grams 1,000 mL "1 x 10 liter (1 uL) x x L = 1.47 mg

(TCE spike volume TCE density TCE spikek"J\at 680F / weight

TCE loss = 1.47 mg x 0.422= 0.62 mg lost

Total sample at time of sampling = 0.85 mg

Sample weight = 4.5 g

= 0.85 mg .Initial concentration = 0.85 mg

(dry weight basis) 0.0042 kg soil

202.4 mgkg

- 202.4 ppm

Al Area 0.85 mg TCE = Amount removed duringTtal Area first 20 minutes at

operating temperature

or

8,114 " 011,858.1 0.85 mg = 0.58 mg TCE removed

Residual TCE = 0.85 - 0.58= 0.27 mg residual TCE

(detected by ramping ,temperature)

Residual concentration = 0.27 mgC .

(dry weight basis) soil

(detected by ramping 64.3 mg TCEtemperature) kg soil

= 64.3 ppm

a, 33

5389A

..

1.13

1.63- , , z. L3.'.,

2.13

~3.64

4.14

4.64

5.14

5.64

6.14

6.64

7.15

7.63

9.14

RT ARER TYPE MIDTH HEIG14T BASELINE AREA %9.0* BASELINE 9 START RUN - -6.67".00 THRESHOLD 0 START RUN -2..66 PERK MIDTH 0 START RUN 0.846.63 2379.62 BY •.05 * 752.62 6.96 11.5891.t3 5936.87 OV *6.65 * 189".76 11.89 28.9261.63 4569.48 VYV 0.5 * 1372.17 11.71 22.2642.13 3689.16 VY 6.06 9 969.36 11.53 15.0512.64 1767.16 Vy 6.6 463.53 11.35 9.6163.13 1654.16 VY 6.86 * 256.78 11.19 5.1363.64 687.46 VY 8.07 162.21 11.46 3.3564.14 424.56 BV 6.67 • 92.96 16.56 2.6694. 64 265.05 PY .... 51.56 8.76 1.2915.14 152.91 By ----- * 29.36 7.50 0.7455.64 95.86 BV ----- * 17.95 6.29 0.4676 .14 3 6 .9 ; B e .. ..- -.- - 6 o ) * O6.4 3.97 By -- - 8.5 6.56 0.1086.64 35.89 Y .6.32 4.73 .?1757.15 16.65 Be 3.36 4.37 6.6817.63 7.73 a - 2.43 4.26 0.638a.14 5.66 as 1.95 3.69 6.629

TOTAL AREA -20524.30

Figure 9. Standard curve for 1 pL spike of TCE performedon 10 November 1984.

34

I, ) .

- -

4.2 Experimental results.

4.2.1 Experimental moisture levels. Moisture levels ofIthe soils at ambient and dried at 600C were determined usingstandard procedures. Table 5 lists those results.

4.2.2 TCE removal rate at operating temperature - kinetics.* This was the time (in minutes) when 90 percent of Al was ob-

served. This was calculated using the data described in Sub-section 4.1. These data are presented in Figure 10.

4.2.3 TCE residual percentage. This variable was calcu--. lated as (Area of A2/Total Area) x 100. These data are pre-

sented in Figure 11.

4.2.4 Temperature profiles. Figures 12, 13, and 14 illus-- trate temperature versus time profiles for soils at Level 2

moistures for 270C, 900C, and 1200C, respectively.

4.2.5 TCE verification test. Data generated from the exe-* cution of the test plan indicated that TCE volatilization oc-

curred at the operating temperature and during temperature ramp-* ing. TCE was verified as the material comprising A2 by GC/MS

and GC/IR analysis. Water vapor and carbon dioxide were alsoidentified, but neither contributed to the Hall cell trace.

- ~ 4.2.6 Soil characterization. Volatile organic scans were-, performed on all soils prior to TCE spiking to identify any

compound(s) which may cause interferences in the execution ofthe test plan. These results indicated that no volatile organics

* were present above detection limits (0.1 ug/g) in the background -

soils.

Following the initial volatile organic scans soils were* characterized per original test plan. Table 6 presents these

results.

355389A

TABLE 5. EXPERIMENTAL MOISTURE LEVELS FOR EACH SOIL TESTED

Soil Moisture Levels Used During ExperimentationLevel 1 Level 2

Dried at 600C AmbientSoil Type (% Moisture) (% Moisture)

Sharpe 0.3 4

Twin Cities 0.5 7

Letterkenny 1.0 15

-?

% .4

? i.

.:

36p,," 5389A

V V .

T0 T

0-. - ER

ccS nc

e t n,

N, i N i cn

U 3 -

co I

N N N

a C1.0W

.

t2 E 00-. N N

Nc

Go In

t2 No NP

N - m .- . - -I-.

2c. S, co N,2

(0 (037

C,

0I C.)

Ecy C.LJ

9 0

90

I IF

N x38

S..

-E

o a)

0 Cfl

E D20) N E.

*0. 0)-. 7 is-c 0

C CM

*0-

S0 Z

~4 x

Sm ce C

40 E.me C4

000

-4.

o 0 0 0000 0 o 0 0 0 0Go r, w in Iq -n CM - 0i1-L v

(3,) sinjejedws± C

39

*-'* ** * ~ pa

CL 4

jE4)

= CL

o) CV)

oc,,"0. -L r - . -

E C4 4)C

.4. 4 ) ~ E

0 - 0

00

c~cc6 b. C

~0

CM

*4e 0-

COS

400

E-E

c (D0V .. - -c E

*C E>v m

3 n. (DDc~0

.m 00)04

*A 6

*Ib.

110.*

CL

0 0-O. *-* (D L 1 C)0i ) C) 0 * ( C W

(3)oI.jew.

I.L

.m41

.mea. %

IV I ,

TABLE 6. SOIL CHARACTERIZATION

Sharpe Twin Cities Letterkenny

Concentration of mineralsa

Phosphorous (total) 81 72 4Potassium 200 50 60Magnesium 424 192 85Hydrogen (meq/l00 g) 0 0 0.5Calcium 1,020 1,560 1,920Sodium 260 35 47Aluminum (total) <1 <1 <1Percent calcium carbonate 1.79% 1.96% 6.07%Percent total organic carbon 0.06% 0.17% 0.15%Cation exchange capacity 10.3 9.7 11.2(meq/100 g)

Computed percent base saturation

Percent potassium 5.0 1.3 1.4Percent magnesium 34.4 16.5 6.3Percent calcium 49.6 80.6 86.U

. Percent hydrogen 0 0 4.5Percent sodium 11.0 1.6 1.8

Soil structure

Percent sand 76.4 70.4 24.4Percent silt 18.4 16.4 34.4Percent clay 5.2 13.2 41.2

Dominant minerals(Greatest % to Least %)

X-ray diffraction

Sharpe Quartz, feldspar, calcite, illiteTwin Cities Quartz, feldspar, calcite, illite

• Letterkenny Quartz, illite, feldspar

Dominant clay for all soils: illite

aConcentration in ppm unless otherwise noted.

425389A

4.3 Data analysis.

4.3.1 General. Data were generated in two phases. Thefirst was shakedown of the apparatus and investigation of be-havior of soils with and without TCE under experimental condi-tions. This phase required significantly more effort than orig-inally planned because of the nature of the soil/solvent system.Data and observations collected during shakedown were useful inredesigning the experimental program from what was originallyplanned to what was necessary.

Data generated in the second phase resulted from execution*of two factorially designed experiments which were subsets of

the original factorial design. Second phase experimental designsreflected observations from shakedown andl incorporated experi-mental procedures refined from the original test plan.

Preliminary data analysis was by means of multi-way Analysisof Variance (ANOVA). Results of the analysis were evaluated at

* the 5 percent level of significance. The purpose was to find,from second phase data, which of the experimental factors (or -

combination of factors) contributed most to variations in vola-* tilization kinetics and residual levels.

4.3.2 Volatilization kinetics. The operational effect ofinterest was volatilization of TCE at the operating temperature.Some amount of TCE was volatilized at each of the three testtemperatures. The amounts differed, and review of the chromato-grams indicated that the amount was generally finite. Volatili-zation was completed at the operating temperature by at most 20minutes. The measure of interest was the time required for 90percent of that amount of TCE to be volatilized.

Two experimental designs were used. The first maintained aconstant soil moisture level (Level 1) and varied the threefactors:

(a) Operational temperature.(b) TCE concentration.(c) Soil type.

ANOVA results are shown in Table 7. Operating temperature wasthe statistically significant contributor to variation. Therelationship between kinetics and operating temperature is

* shown in Figure 15.

a. i

435389A

TABLE 7. ANALYSIS OF VARIANCE FOR VOLATILIZATION KINETICS AT -

CONSTANT MOISTURE (MOISTURE LEVEL 1) CONDITIONS

1ijSource Degreesof Sum of of Mean Calculated

variation squares freedom square F

Replication 0.4334 1 0.4334 0.0072

A Operatingtemperature 1,460.435 2 730.2175 12.1703

>3.59*

B TCE concen- 2.816888 1 2.816888 0.0475

tration

C Soil type 81.8684 2 40.9342 0.6903

AB 1.8492 2 0.9246 0.0469

AC 118.9094 4 29.7262 0.4954

BC 265.3115 2 132.6558 2.2109 4ABC 694.396 4 173.5990 2.8933

Almostsignifi-cant

Error 1,019.9923 17 59.9995

Total 3,258.9376 35

F0 .0 5 (1, 17) = 4.45F0 .0 5 (2, 17) = 3.59F0 .0 5 (4, 17) = 2.96

*Significant at the 5 percent level.

