-------
 the  cover  topsoils and the clay bar-
 rier layer.  Note that the greater
 frost depth corresponds to sand with
 a  lower water content.  The poten-
 tial  for cracking includes the full
 frozen soil depth.

      Frozen soils behave elastically
 in response to rapid deformations.
 If one assumes the same behavior for
 the  slower natural thermal contrac-
 tion, the tensile stresses generated
 would be proportional to the amount
 that the temperature drops from some
 reference temperature.  To provide
 more  information on the rate of tem-
 perature drop at various depths,
 computed temperatures in clay are
 summarized in Figure 5 for Fargo,
 North Dakota, on 2-3 February 1974.
 Using a step change in surface
 temperature and the LaPlace-transform
 technique (5), soil  temperatures were
 computed as a function of time at the
 depths shown in Figure 5.   The zero
 depth represents the ground surface.
 Lachenbruch (6) reported that signi-
 ficant tensile stresses would be
 generated in frozen soil  with a tem-
 perature drop of only 2 to 3°C.  The
 temperature change in 9 hours at the
 24 cm depth falls in this  category.
With more time, similar temperature
changes would occur at greater depths.

THERMAL CONTRACTION, TENSILE
STRESSES, AND SOIL STRENGTH

     On cooling,  the frozen soil
cover surface would  contract if it
were not constrained.   Horizontal
tensile stresses  are generated with
no observable horizontal  strains.
The horizontal  thermal  strain is
given by the thermal  contraction,
thus,
        = e.
                                 (3)
where a is the coefficient of linear
contraction (about 5 x 10~5oc-l for
                                                                             JZ

                                                                            CM »
                                         Figure 5.
Computed ground tempera-
tures in sand for Fargo,
North Dakota, on 2-3 Feb-
ruary, 1974.
                                         frozen sand), L0 is the length at
                                         some reference temperature, and AL is
                                         the change in length due to a temper-
                                         ature change AT.  With the soil con-
                                         strained and if the frozen soil is
                                         assumed to behave elastically the
                                         tensile stress becomes
                                                 =  E
                                                   T^T ey
                                                          _  E
                                                          "
                                                                           (4)
                                         where E is Young's modulus and u is
                                         Poisson's ratio.   The stress-strain
                                         curves for sand (Figure 6} in tension
                                     279

-------
  1.0
        0.01
              0.02   0.03
                AXIAL STRAIN
                          0.0ft
                                0.05  0.06
Figure 6.  Stress-strain curves for
           frozen sand in compres-
           sion (c) and tension (t)
           with test duration (t) in
           hours (8).

at 45% gives an E close to 435
kN/cm2.  Using this value with

U = 0.28 gives ^-=600 kN/cm2 or

Oy-0.03  J (kN/cm2) for equation (4).
The tensile strength of the saturated
frozen sand in Figure 6 is at most
only 0.4 kN/cm2 indicating that
frozen sand will be very sensitive to
a temperature decrease.  Note that
failure stresses in tension are much
smaller (Figure 6).  With only par-
tial saturation for frozen cover
soils, tensile strengths would be
further reduced since only the ice
matrix is able to transfer tensile
stresses.  Considering the data sum-
marized in Figure 5 and assuming
elastic soil behavior, the frozen
soils would be close to failure in
tension before the rapid air tempera-
ture drop occurs.  The additional
temperature decrease shown in Figure
5 would cause additional thermal
contraction and larger tensile
stresses.  The total tensile stress
would then be significantly greater
than the tensile strength of the
frozen saturated sand and open
cracks would be predicted.  For the
given assumptions Lachenbruch (6)
states that the crack spacing would
be no greater than the crack depths.

     The example with sand illus-
trates that simple elastic behavior
does not fully represent the frozen
soil behavior.  Frozen soil under a
constant stress will deform with time
in a viscous manner.  This creep be-
havior would serve to reduce the ten-
sile stresses over the period of time
in which temperature is decreasing.
If a viscous deformation law is as-
sumed the stress would be propor-
tional to the rate at which the tem-
perature drops.  The rate of tempera-
ture decrease at various soil depths
is represented by slopes of the
curves in Figure 5.  The data in
Figure 6 also show that Young's modu-
lus increases with a decrease in tem-
perature so that the problem becomes
more complex.  Experimental data are
needed which describe the rate of
increase and magnitude of tensile
stresses which develop in typical
cover soils for freezing temperatures
and cooling rates representative of
the northern tier of states.  The
lack of information on linear thermal
contraction behavior for these same
soils indicates additional research
needs before an accurate analysis of
thermal cracking in the cover soils
can be made.

CONCLUSIONS

     1.  A review of the mechanics of
thermal contraction indicates that
cracks propagating unstably through
the frozen cover soils may extend
deeper than the tensile stresses to
which they owe their growth.  Under
suitable conditions it appears that
the cracks may penetrate the frozen
cover creating an opening for water
                                     280

-------
movement and escape of gases.

     2.  Information on the linear
thermal contraction behavior of
frozen soils needed for computation
of thermal  strains is lacking.   The
dependence of the coefficient of
linear contraction on various soil
parameters should be determined.

     3.  Simple elastic soil beha-
vior used with thermal strains  does
not provide the information needed
for accurate prediction of thermal
contraction and crack formation.
Information is needed on the rate
of increase and magnitude of tensile
stresses which develop in frozen
cover soils for freezing tempera-
tures and cooling rates representa-
tive of the northern tier of states.

     4,  Using information from
items 2 and 3, criteria need to be
formulated which will permit predic-
tion of cracking and the techniques
suitable for preventing crack forma-
tion.  New techniques may be more
economical  than placement of addi-
tional cover soils to prevent freez-
ing and possible cracking of the
clay barrier.

REFERENCES

1.  Lutton, R.J., "Evaluating cover
    systems for solid and hazardous
    waste," Publication SW-867,
    USEPA-MERL, Cincinnati, Ohio,
    September 1982.

2.  Terzaghi, Karl, "Permafrost,"
    Journal of the Boston Society of
    Civil Engineers, The Engineering
    Center, Suite 1110, Boston,
    Massachusetts, Vol. XXXIX,  No.
    1, January 1952.

3.  Smith,  D.W., Reed, S., Cameron,
    J.J., Heinke, G.W., James,  F.,
    Reid, B., Ryan, W.L., and
    Scribner, J., "Cold climate de-
    livery design manual," Publica-
    tion EPA-600/8-79-Q27, USEPA-
    ERL, Corvallis, Oregon, Septem-
    ber 1979.

4.  Aldrich, H.P., Jr., and Paynter,
    H.M., "Analytical studies of
    freezing and thawing of soil,"
    Technical Report No. 42, Artie
    Construction and Frost Effects
    Laboratory, New England Divi-
    sion, U.S. Army Corps of
    Engineers, Boston, Massachusetts,
    1953.

5.  Arpaci, Vedat S., CONDUCTION HEAT
    TRANSFER, Addison-Wesly Publish-
    ing Company, Reading, Massa-
    chusetts, 1966.

6.  Laehenbruch, Arthur H., "Mecha-
    nics of thermal contraction
    cracks and ice-wedge polygons in
    permafrost," Special Paper No.
    70, The Geological Society of
    America, 1962.

7.  Local Climatological Data for
    Fargo, ND, Lansing, MI, and
    Madison, WI.  National Oceanic
    and Atmospheric Administration,
    National Climatic Center, Ashe-
    ville, North Carolina.

8.  Eckardt, H., "Creep tests with
    frozen soils under uniaxial ten-
    sion and uniaxial compression,"
    The Roger J.E. Brown Memorial
    Volume.  Proceedings of the 4th
    Canadian Permafrost Conf.,
    Calgary, Alberta, March 2-6,
    1981.  H.M. French (ed.), Nation-
    al Research Council of Canada,
    Ottawa, pp. 365-373, 1982.

               D1sclaimer

 The work described in  this  paper was
 not funded by  the U.S.  Environmental
 Protection Agency.   The contents do
 not necessarily reflect the views of
 the Agency and no official  endorse-
 ment  should be inferred.
                                    281

-------
SYNTHETIC I»INER SELECTION AND APPLICATION TO GROUNDWATER PROTECTION

       John D. VanderVoort, Schlegel Lining Technology, Inc.

                             ABSTRACT
     This paper discusses the use of synthetic liners as a
means of controlling groundwater pollution.
     It discusses the most common types of raw materials for liners:
chlorosulfonated polyethylene, polyvinyl chloride and polyethylenes,
and other liner materials: butyl rubber, chlorinated polyethylene,
ethylene propylene diene monomer and chloroprene.
     The manufacturing processes are described for producing the final
synthetic liner including calendering, blown film and sheet extrusion.
     The various seaming methods for factory and field seaming are
defined and include adhesive seams, extrusion, hot air, bodied solvent
and solvent seams.
     Causes of failure are presented such as chemical attack, puncture
and tear, and biological.  Also, the paper presents information on how
to select a lining system for a particular application.
     The emphasis throughout this paper is on specifications for all
segments of the synthetic liner industry, and guidelines for the end
user on how to determine if the product meets those specifications.
     A flow chart for quality control and selection for various synthetic
liners is presented.
Introduction

    In recent years, the need for
containment of hazardous wastes has
received much attention as the
potential dangers to public health
from these substances receive closer
scrutiny from both the public and
private sector.  This paper deals
specifically with the use of extruded
polymers as perhaps the most
effective means of containing
hazardous wastes.  Also it describes
some lining materials in use today,
methods of manufacture and some
common failure mechanisms.
Background and Materials Available

    Throughout history, soils and
clays have been the most common
lining materials.  These substances
have never been considered perfect
liners as they do not provide zero
permeability.  Since both organics
and inorganics dissolve parts of clay
structures, permeability usually
increases over time.

    Many polymeric materials will
exhibit essentially a zero
permeability rating when used in a
film or sheet form.  As a result,
                                     282

-------
these materials offer long-term
containment possiblities for ground
water protection.  The first
polymeric materials used for this
purpose were synthetic rubbers in the
1940s,  By the 1970s, many new
synthetic lining materials,
particularly thermoplastics, were
available.

    Today synthetic liners are used
to contain hazardous wastes, protect
groundwater and control erosion.
There are a variety of applications
and a variety of lining materials.

Liner Materials

    The most commonly used base
materials for synthetic liners are:

    Chlorosulfonated polyethylene
(CSPE), a polymer produced by
reacting polyethylene in solution
with chlorine and sulfur dioxide.
CSPE contains 25 to 45 percent
chlorine and 1.0 to 1.4 percent
sulfur, both by weight.  The presence
of chlorine and sulfur side-chains in
the end product has a noticeable
effect on its chemical compatibility
with hydrocarbons.  CSPE exhibits
poor resistance to both aliphatic and
aromatic hydrocarbons, chlorinated
solvents, oxygenated solvents and
alcohols.

    Polyyinyl chloride (PVC), a
polymer produced by the polymeri-
zation of vinyl chloride monomers,
with plasticizers added to improve
flexibility.  Because it also contains
side-chains of chloride, it offers
limited resistance to hydrocarbons,
PVC blended with ethyl acetate also
has side-chains.

    Polyethylenes, because of their
simple chemical structure, exhibit
the best resistance to hydrocarbons,
as well as to other chemicals.
Laboratory tests have shown that the
higher the density, the greater the
resistance to chemicals, including
hydrocarbons.

    Other, but less commonly used,
base materials are:

    Chlorinated polyethylene (CPE),
produced by chemical reaction between
chlorine and polyethylene.  The
resulting polymer is 25 to 45 percent
chlorine by weight and contains
side-chains of dissimilar members.
CPE has excellent flexibility but
limited resistance to hydrocarbon
environments.

    Ethylenepropylene diene monomer
(EPDM), a terpolymer of ethylene,
propylene and a small amount of
nonconjugated diene to furnish
vulcanization sites.  The material
has good chemical resistance except
with hydrocarbons.

    Epichlorohydrin rubber, a
synthetic rubber that includes two
epichlorohydrin-based elastomers,
which are saturated, high molecular
weight, aliphatic polyethers with
chloromithyl side-chains.

    All of the above polymers
mentioned have been successfully used
in a variety of containment
applications.  However, some of them
have also failed.  These failures can
be attributed to poor selection, poor
installation and/or changes in the
containment application at a later date,

    Butyl rubber, elasticized
polyolefin and polychloroprene are
the least commonly used base
materials.

    In addition to base polymers,
there are blends and combinations of
the above which are used either alone
or with other additives.  These
additives include: plasticizers for
improved flexibility; crosslinking
                                       283

-------
agents for promoting the formation of
chemical bonds; fillers to give body;
pigments for coloring purposes; bio-
logical inhibitors to reduce suscep-
tibility to attack by microorganisms;
and scrim reinforcement for improved
mechanical strength.

    These additives are used to
overcome an inherent weakness in a
base material, but their use must
also be suitable in contact with the
contents of a containment system.

Manufacture of Synthetic Liners

    Flexible membrane liners are
produced into a film or sheet by one
of three processes; blown film,
calendering or sheet extrusion.  In
the blown film process, molten
plastic is extruded through a. tubular
die in a vertical direction.  Air is
blown through the die to form a
bubble.  After the bubble is cooled
from the outside, it is flattened at
the top of its travel by a collapsing
frame.  The resultant film or sheet
is then passed through a wind—up
system.  The blown film process is
most commonly used to make thin
polyethylene film, 10 to 30 mil in
thickness.  With this process the
size of the sheets is limited.  It
must be noted as well that some
blown-film liners contain a lubri-
cant, typically calcium stearate, to
facilitate extrusion through the
circular die.  This peculiarity of
some blown—film liners must be taken
into account when the seaming method
is specified; a process that employs
preheating will likely cause the
lubricant to migrate to the surface
and interfere with seam integrity.
The collapse and wind-up process may
also subject the liner to unaccept-
able tensile stressing.

    Most lining materials are
produced by the calendering process,
with the exception of some of the
polyethylenes.  Calendering is really
a form of extrusion with rotating die
nips.  A typical operation contains
four rolls forming three nips.  The
first is the feed pass, the second is
the metering pass and the final nip
is the sheet forming, gauging and
finishing pass.  The most typical
liner width produced by calendering
is five feet, although some calenders
are 10 feet wide.  Thickness of
lining material produced by this
process is generally limited to 10 to
60 mil.

    With the sheet extrusion process,
molten plastic is forced through a
die to form the fianl width of sheets
desired.  Sheets made by the extru-
sion process can be made in thick-
nesses as great as 250 mil, although
80 to 100 mil is more typical.  One
manufacturer is able to extrude
sheets that are 34 by up to 900 feet
in length for few field seams.

    It is important to consider what
services the liner manufacturer will
provide in addition to producing the
liner panels.  Figure I is a flow
chart depicting the various functions
that should be examined when con-
sidering a polymeric material.

    In the majority of liner
applications, site engineering work
is done for the operator of the final
facility by a consulting or in-house
engineer who recommends the required
amount of lining material to cover
the application.  The manufacturer in
those cases only supplies the lining
material.  After manufacture, the
liner is turned over to fabricators
for finishing and factory seaming and
then to installers, or contractors,
who complete the project.  Each party
in this typical factory-to-field
scheme is an independent firm.
                                      284

-------
    When designing a secure
containment application, numerous
factors involved in the complete job
should be considered, with particular
attention to seaming methods,
chemical resistance, puncture and
tear resistance and resistance to
biological attack.

Seams and SeamingMethods

    All liners are seamed, either in
the factory or the field or both.
Because seams are the most obvious
weak points of a lining system, they
must meet all the specifications for
the liner itself.  Seaming methods
include:

    Adhesive seams are joined
together with a chemical adhesive
system that bonds together two
separate membrane surfaces.
Generally, a two-component system is
employed that requires care in
mixing.  When the adhesive is
applied, the residence time before
bonding, pressure applied and timing
of pressure are all critical.  These
variables may be difficult to control
in a field situation.

    With a bodied solvent seam,
lining material is dissolved in a
solvent used to soften and bond liner
surfaces together.  This method is
essentially an adhesive seam made up
of the parent material.  Because
problems with application, timing and
pressure do exist and the amount of
solvent used is critical, field
application requires extreme care.

    Dielectric seams are made when a
high frequency current is used to
melt the surfaces of the membrane
material so that they can be
homogeneously bonded together under
pressure.  This process is used most
often for factory seams, because the
use of high frequency current is
difficult to control in the field.
     Forextrusion welded seams, a
bond is obtained between two flexible
membranes by extruding a molten
ribbon of the parent material between
overlapped pieces of liner material
followed by applied pressure.  This
is a relatively straightforward task
in the field.

     Solvent seams use solvents to
softens the surfaces to be bonded.
Generally pressure is then applied to
the seam.  Under field conditions,
the application of the solvent, and
timing and pressure can be a
difficult task.

     A chemically adhesive tape seam •
may be used to bond liner surfaces
together.  The tape system adds an
additional element to the seam system
that must be assessed for the
requirements of the application.
Taped seams are commonly made in the
field and are subject to all the
hazards associated with adhesive
seams.  The adhesive is different
material and must meet criteria and
needs of application.

     To make thermal or fused seams,
high temperature air or gas is
applied between the two surfaces to
be bonded until the membrane surfaces
melt.  At that'time pressure is
applied to create a homogeneous bond
between the two surfaces.  Timing and
pressure are critical and must be
carefully controlled.  The recent
development of a process that grinds
material from the bonding surfaces
and incorporates it into the seam is,
in this author's opinion, one to be
avoided because of the risk of
encapsulating sand or dirt
particles.  Foreign material within a
seam likely signals a void.
                                       285

-------
     Vulcanized seams are formed when
the areas to be bonded are unvulcan-
ized material that are cured together
with heat and pressure.  This method
is applicable only for thermoset
materials.

Chemical Resistance

    The primary purpose of a
synthetic liner is to protect
groundwater from contamination.
Obviously the liner itself must be
resistant to the chemicals being
contained.  When considering chemical
resistance, the operating temperature
of an impoundment as well as the
potential for exothermic chemical
reactions must be taken into
account.  Whenever possible, chemical
compatibility tests are recommended,
even urged, to assure that a liner
material is an appropriate one for
the operating uses in question.

Puncture and Tear Resistance

    It is impossible to over
emphasize the importance of liner
thickness to the integrity of the
overall lining system.  The puncture
and tear resistance of a given
plastic material is related directly
to its thickness and approximates a
geometric function.  In tests to
determine the relationship of high
density polyethylene (HOPE) liner
thickness to puncture resistance, DIN
16 727 describes a single point
application of high physical stress
at high deformation speed, a
stressing mode that can occur easily
at an installation site.  Results
showed that the height of fall from
which no damage takes place triples
when liner thickness is doubled, from
1,5mm to 3.0mm for example.

    In tests of a HOPE liner for tear
resistance in accordance with DIN 53
515, a Graves angle test specimen is
subjected to tensile stressing at a
speed of 500mm (20in.) per minute.
The test simulates rapid tearing as
would likely be found in actual site
conditions, and shows that as liner
thickness is doubled, tear resistance
also doubled.

    Another critical aspect of
specifying liner thickness is the
ability of a given thickness to
accommodate the almost inevitable
subsoil settling that can be expected
under most impoundments.  The design
engineer must ensure that the liner
material is sufficiently strong and
thick to withstand the multi-axial
tensile loading and to alleviate
localized stress points with
sufficient elongation behavior.

Resistance to Biological Attack

    A potential failure mechanism for
lining systems and one that is often
overlooked is biological attack.  Any
polymeric material containing plasti-
cizers will likely be subjected to
microbiological attack.  Microbes
actually consume certain organic
plasticizers; as the plasticizer is
lost, the liner becomes less flexible.
Eventually it will become brittle and
crack.

    Small rodents and burrowing
animals sometimes use a liner as a
food source, even to the point of
becoming addicted to certain
plasticizers.  Linings containing
rodent food sources should not be
specified for areas where rodent
populations are known to be active.
Rats can gnaw through a liner in
unexpected locations making it
difficult to isolate and repair any
leaks.

    Thin (60 mil or less) synthetic
liners are also susceptible to
peneration of roots and plant life
through them.  Soil sterilization is
a distant alternative to the better
                                       286

-------
 choice of  specifying a thick  liner
 that  is  impenetrable to  such  plant
 life.

 Conclusion

    The  use of extruded  polymers for
 the containment of hazardous  wastes
 is a  vitally  important and beneficial
 use of plastic and rubber materials.
 The details that factor  into  a,well
 thought-out specification, however,
 are numerous  and sophisticated.  This
 paper has  highlighted some of the
 fundamental design and specification
 issues that must be considered so
 that  a lining system services both
 the user and  the public.
              Di sclaimer
The work described in this paper was
not funded by the U.S. Environmental
Protection Agency.  The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
                                     287

-------
                                  SITE SELECTION
                                              SITE PREPARATION
ro
O3
CO
TYPE OF APPLICATION
                      Engineering  Design—^
                      Access  Road	
                      Unloading  Area	
                      Storage Area-
                                     Transport Vehicles	
                                     Water Drainage System-
                                     Securement Trenches-
                                     Surface Preparation for Liners
                                     Compacted Sub-Soil Surfaces
                                     Pipes, Sumps, Concrete Supports

                                              Installation Equipment
                                              Liner Delivered	
                  Basins
                  Canals
                  Curtain Walls
                  Dams
                  Renovations
                EVALUATION OF NEEDS
                     LINER SELECTION
                                                                                               FIGURE  I

                                                                                              FLOW CHART
            LINER INSTALLATION
Ventilation-
Cover Embankments^
Cover Bottom
                      —Physical  Properties
                      —Chemical  Resistance
                      -WeatherablHty
                      —Temperature Extremes
                      —Bacteria  Attack
                        Rodent Resistance
                      —Tear Resistance
                      —Long Term Aging
                      I—Repa1rab1lity
Secure Top Edges—
Seal to Pipes, etc«
Drainage-
                                                                          Leak Detection System-
                                                                          Extracting System-
Testing of Installation-
Repairs if Necessary	

-------
                       SLURRY WALL MATERIALS EVALUATION
                      TO PREVENT SROUNDWATER CONTAMINATION
                          FROM ORGANIC CONSTITUENTS
                Ken E. Davis, Marvin  C. Herring and J. Tom Hosea
                           KEN E.  DAVIS ASSOCIATES
                               3121  SAN JACINTO
                            HOUSTON, TEXAS  77004
                                   ABSTRACT

     The Installation of earthen  liners  and slurry wall trenches  constructed of
soils treated with bentonite and/or  bentonite/cement mixtures are frequently  used
today  as a means to retard or control the  flow of  contaminants from  surface
impoundments and landfill disposal  areas.  Although effective in reducing the  rate
of flow  of water and some contaminants, standard bentonite treatment is not always
effective  in controlling  some types of  contaminants,  such as chlorinated
hydrocarbons.

     Past handling and disposal of  liquid chlorinated hydrocarbon waste in earthen
impoundments and a landfill at an existing plant resulted in shallow ground water
and  soil contamination over  a broad area.  Following field investigations,
extensive  laboratory tests  were conducted to  investigate  the feasibility  of
confining  the  plume of  contamination through the installation of  a slurry trench
barrier.  Specially designed methods  of testing  were  used during  the
investigation.

     Standard bentonite soil mixtures using up to 6% or more bentonite were shown
to be ineffective in restricting the flow of the concentrated wastes found  in  the
subsurface strata.   However,  a 12%/12% bentonite/cement dust soil mixture was
shown to completely restrict the flow or movement of the concentrated waste while
reducing the flow of ground water and contaminated leachate.
INTRODUCTION AND PURPOSE

     Bentonite, or more specifically
the clay mineral montmorillonite, has
been  used for decades to impede  or
reduce  the flow  of water  through
permeable soils.  This characteristic
has been widely used in the drilling
and construction industries.

     The use  of bentonite began  in
the petroleum  exploration  industry
during  the 1920's in drilling muds
used to seal the  sides of  the hole
and to provide  other  benefits which
aided the drilling  operation.   During
the 1940's, it was found that borings
and trenches could be kept open using
a slurry of bentonite, and that  the
resulting trench when  backfilled
exhibited a reduced  permeability to
the horizontal flow  of ground water.
More recently, bentonite  slurry wall
trenches have found  application to
control  pollution migration   for
improperly designed  waste  landfills
and impoundments.

     The design  of  soil bentonite
slurry  walls  for  conventional
groundwater control  applications  has
been widely documented.  Other than
                                     289

-------
normal  standard testing, very little
Investigative work is required.  The
design  of  a  soil-bentonite slurry
wall  for waste  containment however,
requires  site specific studies in
order to determine the effectiveness
of the proposed  barrier.

APPROACH

Test  Program

     The presence  of  organic or
inorganic  compounds  in   the
groundwater can  have a detrimental
effect on the  bentonite slurry wall
and its  ability to  contain contami-
nants.   Certain  chemicals can affect
the chemical/physical properties of
the bentonite  and cause  ailure of
the barrier either  during construc-
tion  or  at  some time in the future.
Therefore,  the  effects of the  ground-
water, waste leachate or liquid waste
on the soil/bentonite system  must be
determined  prior to the preparation
of final design specifications for
the slurry wall.

     Prior  to the start of laboratory
investigations,  detailed field
investigations  are essential to
determine the  geohydrolpgy  of the
proposed slurry trench site.   Unless
detailed information is available!,
sufficient soil  borings should be
made  of  the entire area to define the
geology, particularly within  the area
in which the proposed trench will be
located.   In  addition, sufficient
pump  tests  and field permeability
tests should  be made to define the
hydrology of the area.

     During the course of the field
investigations, chemical analysis
should be made  on the groundwater and
core  samples to  define the extent of
the contamination plume.  Samples of
soil, the  waste, groundwater and
leachate should be  collected  for use
in subsequent laboratory evaluations
of the  proposed  slurry  trench
composition.
     In  this  paper,  the  authors
describe the laboratory  investiga-
tions conducted during  the evaluation
of a proposed  slurry trench barrier
to  contain  contamination  from an
existing industrial disposal area.

Site Description

The  site  for  the proposed slurry
trench barrier was an  old disposal
area within a large  petrochemical
manufacturing  plant that had  been
used for various  waste activities in
the  past.   This included  the
collection  and storage  of liquid
chlorinated  hydrocarbon wastes in
earthen  pits or ponds  and  the
landfill disposal  of solid  organic
and  inorganic wastes, drums of
miscellaneous liquid organic wastes,
plus construction debris  such as
wood, concrete, etc.

     The  field  investigations
indicated that  extensive  shallow
groundwater and soil  contamination
had occurred.   In  addition,  immis-
cible and  relatively pure undiluted
waste organics were found to exist at
points more  than  300 feet away  from
the  original  source  and  down to
depths as  great as 50-60 feet below
the surface.

Site Geology

     The geology of  the area  was
characterized  by  a combination of
hydraulic  and  mechanical fill over-
lying deposits  of  recent alluvium.
Extensive  geological investigations
of the  area were conducted  and a
detailed  description  of this area
developed.

     The proposed area where  the
slurry  trench barrier  would be
constructed  was on the eastern  edge
of the disposal area.  Several  core
borings were  drilled in  this area.

     The geology of  the proposed
slurry trench area was  characterized
                                   290

-------
by several distinct layers of soil.
The area is  covered with a layer of
silty clayey sand fill material from
the surface to a depth of approxi-
mately 12  to 14  feet  below  the
surface.   Beneath  this on the
northern end of the proposed trench
area is a soft black  organic clay
layer  approximately 4 to 6 feet thick.
Beneath this is a layer of fine  to
medium silty sand approximately 25 to
 5 feet  thick that  extended downward
to a depth of 50 to 60 feet below the
surface.   Underlying these soil
layers is a stiff light gray to tan
clay found at depths of 50 to  60 feet
below  the surface throughout the area.
Electric logs of wells in the area
indicate the thickness of this base
clay layer to be 75 to 100 feet.

    Samples  of  soil  and groundwater
taken  from core borings show the soil
and water in the area of the proposed
slurry trench to  be contaminated with
chlorinated hydrocarbons. Data from
samples show  contamination was
present at various depths down to the
base clay  encountered at a depth  of
57 feet below the surface.  Contin-
uous samples were  not  taken for con-
tamination analysis from all borings.
However, a composite of samples taken
at  2  feet  intervals showed high
levels  of  contamination.   Visual
evidence  of  high  concentrations  of
phased  organic contaminants was
noticed during  the drilling  opera-
tion,  especially at  depths   of
approximately 12  through 20 feet
below the  surface.   An composite
analysis of samples taken at 2 feet
intervals from the  depths of  2 to 65
feet  below the surface showed  the
soil  contained  0.59 wt.%  total
chlori-nated hydrocarbons.   These
results are presented on Table  1.
Groundwater  samples  taken from  the
same  location contain 942  ppm  of
total chlorinated  hydrocarbons  as
shown  on Table 2.
    TABLE 1.  CHEMICAL  ANALYSIS OF
      COMPOSITE SOIL MIXTURE  B
Compound
       Wt.
Methyl  Chloroform             0.02
Trichloroethylene            <0.01
Trichloroethane              0.03
Perchloroethane              0.04
Tetrachloroethane             0.01
Tetrachloroethane  (Sym)       0.06
Pentachlorobutudiene          0.10
Hexachlorobutane             0.05
Hexachlorobutadiene           0.26
Unknown                      0.02
             Total           DTW

     TABLE 2. SROUNDWATER ANALYSIS
Component
Concentration
    mg/1
Chloroethylene             0.80
Chloroethane               0.04
Methylene chloride          0.15
1,1 -  Dichloroethene       25.9
1,1 -  Dichloroethane        4.08
1,2 -  Dichloroethene
Trichloromethane         513
1,2 -  Dichloroethane        4.82
1,1,1, - Trichlorethane     4.82
Tetrachloromethane
Trichloroethene            7.12
1,1,2  - Tricholrothane    260
Tetrachloroethene         126
             Total       942

Method of Test

    A standard method of test for
evaluating the inhibiting effects of
soil  additives  to migration  of
organics  through  soils  was  not
available.  Therefore, modification
of standard  permeability testing of
soils  was required for this work.

    A constant head permeability
method of test was  utilized through-
hout  the  laboratory investigation.
However,  instead of  using  the
standard up-flow of fluid through the
core to be tested, a downward  flow
was  used  in  all  cases.  A  flow
                                   291

-------
diagram of the  test apparatus is
presented on Figure 1.

Figure 1.  Permeability Test Unit
                    PRESSURE SOURCE
       Ti»T UNIT      FLUIO BEaEBVOIR
     The  test apparatus was  con-
structed from major component parts
purchased from Soil Test, Inc.  This
included  one high  pressure K-670
Miniature Permeater Unit plus two
additional K-620  Soil Test Permeater
Units.   These were connected  to
enable three simultaneous permeabi-
lity tests.  Subsequently, a second
test unit  consisting of two permea-
ters and  pressure  reservoir was  added
to  expedite  the  test  work  with
contaminated water  as a fluid.

     Operating pressures from 6 psi
to 20 psi  were used to accelerate the
tests.   Constant  pressures  were
maintained through  the use of an air
compressor attached to  the  fluid
reservoir.  The volume of fluid flows
measured  in a buret attached to each
test unit  (Figure  2), or by flowing
into a graduated cylinder through the
bottom drain valve.
Figure 2.  Test Permeater


                  • 1/4" TUBING FITTING
                                                                      POROUS
                                                                      STONES
                                           FLOW OUTLET *
                     200 MESH SCREEN
     Pretreatment of the soils with
the  additives  to  be tested  was
accomplished by direct addition  on a
weight  percent basis.  Approximately
lOOg of soil mixture was prepared for
each test.  Following the addition of
the  additives,  water was added as
required during  mixing to achieve a
medium to stiff consistency prior to
compression in the test mold.

     The unconsolidated soils were
prepared for testing  using parts from
a Harvard  Miniature  Compaction
Apparatus.  The tare weight  of  the
mold or  sleeve  from the permeater
plus a  200 mesh  screen and a 3.0 cm
porous  stone was obtained.  After
clamping the mold in  the mold holder,
the screen was placed in the  bottom
of the mold.  Approximately 30 cc of
the soil to be tested was  added and
gently tapped or  pressed  into  the
mold.  This amount of sample occupied
approximately one-half of the total
capacity of the  mold.  Water  was
added to maintain a saturated test
core during compression.  The porous
stone was then  placed on top of the
                                    292

-------
test sample in the mold and pressure
applied with the collar remover press.
This unit is equipped with  a spring
loaded  plunger with  a base plate
approximately 3.0  cm in  diameter.
The pressure was  applied by pushing
down on  the compression  rod.   In
total,  approximately 50 pounds  per
square inch was applied for a period
of 60 seconds.   After compressing the
sample  into the mold, the mold was
removed and  the final  weight of the
unit with the sample recorded.  Prior
to assembling the mold in the permea-
ter  the  height and  diameter of the
test sample  was  measured   and
recorded.

     Following  assembly of the  test
units,  water was  added to  the  fluid
reservoir.  The reservoir  was raised
to the  desired test pressure through
the use of an air compressor.  Prior
to beginning the tests, sufficient
water was added to each permeater to
displace the air trapped above  the
test core through the upper bleeder
valve.  Once the test unit was filled
with water,  the tests were  begun by
measuring the  flow through the  core
at various intervals of time.

     The  organic  permeability  tests
were also prepared  as described above.
Following measurement of  the  water
permeability,  the  inlet line at the
top of each unit was disconnected and
excess water above  the test cores
removed  (approximately  25 ml).
Approximately 25 ml  of chlorinated
hydrocarbon waste  was then added
through the inlet  opening at the top
of each  test  unit.   The water  line
was then reconnected  and   the  test
resumed  by  measuring the  flow of
water or organics at  various  time
intervals.  Following passage of the
organic wastes through the test  core,
the tests were continued to measure
the permeability of water through the
test core after exposure or contract
with the organic waste.
Soils Tested

     Various  soils were tested for
water  and/or  organic permeability
during the course of  the study.   Some
of these soils were  samples  taken
from specific  stratas encountered
during the core borings.  Some were
composite mixtures of soil represen-
tative of the overall stratas  that
would  be encountered  in an excavation
in the area.  These  included  soils
free from organic contamination and
those which are currently contami-
nated  in the  proposed slurry trench
area.

     Several  samples of  different
types of  uncontaminated soil  stratas
found  in the  are*. ^w" *« a  depth or
55 to  60 feet were  evaluated for
water  permeability characteristics.
Jhe  data were used  to confirm the
accuracy of the special test method
used  throughout  the study  and to
develop background data for hydraulic
flow in this area.

     In addition to the above, a com-
posite mixture of  uncontaminated
soils was prepared in proportion to
the soils that would be encountered
in the proposed slurry trench  area.
According to core borings the percen-
tage of various types of  soils  down
to a depth of 64 feet are presented
on Table 3.

    TABLE 3. COMPOSITION OF SOIL
   MIXTURE A (UNCONTAMINATED SOIL)

SoilDescription & Depth   Percentage

Fine to medium tan
   silty clayey sand
   (0-16)                    25.0%
Firm light gray and
   brown  silty clay
   (16-19)                    3.7%
Soft to very soft
   black organic clay
   (19-23)                    6.2%
                                    293

-------
         TABLE 3 CONTINUED
Light gray fine to
   medium sand
   (23-59)
Stiff light gray
   and tan clay
   (59-63)
56.5%
 5.5%
     A mixture of  these types  of
soils was prepared from samples  taken
previously from the core borings.
This composite sample, referred  to  as
"Soil  Mixture A"  during the study,
was tested for water permeability
both with and without the addition  of
bentonite sealants.

     A contaminated soil mixture,
referred  throughout the study  as
"Soil Mixture B",  was also prepared
from  samples taken  from the  core
borings in the area of the proposed
slurry  trench.  This  mixture was
prepared  by taking  equal  portions
from samples taken  every two  feet
during the drilling operations. The
composition of soil  mixture  B  is
shown on Table 4.

    TABLE 4. COMPOSITION OF SOIL
    MIXTURE B (CONTAMINATED SOIL)

SoilDescription & Depth   Percentage
Tan silty clayey sand
   0-12
Tan and dark gray
   sandy (12-16)
Dark gray to black
   silty organic clay
   (16-20)
Silty, sandy clay
   (20-24)
Tan fine to medium
  sand (24-60)
Stiff gray and tan
   clay (60-62)

Fluids Tested
20.8%

 6.9%


 6.9%

 6.9%

55.2%

 3.4%
     Four  different fluids were
utilized during the investigation.
These included uncontaminated ground
water, contaminated  ground water.,
ground water which was saturated with
chlorinated hydrocarbon wastes  in  the
laboratory,  and  concentrated
chlori-nated hydrocarbon wastes.

     Two five  gallon containers of
contaminated water from the  area  were
contained for testing.  An analysis
of  this  water  was previously
presented in Table  2.  This water
sample was used throughout most  of
the test work and is referred to as
water in  the  balance  of  the report.
The water is of a quality  that could
be  expected to flow through  the
proposed slurry  trench.

     The concentrated chlorinated
hydrocarbon wastes used throughout
the investigation was a sample of API
separator organic waste  from  the
plant.  Although the exact composi-
tion of  this material  was  not
determined, it is believed that  this
is similar  in  composition to  the
waste materials processed  in earthen
ponds in the past that was the  source
of the plume of contamination in the
area.

Additive Materials Tested

     Several  oil  additive  materials
were evaluated  during the  course of
the  investigation.   These  included
two bentonite materials and cement
flue dust.  Although other possible
additive materials are available,  the
scope of the project was limited to
the first material found to  provide a
successful barrier to the flow of the
organic waste.

     The  bentonite  materials used in
the  investigation  were Volclay
Bentonite 125 (BENT125)  and Volclay
Saline  Seal  100  (SS100).   These'
materials are manufactured  by  the
American Colloid Company, Skokie,
Illinois  60077.   Bentonite  125  is
specially formulated for use  in
                                    294

-------
slurry trench construction to reduce
the flow  of water.  Saline Seal 100
is specifically formulated for use
where the fluid to be  controlled
contains relatively  high  concen-
trations of  salts and/or  other
contaminates.

    The  cement flue  dust utilized
throughout the investigation is a by-
product manufactured  by  several
cement plants  located in  Houston and
Dallas,  Texas.  The  sample  of the
cement flue  dust tested was manu-
factured by  Gulf Coast Cement in
Houston.

PROBLEMS  ENCOUNTERED

    The  only problems  encountered
during this series of tests involved
the method  of testing. Since no
standard method was available for
testing the  permeability of  organic
constituents  to  slurry walls  a
modification of ASTM  D-2434-68 was
incorporated.

    The  first  problem involved the
density differences between water and
chlorinated  hydrocarbon  waste.  The
ASTM method calls for  flow from the
bottom to the top of the  sample cell.
However,  the  density of  the chlori-
nated  hydrocarbon  caused water, the
displacement  fluid,  to pass  by the
waste  material and  directly  through
the soil.  To  alleviate this problem
flow was introduced from the top of
the cell  to the bottom.

    The  chlorinated hydrocarbon was
introduced directly into the  sample
cell before  water  was  added from the
top, as  a driving fluid.   This
reduced  the  possibility of phase
separation  and allowed a  slug of
contaminants  to be  injected into the
soil mixture.

    The  other  problem involved the
consolidation of soil  into a  stable
form.   As fluid was passed through
the  soil  mixture,  mobile  particles
traveled to the confining screen  and
plugged the face of the sample cell,
restricting flow.  Porous stones were
used as supplements to the screen to
reduce  the amount of face plugging by
distributing mobile particles over a
broader area.  This eliminated  the
flow restriction  and allowed long
term testing.

RESULTS

On-Site Soil Permeabilities

     The  water  permeability  of
various uncontaminated soil  strata
found in the area was determined in
the laboratory using  the  modified
test method.  These  data are pre-
sented on Table 5.  When  compared
with the field permeability  data of
similar type soils in the area, these
data indicate the modified test
method, used throughout the investi-
gation,  produces  results that  are
comparable with  standard methods of
test  for  field  permeability
measurements.
    TABLE  5  - WATER  PERMEABILITY
       UNCONTAMINATED SOILS
         OF
     Core DepthJ

     3-5 feet
     5-7 feet
     20 - 22 feet
     50 - 52 feet
     62 - 63 feet
     Soil Mixture A
     0-63 feet
Permeability,
    cm/ sec

  5.0 x lO-5
  2.3 x 10-4
  9.0 x ID"5
   .0 x 10-3
     z
2
1.2
     ID'8
3.7 x ID'7
1.9 x 10-7
     Several  samples of contaminated
soil from  the borings  were  also
tested for  water permeability.  These
included strata of  soils encountered
at various  depths  plus a  mixture  of
soils found at various  depths in the
area of the proposed slurry trench.
The  data  from these tests  are
presented on Table 6.
                                    295

-------
     TABLE 6. WATER PERMEABILITY OF
         CONTAMINATED SOILS
     Core Depth

     0-12 feet
     12 to 25 feet*
     25 to 60 feet
     62 to 65 feet
     Soil Mix B
     0-62 feet

     * Zone  of highest
      of contamination
Permeability,
    cm/sec,

  5.0 x 10-5
  2.1 x 10-7
  4.4 x ID'3
  1.8 x ID'8
  2.4 x ID'6
  1.9 x lO-6

 concentration
     The  fi11 material  (0-12 ft)
above the contaminated area  had a
permeability of 5.0  x 10~5 cm/sec.  A
mixture of  soil  in the zone  of
highest contamination  (12 to  25 ft)
had  a  permeability of 2.1  x 10"'
cm/sec.  Contaminated fine to  medium
sand (25 to 60 ft)  had a permeability
of 4.4  x  10~3 cm/sec, or approxi-
mately  the same as  the uncontaminated
sand from the same strata.  A mixture
of contaminated and uncontaminated
soils throughout  a  depth  of 0  to 65
feet below the surface (Soil Mixture
B) had permeabilities of  1.9  x 10~6
cm/sec and 2.4 x  10~6  cm/sec. The
difference between the permeabilities
of the  contaminated soil mixture
(Soil Mixture B)  and the  uncontami-
nated soil  mixture (Soil Mixture A)
of  approximately one  order  of
magnitude is probably a result of the
contaminated soil mixture containing
a higher percentage  of  sand (60% vs
56%) plus a lower  percentage of  dense
clay (2.6% vs 5.5%).

     Permeabilities  to water and
organic  liquids were  also determined
on several  samples  of  contaminated
soils  as  shown on  Table 7.  The
permeability of  Soil  Mixture B to
organics was 2.4  x  10~5 cm/sec  as
compared  with an average  initial
water permeability of 2.7 x  10~6
cm/sec. The contaminated zone mixture
had a permeability to organics of 3.1
x 1Q-5 cm/sec as compared with  an
average initial  water permeability of
3.3 x  10~7 cm/sec.  It is interesting
to  note  that although the initial
permeabilities to water  of these two
soil mixtures differ by  a factor of
approximately 10,  the permeability to
organics was approximately the  same.
An attempt was also made to determine
the permeability to organics of the
stiff clay found at  a  depth of  60
feet below the surface. Although the
clay exhibited an initial permea-
bility of 1.8 x 10~8 cm/sec to  water
and continued to  flow  water through-
out the  test,  there  was no visible
flow of organic through the test  core
as experienced during the other soil
tests.

TABLE 7. PERMEABILITY OF CONTAMINATED
               SOILS

Soil Sample    Ml    £      M£

Mixture B      10~6   10~5   10~7
(0-62  feet)
Contaminzted   10~7   lO"5   10~7
(12-26 feet)
Base Clay      10-°    0     lO"8
(60-62 feet)

Evaluation ofBentonite  125  and
Saline Seal 100

    A series of water  permeability
tests were conducted on  samples  of
contaminated Soil Mixture  B  when
treated with from 1% to  6% Bentonite
125 or Saline Seal  100.  These  tests
were conducted to compare the  per-
formance of  the two  different
bentonite  materials  in  the presence
of the contaminants contained in the
on-site soil. The results from these
tests  are presented  graphically on
Figure 3.
                                    296

-------
Figure 3. Permeability to Water vs.
      Percent  Bentonite
   . r
   10
  - -•
 <10
 UJ
 2  .
 a.
    -8
   10
BENTONITE 125

BENTONITE SS100
          12345
          PERCENT BENTONITE
                        organics was observed.  The average
                        permeability to water was reduced
                        from approximately 2.7 x 10-° cm/sec
                        down to 7.9 x 10"' cm/sec with  the
                        addition of 16% Bentonite 125.  The
                        maximum permeability to organic  was
                        reduced from 2.4 x 10"5 cm/sec down
                        to approximately 1.3 x 10~6 cm/sec at
                        a 16% Bentonite 125 treatment level.
                        In most instances,  the permeability
                        to water  was shown  to be approxi-
                        mately  the same as it was immediately
                        following passage  of the organics
                        through the test  core.  A plot  of  the
                        permeabilities vs percent Bentonite
                        125 is  presented  in Figure 4.

                        Figure  4.  Permeability vs.  Percent
                               Bentonite  125
    The  data  from  these  tests  show
the addition of of bentonite  to  Soil
Mixture B  results in  a gradual
decrease  in  the permeability of the
soil to  water.  With 6%  bentonite,
the permeability was reduced  from
approximately 2 x  10~6  cm/sec to 3  x
ID'7 cm/sec. Very little  if any
difference was observed in the
performance of the Bentonite  125 and
the Saline Seal  100 with the  contami-
nated soil. However, Saline Seal 100
was selected  for use in  all  subse-
quent test  work due to  its reported
resistance to high  concentrations of
salts.
Permeab.11 ity of
Contaminated Sol1
Bentonite Treated
     A  series of  tests  were  run  to
determine the water  and  organic
permeability of  the  contaminated
in-place soils (Soil  Mixture B)  when
treated with up to 16% Bentonite 125
or Saline Seal 100.

     When  treated with increasing
amounts of Bentonite 125,  a propor-
tional decrease in the permeability
of  Soil  Mixture B  to water and
                          « 10

                          o  .,
                          > 10

                          i  •*
                          < 10
                          UJ
                          S  -,
                          .S 10
                             »s
                            10
                                                • WATER
                                                • ORGANIC
                                   4   6   8  10  12  14
                                    PERCENT BENTONITE 125
                                                      16
                             Treatment with increasing
                         amounts of Saline Seal 100 resulted
                         in similar reductions in the permea-
                         bilities.  To  water, the  average
                         permeability was reduced from 2.7 x
                         10"6 cm/sec down to  approximately 9.0
                         x  ID'9 cm/sec with 16% SS100.  The
                         maximum permeability to organics was
                         reduced from 2.4 x 10-5 cm/sec down
                         to approximately 1.1 x  ID'7 cm/sec at
                         the 16% treatment  level.  A  plot of
                         the permeabilities vs percent SS100
                         is presented in Figure  5.
                                    297

-------
 Figure 5.  Permeability vs. Saline
           Seal 100
          4  8  8  10  12 14
          PERCENT SALINE SEAL 100
16
    The data  from  these tests
indicate treatment of  a mixture of
in-place soils in the  area of the
proposed slurry  trench  would require
the addition of  a minimum of 12%
bentonite (Bentonite  125 or Saline
Seal 100) to achieve  a permeability
of  1  x 10-7  cm/sec  to  water.
Treat-ment  with  a minimum of 16%
bentonite (Saline Seal  100)  would be
required to achieve a  permeability of
approxi-  mately 1Q~7 cm/sec or  less
to  organics. Although limited  data
are available on a  16%  Bentonite 125
treated soil, Saline Seal 100 appears
to  be  slightly more  efficient in
reducing the permeability of the
organic.  At all treatment levels,
however, the permeability to organics
was shown  to  be  10  to 100  times
greater  than the permeability to
water.   A positive flow  of
concen-trated  organics through the
test cores was observed in all of the
tests  regardless of  the level of
treatment.

Cement Flue Dust/Bentonite Treated
loTT       ~~

     Several samples  of Soil Mixture
B treated with 6% and 12% Saline  Seal
100 plus 1% up to a  maximum  of 18%
cement flue  dust were evaluated.

     The addition of  increasing
amounts of cement flue  dust  to  the 6%
Saline Seal  100 treated  soil resulted
in a  relatively small but correspond-
ing increase in  the permeability of
the mixtures  to water.  The average
water  permeability  increased from
approximately 3.8 x  10~7 cm/sec with
no flue dust up  to  5.6  x 10~7  cm/sec
with  12% flue dust.  The addition of
flue dust  failed to reduce the
permeability of the mixtures  to
organics.  Permeabilities within the
range of  1.7  x 10'6 cm/sec to  1.9 x
10"5  cm/sec were  measured for
organics.  In all cases, the permea-
bility to organic was shown  to be
higher  than the permeability to water.
Essentially all  of the organic
charged during each test  was observed
to pass through  the test cores. The
results from these test  presented on
Figure  6.

Figure 6. Permeability  vs. Flue Dust
         (With 6% SS100 Treated Soil)
               10
              o  -e
              jf 10

              3  ->
              m 10
             Ul
             a.
               10
               10
                         ' WATER
                         k ORGANIC
                          6  8  10  12  14
                         PERCENT FLUE DUST
                             16
                 The  addition  of  increasing
            amounts of cement flue dust to the
            12%  Saline Seal  100  treated  soil
            resulted in a similar increase  in the
            permeability to water.  The average
            permeability to water  increased  from
            1.7  x 10~8 cm/sec with no flue dust
            up to 9.5  x 10~7 cm/sec with  18%  flue
            dust.   The addition  of cement  flue
            dust to the 12% SS100 treated  soil
            however had a  significant effect  on
            the  permeability to  organics  at
            concentrations  of  6% flue dust  or
            greater.   The results  are presented
            on Figure  7.
                                    298

-------
 Figure 7. Permeability vs.  Percent
          Flue Dust  (With  12% SS100
          Treated Soil  Mixture)
o
a>  -s
5 10

0  -«
>: 10
=!  -»
m 10
  10
• WATER
»ORGANIC
       2   4   6  8  10  12  14 16

            PERCENT FLUE DUST
     The addition  1% to 6% cement
flue dust to the  12% Saline Seal 100
treated soil had little if any effect
on the  permeability  to the organics.
Essentially all   of  the  organics
charged to the units were observed to
flow through the test cores.  Permea-
bilities in the range  of  1  x 1Q~7
cm/sec to  5 x 10"'  cm/sec were
measured  for water  and 5  x  10~6
cm/ sec  for the organic.

     Using concentrations  greater
than 6% flue dust mixed with  the 12%
Saline Seal 100 treated soil, the
addition of organics to the  system
ultimately resulted  in  a sharp
decrease  in permeability  with
ultimate  plugging of the core.
Similar results were obtained with
12% and 18% flue dust.  In most cases
a small  quantity of water (1 to  3 ml )
was observed to pass through the core
following addition  of the  organic.
This was followed by a sharp decrease
in permeability to less than 1 x 10~9
cm/sec with ultimate plugging of the
test core.

Conf 1 rmat i on Tests

     Several  tests  were conducted in
the laboratory to confirm the  initial
positive  test results on  the 12%
SS100/12% flue dust treated  soils.
Tests  were also run  to investigate
the effects of various  other factors
on the  efficiency  and long  term
stability  of these systems.  In
addition,  a  test was run to confirm
the initial  results  obtained  on  the
stiff clay that underlies the area.

Permeability of 12%  SS100/12% Flue
Dust Treated Soil
                       A  sample  of Soil Mixture B
                   treated with 12% SS100  and 12% flue
                   dust  was prepared and tested in
                   accordance with the standard proce-
                   dures  used  throughout the investiga-
                   tion. The resulting mixture exhibited
                   an average  permeability  of 6.1 x 10~'
                   cm/sec to water.  Upon  contact with
                   the organic  waste,  a  sharp  decrease
                   in permeability was observed that
                   ultimately resulted  in pluggin
                   (permeability  less than 1  x 10"
                   cm/sec)  of  the core following passage
                   of approximately 2.5 ml of water.

                   12%/12% Mixture with Contaminated
                   Water
                       A  sample  of Soil  Mixture B
                   treated with 12% flue dust  and 12%
                   Saline  Seal  100 was  prepared for
                   testing using  contaminated water for
                   hydration  of the mixture.  This test
                   core exhibited  an average  permea-
                   bility to  water of 8.1 x 10"' cm/sec.
                   Following addition of  organics to the
                   unit,  the  permeability increased to
                   approximately 1.1 x 10~6 cm/sec for
                   the next  5  ml  of  fluid  to  pass
                   through the core.    During  this
                   period, approximately 1.5 ml of the
                   25 ml of organic charged to the  unit
                   passed through  the core. Subsequent-
                   ly,  a sharp  decrease  in permeability
                   occurred with ultimate plugging and
                   no flow of  fluid through the core.

                   12% Flue  Dust/12% SS100 Time-Set
                   Compact fofT

                       A  sample  of Soil  Mixture B
                   treated with 12% flue dust  and 12%
                   Saline  Seal  100 was  prepared for
                                    299

-------
testing. In this test, the mold was
filled with contaminated water  and
allowed to cure over a period of 12
days prior  to  beginning the permea-
bility testing.  This core  exhibited
a permeability  of 5.2 x 10~?  to water.
Contact with the organics resulted in
a sharp  decrease in permeability  with
plugging  of   the  core  following
passage  of approximately 3 ml  water.
No flow of organic through the core
was observed during the test.

W a t e r  S t r i p p i n g  of Avail a b1e
Alkalinity

     Analysis  of the effluent water
from previous tests using cement  flue
dust indicated an increase in pH from
7.6 up to 12.0  could  be expected due
to leaching of the available alkali-
nity in the  flue  dust from  the
treated soil  mixture.  A  test was
therefore  begun to evaluate  the
long-term effects  of water  stripping
of the available alkalinity  from the
12%/12% system on  the efficiency of
the system  to  retard or prevent the
movement of phased organics.

     A  sample of Soil  Mixture B
treated with 12% flue dust and  12%
SS100 was prepared in accordance with
the standard procedure for  testing.
The  pH  of  the feed water and  the
effluent was 6.9 and 12.0,  respect-
ively at the  start  of the  test. A
gradual  decrease in  the pH of  the
effluent occurred  during the course
of the test.  After  three months,
7,300 ml of water  had passed through
the core and the pH  of the  effluent
had decreased  to 7.0.  At that time,
22 ml of organics were charged  to the
unit.

     The permeability of  the  test
core to water  remained relatively
constant at approximately  3  x  10~7
cm/sec during  the first 2,000 ml of
water to pass  through the core.  The
pH of the effluent at this point had
decreased to approximately  9.8   A
gradual  increase in  permeability  to
approximately 3  x  10~6  cm/sec
occurred during  passage  of  the next
5,300 ml  of water.   Following the
addition   of   the  organics,
permeabilities  within the  range  of
1.0 x ID-6  cm/sec  to 6.0  x 10-6
cm/sec were measured for the next  30
ml of fluid to pass through the core.
This flow consisted of approximately
20 ml of phased organics  of  10 ml  of
water.    Subsequently, the  flow
consisted  of  water free  from  phased
organics  at a  permeability  of 5  x
10~6 cm/sec  for  the balance of the
test.   The pH of the effluent water
at the end of the test was 6.9.

E f f e c t s  o f Con t am in at e d W at er on t h e
12%/12% Mixture

    A  sample  of Soil   Mixture  B
treated with 12% flue dust  and 12%
Saline  Seal 100 was prepared  in
accordance with   the  standard  pro-
cedure  for  testing.   Saturated
contaminated water prepared in the
laboratory were  used to determine the
permeability  to water.  Approximately
2,000 ml  of  contaminated water was
passed through the test core  prior  to
contact  with the concentrated organic
waste.   During this  period, the  pH
decreased from 11.9 to 8.3.

    An  average  permeability of 3.0 x
10~7 cm/sec was measured for  contami-
nated  water.   On contact  with concen-
trated organic,  a sharp  decrease  in
permeability was observed following
passage  of an additional 2.5 ml  of
water  down to approximately 1 x 10~9
cm/sec.   No  phases  organic passed
through the core  during the test.

Base Clay Permeability

    An  undisturbed sample  of the
stiff  gray base  clay that is found
throughout the area was  prepared for
long-term testing  for  water and
                                   300

-------
organic permeability.  The test  core
exhibited an average  permeability to
water of 4.9 x 10~9 cm/sec. Following
passage of 4.1  ml  of water  through
the core, 25 ml  of organic was added
to the  unit.  During the next 80  days
approximately 25 ml of  additional
water passed through the  test core
with  no visible organics  in  the
effluent.  An initial increase in
permeability to  approximately 1 x
10"^ cm/sec was measured immediately
after the organic was added, followed
by a decrease to within the range of
5 x ID'10 cm/sec  and 6 x  10~9 cm/sec
during  the balance of the test.

DISCUSSION

     The  permeability of concentrated
or phased chlorinated hydrocarbon
wastes through the type of soils
found above the stiff base  clay in
the  area  was  shown  to  be
substan-tially greater  than  the
permeability of water through the
same soils.  A mixture of soils  in the
zone of maximum contamination (12 to
25 feet below the surface)  in  the
area of  the  proposed slurry  trench
exhibited a permeability of 3.3 x
1Q~7 cm/sec to  water and  3.1  x  10~5
cm/sec to the concentrated  organic
waste.   A mixture  of soil found  from
the surface down  to a depth of 62
feet had a  permeability of 2.6 x  10~7
cm/sec  to water and 2.4  x 10"5  cm/sec
to organic.

     The  stiff gray and tan base  clay
that extends throughout the area
which was encountered  at  a  depth of
approximately 50 to 60 feet below the
surface in  the  area of the  proposed
slurry trench  was shown to be an
effective  barrier to the downward
migration of  the waste.   This stiff
clay was shown to have  a permeability
in the  range of 4.9 x 10~'  cm/sec and
1.8 x 10"°  cm/sec to water.  The  clay
was impermeable to the flow  of the
concentrated  or  phased  chlorinated
hydrocarbon.

     Treatment  of  the  in-place soils
to produce an effective slurry  trench
barrier to the migration of the
phased organic waste  was  shown  to  be
feasible using one or two different
compositions.  In-place contaminated
soils when  treated  with  16%  Saline
Seal  100 exhibited  a  permeability  of
approximately 9  x  10~9 cm/sec  to
water and  3 x 10-7  Cm/sec to the
concentrated organic.  In-place soils
treated with 12% cement flue  dust and
12%  Saline Seal 100 exhibited a
permeability of approximately 4 x
10"7 cm/sec to water but  was shown  to
be essentially impermeable to the
concentrated organic  waste.   On
contact with the organic,  a sharp
decrease in permeability  occurred and
ultimately  after  passage  of
approximately 1-2 ml  of  water, the
test  cores  plugged with  essentially
no further flow of water or organics.

     The 12% flue  dust/12% Saline
Seal  100 treated soil mixture was
shown to  have a  limited effective
life  time to prohibit or restrict the
flow of phasedorganics.  Water
flowing  through the  mixture permea-
bility in  the range of 3 x 10-7
cm/sec  to 4  x  10"6  cm/sec)  resulted
in stripping of alkaline  constituents
from the  system and subsequent
failure to restrict the  flow  of
phased organics.   According to the
test data available from  the contami-
nated water flow test and the long
term water stripping test failure
should occur following  passage  of
from 73  to 225 trench  volumes  of
water through the barrier.  Under
normal  conditions, the amount of time
required for this  volume  of  water  to
move through  the  slurry  wall  would
range from approximately 1000 to 3000
years.
                                   301

-------
REFERENCES

1.  Anderson,  D.  C., Brown, K.W. and
    Green,  J. Organic Leachate
    Effects on the Permeability of
    Clay Liners.  Remedial Response,
    p. 223.

2.  D'Appolonia, Davis Jo. April 1980.
    Soil  Bentonite Slurry  Trench
    Cutoffs.   Journal  of the
    Geotechnical  E n g inee r i n g
    Division"

3.  Daniel, Davie E., Foreman, David
    E. Effects  of Hydraulic Gradient
    and Method  of  Testing on  the
    Hydraulic Conductivity  of
    Com-pacted  Clay  to  Water,
    Methanol, and Heptane, University
    of Texas, Austin, TX.

4.  Hughes, John. September,  1975.
    Use of  Bentonite as a  Soil
    Sealant for  Leachate Control
    Sanitary Landfills.  Void ay Soil
    Laboratory E n g i n e er _i n g  Report.
    Data 280-E.

5.  Kugs,  Charles, Rogoshewski, Paul,
    Repa,  Edward.  Alternatives  to
    Ground  Water Pumping  for
    Controlling  Hazardous  Waste
    Leachates,  JRB Associates.

6.  Laboratory Soil  Testing.
    Engineer Manual  EM 1110-2-1906,
    Department of the Army, Corps of
    Engineers.

7.  Ryan, Christopher.  February 1977.
    Slurry Cut-Off Walls  Design
    Parameters and Final Properties
    an  Interim Report,  Technical
    Course Slurry Wall  Construction,
    Design,  Techniques,  and
    Procedures, Miami, Florida

8.  Slurry Cut-Off Walls Methods and
    Applications, Geo-Tec  '80,
    Chicago, Illinois, March, 1980.
      The  work  described in this  paper
      Protection Agency.  The contents
     9.   "Standard  Test  Method for
         Permeability of Granular Soils
         (Constant Head)"  ASTM D-2434-68
         (Reapproved 1974)

     10.  T e c h n i c a 1  C at a 1 o g  f g r S o i 1
         S e a 1 a n t s, S1u r r y Tr en c h i n g an d
         Sanitary  L andf 111 ,  American
        Colloid Company,.
     11. T s h e b o t ar i f f ,  Gregory
   P.
Earth
and
      Agency and no official endorsement should be inferred.
        Foundations Retaining  and
        Structures,  Seepage
        Capillarity, p. 316.

     About the Authors
     Ken E.  Davis  is president of Ken E.
     Davis Associates,  3121 San  Jacinto,
     Suite  102,  Houston, Texas 77004. He
     has extensive experience in the field
     of  injection well  systems  and
     groundwater  protection.  His  current
     activities  include  the study of
     shallow groundwater  contamination
     from industrial hazardous waste areas
     and the development  of plans to
     contain and  remove  the contamination.

         Marvin C. Herring is  a Senior
     Scientist  for  Ken  E.  Davis
     Associates, 3121  San Jacinto,  Suite
     102, Houston, Texas 77004.   For the
     past  5 years he has worked as a
     consultant to industry  in the design
     of new systems for the disposal of
     industrial  wastes  and remedial
     actions for improperly designed waste
     activities.

         Tom  Hosea is  the Laboratory
     Supervisor  for Ken E.  Davis
     Associates, 3121  San Jacinto,  Suite
     102, Houston,  Texas 77004.   He has
     considerable  experience in core flow
     testing design and operation and in
     analytical  chemistry.  His  current
     activities  include  laboratory and
     field  applications  chemistry and
     continued slurry wall research.

was  not funded by  the U.S. Environmental
do not necessarily reflect the views of the
                                   302

-------
                     ADVANCED SECURE LANDFILL DESIGN
                        Randolph W. Rakoczynski,  P.E.
                      Waste Resource Associates,  Inc.
                          Niagara Falls, NY  14305
                                 ABSTRACT

     Geologic and  hydrologic conditions play an important role in determin-
ing the suitability of a particular site for  creating a landfill.  Often-
times,  poor  sub—surface conditions  will  prohibit  use  of  certain  areas
forcing waste generators to transport their residuals great distances for
proper  handling.   The advanced  secure  landfill presented in  this  paper
allows  siting in  less  than optimal  geologic/hydrologic conditions without
adversely impacting the environment.  The  advanced  design relies  for its
basis on many of the design elements Incorporated into many of the conven-
tional  secure landfills being  operated  in the  nation today.   It expands
these conventional designs  by  introducing  innovative concepts such as dual
underdrains and an in—situ leachate treatment system  to insure groundwaters
cannot  become contaminated.   It  provides  an opportunity to  establish  a
landfill operation in a less than optimal setting without adversely impact-
ing environmental  quality.
INTRODUCTION AND PURPOSE

     Landfills  have  been used
historically  for disposing of  a
multitude of  different  types  of
chemical and industrial wastes.   As
the  volume of  waste  production
increased,  many  landfills  became
over-burdened.    The  appearance  of
many  new and  often  highly  toxic
constituents in these  wastes  com-
bined  with  landfill containment
failures due to  poor  design and/or
less  than optimal site  selection
led  to  public  outcry.  This public
outcry  continues  to  grow  as  more
and more locations that are seeping
contaminants into  our groundwaters
are identified.

     The  advanced secure  landfill
design  presented,  is  an attempt  to
stem  this growing tide of public
opposition  to  landfill  disposal  by
providing  a  technologically-defen-
sible  approach  toward  containing
hazardous wastes.   It is  a care-
fully engineered design which
provides various  back—up systems to
insure  that  possible  escape  of
leachate  will  not contaminate
groundwaters.

APPROACH

Conventional  Design  Elements

There  are  certain  conventional
design  elements  which have  been
developed and  are currently util-
ized by some of  the major regional
HWMF's  in the   nation.   These
conventional  design  elements  are
the cornerstone  and  basis  of  the
advanced  design  which was  formu-
lated .

Liners

The use of both natural and  synthe-
                                   303

-------
CO
o
            DRAINAQE SWALE
                                                                                                      c
KEY (USED TO HOLD LINER IN PLACE

  AND FOR CLOSURE 8EALINO WITH

  SYNTHETIC LINER IN CAP)
                        SECONDARY UNDERDRAM
                                        Z2>—'
                                                                FOUNDATION OR BAM

                                                         (IDENTIFIED BY BUB-SURFACE EXPLORATION)
                                                                                                                   MONrroniHO WELL
                                                                                                                               z.
                                                                                                                 <—- WATER TABLE
                                                     ADVANCED SECURE  LANDFILL. DESIGN


                                                            (CROSS-SECTIONAL VIEW)

-------
tic  lining  materials  are incorpor-
ated into this design.  The natural
materials used  are  clay-type  soils
which given proper moisture content
and  compaction  can  be installed  to
provide  a liner  with a. hydraulic
conductivity of 1 x 10   cm/sec  or
less.   In order  to insure proper
compaction  it  is  recommended  that
clay liners  be  installed in  6"
lifts.

     There are a variety of synthe-
tic lining materials placed over the
entire interior  surface of the
landfill  which  can  be  used  to
augment the containment provided  by
a  clay  liner.   Among  the materials
currently available are  PVC  (poly-
vinyl  chloride),  CPE  (chlorinated
polyethylene)  and HDPE (high density
polyethylene).   The types  of  waste
which are intended  to  be   disposed
of and the resistance  of the lining
material  to  the range of  constit-
uents in  those  wastes  will dictate
the best choice of synthetic liner.

     In testing the various synthe-
tic lining materials,  a leachate  of
the  treated  (chemically-fixed)
residues to  be  disposed of, would  be
used.   The  leachate  would be  an
alkaline aqueous solution with trace
levels  of various  organic  contami-
nants .

     The lining system as presented
is a  "laminate"  which uses a  com-
pacted  clay  lining as a base beneath
the synthetic  liner.  Overlying  the
synthetic liner  is  another clay
lining which serves a  two-fold
purpose of  providing   a  protective
cover for the synthetic material  in
addition  to   supplying  further
containment.
 Waste Segregation

     The landfill  cell design  pre-
 sented provides for  waste  segrega-
 tio by  establishing  individual
 subcells  or  "compartments",  within
 the  landfill  itself  for  various
 waste types.   There are five  sub-
 cells in  the design:

    -  amphoteric
    -  heavy metal
    - general
    - halogenated
    - toxic

     The  ability to separate these
generic  waste  types  from  one
another enables  various  cover
materials  to be  utilized to immobi-
lize many of  the  constituents  in
the  waste.   For instance,  in  the
amphoteric subcell, waste carbonate
tailings can be  either admixed  with
the  waste or  used  as intermediate
cover material to create a buffered
pH  environment  between  7  and  8
within the subcell.   At  these  pH
values,  pseudo—metaIs  such  as
arsenic  and  selenium and  metals
such as aluminum will  tend  to
remain insoluble and  less mobile.
Application  of  alkaline  cover
materials  such as  hydrate  lime
tailings  in  the  heavy  metals
subcell  will create an  alkaline
pH environment tending to  keep the
heavy metals  insoluble.

     Separating the  halogenated
organic wastes  from other organics
in the general subcell enhances the
long-term  biodegradability  of
wastes in the  general subcell,  as
does the  segregration  of  toxic
heavy  metal  containing  wastes.
Further,  those halogenated residues
where breakdown  has begun to occur
creating  corrosive conditions
can  be neutralized  with  alkaline
wastes to prevent the formation of
acidic leachate.
                                   305

-------
Toxic Subcell

     This  subcell uses  a double
synthetic liner  system.  The "lami-
nate" ,   clay-synthetic  liner-clay
system,  is overlain with  a second
synthetic liner  of  HDPE.   This
material has  the greatest universal
resistance  to the  halogenated
and/or  corrosive  waste types
anticipated to  be  handled  in  such
a  subcell.   Once  again the lining
of HDPE  is covered with a liner of
compacted  clay  for protection and
additional containment.
Innovative Design Elements

     There are features  of this
advanced   secure  landfill  design
which  are  innovative  and represent
definite  improvements  over  land-
fills  constructed  solely with  the
conventional design elements already
discussed.  Specifically, the use of
a  dual underdrain system and  an
in-situ  leachate   treatment  system
are innovative design  features which
make  this  advanced  secure landfill
superior  to  prior  conventional
designs.
Internal  Leachate Collection System  Double Underdrain System
     The upper  clay lining installed
in  each subcell  provides a  rela-
tively  impervious base.   Any  rain-
fall which accumulates  in  the
landfill  and  becomes  leachate
through contact with waste comes to
rest on this base.    The   internal
leachate  collection  system is  the
mechanism  established  to  withdraw
this leachate for processing before
it  is allowed to percolate signifi-
cantly  down  through  the initial
clay lining. The system is composed
of  a tile underdrain,  covered with
crushed  stone,   that   hydraulically
connects two  vertical monitoring
standpipes.   These  vertical  stand-
pipes  are  2'-3'  diameter concrete
sewer  pipes with  bell-ends  placed
one  atop another as vertical filling
of  the  landfill proceeds.   The
initial  section of  each stand-
pipe is  set in a concrete base and
perforated  around its  circumference
to  allow for  inflow of  leachate.
The perforated portion of  each
standpipe  is covered  with crushed
stone  which  acts as  a  filter to
prevent solids  inflow and  subse-
quent  plugging  of  the standpipe.
As  leachate accumulates  within a
standpipe,  it is removed  by pump or
vacuum truck for processing.
     Although  a  single underdrain
 system has  been  in prior  use in
 various  conventional  secure land-
 fill  designs, this is the first
 time a dual (primary and secondary)
 underdrain  system  has been  pro-
 posed.   The primary underdrain is
 located  below  the  synthetic liner
 and  actually  serves  as the  base
 for the  in-situ leachate treatment
 system.    It serves  as a  device
 to  both  monitor  the  integrity of
 the  synthetic  liner and also
 remove  treated  leachate should  a
 leak in the liner occur.

