AD-A244 010
li lu liil! Llli iiiii lilijii
Rtport No. AIIXTH-TE-Cfi>6801«U^
Mm tfp on
U SATHAM A
U.S. Aftny Toxic ami Hazarctoua Matartaia Agency
LABORATORY-SCALE SOIL
WASHING TEST ON
ROCKY MOUNTAIN ARSENAL
BASIN F MATERIAL
(TASK OflOER NO. 8)
August 1968
Contrset No. DAAK11-89-04)008
Arthur 0. Uttlo, Inc.
Acom Parte
CamhrMgs, ltassa«husstts 0214g>2390
MTA Rsmsdial Rtsourcst, tno.
Qoldtn, CotoraCo 90401
U.S. Army Toxic and
Htzardoua Matorlais Agency
Procats Dtvak^smsnt Brar^
Absfdaan Prov^ Qround, 10} 21010-9481
DISTRiBUTION UNUiyBiTED
The views, opink)ns, nx^or flrejtnQS contak^ in th(S report shoutd not b« constrimd
as an official Dapartmsnt of tn# Amry position, policy, or dsdsion, unlass so dssig-
naM by other docunwmation.
The use of trade names in tttis report cfoes not constitute an official endorsement or
approval of ths use of such corrsnercial prcducts. Thte report may not be citsd for
purposes of advertisement.
Final Report to
United States Army
Toxic and Hazardous
Materials Agency
August 1988
Laboratory-Scale Soil Washing
Test on Rocky Mountain Arsenal
Basin F Material
(Task Order Number 8)
Final Report
A.A. Balasco
Program Manager
J.l. Stevens — Task Leader
J. W. Adams
D.L. Cerundolo
S. Rickard (MTA Remedial Resources)
P.B. Trost (MTA Remedial Resources)
Principal Investigators
K « •
n
» -
1
Distribution Unlimited
91-18796
'Mu* iM iMM M. ^ .
Arthur D. Little, Inc.
Contract No. DAAK11-85-0-0008
Reference 54148
USATHAMA Reference AMXTH-TE-CR-88016
k/i .J.^
UNCLASSIFIED
StCuRiry CLASSIFICATION OF '4.$ faGE
REPORT DOCUMENTA’ilON PAGE
la. REPORT SECURITY CLASSIFICATION
UNCLASSIFIED
lb. RESTRICTIVE MARKINGS
2a security CLASSIFICATION AUTHORITY
i OISTRIBUTIOW/ AVAlLAaiUTY OF REPORT
I’nT
20 DEC’wASStfiCATiON / (XJWNGRAOiNG SCHEDULE
4 PERFORMING ORGANIZATION REPORT NUMBERIS)
Reference: 54148
S MONITORING ORGANIZATION REPORT NUM8ER(S)
AM3CrH-TE-CR-88016
6a NAME OF PERFORMING ORGANIZATION
Arthur D. Little, Inc.
66 OFFICE SYMBOL
{If
7a. NAME OF MONITORING ORGANIZATION
U.S. Army Toxic and Hazardous Materials Agency
6< ADDRESS (Oty, Star*, tnd ZlfCodti
Acorn Park
7b ADDRESS (Cty, Sratr and ZIP Code)
Attn: CETH-TE-D
Cambridge, Massachusetts 02140-2390
Aberdeen Proving Ground, Maryland 21010-5401
3a. NAME OF FUNDING, SPONSORING
8b OFFICE SYMBOL
9 PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER
ORGANIZAT'ON U.S, Army Toxic i
Ilf ipplkibl*)
Contract Mo. DAAK11-85-D-0008
Hazardous Materials Agency
CETH-TE-D
Task Order Mo, 8
ADDRESS (Gfy, Sfjf#, jnd/;RC oO*) 'Q SOURCE OE F 'JNOING NUMBERS
Attn: CETH-TE-D PROGRAM
Ab«irdeen Proving Ground. Maryland 21010-5401
! ' Title (intluot SKunty CUtiitKition)
Laboratory-Scale Soil Washing Test on Rocky Mountain Arsenal Basin F Material
PROJECT
task
NO
NO.
8
WORK UNIT
’2 ’ERSONAc aothor(S) a.a. Balaaco, J.I. Stevens, J.W. Adana, D.L. Cerundolo, S. Rickard and
P.'
’ 3a type of report
F inal
14 DATE OF REPORT {Y«*r, Mwitfi, Diy) tlS PAGE COUNT
31 August 1988 I
' T COSATI COOES _ I '9 SUBJECT TERMS (Continue on reverj# if neceoaty *nd identify by b/<xk number)
GROUP I SUB-GROUP |* Rocky Mountain Arsenal • Basin F • Soil Washing
• Technology Evaluation • Technology Development/Testing
• Soil Decontamination • Site Remediation _ _
'■■f ABSTRACT (Continue on reverse if neceiwry tnd identify by bi<xk number' The U.S. Amy Toxlc and Hazardous
Materials Agency (USATErVMA) under its program for Innovative Technology Development for
Rocky Mountain Arsenal (RMA) issued Task Order Nc. 8 under Contract No. DAAK11-85-D-0008 to
Arthur D. Little, Inc, to evaluate and rank Innovative technologies for applicability in
treating Basin F Materials at the Arsenal. As a result of that ranking (Final Report on
Evaluation/Selection of Innovative Technologies for Testlrg with Basin F materials prepared
by Arthur D. Little), soil washing was among the technologies chosen for laboratory-scale
testing and MIA Remedial Resources, Inc. (MTARRI) was awarded a subcontract to perform the
work.
To initiate the evaluation of the soil washing process, MTARRI designed and carried out a
laboratory program to determine; the applicability of the process; and the conditions that
would remove both the organic and Inorganic contaminants from Che Basin F materials to yield
a clean soil that could be placed In a fill on-site. The process was Chen proven by a
demonstration run, at the bench-scale, with Arthur D. Little personnel observing and
21 ABSTRAa SECURITY CLASSIFICATION
IDICLASSIFIED
22C. OFFICE SYMBOL
CFTH-TE-0
JO aiSTP'SUT ON / availability of abstract
□ UMCLASSIFIED/'JNLIMITED □ SAME AS RPT □ ODC USERS
,:2j name of RESPONSIBLE INDIVIDUAL
Craig W. MacPhee
DO FORM 1 473, 94 mar 33 APR edition mty b* U5*d until «»h»uJ1*d
All oth*f •ditiom *rt ob«>l«l*
UNCLASSIFIED
UNCLASSIFIED
SECURITY CLASSIFICATIOM Or THIS PAGE _
sampling che demonstration run product screams. Using Che laboratory and deii»nstration run (
data, a process flowsheet and material balance was produced for a plant to treat approximate!}
950 tons per day of contaminated Basin F material.
Laboratory Developiwnt Program - MT^RI had previously shown that the soil washing process I
coTild remove organics and inorganics from soils; however, no work had been done with a mater¬
ial having the particular contaminants contained in Basin F. Therefore, a laboratory develop¬
ment program was required to establish the necessary physical and chemical conditions that |
would remove these contaminants from the Basin F material.
At the time the scope of work for this program was developed, there were no guidelines avail- ,
able on the degree of contaminant removal that the soil wash should achieve. Therefore, the |
laboratory program's objectives were to: develop a process to remove as much as possible of
all the contaminants (both organic and inorganic); establish the technical feasibility of the
process; and determine the operating parameters within some ranges. Therefore, the labora¬
tory's sope of work was limited to process development.
The results of the laboratory development program established a process that can eliminate '
the majority of the aldrin, and presumably the other organic contaminants of concern. To
accomplish this removal, an organic prewash of an aqueous slurry of Basin F material is needed
prior to the flotation.
During the laboratory test work no unusual problems or conditions were evident that would
cause difficulties when the process is applied on a large scale. Overall, the laboratory pro¬
gram was successful in developing a process to clean up Basin F material. It now only remainf
CO demonstrate this when Che conditions established, from the prior test work, are employed
in a cesc run continuously from start to finish. The acid wash section that was initially
I assumed not to need testing is also a part of this complete demonstration run. \
As part of MTARRI's task, a laboratory demonstration run of the process developed during the I
laboratory test program phase was carried out. Arthur D. Little personnel observed the demon¬
stration run and were responsible for the collection of samples and their analysis. Data
generated during the demonstration run were used as the basis for developing the process flow¬
sheet and material balance. Sample collection and analysis by Arthur D. Little was to be
detailed, in that major compounds of concern were to be tracked, as far as practicable, I
throughout the entire process. Analytical methods used were approved and certified by
USATHA.MA.
In addition, the sampling and analytical program was performed to obtain sufficient data to
confirm certain aspects of the process chat had not been studied extensively during the lab-
orator;/ development program. For example, the number of stages in Che organic wash section
and organic flow requirement were to be evaluated from Che demonstration run data; as was Che
need for a final acid wash of the Basin F material.
At the time this program was developed, there were no guidelines available on the degree of
contaminant removal that Che soil washing process should achieve. Therefore, our objective
was to remove as much as possible of all Che organic and inorganic contaminants. This caused
us CO use a more extensive process during the demonstration run than was necessary based upon
the data subsequently obtained from the demonstration run. Therefore, the soil washing pro¬
cess for the full-scale treatment of Basin F material has fewer unit operations than were
employed in the demonstration run. It has been assumed that if the clean washed Basin F
material meets Che criteria set forth in the EPA's proposed toxic characteristic leaching
procedure (TCLP), we would have achieved Che required goal of contaminant removal.
Overall, the demonstration run showed that soil washing of Basin F material can eliminate the
contaminants, both organic and inorganic,and yield a final clean soil that passes or exceeds
the proposed TCLP criteria sec by the EPA.
During Che demonstration run no problems were encountered that were insurmoutable or would
make this a difficult process to implement on a large scale. The required equipment is
currently manufactured so no new equipment design or development is required. Reagents used
„ are, all, available in large quantities.___ (Continued on separate page) _
I'NCLASSIFTED
5CCU«fTY classification OF THIS PAGE
: UNCLASSIFIED
CLAssir.CAT.o,, or THIS P*=g OMO rnnMnupH^ _
Although some data was not obtained during the demonstration run aad some problems with the
I calculated material balances were observed, these were resolved by the described assumptions
and adjusting the mass flow and analysis. These adjustments were necessary so that material
balances could be developed and equipment sized, but in no way detracts from the conclusion
that this process will clean up Basin F material. In addition, analysis of the data from the
demonstration run showed where some process simplifications could be made. These changes
were incorporated into the full-scale process flowsheet.
Full-Scale Basin F Soil Washing Process - Using the data collected and numerous flowsheet/
material balance studies, a processing plant was designed that will produce clean soil (as
defined by the EPA's TCLP procedure) chat can be returned to a fill an-site. This plant
would employ equipment currently available and reagents that are readily available in large
quantities.
The largest flow stream in the plant is about 800 gpm (slurry to the organic wash area) and
the largest piece of equipment, other than tanks, are the four flotation cells at 1500 cu.
ft. each.
Overall, a full-scale soil washing plant would be easy to operate, require a minimum of
maintenance, and would have an above average on-line factor.
Instrumentation and control loops would be minimal. The major control would be tank levels, j
I
The plant would be relatively safe to operate since neither high pressure or high tempera- j
tures are employed. The plant would be environmentally acceptable since it is temporary and
all emissions, air, water, and soil would comply with current emission standards.
If in the final evaluation of Innovative technology to clean up Basin F, soil washing con¬
tinues to be a viable technology both from an environmental and cost standpoint, additional
work is needed to finalize Che process, engineer the plant and finalize the costs for the
overall project. This additional work would be in three steps: laboratory studies, pilot
plant demonstration, and engineering design.
Additional laboratory studies would consist of two parts. First, the process would be
defined in detail. For e.xample, can toluene be eliminated from the wash solvent, and is
there a better solvent system that would make the distillation system more effective in
I eliminating the organic contaminants. Second, data for engineering design needs to be
I gathered to more accurately size the equipment. Such items needed are size distribution of
j the feed material to a detailed analysis of flotation variables for scale-up design.