445389A

---

- , , '. ' x . . ,. ,x , - :. . . . -, - • - " - ".. . ." ... - . . .....

*~~~~~~~~- --- -. . .. . . . . . . . . . . . . . . - v

.4'j

4C'

4co

0-

0

0 IMC)1

o) 0uiLO EO

0~~~~~ CO( t C1 ) c C C

CC\J

(Ulw) WV %06 JOI Owli

45

The second experiment maintained a constant TCE concentra-tion and varied the three factors:

(a) Operational temperature. t.'(b) Soil moisture.(c) Soil type. '

ANOVA results are shown in Table 8. Significant factorswere:

(a) Soil moisture.(b) Operational temperature.(c) Interaction between (a) and (b).

Of these, soil moisture was the dominant factor. Figure 16 dem-onstrates moisture and temperature effects and interactions on90 percent removal times.

It can be concluded that soil moisture and operating temper-ature were the two experimental factors which most significantlyaffected the rate of TCE volatilization.

4.3.3 Residual TCE, percentages. After all the TCE whichwould volatilize at the operating temperature was collected,soil temperature was ramped to drive off any remaining TCE.Residuals were observed during temperature ramping undercertain experimental conditions.

The measurement of interest was the proportion of total TCEvolatilized represented by these residuals. Data analyzed werethe percentage of residual TCE to total TCE volatilized.

Data were collected from the same two sets of experimentsused for volatilization kinetics. The first set of data werefor conditions of constant soil moisture and variable operatingtemperature, TCE concentration, and soil type. ANOVA resultsare summarized in Table 9. Statistically significant factorswere:

*(a) Operating temperature.(b) Soil type.(c) Interaction between (a) and (b).

Operating temperature was the dominant factor. Interaction ef-fects were statistically significant at the 5 percent level,but they were marginally significant compared with operatingtemperature. Figure 17 displays individual and combined signif-icant factor effects on percentage residuals.

465389A

TABLE 8. ANALYSIS OF VARIANCE FOR VOLATILIZATION KINETICSAT CONSTANT TCE CONCENTRATIONS

Source Degreesof Sum of of Mean Calculated

variation squares freedom square F

Replication 0.9344 1 0.9344 0.0237

A Soilmoisture 1,078.2467 1 1,078.2467 27.3739

>4•45*

B Operatingtemperature 568.4723 2 284.2362 7.2160

>3•59*

C Soil type 15.7096 2 7.8548 0.1994

AB 368.1472 2 184.0736 4.6732>3.59*

AC 242.6701 2 121.3351 3.0804

BC 265.8560 4 66.464 1.6874

ABC 171.6655 4 42.9164 1.0895

Error 669.6223 17 39.3895

Total 3,381.3241 35

F 0 . 0 5 (1, 17) = 4.45F0 . 0 5 (2, 17) = 3.59F0 .0 5 (4, 17) = 2.96

*Significant at the 5 percent level.

.~f

475389A

- .. . .-. ._ . ." . ". . . , ." '.", :. .- . . .- . . "", "" '2 ''..'',- " . :.- '-." " : ".." i ." .< 4

.

F4

Level 3(1200C) 8.17 1.67

* a.

E0- Levelo 2 16.67 10.38

0~Level 1LevelC 118.18 3.23(2700)

Level 1 Level 2(Dried at 600C) (Ambient)

Soil Moisture

Note: Higher temperatures and increased moisture levels -S-promote iigher stripping rate of TCE from soil.

'. Figure 16. Average time (minutes) for 90 percent removal (of Al) of TCE at afixed TCE concentration at specified operating temperatures andsoil moisture levels prior to ramping operating temperature to 175 0 C.

48

* ~.A'* ~ ~ .............. ... ,.......... ........... ................ . . . .... •- • - -i i . ... i'm -' A.. .. * .. . ... , .L • - " " . _, , : ;. . . ";... . *.-*.* - ,.,

3 TABLE 9. ANALYSIS OF VARIANCE FOR PERCENTAGE RESIDUAL TCE,CONSTANT MOISTURE CONDITIONS

Source Degreesof Sum of of Mean Calculated

variation squares freedom square F

Replication 15.2881 1 15.2881 0.134

A Operatingtemperature 51,938.4829 2 25,969.24 228.34

>3.59*

B TCE con-centration 182.9707 1 182.9707 1.61

C Soil type 2,281.92 2 1,140.96 10.03>3.59*

AB 256.8456 2 128.4228 1.13

AC 1,465.8828 4 364.2207 3.20>2.96*

BC 417.6272 2 208.8136 1.84

S ABC 671.2563 4 167.8141 1.48

Error 1,933.3903 17 113.7288

* Total 59,163.6639 35

F0 .0 5 (1, 17) = 4.45* F0 .05 (2, 17) = 3.59F0 .05 (4, 17) = 2.96

%'. ,

*Sgnifcan atte5pr etreel.

49538 9A

% %1

,-4.'s--I - - . - V T

.,A, Letterkenny .- 99.5 52 0.13

-- ,-_ Twin Cities -- 99 45.25 2.00

°";"Sharpe .- 81.5 16-25 0.85 I

I II

Level 1 Level 2 Level 3

"- i(27- °C) (90-° C) (120-oC)

. .'; "Temperature (°C)

-" ," Note. Moisture levels - SH: 0.5%; TC: 0.5%; LK: 1.0%. ." ""Percent residual TCE increased withdecreasing operating temperatures.

ItI

Figure 17. Percent residual TCE at a fixed moisture level compared to soildceig type and operating temperature.

i- so

The concentration of TCE did not appear to significantlyinfluence either the existence or the relative magnitude of apost-operating temperature residual under constant moistureconditions.

The second set of data were for the factors of soil mois-ture, operating temperature, and soil type with TCE concentra-tion held constant. ANOVA results are shown in Table 10. Theseindicated that the statistically significant factors were:

(a) Soil moisture.(b) Operating temperature.(c) Interact ion between (a) and (b).

Operating temperature was the dominant factor. Figure 18 dis-plays individual and combined significant factor effects on

*residual percentages. Soil type was not significant under con-ditions of constant TCE concentration.

These results indicated that operating temperature, soiltype, and soil moisture significantly influenced the existenceand proportion of TCE residuals which remained after maximumpossible volatilization at an operating temperature.

4.3.4 Residual TCE, concentr~ations. The calculation ofresidual percentages, although ar important engineering param-eter, eliminates the physical ditferences in density and testvolumes of the test soils. It was necessary, therefore, tocompute residual concentrations since these are the quantitiesmore relevant to regulatory agencies. Individual soil volumesand weights were used which introduced additional variations inthe set of experimental results.

Figure 19 presents estimated residual concentrations in ppmIfor the same two sets of experiments described earlier. Valuesrange from not detected to more than 1,300 ppm.

The data in Figure 19 were subjected to the same type ofANOVA as were the data already described. ANOVA results for

*constant moisture conditions are presented in Table 11. Allfactors and their paired interactions were statistically sig-LA

Anificant at the 5 percent level.

ANOVA results for constant TCE concentration are presentedin Table 12. Ag-ain, all individual factors were statisticallysignificant. In this case, only the interaction between soilmoisture and operating temperature was also significant.

515389A

i7 -. ,;,R

TABLE 10. ANALYSIS OF VARIANCE FOR PERCENTAGE RESIDUAL TCE,CONSTANT TCE CONCENTRATIONS

Source Degreesof Sum of of Calculated

variation squares freedom Mean square F

Replication 236.2881 1 236.2881 0.688

A Soilmoisture 1,967.9574 1 1,967.9574 5.73 >4.45*

B Oven Tem-perature 32,042.9885 2 16,021.494 46.67 >>3.59*

C Soil type 1,482.0763 2 741.038 2.16

AB 6,394.8630 2 3,197.432 9.31 >3.59*, 2. b4', •

AC 673.5140 2 336.757 0.98

BC 3,689.6142 4 922.404 2.69

ABC 2,756.1622 4 689.041 2.01

Error 5,836.5218 17 343.3248

Total 55,079.9855 35

F0.0 5 (1, 17) = 4.45F0 . 0 5 (2, 17) = 3.59F0 .0 5 (4, 17) = 2.96

*Significant at the 5 percent level.

.. ,

525389A

-" -... ". "..'.," ,, * .... ,, ,, .:..' ." ,, . .., ,. .:\ .,. v..' -.. , .4

Level 3 0.83 0.33(120°C)

I.- Level 2Cm (90C) 38.83 47.17S

0.

0

Level 3

IW(27oC) 99.33 47.17• :(9000

Level 1 Level

(Dried at 600C) (Ambient)

Soil Moisture

b..

Note: TCE removal is increased by increasingmoisture levels in soils. Level 2 data resultedfrom experimental error.