     The secondary  underdrain is
 located  at  the  bottom of  the
 excavation  beneath  the  compacted
 clay  liner which  is  below  the
 synthetic  liner.   It  allows  the
 integrity of  the clay  liner to be
monitored  while also enabling
 any elevated water table  levels
 to be withdrawn before infiltration
 into the clay  lining  and  possibly
 into  the landfill.   Fluctuating
 and/or seasonally  high  water table
 elevations  can be  managed by using
 the secondary underdrain.
                                  306

-------
 In-situ Leachate  Treatment  System
                                     PROBLEMS ENCOUNTERED
      This is  probably the  most
 unique  feature of  the  design  since
 it  has  not been  used  to  any extent
 in  any  other conventional  designs.

      It  relies on  straight  forward
 chemical/physical   principles
 associated  with  immobilizing  the
 constituents  in .the  waste  types
 disposed of in the various  subcells
 of the  landfill.   Beneath the
 synthetic  liner  in  each  subcell  is
 a fixed  bed of different porous
 media capable of   sequestering  or
 immobilizing  contaminants   associ-
 ated  with  the  leachate from  that

subcell.   For instance, beneath the
amphoteric subcell uniformly crushed
limestone is  placed while  granular
lime is used under  the heavy metals
subcell.   The metallic  cations which
might be present  in   the  leachate
from  the  heavy metals  subcell  are
precipitated  by the alkaline condi-
tions  and enmeshed in the lime.  The
heavy  metal-free  aqueous liquid can
then  be  withdrawn   via  the  primary
underdrain system.

     Beneath  the  halogenated,
general  and  toxic  subcells,  the
fixed  bed  of porous media  is com-
prised of granular activated carbon.
A bed depth of 3"  to  4" is suffic^-
ient  to  provide enough carbon  to
adsorb the  soluble,  high molecular
organic  contaminants in the leachate
in these subcells.   As long as this
activated  carbon  is exposed  to
continually increasing  concen-
trations  of  adsorbable  organics  in
the  leachate which  it contacts
(which should be  the case during any
post-closure  liner   failure/leachate
leakage),  it  will  function  effec-
tively.   Once again treated aqueous
liquid can  be withdrawn using  the
primary  underdrain system.
     In order  to determine  if the
advanced   secure  landfill  design
is  competitive  with conventional
designs,  an economic analysis  of
the cost  of  construction  of  both a
"shallow"  and   "deep"  landfill
for both  the advanced  and conven-
tional designs was undertaken.   The
"shallow"   landfill  has  a  depth  of
22.5 feet while  the  "deep" landfill
has  a depth of  47.5  feet.    An
additional variable,  the  location
of the landfill  with respect to the
existing  grade,   was also considered
in  the comparative  cost   analysis.
In  the situation where  100%  of the
landfill  is located above  grade,
the floor  of the  landfill  cell  is
actually  at  ground level.   In all
cases the size of the landfill  cell
considered was 500' x 500*  measured
along  the  top  of the  perimeter
berms.   This gives a maximum  dis-
posal  capacity of 147,200 cu.yds,.
for the   shallow landfill  and
260,000  cu.yds. for  the.deep.

FixedCosts

     There are  certain cost ele-
ments in both the  conventional and
advanced designs  which are  indepen-
dent of  the  depth ' below the grade
we  choose  to  locate the  landfill.
For  the   "shallow"  and  "deep"
conventional  and  advanced  landfills
these  elements   are  the  synthetic
liner and  compacted  clay  to  cover
the liner:

         Synthetic liner

"shallow"      275,000 sq.ft.
"deep"        325,000 sq.ft.
       Primary clay liner
"shallow"
 "deep"
21,000 cu.yds.
30,000 cu.yds.
                                   307

-------
                            Table 1:  Construction Volumes  ("Shallow" Landfill)

                                              in thousand cubic yards
CO
o
CO



Excavation
Perimeter Berras
Cover Material
Primary Clay Liner
Secondary Clay
Liner
Stockpile
(Shortfall)
Table




Excavation
Perimeter Beras
Cover Material
Primary Clay Liner
Secondary Clay
Liner
Stockpile
(Shortfall)
% Above Grade
0% 50%
Con. Adv. Con. Adv.
150.0 275.0 75.0 244.0
34.0
36.8 36.8
21.0 21.0
35.0 - 35.0

92.2 182.2 (16.8) 116.2

2; Construction Volumes ("Deep" Landfill)
in thousands of cubic yards
% Above Grade
0% 50%
Con. Adv. Con. Adv.
295.0 385.0 108.0 485.0
134.0
65.0 65.0
30.0 30.0
45.0 - 45.0

200.0 245.0 (121.0) 271


100%
Con . Adv .
150.0
120.0
36.8
21.0
35.0

(177.8) (62.8)




100%
Con . Adv .
232.0
508.0
65.0
30.0
45.0

(603.0) (416.0)

-------
Excavation
Perimeter Benns
Cover Material
Primary Clay
Synthetic Liner
Primary Under-
 drain
Secondary Clay
Secondary Under-
 draln

Total Cost:

Unit Cost:
(per cu.yd.)
                  Table 3:  Cost Comparison ("Shallow" Landfill)
                               in thousands of dollars

                                   % Above Grade
0% 50%
Con.
$375.0
—
— ,
$105.0
$165.0
*_«.
—
—
Adv.
$687.5
—
—
$105.0
$165.0
$850.0
$175.0
$65.0
Con.
$187.5
$85.0
$4.0
$105.0
$165.0
_—
_
—
Adv.
$610.0
$85.0
—
$105.0
$165.0 '
$850.0
$17.50
$65.0
100%
Con.
-
$900.00
$184.0
$210.0
$165.0
—
_
—
Adv.
$375.0
$195.0
$184.0
$105.0
$165.0
$850.0
$175.0
$65.0
  $645.0  $2,047.5

   $5.84    $18.55
  $626.5 $2,055.0

   $5.67   $18.61
$1,459.0  $2,114.0

  $13.22    $19.15
Excavation
Perimeter Benns
Cover Material
Primary Clay
Synthetic Liner
Primary Under-
 drain
Secondary Clay
Secondary Under-
 drain

Total Cost:

Unit Cost:
(per cu.yd.)
                  Table 4:  Cost Comparison ("Deep" Landfill)
                               in thousands of dollars
                                   7, Above Grade
        0%
    Con.     Adv.
  $737.5    $962.5
$1,082.5  $2,132.5

   $5.55    $10.94
        50%
    Con.    Adv.
  $270.0 $1,212.5
  $612.5   $335.0
  $325.0     —
$1,552.5 $2,717.5

   $7.96   $13.94
      100%
   Con.     Adv.
           $580.0
$3,810.0 $2,632.5
  $325.0   $325.0
$150.0
$195.0
—
$150.0
$195.0
$550.0
$225.0
$50.0
$150.0
$195.0
—
$150.0
$195.0
$550.0
$225.0
$50.0
$300.0
$195.0
—
$150.0
$195.0
$550.0
$225.0
$50.0
$4,630.0 $4,707.5

  $23.74   $24.14
                                      309

-------
Variable Costs

     There  are  various  site-spec-
ific  factors which will  have a
dramatic impact on the actual costs
associated with the construction of
a  landfill  facility using  either
the conventional or advanced secure
landfill design.

     There  is a distinct possibil-
ity  that  regenerated  activated
carbon which is no longer suitable
for conventional wastewater treat-
ment  applications can  be  used  in
the  in-situ treatment  system,
drastically   reducing  the  overall
capital cost of  the advanced secure
landfill facility.

     The tables  which  follow
related  the  construction  volumes
and  associated  costs  for  the
various scenarios considered.
     It is understood that construc-
tion costs may  vary  widely between
different  geographic  regions and
many  site-specific factors  will
influence  actual  construction
costs. The cost comparison presented
is not  meant  to  establish  an abso-
lute  capital cost estimate for
constructing  a  facility  since
specific costs associated with site
acquisition/  development,  hydro-
geological   investigation,  permitt-
ing,  monitoring and  ancillary
equipment/structures  were  not
included.    The following unit
costs were  used  to  enable a  valid
comparison  to be  drawn  between  the
two  designs  and to evaluate  the
impact the degree below grade would
have on a proposed design.
 Unit Costs

 Excavation (assuming   $2.50/cu.yd.
 a  uniform deposit at
 less than saturated
 conditions with a
 stable sub—base)

 Perimeter Berms:
 -obtain off-site soil  $7.50/cu.yd.
 and compact
 -compact on-site       $2.50/cu.yd.
 soil

 Compacted Clay
 -obtain off-site      $10.00/cu.yd.
 clay and compact
 -compaction on-site    $5.00/cu.yd.
 clay
 Cover Material
 -obtain off-site       $5.00/cu.yd.

 Results

     The  comparative  cost  analysis
 bears  out the  fact for all  cases
 with  the  exception of a "deep"
 landfill  located   entirely  above
 grade,   the  conventional design  is
 considerably less expensive than the
 advanced  design.   Although the
 conventional  design presents  a

 less costly alternative in landfill
 construction,  the costs associated
 with remedial  action  necessary  to
 rectify  groundwater  contamination
 that might occur with  this design
 can be several  multiples  of  the
 total  construction  cost  listed.
 The magnitude  of  these remedial
 action  costs  and  the  long  term
 liability it represents to facility
owners  and  to  the  government  (if
 remedial action is  required during
or  following  the  post closure
 period bears out  the  economic
 feasibility  of the additional
 expense associated with the  use of
 the advanced design.
                                   310

-------
     Additionally  economic  feasi-
bility  for  the  advanced  secure
landfill design can be demonstrated
if  the  distances  from  the  source
of waste generation to a permitted
facility  using  a conventional
design are considered.  If one-way
transportation charges of $3.75 per
mile for a bulk, 20 cu.yd. shipment
of waste are assumed,  the following
table  shows   the  "breakeven"  dis-
tances  in  miles.   If the distance
between source of generation and a
permitted landfill facility employ-
ing  the conventional design are
greater than those  listed,  the
advanced secure landfill design can
be  justified  on  the  basis  of
savings in  transportation charges.

"Breakeven  Point Transportation
 Distances (one-way miles)

          % Above  Grade
            0%    50%    100%
"Shallow"   68    69

"Deep"      29    32
32
             Acknowledgements

                  The advanced  secure landfill
             design presented  is the property of
             Environmental  Design,  Inc.  (Hwy.l,
             Jupiter, FL 33458)  and is protected
             under  U.S. Patent No. 4,430,021.
             Waste  Resource  Associates,  Inc.
             acted  as technical consultant  to
             Environmental  Design,  Inc.  in  the
             development  of   the  design  and
             Randolph W. Rakoczynski is  a  co-
                          the  design  along with
                          Wagner and  Mr. Harold
                           Flannery & Esz, Inc.
                          OH)  have had consider-
             able experience  in  the construction
             of  secure  landfill facilities.
inventor of
Mr. Louis E.
F. Flannery.
(Cincinnati,
     Mr.  Gary W.  Catlin,  Vice
President  of   SLC  Consultants/Con-
structors,  Inc., an individual with
considerable actual field experience
in  secure  landfill  construction,
assisted  in establishing  the  unit
costs used  in  the  cost comparison.

     Currently,  SolidTek  Systems,
Inc. is a licensee of Environmental
Design,  Inc. for the advanced secure
landfill in  the states of Georgia
and Pennsylvania.
                               Disclaimer

                 The work described in this paper was
                 not funded by  the U.S. Environmental
                 Protection Agency.  The contents do
                 not necessarily  reflect the views of
                 the Agency and no official endorse-
                 ment should be inferred.

-------
                  DE-GASIFICATION OF EXISTING LANDFILLS
           Paul C. Rizzo
           Paul C. Rizzo Associates, Inc
           Post Office Box 17180
           Pittsburgh, PA  15235

                                ABSTRACT
      Carl M. Rizzo
      R & R Petroleum, Inc.
      206 Rodi Road
      Pittsburgh, PA  15235
    De-gasification programs at waste landfills are undertaken for two
purposes; resource recovery and landfill safety.  This paper will
describe a specific methodology to address safety problems posed by
pressurized toxic gases at an uncontrolled waste disposal site.  A
conceptual design for a system to both investigate and relieve
pressurized, and possibly toxic, gas problems is presented.  The
hypothetical case is for a landfill located in a populated area in which
"zero release" of gases Is a project constraint.
INTRODUCTION AND PURPOSE

    This paper deals with a
subset of sites that for one
reason or another have (a) a
combination of chemical con-
stituents capable of producing
toxic and/or volatile gases
such as hydrogen sulfide,
methane, chlorine, etc.; and
(b) a geometrical config-
uration that would allow for
the build up of pressure.
Buried drums are the first
such configuration that might
come to mind.  Other typical
geometries are illustrated in
Figures 1 and 2 where a "hard"
crust forms a buried cap over
a. zone of waste materials
capable of producing a gas.
The crust, being relatively
Impermeable and of relatively
high shear strength, allows
for the build up of gas
pressure.  The gases may build
up in pockets or in a
relatively uniform zone,
depending on the homogeneity
and thickness of the crust,
and the nature and quantity of
the gas-producing waste
material.
    Upon penetration of a
buried drum or a crust by a
conventional geotechnical
drilling rig, or by the unwary
construction equipment
operator during cleanup of the
waste, the gas can be released
in an uncontrolled manner,
thus endangering the drillers,
construction equipment
operators, and possibly the
public as well.  Of course, if
the concentrations are high
enough and the pressure and/or
quantity is high enough, the
problem could have relatively
serious implications.

    In the oil and gas
industry, the handling of
unknown quantities of natural
gas and occasionally hydrogen
sulfide are practical problems
encountered in the normal
course of business and
operations.  These gas "kicks"
are handled routinely, and in
the case of hydrogen sulfide,
there are established safety
protocols for dealing with the
                                    312

-------
toxic nature of the gas.  This
industry has built up a wealth
of experience, equipment,
monitoring systems and safety
procedures.  It Is the purpose
of this paper to illustrate
how the experience and back-
ground of the oil and gas
industry can be used in
conjunction with conventional
geotechnical Investigation
techniques to Investigate,
relieve, or vent toxic gases
from a shallow landfill In a
safe controlled manner—either
on an emergency basis or as
part of a remedial action
program.

APPROACH

    The installation of toxic
gas venting wells at a site
such as shown In Figures 1
and/or 2, requires a
combination of highly
specialized skills that are
not easily combined.  For
example, the depth is rather
shallow, generally less than
50 feetj this is quite
suitable for a typical
geotechnical drilling rig.
However, the presence of a
toxic or volatile gas such as
chlorine, hydrogen sulftde or
methane at unknown pressure
and concentration, requires
the use of oil field tech-
niques including extra heavy
mud, possibly a blowout
preventer (BOP), some type of
mud-gas separator, shaker, and
a degasser such as shown In
Figure 3.  The degasser
removes the gas from the
drilling mud and allows for
transmission to a treatment
unit such as a scrubber, or in
limited cases to a flaring
system.  The extra heavy mud
and possible BOP are both
required to assure a margin of
safety consistent with the
margins used In other parts of
an EPA-sponsored cleanup.  The
heavy mud and the BOP require
an oil-field type rig because
of the need to use larger
diameter drill pipe and
casing, and heavy duty mud
pumps.  BOP's are not always
readily available for small
diameter drill pipe; and
furthermore, the use of extra
heavy mud (possibly 20 pounds
per gallon) requires larger
diameter drill pipe and
adequate pump capacity.  These
are key factors in the
technique which may preclude
the use of a conventional
geotechnical drilling rig.

    The handling of the toxic
gas requires a closed-mud
system, also common to an oil-
field driller but not to the
geotechnical driller.  As
mentioned above, the system
might Include a shaker,
possibly a mud/gas separator,
and a degasser to allow for
removal of the gas and
transmission to some sort of
treatment unit such as a
scrubber.  Mud tanks are used
for mixing and handling the
mud as opposed to mud pits.
Here Is an example where the
experience of the geotechnical
driller comes Into play.  The
oil and gas driller Is used to
handling large quantities of
mud, and consequently, the use
of mud pits Is the common
practice.  Shallow drilling
for waste sites requires
relatively small quantities of
mud; and, therefore, mud
tanks—the type that a
geotechnical driller might use
In a confined environment are
more appropriate.
                                   313

-------
    It is also noted in Figure 3
that we show an emergency
"kill" line similar to that
used for oil and gas drilling.
This line allows for the
addition of drilling mud under
higher pressure should a fire
or an unusual gas"kick" occur
during drilling or development
of the well.  We also show the
mud pump as being separated
from the drill rig; however,
the mud pump may very well be
part of the drill rig in many
applications.

TYPICAL INSTALLATION SEQUENCE

    Generally speaking, the
operation begins with the
mobilization of a crew and
equipment to the site.  The
field crew normally would have
a project safety review before
leaving the yard, including
physical examinations and a
review of the safety equipment
contemplated for use at the
site.  The crew normally con-
sists of five experienced
members (superintendent, tool
pusher/driller, driller/helper,
power swivel operator, and
well-site health and safety
officer).  Me mention a power
swivel operator as being a
member of the crew because an
oil well service rig equipped
with a power swivel as opposed
to a rotary mud rig is often
the preferred type of equip-
ment for this application.
The crew is usually programmed
to work an extended single
shift as work during daylight
hours is preferred; however,
in many applications, round-
the-clock operations are
necessary.  This is the norm
in the oil and gas industry
and does not present unusual
operating problems.
    Along with the service rig
and power swivel, a mud-mixing
system including a shaker and
degasser are mobilized to the
site.  Material such as
casing, drilling mud, cement,
and so forth are ordered in
quantities sufficient to
complete a group of wells with
the mud coming from a supplier
retained to specify the mud
program including mud-testing
procedures.  Cementing is
generally delegated to a
specialty oil field cementing
service company as this
requires special equipment not
normally operated by the
drilling crew.

    The detailed technique for
the venting well installation
will vary from site to site,
depending on particular
conditions.  The following
steps would apply to a site
having a geometry as shown in
Figures 1 and 2 where the
crust might be about 20 feet
down from the surface and
about two feet thick;

1.  Construct an earth-
    drilling pad to support
    the drilling rig.
    Generally speaking, the
    surface will be relatively
    soft, too soft to support
    the drilling rig; and
    therefore, a drilling pad
    will have to be
    constructed with earth-
    moving equipment.  In
    Figure 4 we show the use
    of geotextile material
    between the surface of the
    landfill and the earth-
    drilling pad.

2.  Rotary drill an open hole
    with heavy mud to a point
    within about 4 feet of the
                                  314

-------
crust.  Install steel
surface casing and cement
(the first stage) to
surface using a specialty
cementing service sub-
contractor.  Casing Is
sized to permit the
circulation of extra heavy
mud, which is critical in
subsequent steps in main-
taining a closed system to
mitigate random or unpre-
dicted gas releases
(Figure 4).

Connect the drilling tree
(with blowout preventer,
degasser and drill line
connections) to casing and
re-insert rotary drill
tools (Figure 5).

Advance the hole using the
rotary techniques through
crust to a point about 4
feet below "crust."  Use
extra heavy mud and a mud
program designed by the
supplier.  Circulate the
mud through the shaker and
degasser which is in turn
connected to the scrubber
(Figure 6).  It is noted
that during the drilling
process the circulatory
mud might bring waste
liquids, such as organic
solvents, from "old" drums
to the surface.

If no gas kick occurs or
after a. kick is fully
relieved, trip out the
drilling tools and install
production casing with
closure valve through the
surface casing.  The
bottom four feet of the
production casing will be
slotted or screened.
Using a combination of
external and internal
    packers such as a for-
    mation packer collar
    and/or baffle plates set
    in at the crust,.remove
    the mud above the packers
    and cement (second stage)
    the production casing to
    the surface (Figure 7).

6.  Re-Install a drilling tree
    over main closure valve.
    Trip in the small diameter
    drilling tools.  Refill
    production casing with
    mud.  Drill out the
    internal packer or knock
    out baffle plates.

7.  Dilute the mud mix by
    washing through tools
    until mud is practically
    water.  "Develop" the well
    to extent practical.

8.  Trip out the drilling
    tools through BOP and
    stripper.  Close the main
    valve on the production
    casing.  Remove the
    drilling tree and install
    production tree (Figure
    8).

    At this point, the well is
available for venting of the
gas and its subsequent
treatment at a scrubber or
treatment unit.

HEALTH AND SAFETY

    The members of the
drilling crew must be
thoroughly knowledgeable of
the personal protection
procedures dictated by the oil
and gas industry and hazardous
waste emergency response
health and safety programs.
Normally this type of project
will employ EPA "Level B"
protection with some minor
                              315

-------
modifications  to suit  field
conditions.  For the example
shown in this  paper where  the
toxic gas is generated in  the
waste beneath  the crust, the
Level B protection would be
employed when  the drill tools
are re-inserted into the
cemented surface casing.
Level B is not necessary for
work above the crust because
any gas encountered above  the
crust would be relatively  low
pressure and would be counter-
acted by the heavy mud with
adequate margin.

    Monitoring systems are set
up around the  rig and around
the site fence line.  For  a
hydrogen sulfide venting
project, we would employ a
battery-operated monitoring
system sensitive to one PPM.
This would be operated during
drilling operations after  the
surface casing is installed if
we are reasonably sure that
the gas content of the waste
material above the crust is
minimal.  Alternately, one
might utilize an organic
monitor and/or an explosive
meter, or other appropriate
real-time monitoring
equipment.

SUMMARY

    Remedial action programs
and emergency response pro-
grams occasionally dictate the
need for the relief, handling
and treatment of pressurized
toxic or volatile gases in
relatively shallow lagoons
and/or landfills.  This paper
presents a technique which
combines the knowledge and
experience of the oil and gas
industry with  that of the
conventional shallow geo-
technical driller.  This
procedure results in an
effective and safe means of
installing relief wells with-
out undue risk to drillers and
the general public.  The key
factors in the operation
include the use of an extra
heavy drilling mud, a closed
mud system, and the possible
use of a blowout preventer.
The techniques used herein are
not new, and in some cases,
would be viewed as relatively
conservative applications of
equipment that are used for
much more dangerous situations
such as those encountered in
oil and gas drilling.  The
application in such shallow
deposits as encountered in
lagoons and landfills, how-
ever, does require some mod-
ification because of the shal-
low nature of the deposits.

ABOUT THE AUTHORS

Dr. Paul C. Rizzo
    Dr. Rizzo is founder and
President of Paul C. Rizzo
Associates, Inc., a consulting
organization based in
Pittsburgh that works inter-
nationally on the solution and
cleanup of waste problems—
both nuclear and chemical
hazardous waste.  Dr. Rizzo
received his Doctorate from
Carnegie Mellon University,
has published numerous papers,
and has served on a number of
international panels and
committees dealing with
environmental issues.

Mr. Carl M. Rizzo
    Mr. Rizzo is co-founder
and President of R & R
Petroleum, Inc., a Pittsburgh
based independent oil and gas
                                  316

-------
producer working in the
Appalachian Basin of the
United States.  Mr. Rizzo has
designed, supervised, and
operated oil and gas drilling
programs and production
systems for wells as deep as
5,000 feet—the normal depth
in Appalachia.  Mr. Rizzo
holds a Masters degree from
the University of Pittsburgh
and has worked both onshore
and offshore on shallow
drilling projects.
                               Disclaimer

                 The  work  described  1n  this  paper was
                 not  funded by  the U.S.  Environmental
                 Protection Agency.   The contents do
                 not  necessarily  reflect the views of
                 the  Agency and no official  endorse-
                 ment should be Inferred.
                                   317

-------
                                      -cxsiui. -Pipe.
                   IPO*; /
                                      X
                                                          T*s.o-
                                                  \\\\
318

-------
4'
                        (61.1. tlJt-
                \~\\\\V\A\ \ \
          FIGUR6-C.
L
                                                                ou rr
                                                                                       K.IUL uae.
                                                        T=teocuet"'»J
                                                                       1
                                                                               /"-
                                                319

-------
                FACTORS AFFECTING STABILIZATION/SOLIDIFICATION
                              OF HAZARDOUS WASTE
                        Jerry N. Jones, R. Mark Bricka,
                    Tommy E, Myers, and Douglas W. Thompson
               U. S. Army Engineer Waterways Experiment Station
                           Vicksburg, MS  39180-0631

                                   ABSTRACT

     The stabilization/solidification of toxic wastes involves a series of
chemical and/or physical treatment procedures.  The waste is normally treated
so as to complex or bind the toxic elements in a stable, insoluble form or to
entrap the waste material in a crystalline matrix.  A hazardous waste may con-
tain many constituents that could interfere with the binding process.  This
recently initiated project is concerned with identifying possible interfering
mechanisms between particular compounds and waste binding systems.

     A synthetic sludge was produced containing parts per million concentra-
tions of cadmium, chromium, mercury and nickel.  The sludge is mixed with
increasing concentrations of interfering compounds and then with selected
waste binders.  Various waste mixtures will be evaluated to determine the par-
ticular interfering effects on stabilization/ solidification processes.  This
paper provides an interim report on the progress of the study.
INTRODUCTION

     The Environmental Protection
Agency (EPA) is responsible for eval-
uating the suitability of hazardous
waste for land disposal and for the
examination of hazardous waste de-
listing petitions under the Resource
Conservation and Recovery Act (RCRA)
regulations.  A thorough understand-
ing of the potential behavior of
stabilized/solidified waste is neces-
sary to make judgements as to the
effectiveness of contaminant contain-
ment.  There are several methods
available for the stabilization/
solidification (S/S) of many hazard-
ous wastes.  The complexity of some
wastes is such that some of the chem-
ical components of the waste may in-
terfere with the proposed S/S process
and cause an undesired phenomenon
(e.g. flash set, set retardation,
spalling, etc.).  There is a lack of
quantitative data concerning the
effects of these interfering compo-
nents upon the particular S/S pro-
cess.  A study of the factors
affecting S/S of hazardous waste is
designed to help fill that data void.

Background

     Experience from the cement and
asphalt industries, as well as radi-
oactive waste solidification, has
demonstrated that small amounts of
some compounds can seriously reduce
the strength and containment charac-
teristics of binder/waste mixes used
in S/S technologies.  The common
binding materials in waste S/S sys-
tems are derived from industrial
cement and asphalt materials and some
data are available on the effects of
impurities on strength, durability,
and permeability of structural cement
mixtures.  The cement industries have
had to specify the types of cement,
aggregate, and accelerators or re-
tarders that will be permitted in
Portland concrete specifically
because additives affect performance.
Radioactive waste processors have had
to develop limits for particular com-
ponents that reduce the effectiveness
of S/S of their wastes.  In some
                                      320

-------
cases these interfering materials can
produce set retardation so that hard-
ening does not occur.  In other
cases, a waste constituent may cause
a flash set or flash hardening so
that effective mixing of the waste
and binder cannot occur.  Some waste
constituents can react to cause
swelling or disintegration of the
solidified mass after setting.  The
effects of these interfering materi-
als are often disproportionate to the
amounts present in the waste.

     Research that clarifies the re-
lationship between the composition of
the waste and performance of the
binder can greatly help in evaluating
the hazards of the materials.  For
example, minor organic components in
a waste stream can significantly
retard the cementing reaction, but
these specific materials can be lim-
ited in a processor's treatment per-
mit to assure that binding reactions
occur, and that waste containment is
not compromised.  Similarly other
deleterious compounds that cause
flash sets or expansion reactions can
be excluded where reactive binders
are proposed.

     When waste is placed for final
disposal, there is always a possibil-
ity of the waste being unearthed by
erosion or later excavation.  The
impact of stabilized/solidified mate-
rial on its surroundings after expo-
sure can be significant.  The effect
that interfering substances have on
the durability of stabilized/solidi-
fied waste is a question of paramount
importance.

PURPOSE

     The purpose of this study is to
develop technical background data on
the compatibility of critical waste
constituents with various waste bind-
ing agents (portland cement, portland
cement with additives, lime-pozzolan
cement, gypsum cement).  Specific
supporting objectives are as follows:
     (1)  Perform physical tests and
chemical leaching tests on waste/
binder mixtures.

     (2)  Investigate the relation-
ship between the amount of interfer-
ing substance and the significance of
the interference to waste S/S
reactions.

     (3)  Evaluate the durability of
specific waste/binder mixtures to
long-term outdoor weathering.

     (4)  Participate in the joint
Alberta Environmental Center/Envi-
ronment Canada/USEPA/Industry project
to investigate test methods for use
in evaluating solid wastes.

This paper will report on the results
and progress made as of May 1985.

APPROACH

     The study is being conducted in
two phases (Figure 1).  The first >
phase (Phase I) was completed in
December 1984 and included the fol-
lowing tasks:

     (1)  Conduct a literature search
and report on potential factors af-
fecting solidification processing of
hazardous industrial and radioactive
wastes.

     (2)  Selection of four waste/
binder systems and ten potential
interference materials.

     (3)  Formulation and processing
of stock synthetic waste to be used
in preparation of test specimens.

     (4)  Screening for various
waste—to—binder ratios.

     Phase II of the study will be
conducted over a 24-36 month period.
Samples of the selected waste/binder
system and interfering materials will
be tested and evaluated by the U. S.
Army Engineer Waterways Experiment
                                      321

-------
                                        (PHASE I)
                    REVIEW LITERATURE •
                                                 SELECT WASTE/BINDER SYSTEMS
                    FORMULATE STOCK
                    SYNTHETIC WASTE
                                                 SELECT INTERFERENCES
                               SCREEN FOR WASTE-TO-BINDER RATIOS
                                       (PHASE II)
                       PREPARE SAMPLES
                    FOR TESTING AND EVALUATION
                    WES TESTING PROGRAM

                        LABORATORY TESTING

                        FIELD EXPOSURE TESTING
      SPLITS TO ENVIRONMENT
  CANADA FOR COOPERATIVE TESTING
                    REPORTING OF RESULTS
                      Figure  1.   Simplified Study Approach,
Station  (WES).   Samples of stabi-
lized/solidified waste will be  tested
for various  physical properties and
chemical  leaching quality.

PROBLEMS  ENCOUNTERED

     The  first  experimental diffi-
culty encountered in this study was
the requirement to produce a  syn-
thetic waste sludge that was  both
reproducible and had stable physical
and chemical characteristics.   A
hydroxide metal sludge filter cake
has been  formulated that meets  this
study requirement.  The cake  is pre-
pared by  taking a solution of metal
salts, precipitating the metals with
lime, and then  vacuum filtering to
produce a consistent filter cake of
approximately 25 percent solid
content.
RESULTS

Literature  Survey

     A review of the literature  on
solidification processing of waste
showed that published information is
limited which specifically relates to
organic and inorganic interference
phenomena in waste binding processes.
Since most  waste S/S systems incor-
porate various cement configurations,
some inferences may be made between
admixtures  in cement chemistry and
certain Interferences in S/S
processes.

Additives In Portland CementConcrete

     In the production of portland
cement concrete, the use of chemical
additives to control setting times,
to reduce water requirements,  to en-
train air,  and to create many  other
beneficial  effects is common prac-
tice.  The  changes in properties
                                       322

-------
affected by additives are assumed to
reflect fundamental changes in the
hydration products, especially the
crystalline calcium silicate hydrate.
The chemical and physical development
of concrete is complicated.  The in-
terfering effects of waste constitu-
ents on setting reactions compound
the complicated nature of cement
chemistry.

     The influence of a limited num-
ber of organic compounds on the mic-
rostructure and engineering proper-
ties of hydrated cement has been
investigated (Ramachandran, 1971).
These compounds include sugars, lig-
nosulfates, carboxylic acids, tri-
ethanolamine, and others.  One inor-
ganic substance, calcium chloride, an
inexpensive and effective accelera-
tor, has been extensively studied
(Ramachandran, 1971).  Calcium chlo-
ride accelerates hydration of trical-
cium silicate.  Morphological altera-
tion of the hydrated calcium silicate
gel and calcium hydroxide, along with
changes in CaO/SiO  and HO/ SiO
ratios, surface area, and pore-size
distributions have been documented
(Ramachandran, 1971; Odler, 1971; and
Collepardi, 1972).

Interfering Mechanisms

     In this section attention is
given to conceptual models of inter-
fering mechanisms.  These models
include interference via adsorption,
complexation, precipitation, and
nucleation,

Adsorption

     One possible interfering mecha-
nism is adsorption of additive mole-
cules by crystalline particles
thereby blocking the normal hydration
reactions.  Studies (Young, 1970)
have shown that the retarding effect
of organic compounds are related to
their adsorption on metastable hexa-
gonal calcium aluminate hydrates.
The organic compound inhibits crystal
growth and conversion to calcium
aluminate hydrates.  The inhibiting
effect roughly correlates with the
number of hydroxyl, carboxylic, and
carbonyl groups in the organic mole-
cule.  Hansen (1952 and 1959) noted
the effect of two particular families
of organic compounds, lignosulfonic
acid derivatives and hydroxylated
carboxylic acids, on setting reac-
tions.  Lignosulfonates are strongly
adsorbed onto tricalcium aluminate
(Blank, et al., 1963).  The adsorp-
tion of calcium lignosulfonate onto
tricalcium aluminate results in a
relatively thick film or layer.  The
strong adsorption and thick layering
of lignosulfonates onto tricalcium
aluminate is indicative of a chemical
reaction involving the organic and
tricalcium aluminate hydration pro-
duct.  Taplin (1962) found retarding
effects from aliphatic and aromatic
dicarboxylic acids (e.g. maleic
acid).  In alkaline solutions where
maleic acid has no hydroxyl group for
a hydrogen adsorption bond, chela,tion
may be the interference mechanism.

     Although adsorption of organic
retarders is primarily on tricalcium
aluminate, retardation is due to ad-
sorption on tricalcium silicate.
There is no evidence of adsorption
onto anhydrous surfaces.  Organic
additives can have an important bear-
ing on reaction rates during cement
hydration.

Complexation

     Taplin (1962) related the re-
tarding activity of organic compounds
to the proximity of oxygen atoms to
carbon atoms.  He observed that the
compounds with oxy-functional groups,
in close proximity to each other,
were also indicative of retarder ef-
fectiveness, and that chelation to
metal ions could be an important fac-
tor in set retardation.  Calcium ions
can chelate with various hydroxyl or
carboxylic acids, but the retarder or
accelerators (respectively) are so
                                      323

-------
dilute that the complexation of the
calcium is not an important factor
(Young, 1972).

     The effects of complexing cal-
cium are more significant when the
additive to affected-ion ratio is
very large and when the affected ion
is important to the setting system.
Such would be the case for the alu—
minate and ferrite ions.  Researchers
(Kalousek, 1943, and Roberts, 1967)
have shown that the addition of su-
crose increases the concentration of
alumina and calcium ions to above
normal levels.  Experiments (Young,
1972) with tricalcium aluminate indi-
cate that 1 percent by weight addi-
tions of sucrose, succinic acid and
tartaric acid increase the amounts of
calcium and alumina in solution at
first, but concentrations later de-
crease to normal or below normal.
Silica concentrations are also in-
creased when additives that affect
alumina concentrations are used.
Apparently, conditions in a cement
paste are favorable to aluminate,
ferrite and silicate ion complexa-
tion.  It is possible that complexa-
tion delays the formation of hydra-
tion products.  When cement forming
ions are kept in solution by complex-
ation, hydration barriers are estab-
lished that retard the set.  Waste
components that chelate or complex
toxic constituents may also acceler-
ate their leaching even if the waste
is successfully solidified.

Precipitation

     The formation of insoluble pre-
cipitates by additives reacting with
cement compounds is conceptually not
a realistic mechanism of admixture
interference.  Certainly the forma-
tion of insoluble compounds could
impede water transport, solubility,
and subsequent hydration reactions.
However, if retardation is due to
precipitation, then the process
should be non-selective.  In this
case, tricalcium silicate and
alurainate would both release calcium
ions and the resulting effect on set-
ting should be equally weighed among
both compounds.

Nucleation

     The inhibition of nucleation of
crystalline calcium hydroxide by sol-
uble silica is believed to be the
self-retarding set feature of trical-
cium silicate hydration.  Growth of a
crystalline matrix is retarded by the
adsorbed silica ions when a hydrated
calcium silicate gel layer results in
a diffusion barrier to calcium hy-
droxide.  Eventually crystal growth
results in the adsorbed silica being
trapped in the crystalline matrix as
the hydration process continues.
Prismatic growth of calcium hydroxide
results from differential adsorption
of silica on calcium hydroxide crys-
tal faces (Young, 1972).

     It is postulated that organic
retarders act much the same as silica
ions being adsorbed onto the calcium
hydroxide nuclei.  However because of
more retarders being'solubilized, the
organic retarders are much more ef-
fective in being adsorbed and more
completely cover crystal growth sur-
faces.  The resulting retardation of
crystal growth causes the formation
of more crystallite nuclei in the
saturated solution.  The net effect
of crystal growth on this many nuclei
is acceleration of tricalcium sili-
cate hydration following the initial
retardation period (Young, 1972).  A
one percent (by weight) of a strong
retarding agent, completely inhibits
tricalcium silicate hydration.  Addi-
tion of prehydrated tricalcium sili-
cate has been found to overcome the
effect of the organic retarder indi-
cating the merit of an adsorption/
nucleation model.
                                     324

-------
Effects of Organic Compounds cm Phys-
ical Properties

Accelerators

     Any compound that promotes the
nucleation and growth of calcium
hydroxide will be an effective accel-
erator.  The most commonly used or-
ganic accelerator of hydraulic cement
is triethanolamine (TEA).  At addi-
tion rates less than 0.06 percent by
weight of cement, TEA is an accelera-
tor.  At greater dosages, TEA is a
retarder.  Calcium formate is another
common accelerator.  The addition of
calcium chloride to the cement mix-
ture accelerates crystal growth by
eliminating the necessity for ions to
move from the Tricalcium silicate
structure into solution.  Both TEA
and calcium formate have been shown
to improve the 28 day compressive
strength of Portland cement concrete.
Formaldehyde and paraformaldehyde de-
crease the 28 day compressive
strength at higher dosage rates
(Rosskopf, et al., 1975).

Retarders

     Early strengths are lower in
concrete specimens with retarders
than those without.  However, as the
age of the specimens increase, those
containing retarders generally have
higher compressive strength and com-
parable flexural strength.  The
drying-shrinkages of specimens con-
taining retarders are also comparable
to those without retarders.  It is
interesting to note that soluble cal-
cium salts which provide anions that
adsorb onto the calcium hydroxide
crystal surfaces cause a retarding
effect (e.g. calcium nitrate).

Waste Binding Systems and Interfer-
ences Selected for Study

     Portland cement, lime-pozzolan,
and gypsum cement have been used in
conjunction with other additives
(e.g. flyash) to stabilize/solidify
industrial waste sludge.  Lime-poz-
zolan mixtures are frequently used to
convert liquids and semi-solids to a
solid form.  Systems selected for
testing are those found in use in
waste treatment and judged to have
potential widespread application.
Type I Portland cement and Type I
Portland cement with flyash were
chosen for immediate investigation.
A lime-pozzolanie cement and a gypsum
cement formulation will be used in
subsequent experimentation.

     Ten (10) materials with proper-
ties known or suspected to interfere
with the S/S process were chosen for
study.  The frequency and concentra-
tion with which these materials are
commonly found in waste streams con-
sidered for S/S processing were also
considered.  The interfering materi-
als selected for study are:  (1) oil
and grease| (2) light weight oil;
(3) phenol; (4) sulfates; (5) strong
base; (6) pesticide; (7) degreaserj
(8) lead; (9) copper; and (10) zinc.

Synthetic Waste Sludge Deyelopment

     The synthetic sludge to be used
in this study was generated by treat-
ing a synthetic wastewater containing
cadmium, chromium, mercury, and
nickel with hydrated lime to yield a
hydroxide sludge containing these
metals in concentrations at least
100 times that necessary to cause
rejection on the basis of the EP Tox-
icity Test (eg. 100 ppm Cd, 500 ppm
Cr and Ni, 20 ppm Hg).  Synthetic
wastewater was prepared in 500-gallon
batches using the nitrate salts of
the above metals.  After allowing
sufficient mixing time for equilibra-
tion, hydrated lime was added in a
sufficient quantity to produce a pH
of at least 10.  Following a period
of rapid mixing and slow mixing to
allow for floe formation, the mixture
was maintained under quiescent condi-
tions for a minimum period of 24 hrs
to provide for settling and sludge
accumulation.  Clear water above the
                                     325

-------
   TABLE 1.   REQUIRED SAMPLES  FOR STABILIZATION/SOLIDIFICATION PROJECT-FY85


   FroctiiMt Foctl*ed ecmvnt and pott land ceE«nt/f lyaih
   In£*rf«ift|t Afuscit Oil mud Crease, oil, ph«nol, sulfate, caustic* pesticida, degreaser, lend) copper, and zinc

test
» Unfit
Contt Indtx
(tell)
daralty
Wst/Bry
(IS cycles)
(dual sxcept
(or control)
X-ray mat SOt
(sins.!* except
for control.

Test Period
24 ? 28
hr day day
XXX
XXX
3t
X
X

test Period
24 7 28
Speciswns hr dsy day Specimens
1 X 2
6 XXX 6
(2 x 3)
9 XXX 9
(3 x 3)
2 X 2
1 X 2
0 or 1 X 1

24 7 28
hr dsy day Specimens
X J
XXX 6
XXX 9
X 2
X 2
X I

24 7 28
hr day day Specimens
X 2
XXX 6
XXX 9
X 2
X 2

Total Required
Specimens Sarcples
8 0*
24 24
36 36
8 0*
7 7
(4 in. dian)
3 or 4 0*
   third control)
                                                                       (2.8 in, dlas)
    Tile vtfghe/dansit? measurement will be obtained using the 28-day Cone Index (CX) saaples. After the CI is obtained, the samples Hill be
    broktn iaco pieces and subsavples will be taken £or the EP and X-r&y/SEM tests.
sludge was  decanted and the  sludge
transferred to a holding tank In
preparation for dewatering.   Dewater-
Ing to approximately 25% solids  was
achieved  using a drum vacuum filter.
Approximately 150 pounds of  sludge is
produced  in each batch.  Fresh sludge
will be prepared on a regular basis
so that all sludge used in the inter-
ference portion of the study will
have a maximum age of seven  days.

Work to be  Conducted in FY 85

     The  initial laboratory  work to
be conducted in FY 85 will involve
development of the required  amounts
for the S/S additives used in the
Portland  cement and portland cement/
flyash processes.  These amounts will
be established by preparing  serial
batches of  solidified sludge, allow-
ing representative samples to cure
for 28 days, and determining the un-
confined  compressive strength (UCS)
of each sample.  A baseline  dosage
rate will be determined.  The base-
line dosage is the minimum dosage
required  to  produce a UCS of  100  psi
using a minimum amount of additives.
Work will also be conducted on  the
installation and calibration  of
equipment to be used later in the
study, documentation of sample  prep-
aration and  testing procedures, and
the development and verification
techniques for preparing and  adding
interfering  agents to the synthetic
waste sludge.

     Work on preparation of solidi-
fied specimens of sludge containing
the interfering agents is scheduled
to begin  in  January 1985.  Work will
be conducted on one S/S process at a
time.  A  list of the samples  to be
prepared  and corresponding testing to
be conducted is presented in  Table 1.
As indicated, aliquots of four
sludge/interfering agent mixtures (0,
2, 5, and 8  percent interfering agent
by weight) will be solidified and
test samples prepared.  All four
aliquots  will be obtained from  the
same batch of sludge.  Testing  will
include EP,  cone index, UCS,
                                       326

-------
weight/density, wet/dry durability,
X-ray and scanning electron micro-
scopy, and permeability.  Cone index
and UCS testing will be conducted on
samples aged for 24 hrs, 7, and
28 days.  Other testing will be con-
ducted on 28-day samples.  Sample
preparation for each S/S process will
require approximately 20 weeks.  Data
evaluation will be conducted on a.
continuing basis as test results
become available.

ACKNOWLEDGEMENTS

     This study is part of a major
research program that is now being
conducted by the U. S. Army Engineer
Waterways Experiment Station and
funded by the Environmental Protec-
tion Agency, Municipal Environmental
Research Laboratory, Solid and Haz-
ardous Waste Research Division, Cin-
cinnati, Ohio, under Interagency
Agreement DW96930146-01.  Carlton C.
Wiles is the EPA Project Officer for
this research area.

REFERENCES

Blank, B., D. R. Rossington, and
     L. A. Weinland, 1963.  "Absorp-
     tion of Admixtures on Portland
     Cement, American CeramicsSoci-
     ety, Vol. 46, p. 395.

Collepardi, M., and B. Morchese,
     1972.  "Morphology and Surface
     Properties of rtydrated Trical-
     cium Silicate Pastes," Cement
     and Concrete Research, Vol. 2,
     p. 57.

Hansen, W. C., 1952.  Proc. Third
     Intern. Symp. Chemistry of
     Cements, Cement and Concrete
     Assoc., London, pp 598,

Hansen, W. C.  1959, Symposium of
     effect of Water-Reducing Admix-
     tures and Set-Retarding Admix-
     tures on Properties on Concrete,
     Amer. Soc. Test, Mat. Spec. Pub.
     266, p. 3.
Kalousek, G. L., Jumper, C. H. and
     Tregoning, J. J., 1943.  "Com-
     position and Physical Properties
     of Aqueous Extracts from Port-
     land Cement Clinker Pastes Con-
     taining Added Materials," Jour.
     Res. Natl. Bur. Stand., V. 30,
     p. 215.

Odler, L., and Skalny, J., 1971.
     Jour, of Amer. Ceram. Soc.,
     V. 54, p. 362.

Ramachandran, V. S., Mater. Struct.
     (R1LEM), 4, 3 (1971).

Roberts, M. H., 1967.  "Effect of
     Admixtures on the Composition of
     the Liquid Phase and the Early
     Hydration Reactions in Portland
     Cement Pastes, RILEM Symposium
     on Admixtures for Mortar and
     Concrete, Brussels.
     Station (England) _
     61/68.
 Bldg.  Res.
	 Paper
Rosskopf, P. A., Linton, F. J., and
     Peppier, R. B., 1975.  "Effect
     of Various Accelerating Chemical
     Admixtures on Setting and
     Strength Development of Con-
     crete," Journal of Testing and
     Evaluation, Vol. 3, No. 4^
     pp. 322-330.

Suzuki, S., and Nlshi, S., 1959.
     Semento Gijutsu Nenpo., V. 13,
     p. 160.

Taplin, J. H., 1962.  Discussion of
     paper by H. E. Vivian, Fourth
     Intern. Symp. Chemistry of Ce-
     ments , Wash. D, C., U. S.
     National Bureau of Stds. Monogr.
     43, Vol II, p. 924.

Young, J. F., 1970.  J. Amer. Ceram.
     Soc. V. 53, p. 65.

Young, J. F., 1972.  A Review of the
     Mechanisms of Set-Retardation in
     Portland Cement Pastes Contain-
     ing Organic Admixtures, Cement
     Concr. Res., Vol. 2, pp. 415-433.
                                     327

-------
        A PROCEDURE FOR CHARACTERIZING INTERACTIONS OF ORGANICS
  WITH CEMENT:  EFFECTS OF ORGANICS ON SOLIDIFICATION/STABILIZATION

                  M, E. Tittlebaum, F. K. Cartledge,
                  D. Chalasani, H. Eaton and M. Walsh
                 The Hazardous Waste Research Center,
            The Colleges of Engineering and Basic Sciences
                      Louisiana State University
                        Baton Rouge, LA 70803
                               ABSTRACT

     The effects  of  organics on the setting  reactions  of hydrating Port-
land cement  and  on  the eventual  structure  of the  hydrated cement paste
have been  studied by  a combination of  techniques  including  solvent  ex-
tractions with  solvents of  varying polarities,  scanning electron micro-
scopy,   energy  dispersive  X-ray  analysis  and  X-ray powder diffraction.
When ethylene  glycol  (EG)  is solidified with  cement,  the  EG appears to
occupy  at  least  3  different kinds  of  sites, characterized  by differing
extractability.   Gross  alterations of  the morphology of the cement matrix
do not  become  apparent until rather large amounts  of  EG (EG/cement =0.1
by weight) have been added to the  hydrating cement mixture.  At that point
the structure  is  clearly weakened, and the  EG  is more  readily extracted.
However,  at  concentratons   below  the  point  at  which  major structural
changes  take  place,  there  is evidence from  X-ray  diffraction that EG is
entering the  semi-crystalline calcium silicate hydrate (C-S-H) gel phase
and altering its  structure.   Since the C-S-H gel phase comprises more than
half of the hardened cement paste, alterations to that phase can be expec-
ted to alter the  ability of the cement to immobilize wastes in solidifica-
tion/stabilization processes.
INTRODUCTION
                   work  was  under-
                  a larger study of
     The  present
taken as  part  of
techniques    of    solidification/
stabilization  used  for immobiliza-
tion  of wastes  prior  to  landfil-
ling,  roadbed  construction,  etc.
The  technology has  been  most com-
monly studied  as  a  potential tech-
nique  for immobilization  of toxic
metal  ions  or  radioactive  waste.
Complete  detoxification of  metals
by  chemical,   biological  or  other
means  is  impossible short  of ele-
mental     transmutation;     hence,
methods which dilute and/or isolate
the metals are necessary as part of
waste  management   schemes.    Many
vendors  of solidification/stabili-
zation  technology  have  presented
leaching data  claiming  to  show the
efficacy of  various immobilization
techniques;  although  claims  con-
                                   328

-------
earning  the   nature   of  chemical
interactions  between  waste  metal
ions  and  the  fixing  agents  have
remained unsubstantiated(l).

     The two principal  constituents
of  most commercial  fixing agents
are  cementitious materials  (Port-
land  cement,   fly  ash,  etc.)  and
soluble  silicates.    A  limitation
often  cited  is that the processes
are incompatible with organics, but
details of such incompatibility are
generally   not  available(2).    A
potential concern is that organics,
even  in. small  amounts,  can  alter
the  process  sufficiently   to  sub-
stantially decrease the  ability of
the  fixing  agents  to  immobilize
metal  ions.    Indeed,  it   is  well
known  that  organics  can alter the
setting characteristics of Portland
cement.  It  is not clear  whether,
and at what concentration,  organics
interfere   with   the   relatively
complex setting reactions resulting
in  a  significantly altered cement
matrix.   The   question  is  of  in-
terest,  not just  with  respect  to
solidification/stabilization  tech-
nology,  but  also  with  regard  to
fundamental understanding of cemen-
ting reactions.

     Polyhydroxy   compounds    are
among the classes of organics which
alter  the  setting  characteristics
of Portland cement.   The effects of
relatively small  amounts (up to 0.5
percent)  of triethanolamine  (TEA)
on  the setting  characteristics  of
Portland cement  have  been  studied,
mainly  using   differential  thermal
analysis,  thermogravimetric analy-
sis  and  conduction calorimetry(S).
At  amounts  from  0.17  to  0.5  per-
cent,  TEA  greatly accelerates the
initial set, but  retards the final
set  and produces  a weaker cement
structure.   The  results have  sug-
gested  "a  complex  formed  between
TEA and the hydrating components of
Portland cement."

     The   present   study  involves
ethylene  glycol  (EG)  and presents
data relating to  the presence of a
chemical   interaction  between  or-
ganics  and Portland cement.   Ethy-
lene glycol was  chosen for initial
study  for  several   reasons.   It is
completely miscible with water, and
hence cement pastes can be prepared
with  any  concentration  of  EG  in
water and  in  all  cases a homogene-
ous  liquid  phase   is  being  mixed
with  the  unhydrated  cement.   Fur-
thermore,  there  is  evidence  in the
literature  that  EG does  have  pro-
found effects on already set cement
and the mineral  constituents there-
of (4, 5).   When  hydrated  cement or
tricalcium silicate (C~S) is treat-
ed with  EG,  etching occurs,  but it
is  unclear whether  mainly calcium
silicate  hydrate "gel"  (C-S-H) or
calcium  hydroxide   (CH)   is  dis-
solved^).

APPROACH

     The  presence  of  a  chemical
interaction  between  water-soluble
organics  and Portland cement  has
been  investigated  by  solidifying
pure ethylene glycol (EG) with type
I  Portland cement  and  water.   The
cement  used  had  the  composition
shown  in  Table  I.   Samples  were
prepared by weighing EG into a 20mL
borosilicate   glass,   screw   cap
scintillation    vial,     Portland
cement and water were added and the
mixture  stirred  to  apparent homo-
geneity  with  a  glass  stirring rod
(ca.  1 minute of  stirring).   Sam-
ples  were  allowed  to  cure  for
variable lengths of time.   Most of
the  samples in  the  present  study
were prepared using 4.0ml_ of water
and  10.Og  of  cement and sufficient
EG to  give EG/cement weight ratios
                                   329

-------
Table 1  Percent   Composition   of
         Portland  Cement  Used  in
         the Present Study
               Si0
               A1203
               CaO

               MgO

               K20

               SO.
20.6%

 5.2%

 2.4%

64.9%

 3.3%

 0.6%

 2.9%
of  0.02,  0.04,  0.1, 0.2,  0.5  and
1.0.

     Extractions  of ground  cement
pastes were carried  out with three
solvents:    dichloromethane   (DCM)
(dielectric   constant = 8.9,   po-
larity  index(6)  = 3.1);   dimethyl
sulfoxide (DMSO)(D =47, PI = 7.2);
and   water  (D = 80;   PI = 10.2).
After the  appropriate  curing time,
the  vials  were   broken   and  the
samples were  ground with  a  mortar
and  pestle  to  pass  a  100  mesh
screen.   The powder  and the pieces
of broken glass  from the vial were
transferred to a  125mL Erlenmeyer
flask, mixed with 120mL of solvent
and  shaken  for   0.5   hour.    The
mixtures were  filtered  under suc-
tion  using  Whatman  No. 41  filter
paper and  analyzed  by  gas chroma-
tography  using  a Hewlett Packard
HP5790A  gas chromatograph coupled
to a HP3390A integrator and a flame
ionization detector.  A 6 foot by %
inch OD (2mm  ID)  80/100 mesh TENAX
glass column  was  used.   Quantita-
tion  was  performed  using  internal
standards.     For  water   and  DCM
extractions,   1,3-propanediol   was
the  internal   standard;  for  DMSO
extractions   2-propanol   was   the
internal  standard.    In  some  pre-
liminary work  on solidification of
decanol, gas chromatographic analy-
sis  was done  using  a 25 meter by
0.31mm  ID  5%  phenylmethylsilicone
column  and  1-octanol  as  the inter-
nal standard.

     Examination   of   unextracted
samples  was  also  carried  out  by
scanning electron  microscopy (SEM)
and X-ray powder diffraction (XRD).
Each  sample was  cooled  to  liquid
nitrogen   temperature   and   then
fractured  into  several  small  pie-
ces,   each   approximately   1mm  in
diameter.   A  representative  piece
was  mounted onto an  aluminum  stub
for scanning electron microscopy in
an   ISI  60-A   Scanning  Electron
Microscope  (SEM).    Prior  to  ex-
amination,  the  mounted  sample was
coated  with a 20 nm  film  of Au-Pd
in  a   Hummer  VI  Sputter  Coater.

     The SEM was operated  at 15-30
keV  accelerating voltage with the
vacuum  chamber _at  a  pressure  of
less than 5 x 10   torr.   When fine
surface  textures were observed at
high   magnifications,   the   lower
voltages were necessary in order to
reduce  beam  penetration.    Micro-
chemical analyses were made with an
EDAX-ECON 2 Energy Dispersive X-Ray
Analyzer and  x-ray  powder diffrac-
tion  analysis  by  a GE  X-Ray  Dif-
fractometer  with a magnesium  tar-
get.

PROBLEMS ENCOUNTERED

Sample  Preparation and
Reproducibility

     At the outset of the  project
we  anticipated using  organics  of a
number  of  different types  and in a
wide  range  of concentrations.   The
result  of this is that the "solidi-
                                   330

-------
tied"   samples   vary   greatly  in
character  from  solid   masses  much
like  set  cement,  to non-solidified
"soups",  to  solid mass  with  free-
standing liquid.
     None  of the  standard  methods
of specimen preparation is strictly
applicable to the  conditions asso-
ciated  with  the present  research.
Due to  the heterogeneous  nature of
some   of  the   samples,   reliable
partitioning  of  large  solidified
samples  would   not   be  possible.
Hence   each  sample   needs   to  be
prepared  separately  and  must  be
capable  of being transferred quan-
titatively   at   each   stage  from
mixing  to  curing to  analysis.   An
additional concern in  our work has
been to limit the volatilization of
the   organic   admixture.    These
experimental  restrictions  have led
us  to   choose  to prepare  and  cure
samples  in  the  same  container and
to  use sealable  glass vials.   In
the sealed  vial  the  partial  pres-
sures  of water  or organic  vapors
above  the solid probably  rapidly
reach  equilibrium  values, although
it  is  possible  that  as  setting
reactions continue,  those equilib-
rium partial  pressures may  change
slightly.    Special    temperature
control  is  not  being  applied,  and
room temperature is  in  the  range
23°±2°C,  Samples  are  being  stored
in  the  dark to  avoid any  possi-
bility  of photochemical  reactions
of the organics.

     We were concerned with  repro-
ducibility   from   the   beginning
because  of  the  large  number  of
samples  being  prepared.  The  best
indicator we  have  that assures  us
of  reliable  sample  replication  is
the   extraction   behavior.    With
occasional exceptions we can  repro-
duce the amount of organic recover-
ed from solidified samples prepared
by our standard procedure and cured
for the same length of time so that
the   standard   deviation   is  <10
percent  of  the   average   of  the
values in  three  to six replicates.
Usually  it is  <5  percent.   These
results give us confidence that the
trends  we  are  seeing  are  real.

RESULTS

     Some  differences  in  the sam-
ples   prepared    with   differing
EG/cement ratios are obvious to the
casual  observer.    At  the  weight
ratios  from 0.02  to 0.2  the mix-
tures  harden,  but  the  0.5 and 1.0
ratio mixtures do not.   Even at the
0.2  ratio,  the  surfaces  of  the
solidified  material appear  moist.

     The extractions  with  solvents
of differing polarities give infor-
mation  regarding  the  extent  and
                   binding
                    cement
reversibility  of
ganics   to   the
Changing solvent polarity
expected to  have
                            of  or-
                            matrix.
                           would, be
                     number of pos-
sible  effects  on  the  extraction
behavior.   Intermolecular interac-
tions between  EG  and the solvents,
particularly    dipole-dipole    and
hydrogen  bonding  will  increase in
strength  in  the order DCM < DMSO <
FLO.  The solvent interactions with
tne  cement  matrix will  also  vary.
The  ability  of  HpO  to  dissolve
Ca(OH)?, and perhaps other constit-
uents as well,  out of  the  cement
paste may mean  that water penetra-
tion  into  the  material  will  be
greater  than  that  of  the  other
solvents, thus having the effect of
freeing  more  of any  organic  which
was  simply  physically  trapped in
areas of the matrix inaccessible to
the  solvent.  We  are not convinced
that the latter argument  is  valid
under   our   conditions.    When  1-
decanol,  a  quite  nonpolar organic,
is   solidified  with   cement  and
                                  331

-------
extracted with DCM,  >85  percent of
the  decanol   is   recovered.    The
matrix  may  be  different  in  the
presence  of  decanol  than  in  the
presence  of EG,  but our  grinding
and  extracting  procedure  clearly
does free most  of  the  physically
entrapped organic.

     Some results of extractions of
1  month old  samples are  shown in
Table 2 and are  expressed as per-
cent recoveries of  EG.   Our expec-
tation   that  increasing   solvent
polarity would result in increasing
recovery of  EG was  borne out.  The
results are  clearly  not  due simply
to solubility effects, since  EG is
very soluble in all three solvents.

Table 2  Percent  Recovery  of  EG in
         Extractions  with  Various
         Solvents at Differing EG/
         Cement Weight Ratios

                       %   Recovery
                      dil*    dil**
Wt. Ratio
0.02
0.04
0.10
0.20
0.50
1.0
DCM
3.7
3.4
6.5
7.0
5.1
4.8
DMSO
12
11
15
43
54
-
base
79
79
89
83
79
84
acid
95
87
76
79
79
81
* The water  solution was initially
0.05M   in   Na7CO,,   and   0.05M  in
NaHCOg, with  Sf\  fnitial  pH of 9.9.
The  pH  after extraction was  13.

**The water  solution was initially
0.05M  in acetic  acid and  0.1M in
sodium  acetate,  with an  intial  pH
of  5.0.   The  pH  after  extraction
was 12.
Since  DCM and  DMSO extract  rela-
tively  small  amounts  of EG  (com-
pared  to  water),  and  since  our
procedure  appears   to  free  most
physically  entrapped  organic,  the
reasonable  conclusion   to  draw  is
that there is  some  chemical  inter-
action  between  EG  and  the  cement
matrix  that  is  stronger than  the
EG-DCM  or  EG-DMSO  intermolecular
interactions.   The interaction that
is  certainly present   is  hydrogen
bonding, and that interaction alone
is  probably  sufficient  to  explain
the  relative extracting abilities
of the 3 solvents.

     The  loading  of  EG on  cement
might   be  expected  to  influence
extractability  very substantially,
since  bound  EG  could  be  in quite
different  environments,   and  the
cement  matrix  might be  quite dif-
ferent  at different EG concentra-
tions.   Surprisingly, any effect is
obvious  only  in  the case of DMSO
extraction.  In that case  there is
a  major break in  behavior  between
weight  ratios  of 0.1  and  0.2.   We
have concentrated on  that  concen-
tration  range  in  the  microstruc-
tural  characterization  to   be  de-
scribed below.

     The extent to  which any EG is
irreversibly  bound  in  the  cement
matrix  has  been  investigated  by
repeated  extractions  of the same
sample  with  water and  DMSO  (Table
3).  In these experiments the solid
residue  after the  initial  extrac-
tion  was  treated  with  a  second
120mL  of  solvent, reshaken,  refil-
tered  and  analyzed.    The  samples
were only  ground  one time.   All of
the EG  has not been recovered after
5  extractions with  the most agres-
sive  solvent, water.    Evaporative
losses  are  possible,   but  we  are
using  very  non-volatile organics;
and   deliberately   allowing   the
                                    332

-------
Table 3  Percent  Recovery  in  Re-
         peated Extractions of
         0.04   EG/cement   Samples
% Recovery*
Extraction No.
1
2
3
4
5
TOTAL
H 0**
£.
88
5.2
0.53
0.16
0.13
94.2
DMSO
12
1.8
0.4
-
-
14.2
*  Expressed  as   percent  of  the
original  amount   of  EG   in  the
sample.

** Deionized water.
solids to stand open to the air for
an hour or more prior to extraction
did  not  appear  to affect  the  re-
sults.   A  similar phenomenon  was
observed   in   extractions   of  0.1
weight ratio  1-decanol/cement with
DCM.   After  3 extractions  97 per-
cent  of  the  decanol  was recovered,
and  decanol  could not  be  detected
in   the   fourth   extraction.    We
conclude  from the  decanol  results
that  there are  sites  in the matrix
where physically trapped organic is
isolated   from   contact  with  the
extracting  solvent.   Water  is  a
much  more  aggressive  solvent than
DCM and can be expected to etch the
cement matrix, and yet even after 5
extractions  there  are  significant
amounts  of EG unaccounted  for.   We
conclude that there are sites where
EG  is bound  more tightly  than by
surface  hydrogen   bonding  and/or
that  there are  sites  in the cement
matrix that  are  very poorly acces-
sible to water.

     The  age  of  the  sample  does
affect  extraction  behavior  (Table
4).   Again   the  DMSO  extraction
seems most sensitive to  changes in
the  environment  of  the  EG.   The
hydration  reactions  will  be  only
partially  complete  at 7  days  of
cure.  The DMSO extraction results,
in  comparison to  the hLO  extrac-
tions,  suggest that theTe  are  at
least  two qualitatively  different
                                       Table 4  Percent Recovery  of EG in
                                                Extractions  with  Various
                                                Solvents at  Various Times
                                                of Cure
                    % Recovery

Wt. Ratio*  7-day Cure  28-day Cure
   0.04

   0.1



   0.04

   0.1
  HpQ Extraction

79           79

74           78

  DMSO Extraction

II            7.4

27           16

  DCM Extraction
0.04
0.1
4.8
6.8
3.6
7.1
*  The samples  in this  table were
prepared in a different manner from
those cited elsewhere in the paper.
In all these samples, regardless of
the  EG/cement  ratio, the  ratio of
volume  of   liquid  to   weight  of
cement  was   kept  constant  at 0.4.
                                  333

-------
environments   in   which  EG  finds
itself,  a more  tightly  bound  one
and  a  less tightly  bound one;  and
the  proportion of  EG  in  the more
tightly bound environment is higher
at  lower EG concentrations  and at
longer times of cure.

SEH Observations

     A   calcium   silicate  hydrate
(C-S-H)   gel   phase   constitutes
approximately   one-half   to   two-
thirds  of the  volume  of hydrated
pastes,  and  consequently, strongly
influences their behavior.

     Calcium   hydroxide   crystals
constitute  20-25  percent  of  the
paste  volume.   As the  paste ages,
the  calcium  hydroxide  becomes  an
increasingly  prominent  constituent
of the microstructure.  Researchers
utilizing  the electron microscope
to  characterize  Ca(OH),,  commonly
observe   hexagonal   places.    The
scanning    electron    micrographs
reveal   the  presence  of this phase
in both  plain  and waste containing
cements.

     Of    considerable    practical
interest  is  the   concentration  of
organic  required  to  make signifi-
cant changes  in the  cement matrix.
It is  reasonable  to  assume that if
the matrix is considerably altered,
the  ability  of  the  matrix  to  im-
mobilize  both metal  ions  and  or-
ganics  will  be  altered.    SEM  ob-
servations of  the gross morphology
of fracture surfaces  of the solidi-
fied  samples  reveal  few  obvious
differences  between  specimens  of
hydrated  cement alone and hydrated
cement  containing "low" concentra-
tions  of  EG.   At  the  0.1 weight
ratio,  however, significant differ-
ences are obvious.   Figure 1 shows
micrographs taken at  relatively low
magnification  under  conditions  as
nearly  identical  as  possible.   In
the  figure,  a  contains no  EG,  b:
0.04  EG/cement,  c:   0.1 EG/cement,
d:    0.2  EG/cement.    At  higher
magnifications  it  is  possible  to
find  features  that  have  similar
appearances,  particularly  in  a,  b
and c.  However, the differences in
appearance  are  strikingly evident
in  the  general  views  at 300X.   In
particular, massive  structures and
regions   of   crystal!inity  become
much  less common as  EG concentra-
tion  increases,  until  finally  d
bears little resemblance to a.   The
differences between  a  and  b might
well  go  unnoticed  if  the  other
members   of  the  series  were  not
available for comparison.

X-RayPowder Diffraction Observa-
tions
     Each  sample  was  analyzed
x-ray  powder diffraction  prior
each    microstructural
This  has  been  found
useful   since   changes
crystalline   components
waste/binder  sample   can
               by
               to
        analysis.
      to  be  very
          in   the
          of   the
        be deter-
mined  if  the  quantities  of  the
phases  are  sufficient.   Figure  2
shows XRD  data  in a graphical form
for  the  same 4  samples shown  in
Figure  1,  also identified  here  as
a-d.  The  trend  away from crystal-
1inity   as  EG   concentration  in-
creases is evident in these graphs,
as  it  is  in  the  micrographs.   The
trace  for sample  d is  quite  fea-
tureless  compared to that  for the
other samples.  It  is  worth noting
that in the two samples at lower EG
concentration,  the   same  general
features  are  apparent  as  in  the
cement    control     sample.     The
crystalline  component of  hydrated
cement that  is  easiest to identify
is  Ca(OH)?.   One
Ca(OH)2 peaks  is
plots along with
 of  the prominent
identified on the
peaks ascribed to
                                    334

-------
C-S-H  gel.    Even  sample  b  shows
distinct changes in the fine struc-
ture of  the peaks associated  with
C-S-H  gel.   It  thus appears  that
while  the  crystalline,  or  semi-
crystalline  phases  characteristic
of  hydrated  Portland   cement  con-
tinue to be  present  as  significant
quantities of organic are  added to
the  hydrating  mixture, there  are
significant  changes  in the  amor-
phous or semi-crystalline C-S-H gel
phase  at  relatively  low  organic
concentrations.

CONCLUSIONS

     When  EG  is  solidified  with
cement, the EG appears  to occupy at
least  three  different  kinds  of
environments.   At high concentra-
tions of EG, there is a significant
amount  of  very  loosely bound  EG,
which  can  be extracted  readily by
either  DMSO  or HJ3.   The  second
environment  contains more  tightly
bound  EG,  probably  bound to  sur-
faces by hydrogen bonding, and such
EG can be  readily  extracted by H?0
but  not  by  DMSO.   The  EG  still
remaining  after  repeated  extrac-
tions  with  HpO  or  DMSO  can  be
considered  to  occupy  a third  en-
vironment.    Conceivably  this  en-
vironment could  be one  in which EG
replaces  water   of  crystallization
in some  of the  crystalline  struc-
tures.    However,  the   alterations
that are  obvious by XRD  are those
in the semi-crystalline  C-S-H  gel
phase,  and  we suggest  that  the EG
is  located  there.    The  Feldman-
Sereda model(7)  of  hardened cement
paste  includes   an environment  for
water between layers of C-S-H.   EG
could substitute for water in such
an environment   and  be  both  chem-
ically  bound,   probably mainly  by
hydrogen  bonding,  and  also  physi-
cally  inaccessible  to  extracting
solvents.
     The EG  being  added to solidi-
fying  cement   does   not  have  an
effect  on  the  developing  cement
matrix  which  is  obvious  to  the
naked  eye  until  the concentration
of  EG  is rather  high (EG/cement =
0.2).  At that point the changes in
structure can  be seen  in the SEM,
but  are  also obvious  to the naked
eye.    At   lower   concentrations,
however,  there   are  more  subtle
changes, which  are clearly evident
to  XRD.   The  changes  appear  to
involve  the semicrystalline  C-S-H
gel  phase, which  normally makes up
more  than  half of  the  volume  of
hydrated cement structure.  Changes
in  that  phase will  clearly influ-
ence  the physical  characteristics
of the hardened material as well as
its  ability  to  immobilize  metal
ions.

ACKNOWLEDGMENTS

     We wish to  thank River Cement
Company,  St.  Louis,  Mo.  for  pro-
viding  the  cement   used  in  the
present  study.   Financial  support
was  provided by the  USEPA through
the Hazardous Waste Research Center
of Louisiana State University.  The
USEPA   Project   Officer  for  this
study  is  Dr.   Carl ton   Wiles.   We
also wish to thank Dr.   Roger Seals
for  helpful  discussions during the
course of this research.

REFERENCES

(1)  (a) Mishuck, E., 1984, "Encap-
     sulation/Fixation Mechanisms",
     U.S. Army  Toxic and Hazardous
     Materials  Agency   Report  No.
     DRXTH-TE-CR-84298,   June  18,
     1984.
                        G.
R.   J.
(b)  Malone,  P.
Larson,    1982,    "Scientific
Basis   of   Hazardous   Waste
Immobilization", Second Annual
                                   335

-------
     ASTM  Symposium on  Testing of
     Hazardous and Industrial Solid
     Wastes,  Orlando,   FL,   June,
     1982.