Pilot/demonstration testing would be required since it would not bs prudent to scale up the
proposed process from laboratory bench-scale studies to the full-scale plant (950 cons per
day). Therefore, a pilot/demonstration plant should be built and operated for two to four
months. This small-scale plant (1000 Ib/hr) would be built based upon additional detailed
laboratory work which would fi.x the flowsheet so little or no equipment arrangement testing
will be needed. Therefore, this would be more a demonstration plant with only limited test¬
ing on Che effect of process water recirculation, distillation and quality of the clean soil
produced. All data needed for scaling up to a full-scale plant will be obtained by a month's
run at steady state conditions.
During the preceding two steps, the work should be subjected to engineering optimization as
Che data becomes available. This would require establishing an engineering team to assist
in program design prior to the start of either phase of work. During the additional data
collection the engineers could evaluate data and make recommendations for changes, new tests,
additional data so that upon completion of the pilot/demonstration mn there would be
sufficient and complete data for the final and detailed engineering of the full-scale soil
washing process plant.
UNCLASSIFIED _
SeCUSITY CS.»SSIF1C*T10M Of THIS e»GE
This report was prepared by MTA Remedial Resources, Inc. for
Arthur D. Little. Inc. in fulfillment of a requirement for
Task Order Number 3 under Contract DAAK11-85-D-0008.
SUMMARY
TABLE OF CONTENTS
Page No.
S-1
1. INTRODUCTION.1
1.1 Laboratory Development Program...1
1.2 Demonstration Run.....2
1.3 Program Analysis and Engineering Design.3
2. DESCRIPTION OF TEST FACILITIES.4
2.1 Analytical Support.4
2.2 Health and Safety. 4
2.2.1 Worker Protection.4
2.2.2 Environmental Protection...6
3. LABORATORY DEVELOPMENT PROGRAM PROCEDURES AMD RESULTS.7
3.1 Introduction.7
3.2 Sample (Basin F Material)...8
3.3 Surfactant Scoping Tests. ...8
3.3.1 Results of Surfactant Scoping Tests..... 9
3.4 Flotation/Chemistry Variables Evaluation. 9
3.4.1 Discussion of Results of Flotation Variables Evaluation.... 9
3.5 Basic Flotation Process Modification Studies...12
3.5.1 Discussion of Results for Flotation Process Modification..14
3.6 Solvent Prewashing Process Evaluation.14
3.6.1 Solvent Prewash Test Procedure.14
3.6.2 Discussion of Results of Solvent Washing Evaluation.16
3.7 Summary and Conclusions of the Laboratory Development Program.... 17
4. DEMONSTRATION RUN.19
4.1 Introduction.19
4.2 Demonstration Run Procedure. 20
4.2.1 Organic Wash Steps.20
4.2.2 Flotation Step. 22
4.2.3 Acid Wash Steps. 23
4.2.4 Demonstration Run Samples. 23
4.2.5 Problems During Demonstration Run. 24
4.3 Demonstration Run Product Analysis and Material Balance...2A
4.3.1 Analytical Results.24
4.3.2 Material Balance Calculations.26
4.3.3 Discussion of Sampling, Analytical and Material
Balance Results.29
4.1 Demonstration Run Data Adjustments.29
4.4.1 Assumptions Used to Adjust Data.29
4.4.2 Discussion of Adjusted Results.32
4.5 Process Calculations and Flowsheet Analysis.33
4.5.1 Acid (Counter-Current Decantation) Wash Section
(Inorganic Contaminant Removal).33
4.5.2 Organic Wash Section.33
4.5.3 Non-Pesticide Organic Contaminant Removal.35
4.5.4 Heat Balance. 35
4.6 Summar-y of Demonstration Run. 36
1
5. BASIN F SOIL WASHING PROCESS.37
5.1 Introduction.37
5.2 Process Description.39
5.2.1 Feed Preparation Area.39
5.2.2 Organic Wash Area.41
5.2.3 Organic Filtration.41
5.2.4 Organic Distillation.42
5.2.5 Flotation Area. 42
5.2.6 Froth Liquid/Solid Separation.42
5.2.7 Clean Soil (Tails) Liquid/Solid Separation.43
5.2.8 Carbon Adsorption.43
5.2.9 Wastewater Treatment.43
5.2.10 Volatile Organic Control.44
5.3 Summary of Basin F Soil Washing Process.44
6. RECOMMENDATIONS FOR FUTURE WORK.46
6.1 Additional Laboratory Studies.46
6.2 Pilot/Demonstration Testing. 46
6.3 Engineering Optimization.47
APPENDIX A
• Soil Wash Plant Flowsheet for Basin F Material
• Table A - Material Balance Basin F Soil Washing Plant
LIST OF TABLES AND FIGURES
TABLE Page No.
1 Surfactant Screening Test Results...10
2 Flotation Process Variable Tests. 11
3 Organic Addition to the Basic Flotation Process.13
4 Solvent Washing-Flotation Process Parameter Studies.15
5 Important Process Parameters for Solvent Prewasli/Flotation
for the Removal of Pesticides from Basin F Soils.18
6 Demonstration Run Product Analysis Data.25
7 Pesticides Material Balance Calculations.27
8 Inorganic Material Balance Example Calculation........28
9 Calculated and Estimated Analysis and Material Wts.31
10 Generalized Material Balance for Basin F Soil Washing Process.40
FIGURE
1 Manufacturer' s Drawing of Flotation Cell.5
2 Demonstration Run Process Steps and Material Flow.21
3 Generalized Flowsheet for Basin F Soil Wash.38
SL’WARY
The U.S. Army Toxic and Hazardous Material Agency (UflATHAMA) under its program
for Innovative Technology Development for Rocky Mountain Arsenal (RMA) issued
Task Order No. 8 under Contract flo. DAAK-ll-85-00008 to Arthur 0. Little, Inc.
to evaluate and rank innovative technologies for applicability in treating
Basin F Materials at the Arsenal. As a result of that ranking (Final Report on
Evaluation/Selection of Innovative Technologies frr Testing with Basin F
Materials prepared by Arthur D. Little) soil washing was among the technologies
chosen for laboratory-scale testing and MTA Remedial Resources. Inc. (MTARRI)
was awarded a subcontract to perform the work.
To initiate the evaluation of the soil washing process. MTARRI designed and
carried out a laboratory program to detenmne: the applicability of the
process: and the conditions that would remove both the organic and inorganic
contaminants from the Basin F materials to yield a clean soil that could be
placed in a fill on-site. The process was then proven by a demonstration run,
at the bench-scale, with Arthur 0. Little personnel observing and sampling t!ie
demonstration run product streams. Using the laboratory and demonstration run
data a process flowsheet and material balance was produced for a plant to treat
approximately 950 tons per day of contaminated Basin F material,
Labcretcrv Teve I comeri'f -rccrem
MTARRI had previously snown that the soil washing process could remove
organics and inorganics from soils; however, no work had been done with a
material having the oarticular contaminants contained in Basin F.
Therefore, a laboratory development program was required to establish the
necessary physical and chemical conditions that would remove these
contaminants from the Basin F material.
At the time the scope of work for this program was developed, there were no
guidelines available on the degree of contaminant removal that the soil
wash should achieve. Therefore, the laboratory program's objectives were
to; develop a process to ranove as much as possible of all the
S-1
contaminants (bovh organic and inorganic); establish the technical
feasibility of the process; and determine the operating parameters within
some ranges. Therefore, the laboratory’s scope of work was limited to
process development.
The results of fhe laboratory developrent program established a process
that can eliminate the majority of the aldrin, and presumably the other
organic contaminants of concern. To accomplish this removal an organic
pre-wash of an aqueous slurry of 3asln F material is needed, prior to
the flotation.
Turino the laboratory test work no unusual problems or conditions were
evident *hat would cause difficulties when the process is aoplled on a
large scale, "verall, ti^e laboratory orogram wrss successful in developing
a process to clean jo Basin ~ material. It now only '“emains to demonstrate
this when the conditions established, from the prior test work, are
employed in a *es* run continuously from start to finish. The acid wash
sec*ion -^hat was ini-^ia!!/ assumed no* to '^eed testing Is also a Dart of
this ccmplete Semen s-'-at i on r^n.
Tem<:;nst'~a'*’Icn "“est
As 030"^ of '.'■'iP^l's Tasn, 3 130003*00'/ demonstration run of the process
developed dur't-; *ne laccrat'-rv ■"est pregran phase was carried out.
■*rThur T. Lit'^ie personnel reserved the demonstration run and were
responsible ‘cr the ooMec^ion of samples and their analysis. Data
cenerateJ durin-; rne demens-'ra'^ior. "-un were used as th-e basis for deve I op¬
ine ■!'ne process flowsheet a^d '"a'eo’al balance. Sample collection and
analysis bv J. Li'^tle was to be detailed, in that major compounds
of concern we^e po fracked, as far as practicable, tnrouchout the
entire process. Anaivtical methods used were approved and certified by
In ’dditicn, toe sampling and analytical program was performed to obtain
sufficient da'fp confirm certain aspects of the process that had not
been studied ex+ensively during the laboratory development program, "or
exan’ole, **'e 'iurte'" of staces 'f’f'.e organic was ssctior and organic
‘icw r^o'j i rereor fo Ce evaluated from the deftcnstrat ion run data; zs
was t-.e need ^zr a final acid was^ of the Basin F nateriai.
At the time this program was developed, there were no guidelines available
on the degree of contaminant removal that the soil washing process should
achieve. Therefore, our ooiective was to remove as much as possible of all
the organic and inorganic contaminants. This caused us to use a more
extensive process during the demonstration run than was necessary based
upon the data subseguently obtained from the o monstration run. Therefore,
the SOI' washing orocess for the full-scale t' ?atment of Basin F material
has fewer unit ocerat’ons than were eniployed ’n the demonstration run. It
has been assumed that ’f the clean washed Rasm F material meets the
'-o i~‘'- rr-ocsed ’•’cv I * c‘'arac*eni s*ic ’e-ac"'''g
jc^'evad -“cuireJ xal ccr^omi-ent
'•-•meva .
Ove'^a'', the de'^onstra"'o" '•un showed that so’l washing of Basin F material
can eliminate the contamiriants, both organic and inorganic and yield a
final e'ean soil passes O'- mceeds the proposed ’CLP criteria set by
the F ^ A,
During the demonstraticn nun no proo'ems were encountered that were
1 nsurnountao> or wou’d -axe th’s a difficult process to implement on a
large snx’e, ’f'e negjtred »gu'pmen; is Curnently manufactured so no new
itdu'oment ies'gn or ieve’ocmem*. is n»quired. Reagents used are all
avai'ab'e ’n ' juan ’ t ••.
Althougn some iatJ was ''Ot ob'ained during the demonstration run and Seme
oroo'iJ-'S wi'.n t’-u :a'cj'at»‘d -atenM’ balances were observed, these were
*y .'i-sr ■ Oe',' »s , jcip'i ,;ns an.} adjust'ng the mass flow and
ina’ys’s. sdjustm^n*.-; were necessary so that material balances could
be de..-ioDed md eouipmer* s'/ed, but m np »ay detracts from the
conclusic<^ that this process M dean jp rfasin P m3*eri3!. In addition,
analysis of the data from the dercnstration run showed where sens process
simplifications could oe made, ''hese changes were incorporated into the
full“scale process ♦lowsheet.
Full-Scale oasi^ ~ 'dl iVasni^o ^-ocess
Using the data collected and numerous flowsheet/matenal balance studies, a
processing plant was designed that will produce clean soil (as defined by
the EPA's TCLP procedure) that can be returned to a fill on-site. This
plant would employ equipment currently available and reagents that are
readily available in large quant'l'es.
The largest flow stream in the Dl.jnt. is about BOO gpm (slurry to the
organic wash area) and t.he largest pi>'ce of equipment, other than tanks,
are the four f’otat'on cells at 1500 cu. ft. each.
Overall, a full-sca'e soi' washing plant would be easy to operate, require
a minimum of maintenanre, and would have an above average on-lme factor,
.nstrumenta 11 on and control ooos would be minimal. The majo** control
would be tank levels.
^he plant would be rwlatwely safe to operate since neither high pressure
or high tempe'-atures are employed. The plant would be environmenta1ly
acceptable since it 's temoorary and all emissions, air, water, and soil
would comply with Current emission standards.