~Figurel18. Percent residual TCE at fixed TCE concentration (1,ulL) compared

to operating temperature and soil moisture level independent ofsoil type.

53

%. . . . . . .~"r~.~-~*.................•

00

I- E

v CLCS, C.

04

CYC

- -

N NCN0 'o

ULC4 Of ,

a ,a N z

""'C CY c

cm . m CM

4.,54

4-y 'W

TABLE 11. ANALYSIS OF VARIANCE FOR PERCENTAGE RESIDUAL TCE,CONSTANT MOISTURE CONDITIONS

Source Degreesof Sum of of Calculated

variation squares freedom Mean square F

Replication 7,321.6545 1 7,321.6545 0.8924

A Operatingtemperature 2,489,293.351 2 1,244,646.676 151.71 >3.59*

B TCE con-centration 1,494,343.255 1 1,494,343.255 182.14 >4.45*

C Soil type 348,568.3177 2 174,284.1589 21.24 >3.59*

* AB 980,809.627 2 490,404.8135 59.76 >3.59*

AC 176,736.1173 4 44,184.0293 5.39 >2.96* ~

BC 275,974.2063 2 137,987.1032 16.82 >3.59*

ABC 82,087.7717 4 20,521.9429 2.50

Error 139,470.6755 17 8,204.1574

Total 5,994,604.976 35

F0.0 5 (1, 17) = 4.45FO. 05 (2, 17) = 3.59F0*05 (4, 17) = 2.96

*Significant at the 5 percent level.

538 9A 5 ~

TABLE 12. ANALYSIS OF VARIANCE FOR TCE RESIDUAL CONCENTRATION,CONSTANT TCE CONCENTRATIONS

Source Degreesof Sum of of Calculated

variation squares freedom Mean square F

Replication 1,241.3878 1 1,241.3878 1.2599

A Operatingtemperature 33,403.6545 1 33,403.6545 33.901 >4.45*

B Moisturecontent 197,854.3739 2 98,927.1870 100.40 >3.59*

C Soil type 20,238.8123 2 10,119.4061 10.27 >3.59*

AB 20,965.7772 2 10,482.8886 10.64 >3.59*

AC 6,121.8088 2 3,060.9044 3.11

. BC 11,040.6294 4 2,760.1574 2.80

ABC 5,505.3395 4 1,376.3349 1.40

Error 16,750.6922 17 985.3348

Total 313,122.4756 35

F0 . 0 5 (1, 17) = 4.45F0 . 0 5 (2, 17) = 3.59F0 .0 5 (4, 17) = 2.96

*Significant at the 5 percent level.

* .- ,

L%.%~i56

5389A

5. DISCUSSION

5.1 Volatilization kinetics. Statistically significant fac-tors were:

(a) Constant moisture:-Operating temperature.

* (b)Constant TCE concentration:- Operating temperature.- Soil moisture (dominant factor).- Interaction between them.

Relationships indicated by Figures 15 and 16 were:

(a) Kinetics is a function of operating temperature.(b Kinetics is a function of soil moisture.(c) Kinetics is a function of a combination of temperature

and moisture.

It may be expected that the rate of volatilization of a com-p ound such as TCE would be a function of soil temperature. Fig-ure 15 indicates that there may be a critical temperature in thevicinity of the boiling point of TCE (87 0 C) above which vola-tilization rates increase rapidly. This is a nonlinear relation-ship which most likely involves sorption/desorption thermody-namics.

The significant behavior of soil moisture and its interac-tion with operating temperature precludes a simple temperaturerelationship. Increased volatilization rates with increasedmoisture (within the bounds of moisture levels used in thisstudy), and the interaction of moisture with temperature suggestavolatilization mechanism which involves sorption/desorption r

- ~ with soil particles and solution/desolution with soil moisture.This leads to the conjecture that TCE may follow a model formu-lated for certain pesticides.

-. 4'According to this model, the volatilized compound rarely de-sorbs directly from soil particles. The reason is that in thecompetition for sorption sites on and within soil particles, TCEand similar low polarity compounds may be subordinate to soilwater. That portion of TCE which succeeds in finding soil par-

- -. ticle sorption sites may be difficult to dislodge directly.Therefore, the compound is more easily volatilized from eitherpockets of bulk material or from solution in soil water. Vola-

tilization from bulk material is possible if the material is

5389A

present far in excess of its solubility in available soil mois-ture and few sorption sites are available. This is most likelythe case for the concentrations of TCE used in this study. Vola-tilization from soil moisture follows Henry's Law. Material al-ready in solution is simply air stripped according to estab-lished mechanisms. Material sorbed in or on soil particles ismore easily dissolved in soil moisture than stripped directly.Once in solution, air stripping proceeds.

This model holds until the amount of soil moisture drops be-low some critical value. It is thought that this critical valuerepresents the moisture level at which the volatile material be-comes the dominant sorbed species. It is now necessary to desorblarger amounts of the material directly from soil particles,something which requires significantly more energy than airstripping from solution.

The rate of desorption (volatilization) decreases propor-tionally.

Data generated in this study was sufficient to indicatestatistical significance. This was necessary and sufficient forengineering preparation of pilot studies. They were not suffi-cient to verify the volatilization model outlined above. Thedata do fit the model conceptually, and it is recommended thatmechanisms analogous to those of the model are operative in thissituation.

Temperature and moisture are critical to the rate of vola-tilization of TCE. It is important that the moisture level ofsoil being air stripped of TCE not be decreased prior to place-ment in the volatilization chamber.

5.2 TCE residuals, percentage. The existence of a TCE re-sidual, verified by separate analysis, was a surprise. Formula-tion of the original Test Plan, Appendix A, was based on the as-sumption that TCE would be sufficiently volatile to permit es-sentially complete stripping. Time for completion was anticipat-

-.'- ed to be the primary dependent variable.

.- .

585389A

A residual did exist under most of the experimental condi-tions of this study. Evaluation of residuals was first made onthe basis of residual proportion to total TCE volatilized sothat direct comparison of data could be made without referenceto specific soil properties (e.g., sample weight and soil dens-ity). Summary results of statistically significant factors were:

(a) Constant moisture:- Operating temperature.- Soil type.- Interaction between the above.

(b) Constant TCE concentration:__ - Operating temperature.

- Soil moisture.- Interaction between the above.

Operating temperature was the dominant factor in both cases.

Relationships indicated by Figures 17 and 18 were:

(a) Residual fraction decreases as temperature increases.(b) Residual fraction decreases as soil moisture increases.(c) Residual fraction inversely proportional to an unspeci-

fied interaction between soil type and operating tem-perature.

(d) Residual fraction inversely proportional to an unspeci-fied interaction between soil moisture and operatingtemperature.

TCE concentration was not statistically significant to thebehavior of residual fractions. This was not surprising sinceproportions were used rather than absolute values. In suchcases, the proportions tend to normalize data and eliminatevariations due to absolute values. Proportions permit wide com-parability of data when searching for trends, patterns, and re-lated behavior.

A model is not proposed for TCE residuals. It is conjecturedthat the existence and relative magnitude of a residual may bedependent upon soil parameters such as clay content (related tospecific surface area) and organic content (related to sorptionof organic compounds in the organic fraction). These are keyparameters in modeling transport of materials through soils.

.°v 59

5389A 59

4W1

The primary characteristic of the TCE residuals observed inthis study is that they existed at stripping temperatures below1200C for processing times less than 20 minutes.

- - 5.3 TCE residuals, concentration. When sample weights wereused to compute absolute values of residual concentration, in-creased variability was introduced to the residual data because

-: of differences in individual soil densities and va-iat ions inexperimental sample weights. Under these conditions, all experi-mental factors and most of their paired interactions were sta-tistically significant. These results followed observations forresidual fractions.

Important to this discussion are the residual concentrationvalues. These values ranged from not detected to more than1,300 ppm. It is not clear if these concentrations were trueproperties of the soil/solvent system under study or were arti-facts of the experimental apparatus and procedures. Residualpatterns were reasonably consistent, however, and their exist-ence at temperatures below 1200 C have strong implications fordesign of pilot studies for heated air stripping of contaminatedsoils. Other implications, such as those for regulatory compli-ance, should be noted but placed in abeyance since only this setof data are available at present. Verification studies for theexistence, magnitude, and conditions of TCE residuals are re-quired. Some verification will come from an in situ pilot studyunder way at this writing.

5.4 Carrier gas flow rate. The experiment performed was de-signed with a fixed carrier gas flow through the soil containedin the test cell. This restriction (20 cubic centimeters per

-minute) was implemented during shakedown since higher carriergas rates resulted in poor performance of the Hall cell.

*In any gas stripping system the carrier gas rate is impor-tant to the stripping rate of a volatile compound from another

- . media. Increasing the rate of stripping gas will increase theamount of contaminant stripped (based on mass transfer theory).Future pilot studies must utilize gas flow rate as an experimen-tal parameter to observe the effects on TCE removal rate effi-ciency.

6015389A

.. . . . .

6. CONCLUSIONS AND RECOMMENDATIONS

6.1 Volatilization kinetics. Operating temperature (thetemperature of the soil being treated) and the level of soil

.- moisture are two factors in determining the rate of TCE vola-tilization from the three soils under study. Volatilizationrates can be increased by almost a factor of 3 when the operat-ing temperature is raised from 270 to 1200C. Volatilizationrates can be increased by almost a factor of 8 when soil mois-ture is raised from a 600 C equilibrium to an ambient equilib-rium level.

