x>EPA
            United States
            Environmental Protection
            Agency
           Industrial Environmental Research
           Laboratory
           Research Triangle Park NC 27711
EPA-600/2-78-125
June 1978
            Research and Development
Development of a
DDT Manufacturing
and  Processing
Plant Waste
Treatment System

-------
                 RESEARCH REPORTING SERIES


Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional  grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

    1. Environmental Health Effects Research

    2. Environmental Protection Technology

    3. Ecological Research

    4. Environmental Monitoring

    5. Socioeconomic Environmental Studies

    6. Scientific  and Technical Assessment Reports  (STAR)

    7. Interagency Energy-Environment Research and Development

    8. "Special"  Reports

    9. Miscellaneous Reports

This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
                        EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.

This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                         EPA-600/2-78-125
                                                  June 1978
Development of a DDT Manufacturing
        and Processing  Plant Waste
              Treatment  System
                            by

               M. Sobleman, K.H. Sweeny, and E.D. Calimag

               Montrose Chemical Corporation of California
                         P.O. Box 147
                     Torrance, California 90507
                       Grant No. 804293
                    Program Element No. 1BB610
                 EPA Project Officer: David K. Oestreich

               Industrial Environmental Research Laboratory
                 Office of Energy, Minerals, and Industry
                  Research Triangle Park, NC 27711
                         Prepared for

               U.S. ENVIRONMENTAL PROTECTION AGENCY
                  Office of Research and Development
                     Washington, DC 20460

-------
                       ABSTRACT
     A process  for treating DDT manufacturing waste has
been investigated.  In this process, concentrated wastes
containing  a  few hundred to a few thousand mg// of DDT,
ODD and DDE as  well as tars, ~20% Na2SO4, and ~3% sodium
p-chlorobenzenesulfonate were extracted in various con-
tinous liquid-liquid extractors with monochlorbenzene
 (MCB) and heptane as extractants.

     A primary  objective of this study was to determine
the existence of a suitable method for making the ex-
traction, and three devices were shown to make good ex-
tractions of  the caustic waste.  The Ross in-line homo-
genizer and the Podbielniak continuous contactor were
both able to  reduce the DDT and DDE content to as low
as ~2 rag/a , and a baffled turbine mixer in a 3-stage
array reduced the DDT + DDE level to as low as 0.2 mg/£ .

     A second-stage extraction of the aqueous effluent
from the Podbielniak unit reduced the DDT content to
detection limits, and the DDE to ~0.1 mg/£.

     Monochlorbenzene  (MCB) has been shown to be a suit-
able extractant, and appears superior to heptane, used
in earlier  studies.  The use of MCB provides a less flam-
mable solvent,  and potentially permits the MCB laden
with DDT, DDE to be recycled to process, thus avoiding
DDT, DDE disposal by incineration or other processes.

     Major  problems have occurred relative to adequate
phase separation with excessive amounts of solvent being
carried over  into the aqueous phase.  Reasonably good
coalescence can be achieved on a glass wool mat, but the
treated liquor  contains 200 to 4000 mg/& of MCB which is
equal or greater to that contained in the untreated waste
(reported solubility of MCB in water is 382 mg/£).  The
formation of  emulsions between the aqueous phase and the
solvent causes  confusion on the efficiency of extraction.
The chemical  analysis of the aqueous may not be accurate
because of the  suspended or dissolved MCB in the aqueous
phase,  which  could solubilize DDT and homologs.

-------
     Chemical reduction of the MCB was shown to have  limited  applicability,
with six passes through a catalyzed iron reductive column degrading MCB
from 225 mg/Jl to 22 mg/H  (90% reduction) .

     A further problem has been that the densities of the aqueous waste
and of MCB are virtually the same, so that gravity separation of the phases
is impossible.  Dilution of the waste allows phase separation and was used
in these studies, but it also requires the treatment  of a greater volume
of waste.

     The cost of an effective, practical process cannot be estimated because
the coalescence/phase separation process remains unsolved, as are the methods
for removal of residual MCB and sodium salt of monochlorbenzene sulfonic
acid  (NaMCBS).

     The purpose of this study was to determine whether an alternative
method to disposal of DDT manufacturing wastes could  be developed.  The
method which was explored was judged to be unworthy of further development.
                                   111

-------
                    ACKNOWLEDGMENT
     We would like to acknowledge the kind assistance of
Mr. D. K. Oestreich, Project Officer, for his aid in many
aspects of this program.

     This program has been carried out as a cooperative
effort between the Montrose Chemical Corporation of
California, and Envirogenics Systems Company as a major
subcontractor.

     Participating from Montrose have been Messrs.
Max Sobelman, Vice President - Operations and Grant
Director; John Kallok, Plant Manager as Assistant
Grant Director; Guy Dimichele, Chemical Engineer;
Walter Carey, Chief Control Chemist; Bernard Bratter,
DDT Plant Superintendent; and Rick Wilson and Emanuel
Maya, Chemical Engineers.  The Envirogenics effort
was directed by Dr. K. H. Sweeny and carried out by
Mr. E. D. Calimag, both under the direct supervision
of Dr. C. W. Saltonstall, Jr., Director of Research
and Development.  The assistance of all participants
in the project is appreciated.

     Valuable assistance has also been provided by
Mr. Joseph Lesnefsky and Mr. Gordon Davies of Baker
Perkins, and Mr. Robert Emlick and Mr. Ronald Reid
of the Charles Ross Company.
                           IV

-------
                               CONTENTS
Abstract	ii
Acknowledgment	iv
Figures	vi
Tables	vii

     1.  Introduction  	   1
     2.  Conclusions	   4
     3.  Recommendations	   6
     4.  Podbielniak Continuous Contactor  	   7

             A.  Description of Podbielniak Contactor  	   7
             B.  Experimental Results  	   7
             C.  Summary and Analysis  of the Podbielniak
                 Contactor Data	27
             D.  Second-Stage Extraction.  ........  	   29

     5.  Ross In-Line  Homogenizer	31

             A.  Description of Ross In-Line Homogenizer.  ......   31
             B.  Experimental Results  	   31
             C.  Summary of Tests  With Ross In-Line
                 Homogenizer	58

     6.  Baffled Stirrers  	   59

             A.  Description of Stirrer System	59
             B.  Experimental Results  .	59
             C.  Summary of Baffled Stirrer Tests  	   63

     7.  Other Mixing  Systems  	   64

             A.  Waring Blendor	64
             B.  Ross  Laboratory Homogenizer	65
             C.  Polytron  Homogenizer  	   65

     8.  Chemical  Reduction of MCB in  the  Aqueous  Extracts	68

     9.  Analysis  of Results	71

References	74

Glossary	75
                                    v

-------
                        FIGURES


Number                                                   Page

  1    Experimental Setup for Potfbielniak
       Contactor Tests	8

  2    Podbielniak Continuous Contactor 	  9


  3    Ross In-Line Homogenizer ..... 	 33
                             VI

-------
                          TABLES
Number                                                   Page

   1   Typical Composition of Waste Discharge
       to Class I Landfill	2

   2   Analysis of Typical Aqueous Waste	10

   3   Operation of Podbielniak Contactor at 3000 RPM . . 11

   4   Operation of Podbielniak Contactor at 4800 RPM . . 11

   5   Operation of Podbielniak Contactor at 4800 RPM
       and 250 m£/min MCB Solvent Flow	12

   6   Operation of Podbielniak Contactor at 7000 RPM . . 12

   7   Operation of Podbielniak Contactor at 7000 RPM
       and 2.0:1 Ratio of Waste-to-Solvent Flow	13

   8   Operation of Podbielniak Contactor at 7800 RPM
       and 1.5:1 Ratio of Waste-to-solvent Flow 	 14

   9   Operation of Podbielniak Contactor at 9500 RPM . . 15

  10   Operation of Podbielniak Contactor at 9500 RPM
       Effect of Increased Outlet Pressure of Waste Flow
       and Decreased Pressure Solvent Phase 	 16

  11   Operation of Podbielniak Contactor at 9500 RPM
       0.3:1 Ratio of Waste-to-solvent Flow	16

  12   Operation of Podbielniak Contactor at 9500 RPM
       0.04:1 Ratio of Waste-to-solvent Flow	17

  13   Operation of Podbielniak Contactor at 9500 RPM
       3.0:1 Ratio of Waste-to-solvent Flow	18

  14   Operation of Podbielniak Contactor at 9500 RPM
       3.0:1 Ratio of Waste-to-solvent Flow
       Effect of Pressure Variations	18

  15   Operation of Podbielniak Contactor at 9450 RPM
       4.0:1 Ratio of Waste-to-solvent Flow	20
                            vxi

-------
                          TABLES (Coht'd)
Number                                                   Page

  16   Operation of Podbielniak Contactor at 9450 RPM
       2.8 Ratio of Waste-to-Solveht Flow . . ;	21

  17   Operation of Podbielniak Contactor at 9450 RPM
       3.7:1 Ratio of Waste-to-Solveht Flow	22

  18   Operation of Podbielniak Contactor at 9450 RPM
       2.5:1 Ratio of Waste-to-Solvent Flow	22

  19   Operation of Podbielniak Contactor at 9450 RPM
       4.5:1 Ratio Waste-to-solvent Flow	23

  20   Operation of Podbielniak Contactor at 9450 RPM
       4-0:1 Ratio of Waste-to-solvent Flow	24

  21   Operation of Podbielniak Contactor at 9450 RPM
       2.0:1 Ratio of Waste-to-solvent Flow ....;.. 24

  22   Operation of Podbielhiak Contactor at 9450 RPM
       2.0:1 Ration of Waste-to-solvent Flow
       Total Flow 990 m /tain.	 25

  23   Operation of Podbielniak Contactor at 9450 RPM
       4.0:1 Ratio of Waste-to-solvent Flow
       Total flow 500 mt/min.	  . 26

  24   Operation of Podbielhiak Contactor at 9450 RPM
       1.0:1 Ratio of Waste-to-Solvent Flow
       Total Flow of 500 mi/mih	26

  25   Operation of Podbielniak Contactor at 9450 RPM
       1.5:1 Ratio of Waste-to-Solvent Flow
       Total Flow of 550 m£/min	27

  26   Summary of Podbielniak Contactor Evaluation Tests.28

  27   Second-Stage Extraction of Podbielniak-
       Extracted Waste	30

  28   Ross In-Line Homogenizer - Coalescers Evaluated.  .35

  29   Ross In-Line Homogenizer Operation - Evaluation
       of Coalescers	37
                            Vlll

-------
                          TABLES (Cont'd)
Number                                                   Page

  30   Ross In-Line Homogenizer Operation - Effect
       of Mixing Time on Extraction	38

  31   Ross In-Line Homogenizer Operation - Pressure-
       Aided Flow Through Coalescer  	 39

  32   Ross In-Line Homogenizer Operation - Second-Stage
       Extraction of Homogenizer - Extracted Waste.... 42

  33   Operation of Ross In-Line Homogenizer - Effect
       of Homogenizer Speed of Rotation on Extraction
       Silicone-Treated Glass Wool Coalescer	45

  34   Operation of Ross In-Line Homogenizer - MCB
       Recovered When 375 m£ Mixed with Waste and
       Passed Through Coalescer 	 46

  35   Effect of Mixing Time on the  Extraction With
       Ross In-Line Homogenizer 	 47

  36   Operation of Ross In-Line Homogenizer Tests With
       Pressure-Aided  (10 psi) Flow  Through Coalescer  .  . 48

  37   Operation of Ross In-Line Homogenizer - Evaluation
       of Sand and Gravel as Coalescers	50

  38   Operation of Ross In-Line Homogenizer - Evaluation
       of Fine Gravel Coalescer	51

  39   Operation of Ross In-Line Homogenizer - Evaluation
       of Silicone-Treated Sand Coalescer 	 52

  40   Operation of Ross In-Line Homogenizer - Tests
       With Glass Wool Coalescer	53

  41   Operation of Ross In-Line Homogenizer - Tests
       With 3 Inch Dia x 30 Inch Deep Glass Wool
       Coalescer	54

  42   Operation of Ross In-Line Homogenizer - Tests
       With Pressure-Aided Flow Through Glass Wool
       Coalescer	54

-------
                          TABLES (Cont'd)
Number                                                    Page

  43   Operation of Ross In-Line Homogenizer  - Second
       Extraction of Waste With Homogenizer  .......  55

  44   Operation of Ross In-Line Homogenizer  - Extraction
       of Waste With Heptane at 7000 RPM .........  56

  45   Operation of Ross In-Line Homogenizer  - Extraction
       of Waste With Heptane at 10,000 RPM ........  57

  46   Operation of Ross In-Line Homogenizer  - Extraction
       of Waste With Heptane at 10,000 RPM -
       Evaluation of Ant i foam Agent  ...........  57

  47   Extraction of Montrose Waste  in Baffled Tank
       Using Straight Blade, Curved  Blade and Tapered
       Blade Turbine Stirrers ..............  60

  48   Effect of Blade-Baffle Clearance on Extraction
       Efficiency - Analysis of Aqueous Phase After
       Extraction, fig/H .................  61
  49   Multiple Extraction of Montrose Waste With
       Baffled Straight-Blade Turbine Mixer  .......  62

  50   Temperature of Waste-MCB Mix as Function  of
       Mix Time in Waring Blendor  ............  64

  51   Extraction Tests With Brinkman Polytron.
       Homogenizer ....................  66

  52   Analysis for Monochlorbenzene DDT  and DDE
       After Treatment in Reduction Column.  .......  69
                            x

-------
                        SECTION 1

                      INTRODUCTION
     The objective of these studies was to investigate a
process for the removal of DDT and related products from
DDT manufacturing wastes currently discharged to a Class
I landfill, and the subsequent treatment or disposal of
the removed DDT if necessary.  The successful development
and demonstration of the subject process would lead to
very low discharge of DDT* and related compounds since
process recycle previously installed has already elimi-
nated all but the landfilled discharge.