""‘r: r- ' C sr
!f in the final evaluation of innovative technology to dean up Basin f, soil
washing continues to be a viab'e technology both from an environmental and cost
standpoint, additional work is needed to ^maliie the process, engineer the
plant and final ue the costs for the overall project. This additional work
would be in three steps: laboratory studies, pilot plant demonstration, and
engineering design.
Addlticnal laboratory studies «ou!d consist of two parts. First, the process
would be defined in detail, "or example, can toluene be eliminated from the
wash solvent, and is t'-ere a better solvent system that would make the
distiNation system "ore effective in eliminating the organic contaminants.
Second, data for engineering design needs to be gathered to more accurately
size the equioment. Such items needed are size distribution of the feed
material to a detailed analysis of flotation variables for scale-uo design.
Pilot/demonstration *e 5 ming would be required since It would not be brudent
*0 scale ub tne brcoosed Drqqess ^rcm laborator"/ bench-scale studies co *ne
‘j!!“SC 3 le bian"*" *cns cer davN ’"hena^cre, a bi iqt/demonstrat iqn slant
should be bui't and ;bera'*ed ‘or •■w" *c *cur mcntn$. "his s^all-scale blar*
viSCC ib/hr) would be bui'* based ubcn additional detailed labori*crv work
wn I :h would *'< “"a <;'ws'-eet ';*r!a or no eculomenr .jrranoemen* *e3*lno
*N i o.v ■'eeded. t'-.a-a^ore, t-’s would be mcra a ’demonstration oian* wl*'* d-n i
:;-;*a.d *est'"': “"e ef*'?:* crooess water reo i rcu I at ion, distillation an
duaiitv o‘ t-a -lean soil qroduced. A|i data "ee^ded for scaling uc *0 a full
s:.a'e clan* wNi be cb*'a;'ed bv -a ~cnth'j run at 3*eadv s^ate condit!o"s.
Ounng the preceding two steos, the work should be subjected to engineering
optimijation as the data becomes available. This would rggmre
establishing an engineering team to assist in program design prior to the
start of either phase of work. During the additional data collection the
engineers couM evaluate data and make recommer.dations for changes, new
.eSvS. additional data so that uoon completion of the odlot/demonstration
run there would be sufficient and rnmpleta data for the final and detailed
engineering of the fuH-sca’e soil washing process plant.
7. INTRODUCTION
The U,S. Army Toxic and Hazardous Material Agency (USATHAMA) under its program
for Innovative Technology Development for Rocky Mountain Arsenal (RMA) issued
Task Order No. S under Contract "Jo. DAAK-11-85-00008 to Arthur 0. Little, Inc.
to evaluate and rank innovative technologies for applicability in treating
Basin F Materials at the Arsenal. As a result of that ranking (Final Report on
Evaluation/Selection of Innovative Technologies for Testing with Basin F
Materials prepared by Arthur D. Little) soil washing was among the technologies
chosen for 1aborato'"y-sca!e testing and MTA Remedial Resources, Inc. (MTARRI)
was awarded a subcontract to perform the work.
To initiate the evaluation of the soil washing process, MTARRI designed and
carried out a laboratory program to determine: the applicability of the
process: and the conditions that would remove both the organic and inorganic
contaminants '^rom the Basm F materials to yield a clean soil that could be
placed in a fill on-site. Tne process was then proven by a demonstration run,
at the bench-scale, with Arthur D, Little personnel observing and sampling the
demonstrat ion run product streams. 'Jsmg the laboratory and demonstration run
data a process flowsheet and material balance was produced for a plant to treat
approxmately 950 tons per day of contaminated Basin F material.
"rom the laboratory data, demonstration run results and the flowsheet and
material balance calculations, capital and operating costs were then
developed. This cost data is reported m a separate memorandum report for
incorporation into the Final Firoject Report being prepared by Arthur 0. Little.
Inc.
^^ L aborcatpry Development ^n ggram
"■^TAPRI had previously shown that the soil washing process could remove
organics and inorganics from soils: however, no work had been done with a
material having the particular contaminants contained in Bas in F.
’herefcro. a laboratory development program was required to establish the
necessary phys’cal and chemical conditions that would remove these
contaminants from the Basm F material.
At the time the scope of work for this program was developed, there were no
guidelines available on the degree of contaminant 'emoval that the soil
wash should achieve. Therefore, the laboratory program's objectives were
to: develop a process to remove as much as possible of all the
contaminants (both organic and inorganic); establish the technical
feasibility of the process: and determine the opera*”'ng parameters within
some ranges. Therefore, the laboratory's scope of work was limited to
process development.
The data from the laboratory program and demonstration run along with
engineering judgment were used in the preparation of the preliminary
process flowsheet and the specification and selection of equipment for a
full-scale (950 ton per day) treatment plant. The flowsheet and equipment
specified were then used to estimate the capital and operating costs for
the full-scale soil washing plant.
Process conditions established during laboratory development program are
oresented in Section 3 of this report. The optimum conditions were then
used in the subsequent demonstration run.
1 .2 Oemonstraticn
During the week of April 6, 1987, 'irARPI carried out the demonstration run
using the techniques and reagents which pruvided the best removal of the
contaminants from Basm F material based upon the laboratory program.
During this demonstration period, Arthur 0. Little, Inc. personnel observed
the run and collected the samples to be analyzed to determine clean up
effectiveness and to obtain data for materia! balance. These samples were
analyzed in the Arthur 0. Little. Inc. laboratory which has been certified
by dSATHAMA for the chemical compounos of concern. The results were
reported to ‘''TARRI, and are incorporated in the discussion of the
demonstration run (Section A),
- 2 -
1.3 Program Analysis and Enqineering Design
The scope of the laboratory program did not provide for the collection of
all the data necessary to design a full-scale plant since this program was
a preliminary technical assessment of the soil washing process as applied
to Basin F material. In addition, some data was not obtained during the
demonstration run. These factors created the need for extensive treatment
of the data, flowsheet analysis, and engineering estimates to complete the
process evaluation. The calculations and assumptions used for this data
treatment are discussed briefly as part of the evaluation of the
demonstration run. The results of the overall analysis of the program data
were used to complete the detailed flowsheet and to develop the capital and
operating costs for a Basin F material washing plant.
- 3 -
2 . DESCRIPTION OF TEST FACILITIES
The laboratory development program and the bench-scale demonstration run were
carried cut in a 300 square foot laboratory. For health and safety concerns
this laboratory, with its contained equipment, was totally dedicated to this
single project.
Other than the flotation machine, all laboratory procedures were carried out in
standard glassware and with conventional laboratory equipment, such as pH
meters and balances. The flotation machine used was a Denver Equipment D-12
Lab float machine (Figure 1 is the manufacturer's drawing). In all tests, a
1000 gram tank was used. This machine provides agitation and aeration, to
separate hydrophobic materials from the bulk of a slurry.
2.1 Analytical Support
The laboratory develcoment program was supported by three different levels
of analysis. Initial analyses used methylene chloride extraction in
Soxhlet extractors to produce data to determine the degree of removal of
the gross organics. This was followed with detailed analysis, by an
indeoendent analytical laboratory, to determine the disposition of aldrin
in the test products. Also. ’^TAPPI used its own analytical capabilities to
aid in the laboratory program.
For the demonstration run, analytical work was performed by Arthur D.
Little using USATHANA certified procedures.
2.2 '-lea 1th and Safety
Handling and containment of 3asin F material was necessary to safeguard
workers and the environment. Specific procedures followed are summarized
below.
2.2,1 .-iorker P>'otection
When working with, and around 3asin F material, several steps were
taken to protect the employees. Medical surveillance was established
for the employees who would be in contact with the material by
- 4 -
FIGURE 1
TT!
' “I^rt
X
pre-exposure physical examinations, including extensive blood
chemistry analysis. Upon completion of the work, they were examined
once again to make sure they had not been exposed to excessive levels
of chlorinated organics.
While working in the dedicated laboratory, complete changes of fresh
clothing were provided daily. The clothing was covered by Tyvek*
coveralls and Tyvek*shoe covers. Hands were protected by two gloves,
both of which were solvent resistant. Respirators approved for
organic vapor and dust protection were utilized at all times in the
laboratory, as were safety glasses.
2 . 2.2 Environmental Protection
The dedicated laboratory was provided with a negative pressure
ventilation system to prevent release of any toxic materials outside
the laboratory. The heating and cooling system was isolated to
eliminate diffusion. All equipment in the laboratory remaitied in the
laboratory for the duration of the entire program. Lab wastes and
cleaning materials were placed in a sealed container in the
laboratory; when the laboratory is decontaminated all waste will be
drummed and shipped back to RMA.
- 6 -
3. LABORATORY DEVELOPMENT PROGRAM PROCEDURES AND RESULTS
This section discusses the work completed during the laboratory development
program to evaluate the soil washing process on Basin F material.
3.1 Introduction
MTARRI's prior work on other contaminated soils provided base line
information on the soil washing process and reagents which had predicable
pronability ot success on Basin F material. Initially, it was assumed that
the inorganic contaminants would be easily removed using a counter-current
acid wash: this assumption was based upon previous work with soils
contaminated with the same type of inorganic contaminants. Therefore, this
step was not examined during the laboratory phase of the program but would
be te'ted during the demonstration run.
Previous work by MTARRI on other projects indicated that organic
contaminants could be freed from soil particles and subsequently separated
using froth flotation by reacting an aqueous slurry of contaminated soil
with a mixture of caustic, silicate and a surfactant. This served as the
starting point of this investigation. In order to expedite the program and
control costs, it was assumed that if the major contaminant (aldrin) could
be removed from the contaminated material then the other organic
contaminants could be also. Therefore, during this laboratory program
aldrin was the only contaminant monitored.
Since gathering engineering data was not one of the objectives of this
assignment, we did not, for example, study in detail the settling rates of
the slurry after flocculation to carefully size thickeners. We only
determined if the material could be flocculated, and by observation
determine if settling was within acceptable rates. Consequently,
engineering judgments were used in the selection and sizing of much of the
equipment for the preliminary process flowsheet design.
- 7 -
To summarize: the objective of the laboratory program was simply to
determine the physical/chemical conditions that would remove the
contaminants from Basin F material using equipment that, based upon our
professional engineering judgments, could be applied on a large scale.
3.2 Sample (Basin F Material)
The contaminated Basin F material used in this program was received in two
5 gallon sealed plastic pails, early in January 1987. from Rocky Mountain
Arsenal. This sample was a wet mass with the consistency of moist modeling
clay. The two pails were mi.-'ed together by emptying them onto a plastic
sheet and combining them into a single pile. This pile was remixed five
times and then split in half and returned to the original plastic pails.
The two pails were then stored at room temperature.
Several observations were made during the mixing operation. First, the
sample appeared to be homogeneous. Second, there were no coarse rocks or
sand. Third, there were lumps of a black material (up to 1/2 inch in size)
which resembled asphalt.
Four samples of the Basin F material were taken during the mixing process.
One of these samples was sent to Arthur D. Little. Inc. for analysis. Two
of the samples were sent to an independent laboratory for aldrin analysis,
while the fourth was held in reserve.
3.3 Surfactant Scoping Tests
Of the numerous surfactants available, three (each of a different type)
were selected to be tested. The selection of these three was based upon
past experience with surfactants that have performed well, each for
differing types of contaminants. These three were: 1) Biosoft EA4*, an
alkyl ethoxyelated alcohol, nonionic, soluble in water and organic
solvents, which has seemed to perform well for a wide range of
contaminating materials, 2) Makon NFS®, an alkyl aryl ethoxyelated
surfactant, oil soluble, which removes heavy oils, and 3) Stephanflo 20*,
an anionic olefin sulfonate, which removes light oils.
In these scoping tests, 700 g of wet feed (heads) was mixed with one liter
of water to which sodium hydroxide (to pH 9.5), sodium silicate (7 Ib/ton)
and surfactant (3 Ib/ton) had been added. This slurry was then mixed at
room temperature for twenty minutes. The slurry was transferred to the
1000 g float cell, diluted to 2.5 liters and floated for 30 minutes. The
tails (washed material) from tests 1, 2 and 3 were then submitted for
aldrin analysis.