There is an undefined interaction between temperature andmoisture which is also significant.

6.2 TCE residuals, percentage. TCE residuals occurred inthis study for all experimental conditions at operating tem-peratures less than 1200C.

Residual fractions were signifcantly dependent upon all ex-*' perimental factors except TCE concentration. Increased operating

temperature, increased soil moisture, and change from Letter-kenny to Sharpe types of soils all decreased residual fractions.Paired interactions between the significant factors were notedto exist but were not definable.

6.3 TCE residuals, concentration. All experimental factorsand all paired interactions between them were statistically sig-nificant to TCE residual concentration levels.

6.4 Engineering implications.

6.4.1 Assumptions.

(a) The significant factors identified in this study willalso be significant factors in a pilot study.

(b) The conclusions from this study are correct and based 9upon actual phenomena.

(c) Increased carrier gas flow rate will be a significantfactor in increased TCE removal rates.

(d) Engineering extrapolations can be made from the data ofthis study for the purpose of designing subsequentpilot-scale studies or for the purpose of suggestingmodifications to study plans for pilot-scale studiesin progress.

7

615389A

6.4.2 Implications for ambient temperature volatilization.The obvious and consistent patterns in temperature dependency ofboth kinetics and residuals implies that significant residualswould remain from attempts to air strip TCE from soils at am-bient temperatures greater than 90 percent of original TCE spikewith soils at ambient moisture. The kinetics factor must betaken into account. Pilot studies may reveal that ambient tem-peratures merely prolong the stripping process and the end re-sult would be the same regardless of temperature. This has tobe verified.

6.4.3 Implications for thermally induced volatilization.Temperatures above ambient are required to improve strippingkinetics and to minimize residuals. The operating temperatureshould reach 110 to 1200C for complete stripping of TCE. Co-volatilization of water vapor with the TCE may assist in com-pletely scrubbing the TCE residual. This has to be verified.

Machinery for stripping TCE from soil may be of two types.The first is a single chamber device operating at the appropri-ate temperature throughout. The second is a two-stage devicehaving a low temperature first stage and a high temperature sec-ond. A comparative evaluation is required to be able to recom-mend which generic type may be most applicable.

6.4.4 Options for pilot-scale testing of thermally inducedvolatilization. Equipment most readily adaptable to pilot test-ing of thermally induced volatilization comes from the field ofcommercial drying.

The protocol for pilot testing should be designed to verifythe conclusions of this study and to provide answers to engi-neering questions of optimization.

Machinery employed for pilot testing should have sufficientcontrols to permit parametric testing of both design and opera-tional parameters. Compromises will be required to avoid thenecessity of custom construction. Parameters include feed rate,TCE concentration, operating temperature, soil moisture, feedpreparation, agitation rate, and air recycle rate.

Options for equipment include actual pilot-scale machineryand small-scale commercial machinery. A brief survey of theequipment market revealed that both options are conceptually vi-able. Actual selection will depend upon availability, configura-tion, cost, and dependency on proprietary designs.

625389A

b<';'<:" : :':: ": ".'"":"" "'-"- .'. ..........-..... . ........ ".....".-"""-.".-. -". " . " """-'

W7 7- -- 7-

APPENDIX A

TEST PLAN FOR A BENCH-SCALE INVESTIGATION OF

LOW TEMPERATURE THERMAL REMOVAL OF TCE FROM SOIL

Contract No. DAAKl-82-C-0017Task Order 4

INSTALLATION RESTORATION GENERAL ENVIRONMENTAL

TECHNOLOGY DEVELOPMENT

Test Plan for a Bench-Scale Investigation of

Low Temperature Thermal Removal of TCE from Soil

Walter L. Lambert, Ph.D.

Lawrence J. Bove

Peter J. Marks

ROY F. WESTON, INC.West Chester, Pennsylvania 19380

January 1984

Prepared for:

U.S. ARMY TOXIC AND HAZARDOUS MATERIALS AGENCY

Aberdeen Proving Ground (Edgewood Area), Maryland 21010

A-1................ r * -~~**4* .I- *-'- -

S:". *" . "- . "'" .'.,.,.-.-" . . '.I% %' ', "

?

CONTENTS

Page

Paragraph EXECUTIVE SUMMARY. ..............1 PURPOSE AND OBJECTIVES ......... 51.1 Purpose ................ 51.2 Programmatic framework ........... 51.3 Objectives..................61.4 Criteria for positive test ofcoep62 PARAMETERS AND TEST CONDITIONS ........ 72.1 Test parameters. .............. 7

'32.1.1 Definitions ................. 72.1.2 Relevant test parameters .......... 72.2 Test design. ................ 102.2.1 Discussion, independent variables 102.2.2 Data analysis . .. .. .. .. . .. . 122o2.3 Experimental designo.............123 TEST APPARATUJS...............143.1 Background ................ 143.2 Test apparatus. ............. 154 SAMPLING AND ANALYSIS PLAN..........184.1 Sample preparation and handling . . . . 184.1.1 Soils. ............ ... .. 184.1.2 Bulk samples... ...... ..... 204.1.3 Experimental samples, soil. ....... 204.1.4 Experimental samples, off-gas .. ..... 224.2 Baseline information requirements and

procedures............. . . 234.e.l Bulk soil. ................. 234.2.2 Experimental samples, soil adofgas. 254.3 Experimental information requirements

and procedures.............254.3.1 Experimental information ....... 24.3.1.1 TCE volatilized .............. 254.3.1.2 moisture content.. ...... . . . . 254.3.1.3 Time . . . . . ... . . . 274.3.1.4 Temperature, oven..............274.3.1.5 Temperature, soil sample..........274.3.1.6 Purge gas flow rate ............. 274.3.2 Experimental protocol ............ 27

4576A

A- 2

V 7 -0

CONTENTS

Paragraph 5 DATA ANALYSIS. ............. 29

5.2 Data analysis . . . . . .. .. .. .. .. 296 SCHEDULE. .. ............... 336.1 Project schedule . . .. .. .. .. .. 336.1.1 work breakdown structure. ........ 336.1.2 Task 1I- Setup and coordination . . . 336.1.3 Task 2 -- Shakedown .............336.1.4 Task 3 -- Experimentation .. ....... 336.1.5 Task 4 -- Data analysis ........... 346.1.6 Task 5 -- Final report...........346.2 Project schedule.............34

4576A

A- 3

.4 °.

FIGURES

Page

FIGURE 1. Schematic of the apparatus to measurethe volatilization of TCE from soil. . 3

2. Experimental design for investigatinglow temperature thermal removal oftrichloroethylene from soils ..... 13

3. Schematic of the apparatus to measurethe volatilization of TCE from soil . 16

4. One type of drying curve . ........ .. 245. Type of TCE concentration curves antic-

ipated . . .............. 266. Anticipated relationship between TCE re-

siduals in soil and total organic car-bon of the soil .... ............ . 30

7. Anticipated relationship between TCEvolatilization and drying temperature 31

8. Projected technical schedule ...... .. 35

TABLES

Page

TABLE I Analysis of experimental variables . . . 82 Amount of soil of each type required 193 Characterization parameters for each

soil type ............. 214 Simplified experimental protocol . 28

4-r~

iv

4576A

A-4

EXECUTIVE SUMMARY

The purpose of the test is to prove the concept of low tem-perature ambient, i.e., 100+ OC) thermal removal of tri-chloroethylene (TCE) from soils.

The objectives of the bench-scale study are the following:

(a) Identify statistically significant parameters that af-fect removal rate and residual levels.

(b) Determine the sensitivity of residuals to key opera-tional parameters.

(c) Bracket the values of key design and operational param-eters.

(d) Provide a data base for deciding whether to employ pi-lot- or demonstration-scale testing.

The variables and parameters to be controlled and/or meas-ured are the following:

(a) Independent experimental variables:

- Moisture content of soil at two levels.- Temperature of the test environment at three levels.- TCE concentration of the test soil at three levels.- Flow rate of purge gas (nitrogen flow used to carry

TCE from heated soil), fixed by preliminary experi-ments at one level.

- Soil type (uncontaminated soil from actual installa-tions) fixed at three levels (from three sites).

(b) Dependent experimental variables are the following:

- TCE concentration of purge gas (leaving heated soilsample), continuously sampled and measured.

- TCE concentration of the soil during an experiment,computed by mass balance.

- TCE concentration of the soil, measured immediatelyupon completion of an experiment.

1 "

4576A

A-5

"%N .. - t;T

(c) Parameters to be monitored as part of an experiment arethe following:

- Temperature of soil sample at start of experiment.- Temperature of soil sample during an experiment.- Temperature of soil sample immediately upon comple-

tion of an experiment.- Temperature of purge gas entering sample container.- TCE concentration of purge gas before entering sam-

ple container.- Dry weight of soil sample.- Time.