     (c)  Anon.,  1979,  "Comparative
     Investigation   on  Four   Im-
     mobilization      Techniques",
     Institute  for Waste  Research
     Publication   39,   Amersfoort,
     The   Netherlands,   September,
     1979.

(2)  Tittlebaum,   M.   E.,   F.   K.
     Cartledge,   R.  W.  Seals,  S.
     Engels,    1985,    "Technical
     Feasibility  of  Stabilization
     of Hazardous  Organic Liquids"
     CRC Critical Reviews,       In
     Press.

(3)  Ramachandran,  V.   S.,   1976,
     "Hydration of Cement - Role of
     Triethanolamine"  Cem.  Concr.
     Res.,  Vol.  6,  pp.  623-631.

(4)  Schlapfer,   P.,  P.   Esenwein,
     1936, "Untersuchungen uber die
     Einwirkung  von Aethylenglycol
     and  Glycerin  auf verschiedene
     Kalzi urn      alumi nathydrate",
     Zement, Vol.  25,  pp.  814-816.

(5)  Ciach, T.  D., J. E.  Gillott,
     E.  G.  Swenson, P.  J.  Sereda,
     1971,    "Microstructure    of
     Calcium  Silicate  Hydrates",
     Cem.  Concr. Res., Vol.  1, pp.
     IF25I

(6)  Snyder,   L.    R. ,   1978,   J.
     Chromatogr. Sci., Vol.  16, p.
     223.

(7)  Feldman,  R. F.,  P.  J.  Sereda,
     1968,  "A  Model   for  Hydrated
     Portland Cement from Sorption,
     Length  Change and  Heehanieal
     Properties",     Materiaux et
     Constructions. Vol.  1,  p.  509.
              Disclaimer

This paper has been reviewed in
accordance with the U.S. Environ-
mental Protection Agency peer and
administrative review policies and
approved for presentation and publi-
cation.
                                     336

-------
»¥. - •' -;.- r: *!**,**;?' -:- -' --X /•*••* -C.
^ <*'> ^H^if ^ - ' V •• ua -v '^?
5C^>/if'^i *'.*^c^v^^'-  ^*

               ^ - "tf.?'

     ? ^^f&P**'"' i^S** ^W-*i-





       a. CONTROL
                           b. 0.04 EG/cemtnt
          . .  »4   nr •
          .-*  '>.  -• V • -

               '''
                                     ^^,%M''4f
     ^"> •&«».•* J*y--«-- *-*r-  TIP- i^rvvw^y, ^f-A8' "••*'/*i-*'^:jr.
     ft i m^-- -^ - •>  «4 4W x-i7^• • '^; /fe
     ^ Ci  i %> l%  W$W&*i*'&&
     &Mf&.*.  &*$n'$i!m
                  (A
    c. 0.1 EG/c«mtnt            d. 0.2 EG/c«m«nt


           FIGURE I.  MICROGRAPHS OF

VARIOUS ETHYLENE GLYCOL/CEMENT SAMPLES AT 300 X
                    337

-------
                   CSH
                   CH

                 *    *
                 * *  *.*."_»
                 •*"•.""• „*"-* •
                   *>
                 • • .v..
                               X\ .  a. CONTROL
                                * « »  *
                                             •> '."/•'.••*•
                                              • '• ^•••".'V- ••.
                                                  * *   »*« **
                                                     *




 i I i 1 i i i i i i i i i i i i i i i I i i ii i i i i i 1 i i i i i i i i i I i i i i iii i i I i ii i II I V
                                 -CH
                                 '.... b. 0.04
                                "•*   c. O.i
                                 • •..  •
                                 «*•*.*.•*.
                                  -.' •-•
                                ...."  d. 0.2

                                                 '*•*-_•
0
10         15        20

       DEGREE 2 THETA
25
30
        FIGURE  2. X-RAY POWDER DIFFRACTION OF

      VARIOUS ETHYLENE GLYCOL/CEMENT SAMPLES
                            338

-------
             THE RATIONAL USE OF  CEMENT-BASED STABILIZATION
             TECHNIQUES FOR THE DISPOSAL OF HAZARDOUS WASTES

               Alistair I. Clark, Chi S. Poon, Roger Perry
                     Department of Civil Engineering
                 Imperial College of Science & Technology
                            London SW7 2BU, UK
                                 ABSTRACT

    The mechanism of zinc and mercury fixation by a cement/sodium  silicate
stabilization process has been assessed from leaching, scanning electron
microscopy. X-ray diffraction and porosimetry studies.  The results of
these tests correlate closely and suggest the operation of two separate
fixation mechanisms.  The presence of zinc has a significant effect upon
the hydration and final physical properties of the final product.  Mercury
and related metals which do not form precipitates at elevated pH levels
are held in pore solution.  The importance of microstructure in metal
fixation and also metal leaching from the cementitious matrix has  also  ,
been demonstrated.  The calcium aluminate hydrate structure, ettringite,
has been identified as to be related to the structural integrity of the
solidified product.  The understanding of the mechanisms of action can
enable a rational, cost effective approach to process design and operation
to be achieved.
INTRODUCTION AND PURPOSE

    A number of pretreatment pro-
cesses have been developed to
render industrial wastes suitable
for final disposal.  Waste stabili-
zation methods based upon ordinary
Portland cement (OPC) or other poz-
zolanic materials provide one form
of pretreatment with landfill as an
ultimate disposal route.  The mobi-
lization of wastes into water has
always been a major consideration
in containing toxic waste, and as
such, much of the emphasis or sta-
bilization processes has been
placed on preventing the waste from
coming into contact with water and
controlling the chemical conditions
of the aqueous environment in order
to minimize solubility.
    The mode of interaction of the
inorganic waste and the stabiliza-
tion system will clealy influence
the leachability of the metal (1).
Leaching studies of certain metals
stabilized in a cement matrix have
shown that the amount present in
the leachate is often considerably
lower than the calculated value
based upon the theoretical solubi-
lity product.  A variety of fixa-
tion mechanisms have been
postulated to account for this,
involving absorption by cement
hydrates, substitution and solid
solution in the hydrate structure,
and formation of insoluble com-
pounds.  However, many of these
claims are related to semi-
quantitative observations and
interpretations, leaving many of
                                   339

-------
the fundamentals to be resolved
although the  importance of the
mlcrostructure of the cement  hyra-
tion process  in relation to the
macro properties of stabilized
waste has been emphasized  (2).

    The aims  of this study were
therefore to  identify the mecha-
nisms operative in the fixation of
hzardous heavy metal wastes by
cement-based  processes and also to
improve the existing mix criteria
of these cement-based processes."

APPROACH

    The approach adopted in this
study was to  examine certain  micro
and macro properties of stabilized
wastes using  the techniques of; (i)
X-ray diffraction (XRD), scanning
electron microscopy {SEM)t mercury
intrusion porosimetry (KIP),  and
(il) leaching tests and compressive
strength tests.  This paper reports
the results of a study of an
OPC/sodium silicate formulation
although similar work is being
undertaken by the authors at  pre-
sent on the alternative OPC/PFA
process.  Both these principal
cement-based  fixation techniques
are being marketed in Europe  and
North America.

    A simulated inorganic
industrial waste containing Zn and
Hg (both group lib elements)  was
used in this  study.  These waste
compounds were selected because of
known industrial and experimental
experiences which have indicated
that their leaching potential is
very different.  The sample pre-
paration varied according to  the
analytical procedure used and can
be summarised as follows:

Leaching Test : A solution (200 ml
of 2000 ppm)  of Zn and Hg was soli-
dified by 50 g of OPC and 12 ml 40%
Na.SiO., cured at room temperature
for 28 days, crushed into small
lumps and transferred to a. con-
tainer.  Buffered acetic acid
(100 ml of 0.15 M) was added and
the mixture agitated using a rota-
tional shaker.  After 24 h, the
slurry was filtered through a
0.45 ym membrane.  A fresh portion
of acetic acid was added and the
process repeated over a period of
time.

SEM, XRD and MIP : The composition
of the four samples analysed by
these techniques were as follows:
Sample A = 10 g OPC + 10 ml H^O?
Sample B = 10 g OPC + 10 ml H_O +
1 ml 20% Na2siO,-j Sample C = 10 g
OPC + 10 ml 2% In solution + 1 ml
20% Na2SiO3? Sample D ~ 10 g OPC +
10 ml 2% Hg solution + 1 ml 20%
    Samples were prepared by
shaking either water or metal solu-
tions with cement for 3 min in a
plastic container? Na2SiO3 solution
was then added as needed and the
mixtures shaken for a further 30 s.
All samples were allowed to cure at
room temperature.  For SEM and XRD
studies 1-day samples were oven
dried at 105°C for 15 min.
Fracture specimens were prepared
and coated with gold or carbon film
prior to SEM examination using a
Jeol 35CF + EDAX system (Energy
Dispersive Analysis of x-rays)
while powdered samples were ana-
lysed by a Philips Powder X-ray
diffractometer.  The porosity stu-
dies were performed on 7-day
samples using a Carlo Erba Mercury
Intrusion Porosimeter.

    In addition, leached solidified
waste samples (2000 ppm Zn and Hg)
of every leaching period from a
leaching experiment using 0.5 M
                                   340

-------
Buffered acetic acid and a pre-
leached solidified waste sample
were oven dried at 105°C and exa-
mined by SEM.  Powdered XED pat-
terns were obtained for solidified
materials with and without dosage
of Zn and Eg  (at the 0.2% level) as
before.

Strength Testing:  Various mixes
were prepared with different water
to cement (W/C) and silicate to
cement (Si/C) ratios.  Additional
mixes were prepared by using solu-
tions containing 2% Zn and 2% Hg.
Compressive strengths were tested
at intervals of 1, 7 and 28 days
using a standard compression test
instrument conforming to BS 4550.

RESULTS

    The results from the SEM, XRD,
MIP and leaching studies of the
four samples are summarized in
Table 1 and Figure 1.  The porosity
data for the four sample types
appear to suggest three basic dis-
tribuations: one centered at 370 A,
one centered at 7500 A, and the sum
of these two.  Such an interpreta-
tion is suggestive of at least two
separate mechanisms operating in
the interaction between these
metals and the OPC/silicate system.
Increased pore volume and pore size
of the Zn-containing samples
occurred because of the extensive
growth of ettringite crystals in
the hydrated paste due to the acce-
lerated hydration of C,A as
observed in the SEM ana XRD analy-
ses.  Despite the higher porosity
the leachability of Zn is low which
indicates that permeability is not
an important factor in determining
movement of this metal through the
matrix and that chemical stabiliza-
tion rather than physical encap-
sulation is the controlling factor
in reducing metal mobility.
    It has been claimed that stabi-
lization of metal involves the for-
mation of insoluble metal silicates
but the SEM and XRD examination did
not reveal any identifiable
crystalline zinc silicate, though
amorphous gel of calcium silicate
was observed in both pure
OPC/silicate and metal-dosed
OPC/slllcate matrices.  Under these
conditions it is thought likely
that most of the Zn would be preci-
pitated on the hydroxide or would
react with the calcium hydroxide
(OH) to produce possibly calcium
zincate (3) although no evidence of
either of these zinc compounds has
been found by the SEM or 15U) tests.
The absence of C-H crystals in the
sample containing zinc indicates
that C-H plays an important role in
the fixation of Zn as confirmed by
the semi-quantitative XRD study
which demonstrated the presence of
the more crystalline phases (Table
2).  Indeed, a recent study on
cementitious solidification of
electroplating waste confirmed the
presence of amorphous metal
hydroxides in the solidified pro-
duct (4).

    By contrast, Hg did not
seriously affect the normal hydra-
tion process as evident from the
SEM, XRD and MIP studies.  The for-
mation of calcium silicate hydrates
(..O-S-H), calcium silicate and C-H
all proceed in the same manner as
in the OPC/silicate system which
indicates that there is little or
no interaction of this metal with
OPC or sodium silicate.  This ina-
bility of Hg to form an insoluble
hydroxide or silicate with the
solidifying material means that the
metal remains In pore solution or
at most is only loosely bound to
the hydrated products through sorp-
tion.  The metal is therefore phy-
sically encapsulated within the
                                   341

-------
TABLE 1.
S*H?1£

A. OPC *
H20

1. OPC/


c. OPC/
Solution

D. OPC/
* H9
Solution


, SUMMARY OF CORRELATION OF FOUR STUDIES
SEH
< 1 day sample )

C-S-H
C* tOH)2
*ttringite
C-3-M
ettrinqite
c*-Biiicate
c-s-s
C« CQ8J2
ettringite
Ca~silicate
C-S-H
Ca IOH>2
ettritigite
Ca~*ilic*te

flbroua
large crystals
small rods
fibroua, hydro ted
•hell C Bad ley grain}
snail crystals
snail rods
gel massive
a little,
reticulated
absence
large hexagonal
prism with AFm
gel massive
fibrous t hydra ted
shell CHadley grain)
small crystals
snail rods
gel massive
Powder XRO PG cosine try Total Pore
C1 day sample) <7 days sample) Volume
Con3.?"*1 )
Unhydrated cement medium strong single 0. 136
Ca (OHJ2 strong
ettringite weak 3?QA
Unhyd rated cement medium strong double 0.416
Ca (OK)2 cMsdiun! 370ft,
ettringlte weak 7SOOA


y 99
Ca EQK ) 2 absence low
ettringite medium 720 OA

Unhydrated cement medium strong double 0.376
Ca (QH>2 medium 370R high
•ttsingite weak 7500A

Leachability
Values


-


-


low

high


 cement structure  and not chemically
 stabilized to the extent observed
 for zinc.  Mobilization of Hg
 leachability of the  fixed product
 is high and probably dependent on
 the permeability  of  the solidified
 product.

     The use of SEM has elucidated
 in what form this leaching fluid
 acts on the cementitious material
 (e.g.  Figures 2,  3  and 4).  In an
 unleached OPC/silicate sample with
 the simulated metal  waste, normal
 hydration products of C-S-H, C-H,
 calcium aluminate hydrates (: Xt and
 Afm phases) and unhydrated cement
 grains together with calcium sili-
 cate gel were observable.  In com-
uarison with OPC/silicate without
 metal addition, increased growth of
 the Aft phase with less C-S-H
 (Mostly Type I) was  noticeable.
 The SEM study showed that the most
TABLE 2. SEMI-QUANTITATIVE XRD DATA
   Sample
                               Relative

                               Int«n»lty*
 A 50 g OPC

   12 Bl KB.
 B 50 g OPC
   12 Kl N«2Si03
                     C-H
                               2.45
                    bellte (C2S)  1.03

                    •lite  {C3SJ  0.87

                    C-B        1.64
   200 al 2000 ppct Hg,Zn  belite (C^S)  1.16

                    •Xite (CjS)   0.95



 * normaliaed to HgCOH)
                                     342

-------
   -  0-14 r
    e»
   -E 0-12 -
    •g 010 -
    JJ
    6 D oe-
    § 004-
    I 0-021-
                          I*)
7500   3700    1500   750    370
      08
      0-6
      0-i
   §0.2
   1
                                          1SO
"n"
                                                i
                                               37
—T~
 15
"rT
                                                        Pore rotJiui nin
                                                                          3-7
                                                                         1C!
                                                                          IB)
7500  3700    1500   750   370
                                          150    75     37       15
                                                  Pert rediu* nm
                    7-5
If
i i
*
* »
i:
i;
; j
-
-

-
-

-
™
-
nm mo noo »>» m no n r »
Swpl* *•
•
s
» •
•
1 *
* I
hi J"
r
-
—
-
—
-
-
]_, _
NlrTT "i
TI it noc STDD two rMtTD tsonif nti»
mm fNp* *•«>•«» *»
       MD  no  in
       Siopl* e
Figure  1. Cumulative pore volume and pore-*.M.ze distribution.
                                       343

-------
Figure 2.  Micrograph showing pre-leached sample with long rods of
           ettringite, C-S-H, calcium silicate gel.  Corresponding
           EDAX analysis showing major peaks at 1.74 and 3.69 ev
           (silicon and calcium respectively - more Ca than Si).
Figure 3.  Micrograph of 1-day leached sample showing a grain with
           ettringite crystals and C-S-H.  The structure of C-S-H is
           less defined.  Sample still contains more Ca than Si.
 Figure 4.  Micrograph of 5-day leached sample showing smooth hydrolysed
            surface of grain and areas of gel-like morphology.  EDAX
            analysis indicates sample contains more Si than Ca.
                                344

-------
easily hydrolysed  structure was the
C-S-H phase which  after the 3  day
leaching period was  no  longer
observable.  However Zn and Hg
leachate concentrations did not
change appreciably which indicates
that C-S-H is unlikely  to be
involved in the fixation mecha-
nisms .

    Further leaching hydrolysed the
matrix and was associated with con-
tinued Ca removal  as demonstrated
by the EDAX analysis.   Possible
sources of Ca are  AFt and Mm,
calcium silicate gel, C-H and
unhydrated cement  grains.  After 3
days, less Aft (the  most iden-
tifiable phase) and  Afm were obser-
vable with increased calcium
silicate gel type  structures.
Concentrations of  Si and Ca were
similar at this stage.   However,
the Aft phase was  no longer visible
after the 5 day leaching period and
this coincided with  the dramatic
rise in Zn and Hg  leachate con-
centrations which  occurred sub-
sequently.  This indicates that the
disappearance of this phase marks
the beginning of break  down of the
structures that "fix" the metal
though there is no direct evidence
to prove that this phase is
directly responsible for such fixa-
tion.  After this  phase was hydro-
lysed the cementitious  matrix has
been sufficiently  broken down for
the fixed waste material to be
easily leached.  This occurs after
approximately 55%  of Ca removal and
corresponds to a massive breakdown
of structure.

    Figure 5 and 6 demonstrate that
high Si/C ratios increase initial
strength {within a few  hours) but
that the final strength is more
dependent on the w/c ratio).  The
increase in w/c not  only increases
the total intruded volume but also
        60-1
        so-
        40-
       o>
       c
      §30-
       *
       >
        2O-
      O 10.
                    -i	r W/C
          0 0-2 0-4 0-6 08 1-0
Figure 5. Strength vs W/C  ratio.
    O-l
    2H
   o
   o
    1-
                          28d
     o   O-04 o-es o-oe o-o?  o-oe
        1-04 1-05 *O6 1-07  t-08   W/C
Figure 6. Strength vs Si/C and
          effective W/C ratios.
                                   345

-------
shifts the pore size distribution
to a large pore radii.  This  is
considered to be the most critical
factor in determining the mechani-
cal strength of a cementitious
system (5).  The role of sodium
silicate in accelerating the
setting of the product is crucial
in normal stabilization practice in
order to minimize runoff and
leaching of waste material during
the critical initial setting  stage.
The addition of sodium silicate
does not promote a significant
improvement in final strength which
is in accord with the findings of
Nelson and Young (6).

    Material costs contribute the
majority of the overall cost  of
solidification processes and  thus
to a waste disposer, the addition
of additives such as sodium sili-
cate which accelerate setting may
increase the volume of waste  that
can be treated for the same amount
of material.  However, such an
increase in the water content
results in reduced physical strength
which has important implications
for the disposal of solidified
wastes in landfill sites.  It must
be noted that the objective of
these processes is to reduce  both
leachability and permeability and
to improve compressive strength
with minimal material costs.

ACKNOWLEDGEMENTS

    One of us (CSP) wishes to
acknowledge the support of Wimpey
Waste Management and Wimpey
Laboratories.

REFERENCES

1.  Poon, C.S., C.J. Peters,  R.
    Perry and C. Knight, 1984,
    Assessing the leaching charac-
    teristics of stabilized toxic
    waste by the use of thin layer
    chromatography, Environ.
    Techno1. Lett., Vol. 5, pp.
    1-6.

2.  Poon, C.S., C.J. Peters and R.
    Perry, 1983, Use of stailiza-
    tion processes in the control
    of toxic wastes. Effluent and
    Water Treament J., Vol. 23, pp
    451-459.

3.  Tashiro, C. and S. Tatibana,
    1983, Bond strength between CjS
    paste and iron, copper or zinc
    wire and microstructure of
    interface.  Gem, and Concr.
    Res., Vol. 13, pp. 377-382.

4.  Malone, p.G., C.w. Jones, J.P.
    Burkel, 1984, Application of
    solidification/stabilization
    technology to electroplating
    wastes.  In Land Disposal of
    Hazardous Wastes. Proc. Annual
    Research Symposium (9th), Ft
    Mitchell, Kentucky,
    PB-84-11877.

5.  Ramachandran, V.S., R.F.
    Feldman and J.J. Beaudoin,
    1981, Concrete Science,
    Treatise onCurrentResearch,
    Heydon and Sons Ltd., London,
    UK.

6.  Nelson, J.A. and J.F. Young,
    1977, Addition of colloidal
    silicas and silicates to
    Portland cement paste.  Gem.
    Conor. Res», Vol. 7,
    pp. 277-282.
               Disclaimer

 The work described in this paper was
 not funded by the U.S. Environmental
 Protection Agency.  The contents do
 not necessarily reflect the views of
 the Agency and no official endorse-
 ment should be inferred.
                                   346

-------
SAMPLE
h* OK: *
H2°

B. GPC/


c. OIK:/
solution


D. QPC/
+ Hg
Solution

SEH
{ 1 day sample)
C-S-H
Ca (QH)_
ettringite
C-S-H
Cm (Otn 2
ettringite
Ca-silicate
C»S-H
Ca 2
ettringite

Unhydrated
Ca { OH ) »
ettringite
Powder XRD
day sample)
cement medium strong
strong
weak
cement medium strong
week

cement strong
absence
medium

cement raedium strong
medium
weak
ftorosimetry Total Pore Leachatoility
C? days sample) Volume Values
single 0.136
370ft
double 0.416
370A
7500&

single 0.684
low low
7200ft

double 0.376
370A high high
7500A
Ca-silicate gel  raasaivn
                                         347

-------
            SORBENT ASSISTED SOLIDIFICATION OF A HAZARDOUS WASTE
       Tommy E. Myers, Norman R. Francingues, Jr., Douglas W. Thompson
                      USAE Waterways Experiment Station
                            Vicksburg, MS  39180
                                     and
                               Donald 0. Hill
                        Mississippi State University
                        Mississippi State, MS  39762

                                  ABSTRACT

     Sorbent assisted solidification as discussed in this paper is a treat-
ment technique for hazardous wastes containing toxic metals.  The basis of
the technique is adsorption/ehemisorption of metals by a sorbent that is used
in conjunction with a solidification process to immobilize metal contaminants.
The sorbent-metal complex is thought to become incorporated (not simply
entrapped) into the crystalline matrix provided by solidification processing.

     Sorption isotherms and chemical leach tests were used to investigate the
capacity of various sorbents to seize and hold metal ions, specifically
copper ions.  The sorbents investigated were flyash, soil, and organosilane
conditioned flyash and soil.  Organosilanes are chelating agents that have
the ability to seize and hold metal ions.  Adsorption isotherms were run in
order to determine the ultimate capacity of the various sorbents for copper.
The Langmuir equation was used to model the adsorption process.  Desorption
isotherms were run in order to obtain distribution coefficients related to
the release of copper from solidified waste.

     The data show that N-(B-aminoethyl)-Y-aminopropyl-trimethoxysilane, a
commercially available organosilane, can be used to improve the adsorption
properties of soil, and that sorbent assisted solidification reduces copper
leaching.  With the proper development and application, sorbent assisted
solidification could provide the technology needed for improved land disposal
of solidified hazardous waste.
INTRODUCTION

     Solidification is a treatment
technology that is sometimes
applied to liquids and semi-solids
which are too toxic for biological
treatment, too low in energy value
and/or too corrosive for thermal
processing, and too dilute for
landfilling.  Solidification typ-
ically involves mixing a setting
agent(s) with a waste to form a
hard, durable product that is sub-
stantially insoluble in water and
in which the waste contaminants
are entrapped in the solidified
mass.  There are several commer-
cially available solidification
processing systems  in use  in  the
United States  (12).  The most common
setting agents are  Portland cement
and pozzolans such  as flyash, kiln
dust, lime, soluble silicates,
gypsum, and combinations of these
materials.  (Pozzolans are materials
other than Portland cement that have
cementious properties.)  Generic
descriptions of the commercially
available solidification processes
have been published by Malone and
Jones (11).

     Solidification typically pro-
vides three major advantages over raw
waste disposal| these are 1) removal
of free liquid, 2) development of
                                      348

-------
structural integrity, and 3) improved
contaminant isolation and containment
(11, 12, 14,).  Isolation and con-
tainment of hazardous constituents
are accomplished by waste entrapment
in a cemented matrix and by con-
version of waste constituents to less
soluble compounds (precipitation).
Unfortunately, these mechanisms do
not always prevent the leaching of
hazardous constituents.  The
effectiveness of the entrapment
mechanism depends on the permeability
and durability of the solidified pro-
duct.  Because waste constituents can
interfere with the setting reactions
responsible for the development of a
hardened mass, there are problems
with durability.  Insoluble products
from precipitation can be resolu-
bilized if the leaching conditions
are different from the conditions in
which precipitation took place.  In
particular, pH and oxidation-
reduction potentials can be altered
by percolating water to resolubilize
toxic metals.  In addition, not all
materials are insoluble under the pH
and oxidation—reduction conditions
in moist concrete or pozzolan.
Hence, contaminants that have been
simply entrapped or precipitated can
be leached from solidified waste in
varying degrees, depending on the
type of waste and the kind of addi-
tives used (14).

     Adsorption is one means by which
soluble metals can be removed from
contaminated aqueous systems
(2,3,7,8).  Since adsorption is
reversible, desorption will occur to
some extent depending on the relative
affinity of the contaminant for the
aqueous phase versus the sorbed
phase.  The thermodynamics of the
sorption system partitions the con-
taminant mass between aqueous and
adsorbed phases so that the con-
taminant is never all in one or the
other phase, and thus, cannot be
released all at one time.  Con-
sequently, the amount released and
especially the release rate is
reduced if the contaminant is
adsorbed to a solid phase, rather
than simply entrapped in a lattice
work as a soluble or potentially
soluble salt.

     Unfortunately, the solid matrix
provided by cement and pozzolan pro-
cessing has little or no sorption
potential.  However, if a sorbent
that becomes incorporated into the
crystalline matrix provided by solid-
ification is included in the process
formulation, then the pollutant
potential of the solidified waste
should be significantly reduced.

     The metal sorption properties
of materials can be improved by bond-
ing certain organosilanes to them
(9, 13).  Organosilanes are silicon
compounds derived from silane, SiH,.
As chelating agents some organosilanes
have the ability to seize metal ions
and sequester them from further
reactions.  Leyden and Luttrell (9)
used the metal chelating properties
of organosilanes to preconcentrate
dissolved metals prior to chemical
analysis by X-ray fluorescence.
Malone and Karn (13) reported the
usefulness of an organosilane-silica
gel sorbent in removing cadmium,
chromium, copper and zinc from con-
taminated wastewater.

PURPOSE

     The objective of this study was
to investigate the feasibility of
reducing the metal leaching potential
of solidified waste by conditioning
of selected solidification additives
with N~(g-aminoethyl)-Y-aminopropyl-
trimethoxysilane.

APPROACH

     The research approach consisted
of 1)  conducting adsorption isotherm
tests on organosilane-additive pre-
parations in order to determine the
sorption characteristics for copper
and 2) conducting desorption isotherm
                                    349

-------
tests on solidified waste prepared
with and without silane conditioning
of selected solidification additives
in order to obtain distribution coef-
ficients for copper release,

Waste Solidification

    A concentrated brine containing
approximately 12 percent by weight
chloride, 8 percent by weight organic
carbon, 17 percent by weight dis-
solved solids, 4040 mg/£ copper, and
various toxic substances such as
aldrin, arsenic, and cyanide at parts
per million levels was collected from
a hazardous waste impoundment for
study and testing.  The liquid was
solidified in 1 liter batches with
equal weights of soil, flyash, lime,
and waste.  This solidification pro-
cess was chosen for investigation
because it included two materials
with sorbent potential (soil and fly-
ash) .  The flyash was obtained from
a local power plant.  The ash had a
specific gravity of 2.07 and con-
tained 4,7 percent lime as CaO.  The
soil, also available locally, was a
sandy clay (CL by the Unified Soil
Classification System).  The solidi-
fied waste was cured at room tem-
perature for 1 days prior to testing.

Sorbent Preparation

    The preparation of sorbents
involved the bonding of
N-(8-aminoethyl)-Y~aminopropyl-
trimethoxysilane to either flyash or
soil.  The organosilane (Dow Z-6020)
was obtained from SCM Speciality
Chemicals, Gainesville, Florida.  The
bonding procedure adapted from Leyden
and Luttrell (9) incorporated the
following steps:

    _a_.  A 10% aqueous silane solution
was prepared by adding silane to
deionized-distilled water with
stirring.  The solution was acidified
to pH 5-6 by adding reagent grade
glacial acetic acid.  The pH was
checked with pH paper.  Plastic con-
tainers were used to avoid reaction
with glass.
                                   350
    b_.  One gram of either flyash or
soil was contacted with aqueous
silane on a mechanical shaker for
30 minutes at a liquid to solids
ratio of 1.5 ml of aqueous silane to
one gram of substrate.

    _c_.  The mixture was reacted at
70°C in a vented oven for 24 hours.
In one bonding procedure the loss of
water was controlled so that the mix-
ture was not allowed to dry.  In
another the mixture was taken to com-
plete dryness at 70°C.

    d_.  The sorbent preparations that
were taken to complete dryness were
washed with water to remove excess
silane and then dried at 40°C.

    _e_.  The sorbent preparations that
were not taken to complete dryness
at 70°C were washed with water to
remove excess silane and then air
dried in a hood at room temperature.

Adjs orp t ion I so therms

    Adsorption isotherms were run in
order to determine the ultimate
capacity of the various sorbents for
copper and to determine equilibrium
adsorption constants.  In these tests
one gram quantities of sorbent were
contacted on a mechanical shaker for
24 hours with 100 ml of liquid waste
in various dilutions.  Each mixture
was then filtered and analyzed for
copper.  Six dilutions of liquid
waste were used as follows;  1000/1,
100/1, 50/1, 10/1, 5/1, and 1/1 (no
dilution).  Blanks consisting of
sorbent and deionized water were run
with each adsorption isotherm test.
Liquid-solid separation was by filtra-
tion using Gelman No, 61631 glass
fiber filters.  Adsorption isotherms
were run on soil, soil with organo-
silane, flyash, and flyash with
organosilane.

Desorption Isotherms

    Desorption isotherms were run by
contacting solidified waste samples

-------
with deionized-distilled water on a
mechanical shaker for 24 hours in
liquid to solid ratios as follows:
50 ml:5g, 50 ml:2g, 50 ml:lg,
100 ml:lg and 100 ml:Q.5g.  The mix-
tures were filtered and analyzed.
Blanks consisting of deionized-
distilled water were included in each
set.  Liquid-solids separation was
by filtration using Gelman No. 61631
glass fiber filters.  Desorption iso-
therms were run on waste solidified
using soil without organosilane condi-
tioning and on waste solidified using
organosilane conditioned soil.

Chemical Analysis

    Isotherm samples (aqueous phase)
were analyzed for copper on a
directly-coupled plasma arc spectro-
photometer by the Analytical Labora-
tory Group (ALG), Environmental
Laboratory, USAE Waterways Experiment
Station, Vicksburg, MS.
the soil surface are forced into a
high collision probability by taking
the mixture to dryness.

RESULTS

Adsorption Isotherms

    In an isotherm test the amount
of contaminant removed is determined
as a function of the aqueous con-
centration at a constant temperature.
The resulting set of data is called
an adsorption isotherm.  From the
data a table of aqueous phase con-
centrations, C, and corresponding
sorbent phase concentrations, q» can
be prepared.

    The Langmuir equation is often
used to model adsorption processes
at equilibrium (15).  The Langmuir
equation is

        q - K. Q C/(l + K. C)    (1)
PROBLEMS ENCOUNTERED

    The most significant problem
encountered involved the
soil/organosilane bonding procedures.
When the soil organosilane mixture
was not taken to dryness, the sorbent
produced was inferior to that pro-
duced when the mixture was taken to
dryness.  This was indicated by
reduced adsorption coefficients as
discussed in the results section.
One possible explanation for this
relates to the collision frequency
between the silane molecules and
active hydroxyl sites on the soil
surface.  The reaction of organo-
silane with a substrate involves
hydrolysis of the methoxy groups,
polymerization by condensation,
hydrogen bonding between the con-
densed organosilane polymer, and
finally formation of a silicon-
oxygen bond (siloxane) between the
substrate and the silane (1).
Apparently the final step is not
accomplished until condensed organo-
silane polymer and active sites on
where
     q = Concentration of contaminant
         in the adsorbed phase, m/m
                                   5

     Q = Ultimate monolayer capacity
          of the sorbent, m/m
     C = Concentration of contaminant
         in the aqueous phase, m/L
    K. = Langmuir distribution
         coefficient, L /m
    m  = Mass of sorbent
     s
    The adsorption isotherm data
obtained in this study were analyzed
by least squares fitting of the data
to the linear form of equation (1)
below.
         C/q
l/(QKd)
C/Q
(2)
The coefficients 1/(QK.) and 1/Q were
determined by least squares analysis
and from these Q and K, were obtained.
These coefficients were then used to
produce the adsorption isotherms
shown in Figure 1.
                                    351

-------
      0       500      1000     1500
                 C, mg/K


               LEGEND

  (1);  SOIL
  (2);  SOIL/ORGANOSILANE TAKEN TO
        DRYNESS
  (3):  SOIL/ORGANOSILANE, WET
  (4):  FLYASH/ORGANOSILANE TAKEN
        TO DRYNESS
  (5):  FLYASH

Figure 1.  Adsorption Isotherms

    The curves show that con-
ditioning of the soil with organo-
silane enhanced the ultimate
adsorption capacity, Q.  (Q is the
value each curve asymptotically
approaches.)  On the basis of ulti-
mate adsorption capacities, the
organosilane conditioned sorbents
were superior to the unconditioned
sorbents, and the soil based
sorbents were superior to the flyash
based sorbents.  The adsorption
isotherm data also show that the
soil even without organosilane
conditioning is more than just a
filler in the solidification
process.  Copper is adsorbed by the
soil.

    The data also indicate that the
soil/organosilane bonding procedures
affects sorbent performance.  The
best copper adsorbent investigated
in this study, as indicated by the
isotherms shown in Figure 1, was the
soil/organosilane sorbent prepared
by taking the soil/organosilane mix-
ture to dryness.  Soil was generally
less effective than the
soil/organosilane sorbents, espe-
cially when the soil/organosilane
mixture was taken to dryness during
sorbent preparation.  The flyash
based sorbents were inferior to the
soil based sorbents.  The poor per-
formance of the flyash preparations
was probably due to the lime com-
ponent in the flyash.  Base
hydrolysis of the siloxane bond can
occur under certain conditions (1).

Desorption Isotherms

    Desorption isotherms can be used
to provide fundamental information
on the interactions between ground-
water and contaminated soils and
sediments (4, 5, 16).  If the sorp-
tion processes of adsorption and
desorption have taken place under
identical conditions, then the
desorption of a contaminant back into
the aqueous phase should proceed down
the adsorption isotherm and follow
it exactly.  However, the conditions
under which, adsorption takes place
is usually significantly different
from those under which desorption
isotherms are developed.  If the
sorption systems differ significantly
in pH, ionic strength, and chemical
activity, the desorption process will
not simply be a reversal of the
adsorption process.

    Jeffe and Ferrar (6) suggested
the following simple mass action
desorption model:
dC/dt
                         Kq
(3)
where K| is the adsorption rate and
K_ is tne desorption rate.

For steady state conditions dC/dt =
0, and equation 3 becomes equation 4
below:
                                     352

-------
                                            4 p
              q-K' C
(4)
where Kt = K!/K
       a    1
Equation 4 describes the
relationship between sorbed and
aqueous phase concentration for a
desorption-dominated equilibrium by
a simple distribution coefficient,
K'.  Each value of contaminant
loading in the solid phase, q,
supports a unique aqueous phase
concentration, C, that at equilibrium
is directly proportional to q by the
distribution coefficient, K'.  In
terms of leaching potential, the
higher the distribution coefficient,
Kj, the lower the aqueous phase con-
centration, C, that a given sorbent
loading, q, will support.

    The desorption isotherm data are
presented in Figure 2.  Sorbed versus
aqueous phase concentrations are
plotted for waste solidified using
soil without organosilane condition-
ing and for waste solidified using
soil with organosilane conditioning.
The desorption isotherm data fit the
linear model of q versus C presented
in equation (4).  The desorption
isotherms show that sorbent assisted
solidification using organosilane
conditioned soil produced a product
with a higher Kl than did sorbent
assisted solidification using soil
alone.  The affinity of copper for
the organosilane reduced the amount
of copper in the aqueous phase.  The
organosilane used in the desorption
isotherm test was not prepared by
taking the soil/organosilane mixture
to dryness.  The adsorption isotherms
in Figure 1 indicate that had sorbent
preparation involved taking the
soil/organosilane mixture to dryness,
the distribution coefficient would
have been even higher.  The problem
with bonding procedures was not
recognized early enough in the test-
ing program to include sorbents pre-
pared by taking the mixture to
dryness in the desorption isotherm
testing.
                                    353
                                                           0.6829^'? °'3562
                     l.  I   I	L
                                   LEGEND

                                 SOIL/ORGANOSILANE
                               I  SOIL
                               	1,  I	1	1
                     3456
                         C, mg/8
                                7  8  9  10
         Figure 2.  Desorption Isotherms

         PotentialFor Field Application

             Several important aspects of
         field application were not addressed
         in this laboratory study of sorbent
         assisted solidification using organo-
         silane.  Topics beyond the scope of
         the investigation include scale-up
         factors, organosilane compatibility
         with alternative binder/substrate
         systems, long-term stability of £he
         solidified product, and engineering
         economy.  Additional testing and
         evaluation is needed before organo-
         silane assisted solidification can
         be applied in the field.

             In addition, organosilanes are
         produced as speciality chemicals that
         are expensive in small quantities.
         Even with a significant cost break-
         through between laboratory and bulk
         quantities, full-scale application
         of sorbent assisted solidification
         using organosilane may be limited to
         small volumes of highly contaminated
         metal wastes that would otherwise
         pose a serious environmental hazard
         if landfilled.

         CONCLUSIONS

             Sorbent assisted solidification
         using organosilane is an innovative
         treatment technology that could be
         applied, depending on technical and
         economic factors associated with
         full-scale application, to highly
         contaminated metal wastes prior to

-------
land disposal.  Specific conclusions    3.
drawn from the results of the study
are as follows:

1.  Organosilane, specifically
N-(0-aminoethyl)-Y-aminopropyl-
trimethoxysilane, can be used to
improve the natural adsorptive capa-    4.
city of soil for copper.

2.  Organosilane, specifically
N-(0-aminoethyl)-y-aminopropyl-
trimethoxysilane, can be used to
reduce the copper leaching potential
of solidified waste when Organosilane
conditioned soil is used as a solidi-
fication additive.
                                        5.
3.  Soil/organosilane bonding pro-
cedures affect sorbent performance.
When the soil/organosilane mixture
is not taken to dryness during
preparation, the sorbent produced is
inferior to that produced when the
mixture is taken to dryness.            6.

ACKNOWLEDGEMENTS

    The tests described and the
resulting data presented herein were    7.
obtained from research conducted
under the Department of the Army In-
House Laboratory Independent
Research (ILIR) Program, 1LIR project
No. 4A161101A91D, by the US Army
Engineer Waterways Experiment Sta-
tion, Vicksburg, MS  39180.             8.

REFERENCES

1.   Arkles, B.  1977.  "Techniques
     for Silylation," Chemtech,
     7:713-717.
                                        9.
2.   fuller, W. H., Amoozegar-Fard,
     A., Niebla, E. E., and
     Boyle, M.  1980.  "Influence of
     Leachate Quality on Soil
     Attenuation of Metals."  Proc.
     Sixth Annual Res. Symp. on
     Disp. on Haz. Wastes,
     EPA-600/9-80-010, U. S. Environ-
     mental Protection Agency,
     Cincinnati, OH  45268.
Griffin, R. A., et al.  1976.
"Attenuation of Pollutants in
Landfill Leachate by Clay
Minerals."  Envir. Geology
Notes No. 78, Illinois Geo-
logical Survey, Urbanna, IL.

Houle, M. J. and Long, D. E.
1980.  "Interpreting Results
from Serial Batch Extraction
Tests of Wastes and Soils."
Proc. Sixth Ann. Res. Symp.,
Chicago, IL,
EPA-600/9-80-010, US Environ-
mental Protection Agency,
Cincinnati, OH  45268.

Isaacson, P. J. and Frink, C. R.
1984.  "Nonreversible Sorption
of Phenolic Compounds by
sediment Fractions:  The Role of
Sediment Organic Matter," Envir.
Sci. &Tech.. 18:43-48.

Jaffe, P. R. and Ferrara, R. A.
1983.  "Desorption Kinetics in
Modeling of Toxic Chemicals," _J._
of Env. Eng.. 109:857-867.

Korte, N. E., Skopp, J., Fuller,
¥, H., Niebla, E. E., and Alesii,
B. A.  1976.  "Trace Element Move-
ment in Soil:  Influence of Soil
Physical and Chemical Properties,"
Soil Sci.. 122:350-359.

Korte, N. E., Skopp, J., Niebla,
E. E., and Fuller, W. H.  1975.
"A Baseline Study on Trace Metal
Elution From Diverse Soil Types,"
Water, Air and Soil Pollut.,
5:149-156.

Leyden, D. E. and Luttrell,
G. H.  1975.  "Preconcentration
of Trace Metals Using Chelating
Groups Immobilized via
Silylation," Anal. Chem., 47:9,
1612-1617.
                                    354

-------
10.  Maloch, J. L., Averett, D.  E,
     and Bartos, M. J., Jr.   1976.
     Pollutant Potential of  Raw and
     Chemically Fixed Hazardous
     Industrial Wastes and Flue Gas
     Desulfurization Sludges.
     Interim Report.  EPA-600/2-76-
     182, U. S. Environmental Pro-
     tection Agency, Cincinnati, OH
     45268.

11.  Malone, P. G. and Jones, L. W.
     1979.  Survey of
     Solidification/Stabilization
     Technology for Hazardous
     Industrial Wastes.  EPA-600/2-
     79-056, U, S. Environmental Pro-
     tection Agency, Cincinnati, OH
     45268.

12.  Malone, P. G., Jones, L. W., and
     Larson, R. J.  1980. Guide to
     the Disposal of Chemically
     Stabilized and Solidified Waste.
     SW-872, U. S. Environmental Pro-
     tection Agency, Cincinnati, OH
     45268.

13.  Malone, P. G. and Karn, R.  A.
     1982.  Toxic Metal Removal From
     Electroplating Wastewater Using
     Silylated Silica Gel, Misc. Paper
     EL-82-3, USAE Waterways Experi-
     ment Station, Vieksburg, MS
     39180.

14.  Malone, P. G. and Larson, R. C.
     1983.  "Scientific Basis of
     Hazardous Waste Immobilization,"
     Hazardous and Industrial Solid
     Waste Testing.  Second  Symposium,
     STP 805, ASTM, Philadelphia, PA
     19103.

15.  Metcalf and Eddy, Inc.   1979
     WastewaterEngineering;  Treat-
     ment, Disposal, Reuse,  2nd ed.
     McGraw Hill Book Co., New York,
     NY.
16.  Myers, T. E. and D, 0. Hill.
     1985.  "Extrapolation of Labora-
     tory Leach Data to the Field
     Situation," Paper presented to
     the Annual Meeting Mississippi
     Academy of Science, 21 Feb 1985,
     Jackson, MS.
                 Disclaimer

   The work described  1n  this  paper was
   not funded by  the U.S.  Environmental
   Protection Agency.   The contents do
   not necessarily  reflect the views of
   the Agency and no official  endorse-
   ment should be Inferred.
                                     355

-------
           THE EFFECT OF PARTICLE SIZE ON THE.LEACHING OF HEAVY METALS
                        FROM STABILIZED/SOLIDIFIED WASTES

                        Todd M, Brown and Paul L. Bishop
                           University of New Hampshire
                           Durham, New Hampshire 03824

                                    ABSTRACT

     Stabilization/solidification of hazardous liquids and sludges by a
variety of techniques has been investigated as a method of treatment that
will bind hazardous materials in a form that minimizes adverse effects on
the environment after landfilling.  One method of stabilization/solidifi-
cation that has been investigated is the combination of the waste with
Portland cement to form a cement paste.  The structure of the paste re-
sists chemical attack, making the technique attractive as a disposal alter-
native for heavy metal sludges.  This paper presents results concerning the
effect of particle size on the leaching pattern of metals from these
wastes.  Both batch and upflow column leaching tests were investigated.

     The results indicate that in the early part of the column test, small
particles at the bottom of the column release both metals and alkalinity.
As the leachant moves upward through the column, the alkalinity leached
neutralizes the acid leachant causing the pH to rise.  This forces the
metals that were leached below to either be re-precipitated or sorbed onto
a particle surface.  As the leaching zone rises through the column, metals
are concentrated near the top of the column.  When the leaching zone breaks
through the top of the column, metal concentrations in the leachate increase
rapidly.  As the extraction continues beyond this point, the concentrations
of metals in the leachate gradually decline.  In the columns packed with
larger particles, however, the rate of alkalinity release from the particles
is not rapid enough to neutralize all the acid in the leachant by the time
the leachant reaches the top of the column.  This results in an immediate
release of metals into the leachate. As extraction continues, the rate of
metals release gradually increases, indicating that the alkalinity that is
being released is exposing  more surface area to the acid leachant.

INTRODUCTION AND PURPOSE                    One method of disposing of
                                       some of these wastes.in a more en-
     The disposal of hazardous         vironmentally acceptable fashion is
waste on land has almost universally   to stabilize/solidify them before
led to environmental problems due      landfilling.  In this process the
to leaching of waste constituents      waste sludges are combined with
into groundwater.  Properly designed,  various additives that both chemically
constructed and managed "secure"       bind and physically solidify the
chemical landfills, in which clay      hazardous materials, thus making them
or synthetic membrane liners and       less, susceptible to leaching.  Stabi-
leachate collection systems are        11 zed/solidified wastes may still
used, can delay the onset of ground-   leach with time, but the rate of
water contamination, but these         leaching should be very low so that
liners may eventually break down.      the pollutants will disperse into the


                                      356

-------
environment without adverse con-
sequences. 	

     One method of stabilization/
solidification which has been stud-
ied extensively at the University
of New Hampshire consists of the
combination of inorganic waste
sludges with port!and cement to form
a cement paste.  The structure of
the Solidified paste resists physical
attack and gives the product strength
and durability, and the highly
alkaline nature of the material
resists chemical attack, making the
technique attractive as a disposal
alternative for heavy metal sludges.

     The environmental acceptability
of hazardous wastes is partially
determined by use of the U.S. En-
vironmental Protection Agency Ex-
traction Procedure (EP) test, in
which the waste is contacted with a
mild acetic acid leachant for 24
hours, after which the leachant is
analyzed for specific metals which
may have leached.  This test proce-
dure is not valid, though, for
stabilized/solidified wastes because
the alkalinity present in the waste
quickly neutralizes all of the acid
present so that leaching occurs
under highly alkaline conditions
rather than acific ones, and because
the solidified waste is a monolithic
mass rather than in particulate
form.  These properties are bene-
ficial from the standpoint of environ-
mental protection as they greatly
decrease the tendency of the material
to leach, but.they do cause problems.
relative to determining the.actual
Teachability of:the waste.  Diffi-
culties associated with the al kali.ne
nature of the waste are addressed
elsewhere U»3»4,7}.  This paper is
concerned with, the effect of part-
icle size on leaching properties of
solidified/stabi 1i zed hazardous
wastes.
     Stabi 1 tzied/s.olIdifled wastes
are placed,in the environment as
monolithic blocks. With low •perme-
abilities.  Over.time».tfi.Qugh, these
blocks may be physi cally and chemi =
cally attacked in such a way as to
create smaller particles. It has
been:proposed that the amount of
metals, leached:should be directly
proportional to the surface area
available for leaching (5,6,8).
Early research at the University of
New Hampshire, though, found that
smaller particles leached less metals
than larger particles (2).

     The purpose of this paper is to
evaluate this phenomenon and to ex-
plain how particle size affects the
release pattern of heavy metals from
stabilized/solidified hazardous
wastes during the conduct of both
batch and column extraction tests.

APPROACH

Synthetic hazardous waste sludges wfere
prepared in the laboratory, solidified/
stabilized with type II port!and
cement, ground to specified size
ranges, and the ground material sub-
jected to various leaching tests.
Batch extractions were performed us-
ing both the U.S. EPA EP Toxicity
test procedure (9) and a modified
test developed by Bishop et al. (2).
Column extraction test procedures
were developed by the authors.

     The sludges were formulated to
contain Q.04 moles/I each of cad-
mium, chromium, and lead.  Metals
were added as chromium chloride,
cadmium nitrate and lead nitrate and
converted to thei r hydroxide precipi tate
forms in the sludge by neutralization
of the solution to pH 8.5 with sodium
hydroxide. .The sludges were mixed
with type!! Portland cement to form
cement pastes with water/cement (w/c)
                                     357

-------
ratios of Q.5 and 1.0 using ASTM
procedure C30582.  The freshly
mixed pastes were placed in 3,8 cm
diameter by 7.6 cm tall PVC cy-
linders and allowed to set for
24 to 48 hours at 100% humidity
before they were removed from
their molds and returned to the
100% humidity environment.

     Cylinders to be used in the
extraction tests were broken and
pulverized with a mortar and
pestle.  The particles were
mechanically sieved and the
resulting particle sizes were
collected for testing.  Particle
size ranges used were 4.8-9.5 mm
(large particles) and 0.3-0.6 mm
(small particles).

      Batch extraction of the
particle fractions were performed
using the U.S. EPA-EP toxicity
test and a modified version
developed in our laboratories.
The two main differences in
these tests are: 1) The EP-
toxicity test uses 0.5N acetic
acid to enhance leaching while
the modified test use 17.4 N
glacial acetic acid, and 2) the
EP-toxicity test limits the
amount of acid added to keep the
pH at 5.0 +. 0.5 to 40 ml so that
the actual pH of the leaching
medium may be well above pH 5.0,
while there was no limit to how
much 17.4 N acetic acid could be
added to keep the pH at 5.0 +
0.5 in the modified test.  TRe
samples were placed on a shaker
table in a contolled temperature
room (20 C); the pH was monitored
and adjusted over a 24 hour
period as specified in the EP-
toxicity test procedure (9).

     The samples were removed
from the shaker table after 24
hours and filtered through a
0.45 vm filter. Metal analysis
was performed on the filtered
leachate by atomic absorption
spectrophotometry.

     Column leaching tests were
performed in the upflow mode.  The
columns were 46 mm diameter by 139
mm high Buchner type funnels that
were modified by a glassblower.  The
modifications included addition of a
side arm for effluent flow and taper-
ing the neck of the funnel to a pi pet-
like tip to accomodate narrow bore
tubing.

     All leachant flows were set at
0.2 ml/min.  Acid strengths of Q.05N
and 0.1N were used to create acid
fluxes of 1.0 and 2.0 meq/g sample/
day.  Leachates from the columns
were collected in 500 ml Erlenmeyer
flasks.  Leachate was collected every
24 hours, pressure filtered through a
0.45 ym filter, and the metal concen-
trations determined. The alkalinity
of the leachate was measured with a
recording titrator.

RESULTS AND DISCUSSION

     The results of batch leaching
tests for cadmium using the modified
EP procedure are presented in Figure
1.  Leachate lead concentrations
followed similar patterns. Leachate
metal concentrations in the EPA-EP
extracts were much lower, being at or
near minimum detection limits.  The
pH of these leachates were generally
between 10 and 11, considerably
higher than in the modified EP leach-
ates.  Thus, little of the metal in
the samples subjected to the EPA-EP
test was in a soluble form.
                                     358

-------
                                       ate metal  concentrations.
 100
i
o
 O.I I	
   O.OOI
             O.OI         O.I
            PARTICLE SIZE (IncMl)
 Figure 1.  Effect of Particle Size,
 Water: Cement Ratio and Sludge Con-
 centration on Cadmium Leaching -
 Batch Tests,
     Column leaching tests were
then developed to investigate this
phenomenon further.  It was observed
that as the extractions proceeded,
the small size particles that were
exposed, to. the .'acid turned from their
natural cement-gray color to a dark
gray and then to an orangish-brown
color and finally to a white color.
By observing the boundary between the
cement gray and the dark gray, the
progression of the acid attack could
be closely monitored. The larger
particles also turned an orangish
brown, but never turned to the white
color of the smaller particles.  The
progression of acid attack in the
columns containing large particles
was faster than in columns with small
particles.  All the large particles
turned  to an orangish brown color
in  the  first two days of extraction,
while  the  color change of the small
particles  was  a slower progression  up
through  the column.
      It was initially hypothesized
 that as the particle size decreased,
 thereby increasing the surface area,
 the magnitude of metal leaching would
 increase.  Surprisingly, the reverse
 of this was found to occur.  This led
 to the question as to whether the
 heavy metals are actually "locked up"
 in the smaller particles or whether
 some other property of the smaller
 particles enhances metal binding.
 Total digestions were performed on
 representative particles of each size
 range using boiling nitric acid.
 These indicated that heavy metal
 concentrations were:the same in all
 particle sizes before leaching.  One
 explanation for these results is that
 the heavy metals are bound to the
 particles by sorption mechanisms.
 Increased surface areas occurring in
 the smaller particles would cause
 greater  ion exchange and adsorption
 of heavy metals and thus lower  leach-
      The  difference  in  behavior  of
 the  large and  small  particles may be
 explained by the difference in avail-
 able surface area.   The small particles
 have more available  surface to  leach
 alkaline  species from,  neutralizing
 the  acidity of the  leachant before  it
 reaches  the particles near the  top  of
 the  column. The large  particles, on
 the  other hand, do  not  have as much
 available surface alkalinity and the
.acid attacks the entire column without
 being completely neutralized.   In all
 cases, the time of  break through of
 the  orange front at the surface  of
 the  layer of particles  corresponded
 to the time when the pH of the  leach-
 ate  dropped below 6, and to the time
 When the ratio of total alkalinity
 leached from the particles to the
 total amount of acid passed through
 the column fell below 1.0.  This is
 the point at which the acid being
                                    359

-------
added is no longer being  com-
pletely neutralized by  the  alka-
linity of the cement paste.

     Figure 2 shows the concen-
trations of metals leached  from  the
particles as a function of  the
total amount of acid passed through
the column for one experimental
run.  The release of metals into
the leachate water leaving  the
columns containing small  particles
did not occur until after addition
of approximately 8 meq  acid/g waste.
   2,3
   Z.O
  X 0,3
      A Lor 54 porllcw,C4
      * Lorn* parlkM,Cr
      & Smol pattid»,Cd°
      O Srocfl portid»,er
               10    IS    20
               TOTAL ACID (nwq/g)
Figure 2. Metals  Leached as a
Function of Total  Acid Added -
Column Tests.

Metal concentrations  in the leachate
then increased  rapidly for a time
before falling  back to lower con-
centrations. The  rapid rise in
concentrations  corresponded to the
time when the ratio of alkalinity
leached/ meq of acid  passed through
the column fell below 1.0 (Figure
3), indicating  the time when the
alkalinity  leached from the solid
could no longer neutralize all of
the acid being  added.  The period
of declining concentrations is
probably due to a bulk diffusion
limited release, which would  pre-
dict similar declining rates  of
release as time progresses  since
it takes more time for ions to
e 0.8
                                        "0.6
                                        jO.4
                o Small partlcta— Ratio
                a Small particl*— gH
                * Lorgt partid* ~Ro1lo
                » Latg» patUctt—pH
         5     IO     19     20    28
             TOTAL ACID (m»q/g)

Figure 3. Alkalinity Leached to  Acid
Added Ratios - Column Tests.

diffuse from further within  the
particle.

     The metal leaching patterns for
large particles, shown  in  Figure 2,
were quite different than  those  for
the small particles. The release of
metals from the large particles  began
immediately and concentrations  in-
creased throughout the  test, with only
a slight decrease near  the end of the
test.  Cadmium appeared to
more readily than chromium
This rise in metal release
direct function of surface
for leaching. As alkalinity  is  leached
from the particles, more pores  are
opened, leading to additional surface
area available for acid attack  and
metals release.  Alkalinity  leached/
acid added ratios were  always below
1,0 (see Figure 3} indicating that
sufficient alkalinity  to neutralize
acid was not immediately available to
the leachant due to  the reduced surface
area provided by larger particles.  As
                            leach much
                            or lead.
                            could be a
                            area available
                                      360

-------
a result, the pH of the  leachant
remained low, allowing any metals
leached to remain in solution.

     Figure 4 shows curves of alka-
linity leached as a function of acid
          5      IO     19
           TOTAL ACID (|MD/|)
Figure 4. Alkalinity Leached  Funct-
ion of Total Acid Added - Column
Tests.

passed through the column for both
small and large particle columns.
The rate of alkalinity leaching from
the large particles dropped as the
extraction progressed, but not as
rapidly as that from the small part-
icles.  This indicates that the
available surface area for leaching
was increasing at a decreasing rate,
which corresponds with the metal
releases.

     These results are in general
agreement with those presented earl-
ier for batch leaching tests, where
it was found that smaller particles.
leached less metals than larger
particles, particularly for low
acid doses which would correspond
to conditions in the batch tests.
However, plots of cumulative metals
leached as a function of total acid
addition (Figure 5) indicate  that
eventually cumulative metals  leach-
ed from the small•particles.be-
comes greater than from the large
particles.  This generally"occurs
at the point at which the acid
flowing by the small particles is
no longer totally neutralized and
leachate pH drops.
                                              Larg* parflclf*, Cd

                                             • large pgrticlt*, Cr

                                             * Larga particle*, PS

                                              Smell ptjrllcltl, Cd

                                             o Small partlclu, Cr

                                             D Small portion*. Pb
Figure 5,  Cumulative Metal  Leached
as a Function of Total Acid  Added  -
Column Tests.

     Chromium and lead were  leached
in much higher concentrations from
the small particles than from the
large particles.  The cement mat-
rix in the small particles was
probably breaking down due to the  lack
of alkaline species available to neu-
tralize the acids containing them.   If
the chromium and lead were bound into
the cement matrix, the higher rate of
leaching from the small particles  would
be explained.  To determine  whether  the
cement matrix of the small particles was
actually being broken down at a higher
rate than for the larger particles,
representative samples were  analyz-
ed for silicon.  These analyes
showed that.the smaller particles
released three times as much sili-
con as the large particles after
the initial alkalinity was reduced.
                                     361

-------
     Figure 6 shows a pC-pH dia-
gram for hydroxides for lead, cad-
mium, and chromium,  Hicjher pC
Values mean less, metal in tfie
soluble phase since pC is the nega-
tive logarithm of the concentration.
                B
                PH
Figure 6.  pC-pH Diagram for solu-
bility of Metal Hydroxides.

The high pH of the paste as it is
being mixed (around pH 12) causes
lead and chromium to form soluble
hydroxide complexes.  Availability
of these anions in solution allows
them to participate in the formation
reactions, possibly much like
aluminum.  The insoluble cadmium hy-
droxide is not available to take part
in solution reactions and is trapped
in the pores as the solid is formed.
This could explain why Teachability
of chromium and lead appears to be
dependent on the dissolution of the
silicate matrix while release of
cadmium seems to occur as pores are
opened.  Although the curve shows
that only about 10% of the chromium
and lead orginally present would
form the anionic hydroxide complex,
more could be formed as that init-
ially present is depleted by parti-
cipating in solid formation re-
actions.  The hypothesis cannot
rule out the possiblity of an
encapsulation type reaction where
lead and chromium hydroxide are
 surrounded  by  the  silicate fibrils
.that'are  formed.   However, there is
 no  apparent reason.why,encapsulation
 would.occur for chromium and  lead
 and not for cadmium.

      The  results of the column
 tests indicate that cadmium is
 primarily bound into  the solid
 matrix  by surface  related mechanisms
 such as ion exchange  and adsorption.
.The lead  and chromium, however, are
 primarily bound into  the silicate
 matrix  of the  paste.

      The  size  of the  particles
 being leached  controls the extent
 to  which  acidic leachant is neutraliz-
 ed  in the beginning of the extract!on,
 with small  particles  having more
 alkalinity immediately available.
 When the  leachant  is  neutralized,
 metal concentrations  in the leachate
 are low.   However, when the leachate
 can no  longer  be neutralized  be-
 cause much of  the  available alka-
 linity  has been  leached, the  metal
 concentrations in  the leachate  in-
 crease.  The small particle size
 leached more metal than the larger
 size, but release  was delayed by
 the particles' ability  to  neutralize
 the leachant before  it  left the
 column  in the  early  part of the
 extraction.

      The  small particles  in the
 lower portions of  the column  began
 to dissolve early  in the  extraction,
 as buffering capacity of  these
 particles was quickly reduced.   The
 rapid reduction of buffering  capa-
 city was  due to immediate  avail-
 ability of alkaline species  from
 the  large exposed  surface  area.
 This dissolution of the small part-
 .ides.led to the release of lead
 and  chromium in higher concen-
 trations than were observed in
 the  large particle leachates  where
 alkaline species available from
 deeper within the particles pre-
                                   362

-------
vented dissolution from oecuring
as rapidly.  Final cumulative
cadmium concentrations and total
alkalinity 1 ea.ched, f rom large'and
small particles were similar. -The
similarity in alkalinity leached
indicates that approximately the
same amount of surface area Bad
been exposed to th".e leachant by
the end of the extraction.  This.
led to nearly equal amounts of
cadmium leached, since the cadmium
is primarily bound to the surface
or held as insoluble hydroxides in
the pores.

     It is apparent that batch leach-
ing tests or short duration column
tests do not provide a true indi-
cation of the leaching potential of
solidified/stabilized hazardous
wastes or of the metal binding
mechanisms present.  It is only
through the use of longterm column
leaching tests or possibly with
multiple extraction batch leaching
tests that meaningful data can be
obtained.

ACKNOWLEDGMENTS

     This research was supported in
part by the Office of Sea Grant,
National Oceanic and Atmospheric
Administration, and by New Hampshire
Division of Public Health Services,
Office of Waste Management.

REFERENCES

1.   Bishop, P. and D. Gress, 1982,
Leaching from Stabilized/Solidified
Hazardous Wastes., Proceedings, 1982
National Conference on Environmental
Engineering, Minneapolis, MN, 423-429;-

2.  Bishop, P., S. Ransom and D.
gress, 1983, Fixation Mechanisms in
Solidification/Stabilization of In-
organic Hazardous Wastes, Proceedings,
38th Purdue Industrial Waste Confer-
ence, .West Lafayette, IN, 473-480.

3.   Brown, T.:and p. Bishop, T985,
Alkal i nity;Releas.es..and _th.e Leach.i.ng
of Heavy- Metals from Stabilized/
Sol i di fied Was tes. Proceedings, Fifth
International Conference on Chemistry
for Protection of the Environment,
Leuven, Belgium.     •

4.  Brown, T.» W. Shively, P. Bishop
and D.  Gress, 1985, Use of.an Upflow
Column Leaching Test to study the Re-
lease Patterns of Heavy Metals from
Stabilized/Solidified Heavy Metal
Sludges, Proceedings, International
Symposium on Industrial and Hazard-
ous Waste, Alexandria, Egypt.

5.   Ham, R., M. Anderson, R. Steg-
mann and R. Stanforth, 1979, Back-
ground Study on the Development of
a Standard Leaching Test, EPA 600/
2-79-109, USEPA, Cincinnati,  OH.

6.   Lowenbach, W., 1978, Compilation
and Evaluation of Leaching Test Me-
thods, EPA 600/2-78-095, USEPA, Cin-
cinnati, OH.

7.   Shively, W., T. Brown, P. Bishop
and D. Gress, Heavy Metal Binding
Mechanisms in the Stabilization/
Solidification Hazardous Waste
Treatment Process 1984, Industrial
Waste, .57th Annual Conference of the
Water Pollution Control Federation,
New Orleans, LA.

8.   Thompson, D., 1979, Elutriate
Test Evaluation of Chemically
Stabilized Waste Materials, EPA
6QQ/2-79-154, USEPA, Cincinnati,, OH,

9.   U.S. Environmental Protection
Agency, T98Q, Hazardous Waste and
Consolidated.permit Regulations,
Federal Register. Vol. 45, No. 98,
33Q63-33285, May 19, I960.
The work described In this paper was not funded by the U.S. Environmental
Protection Agency.  The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
                                    363

-------
                     DEVELOPMENT OF A METHOD FOR MEASURING
                          THE FREEZE-THAW RESISTANCE OF
                          SOLIDIFIED/STABILIZED WASTES

                            P.  Hannak,  and A.  3.  Li em
                          Alberta Environmental Centre
                           Vegreville,  Alberta, Canada
                                     TOB 4LO
                                    ABSTRACT

    A description is given of the development of a method for measuring the
freeze-thaw resistance of solidified/stabilized wastes.  Various features of
the existing ASTM methods have been adopted, such as the use of weight loss as
a measure of freeze-thaw resistance, and the media and durations of the
freezing and thawing phases.  New features are also introduced, including
loose particulate removal method and the use of control specimens to isolate
the effect of freezing and thawing.

    The initial results show that a relative standard deviation (ratio of
standard deviation to mean) of better than 11% could be obtained in the
cumulative weight loss measurements.  The method was applied to a synthetic
waste treated with five different additives that are commonly used in
solidification/stabilization processes.  Differences in freeze-thaw resistance
could be readily observed, thus indicating the potential use of the method for
comparing different processes on the basis of freeze-thaw resistance.

    Modification to simplify the method are suggested, and further tests are
planned to assess its applicability to a wide variety of wastes treated by
commercially used processes.  Adoption of the method for measuring wet-dry
resistance will also be assessed.
INTRODUCTION

    Leaching of contaminants from
land-disposed  wastes is a well
recognized environmental concern.
To reduce such leaching and its
consequent potential for ground-
water contamination, solidi-
fication/stabilization (s/s) pro-
cesses are widely used to treat
wastes prior to disposal.  Typ-
ically, aqueous wastes containing
heavy metals are treated in this
manner.
    In brief, s/s processes involve
the use of additives to convert a
liquid waste into a solid mon-
olithic-matrix.  The contaminants,
which are then  contained within
such a matrix, will therefore
become less accessible to potential
leaching agents.  In some cases,
the additives are selected to also
react with and reduce the sol-
ubility of the contaminants.
                                    364

-------
Hence, an additional measure is
provided to further reduce the
potential for groundwater con-
tamination.

    Clearly it is important that
the physical integrity of such
treated wastes be maintained.
Disintegration, resulting in
generation of small particulates,
or formation of cracks, thereby
increasing the apparent per-
meability of the solid matrix,
would defeat the purpose for which
s/s processes are used.

    Such deterioration in physical
integrity could be caused by
adverse climatic conditions, such
as changes in temperature which
cause repeated freeze-thaw cycles.
Even if the treated wastes are
eventually buried under soil
layers, which would minimize such
effects, there is still an in-
termediate period during which
these wastes are exposed to such
adverse conditions.  Therefore, at
locations where freezing and
thawing occur, measurements of
ability to withstand these
conditions are an important and
integral part of s/s process
evaluation.

    This paper describes the
development of a new method for
quantifying the freeze-thaw re-
sistance of solidified wastes on
the basis of weight loss
measurements.
EXISTING METHODS

    To our knowledge, there are
only two existing methods for
measuring freeze-thaw resistance.
Both are ASTH standard methods,
developed for concrete and
soil-cement mixtures.  These are
briefly described as follows:

ASTH C 666-80. "Resistance of
Concrete.to Rapid Freezing and
Thawing" '

    The test specimens  are prisms
of 10 x 10 x 38 cm nominal di-
mensions, which are immersed in
water at 4.4°C during the thawing
phase.  For the freezing phase, one
of two procedures could be
adopted.  In Procedure A, the test
specimens remain in the water used
in the thawing phase, and in Pro-
cedure B,  they are exposed to
air.  In both cases, the freezing
phase temperature is -18°C,

    Several constraints are imposed
on the duration of the transition
period, on the rate of temperature
change during this period, and on
the minimum duration of the thawing
phase relative to that of one
freeze-thaw cycle.  However, strict
requirements are not specified for
the freezing and thawing periods.
Test specimens could be stored in a
frozen condition "indefinitely"
when interruptions occur.

    The freeze-thaw resistance is
expressed as durability factor
(OF), ranging from 0 to 100, which
is computed from measurements of
dynamic modulus of elasticity.
These measurements are carried out
at least once every 36 freeze-thaw
cycles.

    The prescribed number of
freeze-thaw cycles is 300, or that
when the dynamic modulus elasticity
reduces to 6035 of the initial
value.  The acceptable, within-
laboratory precision values are
specified in terms of standard
deviation and actual difference in
                                    365

-------
Individual OF values.  These vary
depending on the average DF: in the
range of 1-15 in standard deviation
and 2-43 in difference between test
results.

    This method was applied to
samples of Uranium mine-mill
tailings treated by a number of s/s
processes.2  Experimental dif-
ficulties were encountered, such as
surface scaling, sample dis-
integration and warping, and dif-
ficulties in carrying out
quantitative measurements were
reported.  From visual obser-
vations, however, the different
processes were ranked in terms of
freeze-thaw resistance.

ASTH D560-57 (Re-approved 1976).
"Freezingand ThawingTests of
Compacted Soi1-Cement Mixtures" 3

    The test specimens are cylind-
rical, with nominal dimensions of
10 cm diameter and 12 cm long.
Both the freezing and thawing are
carried out in air, with 100% re-
lative humidity and at temperatures
of -23°C and 21°C, respectively.
The freezing and thawing periods
are specified at 24 h and 23 h,
respectively.  As in the previous
method, during interruptions, test
specimens are stored in a frozen
condition.

    The freeze-thaw resistance is
expressed in terms of weight loss.
Measurements of weight loss involve
the application of "firm strokes"
on all surface areas of the test
specimen with a wire scratch
brush.  A firm stroke is defined to
correspond to approximately 13.3 N
(3 Ibf) as measured with a
balance.  For specimens that form
scales, a sharp-pointed instrument
such as an ice pick is to be used
instead of a brush.  These weight
loss measurements are usually made
after each freeze-thaw cycle.

    The prescribed number of
freeze-thaw cycles is 12.  No
specification is given on the
acceptable precision value for the
weight loss measurement.

    This method was applied to four
treated wastes with different types
of physical characteristics:  low-
strength concrete, rubber-like
solid, plastic-encased block and
soil-like material.4  With the
exception of one plastic-encased
sample, all samples disintegrated
after 12 freeze-thaw cycles.  In
fact, 62% of the samples disinte-
grated after 2 cycles.  The weight
loss measurement results were not
reported.

DESCRIPTION OF PROPOSED METHOD

    Various features of the
existing ASTM methods have been
adopted.  These include:

    (1)  Test specimens  are frozen
in air and thawed in water
(C666-80, Procedure B).