3.3.1 Results of Surfactant Scoping Tests
Table 1 presents the results of the surfactant scoping tests (tests 1
through 3). As can be seen, the Biosoft EA4 achieved greater removal
of aldrin than the other two and was therefore selected to be used
throughout the remainder of this program.
3.4 Flotation/Chemistry Variables Evaluation
The next series of tests was directed at evaluating the process variables
for flotation removal. Table 2 presents the data for these tests (tests 4
through 10).
These tests were similar in nature to the scoping tests, in that a known
amount of wet feed (heads) and 2.5 liters of water were added to the 1000 g
flotation cell. This slurry was mixed using the flotation machine, without
air, during which time reagents were added. After a suitable time of
mixing (20 to 30 min.), air was introduced and the froth collected. The
variables tested are shown in Table 2.
In tests 8 and 9, the Basin F feed material was slurried with water then
the solids flocculated and the liquid decanted. This was done to see if
removal of soluble salts would aid in the flotation removal of the
pesticides (aldrin, etc,).
3.4.1 Discussion of Results of Flotation Variabli s Evaluation
Variables examined were not all inclusive, but were the ones that were
believed to have the greatest effect upon organics removal from Basin
F material. As can be seen from Table 2, flotation alone was not
effective in eliminating aldrin to a level of more than a few hundred
- 9 -
TABLE 1
SURFACTANT SCREENING TEST RESULTS
TEST MO.
SURFACTANT USED
TAIL (WASHED FEED MATERIAL) ANALYSIS (pom)
Aldrin
Oieldrin
Isodrin
1
BioSoft EA 4
465
203
62
2
Makon NF 5
875
415
31
3
Stephanflo 20
711
277
6
Head
(Contami
1 nated Feed ’’’aten a 1 )
1190 ± 200
460 ± "
42 ± 7
Source: f'H'A Remedial Resources, Inc.
table 2
FLOTATION PROCESS VARIABLE TESTS
TEST
■JO. variable ’ES’ED
: SOLIDS
Pi DRY TAIL
ppm ALDRPi
i?J DRY tail
: DISTRIBUTION
OF ALDRIN IN TAILS
4
Repeat "est 1
91
not analyzed
—
5
'•■otation Time
73
310
20.3
6
'^lotat'cn "ime
91
520
39.7
Ong . • f 1 (It-
37
570
41.8
3
^'■ewa'ih
33
420
30.9
9
P<'ewa$h-Car(;c'i
92
660
51.1
:0
Caustic Atjd't'on
33
350
62.8
' Cantamtrated "eed '^atenial'
1190 1 200
'^O'jrce: '"^TA Inc.
n
parts per million. None of the process variables studied reduced
aldrin below a few hundred parts per million so it appeared that
flotation alone would not effectively dean up Basin f material.
However, aldrin removal achieved by flotation was greater than the
solids in the froth: therefore, it was concluded that if aldrin could
be freed from the substrate it could be selectively removed. During
these tests, it was observed that some black, asphaltic type particles
were not removed by flotation, and the assumption was made that these
black organic lumps were holding the balance of the pesticides.
Therefore, a technique to cause the removal of this black, asphaltic
material needed to be developed. This led to the next phase of the
study, in which organic solvents were used to dissolve these black
particles and/or to cause them to float,
3.5 Basic Flotation Process ‘■Modification Studies
The third senes of tests used organic solvents to determine whether the
black, asphaltic material, observed in prior tests, could be made to float
or dissolve so as to release the pesticides (aldrin) and thus imorove the
degree of decontamination of the Basin F material. Various solvents and
techniques were tried m tests 11 through 15 to examine this modification
to the basic soil washing process. Table 3 presents the summary data for
these tests.
For tests 11. 12 and H, the Basin F material was combined with water and
reagents, as before, m the flotation cell during anitation. Organic
solvent was added to the agitated slurry and mixed for 10 to 30 minutes.
Once thoroughly mixed, air was introduced and the resultant froth removed.
In tests 13 and 15 the Basin F material, water, and the reagents used in
the flotation chemistry studies were mixed together. After approximately
30 minutes of mixing, the organic solvent was added and agitation continued
for an additional 30 minutes. The agitation was then stopped, the mixer
removed and the slurry allowed to settle during which time the organic
phase floated to the top and was removed by decantation. The residual
slurry was transferred to the flotation machine and floated as in all
previous tests.
1 ?
__SSJ3(W<I NOIlVldlj_
Oisva 3H1 01 NOIIIOQV DINVOW
z —
o <
H*
CO «
<—•
cr
m
•■4
•
•
0£. SS
CO
o
c^
CM
1—•
C/^ QC
00
CM
4q>
<
u.
O
8
CM
♦4
O
(Tv
ftm
13
Source: MTA fteraedidl Resources, Inc
3.5.1 Discussion of Results For Flotation Process Modification
Data from tests 11, 12 and 14 indicate that the addition of organic
solvent and the recovery of the solvent by flotation, was not an
effective way to reduce the aldrin concentration in the tail solids.
However, addition of the organic solvent with removal prior to
flotation (tests 13 and 15) produced better removal of aldrin from the
soil (tails) tnan all previous tests. (Mote also that with the pre-
organic wash followed by flotation, the amount of solids reporting to
the tails was greater than when flotation was used alone.)
These five tests suggested that pre-treatment of Basin F feed material
with an organic solvent prior to flotation would improve the
effectiveness of the process: that is, a greater degree of aldrin
removal was achieved. The next series of tests were performed to
gather data on organic solvent washing in combination with flotation.
3.6 Solvent Prewashmq Process Evaluation
Twelve additional tests (16 through 27) were conducted to define the
process operating conditions pnior to the scheduled demonstration run.
These tests centered mainly upon the organic solvent pre-wash section.
Table 4 presents the data from these tests.
3.6.1 Solvent Pre-wash Test Procedures
Organic solvent pre-wash was accomplished by mixing varying amounts
and types of organic solvents with an aqueous slurry of Basin F feed
material. Subsequent to mixing, the bu’k of the solvent was removed
by settling and decanting of the floating solvent/emulsion. In the
staged tests, this step was repeated two or more times. Following the
last decantation, the required flotation reagents (caustic, silicate
and surfactants) were added to the slurry and this slurry subjected to
flotation to remove any trace of the added organic solvents and
additional Basin F contaminant. In one test (test 12), flotation was
employed between each stage of solvent washing to enhance solvent
remova1.
- 14 -
Total including intorstage
3.6.2 Discussion of Results of Solvgnt Washing EvaTuation
Tests 16 through 27 were scoping in nature and as such have certain
limitations in determining the extent of removal of aldrin that can be
achieved using the solvent pre-wash; however, even with this
limitation, several important conclusions and observations could be
made. The major conclusion is that organic solvent prewashing does
effect a good removal of aldrin.
Test 16 was run to compare the kerosene-toluene mixture with the
kerosene-octanoI mixture used in test 15 (Table 3). The results
appeared to be the same but better than diesel (kerosene) alone (test
13. Table 3). Remaining unanswered was the question "Did these two
mixtures remove aldrin attached to different constituents in the Basin
F material'i'"
Therefore, a toluene, kerosene, octanol-1 (TKO) solvent was tried in
test 18. The three component solvent removed more aldrin, therefore
in all subsequent tests this TKO mixture was used. No other solvent
mixtures were tried but it is very likely that other combinations
could be found that work as well or better.
Test 17 was carried out with a very small amount of organic solvent
mixture, with flotation to recover it between mixing stages. The
results show that this technique yields the same results as decanting,
but uses less solvent and fewer stages. This may be the preferred
method to be employed in a full-scale plant. This method was not
employed in the laboratory because of testing difficulties at this
small scale.
Test 19 was run to see if staging could be eliminated using a larger
volume of solvent. It seemed this was true. However, a repeat of
single stage test (test 27) indicated poorer results, which could have
also been the result of a shorter flotation time used in this test.
- 16 -
Tests 18 and 22 were comparable tests but the level of agitation used
in test 22 was higher than test 18. This indicated that mixing speed
was an important variable. However, time of mixing, at least beyond
20 minutes, does not appear to offer any advantage.
By comparing data from tests 19, 20, 22, 26 and 27 it appears that the
total flotation time was an important process parameter. To achieve
good removal of pesticides flotation times in excess of 60 minutes or
more will be required. This data also indicates that heating during
the solvent washing is not necessary to achieve good pesticide
(aldrin) removal.
This series of tests did elicit the major important process variables
for treating Basin F material via a soil washing process. Table 5
presents these variables in an order of importance over the ranges
tested.
3.7 Summary and Conclusion of the Laboratory Development Program
The results of the laboratory development program established a process
that can eliminate the majority of the aldrin, and presumably the other
organic contaminants of concern. To accomplish this removal an organic pre¬
wash of an aqueous slurry of Basin r material is needed, prior to the
flotation.
During the laboratory test work no unusual
evident that would cause difficulties when
large scale.
Overall, the laboratory program was successful in developing a process to
clean up Basin F material. It now only remains to demonstrate this when
the conditions established, from the prior test work, are employed in a
test run continuously from start to finish. The acid wash section that was
initially assumed not to need testing is also a part of this complete
demonstration run.
problems or conditions were
the process is applied on a
- 17 -
TABLE 5
IMPORTANT PROCESS PARAMETERS FOR
SOLVENT PREWASH/FLOTATION FOR THE
REMOVAL OF PESTICIDES FROM BASIN F SOILS
(in apparent order of importance)
PARAMETER
COMMENT
RANGE STUDIED
PROBABLE MINIMUM
REQUIRED
Flotation Time
Needed after solvent wash
to achieve removal of all
orgamcs.
30-130 min
> 60 min
Stage Addition
Will reduce amount of
solvent required
(counter current).
1-7 stages
2 to 4
Amount of Solvent
With better interstage
removal can reduce amount
used.
30 ml-400 ml
per 2 liter
slurry
< 30 ml
per 2 liter
slurry
Interstage
Separation
Can be achieved by long
settling time or flotation.
1-48 hrs.
use flotation
Mixing
Greater mixing energy will
improve removal of
contaminants.
very low to
low
unknown
Mixing Time
Mixing energy input and
mixing time must go hand
in hand.
20-130 min
unknown
Solvent Mixture
Mixture used probably un¬
necessary. Maybe, with what
is now known, use kerosene
alone or with 1-2% Octanol-1.
up to 40%
other than
kerosene
kerosene alone or
or 1-2% octanol
Temperature
Unimportant in solvent wash,
but needed in flotation.
Room to 75’C
none in solvent
wash, approx.
50°C
in flotation.
Source: MTA Remedial Resources. Inc.
- 18 -
4. DEMONSTRATION RUN
This section describes the demonstration run procedures, presents and discusses
the results and how they were used to develop the process flowsheet and
material balance for the "Soil Wash Decontamination Process for Basin F
Materials."
4.1 Introduction
As part of MTARRI's task, a laboratory demonstration run of the process
developed during the laboratory test program phase was carried out, Arthur
D. Little personnel observed the demonstration run and were responsible for
the collection of samples and their analysis. Data generated during the
demonstration run were used as the basis for developing the process
flowsheet and material balance. This data In turn led to the development
of the capital and operating cost estimates (presented in a separate
memorandum report). Sample collection and analysis by Arthur D. Little was
to be detailed, in that major compounds of concern were to be tracked, as
far as practicable, throughout the enti-e process. Analytical methods used
were approved and certified by USATHAMA.
In addition, the sampling and analytical program was performed *o ootain
sufficient data to confirm certain aspects of the process that had not
hoen studied ey+onsively during the laboratory development program. P'or
example, 'I’he number of stages in the organic -ash sec+ion and organic
flow reauirement were to be evaluated frcm demonstration run data; as was
the need for a final acid wash of the Basin F material.
At the time this program was developed, there were no guidelines available
on the degree of contaminant removal that the soil washing process should
achieve. Therefore, our objective was to remove as much as possible of all
the organic and inorganic contaminants. This caused us to use a more
extensive process during the demonstration run than was necessary based
upon the data subsequently obtained from the demonstration run. Therefore,
the soil washing process for the full-scale treatment of Basin F material
has fewer unit operations than were employed in the demonstration run. It
has been assumed that if the clean washed Basin F material meets the
- 19 -
criteria set forth in the EPA's proposed toxic characteristic leaching
procedure (TCLP), we would have achieved the required goal of contaminant
removal.