     Use of DDT in the U. S. has been eliminated by Federal
action, and domestic use is allowed only for emergency
situations and under special permit.  Examples of such
use recently was for control of Bubonic Plague vector
in Colorado, and control of the spruce budworm that
threatened the forests of the Pacific Northwest.

     DDT is still manufactured for export at one plant
in the U.S.  The majority of exports are made for the
purpose of controlling malaria vector in Asia, Africa,
Central and South America.  Consequently, the only pos-
sible release of DDT to the environment in the U.S. could
be from the manufacturing process.  The manufacturer has
taken considerable effort to minimize his water discharges
by employing extensive recycle of process water.  Thus,
the only DDT-containing wastes leaving this plant are
approximately 30,000 gallons per day of caustic waste
which go to a Class I landfill.

     This DDT manufacturing plant makes use of an extensive
recycle system which eliminates all but about 2500 gal/day
of steam boiler blowdown and sanitary waste.  These are
discharged to the county sewer.  The current operation
insures that all government regulations for water standards
are being met.
*DDT + ODD + DDE; see Glossary for definition of terms.

                            1

-------
     The principal waste discharge to the Class I landfill is neutralized
caustic liquor from DDT washing operations, tar pot drainings, spills  and
tank drainings.3 The analysis of these wastes is typically shown in
Table 1.
                                  TABLE 1

                  TYPICAL COMPOSITION OF WASTE DISCHARGE
                            TO CLASS I LANDFILL
            Product                                Pound/Day
Sodium Sulfate                                       21,615

Sodium Salt of Monochlorbenzene                       3,670
SulfoniczAcid  (NaMCBS)

Caustic  Soda                                             50

DDT  (+ DDE, ODD)                                        119

Miscellaneous  (tars, etc.)                              139

Water                                               255,550
                                    TOTAL           281,143
 The  total volume of waste may vary somewhat as additional water is added to
 the  landfill discharge.  It is this waste for which treatment is considered
 in this report.

     The use of solvents to extract and recover pollutants from waste waters
 is not a new concept and has been successfully applied on a number of
 different industrial waste in the extraction of waste water from styrene
 manufacture; butyl acetate and isopropyl ether for the removal of phenolics
 in the byproduct coke industry; and 2-ethylhexanol in the removal of ethanol,
 ethylene dichloride, ethylene chlorohydrin and chloral hydrate from neutralized
 oxychlorination waste water.
aEarhart, J. P., and K. W. Won, et al.  University of California,  Recovery
of Organic Pollutants Via Solvent Extraction.  Chemical Engineering  Progress,
May 1977.  pp. 67-73.

-------
Witt and Forbes  describe several additional extraction
processes successfully employed in the petroleum industry,
also noting a few caveats:  "This solvent must be inex-
pensive. .. (and) it must be almost completely insoluble
in water, since the portion that dissolves is lost to the
operation and may be a cause of pollution in the water
effluent."

     Earlier studies under EPA Contract 68-01-0083
(Reference 1) demonstrated that the DDT, ODD and DDE
could be effectively removed from the caustic waste pro-
vided a high-shear mixing by multiple passage through
a centrifugal pump was found effective for DDT removal.
Although the principle of high-shear mixing for continuous
extraction of DDT mixing process was not considered suit-
able for ready scale-up to full plant operation.

     The primary emphasis of this program was the evalu-
ation of a series of high-shear mixers which could po-
tentially provide a practical means for the requisite
high-shear mixing and be readily scaled-up to useful
plant treatment capacities.  Among the mixing devices
considered have been the Podbielniak continuous contactor
and several types of continuous and batch-type mixers.
Further consideration was given to the solvent employed
for the extraction and ultimate disposal systems that
could be utilized.

     The ultimate disposal of the DDT and DDE could pos-
sibly be handled in either of two ways.  If a solvent,
such as heptane  (used in the earlier studies under Con-
tract 68-01-0083, Reference 1), is employed, the removal
of the heptane and its return to the system, and a means
for disposal of the DDT and DDE must be provided.  The
heptane can be removed by distillation, and the DDT and
DDE in the still bottoms can be destroyed by Friedel-
Crafts reaction, incineration, etc.  However, a more
attractive approach would appear to involve the use of
a suitable solvent so that the DDT and DDE can be re-
turned to the process.  The use of monochlorbenzene  (MCB)
as the solvent, with the return of the MCB laden with
DDT and DDE to the DDT condensation step of the reaction
would eliminate the need for a degradation process to
destroy the extracted DDT and DDE, as well as conserv-
ing these two materials in the process.  Major effort
has therefore been directed towards using MCB as the
extractant.
aWitt, P.A., Jr., and M. C. Forbes.  By-Product Recovery
 Via Solvent Extraction.  Chemical Engineering Process,
 October 1971.  pp. 90-94.

-------
                       SECTION 2

                      CONCLUSIONS
     It has been demonstrated that high-shear mixing of
DDT plant aqueous alkaline waste with monochlorbenzene
(MCB) solvent is effective for the extraction of DDT, ODD
and DDE from the waste.  Levels as low as ~2 mg/£ have
been obtained (total DDT + DDE) when the Ross homogenizer
or Podbielniak contactor are employed, and as low as 0.2
mg/£ for a three-stage baffled turbine stirrer.  A second-
stage extraction (without high-shear mixing) of the Podbiel-
niak-treated waste has reduced the DDT + DDE to 0.1 -
0.2 mg/£.  However, the process has not been shown to
give adequate coalescense of the phases, so that clear-
cut phase separation has not been regularly obtained.
Extraction ratios as high as 4 volumes waste/1 volume
solvent appear satisfactory for extraction of DDT, ODD
and DDE.  Of the various high-shear mixers tested, the
three-stage baffled stirrer with the 3 inch diameter x
42 inch deep coalescer was the best system.

     The wastes were shown to contain ~200 - 1000 mg/£
MCB as received and treatment did not reduce, but in some
cases (homogenizer) increased, the MCB level, apparently
because of ineffective coalescence.  Attempted degradation
of the MCB by chemical reduction led to regular but slow
destruction of the extractant.  Six passes through an
iron reduction column was necessary for a 90% removal
of MCB.  The degradation product was cyclohexanol.

     The use of MCB as a solvent/extractant is apparently
more effective and less flammable than heptane.  Use of
MCB as the solvent potentially permits recirculation of
the extractant to the condensation step of the process
so that further disposal process would not be necessary.

     It was found that the density of the aqueous waste
is regularly equal or nearly equal to MCB.  Consequently,
dilution of the waste to give a density difference suffi-
cient for gravity separation of the phases is necessary.
Such dilution would require the handling and processing
of an additional volume of waste.
                           4

-------
     The less expensive baffled stirrers and in-line
homogenizer appeared to function at least as well as
the Podbielniak, so that the large capital investment
for the continuous contactor is probably not necessary.
Cost estimates for an effective practical process cannot
be formulated because coalescer and phase separation
problems remain unsolved.

-------
                        SECTION 3

                     RECOMMENDATIONS
     The feasibility of removing DDT and DDE from aqueous
DDT manufacturing waste using a high shear mixing solvent
extraction process has been demonstrated by tests with the
Podbielniak continuous contactor, the Ross In-Line Homogen-
izer, and baffled turbine stirrers.  Reduction in DDT + DDE
levels to 0.2 to 2.0 mg/£ were regularly obtained.  The
use of MCB (monochlorbenzene) as a solvent/extractant was
deemed attractive because the extracted DDT, ODD and DDE
could potentially be returned directly to the process with-
out the added complexity of ultimate disposal of these
materials.

     However, several serious problems remain.  The
aqueous raffinate contains 200 to 1000 mg/£ of MCB after
processing, an amount that is both economically and en-
vironmentally not acceptable.  Two possibilities exist
to remove the MCB from the extracted aqueous waste.  Tests
with catalytic reduction indicated a reaction rate too
slow for practical use.  Removal by sorption, either by
carbon or macroreticular resins, would be expected to be
expensive and difficult because of the viscosity, high
solids content, and high amount of materials to be removed
from the raffinate.  The densities of the waste and the
preferred extractant, MCB, are virtually identical so
that dilution of the waste is necessary before gravity
phase separation can be achieved; this requires handling
a larger volume of waste.  Further, adequate coalescence
and phase separation were not obtained, also leading to
excessive and unacceptable MCB loss.

     With these factors in mind, it is recommended that
scaleup of the process not be attempted.
                             6

-------
                       SECTION 4

           PODBIELNIAK CONTINUOUS CONTACTOR


A.   Description of Podbielniak Contactor

     The Podbielniak contactor is constructed with a series
of concentric "bands" perforated with many holes.  The
unit is rotated at a high speed (up to ~10,000 RPM) ,  and
the heavier liquids are thrown outward by centrifugal force,
while lighter liquids are displaced inwards.  This action
causes a series of counter-current contacts as the fluids
pass through the perforated bands, thereby producing a
multiple stage extraction.  Either the heavy or light liquid
can be made predominately continuous or dispersed by con-
trol of the position of the principal interface; this is
done by regulating the pressure on the liquid streams.
The centrifugal force (~5000 x g) and the short residence
time generally give conditions whereby liquids do not
emulsify, effluents are clear and entrainment and settl-
ing problems are largely eliminated.

     The experimental set-up used in these tests is shown
in Figure 1, and a view showing the construction of the
Podbielniak contactor is shown in Figure 2.

B.   Experimental Results

     Two groups of experiments will be described in the
section to follow.  In the first group, a variety of con-
ditions were explored leading to effective operation of
the Podbielniak contactor.  In the second group, the opera-
tion of the "Pod" at different solvent-to-waste ratios,
etc., at the discovered operating conditions is described.

     1.  Exploratory Runs - Range Finding Experiments

     The operating variables for the Podbielniak unit are
the input flow rates of the aqueous waste and extractant,
the input and output pressure of the heavy and light liquids,
and the rate of rotation of the contactor.

-------
00

i
MCB
Extractant







1
Waste
















P, , .. 1
1
Pump
Ini
in]


1 —
•*^l )
Pump
Flowmeter
m
b
(7\-
(*J~-
Extractai
Input
Pressure
Waste
QUt P*T ff* fl^l TIT* f*~
T


Flowmeter




it


\











/








^ /T "1
1 T n 1
fk
1
, , '
1 '
1 ! l !
', ^ '-
~r"Y4_ _.



PadhielnuLlc C rnnt.n.c. to


MCB 4- Extracted
DDT. ODD. DDE
A
Extractant
/~\ Output
V_/ Pressure

* Waste
	 1I» /7s Output
^•p Pressure
-.^ 1 -^, Extracted
•>^ "^ Waste

-
p

                Figure 1.   EXPERIMENTAL SETUP FOR  PODBIELNIAK CONTACTOR TESTS

-------
Heavy liquid irv-»-a
Light liquid out
                                              —Rotor
                                              ••-Rotor guard
Rotating shaft
   Heovy liquid out
   Light liquid in
                                                     .-Base
  Figure 2.   PODBIELNIAK  CONTINUOUS  CONTACTOR

-------
 First Test Series

      In the first test series,  the waste was diluted to a
 density of 1.045 while the density of the chlorobenzene
 (MCB)  was 1.096.  The densities of the waste and MCB were
 found to be virtually identical when the waste was un-
 diluted.  The analysis of a typical waste sample is shown
 in Table 2.
                         TABLE 2

	ANALYSIS OF TYPICAL AQUEOUS WASTE*

                         Analysis,  mg/jf


 p,p'-DDT                     56

 o,p'-DDT                     14

 p,p'-DDE                     73

 o,p'-DDE                     18

 chlorobenzene               410

 density, g/m£                 1.045


 *Diluted for test
      In the initial test,  aqueous waste at 1690 m /min
 at a pressure of 10 psig,  and MCB at 200 ra£/min were fed
 to the contactor.   The outlet valves were opened to give
 zero pressure for  both phases,  and the contactor was ro-
 tated at 3000 RPM.  No clear separation of phases was ob-
 tained, with both  layers being dark colored.  The analysis
 of the phases gave the results as shown in Table 3.
                            10

-------
                        TABLE 3

    OPERATION OF PODBIELNIAK CONTACTOR AT 3000 RPM
                    ANALYSIS, mg/£
                   Aqueous Phase      Solvent Phase
p
o
p
p
o
.p
/p
/p
/p
/p
'-DDE
'-DDE
'-ODD
'-DDT
'-DDT
75.
36.
—
180.
125.
2
7

0
0
39
11
—
—
132
.7
.5


.0
     The extraction was poor, probably because a clear
separation of phases was not obtained.

     In the second test, the rate of rotation of the con-
tactor was increased to 4800 RPM, and the inlet pressure
of the aqueous phase increased to 20 psig (flow rate 3808
m£ /min).  Again no clear separation of phases was obtained,
and the extraction efficiency was similar to the first
test.  The results are shown in Table 4.