    (2)  The freezing and thawing
periods are 24 ± Ih, and the
maximum number of freeze-thaw
cycles is 12.  The freezing and
thawing temperatures are -20 +3°C
and 20 i 3°C respectively (0560-57).

    (3)  Weight loss is used as a
measure of freeze-thaw resistance
and measured after each cycle
(D560-57).

    Modification to the existing
methods have also been made, as
described below:
                                    366

-------
    (1)  The test specimens are 5
cm cubes.  This decrease in volume
is desirable to reduce both mat-
erial and space requirements for
sample preparation and for carrying
out the tests, respectively.  More-
over, it was originally planned to
use unconfined compression strength
(UCS) as a measure of freeze-thaw
resistance.  The standard method
for UCS measurements prescribes
test specimens of such shape and
dimensions.5

    (2)  Measurements of weight
loss are carried out on the removed
and dried particulate matters, and
not on the test specimens.  The
weight loss is then expressed on a
dry basis relative to the original
test specimen weight.  This
approach improves the precision of
the weight loss measurements, and
more importantly, eliminates the
necessity of accounting for moist-
ure changes in the test specimens
during the freeze-thaw cycles.

    (3)  "Control" specimens are
used to correct the weight loss
measurements.  These undergo iden-
tical treatment as the test speci-
mens except exposure to a low tem-
perature during the freezing
phase.  This approach has been
adopted to isolate the effect of
freezing and thawing from others
that might contribute to weight
loss, such as matrix dissolution.

    (4)  The removal of loose part-
iculates is carried out by
ultrasonic application followed by
rinsing, and not by a wire scratch
brush.  This approach has been
adopted to improve reproducibility
and to enhance differences in
weight loss, either between control
and test specimens or between
different test specimens.
    A brief description of the
procedure is as follows:
- Prepare and weigh test and
control specimens (see next
section).
- Prepare identical specimen for
measurement of moisture content and
computation of oven dry weight.
- Each specimen is placed in a
covered beaker with known tare
weight.
- Beakers containing test speci-
mens are placed in a freezing
cabinet, while control specimens
are stored in a moist container at
room temperature.
- At the end of the freezing phase
all of the beakers are filled up
with  distilled water to cover the
specimens and thawed in water.
- At the end of the thawing phase,
the specimens are exposed to
ultrasonic for 4 minutes, rinsed
with distilled water to the beakers
of origin and transferred to a new
set of beakers.  The water
containing the removed particulates
is evaporated and the dry solids
are measured.
- The cycling is repeated 12 times
unless specimens lose integrity
earlier in the process.

MATERIALS

    An aqueous solution containing
0.1 molar of As, Cd, Cr and Pb was
used as the waste.  Table I is a
summary of the types and quantities
of the additives used to treat the
waste.  These have been selected to
represent inexpensive s/s processes
that are at present widely
used,6.7 though the composition
of the matrix may not reflect the
best available s/s technology.  For
each set of treated waste specimens
a corresponding "blank™ set
(containing no heavy metal) was
                                     367

-------
 also  prepared  and tested.               Table I Summary of Additives Used

    Preparation  of test specimens,      Deslg,  Additives(b)   Ratio*3)
 Including mixing and molding, was	
 carried  out  using equipment as
 specified in the ASTH C305-828          A      Cement, Fly ash   1:0.45:2,1
 and C192-81.9  Modifications to         B      Lime, Fly ash     1:0.55:1.85
 the prescribed procedures were          C      Cement, Soluble   1:0.65:0.60
 however  made,  such as lengthening              Silicate
 the mixing time  to ensure homo-         D      Cement, Bentonite 1:0.55:0.85
 genelty.  All the test specimens        E      Gypsum            1:1
 were  cured for 28 days at room
temperature and in excess of 96%        (a) Weight ratio, waste:additive:
relative humidity.                      additive, in the order as shown in
                                        the second column.
               Table II Replicate Weigh Loss measurement results,
                   showing reproducibility of proposed method

                Cumulative relative weight loss after Indicated freeze-thaw
                cycle number (% dry basis)(a)


                1    2    3    4    5    6    7    8    9    10   11    12  Av.


                                TEST SPECIMENS**3)

Hean*c)        0.59 1.18 2.07 2.73 3.74 4.42 5.29 6.39 7.19 7.98 9.00 10.39

RSDX           0.74 7.23 13.1 7.37 4.70 8.04 12.2 16.3 13.8 11.5 12.2  7.65  9.5


                              CONTROL SPECIMENS*0)

Mean*c>        0.57 1.11 1.66 1.18 2.71 3.22 3.71 4.25 4.78 5.30 5.82 6.34

RSDX           1.24 1.42 1.84 1.09 0.61 0.85 0.40 0.39 0.17 0.13 0.07 0.20  0.7


(a)  Relative to specimen initial weight.
(b)  Gypsum process, see Table I;
(c)  Arithmetic mean of 4 measurements, specimens prepared from two batches;
     RSD-ratio of standard deviation to mean in %,
                                    368

-------
(b) Sources of materials

Cement (Portland Cement Type II) -
Genstar, Edmonton, Alberta
Flyash - Power Plant, Ontario
Soluble silicate (Type N) -
National Silicate, Toronto, Ontario
Bentonite (Voloclay No. 200) -
Western Bentonite, Edmonton, Alberta
Gypsum (Envirostone A, B and C) -
US Gypsum  Co., Libertyvilie, 111.

RESULTS AND DISCUSSION

Reproducibility

    An indication of the reprod-
ucibility of the proposed method is
shown in Table II, summarizing the
weight loss measurement results for
both the test and control specimens.

    In terms of weight loss
measurement, a relative standard
deviation of better than 2% could
be achieved, as shown by the
results for the control specimens.
Freezing and thawing, combined with
batch-to-batch variation, in-
troduced more scatter, which is not
unexpected.  A maximum  relative
standard deviation of 17.OS could
be achieved for cumulative weight
loss measurements after freeze-thaw
cycles.
    The results summarized in Table
II also show the importance of
using control specimens, especially
when matrix dissolution is con-
siderable.  By this means, the
contribution to the total weight
loss which is attributed to
freezing and thawing could be
isolated.

Applicability to Different S/S
Processes

    Figure 1 shows a comparison
amongst the five different
additives listed in Table I.  The
vertical axis represents cumulative
weight loss, relative to the
initial weight, expressed on a dry
basis and corrected for control
specimen weight loss.  The hori-
zontal axis shows the number of
freeze-thaw cycles.

    The results show marked dif-
ferences in freeze-thaw resistance
amongst the systems tested, ranging
in cumulative weight loss from 0.1%
after 12 freeze-thaw cycles to
about 3% after 5 cycles.  It is our
contention therefore, that the
proposed method provides a sen-
sitive means of comparing different
s/s processes on the basis of
freeze-thaw resistance.

    It is also interesting to note
the progression of the weight
loss.  For example, specimen  E
(Envirostone A) showed very little
weight loss up to 8 freeze-thaw
cycles, but then there was a
dramatic increase in weight loss,
followed by formation of cracks.
Figure 2 shows an example of a test
specimen  with cracks.

    It should be emphasized that
the proposed test method was not
developed for extrapolation
purposes, to predict freeze-thaw
resistance under actual field
conditions.  No consideration has
been given to the various
similarity factors that must be
accounted for in  order to develop
predictive capabilities.  Never-
theless, the proposed method could
provide a rapid indication of the
relative freeze-thaw resistance of
different s/s processes.
                                    369

-------
    34-^
         in
         o
                                              cement  & sol.  silicate
         .c
         en
         •r—
         0)



         0)
         3
         E
         3
         O
CO
*«J
o
cement &
Bentonite

                       gypsum "A1
    fig.  1
                                                                                                 .' lime flyash
    0 --

                                                                      y
                                                                      /
                                                                                                 — cement  & flyash
                                                                                        cycles
                                                                                   10
                             11
12

-------
FIGURE 2
Effect of Contaminants

    Figure 3 (Matrix C) and 4
(Matrix B)show the differences in
weight loss between test and blank
specimens that were exposed to
freeze-thaw cycles and also those
used as control specimens.  Also
shown in Figure 3 are the number of
freeze-thaw cycles at which speci-
men disintegration occurred.

    Somewhat curiously, some of the
test specimens which contained
contaminants displayed a higher
freeze-thaw resistance than the
blank specimens in which no con-
taminant was present.  It could be
speculated that contaminant-
-additive interactions or reactions
occurred, which resulted in im-
provements in freeze-thaw
resistance.  Investigation of the
possible mechanisms is, however,
beyond the scope of this paper.

Modifications and Extension of
Proposed Method

Ultrasonic Application

    A series of tests were carried
out to assess whether ultrasonic
application was necessary.  Elim-
ination of this means of removing
loose particulate matters would
simplify the method, and moreover,
concerns over standardization of
the ultrasonic equipment and its
method of operation would be
eliminated.  The results for both
test and control specimens are
summarized in Table III.

    Standard conditions were used
by centering samples immersed in
                                     371

-------
        o
        S-
        o
        o
        O)
                                                                                =»=
                                                                                tfi

                                                                                
-------
          OJ

          Q.

          at
                                                                                oo
 % sso[ 5.1)6 IBM
4	1	i	1
                                   373
                                                                             O     M~
-------
230ml water containing beakers
(type Klimax No. 1400) in an
ultrasonic bath (type Bransonic
221).

    For the control specimens a
marginal increase in weight loss
was indicated as a result of ultra-
sonic application.  However, for
the test specimens  no apparent
change was observed.  In fact, in
terms of corrected weight loss
(ie. test-control), the effect of
                            ultrasonic  application  masked  that
                            of freezing and  thawing.   It would
                            seem that when freezing and thawing
                            alone were  sufficient to  effect
                            particulate removal,  ultrasonic
                            application produced  little
                            additional  effect.

                                For the above reason  and also
                            for simplicity,  ultrasonic
                            application is judged unnecessary
                            and hence will be abandoned in
                            future tests.
Duration
(Minutes)
Table III   Effect of Duration of Ultrasonic Application


                                           Control	
           Weight Loss, %  (a)
  Test Specimen (b)   	
Mean(c)
                                            Range
                                             Meantc)
SP(c)
0
4
8
0.55-0.80
0.55-0.80
0.60-0.80
0.68
0.70
0.70
0.08
0.08
0.08
0.45-0.85
0.60-0.80
0.65-0.80
0.61
0.66
0.70
0.12
0.07
0.05
(a)  Relative to initial weight, dry basis
(b)  Gypsum system E, see Table I
(c)  Computed arithmetic mean and standard deviation (SD) of weight loss
     measurements over 12 freeze-thaw cycle.
Test Specimen Dimensions

    As previously mentioned, 5 cm
cubes were used since it was
originally planned to incorporate UCS
measurements to characterize freeze-
-thaw resistance.  UCS measurements
are destructive, hence a large number
of specimens need to be prepared.
Because of this disadvantage, and
furthermore, because weight loss
measurements are sufficient to
characterize freeze-thaw
resistance, in future tests cylin-
drical specimens, 4.4 cm  diameter and
7.4 cm long, will be used.  These
specimens can be prepared using
readily available plastic vials.  A
                            further advantage is thus  gained  in
                            that these vials are considerably less
                            costly than the metalic molds  required
                            to prepare cubical  specimens.

                            Future Plans

                                A cooperative project, involving
                            US EPA, Environment Canada,  the
                            Alberta Environmental Centre and
                            vendors of s/s processes,  will  be
                            undertaken to assess the applicability
                            of various short-term laboratory  test
                            methods.10  The proposed test  method
                            is included in this project.  Thus,
                            further information will be  gathered
                            on its applicability to a  wide variety
                            of wastes, including "real wastes",
                            treated by commercially used s/s
                                     374
-------
processes.  Information on inter-
-laboratory reproducibllity will also
be pursued by conducting a round-robin
study under of ASTM D-34.  Development
of the method to an ASTM standard
level will then be considered.

    Also included in the co-operative
project is a test method for measuring
wet-dry resistance.  The approach used
in the proposed method for measuring
freeze-thaw resistance will be
adopted:  the use of weight loss as a
measure of wet-dry resistance, control
specimens to correct test specimen
weight loss measurements, and the same
particulate removal method.  The
wetting and drying phases will be
carried out at 20 ± 3°C, 60±3°C for 24
hrs and 23 hrs, respectively.  The
weight loss measurements  will be
carried out after each complete
dry-wet cycle.
CONCLUSIONS

    The development of a method for
measuring the freeze-thaw resistance
of solidified/stabilization wastes has
been described.  Novel features are
introduced, including methods for
particulate removal and weight loss
measurement, and the use of control
specimens to isolate the effect of
freezing and thawing.

    The initial results obtained show
promise in terms of reproducibility
and ability to detect differences
amongst different s/s processes.
Various modifications have been
introduced to simplify the method, and
further tests are planned to assess
the applicability of the method to a
wide variety of soldified/stabilized
wastes and to obtain inter-laboratroy
reproducibility values.

    Adoption of the proposed method
for measuring wet-dry resistance will
also be assessed.
ACKNOWLEDGMENTS

    The authors  greatfully acknow-
ledge the scientific support of
Environment Canada Wastewater
Technology Centre, and personally Mr.
Pierre Cote.

    Obtaining data on gypsum became
possible by the assistance of
Mr. Ken Watkins and the US Gypsum Co.

REFERENCES

1.  ASTM  C 666-80.  American Society
    for Testing and Materials, Annual
    Book of ASTM standards, Part 14,
    1980.

2.  Bruce R. B., at al.  Physical and
    Chemical Properties of Chemically
    Fixed uranium Mine-Mill Tailings,
    Ontario Research Foundation Final
    Report, May, 1981.

3.  ASTM  D 560-57.  American Society
    or Testing and Materials Annual
    Book of ASTM Standards, Part 19,
    1980.

4.  Physical and Engineering
    Properties of Hazardous Industrial
    Wastes and Sludges,
    EPA-600/2-77-139, August, 1977.
    US EPA Cincinnati, Ohio  45268.

5.  ASTM  C 109—80  American Society
    for Testing and Materials, Annual
    Book of ASTM Standards, Part 14,
    1980.

6.  Cote, Pierre and Donald Hamilton.
    Leachability Comparison of Four
    Hazardous Waste Solidification
    Processes, presented at 38th
    Industrial Waste Conference
    May 10-12, 1982.
                                    375
-------
EPA 600/2-79-056, July, 1979.       9.  ASTM  C 192-81.   American Society
Survey of Soldification/-               for Testing and  Materials, Annual
Stabilization Technology for            Book of ASTM Standards, 1981.
Hazardous Industrial Wastes, by US
Army Engineer Wasteways Experiment  10. "Proceedings of  The Environmental
Station, Vicksburg, MS  39180.          Assessment of Waste Stabilization-
                                        /Solidification  Workshop."  Nov.
ASTM  C 305-82.  American Society       21-22, 1983, Alberta Environmental
for Testing and Materials, Annual       Centre, Vegreville, Alberta (in
Book of ASTM Standards, 1982.           press).
                             Disclaimer

               The work described in this paper was
               not funded by the U.S. Environmental
               Protection Agency.  The contents do
               not necessarily reflect the views of
               the Agency and no official endorse-
               ment should be inferred.
                                376
-------
                              PARTITIONING ANALYSIS
                              OF CHEMICAL SUBSTANCES
                   AS A TOOL FOR MANAGING HAZARDOUS WASTE STUDIES

                        Surya S. Prasad and James S, Whang
                                    AEPCO, Inc.
                                Bethesda, MO 20816
                                     ABSTRACT
     Partitioning  analysis  schemes  were applied to determine distribution of repre-
sentative chemical pollutants within water  to  air and soil  to air pathways.  Esta-
blished theoretical and/or empirical mathematical representations and models;  and
available  chemical  property data were used.   Information  elements  for the schemes
included the  physical, chemical, biological,  and environmental  characteristics of the
chemicals and  the  partition  coefficients.   Relative  fitting of the parameters into
the existing mathematical representations or models was assessed.  Certain data
inadequacies  were taken into  account.

     The findings  suggest that within hazardous waste  management  studies, partition-
ing analysis  is useful  in:

     o Identifying the  predominant  environmental  compartment(s) where a given chemi-
       cal  will reside.
     o Assisting in the design of optimal  field sampling programs.
     o Estimating half-life and  mobility, degree of bioaccurnulation,  residence time
       for  degradation  and  transformation, time of risk exposure,  and  toxic  effects.
     o Establishing bases for the selection of remedial measures.
     o Developing  representative conditions for laboratory or pilot-scale treatabili-
       ty studies.
     o Providing guidance for designing  study protocols.
     o Assisting engineers  in designing remedial disposal  methods  minimizing adverse
       environmental  impacts.
     o Providing guidance in the formulation of environmental regulations.
INTRODUCTION AND PURPOSE

  After a pollutant is released  into the
environment, it is distributed  chiefly
among  three compartments:   air (atmos-
pheric),  water (aquatic), and  soil  (ter-
restrial).   The concentration of a che-
mical present initially  in each compart-
ment is  a function of both  its  proper-
ties and type of release. Once  present
in the environment, the chemical under-
goes dynamic changes which are a fun-
ction  of intercompartmental  transfers,
transformations, and degradations.  Lf
the pollutant is persistent, the poten-
tial  for eventual  uptake by aquatic and
terrestrial  organisms  increases.  Conse-
quently through  biomagnification,  it is
                                       377
-------
possible for concentrations  in  biologi-
cal tissues to reach a  level, which may
cause  significant toxicological  hazard
to humans and other populations.

  Health and  environmental  hazard or
risk assessments can be initiated once
the pollutant-favored environmental com-
partment is  delineated.   One of the most
difficult tasks in performing  these risk
assessments is adequate screening and
testing of  large numbers  of  chemicals
under stringent time  and budgetary con-
straints.   To alleviate this problem,
partitioning  analysis  can   assist  in
focusing limited  resources to key media
and elements of the risk assessments.

  Water-to-alr  and  soil-to-water  parti-
tioning schemes involve the use of ma-
thematical  representations  or  models
that are predominantly based on chemical
partitioning  properties  such as  the
molecular  formula, structure,  and
weight; specific gravity and density;
vapor pressure and density;  solubility
in water and certain organic solvents;
octanol/water partition  coefficient;
soil adsorption  coefficient;  half-lives
related to  evaporation,  photolysis,
hydrolysis, and other physical, chemi-
cal, and biochemical  reactions.  Predic-
tive mathematical models based  on the
above properties  can  either be  obtained
from  the  available literature  or be
specifically developed.

APPROACH

The following general categories  of the
potential  transformation and transport
pathways were studied.
  o Intercompartmental transfers
  o Transformation,  bioaccumulation,
    biomagnification
  o Environmental  degradations
        Each environmental  pathway  was  consi-
       dered separately as a function  of  the
       physical  and chemical properties of  a
       given  chemical pollutant in relation to
       the  intercompartmental  transfer (the
       transport and redistribution processes
       among the environmental solid,  liquid
       and  gaseous  phases.)

       Hater-to-AIr Models

        The  two models employed  in this phase
       of the study were originally proposed by
       Uilling (Model I,  1977)  and Mackay  and
       Leinonen  (Model  II, 1975).
       Model  I
         H =
        Kl «
               16.U4 x P x M
                   T x S
                    221.1
              (1.042/H + 100) X M0*5

       t1/2 =  (0.6931) x d / K,

       where,
         H   =  Henry's law  constant (dimension-
              less)
         P   =  Vapor pressure (mm Hg)
         M   =  Molecular weight (gm/mole)
         T   =  Absolute temperature  (°K)
         S   =  Solubility in water (mg/L)
         Ki  =  Overall  liquid exchange constant
              (cm/min)
       ti/2 =  Evaporation half-life (min)
         d   =  Solution depth  (cm)
and
       Model II
          H =
                VP
                ws
                                      378
-------
n/2 '

where3
  H  •

  VP =
  WS *
 kiLs
 kiG
 MW  -
1/2  *

 R   -

 T   .
        U.2 x (MU of C02)
                         0.5
        (MW of Compound)0*5
        30 x (MW of H2U)

        (MW of Compound)
                        0.5
                        0.5
K1L
              H  x
             x L
                      /(R x T)
Henry's  law constant (atm-m^/
mole)
Vapor pressure  (atm)
Water solubility (moles/gm)
Liquid film  mass transfer coeffi-
cient (m/hr)
Gas  film  mass transfer  coeffi-
cient (m/hr)
Overall  liquid  coefficient (m/hr)
Molecular weight (gm/rnole)
Volatilization  half-life  (hr)
Liquid depth  (assumed 1 m)
Gas constant  =  8.2 x 10"6 (m3-
atm/tnole-°K)
Absolute temperature  (°K)
  Water-to-air partitioning was repre-
sented by the  half-life of a chemical  in
a water  body  of given depth.   It was
assumed that the water depth was 1  m and
the water temperature was at 20°C.   The
partitioning procedures were:

  1. Obtain vapor  pressure  and water
     solubility data from reliable  sour-
     ces.
  2
  3.
     Calculate  Henry's law constant.
     Apply these values to the Oil I ing
     or Mackay  and  Leinonen models to
     calculate  coefficients/constants
     and the ha If- life.
                                           Soil-to-Water Models

                                             The two models adopted for this phase
                                           of the study  were originally proposed by
                                           Kenaga  and  Goring (Model  I,  1980)  and
                                           Briggs (Model II,  1973).
                                           Model I
                                            Log K
                                                 oc
                                             where,
                                                Koc
                                                 WS «
                                                     3.64  -  0.55 x (Log WS) ± 1.23
                                                     order of magnitude
                                                     [95% confidence level and
                                                     correlation coefficient of
                                                     0.84]

                                                     Soil/water partition coeffi-
                                                     cient per unit  organic  matter
                                                     (dimensionless)
                                                     Water solubility in (mg/L)
                                           Model II
                                              Log Q = (0.524 x  Loy  P) + 0.618
                                              where,
                                                  Q
                                                  P =
                                                      Soil organic  matter/water
                                                      partition coefficient (dimen-
                                                      sionless)
                                                      Octanol-water partition coef-
                                                      ficient (dimensionless)
                                    For the definition of soil adsorption
                                    capabilities, we adopted KQC (Kenaga  and
                                    Goring,  1980) and Q  values  (Briggs,
                                    1973).   We assumed that the soil/water
                                    matrix  was at an equilibrium temperature
                                    of 20°C.   The procedures used for  calcu-
                                    lating  KQC and Q were:

                                      1.  Obtain water solubility data from
                                         rel1able  sources.
                                      2.  Calculate the relevant  transfer
                                         coefficient(s) in  the Kenaga  and
                                         Goring's and Briggs1 models.
                                      3.  Calculate KQC  and Q values.

                                       Mobility of a given  compound  from  soil
                                     to water  (e.g., by leaching) is a  fun-
                                       379
-------
ction of soil  adsorption coefficents.
The mobility  rating scheme proposed by
Kenaga (1980) is as follows:

  o Immobile (KQC  >  1,000)
  o Moderately  to highly mobile
    (Koc <  1UU)

  The mobility  rating scheme proposed by
Brlggs is as follows:

  o Immobile (Q >  398)
  o Low (398 >  Q >74)
  o Intermediate (74 > q > 29)
  o Mobile (29 > Q > 4.5)
  o Very mobile (Q < 4.b)

PROBLEMS ENCOUNTERED

  The  applicability  of  the proposed
partitioning  schemes  was  generally
limited by the  following;

  o They are less effective for chemical
    pollutants  having  inadequate  or in-
    complete data.
  o The  analysis was  based on  purely
    empirical  data and detailed experi-
    mental  protocols  were  not esta-
    blished.
  o The schemes were generally  appli-
    cable to known  and pure compounds
    only.
  o The schemes served  as  a  guide to the
    partitioning character  of the indi-
    vidual  compounds studied  individual-
    ly but  yielded no information on the
    environmental  partitioning of heter-
    ogeneous groups of toxic  chemicals.

  Furthermore, information was  inade-
quate on the environmental transforma-
tion or transport  pathways  of pollutants
for which  mathematical models are not
readily available.   Some of the availa-
ble models are formulated  in  terms of
kinetic parameters rather than  simple
physical and chemical  data.   Application
 of these  models  is  limited  because they
 lack published  kinetic data for  most
 compounds.   Availability  of  relevant
 field data  are  scarce  for environmental
 partitioning  scheme  development  and
 validation.
RESULTS

Hater-to-Air Partitioning

  The results  of  calculations for the
selected  compounds are  summarized in
Table 1.   The calculated half-life
values range from 4.8 to 14.U hours for
all compounds,  except  for aniline, iso-
phorone, and phenol having a predicted
half-life  greater  than  382.2  hours.
Figures  1,  2,   and  3  are plots of the
results.   The following  conclusions can
be generalized from Table 1 and these
plots:

1. Effects of  vapor pressure  on water
   solubility and  half-life (Figure  1):

   o Water  solubility increases  with
     increasing  vapor pressure
   o Half-life  decreases  with increasing
     vapor  pressure

2. Effects  of water  solubility on half-
   life  (Figure  2):

   o Half-life decreases  with increasing
     water  solubility

3. Effects  of molecular weight on vapor
   pressure, water solubility, and half-
   life  (Figure  3):

   o Vapor  pressure and water solubility
     decrease with increasing molecular
     weight
   o Half-life increases  with increasing
     molecular weight
                                       380
-------
              Table 1.   Summary of  Water-to-Air  Partitioning  Analysis
MOOCL i
(OMIIng, 1977)
IIPUT PARAMETERS*
VAPOR PRESSURE
COfCUND
BEKENE
TOLUENE
CHLOROFOW
NOWCHLOR08ENZENE
TRICHLOROETHYLENE
t,5-OICHl«OBENZENe
TETHACHI.OROETHYLENE
1 ,2-DHXLOROeEKZEKE
l,4~0!CHLO«Qe£NZE«E
I,2,4-TRIO4LOR08£NZENE
1 , 2,4 , S-TETRACHLOROBt«ENE
MEXAOILOHOCrCLOPEHTAO 1 EHE
•NILINE
ISQPHORONE
PMEWL
'Nacuy ana L»Jnon»n (1975
{mm Ho)
95. ZO*
28.40°
200.00*"
8.80C
74.00°
l.89d
15.80"
1.00=
0.60C
0.2911
0.05d
0.08'
i.oo"
0.20^
1 .00^
Jj filling.
(»"•>
1.25E-Q1
3. 74E-02
2.63E-01
1.16E-02
9.74E-02
2.4M-03
2.08E-02
1.S2E-03
7.89E-04
3.82E-04
6.56E-OS
I.05E-04
1.32E-03
2.63E-04
I.52E-05
MOLECULAR
•EIGHT
10«/rol.l
78.11
92.15
119.39
112.56
131.39
147.01
165.82
147.01
147.01
181.45
215.90
272.77
93. 1J
138.21
94.11
_ _n_
•t «l. (HJ5H ev»r»eli
WATER SOLUBILITY
(1I9/L1
1.78«>03*
5.15E*02°
7.9SE*03b
5.00f«02!:
I.IOEKIS"
1 .23£*02e
4.00E»02'>
1 ,OOE*02e
7.90£»0l'!
3.00E*OI*
6.00E-KJO*
1.80E«CO°
3.6M»04*
f .20E-*O4
8.20E»(X*
u»r«n I197I); '
l.ol. /mJ)
22.79
5.39
66.S9
4.44
8.37
0.84
2.41
0.68
0.54
0.17
0.05
0.01
293.00
86.82
•71.J2
3R.«.r«=,
HEWW'S
CONSTANT
Kit)
2.29E-01
2.78E-01
I.64E-01
1. 081-01
4.84E-01
1.24E-OI
3.58E-01
8.05E-02
6.IIE-02
9.60E-02
9.84E-02
6.64E-01
1.39E-04
I.26E-04
«.2»6-05
1 tt«B8); *K»
LIQUID
EXCHANGE
CONSTANT MALF-
«, LIFE
met. 11
(MacKav *nd L*lnon*n
HE MIT'S
CONSTANT
(c*/nln) (hr) /f«ol>
0.239
0.222
0.190
0.190
0.189
0.1(8
0.167
0.161
0.156
0.141
0.156
0.132
0.003
0.002
0.001
naga and
4.8
'.2
6.1
6.1
6.!
6.9
6.9
'.2
7.4
'.8
8. 5
8.8
382.2
513.7
B4S.7
5.50E-05
6.68E-03
3.9SE-03
2.61E-03
l.l'E-02
2 97E-03
8.62E-03
1 .9JE-03
I.47E-OS
2.31E-05
2.J7E-03
I.60E-02
3.35f-0«
3.05E-06
1.51E-06
Borlnj CI980); 'll-tin
OKEHALL
TRANSFER
OOEFF.
(B/BD
0,144
0.133
0.114
0.114
0.113
O.I 01
0.100
0.0»7
0.093
0.089
0.082
O.OJ9
0.002
0.001
0.00 1
S196J>! 'o
. 19751

H»LF-
LIFE
inn
4.8
5.2
6.1
6.1
£.1
6.9
6.9
7.2
7.4
7.8
8.3
8.8
382.3
313,8
845.9
,,„„,„
(1979); "Sax (1979);  Ha»l»y (19771;
                             (1980); and frowning (1965).
   2-
   1-
   -1.5      -0.5  0   0.5       1.5
           Log (Vapor Pressure, ran Hg)
                                      2.5
      D MS  « Hater Solubility (rag/L)
      f ti,  " Half-Life {hr)
      ^ KQC (Dimensionless)

Figure 1.   Effects of Vapor Pressure
             (VP)  on Water Solubility,
             Half-Life, and
                                                         8-
                                                         7-
                                                         6-
                                                         5-
                                                         4-
                                                         3-
                                                         2-
                                                          0123
                                                                Log (Water Solubility, mg/L)

                                                           D tij  » Half-Life (hr)
                                                           "*" 'Sec (Dimension!ess)
                                                      Figure  2.  Effects  of Water Solubility
                                                                  on Half-Life  and
                                               381
-------
    9-

    8-

    7-
 o.

 ?  ^
 «J
 u  c,
 o  3

 sg  4'

 -•  3


 3  2-
 «JI  1,


 £.  0-

    -!•
    -2.
     70   100     150     200     250
            Molecular Height (gin/mole)
       0 t, • Half-Life (hr)
       + K3C (Olmensfonless)
       $ HS « Hater Solubility (ng/L)
       A VP • Vapor Pressure  {ran Hg)
                                280
Figure 3.  Effects of Molecular Weight
          on Half-Life (tj,),  Kg,.,
          Water Solubility and va
          Pressure
                               apor
  These findings  are most pronounced for
the chlorinated  benzenes, probably due
to their similarity in molecular struc-
ture.   Extrapolating  this  finding,  par-
titioning  analysis  may  be useful  in
predicting  the volatilization potential
of exotic compounds of unknown physical
properties by  drawing  analogies  from
closely  similar compounds within the
same family.   An increase  in chlorine
atoms per molecule causes an increase in
molecular  weight, a decrease  in vapor
pressure, a decrease in water solublity,
and  an  increase  in  half-life.   Com-
pounds of  higher molecular weight and
longer  half-life  tend to favor the aqua-
tic over the atmospheric compartment.
Other  observations, awaiting further
confirmation,  are that the structural
positions of chlorine  atoms  on  the  ben-
zene rings  somehow dictate the values of
 vapor pressure and water solubility, and
 hence  the  half -life  of the compound
 within the chlorinated benzene family.

  Aniline,  isophorone, and  phenol  nave
 vapor pressure values  very close to that
 of most compounds listed in  Table 1, but
 they also have  nigh water solubilities.
 The  latter  property  hinders the  vola-
 tilization process.

  Focusing on  the volatilization poten-
 tial or  half-life  of various  benzene
 derivatives, some  guidance information
 relevant to  hazardous  waste  management
 can be preliminarily deduced.  For exam-
 ple, high volatile  substances, which
 upon their  release  to the environment
 favor air pathways,  are more amenable to
 removal by air or steam stripping.  Fur-
 thermore, half- life data can be  used to
 estimate the emission  strength  of sour-
 ces and to assess their  risk potential
 within  hazadous waste sites or produc-
 tion facilities.

 So i 1 -to- Mate r Part i t i on i n g

  The results of the calculations of the
 soil adsorption capabilities  for  the
 compounds studied  are  summarized  in
 Table  2.    The calculated  KQC  values
 range from 9 to 3,lb9.  The calculated Q
 values  range from 5  to 1,162.  In gener-
 al, KQC and Q values and their  respec-
tive mobility  ratings agree with each
other very  well.  Figures 1, 2, and  3
contain plots  of results for all com-
pounds except for aniline, isophorone,
and phenol.  The following conclusions
can  be  generalized  from Table 2  and
these figures:
1.  KQC decreases with increasing  water
              and  vapor pressure (Fi-
     solubility
     gures 1 and
                                                              vapor  pressure
                                                          2, respectively)
                                       382
-------
                Table 2.   Summary of Soil-to-Air Partitioning Analysis
                          MODEL I (Kenaga and Goring, I960).
               HQOB. II, (Brlgas, 1973)


COMPOUND
PHENOL
ACRYLQNITRIIE
ANILINE
1 SOPHORONE
CHLOROFORM
BENZENE
raiCHLOROETHYLENE
TOLUENE
HONOCMLOROBENZENE
TETRACHLORQETHYLENE
,3-OICHLOR06ENZENE
,2-OICHLOROBENZENE
,4-OICHLCROBENZEN£
, 2 ,4 -TR 1 CHLOROBENZENE
,2,4, 5-TETRACHLOR08ENZENE
HEXACHLOROCYCLOPENTAD 1 ENE
WATER
SOLUBILITY
(mq/L!
8.20E«4a
7.50E»04°
3.66£+04a
1.20E+04C
7.95E+05d
I.7BE+038
I.IOE«3d
5. I5£*Q28
S.OOE+02f
• 4.00E+0211
I.23E-HJ2*
I.OOE+02f
7.90E+OI*
3.00E+OIa
6.00E+008
I.BOEtOO9
SOIL
ADSORPTION
COEFFICIENT
Koo
(dimension- RELATIVE
less)
9
9
U
25
31
71
93
141
143
162
309
347
395
672
1,629
3, (59
MOBILITY*
Highly mobile
Highly mobile
Highly mobile
Highly mobile
Highly mobi la
Moderately mobile
Moderately mob Me
Moderate! y mobi *e
Moderately mobl U-
HoderaTety mobile
Moderately nubl 1*
Moderately mobile
Moderately mobile
Imtobl le
Innmobi le
1 fflmob 1 1 e
OCTANOL-WATER SOIL
PARTITION ADSORPTION
COEFFICIENT COEFFICIENT
Log P 0

(dimension- (dimension- RELATIVE
less) less)
0.23*
0.26«
0.5I«
2,26"
1.90«
2,13"
l.72«
2.2t«
5
6
8
63
41
54
33
60
2. 13* 58
2.W" 51
3.44* 263
3.40* 251
3.39* 246
4.17* 635
4,67* 1,162
3.99* 511
MOBILITY*"
Mobile
Mobile
Mobile
n termed late
fitermedlate
ntermad 1 ate
n termed late
n termed late
ntarfflediate
n termed la to
LOM
Lo«
LOW
irmoblle
Imioobl le
Imobl le
      * Based on the mobility rating method proposed by Kanaga, 1980.
      •• Calculated using the equation suggested by Kanaga and Goring, 1980.
      * Values reported by Hanscn and Leo, 1979.
     *** Based on the rnobltlty rating method proposed by Briggs, 1973.
      °Kanaga and Goring (1980); bKenaga U9aO); qBro«tn§ (1965)i tfOI 11 Ing et al. (1975); 8Mackay and Lelnonen (1975);
      fverscnueren (1977);  and 9Callanan (1979).
  2. KQC increases with increasing mole-
     cular weight  (Figure  3)
  These general findings  are also most
consistent for  the family  of  chlorinated
benzenes, probably  due to their simi-
larity in molecular  structure.   Extrapo-
lating tnis finding, partitioning analy-
sis  may  be useful  in  predicting the
mobility  of exotic  compounds of  unknown
physical  properties  by drawing  analogies
from  closely similar compounds within
the same family.  An  increase  in chlo-
rine  atoms per molecule causes a de-
crease in water solubility (which corre-
lates well with an increase in  molecular
weight), an increase  in KQC  or Q, and a
decrease in mobility from  aquatic  to
terrestrial compartments.   Compounds  of
higher KOC or  Q tend to  resist leaching
and stay  in the terrestrial compartment.
Other tentative observations,  awaiting
further  confirmation,  are  that the
structural positions  of chlorine atoms
on the benzene rings somehow  dictate the
values of  water solubility,  and,  hence,
the mobility  of the compound within the
chlorinated benzene family.

  Regarding guidance information for the
management of hazardous waste,  prelimi-
nary  findings  suggest that  compounds
having high mobility tend to escape from
their sources of release by leaching and
to contaminate surface  water  and ground-
water systems.  These compounds  are more
effectively recovered  from water media
(e.g.,  via groundwater pumping  or lea-
chate  interception)  for central onsite
or offsite treamtent.
                                        383
-------
  Compounds  having low mobility tend to
remain at  the source,  potentially
causing bioaccumulation and biomagnifi-
cation problems.  However, these com-
pounds  can  be effectively removed by
exacavation  or controlled  by  source
containment  and  encapsulation.

ACKNOWLEDGEMENTS

  Information  in this paper was obtained
in part from a study project  sponsored
by the United  States Environmental Pro-
tection Agency.

REFERENCES

 1. Briggs,  G.6., 1973, A  simple rela-
    tionship  between soil adsorption
    coefficients.  Proceedings 7th Bri-
    tish insecticide  and Fungicide Con-
    ference, pp. 83-86.

 2. Browning  E.,  1966,  Toxicity  and
    metabolism  of  industrial solvents.
    New York,  NY:  Elsevier Publishing
    Co.

 3. Callahan,  M.A.,  H.W.  Slimak,  N.W.
    Gabel, et  al., 1979, Water-related
    environmental fate of 129 priority
    pollutants: II. Halogenated  alipha-
    tic  hydrocarbons,   halogenated
    ethers, monocyclic aromatics,  phtha-
    late  esters,  polycyclic aromatic
    hydrocarbons, nitrosamines,  miscel-
    laneous compounds.  Monitoring and
    Data  Support  Division,   Office of
    Water Planning  and Standards, U.S.
    EPA, Washington,  D.C.,  EPA-44U/4-79-
    029b.
 4, Oil ling,  W.L.,  N.8. Tefertiller and
    G.J. Dallos, 1975, Evaporation rates
    and  reactivities of  methylene chlo-
    ride, chloroform, 1,1,1-trichloro-
    ethane, trichloroethylene,  tetra-
    chloroethylene, and other chlori-
    nated compounds in dilute  aqueous
    solutions.   Environ.  Sci.  Technol.
    9:833-B3B.

 b.  Dill ing W.L.,  1977,  Interphase tran-
    sfer processes.   II.  Evaporation
    rates of  chloro methanes,  ethanes,
    ethylenes, propanes, and  propylenes
    from dilute aqueous solutions.  Com-
    parisons  with theoretical  predic-
    tions.  Environ.  Science &  Technol.
    11:4U5-4U9.

 6.  Hansch, C. and A, Leo,  1979,  Substi-
    tuent constants for  correlation
 .   analysis  in chemistry and  biology.
    New  York,  NY:   John Wiley and Sons.
 7.  Hawley, G.G,  (Editor), 1977, Conden-
    sed  chemical  dictionary, 9th ed. New
    York,  NY:   Van Nostrand Reinhold Co.

 8.  Irish, D.D.,  1963, Hydrogenated
    hydrocarbons: hexachlorocyclopenta-
    diene.   In:  Patty F.A.  (Editor),
    Industrial  hygiene and toxicology,
    2nd  rev. ed., New York: Johy Wiley &
    Sons, Inc.  pp. 1333-1363.

 9.  Kenaga E.E.,  1980,  Predicted biocon-
    centration factors  and soil  sorption
    coefficients  of pesticides and other
    chemicals.   Ecotox.  Environ. Safe-
    ty.  4:26-.38.

10.  Kenaga E.E. and C.A.I.  Goring, 1980,
    Relationship  between water  solubil-
    ity, soil sorption, octanol/water
    partitioning, and  bioconcentration
    of chemicals  in biota.   In:  Eaton
    JC,  Parrish  PR,  Hendricks AC, eds.
    American  Society  for Testing and
    Materials.  ASTM STP  7U7.   pp. 78-
    115.

11.  Mackey I), and P.J. Leinonen, 197b,
    Rate of evaporation of low-solubil-
    ity  contaminants  from water bodies
                                      384
-------
    to atmosphere. Environ.  Science &
    Technology 9(13): 1178-1180.

12.  Richardson,  L.T.,  1968,  Selective
    vapor phase activity  of chloronitro-
    and  chlorobenzene in  soil.  Phytopa-
    thology.   58:316-322.

13.  Sax, N.I., 1979, Dangerous  proper-
    ties of industrial  materials. 4th
    ed.  New  York, NY:   Van  Nostrand
    Reinhold  Co.

14.  U.S. Environmental  Protection Agen-
    cy,  1984,  Effluent Guidelines Divi-
    sion.  Combined Sewer Overflow Toxic
    Pollutant Study.    EPA  440/1  -
    84/304. April   19b4.   p.  38. yashing-
    ton, O.C.

15.  Vershueren,  K., 1977,  Handbook of
    environmental data on organic chemi-
    cals.   New York, NY:  Van Nostrand
    Reinhold  Co.
                                     Disclaimer

                       This paper has been reviewed in
                       accordance with the U.S. Environ-
                       mental Protection Agency peer and
                       administrative review policies and
                       approved for presentation and publi-
                       cation.
                                    385
-------
                      RECYCLING AND CLEANER TECHNOLOGY
                  AS A MEANS  OF HAZARDOUS WASTE MANAGEMENT

                             Dr. Klaus Muller
                National  Agency of Environmental Protection
                           Copenhagen, Denmark


                                 ABSTRACT

    Increasing amounts of wastes, new  hazards,  new types of wastes,  and
limited possibilities for disposal or  incineration  of hazardous wastes create
the need for new ways in hazardous waste management.

    Great progress  has  been made in the  development of new technologies  for
treating hazardous wastes, but the so-called "end-point-solutions" will not be
sufficient to solve the problem.

    Recycling and cleaner technologies are presented as new concepts and means
for hazardous waste management.

    Recovery methods  as a part of recycling activities  have great importance
and are demonstrated by some examples.

    Cleaner technologies  try  to avoid  hazardous waste problems by regarding
the whole  production and product-life  cycle.  Problems  and examples  of  the
diffusion of cleaner technoloiges are discussed.

    Denmark has established  a system for promoting recycling and  cleaner
technologies acitivites; some  experiences are presented.
                                    accept  new industrial installations  at
INTRODUCTION                         various  distances  from the  homes.
                                    Source:/U.S. Council.../)
    Hazardous  wastes  are the  new
challenge of our time—all over the      Figure 1 shows  the lowest percentage
world.    We  are   all  aware  of  of people willing to accept new disposal
increasing  amounts  of wastes;  sites for  hazardous  waste  chemicals  in
household  wastes,   post-consumer  their neighborhood—even  nuclear  power
wastes,   industrial   wastes,  plants  are  not as  threatening as  those
agricultural  wastes and—last,  but  disposal sites!   It is, furthermore,
not least—hazardous wastes.   This  becoming more and more difficult—at
is a threat to all  nations in that,  least in Europe — to  find suitable  and
not only are the amounts of  wastes  accepted disposal  sites for all kinds of
increasing,   but  also  the'wastes.    We  are  confronted  with
dangerousness  and  the   risk,  increasing  amounts  of hazardous  wastes
especially of  hazardous  wastes,  and we need new solutions for hazardous
People feel this and, as  an example  waste management!
of their opinion and their feelings,
I will  show a  result of  a national      in  the  future  we will have  to deal
opinion survey which demonstrates  with new production methods,  with  new
the need for active work to find new  products, and  with new environmental
ways  of  dealing  with  hazardous  problems:  this strengthens the need not
wastes.  (See  Figure  1:   Cumulative  only for new technologies, but even more
percentage of  people  willing   to  for  new  strategies  and concepts  for

                                  386
-------
treating  hazardous wastes—and
recycling and  cleaner  technologies
are the  approaches for solving these
problems in the future!

Figure 1.   Cumulative percentage  of
people   willing  to accept  new
industrial  installations  at  various
distances from their  homes.
         Hazardous Waste Treatment

         Hazardous waste is  a very  wide  and
         unprecise term—it covers generally a-11
         wastes which  are  regarded as  toxic,
         reactive,  corrosive  or  radioactive.
         Just the fact that you  can find two such
         divergent numbers as  150 million tons
         versus 67 million tons, as the amount of
          100
           10
         Wouldn't Less
          m»tt*r  Than
         on* way one
          or the  mile
          other
*-S »-9 1O- IS- 20- 30- 4O- SO 51-  100 1O1*  Don't
      14  19  29  39 49    39         want

        Miles                     .-•»«•»
                                 distant*
 U.S.   Council  on  Environmental
 Quality,  Department of  Agriculture,
 Department  of  Energy,    and
 Environmental  Protection  Agency:
 Public  Opinion   on  Environmental
 Issues:   Results  of  a  National
 Public  Opinion Survey,   Washington,
 Government Printing Office, 1980.
          hazardous waste production projected for
          the  United States  in  1985, shows the
          problem  of defining and of controlling
          hazardous wastes!

              I  do   not  want   to  add  a   new
          definition  to the existing ones; I just
          take hazardous wastes as a problem which
                                   387
-------
has to be solved!   In the past it
was easy to  dump those wastes, but
an Increasing  understanding of our
environment,  a rise  in  public
consciousness,  and  more  complete
environmental  legislation  supported
the  development   of  physical,
biological and chemical  treatment
methods for hazardous wastes.  There
is still  a  need,  and this  will
always  exist,  to  find  new  ways
because the most predominant present
methods of hazardous  waste treatment
(listed below),  as,  for example,
they  are used  in  the  United States,
will   be  neither sufficient nor
allowed  in a  couple of  years, or
they  will become too  costly:

   - indiscriminate  dumping
   - drum dumping
   - ocean dumping
   - pooling for evaporation
   - encapsulation and fixation
   - placement into lined  disposal
      sites
   - spraying into the ground
   - mixing with soil
   - deep well injection
   - ocean burning
   - incineration
    Emphasis  is currently placed  on
developing "new" advanced methods,
still dealing  with the three so-
called "end-point"-solutions,  which
are as follows:

a)  Destruction of the hazard or the
    hazardous  waste.    This  is
    achieved either by  biological
    degradation   methods   or  by
    thermal  processes such  as  high
    temperature fluid wall  (HTFW)-,
    fluid bed-,  or  molten  salt-
    incineration;   pyrolytic
    processes; plasma arc reactions;
    microwave plasma  technologies;
    wet  air oxidation  (WAO);  or
    supercritical   fluid  (SCF)
    oxidation;   among   other
    destruction technologies.

b)  Reduction of the hazard or the
    hazardous waste.  There  is  much
    work done in developing new chemical
    or biological methods.  The  most
    outstanding results seem to be found
    in the field of enzyme technologies
    and genetic  engineering  ('gene
    splicing')  for development  of new
    microbial      strains which  are
    capable   of degrading  previously
    recalcitrant organic chemicals; but
    progress  has  also been achieved in
    development of chemical  processes
    that reduce the hazard and danger of
    heavy metal wastes (as, for example,
    the conversion  of  hexavalent-
    chromium  waste  to the  trivalent
    state) or that involve  the oxidation
    or hydro!ization  of organic
    compounds.  Other examples could be
    the  use  of  reduction  agents,
    different  chemical  treatments,
    photolysis, and gamma irridation
    methods.   But physical  processes can
    also be  used  to reduce the hazard—
    let us   just take  as  examples
    distillation,  absorption,   and
    extraction  by  supercritical  fluids;
    as well  as  different  membrane
    processes such as  ultrafiltration,
    reverse  osmosis,  dialysis  and
    electrodialysis,  and  freezing and
    crystallization processes.

c)  Isolation  of hazardous wastes.
    Progress  has  also been achieved in
    the  isolation  of hazardous   wastes,
    which has to be regarded as the most
    traditional  and coventional  of end-
    point treatment methods. There have
    been achieved  better knowledge and
    safer methods  of disposing  wastes,
    either  in  secure landfills  or by
    neutralizing,  solidifying,  or
    encapsulating the  wastes  in  cement
    or pozzulano-based  materials,
    thermo-plastic materials,  organic-
    polymers,   or  glassification
    materials;  but  there is still  a
    great deal  of uncertainty about the
    long-term effects of these methods.

    But even in  achieving these end-
points there  will, in almost all  cases,
remain the need for  disposal  or
isolation of  residuals,  and there will
always be an  emission of stack gases and
                                  388
-------
effluents,  because  none  of  the
treatment methods can obtain a zero
level  of pollution.

Recycling and Cleaner Technologies  as
Farsighted Means of  Hazardous Waste
Management

    Many  of  the above  named  new
methods  for  the  treatment of
hazardous wastes  are  the result of
the development or  adaption of  new
industrial production methods or
reactor designs, as in the case of
fluid  bed reactors.  By economically
and  technologically  improving
production methods, technologies,
and systems, people became more and
more aware of minimizing production
spills and wastes—and  one of the
means  to achieve this  was recycling.

Recycling

    Recycling has for a long  time
been an  integral part of  resource
management,  adding  to materials
supply  and  alleviating  resource
depletion.  But in the recent years
recycling has had an important role
as a means in waste management, and
there  are  good   chances   for
establishing recycling  as a means of
hazardous waste  management—though
there are still  a  lot of  barriers
and constraints.

    But  what  is  recycling?   Well
knowing  that there are a lot of
diverging definitions,   I  will
explain,  by the next figures,  which
possibilities  there  are  to  use
materials physically again,  to
recycle them.
Figure  2.    Recycling Possibilities.
             Source:  Muller
   raw material A
    raw material BI
:              reduction B[   4use BI  »
                      I    i	 __    _ _ f1"'""" ""••
    We all  know the principle of reusing
a product for its original purpose,  such
as a returnable bottle, or a material
for its original  purpose (glass from
broken bottles),  and we also  know a  lot
of other materials which are used  in
such  recycling  processes.   Scrap,
rubber, plastics, paper and glass  are
examples   for   these  well-known
processes—but we all know as well  that
much  more  could be  done in  these
traditional  fields  of   recycling
activities!

    But recycling  has  also  become  an
important  part  of  hazardous  waste
management  activities — either  for
directly regaining economic  value or by
combining  environmental  needs with
recovery activities.   Recycling of waste
oils,  asphalts,  solvents  and acids  are
well  known,  but not fairly well used
examples for recycling activities in the
field  of hazardous wastes.

    It is obvious that the possibilities
for a direct recycling  of hazardous
wastes,   both  technologically   and
economically, seem to be rather limited
because of  the  increasing complexity of
many production processes and production
wastes.  But there  are still  a  lot of
often  surprising  possibilities—for
example the  recycling of petrol from the
petrol-laden air which  is emitted under
each tankfilling operation.

    Recycling as a means of hazardous
waste management  has become of greater
                                 389
-------
importance  for  those activities
which have been called "recovery".
Let us just look at some examples
which show that there  are a  lot  of
possibilities which  can be  used  in
an economic way:

  - In  the electronics industry you
    can  find  a  lot  of  copper
    containing wastes from  excess
    copper which is  removed by means
    of   hydrochloric  acid.    This
    waste liquid could  go to special
    hazardous waste  treatment plants
    to  end up as filter cakes which
    have   to   be  deposited   in
    controlled landfills.   But  on
    the other  side is copper,   a
    micro-nutrient  in  fertilizers,
    where  the  Cu-requirement  is
    approximately 01%—and therefore
    some fertilizer producers are
    recycling  the copper  from
    copper-containing wastes from
    the  electronics  industry.

  - Another example  could   be the
    recirculation of process  waters
    from  the  production  of  glass
    wool.   These  waters contain
    hazardous organic compounds and
    are now  recirculated for the
    production of  a binding agent
    which is used for the production
    of  glass wool.

  - Acid wastes from sulphate based
    titan-dioxide production  can  be
    recycled in  an energy  saving
    manner.  Thus it can be achieved
    that no acid  wastes have to  be
    dumped in the ocean.

  - Bottom  products from  the
    distillation  of  solvents
    containing wastes from the paint
    and  colour  industry  can  be
    recirculated  as  fillers  in
    certain paint types.

  -Pickling  acid  wastes  and
    electrolytic pickling baths from
    the   aluminum   industry  are
    hazardous wastes  which  can  be
    recovered  and converted  by
    special  combinations of these two
    waste  streams,  whereby  sodium
    aluminate  (NaAlC^h   aluminum
    sulphate-Al9(804)3- and   Glauber's
    salt (miraEil i'te",
    can  be  produced.
NaS04.10
H2S04)
    There are a lot of other examples
for the recycling of materials by the
recovery of  hazardous wastes—but it
must be said  that there is still a need
for further development.

Cleaner Techno!ogies

    The  development  of  extended
recycling and recovery  processes  was the
first step on the way  to  a  new  concept
in  managing  waste  problems--cleaner
technologies,  known  also  as  low-waste-
technologies.

    Though there are many  differences in
defining this term,  cleaner technologies
should be regarded  as  those production
measurements  which  reduce the  quantity
and the hazard of all types of emissions
of  a  production cycle,  which  can be
illustrated by the following figure:

    It  is  obvious  that  a  recycling
policy just  covers  those  production or
product wastes which have occurred,
while cleaner technologies try  to avoid
the origin of wastes and  emissions,  as
far as possible, by regarding the whole
production and product  cycle.

    How can this be  achieved?

    There are three principle ways of
reducing the level of emissions and
wastes:

Figure 3:Cleaner Technologies and the
           Production  Cycle.
           Source:   Miljostyrelsen

a)  Choice of other  or alternative raw
    materials
b)  Development of new  processes
c)  New or alternative  product design
    ada:   It  is  a  basic principle  that
those raw  materials which contain fewer
                                  390
-------
    Figure 3
                          frecyc-^
                            ling
V7c.t. = cleaner technology
       "Environment"
      391
-------
pollutants   result  in  fewer
environmental emissions;  but also
the choice of alternative materials
with  the same  function  in  the
production process will make a given
technology   cleaner.     These
substitution processes  can  be
divided  into physical  (such as
replacement of transformer  oils by
PCB-free  oils),  quantitative  (such
as  using thinner  coatings)  or
functional  (as in replacement of
solvents in  paints  by  water)
substitutions.

    ad  b:  The development of new
processes is a very  important part
of   a   cleaner   technology
conceptualization,  and  there  are
numerous examples  of  cleaner
technologies  avoiding  airborne,
liquid or solid hazardous wastes:

  - Development of a non-polluting
    process for recovering 99.99%-
    pure  lead  from  discarded
    batteries.   The  method  reduces
    lead  and  sulphur  dioxide
    emissions  by avoiding smelting,
    the  conventional   way  of
    recovering  lead from spent
    batteries.  The new process  uses
    electrolytic   melting  and
    subsequent chemical/electroly-
    tical operations.

  - Development  of  an   (as  yet
    unpublished)  ion-exchange
    process,  which converts waste
    gypsum   (CaSO^) and  cheap
    potassium chloride  into  the
    valuable  fertilizer component
    potassium  sulphate  (KgSO^) and
    harmless  calcium   chloride
    (Cad?)-   This process  gives a
    profit   compared   to   the
    conventional  production  and, at
    the  same time, gypsum  waste is
    avoided and  the  heavy metals
    from  the  proceeding  rock-
    phosphate  treatment process--
    especially copper, lead, zinc,
    cadmium   and  mercury--can  be
    precipitated.
  -  Development of alternative spray-
    coating  processes or  production
    technologies  for fly-ashes  or
    different  filter dusts in order to
    minimize solid hazardous wastes.

    ad  c:  New or alternative product
design   includes   substitution   of
function,  materials,   and construction.
These substitution methods are important
in hazardous waste  management,  as  the
following examples show:

  -  substitution for mercury-containing
    batteries  of zinc-oxygen  batteries
    to  avoid  mercury emissions from
    refuse incinerators

  -  development  of  water-based  and
    solvent-free paints and colours

  -  development  of  chlorine-free
    bleaching  processes in the pulp and
    paper industry

  -  replacement of cadmium in colouring
    pigments

  -  substitution for PCB-containing oils

  -  substitution  for  halogenated
    hydrocarbons as  propellent, sponging
    or foaming agent

  -  development  of new  construction
    principles and materials

    It    is   apparent   that  the
conceptualization  of  cleaner
technologies  is  very complex, and that
the  development  of new raw  materials,
production processes or products will
take time, as the following table shows.
                                 392
-------
Figure  4:   Levels of Substitution
             and Associated Develop-
             ment Times.
             Source: Schlabach
                            Development
                              time
       Level
1  Noninteractive material substitutions       1 - 3
  Assembly or component change
2  Development of new material or chemical    3-4
   process
3  Interactive materials substitution         4 — 5
  Electronic technology change
  Subsystem or small system development
4  Systems of reasonable complexity         ~7
5  Complex weapon systems             ~10
6  Telephone exchange technology          ~13
7  Time for scientific discoveries to find large-   ~15
   scale technical application


    Many  will  claim that all new
developments concerning  products,
production processes,  and  emission
equipment can be  regarded as cleaner
technologies  because of their effect
on reducing emissions  and  spill
products,  and as a consequence  of
minimizing production  costs—but
isn't that  too  easy to  ease our
consciences?

    Cleaner  technologies  should  be
regarded as those technologies which
are a real and jntended  alternative
to existing  technologies, from  an
environmental  point of view; the
environmental benefits should not  be
regarded  as an accidental by-
product!

RecyclIng andCleaner Techno!ogy  in
Denmark
                                     law,  which was passed  in the extended
                                     form  including  cleaner  technologies  by
                                     October 1, 1984,  can be  given:

                                     a)  We feel  that recycling activities
                                         have been promoted very much by this
                                         law, and that information activities
                                         are  of  great  importance  to  the
                                         attitude of  the  population  and  the
                                         result of the recycling activities.

                                     b)  The importance of easy,  consumer-
                                         oriented  collection systems  is
                                         evident  both  for  households  and
                                         industries.    This  is  a  very
                                         important factor for separation  and
                                         recycling  activities  concerning
                                         hazardous wastes.

                                     c)  It  takes  time  to  develop and
                                         introduce  cleaner technologies  in
                                         the industry,  especially in  small
                                         and medium sized enterprises.

                                     d)  There  is  a  great  interest   in
                                         participating in  our  recycling-and
                                         cleaner technology program,  and  we
                                         feel  that  we  are  promoting this
                                         concept a lot.  We  hope to  be able
                                         to  extend  our  activities  more
                                         towards hazardous  wastes.

                                     e)  A great incentive, or press,  for  the
                                         promotion of  cleaner technologies is
                                         given  by   environmental   law
                                         activities—and it always turned  out
                                         that cleaner  technologies  are paying
                                         for themselves.

                                     SUMMARY AND CONCLUSIONS
                                     1.
    Denmark has  passed  a  new  law on
recycling  and  cleaner technology
which enables the state,  to support
all types  of  recycling activities
concerning  materials.   On the other
hand,  it  is  possible to promote
cleaner technology by  financing  or
supporting  research  and information   2.
activities  or  research  and
demonstration  projects.

    It would be  inappropriate to try
to  explain and  discuss  this  law   3.
here.  Some experiences with  this
                                    393
Increasing  amounts  of  wastes,
increasing  hazards,   and  new
technological   developments  make
solving future  problems by new  way
of  managing   hazardous  wastes
inevitable.

Conventional  ways of treatment will
become more  and more insufficient
both    economically   and
environmentally.

Recycling is  becoming an  integrated
part of the industry's activities,
-------
but  there  are  still
unsolved  problems.
                    a  lot  of
4.
Cleaner  technologies  offer  new
and non-traditional  possibili-
ties for solving hazardous waste
problems  by  regarding   the
complete  product life cycle.
5.
It
cleaner
   seems  that  all recycling and
        technology activities
can  be ensured  by   law  and
financial support measurements
by state  authorities.
REFERENCES
1.
Curran, Linda H.:  New Solutions
To Industrial Waste Management,
in:   Environmental  Progress,
Vol.  3, No. 2,  May 1984.

Lovgren,  Peter:   The  Danish
System,  in:  Proceedings of the
1st  International Symposium  on
Operating  European  Centralized
Hazardous  (Chemical)  Waste
Management Facilities,  Odense,
Denmark, September 1982.
    M a c k i e,
    Kathleen:
    Chemical
    1984.
           Jay  A.;   Niesen,
          The Alternatives, in:
         Engineering, Auag. 6,
4.   Miljostyrelsen (NAEP): Recycling
    and  Cleaner  Technologies,
    Copenhagen, Oct.  1983.

5.   Muller,  Klaus:  Altolverwertung
    (Reuse  of  used  oils),  Erich-
    Schmidt-Verlag,  Berlin,   W.
    Germany,  1982.

6.   N.N.:    Lead-recovery  Route
    Eliminates   Smelting,  in:
    Chemical Engineering, Aug. 20,
    1984.

7.   N.N.:    Recycling   in   der
    Aluminiumindustrie (Recycling  in
    the  Aluminum  Industry),  in:
    Umweltmagazin, Oktober 1984.

8.   Schlabach, T.D.:   Materials
                                   9.
                                   10.
                                   Conservation:
                                   Substitution:
                                   Conservation  and
                                   No.  1,  1984.
                Technology,   and:
                 Technology, in:
                Recycling,  Vol.  7,
Tucker,  Samuel  P.:  Deactivation of
hazardous  chemical  wastes,  in:
Environmental   Science   and
Technology, Vol. 19,  No. 3, 1985.

U.S.   Council  on  Environmental
Quality,  Department of Agriculture,
Department   of   Energy,    and
Environmental  Protection Agency:
Public  Opinion on  Environmental
Issues:   Results  of a  National
Public  Opinion  Survey,  Washington,
Government  Printing Office, 1980.
                                                 Disclaimer

                                    The work described in this  paper was
                                    not funded by  the U.S. Environmental
                                    Protection Agency.  The contents do
                                    not necessarily reflect the views of
                                    the Agency and no official  endorse-
                                    ment should be inferred.
                                   394
-------
        DEWATERTNG OF HAZARDOUS WASTES USING REVERSIBLE GEL ABSORPTION
                         W. J.  Maier and E.  L. Cnssler
                            University of Minnesota
                             Minneapolis, MN 55455
                                   ABSTRACT

     We have studied the technical feasibility of using gels to concentrate
aqueous waste SEX earns.  Potential applications focus on extracting pure water
from wastewaters that contain low concentrations of hazardous solutes in order
to reduce volumes and concentrate tne hazardous material.  It has been demon-
strated that water sensitive cross-linked gels absorb water while excluding
macromolecules tnat are larger tnan the gel pore size.  Exclusion of ionic sol-
utes can be obtained using gels that contain potentially ionizable groups such
as the cross-linked, partially hydrolyzed polyacrylamide gels.  Gels have been
contacted with solutions containing a wide range of different molecular weight-
size solutes.  The gels expand by imbibing water.  The expanded gel is separat-
ed from the concentrated raffinate by draining (filtration) and regenerated.
Gel volume is sensitive to pH.  It collapes when acidified, yielding an extract
of water.  A cyclic process of swelling, acidification to release water, and
gel  reuse nas been demonstrated.  The process is similar to solvent extraction
with the added feature that it separates by molecular size.
INTRODUCTION

     Innovative methods for managing
aqueous hazardous wastes are needed.
Presently available methods of treat-
ment are costly, particularly when
the pollutants are dispersed at low
concentrations in large volumes of
water.  The concept of concentrating
the waste stream by removing water is
attractive because it facilitates
subsequent reuse, disposal or
destruction of the waste material.
The alternative of extracting the
pollutant from the water phase can
involve high costs because of the
need for essentially complete capture
of hazardous materials.  Research was
therefore initiated to assess tne
feasibility of removing water using
water sensitive cross—linked gels
that can be easily regenerated.
APPROACH

     Polymeric materials possessing
the ability to swell in water without
dissolving have been described in the
literature (1,20).  Swelling is a
consequence of the affinity of the
chemical structure of the gel for
                                    395
-------
water.  Swelling is also atfected by
the degree of cross-linkages and the
potential for ionization.  Cussler et
alj (3) suggested that appropriately
cross—linked, partially hydrolyzed
polyacrylamide gels could serve as
selective extraction agents because
these gels have unusual swelling char-
acteristics.  Such gels absorb water
(a low molecular weight solvent) but
reject high molecular weight solutes
that are larger than the average pore
diameter of the hydrogel.  The gels
can be regenerated by imposing a small
change in process conditions such as
pH (11), temperature (17,18), elec-
tric field (26), solvent composition
(17, 18,24), or salt concentration
(7). These changes cause the gels to
collapse, thereby releasing much of
the water phase.

     It has been shown that drama-
tic changes in volume occur as a re-
sult of small changes in process con-
ditions when some of the side groups
of the gel network undergo ioniza-
tion.  lonization sets up localized
negative charges.  Small numbers of
charge sites, around 1% (15), are
erfeotive.  The resultant ionic ma-
trix therefore expands and imbibes
water while excluding macromolecules
by steric exclusion and ions by Don-
nan exclusion.

     The manner in which the gels are
used in an extraction process is
shown in Figure (1).  Gel material is
added to the wastewater solution.
The gel swells by absorbing water but
excluding high molecular weight sol-
utes.  The raffinate solute concen-
tration increases accordingly.  It
can be separated from the swollen
gel by draining or filtration.  The
swollen gel is now regenerated by
adding acid.  Since the gel volume is
very sensitive to pH change, it col-
lapses and releases the absorbed
water.  The collapsed gel is again
drained, filtered to remove water.
Finally, the acidified gel is neu-
tralized with a small amount of base,
enabling it to swell again when added
to another batch of solution.

     The first part of this process,
concentrating macromolecular solu-
tions by allowing gels to swell in
them was first reported by Flodin
et al.(8) .  Specific applications
have developed sporadically since
then (1,4,7).  Despite its success in
concentrating solutions, the techni-
que was limited in application due to
the lack of an efficient method for
regenerating the gel (drying was the
only means).  The second part of this
process, the drastic swelling and
collapse behavior of the gels in sol-
vent, has been studied primarily as a
type of phase transition behavior(5,
15,25).  Interest in this area was
accelerated by Tanaka's 1978 (18)
report of a discrete change in gel
volume with an infinitesimal change
in solvent composition.

THEORY

     The proposed separation process
relies on regenerating the gel by
changing its volume via decreasing
the pH.  This variation of gel volume
with pH can be estimated using a mod-
ification of Flory-Huggins theory for
polymer solutions (6,11,12,15,16).
Theoretical models that describe
volume changes in ionic nature of the
gel have been described by several
investigators (9,13,15,25). Of these,
the model developed by Cnssler ejt
al. (3) . is the simplest.
(1)
oc
                  R
          K
                     B
                                     396
-------
where
   V =
   R =
   K
   B =
volume of gel
concentration of potentially
ionizable groups
[R COO H] + [R COO-]
ionization constant
K = [R COO-]/[R COO HHOH-]
total base added to an acid
gel
Equation 1 shows that gel volume to
the -2/3 power should vary linearly
with the reciprocal of the amount of
base added.
EXPERIMENTAL PROCEDURES

     The gel used in initial experi-
ments was made by hydrolyzing cross-
linked polyacrylamide beads produced
commercially as packing for gel per-
meation  chromatography (BioRad La-
boratories, Richmond, CA).  It is
sold as Bio-Gel P-6 and has a part-
icle size in water of 150-300 nm.
The gel was hydrolyzed for 24 hours
at 323°K in 0.50 M Na2C03<19).  Later
work has focused on polyacrylamide
gel produced in our laboratory.  De-
tails of preparation will be present-
ed at the conference.

     Model solutes included a variety
of chemicals.  Sugar, urea, and sod-
ium chloride were reagent grade
(Fisher) and were use as received.
Polystyrene latex particles 910 na in
diameter were purchased commercially
(Duke Scientific, Palo Alto, CA).
Polystyrene latex particles 34.6 nm
in diameter were a gift of Dr. Hamish
Small (Dow Chemical Co., Midland,
MI).  Colloidal silica particles
(Nalco Chemical, Chicago, ID were 5
nm in diameter.  Bovine serum albumin
(Sigma Chemical, St. Louis, MO) was
diaJLysed and then recrystalized.
Hemoglobin was a highly purified sam-
ple obtained from Dr. Rufus Lumry
(University of Minnesota).  The sam-
ple of polyethylene glycol used (J.T.
Baker, Phillipsburg, NJ) is sold as
having 68-84 monomer units and a mo-
lecular weight of 3000-37000 daltons.

     Starting in late 1984, a series
of tests using potentially hazardous
chemicals were initiated.  They in-
clude pentachlorophenol, heavy
metals, and their organic complexes.
While the results of this work were
not available at the time of writing,
they will be reported at the confer-
ence.

     Experiments were carried out by
contacting 0.5 g of dry gel with 20 g
of solution.  The materials were
shaken to allow swelling and the
swollen gel was recovered by decant-
ing or centrifugation.  Raffinate and
gel volumes were calculated from
weight measurements.  The concentra-
tion of chemicals in the feed solu-
tion and raffinate were measured,
analytical procedures are described
elsewhere(22).
                                   RESULTS AND DISCUSSION

                                        Three kinds of experiments have
                                   been carried out to show that gels
                                   can function as size—selective ex-
                                   traction solvents.   Batch  tests were
                                   made to show that gel absorption is
                                   selective for water while rejecting
                                   high molecular weight solutes.  Batch
                                   test acidification tests were design-
                                   ed to show how the  gels can be regen-
                                   erated and reused.   The third kind of
                                   testing was aimed at demonstrating
                                   that tne gels are sufficiently strong
                                   to be used repeatedly.
                                    397
-------
      That  the  gel  can function  as  a
 size-selective extraction solvent  is
 shown by the experiments  reported  in
 Table I.   Gel  was  contacted with a
 series of  water solutions containing
 different  size solutes described in
 columns 1-3.   Columns 4-5 give  the
 feed  and final (raffinate) concentra-
 tions,  the increase  in concentration
 reflects removal of  water with  the
 small amount of gel  used.

      The last  column in Table 1 is an
 overall measure of extraction effici-
 ency  (E).   As  shown  in equation 2  it
 is  defined as  the  concentration dif-
 ference AC between the initial  solu-
 tion  and the raffinate divided  by
 the maximum concentration difference
A<2i8X vhich would  be attained if all
 the solute in  the  initial  solution
 were  recovered in  the raffinate.
 From  this  definition it follows that:
 (2)
where
E =  &c   — Cr/Ci — 1
      &Cm    Ml/Mr—1

 ratfinate  solute  concentration
  Cj =*  initial  solute concentration

  Jig —  mass of  ralfinate

  Mj =*  mass of  initial  solution

The quantity E  has been termed the
efficiency, it  is really a measure of
the degree to which a solute  is ex-
cluded  from the gel as  water  is ab-
sorbed.  For practical  applications,
the dgree of swelling is also import-
ant.  However,  the degree of  swelling
of polyacryl amide gels  is so  large
that applications are likely  to be
limited by exclusion, not by  insuf-
 ficient  swelling.  Therefore, E is a
 useful parameter for characterizing
 performance.