4.2 Demonstration Run Procedure
Figure 2 shows the demonstration run steps used and the material flows.
4.2.1 Organic Wash Steps
In step 1, 686.2 grams of mixed wet feed sample was taken from the
5-gallon storage container and put into the round bottom mixing tank.
To this 2 liters of tap water from Golden, Colorado municipal water
•system was added. Mixing was begun using a 2 1/2" diameter three
blade marine type propeller turning at 900 rpm. When the solids had
been dispersed the organic solvent mixture (TKO) was added. This
organic mixture consisted of 69. kerosene, 20.0* toluene and 10.6%
octanol on a weight basis. The slurry was heated, and the temperature
reached ^1*C. (Due to the ^act that there were tight time constraints,
not all of the data obtained during the Iaborator/ deveiopment phase
of the program was comolately analyzed prior to the demonstration
phase testing, this is true with respect to the data indicating that
heating during solvent washing is not a necessity *o achieve good
aldrin r-emovaI. As a result, heating was used during the demonstration
phase testing.) Mixing was continued for 60 minutes. Mixing was
stopped and the agitator removed from the slurry and the slurry
settled for 60 minutes, 'i’he organic layer on taop was then carefully
decanted (158 ml) from the aqueous slurry.
In step 2, 200 ml of fresh TKO mixture was added to tne mixing tank
containing the aqueous slurry from the previous wash step, the
agitator replaced, heaters turned on and mixed for 60 minutes. During
this second period of mixing the temperature reached As before,
the mixer was stopped, removed and the slurry settled for 60 minutes
and the organic layer (90 ml) carefully decanted off. In this step
the recovery of organic solvent was poor.
- 20 -
cess Steps
w
0 0
0 0
)(«■
lartt
Floccufcr}
Wol»f
Wcler
lUttic
Acid
Acid
rfoclonl
Roc oj lord
riocojlcnl
Wcter
nocojlor.)
fixi(h
Skrrj
850 ml
rma kko
1*1 Acid
Wosh
zceomi
RMA lOAl
2 nd Aod
Woih
2000 fH
Rma ICM2
3rd Acid
Wosh Fillrcle
2600 rrJ
RMA ICK3
Toll
(Cleon Soil)
Not Meosured
RMA 1038
FTotoliofi
Sect'
Toil Wash Section
Step 3 was started similar to step 2 but it was observed that the
system was flocculated. Therefore, 2.7 g D grade sodium silicate was
added which broke up the flocculants. Mixing was continued for 60
minutes and in this step the temperature rose to 44‘’C. Again the
mixer was stopped and the agitator removed and the slurry allowed to
separate for 60 minutes. After settling, the organic layer was
decanted off and 356 ml of organic solvent were recovered.
Step 4 was run the same as step 2 and no additional silicate was
required. In this step, the temperature reached 68°Cduring mixing;
205 ml of organic solvent was decanted.
In step 5. 200 ml of TKO mixture was again added to the aqueous slurry
from the prior step and mixed for 60 minutes as in the previous steps;
in this step the temperature rose to 69°C. After the mixing the mixer
was stopped and removed; the slurry was allowed to settle overnight
before the organic solvent (218 ml) was decanted.
4,2.2 riotation Step
After as much organic solvent as possible was removed, the entire
volume of aqueous slurry was transferred into the 1000 g flotation
cell and placed in the laboratory flotation machine. About 1 liter of
water was added to this mixture and the flotation machine agitator was
started and the slurry heated with immersion heaters. Then the
following reagents were added:
3.50 g Caustic (^JaOH)
2.00 g 0 grade Sodium Silicate
0.19 g Biosoft EA4*
The slurry was mixed and heated for 15 minutes prior to the start of
flotation. After mixing, the slurry was at 49°C and had a pH of 11.
Flotation was started by the injection of air with the machine rpm at
1200. The froth was continuously removed for 30 minutes at which time
additional surfactant (0.16 g Biosoft EA4*) was added and the rpm
increased to 1500. Flotation and froth removal was continued for a
total of 60 minutes. The froth volume was measured at 350 ml, and the
tail (clean soil) slurry at 2200 ml.
- 22 -
The tail slurry was transferred to a 4 liter beaker where flocculants
were added so the solids could settle and dear water recovered. The
flocculants used in this step consisted of both organic and inorganic
compounds: 100 ml of 0.1 g/1 Superfloe 84* solution and magnesium
chloride. Recovered water was 850 ml after settling for 1 hour and 22
minutes.
4.2.3 Acid Wash Steps
The settled solids remaining were then acid washed in three stages
using hydrochloric acid (steps 8 through 11).
In the same beaker used to initially decant off tail water, tap water
was added to bring the total slurry volume to 3600 ml. HCl was then
added during mixing to bring the pH down to 5.0; this required 7.3 ml
of reagent grade acid (37!t HCl), The solids were flocculated again
using Super-^loc 84* and allowed to settle for 30 minutes. After
settling. 2.05 1 of clear solution was recovered.
For the second wash (step 8) tap water was added in the same beaker to
bring the slurry volume back to 3600 ml. Again HCl was added to pH
4.3 (1.3 ml reagent grade acid, 37^ HCl) during mixing. The slurry
was flocculated using the same reagents and allowed to settle; 2.0 1
of clear solution were recovered.
The last wash (step 9) was carried out as before, except no acid was
added: the volume was brought up to 3600 ml with tap water and mixed
briefly. The pH was 4.9. Flocculant was added and allowed to
settle. In this step, the clear liquid was decanted off and the
settled solids were transferred to two Buchner filters to remove
additional solution. The filtrate and decanted solution were combined
for a total volume of 2300 ml. The wet filter cake was transferred to
sample jars.
4,2.4 Demonstration Run Samples
During tms demonstration run samples of the products were collected
by the Arthur D. Little observer. The Arthur 0. Little laboratory
numbers assigned to these samples are shown in Figure 2.
The organic solvent material and tails (washed Basin f material) were
taken m their entirety as samples. In the case of the tails, the
sample was nlaced in two containers. Only a portion of the other
streams wc'“f taken for a sample and the remainder discarded; this
includes the froth slurry, 1st, 2nd, and 3rd acid wash solutions. The
tail dec*.-’* liquid was not sampled.
4.2.5 *y~.blems During Demonstration Run
One process problem which arose during the demonstration run was
flocculation that occurred during the organic wash (step 3) and
prevented complete recovery of the TKO (organic wash solution). This
pr-^'-'. was resolved by adding sodium silicate which improved TKO
so .* on recovery in the last three wash steps.
Several problems arose 'n data acquisition for the demons*r’it!cn run
that created gaos ’n the data, first, the failure to sample the tail
decant ' iquid. Second, failure to measure the volume and weight of
tail slurry. Third, failure to analyze all the products for all
the contvfm-nants of concern, especially the washed Basin F material
vtail). f.urth. the failure to determine the solids in the test
product samples. *0 comp sate for these data gaps, some assumptions
and calculations were m- to fill in the missing data. This is
discussed more m the following sections.
4,3 Demonstration hun Product Analysis and 'iaterial Balance
Analysis of the products and the material balances arising from the
demonstration run data are presented in this section.
4.3.1 Anal yt i cal >?esu Its
Samples of the various output strears from the demonstration run were
analyzed by Arthur 0. Little using USATHAMA-approved and certified
ire r'"“sentad |n 'able 6. Also, Included
"ao'e r ir.^ **■0 anal/s‘s ~f **“« freer "*'0 organic ■"Uture and tap
wa*er ijced. l- -ao’e t: ♦'■e of a 'TL= laacrahe ‘est ryn cn 'ha
- 24 -
m
tails (washed Basin r material) are given in the column labelled
Leachate. The blanks in this table are where no analyses were
performed.
4.3.2 Material Balance Calculations
A measure of how carefully laboratory tests were run and analyses were
performed is how well the amount of material in the feed can be
accounted for in the test products, that is. a material balance.
Using the analytical data in Table 6 and the volumes of outflowing
streams given in Figure 2, material balance calculations were made.
Material balance calculation for the pesticides is shown m Table 7.
For all the pesticides there is excess (15 to 2311) material, when the
five organic wash solutions are compared to the feed. The cause for
this discrepancy is unknown; possible sources contributing to this
excess include imprecise measurement of volumes and/or weights or lack
of precision and accuracy of the chemical analyses.
'Material balance calculations for some of the inorganic materials are
shown in Table 3. In these calculations, the concentration of
material in the solution from the tail decant (step 7) and the 2nd
Acid Wash solutions had to be estimated as part of the material
balance calculations.
‘lotice that with the exception of fluorine, the material balances for
inorganics are poor. In these cases, the discrepancies in the
material palance are considerable such that neither errors in sample
volume measurements or analytical results could explain them; the
unaccounted for material is either in the froth, or tail slurry which
were not analy7''d for these inorganic compounds.
I
WA Raat«iia| fe*.
IVIHlikV
« 3t(Vl
1
$,P;§g5SSgSSS3S3S
••• ••«
p'* «N rsj ^ ^ ^
jiNcNiN
c 3 ? a s? ^
§* o* (N
:;]« l,T.
•“ ; •ji
^ ' . . . ~ Oi ( 7 >
J - - ~s
5i ^ ^ S - 1 a c ^ 3 a R :; 3 3
- ci- :1^ ^ J 2 ^ ^ i'
- ?R -
Souicei WA Fd-idi,i Siij^rcts
In the case of fluorine, the material balance approaches 90^ when the
amount that could have been contained in froth slurry liquid is added
to the total.
4.3.3 Discussion of Sampling, Analytical
and Material Balance Results
As in most research programs where considerable amounts of data are
collected, some data is overlooked and some does not fit because of
inaccuracies made in measurements or in analysis. When this occurs,
the data that has been obtained must be normalized for use in
developing overall material balance.
The data obtained was adjusted using assumptions in order that a
process material balance could be developed. This normalization and
the assumptions used are discussed in the following section.
4.4 Demonstration Pun Data Adjustments
In the preceding section specific data gaps were identified in the data to
be used for material balance calculations. Therefore, adjustments were
made to the data collected so as to compile a consistent data set for
subsequent material balance calculation and process equipment sizing.
4.4.1 Assumptions Used to Ad.iust Data
The data gap of most concern involved the lack of information
regarding the distribution of solids in the froth and tails. To
resolve this issue, several assumptions were made based upon previous
test data. These assumptions included:
1) the feed sample was 75% solids:
2) the specific gravity of the froth slurry was 1.02:
3) the tail slurry was 1 liter in volume and contained 36% solids.
The first assumption was based uoon the percent solids determined in
the preliminary laboratory program. The specific gravity if the froth
slurry estimate was based upon a measurement made by Arthur D. Little
which was measured in the laboratory and found to be about C.98. It
- 29 -
was assumed that the technique used had an error of about ±0.04 points
and so 1.02 was used since the slurry density had to be greater than
water. Using a solids density of 2.65, the weight of solids in the
froth slurry was then calculated. It was necessary for the last
assumption to be made since the tail slurry was collected in two
containers and was not remixed (homogenized). Consequently,
measurements on a single container alone (as was done) was not
sufficient to accurately determine the percent solids. However, a
measuremen*' on one of the containers (believed to ha*e the lowest
percent solids) determined the percent solids to be 32%: so the
assumption of 36% solids seems reasonable.
The results of these calculations and assumptions are presented in
Table 9.
The next assumptions made were that the solution in the tail slurry
had the same concentration as the 3rd Acid Wasn solution and that the
TCLP leaching process removed all of the compounds of interest
remaining adsorbed on the solids. The first assumption is reasonable:
the second is questionable. Using the volumes of filtrate and volumes
of leach solution, the total amount of the compounds of interest were
then calculated and the excess over that in solution was assumed to be
associated with the solids.
The solids may have more of the compounds of interest than calculated
by this method: however, this does not affect the material balance
used for flowsheet development.
Results of this calculation of the tail slurry, tail (clean soil) and
tail solution analysis are presented in Table 9 in the columns headed
Tail Slurry (Total, Solids and Soln).