                        TABLE 4

    OPERATION OF PODBIELNIAK CONTACTOR AT 4800 RPM

p , p ' -DDE
0 , p ' -DDE
p , p ' -ODD
p/P'-DDT
0 , p ' -DDT
ANALYSIS, Tag /H
Aqueous Phase
72.4
32.1
—
66.7
110.0
Solvent Phase
65.2
28.5
—
13.3
147.0
                           11

-------
      In the third test,  the solvent flow was increased
 to ~250 m£/min,  and the  output pressure increased from
 a nominal 0 psi  to 1.0 psig.  Other conditions were the
 same.   In this test, a clear separation of phases was ob-
 tained, but analyses of  the phases did not show appreci-
 ably better extraction.   The analyses are shown in Table 5,
                         TABLE 5

   OPERATION OF PODBIELNIAK CONTACTOR AT 4800 RPM AND
	250 ml/min MCB SOLVENT FLOW	

                     ANALYSIS, mg/£

                    Aqueous Phase        MCB Phase


 p,p'-DDE                44.0              64.5

 o,p'-DDE                11.2              24.3

 p,p'-DDD

 p,p'-DDT               133.0

 o,p'-DDT                96.8              76.1
      Following this,  the speed of rotation of the contactor
 was increased from 4800 to 7800 RPM.  The other conditions
 remained the same, except that the flow rate of the aqueous
 waste decreased from 3800 mJ^/min to 1674 m£/min.  A clear
 separation of phases  was obtained, and substantially better
 extraction was shown.  The analyses are shown in Table 6.


                         TABLE 6

   OPERATION OF PODBIELNIAK CONTACTOR AT 7000 RPM
ANALYSIS, mg/1
Aqueous Phase Solvent Phase
p,p'-DDE
o , p ' -DDE
p , p ' -ODD
p , p ' -DDT
o , p ' -DDT
59.3 72.1
26.1 36.6
23.3
133.0
110.0
                            12

-------
      In the next test,  the solvent flow rate was 822 m£/
 min,  while  the solvent  pressure was 20 psig input and out-
 put,  and the waste 35 psig input and 5 psig output.   In this
 test,  a clear separation of phases was not obtained.

      The aqueous phase  effluent contained about 50% solvent
 phase and the MCB solvent phase contained about 10% aqueous
 phase.   In  this test, and succeeding tests, there was no
 sharp separation of phases, but the. MCB outlet would have
 some  aqueous phase which separated rapidly, and the aqueous
 waste outlet some MCB which separated quickly.  Large-scale
 operation of the Podbielniak would appear to require a con-
 tinuous decanter to separate the phase in both the heavy
 and light liquid outlet streams, though it is possible that
 with  fine tuning,  a larger unit may provide better separa-
 tion.   The  phases were  allowed to separate and the layers
 tested.  The analyses showed about 98% extraction.  The
 results are shown in Table 7.

                          TABLE 7

 OPERATION OF PODBIELNIAK CONTACTOR AT 7000 RPM AND
	2.0:1 RATIO OF WASTE-TO-SOLVENT FLOW	

                     ANALYSIS, mg/£
                    Aqueous Phase     MCB Solvent
p , p ' -DDE
0,p'-DDE
p , p ' -ODD
p , p ' -DDT
2.5
0.5
- -
- -
20.5
54.5 ' \.
- -
87.0
 o,p'-DDT                - -              21.0
      This test showed good extraction after the separation
 of phases,  with no detected DDT or ODD,  and only 3 mg/J?  of
 the DDE isomers.

      As a further test,  the flow rate of the aqueous phase
 was reducted from 1865 to 1234 m£/min, the inlet pressure
 of the MCB solvent increased from 20 to 25 psig, the outlet
 pressure increased from 20 to 35 psig, the inlet pressure
 of the aqueous phase increased from 35 to 50 psig.  The rate
 of flow of the MCB solvent remained at 822 mJ2/min, and the
 speed of rotation 7800 RPM.  The results of this test are
 shown in Table 8.
                            13

-------
                         TABLE 8

   OPERATION OF PODBIELNIAK CONTACTOR AT 7800 RPM AND
	1.5:1 RATIO OF WASTE-TO-SOLVENT FLOW	

                     ANALYSIS, mg/g

                    Aqueous Phase     MCB Solvent


 p,p'-DDE                1.0              40.0

 o,p'-DDE                •—                7.5

 p,p'-DDD

 p,p'-DDT                —               16.5

 o,p'-DDT                —                4.5
      In this test,  the aqueous phase contained about 30%
 MCB (after phase separation),  while the solvent stream
 appeared to be free of the aqueous phase.

      At this point, a pressure regulator was installed
 in the solvent effluent stream, the rate of rotation of
 the contactor increased from 7800 to 9500 RPM (full-speed),
 and the pressures adjusted on the basis of the manufac-
 turer's recommendations for the phase densities and speed
 of rotation.  These values were:  aqueous phase input 35
 psig,  output 10.7 psig; and MCB solvent 35 psig input and
 20 psig output.  In this test, both output streams were
 about 50% solvent,  50% aqueous.  The analysis of the
 separated phases is as shown in Table 9.
                            14

-------
                        TABLE 9

    OPERATION OF PODBIELNIAK CONTACTOR AT 9500 RPM
                    ANALYSIS, mg/£

                   Aqueous Phase     MCB Solvent
p / p
o,p '
P'P1
p/p1
o,p'
-DDE
-DDE
-ODD
-DDT
-DDT
44
6
-
62
13
.5
.5
-
.0
.0
31
3
-
11
3
.5
.5
-
.5
.5
     The extraction was not effective, and indeed, there
was apparently more of the DDE and DDT in the aqueous layer
than in the solvent phase.  This anomalous result is
probably due to solvent carry-over into the aqueous phase.

     In the next test, the outlet pressure on the aqueous
phase was increased from 10.7 to 20 psig, while the input
pressure remained 35 psig; and the inlet pressure for the
solvent phase was decreased from 35 to 20 psig, and the
outlet pressure decreased from 20 psig to zero psig.  Other
conditions remained the same.  In this experiment, the
light phase output, which should have been the extracted
aqueous phase was MCB solvent, and the supposed heavy
phase was the less dense aqueous phase.  The analysis of
the phases  is shown in Table 10.
                            15

-------
                       TABLE 10

    OPERATION OF PODBIELNIAK CONTACTOR AT 9500 RPM
       EFFECT OF INCREASED OUTLET PRESSURE OF
    WASTE FLOW AND DECREASED PRESSURE SOLVENT PHASE
ANALYSIS, mg/i
Aqueous Phase MCB Solvent
p , p ' -DDE
o , p ' -DDE
p , p ' -ODD
p , p ' -DDT
o , p ' -DDT
3.5 90.5
32.5
—
100.0
20.0
     Prior to the next test, a back-pressure valve was
installed in the aqueous discharge line, the MCB solvent
flow was adjusted to 2500 m^/min, and the aqueous flow
to 702 mf/min.  The MCB solvent inlet pressure was 60
psig, while the outlet pressure was 40 psig and the aque-
ous phase 60 psig inlet and 30 psig outlet.  Other con-
ditions were the same as previously.  The analyses are
as shown in Table 11.

                       TABLE 11

    OPERATION OF PODBIELNIAK CONTACTOR AT 9500 RPM
           0.3:1 RATIO OF WASTE-TO-SOLVENT FLOW

p,p'-DDE
0 , p ' -DDE
p , p ' -ODD
p , p ' -DDT
o , p ' -DDT
ANALYSIS, mg/£
Aqueous Phase
73.5
13.5
--
172.0
62.0
MCB Solvent
76.5
47.5
--
144.0
32.0
                           16

-------
      Despite the high ratio of solvent-to-aqueous waste,
 the extraction was poor.  In this test, the "heavy" phase
 contained about 80% aqueous phase (should have been all
 MCB)  and the "aqueous" outlet phase contained about 60%
 MCB.

      The ratio of solvent to aqueous waste was further
 increased in the next test to 8000 m//min MCB extractant
 and 308 mf/min aqueous phase.  The MCB inlet pressure was
 decreased from 60 psig to 40 psig, while the outlet pres-
 sure of the MCB was decreased from 40 to 10 psig and the
 aqueous outlet from 30 to 20 psig.  In this test the
 "aqueous" outlet contained about 98% solvent and the
 "solvent" outlet 60% aqueous waste.  The analysis of the
 phases again indicated a poor extraction.  The results
 are shown in Table 12.
                        TABLE 12

     OPERATION OF PODBIELNIAK CONTACTOR AT 9500 RPM
	0.04:1 RATIO OF WASTE-TO-SOLVENT FLOW

                     ANALYSIS,  mg/£
                    Aqueous Phase     MCB Solvent


 p,p'-DDE               61.5              83.5

 o,p'-DDE                8.0              30.0

 p,p'-DDD

 p,p'-DDT              184.0             170.0

 o,p'-DDT               66.0              48.0
      Changing the ratio of aqueous waste to solvent to
 3.0:1 (3300 m//min aqueous waste,  100 m£/min MCB solvent),
 and the pressures to MCB:inlet 20  psig,  outlet 40 psig;
 aqueous waste:inlet 45 psig,  outlet 10 psig, led to an
 effluent of which the solvent phase appeared free of aque-
 ous waste,  and the aqueous waste contained only about 10%
 MCB.  However,  the extraction was  still poor, probably
 because of  the solvent remaining in the aqueous phase.
 The results are as shown in Table  13.
                            17

-------
                       TABLE 13

   OPERATION OF PODBIELNIAK CONTACTOR AT 9500 PRM
        3.0:1 RATIO OF WASTE-TO-SOLVENT FLOW
                    AQUEOUS, mg/£
                   Aqueous Phase     MCB Solvent
p
0
p
p
o
/p1
,p'
/p1
/p1
/p1
-DDE
-DDE
-ODD
-DDT
-DDT
72
49
-
148
26
.5
.0
-
.0
.5
49
49
-
23
5
.5
.9
-
.5
.0
     In two additional tests, an adjustment of the pres-
sures, did not lead to significantly improved results, nor
a clear separation of the phases.  The test conditions and
the results of these tests are as shown in Table 14.
                       TABLE 14

    OPERATION OF PODBIELNIAK CONTACTOR AT 9500 RPM
         3.0:1 RATIO OF WASTE-TO-SOLVENT FLOW
  	EFFECT OF PRESSURE VARIATIONS	

                    TEST CONDITIONS
                                     1st       2nd
  	Test	Test

   Aqueous Waste Flow, m//min        3300      3300

   MCB Solvent Flow, mf/min          1100      1100

   Aqueous Waste Pressure, psig
     Inlet                             50         50
     Outlet                            20        ~0

   MCB Solvent Pressure, psig
     Inlet                             50         50
     Outlet                            30         30
 (Continued)                18

-------
                 TABLE 14 (Continued)
                    ANALYSIS, mg/£

                     1st Test             2nd Test

P'P
O f P
p 1 p
p/p
o,p

'-DDE
'-DDE
'-ODD
'-DDT
'-DDT
Aq
3.
30.
—
168.
49.

5
0

0
5
MCB Aq
80.5 5.0
17.5
— —
121.0 56.0
25.5 9.5
MCB
84.
—
—
104.
24.
5',


0
5
Second Test Series

     At this point, several changes were made in the pro-
cedure and method of operation, and significant improve-
ment is phase separation and extraction efficiency was ,
obtained.  These changes included a shortening of the ,/ -
piping to provide reduced resistance, particularly for   .
the outlet pressures; use of damped pressure gauges and
lower range flowmeters for more accurate measurement and
control of these variables; and starting the contactor
with solvent before turning on the aqueous waste pump,
rather than starting the flows simultaneously.

     The first test with the revised apparatus was made
with an aqueous waste flow rate of 800 m^/min and a MCB
solvent flow of 200 m//min.  The solvent and aqueous waste
inlet pressures were 13 psig, while the outlet pressures
were ~0 psig.  The rate of rotation of the contactor was
9450 RPM.  The results of this test are as shown in Table 15
                           19

-------
                       TABLE 15

    OPERATION OF PODBIELNIAK CONTACTOR AT 9450 RPM
         4.0:1 RATIO OF WASTE-TO-SOLVENT FLOW
                   ANALYSIS, mg/g
                   Aqueous Phase     MCB Solvent
p,p'-DDE               5.5              57.0

o,p'-DDE               1.0               19.5

p,p'-DDD

p,p'-DDT               —               98.5

o,p'-DDT               —               30.0
     Even though the aqueous phase outlet contained about
20% MCB after phase separation, the analysis of the aqueous
layer indicated good  extraction of DDT and homologs.
The efficiency of extraction was 97.0%.  As was explained
earlier, phase separation of MCB from aqueous layer appears
to take place quickly, but the aqueous outlet generally
contains an MCB layer, and the MCB outlet an aqueous layer
which must be separated before analysis.

Evaluation Tests

     With the results shown in the previous test, further
tests were made to confirm these findings, and to investi-
gate the effect of solvent-to-waste flow ratio and total
flow.

     In the first of these evaluation tests, the inlet
and outlet pressures remianed the same; and the flow of
aqueous waste was 453 m//niin, with solvent flow at 160
m//min, giving a waste-to-solvent ratio of 2.8:1.  In this
test, the aqueous waste raffinate contained about 15% sol-
vent, and the solvent effluent about 3% aqueous waste.
The results are as shown in Table 16.
                           20

-------
                       TABLE 16

    OPERATION OF PODBIELNIAK CONTACTOR AT 9450 RPM
          2.8 RATIO OF WASTE-TO-SOLVENT FLOW
                  ANALYSIS, mg/J2
                  Aqueous Phase      MCB Solvent
P/P
o,p
p/p
p/p
o,p
'-DDE
'-DDE
'-ODD
'-DDT
'-DDT
3.5 39.0
0.2 13.5
—
105.5
21.5
     The DDT isomers and ODD appear to be extracted more
efficiently than DDE (both isomers).  This difference is
believed to be due to the greater solubility of DDE in
water.                                                   /  .

     The extraction efficiency was 98.0% in this test,
with the DDE isomers analyzing 3.7 ing// and the DDT and
DDD undetected in the aqueous effluent.