The test design is a full factorial, as follows:

(2 levels of soil moisture) x (3 temperatures) x (3 TCEconcentrations) x (3 soil types) x (3 replications) = 162experiments

The levels of the experimental variables are as follows:

Variable Level 1 Level 2 Level 3

Soil moisture 20 percent 40 percent ---Temperature 250C 900C 150 0CTCE concentration 10 mg/kg 100 mg/kg 1,000 mg/kg

in soil (dry) (dry) (dry)Soil type Location 1 Location 2 Location 3

The test apparatus is a modified Hewlett-Packard Model5880A gas chromatograph (GC) with a Hall detector. Flow-throughsample tubes will be held under isothermal conditions in the GCoven. Dry nitrogen purge gas will flow through the tube andcarry off TCE through a programmed sampling valve to a Hall de-tector. The rate of TCE volatilization will be recorded on anindividual grab sample and cumulative basis. A mass balancearound the soil sample will provide interim values for TCE re-siduals. Figure 1 is a diagram of the test apparatus.

Data analysis will be step-wise linear regression and analy-sis of variance. Models have been identified from the areas ofsoil volatilization of organic chemicals, commercial drying op-

.€. erations, and dimensional analysis for investigating strippingeffectiveness and kinetics.

&. 2'., 4576A

A-6

00

a.

0

E5

00

C 0 0

miC- r~ d ; C . U ' .4

-IA--

Details concerning the rationale for the selection of exper-imental variables, test apparatus, and the most likely commer-cial-scale unit processes are provided in this test plan.

Execution of the plan will involve continual interaction

between WESTON and USATHAMA. It is anticipated that a mid-streamdecision will be required on the final choice of the third tem-perature to be used. Data analysis will begin soon after thereceipt of initial experimental data. The results of that anal-ysis may indicate that a temperature other than 150 0 C shouldbe used. This will be presented prior to the start of experi-ments at the third temperature level.

4 .:

. .

4

4576A

A-8

" " 2-

1. PURPOSE AND OBJECTIVES

1.1 Purpose. Contamination of soils with trichioroethylene(TCE) is an identified concern at several U.S. Army Developmentand Readiness Command (DARCOM) installations. The purpose ofthis test plan is to provide detailed guidance and specific pro-cedural information for execution of bench-scale testing of a

* concept for low temperature removal of TCE from soils.

Low temperature stripping of TCE from aqueous solutions hasbeen demonstrated to be an economical and practical process.However, information concerning low temperature removal from a

* soil medium is scarce and limited. The objective of this inves-tigation will be to develop design parameters for this process,and to determine the factors that would affect removal effi-ciency.

The purpose of the investigation is to determine if theconcept is feasible.

1.2 Programmatic framework. This test plan is an interme-diate product within a general program for the development ofpollution abatement technology. Preceding the test plan were ex-tensive efforts in the following areas:

(a) Defining soil contamination problems at selected DARCOM-. installations.

* (b) Identifying technologies that were conceptually feasi-ble and economically attractive in mitigation butthat were not necessarily state-of-the-art for suchapplications.

(c) Assessing those technologies to determine which, if* any, were promising candidates for further research

and development.

Experimental evaluation of promising technologies is to be-~conducted in a phased program that includes bench-scale inves-

tigations for proof of concept and pilot-scale investigationsfor verification of concept and for evaluation of engineeringdesign and performance parameters. Each phase of the testingprogram is governed by detailed test plans of which this docu-ment is one.

~ 4576A A- 9

1.3 Objectives. The primary objective of the investigationis to decide if the concept of low temperature thermal removalof TCE merits pilot-scale testing.

Secondary objectives include the following:

(a) Identification of process sensitive parameters, includ-ing an analysis of sensitivity.

(b) Indications of optimum ranges of operational parame-ters.

(c) Indications of the type of pilot- and full-scale equip-ment that may be most applicable.

(d) Preliminary (order of magnitude) cost analysis of theconcept.

(e) Cost/benefit analysis to deterine applicability of fur-ther technology development.

1.4 Criteria for positive test of concept. A positive testof concept indicates that the concept is applicable. The crite-rion for judging whether the test is positive is a treated soilcontaining barely detectable trichloroethylene concentrationsafter treatment.

This criterion is based on anticipated requirements for vol- .ume production through commercial equipment. The availableequipment itself demands an operating temperature in the rangeof ambient to 4000F. Economical throughput requires a soilresidence time measured in minutes. The absence of a universallyacceptable residual level of TCE in soils, and the exploratorytesting for potential engineering performance (mass balance onTCE around the test system) require that essentially all of theTCE be removed.

6

4576A A-10

* 2. PARAMETERS AND TEST CONDITIONS-

2.1 Test parameters.

* 2.1.1 Definitions.

*(a) Parameter. This is a measurable property or character-istic that may be quantified as part of bench-scaletesting. The word is also used for constants, coeffi-cients, and exponents that describe statistical popu-lations. Both uses of the word will be applied inthis test plan.

(b) Experimental variable. This is a parameter that is un-der investigation. Experimental variables are eitherdependent or independent. The latter are controlled,and their values arc predetermined. The former areuncontrolled, and their values are monitored andmeasured.

2.1.2 Relevant test parameters. Thirty-seven parameterswere identified as having relevance to this study. Of those, 18were soil characteristics that were fixed by the source and typeof soils to be used. The remainder were associated with either

* experimental or full-scale operation of potentially adaptable _

commercial equipment. Each of the parameters was considered tobe a candidate experimental variable. An analysis of each wasconducted, and the results are shown in Table 1.

All of the parameters in the list were either controllableor not. Some had relevance to bench testing, and all had rele-4vance to full-scale operations. All were measurable. If used asan experimental variable, some were judged to be dependent inthat they were Judged to be products of a commercial operation.Others were judged to be independent because they were eitherf ixed by materials to be used or could be controlled independ--

__ ently of one another.

Experimental variables were identified as parameters meetingthe following set of criteria:

(a) Controllable.(b) Relevant to bench-scale investigations.(c) Measurable.(d) Not fixed by preservation method, source of material,

or condition of supply, e.g., TCE-free purge gas sup-plied in that condition in cylinders.

4576A A-1

77

X X

-- x

Wic -

7,1

-2 r2

-~77-

4I2 ~ AC AC CA - . C 12.

x =

I

x "

02

.l- ..c O

E- 0 t,

Igo

A. l .I - ... . -.-

A-1

. .....

,..02 202"

i .

[ A-13

Independent experimental variables for consideration in thisstudy are the following:

(a) Moisture content, soil, initial.(b) Temperature, oven.(c) TCE concentration, soil, initial.(d) Flow rate, purge gas.(e) Soil characteristics (18 characteristics).

Dependent experimental variables are the following:

(a) TCE concentration, purge gas, outflow (measured), con-tinuous.

(b) TCE concentration, soil, interim (computed).(c) TCE concentration, soil, final (measured).

Experimental parameters for monitoring are the following:

(a) Temperature, soil sample, initial.(b' Temperature, soil sample, interim.(r) Temperature, soil sample, final.(J) Temperature, purge gas, inflow.(e) TCE concentration, purge gas, inflow.(f) Weight, soil sample, dry.(g) Time.

2.2 Test design.

2.2.1 Discussion, independent variables. The moisture con-tent of the soil to be stripped of TCE is a key parameter forseveral reasons. First, the various forms of water found in

* soils are energy sinks. It is anticipated that water will bevaporized jointly with TCE, with energy consumed in the process.On the other hand, the microscopic behavior of TCE toward soilparticles, interstitial water, and their interface is unknown. -

It may be that free water will be required to expedite TCE re-moval because TCE may be more easily distilled from solutionthan stripped from dry soil particles. For full-scale opera-tions, it may not be feasible to reduce the moisture content ofcontaminated soils prior to heating; however, it may be possibleto increase the moisture. Therefore, two moisture levels areconsidered adequate for bench testing. One level will reflectcurrent past experience with actual contaminated soils (12+ per-cent). The second will be at a higher level to determine thesensitivity of the thermal process and to simulate conditionsthat may be feasible at full-scale. Both levels need to be con-trolled. The lower was chosen to be 20 percent, which is a lit-tle above ambient to allow control in the laboratory. The secondwas 40 percent.

10

*4576A A-14

-.

Stripping temperature is a second k3y parameter. Attainingand maintaining temperature in a commercial unit will depend onthe heat capacity and initial temperature of the feed material,heat losses of the equipment, and amount of water costripped.Maintaining the stripping temperature is directly related to fu-el costs. The objective of a commercial operation will be to op-erate at the lowest temperature that provides an acceptableproduct. Bench testing should narrow the range of preferredtemperatures as much as possible. Three benchmark temperaturesare the following:

(a) Ambient (250C).(b) Greater than ambient and less than the boiling point of

- water (25oC>T<100 0 C).(c) Greater than the boiling point of water (above 100 0C).

The level of TCE in the feed material is an obvious keyparameter. It will be interesting to determine if, however, therate of removal of TCE is a simple function of temperature. If

* so, operating parameters such as soil residence time may be pre-dicted by knowing the TCE level. Making such a determinationrequires at least three levels of TCE in the feed soil. Threelevels considered adequate for investigating TCE concentrationeffects are 10, 100, and 1,000 mg TCE/kg dry soil. These concen-trations were selected based on WESTON's experience with TCElevels in contaminated soils on DARCOM installations.