The balance of the adjustments were made based upon volume of solution
added and removed, and the assumption that no inorganic salts were
contained in the organic wash solutions. The final normalized data is
- 30 -
31
shown in Table 9. Where ND occurs, zero values were assumed rather
than detection limit values in performing the material balance
calculations.
4.4,2 Discussion of Adjusted Results
Comparison of Table 9 with Table 5 indicates that even though
assumptions and estimates had to be made, there is considerable
agreement between the data. Consequently, the assumptions and
subsequent calculaticrs oased on these assumptions did not result in
any major disagreement with the actual data that was obtained. During
the course of these calculations, no cases were encountered where the
assumptions were shown to be invalid.
Assumptions used and subsequent calculations do imply, however, that a
considerable quantity of inorganic material went with the froth
solids. Flotation was carried out at a pH of 11, where even magnesium
would form a solid hydroxide: consequently, the assumption and
resulting material balance calculation may not be too far from what
actually happened. Hydroxides are known to gather organics from
solution and it is postulated that this is what occurred.
The actual distribution of solids between all the products from the
test is unknown since no accurate determination of the percent solids
was made at the time of analysis of the demonstration run samples by
Arthur D. Little. If the real distribution is considerably different
than that developed using the assumption then there may be some impact
upon the oper=i'’ing costs of the process.
Overall, the adjustments made to the data do not detract from the
results and conclusion of the demonstration run. The washed soil was
shown to meet the criteria of EPA's TCLP test. The material balance
and flowsheet developed using these data are reasonable and no major
changes will occur even if some changes in analytical or mass
distribution were made. This is due to the fact that all of the
assumptions and calculations were reasonable and conservative.
- 32 -
4.5 Process Calculations and Flowsheet Analysis
The analysis of the compounds of concern indicated that some process
simplifications could be made from those tested in the demonstration run
and would not effect the final results. These changes reduced the capital
and operating costs (including the cost of ancillary functions such as
organic distillation and wastewater treatment) for a full-scale treatment
system.
4.5.1 Acid (Counter-Current Decantation) Wash Section
(Inorganic Contaminant Removal)
Material balance calculations indicated that the majority of the
inorganic contaminants were eliminated in the flotation section and
any additional elimination of heavy metals accomplished by acid
washing was insignificant, for example, only 1.9% additional arsenic
was eliminated by acid washing. Flowsheet/material balance studies
indicated that this same reduction could be achieved by a single
filtration step of the tail slurry with adequate water wash on the
filter. This change reduced the number of equipment items as well as
reduced the volume of wastewater that eventually had to be treated.
Therefore the acid, counter-current decantation section was
eliminated.
4.5.2 Organic Wash Section
The laboratory development program had not defined either the minimum
number of stages or the organic to slurry ratio required in the soil
washing process. The analytical results from the demonstration run of
the contaminant organics in the organic wash operation produced data
that defined this area of the process.
Using the unadjusted data for the pesticides it can be seen that the
distribution coefficient of the pesticides between the TKO organic
phase and the aqueous slurry is very large. For example, for the
first step there was 200 ml of TKO with a concentration of 3100 ug/ml
aldrin, which accounts for all the aldrin in the feed (620,000 ug in
TKO vs, 570,000 in the feed). This suggests that aldrin may be
completely absorbed into the TKO mixture. The other pesticides show
similar results. Therefore, the removal of pesticides, at least
theoretically, can be accomplished in a single stage if the organic
phase can be completely separated from the aqueous slurry provided the
solubility of the pesticides in the TKO mixture is not exceeded.
Since the separation of the organic phase from the aqueous slurry will
not be complete due to the solubility of the TKO and very fine
droplets in the slurry, three stages of counter-current organic wash
have been provided. To achieve the maximum organic removal,
coalescing type oil/water separators would be employed. These units
typically reduce the level of entrained organics to 10 mg/1; which for
the flows in this process amounts to a 99.9% removal per stage.
No data was obtaineo from the laboratory tests or demonstration run to
determine the minimum organic to slurry flow ratio. Solubility of the
feed organics in the organic solvent was at least 3100 ug/ml for
aldrin, as determined in the demonstration run. However, thiu did not
appear to be the maximum based upon published data of solubilities in
organic solvents. Therefore, a value of 15,000 ug/ml was used as the
solubility of the feed organics in the TKO organic solvent to set the
organic (TKO) to aqueous slurry flow ratio at 0.023.
One other point about the organic wash section that should be
discussed is the fate of the organic contaminants other than
pesticides. The TKO solvent loaded with pesticides and other organic
contaminants is to be distilled to recover the TKO. Analysis of the
distillation unit operation by Arthur 0. Little personnel pointed out
the fact that the majority of the organic contaminants, other than the
pesticides, would report in the returned organic solvent. The
demonstration run data shows that the distribution coefficient between
the TKO solvent and the aqueous slurry is low for these compounds, so
that they would be washed out of the TKO solvent into the aqueous
phase and would have to be removed by the following processing steps.
- 34 -
4.5.3 Non-Pesticide Organic Contaminant Removal
The developed flowsheet, based upon the preceding information,
required that all of the dithiane, sulfoxide, sulfone, and OMMP would
have to be removed either in the flotation area or by other means.
The adjip'^ted Hata presented in Table 9 suggests that about two thirds
of these other organic contaminants would be removed by flotation: the
balance remain in solution. Therefore, an activated carbon adsorption
system was added to the process to treat a bleed stream of water to
eliminate the balance of these other organic contaminants. The loaded
carbon would be disposed of with the flotation froth solids and
distillation bottoms.
4.5.4 Heat Balance
Data from the laboratory development program indicated that heat was
not required during solvent washing and it appeared that heat was not
an important variable during flotation. However, no studies were done
to establish the effect of heat in the reactor prior to flotation. In
addition, heat was used in every step in the demonstration run. A
compromise was used to estimate the heat required for the soil washing
plant. It was assumed that a temperature of 180°F would be required
in the pre-flotation reactor only.
Therefore, in the process flowsheet provisions were included to heat
the pre-flotation reactor to 130°F and to recover heat from the slurry
exiting the tank in a single pass heat exchanger to heat the incoming
slurry. In addition, heating panels are incorporated in the reactor
to add the additional heat. Overall, the heat balance calculation
showed that 30 x 10^ BTU's per hour would be required during very cold
periods. This heat load allows for heat losses, boiler efficiency and
will be used throughout the year to calculate the operating costs.
The flotation feed slurry will be maintained at 78°F.
- 35 -
4.6 Summary of Demonstration Run
Overall, the demonstration run showed that soil washing of Basin F material
can eliminate the contaminants, both organic and inorganic and yield a
final clean soil that passes or exceeds the proposed TCLP criteria set by
the EPA.
During the demonstration run no problems were encountered that were
insurmountable or would make this a difficult process to implement on a
large scale. The required equipment is currently manufactured so no new
equipment design or development is required. Reagents used are all
available in large quantities.
Although some data was not obtained during the demonstration run and some
problems with the calculated material balances were observed, these were
resolved by the described assumptions and adjusting the mass flow and
analysis. These adjustments were necessary so that material balances could
be developed and equipment sized, but in no way detracts from the
conclusion that this process will clean up Basin F material.
Analysis of the data, from the demonstration run, showed where some process
simplifications could be made. These changes were incorporated into the
process flowsheet presented in the following section.
- 36 -
5. BASIN F SOIL WASHING PROCESS
From the information obtained during the laboratory program, a soil washing
proces-s was developed. This process was further refined based on the
analytical results and flowsheet analysis using the demonstration run data.
5.1 Intrcduction
Soil washing of Basin F contaminated material was studied in the laboratory
with the emphasis on pesticide removal. It was initially assumed that:
1) the other organic contaminants would follow the pesticides; and 2 ) the
inorganics would have to oe eliminated by an acid wash of the organic free
soil. The data from the demonstration run showed both of these
pre-conceived ideas to be incorrect. The other organic contaminants could
not be eliminated along with the pesticides: therefore an activated carbon
adsorption unit was added to the process for the elimination of these other
organic contaminants. The inorganics, principally the heavy metals were
found to be concentrated and removed with the froth solids, probably as
metal hydroxides. Since the anions and the cations such as chloride and
sodium can not be eliminated by this method, a water bleed stream will be
needed to control these contaminants.
With all these factors considered, a process flowsheet was developed along
with a material balance for soil, water, pesticides, other organic, and
inorganic contaminants.
The generalized process flow diagram is illustrated m Figure 3. This
diagram shows the individual operational process areas and how they are
interconnected by the material flows, the inputs (feed, reagents.water,
etc.; and outputs of the process. The process has five output streams
containing various contaminants that currently are anticipated to be
disposed of by incineration. One aqueous output, free of organics, must be
treated to eliminate the dissolved salts. It is anticipated that a portion
of this aqueous output stream can be used, as required, as process water in
the incinerator and the balance evaporated with the recovered water being
returned to the soil washing plant.
- 37 -
xjurce! wifi ffe-TEdial , hiC.
Table 10 summarizes the material balarce for the various constituents for
the generalized flow diagram (Figure 3).
The generalized flowsheet and material balance resulted from a more
detailed flowsheet which was completed to size the various pieces of
equipment. While these details are not pertinent to the understanding of
this reoort. the detailed flowsheet and material balance are included in
Appendix A.
5.? Process Description
The following is a brief discussion of what is to be accomplished in each
process areas, shown m the generalized flowsheet, and how it is to be
accomplished, and with what equipment.
5.2.1 Feed Preparati o n Area
In this area the feed is received by dump trucks at a rate of
approximately 20 tons every 30 minutes and dumped onto a 3" opening
fixed bar grizzly to remove large rock to protect the log washer
downstream. The sticky material is washed through the grizzly using
return process water. Material passing through the grizzly drops into
a log washer to break up the material. The log washer levels out the
feed surges. The pulp discharged from the log washer is passed over a
screen to remove the coarse material. Screen and grizzle oversized
material is crushed m a jaw crusher that will accept 6" rocks. The
crushed material is returned to the feed end of the log washer.
Screen undersized material goes to a large holding tank where it Is
adjusted to the correct slurry density in preparation for the organic
wash section, ’he combined holding capacity of the log washer and
feed surge tank is about 34 tons of solids or a little over 2 hours
operating time.
Because of the nature of the material, sticky and plastic. It was
deemed necessary to store slurry r.itner than excavated feed material.
- 39 -
SS33IHd ONIlISm 1IOS i PUSVH MU IMV WH NIHllVW WHWO
Sources PITA Remedial Resources
Although no coarse material (+ 1/4") was seen in the Basin F material
sample received, MTARRI has provided this oversize protection and size
reduction as an insurance against large material being present. The
downstream process, especially the flotation section will operate
better on a sized feed.
5.2.2 Organic Wash Area
The feed slurry at 20% solids is pumped to the first mix tank in the
organic wash area at 596 gpm where it is mixed with 12,7 gpm of TKO
(organic solvent) moving counter-current to the slurry from the second
stage of organic wash. The organic/aqueous slurry mix is then pumped
to a settling tank to allow the majority of the solids to settle out.
From this tank the top portion of the slurry consisting of water, TKO
and fine solids is pumped at 300 gpm through an oil/water coalescing
separator. Leaving the separator are two streams: 1) the pesticide
bearing TKO solvent; and 2) the aqueous slurry, which contains about
10 mg/1 TKO. The aqueous slurry is combined with the unde'‘f1ow solids
being pumped from the settling tank and becomes the feed for the
second organic wash. The organic phase containing the pesticides is
pumped to the organic filtration area.
The second organic wash tank receives the combined aqueous slurry from
the first wash stage and TKO solvent from the third organic wash stage
and is processed as in the first stage of organic wash.
Third and last organic wasJi stage r-coives the cembinsd aqueous slurry
from the second wash step and TKO solvent which has had the pesticides
removed by distillation (plus make-up). This step operates the same
as the first two. The TKO is advanced to the second stage of organic
wash and the aqueous phase slurry is sent to the flotation section for
additional cleanup.
5.2.3 Orqanic Filtration
From the first step of organic wash the pesticide containing TKO
organic solvent, which contains some entrapped solids, is filtered
with a recessed plate and frame f;Iter to eliminate these solids.
- 41
This is necessary to prevent problems in the organic distillation
unit. The TKO is then sent to distillation while the solids removed
from the press would be and comingled with other contaminated streams
from the soil washing plant and incinerated.