     The results in Table I show that
 solutes which are greater tnan 3 nm
 in diameter can be concentrated with
 an efficiency of at least 80%.  These
 efficiencies are compromised by weak
 solute adsorption on the surface of
 the gel  spheres.  For example, for
 the 34.6 nm latex, some latex adhered
 weakly to the gel.  When this latex
 was removed by washing, the extrac-
 tion efficiency increased to 97%.  In
 this sense, the gel extractions are
 similar  to freeze-concentration tech-
 niques used in the food industry,
 where solutes can adhere to ice
 crystals.

     The molecular size cut-off for
 rejection is dependent on the degree
 of cross-linking of the gel.  In-
 creased  cross-linking reduces average
 pore size openings and hence particle
 rejection size range.  Studies using
 different degrees of crosslinking
 will be presented at the conference.

     One limitation on gel exclusion
 that is not shown in Table 1, in-
 volves the behavior of positively
 charged  species.  Cationic proteins
 like lysozyme and cytochrome C, pre-
 cipitate with the gel.  Thus cationic
 interaction may compromise separa-
 tions.   Interestingly, the gel's
 swelling seems to be largely unaf-
 fected by these reactions,

 Gel Regeneration

     Regeneration depends on collaps-
 ing the gel, thereby releasing the
 bulk of the absorbed water. The water
 retention properties of hydrolyzed
 polyacrylamide gel were tested at a
•series of pH levels in order to es~
                                     398
-------
 tablish  its volume vs.  pH  relation-
 ship.  As  shown  in Fig. 2,  the  gel
 volume increases  sharply at pH  5-6
 and  goes through  a soft maximum at
 higher pH.  These pH values represent
 the  bulk water phase and pH inside
 the  gel may be somewhat different
 (27).

     For a separation to be effec-
 tive, the  sudden  increase  in gel vol-
 ume  should occur  at a lower pB  than
 that of the solution being  separated.
 However, the  separations need not in-
 volve changing the pH of the solu-
 tion.  As  shown in Fig. 1, the  gel
 can  be added  to the solution and re-
 moved from it at  the solution's pH.
 It is only the gel regeneration which
 involves adding acid or base, and not
 the  separation itself.

     The changes  of gel volume  shown
 in Fig. 2 result from a variety of
 thermodynamic non-idealities (21).
 We believe that gel ionization  is the
 most important factor involved, equa-
 tion (1), which predicts that the
 (-2/3) power of the gel volume  should
 vary linearly with the reciprocal of
 the  total amount of sodium hydroxide
 added to the gel, should apply.  This
 prediction is verified by the results
 shown in Fig. 3.  At the same time,
 we expect that experimental results
 studying a broader range of variables
 will uncover cases where other  fac-
 tors are significant.

 Gel  Reuse

     Data on repeated use of poly—
 acry1amide gel are not currently
 available.  However,  the results ob-
 tained in earlier work using a  dex-
 tran gel indicate that repeated use-
 age  is possible.  Dextran gel (CM
- Sephadex C-SO, Pharmacia Fine Chem-
 icals, Pi scataway, NJ, which is used
 as a packing for gel chromatography)
was used.  This material has a dry
particle  size of 60-120 pm, it is
weakly  ionic and was used as receiv-
ed.  A  dilute suspension of the 34.6
nm polystyrene latex solution (Table
1) was  contacted with the dextran
gel.  After the gel had expanded it
was removed and raffinate concentra-
tion was measured.  Acid was added to
shrink  the dextran gel and filtered
to remove released water.  A drop of
base was added to the gel filter
cake, and again contacted with the
raffinate.  From a mass balance, the
final raffinate concentration c after
n treatment cycles, assuming perfect
exclusion of the latex particles is
given by equation 3.
(3)
C =
               m
where m is the initial mass of latex,
V0 is the initial volume of solution,
and v is the volume removed by one
treatment cycle of gel absorption.
Thus the reciprocal of concentration
c should vary linearly with the num-
ber of cycles n.  The results for
12 cycles given in Fig. 4 show that
this is the case for latex solutes.
These results have two important cor-
ollaries.  First, there is little
cumulative loss due to latex adsorp-
tion on this gel.  This suggests that
adsorption, which is responsible for
the lower separation efficiencies
reported in Table I, is apparently
most significant for the first cycle,
and is less important as the gel is
reused.

     The second corollary of the re-
sults in Fig. 4 is that the gel is
removing the same amount of water on
the tenth cycle as on the first
cycle.  This implies that the gel's
                                     399
-------
swelling capacity remains essentially
constant over all cycles, and hence
can be reused in cyclic operation.
To be sure, reuse in practice will
involve hundreds of cycles and such
extensive long term tests nave not
been aade.
ACKNOWLEDGEMENTS

     This work was supported by EPA
Grant EPA 168-03-1957 and by National
Science Foundation Grants CPE 80-
25304 and CPE 82-07017.
1EFEEENCES

 1.  Andrade, J.D., Hydrogels for
     medical and related applica
     tions, ACS Symposium Series
     No. 31, Washington, 1976.

 2.  Cussler, E.L., Diffusion, Cam-
     bridge University Press, London,
     1983. Section 5.2.

 3.  Cussler, E.L., M.R. Stokar and
     I.E. Varberg, Gels as Size sel-
     ective Extractive Solvents,
     Submitted for publication, 1984.

 4.  Determan, H., Gel chromatogra-
     phy. Springer-Verlag, NY, pp.56-
     60, 1980.

 5.  Dusek, K. and D. Patterson,
     Transition in swollen polymer
     networks induced by intramolecu-
     lar condensation, J. Poly.
     Science A2, 6, 1209-1216, 1969.

 6.  Dusek, K, Inhomogeneities
     induced by crossiinking in the
     course of crossiinking copoly—
     merization, in Polymer Networks:
     Structural and Mechanical Pro-
     perties, A.J. Chompff, ed.,
     Plenum Press, NY, pp. 245-260,
     1971.

 7.  Fisher, L., Gel filtration
     chronatograph, Elsevier Press,
     NY, pp.205-231, 1980.

 8.  Flodin.P., B.Gelotte, and J.Po-
     rath, A method for concentrating
     solutes of high molecular weight.
     Nature, 188, pp.493-4, 1960.

 9.  Flory, P.J., Principlies of
     Polymer Physics, Cornell,
     Ithaca, 1979.

10.  de Gennes, P.G., Scaling
     Concepts in Polymer Physics,
     Cornell, Ithaca, 1979.

11.  Hasa, J. and M. Ilavsky, De-
     format ional, swelling, and po-
     tent iome trie behavior of ioniz-
     ed poly (methacrylic acid) gels,,
     II Experimental results, J.
     Poly. Science, 13, pp.253-262,
     1975.

12.  Hasa, J.M., M. Ilavsky and K.
     Dusek, Deformational, swelling,
     and potentiometrie behavior of
     ionized poly (methacrylic acid)
     gels.  I, Theory, J". Poly.
     Science, 13, pp.253-262, 1976.

13.  Hill, T.E., Statistical thermo-
     dynamics, Addison-Wesley, Read-
     ing, MA, pp.410-419, 1960.

14.  Hochberg, A., T. Tanaka, and
     D, Nicoli, Spinodal line and
     critical point of an acrylamide
     gel, Phys. Rev. Letters, 43,
     pp.217-219, 1979.

15.  Ilavsky, M., Effect of Electro-
     static  interactions on phase
     transition in the swollen poly—
                                      400
-------
     meric network. Polymer, 22, pp,,
     1687-1692, 1981.

16,  Ilavsky, J., J. Hronz and K.
     Dusek, Inhomogeneitics and de-
     viations from the gaussian
     photoeleastic behavior of net-
     works, J. Macromol. Sci., BL9,
     pp.227-236, 1981.

17.  Ilavsky, M., J. Hronz and K.
     ulbrich. Phase transition in
     swollen gels. III. The tempera-
     ture collapse and mechanical be-
     haviour of poly (N,N~di-ethyl-
     acrylamide) networks in water.
     Polymer Bull., 7, pp.107-113,
     1982.

18.  Ilavsky, H., Macromolecules, 15,
     pp. 782-788, 1982.

19.  Jacobson, B.S. and D. Branton,
     Plasma membrane:  Rapid isolation
     and exposure of the cytoplasaic
     surface by use of positively
     charged beads, Science, 195,
     pp.302-304, 1977.

20.  Ratimer, B.D. and A.S. Hoffman
     in Hydrogels for Medical and
     Related Applications, edited by
     J.D. Andrade, ACS Symposium
     Series, No.31, Washington, 1976.

21.  Rogovina, L.Z. and G.L. Sloninr-
     skii, Formation, structure, and
     properties of polymer gels, Rus-
     sian Chem. Revs. 43(6), pp. 503-
     523, 1974 (translated from
     Uspekhi Khimii, 43, pp.1102-1135,
     1974).

22.  Stokar, M.R., M.S. Thesis, Gels
     as a size-selective extraction
     solvent. University of Minnesota,
     1982.
23.  Tanaka, T. Physical Review Let-
     ters, 40(12), pp.820-823, 1978.

24.  Tanaka, T., D. Fillmore, S.T. Sun,
     I. Nishio, G. Swislow, and A.
     Shah, Phase transitions in Ionic
     Gels, Phys. Rev. Letters, 45,
     pp.1636-1639, 1980.

25.  Tanaka, T., Gels, Scientific
     American(l), vol.244(1981),
     pp.124-138.

26.  Tanaka, T., I. Nishio, S.T. Sun,
     and S. Deno-Nishio, Collapse of
     gels in an electric field.
     Science, 218, pp.467-469, 1982.

27.  Tanford, C,, The physical Chem-
     istry of Hacromolecules, Wiley,
     NY, I960.
                Disclaimer

  This paper has been reviewed in
  accordance with the U.S. Environ-
  mental Protection Agency peer and
  administrative review policies and
  approved for presentation and publi-
  cation.
                                      401
-------
Table I.  Concentration of Dilute Aqueous Solutions Using Hydrolyzed
          Polyacrylamide Gels.


            Mol.Wgt.,  Solute       Feed          Raffinate
Solute      Daltons   Size,nm  Concentration(a) Concentration(a) Efficiency(b)
Polystyrene —
Latex
Polystyrene —
La tea:
Silica
Bovine
Scrum 66,000
Albumin
Hemoglobin 64 ,500
Poly- 3000-
ethylene 3700
Glycol
Sucrose 342
Urea 60
990. 0(c)
34.6(c)
S.O(c)
7.2
-------
               STEP t :  Add solution to b*«tc
                       form of g*l.  Gel swells,
                       preferentially absorbing
                       solvent.

               STEP 2 :  Withdraw.non-absorbed
        rafHnate        rtffin*te. now a concentrated
                       solution.
               STEP 3 : Recover ewollen s*l by
                       filtration or centrlfugatlon.
               STEP 4 : Add acid to gel, which
        extract        . shrink* drastically.
        1111 •            Withdrew released solvent.
               STEP S :  Add baae to gel eo that
                       It I* ready for re-use
                       in Step 1.
                 FIGURE 1
  FIGURE  2
$  0.30

 Ul
 z
    0.20
              ZO -
                      o.to            o.ao
                 [ADDED  BAse]"1

                   FIGURE 3
  4             8
NUMBER OF CYCLES

    FIGURE  4
12
                                                  403
-------
      STUDIES ON THE BIODEGRADATION OF ORGANOPOLLUTANTS
                    BY A WHITE ROT FUNGUS

              John A. Bumpus and Steven D. Aust
 Center for the Study of Active Oxygen in Biology & Medicine
                 Department of Biochemistry
                  Michigan State University
                East Lansing* MI  48824-1319
                          ABSTRACT

     The  lignin  degrading white  rot  fungus,  Phanerochaete
chrysospgrium degrades a broad spectrum of organopollutants to
CO2-   We  have studied the ability of this fungus to  degrade
organopollutants      using    1»1'-fois(4-chlorophenyl)-2/2/2-
trichloroethane  (DDT)  as  a  model  compound-   Like  lignin
degradation*  the degradation of DDT requires the presence  of
another carbon source.   in general,  the total amount of J-^C-
DDT  degraded  to  14CO2 during the 30 day  incubation  period
increased when the glucose concentration was increased from 23
mH to 112 mM-   A further increase to 224 mM did not result in
a further increase in 14CO2 production.  Studies on the effect
of   organopollutant   concentration   showed   that   l^c-DDT
degradation  to  14CO2  exhibited first  order  kinetics  with
respect   to  14C-  DDT  concentration.    Metabolism  studies
demonstrate  that,  under nutrient nitrogen limiting (2-4  mM)
conditions  DDT  is converted to CO2 via a  pathway  in  which
I/1' — bis ( 4— chlorophenyl )-2, 212-tr ichloroethanol (Dicofol) is *•
major   intermediate.    Under  nutrient  nitrogen  sufficient
conditions  (24 mM)/  CO2 production is suppressed and DDT  is
degraded via a pathway in which  2,2 *-bis(4-chlorophenyl)-l*1-
dichloroethane (DDD) is a major intermediate-


INTRODUCTION AND PURPOSE       degraded by this  microorganism
                               are:   DDT/  2,3>7,8-tetrachloro-
     Studies  in our  labora-  dibenzo-p-dioxin (2>3»7»8-TCDD>,
tory  have demonstrated  that  3/4,3',4' - tetrachlorobiphenyl
the    white   rot    fungus,  (3,4,3',4'-TCB), 2,4,5,2',4', 5'
Phanerochaete   chrysosporiurn  -hexachlorobiphenyl   (2*4/5,2'f
has the ability to degrade  a  4',5'-HCB),   1 / 2 , 3,4 , 5, 6-hexa--
number     of    recalcitrant  chlorocyclohexane  (Lindane) and
organopollutants to CO2  (1)-  benzo[a]pyrene    (Fig.      1),
Included  among the compounds  Evidence   suggests   that   the
                             404
-------
                     IV
                     VI
Figure  1.  Structures of re-
calcitrant   organopollutants
which are degraded to C<>2  by
ILi siiEiis.SiSizsyLiyjs • * • DDT , n.
BenzofaJpyrene, III- 2,3,7,8-
TCDD,    IV.    Lindane,   V-
3,4,3*,4'-TCB/  VI-    2,4,5,
2',4',5'-HCB.
highly  non-specific and non-
stereoselect1ve   lignin  de-
grading system of this fungus
may  also  be active  in  the
degradation of these  organo-
pollutants (1} .

     In order to be useful in
systems used in the biotreat-
ment  of  hazardous   organic
chemical  wastes,   a  micro-
organism   must  possess  the
demonstrated    ability    to
degrade, or aid in the degra-
dation    of,    recalcitrant
organopollutants.     It   is
desirable that degradation be
complete  and independent  of
the  concentration  of   sub-
strate.   In  this  study  we
present  our initial findings
which focus on the  optimiza-
tion   and  control  of   the
                                degradation   of   recalcitrant
                                organopollutants      by     P_._
                                Sl!Ey_§.o.§J2oJliy_m.*   Because much is
                                known concerning its  biodegra-
                                dation,   we have elected to use
                                DDT  as   a  model  compound  in
                                these studies.
APPROACH

   E.I.   chrysospor ium   (ME-446)
was   obtained  from  the  USDA
Forest   Products   Laboratory,
Madison,  WIS.   Cultures  were
incubated  at 39°C  under  100%
oxygen.   The  culture  medium,
consisting  of glucose  (56 mM),
ammonium  tartrate  (12  mM)  and
2,2-dimethylsuccinate.   buffer,
pH  4.2  (100 mM)   and  supple-
mented  with thiamine and trace
metals   has  been
described (2).
                     previously
    Evolut ion
radiolabeled
was  assayed
culture with
intervals,
                of  14CO2   from
               organopollutants
              by  flushing  the
              oxygen  at 3  day
              and  forcing  the
cultur®  atmosphere  through  a
scintillation  cocktail   which
contained ethanolamine as a CC>2
trap (1,2).   Typically, 50 nCi
(110,000  dpm)  of  the  radio-
labeled  compound was  included
in  each culture.   Time points
in   14CC»2  evolution   studies
represent  the average of   four
replicate cultures.

    DDT  disappearance and  me-
tabolite  formation was assayed
by  GLC (1).   The identity  of
metabolites  was determined  by
comigration of metabolites with
authentic standards and by  GC-
MS.    Mycelium dry weights were
determined after collection and
drying   on   tared   millipore
filters.     The    data    was
expressed as the average + S-D.
                              405
-------
of three replicate cultures -
PROBLEMS ENCOUNTERED

     Because the organopollu-
tants examined in this  study
may  not be uniformly distri-
buted   in   aqueous   media/
sampling   errors  may  occur
when aliquots are taken  when
monitoring   degradation   or
metabolite synthesis via GLC.
To   prevent  such   sampling
errors entire cultures rather
than aliquots of cultures are
extracted and quantitated  in
such studies.
RESULTS

     Under  conditions  known
to promote lignin degradation
ky Hr  StULY-SSSESEiHIS' we have
shown that a number of recal-
citrant  organopollutants are
also  degraded  to  CO2  (D-
Initial 3-4CO2 evolution rates
varied  from  0.8  pmoles  of
substrate    converted     to
14CO2/day  for 3/4,3 */4'-TCB,
a  PCB  congener/   to   13-9
pmoles of substrate converted
to  *4CO2/day for Lindane/  a
moderately recalcitrant alkyl
halide  (1)-   Over a 30  day
period    13-3   pmoles    of
              ensure  the timely
              of   the  compound
              degradation.
                                                   destruction
                                                    undergoing
3,4,3"/4'-TCB
and
                        190.8
pmoles   of   Lindane    were
degraded    to   14CO2   (1).
Although   these  rates   may
appear low/  two facts should
be considered;  1) the condi-
tions    used    were     not
necessarily optimal and 2)  if
even these rates can be main-
tained   in  a   biotreatment
system/ aerobic composting  or
land  farming/  for  example/
they  would be sufficient   to
                   Because  lignin   degrada-
              tion  and  the  degradation  of
              organopollutants     by      P_._
              £ll£y_®2§.£2!:Liyj5  both  appear  to
              occur  via   cometabolism/   we
              examined  the effect of glucose
              concentration     on      *4CO2
              evolution  from  14C-DDT  (Fig.
              2).  Standard  glucose  concen-
               CJ
               O
               o
               O
                                S
               cc
               o
               IU
               o
               o
               Q
                                     100


                                     80

                                     60


                                     40


                                     20


                                      0
                       06    12   18   24  3O

                             TIME (DAYS)
Figure  2.  Effect  of  glucose
concentration      on       the
degradation of 14C-DDT to 14CO2
ky  PJL   chEy^SSESEiy.!!* *    Open
circle - 224 mM glucose/ closed
circle  - 112 mM glucose/  open
triangle - 28 mM  glucose/  and
closed    triangle   - 56    mM
glucose.   The concentration of
*4C-DDT was 0-125 UM (44 ppb).
                               tration   in our  studies  was   56
                               mM-   By  increasing  the  concen-
                               tration from  56  mM to  112 mM/  a
                               270%   increase   in  the total
                               amount  of ^4CO   evolved   from
                             406
-------
         was  observed  (Pig.
2).     However/   when   the
glucose   concentration   was
increased   to  224  mM*   no
further  increase was  noted.
In   fact/   9   small*   but
statistically   insignificant
decrease    was     observed.
Similarly,  when  the glucose
concentration   was   reduced
from  56 mM to 23 mM*  a  23%
decrease occurred.   When the
rates of Hcc>2 evolution  are
compared  it  is  noted  that
cultures    containing    the
higher   concentrations    of
glucose continue 14CC>2 evolu-
tion  longer than those  with
lower   concentrations   and/
generally* at greater rates.

    An  exception to this was
observed     in      cultures
incubated  with 23 mM glucose
which has the highest initial
rate   of  14CO2   evolution.
This  apparent  inconsistency
is in agreement with Jeffries
ejt a_l-  (3)* who demonstrated
that* in addition to nitrogen
starvation/   low  levels  of
glucose are also effective in
initiating  lignin   degrada-
tion.   These  results  would
suggest  that lignin and  DDT
degradation  are catalyzed by
the same enzyme system or are
at least under the same  type
of metabolic control.

    In a typical 14CC>2 evolu-
tion  study*  the  3-4C-radio-
labeled organopollutant under
study is added to the incuba-
tion mixture when the culture
is innoculated with fungus at
day  0.  During  the first  3
days of incubation a mycelial
mat is quickly  formed*  how-
ever 1*CO2 evolution does not
occur at this time (Fig-  3).
                        2000
                         1500
                         1000
                         500
                            o
                            I-
                            o
                            at
                            o
                            o
                               *>
                               o
                               Q.
        6  12   18  24

          TIME  (DAYS)
30
Figure 3.  Time course for  the
growth (mycelium dry weight) of
E.I    gbry_sosporium   and   its
degradation   of   14C-DDT   to
1^CO2-   The  concentration  of
14C-DDT  was 4-92 UM (1.7 ppm)-
Triangles   -   mycelium    dry
weight*     circles    «
evolution.
Between day 3 and day 6*  14CO2
evolution begins and is maximal
between  day 3 and day 18 after
which   it   continues   at   a
decreasing  rate until the  end
of   the  30   day   incubation
period.   In  studies in  which
the     cultures    were    not
terminated at 30 days* but were
instead     fortified      with
supplemental  glucose (56  mM)*
the  rate  of  14CC>2  evolution
increased  and  continued at  a
rate of no less than 4.5 pmoles
14C-DDT converted to Hc<>2  per
day for the duration of another
                             407
-------
                           TABLE 1

     EFFECT OF SUBSTRATE CONCENTRATION ON DDT DEGRADATION
              Initial Rate of  14C-DDT degraded
   initial      Degradation     to 14CO2 during
Concentration    of 14C-DDT         the 30 day     % 14C-DDT
   of DDT        to 14CO2       incubation period  converted
(ppb)  (uM)    (pmoles/day)         (pmoles)        to
  44 (0.125)

 17? (0.5)

1742 (4-92)
  2-8

 10-5

108.5
  48.0

 207-1

2047-0
3.8

4-1

4.2
30   days   of    incubation.
Similar   results  have  been
observed   with   the   other
organopollutants   we    have
studied.

     One obvious method which
might  be  used  to  increase
biodegradation entails merely
increasing   the  amount   of
organopollutant  presented to
the  microorganism.   Studies
in which P_-_ chrysosgorjium was
incubated  in the presence of
increasing amounts of •L4C-DDT
are  presented  in  Table  1.
These  results indicate  that
increasing  the concentration
of   J-4c—DDT  increased   the
amount  of ^-4c—DDT which  was
degraded.   When the  initial
rate of *4CO2 evolution (i.e.
the  nearly linear rate which
usually occurs between day  3
and   day  18)  was   plotted
versus   3-4c-DDT   concentra-
tion/   it  was  demonstrated
that the increase in degrada-
tion  exhibited  first  order
kinetics  with   respect   to
                 —DDT  concentration over the
              concentration  range   assayed.
              DDT  concentrations as high  as
              1.7   ppm  (4-9  uM)  did   not
              inhibit  14C-DDT degradation to
              *-4CO2 -    when  the  data  were
              expressed  as per cent  I4c—DDT
              degraded  to 14CO2 per 30  days
              (Table  1),  it was found  that
              the  efficiency of  degradation
              was   approximately   4%/   re-
              gardless of the initial concen-
              tration.  Organopollutants  are
              often  present in  contaminated
              soils  and  sediments  in   the
              parts   per  billion  range  or
              less-   Thus it is  significant
              that the efficiency of degrada-
              tion  does  not  appear  to  be
              concentration  dependent.    It
              should  also be noted that  th«
              chemical  is  not  required  to
              induce  the  synthesis  of  the
              degrading  enzymes.    In  fact
              their  presence is not required
              at  all  because  this   enzyme
              system  is produced in response
              to nitrogen starvation.
                             408
-------
     Whereas nitrogen starva-
tion promotes both lignin and
DDT    degradation/    excess
nitrogen suppresses  degrada-
          both  chemicals  as
          by  CC»2   evolution
         Dicofol was found as
         intermediate of  DDT
             during  nitrogen
            and DDD   appears
tion  of
assayed
(1,2) .
a  major
metabolism
starvation
to be the major  intermediate
when  nutrient  nitrogen  was
not limiting (Fig.  4).   The
Dicofol   pathway  predominates
when  degradation  to   CC>2    is
favored  while the DDD   pathway
predominates  when  degradation
to  CC>2 is  suppressed.   These
results  seem to have   environ-
mental  relevance for DDD is   a
very recalcitrant environmental
pollutant   (4).   In  contrast/
Dicofol  does not appear to   be
persistent  in the  environment
(5).
            2.4 raM NUTRIENT NITROGIN
    1
    ASCO
           STANDARDS

           A - DBF + DICOFOL
           B - DDE
           C - ODD
           0 - 00T
            24n*i NUTRIBfT NITROGEN
     M1N
Figure  4.  Gas chromatograms
of  hexane extracts  obtained
from    cultures    of      P_-_
£h£yjsosp^rijijm  grown   in   the
presence  of  DDT   (4.92   uM)
under  high  (24 mM) and   low
<2-4  mM)  nutrient  nitrogen
conditions.   (DBF   -  4/4'~
dichlorobenzophenone).
                                SUMMARY

                                    These  studies  demonstrate
                                that   the   ability   of    P_-_
                                £?l£v.so§]22E.LEm.     to    degrade
                                recalcitrant   organopollutants
                                may  be enhanced by  increasing
                                the carbohydrate  concentration
                                of  the culture/  is independent
                                of   substrate concentration and
                                is  stimulated by nitrogen star-
                                vation,   it  also appears that
                                there  may be two pathways  for
                                the metabolism of  DDT.   Under
                                low  nutrient  nitrogen  condi-
                                tions   DDT  is  converted   to
                                Dicofol    which   is   further
                                degraded  to C<>2-   In contrast
                                high  nutrient nitrogen  condi-
                                tions promote formation of  DDD
                                and  further degradation to C(>2
                                is  suppressed.
                                ACKNOWLEDGEMENTS
                                     Th i s work
                                Cooperative
                                tCR811464,
                                Protection
                                Research
                                Hazardous
                                Research
                                Cincinnat i
                                Project Officer.
                                wish   to   thank
                                               was
                   supported  by
                      Agreement
            U.S.  Environmental
            Agency/  Office   of
            and    Development,
            Waste   Engineering
                    Laboratoryf
             OH/  p.R.  sferra/
                   The  authors
                 Ms-  Cathy   M-
                                Custer
            for
secretarial
                             409
-------
assistance  in  the  preparation
of this manuscript.
                           as a Miticide/  1974-  Pestjcides
                           !§2£li£2Ei!12  JSJiEUJ-i'   Vol.   13 /
                           No- 2t pp- 72-74-
REFERENCES
     Butnpus/   J.A./   M-  Tien/
     D.      Wright,  and S.D-
     Aust,  1985.  Oxidation of
     Persistent  Environmental
     Pollutants   by   a  White
     Rot Fungus/      Science*
     In Press-

     Kirk,  T-K-,  E.  Schultz,
     W.J.    Connors/    L.F.
     Lorenz/ and %J-G. Zeikus,
     1978-  Influence of Cul-
     ture    Parameters     on
     Lignin   Metabolism   by
     Phanerochaete      SilSLY^:
                         Arch_-_
                   Vol.   117,
sosporium/
Microbiol./
     pp. 277-285-

3.   Jeffries/  T-W-,  S.  Choi/
     and  T-K-   Kirk,   1981.
     Nutritional    Regulation
     of Lignin  Degradation by
     Phanerochaete    chryso-
     sporium-   Appl.  Environ.
     Mtcrobiol./ Vol. 42, NO.
     2, pp.  290-296.

4.   Menzer*  R..E-,  and J.O.
     Nelson/  198O. Water and
     Soil    Pollutants/    jEn:
     Casarett   and   Doull=*_s
     Toxicology?    The  Basic
     science of Poisons/  2nd
     Edition (J.   Doull/ C-D.
     Klaassen/     and    M.O.
     Amdur/  eds-)/ MacMillan
     Publishing Co./   Inc./
     New York/  pp. 632-658.

5.   Lyman/  W.R-/  and  R.J.
     Anderson/  1979. Dicofol
     Residues    in    United
     States  Soils  Having  a
     Known History of Its Use
                                            Disclaimer

                                This paper has been reviewed in
                                accordance with the U.S. Environ-
                                mental Protection Agency peer and
                                administrative review policies and
                                approved for presentation and publi-
                                cation.
                               410
-------
               ENVIRONMENTAL VAULT - A NEW CONCEPT IN LAND STORAGE
                              William B. Philipbar
                      Rollins Environmental  Services,
                           Wilmington, Delaware 19899
            Inc.
                                    ABSTRACT

     A large volume of hazardous residues can only be treated and/or reduced
in volume to a certain degree and then must be stored or disposed of in a man-
ner that will isolate the residues from the environment for a period of time.
Presently, these materials are, for the most part, being deposited in conven-
tional below ground landfills which can create potential problems that are
associated with this type of disposal.

     The Environmental Vault has recently been developed as a long term storage
replacement for the below ground landfill (U.S. Patent 4,464.081, August 7,  1984).
The Vault is an above ground structure designed to isolate wastes from the sur-
rounding environment and the environment from the wastes.

     The major advantages of the Vault are;  .1. Location is not dependent on
the geology and hydrology of the site;   2. Triple liners provide failsafe
groundwater and surface water protection;  3. All leachate and monitoring con-
trols are dependent on gravity flow rather than mechanical pumping;  4. The
above ground construction allows: visual observation of sides and top to check
for possible malfunction of the system, easy remedial response to correct Vault
problems, retrieval of the stored wastes if at some future date new technology
or changing economies make recovery or future treatment possible;  5. Exemption
under the Federal Act, "The Hazardous and Solid Waste Amendments of 1984" from
groundwater monitoring requirements under certain circumstances.
INTRODUCTION

     The need for long term storage of
hazardous waste residuals is going to
be with us for a long time to come.
Today these residuals are being stored
in and on the ground utilizing below
ground landfills, pits, ponds and la-
goons, as well as waste piles.  The
"Hazardous and Solid Waste Amendments
of 1984" - RCRA Reauthorization, is
going to greatly restrict what hazar-
dous waste can be land disposed and it
will also set rigid standards on the
construction and operation of disposal
facilities.  Because of these regula-
tions and because of the past problems
with below ground landfills and the
possible liabilities that these fills
may present, it is doubtful whether any
new below ground landfills will be per-
mitted in the United States.  This,
along with the fact that many presently
permitted landfills are either nearing
capacity or are being phased out be-
cause of the new forthcoming RCRA regula-
tions, presents a major problem in hazar-
dous waste management.
                                      411
-------
     Inorganic residuals, some organic
residuals, contaminated soils and in-
cinerator ash and sludges require long
term storage, and if the present stor-
age options are restricted and/or eli-
minated, new techniques are going to be
required to manage these types of waste.
The need for secure long term storage,
in fact, is going to increase over the
next decade as we remediate our problems
of the past through Superfund and the
private sector.  A large percentage of
the contaminated material at these sites
is soil, with relatively small concentra-
tions of toxic pollutants.  These soils,
and the possible leachate from these
soils, must be isolated from the ground-
water.  However, it makes no sense to
transport this contaminated soil, possi-
bly hundreds of miles, and deposit it
in an existing below ground landfill
that could be the next generation of
Superfund activity.  What is needed is
a secure storage design that can be
easily constructed at the site that is
to be remediated, that is tolerant of a
variety of hydrogeological conditions,
is cost effective, and will protect the
environment from the waste and the
waste from the environment.
PURPOSE

     The Environmental Vault is a new
development that satisfies these needs.
It is an above ground structure that is
designed for long term storage of hazar-
dous wastes.  It protects the wastes from
precipitation, ground and surface waters,
and it protects the ground and surface
waters from the wastes and any leachate
that is produced.  Because the Vault is
above ground, it can be constructed at
sites with a variety of hydrogeological
conditions.

     Other advantages of the Vault are:

     * A triple liner system for maximum
protection.
     t All monitoring and leachate systems
are gravity fed and not dependent on mech-
anical devices.

     a Visually inspected for possible
problems.

     a Stored wastes can be retrieved in
the future if new technology or economic
considerations make further treatment or
recovery practical.

     In constructing the Vault, the site
is first prepared as you would in build-
ing a house on a concrete slab foundation.
Next a concrete base is constructed with
concrete walls using state-of-the-art
civil engineering construction techniques
to insure a secure structure.  This outer
concrete construction forms the contain-
ment structure and, after it has been pro-
perly sealed, also becomes the tertiary
liner.  Secondary and primary liner sys-
tems are placed over the base, with inter-
spaced drain zones between the liners to
catch and drain any leachate formation.
The materials of construction of the pri-
mary and secondary liners will be deter-
mined by the nature of the wastes being
stored, however, as a general rule a 100
mil high density polyethylene liner will
be used.

     The drain zones between the waste
and the primary liner, and between the
primary, secondary, and tertiary liners
incorporate a monitoring pipe network
that through gravity flow collects and
delivers any leach to leachate collection
tanks that can be instrumented to measure
the volumes of any liquids that are col-
lected.  This system will not only col-
lect and store leach contained by the pri-
mary liner, but monitor the integrity of
the liners by detecting any liquid flow
between the liners.
                                         412
-------
                  Environmental
Vault
       Cover System
Containment Structure

      Waste Material
   Multiple Liners and—
  Monitoring Systems
             Grade
                                             Rolled, Compacted Sub-Grade
                                                              Umbrella Cap
                                                              Storm Water
                                                              Collection

                                                              Cap Monitoring
                                                              System

                                                              Secondary Cap
                                                             Compacted, Stabilized
                                                             Waste
                                                             Containment Wall
                                                             Drain/Protective
                                                             Layers

                                                             Leachate
                                                             Collection System

                                                             Monitoring Systems

                                                             Original Grade
                                                             Clay or Concrete hrtiary
                                                             Liner (Base)
                                           413
-------
     The Vault cap includes two polymer
liners, namely an umbrella cap and a
secondary liner.  A collection system
is installed above the umbrella cap for
stormwater collection and a second col-
lection system is installed between the
umbrella cap and the secondary liner so
as to monitor the integrity of the um-
brella cap.  A ballast material or top-
soil and cover combine to serve as an
excellent water runoff system and also
provide an aesthetically pleasing
appearance.  The surface of the cap
could, however, be paved if this would
serve a purpose.  A vent system is de-
signed to collect any gases produced by
the wastes and vents them through ab-
sorption mediums such as carbon and then
to the atmosphere.

     The wastes that are deposited in
the Vault are first treated to stabi-
lize and reduce their Teachability.
They are further compacted when they are
placed into the Vault.

     The Vault can be constructed in a
variety of sizes.  An optimum size is an
acre to one and a half acres in area and
20 to 25 feet in height, which provides
a volume of 30,000 to 45,000 cubic yards.
If the demand warrants it, several of
these Vaults can be constructed at a
given site.
storage .   With the use of the Rollins
Environmental Vault the problem hazar-
dous waste site can be cleaned up and
made available for other uses.  In con-
trast, when groundwater containment
barriers or similar remedial  actions  are
used, the site remains unusable for othe
purposes.   Using vegetation and land-
scaping techniques, the land  is freed fo
alternative uses.
CONCLUSION

     In review, long term storage of
hazardous wastes will be needed in this
country for many years to come.  In fact,
with the remediation of sites that have
been polluted in the past, needs for this
type of storage will increase over the
next decade.  New RCRA regulations are
going to restrict the construction and
operations of land disposal/storage faci-
lities.  The Environmental Vault is a new
technique for environmentally secure
                 Disclaimer

   The work described in this paper was
   not funded by the U.S. Environmental
   Protection Agency.  The contents do
   not necessarily reflect the views of
   the Agency and no official  endorse-
   ment should be inferred.
                                       414
-------
                       IN SITU TREATMENT TECHNOLOGIES AND SUPERFUND
                  Michael Amdurer, Robert Fellmaa and Salah Abdelhamid
                              Ebasco Services Incorporated
                                 Two World Trade Center
                                   New York,  NY  10048


                                        ABSTRACT
    In situ systems for treating waste deposits involve three essential elements:
 selection of a chemical or biological agent which treats the waste,  a method for
 delivery of the reactant to the deposit and a method for recovery of the reaction
 products or mobilized waste*

    Four methods for in situ treatment (or extraction) of subsurface organic wastes are
 reviewed: blodegradation, surfactant-assisted flushing, hydrolysis and oxidation.  These
 methods have potential application to certain organic wastes, and each has potential
 drawbacks.  Biodegradation and surfactant-assisted flushing are among most promising
 insitu treatment methods.  An example of the use of in situ biodegradation is provided.

    The underlying limitation to use of the technologies associated  with in situ treat-
 ment of or deposits is that these methods are often in an experimental stage of
 development.  This finding frustrates efforts to systematize selection of reagent and
 delivery and recovery systems and emphasizes the need for laboratory simulation  and
 testing prior to implementation.  Each site at which in situ treatment is considered
 becomes, in essence, a research project.  Although the need for in situ treatment
 methods is underscored by the expense and impacts associated with such practices as
 excavating, transporting and landfilling waste deposits, the use of  in situ methods on a
 large scale is hampered by the lack of field experience.  There is little incentive for
 site managers to consider their use.  The tasks which could be undertaken at the
 programmatic level to enhance the development and implementation of  in situ treatment
 methods for the remediation of hazardous waste sites are discussed.
INTRODUCTION

    In December 1982, Envirosphere Co.
a division of Ebasco Services Inc. was
selected by JRB Associates to undertake
a project entitled "Evaluation of Systems
to Accelerate Stabilization of Waste
Piles or Deposits".  The EPA project
officer is Dr Walter Grube of HWERL.
The EPA project number is 68-03-3113,
Task 37-2.
    This project represents Phase 1 of a
two phase scope of work whose purpose
was to document the feasibility of engi-
neered approaches to treating subsurface
waste deposits via the application of in
situ methods.  Phase I concentrated on
applications of available technology and
examined the limitations imposed on
their use by site and waste specific
characteristics.  A future phase of the
project, Phase II, is expected to under-
                                      415
-------
 take beach scale or pilot studies to
 expand the data base available to
 potential users of in situ methods.

 APPROACH

     Envirosphere pursued a six part
 program of investigation which consisted
 of a literature review,  a definition of
 the capabilities and limitations  of
 delivery,  recovery and treatment  tech-
 nologies,  visits to sites where remedial
 activities were underway, a definition
 of laportant site and waste characteris-
 tics, and  an evaluation of remedial
 technologies.   The available data were
 evaluated  to determine classes of organic
 chemicals  amenable to treatment by
 various potential In situ treatment
 methods.   Potential delivery and  recovery
 systems for these treatment agents were
 then evaluated  with respect to site
 hydrogeologic characteristics.

     Bavirosphere then developed a guid-
 ance manual which identified combinations
 of delivery/recovery technologies and
 reagents which  have a reasonably  high  or
 clearly low probability  of success.

 PUREOSE

 1}  To  present  a concise summary  of
     Envirosphere's  findings  relative to
     selection of an in situ  treatment
     option,

 2)   To  illustrate an example of a
     successful  application of in  situ
     techniques,

 3)   To analyze  the programmatic barriers
     to implementing  In situ  treatment
     technologies within  the  Superfund
     program.

     With respect  to  the  Phase  I effort,
 EaviroBphere has  submitted its  draft
 final report covering the  evaluation of
 in situ treatment systems  to EPA  for
final approval.   This document  is
 expected to be  published late in  1985.
 The  report Is a guidance document which
will assist  to  site  managers in choosing
 remedial action, involving in situ
 treatment especially in  the  early phases
of technology selection within  the FS
 process.
 EROBLEMS ENCOUNTERED

     In the  study  of in situ  treatment of
 hazardous waste piles or  deposits  one
 encounters  a  significant  difficulty
which  prevents the drawing of many
 generalizations from the  extant liter-
 ature.   Each  in situ application
 resembles a research effort  which must
 be customized to  the site, and waste
 characteristics.  The essence of a
 successful  application of an in situ
 method is the performance of a treat-
 ability study designed to account for
 the  peculiarities of the  waste and treat-
ment reagent  combination as well as the
 unique  geohydrologlcal characteristics
 of the  site.  Since treatabillty studies
 cannot  exist  for  the generalized case,
almost  all  conclusions to be drawn from
 the literature survey were necessar-
ily based upon engineering judgement.
 Verification  of hypotheses by reference
 to documented field experience was not
 feasible in most  situations.

    Another problem encountered was the
degree  of homogeneity of  the waste
deposit.  Subsurface deposits contained
in drums or within non-uniform forma-
 tions which Impede the flow of water-
borne reagents cannot be considered as
realistic candidates for in situ treat-
ment.   The experience that exists
strongly suggests that the greatest In
situ success will be with a plume or a
spill situation rather than with a
source  deposit itself.

RESULTS

    There are many different viewpoints
as to what constitutes application of in
situ methods.  A workable definition Is
that insitu treatment means treatment of
contaminated soil without excavation.
This definition encompasses four dif-
ferent  possibilities for in situ methods:
Treatment Involving the  inground trans-
formation of contaminants to harmless
products; Stabilization  involving the
inground chemical  immobilization of
contaminants; Solidification Involving
the in—ground physical immobilization of
contaminants and Extraction involving a
subsurface displacement  followed
(usually) by chemical treatment on the
surface.
                                           416
-------
     This paper reports on in situ treat-
 ment methods and certain extractive
 methods applicable to target organic
 contaminants.

 Biological Renovation of Waste Deposits

     Aerobic and anaerobic bacteria,
 fungi, actlnomycetes, algae and
 cyanophytes (blue-green algae) have all
 been shown capable of degrading many
 classes of organic chemicals.  These
 microbes include natural microbial
 populations, adapted microbial cultures
 and potentially bioenglneered microbial
 strains (1,2),  Once the extent of the
 contamination and its chemical character-
 ization have been determined, the proper
 microorganisms (or groups of microbes)
 may be identified and developed.  The
 identification of the proper agents for
 •Haste  site renovation is based upon past
 experience,  laboratory screening, and
 onsite pilot-scale tests.

     To date, aerobic bacteria such as
 pseudomomas  have been most commonly used
 for in situ  biodegradation of contamin-
 ants.   These organisms can potentially
 completely convert the organic compounds
 to  C02 and water,  and do not produce
 H2S or methane as  reaction products.
 However, anaerobic bacteria are
 important  for  the  biodegradation of
 pesticides and halogenated organics.
 Organic contaminants that have been
 successfully treated by biodegradation
 include phenols, gasoline and other
 petroleum  products,  methyleae chloride,
 alcohols and acetone (3,  4).

     In the process of designing the
 microbial waste treatment system,  one
 must determine the oxygen,  emulsifier
 (if the wastes are  insoluble)  and
 fertilizer requirements  for optimum
 waste  treatment rates.  Microbial  agents
 require the  maintenance of  sufficient
 concentrations of  nitrogen,  phosphorus
 and trace elements,  and a  pH  range that
 will support their growth.   The  levels
 of  these factors at  the site  should be
determined during  the  site  investiga-
 tion;  the need for additional fertil-
 izers  or buffers required  to  support
microbial growth can  then  be  identified.
     Biological renovation of subsurface
 waste deposits poses problems relating
 to oxygen supply, temperature, perme-
 ability and accessibility not encountered
 with surface disposal sites.  Injection
 wells may be established into and below
 the waste site to deliver a fertilizer
 and oxygen supply (5).  Oxygen sources
 would include injectable solutions of
 peroxides, oxygen-charged water produced
 by ozonation (6), or direct sparging of
 air into the groundwater (3).  Recovery
 wells or trenches should be situated at
 points peripheral to or downgradient of
 the waste deposit.  Flow patterns estab-
 lished between injection and recovery
 wells should be planned to aid in confin-
 ing the waste during the renovation
 process.   la this way groundwater plumes
 that may be migrating from the site can
 be renovated as well.

 Application of Hydrolysis  toWaste
 Deposit Stabilization

     Ifydrolysis is a  chemical reaction
 involving the cleavage of  a molecular
 bond by reaction with water.   The  rates
 of hydrolysis for some compounds  can be
 accelerated  by altering  the  solution pH,
 temperature,  solvent composition,  or by
 introducing  catalysts.   For  in situ
 treatnent, alteration  of pH,  particu-
 larly raising the pH (base-catalyzed
 hydrolysis),  is the  most promising
 approach.   The range of  chemical classes
 potentially  treated  by base-catalyzed
 hydrolysis includes  amides,  esters,
 carbamates, organo-phosphorus compounds,
 pesticides and herbicides.   Base
 catalyzed hydrolysis has been success-
 fully used for treatment of  surface
 spills  of acrylonitrile and pesticides
 (7).

    The primary design concern for
 implementation of  base—catalyzed
 hydrolysis within a waste deposit will
 be  the  production and maintenance of
 high pH (9 to 11) condition with
 saturation or high moisture content in
 the waste deposit.  For shallow sub-
 surface or surface deposits, surface
application of lime,  sodium carbonate or
 sodium  hydroxide followed by surface
application of water may be appropriate.
                                           417
-------
 For deeper deposits,  subsurface delivery
 or Injection of alkaline solutions may
 be required.

 Potential for In Situ Oxidation of Waste
 Deposits

     Ihe  potential application of three
 oxidants (ozone,  hydrogen peroxide,  and
 hypochlorites)  to waste deposits was
 evaluated.   Although  in widespread use
 in surface  water treatment applications,
 significant problems  nay preclude their
 effective implementation as in situ
 treatment agents  for  waste deposits.

     Hypochlorite  reacts with organic
 coapounds as both a chlorinating agent
 and an oxidizing  agent.   Documentation
 on the effectiveness  of nypochlorite as
 an oxidizing agent for organic Materials
 IB extremely limited.   Hypochlorite  addi-
 tions nay lead  to production of undesir-
 able chlorinated  by-products (e.g.,
 chloroform)  rather than  oxldative degra-
 dation products.   Therefore the use  of
 hypochlorite for  in situ treatment of
 organic  wastes  is not  recommended.

    While ozone is an  effective oxidiz-
 ing agent for many organic compounds in
 wastewater  treatment applications, its
 relatively  rapid  decomposition rates in
 aqueous  systems,  particularly in the
 presence of certain chemical contami-
 nants or other  agents  which catalyze its
 decomposition to  oxygen,  preclude its
effective application  to subsurface
waste deposits.   The half-life of ozone
 in groundwater  la less than one-half
 hour (10).   Considering  that flow rates
of  water through waste deposits  are
likely to be on the order of inches/hour
or less, it is unlikely  that effective
oxidant doses of  ozone can be delivered
outside  of the  immediate  vicinity of the
 point of application.  Successful use  of
ozone for in situ  chemical oxidation is
unlikely.  However, ozonation has been
used successfully  to supply oxygen for
aicrobial biodegradation,  and to
chemically oxidize complex organics in a
surface  reactor to simpler compounds
that are more readily biodegradable
 (6).  Ihis use of ozone as  a supplemen-
tary treatment for biodegradation seems
promising.
     Hydrogen  peroxide is a weaker oxidiz-
 ing  agent  than ozone; however, its sta-
 bility in  water  is considerably greater.
 Since  decomposition of hydrogen peroxide
 to oxygen  may be catalized by Iron or
 certain other metals, effective delivery
 of hydrogen peroxide throughout an entire
 waste  deposit may be difficult or impos-
 sible  because of the relatively low
 transport  velocities achievable in waste
 deposits.  Prior to consideration of
 hydrogen peroxide as an in situ treatment
 method, It will be necessary to investi-
 gate the stability (or rate of decompo-
 sition) of hydrogen peroxide In the
 specific waste deposit matrix.  Hydrogen
 peroxide may  also be used as an oxygen
 source  for mlcrobial biodegradation (9).

 Surfactant-Assisted Flushing or Solu-
 blllzatlon of Wastes

     Flushing  or mobilization of wastes
 can  serve  two purposes:  to promote the
 recovery of wastes from the subsurface
 for  treatment on the surface, or to
 solubillze adsorbed compounds in order
 to enhance the rate of other in situ
 treatment techniques (such as biodegra-
 dation  or  hydrolysis).  Flushing or
 mobilization using water alone may be
 sufficient for relatively soluble
 compounds such as phenols;  however,  the
 use  of  surfactants will be required for
 significant solublllzation of insoluble
 (hydrophobic) compounds.

     Surfactants (surface active agents)
 are  a class of natural and synthetic
 cheoicals which promote the wetting,
 solubilizatlon, and emulsificatlon of
various types of organic chemicals.   A
 simple approach to evaluating the
 potential use of surfactants in organic
waste recovery involves consideration of
 the aqueous solubility or octanol-water
 partition coefficient, K^,,.  Sur-
 factants would be most effective in
 promoting the mobilization of organic
 compounds of relatively low water
 solubility and high Kow values.

    Laboratory tests  suggest that sur-
 factants nay enhance the recovery of
subsurface gasoline leaks by groundwater
pumping, and promote the mobilization of
 crude oil and PCBs fron soils (10).
                                            418
-------
 However,  certain environmental factors
 may  reduce  the in situ effective-
 ness of surfactants.  These include
 precipitation of the surfactant by
 groundwater with high TDS or alkaline
 earth cation concentrations (Ca, Mg);
 reduction of surfactant effectiveness
 due  to non-optimal pH or temperature; or
 adsorption  of the surfactant by soil
 particles,  negating its solublliziag
 properties.  Nevertheless, the use of
 surfactants either alone (to flush
 otherwise insoluble organics) or in
 combination with other treatments (to
 solubilize  the waste materials and
 thereby promote biodegradation) is a
 promising avenue for further research.

     The potential applicability of the
 four above  treatment technologies versus
 a wide array of target organic compounds
 is presented in Table 1.

 Case History - Biodegradation

     Probably the most successful insitu
 method for  remediation of soil contam-
 ination and groundwater restoration is
 biodegradation.  Success has been
 reported  (4) in treatment of spills and
 groundwater contaminated by methylene
 chloride, ethylene glycol, isopropanol,
 phenolic  compounds, acetone and tetra-
 hydrofuran.  Other references cite
 successful attempts using biodegradation
 in treatment of methylene chloride (3),
 and  petroleum based hydrocarbons and
 nexaeyanoferrate ion (4).

     One of the best documented case
 histories on biodegradmtion is the case
at Biocraft Laboratories in Waldwlck,
N,J. (11).  Biocraft Laboratories is a
 small synthetic penicillin manufacturing
 plant located on a 4.3 acres site in an
Industrial park In Waldwick,  N.J.   In
August 1975, contamination was observed
in a small creek which was traced to a
 storm sewer draining the site and
subsequent investigations revealed a
leak in underground storage process
lines connecting solvent storage
facilities.  The Inventory of contam-
inants estimated to have escaped  into
 the environment is presented  in Table 2.
 TABLE 2.   BIOCRAFT SITE ESTIMATED
     CONTAMINATION INVENTORY IN SOILS  (3)
 Methylene  Chloride   CH2C12     181500 Ib
 N-Butyl Alcohol      C^gO      66825 Ib
 Dimethyl Aniline     %H11N      26300 Ib
 Acetone              C3H6°       10890 Ib
    An approximately 1.75 acre area and
1200 yd  of soil were contaminated.  A
municipal well located about 1000' east
of  the property was believed to be in
jeopardy of contamination.  Blocraft
undertook an extensive groundwater pro-
filing, sampling and analysis program.
A total of 22 wells with continuous
recorders were constructed on the site
and a  detailed description of the
groundwater regime was generated.

    A  number of alternative response
technologies were considered including
collecting and treating all discharge
from the storm sewer; isolating the
storm  sewer from the contaminated flow
by  grouting pipe joints, pipe resleeving
or  pipe replacement! surrounding the
area with grout or a slurry cutoff wall
or  excavating the entire contaminated
soil column under the site.  Selective
pumping of groundwater and offsite
disposal was chosen as the remedial
method.  However, to undertake this
remedial measure, contaminated ground-
water  had to be shipped off-site for
disposal at an average estimated cost of
$0.35  per day (12).

    Driven by the high cost and ongoing
uncertainty associated with the effec-
tiveness of the initial remedial method
in  treating the contaminant, Biocraft
developed the selected alternative:
collecting the contaminant plume down-
gradient In a slotted pipe collection
trench, treating the collected ground-
water in a surface aerobic biological
treatment system, injecting the treated
"bioactive" water upgradient in two
slotted pipe recharge trenches and
enhancing insitu biological treatment in
the subsurface through Injection of  air
in equally spaced air injection wells.
                                          419
-------
    A detailed description of the system
which has  been patented by Groundwater
Decontamination System, a subsidiary of
Biocraft Laboratories,  appears In
reference  11.   Of considerable interest
is  the  performance of the system reported
to  date.   Tables 3 and 4 provide
summaries.
TABLE  3.   AVERAGE REMOVAL EFFICIENCY
        OF AEROBIC TREATMENT SYSTEM (3)
Target  Contaminant

Methylene  Chloride
n-Butyl Alcohol
Acetone
Dimethyl Aniline
Reported Removal
   Efficiency

     98%
     98%
     97%
     93%
TABLE 4.   PRE-ANB  POST-TREATMENT RESULTS
          FOR METHYLENE  CHLORIDE
          AT BIOCRAFT SITE (3)

   Well.   1981  (Ere)   1983 (Post)

   4A      67 ppa      ND ("Not  Detected)
   25     260 ppm      ND
   26     360 ppm        8 ppb
   31      78 ppm      10 ppb
   P-13   989 ppm      ND
   P-30   880 ppm      23.5-64 ppm
   B-32A  305 ppm      305-182 ppm

    It is reported that  the overall
remediation at  the site  is proceeding
successfully (13)  and that 60%  of the
treatment occurs above ground (in the
reactor) and that  40% occurs below the
surface.  It is estimated that  the
system will require 5 years to  com-
plete the cleanup  at  the Biocraft Site
compared with 15-20 years using the
Initial groundwater withdrawal  and off-
site treatment  alternative.

    Capital and Operating and Maintenance
costs are documented  at  &926K and $2267
day respectively (11).   The total cost
at the Biocraft Site  using the  system
now in operation is estimated to be 1/4
of the total cost  which  would be
Incurred via the initial remedial
measure (pumping and  offsite disposal)
(13).
    This one example points up the poten-
 tially  significant advantages of in situ
 treatment.  However, an important point
 to consider is that the preliminary
 research and pilot studies aspect of the
 overall capital cost was about $450,000
 or 1/2  of the total capital cost.  Thus,
 in order to realize the ultimate savings,
 a significant investment was required in
 site specific research.  The necessity
 for significant treatability studies and
 research is a fundamental element in
 virtually all in situ remediations and
 represents a significant risk In
 potential time and money to the
 prospective user of in situ methods.
 Some of the barriers to employing in
 situ technologies for remediation at
 hazardous waste sites are now discussed
 along with suggested solutions to these
 implementation problems.

 NOH-TECHNICAL CONSIDERATIONS

    The preceding sections presented
 several ambitious in situ treatment sys-
 tems for application at hazardous waste
 sites.  The primary objectives behind
 EPA's funding of this type of research
 and development program is to accelerate
 the rate of technology development and
 transfer in this area and also to pro-
 mote its adoption at actual sites.  How-
 ever, in addition to the numerous tech-
 nical challenges to be met in developing
 any of  the technologies discussed in this
 article, there are others which are non-
 technical in nature that must be met to
 enhance the chances of developing and
 applying these energing technologies to
 hazardous sites.  The following para-
 graphs  briefly outline these major
 challenge areas and suggest general
 approaches to dealing with them.

 Funding

    Few technologies have been evaluated
 in the  context of a site-specific
 application.  Because of the uncertainty
 involved in applying an innovative
 remedial approach,  there will always be
 reluctance to test their feasibility.
 In the case of Superfund sites,  the
 skepticism is enhanced by the political
and economical restraints which would
make it impossible for a regulatory
                                          420
-------
 agency,  the public, and  engineers  to
 risk pursuing what is  perceived as an
 "R&D study of an unproven technology".

     One  major area of  Improvement  is
 funding.   Funding by EPA and other
 government agencies (e.g., NSF), as well
 as industrial trade organizations  (e.g.,
 CMA) is  recommended to accelerate  the
 development and  application of the
 subject  technologies.  Specifically, the
 following  are suggested:

     Funding of Superfund  site-specific
 R&D projects should be initiated and
 expanded.   It is recommended that  con-
 siderations be given to hazardous waste
 sites that are not on  the NPL.  This
 would enhance the chances of a feasi-
 bility study being conducted on sound
 technical  bases  that are not complicated
 by the typical financial, political and
 institutional considerations that are
 characteristic of NPL  site work.

     Congress and EPA should work out a
 formula  to ensure that an R&D budget for
 in-situ  treatment technologies is
 established in linkage with the budget
 for  the REM programs which have the
 overall responsibility for ascertaining
 the  need for, and utility of, these
 innovative  remedial technologies.

 Institutional Framework

     The objectives and approaches to site
 remediation within the context of Super-
 fund, is defined by EPA in the National
 Contingency Plan (NCP), and substantially
 shaped by various articles in the NEPA,
 CWA, RCRA and TSCA regulations.  These
 environmental regulations and guidelines
 form the basis for assessing environ-
mental and health risks,  development and
evaluating remedial measures,  but most
 importantly ranking and selecting the
 remedial alternatives.  The general
 scope, specific contents, and enforce-
ment status of these regulations affect
 in many ways the perceived urgency of
applying in situ treatment methods at
 hazardous sites and,  in most  cases, the
 feasibility of their design,  permitting,
 Implementation and operations.   Cases in
point; In the early years of  Superfund,
little emphasis was placed on considering
 in situ treatment techniques at NPL
 sites.   This is attributed primarily to
 the perception that  the RCRA regulations
 regarding secured landfilling of hazard-
 dous waste were reasonably achievable.
 As additional experience with such
 landfills was gained, technical guidance,
 permitting,  and later on RCRA amendments,
 tended  to impose substantial constraints
 on land disposal of  hazardous wastes.
 As a result, a great deal of research,
 development  and implementation is needed
 regarding in situ treatment methods.  It
 is believed  that EPA and state agencies
 will, and should,  continue to limit
 secured landfilling. This, coupled with
 an intensified enforcement of the regula-
 tions,  should increase the attention
 given to the development and application
 of the  in situ treatment technologies.

     The experience with remedial site
 projects Indicates that permitting of an
 in situ treatment  operation and associ-
 ated field pilot  studies can be critical
 to the  successful  development of a
 process.  The uncertainty associated •
 with subsurface geochemical and
 contamination conditions pose real
 difficulty with regard to providing
 specific  effluent  and other data
 typically required in a permitting
 process.  Moreover, the complexity of
 these technologies and the permitting
 process itself, are expected to result
 in delays that can substantially hinder
 cost-effective feasibility demonstra-
 tions, particularly at NPL sites.   It is
 therefore recommended that consideration
 be  given  to develop and implement  a  more
flexible  permitting policy with regard
 to  field  testing and operation of  in
 situ treatment technologies.

    The financial resources,  or some-
times the perception thereof,  required
for the development,  and tailoring to a
site condition, of an in situ treatment
process can be great. This can be a
hindrance with respect to getting
private responsible parties (PRPs) to
"take a chance" on trying to  develop  and
apply an in situ treatment method at  a
given site.  It is suggested  that
efforts be made to creating effective
incentives for PRPs to pursue  innovative
remedial approaches.
                                       421
-------
 agency, the public, and engineers to
 risk pursuing what Is perceived as an
 "RSD study of an unproven technology".

     One major area of improvement is
 funding.  Funding by EPA and other
 government agencies (e.g., NSF), as well
 as industrial trade organizations (e.g.,
 CMA) is reeoniaended to accelerate the
 development and application of the
 subject technologies.  Specifically, the
 following are suggested:

     Funding of Superfund site-specific
 BSD projects should be initiated and
 expanded.  It is recommended that con-
 siderations be given to hazardous waste
 sites that are sot on the NFL.  This
 would enhance the chances of a feasi-
 bility study being conducted on sound
 technical bases that are not complicated
 by the typical financial, political and
 institutional considerations that are
 characteristic of NFL site work.

     Congress and EPA should work out a.
 formula to ensure that an R&D budget for
 in-situ treatment technologies is
 established in linkage with the budget
 for the HEM programs which have the
 overall responsibility foz ascertaining
 the  need for,  and utility of, these
 innovative remedial technologies.

 Institutional  Framework

     Ihe  objectives and approaches to site
 remediation within the context of Super-
 fund, is  defined by EPA in the National
 Contingency Plan (NCP),  and substantially
 shaped by various articles in the NEPA,
 CHA, RCRA and  TSCA regulations,   these
 environmental  regulations and guidelines
 fora the basis for assessing environ-
mental and  health risks,  development and
 evaluating  remedial measures,  but most
inportantly ranking and selecting the
 remedial  alternatives.   Hie general
 scope,  specific contents,  and enforce-
ment status  of these regulations affect
 In many ways  the perceived urgency of
applying  in situ treatment methods at
hazardous sites and,  in most cases,  the
 feasibility of their design,  permitting,
 implementation and  operations.   Cases in
point:

     In the early years  of Superfund,
little emphasis  was  placed on considering
 in situ treatment  techniques  at  NFL
 sites.   This is attributed primarily to
 the perception that  the RCRA  regulations
 regarding secured  landfilling of hazard-
 dous waste were reasonably achievable.
 As additional experience with such
 landfills was gained, technical  guidance,
 permitting,  and later on RCRA amendments,
 tended  to Impose substantial  constraints
 on land disposal of  hazardous wastes.
 As a result,  a great deal  of  research,
 development and implementation is needed
 regarding in situ  treatment methods.  It
 is believed that EPA and state agencies
 will, and should,  continue  to limit
 secured landfilling. This,  coupled with
 an Intensified enforcement  of the regula-
 tions,  should increase the attention
 given to the  development and  application
 of the  in situ treatment technologies.

     The experience with remedial site
 projects indicates that permitting of an
 in situ treatment  operation and associ-
 ated field pilot studies can  be critical
 to the  successful  development  of a .
 process.   The  uncertainty associated
 with subsurface geochenical and
 contamination  conditions pose  real
 difficulty with regard to providing
 specific effluent and other data
 typically  required in a pernitting
 process.  Moreover, the complexity of
 these technologies and the permitting
 process  itself, are expected  to result
 In  delays  that  can substantially hinder
 cost-effective feasibility demonstra-
 tions, particularly at NFL sites.  It is
 therefore  recommended that consideration
 be  given to develop and implement a more
 flexible permitting policy with regard
 to  field testing and operation of in
 situ treatment  technologies.

    The financial resources, or some-
 times the perception thereof,  required
 for  the development,  and tailoring to a
 site condition, of an in situ  treatment
 process can be great.  This can be a
 hindrance with respect to getting
 private responsible parties (PSFs) to
 "take a chance" on trying to develop and
apply an in situ treatment method at  a
 given site.  It Is suggested  that
efforts be made to creating effective
 incentives for PRPs to pursue innovative
remedial approaches.
                                        422
-------
 Superfund's Programmatic Considerations
 CLOSING
     Several issues related to the scope
 and approach of the Superfund Program
 are noteworthy as they relate to develop-
 ment and application of innovative
 remedial technologies.  The following
 paragraphs focus on the remedial
 investigation/feasibility study (RI/FS)
 stage of a site's remedial response
 chronology.

     The  basic planning process of an
 RI/FS needs to be reconsidered to enable
 the early identification and detailed
 evaluation of innovative in situ treat-
 ment technologies.   These are currently
 eliminated at the initial screening
 stage, primarily on the grounds of
 inadequate definition of engineering and
 environmental performance character-
 istics.

     Exceptions must be taken to the
 established "rules  of thumb" applicable
 to  "guesstimating"  preliminary budgets
 and schedules for RI/FS tasks when an
 innovative technology is considered.
 Depending on the site conditions and the
 particular technology,  both budget and
 schedule of  an RI/FS would be extended
 to  allow for the requisite site  and
 bench scale  studies.

    A practical rational approach needs
 to  be developed to  perform cost-
 effectiveness  evaluation of  competing
 remedial alternatives which does not
 penalize emerging in situ treatment
 technologies on the  basis of cost and
 lack of  track  record.

    Community Relations  Program  (CRP)
 for a RI/FS  that considers an innovative
 in  situ  remedial technology will  involve
 a higher level of sensitivity and
 participation  from the public.  This
 must be carefully planned  for in advance
 since the public perception of these
 remedies will often be that  "the waste,
and associated  hazards, are  staying in
 the community."  The program should also
 highlight the growing limitations and
 restrictions associated with secured
landfilling and other conventional
 remedies.
     The application of in situ treatment
 technologies  to hazardous waste site
 represents a  technical, programmatic and
 economic challenge ia which significant
 research expense must be incurred at
 early  stages  in the remedial action.
 The  existing  early experience points to
 cost effective possibilities especially
 for  plume and spill remediation.  In
 situ treatment represents one of the
 possibilities for escaping from the
 secured landfill as the ultimate
 repository for hazardous waste.  Changes
 as suggested  above in the approach to
 remediation at the program and regulatory
 level are needed to nurture technical
 advances in situ technologies.

 REFERENCES

 1.  Atlas, R.M., 1981.  Microbial
    Degradation of Petroleum Hydro-
    carbons:  An Environmental
    Perspective.  Microbiological
    Reviews 45:180-209.

 2.  Kobayashi, H. and B.E.  Rittman,
    1982.  Microbial Removal of
    Hazardous Compounds.  Environmental
    Scienceand Technology.  16:17QA-183A.

 3.  Jhaveri, V.  and A.J. Mazzacca,  1983.
    Bio Reclamation of Ground and Ground-
    water - Case History.   4th National
    Conference on Management of
    Uncontrolled Hazardous Waste Sites,
    310 ct - 1 Nov 1983, Wash.  D.C.

4. " Flathman, P.E. and J.A.  Caplan,  1985.
    Biological Cleanup of Chemical
    Spills.   Proceedings of  Hazmacon 85,
    23-25 April 1985,  Oakland,  Calif.

5.  API Committee on Environmental
    Affairs, 1982.   Enhancing  the
    Microbial Degradation of Underground
    Gasoline by  Increasing Available
    Oxygen.   Texas Research  Institute.

6.  Rice, R.G.,  1984.   Purification and
    Recycling of  Groundwater Contaminated
    With Petroleum Products  and Cyanides.
    5th Conference on  Management of
    Uncontrolled Hazardous Waste Sites,
    7-9 Nov  1984, Washington, D.C.
                                          423
-------
 7.  Harsh, K.M., In Situ Neutralization
    of an Acrylonitrile Spill.
    Proceedings of1978 Con£. on
    Control or Haz. Materials Spills,
    April 1981, Miami, Fla.

 8.  Rice, A., and A. Netzer, 1982.
    Handbook of Ozoae Technology and
    Applications, Vol. 1. Ann Arbor
    Science.

 9.  Wetzel, R., S.M. Henry, B.A, Spooner,
    S Janes and E. Heyae, 1985.  In Situ
    Treatment of Contaminated Ground-
    water and Soils.  Proceedings of llth
    Annual Research Symposium.  Land
    Disposal Remedial Action Incineration
    and Treatment of Hazardous Hastes,
    29 Apr - 1 May, 1985, Cincinnati, OH.

10. Ellis, N.D. and J. R. Payne.  The
    Development of Chemical Counter-
    measures For Hazardous Waste
    Contaminated Soil, EPA-600/D-84-039,
    PB84-148840, Jan. 1984.

11. Remedial Response at Hazardous Waste
    Sites,  Case Studies 1-23.  EPA 540/2-
    84-002b, March 1984.

12. Jhaveri, V.  and A.J.  Hazzacca,
    Digest and Supplement to Bio-
    reclamation of Ground and Ground-
    water.   Literature supplied by
    Croundwater Decontamination Systems
    Inc,  June 1985.

13. Personal communication.  R.  FeTlnan,
    V  Jhaveri, and A.  J.  Mazzaaca.
    June  4, 1985
                 Disclaimer

This paper has been reviewed 1n
accordance with  the U.S.  Environ-
mental  Protection Agency  peer  and
administrative review policies and
approved for  presentation and  publi-
cation.
                                         424
-------
        TABLE 1.  POTENTIAL APPLICATIONS OF TREATMENT METHODS TO WASTE CONTAMINANTS
          _,   ,   ,
          Chemical
           Class
                                 _.
                                 Bio-
                                                                     Water
                                                                                Surfactant
Aliphatic Hydrocarbons
Alkyl Halides
Ethers
Halogeaated Ethers
  and Epoxides
Alcohols
Glycols/Epoxide s
Aldehydes, Ketones
Carboxylic Acids
Amides
Esters
Nitriles
Amines
Azo Compounds,
  Hydra'zine Derivatives
Nitrosa mines
Thiols
Sulfldes, Bisulfides
Sulfonic Acids, Sulfoxides
Benzene & Substituted
  Benzene
Halogenated Aromatic
  Compounds
Aromatic Nitro Compounds
Phenols
Halogenated Phenolic
  Compounds
Nitrophenolic Compounds
Fused Polycyclic
  Hydrocarbons
Fused Non-Aromatic
  Polycyclics
Heterocyclic Nitrogen
  Compounds
Heterocyclic Oxygen
  Compounds
Heterocyclic Sulfur
  Compounds
Organophosphorus Compounds
Carbamates
Pesticides
                                                 (l)         (2>        (3)
                         degradation   Hydrolysis   Oxidation   Flushing   Flushing
4-
4- 4-
4- -
4- 4-
4-
4- -/+
4- -
4-
4- +
4- 4-
4- 4-
4- 4-
4- -
+
+ _
4-
+ 7
+ _
4- . -
4-
1
7
J
+
+
+
4-
4-
?
4-
4-
+
4-
+
4-
?
+
?
4-
- ?
4-?
? ?
4-
? 7
4-? -?
? ?
? 7
? ?
? ?
4-? -7
? 1
? 7
7 7
7 7
7 7
- 4-
4-
                              4-
                              4-
                              4-
                              4-

                              4-

                              4-
                              4-
                              4-
4-

4-
             4-
             +
             4-

             4-*
             +

             4-

             4-

             4-
4-


4-




4-?


4-?
(1)

(2)

(3)
Based upon calculated half-lives for base catalyzed at pH 9 to 11.

Based on oxidation of chemicals in water and wastewater by H^O.,.