     In another test, a flow rate of 563 rai/min of aqueous
waste and 151 m//min of MCB solvent was used, giving a
waste-to-solvent ratio of 3.73.  In this experiment, the
aqueous output contained about 13% MCB, and the solvent
outlet 2% aqueous waste.  The extraction was again effect-
ive, with an extraction efficiency of 98.0%.  The results
are as shown in Table 17.
                           21

-------
                       TABLE 17

    OPERATION OF PODBIELNIAK CONTACTOR AT 9450 RPM
         3.7;1 RATIO OF WASTE-TO-SOLVENT FLOW
                  ANALYSIS/ mg/£

                  Aqueous Phase      MCB Solvent
p,p'-DDE               3.0              44.5

o,p'-DDE               1.0              10.0

p,p'-DDD

p,p'-DDT               ~              112.5

o,p'-DDT               —               25.0
     In a further test, a flow rate of 500 ml/min of aqueous
waste and 200 m//min of MCB solvent was employed  (waste-
to-solvent ratio 2.5) with the inlet pressure  (both fluid
streams) at 12 psig.  The extraction was not as effective
as the previous tests.  Extraction efficiency was 94.4%.
The results are as shown in Table 18.
                       TABLE 18

    OPERATION OF PODBIELNIAK CONTACTOR AT 9450 RPM
         2.5:1 RATIOOF WASTE-TO-SOLVENT FLOW

p , p ' -DDE
0 , p ' -DDE
p , p ' -DDD
p,p'-DDT
0 , p ' -DDT
ANALYSIS, mg/£
Aqueous Phase
7.0
1.5
—
2.5
1.5
MCB Solvent
44.5
12.5
—
125.0
28.0
                           22

-------
      The  ratio of aqueous  waste-to-solvent was increased
 to  4.5  in the  next test, in which the  MCB solvent flow
 rate  remained  at  200  m//min,  while the aqueous waste flow
 rate  was  increased to 900  m^/min.  A 97.4% extraction
 efficiency was obtained.   The results  are as shown in
 Table 19.

                        TABLE 19                          '

    OPERATION  OF  PODBIELNIAK CONTACTOR AT 9450 RPM
	4.5;1 RATIO  WASTE-TO-SOLVENT  FLOW	

                   ANALYSIS,  mg/£

p , p ' -DDE
0 , p ' -DDE
p , p ' -ODD
p , p ' -DDT
0 , p ' -DDT
Aqueous Phase
3.5
1.0
—
1.0
0.35
MCB Solvent
51.0
12.0
—
120.0
35.0 /,
      The ratio of aqueous waste to solvent was adjusted
 to 4.0 in a test in which 800 mf/min of the aqueous waste
 was extracted with 200 m£/min of MCB solvent.   The aqueous
 waste effluent stream contained about 18% MCB, while the
 solvent stream appeared to be free of the aqueous waste.
 Extraction was excellent, with a 99.5% extraction effi-
 ciency,  and only 0.6 mg/l  of p,p'-DDE detected in the
 extracted waste stream.  The results are as shown in Table 20,
                            23

-------
                        TABLE 20

     OPERATION OF PODBIELNIAK CONTACTOR AT 9450 RPM
          4.0:1 RATIO OF WASTE-TO-SOLVENT FLOW	
                   ANALYSIS,  mg/£
                   Aqueous Phase      MCB Solvent
 p,p'-DDE               0.6               44.0

 o,p'-DDE               —                 9.6

 p,p'-DDD               —                 —'-'•

 p,p'-DDT               —                55.6

 o,p'-DDT               —                16.0
      The aqueous waste-to-solvent ratio was adjusted to
 2.0 in the next test,  with an aqueous waste flow rate of
 1000 m//min,  and the MCB solvent flow rate 500 mJ?/min.
 In this test  the inlet pressure of the solvent was in-
 creased from  14 to 18 psig,  and the aqueous waste to 16
 psig.   The effluent aqueous  waste contained about 27% MCB,
 while the solvent stream contained  is less than 1% of
 aqueous waste.   Again, the extraction was excellent, with
 only 1.0 mg/£  of p,p'-DDE in the extracted waste and a
 99.1% extraction efficiency.  The results are as shown
 in Table 21.

                        TABLE 21

     OPERATION OF PODBIELNIAK CONTACTOR AT 9450 RPM
	2.0:1 RATIO OF WASTE-TO-SOLVENT FLOW	

                   ANALYSIS,  mg/£
                   Aqueous Phase      MCB Solvent
p , p ' -DDE
o , p ' -DDE
p , p ' -DDD
p , p ' -DDT
0 , p-DDT
1.0 37.0
9.0
--
55.0
13.0
                           24

-------
      The efficiency of extraction was next investigated
 at the same 2:1 aqueous waste to MCB solvent flow rate,but
 with the total flow reduced from 1500 nU/min to 990 m//min.
 In this test the aqueous waste feed rate was 660 ml/min,
 and the MCB solvent was fed to the contactor at 330 m//
 min.  The pressure for the two inlet streams (waste and
 solvent)  was 15 psig.   In this test, the aqueous effluent
 stream contained about 34% MCB solvent while the solvent
 stream appeared to be  clear of aqueous waste.  The results
 of this test indicate  a 94.9% extraction efficiency.  The
 results are as shown in Table 22.

                        TABLE 22

     OPERATION OF PODBIELNIAK CONTACTOR AT 9450 RPM
          2.0:1 RATIO OF WASTE-TO-SOLVENT FLOW
	TOTAL FLOW 990 m£/min	

                   ANALYSIS, mg/£
                   Aqueous Phase      MCB Solvent
p,p'-DDE
o,p'-DDE
p , p ' -DDD
p,p'-DDT
o , p ' -DDT
1.0
0.35
—
0.15
1.0
26.5
7.0
—
9.0
4.5
      In the next test, the aqueous waste-to-solvent flow
 was increased to 4.0, with 400 mf/min aqueous waste and
 100 mi/min MCB solvent, for a combined total flow of 500
 mf/min.  The results did not produce satisfactory extrac-
 tion, possibly because of the low solvent flow rate.  The
 extraction efficiency was 78.1%.  The aqueous effluent
 appeared to contain no MCB, and the solvent stream had
 about 1% aqueous phase in it.  The analysis is as shown
 in Table 23.
                            25

-------
                       TABLE 23

    OPERATION OF PODBIELNIAK CONTACTOR AT 9450 RPM
         4.0:1 RATIO OF WASTE-TO-SOLVENT FLOW
   	TOTAL FLOW 500 ml/min	
                  ANALYSIS, mg/£
                  Aqueous Phase      MCB Solvent
p,p'-DDE               34.5              81.0

o,p'-DDE       '        10.5              42.0

p,p'-DDD

p,p'-DDT               17.0              92.0

o,p'-DDT                7.5              33.0
     Equal quantities of solvent and aqueous waste  (250 m//
min) were used in the next test, for a combined flow rate
of 500 nU/min.  Pressures of 17 psig on the solvent stream
and 7 psig on the aqueous waste inlet stream were required.
Both the solvent stream and aqueous effluent appeared to
be clear of the other phase.  The extraction efficiency
was 82.1%.  The results are as shown in Table 24.

                       TABLE 24

    OPERATION OF PODBIELNIAK CONTACTOR AT 9450 RPM
         1.0:1 RATIO OF WASTE-TO-SOLVENT FLOW
                TOTAL FLOW OF 500 m£/min

p , p ' -DDE
0 , p ' -DDE
p,p'-DDD
p , p ' -DDT
o , p ' -DDT
ANALYSIS, mg/g
Aqueous Phase
9.5
2.5
—
2.0
1.5
MCB Solvent
24.0
6.5
—
31.5
9.0
                           26

-------
     The ratio of waste to solvent was increased to 1.5
in a test in which 300 m//min of the aqueous waste and
200 m//min of MCB solvent were employed.  The extraction
efficiency in this experiement was 95.4%.  The results
of the analysis of the separated phases are as shown in
Table 25.

                       TABLE 25

    OPERATION OF PODBIELNIAK CONTACTOR AT 9450 RPM
         1.5:1 RATIO OF WASTE-TO-SOLVENT FLOW
                TOTAL FLOW OF 550 ml/min
ANALYSIS, mg/e
Aqueous Phase MCB Solvent
p , p ' -DDE
o,p'-DDE
p,p'-DDD
p,p'-DDT
0,p'-DDT
5.0 49
1.0 12
__
0.50 64
0.35 18
.0
.0
-
.5
.0
     It was intended that tests also be made with a hydro-
carbon solvent as a comparison with the chlorinated MCB.
However, the contactor in operation heated to about 60°C,
and it was not considered advisable to use a flammable
solvent, such as heptane, under these conditions.

C.   Summary and Analysis of the Pgdbielniak Contactor Data

     The Podbielniak contactor was shown to be effective
for the extraction of aqueous DDT plant waste with mono-
chlorbenzene (MCB) extractant under defined conditions.
Data summarizing a series of evaluation runs are as shown
in Table 26.
                           27

-------
            TABLE 26

SUMMARY OF PODBIELNIAK CONTACTOR
        EVALUATION TESTS
Ratio
Aq.
Waste/
MCB
1.0
1.5
2.0
2.0
to
2.5
2.8
3.73
4.0
4.0
4.0
4.5
Flow,
Aq.
Waste
250
300
660
1000

500
453
563
400
800
800
900
m£/rain
MCB
250
200
330
500

200
160
151
100
200
200
200
Total
500
500
990
1500

700
613
714
500
1000
1000
1100
Aqueous Effluent Analysis, mq/0
p , p ' -DDE
9.5
5.0
1.0
1.0

7.0
3.5
3.0
34.5
0.6
5.5
3.5
o , p ' -DDE
2.5
1.0
0.35
0

1.5
0.2
1.0
10.5
0
1.0
1.0
p,p'-DDT
2.0
0.50
0.15
0

2.5
0
0
17.0
0
0
1.0
O , p ' -DDT
1.5
0.35
1.0
0

1.5
0
0
7.5
0
0
0.35
Total
15.5
6.85
2.5
1.0

12.5
3.7
4.0
69.5
0.6
6.5
5.85
Exten .
Efficiency
°/o
82.1
95.4
94.9
99.1

94.4
98.0
98.0
78.1
99.5
97.0
97.4

-------
     From this tabulation, it appears that greater than
95% extraction will be obtained with maximum rate of con-
tactor rotation, a waste-to-solvent flow rate of 1.5 to
4.5, and'a total flow of 500 to 1500 m£/min; the only very
poor extraction was achieved when only 100 mf/min solvent
was employed  (aqueous waste 400 mf/min, total flow 500
     As stated earlier, in a number of the tests, a small
to moderate quantity of MCB solvent was found in the
aqueous stream, or aqueous waste in the solvent stream,
but the phases separated rapidly, and possibly clean layers
could be obtained in a larger operation with continuous
decanters .

     In summary, the Podbielniak continuous contactor
appears capable of giving greater than 95% extraction
of the DDT, ODD and DDE from aqueous alkaline waste, to
give an aqueous phase generally with 5 mg/£  or less pesti-
cide components.  Waste to solvent ratios of ~4:1 and
flows of 1-1.5  £/min appear satisfactory.  No evidence
of phase emulsification was seen, though the conditions
of operation to give a waste stream without significant
MCB entrainment appear to be very narrow? that is , the
operation of the unit must be carefully controlled if,
indeed, it is possible to keep MCB to a practical minimum.
Except for this problem of solvent entrainment in the aque-
ous layer, the Podbielniak under very carefully controlled
conditions, was shown to give effective extraction of the
DDT from Montr os e waste — but still not good enough to
permit discharge of the treated aqueous waste to the public
server.

D.   "Second-Stage" Extraction

     Earlier work carried out under EPA Contract 68-01-0083
 (Reference 1) showed that the levels of DDT, ODD and DDE
could be substantially reduced from ~10 - 20 mg/l to a
few fig/I by a further extraction of the samples, follow-
ing the initial high-shear mixing.  This concept has been
further examined with the extraction of aqueous effluent
from the Podbielniak contactor.

     In these tests, the aqueous effluent was shaken for
one minute with 1/4 its volume of MCB solvent, and the
separated layers analyzed.  This process was repeated with
the aqueous layer from the above extraction, if any pesti-
cides remained after the first extraction.  This test was
carried out with the aqueous effluent from the preceding
six Podbielniak contactor runs.  The results of the tests
are as shown in Table 27.

                           29

-------
                                  TABLE 27

          "SECOND-STAGE" EXTRACTION OF PQDBIELNIAK-EXTRACTED WASTE
	 r n 	 __ _i |
1. Pod Effluent**
1st Extn
2. Pod Effluent
1st Extn
3. Pod Effluent
1st Extn
4. Pod Effluent
w 1st Extn
0 2nd Extn
5. Pod Effluent
1st Extn
6. Pod Effluent
1st Extn
2nd Extn
in • •-•— • i— • 	 	 «_
ANALYSIS
p,p'-DDE
0.6
~ 0
1.0
~ 0
1.0
~ 0
34.5
2.0
0.35
9.5
~ 0
5.0
0.4
0.2
—•— ^^^fc-^MfcM.^"— *»™"«— ••'^^"•fc—IP"
OF AQUEOUS
0 , p ' -DDE
~o
~ 0
~o
~o
0.35
~ 0
10.5
0.65
~ 0
2.5
~ 0
1.0
~ 0
~o
• 	 •'
PHASES , mg/£ '
p,p'-DDD
~ 0
~ 0
~ 0
~ 0
~ 0
~0
~ 0
~ 0
-0
~ 0
~ 0
~ o
~ 0
~ 0
	 — -
*
p,p'-DDT
~ 0
~ 0
~ 0
~ 0
0.15
~ 0
17.0
0.6
~ 0
2.0
~ 0
0.50
~ 0
~ 0
0 , p ' -DDT
~ 0
~ 0
- 0
~ 0
1.0
~ o
7.5
0.8
~ 0
1.5
~ 0
0.34
~ 0
~ 0
* Approximate Sensitivity Limits:  p,p'-DDE 0.25 jug// ; o,p'-DDE 0.33 ug/l ;
  p,p'-DDD 0.38 V<3/1 ; p,p'-DDT 0.37 ug/t ; o,p'-DDT 0.41
**The samples labeled "Pod Effluent" give the analysis of the  aqueous  phase
  exiting the Pod and represents the input to the  "second-stage"  extraction.