5.2.4 Organic Distillation
Solid free TKO organic solvent is distilled to recover the majority of
the TKO organic solvent and to concentrate the pesticide in a
distillation column bottoms stream. The pesticide free organic stream
is returned to the soil washing plant where fresh organic reagents are
added to make up for distillation losses. This TKO organic solvent is
used in the third organic wash stage. The bottoms consisting of
pesticides, tars, and kerosene (about 13,000 Ib/day) is sent to the
incinerator for destruction.
5.2.5 Flotation Area
The pesticide free slurry from the organic wash area is heated and
sent to a mix tank reactor where reagents are added to free additional
organic contaminants to be recovered by flotation. In this reactor,
sodium silicate, caustic, and a surfactant are added and held for 30
minutes at ISO^F, with high agitation.
The reacted slurry is pumped through a heat exchanger to recover heat
and sent to the flotation cells. In these cells air is blown th’-ough
the pulp and the hydrophobic material (organics) is collected in a
froth which floats to the top and is mechanically removed. The bulk
of the slurry (tails) passes through the cells and is now free of most
of the organic and inorganic contaminants.
5.2.6 Froth Liquid/Solid Separation
Froth from the flotation section contains about 3% solids so water is
*
recovered and returned to the process. Also water removal reduces the
heat load in the incinerator. Water removal is accomplished by adding
a non-ionic polyacrylamide flocculant reagents to increase the
settling and filtering rate of the solids. Flocculated slurry is
settled in a high efficiency thickener; the overflow water being
returned to the process. The underflow solids still contain a lot of
water (± 80%) so this stream is filtered on a small belt press
filter. The filtrate water is returned to the wash process and the
solid cake (50% solids) sent to incineration.
5.2.7 Clean Soil (Tails) Liquid/Solid Separation
The slurry containing the clean soil (tails) must also be separated to
recover water and produce a solid cake with no apparent free
moisture. Flocculant was added to the slurry which aided in settling
the solids in the thickener, and the thickener underflow was
filtered. The thickener overflow and filtrate water are recycled back
into the process. The clean soil filter cake is washed with fresh
water to eliminate the final contaminants remaining dissolved in
solution in the cake. Finally, the clean soil, containing about 65%
solids is placed in a fill on-s’te.
5.2.8 Carbon Adsorption
Several of the organic contaminants are not completely removed by the
preceding process steps. Therefore, these remaining organic
contaminants are eliminated by activated carbon adsorption of a bleed
stream of water from the plant, A two stage fully automated counter
flow carbon adsorption system is proposed to accomplish this final
removal. The water leaving this unit will be free of organics and is
sent to a wastewater treatment system because of the inorganic salts
still remaining in it. The carbon, when fully loaded, is sent to
incineration along with the other contaminated streams from the
washing plant.
5.2.9 Wastewater Treatment
Arthur D. Little is developing the wastewater treatment process. At
present, about half of the water in the bleed stream will be treated
to eliminate dissolved salts and returned to the soil washing plant as
fresh water with the balance of the bleed stream water being used in
the incineration unit.
- 43 -
5.2.10 Volatile Organic Control
To control emissions of volatile organics from the plant,' several
designs were incorporated in the process flowsheet to reduce the
amount of air to be scrubbed. Where possible, process equipment will
be sealed so no volatile organic will escape. Emission from equipment
that cannot be sealed will be enclosed in a building which will be air
swept.
Areas where v-'latile organic control will be required are principally
in the truck unloading, screening, and filter presses. Truck
unloading and screening operations will be in a building with air
sweep. The organic filter press and the froth belt press will be
housed in another enclosure with an air sweep. The flotation cell
will need to be covered with a hood. It is reported that the
flotation cells will leak at 0.1 SCFM per barrel of material
processed, or in this case about 2.4 SCFM.
Overall, the volatile organic control would be small, about 100 SCFM.
This assumes 5 air exchanges per hour of two buildings having a total
volume of 10.000 cu. ft. plus that from the flotation cell, 70 SCFM.
5.3 Summary of Basin F Soil Washing Process
Using the data collected and numerous flowsheet/material balance studies, a
processing plant was designed that will produce clean soil (as defined by
the era's TCLP procedure) that can be returned to a fill on-site. This
plant would employ equipment currently available and reagents that are
readily available in large quantities.
The largest flow stream in the plant is about 800 gpm (slurry to the
organic wash area) and the largest piece of equipment, other than tanks,
are the four flotation cells at 1500 cu. ft. each.
t
Overall, a full-scale soil washing plant would be easy to operate, require
a minimum of maintenance, and would have an above average on-line factor.
- 44 -
Instrumentation and control loops would be minimal,
would be tank levels.
The pl«nt would be relatively safe to operate since
or high temperatures are employed. The plant would
acceptable since it is temporary and all emissions,
would comply with current emission standards.
The major control
neither high pressure
be environmentally
air, water, and soil
- 45 -
6. RECOMMENDATIONS FOR FUTURE WORK
If in the final evaluation of innovative technology to clean up Basin F, soil
washing continues to be a viable technology both from an environmental and cost
standpoint, additional work is needed to finalize the process, engineer the
plant and finalize the costs for the overall project. This additional work
would be in three steps: laboratory studies, pilot plant demonstration, and
engineering design.
6.1 Additional Laboratory Studies
These studies would consist of two parts. First, the process would be
defined in detail. For example, can toluene be eliminated from the wash
solvent, and is there a better solvent system that would make the
distillation system more effective in eliminating the organic
contaminants. Second, data for engineering design needs to be gathered to
more accurately size the equipment. Such items needed are size
distribution of the feed material to a detailed analysis of flotation
variables for scale-up design.
This program could be completed within three to six months. The
controlling factor for completing this program will be analytical data
requirements. Analytical requirements will also be the major cost to
complete the work.
6 .2 Pilot/Demcnstration Testing
It would not be prudent to scale up the proposed process from laboratory
bench-scale studies to the full-scale plant (950 tons per day) even with
additional laboratory data. Therefore, a pilot/demonstration plant should
be built and operated for two to four months. This small scale plant (1000
Ib/hr) would be built based upon additional detailed laboratory work which
would fix the flowsheet so little or no equipment arrangement testing will
te needed. Th’erefore, this would be more a demonstration plant with only
limited testing on the effect of process water recirculation, distillation
and quality of the clean soil produced. All data needed for scaling up to
a full-scale plant will be obtained by a months run at steady state
conditions.
6.3 Engineering Optimization
During the preceding two steps, the work should be subjected to engineering
optimization as the data becomes available. This would reguire
establishing an engineering team to assist in program design prior to the
start of either phase of work. During the additional data collection the
engineers could evaluate data and make recommendations for changes, new
tests, additional data so that upon completion of the pilot/demonstration
run there would be sufficient and complete data for the final and detailed
engineering of the full-scale soil washing process plant.
- 47 -
/
APPENDIX A
• Soil Wash Plant Flowsheet for Basin F Material
• Material Balance Basin F Soil Wash Plant
/
3t
L/> lO
O 1
p-« H*
VO VO OC
lO
1— o
VO
v*p to
Oj f
vr Oj
r*- ov
^ 1
*— O 00
o
U
P^ OJ f—
m
Oj 0-)
^ I
0} E
OJ ro
VO Oj
VC
c 1
c
■o O
' •
•
c w
CJ
3 Li.
oc
3 ^
o
oO
L.
<}) E
> O
O i-
00 •IT
on o r^*
c c^
CTv c\j
c «—
vf) un
CO ^ o
tn c\j
m
».C CNj
C\4 <Vj C\J *“ O CC
<7^ ^ r-
a>
c p*
o '
w c
0)
00 h-
^ cr c\j
VC a> <T
•- c
O r«-
P\i uo
^ p—
04 lO
cc
OsJ 04
VO f—
VO ro p-
p— C VO
VO
C
0*5
<
o c
-*-> <t3
C I- X
vo o ^
“C O yi|
a> *j
5 -c “■ ‘
C! <y .
i- CT
U I
vO
c!
o u
■rr t5 vo
a> cr*
a> 5- 3:
•-W o
VC
Oi
Oo
o vO
—* 04 o
C O oo
cr o vc
p- 04 o
^ lD
p— ^
r- •sr
oj uo
c
Oj vD
<y\ O,
Oj* O
0 ^ I
c\j vO
CC
O
Oj O/
Oj .—
O. UO
00
ur m p- V45
^ O VO o
L.
1)
lt m p—
•— O VO
VO VC C
r—’ O f-*
TT
OJ t/1
r~ to i.
u o; OJ
Oj >> U 4 -^
U O <T3
(DUS
a: CL
o. O ^
rr Oj
*— tO
«r
Oj
1— LO
c
o
Oj vn OJ
VO CC
CTv
m O
OJ Oj
OJ CO
CD
lo m <—
— o VO
i/i
-a
>> r-
U Q r-
U CL O
O ►— CT
IJO
8^?
E u Q
CL CT <D CL
cn 4-) h-
CL -tS
(/> 3:
s tc o
CL «- a.
o
CO
Q
Q.
h-
(/»
C
u
.
a>
c
O
u
'O 1/1
t/^
3
(D
o> u
4 .>
"O
U p>
0 ) Ul
X
•»—
o c
*J XJ
c
u
Its p-
E S O
E
i- d
u <—
r—
Q. CL
a.
4 .)
o u
r- O
o
CJ >0 f—
a>
l/i
-c o
C /1
(jO
(V
^ C
■ (>.
h-
a.
o —
(jO
HAftKlAl BAIANCI MSIN J SOU WASlilliG PIAHT
f— w gn c
01 <c
S S’-
^NJ C
O —
O csj
ircNj r^rvj csicvj ^COC
^ fMM ^
sC fVi •— C
cr>
CT' X fVi
— sC
X i?'
c
• • •
*- ->il
V ^ C‘
— c!
^ 0
fVw uT
<*«» rsi
fSi
X
W 1
•
•
^ i
rr
W;
0»l
0 K!
W
(T <T
O' <Vi
3e:
• •
c
^ cv
0
^ Tr(
sC
C
-C < ^
^ c £‘'
z z: c
V, u*' -
2 : c
^ ^ >4-
C ^
C X
c
tv. 'll —' —
a .r V a.
a •?
I? -r S r Cli S
ol - a a A
•—
O ’ i
C
^ ^
V r* w
t: w •
- c c
IS
- w
w V w
^ 0
a» w c
a o —
0/ V'
•-- ‘t? c
V —
-
i
1
I
I
I
I
I
i/>
>>
•Q
c
0) a
TO
cn 3
«o
fM*
04
0) <v
CO
VO 04
Ck:
•
• •
•
'Q
VO
VO
VO
o ac
Csi
3
Ck: O
o> O
w. k.
O
04
iT)
Og
in —
iT)
c
(/Vj
h-
^ §
r“
VO
in
Z
Ou wi -»->
1
•
<
Oj •»“ -w
! ®
vd
»—
O
1
&
COI
1
i
MM
u
D -C
C Q
a
cn —
vyo
K D
- a;
d O d
M- ^
•—
— O vO
4^
— Z U- i
*— CNi C'
o».
r\^ tr
•“
M !«•
-D
1 ^
Oj n
QC
c
c
0)
mJI
tJ oi
•
•
a
•— 1
i- <
c
Si
’
w
%
^ 1
z<
V 1
MN 1
>.
n \n
in\
yi U !
^ ^ n
c
04 Og
*- c
<
i\
*J W
• « •
t •
C — 3
O ^
U~ Og
^ Og
««•
^ d d
UJ
<\, “C — 3-
0-, ^ CC
r*..
—
•mi
^1
0 54^ C
a fv
vO Og
c
c
i
O
:r
o. <
mjI
a|
^1
V '
" wl
«|
'.1 r!
Uilj
11 c
;T- n O'
(yv Og
O o.
<1
o ■—
— -TJ .,
•t •T Og
mm t
oJ d
Og Og
M* •» .->
X'
o cl
vO —
m t
VO —
M M -M
O 1-
C '
c c c
c:
S\j
c
0—0.:
— *4 «*
*4 I
W 4^ ^ t
V C|
"O
£ I"':
^ a:
Og -^ —
r— Oj <-~
Oj ^
oj
— m
^ Oj
rr-, r-o.