Based upon aqueous solubility and octanol/water partition coefficient
4-  ~ caa be used
-  =" cannot be used
4-? « probably can be used
                                    further research needed
                                    probably cannot be used
                                         425
-------
            PUMPING TOXIC AND RADIOACTIVE FLUIDS WITH AIR LIFTS
                             Nigel N.  Clark
            Particle Analysis Center, ¥est  Virginia University
              223 White Hall,  Morgantown, West Virginia 26506
                                ABSTRACT

     Extreme reliability is often required of pumps handling highly toxic
or radioactive liquids because of  the problems associated with maintaining
or repairing the pumps.  Moreover, many of these pumps  will be in remote
locations,  and  will be used  interraittantly, so that simple control and
start-up is  an advantage.  Air lifts offer  an attractive solution to such
pumping problems, but few theoretically sound design  techniques for air
lifts have appeared in the literature.  Operating curves show the delivered
rate  of fluid as a function of the air flowrate used; for relatively short
pumps, these curves can be constructed  by assuming an average air flowrate
in the pump,  and performing a momentum balance over the height of the pump,
equating the sum  of the hydrostatic head, frictional losses and
acceleration terms with the pump submergence. This approach is valid for
typical pumps of up to 30 meters in height,  and can be used to analyze the
effect of design variables on the liquid flowrate.
INTRODUCTION

Uses ofair lifts

     Air lifts provide a simple,
reliable method  of pumping radio-
active or highly  toxic liquids
because they require remarkably
little maintenance.   To date these
pumps  have  been  used  for  mine
dewatering, ocean mining,  dredging,
effluent aeration and for  flotation
of coal. Having no moving parts, an
air lift is  resistant to damage by
solids or  particles in the  liquid,
unlike centrifugal  pumps which
suffer impeller erosion, so that air
lifts can be  used to raise difficult
slurries or  to  desludge effluent
ponds.
Air lift  description

     An air lift pump is simple and
inexpensive to construct, consisting
of a vertical pipe  partly submerged
in  liquid, or  equivalently  fed
liquid at some head from a tank.  In
figure 1  the distance (xi~xo)  is
called the submergence, and the
distance  (x2-x-|) is termed the lift.
Usually  the pressure in the tank,
P-j , and pressure in the header pot,
?2  are equal (often being atoms-
pheric pressure).   In cases where
they are not equal,  the effective
submergence  of the pump will be
altered.   Air is introduced near the
base of the pipe, and a bouyant air-
liquid mixture  is  carried up the
pipe to a header pot, where the gas
and liquid disengage.  Generally the
air  and  liquid  flowrates are such
that the  mixture is  in  "slug  flow"
                                  426
-------
             AIR OUT
HEADER
POT



-iff-

TANK











HEIGHT
= xl






1
^-^_
^T-4o «. «K ""7" _ t













0
Q
0


(


_w
f\
u
.
p-


1




TTOTTin FTTM

AAK bLUijr


J£R. f
IN l
                                         HEIGHT = x2
                                         PRESSURE = Pf
                                         HEIGHT =
                                         PRESSURE = P
                                                      o
Figure 1:  Air Lift Pump,
                         427
-------
 ( 1 , 2 , 3 ) >  which  consists of
alternating plugs of liquid and
large air bubbles  spanning almost
the whole  pipe diameter.  The energy
source  for  the pump  is the  com-
pressed air, which expands over the
pump  height}  liquid flowrate is
controlled by varying  the air rate,
so that remote operation is  readily
implemented.
PURPOSE

     This paper seeks to present a
reliable  design procedure to
encourage the use of air lifts in
handling dangerous liquids and
slurries.  A well designed air lift
pump can  offer energy efficiencies
of over 50%, which many  specialist
pumps  cannot match.  Despite the
fact that air  lifts have already
been used at nuclear reprocessing
plants (4). and that  air lifts with
total  heights from a meter to
several  hundred  meters  have been
demonstrated successfully in a wide
range of industries, air lifts do
not currently enjoy the application
that they ought.  This  is because
reliable design  techniques have
received  insufficient attention in
the literature.  Current engineering
handbooks still recommend empirical
or semi-empirical equations, and
fail to  apply  an approach  with a
sound  phenomenological basis.
Energy balance designs  (5)  remain
inadequate because they require a
prior  knowledge  of  the operating
efficiency,  which is  seldom
available.  Advances  in multiphase
flow theory have permitted design by
an accurate momentum balance, as
described below.
APPROACH

Pressure  requirements

    For a particular application,
the  required lift and volume
flowrate of liquid are  generally
specified, and the air pressure and
flowrate requirement  must be
predicted to size the compressor or
blower.   Air pressure  is simply
determined by calculating the hydro-
static  head of liquid  at  the
proposed point of air introduction,
usually near the  bottom of the  pump
and adding this  to  the  absolute
pressure in the feed tank,  ?•) .
Referring to figure 1, this pressure
is given  by the equation
po =
1
where PJ, is the fluid density.   All
calculations for  air  lift pumps
should be conducted using absolute
pressures.
Momentum balance

    Volume flowrate  of air is not as
readily predicted, and must be found
by  performing a momentum balance
over the length of the pump.  The
difference in  pressure over the
length of  the pump, PO~P2»  a*
equilibrium must equal the  sura of
the hydrostatic head in the pump,
pressure change due to acceleration
of the fluid, and frictional losses
at the pipe wall.  For the simplest
case, where the tank and header pot
are at atomspheric pressure,
P0-P2
                                        Neglecting   the   small
                                    contribution of the air in the pump,
                                    the hydrostatic head is given by
                                 428
-------
APH =
3   ¥G = QV/A(0.5P2+0.5P0)
where e  is the void fraction of air
present  in the pump.  The air  void
fraction is smaller  than the ratio
of air flowrate to total flowrate
for two  reasons.  Firstly,  the air
is present in large  bubbles or slugs
which are axially situated,  and
which are separated from the  pipe
wall  by a film  of  liquid.   These
slugs are  therefore more con-
centrated near the  pipe  center,
where the flow velocity is  higher,
so that  this "profile effect" serves
to reduce the air void fraction.
Secondly,  the  air  slugs rise
relative to the liquid, which flows
back  aroud the slug in the annular
film,  so that there is a local slip
effect which further increases the
gas velocity.  Both  of these effects
are accurately modelled by a drift-
flux  analysis  (1,6,7,8), which
predicts the  gas void fraction, c.,
by the equation
where WQ and WL  are the  gas and
liquid  superficial velocities, Co  is
a  constant accounting  for the
profile effect,  generally  1.2  in
turbulent flow (1,2,6) and ¥v is the
drift velocity accounting for local
slip, given by the equation

Vv = 0.35(gD)°-5                 5

where D is pipe  diameter,  for low
viscosity liquids.  Compensation for
high viscosities is discussed  by
Govier & Aziz  (?)•   Since  the
pressure changes over the height  of
the  pump,  the  gas superficial
velocity is not constant throughout
the device.  As an accurate approx-
imation for short pumps,  the gas
superficial velocity, ¥Q, used  in
equation [4],  is evaluated  at the
mean pressure in the pump
    where Q'Q is the free air  volumetric
    flowrate, P*  is  atomspheric
    pressure,  and A is  the pipe  cross
    sectional area.
        Prictional  losses can be
    evaluated by assuming that there are
    no  losses at  the wall  from the
    annular film surrounding  an air
    slug,  and that all losses occur in
    the liquid zone, which is in contact
    with  a fraction of  approximately
    (1-e)  of the  pipe wall.  Since, by
    continuity arguments,  the  liquid
    zone must on average travel at the
    total  superficial velocity,  Vg+W^,
    the frictional losses, by analogy
    with single phase turbulent flow,
    are
    where the friction factor f is found
    from a conventional single phase
    diagram, using the Reynolds number,
    Re = PL(¥G+¥L)D/ ML
8
    and s is evaluated using equation
    [4]-
         An additional pressure term
    arises due to the acceleration  of
    the liquid from a standstill at the
    base  of the  pump,  to a velocity
    equal  to the total superficial
    velocity  at the top of the pump.   At
    the top  of  the  pump all of this
    liquid momentum is irreversibly
    lost.  Upon entrance into the pipe,
    the pressure change due to momentum
    increase  must be
    and  pressure  loss  due  to
                                 429
-------
 acceleration from the point  of air
 introduction to the top  of the pump
 is given by
                                10
where W02  is  the air superficial
velocity at the top of the pump,
given by
                                11
    Although  there  are  some
additional irreversible  losses  as
the liquid enters the base of the
pump,  these are small and can  be
neglected.  Thus the  total momentum
balance over the pump  is given by
                                12
where the terms on the right hand
side  are given by equations [3],
[7],  [9]  and [10]  respectively.
Equation [12] must be satisfied to
predict the free air  flowrate, Q'Q,
required to  produce a given liquid
superficial velocity, W^,  in  the
pipe.   In practice this can be done
by  rearranging  the  relevant
equations into a simpler solution
(9) » or by writing a  short computer
program, and finding Q'G by trial
and error, which  is  still  a very
rapid  procedure.   As few as six
points determined  in  this way will
define an operating curve  quite
accurately.
PHOBL1MS ENCOUNTERED

The case of a tall pump

     The equation developed above is
for  the  case  of a short  pump,
because an average air  superficial
velocity has  been assumed  in
developing all of the equations.
The air superficial velocity changes
in a non-linear fashion over the
pump length, so that this assumption
causes  overprediction  of  the
required air flowrate when applied
to very tall pumps.  However, Dabolt
&  Clark  (10)  have demonstrated
recenty  that  the short design
approach is quite accurate for
typical pumps up to JO meters in
height,  when  compared to a more
complex tall  pump momentum balance
(2,11).   Thus the  design approach
presented  above   should  be
sufficiently accurate for most  toxic
and  radioactive  liquid  pumping
installations.
Practical design problems

     Although the momentum  balance
solves the problem of relating the
air  and liquid velocities in the
pump, there are  some practical
problems to the pump design which
must  be  discussed.  Firstly,  the air
must  be  introduced into the pump a
short  distance  up the  pipe,
otherwise there is  the  danger that
the air  will bubble back into the
tank, so that the air lift pumps at
a reduced  rate or  not at  all.
Secondly,  the mixture leaving the
top of  the tube must be separated
into air  and  liquid  streams.
Although fairly sophisticted header
pot designs have been proposed for
this  purpose (4), invariably some
liquid droplets are  entrained in the
exit air, and  some air leaves  as
bubbles  in the  liquid  stream.
Provision for  this  problem  must be
made, because the contaminated air
cannot  be discharged immediately to
atomsphere.  In addition,  the air
lift pump must be supplied with
sufficient submergence to  operate
effectively: the strong effect that
submergence has  on flowrate  is
                                 430
-------
discussed  in the section below.
RESULTS

     The validity of  a  momentum
"balance design approach has been
demonstrated  previously by
comparison with data from  air lifts
ranging from  6  to 250  meters  in
height (1,2,9,11).  In this paper,
the momentum balance will  be used to
demonstrate the strong effect of
submergence on the air  lift
operating curves.   Figure  2 presents
operating curves for a pump 10
meters in height and 100 millimeters
in diameter lifting water at sub-
mergences of 3-5,  5 and  7 meters,
using a typical friction factor of
0.01.   Both the  tank  and header pot
were  taken as being at atomspheric
pressure.   The submergence  has a
very strong effect on  the maximum
flowrate which can be  expected from
the pump, and also on  the air rate
required to lift the volume  of
water.  At very low submergences the
pump will be incapable of lifting
any liquid.
    Where  the efficiency of the pump
is given by  (1 )
n =
figure  2  also  demonstrates  that
pumps with a very low submergence
are less energy  efficient.  Nicklin
                                       SUBMERGENCE = 7 meters
                     1-0      2.0      3.0      4.0     5.0     6.0
                         FREE AIR SUPERFICIAL VELOCITY (m/see.)
Figure 2:  Operating curves for an air lift of 10 meters height at three different
           submergences.  Friction factor =0.01, Diameter = 100 mm.
                                431
-------
 (1)  has showii that  there  is an
 optimum submergence for an air lift.
 Except in cases  where  there are
 specific  limitations on the sub-
 mergence available, a depth of 50  to
 60%  of the  overall  pump length
 proves most practical, although far
 lower and higher submergences have
 been used in practice.
REFERENCES

1. Nicklin, D.J., 196?, The  Air Lift
Pump:  Theory and Optimisation.
frans.  I.Chem.'l. Vol.- 41 ,  p.  29.

2.  Clark, N.»., Meloy, T.P.  and
Plemmer,  B.L.C.,  1984, Air Lift
Pumps  for Hydraulic Transport  of
Solids, Pro.c.  9th..  Annual  Powder &
Sulk Solids  Conf., Rosemont,
Illinois,  pp.  446-456.

3- Govier,  G.W. and Aziz,  K., 1972,
The  Flow  of  Complex Mixtures in
Pipes,  Van Nostrand  Heinhold,  New
York;

4» Babolt,  R. J. and Plummet, K.E. ,
1980, Design of Air Lift Systems for
Transfer  and  Measurement of
Radioactive  Liquids.,  Dept.  of
Energy  Report AGNS-35900-3.2-77.

5. Shaw, S.I., 1920,   tinwatefing the
Tiro General Mine  by  Air Lift.
Trans.  AIMS, pp.421-455

6. Nicklin, D.J., Wilkes,  J.O. and
Davidson;  JiP., 1962,  tfwo  Phase Plow
in Vertical Tubes. Trans.  t>.Chem.E.
Vol.40, p;61.

7. Zubef, N.  and Fihdlay, J.A.,
 1965,  Average   Volumetric
Concentration in,Two Phase Plow
Systems. A.S.M.E. J.  Heat Transfer
Vol.87, p.453-
8. Clark,  N.N. and Plemraer, R.L.C.,
1985, Predicting Holdup in Two Phase
Bubble Upflow and Downflow using the
Zuber and Findlay  Drift-Flux Model.
A.I . Ch _. E .  Journal ,  Vol.31,
pp.500-503.

9- Stenning, A.H.  and Martin, C.B.,
1968, An Analytical and Experimental
Study of Air-Lift  Pump Performance.
A.S.M.E.  Jour.  Eng.  for Power,
Vol.90,  p.106.
10.   Dabolt, R.J. and Clark,  I.N.,
1985, Pumping  radioactive slurries
by air lift"  10th. Powder and Bulk
Solids Conference,  Ro a emon t,
Illinois, pp. 779-788.

11.  Clark, N.W. and Dabolt, R.J.,
1985, A general  design  equation for
air lift pumps operating in the slug
flow regime  A.I.Ch.E. Jour., In
Press.
              Disclaimer

 The work described  1n this paper was
 not funded by the U*S. Environmental
 Protection Agency.  The contents do
 not necessarily reflect the views of
 the Agency and no official endorse-
 ment should be inferred.
                                  432
-------
               GREATER-CONFINEMENT DISPOSAL OF LOW-LEVEL RADIOACTIVE WASTES

                 LaVerne E.  Trevorrow, Thomas L.  Gilbert, Charles Luner,
                Pamela A. Merry-Libby, Natalia K.  Meshkov, and Charley Yu
               Environmental Research Division, Argonne National Laboratory
                                 Argonne, Illinois  60439
                                        ABSTRACT

     Low-level radioactive wastes  include a broad spectrum of wastes that have different
radionuclide  concentrations, half-lives, and physical and chemical properties.  Standard
shallow-land  burial  practice can provide adequate protection of public health and safety
for  most low-level wastes, but a  small  volume fraction (~1%) containing most  of  the
activity  inventory (~90%) requires  specific  measures  known  as "greater-confinement
disposal" (GCD).   Different  site  characteristics and different waste characteristics—
such as high  radionuclide concentrations, long radionuclide half-lives,  high radionuclide
mobility, and physical or chemical  characteristics that present exceptional  hazards—lead
to different  GCD  facility design  requirements.  Facility design alternatives considered
for  GCD  include the  augered shaft,  deep  trench, engineered structure, hydrofracture,
improved waste  form,  and high-integrity container.   Selection of an appropriate design
must also  consider the  interplay  between  basic risk limits for protection of  public
health and  safety,  performance characteristics and objectives, costs, waste-acceptance
criteria, waste  characteristics,  and  site  characteristics.   This  paper  presents an
overview of the  factors  that must be  considered in planning the application of methods
proposed for  providing greater confinement of low-level  wastes.
INTRODUCTION

     Low-level radioactive wastes include
a  broad  spectrum  of wastes  that  have
different  radionuclide  concentrations,
half-lives,  and  physical  and chemical
properties.   These  wastes  range  from
wastes  containing  naturally  occurring
radionuclides to  mixed  wastes containing
both  radioactive  and chemical  contami-
nants.  Standard shallow-land burial (SLB)
and  other  near-surface  stabilization
methods are most commonly used for disposal
of these wastes.   In a typical SLB facility,
the waste  is buried in shallow trenches
about 8 meters in depth.  Wastes containing
long-lived,  naturally  occurring  radio-
isotopes have, in many  cases, been  tempo-
rarily stabilized by simple means such as
confinement in pits or covering with soil.

     A small  fraction of  low-level  wastes
from both  U.S. Department of  Energy (DOE)
and commercial sources  contains radionu-
clides in sufficiently high concentrations
or with  sufficiently long half-lives to
require   greater-confinement   disposal
(GCD), defined as "a technique for disposal
of waste  that uses natural and/or engi-
neered barriers which provide a degree of
isolation greater  than  that  of  shallow
land burial  but possibly less than that of
a geologic repository" (DOE Order 5820.2).

     In anticipation of the  need for new
land disposal facilities to better accommo-
date  low-level  wastes generated by DOE/
defense  and  commercial  activities,  a
Low-Level Waste Management  Program  was
established  within  DOE  to  initiate  and
coordinate research and development activi-
ties for safe and cost-effective means for
disposal of  low-level wastes.   The types
of wastes that are being considered  for
GCD include:  (1) wastes with high concen-
trations  of  short-lived  radionuclides;
(2) wastes with  long-lived radionuclides;
and  (3) wastes  co-contaminated  with
hazardous chemicals or  chemicals that
increase  the mobility of  radionuclides.
                                          433
-------
Wastes of  the  first type include some of
the wastes generated by  DOE and other
government agencies as a result of defense
activities, uranium-enrichment activities,
and research and development activities as
well as  some  of the wastes generated  in
commercial  activities such  as  nuclear
power  production,  manufacturing, medical
applications, and research.  Wastes of the
second type are largely  those containing
naturally  occurring radionuclides—e.g.,
mill tailings  or raffinates,  equipment,
contaminated  soils, and  decommissioning
rubble that remain at sites that were used
for processing  or  storage of uranium and
thorium ores and compounds.

     The major  reason that GCD  is  being
considered  for  these types of wastes  is
the potentially unacceptable risks associ-
ated with  releases  to the  environment  and
with human intrusion into the wastes  if
government control  of the disposal  sites
were to  cease  in  the future.   Possible
reasons  for cessation of  control  are  loss
of funds and catastrophic events.

     The methods  for greater confinement
can be grouped  according  to  modifications
to the disposal  cell  or modifications to
the wastes or packaging.   (The term "cell"
is a general term indicating an individual
hole, trench, shaft, or structure in which
wastes are emplaced for disposal.)  Modifi-
cations  to the  disposal   cell  include
augered shaft, deep trench, and engineered
structure.  The special  disposal technique
of hydrofracture  is also  being considered
as  an  example  of greater confinement.
Modifications to  the  wastes  and packaging
are commonly referred to as improved waste
form and high-integrity  container (HIC),
respectively.

PURPOSE

     The purpose is to present an overview
of the factors  that must  be  considered in
planning the application  of  methods pro-
posed for providing greater confinement of
low-level wastes, to  present methods for
evaluating existing and conceptual disposal
units that would provide  greater confine-
ment of  low-level  wastes, and  to review
the characteristics of  a limited set  of
designs  that  have  emerged  from these
several efforts as  the most  promising  for
providing  the  confinement  that may  be
required for these waste types.
APPROACH

     The  characteristics  and  expected
volumes  of  wastes  that might  require
greater confinement were derived from data
bases  that  have been  collected  by DOE
contractors.   In  general,  these  wastes
could  not be  disposed by  technologies
referred  to in  10 CFR 61  regulations.
Greater  confinement  for  even  larger
volumes than these  is  being"requested by
some citizen groups despite the fact that
the  characteristics  of the wastes might
permit less sophisticated  and less costly
disposal technologies.   Current regulations
on low-level wastes are purposefully not
prescriptive with  respect  to  technology;
however, they  do  refer to SLB but not to
any  of the  designs  discussed  here.   Cri-
teria  expressed  by the  International
Commission on  Radiological  Protection  (9)
are  used  in this  work as minimum indica-
tions of  the performance objectives  that
greater  confinement must  achieve.   The
possibility is explored  of expressing
performance assessment,  i.e., analysis of
the  behavior  of the  technology  and  its
compliance with performance objectives, in
terms of  risk  analysis.  The role of cost
and  benefit in  selection  of  disposal
technology by  potential  operators is also
considered.   Several  design options  were
selected  for  this assessment,  and these
options are examined to identify the basic
elements  in each option that will support
the  performance  objectives.  The advan-
tages  and disadvantages associated with
each option are also discussed.

PROBLEMS ENCOUNTERED

     One problem encountered in this work
was  the  brevity  of  information  available
for some disposal unit concepts, especially
the  engineered structures.  The expected
performance of some disposal unit alterna-
tives was difficult to evaluate because of
the  incomplete state   of  development of
some  of  the concepts.   Regulations on
low-level waste disposal such  as 10 CFR 61
and DOE Order 5820.2 are not restricted to
SLB; for example, the U.S. Nuclear Regula-
tory  Commission  (NRC) maintains that  it
could assess the licensability of alterna-
tive  disposal  units  by the  10 CFR 61
guidelines.   Nevertheless,  these regula-
tions—although only  recently finalized—
were developed before the even more recent
surge of  interest  in alternatives to SLB.
                                           434
-------
Performance  assessment  of  GCD by  the
methods of risk analysis and cost-benefit
criteria could not be carried to the point
of quantitative  results because  of  the
lack of parameters needed as input to the
calculations.

RESULTS

Waste Characteristics

     The  expected  characteristics  of
wastes—especially  their  radiological,
chemical, and physical properties—will  be
the most  important determinants  not only
of whether  GCD is required  but  also of
which GCD technique will be applicable for
a given disposal  site.  The concentration
of radionuclides  in  the low-level wastes
will  be the  primary  index of whether they
must be  managed  by GCD techniques.  For
commercial wastes, it is expected that any
wastes exceeding the Class C radioactivity
concentrations defined  by  10 CFR 61 will
require  management by  GCD,  but  site-
specific  criteria may also require that
wastes of concentrations  lower than the
limits of 10 CFR 61 be treated by GCD.  For
example, at the Savannah River Plant, some
wastes that  do not exceed concentrations
corresponding to  the  limits of Classes  C,
B, or even A are managed by GCD techniques.

     Our knowledge of the characteristics,
volumes,  and  properties of  wastes in the
United States  that will  require disposal
is continually being  improved  by several
waste inventory systems (11).  The annual
review by DOE (20) also presents  informa-
tion on U.S.  waste  inventories that will
aid decisions  on  which  wastes are likely
to require GCD.  Current estimates indicate
that about J% or 1,600 nrVyr (2,100 yd3/yr)
of all  low-level  wastes may require this
special treatment (7,20).   In  addition,  a
total  of  2.3 x 107 m3  (3  x  107  yd3)  of
materials  contaminated  with  long-lived,
naturally occurring  radionuclides awaits
permanent disposal,

Regulations

     Regulations  specifically  for manage-
ment of GCD wastes have not been expressed
in detail,   but  general  guidelines are
given  in  DOE Order  5820.2,  10 CFR 61
(NRC), and the criteria,  rules,  and laws
being developed in the  formation  of  state
compacts.   The policies of DOE Order 5820.2,
Chapter III, that apply  to  wastes gener-
ated at DDE-controlled sites give general
guidelines  for  the  waste-acceptance cri-
teria that  must be  developed by each DOE
disposal site.   The criteria of 10 CFR 61,
which apply to commercial wastes, indicate
the  limits  above which  wastes  require
greater confinement than conventional SLB.
Although the  regulation  of  GCD has not
been  explicitly defined, at least  the
concentrations  of  radionuclides  at the
site boundaries  of a facility are defined
by the  concentration limits  of  10 CFR 20;
DOE  Order 5480.1A,  Chapter XI;  and the
drinking  water  limits   of  40 CFR 141.
These represent  goals  for performance of
GCD techniques.  The 10 CFR  61 regulations
indicate that more  specific guidance for
alternatives to near-surface disposal  of
low-level  wastes, e.g., 10 CFR 61.50(b) on
site  selection,  will  be developed.   In
addition to these regulations and criteria,
plans for design and construction of a GCD
facility may  require  an environmental
evaluation in compliance with the National
Environmental  Policy Act of  1969.

     Each GCD  alternative carries with  it
certain waste-acceptance restrictions and
thus  imposes  some  restrictions on  waste
generators.   Although  the  site-selection
criteria for application of  improved waste
form or HIC may not differ  from those for
SLB,  there  will be additional  criteria
relative to the GCD techniques of deep
trenches,  engineered  structures, augered
holes,  and  hydrofracture—of which  some
criteria will  be unique  to  each method.

PerformanceAssessment

     The importance of assessing the tech-
nical performance of a disposal facility—
before, during,  and after its operational
lifetime—is emphasized  in  10 CFR 61 and
DOE  Order 5820.2, and  it is likely  that
performance assessment  of  a  GCD system
will also be treated with importance.  The
performance of  a disposal  facility  is
customarily assessed  against performance
objectives.   Although  federal regulations
(10 CFR 61, 10 CFR  20,   and  DOE Order
5480.1A) imply  performance  objectives and
although each regional  compact is expected
to express  its  own  set of objectives, the
clearest current statement  of performance
objectives  is presented  in the basic rules
of the  International Commission on  Radio-
logical Protection  (9).   The essence of
                                          435
-------
these  rules  is that  risk to  both the
general public  and occupational workers
should be  limited.  The basic  dose limits
are  500 mrem/yr  for short-term exposure
and  100 mrem/yr  for  lifetime  exposure.
The  occupational limits are  greater by a
factor of  10.  Decisions  among design
alternatives  should  be based  on (a) the
expected technical  performance that will
permit achievement  of these performance
objectives, and (b) the cost of achieving
a  given  level of technical performance.
Thus,  ideally,  the choice of disposal
techniques should be made  on the basis  of
benefit-cost-risk  (BCR)  analysis.   The
application of this type of assessment  to
GCO  techniques  has  been   described by
Gilbert and Luner (7).

     Planners and waste generators antici-
pate that  the cost  of  GCD  will be  greater
than the cost of SLB.  Based on some cost
estimates  for GCD  designs  that have been
made (7),  costs are expected  to increase
in the order SLB < deep trench or  SLB with
intruder barrier < augered shaft < concrete-
walled trench < improved waste form.  Costs
are  strongly  dependent on  site character-
istics; hence, site-specific considerations
could alter this order.  Hydrofracture  is
too  dependent on site-specific factors  to
permit  inclusion  in  this  ranking.  If
geologic  conditions permit  the  use of
hydrofracture, it  is   probably the most
cost-effective disposal method for liquid
waste and a comparable disposal method for
solid wastes  that  can readily be  formed
into a slurry (e.g., ash from  incinerated
low-level wastes).   The cost categories in
which GCD  is  expected  to  differ signifi-
cantly  from  SLB are   labor, materials,
post-operational stabilization,  and pur-
chase and replacement  of equipment.

     Although a BCR analysis can be mathe-
matically expressed, the lack  of parameters
with which to obtain quantitative  results
has  led to the proposal to make decisions
among  GCD  alternatives  by  a  two-part
assessment   method:    (1) quantitative
estimation  of risk  associated with  a
disposal method by  modeling  the migration
of  radionuclides  from the disposal site
(pathway  analysis),  and  (2) qualitative
comparison of the  attributes  of   alter-
native disposal  techniques.   The  results
of  calculating  the concentrations of
radionuclides  at  various  distances in
pathways leading from  a disposal cell  can
be  used  to  compare the risks associated
with alternative designs. The qualitative
comparison of technical performance can be
based on an evaluation of the contribution
to  realizing  the performance objectives
that would be  made by performance attri-
butes such as  the  characteristics of the
waste  form,  container,  design  of  the
disposal cell, emplacement procedures, and
emplacement equipment.  The assessment can
be carried into further detail  by determin-
ing which designs  provide the even more
basic elements that are ultimately respon-
sible  for  those performance attributes:
intrusion resistance, compressive strength,
corrosion resistance, radiation  stability,
drainage control, infiltration resistance,
leach resistance, biodegradation resistance,
ion-exchange capacity, thermal stability,
distance  from  surface,   distance  from
hydro!ogic movement, permeation  resistance,
distance from  radiation  sources,  minimum
time  of  exposure  to  radiation  sources,
shielding,  structural  stability,  and
chemical inertness.

Disposal Cel] Design Alternatives

     A  few  disposal  cell  concepts  have
been  considered to  be practicable  by
several  evaluations.   These  concepts are
being  catalogued in an overview of GCD,
currently in  preparation, that will  be
published as  one of the DOE handbooks on
management  of  low-level  wastes.   These
concepts  include  augered shaft,  deep
trench, and engineered structures.

     The augered shaft consists of a  hole
in  the  ground  with a diameter of  3 to 4 m
and a  depth of 10  to 35 m, as exemplified
by  demonstrations   at the Nevada Test
Site (13) and the Savannah River Plant  (14).
Smaller,  shallower  boreholes have  been
used  for waste  disposal  in  the  United
States and other parts of the world.   The
advantages of  the  augered shaft include a
geometry  that  shields operators  from
emplaced radioactivity; compatibility with
remote-handling  techniques;  remoteness
from  plant  and animal intrusion; easy
closure, both temporary and final; and  low
susceptibility  to  erosion.  A disadvantage
of  the  augered shaft is the limited  size
of  the  waste  items that are  acceptable to
the  typical  diameters  of the  shafts.

     The deep  trench disposal  unit is  an
excavation that is deeper than  the normal
                                         436
-------
8-m depth of  the  SLB trench.  The wastes
are surrounded with  soil  material  in the
deep trench,  as  in  SLB.   The deep trench
has not  received much attention either  in
design or  in use.   One  of  the earliest
references to the concept suggested that
it would be  quite  similar to an SLB unit
except twice  the  depth  of a conventional
SLB (10).  In addition to the advantage of
placing wastes beyond the depth of pene-
tration  of  roots  and animals,  the  deep
trench offers simplicity, flexibility in
acceptance  of waste  types,  and little
vulnerability to erosion.  The deep trench,
however,   requires  a  site that has  an
unusually thick layer of soil  and uncon-
solidated materials  over  the water table.
Furthermore, unless  special shoring tech-
niques are used, the wide opening required
to excavate a deep  trench will involve a
relatively  large  area and may restrict
emplacement  techniques  to  unloading  a
waste-carrying vehicle  at the bottom of
the trench.

     The engineered structure is a disposal
cell in  which one  of the most  important
barriers  to  intrusion  and release  of
radionuclides is a chamber typically built
of concrete.  A  large variety of designs
have been described,  involving placement
of wastes both above and below grade level
(2,8).    These  include   the  Canadian
concrete-walled trench (5),  the Savannah
River  Plant  concrete-shored  trench (14),
the French tumulus (24),  the NRC concrete-
walled  trench (21),   and the  concrete
chamber of the University of Arizona (25).
The main  advantage  of engineered struc-
tures  is their  potential barriers  to
infiltration  and  intrusion.    Structures
that are initially  roofed will also pro-
vide  protection  from  adverse  weather
during emplacement  operations.   However,
because  concrete  is  prone  to eventual
cracking, the engineered  structure elimi-
nates  neither  infiltration of water nor
release  of  leachate  over the  long term.
Recently, a  preference  for  engineered,
above-grade  disposal units   has  been
expressed in planning facilities that will
be  operated  by  state governments  for
disposal  of low-level waste (23).  Above-
grade  structures have received consider-
able attention because they are perceived
to offer  advantages  for protection from
groundwater and also for  ease  of surveil-
lance, maintenance,  and  remedial  action.
Hydrofracture for Greater Confinement

     The  waste-emplacement configuration
of the  hydrofracture  alternative  consists
of a stack, several  hundred meters  in
diameter, of thin  sheets of grout incor-
porating  the wastes;  the grout  sheets  are
interleaved  between  underground   shale
layers.   This   unique waste-emplacement
configuration  sets hydrofracture apart
from the  category  of  disposal  cells dis-
cussed in the foregoing sections.   Disposal
by hydrofracture (mixing wastes in liquid
or slurry form  with cement and injecting
the  mixture  into horizontal  fractures in
rock formations  located several  hundreds
of meters below  the earth's surface) has
been practiced  successfully  over  a period
of many years at Oak Ridge National Labora-
tory (26).  The advantages of hydrofracture
include  a high  degree of isolation from
the  environment and  from intruders,  a
small  commitment  of  surface land  area
above  the disposal  zone, and a relative
insensitivity to weather during  emplace-
ment and  to erosion  after emplacement.
The disadvantages of hydrofracture include
applicability only  to wastes in liquid or
slurry  forms  or to wastes that  can be
converted to such  forms, the  possible
stimulation of minor  seismic effects,  and
the  requirement  that the  disposal site
have  special geologic  characteristics.
Also,  if  contamination of deep aquifers
did  occur,  remedial  action would  not  be
feasible.

Improved Waste Forms for
Greater Confi nement

     Whereas the GCD technologies described
in the  foregoing sections have  emphasized
confinement by geologic media,  the concept
of  improved waste  form emphasizes the
capability of confinement derived from the
physical  and  chemical properties  of the
waste  form.   Improved  waste forms are
generally solid media  into which  primary
waste forms  are  incorporated.   Among the
advantages of improved  waste forms is  the
potential for their use in an ordinary SLB
trench to provide GCD.  Also, they provide
some attenuation of penetrating radiation,
are  independent of  site characteristics,
permit retrievability in  case of  need  for
remedial actions,  limit dispersion in case
of accidents,  and  reduce migration of
radionuclides caused  by leaching.  Among
                                         437
-------
the disadvantages of  improved waste forms
is  the involvement with  chemical pro-
cessing  equipment—with  the  attendant
needs  for maintenance,  decommissioning of
contaminated equipment, and  costs.   Some
solidification agents are  unable  to  com-
pletely  incorporate   all  waste  forms,
particularly oils and  organic  liquids.
The solidification agents  used to produce
Improved waste forms  can  be  grouped into
three  types:  cement, organic solids,  and
glass  (6,15).  Cement is the most commonly
used solidification agent in management of
radioactive  low-level  wastes.   Additives
such as organic  polymers,  silicates,  and
gypsum improve such properties as ability
to  incorporate oils,  mechanical strength,
and  leach resistance;  they  introduce,
however, some chemical-processing compli-
cations.   Some   organic  solidification
agents that  have been  investigated  and
used to varying degrees in actual  practice
are  urea-formaldehyde,  bitumen,  epoxy
resins, and  vinyl ester-styrene.  Urea-
formaldehyde, once widely used,  has now
been  rejected,  mainly  because  of  its
release of  contaminated water; bitumen,
used  frequently   in  Europe,  but  infre-
quently in the United States, still seems
acceptable for  some   applications; epoxy
resins are  offered in  commercial waste-
solidification systems; and a vinyl ester-
styrene process  is available in  another
commercial solidification system.   Although
glass  waste  forms have been developed
mainly with  the  intention of application
to  high-level waste,  their application to
other wastes also seems feasible according
to a recent evaluation  (1).

High-Integrity Containers for
Greater Confinement

     Another GCD  technique that relies on
factors other than those of geologic media
to provide confinement  is the high-integrity
container (HIC).   Its confinement capabili-
ties are  based  on its  design and  on  the
physical  and chemical  properties of  the
material  from which  it is fabricated.  A
high-integrity container  is  a vessel  that
is  intended  to provide structural  sta-
bility  and  containment of radionuclides
for a  long period; characteristics of  the
HIC have been more specifically defined in
criteria  formulated  by regulatory bodies
such as the  NRC  and  the  state of South
Carolina (4).   Designs  of  containers
intended  to  meet criteria for HICs  have
been developed  by several organizations
(3,12,16,27).   The  favored materials of
construction are polyethylene and concrete.
Sizes vary  from 55-gal  drums  to large
units that can be handled  only by powered
cranes.   In many cases,  emplacement of an
HIC  in  an  ordinary SLB trench  should
provide  the  security  required  for GCD
without the cost or trouble of construc-
ting any more sophisticated disposal  unit.
Probably the most serious  disadvantage of
the HIC  is  its  inability to accept large
items,  e.g.,   those  that   may  occur
occasionally as  a  result  of decommis-
sioning  activities   such  as  vehicles,
cranes,  processing  equipment,  and rubble
from the demolition of buildings.

Confinement of Low-Level,
Long-Lived Wastes

     Because the radioactive constituents
of long-lived wastes—raffinates, tailings,
rubble,  and contaminated soil material—are
mostly  isotopes  of  uranium  and  thorium
with their daughters,  a  major  concern is
the control  of  radon  release.   Thus,  a
diffusion barrier that slows the escape of
radon to permit most of it to decay before
reaching the atmosphere—consisting  of a
medium of low permeability, e.g., clay—is
common  to most  designs of  disposal units
for this type of waste.   Such a  barrier
can do  triple  duty  if it  also has the
capacities to slow  the migration  of ionic
radionuclides,  as some clays do, and  to
resist the infiltration of water.   Designs
for these disposal  units place the wastes
either above or below the earth's surface.
Design  criteria  for such  disposal units
include  multilayered  caps  of  natural
materials that provide—with little main-
tenance—drainage,   physical  stability,
erosion  resistance,  and  intrusion resis-
tance.  Examples of the latest designs for
such units are given in plans for handling
wastes at West Chicago (22), Weldon Spring
(18), Niagara Falls (19), and Cannonsburg,
Pennsylvania (17).   More recently, however,
groups  concerned with safe  disposal  of
wastes  are demanding  that  design  elements
similar to those of engineered structures
be evaluated for confining these  types of
wastes,   including man-made materials  for
drainage and resistance to water  infiltra-
tion.   The augered  shaft technology might.
be applicable to this type of waste, but
it might not satisfy the  current prefer-
ence  of citizen  action groups and state
planners for  above-grade  emplacement  of
wastes.
                                         438
-------
Conclusions

     The  need  for  disposal   technology
offering  greater confinement  than SLB
arises  not  only from  the existence  of
wastes exceeding the regulatory limits for
SLB, but  also  from  individual  policies of
organizations  and  demands of concerned
citizens.    The  number  and variety  of
technologies judged  capable  of providing
greater confinement  than SLB  have been
expanded.   Several  of  them that have  been
perceived to be  technically  feasible  and
applicable  to  the types  of  wastes for
which greater confinement is being demanded
have been briefly described and evaluated
here.  Applicability of  any  one of these
techniques  to  an   individual  disposal
problem will depend on the characteristics
of the  wastes  and  the  disposal site.   It
is expected that not only the selection of
the  technology  to be applied  but also the
need for  GCD  will  have to be established
on a case-specific basis.

ACKNOWLEDGMENT

     This work was  supported by the  DOE
Low-Level Waste Management Program.

REFERENCES

 1.  Armstrong,  K.M.,  and L.M. Klinger,
     1984,  Evaluation  of  a  Processing
     Technique  for  Immobilization  of  Low
     Level Radioactive Waste, Mound Labora-
     tory, Miamisburg,  OH (MLM-3149).

 2.  Bennett, R.D., W.O.  Miller, J.B.  Warri-
     ner, P.G.  Malone,  and C.C. McAneny,
     1984, Alternative Methods for Disposal
     of  Low-Level   Radioactive Wastes,
     U.S. Nuclear  Regulatory  Commission
     (NUREG/CR-3774, Vol. 1).

 3.  Chapman, R.L., A.L.  Ayers, Jr., R.T.
     Haelsig,  and  R.E. Holzworth,  1982,
     Development  of  a  High-Integrity
     Container  for  Disposal  of Special
     Types of  Radioactive Wastes,  Ijn  Low
     Level Waste Packaging and Shipping,
     Proceedings of  the  American  Nuclear
     Society Workshop.

 4.  Ensminger,  D.A.,  and  M.F. Kaplan,
     1983,  Criteria Definition for LLW
     Solidification and  Containerization,
     Electric  Power Research Institute
     (NP-3112, Final Report).
10.
11.
12.
Feraday, M.A., 1982, Canadian Experi-
ence with  the Storage and Disposal
of Low- and Intermediate-Level Wastes,
In Proceedings  of the Symposium on
Low-Level Waste  Disposal:   Facility
Design,  Construction,  and Operating
Practices, pp. 411-429 (NUREG/CP-0028;
CONF-820911, Vol. 3).

Fuhrmann, M., R.M. Neil son, Jr., and
P. Colombo, 1981, A  Survey of Agents
and  Techniques  Applicable  to  the
Solidification  of Low-Level   Radio-
active  Wastes,  Brookhaven National
Laboratory (BNL 51521).

Gilbert, T.L.,  and  C. Luner, 1984,
Analysis of Alternatives for Greater-
Confinement  Disposal  of  Low-Level
Radioactive  Waste,   Final  Report,
Prepared by  Argonne  National  Labora-
tory for DOE  Low-Level Waste  Manage-
ment Program.

Gilbert/Commonwealth  Engineers  and
Consultants,  1979,  State-of-the-Art
Review  of  Alternatives  to  Shallow
Land Burial  of  Low-Level Radioactive
Waste, Oak Ridge  National Laboratory
(ORNL/SUB-79/13837/1).

International Commission  on  Radio-
logical  Protection,  1977,  Limits  of
Inhalation  of Radon  Daughters  by
Workers (ICRP Publ.  26).

MacBeth, P.J., G. Wehmann, B.J.  Thamer,
and  D.H. Card,  1979,  Evaluation of
Alternative Methods  for the Disposal
of Low-Level Radioactive Wastes, Pre-
pared by Ford, Bacon & Davis Utah Inc. ,
for U.S. Nuclear Regulatory Commission
(NUREG/CR-0680;  FBDU-209-03).

Notz, K.J.   (ed.),  1982,  Radwaste
Inventories  and  Projections:  An
Overview, Oak Ridge  National  Labora-
tory (ORNL/TM-8322).

Reno, H.W., and R.L, Dodge,  1982,  Low-
Level Waste  Packaging and Shipping:
the  EPICOR-II Research and  Disposi-
tion Program  at the Idaho National
Engineering Laboratory, lr± Low  Level
Waste Packaging  and Snipping,  Pro-
ceedings  of the  American Nuclear
Society Workshop.
                                          439
-------
13.   Reynolds Electrical  and  Engineering
     Co., Inc., 1983, Greater Confinement
     Disposal Test  at the  Nevada  Test
     Site, Prepared for U.S. Department of
     Energy (DOE/NV/Q041Q-79).

14.   Towler, O.A. ,  J.R. Cook,  and  D,L.
     Peterson, 1983,  Greater  Confinement
     Disposal Program  at the  Savannah
     River  Plant,  Iji  Proceedings  of the
     Fifth  Annual  Participants'  Informa-
     tion Meeting,  DOE Low-Level  Waste
     Management  Program   pp. 308-315
     (CONF-8308106).

15.   Tucker, Jr.,  R.F., R.A. Nelson,  and
     C.C. Miller,  1983, Low-Level  Radwaste
     Solidification,   Electric  Power Re-
     search Institute  (NP-2900).

16.   Tuite,  P.T.,   1983,  High  Integrity
     Container Development,  J_n Proceedings
     of  the Fifth Annual  Participants'
     Information Meeting,  DOE Low-Level
     Waste Management Program,  pp.  504-512
     (CONF-8308106).

17.   U.S. Department of Energy,   1983,
     Remedial Action  Plan  for  Stabili-
     zation of the Inactive Uranium Mill
     Tailings Site  at Canonsburg,  Pennsyl-
     vania,  Albuquerque Operations Office
     (UMTRA-DOE/AL-140),

18.   U.S. Department of Energy, 1984a,  Engi-
     neering  Evaluation  of  Alternatives
     for  the  Disposition  of the Weldon
     Spring Raffinate Pits Site,  Oak Ridge
     Operations Office (DOE/OR/20722-5).

19.   U.S. Department of Energy,   1984b,
     Draft Environmental  Impact Statement,
     Long-Term Management of the Existing
     Radioactive Wastes and Residues at
     the Niagara Falls Storage Site (DOE/
     EIS-0109D).

20.   U.S. Department of Energy,   1984c,
     Spent  Fuel  and  Radioactive  Waste
     Inventories,  Projections,  and  Charac-
     teristics,  Office of Civilian Radio-
     active Waste Management (DOE/RW-0006).
21,  U.S. Nuclear Regulatory Commission,
     1981,  Draft  Environmental  Impact
     Statement on 10 CFR Part 61,  "Licens-
     ing Requirements for Land Disposal of
     Radioactive  Waste"  (NUREG-0782).

22.  U.S. Nuclear Regulatory Commission,
     1983,  Final   Environmental  Impact
     Statement  Related  to the Decommis-
     sioning of the  Rare  Earths Facility,
     West  Chicago,  Illinois,  Docket
     No. 40-2061 (NUREG-0904).

23.  U.S. Nuclear Regulatory Commission,
     1984, State Workshop on  Shallow Land
     Burial  and  Alternative  Disposal
     Concepts,  Proceedings  of Symposium
     (NUREG/CP-0055).

24.  Van Kote, F.,   1982,  12 years  of
     Experience of  Shallow  Land  Disposal
     of  Low and Intermediate  Level Radio-
     active Waste in France,  _In  Proceed-
     ings  of  the Symposium on Low-Level
     Waste  Disposal:   Facility  Design,
     Construction,   and  Operating  Prac-
     tices,  pp.  177-200   (NUREG/CP-0028;
     CONF-820911, Vol.  3).

25.  Wacks, M.E. ,  1981,  Alternatives to
     Shallow Land Burial  for  the Disposal
     of  Low-Level Wastes,  Final  Summary,
     Prepared  for   Los Alamos National
     Laboratory by  University of  Arizona
     Nuclear Fuel Cycle Research Program.

26.  Weeren, H.O.,   G.D.  Brunton,  W.  De
     Laguna, and J.G. Moore,  1974, Hydro-
     fracture Site  Proof Study at Oak Ridge
     National   Laboratory  (ORNL-TM-4713).

27.  Williams, J.D., 1982, Identification
     of Possible High Integrity Containers
     for Low-Level  Nuclear Waste  Disposal,
     jtn Scientific  Basis for Nuclear  Waste
     Management, Proceedings  of the Mate-
     rials Research Society Annual  Meeting,
     p. 449 (CONF-811122).
                                                             Disclaimer

                                             The work described in  this paper was
                                             not funded by  the U.S.  Environmental
                                             Protection Agency.  The contents do
                                             not necessarily reflect the views of
                                             the Agency and no official endorse-
                                             ment  should be inferred.
                                         440
-------
                     DESIGN OF  RADIOACTIVE  TAILINGS  DISPOSAL
                            SITES TO LAST 1,000 YEARS

                                Christopher  M. Tinra
                          Jacobs Engineering Group Inc.
                             Albuquerque, N.M.  87108
                                     ABSTRACT

     Cleanup and stabilization is currently being conducted by  the Department
of Energy at 24 inactive uranium mill sites involving from 100,000 to more than
3,000,000 cubic yards of radioactive tailings and other contaminated materials.
The applicable EPA standards provide standards for radiation control and stress
the use of passive controls such as thick, earthen covers, below ground disposal
or rock covers to achieve a mtniaiuB maintenance free design life of 200 years
and, ideally, 1,000 years.

     The 24 mill sites in question are principally in the semi—arid areas of the
Western USA.  The majority were built on the banks of or very close to rivers
so most of the tailings piles are in floodplains and vulnerable to flood erosion.
Another common problem is ground—water contamination.  Since most ponds were
not lined, large quantities of the tailings water percolated into the ground
and contaminated the near surface aquifers with uranium, selenium, arsenic,
sulfates, and similar toxicants.

     These conditions, along with the radioactivity required that the design
consider radiation reduction, prevention of wind, or water erosion, protection
from catastrophic events and minimization of water infiltration into the stabilized
tailings.

     This paper discusses the selection of design criteria sufficiently conser-
vative to assure a high probability that the embankment will last 1,000 years
or more essentially intact.  Among the criteria utilized in the designs are the
application of Maximum Credible Earthquake, Probable Maximum Precipitation and
Probable Maximum Flood to the location, and configuration of the proposed
disposal site.
INTRODUCTION AM) PURPOSE

     Title I of the Uranium Mill Tailings
Radiation Control Act of 1978 authorizes
the Department of Energy (DOE) to cleanup
and isolate from Che environment the
radioactive tailings and other contaminated
materials at 24 inactive uranium mill
sites.  The principal purpose of this act
is protect the public health from the"
radionuclides that are either emitted from
the tailings as radon gas or carried away
from the sites by wind or water movement.
     The D.S. Environmental Protection
Agency (SPA) established health and
environmental standards to govern the
cleanup and control of these radio-
active materials In 1983.  In developing
the standards, EPA determined "that the
primary objective for control of tailings
should be Isolation and stabilization to
prevent their misuse by man and dispersal
by natural forces such as wind, rain
and flood waters.  Consequently, the
                                       441
-------
 standards set forth the following
 design objectives:

 *  Reduction of the radiation emissions
 (in the form of radon gas) to less
 than 20 pCi/n2

 *  Effective Isolation of the tailings
 from the environment for at minimum
 of 200 and ideally  1,000 years

 *  A "maintenance free" design

      As prime contractor to DOE,  the
 first priority of Jacobs Engineering
 was to establish design criteria  and
 an approach that fulfilled the standards
 and objectives and  was acceptable to
 the Nuclear Regulatory Commission (NRG),
 EPA and States.

      The design of  any engineered system
 to oeet a maintenance  free design life
 of 200 to 1,000 years  is complicated from
 the start by the absence of any histor-
 ical (or empirical)  examples.   The more
 famous  historical engineered systems in
 the range of 200 -  1,000 years old -
 forts,  castles,  earthworks,  canals,
 irrigation systems,  etc.  all either
 show the effects of  wind/water erosion
 or require  constant  care.   Further,
 research and development of engineered
 systems  to  withstand the vagaries of
 nature  is relatively new activity with
 only SO to  100 years of  data and  obser-
 vations.   Finally,  the record of  the
 major natural  forces,  wind  velocities,
 rainfall intensifies and duration,  flood
 velocities  and depths,  etc.  only  cover
 about 70-80 years maximum in most of the
 localities  and the measuring stations
 usually  are not  very near to the
 slte(s)  being evaluated.  Consequently,
 the  selection of 200-1,000  year design
 criteria involves coupling  scientific
 data, statistical methods and  an  evalua-
 tion of  risk.

APPROACH

     Developing design criteria to
Beet the  first of the  objectives,
 radiation protection, was relatively
easy.  Through research  sponsored  by
 DOE,  NRC and  others,  radiation  reduc-
 tion  was demonstrated to be effectively
 achieved by the  use of a soil cover
 system to trap the radon gas until it
 decayed to solid radionuclides.  This
 system ranged from approximately 1-1/2
 to 3  meters thick depending upon
 several factors  including the strength
 of the radioactivity  (source term), types
 of soils available for cover materials
 (sandy/sllty  clays attenuate radon better
 than  sands or silts), the expected  long-
 term  moisture content and the degree of
 of compaction.

      Protection  of the radiation barrier
 or soil  cover system  required much more
 extensive engineering analysis.  The most
 critical design  concerns Include differ-
 ential settlement which could rupture the
 cover system, slope failure which could
 also  result in the exposure of tailings,
 prevention of erosion from both normal
 weather  events and large, intensive rain-
 falls  and  the associated runoff.  Coupled
 with  the concern about embankment erosions
 is  the related concern about regional erosion
 potential  (gully formation, etc.) and river
 channel  migration.  Finally, protection
 of  ground-water  requires an analysis of the
 impacts  of ground water movements up into
 and through the pile,  the effects of the
 cover  system on ground water and the tran-
 sitional impacts of construction water on
 ground-water quality.   These can be sum-
 marized  as four principal design concerns:

      *   Stability
      *   Longevity
      *   Ground-water protection
      *   Cost

     The stability design criteria were
also not too difficult to establish.
 Since  the tailings piles had random sand
 and slime layers, wedges and/or deposits,
 it was recognized quickly that  design
 criteria were required for slope stability
and differential settlement.

     Most of the tailings piles presently
exist at or near their natural  angle  of
repose — someplace between 1.5  and  2.0
horizontal to  1.0 vertical.   These  steep
slopes are very susceptible to  slumps,
                                        442
-------
 slides and erosion.   Drawing upon pre-
 vious  work by DOE and NRC (DOE-1), a
 design criterion of  maximum side slopes
 of  5H:IV was  established.  Further, the
 stability of  this slope was verified by
 establishing  the worst case pile profile
 (thickness and number of sand/slime
 layers plus cover system) and calcula-
 ting the passive factor of safety and,
 using  the Maximum credible earthquake to
 determine the dynamic factor of safety.

     Differential settlement would be
 minimized by  homogenizing any relocated
 tailings  and  by compacting both the
 tailings  and  the cover system to at
 least  95% dry density.

     The  factors which affect the long-
 evity  and their associated effects
 including:

 0  water/wind erosion — cover damage

 *  regional erosion  potential - gully
 ing and under cutting

 °  flooding - catastrophic embankment
 failure

 0  river  erosion/migration - undercut-
 ting and  erosion of  embankment  systems
were more difficult  to develop  accept-
able design criteria  for.

     The  effects  of water/wind  erosion
are related to  the exposed  surface  mater-
ials and  the  exerted  erosive forces.
The selection  of  surface  materials  is
principally dependent  upon  the  areal
climate.  Since most  of the  sites are
located in semi—arid  areas,  it  was  im-
mediately recognized  that a  vegetated
covet  could not  be established  and  main-
tained without some form  of  irrigation
system which  defeated  the maintenance-
free objective.   Therefore,  a rock  or
gravel cover  layer was  incorporated into
the design to withstand Che  wind and
water erosive  forces.

     The  design wind/water  erosive  forces
are dependent on  the magnitude  of the
selected  design  event.  Further,  the
erosive forces of an intense rainfall are
 recognized as far exceeding those  from
 any  windstorm so  it was  decided  that
 design  of  an erosion  protection  layer
 for  runoff would  protect equally well
 against the wind.  (DOE-2)

      Since the Ideal  design life is
 1000 years, it was obvious  that  the wind
 velocity or rainfall  intensity had to be
 at least equal to  that generated by a storm
 with a  recurrence  interval  of 1000 years
 or more.   However, since these are random
 events,  there is  some probability that an
 even larger event will occur within 1000
 years.   This probability (or risk) can be
 calculated for any larger event  (500 year,
 1000 year,  etc.), but there are  two compli-
 cating  factors.  First,  there is a large
 degree  of  uncertainty in extrapolating
 50-100  years of historical meteorological
 records  to larger storms with more infre-
 quent recurrence intervals.  Second,
 there is no established  or agreed on
 "reasonable"  risk (i.e., 102, It, 0.12
 chance  of  a larger than  design event
 occurring  and breaching  the embankment)
 to be used in the design,  therefore,
 design methods  must be adopted which
 incorporate conservatisms into the design.
 The design  criteria for  the stability of
 the UMTRA  Project tailings piles due to
erosive  forces  resulting from rainfall
 runoff  across  the top and down the sides
of the stabilized embankment are based on
 the runoff  from the localized Probable
Maximum  Precipitation (PHP).  The PHP is
 the worst  possible event that could occur
 as a result  of  a combination of the most
severe meteorological conditions occurring
over a watershed at the same time.   Al-
though no  recurrence interval can be
assigned to  this event,   it is felt by most
experts that  the recurrence interval is
 In excess  of  100,000 years.  This roughly
equates to a less than 1Z chance that a  PMP
event will occur in the next 1000 years.
Procedures investigated  for calculating
the required mean rock size needed to pro-
vide a stable rock slope during the PMP
 included:

 0  Bureau of Public Roads (Searcy,  1967).

 °  U.S. Army Corps of Engineers Waterways
   Experiment Station U.S.  Ar*?,  (1970).
                                        443
-------
 *  California Division of  Highways (Cal.
    DPS,  1910).

 *  ASCE  Task  Committee on  Preparation
 of  Sedimentation Manual (ASCE,  1972).

 *  Bureau of  Reclamation (Bur.  Rec»,
 1958).

 *  Lane's Method (Lane).

 *  Campbell's Method  (0.S. Army,  1966).

 *  Wyoming State Highway Department
 (Safety  Factors Method)  (Stevens,  1976).

 *  Rockflll in Hydrologlc  Engineering
 (Stephenson,  1979).

     All of these procedures use one  of
 four basic approaches:

 1)  critical velocity,  2) lift and  drag
 force mechanisms, 3) critical shear
 stress, or 4) empirical  solutions.

     The critical velocity equations  con-
 sider the impact of flowing water  on  the
 particles such that the  material of a
 given size and weight  is just able to
 move.  Inherent in this  approach is the
 lack of good definition  of the  bottom
 velocity and the difficulty in  accurately
 measuring or predicting  this velocity.
 Another difficulty in using this
 approach Is determining  Che relationship
 between bottom velocity  and average
 velocity.

     The lift force mechanism approach
 accounts for a pressure  differential
 caused by Che gradient of  the velocity.
The pressure differences occur  because
of steep velocity gradient, where  the
velocity at the top surface of  a par-
 ticle at rest on a channel bottom is
greater than zero, while the velocity
at the bottom surface Is zero.  This
 method although important,  has  been
difficult to develop.  Additionally,
 the critical shear stress and critical
velocity methods Implicitly include the
 life force effects.
     The  critical  sheer stress equations
 consider  frictional drag of the flow on
 the  particles and  considers the fluid
 shear stress on a  rock layer to the
 mean flow velocity.  This approach is
 the  best  for the UMTRA project because
 mean cross-sectional velocities, depths
 and  flow  can be easily obtained assuming
 sheet flow characteristics.

     The  design method which Is most
 applicable to the design of a rock blanket
 for  erosion protection on the top and
 along the toe of the embankment is the
 "Riprap Design with Safety Factors Method"
 developed for the Wyoming State Highway
 Department by Stevens et al. (1976).  The
 theory and formulation of this method
 are  not discussed as part of this paper
 since they are well documented in the
 published paper.  This method is based
 on the theory of critical shear stress
 and  allows more flexibility in design
 because the designer is able to choose
 the  factor of safety needed for the
 design of a particular site and work
 through the analysis to determine the
 required  rock size.  This flexibility
 is particularly important when consid-
 ering the conservatism associated with
using the PHP as the design storm.

     Either the Safety Factors Method or
 the  Stephenson Method,  which is based on
 critical shear stress,  are suited for the
design of the rock layer on the side
slopes for a PMP centered on the embank-
ment.  The Safety Factors Method is
applicable so long as the depth of flow
la greater than 0.5 tines the mean rock
diameter.  Below this ratio, the Safety
Factors Method is not accurate and the
Stephenson Method must  be used because
the  flow is no longer sheet flow but Is
interstitial among the  rock particles.

     The selection of a design flood event
to achieve the EPA longevity requirement
would appear to be straight-forward.
The standard request from the Nuclear
Regulatory Commission (NEC), and the
standard design approach taken,  Is to
determine the magnitude and potential
                                         444
-------
 Impacts  resulting  from  a  Probable
 Maximum  Flood  (PMF) event.  However,
 the use  of  the PMF as the control  of
 uranium  tailings is not clear—cut.

     The EPA standards do not specif-
 ically state that  a PMF event must  be
 used for design in order  to achieve
 the stated  containment  life.  An ana-
 lysis of exceedence probabilities  for
 various  events with respect to  the
 contaiiutent life (Junge and Dezman,
 1983), suggests that design events  with
 a very long return period  (e.g., 10,000
 years or greater)  must be used  to meet
 a longterm  containment objective.   How-
 ever, the limited statistical record
 available cannot accurately be  extra
 polated  to  such long return periods.
 Therefore,  the generally  accepted alter-
 native,  by  default, is to use maximum
 credible events such as the PMF for
 design purposes.

     By  definition, a PMF  is based  on
 the most severe combination of  critical
 meteorologic and hydrologic conditions
 for a particular area, and has  a very
 small chance of being exceeded.  There-
 fore, a  tailings disposal system designed
 to withstand a PMF would have a very
 small risk  of  failure and thus, meets
 both the intent and containment objective
 of the EPA  standards.

     Although the use of a PMF  meets
 the intent  of  the EPA standards, this
 does not mean that designing for a  flood
 of lesser magnitude would not also  meet
 the longevity requirement.  The use of
 a lesser flood event, however,  is strong-
 ly discouraged by the NRG and concurrence
would be difficult to obtain.

     The simplest approach is to use
 the PMF as  the design flood event but
 recognize that the conservative nature
of its use  is like having a built—in
 factor of safety.   This  fact can be
used as  the basis against having Co
apply additional arbitrary factors
 of safety and overly conservative para-
 menters throughout the remainder of the
design.
      Developing  the design  velocities
 and  depths  related to the PMF was done
 using the HEC-1  and HEC-2 Models (COE,
 1981,  1982).  The principal inputs to
 the  HEC-1 Model, which develops the
 hydrograph  are:

 8  PMP estimates and hydrographs for
 each distinct drainage area in the basin

 0  Hydrograph time interval

 °  Soil  infiltration rate on runoff
 curve  numbers

 0  Lag times for each subbasin.

     The PMP estimates are generally
 derived from the National Oceanic and
 Atmospheric Administration (HOAA) Hydromet
 publication appropriate to the area.
 Since  most of the mill tailings sites are
 west of the 105° meridian, the hydrographs
 had  to be developed and input directly.

     Normally, flood hydrograph time
 intervals are chosen to be either 15
 minutes or one hour.  The 15 minute
 Interval will produce a much higher
 peak flow than using a one-hour interval.
 It was concluded, however, that a time
 Interval of less than one hour is too
detailed for the magnitude of a PMF ana-
 lysis.

     Soil infiltration is more difficult
 to effectively estimate.   Soil infil-
tration rates can be directly input into
the HEC-1 model; or runoff curve numbers
can be specified and used by the model to
calculate the soil infiltration rates
based on empirical formulas.  Either
method requires a knowledge of the soil
and vegetation characteristics of the
drainage basin.  With small drainage
basins, this evaluation can be easily
accomplished.  Unfortunately, most  of
the tailings sites located in or near
flood  plains are Impacted by drainage
basins comprising thousands of square
miles.  The parameters  are highly sen-
sitive and can very the results by
several hundred percent.
                                            445
-------
      If runoff curve  numbers  are used,
 it  should  be  noted  that  this  system was
 originally developed  solely for agri-
 cultural land.  The basis  upon which the
 empirical  equations were developed  is
 not well published  nor understood.
 There is also a wide  variation in curve
 numbers depending on  what  antecedent
 moisture condition  (AHC) is used.
 tower curve numbers are  specified for
 an  AHC-II  condition than an AMC-III
 conditions (BOI, 1977).  The  user may
 feel in his judgement that an AMC-II
 condition  is  sufficient  for use with a
 PMF.   However,  it is  noted that the
 opinion of the NRC  is to use  an AMC-III
 condition.

      It  is  the opinion of  the design
 team that  the nethod  of  specifying  the
 •oil  infiltration rate is  not only
 easier to  develop but also has a more
 established basis.  As an  example,
 through discussions with the  U.S. Army
 Corps  of Engineers  (COE) in Ohaha,
 Nebraska,  a nap of  "Soil Infiltration
 by  Generalized  Soil Groupings"  (MBIAC,
 1966)  was  obtained  for use on the River-
 ton, Wyoming,  site  flood analysis.   The
 COE has  used  it  and recommended its use
 for PMF  analysis.  The HRC, on the  other
 hand,  has  indicated that the  recommended
 values on the  map are  too  high.  This is
 very  important  to note due to the extreme
 sensitivity of  this parameter.   A sensi-
 tivity analysis  of  this  parameter on the
 Elverton site  flood analysis  showed that
 a 0.1  inch  change in  this  parameter
 resulted in over a 200,000 cfs  change
 in  the PMF  flow  rate.

     tag times  for subbasins  throughout
 a watershed are  typically  computed  using
 the lag  time  empirical relationship
 contained  in  Design of Small  Dams (DOI,
 1977).  Experience has indicated  that
 this  relationship generally gives longer,
 less conservative lag  times than what
might occur.  The approach  of  the design
 team is  to  initially calculate  lag  times
with the above stated relationship.  The
 resulting peak  flows  for each subbasin
hydrograph are then used with Manning's
equation to determine a  better  estimate
of channel  velocities.  The velocities
 are  then used  to  recalculate the routing
 lag  times.

     Once the  hydrograph was established,
 the  flood routing was done using HEC-2
 was  relatively straightforward.  The only
 serious concern was selecting appropriate
 Mannings "n" values for the channel and
 overbank areas.

     The most  difficult problem area to
 evaluate was the  potential for river
 migration/realignment.

     The primary  geomorphlc concern with
 long-term stabilization of sites located
 in flood plains is the potential for lat-
 eral movement  of  a stream channel causing
 undermining or erosion of the tailings
 impoundment.   Stream channel migration
 can occur gradually during the period of
 containment.   A more severe situation that
 can occur, however, is a rapid channel
 shift in response to a major flood event.
The potential  for channel migration must
 be carefully evaluated on a site — specific
 basis with all available geomorphic data.
Two examples are given to illustrate the
difficulaty of correlating the possible
geomorphic changes with the PMF analysis.

     The Gunnison, Colorado site is a
tailings pile  located in a flat, terraced
valley about 1 mile upstream of the
confluence of  the Gunnison River and
Tomichi Creek.  The entire valley was
comprised of terrace gravels,  sands and
silts deposited by floods in prehistorical
times and the difference in elevation
between the site and the watercourses is
only about 5 feet maximum, (Figure 1).
The Rifle,  Colorado site (Figure 2) is
different in that the tailings  piles lie
in the modern  floodplain about  1000 yards
downstream of  a 90° bend that  is subject to
continuous severe erosion forces.

     In the case of the Gunnison site,
the Gunnison River was classified as
having only moderate with a high poten-
tial for channel migration.  An evalua-
tion of the existing Gunnison  River
channel alignment would Indicate that
during 8 PMF the alignment would shift
laterally towards the site in  a gradual
                                        446
-------
 curve  Co  the  confluence  with Tomichl
 Creek.  A limitation  of  the  HEC-2  model
 is  that it cannot  predict  or estimate
 changes in channel  configuration that
 might  occur during  a  major flood event.
 Additional HEC-2 analysis  was conducted
 to  check  mean channel velocities assum-
 ing that  major erosion of  the overbank
 area has  occurred towards  the embank-
 ment.  The Gunnison River  channel  was
 assumed to have widened  from its present
 location  all  to the edge  of the embank-
 ment.  The widened  channel was assumed
 to  have a uniform steanbed elevation.
 since  the  widened channel  would have a
 uniform depth then  the flow  velocity at
 any point  would equal  the  mean channel
 velocity.   The resulting velocities and
 assumed depths of undercutting were
 then incorporated into the design  to
 assure that adequate size  and quantity
 of  rock protection  was provided at the
 top and sides of the embankment.

     At the Rifle site,  the  Colorado
 River has  a high potential for channel
 migration.  The evaluation of the
 channel and floodplatn topography  showed
 that under  PMF conditions  the water
would flow  essentially from  east to west
 across the site and the  present pile
alignment  would be  a large blocking
 force.  To mitigate this,  the proposed
design specified realigning  the tailings
 embankment on an east-west axis and
 pointing the upstream end  like the prow
of a ship.  This,  plus a riprap blanket
extending  from the  embankment to the
present dike were felt sufficient to
prevent both undercutting due to daily
erosive forces and  breaching during
catastrophic events.

PROBLEMS ENCOUNTERED

     Interpretation and application of
above design criteria were not without
some problems in design conservatism.
The reviewing agencies,  in particular the
NRG, wanted the safety factors (SF) to be
calculated by both  the circle of failure
and sliding wedge  approach even when the
circle of   failure  showed SF's greater
than 2.0.   Also,  there was a  constant
effort to design to higher SF's.
      Commonly,  a S.F.  of  1.0 was used
 for all  dynamic situations  since it  was
 argued that  the low probability of the
 design events was  in itself sufficiently
 conservative and the use  of a higher S.F.
 was simply overdesign.

      An  additional problem  was in amount
 of  data  required by the reviewing agency.
 It  was standard practice  to provide  the
 regulators with design summaries and the
 supporting calculations and data.  How-
 ever,  in almost all cases,  the reviewing
 agencies, particularly the  NRC,  responded
 with requests for  more data.   At first it
 was justified in that the program was
 still  being  defined,  but after the 4th
 site,  it appeared  that it Bay have been
 more the case of avoiding the  decision by
 requesting "more study".

     Finally, a part of the  problem  was
 the  uncertainties  in achieving a 1000
 year design  life.   Since it  had  never
 been done before,  even the methods were
 subject to a certain amount  of reserva-
 tion.  It often appeared that  personnel
 started out with a preconceived  idea of
 rock size, PMF  size, etc and  then  argued
 for  the methods  or assumptions that  best
 supported their position.   This  made
 agreement on siting and design very dif-
 ficult.

 RESULTS
     Notwithstanding the problems, the
Jacobs design team were able to formulate
an approach that was acceptable for the
design of tailings embankments to last
1000 years.  The basic approach was to
first qualitatively analyze the poten-
tial for castastrophic failure — floods,
earthquakes, etc.  If this analysis in-
dicated that the site was potentially
stable, then a more detailed analysis
was conducted.

     Generally, the potential for flood
damage or river channel migration were
the most severe limitations and required
the most engineering.  Once this was
done, a decision could be made on whether
the tailings could be stabilized-in-place
                                        447
-------
or had  Co  be  relocated.   Based  upon this
decision,  the design  of  the  final  cover
system  proceeded without too much  dif-
ficulty.

REFERENCES

1.  Campbell, F.B.,  1966.  "Hydrologic
    Design of rock Riprap,"  Miscellaneous
    Paper No. 2-777,  prepared by U.S.
    Army Engineers, Waterways Experiment
    Station,  vicksburg,  Mississippi.

2.  COE  (U.S. Army Corps of Engineers),
    1970.  Engineering and Design,
    Hydraulic Design  of  Flood Control
    Channels, EM 1110-2-1601, Office of
    the Chief of Engineers,  Washington,
    D.C.

3.  COE  (U.S. Army Corps of Engineers),
    1981.  HEC-l Flood Hydrograph
    Package. User's Manual,  Computer
    Program 723-X6-L2010, Water Resources
    Support Center, The  Hydrology Engi-
    neering Center, Davis, California.

4,  COE  (U.S. Array Corps of Engineers),
    1982.  HEC-2 Water Surface  Profiles,
    User's Manual, Computer  Program 723—X6-
    L202A, Hater Resources Support Center,
    The Hydrology Engineering Center,
    Davis, California.

5.  Crlppen, J.R., and D.C.  Bue, 1977.
    Design o£ Small Pans, U.S.  Bureau
    of Reclamation, Water Resources
    Technical Publication, Washington,
    D.C.

6.  DOE  (U.S. Department of  Energy),
    1984.  Design Criteria for  Sta-
    bilization of Inactive Uranium
    Mill Tailings Sites. OMTRA-DOE/
    AL-163, Uranium Mill Tailings
    Remedial Action Project,   Albu-
    querque, N.M.