-------
                        SECTION 5

                ROSS IN-LINE HOMOGENIZER


A.   Description of Ross In-Line Homoqenizer

     The Ross homogenizer was selected for tests as a
good representative of the various high shear emulsifiers
and homogenizers which are available.  The mixing head
consists of a 1-3/8 inch diameter set of four blades
rotating within a closely fitting "stator."  The slotted
stator, giving the highest shear, has been used in these
tests.  The speed of rotation of the blades can be varied
up to 10,000 RPM.  The rotation of the rotor within the
stator gives a flow pattern wherein the fluid is pulled
into the cavity of the rotor and forced out through the
slots in the stator.  For the in-line model, the mixing
head is enclosed so that fluid is drawn into the rotor
cavity and expelled radially after passing the stator
assembly.  The pumping and mixing action of the homogen-
izer is thus broadly similar to the action of a centri-
fugal pump, (which had been found useful for mixing in
earlier studies.)  A schematic view of the Ross homogen-
izer is shown in Figure 3.

     In operation, the solvent and waste streams were
pumped into the vessel in which the in-line homogenizer
was contained and were mixed before entering the homogen-
izer.  The mixed phases then passed through the homogen-
izer and gave phases which separated slowly.  The stream
then was passed into a coalescer and continuous decanter
from which samples were withdrawn periodically.

B.   Experimental Results

     The operation of the Ross homogenizer followed earlier
experience (Section 7) using the laboratory model of the
device.  (The in-line model was not yet available.)  This
experience showed the need for (1) mixing at or near the
highest speeds, and (2) a coalescer appeared necessary
for adequate separation of phases.
                            31

-------
     These data show that effluent from the Podbielniak
contactor containing 0.6 to 15.5 mg/£ of DDT, ODD,  and
DDE were reduced to detection limits  (0.1-0.2 fJtg/l) when
extracted one additional time with MCB solvent.  In an
additional test in which 69.5 mg/£ of the combined  toxi-
cants were present, the level was reduced to 4.05 mg//
by the first extraction and 0.35 mg/i by a second extrac-
tion of the aqueous phase.  In another test, Podbielniak
effluent assaying 6.85 mg/f  was reduced to 0.4 mg// by
the first extraction and 0.2 mg/£ by an additional extract.
These data show clearly that a second-stage extraction
of the Podbielniak contactor effluent, as could be  achieved
with a packed tower, would appear to reduce the toxicants
to detection limits (~0.1 v-g/l) .  There appears to  be no
need for high-shear mixing in the second-stage extraction
if the mixing is adequate in the first stage of the extrac-
tion process.
                           32

-------
Output to
Coalescer
                      Mixed Phase
                         Input
        Figure 3.   ROSS IN-LINE HOMOGENIZER
                         33

-------
     In initial experiments, the Ross was operated at
9000 RPM with four volumes of waste (diluted to density
of 1.015) to one volume of monochlorbenzene (MCB) ex-
tractant.  The purpose of these experiments was to evalu-
ate several types of coalescers so as to determine a
possible configuration giving adequate phase separation
after mixing; the mixer can be evaluated adequately only
if the phases can be completely separated after the high
shear mixing.

     Data from the tests are shown in Table 28.
                            34

-------
                                  TABLE 28

          ROSS IN-LINE- HOMOGENIZER OPERATION - COALESCERS EVALUATED
  Coalescer Material
                               COALESCER TESTS

                                    Observation
 Extraction
Efficiency, %
1.  No coalescer


2.  Glass Wool
    1.5" dia x 6" deep

3.  Glass Wool +
    "Scotchguard"*
    1.5" dia x 6" deep

4.  Cotton + "Scotch-
    guard "
    1.5" dia x 6" deep

5.  Cotton
    1.5" dia x 6" deep

6.  Fritted glass
                            Two layers but no clear MCB
                              separation

                            Two layers but no clear MCB
                              separation

                            Two layers formed,  slow flow
                              through coalescer
                            Two layers formed.   Very slow
                              flow through coalescer
                            Two layers formed.   Very slow
                              flow through coalescer

                            Two layers formed but no
                              clean MCB separation
     93


     96


     92



     80



     93


     96
*A silicone producing a hydrophobia coating;  a product of 3M Company,

-------
     The total DDT + ODD + DDE in the aqueous extracted
material under the best conditions was about 2.6 mg/£ ,
or generally equivalent to the results with the
Podbielniak contactor described earlier.  The analyses of
the aqueous layer from the preceding tests are shown in
Table 29.
                            36

-------
                                  TABLE 29

        ROSS IN-LINE HOMOGENIZER OPERATION - EVALUATION OP COALESCERS
                                Analysis of Aqueous Phase
                          	After Extraction, mg/jg	
    Coalescer Material    p,p'-DDT  o,p'-DDT  p,p'-DDE  o,p'-DDE   TOTAL
1.  No coalescer

2.  Glass Wool
    1.5" dia x 6" deep

3.  Glass Wool +
    "Scotchguard"
    1.5" dia x 6" deep
5.  Cotton .     -- -:
    1.5" dia x 6" deep

6.  Fritted glass
1.1


0.5



1.5
4.  Cotton + "Scotcnguard"
    1,5" dia..x,6.'! deep      3.4
1.5

0.7
1.15


0.8



0.7


1.8


1.0

0.45
2.2


1.2



2.65


7.2


2.2

1.6
0.3


0.1



0.45


0.9


0.2

0.1
 4.75


 2.6



 5.3


13.3


 4.9

 2.85

-------
     In the next series, the effect of mixing time was
examined.  The 4:1 ratio of waste-to-MCB extractant was
maintained, and the homogenizer was run at 10,000 RPM.
The coalescer consisted of "Scotchguard"-coated glass
wool in a 1-3/4 inch diameter tube.  The results of the
analyses of the aqueous extracted waste are as follows:
                        TABLE 30

         ROSS IN-LINE HOMOGENIZER OPERATION -
         EFFECT OF MIXING TIME ON EXTRACTION

Mixing Time  Analyses of Extracted Aqueous Waste, mq/l
  Minutes    p, p'-DDT  o,p'-DDT  p,p'-DDE  o,p'-DDE   Total
1 0.7
5 0.35
10 2.0
20 0.7
0.45
0.1
0.6
0.25
1.4
1.0
3.2
2.0
0.3
0.15
0.4
0.4
2.85
1.6
6.2
3.35
     In this test, using a small coalescing mat, clear
separation of the solvent and aqueous phases was not ob-
tained.  Although the results are scattered, it appears
that the amount of pesticides remaining in the aqueous
waste tends to increase with longer mixing times; this
result agrees with earlier data where apparently a more
stable emulsion tends to form as the mixing time is
increased.

     It has also been observed that the flow rate through
the coalescer is slow, especially when the glass wool is
coated with the hydrophobic silicone coating.  A test was
then carried out in which a small applied pressure  (0-10 psi)
was placed on the coalescer to increase the rate of flow
through it since the non-pressurized flow rate was very
low.  In this test, the 4:1 waste:MCB extractant ratio
and 10,000 RPM mixing speed were maintained for a five-
minute mix time, and the coalescer consisted of a lOg glass
wool mat coated with "Scotchguard" in a 2 inch tube.  The
test conditions and test results are shown in Table 31.
                            38

-------
                                          TABLE 31
                             ROSS IN-LINE HOMOGENIZER OPERATION -
                            PRESSURE-AIDED FLOW THROUGH COALESCER
                Coalescer
                                 COALESCER TESTS
                               Retention
                               Time in
                               Coalescer         Observation*
                                     Extraction
                                    Efficiency, %
U)
      1.   No coalescer
2.  Glass Wool +
    "Scotchguard"
    No applied pressure

3.  Glass Wool +
    "Scotchguard"
    Max. pressure

4.  Same as 3. at
    less pressure
          Two layers but no clear       79
            MCB layer

3 hrs     Two layers formed, yellow     78
            MCB layer
20 rain    Three layers formed           85
            with yellow MCB layer
                                     ~ 35 min    Same as 3.                     63
     *The suspended solids sometimes appeared to form a separate layer
      so that it appeared that three layers were formed.
      (continued)

-------
TABLE 31 (Continued)
Analysis

1.
2.
Coalescer
No coalescer
Glass Wool +
P.P1
2.
2.
-DDT
0
0
••••^•••^•.^^^^^^^••^•^^•••••••••••KI^HtfWM-OII^M-MW^^^M^HW^^^^^H
of Extracted Aaueous
o,p'
1.
1.
-DDT
0
0
P,P
7
7
'-DDE
.5
.5
•-. i 	 1 i 	
Waste, mq/
o,p
0
1
'-DDE
.5
.0
4
Total
11.0
11.5
    "Scotchguard"
    No applied pressure

3.  Glass Wool +
    "Scotchguard"
    Max. pressure

4.  Same as 3. at
    less pressure
  1.0
  3.5
1.0
1.5
 5.5
13.0
0.5
1.5
 8.0
19.5

-------
     There does not appear to be a clear correlation between
retention time in the coalescer and extraction efficiency.
Since the same coalescer was used, but the applied pressure
varied, the effect of reduced efficiency of extraction dur-
ing the later runs may be due to increased solids deposited
in the coalescer.

     The effect of a second stage extraction on these
samples was examined by further extracting the aqueous
layer after hand mixing.  The same 4:1 ratio of waste
to MCB was employed.  The results indicate a further re-
duction in toxicant values.  The test conditions and test
results are shown in Table 32.
                           41

-------
                                    TABLE  32

                       ROSS  IN-LINE HOMOGENIZER OPERATION
           "SECOND-STAGE" EXTRACTION OF HOMOGENIZER -  EXTRACTED WASTE
Analysis of Extracted Aqueous
Waste After Second Extraction, mcr/i

1.
2.

3.

H^ A,
to ^* •
Coalescer
No coalescer
Coated Glass Wool,
no pressure
Coated Glass Wool,
max. pressure
Coated Glass Wool,
red. pressure
p , p ' -DDT
0.5
1.5

1.5

14.5
o,p'-DDT
1.0
1.0

1.5

4.0
p , p ' -DDE
1.0
5.0

5.5

20.5
0 , P ' -DDE
0.5
0.5

0.5

2.0
Total
3.0
8.0

9.0

41.0
(continued)

-------
                                  TABLE 32 (Continued)
Analysis of Extracted Aqueous
Waste After Third Extraction, mg/jf

1.
2.
Coalescer
No coalescer
Coated Glass Wool,
p , p ' -DDT
1.5
1.0
o , p ' -DDT
1.0
0.5
p , p ' -DDE
2.5
1.5
o,p'-DDE
0.5
0.1
Total
4.5
3.1
OJ
    no pressure

3.  Coated Glass Wool,
    max. pressure

4.  Coated Glass Wool,
    red. pressure
                                  1.0
                                  1.0
0.5
1.0
1.0
3.5
0.1
2.6
0.45    5.95

-------
    While some improvement was shown, a high correction for
the toxicants in the MCB made precise determinations dif-
ficult; the MCB contained ~90 rag// of DDT + ODD + DDE.
The tests were terminated with that source of MCB and a
new lot analyzing ~3.3 mg/f  (DDT+DDD+DDE) was obtained
from Montrose; the contaminated drum of MCB was used in
this test only.

    In the next series of tests, a coalescer consisting
of a 3 inch diameter glass column packed about 18 inches
deep with glass wool treated with "Scotchguard" hydro-
phobic coating was used.  Flow through the coalescer was
aided with a small applied pressure.  The rate of rota-
tion of the homogenizer was varied, as well as the flow
rate through the homogenizer and coalescer.  Good to poor
extraction was achieved.  The results of the test are
shown in Table 33.
                          44

-------
                                    TABLE 33

                      OPERATION OF ROSS IN-LINE HOMOGENIZER
              EFFECT OF HOMOGENIZER SPEED OF ROTATION ON EXTRACTION
                      SILICONS-TREATED GLASS WOOL COALESCER
(3 inch dia x 18
Flow
Rotation
RPM
6,000
9,000
£ 9,000
10,000
10,000*
10,000*
3,000-6,000
Waste
Ratio MCB
5.7
4.0
4.0
5.15
2.00
4.00
2.81
Rate
m^/min
513
319
-300
299
440
440
242
inch deep)
Coalescer
Pressure
psi
0-6
0-8
0-6
0 6
0-6
0-6
0-6
EXTRACTOR TEST DATA
Analysis of
Aaueous Phase, mq/4
P/P1
DDT
19.0
18.0
4.0
18.0
~ 0
4.0
2.0
o,p'
DDT
4.0
2.5
1.0
5.0
-0
~o
~o
P/P1
DDE
39.5
31.0
16.0
20.5
1.5
5.0
3.5
o,p'
DDE
5.0
2.5
1.5
2.0
~ 0
1.0
0.5
TOTAL
67.5
54.0
22.5 .. . •-
45.5
1.5
10.0
6.0
*Coalescer depth increased from 18 to 24 inches.

-------
      In all tests,  the MCB layer was colored (yellow).
 At the highest speed (10,000 RPM),  and the lowest ratio
 of waste volume to  MCB volume (2.0), good separation of
 the pesticides was  obtained.  An increased depth of the
 coalescer also appeared to increase the extraction.  This
 is probably a strong function of the effectiveness of phase
 separation, and that the apparently poor separations may
 be in fact MCB (laden with DDT,  ODD and DDE) suspended in
 the aqueous layer.