The flow rate of the purge gas moving through or over heatedsoil is a critical operational parameter for removal of water orlarge quantities of solvents. In those cases, the relative hu-midity of the purge gas before and during operations influencesthe rate and amount of water or solvent pickup. A review of thepsychometric behavior of benzene, whose physical behavior shouldapproximate that of TCE, indicated that saturation of the purgegas is unlikely at the levels of TCE used in this study. There-fore, it appears that the flow rate of the purge gas can be

- * fixed at a value approximating that which would be found infull-scale equipment.

.. Eighteen soil characteristics are given for the soil understudy. It is anticipated that widely differing soils from threesites will be used.

4576A A-15

- .- - -iU!. *

77 --- - - -Z .- W

A soil scientist will collect bulk quantities of each soilat each location. It is important to recognize differences be-tween soil horizons with depth, especially organic content. Col-lecting a representative sample of site-specific soil will de-pend on proper sampling and compositing. How each soil is to becollected will depend on an estimate of soil conditions at eachsite. The specific sampling strategy to be used will be deter-mined after the site assessment. 2

'..I

2.2.2 Data analysis. It is desirable to extract as muchinformation as possible from the minimum number of experimentalruns. The method planned for data analysis is step-wise linearregression with two-way analysis of variance.

2.2.3 Experimental design. A full factorial design withthree replicates per test case is planned. The number of runsare:

(2 levels of soil moisture) x (3 drying temperatures) x (3TCE concentrations) x (3 soil types) x (3 replicates) = 162runs. ?

-% Experimental runs made as part of the factorial design donot include preparatory runs for equipment shakedown and explor-atory runs for fixing the flow rate of the purge gas. . .

Identification of each run is provided in Figure 2. Each' run is identified by a five-digit number. Each digit denotes

one of five experimental variables (factors), and the value ofthe digit denotes the level (value) of that factor.

The values shown in Figure 2 for each level of the experi-mental factors are tentative. They are based on best estimates

- at the time of plan preparation. During the shakedown phase of- the project, it may be found that levels for some factors may

have to be changed. Temperature is an example. The standardizedmethod for TCE headspace analysis uses 100oC as the driving

-. temperature for removing TCE from small soil samples, quantita-tively. It may tu.,n out that two of the temperatures originally

*. chosen fit the objectives of the study, but the third may have*. to be adjusted. If such adjustments are necessary, USATHAMA ap-

proval will be solicited prior to making the adjustment.

12

4576A A-16

Vi

ca

SiQ 0c, R 0

C N 0

0.N 00- 0

20

-a. E t ic'2

NN

E a.

cm CC

E~ iC

C-4o

L- " aO-- d, SA Za/4 CaA d/ d AI"ii s los lo 'I 1 l" ls ls I

a10' ul L 554 10Ua10 olU DU011AIU0uo 301uolwiioouO 30 U0l9Au~uO 30

""A-17

- . V . -'-- -. .

3. TEST APPARATUS

3.1 Background. Thought was given to the types of commer-cial-scale equipment that might be used for volatilization ofTCE from soils. A brief review of chemical and metallurgicalprocessing equipment resulted in industrial dryers being theprime candidate for full-scale operations. A brief survey ofdryer manufacturers and vendors was completed. It was evidentfrom discussions with vendors that a bench-scale study using ageneral apparatus could not simulate the engineering performanceof a full-scale unit. Therefore, the thrust of the bench studyis proof of concept.

Considerations for selection, design, configuration, andoperation of bench-scale apparatus included the following:

(a) A soil sample size in the 1-5 gram range. A review ofthermal testing equipment available off-the-shelf,including a rather sophisticated differential scan-ning calorimeter, indicated that instrumentation forthermal analysis of either pure or homogeneous mate-rial is available. However, sample sizes are often inthe 10 to 200 milligram range. This was consideredtoo small. It was anticipated that variability amongsoil samples of such size would be unacceptablylarge, and the resultant masses of TCE evolved wouldbe very small.

(b) Flow-through purge gas. Most of the applicable indus-trial dryer designs subject the treated material toeither extensive tumbling or they force purge gasthrough the material as it travels on slotted beltsand trays. In either case, purge gas passes throughthe material. It was considered important that thismode of gas travel be incorporated in the test appa-ratus design.

(c) Temperature control. Isothermal conditions rather thanramped heating of the sample were considered more ap-plicable and meaningful. For proof of concept, tem-perature control to within a degree Celsius wasjudged adequate.

(d) Mass balance of TCE. The following three methods forclosing the mass balance for TCE around the sampleswere considered:

- Direct weight loss measurement. This is a techniqueused in thermal gravimetric analysis (TGA). Ananalysis of sample size and TCE concentrations in-dicated that TCE weight losses over time would be

144576A

A-16

so small that considerable experimental error wouldbe introduced. This technique was rejected.

- Residuals measurement. A sampling technique was con-sidered by which treated soil samples could bewithdrawn from the heated environment and directlyanalyzed for TCE. A review of sample handling re-quirements, the number of samples required, andthe sources of experimental error caused this tech-nique to be rejected.

- TCE concentration in the purge gas. GC/Hall cellanalysis for TCE appeared to be sufficiently sensi-tive to permit periodic sampling of the purge gasevolved from the heated soil sample. Frequent sam-pling with results integrated over the run timepromised to provide a time-release curve for TCE.Analysis of duplicate soil samples prior to heatingcould provide the initial concentration of TCE inthe soil. Elevating the oven temperature to approx-imately 110 0C at the conclusion of a test runwould drive off any residual TCE in the sample

V. (this is analogous to the headspace method used tomeasure TCE in soils). This method promised to pro-vide a reasonably accurate mass balance over timeand was the method chosen.

(e) A review of candidate bench-scale laboratory equipmentwas made. It was concluded that all of the require-ments for control, sample size, and detection could

.- be satisfied if the test were run using various sub-assemblies of a standard gas chromatograph.

3.2 Test apparatus. Figure 3 is a diagram of the proposedtest apparatus. Major subsystems of a Hewlett-Packard Model

. 5880A gas chromatograph (GC) will be adapted for use as a simu-lation system for through-circulation volatilization of TCE froma variety of soils.

The GC oven (Figure 3, item 9) will provide a controlled,isothermal environment. It can also provide ramped temperaturesfor TCE residuals volatilization, and for drying soil samples atthe end of a run.

15457 6A

A-101~ ~ ~ **''.*,"* *

E'a'

.