7.  DOE  (U.S. Department of  Energy),
    1984.  Plan for Implementing SPA
    standards forUMTRA  Sites,  UMTRA-
    DOE/AL-163Uranium Mill  Tailings
    Remedial Action Project,  Albu-
    querque, N.M.
8.  DOT  (U.S. Department of Trans-
    portation), 1975.  Design of Stable
    Channels with flexible Lining,  Hyd
    raulic Engineering Circular Ho. 15,
    Federal Highway Administration,
    Washington, D.C.

9.  Junge, W.R., and L.E. Dezman,  1983.
    An Analysis of Control Standards
    for che Long-Term Containment  of
    Uranium Tailings, prepared by  the
    Colorado Geological survey and the
    Colorado Division of Water Resources,
    Denver, Colorado.

10. MBIAC  (Missouri Basin Interagency
    Committee), 1966.  Soil Infiltra-
    tion - By Generalized Soil Group-
    ings, Base Map Compiled by U.S.
    Geological Survey,  copy available
    from the UMTRA Project Office,
    Albuquerque, N.M.

11. NOAA  (National Oceanic and Atmos-
    pheric Administration), 1977.   Hy-
    drometeorological ReportNo. 49,
    Probable MaximumPrecipitation
    Estimates - Colorado River and
    Great Basin Drainages,  Silver
    Spring, Maryland.

12. NOAA  (National Oceanic and Atmos-
    pheric Administration), 1978.   Hy-
    drometeorological Report No. 51,
    Probable Maximum Precipitation
    Estimates - United  States East  of
    the 105th Meridian,  Washington,  D.C.

13. NOAA  (National Oceanic and Atmos-
    pheric Administration), 1984.   Hy-
    drommeteorologlcal  Report No. 55,
    Probable Maximum Precipitation
    Estimates - United  States Between
    the Continental Divide and the  103rd
    Meridian, Silver Spring,  Maryland.

14. NRC  (0.S.  Nuclear  Regulatory Com
    mission), 1983.   Staff  Technical
    Position WM-8201, Hydrologlc Design
    Criteria for Tailings  Retention
    System, Low-Level Waste Licensing
    Branch, Washington,  D.C.
                                       448
-------
15.  Stephenson, Bavtd, 1979.  Rockflll
    In Hydrologtc Engineering, Elsevter
    Scientific Publishing Company,
    New York,  New York.

16.  Stevens et al. (M.A,  Stevens, D.B.
    Simons, and G.L. Lewis), 1976.
    "Safety Factors for Riprap Protec-
    tion," in  American Society of Civil
    Engineers  Journalof  Hydraulic
    Engineering.

17.  Walters, W.H., 1982.   Rock Riprap
    Design Methods and Their Appli-
    cability to Long-Tenn Protection of
    Uranium Mill Tailings Impoundments,
    NUREG/CR26-84, PNL-4252, prepared
    by Pacific Northwest  Laboratory,
    Richland,  Washington.
                                   Di sclaimer

                   The  work described in  this  paper was
                   not  funded by the U.S. Environmental
                   Protection Agency.  The contents do
                   not  necessarily  reflect the views of
                   the  Agency and no official  endorse-
                   ment should be  inferred.
                                      449
-------
                     A NOVEL TYPE OF NUCLEAR REACTOR -  THE
                                 HYDRO REACTOR

                                  6e Andlauer
                                   Ener Plan
                              Mundelheim, France


                                   ABSTRACT

     It  consists  of  a solute  uranium oxide  in water.   The nuclear  fission
induces  four  sorts  of ionizing  radiation.   First  the  alpha  particles  from
uranium  self  and  the  trans uranic  nuclei  such  as  plutonium  or  neptunium.
Then  beta  particles   issued  from  the fission  products  such  as  cesium  and
lanthanum.  Third  the  fission  neutrons that  collide  on  H20 molecules so  as
to eject  secondary  protons.  At last  theYrays that ionize the H20  molecules.
These  divide  into a  hydrogen and  a hydroxide  ion  according to H£0  — H+  +
CH~.   The hydrogen atoms  react  together in  order to  form  H£ molecules  that
rise out  of the solution.  The  OH  radicals react on heavy metal cations  such
as U^* in order to form insoluble hydroxides that replace the  H  capture.

     The  reactor  produces  hydrogen  as useful  energy  source and  reduces  the
related radioactive waste  in terms of its  simplicity and life.

References

F.  S.  Dainton  -  Chemical  Reactions  Induced  by  Ionizing Radiations -  Annual
Rep. Progr. Chem.  45 - 5/33 1948

H. Mohler - Chemische  Reaktionen  lonisierender Strahlen  -  Aarau  -  Frankfurt
a M - 1958
                                    D1sclaimer

  The work described in this paper was not funded by the U.S. Environmental
  Protection Agency.  The contents do not necessarily reflect the views of the
  Agency and no official endorsement should be inferred.
                                      450
-------
   EXTRACTION OF PESTICIDES FROM PROCESS STREAMS USING HIGH VOLATILITY SOLVENTS

                                 Stan L. Reynolds
                                      S-CUBED
                                San Diego,  CA 92129


                                     ABSTRACT

    This cooperative agreement research program was initiated with the overall  ob-
jective of determining the feasibility of extraction of pesticides from process
waste streams using a liquid-liquid solvent extraction approach.   A literature  re-
view provided foundational data in support of the technology identified for study
through:  (1) priority-ranking and selection of pesticides for project study and
the identification of optimum solvents for liquid-liquid extraction, (2) compilation
of various engineering design options for liquid-liquid extraction systems culmi-
nating in the selection of a design choice consisting of a counter-current-flow ro-
tary disk contactor (RDC) process, and (3)  application of literature on engineering
design economics to formulate reliable cost estimates for the derived technology.
While liquid-liquid solvent extraction is not new and has, in fact, been used in
the chemical process industry for many years, it has seldom found use as a pollu-
tion control technology.  The novel aspect of the technology described herein is
that it uses volatile organic solvents (within a boiling range of 34°C - 78°C)  with
very low water solubility characteristics.   The bench-scale RDC-type Solvent Ex-
traction of Organic Pesticides (SEXOP) system was used to test both synthetic and
real-world process effluent samples.  Tests using the synthetic sample demonstrated
extraction efficiencies in excess of 99.9 percent.   Subsequent to process optimiza-
tion, bench-scale runs using real-world process effluent samples (consisting of raw
untreated waste) resulted in a clean-up for most constituents in excess 6f 99 per-
cent.  No constituents of the best-case raw untreated waste final run were ex-
tracted at an efficiency level below 90 percent.
INTRODUCTION AND PURPOSE

    Research conducted on the project
was structured such that the following
three areas were addressed:

1.  Solvent-pesticide partition coef-
    ficients were established through
    use of both synthetic pesticide
    mixtures and pesticide manufactur-
    ing wastewaters.

2.  A bench-scale liquid-liquid extrac-
    tion unit was designed and fabri-
    cated and experimental studies were
    conducted to test the capability
    of the system.
3.  Subsequent to the bench-scale runs,
    economic feasibility studies were
    conducted in order to determine
    the potential process viability as
    compared to commercial-scale car-
    bon adsorption systems.

    Research objectives in the above
three areas were met through work in
the following four discrete task areas:

Task 1 - Literature Review
Task 2 - Screening Extraction Studies
Task 3 - Bench-Scale Studies
Task 4 - Economic Analysis
                                        451
-------
    Foundations! concepts and suppor-
tive data for work on this project were
derived from prior in-house S-CUBED
supported research into the feasibility
of liquid-liquid solvent extraction of
pesticides from water.
APPROACH

    The central focus of the approach
described herein rests on the selection
of high volatility, low (water) solu-
bility solvents as extraction media.
Specific solvents targeted for inves-
tigation during the course of this
stuoy (listed in order of descending
polarity) were:

1.  Butyl chloride,
2,  Isopropyl ether,
3.  Diethy! ether,
4.  Hexane, and
5.  Pentane.

    During the course of this study,
each of these candidate solvents were
evaluated as extraction media for use
in the bench-scale experiments.

    In contrast to the above list, the
solvent extraction media generally em-
ployed in commercial liquid-liquid ex-
traction systems are much less volatile
and usually considerably more soluble
in the treated matrix.  Examples of
typical solvents used in industrial
liquid-liquid extraction system include
nitrobenzene, cresol, phenol, furfural,
and blends of benzol-sulfur dioxide.
In such systems extraction frequently
takes place at elevated temperatures
and steam stripping of the raffinate
and/or extract may be required for
solvent recovery or increased purifi-
cation.

    While the theoretical and technical
aspects of this technology are highly
refined on a commercial level, neglig-
ible consideration has been devoted to
applications in pollution control tech-
nology areas.  This is partially due
to the highly specialized nature of
the solvents required and also to the
costs and further environmental  prob-
lems associated with steam stripping,
and solvent reclamation.   Also,  very
little attention has been given  to the
use of solvent extraction for removal
of trace-level constituents in process
effluent streams.

    The SEXQP approach attempts  to ad-
dress these issues through use of sol-
vent systems which can be employed at
ambient temperature and which can be
stripped and recovered easily due to
low water solubility and high volatil-
ity.

Task 1 - Literature Review

    A literature review was conducted
at project initiation with approxi-
mately a two-month term allocated for
the bulk of the search and a low level
of effort allocated over the remainder
of the project to keep the file  cur-
rent.  The purpose of the literature
review was to acquire information con-
cerning:

*  Those pesticides which are the most
   environmental signficant, either
   because of their high toxicity
   ranking or because they are manu-
   factured in large quantities.

•  Candidate solvent systems which can
   be employed effectively in continu-
   ous liquid-liquid extraction  sys-
   tems.

»  Continuous liquid-liquid extraction
   engineering process designs.

•  Economic considerations with respect
   to solvent selection and process
   design.

    At the initiation of the literature
review a special category was set aside
for each of the above four areas of
study.  This resulted in the formula-
tion of an annotated bibliography
                                        452
-------
structured according to:

»  Subject category,
   Reference
   author,
in alphabetical  order by
•  Brief bibliographic statement.

    This annotated bibliography, con-
taining more than 60 references is in-
cluded in the NTIS Report resulting
from this research effort (JJ.  Many
of these citations contributed greatly
to the work discussed herein and ac-
knowledgment to this effect is docu-
mented in the above referenced report.

Task 2 - Screening Extraction Studies

    Based on information derived from
the Task 1 - Literature Review, the
solvents and pesticides identified as
experimental candidates were procured
for screening extraction studies.

    The screening extraction studies
were conducted using a single batch
storage extraction of water containing
a known quantity of dissolved pollutant
under investigation.  The organic phase
was then isolated and the quantity of
pollutant extracted was determined us-
ing electron capture gas chromatography
(EC/GC).  The resulting EC/GC data was
compared to the pollutant concentration
in the aqueous phase and the partition
coefficient (Kp) for each solvent sys-
tem was then calculated.  These calcu-
lated partition coefficient data were
reviewed in order to select the most
promising solvent/pesticide combination
for use in the Task 3 - Bench-Scale
Studies.  Table 1 presents some cri-
teria used for the selection of solvent
for the bench-scale study.

    Viewing Table 1, the kind of trade-
offs required in the solvent selection
process become readily apparent.  While
n-butylchloride is an excellent solvent
from the standpoint of water solubility
and Kp value, the boiling point is
high and the cost in terms of dollars
per gallon is relatively high.  Both
of the ethers have attractive Kp values
and costs but water solubilities are
high and the ever present danger of
peroxide formation and explosion con-
stitutes a substantial safety consider-
ation.

    TABLE 1.   CRITERIA USED  FOR  THE
     SELECTION OF SOLVENT FOR THE
           BENCH-SCALE STUDY
Solvi-nt
n-«ytyl-
eaiortii*


IswrOBYl
EUlir


IMI4WI



NfltiR*



mttnyi
CUtr


SoIufclHty 10111114
tn uattr Point
an (*«
0.02 ?».4



0.2 68



0.03 49



0.01 3t



».» M.S



VSjl K,
3.20 OUT1
TottVMM
Cnlornin*
.wrfl union
2. so OOT'
TOUPtIM*
Cfilortiant
Korfluntofl
2.40 «BT'
Toupntn*
Cftiofdtn*
Korf lunizon
3.00 88T'
Toupnmt
C!itQr3*j9*
mrfl union
2.70 OOT1
Toxcplwi*
CftlQrMll*
Morf 1ur«ion
C^
310.000
ito.ooo
249,000
200
230,000
140,000
84,000
tl
320,000
74,000
11,000
17
270.000
39,000
26,300
27
960,000
1(0,000
130,000
ISO
Id! valut)
Qluran
2,4-fl
Srooiell
^lypnosatt
Olaron
2.4-0
Irowell
SlypnoMt*
Blsron
Z.4-0
3rt»rtc1l
GlyonoJltt
Blyron
2.4-0
IroMCll
Slypnosatt
Ofuron
2.4-0
Irmcll
Slypnoutt

55.1
t.J
21.7
0.03
53.1
84
10.7
0.048
3.S
0.02
1.32
0.111
3.0
0.19
1 .25
0
182
6S.9
18. t
0.109
                                HtT 1i no Iwiqtr Mnnficturad '" tn* U.S.

                                   After consideration of all  of the
                                solvents selected  for study,  it was
                                decided that hexane probably  represents
                                the best overall compromise.   Its water
                                solubility and cost are both  low, most
                                Kp values are on the high  end of the
                                range,  and the boiling point  (while
                                higher than would  be desired  for opti-
                                mum conditions) is tolerable.

                                   Referring again to Table  1, it will
                                be noted that the  eight pesticides sel-
                                ected  for study cover a broad general
                                range  of Kp values.   This  diversity
                                in Kp  values was chosen intentional- -
                                ly in  order to challenge the  scope of
                                applicability of SEXOP.  While it is
                                predictable that extraction efficien-
                                cies would be high for species with
                                large  Kp values, it is not clear at
                                which  point efficiencies become un-
                                acceptable for lower Kp values.  It
                                should be stressed,  however,  that Kp
                                        453
-------
 value was not the only criterion for
 pesticide selection.  The selection
 process also included ranking pesti-
 cides as a function of environmental
 significance either because of high
 toxicity rating or because they are
 manufactured in large quantities.  It
 should be noted that for the purpose
 of brevity the term "pesticide" is used
 herein to denote both pesticide and
 herbicide species.

 Task3 - Fabrication of the Bench-Scale
          Device and Conduct ofExperi-
          mental Process Run?

     The liquid-liquid extraction pro-
 cess design configuration selected for
 the bench-scale system was of the
 "continuous differential contactor"
 variety.  All differential contactor
.designs utilize a two-phase counter-
 current flow technique wherein the
 material of greater density is intro-
 duced at the top of an extractor unit
 and the material of lesser density is
 introduced at the base.  Within this
 category, S-CUBED elected to design
 and fabricate a rotary disk contractor
 (ROC) having determined that this de-
 sign configuration would optimize ex-
 traction efficiencies.

     Figure 1 is an illustration of one
 ROC stage.  A total of four stages
 comprise the column with the outer
 shell of the RDC unit consisting of a
 stainless steel (SS) pipe four feet in
 length and three inches in diameter.
 An explosion-proof motor is mounted
 about one foot above the top end plate
 of the RDC.  Ball-bearings support the
 shaft just above a flexible coupling
 which connects the motor shaft to the
 RDC shaft.  The motor is connected with
 a variable-speed controller in order
 to permit a variable range rotational
 speed with a maximum of 1,750 revolu-
 tions per minute (RPM).  The stator-
 ring cluster, also shown through the
 view port in Figure 1, consists of a
 series of rings which are 0.035-inch
 thick with an outside diameter slight-
ly less than a three inch inside diam-
eter of the RDC shell.  The stator
center!ine holes are two inches in di-
ameter.  The stators are held in posi-
tion by three 3/16-inch rods running
the entire length of the column.  One
inch spacers are used to separate the
stators, and threaded tightening nuts
are used at each end of the rods.  A
mid column bearing is installed to
minimize vibration of the rotating
shaft.  The bearing is held in posi-
tion by three 3/16-inch rods and is
constructed from a center-punched
teflon disk.  Peripheral to the RDC
unit is a counter-current solvent
stripper which, when fully assembled,
is four feet in length and three inches
in diameter.  It is composed of assem-
bly sections one foot in length to in-
crease or reduce the total length, as
needed, depending on efficiency condi-
tions during experimental runs.  The
packed bed of the column consists of
six-millimeter glass beads to simulate
the packings used in commercial-scale
strippers with the intent of providing
a larger surface area with sufficient
void space to permit efficient liquid
and inert gas (nitrogen) counter-cur-
rent flow.
Section Flange Clamp
Rotary D!«c


Stator Ring
                     Stiinloft Staal
                     window Support
                     Cilkat
Rotary Oftc Support.
Shalt   ^^
ft.«=r*.»» Support
Figure 1.
Close-up View of the Rotary
Disc Contract.
                                         454
-------
    Nitrogen and solvent vapor  emitted
from the top of the stripping column
are introduced into a multiple-loop
water cooled condenser which is made
of 0.5-inch SS tubing.  The condensed
solvent is accumulated in a collection
basin which is also fabricated  from
SS.  Reclaimed solvent from the col-
lector is recirculated to the RDC.

    A flow diagram illustrating the
configuration of the current bench-
scale system is presented in Figure 2.
Also shown in Figure 2 are the  compon-
ents, which would have to be added to
the existing bench-scale configuration
in the fabrication of a pilot or  larger
scale SEXOP system.
    Proctsi cMtwiwrrti wftMn tftt dasfttti H«e rtoraitfit tht configuration
    of tftt cyrr*»t B««s-*e*1» systM. toMDoacnts exterior to ta» 4«ft*a
    ?1n» rtpr***At projoctttf requirements for a co«»rci»l scilt lystttn.
Figure 2.  SEXOP Process Configuration.

SEXOP Runs Using Synthetic DDT  Solu-
tions

    Following fabrication of  the  bench-
scale SEXOP device, preliminary check-
out runs were conducted to ensure op-
erability of the integrated system as
a whole.  Once the process check-out
phase was complete and operational  in-
tegrity was assured, the next step in
the research plan involved the  conduct
of SEXOP runs using synthetic saturated
DDT solutions.  These solutions were
formulated in ten gallon  batches  and
were used to simulate extractor feed-
stock until the full volume  was de-
pleted through the course of one  or
more runs.  Individual  runs  were  from
two to three hours in duration  and the
exchange rate per hour  (i.e.,  full  RDC
column volume exchange) ranged  from
about 1.4 to 3.5.  The  total  RDC  col-
umn volume is 1.27 gallons (4.8 li-
ters).  The list of experimental  vari-
ables pertinent to each SEXOP  run is as
follows:

•  Rotary disk shaft motor speed,

»  Nitrogen stripper flow,

•  Solvent to water ratio,

•  Number of RDC column volume  ex-
   changes per hour.

These experimental variables were used
to formulate standardized engineering
run sheets which were subsequently used
for all bench-scale experiments per-
formed throughout the remaining dura-
tion of the project.  A total  of  five
synthetic DDT runs were conducted using
the bench-scale system.   In  all cases
extraction efficiencies in excess of
99.9 percent were realized using  satur-
ated DDT solutions as starting  mate-
rial.  Data for a typical  extraction
run is presented in Table 2  and is ac-
companied by gas chromatographic  data
in Figure 3 to illustrate the extrac-
tion efficiencies graphically,  for
starting material and TS  correspond-
ing to Table 2.

    In light of the excellent extrac-
tion efficiencies attained during the
synthetic DDT runs using  the hexane
solvent, it was decided to proceed to
the next phase of the activity  involv-
ing test runs using real-world  pesti-
cide manufacturer effluent samples.
                                        455
-------
 TABLE 2.   EXTRACTION RUN NO. 4 USING
 SATURATED  SYNTHETIC DDT SOLUTION AND
      A MOTOR SPEED  SETTING OF 70
        Ittmlm  CMtwIiunt    ConUalnant    function*
                      ewKMtritlon (»  tf«1c1my (»!
 lunlnt
 Mt*rUI
1
         10.21
         11.41
         t.H
         11.SI
 11.If
       1t.I4i.7M
        l,Stt,30»
         •72.011
       S4.U7.HI


         "!m
         »»;,»«
         211.110

         2).391
         21.472
                  1,071

                 10,711
 11.«

M 9
-------
due to a high  content of sol utilized
organic and  halo-organic compounds.

    A total  of four runs were  conducted
using this sample with progressively
higher extraction efficiencies for each
run resulting  from optimization of op-
erational variables and engineering
design.  Table 3 presents a  summary of
representative bench-scale runs from
this sequence.  Figure 4 presents gas
chromatograms  from Bench-Scale Run
No. 9 showing  the starting material
and Aliquot  No.  5, respectively.

   TABLE 3.'  BENCH-SCALE EXTRACTION
   RESULTS FOR RUNS 006,009, AND 010
                  Extraction
 009
 010
First ti»fteff**c*!* nifi on at proc*si tffluwt $a*ei«. B«ult» n*r»
only noflifittty luectss/u) (n tn*t wiy conititmmti *«r* not
$1?Blfleant ly rwhutcl.

Vtry poilttv* result*.  All pwk* «1tft ixcvptlen of tn« &.Q2
flrtmttt r«tt«t1en tlM eoitit1tii«at w*r« 1a tftt a1d~to-«pw 98th
p»re«itl1t txtnictloa •fHdwicy rang*.  Tht ptik with tht 6.02
arfnutt rtttntloa tint was txtrtctwi  *t *n *ff1e1**tcy sf 
-------
    In light of the technical results
of SEXOP with the synthetic and process
effluent DDT samples, an economic anal-
ysis of the process was undertaken to
develop an economic basis for the tech-
nology.  This was accomplished using a
two-fold approach:  (1) based on cur-
rent knowledge, an engineering cost
estimate was projected for both a large
and a small commercial-scale SEXOP pro-
cess; and  {2} these cost projections
were compared with the more firmly es-
tablished capital and operational costs
associated with carbon adsorption tech-
nology.  This study revealed that the
cost of SEXOP technology compares fa-
vorably with the carbon adsorption al-
ternative.

    A projected large SEXOP plant
should be able to process 301 million
gallons/year at a cost of $2.02/1,000
gallons.   Information derived as a re-
sult of the literature review effort
has shown that, based on total annual
cost, carbon adsorption may require
from $1.15 to $230.00 per thousand gal-
lons of effluent processed.  The most
common value cited is around $7.00/
1,000 gallons.  Based on the overall
number of citations, in the literature
studied, the most realistic range lies
between $3.00 to $10.00 per thousand
gallons of effluent treated.  There
are a number of reasons for the dis-
crepancies in the literature for capi-
tal equipment and OSM costs for carbon
adsorption.  Some process variables
which play an important role in the
wide range of carbon adsorption efflu-
ent treatment costs include:

*  Concentration of pollutant in the
   stream being treated.

*  Size of the facility ($/l,000 gal-
   lons treated tend to decrease as
   the size of the facility increases).

•  Whether the carbon is disposed of
   or regenerated and if regenerated,
   whether this is done on site or at
   a central location.
•  Location of the facility.

•  Effluent purity objective.

    Bearing these discrepancy factors
in mind, however, and using the lowest
number for carbon adsorption, current
projections indicate that SEXOP still
maintains an attractive economic posi-
tion.


ACKNOWLEDGMENTS

    The author wishes to acknowledge
the constructive guidance provided by
Messers. Robert Hendriks and David
Sanchez of the EPA Industrial Environ-
mental Research Laboratory at Research
Triangle Park, NC.  Of special value
was the comprehensive engineering ex-
pertise provided by Mr. George Hiler
of S-CUBED throughout the course of
the bench-scale process fabrication
activity.  Technical guidance and pro-
ject review provided Dr. E. A. Burns
of S-CUBED throughout the course of
the research effort is also greatly
appreciated.


REFERENCE

(1)  Reynolds, S.L., June 1983, Ex-
     traction of Pesticides from Pro-
     cess Streams Using High Volatility
     Solvents - A Feasibility Study,
     EPA600/2-83-0404, Industrial En-
     vironmental Research Laboratory,
     Research Triangle Park, NC, 144 p.
     (NTIS PB83-209767).
                 Di sclaimer

    The  work described in this paper was
    not  funded  by the U.S. Environmental
    Protection  Agency.  The contents do
    not  necessarily reflect the views of
    the  Agency  and no official endorse-
    ment should be inferred.
                                        458
-------
                BIOLOGICAL REMOVAL OF MERCURY FROM TOXIC WASTE
           Jeffrey W. Williams, Gorily L. Hansen and Anish Jantrania
                           The  Ohio  State University
                             Columbus,  Ohio   43210
                                   ABSTRACT

    Mercury resistant bacteria have been applied to the biological treatment
of mercury-contaminated waste.  The approach was to use mercury resistant
bacteria to reduce Hg   to Hg .  After preliminary batch culture studies, a
pure strain of mercury resistant Esoherichia coli was introduced into a
bioreactor equipped with a pump for adding HgClp in-line.  Using nonsterile,
urban sewage (average BOD_ 135 mg/£) in a continuous operation, it was found
that the bacteria could remove mercuric ion for periods of over 2 weeks at the
rate of 2.5 mg/£-h.  If the effluent was allowed to set at room temperature
for an additional 24 h after exiting the bioreactor, 98/5 of the added mercury
(70 mg/£) was removed.  Examination of the microbial population 24 h after the
initiation of the process indicated that the original strain of E. coli had
been replaced by other species of bacteria.  A random selection of the
predominant organisms were analyzed for antibiotic and mercury resistance.
There appeared to be some organisms which were Hg-resistant but did not
volatilize mercury.  The biological process was applied to an actual
industrial sample from the chlor-alkali industry.  Making only slight
modification to our biological process, we were able to completely detoxify
this material using the mercury resistant bacteria.  These results indicate
the feasibility of the biological approach to the control of industrial
mercury pollution.
INTRODUCTION

    Toxic and potentially hazardous
heavy metals are found in wastewater
from many industrial sources.  Among
these compounds, mercury is one of
the most toxic and best understood
examples of heavy metal pollution.
From years of study, it is now
apparent that the toxicity of mercury
is dependent upon the chemical form
in which it occurs.  In general, the
organic mercury compounds are much
more toxic than the inorganics.  Of
the organomercurials, methylmercury
is perhaps the most problematic since
it is naturally formed in the aquatic
and terrestrial environments from
elemental and mercuric mercury.  Most
important from the standpoint of this
report, it has been shown that the
environmental local load of
methylmercury may be considerably
increased by industrial release of
inorganic mercury (1).  Thus, mercury
represents an example of a pollutant
which is converted into a more toxic
substance when released into the
environment.  Hence, we believe that
serious consideration should be given
to the concept of complete removal
and recovery for reuse, rather than
"secure" burial, when dealing with
compounds such as mercury.

    Mercury is currently removed from
industrial wastewater by chemical
precipitation (2-6).  A widely used
                                      459
-------
 method involves  the  addition of
 sodium sulflde into  mercury laden
 water.  The  insoluble mercuric
 sulfide which  forms  is  then removed
 as  a sludge.   Another method which
 has been employed  uses  sodium borohy-
 dride to reduce  the  mercury to its
 metallic form  which  is  then filtered
 from the effluent.

     Microbiological  methods for the
 extraction and recovery of metals
 have previously  been proposed (7-9).
 Reported methods most often involve
 uptake or binding  to the micro-
 organisms.  Unfortunately, these
 approaches suffer  from  the same
 disposal problem mentioned above.
 Alternatively, we  propose the use of
 mercury resistant  bacteria which are
 capable of converting Hg   to the
 volatile elemental form (10-12).  The
 volatile, metallic mercury thus
 produced could then  be  vented for
 recovery.  Since the bacteria used
 for the reduction of mercury are
 aerobic,  the removal of volatile
 mercury is facilitated  by the
 aeration required to maintain
 bacterial growth.

 APPROACH

 Bacterial Strain and Growth Medium

     The mercury  resistant organism
 used in these  experiments was
 Esoherichia coli KP245  which harbors
 the mercury resistance  plasmid
 pRH30.  Although every experiment
 was  initiated with this organism,
 prolonged growth under  the nonsterile
 conditions used  in our  experiments
 normally resulted in the development
 of a mercury resistant mixed-culture.

 The  bacteria were grown and
maintained on media containing yeast
 extract  (0.5 g/£), sucrose (6 g/£)
 and HgCl- (5 mg/£).  The sewage used
was  collected from a residential
 subdivision of about 300 homes with
 no industry or other commercial
 enterprise and was screened before
 use.

 Experimental Apparatus

    The bioreactor was a New
 Brunswick Scientific Mieroferm MF
 115,  14£  feraentor equipped with
 variable mixing and aeration, as well
 as automatic temperature and liquid
 level control.  The two paddle mixer
 was operated at ca. 120 rpm, while
 the liquid temperature was maintained
 at 37 C.  The aeration rate was 2
 £/min into a 5£  working volume which
 kept dissolved oxygen at over 8 ppm
 for all cultivations reported.  The
 pH remained within an acceptable
 range without adjustment.

 Mercury Analysis

    Mercuric ion concentrations were
 routinely monitored
 spectrophotometrically by the
 dithlzone assay as described
 previously (13).  The assay was
 determined to be sensitive enough for
 the higher concentrations of mercury
 used in this study.  Some samples
 were submitted for wet ashing (1:1
 mixture of HNO^/HCIO.) and analysis
 by inductively coupled plasma (ICP)
 emission spectrometry at the Ohio
 Agricultural Research and Development
 Center (Wooster, Ohio).

 RESULTS

    Our initial studies examined the
growth and mercury removal
 characteristics of shake-flask
 cultures of a strain of E« coli
 (KP24S) selected for its high
tolerance to mercuric ion.  It was
 shown (Fig. 1)  that increasing the
 concentrations of mercuric ion up to
 10 mg/£ increased the lag phase, but
had no apparent effect on the growth
rate.  In other studies, the rate at
which the mercuric ion was removed
from the culture flasks was also
 examined.  In these experiments, an
overnight culture (grown on 10 mg/£
                                     460
-------
Hg   ) was harvested, washed twice,
and  resuspended in fresh growth
medium in the presence of  10 mg/£
Hg   .  Samples were removed at   ~
various times and assayed for Hg
The  results from these experiments
(Fig. 2) demonstrated that
immediately following inoculation the
mercuric ion concentration decreased
very rapidly, during which time  there
was  no observed growth.  However,
after most of the mercury was removed
(within 1.5 hours), normal growth
resumed.

     In order for this process to be
applicable to industrial situations,
it was necessary to determine whether
continuous mercury removal under non-
sterile conditions was feasible.
These studies were performed with the
bench scale bioreactor described
above.  Each experiment was initiated
by adding 5£  of raw sewage, (average
BOD5 = 135 mg/£J to the bioreactor
along with 100 m£ of the starter
culture.  Refrigerated sewage was
then pumped into the bioreactor on a
continuous basis at the rate of 5£/24
hours and a concentrated solution of
      was mixed in-line with the
sewage at an adjustable rate which
allowed us to control the
concentration of Hg   in the
influent.  Under these conditions, we
found that 70 mg/£ Hg + in the
influent was a safe, upper limit for
extended cultivations although the
bacteria were able to reduce as much
as 100 mg/£ over 12-14 hptime
periods.  However, if Hg   was
maintained continuously at this level
the biological reduction process
would eventually slow or collapse
entirely.

    Table I shows the results of .one
of three replications of two week,
continuous cultivations using sewage
aspthe growth medium and maintaining
Hg   at 70 mg/| in the influent.
Results for Hg   removal from the
other replications were similar.
Samples were removed daily and
           60 120 180 240 300
              Time (Win)
Figure 1.  Bacterial growth at~
different concentrations of Hg   .
After a 2% overnight inoculum, growth
curves were determined at Hg
concentrations of:  (o) 0 mg/£, (A) 2
mg/£, (») 4 mg/£, and (n) 10 mg/£.
With permission from Ref. 14.
         0   60   120  180  240
              Time (Uin)

Figure 2.  Reduction of mercury
concentration and bacterial growth.
(•) Hg   concentration, (A) bacterial
growth.  With permission from Ref.
14.

analyzed for mercury content.  The
data indicated that the total mercury
                                     461
-------
concentration decreased to a low
level in unsettled effluent held for
an additional 24 hours after leaving
the bioreactor  (Table 1).  It can
also be seen that during this entire
period the average free Hg
concentration in the effluent was
below the sensitivity of the
dithizone assay (< 1.5 mg/ ).  The
ashing and inductively coupled plasma
(ICP) analysis indicated a low
concentration (1.2 mg/£) of total
mercury in the effluent.

    From these results, it was
calculated that the bioreactor could
remove Hg   from the influent sewage
at the rate of 2.5 mg/£-h with an
efficiency of 88%.  By allowing the
effluent to set at room temperature
for an additional 24 hours after
exiting the bioreactor (there was no
significant settling during this
time) we could remove 98% of the
added mercury.

    As another aspect of this study,
we examined the bacterial population
of the bioreactor 24 h after
inoculation.  The original E. coli
strain was not detected at this time.
If the mercury removal process was
initiated in the absence of the
starter culture of E.coli, there was
no significant removal of mercury
(data not shown).  These results
indicated that under nonsterile
conditions the continuous removal of
mercury was accomplished by a mixed-
population of mercury resistant
bacteria.  Since the starter culture
was required to initiate the process,
the mixed population arose naturally
at the reduced mercury levels
maintained in the bioreactor, or
else, genetic transfer from the
starter culture must have occurred.

    In order to better characterize
the mercury-reducing, mixed-culture,
we isolated 13 colonies at random
from a 10  dilution on agar plates
containing 10  M HgCl_.  The isolated
strains and their antibiotic
phenotype are given in Table II.
Note that none of the strains share a
similar antibiotic resistance profile
with the original E. coli KP245
strain.  This provides indirect
evidence that intact plasmid transfer
from E. coli is not important.  The
isolates were all resistant to
mercury (Table III).  However, the
level of resistance varied
considerably.  Mercury resistance was
also measured as a function of the
chemical oxygen demand (COD) of the
media.  For this study, the COD was
due to varying concentrations of
glucose in the medium and, thus, it
was used to reflect the amount of
oxidizable material.  As shown in
Table III, most of the isolated
strains had decreased mercury
resistance levels at decreased COD
values.  However, for strains 9, 10
and 11, there was no association
between COD and mercury resistance.
Since mercuric reduction is an energy
requiring process, decreased levels
of resistance would be expected in
media of low COD.  Therefore,, strains
9, 10 and 11 are mercury resistant,
but probably do not reduce Hg
Mercuric reductase was not detectable
in these strains.  Therefore, it
appears that in the bioreactor there
exists within the mixed-culture a
mercury-resistant, nonreducing
subpopulation.  Further studies are
required to determine the effect of
this subpopulation upon mercury
volatilization, especially at lower
influent mercury concentrations.

    More recently, we have applied
our biological process to an actual
industrial waste sample from a chlor-
alkali plant in Northern Ohio.  This
sample of a filtrate "mud" from the
electrolysis cells was mainly
composed of potassium salts, •
insoluble metal hydroxide and 100-150
ppm of mercury in the form of mercury
hydroxide.  Mercury was successfully
removed from this sludge using our
process in both batch and continuous
modes.
                                     462
-------
TABLE I.  CONTINUOUS BIOLOGICAL REMOVAL OF MERCURY FROM SEWAGE OVER A TWO-WEEK
          PERIOD
Influent Hg2+,
Concentration11
Average bioreactor Hg
Concentration (24 h
detention time)

Free Hg2+      Total Hgc
                 Average effluent H
                 Concentration (additional
                 24 h detention time)
                 Free Hg*
                                                                  Total Hg
     70
  < 1.5
9.6 + 2.7
< 1.5
1.2 + 1.0
.All Hg concentrations are in mg/£.
 Free Hg   concentration was determined by the dithizone assay.
 Total Hg concentration was determined by wet ashing and ICP  analysis.
    In the batch process, a measured
quantity of mud was diluted in water
and the pH adjusted to 7 by adding
concentrated sulfuric acid.  This
solution was then added to the
bioreactor containing a mixed-culture
of mercury resistant bacteria in
nutrient broth.  After addition of
the diluted sample, the initial
mercury concentration in the
bioreactor was 3.5 mg/£.  As shown in
Table IV, the mercury concentration
in the bioreactor decreased 89$ over
a 12 h period.

    In the continuous process, the
diluted waste sample and the nutrient
broth solution were continuously
pumped through two separate influent
                      lines  to the  bioreactor.   The
                      reservoir of  diluted mud  (6.5 mg/£
                      Hg  )  was mixed continuously to keep
                      undissolved solids  in  suspension.
                      The  influent  flow rate of the diluted
                      mud  into the  bioreactor was 0.18 £/h.
                      The  nutrient  broth  was added from a
                      separate line at the rate of 0.34
                      £/h.  In 3 separate experiments, the
                      bioreactor was allowed to reach
                      steady-state  after  which  data were
                      collected over a 24 h  period.  In
                      each of  these experiments, the
                      mercury  concentration  in  the
                      bioreactor was maintained at 0.25
                      mg/£,  indicating that  mercury was
                      continually removed at an average of
                      0.22 mg/£-h.
                                     463
-------
         TABLE II.   ANTIBIOTIC RESISTANCE  PROFILES  OF  RANDOM
                   ISOLATES FROM THE BIOREACTOR

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
11.
Strain
Pseudomonas
diminuta
Staphylococcus
hgemolyticus
Pseudomonas
diminuta
Staphyloooecus
epidennidis
Acinetobacter
calcoaceticus
Acinetobacter
calcoacetious
Aeromonas
hydrophila
Pseudomonas
.algal igenes
Micrococcus
luteus
Staphyloooocus
hominis
S taphyl oco c cus
haemolyticus
Alcal igenes
odorans
StaphylocoGcus
epidermidis
Escherichia
coli, KP245°

AM3 AN B C E K NA S T
Rb S SSRRB RS
R S SSSRS RS
S R RSRRS RS
S S SSSSS SR
S S RSRSS SS
S S SRRSS SS
R S RSSSS SS
R S RSRSR SS
S S SSSSS SS
S S SSSSS SS
S S SSSRS RS
S S RSRSS RS
S S SSSSS SS
R S SSSSS SS
 Antibiotics used were:  ampicillin (AM),  amikacin (AN),  bacltracin (B),
 chloramphenicol (C), erythromycin (E),  kanamycin (K),  nalidixic acid
, (NA), streptomycin (S),  and tetraeyeline  (T).
 B, resistant; S, sensitive.
 Used for the initial inoculation, but  not recovered from the
 bioreactor.
                               464
-------
       TABLE III.  MINIMAL INHIBITORY CONCENTRATION (MIC) OF MERCURY AS
                   A FUNCTION OF CODa
Strain
1
2
3
4
5
6
7
8
9
10
11
12
13
14
COD
19 600
16
4
12
8 ' •
40
40
40
40
80+
80+
80+
40
40
40
MIC (mg/£ Hg2+) Values
COD COD COD
9 800 4 900 1 200
-
_
30
30
30
30
70
70
70
-
-
30
-
_
4
20
8
20
70
70
70
_
_
20
-
-
4
4
4
16 •
70
70
70
-
-
4
COD
300
mm
• -
4
2
4
16
70
70
70
_
-
4
COD
150
mm.
_
4
2
2
16
70
70
70
-
-
4
f*COD as mg/  0_.
 -, No growth aetected.
                                   465
-------
             TABLE IV
   MERCURY REMOVAL IN A BATCH PROCESS
Time
(h)

0
3
6
9
12
Influent
Mercury
(mg/£)
3.5
2.0
1.5
0.9
0.4
Mercury
Removal
(%)
0
42
56
75
89
ACKNOWLEDGEMENTS

    The technical assistance of John
Busic is gratefully acknowledged.  We
would like to thank Rose Smith for
her assistance in manuscript prepara-
tion.  This work was supported in
part by grants from NOAA and NIH.
REFERENCES

 1.  Swedish Expert Group, 1971.
     Hord.Hyg. Tdskr. Suppl.  4.

 2.  Bhattachargya, D., Jemawon, A.
     B., Jr. and R. B. Grieves, 1979.
     Sep. Sci. Technol. t4_, 441.

 3.  Habashi, F., 1978.  Environ.
     Sol. Teohnol. 12, 1372.

 4.  Findlay, D. M. and R. A. N.
     McLean, 1981.  Environ. Soi.
     Teohnol. 15, 1388.

 5.  Lander, J. and M. M. Cook, 1981.
     Pollut. Engin.,  Dec., pp 36-38.

 6.   Perry, R.,  1974. U.S. EPA Report
     660/2-74-086.

 7.   Dunn, G. M. and A. T. Bull,
     1983.  Bur. Jf. Appl. Mierobiol.
     Biotechnol. 17,  30.

 8.   Kelly, D.  P.,  Norris, P. R. and
     C. L.  Brierley, 1979. In
     Microbial Technology;  Current
     State, Future Prospects, A. T.
     Bull, D. C.  Ellwood, and C.
     Ratledge,  Eds.  Cambridge
     University Press, Cambridge pp.
     263-308.

 9.   Nelson, P.  0., Chung, A. K. and
     M. C.  Hudson, 1981.  J. Water
     Pollut. Control  Fed. 53, 1323.

10.   Summers, A. 0. and S. Silver,
     1972.  jj.  Bacteriol. 112, 1228.

11.   Summers, A. 0. and S. Silver,
     1978.  Annu. Rev. Mierobiol. 32,
     637.

12.   Williams,  J. W.  and S. Silver,
     1984.  .Enz. Microb.  Teohnol. 6,
     530.

13.   Johnson, W. C.,  Gage, J. C.,
     Gorsuch, T. T.,  Johnson E. I.,
     Johnson, E. M.,  Milton R. F.,
     Newman, E.  J., Sharpies, W. G.,
     Thackray,  G. B., Tyler, J. F. C,
     and P. W.  Shallis, 1965.
     Analyst 90, 515.

14.   Hansen, C.  L., Zwolinski, G.,
     Martin, D.   and  J. W. Williams,
     1984.  Biotechnol. Bioengin. 26„
     1330.
                                      Disclaimer

    The work described In this paper was not funded by the U.S. Environmental
    Protection Agency.  The contents do not necessarily reflect the views of the
    Agency and no official endorsement should be inferred.
                                      466
-------
       COMBINED POWDERED  CARBON/BIOLOGICAL  ("PACT") TREATMENT TO DESTROY
                       ORGANICS IN INDUSTRIAL WASTEWATER

  (Case History of Du Pont Chambers Works 40 MGD Wastewater Treatment Plant)

                             Harry W. Heath, Jr.
                   E. I.  du Pont de Nemours & Company,  Inc.
                                Chambers Works
                             Deepwater,  NJ  08023


                                   ABSTRACT

    This paper discusses  the 40 million  gallon per day  industrial  wastewater
treatment plant (WWTP) at Du Pont's Chambers Works site at Deepwater, NJ.
This very large WWTP uses a Du  Pont developed advanced  technology  to  treat
highly colored, acidic wastewater containing a wide variety of  organic
compounds, including priority pollutants.  Many of these compounds are not
susceptible to conventional biological treatment.  The  advanced technology is
the patented Powdered Activated Carbon Treatment (PACT)CD process, which is
a combined secondary/tertiary treatment.  It is accomplished by adding powdered
activated carbon directly to an aerator, and then operating a combined biolog-
ical oxidation/activated  carbon adsorption process.

    Some noteworthy features of the Chambers Works WWTP are reviewed, includ-
ing a unique, 40 tons/day regeneration furnace for powdered activated carbon,
and a "state of the art"  double-lined secure landfill where primary sludge
from the WWTP is deposited.  Some performance advantages of the "PACT" process
over conventional activated sludge treatment will be  shown. Data  on  removal
of priority pollutants is included.  Some of the reasons that have led to the
development of'a rapidly  expanding Du Pont Environmental Services  business
based on available capacity at  the WWTP  will also be  discussed.


INTRODUCTION AND PURPOSE

    This is a case study  of a very        which was developed for  this purpose,
large, successful industrial waste-       did provide combined  secondary/-
water treatment plant which had to        tertiary treatment in a  facility corn-
use innovative technology to meet         parable to  a  conventional activated
required effluent quality at a             sludge plant  in terms of investment
reasonable cost.  The "PACT" process,     and operating cost.   The successful
(l)t«PACT" technology is now owned by Zimpro,  Inc.,  who are making it  available
    under license.  "PACT System" is a registered trademark of Zimpro,  Inc.
                                      467
-------
treatment of Chambers Works highly
colored, variable waste containing
significant amounts of normally
difficult to biodegrade organics by
this technology may indicate
advantages from its use in other
treatment situations.

    It is also a purpose of the paper
to point out some unique advantages
of a large, central, advanced waste-
water treatment facility in treating
a variety of outside wastes which
may be unsuitable for smaller, local,
less sophisticated treatment plants.
DISCUSSION AND RESULTS

1)  Treatment Process

    The waste stream which the WWTP
was designed to treat in 1975 had
the following average charac-
teristics :
Flow

Soluble BOOj

Color

Acidity

Dissolved Organic Cartoon - 65,000 Ibs/day (205 «fl/l)

.Total Dissolved Sollrte  - 2,000 to 5,000 »g/l

Total Suspended Solid*  - 40 tons/day1"* (258 «o/l)
                  - 25,400 gp«

                  - 88,700 Ita/ctay (280 ng/1)

                  - 1,000 APHft

                  - 227 tor*/*** <
 *Acidity concentrations or weights
  are expressed as  CaCOj equiva-
  lents throughout  this paper.

**Jncluded 28 tons/day by-product
  solids from lime  neutralization.
This waste stream represented the
combined discharge from Chambers
Works plus the adjacent Carney's
Point site, which has since closed.
While this was only a medium strength
waste in terms of BOD concentration,
the organic component of the waste
stream was particularly difficult to
treat.  It contained a variety of
aromatic compounds, and the composi-
tion of the organics was continually
changing because of the batch nature
of many of the Chambers Works
operations, particularly in the dyes
area.

    Du Pont constructed a $45 MM
WWTP to treat this waste.  Current
replacement value is over $100 MM.
Figure 1 is a diagram of the
facility.  The WWTP operates 24
hrs/day and 7 days/week.  It is
staffed with 26 chemical process
operators and an approximately equal
number of clerical, maintenance,
electrical, and instrument
personnel.  There are approximately
15 supervisory and technical people
assigned to the WWTP.

    a)  Primary Treatment

        A three stage treatment
process was specified.  In primary
treatment the acidic water is
neutralized with lime in a single
stage neutralization using three,
stirred 200,000 gallon reactors
operated in parallel.  The WWTP
receives powdered lime by rail car
or truck.  Powdered lime is stored
in 4 large silos and is slaked to
8-10X concentration as needed.  Lime
slurry is fed to the neutralizers by
an automatic control system that
maintains pH at any preset level.
The WWTP currently consumes approx-
imately 60 tons/day of dry lime to
treat 100 tons/day acid.  The plant
has enough capacity to treat over
200 tons/day acid.
                                       468
-------
cn
to
    CD
    c
    ns
                                     RR TRACK
                                                         CARBON
u.

! M II 1 I
| f | | ) UNLOAOIN
RR TRACK
1 1 1 1 II 1 1 1
nr
1
J /T
                                                                 KEY: AB
                                                                     A3
                                                                      C
                                                                    CAW
                                                                      F
                                                                     Ff
                                                                     FS
                                                                     FT
                                                                     HP
                                                                      L
                                                                    LST
                                                                     ME
                                                                      P
                                                                    MAS
                                                                    I°ST
                                                  AFTERBURNER
                                                  ACID STORAGE
                                                  CARBON  SLURRY TANK
                                                  CARBON  ACID WASH TANK
                                                  FLOCCULATOR
                                                  FiureR PRESS
                                                  FLOW  SPLITTER
                                                  FUEL STORAGE TANK
                                                  200 PSIO FILTER FEED PUMP
                                                  LIME  STORAGE  SILO
                                                  LIME  SLURRY TANK
                                                  NEUTRALIZATION TANK
                                                  8,000 6PM WASTE WATER FEED PUMP
                                                  RECYCLE ACTIVATED SLUDOE  PUMP
                                                  PRIMARY SLUDOE HOLD TANK
                                                                                                                    TREATED
                                                                                                                    EFFLUENT
                                                                                                                    TO RIVER
                                                             PRIMARY SLUDGE
                                                             TRUCKED TO
                                                             LANDFILL
                                                                                             TO ATMOSPHERE
                                                          CARBON
                                                          REOENER.
                                                          FURNACE
                                                        • CONTROL!
                                                        I ROOM  I
       WASTEWATER
          DITCH
                                                  CARBON
                                                  SLURRY
                                                                   SLUDGE
                                                                   THICK.
                                                                                                             CLARIFIES
                                                                                                              2,500,000
                                                                                                                GAL.
                                                              =» I
                                                            AERATION
                                                              TANK
                                                            4,000,000
                                                              GAL.
4M  CLARIFIER
1.000,000 OAL.
                                          «»Z CLARIFIER
                                          1,000,000 OAL.
                  *2N
                  200,00
                  GAL.
                                                                                             ** 2
                                                                                           CLARIFIER
                                                                                            2,500,000
                                                                                              GAL.
                                          *3 CLARIFIER
                                          1,000,000 GAL.
                                                    *» 3
                                                 AERATION
                                                    TANK
                                                  4,000,000
                                                    OAL.
                                                                         AERATION
                                                                           TANK
                                                                         4,000,000
                                                                            GAL.
 »3N£
200,000
 GAL.
4*4 CLARIFIED
1,000,000  OAL
-------
        The neutralizers overflow to
four primary clarifiers, each one
million gallons in size.  The
rectangular clarifiers operate in
parallel and are 230 by 55 by 12
feet deep.  Solids in the wastewater
feed, as well as by-product solids
formed during neutralization, settle
and are removed as an 6-10% slurry
in the clarifier underflow.  After
filtration to a 45 to 50K solids
cake, the solids are hauled to the
Secure Landfill.  Two high-pressure,
(210 psig at end of cycle) large
recessed chamber filter presses make
approximately 7 tons of wet press
cake per 30 to 60 minute cycle.

        The WWTP currently generates
from 50 to 70 tons/day (dry basis)
solids, equivalent to 50,000 yd3/-
year of landfill volume.  The solids
are mainly inorganic, primarily
calcium and magnesium salts, as well
as silica compounds from river water
silt.  These solids also contain
small amounts of heavy metals and a
variety of organic compounds.  The
solids are an inevitable by-product
of the primary treatment process,
and because they are primarily
inorganic compounds, the only
feasible disposal method is land-
filling.

    b)  Secure Landfill:

        Figure 2 is a diagram of the
landfill.  A double liner of a
chlorosulfonated polyethylene
material, "Hypalon", covers the
entire bottom and part of the sides
of the landfill.  Collection pipes
between the liners serve as leak
detectors and drain to sumps outside
the landfill.  The sumps are sampled
on a regular basis for any signs of
contamination.  Above the top liner
is a similar collection system for
leachate which is pumped back to the
WWTP.  The landfill will ultimately
become a 70 feet high pyramid with a
15 acre base.
        As outer sections of the
landfill are filled, the edges and
top are covered with a 2 feet layer
of essentially impervious, permea-
bility of less than 1 x 10~7
cm/sec, clay and 12 inches of top
soil.

        A controlling factor in the
operation of the primary filter
presses is that the filter cake must
meet rigid soil stability criteria.
These solids have to support the
weight of heavy earth-moving equip-
ment.  Further, the allowable slope
of the landfill sides, and hence the
volumetric capacity of the landfill,
is closely regulated to assure a
large safety factor against any
possible slippage along the slopes.

        As backup protection, there
are 26 monitor wells located around
the landfill; and in the very unlike-
ly event of a double liner failure,
interceptor wells would be utilized
to prevent any harm to the environ-
ment while the problem was corrected.

    c)  "PACT" Secondary/Tertiary
        Treatment&Carbon
        Regeneration

        The neutralized wastewater
that overflows the primary clari-
fiers flows into three, four million
gallon aerators operated in parallel.
These are 185 feet diameter, 20 feet
deep vessels.  Here over 90% of the
BOD is removed by a slurry of sus-
pended solids; the solids consisting
of an approximate 50/50 mix of
bio-mass and activated carbon.
Approximately two million pounds of
solids are under aeration.  Air for
the biological oxidation is supplied
from large blowers that normally
maintain a dissolved oxygen concen-
tration of 1 to 7 ppm in the water
by forcing 5,000 to 15,000 standard
cubic feet per minute air through
                                      470
-------
      BOTTOM
TOP UNER

           I UNER
GRAVEL
'
\
• c
).
1
UIACHATESUMP

                                                          
-------
diffuser pipes in the bottom of each
aerator.  The electric bill to run
the blowers for just one aerator is
over $150M/year.  The water is
decolorized, and because of the
presence of the carbon, organic
compounds that normally are not
susceptible to biological oxidation
by bacteria are removed.

        The aerators overflow to two
2.5 MM gallon parallel clarifiers.
These are circular, 172 feet diameter
by 15 feet deep tanks.  The bio-mass/
carbon solids settle and are pumped
back to the aerators.  This recycle
of bio-mass is the key to the acti-
vated sludge process, as it allows
the WWTP to maintain an average of
five to forty days of residence time
for solids in the aerators.  This
gives the bacteria time to acclimate
to the organics in the wastewater
and then consume them.  By compar-
ison, the hydraulic residence time
of the wastewater in the aerators is
in the range of five to nine hours.

        Up to 3 million pounds per
year of virgin activated carbon are
needed for the "PACT" process at
Chambers Works.  It is delivered as
a dry powder by rail car and stored
as a one Ib/gallon aqueous slurry
before use.  The carbon is added to
the wastewater just before it enters
the aerators.

        As the bacteria consume over
40,000 pounds/day of dissolved
organics they multiply and generate
approximately 6,000 pounds/day of
new bio-mass.  Combined with the
continuous feed of powdered carbon,
this imposes a net solids increase
of over 12,000 pounds/day into the
aerators.  There must be an equiva-
lent purge of solids from the
system, and this is provided by
removing a side stream from the
concentrated solids underflow from
the secondary clarifiers.
        The WWTP has the capability
of regenerating powdered activated
carbon from these wasted secondary
solids.  This is another unique
feature of the WWTP, as the powdered
activated carbon is regenerated in a
multiple hearth furnace.  To our
knowledge this is the only commer-
cial use of a multiple hearth furnace
to regenerate powdered carbon.

        When regenerating carbon,
wasted "PACT" sludge is filtered
prior to being fed to the furnace.
The same type filter press is used
for "PACT" sludge as is used for
primary sludge.  The design solids
concentration of the bio-mass/carbon
filter cake was 35%, but in actual
practice higher solids concentra-
tion, 45%, was achieved with the
"PACT" sludge.  Without the presence
of carbon as an internal filter aid,
it would be difficult to get over
25% solids concentration.

        After filtration to a 45%
solids cake, the bio-mass/carbon
solids are conveyed to the top of  a
40 feet high by 26 feet diameter,
five hearth furnace.  Under con-
trolled conditions, the solids are
fed down through the furnace, being
raked across each hearth by rabble
arms mounted on a large, rotating
center shaft.  Water is first evapo-
rated and then bio-mass and adsorbed
organics are burned.  Finally at the
bottom of the furnace at approximate-
ly 1,000°C, and in the presence of
steam, the powdered carbon is regen-
erated.  The incandescent carbon is
quenched in water, washed with acid
to remove inorganics, and recycled
to the aerators.  Heat to the furnace
is supplied from an external, oil
fired burner.  The heated gases rise
through the furnace countercurrent
to the solids.  Gas flow must be
controlled to avoid blowing powdered
carbon out of the top of the furnace.
                                      472
-------
        To control air pollution,
the furnace off gas is scrubbed with
water and then passed through an
afterburner at about 700°C.

        During the initial years of
operation there were significant
processing problems with the furnace
itself, as well as with the "PACT"
sludge filtration and conveying
systems.  Once modifications to
equipment and process control
instrumentation were completed, the
furnace generated 13 T/0 of
activated carbon with an average
iodine number of 400* over a two
year period of operation.  This
represented a quality yield of about
80%.  During this period the furnace
was fed 50 T/D of wet sludge cake,
and  destroyed over 10 T/D bio-mass,
adsorbed organics, and unrecovered
carbon.

        The furnace is currently
out-of-service.  A combination of
reduced organic loads from the
Chambers Works (hence, lower carbon
dosages are needed to assure adequate
organic removal) plus currently very
favorable prices for virgin activat-
ed carbon make it more economical to
use only virgin carbon on a once
through basis.  Currently, wasted
secondary carbon/bio-mass sludge is
recycled to the inlet wastewater
feed stream.  The wasted secondary
solids then become part of the
primary sludge, are removed in the
primary clarification step,  and are
ultimately deposited in the Secure
Landfill.
2)  Performance of the "PACT" System

    The "PACT" process was developed
in 1970 from technical studies to
develop a treatment process for
Chambers Works wastewater.  As com-
pared with a conventional activated
sludge process for treating Chambers
Works wastewater, the "PACT" process:

    o Gave consistent BOD removals
      of over 94%

    * Raised DOC removal from 62%
      to 85%

    • Reduced color

    • Reduced foaming in the aerators

    * Improved the settling and
      filtration properties of the
      secondary solids

    • Protected the bacteria from
      shock loads of organic
      compounds

Typical comparative data from contin-
uous laboratory test units are shown
in Table I.  As would be expected,
performance improved with increasing
carbon dose up to a certain level.

    The combined second/tertiary
treatment with "PACT" resulted in
large investment and operating cost
savings as compared with a conven-
tional tertiary treatment process
using biological oxidation followed
or preceded by adsorption in granular
carbon columns.  In effect, the
"PACT" process gave tertiary treat-
ment quality effluent in a single
carbon adsorption/bio-oxidation
aeration process.
*An industry measure of adsorptive
 capacity for certain types of
 molecules.  The higher the number
 the greater the adsorptive capacity.
                                       473
-------
                   TABLE I

   TABLE I.  LABORATORY DATA COMPARING "PACT" VS.
                ACTIVATED SLUDGE
Soluble BOD5f mg/1
  BCD Removal

DOC, n>g/l
  DOC Removal

Color, APHA
  Color Removal
                   Feed
111
820
           Effluent
                                 "PACT*' "PACT*
                                 25 PPM 100 PPM
                        Activated Carbon Carbon
                         Sludge    Dose   Dose
 7.5
93.2X

31.9
 63*

 700
 20%
 6.6    6.1
94. IX  94.5X
21.7
 7SX

 320
 63%
13.1
 85%

 170
 80%
Data represents three months operation of continuous,.
l«boratory 7.5 liter units.  Temperature range for
•11 units was 16 to 25% and hydraulic residence
time was 8 hours.
                           All the design goals for the
                       "PACT" process  were achieved in
                       actual plant operation.  Table II
                       shows actual WWTP data during the
                       early years of  operation when the
                       waste loads were higher. "

                            TABLE II.  "PACT SYSTEM" PERFORMANCE

                                   1978 Through 1980
                          Flow
Influent Soluble BOD5
Effluent Soluble B005
X Removal
                                               Influent DOC
                                               Effluent DOC
                                               X Removal
                           Influent Color
                           Effluent Color
                           X Removal
                                   24,800 GPM
175 mg/1 (52,100 Ibs/day)
  7 mg/1
9SK
                                                173 mg/1  (51,500 Iba/day)
                                                 32 mg/1
                                                82X
                                           1430 APHA
                                            475 APHA
                                           67X
                                               Aerator MLSS (sludge age) 25,000 mg/1 (41 Days)
       With the  "PACT" process, a  single
   piece of equipment, the aerator,  is
   used to destroy  most of the organic
   waste by a relatively inexpensive
   biological oxidation treatment.   Much
   more expensive carbon adsorption  is
   only used to  remove organic compounds
   which are difficult to biodegrade or
   completely non-biodegradable.   The
   organic removals achieved in the
   "PACT" process are greater than the
   sum of biological oxidation and
   carbon adsorption removals would
   predict.  This was demonstrated by
   dosing the effluent from a continuous
   laboratory activated sludge unit  with
   carbon.  The  effluent from this batch
   adsorption treatment had a higher
   concentration of DOC than the
   effluent from a  parallel continuous
   "PACT" reactor.   There is a synergism
   in the interaction of carbon and
   bacteria such that the bacteria are
   significantly more effective in the
   presence of carbon than by themselves.
                           Carbon Dose
                            X Virgin
                            X Regenerated
                                           120 mg/1
                                           53*
                                           47%
                                Data on average percent removals
                            alone are not the most valid measure
                            of treatment effectiveness.  Treat-
                            ment plants must meet daily maximum
                            as well as monthly average limits,
                            and consistently meeting these daily
                            maxima can be as demanding of a
                            treatment process as meeting a
                            monthly average.  The treatment
                            consistency of  the "PACT" process
                            has been excellent.  Figures 3 and 4
                            are histograms  showing the distribu-
                            tion of feed and effluent concen-
                            tration for soluble 8005 and DOC
                            over a typical  operating period for
                            the Chambers Works WWTP.  There is a
                            particularly narrow distribution for
                            effluent BOD concentration.
                                           474
-------
                                                         1 TO* SOU.

                                                         J-CLAY  .
TOPUNEH
       BOTTOM
UNER
GRAVEL
7
\
• c
1-
}

UEACMATE SUMP
                                                               LEAK
                                                               DETECTION
                                                               CHAMBER
                         Figure  2,   Secure Landfill
                                       471
-------
diffuser pipes in the bottom of each
aerator.  The electric bill to run
the blowers for just one aerator is
over $150M/year.  The water is
decolorized, and because of the
presence of the carbon, organic
compounds that normally are not
susceptible to biological oxidation
by bacteria are removed.

        The aerators overflow to two
2.5 MM gallon parallel clarifiers.
These are circular, 172 feet diameter
by 15 feet deep tanks.  The bio-mass/
carbon solids settle and are pumped
back to the aerators.  This recycle
of bio-mass is the key to the acti-
vated sludge process, as it allows
the WWTP to maintain an average of
five to forty days of residence time
for solids in the aerators.  This
gives the bacteria time to acclimate
to the organics in the wastewater
and then consume them.  By compar-
ison, the hydraulic residence time
of the wastewater in the aerators is
in the range of five to nine hours.

        Up to 3 million pounds per
year of virgin activated carbon are
needed for the "PACT" process at
Chambers Works.  It is delivered as
a dry powder by rail car and stored
as a one Ib/gallon aqueous slurry
before use.  The carbon is added to
the wastewater just before it enters
the aerators.

        As the bacteria consume over
40,000 pounds/day of dissolved
organics they multiply and generate
approximately 6,000 pounds/day of
new bio-fliass.  Combined with the
continuous feed of powdered carbon,
this imposes a net solids increase
of over 12,000 pounds/day into the
aerators.  There must be an equiva-
lent purge of solids from the
system, and this is provided by
removing a side stream from the
concentrated solids underflow from
the secondary clarifiers.
        The WWTP has the capability
of regenerating powdered activated
carbon from these wasted secondary
solids.  This is another unique
feature of the WWTP, as the powdered
activated carbon is regenerated in a
multiple hearth furnace.  To our
knowledge this is the only commer-
cial use of a multiple hearth furnace
to regenerate powdered carbon.

        When regenerating carbon,
wasted "PACT" sludge is filtered
prior to being fed to the furnace.
The same type filter press is used
for "PACT" sludge as is used for
primary sludge.  The design solids
concentration of the bio-mass/carbon
filter cake was 35%, but in actual
practice higher solids concentra-
tion, 45%, was achieved with the
"PACT" sludge.  Without the presence
of carbon as an internal filter aid,
it would be difficult to get over
25% solids concentration.

        After filtration to a 45%
solids cake, the bio-mass/carbon
solids are conveyed to the top of a
40 feet high by 26 feet diameter,
five hearth furnace.  Under con-
trolled conditions, the solids are
fed down through the furnace, being
raked across each hearth by rabble
arms mounted on a large, rotating
center shaft.  Water is first evapo-
rated and then bio-mass and adsorbed
organics are burned.  Finally at the
bottom of the furnace at approximate-
ly 1,QOO°C, and in the presence of
steam, the powdered carbon is regen-
erated.  The incandescent carbon is
quenched in water, washed with acid
to remove inorganics, and recycled
to the aerators.  Heat to the furnace
is supplied from an external, oil
fired burner.  The heated gases rise
through the furnace countercurrent
to the solids.  Gas flow must be
controlled to avoid blowing powdered
carbon out of the top of the furnace.
                                      472
-------
The tremendous dilution provided by
over 20,000 gpm flow was a major
protection against the organic shock
loads that could be caused by a tank
truck discharge.  Because the WWTP
and "PACT" system had been designed
to handle large, varying loads of
aqueous wastes containing organic
materials, tne process was ideally
suited to treat batch discharges of
almost any organic-containing waste,
including those that would pass
through or be inhibitory to a
conventional biological treatment
system.

    This specific site occasionally
has a competitive, quality advantage
in treating dilute, aqueous wastes
when it is important to the customer
that the waste be entirely treated
on site, with no residual discharge
to the environment or other
treatment facility.

    The market survey indicated
there were significant quantities of
aqueous wastes, both at other Du
Pont sites and from non-Du Pont
facilities that could be treated
successfully and on a cost effective
basis at Chambers Works.  It also
turned out to be cost effective to
ship wastes substantial distances.

    A formal marketing "program was
established in 1980.  Since that
time the outside waste business has
shown steady growth.  Currently over
30% of the organic load to the WWTP
is from non-Chambers Works wastes.
Over 50% of the waste originates in
New Jersey and over 85% comes from
the New Jersey, Pennsylvania,
Maryland, Delaware region.  However,
the WWTP has received wastes from as
far away as Maine, Ohio, and Georgia.
The freight-logical area has been
extended in some cases by shipping
wastes by rail car, although as yet
this is a minor part of the business.
        Wastes from a variety of
     industries have been suitable for
     treatment.  Table IV shows some of
     the industries now  sending
     significant quantitites of aqueous
     waste  to  Chambers Works for
     treatment.
     TABLE IV.  TYPICAL INDUSTRIAL WASTE-
      WATERS  TREATED AT CHAMBERS WORKS
       Tank Truck & Tank Car Washings

       Pharmaceutical Wastes

       Water  from Oil-Water Separation
       Processes

       Textile  Treating Wastes

       Metals Treating Wastes

       Bio-sludges from Industrial Waste-
       water  Treatment Plants

       Electronics Industry Waste

       Landfill Leachates

       Lagoon Clean-Ups

       Chemical Process Wastes

       Latex  Wastes

       Chemical Cleaning Rinse  Waters

       Waste  Acids or Bases

       Food Processing Wastes

       Paint  &  Dye Wastes
         Initially outside waste treat-
     ment was limited to dilute, aqueous
     wastes that could be added directly
477
-------
to the WWTP Inlet waste stream, and
whose presence had no significant
effect on the quality of the WWTP
effluent.  A carefully developed
outside waste acceptance protocol
that includes administrative and
analytical check had to be developed
to assure the waste could not cause
any environmental concerns.  Because
of the nature of the WWTP there are
very few dilute organic wastes that
cannot be treated.  However the
inlet waste stream is acidic, so
wastes that generate toxic or
noxious gases, such as those
containing sulfide or cyanide,
cannot be added directly into the
inlet stream.

The presence of heavy metals and
their removal in primary treatment
must also be evaluated, since the
Chambers Works NPDES permit has
specific discharge limits for a
variety of heavy metals.  Many heavy
metals containing wastes can be
satisfactorily treated with careful
control of process conditions.  No
radioactive wastes or wastes
characterized as PCS wastes are
accepted, nor are those that contain
dioxin.

    The outside waste treatment has
been so successful that Du Pont is
expanding the business "beyond the
capability of the WWTP itself.  One
proposal is to establish pretreatment
facilities that themselves generate
aqueous waste streams which the                          Disclaimer
Chambers Works WWTP would then             Tu      .  .     .,  .  .
treat.  A Du Pont Environmental            Th!  fr*  **sc rl***™ th.ls.PaPer was
Services organization has been             "°*  fu"ded  J*  the u-$; Environmental
established to both market the             Protection  Agency.  The contents do
present treatment capabilities of          "P*  necessarily reflect the views of
the WWTP as well as develop new            the  Agency  and no official endorse-
environmental business.                    ment should be inferred.
                                     478
-------
         Supercritical Extraction of PCB Contaminated Soils

              i                2 by           l                 1
   B. O. Brady ,  R. P. Gambrell , K. M. Dooley ,  and F. C. Knopf4-

                  Department of Chemical Engineering
                     Department of Marine Sciences
                     Louisiana State University
                       Baton Rouge, LA  70803


    Supercritical  fluid  (SCF) extraction of organic hazardous waste
from contaminated soils is  a  promising  new technique for hazardous
waste site  cleanup.    The  ability  of  SCFs  to  solubilize  heavy
molecular  weight organics is well documented.  In this investigation
supercritical carbon dioxide (SC-CO2) was  used to extract PCBs, DDT,
and toxaphene from contaminated topsoils and subsoils.  An attractive
feature of this process is that the CO2, being virtually inert,  will
leave  no  solvent  residue  on the processed soil.  Furthermore, the
ease of separation of the extracted solute from SC-CO2 results in the
creation of a smaller waste  volume  of the now concentrated organic,
improving the efficiency of subsequent treatment  processes  such  as
combustion.
    Typically  in SCF extraction a simple solvent gas, such as carbon
dioxide, is contacted with a  solid  or liquid phase at high pressure
and moderate temperature.   Slight  changes  in  the  temperature  or
pressure  of the system can cause large changes in the density of the
solvent and consequently its ability to solubilize heavy non-volatile
waste compounds from the  solid  or • liquid  phase.   For example, by
manipulation of the system pressure a non-volatile can be  extracted.
Following  a  pressure  letdown,  generally  to  below  the  system's
critical   conditions,   this   same   material   can  be  completely
precipitated from the  solvent.    Thus  the  SCF phenomenon offers a
unique opportunity for  separation  and  recovery  of  "difficult  to
separate" materials in one processing stage.
    Supercritical   fluid  densities  compare  to  liquid  densities,
however their viscosities and diffusivities are intermediate to those
properties for liquids and gases.  Thus SCF have the solvent power of
liquids  with  better  mass  transfer  characteristics  than  typical
liquids.  Consequently  separation  efficiencies  for SCF extractions
can be much higher than for liquid solvent extractions.
    An experimental design involving two soil types, a topsoil and  a
subsoil, contaminated by three organic waste mixtures, PCBs, DDT (and
related  compounds),  and  toxaphene  was  employed  to determine the
effectiveness of SC-CO2 extraction.   For the first tests the topsoil
                                 479
-------
was contaminated with  approximately 900  ppm of DDT  and  400  ppm  of
toxaphene;  this  test  soil   was   obtained from an actual spill site
where the DDT and toxaphene had penetrated  the soil for a period of
10-12 years.  For the  next tests,  the uncontaminated topsoil (from an
area near the contaminated soil) was spiked in  the  laboratory  with
600  ppm  DDT.  In the final  tests,  the  subsoil, spiked with 1000 ppm
PCBs (Aroclor 1254), was used.   The  test soils were supplied by the
3JSU Wetland Soils and  Sedimentation Laboratory.
    Several factors affecting soil character also  affect  extraction
efficiencies, and some of these have been explored.  For example, the
effect  of  water in the soils was investigated by examining both dry
and wet (20% water) test soils.    The effect of long-term exposure of
the soil to the organic waste was  examined via extraction experiments
on both spill-site and lab-spiked  topsoils.   Finally, the  effect  of
solute   interactions    was   explored by  testing  spill-site  soils
contaminated with more than one organic waste.   Approximately 70% of
the DDT and 75% of the toxaphene can be  leached in under ten  minutes
by SC-C02 extraction.
    The  SC-CO, extraction of the  laboratory contaminated (with PCBs)
subsoil proved to be most promising,  with over 90% PCB extraction in
under one minute.
    In addition to the systematic  exploration of  factors  which  may
affect  soil extraction efficiencies, a  more fundamental study of the
extraction  .process  has  recently   been  undertaken.    Desorption
equilibrium constants  [ k =  (fluid phase composition)/(composition in
soil) ] have been measured for the various contaminants on the  soils
in  the  presence  of   a SC-CO2   solvent  phase.   These equilibrium
constants  are  being   used   in a  lumped  parameter  model  of  the
extraction  process  to  yield overall   mass  transfer  coefficients
characteristic of SCF  extraction.
                               Disclaimer

                    The work described in this paper was
                    not funded by the U.S. Environmental
                    Protection Agency. The contents do
                    not necessarily reflect the views of
                    the Agency and no official endorse-
                    ment should be inferred.
                                 430
-------
        MICROBIAL DEGRADATION OF POLY CHLORINATED BIPHENYLS

              Ronald Unterman, Donna L. Bedard, Lawrence H. Bopp
                Michael 3. Brennan, Carl Johnson, Marie L. Haberl

      General Electric Company, Corporate Research and Development Center
                            Schenectady, NY  12301
                                  ABSTRACT

      We have used a rapid assay to screen and characterize the polychlorinated
 biphenyl (PCB) degradative competence of 30 new bacterial strains isolated from
 diverse  PCB-containing soils  and sediments.  These strains differ  widely  in the
 extent, rate, and congener specificity of their  PCB degradative activity, and may
 utilize several different enzymatic  pathways.   We  have defined  at least  three
 groups of organisms based on their ability to degrade  one or more of the following
 PCB structural classes:  blocked at all  2,3 positions;  blocked at 4,4' positions; and
 lacking adjacent unchlorinated  positions.  Several of these bacteria degrade penta-
 and hexachlorobiphenyls. The  strains include many Pseudomonas and Alcaligenes
 species. We have found major  differences in PCB degradative competence among
 species of the same genus, and some  differential competence among strains of the
 same species. In addition, PCB degrading bacteria have been isolated from every
 PCB-containing site examined.   Metabolic intermediates have been characterized
 from many of these bacterial strains and suggest the presence of both previously
 described and novel catabolic enzymes. Current work involves further elucidation
 of the biochemical and genetic basis for PCB degradative competence, as well as
 development  of  technologies  which utilize microbial degradative organisms or
 enzymes.
INTRODUCTION AND PURPOSE

    The accumulation  of  xenobiotic
compounds in  our environment has had
a  profound effect  on  our  physical,
social,  economic  and  political  well-
being.  Some of these  substances  have
been  found to be harmful  to humans
while others, including  PCBs, have only
been implicated as such. In either case,
there has been a growing demand for
their  safe  disposal  or  preferably  de-
struction.