      The effect of  mixing speed  and emulsion strength was
 further demonstrated in a series in which 1.5 £ of a 3:1
 waste-MCB mixture  were treated  for 1.5 minutes  in the
 Ross In-line homogenizer, passed through the same coalescer
 as shown in Table 33, and the volume of the MCB layer
 measured.  Both hold-up of the MCB in water, and water
 in MCB demonstrate  a trend with  mixing speed.  The volume
 of MCB should be 375 m/.  The results of the tests are
 shown in Table 34.

                        TABLE 34

         OPERATION OF ROSS IN-LINE HOMOGENIZER -
             MCB RECOVERED WHEN 375 m£
	MIXED WITH WASTE AND PASSED THROUGH COALESCER	

                             Volume of Organic Phase Separating
   Mixer Speed, RPM                   After Mixing, m£
10,000
8,000
7,000
6,000
5,000
100
350
375
400
500
      It is thus seen that at speeds of 6,000 to 8,000 RPM,
 the volume of extractant separated is approximately correct,
 but at higher or lower speeds,  large deviations occur.

      The time of mixing was examined in an experiment in
 which a 3:1 ratio of waste to MCB extractant was mixed at
 7,000 RPM.  The rate of separation of the mixture is also
 given.  The results of the test are shown in Table 35.
                            46

-------
                           TABLE 35

         EFFECT OF MIXING TIME ON THE EXTRACTION WITH
                	 ROSS IN-LINE HOMOGENIZER

         Phase
Mixing  Separa-           Analysis of Extracted
 Time   tion Rate 	Aqueous Phase, mq/jg
Min
1
2
4
10
m//min
105
73
38
22
p,p'-DDT
3.0
0.5
0.1
0.5
o , p ' -DDT
2.0
0.5
0.2
0.5
p , p ' -DDE
9.5
2.0
0.5
1.5
o , p ' -DDE
1.5
0.05
0.1
0
Total
16.0
3.0
0.9
2.5
         While the best extraction was obtained after four-
    minute mixing, a substantially higher phase separation rate
    was observed after two-minute mixing.

         A similar test  (identical conditions, four-minute
    mixing time) was next carried out with the flow through
    the coalescer aided by 10 psi applied pressure.  Both
    "Scotchguard"-coated and untreated glass wool coalescers
    were examined.  The. analysis is  as shown in Table 36.
                                47

-------
                                        TABLE  36




                          OPERATION OF ROSS  IN-LINE  HOMOGENIZER
— II 1
Coalescer
Glass Wool +
"Scotchguard"
Untreated
..IllUm-H 1 1 MM 1 1
Flow Rate,
128
173
HMI^.BHdM^MIB^BM^.^H--P*-—M«*4^—
Analysis
p , p ' -DDT
2.5
9.5
	 h..— MI in- •• i i n i
of Extracted
o , p ' -DDT
1.5
3.5
— • ' •
Aqueous Effluent
p , p ' -DDE o
8.0
19.0
, ma/1
,p'-DDE
1.5
2.5
Total
13.5
34.5
    Glass Wool
00

-------
     In the period of a 1.5 £test, the coalescer appeared
to be essentially filled with suspended matter, the pres-
sure drop increased and the coalescer became ineffective
from the sludge loading.

     Since the operation of the coalescer appears to in-
volve exposure to a relatively high surface area to pro-
mote phase separation, the use of sand or fine gravel
beds as a low cost, expendable coalescer was examined.
The comparison was made with glass wool beds.  The flow
rate is for gravity flow through the bed.  Some of the
results are as shown in Table 37.
                            49

-------
                                     TABLE 37
                       OPERATION OF ROSS  IN-LINE HOMOGENIZER
                    EVALUATION OF SAND AND GRAVEL AS  COALESCERS







Ul
o


Coalescer
Untreated
glass wool
Silanized
glass wool
Sand

Fine
gravel
Coa-
lescer
Size, cm
2.5x15

2.5x15
-
2 . 5x15

2 . 5x15

Flow
Rate
mi'/min
3.7

5.1

1.6

8.0

Analyses of Extracted
Aqueous Effluent, mq/£
p , p ' -DDT
~ 0

~o

~ 0

~0

o , ' -DDT
~ 0

0.45

~o

~0

p , p ' -DDE
1.0

1.5

0.5

1.5

0 , p ' -DDE
~0

0.3

~0

0.15

Total
1.0

2.25

0.5

1.65

Untreated
glass wool
7.5x60
40.0
0
1.0
0.15
1.65

-------
       Good  promise was shown by both the sand and fine
  gravel  in  this  test.   Therefore, a scale-up was made of
  the  coalescer in which a 7-1/4 inch diameter x 6 inch
  deep bed of fine gravel was employed.  However, a clear
  phase separation was  not obtained.  Several of the samples
  were then  centrifuged, and upon analysis of the aqueous
  phase,  no  detectable  amount of DDT (either p,p'- or o,p'-)
  was  found,  although 3.5 to 4.0 mg/  p,p'-DDE and 0.2 to
  1.0  mg/l o,p'-DDE was detected.  The MCB content of the
  centrifuged aqueous extracts was 577 to 949 (average 831)
  mg/f , while the unextracted waste analyzed 413 mg//  MCB
   (MCB solubility in water is 382 mg/f , Reference 6).   This
  level of MCB (dissolved + suspended/emulsified) is  not
  acceptable in a treated waste.

       The efficacy of the fine gravel coalescer was  further
  checked in a test in which 21 i was passed through the
  coalescer  at a flow rate of 100 m//min, and the later
  samples were analyzed before and after centrifuging.
  The  results are as shown in Table 38.
                          TABLE 38

            OPERATION OF ROSS IN-LINE HOMOGENIZER
             EVALUATION OF FINE GRAVEL COALESCER      	

             Analyses of Extracted Aqueous Effluent Samples
            	mq/l Before Centrifuging	
Effluent,^   p,p'-DDT  o,p'-DDT  p,p'-DDE  o,p'-DDE  Total
18
19
20
21
1.5
1.5
2.0
1.5
1.0
1.0
1.0
1.0
9.5
6.0
9-0
7.5
1.0
1.0
1.0
1.0
13.0
9.5
13.0
11.0
                           After Centrifuqing
   18
   19
   20
   21
0.15
0.04
0.15
0.40
0.15
0.15
0.15
0.25
1.5
1.5
1.0
2.0
0.1
0.1
0.2
0.2
1.9
1.79
1.50
2.85
                              51

-------
     A test has also been made in which sand was treated
with a silicone "waterproofing" agent.*  A 20 mesh silica
sand was treated with the silicone, dried and formed into
a 2 inch diameter x 30 inch deep coalescer bed.  The homo-
genizer was operated at 7000 RPM and a 2 minute mixing
time was employed.  The results of some assays of the
aqueous layer are shown in Table 39.
                       TABLE 39

         OPERATION OF ROSS IN-LINE HOMOGENIZER
     EVALUATION OF SILICONE-TREATED SAND COALESCER
Analyses of Extracted Aqueous Effluent,**mq/£
Effluent,^
1
2
4
6
8
10
15
19
MCB
260
590
1270
1460
1740
1740
4320
3940
p,p'-DDT
3.0
1.5
1.0
1.5
1.5
1.5
li.5
1.5
0 , p ' -DDT
1.5
1.0
0.5
0.5
0.5
0.5
0.5
0.5
p , p ' -DDE
7.5
5.0
3.5
4.5
4.0
5.0
4.5
5.0
o , p ' -DDE
1.0
1.0
0.5
0.5
0.5
0.5
0.5
1.0
Total
13.0
8.5
5.5
7.0
6.5
7.5
7.0
8.0
     The  results  in Table  39 are anomalous in that the
 total pesticides  remains essentially constant after the
 first liter of  flow, while the MCB content of the waste
 increases steadily.  .After centrifuging, the MCB appeared
 to drop to a steady ~500 mg/f, so the very high  (3000-
 4000 mg/t) MCB  levels probably represent suspended MCB
 not separated by  the coalescer.

     A five minute centrifuging of some of the samples
 resulted  in an  approximate 2 mg/f decrease in the total
 pesticides; these results  suggest that DDT-DDE-laden  MCB
 is suspended in the aqueous phase and has not been fully
 coalesced.
* 3M silicone FC-857; a silicone in trichloroethylene,

**Non-centrifuged          __
                           52

-------
         The aqueous  fractions  have also been analyzed for
    MCB content, and  as  the  results in the table show,  the
    solvent content of the aqueous  phase increases  gradually
    to ~4000 mg/£ MCB.

         Further tests were  made with the uncoated  Pyrex  glass
    wool coalescer.   In  these tests the homogenizer was oper-
    ated at about 7000 RPM,  and the solution flowing  out  of
    the homogenizer was  passed  through a 1 inch diameter  x
    6 inch deep glass wool bed.  A  waste-to-solvent ratio of
    4.0 was used.  The results  of samples are shown in Table 40,
                            TABLE 40

              OPERATION OF ROSS  IN-LINE HOMOGENIZER
	TESTS  WITH GLASS WOOL COALESCER	

                        Analyses of  Aqueous  Phase,  mg/g
 Effluent, ml  MCB   p,p'-DDT  o,p'-DDT p,p'-DDE   o,p'-DDE  Total
100
200
300
400
2600
1340 0.2
1380 0.15
1080
1
0.15 1
0.15 2
0
.5
.5
.5
.5

0
0
0
—
.3
.45
.1
1.5
2.15
3.25
0.60
         The  inconsistent DDT,  DDE values,  and  the  excessive
    MCB in the  aqueous  phase  suggested that a larger coalescer
    may be required.  Further tests were  made then  with a 3
    inch diameter x  30  inch deep glass wool packed  coalescer.
    Other variables  remained  the same. The results of the
    test are  as shown in Table  41.
                               53

-------
                       TABLE 41
         OPERATION OF ROSS IN-LINE HOMOGENIZER
TESTS WITH 3 INCH DIA X 30 INCH DEEP GLASS WOOL COALESCER
	 -— 	 1 	 	 • 	 p — • 	 f- t __ • 	 • 1
Effluent,£ MCB
0.75 585
2 3200
4 2400
6 1070
8 1170
• - 1^ • ^^^^^^•-••••••••••••••1 in* 	 • 	
Analyses
p , p ' -DDT
6.5
1.5
1.0
0.1
—
!••• 	 1 	 • !• • • 	 • - 1 	
of Acrueous
o , p ' -DDT
2.5
0.5
0.5
0.1
0.1
__l 	 1 	 T ' 	 	 	 *
Phase, mq/£
p , p ' -DDE o
13.0
4.5
2.5
0.5
0.45
_^^^^tew^^^M»«^MMM««««B
, p ' -DDE
2.0
1.5
0.5
0.1
—
Total
24.0
7.5
4.5
0.8
0.55
     The larger coalescer gave improved effluent quality
as the flow continued.  This test was repeated, with  a
faster flow through the coalescer bed; this was accomplished
ky applying a 2.5 psi pressure to the coalescer, rather
than the gravity feed employed in the preceding test.  The
flow through the coalescer increased from an average  14.5
mf/min in the Table 41 test to 79 nU/min for the data in
Table 42 as shown below.
                       TABLE 42

         OPERATION OF ROSS IN-LINE HOMOGENIZER
            TESTS WITH PRESSURE-AIDED FLOW
             THROUGH GLASS WOOL COALESCER

Effluent,!. .MCB
3 900
6 1740
9 1670
12 1760
15 870
Analyses
p , p ' -DDT
0.5
1.0
1.0
1.5
0.5
of Acrueous
0,p'-DDT
1.5
1.5
1.0
1.0
0.5
Phase , mq/£
p , p ' -DDE
2.0
1.5
1.5
4.0
2.5
o , p ' -DDE
0.5
0.45
0.2
0.5
0.1
Total
4.5
4.45
3.7
7.0
3.6
                           54

-------
         Clearly, the increase in flow rate with the applied
    pressure on the coalescer resulted in a deterioration of
    coalescer performance and raffinate quality, compared with
    the results of Table 41.

         The aqueous effluent samples from the experiments
    in Table 42 were reextracted with MCB solvent in order
    to determine whether a "second-stage" extraction similar
    to the initial extraction was useful.  The same conditions
    were employed as used in the preceding test.  The results
    are as shown in Table 43.
                           TABLE 43

             OPERATION OF ROSS IN-LINE HOMOGENIZER
          SECOND EXTRACTION OF WASTE WITH HOMOGENIZER

                  Analyses of Aqueous Phase, mg/£
Effluent,£  MCB ~p,p'-DDT  o,p'-DDT  p,p'-DDE  o,p'-bDti  Total
3
6
9
12
790
1590
1980
3490
0.5
—
0.2
0.2
0.5
0.25
0.25
0.25
2.5
0.35
2.5
1.0
0.15
—
0.1
0.1
3.65
0.6
3.05
1.55
         The results shown in Table 43 are scattered but show
    no significant decrease in DDT, DDE content of the waste
    on the second extraction.  On centrifuging the same samples
    for five minutes and analyzing, the two DDT isomers and
    o,p'-DDE were absent in all four samples, and the p,p'-
    DDE values were respectively 2.0, 2.0, 1.5 and 1.0 wg/t .
    These results appear to further substantiate the supposi-
    tion that the apparent values for the DDT and DDE largely
    represent uncoalesced MCB containing DDT + DDE suspended
    in the aqueous phase.  The MCB contents of the centrifuged
    samples were respectively 81, 84, 250 and 290 rag//.  These
    are significantly lower than the values obtained with the
    uncentrifuged samples.
                               55

-------
Heptane Solvent Extraction Studies

     Tests were also made with the Ross In-Line homogen-
izer using heptane rather than MCB as the extractant.
Heptane was employed as the solvent in the earlier pump-
loop tests (Reference 1).