cSm

@6

U

-- 5

5'E

0 CU

0 00Z1

o 0 s20~~ 0 0(

~~~> E UO-ti-

- = =6E U IL. >

16

A- 20

.'L! I- A--. . .

Purge gas will be dry nitrogen (Figure 3, item 3). It willbe brought to oven temperature by passing it through a coil of

1/16-inch stainless steel tubing (Figure 3, item 8) prior to en-tering the sample holding tube. The choice of dry nitrogen for *purging was made on a worst-case basis. In a commercial dryer,it would be most economical to operate with unprocessed air

-2 that would have a variable relative humidity. Work on the vola--tilization of pesticides indicated that the rate of volatiliza-tion depended in a complex way on diffusion and mass transportof soil moisture. A dry purge gas will maximize the rate of wa-Iter evaporation, which may provide the slowest rate of TCE vola-tilization. Nitrogen will not interfere with operation of theHall cell.

One to five gram samples of spiked soil will be held inglass-wool stoppered tubing (Figure 3, item 6) at oven tempera-ture. The sample tube will be fitted on either end with swagelocfittings for mating to the purge gas tubing. Inside the sampletube, and imbedded in the soil sample, will be a thermoccuple

* (Figure 3, item 5) for recording the temperature of the soil* sample during the experiment.

Off-gas carrying TCE and moisture will be valved (Figure 3,item 10) to a Hal'l furnace (Figure 3, item 13) and a Hall detec-tor (Figure 3, item 14) for quantification of the TCE. The heat-ed valve assembly (Figure 3, items 10, 11, and 12) will providea series of short duration samples. The series of TCE peaks ex-pected to be generated during a stripping simulation run willprovide a TCE mass evolution rate (i.e., flux). Integration ofthe area under the set of curves will quantify the total mass ofTCE evolved.

* Temperatures in the oven and of the soil sample will be mon-itored by thermocouples connected to the GC's real-time clockwith printer.

17

4576A A- 21

'°2

4. SAMPLING AND ANALYSIS PLAN

4.1 Sample preparation and handling.

4.1.1 Soils. Actual soils from three locations will beused. These soils should be free of TCE contamination (to beverified by analysis) but should be representative of the typesof soils from their respective sites.

A stockpile of each of the three soils will be prepared fromOvirgin' material according to ASTM Method D-346-78. This willprovide material that passes either a No. 60 (0.25 mm) screen(silts, clays, and loams), or a No. 20 screen (sands). Theamount of dry material required in each of the three stockpilesis approximately 10 kg. Table 2 shows the estimates of processedsoil required for each type of application to be made. Given2,600 gm required, the amount to be prepared and stockpiled is:

(2,600 gm) 1k0 ) (1.50) (2.5) = 9.75 10 kg.

1,0 gnLoss Storage

factor factor

The amount of soil to be collected from each of the threesites is at least:

(0 k)2.205 lb) 55.1 A 55 lb..40 kg

FractionforcedthroughNo. 60 mesh

Since this is a small amount of soil to account for unseen on* ~. accidental losses, approximately 200 pounds of uncontaminated,

representative soil (dry weight) should be collected per site toprovide sufficient material to carry through the entire testplan.

A portion of the clean, prepared soil will be allowed tocome to equilibrium with the relative humidity in the laboratoryat ambient temperature and pressure. This portion will be usedfor moisture determination according to ASTM Method D-346-78(ll0OC). This moisture level is denoted We and will be usedin subsequent data analyses.

184576A

A- 22

- * .. -C :M71. .

TABLE 2. AMOUNT OF SOIL OF EACH TYPE REQUIRED

Application for soil Weight per Number of per typesample application applications application 4

1 Equilibrium moisture 10 gm 3 30 gm

2. Headspace verification 10 gm 3 30 gm

* 3. Shakedown moistures 10 gm 5 50 gm

4. Shakedown headspace 10 gm 5 50 gm

5. Experimental moistures 10 gm 54 540 gm

- 6. Experimental headspace 10 gm 54 540 gm

7. Characterizations 1,000 gm 1 1,000 gm-

8. Drying curves 5 gm 18 90 gm

9. Volatilization curves 5 gm 54 270 gm

Total minimum amount of prepared soil 2,600 gm

I %.

19

4576A A-23

.7 -'L --

A second portion will be subjected to tJSATHAMA Method 2J,headspace analysis for TCE. This will verify the presence or ab-sence of TCE or analytical interferences in the soil stockpiles.For example, carbonates and sulfates may cause interference if

special handling is not used.

A third portion of each of the three soil types will becharacterized for the parameters listed in Table 3. The methodsto be used are also listed.

The remaining screened clean soil will be stored at 0 toSoC in appropriate containers as a reserve stockpile. The re-maining material will be used for characterization analyses,methodology shakcedown, and experimentation.

4.1.2 Bulk samples. Eighteen different bulk TCE spikedsoil samples will be prepared. These will reflect the following:

(3 types of soils) x (2 moisture levels)x (3 TCE levels) = 18.

Bulk samples will be prepared in the following sequence:

(a) Air dry to equilibrium moisture.(b) Spike with distilled water to bring up to predetermined

**l ~moisture level.(c) Mix in Teflon-sealed amber jar on rolling mill.(d) Add TCE to predetermined level in bottle and seal.(e) Move immediately to cold room (4 to 60C).(f) Mix thoroughly in Teflon-sealed amber jar on rolling

mill.(g) Store at OOC.

Labels will reflect the factorial design identity of eachbulk sample.

4.1.3 Experimental samples, soil. Experimental sampleswill be approximately 5-gm aliquots f rom their respective bulkcontainers. All handling of bulk containers and experimentalsample tubes will be done in the cola room.

20457 6A

A- 24

Z--c. ..,- ***

- ~ ~ ~ ~ ~ ~ . IF- I *~ . W W .. - - -

TABLE 3. CHARACTERIZATION PARAMETERS FOR EACH SOIL TYPE

Amount of soilParameter Method required per test

Soil pH ASA 60-3 (glass electrode pH meter)a 10 gm

Total organic carbon ASA 90-2 (wet combustion) 2 gm

Total exchangeable bases ASA 59-2 (residual carbonate method) 120 gm(base saturation)

Sand ASA 43-3 (filtration/sieving) 100 gm

Silt and clay ASA 43-3 Same sample

Particle size distribu- Standard sieve analysis 500 gmtion

Kaolinite ASA 49-4 (X-ray diffraction) 100 gm

Illite ASA 49-4 Same sample

Vermiculite ASA 49-4 Same sample

Montmorillonite ASA 49-4 Same sample

i Chlorite ASA 49-4 Same sample

Interstratified combi- ASA 49-4 Same samplenations of 2:1 type

'-" components

Carbonate (CaCO3 ) ASA 91 25 gn

Aluminum, total ASA 67-2 0.1 gm

" Cation exchange capacity ASA 57-3 (sodium saturation) 4 gm

Exchange acidity ASA 59-2 (residual carbonate) ±0 gm

,Heat capacity ASA 25-3 (calorimeter) 5 gm

Bulk density ASA 30 100 gm

Total minimum amount of soil for test battery 500 gm

aC.A. Black, Editor-in-Chief, D.D. Evans, J.L. White, L.E. Ensminger, F.E.Clark, and R.C. Dinauer, Methods of Soil Analysis, Part 1: Physical and Min-eralogical Properties, Including Statistics of Measurement and Sampling, andPart 2: Chemical and Microbiological Properties, Madison, Wisconsin: Ameri-

* can Society of Agronomy, Inc., 1965.

21

4576A

A-25

b " * , ",o° - o " °" •.. '. J ° -" -" J .. .- . " . . . .

AD-A162 528 INSTALLATION RESTORATION GENERAL ENVIROWrLTECHNOLOGY DEVELOPMENT TRS (U) WESTON (ROY F) INC MESTCHESTER PA W P LAMBERT ET AL MAR 85 ANXTH-TE-TR-85684

UNCLASSIFIED DRAKII-82-C-8Oi7 F/G 11/2 NL

,s..I

.2

.J&6

11111 1.12.0

1.251-11111.4

L'- L

MICROCOPY RESOLUTION TEST CHART

NATIONAL BUREAU OF STANDARDS 1963 A

'1

Experimental samples will be prepared and handled in thefollowing sequence:

(a) Sample tubes, including thermocouple and glass wool,preidentified, are weighed after 24-hour residence ina dessicator at room temperature.

(b) In the cold room, approximately 5 gm of sample aretransferred from the bulk container to the *sampletube. Glass wool retains the sample, and filling thetube allows the thermocouple to be surrounded bysoil. Glass wool is fitted into both ends of the .7tube, and the end fittings are attached. Both gasports are capped.

(c) The prepared tube is temporarily stored at 0OC untilthe GC operator begins a run. §'

(d) After the run, the oven temperature is ramped up to1100C for 30 minutes (if that temperature had notbeen exceeded for at least that long), and any resid-ual TCE is recorded. The tube is then disconnectedfrom the GC and placed in a dessicator, with both gasports open, overnight.

(e) The tube, with dry soil inside, is reweighed, and themoisture content of the sample is computed. At 20percent moisture, the amount of water present is, ata minimum, two orders of magnitude greater than theTCE initially present, therefore, any weight loss isattributed to water.

4.1.4 Experimental samples, off-gas. Dry nitrogen is thepurge gas to be used for transport of TCE volatilized from ex-perimental soil samples. The flow rate will be between 10 and100 cu cm per minute, and will be fixed as part of the shakedownprotocol.

Purge gas passing through the sample tube moves directlythrough the Hall furnace and Hall detector for quantitativemeasurement of HCI conductivity, which is proportional to theTCE carried off. This is a direct adaptation of a standard meth-od for TCE analysis.

Time zero is when the purge gas flow through the sample tube __

commences.

224576A

A-26

Since a continuous flow of purge gas is not possible throughthe detector, a sampling valve is present in the tube exit lineso that off-gas samples can be run through the detector at fre-quent, programmed intervals. Each of these gas samples lasts fora finite period of time (to be determined during shakedown). Themass of the TCE recorded for each gas sample over the samplingtime represents a short-duration rate of TCE volatilization.The total mass of TCE or the cumulative mass of TCE volatilizedover time is the integral under the curve connecting each of theindividual sample values.

4.2 Baseline information requirements and procedures.

4.2.1 Bulk soil. The characteristics of each of the soilsis required. (This list was shown in Table 3.) Each parameter inthe list has a bearing on either the rate of volatilization ofTCE or on questions concerning full-scale materials handling ofTCE-contaminated soils.

The equilibrium moisture content of each soil is required.The value, called We, has relevance to theoretical models ofwater volatilization and may have a bearing on the rate of TCEvolatilization.

A drying curve is required for each soil at each moisturebeing tested. This curve is generated using a technique calledthermal gravimetric analysis. The soil is heated at constanttemperature. The amount of water driven off is measured by theloss in weight of the soil sample. The number of drying curvesneeded is:

(3 soil types) x (2 moisture contents)x (3 temperatures) = 18.

- -The need for this information stems from the unpredictableinteraction between water remaining in the soil and the rate ofTCE volatilization. An example of one type of drying curve isshown on Figure 4.