.C O,
O,
VC
in n
— c VO
O OG d
>»
gv
Q
t:
U
Ck
a “-s.
M
wi }—
Ck JO
i;
c
4.4
• i
i
k
wv
</»
c ^
^1
1
a
a»
O'
u
^ c
mm ^
4>
w
■a
k
M
(V
nv V
o c
T?
K
M
o
44
'O iA
c 4»
£
>.
«“
i
M»
u
•n
V o
3
tj
W O
a k k o
e TJ o
H X a
e
M
k
O'
u
QC k
44 — .
1>
k. Cl
0
tTv O' Cl.
Cl. a.
a ck
a
W
i;
k
V
u o
3 »—
<40
W H-
H-
c>o ♦—
a>
o
0
M-»
O kc
O 1-
a ^
o
w
c
M
n
j:
gO
H*
Ck
O
wm
V/T
k-
A-, ^
I
I
I
a> tA
r- GO
r— U
• •
U Oj 0^
CO f'j
CSJ >> U 4J
CO CM
<V) y O
ro cv»
0 ) 3
•
Q£ CL
L.
f— «-•
f- VC
O »n
CO «—
iO
0
• • •
• •
• •
• a
4-> -O
O' ^
crs •—
o ^
U (V
O •- cn
xr
«— m
CO
53
»— LD
CM in
00
0) U.
•
•
oc
CO
I
•0
n
O CO
CM CM
O 3
•
•
• •
• •
O O -t-»
m CM
^ CM
<:> 3
m
00
CO
O' •—
■p^
T3 3 C
oc
fM
iO CM
c
1- fT
*
•
CO <
0mm
— C
• •
O 00
!
I
I
I
I
I
!
I
h-
^ L-»
5
-0 3
o c
o^
<3v CM
C
O'
«
«
• •
d.
CM
•
X--
CM
I—
CM 0
CM CM
mm
•c CT
iC
m
VO
■ mm fm
W L
CO Cl
c
c c
<n
“if!
•2
-0
in vn
O
t: k- '
r*'*
MO
f— CO
CO CO
r- 0
3
mm
/i—
L
4
•
• •
• a
C
5 (
Li.
CO 3^
cr
^ CM
mm
C oc
00 L
C 1
iC
mm
<-o
CM CO
fm
w
CM 1>
0 m,
Cm
CO
vO fvj
c
e
U»:
-3
O f
•
0
W w[
CM
(«i»
u
zl
V4. ^:
■H 1
<
u;
UJ
5
u
z
c;
OI
n in
<
31
-c
c CO
mm ^
C
c
O
L '
*
«
• •
• a
s
<“
ro cr(
<Cv^
r*-.
n Cm
r-^ Cm
CM CM
c cd
r>. ^
d .
C
O'
ro
^ <—
VO •“
CM W
E ••'
o**
Cm
vO Cm
mJ
<V
3 «*'
•
•
<
>
W mJ'
c
V*. wi
QC
UJ
h-
o» j
•c
C 1
c
0
1—
^ r>^
C
ro •—
w
-Xl
•
«
« •
• a
, ,
Vfi CO
c 1
CM
— Xf
•»
CM CM
C VO
CM
W 53 I
C“
CM n
\0
O
w L-:
vn
CM <n
00
h~
<i> 1
*
•
CO
i
f
un
CM
W
V ‘TT
X a>
- ZJ
X u.
"O c
m H- i
<M
>.C
O
<M Osj
vO
(A
E
vi
"C
>>
w o E w w a
^ a. o a O' iij a.
3 h- CT' W H-
a
M 3
<v
w
3
O
o.
era
a a.
C7»0
e
Q.
3'
C
</» -<J v/^
C; C7» u
T3 W »-
- O c
u
^ u. cr
4-» i> u
(/> -c o
a;
a. o —
Of
u
TAULE a (Continued)
HATERj/^ BALANCE BASIN F SOIL WASHING PLAMT
k.
^ O 0)
O 4-» ^
"IT O ’ <T 3
k. ■»“ o
CM 00
*— O fO
c
<D ^
C "O
0) 3 0)
^ O <V
<j r~~
■ fm
CTi -C i- ^
ro )— 0 ) 0 )
-4-»
-C c »—
r- m
OC CM
r-* •— OC CM
CM CM CO
CM CO
c
&.
<U 3
X c o
4J 0) ♦—
GC O
CO 4. U k.
u- 0)1
X >
h- ol
k.
0)
X c
><-> Oi
O O
CO k. u HI
CM
0
CM CTl
CM
X
CM X
F— F—
CO
(T>
00 Ol
■#—
•r» -F*
F“
F— r—
c
c c c
•
•
CM 0
1— ^
CM CC
0
cr* cc
OC CM
^ 0 CO
X
(M
0 ^
CO
CO
CM
ro CM
c
— c
X 3 - lH
ro u — . (Ml
U O .<
O Ci
<— -wi
CM C
CM CM —
C
— p- ^ c
c c c
^ X
, ^ ^
w C
*— ^
CM CC
W w
cc
r«»H X
CO* CM*
M-
F- 0 CO
^ 0
'£! —
o^ —
ro
-W L
0 U.
1 ro (\j
CM CM
c
c
c
f F—•
F— •
1
c
0
0 0 CM
X Oo
0 X
CO ^
c
1
X X 0
X OC
C^ r-
— ex
F—*
43 0 i
^ ^ CM
F- X
F«- X
•w H !
0 U-(
1 X CC
X c^
X
c
c
CM X X
r— X X
E
0
>.
k. 0
»—
e:
L L
H
L CL
0
d
C71 H
L
3 h-
X
cn
w
X
X
5^
a ^
ui 3 :
CL CL
CT'O »“
l/l -TJ VI
a> O' u
T3 i- -
- O c
u
*- U C7'
4^ H U
1/1X0
owe
Ql O ^
O
Ou
Q H-
o. o
H- CL
►— C
O 'TJ
W U W
F- u
J W *0
F- 1/1 ( 4 .
3 L
^ 3
X U X
Caustic 50% Sol'n TPD
Solids TPD
I
I
I
1
I
I
8
Water
ings
ter
I o
n
O VO
^ •— /—
CO
a>
oc ol
0^
^3 ■**“ ‘f“
V) <o U.
VO
*— ^o
c
<0
2
VI (3
o
VO
o
^ ON
oi »—
o> c
»—
•
•
C (3
Vk-
CO
00
^ 00
O Oi
rs.
•»“ ^
L
m
c
vn n
00 ^
1— CJ
o;
CM
CT^
fN.
•f- -f-
“C
•
^ JZ
c
yiMK
► - H-
o
^ O cy> cvj ^
o r- o o c
•r“
C
I
i
I
O 3i
CM
□1 C Ol
C Oi —
1 CO
to
4T vO
••• Lie »4-
»—
ro
r— ro
»— U Ji-
! n
VO ^
c
-- P- o;
c
CM*
CM
I— m
• • •
Or— ^ ^ ^
•»“ f* -f-
c c c
5 51
e
o
un
^tr
w I
i/> a;
c
c Oi -c
f- a»|
r- u a;[
•«- U-(
'a -c
'O
c
3 • a“.
U r— <\(
u o .
O '-0 o
J
CO ^
m
in XT
in \n
c
in ^
CM
tD m 0^ f—
OC ^ o oJ
cr» cc ^
m
o> —* ^
c
r-“ cn o C\J CM o
^ O CM ^ ro o c::
c
c c c
o
c
Uy
o
z.
<
<
tt
0 o
•— Vi oo
ro w *—
^ r- C
'*•’ ‘tj
O h- <3 j
U- O
cn
00
-vO
vC
CC CM
o
c
»— U*'
r—
VO
ro n
•
r-5
{^^ f-
C ctI
00 ^
^ m o CM CM
— C CM »— o '—
^ c
1
e
o
vt ^
VI u. ^
CM a; ^
^ w i- O
O 0) ^
W -y u.
a. 1 ^
n
CM
c^
o
CM
r- m
rv ^ ^
wr o -
CM CM c
c c c
V
-C a; <j
^ -w *-• w.
^ O r— *J
t. ^ ^1
C>
^ CM
O ^
^ CM —
i*«« ^ ^
C
c c c
£
tj
w..
Wl
C/^I
i- Q
w a.
wl
T!
*0
oo
e i- 1. o
a r' 01 Qu
01 W h-
a 13
i/i 2
<];
3
VI
s “c o E as: Q
a ^ Ou a. a.
01»— H- CIO H-
c ^
c/r w-
VI
(j
c
Vi (CP 1/1
0) Oi U
T3 4. -
r- o 3
U 13
r- S. 01
<w o; ^
VI ^ o
a» 4-» c
a. o —
Q
CL
O K-
CL O
h- CL ^
H- C
0 ) Its
■w U W
13 r- o
U >*-> 13
»- V> ‘4-
^ 3 L
13 3
C/1 L-? C/0
O
CL
C
'O
3
U
u
o
lu
•1 .
•D
k.
Ov
o^
(U
0)
•
•
•
VO
(U
CM
fvj
in
oc
d
3:
OD
CVJ
9-^
m
c/^ VO
o
e
o
.5 5 ?
VJ
z
<
•J
<
MM I
«
Ul.
c
fO Ol
r- 3 l-
vn 3 (U
UD (J <D
u ^ !n
O 'd 3S
Ln
un
I
t: c(
o o
t; x5
't3 t.1
o «!
^ UJi
u o
< o
-O
'Z
OJ VJ
k.
c c
O S -O
fVj ^ 3 U
;r) k. »— 0^1
T3 O U-i
O kJ I
ro
CSj
o ^
lH ^
f- VC
<VJ 00
vO ^
CV4
cr*
o
^ vC
O ‘O
rr vn
on
c
o> oi o
c
•*-
-D r-.
Id c
o •
-J o
O og ^ ^ ^
C c c
O' 1/1
m
uo on
r- m
CM
c V/^ V.
a a
a a
a a
p- a? oj
cc
O CM
on ^
^ U 4-1
r*^ C>
<0 <31
■ fm
lO >- O d
ui in
UO uo
c
c
c
d k 2
•
Q-
1 on
m
0^1
1
3> U -t-ii
c 0 ; d
I
fTi
^ o^
vD
<31
O k.
a a
LO ^
in oc
^ OC
a—>
— O
o —
F“ •(— <—•
tn ld
in in
d U-
U- 1
CTi «—
O' *—
c
c
c
c
C3
^ p^
O VC
<31 —
o
CM
C k
• « a
a a
a a
- a; a>
O lO cc
OD <3v
O <31
r—
p—
r- C
CTi <—' 4J
o vO —
— LO
CO ^
't~ ’r~
^ d
CM
'd*
c
c c
C
d - o
»
h— Li-
t—
a>
3
>
o
•o
E
k. k.
o
e xj o
4.)
><
6^0
e
V. CL
o
Q.
cn 0 ^
Q-
GL 1 - tx
Qa O.
Q.
3 h-
in
4->
K-
CV*— K-
cno *—
O'
avo
Q. d
o
VI
ca
C
u
o.
CL
o
K*
c
CL
Q
VI
d
^—
O-
4J
.u
o
a;
o>
u
♦—
c
C
Q-
*T3
k.
■•"
c;
d
d
H-
■»»
c
c
4->
U
4-)
r—
U
d
d
•(»»
u
3
C
k.
O'
u
4->
d
U
o
4-J
Oi
k.
•f-
VI
V*-
u
40
vn
-C
o
3
k.
o
k.
o;
4.)
c
•F-
d
3
r***
d
,
»/-»
t y
/
1 .
t >
source: MTA Remedial Resources, Inc.
DISTRIBUTION LIST
Defense Technical Information Center
Cameron Station
Alexandria, Virginia 22314
Defense Logistics Studies Information Exchange
U.S. Army Logistics Management Center
Fort Lee, Virginia 23801
Commander
U.S. Army Toxic and Hazardous Materials Agency
Attn: AMXTH-CO-P
Aberdeen Proving Ground, .Maryland 21010-5401
Commander
U.S. Army Toxic and Hazardous .Materials Agency
Attn: AMXTH-TE-D
Aberdeen Proving Ground, Maryland 21010-5401