     In the first of these tests, the Ross homogenizer
was operated at 7000 RPM, and the waste and heptane  (3
volume waste: 1 volume heptane) was flowing at 100 m//min
with a five minute mixing time.  No observable emulsion
formed and the separation of the phases was rapid.  How-
ever, the separation of the DDT and related materials
was very poor.  The results are as shown in Table 44.
                       TABLE 44

         OPERATION OF ROSS IN-LINE HOMOGENIZER
     EXTRACTION OF WASTE WITH HEPTANE AT 7000 RPM	

               Analysis of Aqueous Phase After Extraction mq/l
               p,p'-DDT   o,p'-DDT   p,p'-DDE   o,p'-DDE  Total
1st
2nd
3rd
fraction
fraction
fraction
65
69
63
.5
.0
.5
28
32
28
.0
.5
.5
63.
64.
64.
0
5
5
18
20
16
.0
.0
.5
174.
186.
173.
5
0
0
     The heptane content of the aqueous samples was  990,
1280 and 1240 rag/I , respectively.  Since the  solubility
of n-heptane in water is about 50 mg/i , an emulsion  must
have been present, even though it was  not visually
observable.             I )

     In a similar  test iri which the rate of rotation of
the homogenizer was increased from 7000 to 10,000 RPM,
and the flow rate  was increased from 100 to 154 mf/min,
the effectiveness  of extraction remained essentially the
same.  The results are as shown in Table 45.
                           56

-------
                       TABLE 45

         OPERATION OP ROSS  IN-LINE  HOMOGENIZER
    EXTRACTION OF WASTE WITH HEPTANE AT 10,000 RPM
Analysis of Acrueous Phase After Extraction, ma/0

1st
2nd
3rd
p , p ' -DDT
fraction 23.5
fraction 59 . 0
fraction 83 . 0
o,p'-DDT p,p'-DDE o,p'-DDE Total
14.0 32.0 8.0 77.5
29.0 63.5 16.0 167.5
38.5 71.5 24.0 217.0
     The heptane content  of  the  aqueous  extracts was 900,
1210 and 1310 mg/f , respectively.

     In the next test, using heptane  as  the extractant,
a homogenizer speed of 10,000 RPM was again used for a
five minute mixing time,  and the ratio of waste volume
to heptane was 4:1.  A trace of  antifoam A* was added to
decrease the foam formation  and  a coalescer consisting
of a 3 inch diameter bed  of  fine gravel  16 inches deep
was employed to aid phase separation.  Again, the separa-
tion of phases appeared good (heptane 1000 mg// in aque-
ous phase).  The analyses of the separated aqueous phase
after three periods of operation are  as  shown in Table 46.
                       TABLE  46

         OPERATION  OF ROSS  IN-LINE  HOMOGENIZER
    EXTRACTION OF WASTE WITH  HEPTANE AT  10,000 RPM.
             EVALUATION OF  ANTIFOAM AGENT
Effluent
Treatment , I
1
5
9
Analysis of Acrueous Phase After
p,p'-DDT o,p'-DDT p,p'-DDE
38.0 18.0 70.0
11.0 7.0 24.0
26.0 16.0 52.0
Extraction, mq/P
o,p'-DDE Total
11.0 137.0
4.0 46.0
7.0 101.0

*Dow-Corning

                           57

-------
     It was thus apparent that heptane did not offer
improved separations in comparison with MCB, and work
with this extractant was suspended.
C.   Summary of Tests With Ross In-Line Homoqenizer

     The tests with the Ross appear to indicate that good
separation of the DDT and homologs from the waste can po-
tentially be achieved, but that reproducible, efficient,
high-flow coalescence of the phases has not been regularly
obtained.

     Units such as the Ross have promise of providing
effective separation with lower capital costs than the
Podbielniak contactor, assuming that a compatible solvent
system can be identified.  The Ross can only be useful
in cases where effective phase coalescence and separation
are achieved.  This appears to be impossible with the
waste-solvent systems examined.
                           58

-------
                       SECTION 6

                   BAFFLED STIRRERS


A.   Description of Stirrer System

     Another potential system for providing relatively
high-shear stirring involves the use of turbine-type
stirrers in a baffled tank.  This concept was examined
using the Bench-Scale Equipment system.  The experimental
set-up consisted of a seven-gallon glass tank fitted with
four vertical, adjustable baffles.  Stirring was achieved
with 3 inch diameter turbine mixer blades and a 1725 RPM
mixer.  Three different mixer blades were employed:  a
straight blade turbine, a tapered blade unit and a curved
blade turbine.

B.   Experimental  Results

     In the first  test, the three types of mixer blades
were compared.  For each test run, the seven-gallon tank
contained 2 £ total with a volume ratio of 4.00 volume
aqueous waste:1-00 volume MCB solvent.

     The samples after mixing for stated periods of time
were passed through a 1 inch diameter x 6 inch long coa-
lescer filled with glass wool.  The results are as shown
in Table 47.
                            59

-------
                           TABLE 47

         EXTRACTION OF MONTROSE WASTE IN BAFFLED TANK
            USING STRAIGHT BLADE, CURVED BLADE AND
        	    TAPERED BLADE TURBINE STIRRERS	
Mixing Time   Analysis of Aqueous Phase After Extraction, mcr/l
   Min.       p,p'-DDT   o,p'-DDT   p,p'-DDT   o,p'-DDE   Total
                       Straight Blade Turbine
     2
     5
    10
    20
     2
     5
    10
    20
1.00
0.45
0.27
0.37
0.35
0.40
0.45
0.35
0.50
0.40
0.25
2.25
2.50
1.50
1.00
0.42
                       Tapered Blade Turbine
0.25
0.30
0.35
0.30
1.50
2.00
2.00
1.50
0.35
0.50
0.07
0.68
0.20
0.20
0.25
0.20
4.35
2.85
1.59
3.72
2.30
2.90
3.05
2.35
                       Curved Blade Turbine
2
5
10
20
0.40
0.35
0.45
0.50
0.30
0.20
0.30
0.40
2.00
1.50
2.00
3.00
0.20
0.20
0.20
0.30
2.90
2.25
2.95
4.25
         The extraction results, based on residual pesticide
    content of the aqueous phase, appeared comparable for the
    tapered and curved blade turbines.  The amount of MCB which
    separated in a clear layer  (i.e., was not emulsified) was
    largest for the tapered blade mixer, and this unit was
    employed in further tests.  It is of importance to note
    that a two to five minute mixing time in the single  stage
    mixer has given extraction  efficiencies comparable to the
    homogenizers.  The above tests also show that neither
    mixer blade type nor mixing time significantly affect
    extraction efficiency.

         A further test was made using the tapered blade tur-
    bine, a five minute mixing  time and the 4:1 volume ratio
    of waste to MCB.  The coalescer is this test was a 3 inch
    diameter x 30 inch long column of glass wool.  At the
                               60

-------
   conclusion of the mixing period, the 25 £ sample was
   passed through the coalescer.  Three tests were made,
   with the mixer-baffle clearance being varied for each
   test.  The average results of the analysis of the aqueous
   fraction from each test are shown in Table 48.
                          TABLE 48
EFFECT OF BLADE-BAFFLE CLEARANCE
ON EXTRACTION EFFICIENCY
ANALYSIS OF AQUEOUS PHASE. AFTER EXTRACTION, ua/f
Component
p , p ' -DDT
O , p ' -DDT
p , p ' -DDE
0 , p ' -DDE
TOTAL
	 • 	 - i
Blade-Baffle
3.30
0.54
0.46
2.10
0.35
3.45
• i— 	 —
Clearance,
2.20
0.30
0.26
1.21
0.22
1.99
in.
1.50
0.16
0.11
0.69
0.08
1.04
Flow through Coalescer
        m//min           131           88          78
        The results in Table 48 show improved extraction as
   a function of lower flow rate through the coalescer and
   as a function of decreased blade-baffle clearance (produc-
   ing greater shear).

        A possible method for utilizing the baffled stirred
   vessel system could involve multiple stirred baths with
   countercurrent flow of the aqueous waste and MCB extract-
   ant.  With this in mind, an experiment was set up to
   examine this concept.  A single batch of aqueous waste
   was extracted three times with fresh solvents.  In this
   test, the tapered blade turbine stirrer was employed with
   a 4:1 ratio of waste-to-MCB solvent.  The mixing time was
   five minutes, and the mixed phases were passed through
   a 3 inch diameter x 42 inch deep bed of glass wool to
   coalesce the phases.  The results of the test are as shown
   in Table 49.


                             61

-------
                TABLE 49

  MULTIPLE EXTRACTION OF MONTROSE WASTE
WITH BAFFLED STRAIGHT-BLADE TURBINE  MIXER
Volume , H

5
10
15
20
25
30
35
<^
NJ
5
10
15
20

5
10
14
Analysis of Aqueous Phase
MCB

1140
510
250
710
810
880
540


680
570
440
440

450
340
330
P/P1

1.
1.
0.
0.
0.
0.
0.


0.
0.
0.
0.

0.
0.
0.
-DDT

50
50
50
50
50
20
25


05
15
05
04

025
033
033
o,p

1
1
0
0
0
0
0


0
0
0
0

0
0
0
1 -DDT p , p '
1st
.0
.0
.5
.5
.5
.2
.2
2nd

.05
.15
.05
.04
3rd
.025
.033
.042
After
-DDE
Coalescense, mq/£
o,p
'-DDE
Total
Extraction
4.
4.
2.
2.
2.
1.
1.
5
5
0
0
5
0
0
0
0
0
0
0
0
0
.50
.50
.20
.50
.50
.10
.15
7
7
3
3
4
1
1
.5
.5
.2
.5
.0
.5
.6
Extraction

0.
0.
0.
0.

50
50
25
25

0
0
0
0

.05
.10
.025
.013

0
0
0
0

.65
.90
.375
.363
Extraction
0.
0.
0.
083
167
083
0
0
0
.025
.017
.017
0
0
0
.158
.270
.175

-------
     Samples were analyzed after each  1 £ effluent, and
the average of these date indicates  a  95.4% extraction
after the first extraction, 99.5%  after two extractions
and 99.7% removal after three extractions.  The  averages
of the fourteen 1 i sample analyses  after three  extractions
were 0.033 mg/£ p,p'-DDT, 0.042 mg/£ o,p'-DDT, 0.083 mg/£
p,p'-DDE and 0.017 mg/£ o,p'-DDE,  giving a total of
0.175 mg/£ . . Tne average MCB content of the aqueous frac-
tion was 381 mg/£, which is almost identical with the
published solubility of MCB  (382 mg/£, Reference 6).

C.   Summary of Baffled Stirrer Tests

     The baffled stirrer in a countercurrent multiple
extraction arrangement appears to  give good extraction,
ranging from 95.4% after one extraction to 99.7% after
three extractions.  The material appears to coalesce
well and the phases separate readily.  The MCB in the
aqueous phase  is substantially lower than with the homo-
genizer tests.
                            63

-------
                       SECTION 7

                 OTHER MIXING SYSTEMS
     Several mixers were evaluated to a lesser degree  in
the course of the studies.  These include the Ross and
Polytron homogenizers and a Waring blendor.

A.   Waring Blendor

     An early screening test was made with the Waring
blendor as an example of a very high-shear rate mixer.
The undiluted waste and MCB were mixed (ratio 3 volumes
waste to 2 volumes MCB in the blendor.  A marked rise  in
temperature was noted, and the phases did not readily
separate after mixing.  The temperature of the mix as  a
function of mix time are as shown in Table 50.
                       TABLE 50

             TEMPERATURE OF WASTE-MCB MIX
       AS FUNCTION OF MIX TIME IN WARING BLENDOR
Mixing Time, Min
0
1
2
5
10
20
30
45
60
Temperature , °C
25
28
38
40
41
41
55
58
62
                             64

-------
     The poor separation  of  phases  was  traced  in part  to
the finding that the MCB  and aqueous  waste  have  virtually
identical densities  (waste 1.10  g/mt, MCB 1.098  g/mf)  so
that gravity separation is poor.  Following this finding,
the aqueous wastes have been diluted  with water  so  that
a reasonable density difference  exists  between the  waste
and solvent phases.  The  tests with the Waring did  not
give good separations because of the  very high-shear mix-
ing tended to form stable emulsions,  and  the tests  were
discontinued as soon as the  Ross  laboratory homogenizer
was obtained.

B.   Ross Laboratory Homogenizer

     Tests were also made with a Ross laboratory homo-
genizer while awaiting the delivery of  the  in-line  model.
These tests gave an indication that high  shear velocities
were required to give good separation,  that a  short (~2
min) time was needed, and that an adequate  coalescer was
needed.  In an early test, for example, a 9000 RPM  velo-
city was used with gravity separation of  the phases.
These results show about  20  mg//  p,p'-DDE and  10 -  12  mg/£
of the o,p'-isomer while  the p,p'-DDT was 3-4  mg/7 and
the o,p'-DDT 1-2 mg/^.  Tests  with  heptane as  solvent
led to a jelly-like phase.   In a comparison of coalescers,
glass frit was better than overnight  standing  for phase
separation.  A glass wool mat was also  effective.