23

4576A

A- 27

,_V - °

II

!L!

0

0

0

E

.r

•4 ,t, Time

Note: This curve is generated at constant tempera-ture gravimetrically. Values of Ww can bepicked from the curve as raw data if needed formodelling the TCE volatilization process.

Figure 4. One type of drying curve.

24

A-28

_*%* . *,.

The initial TCE concentration of the bulk material prior tospiking is required. There should be no TCE present in the soilobtained from each of the source sites. However, in the eventthere is, its concentration is required. At the same time, itwill be important to determine if there are substances (i.e.,sulfates, carbonates) in each of the soils that interfere withthe TCE analytical method to be employed. In such a case, proc-essing or resampling may be required. However, the appropriatecourse of action will not be clear until the nature of any in- Nterference is determined.

4.2.2 Experimental samples, soil and off-gas. The behaviorof each of the actual soils to be used in the experimental ap-paratus is required. Blanks of TCE-free soil and TCE-free purgegas will be run to determine if materials causing interferencewill appear. This information will be obtained during the shake-down phase of the project and will result from applying exactexperimental procedures to the blanks.

4.3 Experimental information requirements and procedures.

4.3.1 Experimental information.

4.3.1.1 TCE volatilized. The mass balance on TCE around anexperimental soil sample will depend on accurate analysis of TCEin a duplicate sample immediately prior to the experiment andaccurate GC/Hall cell analysis of TCE volatilized with time dur-ing an experiment. The method for determining TCE in a duplicatesample prior to experimentation will be USATHAMA Method 2-J. The! Hall cell will provide a concentration of TCE captured duringperiodic sampling of the off-gas from the sample tube over a fewseconds. It is anticipated that the trace from the GC will be -

reduced to a set of curves similar to those on Figure 5. Thecumulative TCE volatilized will be the integral over the collec-

- -, tion of individual sample curves.

4.3.1.2 Moisture content. A continual check on the actualmoisture content of an experimental sample will be made in thecourse of each experiment by elevating the sample tube (if re-quired) to drying temperature (1100C) at the end of each run.The behavior of water driven from the sample will be derivedfrom drying curves performed previously on duplicate samples.

254576A

A- 29

.2

.o 0-~ -0

)Crz to

000u

IL

262

A-30

4.3.1.3 Time. A real-time clock, which is part of the com-puterized data acquisition system on the GC, will record time asan overlay on the chromatograph output chart.

4.3.1.4 Temperature, oven. A thermocouple in the oven willcontinuously monitor oven temperature. A record will be main-tained as a written overlay on the chromatograph output chart.

4.3.1.5 Temperature, soil sample. A thermocouple imbeddedin the soil sample inside the sample tube will be linked to theGC data acquisition system. Soil temperatures will be overlaidon the chromatograph output chart.

4.3.1.6 Purge gas flow rate. This is a preset value cali-brated prior to each experiment.

4.3.2 Experimental protocol. Table 4 shows a simplifiedprotocol that will be followed. Analytical methods have beenreferenced elsewhere. Data analysis is addressed in more detail

-:i in a later section.

27

4576A A-316A*0.* SO sai ~ '

TABLE 4. SIMPLIFIED EXPERIMENTAL PROTOCOL

Soil acquisition Data or informationand preparation Apparatus preparation obtained

1. Coordinate for the collec- 2. Acquire hardware, valves,tion of 200 pounds of repre- and other items required forsentative, uncontaminated linking the various systemsoil from each of the three components.sites (performed by contrac-tor personnel at each of thesites). These will be nonhaz-ardoua materials.

3. Assemble the experimental 3. a Purge gas flow rate. Sam-a apparatus. Perform hydro- pling frequency of off-gas.dynamic testing for system Sample tube assembly and

* -'integrity. Execute a shake- handling techniques. Tom-

down protocol (internal) for perature ramping for TCEdetermination of purge gas residual and moisture con-

flow rate, sampling fre- tent measurements.quency of the off-gas, andother operational parame-

, ters. "

4. Receive raw samples. Processby screening and crushing.Segregate aliquots for char-

acterization and TGA analy-sis. Air dry bulk quantitiesin preparation for spiking.

5. Spike bulk samples with wa- 6. Complete assembly of threeter and TCE. Label accord- days of sample tubes anding to experimental design. thermocouple assemblies.Perform shakedown protocolwith experimental apparat-us.

7. Perform shakedown runs on

Tektronix graphics termi-nal with stepwise linear re-gression routines for use indata analysis. Confirm for-

.. mat and style of input data.Prepare data tape. Designlaboratory notebook datapages to permit fast datatranscription.

S. Perform characterization 8. Time release rates for wa-and TGA analyses. ter at experimental mois-

ture and temperature lev-els. Permits calculationof residual water content,Ww, relevant to volatili-zation kinetics of TCE.Comparative characteristicsincluding total organiccontent and celative claycontent.

9. Execute experimental design. 9. TCE release rates, moisturecontent, and TCE mass bal-ance.

10. Transcribe data onto floppy 10. Significant and not signif-disk and analyze (with icant parameters; kineticstransformations and multi- model.plicative models).

aMumbers relate to protocol paragraphs of the same number.

4576A 28

A-32

%

5. DATA ANALYSIS

5. 1 Proof of concept. The primary objective of the studyis to demonstrate concept feasibility. Therefore, a plot of thetype shown on Figure 6 may be sufficient to illustrate TCE vola-tilization for the test systems used.

A more meaningful graphic, however, would be one that relat-ed TCE residual in soil over time to various operational parame-ters. Figure 6 demonstrates one such plot that may come fromthis study. TCE residual in soil is related over time to the or-ganic content of the test soil.

Figure 7 is along similar lines. It is anticipated that therate of TCE volatilization may be directly proportional to dry-ing temperatures for a period of time, and inversely proportion-al thereafter. This stems from a complex relationship betweenthe partitioning of TCE among soil particles, interstitial wa-ter, and air. As the dry nitrogen carries off soil moisture,there is a possibility that TCE will strongly adsorb onto soil

K- particles and be driven off more slowly.

Since there are no established allowable residual levels forTCE in soil, one objective of the data analysis will be to dem-onstrate how effective thermal removal of the chemical may beunder specific conditions. It is anticipated that the lowest re-sidual levels will at some point in time be more than sufficientto meet ambient residual levels.

5.2 Data analysis. This study was purposely designed tomaximize the amount of information obtained with the minimumnumber of experiments. The analytical technique projected foruse is stepwise linear regression with two-way analysis of vari-ance. Results from such an analysis should indicate the follow-., , ing: .

(a) Significant and relatively insignificant parameters.(b) Insight into how significant parameters may be relat-

ed.(c) Insight into how the rate of TCE volatilization may be

enhanced.(d) In what direction a pilot or full-scale commercial dry-

er design might be made.

29

4576A A-33

0N

w-

C)~

cc.

03

A--

0EM

W0

I L

Increasing temperature

Time

Figure 7. Anticipated relationship between TCEvolatilization and drying temperature.

31

A- 35

Data analysis will be performed using off-the-shelf softwarefor the Tektronix graphics terminal system. Logarithmic trans-formations are anticipated to be required to linearize the data

* according to the various models.

A preliminary (order of magnitude) cost analysis will bemade. It will be combined with a cost/benefit assessment of thequestion of whether to proceed to pilot-scale testing.

- 32

457 6A

A- 36

-Y 71 -7 7. -7 -71' 7 7.- "7 7,77107

6. SCHEDULE

*6.1 Project structure.

6.1.1 Work breakdown structure. The work breakdown struc-ture anticipated for this project involves the following fivetechnical tasks:

(a) Task 1 -- Setup and coordination.(b) Task 2 -- Shakedown.(c) Task 3 -- Experimentation.(d) Task 4 -- Data analysis.(e) Task 5 -- Final report.

6.1.2 Task 1 -- Setup and coordination. The objectives ofthis task include the following:

(a) Coordination, collection, and shipment of bulk quanti-ties of soils.

(b) Final design and acquisition of hardware for an inte-grated test apparatus.

(c) Establishment of work areas and records for the proj-ect.

(d) Contract administration.

6.1.3 Task 2 -- Shakedown. The objectives of Task 2 in-clude the following:

(a) Full operation of the test apparatus.(b) Preparation of all bulk samples.(c) Characterization of all soils.(d) TGA analysis on all soils.(e) Finalization of format and style of laboratory note-

books.(f) Specification of purge gas flow rate.

This task will be conducted within the bounds of a formalprotocol for those operations requiring close supervision andtimely execution. This protocol will be internal and will not bea contract deliverable. Upon completion of this task, all prepa-rations for experimentation will have been completed.

6.1.4 Task 3 -- Experimentation. This task is the execu-tion of the designed experiment.

33

4576A K.--.>.;&A- 37

..............................................

6.1.5 Task 4 -- Data analysis. Work will begin on thistask during Task 2 and will accelerate upon completion of ap-proximately a third of the experimental design. It may be neces-

* sary to change experimental conditions upon completion of two-thirds of the design to accommodate new information or indica-tions that the original parameter values were not optimum.

* Therefore, partial analysis of experimental data will occurthroughout Task 3.

-. 6.1.6 Task 5 -- Final report. This is a contract delivera-ble and will incorporate a complete record of the project. Botha draft final and a final report will be prepared.

6.2 Project schedule. Figure 8 provides a tentative Gantt* chart showing that the anticipated execution time to produce a

draft final report is approximately 19 weeks.

V3

4576A A-3

-.-

40

.2.

0

0 06

U) U)U

0M CV C

______ 35

-A-3

DISTRIBUTION LIST

Defense Technical Information CenterCameron StationAlexandria, VA 22314 12

CommanderUS Army Toxic and Hazardous Materials AgencyAttn: AMXTH-ESAberdeen Proving Ground, MD 21010-5401 2

Defense Logistics Studies Information ExchangeU.S. Army Logistics Management CenterFort Lee, VA 23801 2

I

FILMED

D .c

2.2.'Tl T~ S7o2'