C.   Polytron Homogenizer

     The Polytron, a homogenizer produced by Bririkman
Instruments, was also evaluated.  This  device  consists
of two sets of closely  fitting cutting  blades  rotating
in opposite directions  at high rates  of speed.  This
device is available  in  an in-line model similar  to  the
Ross.  The speed of  rotation of  the Brinkman may be varied
up to 20,000 RPM, while the  Ross is variable up  to  10,000
RPM.  A series of qualitative runs  were made to  determine
whether separable phases  were formed, and on mixing water
and MCB it was found that at the higher mixing speeds,
very stable emulsions were  formed.  In  a  series  of  tests,
375 ml of MCB was mixed for  1.5  minutes with 1125 m£ of
the aqueous waste, the  solution  passed  through the  given
coalescer, and the amount of MCB which  separated was noted.
The layers were then analyzed for DDT,  ODD  and DDE. The
results are as shown in Table 51.
                            65

-------
                     TABLE 51




EXTRACTION TESTS WITH BRINKMAN POLYTRON HOMOGENIZER
MCB
Recovered
RPM
22,000
22,000
c\ 10 , 000
16,000
Coalescer mi
Glass Wool
Glass Wool +
"Scotchguard"
Glass Wool
Glass Wool
50
- 50
275
150
p , p ' -DDT
2.0
0.5
6.5
4.0
Analysis
o , p ' -DDT
1.5
0.5
1.5
1.0
, ma/0
p , p ' -DDE
6.0
5.5
16.0
11.0

o , p ' -DDE
4.0
0.5
1.5
1.0
Total
13.5
7.0
25.5
17.0

-------
     These results document the previously stated con-
clusion that at the highest speeds, much of the MCB is
formed into a stable emulsion, while at lower speeds of
rotation the extraction is not as efficient.  The evalu-
ation of the Polytron was terminated with these results,
as the system appeared to offer no advantage over the
Ross.
                            67

-------
                       SECTION 8

   CHEMICAL REDUCTION OF RESIDUAL MONOCHLORBENZENE
                IN THE AQUEOUS EXTRACTS
     The earlier tests have shown that the aqueous waste,
either before or after extraction, contains a few hundred
to a few thousand mg/£ of monochlorbenzene (MCB).  These
levels are too high to permit dumping a treated waste in
a municipal sewer system.  A limited evaluation was carried
out to determine the feasibility of removing the MCB by
the Envirogenics reductive degradation process.

     A reduction column was prepared by mixing 600 g of
catalyzed iron powder with 2150 g of 30 mesh sand and plac-
ing the blend in a 3.7 cm diameter column to a depth of
165 cm.  The aqueous MCB emulsion was allowed to flow
through the bed at an average (low) flow rate of 15.6
m /min  (0.36 GPM/sq ft).  In the first test, a 79 mg/£
MCB solution was passed through the column four times.
The MCB analysis after the first pass was 48.1 mg/f,
30 mg/S. after the second pass, 20 mg/f after the third
pass and 11 mg/£ after the fourth pass.

     In the second test, a 225 mg/£ aqueous MCB solution
was passed through the column six times.  The assay after
each pass was as follows:  1st - 154 mg/£, 2nd - 97 mg/£,
3rd - 58 mg/£ , 4th - 39 mgjf, 5th - 28 mg/£ and 6th  -
22 mg/jf.  These data indicate an 86% degradation of MCB
after four passes in the first test, and a 90.2% degrada-
tion after six passes in the second test.  Cyclohexanol
was shown to be a reduction product, on the basis of
chromatography of derivatives, and comparison with
authentic cyclohexanol.

     In another test, an extracted Montrose waste from
a baffled stirrer experiment  (Section 6) was also passed
through the column.  The results of the test are as shown
in Table 52.
                           68

-------
                    TABLE 52

        ANALYSIS FOR MONOCHLORBENZENE
DDT AND DDE AFTER TREATMENT  IN REDUCTION COLUMN
Analysis of Aaueous Phase, mq/f

Sample

Initial
1st
2nd
3rd
4th
Column
Column
Column
Column
Pass
Pass
Pass
Pass
MCB
641
427
163
128
90
P/P
0
0
0
0
0
'-DDT
.20
.20
.12
.20
.15
o,p'
0.
0.
0.
0.
0.
-DDT
20
31
32
25
20
p , p ' -DDE
1.00
1.25
0.83
1.25
1.50
o,p'
0.
0.
0.
0.
0.
-DDE
15
19
13
15
15
Total
1.55
1.95
1.40
1.85
2.00

-------
     While a gradual decrease is shown in the MCB content
of the waste, there appears to be no significant change
in DDT or DDE level of the waste.  Earlier studies (Refer-
ence 1) show that DDT can be reduced, though DDE can only
be reduced under extreme conditions.
                           70

-------
                       SECTION 9

                  ANALYSIS OF RESULTS


     The treatment of the alkaline waste liquor from DDT
manufacturing operations is a complex problem requiring
the solution of several problems.  Important are the
following considerations:

     • Effective removal of DDT, ODD and DDE

     • Environmental effect of other components in
       the waste and their treatment if necessary

     • The ultimate disposal of the DDT and metabolites,
       including a suitable degradation process if
       necessary

     • The capital and operating cost of a suitable
       process to dispose of the waste

Each of these problems can be discussed separately in this
section.

     .The concept of efficient extraction of the DDT and
DDE from aqueous alkaline DDT manufacturing waste appears
to have been demonstrated in principle by the results of
at least three mixing techniques.  Reduction of the DDT +
DDE levels (both p,p'- and o,p'-isomers) to of the order
of 2 mg/£ has been shown for both the Podbielniak continu-
ous contactor and the Ross continuous homogenizer, when
operated at flow rates of the order of 1 i/min.  A three
stage countercurrent mixer using a turbine mixer in
baffled tanks was able to reduce the DDT + DDE level to
about 0.2 mg/m .  However, an important problem remains.
It is particularly evident with the homogenizer results,
but also evident to a lesser degree with the Podbielniak
and baffled stirrer results, that adequate coalescence
of the phases and effective separation of the aqueous
phase from the solvent has not been effectively and re-
producibly shown.  The substantial reduction of the DDT
and DDE in samples which have been centrifuged provides
evidence that the residual DDT (+ DDE) may, to a large
extent, be dissolved in MCB suspended in the aqueous phase.
                           71

-------
     A "second-stage" extraction, particularly with the
Podbielniak system, appeared capable of reducing the DDT,
DDE levels significantly, with the DDT generally being
at or below detection levels, and the DDE 0.1 mg/£ or
less.  A simple packed tower extraction would appear ade-
quate for this step since the high-shear mixing in the
first stage would serve to free the DDT, DDE from carbon-
aceous particles with which it may be associated in the
waste.

     Extraction with monochlorbenzene (jyiCB) appears to
be better than with the hydrocarbon solvent heptane (used
in the earlier Contract 68-01-0083 studies, Reference 1),
and MCB would be preferred on the basis of reduced flamma-
bility and simpler disposal of the extracted DDT, DDE.
It should be noted, though, that dilution of the waste
with water is necessary so that there is sufficient
density difference between the waste and MCB to allow
phase separation.

     In view of the fact that solvent extraction does not
appear to be an acceptable alternative to present practice,
there is no point in attempting to estimate process costs.
It is obvious that the severe coalesence problems would
require considerable time to solve.  If long holding
times are necessary to provide for phase separation,
tanks become unacceptably large, and the process becomes
unacceptably expensive.

     Another area of concern is the ultimate disposal of
the extracted DDT and DDE.  As discussed earlier, it was
hoped that the use of MCB as an extractant would allow
the MCB to be recycled to the condensation step.  Since
it would appear that MCB cannot be used as a solvent due
to excessive solvent loss, then the recycle advantage is
lost.  If another solvent were identified which were ideal
as an extractant, it could not be recycled to the. DDT
synthesis process.  An alternative final disposal system
would be necessary for the pesticide-laden solvent.

     In summary, the effective extraction of DDT and
metabolites by the continuous high-shear mixing system
has been investigated, with effluent levels of ~2 mg/£
DDT + DDE being shown for the Ross homogenizer and
Podbielniak contactor, and as low as 0.2 mg/£ for a three-
stage baffled stirrer.  Further extraction to remove the
residual DDT and DDE to ~0.1 mg/£ has been shown also.
However, adequate coalescence of the phases remains a
problem.
                           72

-------
     The January 12, 1977, Federal Register prohibits
any discharge of DDT from the manufacturing process for
either new or existing sources.  The ambient water
criterion for DDT in navigable water is 0.001 f*g/B_ or
0.001 parts per billion.

     Although the one existing DDT manufacturing plant
would discharge to a line feeding a county sewage treat-
ment plant, and not to navigable water, it is clear that
the technology tested under this grant is incapable of
generating waste which will meet the provisions of the
toxic pollutant standard.  The current operation of this
plant does meet the toxic pollutant standard.  There would
appear to be no alternative to the present disposal practice
which consists of placing the caustic waste in a Class I
landfill.
                            73

-------
                 REFERENCES
Development of Treatment Process for Chlorinated
Manufacturing and Processing Wastes, Envirogenics
Final Report on EPA Contract 68-01-0083.  Submitted
for review.

Water Quality Criteria Data Book.  Volume 1 Organic
Chemical Pollution of Freshwater.  EPA Water Pollution
Control Research Series.  18010 DPV 12/70.  December,
1970.

Mount, D.I., and W. A. Brungs.  A simplified Dosing
Apparatus for Fish Toxicology Studies.  Water Res. 1,
21-29  (1967).

Lane, T. H., and H. M. Jackson.  Voidance Times For
23 Species of Fish, Investigations in Fish Control,
No. 33, 1969.

Gunther, F. A., W. E. Westlake, and P. S. Jaglan.
Reported Solubilities of 738 Pesticide Chemicals
in Water, Residue Reviews, 20.  1-148 (1968).

Stephen, H., and T. Stephen, ed.  Solubilities of
Inorganic and Organic Compounds, Volume I, Part I.
MacMillan, New York, 1963.  960 pp.
                      74

-------
                               GLOSSARY
CHEMICAL, FORMULAS OF PESTICIDES AND  DEGRADATION PRODUCTS
       DDT
              Cl
  H
   i
  C
   I
ci-c-ci
  Cl
       2, 2 bis(p-chlorophenyl)-!, 1, 1-trichloroethane


       ODD  (TDE)    C

                          ci-c-ci
                            H
       2, 2 bis(p-chlorophenyl)-l, 1-dichloroethane
       DDE
                     Cl
                      C
                      (i
                   CI-C-CI
         Cl
       2, 2 bis(p-chlorophenyl)-!, 1-dichloroethylene

                            O
                            ^
                            O
               ci-cu>sci

DOS

bis (p- chlor ophenyl) sulf one

MCB          Cl
       (mono)chlorobenzene

       NaMCBS

                     Cl
                      O SO2O Na
       Sodium p-chlorobenzene sulfonate

-------
GLOSSARY  (Continued)
Material

 DDT



 ODD

 DDE

 DOS

 MCB


 NaMCBS
             SOLUBILITY OF PESTICIDES AND
             DEGRADATION PRODUCTS  IN WATER
  Solubility

14 refs ranging
in sol to 1 ppm
1-3 ppb most probable

Insol, negligible
0.0382% (25.2°C)
488 mg/£ (30 °C)

15.21%  (25 °C)
Reference
                   CONVERSION UNITS
 Gallon

 Inch

 Pounds per square inch  (psi)
                    3.78

                    2.54 cm

                    0.070 kg/cm2

-------
                                TECHNICAL REPORT DATA
                         (Please read Instructions on ike reverse before completing)
 REPORT NO.
EPA-600/2-78-125
                           2.
                              3. RECIPIENT'S ACCESSION NO.
 TITLE AND SUBTITLE
Development of a DDT Manufacturing and Processing
Plant Waste Treatment System
                              5. REPORT DATE
                               June 1978
                              6. PERFORMING ORGANIZATION CODE
.AUTHOR(S)

M. Sobleman, K. H. Sweeny, and E. D. Calimag
                              8. PERFORMING ORGANIZATION REPORT NO
 . PERFORMING ORGANIZATION NAME AND ADDRESS
Montrose Chemical Corp.  of California
P.O.  Box 147
Torrance,  California 90507
                              10. PROGRAM ELEMENT NO.
                               1BB610
                              11. CONTRACT/GRANT NO.
                               Grant 804293
I2. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC  27711
                              13. TYPE OF REPORT AND PERIOD COVERED
                              Final; 2/76-2/78	
                              14. SPONSORING AGENCY CODE
                                EPA/600/13
15. SUPPLEMENTARY NOTES IERL_RTp project officer is David K.  Oestreich, Mail Drop 62,
919/541-2547.
is. ABSTRACT
              report gives results of a study both to test the feasibility of detoxifi-
 cation of DDT manufacturing wastes, using solvent extraction, and to develop a prac-
 tical process , if possible.  Three different liquid-liquid contacting devices were
 tested: all provided reasonably good extraction of DDT and homologs from the caustic
 aqueous phase.  Unfortunately, major phase separation problems resulted in exces-
 sive losses of monochlorobenzene solvent to the aqueous phase. Efforts to improve
 coalescence/phase separation were unsuccessful. Further development of a solvent
 extraction process for detoxification of DDT manufacturing waste cannot be justified.
17.
                             KEY WORDS AND DOCUMENT ANALYSIS
                DESCRIPTORS
                                          b.lDENTIFIERS/OPEN ENDED TERMS
                                           c. COSATI Field/Group
 Pollution
 Industrial Processes
 DDT
 Waste Treatment
 Detoxification
 Solvent Extraction
Chlorobenzene
Coalescing
Pollution Control
Stationary Sources
Monochlorobenzene
Phase Separation
13 B
13 H
07C,06F

06E
07D
07A
13. DISTRIBUTION STATEMENT

 Unlimited
                  19. SECURITY CLASS (This Report)
                   Unclassified
                         21. NO. OF PAGES
                              87
                  20. SECURITY CLASS (Thispage)
                   Unclassified
                                           22. PRICE
EPA r OI-T 2220-1 (9-73^
                 77

-------