    More than 100 forms (congeners) of
PCBs, differing in the number  and posi-
tion of chlorine substituents, were com-
monly used over the last half-century.
The chemical and  physical properties
which formed the  basis for their  use
(thermal stability,  chemical  stability)
have also contributed to their accumu-
lation in the  environment. In addition,
they are highly insoluble in water  and
only slightly  mobilized by aqueous sys-
tems.  Thus,  PCBs  are generally found
in soil and sediments at sites of their
production, use, storage or disposal.

    PCBs,  especially the more highly
chlorinated congeners,  have generally
been considered resistant to biodegra-
                                      481
-------
 dation in the environment.  (For a re-
 cent review see reference 1). However,
 new data indicate that both aerobic and
 anaerobic biotransformations may  be
 much  more  prevalent than  previously
 thought  (2,3).   The  accumulation  of
 PCBs in  the  environment and possible
 effects on  human health have sparked
 an intense interest in devising an eco-
 nomical  means for their destruction. It
 is for these reasons that we have under-
 taken an extensive study of the  micro-
 biology,   biochemistry,  and  molecular
 genetics of bacteria that degrade PCBs.


 APPROACH

    Soil  and sediment samples were ob-
 tained from PCB-contaminated sites in
 New  York,  Massachusetts,  and  Cali-
 fornia, and a non-PCB  containing  site in
 Mississippi.   The  bacterial populations
 in these  samples were then enriched
 and  analyzed for their ability to  de-
grade  PCBs.     From   these   mixed
cultures pure cultures  were isolated and
characterized as to species, substrate
 utilization,  antibiotic  resistance,  PCB
degradative competence (both Aroclors
and pure  congeners),  PCB degradation
 products  (metabolites),   regulation  of
 PCB competence, and the presence  of
 plasmids.

    Specifically, soil  or  sediment was
added to a phosphate buffered minimal
salts medium (PAS) containing biphenyl
(BP) or BP plus PCB as sole carbon and
energy source.   In this way bacteria
were selected for their ability to grow
on  BP.   After several  passages on
PAS/BP  the mixed  bacterial cultures
were  assayed for their ability to de-
grade PCBs.  For these assays biphenyl-
grown  cells  were  harvested, washed,
and resuspended in phosphate buffer  at
an optical density of 1.0 at 615 nm.
One ml  aliquots of this  cell  suspension
were dispensed into screw-capped vials.
Controls  were prepared either by heat-
inactivation  of cells  at 70 C for  20
minutes, or by addition  of HgCl^  to a
final concentration of ImM. The PCBs
(Aroclor 1248  or  defined mixtures  of
pure congeners) were  added to a  final
concentration of 10 ppm, incubated  at
30°C with shaking  for  24  hours,  then
killed by the addition  of 10 ul of  per-
chloric  acid.    The acidified  cultures
were extracted with  2-4  volumes  of
hexane  or ether.  Samples were analy-
zed by gas chromatography  on  a Varian
6000  gas  chromatograph with  electron
capture detector  using  either packed
column  or fused silica capillary techni-
ques.   Results for all  cultures were
calibrated  using a non-degradable  PCB
congener (e.g., 2,4,5,2t,4f,5'-hexachloro-
biphenyl   or   2,4,6,2',4'-pentachloro-
biphenyl) as internal standard.

    Pure  cultures  were  isolated  by
selection  of  individual  colonies from
PAS/agar plates which contained BP  as
the sole carbon source.   These  clones
were  alternately  colony  purified  on
PAS/BP/agar and Luria agar plates. All
pure strains  were initially  characteri-
zed using the N/F system (Flow Labora-
tories,  Inc.).   Selected isolates were
sent to the American Type Culture  Col-
lection (ATCC; Rockville, MD)  for posi-
tive identification.  The pure  cultures
were then assayed for their  PCB degra-
dative competence as  described  above.
An example  of an Aroclor  1248  de-
pletion assay is shown in  Figure 1.

    For metabolite determination,  bac-
terial cultures  were assayed exactly  as
in the substrate depletion assay except
that PCB congeners were analyzed indi-
vidually and  introduced  at concentra-
tions ranging from 1-500 ppm.  Follow-
ing incubation  for  various times  the
cells were acidified (pH 1-2)  and ex-
tracted  with four volumes of anhydrous
ether.  An aliquot of  the ether phase
was  then  reacted  with N,O-bis  (tri-
methylsilyl)  acetamide  (BSA,  Pierce
                                       482
-------
Chemical Co., Rockford,  IL), then as-
sayed using gas chromatographic analy-
sis as described above.

     Six  isolates  were screened for the
presence of  plasm ids  using the  pro-
cedure  of Hansen  and Olsen (4).   For
those cultures where plasmids were de-
tected, the DNA was characterized by
agarose  electrophoresis  and restriction
endonuclease  analysis.
PROBLEMS ENCOUNTERED

     Prior  attempts  to  assay  micro-
organisms  for PCB  degradative com-
petence by measuring disappearance of
Aroclors (commercial PCB  mixtures)
have frequently produced false positive
findings because of volatilization or ad-
sorption losses.  Furthermore, these as-
says have  generally  left the chemical
nature of the competence obscure  be-
cause  of  incomplete  gas  chromato-
graphic resolution and uncertain  identi-
fication of the Aroclor peaks.  We have
avoided these problems by using defined
mixtures  of PCB  congeners  and  by
adopting  incubation  and  extraction
techniques which  prevent physical loss
of the PCBs.  The assay mixtures in-
clude  PCBs  ranging  from  diehlorobi-
phenyls to hexachlorobiphenyls and re-
presenting several structural classes:
chlorinated on a single ring (2,3-dichlo-
robiphenyl);  blocked  at  2,3   sites
(2,5,2',5'-tetraehlorobiphenyl);  blocked
at 3,4 sites (4,4'-dichlorobiphenyl); and
lacking adjacent  unehlorinated  sites
(2,4,5,21,4',5t-hexachlorobiphenyl).

     Packed  column  GC  analysis  of
these mixtures following a  24 hour in-
cubation with  resting  cells  permits
quantitation  of differential  congener
competence.   These  studies have  al-
lowed us to definitively and unambigu-
ously define  the PCB degradative com-
petence of all new isolates.  One such
assay is shown in Figure 2.
                            CONTROL
                        CORYNEBACTERIUM
                             HB1
    TBI
              TETRA    PENTA
                               HEXA
 Figure 1.  MB1 Degradation of Aroclor
 1248 (10 ppm).
     A 6 C D E  F 6
Figure 2.  MB1 Degradation of Defined
PCB Mixture. Lower GC is control
      A - 2,41
      B - 4,4'
      C - 2,4,4'
      D - 2,5,2«,5'
      E - 2,3,2',5'
F - 2,3,2',3'
G - 2,4,3',4f
H - 2,4,5,2',3'
I  -3,4,3',4'
J - 2,4,5,2',4',5'
                                        483
-------
RESULTS

PCB Competence of PureCultures

    We  have isolated PCB  degrading
bacteria which vary considerably.  Our
isolates (Table 1) include a gram posi-
tive Corynebacterium species;  Pseudo-
monads be-ionging to three different
ribosomal  RNA  homology  groups  (P.
putida, P_.    cepacia, and   P. testo-
steroni); an Alcaligenesfaecalis; and a
strain  (A. eutrophus) formerly assigned
to the genus Alcaligenes  but which in
fact probably represents a  new genus
(5).   Our findings illustrate the wide
diversity   of  environmental   micro-
organisms  which have  developed  the
ability to degrade PCBs. In the future
we would like to  determine whether the
genes  specifying the  PCB degradative
enzymes in these organisms are  related.
This information  would give us a better
understanding of the origin of PCB de-
gradative competence.
Corynebacterium sp.
Alcaligenes faecalis
Alcaligenes eutrophus
Pseudomonas:
  Pseudomonas putida
  Pseudomonas cepacia
  Pseudomonas tesiosieroni
  Pseudomonas sp.
  (Acidovorans group)
MB1
Pi434
H850

LB400, LB410
H201, PJ704, RJB
HI28, H336, H430
P1939, H1130
H702, PilOl
Table 1.  Identification of PCB Degrad-
ing Bacterial Strains.
    Table  2  shows  the  PCB  degra-
dative competence of 26 pure cultures
assayed on defined mixtures of 18 PCB
congeners.  As we had previously seen
with the mixed cultures, there is a wide
spectrum of  degradative  competence.
When examined in this  format, it be-
comes immediately obvious that there
is a natural  progression of degradative
capability.   The nine strains  with the
poorest ability to degrade PCBs primar-
ily attack congeners containing a non-
chlorinated ring  or a ring with a single
chlorine in the para position.   The re-
maining  strains  attacked congeners
containing a  ring with  a  single ortho
chlorine and showed varying abilities to
degrade tetra and pentachlorobiphenyls.
These  trends  in degradative compe-
tence may indicate that  the  PCB de-
gradative enzymes of these strains are
closely related.  On the other  hand, the
differences in congener specificity ex-
hibited by strains MB1  and H850 sug-
gest  that there are several distinct
pathways of PCB degradation.  This is
discussed more fully in a later  section.


    Several of our pure cultures easily
degrade  tetra-   and   pentachlorobi-
phenyls.   Even  more significantly,  P.
putida LBW3Q attacked every  congener
in the  defined congener  assay.   Other
congeners which are attacked by LB^fOO
include 3,5,3',5'-tetrachlorobi-biphenyl,
2,4,6,3',5'-pentachlorobiphenyl,     and
2,4,5,2%V,5'-   and   2,4,6,2',4',6'-hexa-
chlorobiphenyls.  It is particularly note-
worthy that none  of these congeners
contain  adjacent unchlorinated  sites.
Possible explanations  for  this  activity
are the presence  of a monooxygenase or
dehalogenase  in  this bacterium.  This
would be a novel  and extremely import-
ant finding in  terms of  basic  bacterial
biochemistry and potential decontami-
nation of environmental PCBs.

    Our varied collection  of  bacterial
isolates has  allowed  us  to  assess  whe-
ther there is any relationship between
the taxonomic classification of a bac-
terium  and its  ability to degrade  PCBs.
The similarities of congener specificity
in completely different species suggest
that the genes encoding the PCB degra-
dative  pathway may reside in diverse
populations, possibly on plasmids.
                                       484
-------
  CONDON
   nine
 IIIBCTITUENf  CONCEHO



1-


2-
2,3-
2,1-
3?i-


2,5-



2,3
,1

,1,1
,5,1
,2
2,5,2'
,3,2 ,3
2,3,2', 5'
2A5,2',3'
*Hj3 jH
2,5,3'. 1'
3,1,3',4'
2,5,2',5'
2,3,1,2',S'
2,1,5,2',5'
2.1,5,2',1',5'






PP P + PPP + + *
+ p + *
+ P P P
P f +
P * +
P P P
+ +
+ +
* +
P
 Table 2.  PCS Degradative Competence
 of Pure Strains.  P = 20-59%, + = 60--
 100% Degradation.
PCS Metabolites

    Our bacteria] strains show a varied
yet substantial capability for degrading
PCBs as demonstrated by our substrate
depletion assays.   We have now  fur-
thered the biochemical analysis of sev-
eral of these strains by using a metabo-
lite assay to characterize the products
of  these bacterial oxidations.    The
demonstration of metabolic  products is
important for several reasons. First, it
.unequivocally confirms that PCS deple-
tion,  as demonstrated  by our  various
depletion assays,  is  due  to bacterial
oxidation.  Second, it allows  us to begin
to elucidate  the metabolic pathway for
these oxidations;  and  third, it  shows
that more than one pathway for  PCB
degradation exists in bacteria since dif-
ferent organisms can produce different
metabolites from  the  same  PCB con-
gener.
    A  time  course  analysis  of  the
metabolites produced  from P.  cepacia
H201  incubation  with 2,4'-dichlorobi-
phenyl is depicted Figure 3. The oxida-
tion of  2,^-dichlorobiphenyl  (50 ppm)
by H201 is  rapid and  essentially com-
plete  in one  day.   It results  in the
accumulation  of  2-chlorobenzoate as
would  be expected from  the  general
observation  that  bacteria  oxidize  bi-
phenyl and PCBs via benzoic acid inter-
mediates, and  often  accumulate  the
chlorobenzoate  products (1).   In addi-
tion, H201 shows the transient  produc-
tion of a high molecular  weight inter-
mediate (Figure 3,B).  This presumably
is  an  initial oxidation product (e.g. di-
hydroxydichlorobiphenyl) which is  fur-
ther degraded to the  chlorobenzoate.
Our GC-mass spectrometry studies with
many  other PCBs have indeed shown
that  the  transient  high  molecular
weight  intermediates of PCB degrada-
tion are  hydroxylated-chlorinated  bi-
phenyls.  The mass spectrum  of  one
such intermediate from the oxidation of
2,3 dichlorobiphenyl by P. putida LBWO
is depicted in Figure 4.
                                       485
-------
—_JL_f^»-'
   '24
   T96
Figure  3.    H201   Oxidation  of  2,4'-
dichlorobiphenyl.  Incubation times are
indicated at the left.   Bottom panel is
heat inactivated control.  A = 2,4'-CB;
B  = transient  high  molecular  weight
metabolite; C = 2-chlorobenzoic acid; D
= internal standard.
                                                                      - SI(CHj),
                                                             Cl Cl
                                       H
                                                             »,   ;!.:   i
                                                             I    ! .   I
                                                             i    P   }

                                                                 L   j
if
\ f
n :'
9
•»S.,T 	
«, T " «•* »..
( fe 1
                                          Figure 4.  Mass spectrum of a deriva-
                                          tized (trimethyl silyl) metabolite from
                                          the LB 400 oxidation of 2,3 diehloro-
                                          biphenyl.  Determined to be a dihydroxy
                                          dichlorbiphenyl.
                                  With  the more highly chlorinated
                              PCBs,  these  high molecular  weight in-
                              termediates  often  accumulate.   This
                              has helped to facilitate our analysis of
                              the complete PCS degradative  path-
                              way.  An example of such an oxidation
                              is shown in Figure 5.  The parent penta-
                              chlorobiphenyl (A) is  only partially de-
                              graded by LB400 and at 20 hours there
                              appear to  be  several metabolites (B-D).
                              We are presently characterizing  these
                              compounds  using  GC-mass   spectro-
                              metry.


                              PCB Degradative Pathways

                                  A comparison of  the PCB  com-
                              petence of Corynebacterium  sp.  MB1
                              with A_. eutrophus H850  reveals  that
                              MB1 is far  superior in  its ability to
                              degrade congeners  substituted  in  both
                              para  positions.   On the other  hand,
                              congeners which are chlorinated at po-
                              sitions 2,5,2'  are preferentially degra-
                              ded  by  H850  (MB1  cannot  oxidize
                              2,5,2%5l-tetrachlorobiphenyl   at   all).
                              MB1 degraded 4,4'-dichlorobiphenyl al-
                              most completely to the 4-chlorobenzo-
                                      486
-------
  "20
 Figure 5. Oxidation of 2t4,5,2',5f penta-
 chlorobiphenyl (50 ppm) by LB 400 rest-
 ing  cells for 0,1,  and 20 hours.  A  =
 2,4,5,2',5f  CBj  B-D  = high molecular
 weight metabolites.
ate,  whereas  H850  shows  only poor
activity on this  congener.   These dif-
ferences are most pronounced at higher
PCB  concentrations.   In  addition  to
exhibiting   distinct  congener  prefer-
ences, MB1 and H85Q produce different
metabolites from the same congener.

     Based on these data, we have pro-
.posed that  the major  pathway  of PCB
metabolism in A^. eutrophus H850 utili-
zes a dioxygenase which  preferentially
attacks  at  carbon  positions  3,%  (6)
(Figure 6).  If this is the case, it will be
the first demonstration of involvement
of a  3,4-dioxygenase in  the  biodegra-
dation  of  biphenyl or  PCBs.   In con-
trast, Corynebacterium sp.  MB1 proba-
bly employs a more common 2,3 dioxy-
genase mechanism. We are currently in
                                          the  process  of  isolating  metabolites
                                          from  both  H850  and MB1  in order to
                                          elucidate these degradative pathways.
Genetics of PCB Degrading Bacteria

    In addition to our microbiological
and biochemical studies of specific PCB
degrading bacteria, we  have initiated
research into  the  genetic basis for this
enzymatic  competence.    An  under-
standing of the structure and expression
of  the  genes  which  encode   these
enzymes will help us to develop superi-
or enzymes, organisms and technologies
for the biological destruction of PCBs.

    Our initial experiments have  in-
volved studies of  the endogenous plas-
mids  harbored by several of our PCB
degrading bacterial strains.  To  date,
we have  demonstrated  that at  least
three of  our  PCB degrading bacterial
strains contain one or more plasmids.
Studies with mutants of H850 show that
the loss of the  ability  to  utilize  bi-
phenyl and  metabolize PCBs is  some-
times associated with loss of its plas-
mid.  Our goal is  to ascertain whether
any of the  plasmids of  H850  and our
other PCB-degrading  cultures,  harbor
the genes for  biphenyl growth and PCB
degradative competence.  We hope to
transfer  this competence, using  either
transformation or  conjugation techni-
ques,  to strains of bacteria that are not
capable of growth on biphenyl.

    This  work is  preliminary to the
eventual isolation  of the specific DNA
sequences which encode  the biphenyl/-
PCB degradative enzymes.  Ultimately,
the cloning  and characterization  of  bi-
phenyl/PCB catabolic genes should lead
to a better  understanding of the enzy-
matic oxidation of  PCBs as well as a
biochemical process for  their destruc-
tion.  It  is toward this goal that our
genetic research is ultimately directed.
                                       487
-------
                  H W
                                   msn
                          B
 Figure 6. Proposed Initial Oxidation of PCBs by 2,3- and 3,4-dioxygenases.

                               Disclaimer

                The  work described in this  paper was
                not  funded by the U.S.  Environmental
                Protection Agency.  The contents do
                not  necessarily reflect the views of
                the  Agency and no official  endorse-
                ment should be inferred.
                                         4.
     Furukawa, K. 1982. Microbial de-
     gradation  of polychlorinated  bi-
     phenyls,   pp.   33-57.     A.M.
     Chakrabarty,  (Ed,),  Biodegrada-
     tion and Detoxification  of  Envi-
     ronmental Pollutants, CRC Press,
     Inc., Boca Raton.                    5.
REFERENCES
1.
Hansen,  3.B.,  and  R.H.  Olsen,
1978. Isolation of large bacterial
plasmids and characterization of
the P2 incompatibility group plas-
mids  pMG-2  amd pMG-5.   3.
Bacteriol. 135:227-238.
2.    Brown,  3.F.,  R.E. Wagner,  D.L.
     Bedard,   M.3.   Brennan,   3.C.
     Carnahan, Ralph 3. May, and T.3.
     Tofflemire,  1984.   PCB  trans-
     formations in  upper Hudson sedi-
     ments!   Northeastern   Environ-
     mental Science 3:167-179.

3.    Bopp,  R.F.,  H.3. Simpson,  B.L.
     Deck, and N.  Kostyk, 1984.  The
     persistence of PCB components in
     sediments of  the lower Hudson:
     Northeastern      Environmental
     Science 3:180-184.
                                              Krieg, N.R.  and 3.G. Holt  (Eds.)
                                              1984.  Bergey's Manual of Syste-
                                              matic   Bacteriology,   Vol.   1.
                                              Williams and Wilkins, Baltimore,
                                              MD.

                                              Bedard,  D.L.,  M.3.  Brennan,  R.
                                              Unterman.   1984.  Bacterial  de-
                                              gradation of PCBs:  evidence of
                                              distinct  pathways   in  Coryne-
                                              bacterjum  sp.   MB1  and Alcali-
                                              genes  eutrophus  H850.   In   G.
                                              Addis  and  R. Komai (Eds.), Pro-
                                              ceedings   of   the   1983  PCB
                                              Seminar,  Electrical  Power Re-
                                              search Institute. Palo Alto, CA.
                                    488
-------
           RADIOLYTIC DECHLQRINATION OF POLYCHLORINATED BIPHENYLS

              Ajit Singh, Walter Kremers and Graham S. Bennett
                   Radiation Applications Research Branch
              Atomic Energy of Canada Limited Research Company
                 Whiteshell Nuclear Research Establishment
                     Pinawa, Manitoba, Canada  ROE 1LO
                                   ABSTRACT

     Sherman  (28)  found  that,  in  alkaline  isopropanol   solutions,  some
chloro-organic  compounds  were  dechlorinated  on  exposure  to  high-energy
radiation,  via  a  chain  reaction.   Later,  this work  was   extended  to the
polychlorinated biphenyls (PCBs) by Shinozaki and co-workers  (24).

     We  have  repeated and  extended  the work reported  by Shinozaki and co-
workers  and obtained  a higher efficiency (S value)  for dechlorination.  The
effects  of various  inorganic  and  organic  additives on  the dechlorination
process  were  also  determined.    The  radiolytic  dechlorination  process has
been used  successfully with drained capacitors  containing adsorbed PCBs, as
well as  with  soil  contaminated  with PCBs.  Some runs were  done on  the  scale
of 20-L  solutions.  Considerations relevant to the  industrial application of
this process are discussed.
INTRODUCTION

     Increasing   concern   about  the
environmental  and  toxicological  ef-
fects of halogenated compounds  (7,12)
led to a ban on the use of one  of the
more prominent members of this  group,
the  polychlorinated biphenyls  (PCBs)
(13), in most countries.

     Polychlorinated   biphenyls  are
stable and mobile  in the environment.
They  have  been found  in  marine life
(14,15),  and  cases  of  PCB-contami-
nated  food  have  been  reported (27,
34).   The toxicity  of PCBs has been
the subject  of many  studies  (7,11,12,
15,18,19,27,34,37)  and they  have been
classified   as  extremely  toxic  and
carcinogenic   by   the  Environmental
Protection Agency  (U.S.A.),  which has
placed  strict restrictions  on their
use  (20).   Similar restrictions have
also been  imposed  in Canada  (4).
     Almost  one  hundred  methods  are
available  for  detoxification,  or
safe disposal, of PCBs (4).  However,
most PCBs that have been taken out of
service  remain  in storage.   At pre-
sent, the most widely used method of
disposal  is  incineration  (4,6) oat
temperatures  greater  than  1100°C.
While the process  is  safe and effec-
tive at  temperatures  > 1100°C (8,16,
17,31),  at  lower  temperatures  the
destruction  of  PCBs  is  incomplete
(16), and accompanied by the produc-
tion  of even  more toxic compounds,
benzofurans  (1,3,4).    Cases  of  un-
satisfactory  incineration  of  PCBs
have  been  reported  by   the  media.
Perhaps,  on-line  monitoring  of  the
temperature  in  the  combustion  zone
and of the organic chlorine (in PCBs,
dioxins,  benzofurans,  etc.)   in  the
exhaust  gases  and ashes  needs  to be
added, or improved,  at some of these
facilities.
                                    489
-------
The use of  high-energy  radiation  for
the dechlorination of PCBs  and other
halogenated  compounds   at   ordinary
temperatures   has   been  intensively
investigated   (2,4,9,11,14,22-26,29,
32,33,35,36).  Since the first report
of  radiolytic  chain  decomposition  of
N20-saturated  alkaline  isopropanol
solutions  (27),  many reports   of
radiolytic  chain  reactions   in  alka-
line  alcohol  solutions  have  been
published  (2,5,10,23-25,28-30).   The
method was  then applied  to  the  de-
chlorination  of  PCBs by   several
Japanese  investigators  (2,23,24,30)
who  established  the following:  (i)
chain  dechlorination  of PCBs,  with
the initial  G(Cl-)  > 600 (G = number
of  molecules formed  or  destroyed  per
100  eV  of  energy   absorbed)  (31));
(ii)  stoichiometric  equivalence  of
the  production of  chloride,  acetone
and  biphenyl,  and   the  loss  of  the
alkali;   (iii) an  inverse  relation-
ship  between  G(Cl-)   and  the  dose
rate;  (iv) inhibition of the dechlo-
rination by biphenyl, acetone, oxygen
and  nitrobenzene;   (v)   stepwise  de-
chlorination of the PCBs.

     Shinozaki  (6)  estimated the cost
of  the deehlorination  of PCBs (~ 541
chlorine)   to  be  890,   95  and
54  yen/kg,  by photolysis,  cobalt-60
radiolysis   and  radiolysis   with   a
3-MeV  accelerator,  respectively.
(Now, 2 yen equals  about $1  US.)   We
found  this  work  promising   and  have
reinvest!"gated and extended  it.
EXPERIMENTAL

     Experimental  details  have  been
published earlier  (31).   Very brief-
ly,  the work was  done  with Arochlor
1254  (trichlorobenzene   removed),
drained  capacitors  and  PCB-contami-
nated soil.  The chloride anion yield
was  determined  with  ion-specific
electrodes and an Orion ion analyzer.
Analyses  for  PCBs,  their  radiolytic
products, acetone and  the cover  gas
were  done with  a  gas  chromatograph
(Hewlett-Packard Model  7620A  with  a
flame-ionization  detector  and  an
electron-capture  detector).    The
formation of hydrogen  and methane  was
determined by mass  spectral analyses;
the yields of both  of  these were very
low (G < 0.5).
RESULTS AND DISCUSSION

     Our  results  with  small   (ml)
samples of solutions of PCBs  in  alka-
line isopropanol were very  similar to
those  reported  by  Shinozaki  and  co-
workers (23,24).   However, the  effi-
ciency of  our  process  is  about  twice
that reported by them.  We  have  opti-
mized  conditions   to   maximize
G(-PCBs)-values  and   have   obtained
values of  G =  75  ± 5 in  small  sam-
ples.   For  the  20-L  runs,  the  best
value for G(-PCBs)  obtained so  far is
80.   In  comparison,  the  values  for
G(-PCBs)  reported   by  Shinozoki  and
co-workers  (23,24)  were < 39.    For
capacitors,  we  have  obtained  values
of G(-PCBs) - 25.

     Irradiation of  slurries of  PCB-
contaminated soil in alkaline isopro-
panol  lead to the  dechlori nation  of
PCBs.    The  yields  were,   however,
lower than for liquid PCBs  by factors
of two to four, as  shown in Table 1.
Table 1.  Initial yield (yield extra-
polated to time 0) of chloride anions
    *"!-))  on   irradiation   of  3.3%
   Is in alkaline isopropanol.
Addi ti ves
G0(C1-)
None
Clay soil
Top soil
   4000
   2300
   1000
                                    490
-------
     The  effect  of  cationic  and
anlonic  additives  on  radiolytic  de-
chlorination was also  examined.   The
presence  of  Fe3+,  Mg2"1",  Ca2+,  Mn2+,
C032-,  SO32-,  SI0,2-  and PO^-  did
not have any effect at concentrations
of  10~3  mol»dm~3.   While  sodium  and
potassium  hydroxide  gave  similar
results,  ammonium  hydroxide  dramati-
cally  reduced  the  dechlorination
efficiency.    Several  organic  addi-
tives  reduced  the  dechlorination
efficiency.

Mechanism

     The  key  reactions  that  bring
about the stepwise dechl on* nation are
as follows:

     On  irradiation,  the isopropanol
radical is formed-
     (CH3)2CHOH
             (CH3)2COH
(1)
     In the  presence  of alkali, this
radical undergoes  ionic dissociation
to produce the acetone am"on.
           .
      (CH3)2COH
               OH-
                            (2)
The  acetone  anion  reacts  with  PCBs
(RC1_)  with  charge  transfer,  which
results  in  the  loss  of one  of the
organic  chlorine  atoms as a chloride
anion.
(CH3)2CO- + RCl
   ?CH3)2CO
                              -  (3)
The  free radical  formed  reacts with
isopropanol, regenerating the isopro-
panol  radical,   and  thus  propagating
the chain reaction.
     •RC1  . -i-  (CH3)?CHOH  *
         CCR3)2COH + HRC1n_j_
                            (4)
The  detailed mechanism has been dis-
cussed  elsewhere  (31).  The ultimate
main products are biphenyl, potassium
chloride and acetone.

Advantages of theRadiolytic Process

     Some  of the  advantages of  the
radiolytic  process  over  the  other
processes (6) are as follows:

(1)  The  process  is  carried out  in
     the  absence  of  oxygen  or  air.
     This  completely eliminates  the
     possibility of  the  formation  of
     benzodifurans or dlpxins.
(2)  On-line  monitoring   of  the  de-
     chlorination  of PCBs  is a  key
     component  of  the process,  thus
     eliminating  the possibility  of
     incomplete detoxification.
(3)  The  process  is  applicable  to
     bulk PCBs as well as to PCB-con-
     taminated items.
(4)  The  process  converts  a  toxic
     waste into useful products.

Further Work

     We  plan to test  the efficiency
of the process using an electron beam
from   an   electron  accelerator  for
irradiations.  Where large quantities
of PCBs are  stored awaiting disposal,
electron  accelerators  may be cheaper
sources  of  radiation than   isotopic
sources.     Preliminary  calculations
suggest  that irradiation  costs  with
accelerators  could be  as  low  as
20<|:/kg,  for  the 10 MeV linear accel-
erator  designed at  the  Chalk  River
Nuclear  Laboratories  by  J.  McKeown
and  co-workers  (21),  provided  the
efficiency of dechlorination  does not
change much  at the higher dose rates.
This figure  does not include  handling
charges,  the costs  of  the chemicals
and  analytical  services,  or  expendi-
tures  for some further  research and
development  required to  optimize the
irradiation  conditions.  We are work-
ing  on a  detailed  cost  analysis for
the  process.  However,  one thing is
                                    491
-------
clear; the  cost would be  highly  de-
pendent on whether the product chemi-
cals, KC1,  acetone  and biphenyl,  can
be marketed.  We  are  also  working on
details  of  detoxifying  transformers
and  large  amounts  of  contaminated
soil.

REFERENCES
1.
2.
3.


4.
5.
6.
7.
8.
Ahling,  B.   and  Lindskog,   A.,
1982.    In  Chlorinated  Dioxins
and  Related  Compounds -  Impact
on  the Environment,   0.  Hutzin-
ger,  R.W.  Frei,  E.   Merian  and
F. PoGGhiari (Editors), Pergamon
Press, Oxford, p. 215.

Arai, S., Matsui, M., Moriguchi,
J. and  Imamura,  M.,  1972.   Rika
Gaku  Kenkyusho  Hokoku,  48,  185;
Chem.   Abstr.,   197J7  78,
158583n.

Buser, H.R.  and Rappe, C.,  1979.
Chemosphere, 8, 157.
Fishbein,  L.   1979.    Potential
Industrial Carcinogens and Muta-
gens,  Elsevier  Scientific  Pub-
lishing Company, Amsterdam.
     taminated  Materials,    Vol.  4.
     Test Incineration of Electrical
     Capacitors   Containing  PCBs.
     EPRI-FP-1207,  Vol. 4.

 9.   Foldesova,  M.,   Piatrik,  M.,
     Varga,   S.,  Tolgyessy,  J.  and
     Cervenka,  I,,  1979.   Radiochem.
     Radioanal. Letters, 40.  73.

10.   Freeman, G.R., 1970.  In Actions
     Chimiques  et Biologiques des Ra-
     diations,  M.  Haissinsky   (Edi-
     tor),   Vol.  14,  Masson et Cie,
     Editeurs,  Paris,  p.  73.

11.   Gilbert,   E.   and Guesten,  H.,
     1974.   Kernforschungszentrum
     Karlsruhe  LBer.J, KFK  1969 UF,
     104;  Chem.  Abstr.  1975,  82,
     63971f.

12.   Hutzinger,  0.,   Safe,  S.  and
     Zitko,  V., 1974.  The  Chemistry
     of  PCBs,  CRC  Press,  Inc., Boca
     Raton,  Florida.
M.H. Dillon Ltd., 1983. Destruc-  13.
tion Technologies for Polychlor-
inated  Biphenyls  (PCBs).    EPS
3-EC-83-1,  Ministry  of  Supply
and  Services  Canada,  Cat.   No.
EN 46-3/83-1E.

Evans,  R.,  Nesyto, E.,  Radlow-  14.
ski, C. and Sherman, W.V., 1971.
J. Phys. Chem. 75_, 2762.

Exner,   J.H.   (Editor),   1982.
Detoxication of Hazardous Waste,
Ann  Arbor  Science  Publishers,
Ann Arbor, Michigan.
                                       15.
                                       16.
Flynn,  N.W.  and  Wolbach,  C.D.,
1980.  Disposal of Polychlorina-  17,
ted Biphenyl (PCBs) and PCB-Con-
     Hutzinger,   0.,   Frei,  R.W.,
     Merian,   E.   and  Pocchiari,  F.
     (Editors),  1982.     Chlorinated
     Dioxins  and Related  Compounds -
     Impact on the  Environment, Per-
     gamon Press,  Oxford,

     Jansson,   B.  and Sundstrom,  G.,
     1982.    In  Chlorinated  Dioxins
     and  Related  Compounds  -  Impact
     on  the  Environment,  0.  Hutzin-
     ger, R.W. Frei, E.  Merian and F.
     Pocchiari  (Editors),   Pergamon
     Press, Oxford,  p.  201.
     Jensen,  S.,  1966.
     tist, 32, 612.
                                                            New Scien-
     Jensen,  S.,   Johnels,  A.G.
     Olsson,  M.   and Otterlind,  G.,
     1969.  Nature,  224, 247.

     Jones,  C.J.,  Hudson,  B.C.  and
     Smith, A.J., 1978.   J.  Hazard.
                                     492
-------
     Mater., 2_,  291.

18.  Komamiya,  K.  and Morisaki,  S.,
     1978.   Environ.  Sci.  Techno!.,
     ii» 1205«

19.  Kimbrough,   R.D.,  1972.    Arch.
     Environ. Health,  25,  126.

20.  McGraw, M.G., 1983.   Electrical
     World, No.  2, 49.

21.  McKeown, J.  and Sherman,  N.K.,
     1985.   Radiat.  Phys. Chem.  (in
     press).

22.  Radlowski,  C. and Sherman,  W.V.,
     1970. J. Phys.  Chem., 74_,  3043.

23.  Sawai,  T.   and  Shinozaki,  Y.,
     1972. Chem.  Lett.,  No.  10,  865.

24.  Sawai,  T.,   Shimokawa,   T.  and
     Shinozaki,   Y.,   1974.     Bull.
     Chem. Soc.  Jpn.,  47.  1889.

25.  Sawai,  T. Shimokawa,  T.,  Sawai,
     T.,  Hosoda,  I.  and  Kondo,  M.,
     1975.   J.  Nucl. Sci.  Techno!.,
     12., 502.

26.  Sawai,  T., Ohara, N.  and  Shimo-
     kawa,  T.,   1978.    Bull.   Chem.
     Soc. Jpn.,  51, 1300.

27.  Scholes,   G.,  Simic,  M.  and
     Weiss,  J.J.,  1963.   Discussions
     Faraday Soc., No. 36, 214.

28.  Sherman, W.V.,  1968.  J.  Phys.
     Chem.,  72_,  2287.

29.  Shimokawa,   T.  and  Sawai,  T.,
     1977.   J.  Nucl.  Sci.  Techno!.,
     14_, 731.

30.  Shinozaki,   Y.,  1977.   Irradia-
     tion  of PCB; a  paper  presented
     at  the  13th  Japan  Conference on
     Radioisotopes, Tokyo.
31.  Singh, A.,  Kremers,  W.,  Smalley,
     P.  and  Bennett,  G.S.,  1985.
     Radiat.  Phys.  Chem.  (in  press).

32.  Southworth,   G.R.  and  Gehrs,
     C.W.,  1976.    Water  Res.,  JLO_,
     967.

33.  Sunada,  T.,  1973.   Genshiryoku
     Kogyo,  19,  33;  Chem.  Abstr.
     1974, 80j~112211c.

34.  Tiernan,  T.O.,   Taylor,  M.L.,
     Solch,  J.G.,   Vanness,  G.F.,
     Garrett, J.H.  and Porter, M.D.,
     1982.   In  Detoxication of  Haz-
     ardous  Waste,  J.H.  Exner  (Edi-
     tor), Ann Arbor  Publishers, Ann
     Arbor, Michigan, p.  143.

35.  Vollner, L.  and Korte,  F., 1974.
     Chemosphere,   3,   271;  Chem.
     Abstr. 1975,  82J~52372k.

36.  Vollner,  L.,   Rohleder,  H.  and
     Korte, F., 1975.  Radiat. Clean
     Environ.,   Proc.   Int.  Symp.
     (IAEA), 285.

37.  Waldbot,  G.L.,  1973.     Health
     Effects of Environmental  Pollu-
     tants,  The   C.V.  Mosby   Co.,
     St. Louis.
                Disclaimer

   The work  described in this paper was
   not funded by the U.S. Environmental
   Protection Agency.  The contents do
   not necessarily  reflect the views of
   the Agency and no official endorse-
   ment should be inferred.
                                    493
-------
              U.S.            OP         MANAGEMENT OP HAZARDOUS WASTE

                                   Yearn H. Choi
                        University of the District of Columbia
                               Washington, B.C.   20004
                                       ABSTRACT

     The U.S. Department of Defense (DoD), as a large buyer of hazardous materials
and a large generator of hazardous wastes, is concerned with the management and
disposal of increasing quantities of hazardous wastes.

     Each DoD installation is required to comply with the requirements of federal
environmental laws and the 50 states' and many local solid and hazardous waste
laws and regulations.  The Office of the Secretary of Defense develops the policy
and monitors the overall programs results of the DoD components - the Army, Navy
and Air Force.  The Defense Logistics Agency (DLA) has the central responsibility
for DoD hazardous waste disposal.

     The DoD hazardous waste site cleamp program encompasses a systematic effort
to cleanup inactive or abondoned DoD hazardous waste sites.  The Department is
making a strong effort to coordinate its programs with all levels of goverrment
and to keep the public Informed concerning the problems and progress being made.

     New waste management operations will significantly reduce the production of
hazardous wastes.  Methods include:  reduction in the amount of hazardous materials
used, substitution of less hazardous materials, better quality control, applications
of Innovative technology, and recycling.
INTRODUCTION AND PURPOSE

     The Department of Defense
(DoD), the purchaser of more than
50,000 hazardous material line items
each year ranging from paint removers
to pesticides, adhesives to fuels, and
propellants to industrial solvents,
has a cradle-to-grave chemical waste
management program as mandated
by Public Law 9^580, the Resource
Conservation and Recovery Act of 1976
(RCRA) and Public Law 94-469, the
Toxic Substances Control Act of 1976
(TSCA).

     DoD generated an approximately
92,000 metric tons of hazardous wastes
In 1981, while the United States
generated about 57 millon metric tons
of hazardous waste in the same year (6).
The DoD wastes are by-products of its
operations and manufacturing processes,
specifically, explosive and propellant
manufacturers, as well as high technology
for specialized military applications.

     Most DoD waste streams are similar to
the civilian sector's in terms of the
Industrial activities generating wastes as
shown in Table 1.

     This paper describes and analyzes
DoD's hazardous waste management and
its research and development programs.
                                        494
-------
                                      Table 1
                          MELITARI HAZARDOUS WASTE STREAMS
Base Industrial-Related
  Operations/Proeess
Hazardous Wastes
Metal finishing
Degreasing
Painting and stripping
Miscellaneous aircraft
  repair wastes
Fuel storage and supply
  operations
Acids
Heavy metals
Caustics

Solvents [e.g.,
  trichloroethylene,
   methyl ketone (MEK)]

Paint strippers and thinner
Waste epoxy (resin)

Brake relining wastes
  (beryllium)
Metal stress and defense
  analysis wastes
  (fluorescent dye)
Welding wastes
  (acetylene sludge)

Fuel and oil wastes
Tank bottom sediment
Tank cleaning sludges
Pest control shops
Battery shops
Waste pesticides
Equipment wash water

Battery acids
Alkaline battery fluid
                                   495
-------
Hazardous Waste Cleanup Programs

     DoD's hazardous waste program
enconpases identification, control,
and cleanup of inactive or abandoned
disposal sites (Installation
Restoration (IR) Program), and
management of hazardous waste
from current and future operations.

     DoD started its IR program
in 1975, five years before the
passage of CERCLA, commonly
known as "superfund").  Under
CERCLA, Section 103(c),
notifications of all known or
suspected closed hazardous waste
sites were to be made to EPA
by June 1981.  Although
82 percent (227 or 274) of DoD
installations reporting in
response to Section 103(c)
were already included in the IR
program, some were not.  This was
because individual DoD installations
frequently reported suspected sites
based only on scant knowledge
in order to err on the side of
safety.  While evaluation of these
sites is still in progress, almost
all sites reported under Section
103(c) and not previously Included
in the IR program have been
confirmed as not to require any
follow-on action.
    The Department's goal is to
complete records searches to
identify hazardous material sites
potentially requiring corrective
actions at every installation likely
to have hazardous material problems
by October 1985.  The Deputy Assistant
Secretary of Defense for Installations
monitors the progress of the military
departments toward this goal.

     Three hundred ninety DoD
installations require records searches.
One hundred ninety-four have been
completed to date.  The remainder will
be finished by September 1985.  DoD
estimates that searches cost approximately
$50,000 each.  To date, remedial actions
have been started at ten installations.

     Probably the most widely reported
DoD installaltion restoration project
is the Army's effort at Rocky Mountain
Arsenal, Colorado.  Rocky Mountain Arsenal
is adjacent to the City of Denver, with
Stapleton International Airport lying
directly south of the Arsenal.  The Arsenal
complex, some 17,000 acres, has groundwater
contamination resulting fran military ehemlc
warfare agent production and Shell Chemical
commercial pesticide production.  The Army's
project at Rocky Mountain Arsenal began
in 1976.  To date, over $45 million has beei
spent to define contaminant migration and
provide remedial actions at the Arsenal.
                                         496
-------
     The data collection and analytical
effort at the Arsenal has been extensive.
Over 1,500 wells have been drilled on
the 25 square mile site.  Pour to six
thousand analyses are completed every
mouth, with over 270,000 data points
on record and over 900 technical
reports published.  The Array has
demonstrated at Rocky Mountain
Arsenal the committment of the
Defense Department to pursue its
cleanup of old waste sites.

Installation Restoration Research and
  Development

     The Army is the lead service for
compiling, refining and coordinating
the development of new and improved
technology and criteria for DoD IR
Program.  As part of this lead
service responsibility, the Army chairs
a Tri-Service Technology Coordinating
Committee.  At the present time, the Army
has three interagency agreements with
Environmental Protection Agency (EPA)
involving research and development for better
hazardous waste management.  The Air Force
and EPA are presently developing an
additional agreanent.  The objective
of these agreements is to ensure
cooperation and coordination between
DoD and EPA in research and development
for pollution abatement and
environmental quality management.

The research efforts include work
in the following areas:

     o  Decontamination and Cleanup
        Technology

     The thrust of this effort is
to identify and develop cost effective
technology for hazard containnent
and soil, groundwater or facility
decontaimination.  The effort Includes
pilot testing of rotary kiln
incineration as an environmentally
safe and cost effective method for
disposing of waste lagoon sediments,
studies of effectiveness of barrier and
liner materials for containing hazardous
materials, basic study of encapsulation
and fixation techniques for isolating
materials, study and evaluation of
techniques for removal of contaminants
from groundwater and lagoon wastewater
including carbon absorption, polymeric
resins, and supercritical fluid methods.
Nondestructive methods for cleanlng-up
buildings and facilities contaminated
with explosive, organ!cs and heavy
metals are also under study and
evaluation.  The successful containment/
decontamination system currently employed
at the north boundary at Rocky Mountain
Arsenal to prevent migration of
contaminated groundwater across the
boundary was an outgrowth of this R&D
program.

    Under its research and development
program, the Defense Department is
developing innovative as well as
conventional methods to resolve
hazardous waste decontamination problems.
DoD installations and Defense
Logistics Agency (DLA) are pursuing
actively these alternative solutions.
The continuously increasing costs for
incineration and secure landfills
provide the greatest incentive to
seek more cost-effective treatment and
disposal technologies.  DoD supported
the first major project for ocean
incineration of toxic chemical wastes
when the Vuleanus was outfitted to
dispose of remaining of supplies of
herbicide orange.  Molten salt
combustion, composting, deepwell
disposal, and soil incorporation (land
farming) are all currently under
investigation by DoD.

    A cooperative project is currently
planned at the Louisiana Army Ammunition
Plant located near Shreveport, Louisiana,
to demonstrate closure technology for
industrial waste lagoons.
                                         497
-------
    •Die' closure of the first waste
lagoon was completed in
PY 83 using solidification
of lagoon sludges and in-place
burial.  Technology for closure
of another, which contains
chemicals from explosive
manufacture, is still under
study.

    o  Criteria Development

    This effort includes environmental
effects studies of contaminants found
at DoD Installations which are not
typical problems for the civil sector.
These efforts will help establish
meaningful decontamination and health
hazard potential criteria.  Vegetative
and wildlife studies are being performed
on the compounds identified during the
assessment phases of the TO program.
Since the inception of the IR program,
over 20 compounds and their
decomposition products have been studied.
Currently, studies are underway on dioxin,
hydrazines, urea, picrates, and
picramates.  The latter two compounds
were found contaminating the soil
at a Naval installation.

    o Analytical Systems

    The Defense Department found early in
the IR program that standardized analytical
methods and sample reference materials
were required.  To promote consistency
and reliability of results, analyses
were performed by various government
and contractor laboratories.  Over 200
analytical procedures and 60 standard
analytical reference materials traceable
to the National Bureau of Standards
have been developed for the IR program.
The analytical procedures and methods
provide the basis for DoD IR program
quality control procedures.
Management of Hazardous Wastes from
    Current and Future Operations

   The Defense Property Disposal Service
(DPDS) of the Defense Logistics Agency
has storage and disposal responsibility
for all hazardous materials turned in to
DPDS by the Army, Navy, and Air Force for
disposal.  Since mid-1980 the DPDS has
had a central responsibility to assure that
DoD hazardous wastes are disposed of in
accordance with Federal, state, and local
laws and regulations.

   These materials are disposed of either
on-site or by contract agents, again In
conformance with applicable federal, state3
and local laws and regulations.

   DoD policy is that the DIA's hazardous
materials disposal cycle must include an
evaluation of waste material salvage and
resale possibilities; i.e., recycle of
materials.  An example of DIA's recycle
approach was in the recent sale of 368 tons
phosgene (manufactured as a chemical warfa
agent by the Army during World War I) to a
civilian firm in New York State.  The finn
used the chemical (carbonyl chloride) as a
feedstock for urethane plastic manufacture

    In two other important areas, studies
recently completed have shown that bases
using only a few drums of solvent per
year can economically recycle these
materials using commercially available
stills.  DoD is now developing a
comprehensive program for waste solvent
segregation collection, distillation,
and recycling.  In the related area of
waste lubricating oil, the Department
is investigating re-refining at regional
centers for recycle to DoD users.

    Collection, storage, and disposal of
DoD's toxic and hazardous materials are
accomplished through 142 Defense Property
Disposal Offices located on military
installations throughout the world, and
74 additional off-installation branches,
which serve as collection, storage,
                                         498
-------
and transfer points.  DPDS's normally
dispose of materials by contract to
commercial firms.  Contract
preparation is handled either at
Defense Property Disposal Service
Headquarters in Battle Creek,
Michigan, or at one of their five
Defense Disposal Regional Offices.

    DoD has many of its own storage
treatment and disposal facilities
across the nation, but it relies
primarily on civilian contract
firms for treatment or ultimate
disposal of hazardous waste.  In
1981, DoD disposed of 51,000 tons
on its own sites and 127,000 tons
off its installation (4).

    Reliance on outside firms is
critical to DoD.  The United
States needs about 50 to 60 new
hazardous waste management sites
over the next several years (7).

     Compounding the problem is the fact
that hundreds of old, envirormentally
inadequate disposal sites are scheduled
for remedial action requiring
removal and relocation of wastes.
Other existJung facilities will
be forced to close as strict RCRA
regulations go Into effect or as they
reach their capacity.  Unfortunately,
many recent attempts to locate sites
for new facilities have not been
successful because of opposition
from local residents.  Pew new off-site
hazardous waste treatment or disposal
facilities have been sited in the
United States since 1978 (1,5).
Florin! suggested interstate
3ompacts for hazardous waste
•nanagement similar to those being
3eveloped for low-level radioactive
.faste (2), under the Low-Level
3adloaetive waste Policy Act of
I960.
     The Array is developing new and
Innovative technology for more effective
management of hazardous wastes and
materials at its industrial facilities
and training and readiness installations.
Research thrusts include development of
technology for control of wastewater and
air emissions from munitions plants
and depots, development of envirormentally
safe disposal techniques for obsolete
or excess munitions, recovery and reuse
of explosives and propellants and
development of computer aided systems
for effective management of hazardous
materials at Army installations.
Development of a sulfide precipitation
method for treatment of electroplating
waste was recently completed and a pilot
system installed at Tobyhanna Army Depot
for operational evaluation.

     Other examples of Defense programs
to resolve hazardous waste problems
are the Navy chrome recycling effort
and Air Force programs for air stripping
of solvents from groundwater and
recovery of paint stripping solvents.

     The Naval Air Rework Facility (NARP),
Pensacola, Florida, was selected as the
test site for a Joint Navy-Department of
Biergy test of closed-loop system for
recycling toxic chrome electro-plating
wastes.  The project Involves an
energy efficient evaporation concept which
concentrates hazardous waste effluents
to a point where they may be returned to the
electroplating process, thereby conserving
resources and avoiding hazardous waste
generation.  Initial results have been
very promising, and there is additional
work underway to apply the concept to
other electroplating processes such as
eadmLurncyanide.  The potential is
considerable, and the Navy is considering
installation of the process at other chrome
electro-plating facilities.
                                         499
-------
    The  clean-up  of  contaminated
groundwater via packed tower alp
stripping and  reclamation of paint
stripping solutions  are two Air
Force  projects with  potential
broad  application.   Trichlorethylene
(TOE), an organic solvent previously
used by  numerous  DoD and  commercial
activities as  a degreasing agent, has
been discovered at various
concentrations in groundwater
aquifers throughout  the United States,
The Air  Force's Engineering and Services
Laboratory, located  at Tyndall Air
Force  Base, Florida, has  been studying
packed-tower air  stripping for its
efficiency and economic soundness
as a method for handling  long-term
clean-up of TOE in groundwater.
Packed tower air  stripping is a
process  involving stripping of
volatile impurities  (TCE)  in
liquids  by exposure  of the
contaminated liquid  to a  turbulent
countercurrent air stream.  The
Impurities are stripped from the
liquid by transfer to the air which
is subsequently exhaused.  Since TCE
concentrations are often  very low,
less than one  part per million, air
emissions are  not likely  to be
significant.   If they are  Important,
carbon absorbers may be added to
the system.  Even this dual treatment
system is likely  to  be cost effective
because  the life of  the carbon absorbers
is much greater when treating air streams
than when treating water  streams.  The
feasibility of air stripping other
volatiles is also being studied.

    The paint  stripping solvent
reclamation project  is directed
toward developing new methods to
extend the life of paint  stripping
solutions.  Deterioration  in paint
stripper performance coincides with
the buildup of paint solids.
Laboratory testing at the  Air Force
Engineering and Service Center
established techniques to remove the
paint solids from the stripping
solution and the recovered product
met performance specifications.  A
full-scale precoat, pressure filtration
system has been installed at Hill Air Force
Base, Utah, and is being evaluated.
Successful operation of the filtration process
will save $50,000 for every month the life
of the paint stripper is extended.  This
project is an excellent example of reducing
waste quantity by material reuse.

       The Army has formalized coordination o:
its research efforts with EPA under a memoran
of understanding and the Air Force and EPA
developed a similar agreement.  The Army
agreement has the objective of promoting
cooperation and coordination of research in
hazardous materials and other areas and
establishes a committee to monitor and promot
coordination efforts.

           FINDING AND RECOWEM3ATIONS

       The DoD hazardous waste site cleanup
program encompasses a systematic effort to
cleanup inactive or abandoned DoD hazardous
waste site.  DoD is making effort to coordina
its programs with all levels of government an
to keep the public informed concerning the
problems and progress being made.

       New waste management operations
will significantly reduce the production
of hazardous wastes.  Methods Include:
reduction in the amount of hazardous material
used, substitution of less hazardous material
better quality control, applications
of innovative technology, and recycling..

       The use of these methods will not
necessarily eliminate the waste problem.
However, waste generation control can greatlj
alleviate the total waste volume and degree c
hazard in many military operations.
                                            500
-------
     Hazardous waste source generation
control should be incorporated as an
integral part of the overall military
hazardous waste management
program.  Installation activities
should be examined by base and
technical support personnel to
determine feasible source reduction
operations and reuse incentives.

    High hazardous waste disposal
costs necessitate a new emphasis
in several phases of the military
system including procurement,
manufacturing, maintenance, and
refurbishing operations.  Economic
analyses can not be limited to
research and development,
acquisition, and operation and
maintenance costs but should
include, the projected costs
and benefits of hazardous waste
management,  For example, the
material acquisition system
should include the costs of
handling and disposing of
any hazardous material safely
and analyzing any hazardous waste
recycling or raise potential (3).

DoD management is also concerned with
technology development for cleanup
operations and hazardous waste
reduction and is working with
SPA in cooperative research and
levelopment efforts in these
ireas.
REFERENCES
1.
2.
3.
6.
7.
EnvironmentReporter, 1981,
Sitting of Hazardous waste Management
Facilities:  A Major Problem Facing
Industry and States.  November 13.

Pilorini, L.K., 1982, Issues of
Federalism In Hazardous Waste Control:
Cooperation or Confusion,  The Havard
Environmental Law Review, Vol. 6,
No. 2, pp. 334 - 336.!

Kawaoka, K.E., M.E. Bfelloy, Q.L. Dever,
and L.P. Weinberger, 1981, Military
Hazardous Wastes:  An Overview and
    Analysist Aerpsgace Report.
    (8376-1), Germantown, MD.
                             No. ATR-81
U.S. Department of Defense
Environmental Policy Directorate,
1982, Bfanagement-By-ObJectlves Report.

U.S. General Accounting Office, 1978,
How to Dispose of Hazardous Waste -
A Serious Question thatNeeds to
be Resolved, II, Document No. CED-79-13.

U.S. General Accounting Office, 1981,
Hazardous Waste Public Health and the
Environment, Document No. CED-81-158.

U.S. Senate Report, 1980, No. 848,
96thCongress, 2d Session, 90.
                                        Disclaimer

                          The work described  in  this  paper  was
                          not funded by the U.S.  Environmental
                          Protection Agency.   The contents  do
                          not necessarily  reflect the views of
                          the Agency and no official  endorse-
                          ment should be inferred.
                                          501
-------
                 CHARTING THE COURSE TO ENHANCED SOURCE REDUCTION

                            Robert B. Pojasek, Ph.D. *

                               Chas. T. Main, Inc.
                                Prudential Center
                          Boston, Massachusetts  02199


                                     ABSTRACT

     The report by the National Academy of Science Committee on Institutional
Considerations in Reducing the Generation of Hazardous Industrial Waste examines
key institutional, or nontechnical, factors that affect the generation of hazardous
waste by industry.  It provides a framework for evaluating public policies, both
regulatory and nonregulatory, to reduce the generation of hazardous waste.  The
report does not itself provide clear-cut solutions; rather, it provides a foundation
upon which improved public policies for hazardous waste management can be built.,

     The report's underlying premise is that waste reduction should be an integral
component of any national waste management strategy.  For the purposes of this
report, "waste reduction" refers not only to in-plant process modifications that
reduce the volume and/or degree of hazard of hazardous waste generated, but also
to reuse and recycling practices.

     This report is the first to deal with non-technical factors affecting the
generation of industrial hazardous waste.  Because little study has been devoted
to this topic, committee members have relied on their own experience and judgment
in formulating their recommendations.  The committee hopes to stimulate public
discussion of this subject as an important component of hazardous waste management.
INTRODUCTION

     To some it may have seemed
like just another tough assignment
for the National Academy of Science.
However waste reduction studies can
be intensely emotional, fueled by
the fires of partisan politics
   The views expressed are those of
   the author and do not reflect on
   those held by his fellow commit-
   tee members or the National
   Academy of Science.
and scientific skepticism.  The final
report by the "Committee on Institutional
Considerations in Reducing the Generation
of Hazardous Industrial Wastes" is his-
tory now.  It will have to stand on its
own merits for those who read it and
attempt to utilize its recommendations
in initiating or improving waste reduction
programs.  This paper will outline some
of the struggles which lay behind this
document.

     The National Academy of Science was
charted by Congress in 1863, i.e., in
Abraham Lincoln's days.  It was established
as a private, nonprofit, self-governing
                                        502
-------
membership corporation.  In 1916 the
National Research Council was added
to the charter for the purpose of
associating the broad community of
science and technology with the
Academy's purpose of furthering
knowledge and of advising the
federal government.  The Council
has become the principal operating
agency of both the National Academy
of Sciences and the National Academy
of Engineering in the conduct of
their services to the government,
the public, and the scientific and
engineering communities.

     The Committee on Institutional
Considerations in Reducing the
Generation of Hazardous Industrial
Wastes was organized in September
1983 to explore the nontechnical
factors that influence decisions by
industrial management to reduce the
generation of hazardous waste.  The
committee, sponsored by the Andrew
W. Mellon Foundation and National
Academy of Sciences Endowment Funds,
was asked to examine the public
policy approaches that may lead
industries to reduce generation of
hazardous waste.

     The committee was comprised of
the following people:

     RAYMOND C. LOEHR. Cornell
     University (Chairman)

     WILLIAM M. EICHBAUM, Maryland
     Department of Health and Mental
     Hygiene

     ANTHONY 0. FACCIOLE, JR.,
     Alexandria Metal Finishers,
     Inc. (deceased)

     SAMUEL GUSMAN, Taos, New Mexico

     ROBERT A. LEONE, Harvard
     University

     MICHAEL R. OVERCASH, North
     Carolina State University

     PHILIP A. PALMER, E.I. duPont de
     Nemours and Co.

     STEFFEN W. PLEHN, Fred C. Hart
     Associates, Inc.

     ROBERT B. POJASEK, Chas. T. Main,
     Inc.

     MICHAEL E. STREM, Strem Chemical,
     Inc.

     Staff

     RUTH S. DEFRIES, Staff Officer

     PAUL SCHUMANN, NRG Fellow

     JOYCE E. FOWLER, Administrative
     Secretary

     The premise underlying the study is
that reduction in the quantities of haz-
ardous wastes which are generated and
need to be treated or disposed is a bene-
fit to society.  The study's goal was not
to provide clear-cut solutions to a com-
plex issue; rather, it is hoped that
this report will provide a foundation
for public discussion from which improved
public policies for hazardous waste
management can be developed.

     Among the "institutional", or non-
technical factors the committee considered
were economic factors, such as capital
costs of waste reduction equipment;
regulatory factors, such as stringency
of standards; and psychological factors,
such as attitudes toward changes.
APPROACH

     Waste reduction is a topic which
almost everyone is talking about.  The
committee sought to collect as much
information as possible that existed on
the subject.  This was done in three
ways.  First an intern was retained to
                                       503
-------
catalog and summarize the vast
amount of information which had
been collected by the author on
the topic of waste reduction.
A literature search was under-
taken and groups active in waste
reduction were contacted directly
to see if there was more recent
unpublished information.  Second,
the Academy staff sent out hundreds
of letters requesting input into
our process.  Many people and
groups availed themselves of the
opportunity to participate.  Several
state governors personnaly responded.
Third, the Committee invited guests
to its monthly meetings to brief
the Committee further on programs
currently in progress.  These
guests included the Massachusetts
Department of Environmental Manage-
ment, Office of Technology Assessment,
and representatives of the California,
Minnesota, Illinois, North Carolina
and New York waste reduction programs.
Throughout this intensive search
and the numerous appeals directed
to industry, very little quantita-
tive, independently verified, and
peer-reviewed information was
gathered on this topic.  Indeed,
this was the first attempt at a
comprehensive work on institutional
considerations.

     Many examples of achievements
in waste reduction were brought to
the committee's attention, and the
committee tried to learn what it
could from them.  It is difficult
to generalize, however, from a
series of examples where there are
limited data to suggest their wider
applicability.  Much of the report
therefore represents the personal
experience and considered judge-
ment of the committee.

     By the third monthly meeting,
the Committee had a draft report to
work with.  New information and the
wide range of Committee members
experience were used to revise the
material.  Because it was difficult to
get a clear picture of exactly what was
meant by the term "institutional con-
siderations" in our charter, we opted to
utilize a more fundamental "barrier"
approach.  A large number of barriers to
waste reduction were formulated along
with the possible means to overcome
these barriers.  The Committee then
decided it could go no further with the
available information and that we had
something that was potentially ready for
release.

     To test this judgement, the committee
organized a workshop in May 1984 at which
a group of highly experienced people from
industry, state and federal government,
and environmental groups were asked to
discuss the issues raised in this report.
Discussion papers prepared by the Com-
mittee were circulated in advance and
served as the focus of the interaction.
The papers discussed the institutional
barriers to more effective waste reduc-
tion in the United States.  The workshop
participants responded that to focus on
the barriers to waste reduction seemed
unnecessarily negative, in that it did
not highlight the achievements that
have been made with waste reduction and
wrongly implied that opportunities for
waste reduction are limited.  The com-
mittee then framed the ideas in this
report in a more neutral tone, focusing
on "factors affecting industrial decisions
about waste generation".

A digest has been prepared to provide wide-
spread, immediate, free dissemination of
the core of the committee's deliberations
to policy makers in all sectors, industrial
leaders, and others who are concerned with
means to reduce the generation of hazardous
industrial waste.  The full committee re-
port, Reducing Hazardous Waste Generation:
An Evaluation and Call for Action, priced
at $4.95 list, is availablefrom the Na-
tional Academy Press at 2101 Constitution
Avenue, Washington, DC 20418,  The present
format of the Report is outlined in Table 1
                                        504
-------
TABLE 1.  NATIONAL ACADEMY. OF SCIENCE
          REPORT OUTLINE ON REDUCING
          HAZARDOUS WASTE GENERATION  •

Summary of Analysis, and Conclusions,

     1.  Introduction
         Scope of the Study
         Definitions of Hazardous
           Waste
         Estimates, of Hazardous
           Waste' Generation
         The Role of Waste Reduction
           in Waste Management
           Strategies-
         Dynamics of Waste. Reduction
           Strategies

     2...  Factors. Affecting Industrial
           Decisions About Hazardous
           Waste Generation
         Introduction-
         Cost of Land. Disposal
         Attitudes Toward Change
         Availability of Information
           About Waste Reduction
           Methodologies-
         Regulatory, Issues, in
           Reducing, the. Generation
           of: Hazardous Waste
         Needs for Research, and-
           Development
         Capital Costs.
         Issues, in Assembling,
           Processing, and Sale
           of Recycled" Materials
         Product Quality Standards

     3-..  Approaches for Encouraging
           Hazardous Waste Reduction
         Approaches for Encouraging
           Firms to Reduce Hazardous
           Waste Generation in the
           Initial Phase.
         Approaches for Encouraging
           Firms to Continue Waste
           Reduction Programs in
           the Development Phase
         Considerations in the
           Mature Phase

References,
Appendixes

A  Hazardous Waste Management Meth-
     odologies
B  A Typical Waste Reduction Program
C  Additional Documents Reviewed by the
     Committee
D  List of Workshop Participants
E  Biographical Sketches of Committee
     Members-
PROBLEMS ENCOUNTERED

     There were some definitional pro-
blems at the start.  The term "waste
reduction" as used In the report, refers
not only to in-plant process modifica-
tions that reduce the volume or degree
of hazard, of the- hazardous waste gen-
erated, but also to the reuse and re-
cycling of hazardous materials.  These
later cases can be accomplished both on
and; off the site of generation*  Many
state waste reduction programs have
chosen to exclude off site activities
in their programs or not distinguish
between recycle and reuse.

     Industrial decisions about waste
reduction are made for varied and complex
reasons.  The committee's task to under-
stand these reasons was constrained by
the lack of data on waste generation and
by the, lack of extensive literature on
the nontechnical aspects of waste reduc-
tion.  Therefore many of the observations
and: conclusions, in: the- report are based
on- the collective experience of the com-
mittee members.  The members bring a broad
spectrum, of experiences to the committee,
from large and small firms, public admin-
istration at the federal and state levels,
and consultancy with industry.

     The report was reviewed by an outside
group of experts under the Academy's strict
peer review guidelines.  The committee is
required to respond to all the comments.
A report review committee consisting of
members of the Academy is assigned to re-
view the committee's responses.
                                        505
-------
     A generally polar response pat-
tern was evident in the comments
as shown below:

     "The report does not impress
     me.  To the contrary, it
     leaves me uninterested."

     "Overall, the draft is clear,
     organized and useful as a
     general treatment of the
     subject.  It is balanced and
     contains numerous practical
     Insights."

     "My first impression was the
     feeling this was basically a
     philosophical or policy
     document but with little
     meat to it."

     "It is a comprehensive docu-
     ment that proposes a complex
     solution to reducing our
     nation's hazardous waste
     management requirements."

     This will provide some indica-
tion of the degree of opinionation
that can develop in such a new,
yet politically active area.  As
the data base increases and we are
able to gain a better understanding
of what is happening and why, this
diversity of opinion should be less
polar.  Certainly there will be more
reports.  Such a worthy goal as waste
reduction demands more than just the
"talk" it has been receiving to date.
RESULTS

     As a result of its deliberations,
the committee arrived at the following
general principles that should govern
efforts to reduce the generation of
hazardous waste.

     No single approach to encourag-
ing waste reduction will be most
effective in all circumstances.
The effectiveness depends on variables,
such as the type and size of the industry
or plant and the amount of waste reduc-
tion that has already been achieved.
The dynamic character of waste reduction
programs provides a framework to explore
the potential effectiveness of public
policy alternatives.

     Reductions in the generation of
hazardous waste can be expected to occur
through a series of loosely defined and
overlapping phases.  Initially, firms
consider changing their current waste
management practices in order to exploit
technically simple, low-cost waste re-
duction opportunities.  Firms then
undertake increasingly sophisticated,
more costly technologies to achieve
furher waste reduction.  Finally, firms
begin to confront the political, economic,
and technical limits to waste reduction
activities.  Different public policies
are appropriate at different stages of
an industry's waste reduction effort.

     It is desirable to reduce the gen-
eration of hazardous waste.  Regulatory
standards, however, should be based on
overall health and environmental consid-
erations and not made more stringent
than necessary solely to encourage waste
reduction.  If properly developed and
applied, standards will be a strong
Impetus to undertaking waste reduction
efforts.

     The costs of alternative methods
of waste disposal should reflect the
social costs of protecting public health
and the environment.  The Incentive for
industrial firms to pursue opportunities
for reducing waste generation will be
inadequate if the disposal option remains
priced below the true costs.

     Regulation will continue to play a
crucial and central role in the overall
waste management effort, but future
waste reduction is more likely to be
forested by non-regulatory methods,
such as information dissemination
                                       506
-------
programs and economic incentives.
waste generation.
CONCLUSIONS

     Most waste reduction efforts
in U.S. industry are still in their
early stages.  Many opportunities
exist for reducing the generation
of hazardous waste.  Efforts should
begin now to encourage industries
to take advantage of these
opportunities -

     At the current stage of
development of industrial waste
management programs across the
nation, substantial progress in
reducing the amount of hazardous
waste generated can be achieved by
employing relatively simple methods
that entail modest capital expense.
Such methods emphasize engineering
or plant specific circumstances.
The amount of waste generation that
can be avoided is, unfortunately,
not known, because of difficulties
in obtaining reliable data.

     The current trend toward in-
creasing costs of land disposal
for hazardous wastes — through
greater liability for generators
and site operators as well as
through restrictions on this use
of land — is an extremely import-
ant impetus to implementing waste
reduction programs.  To encourage
reduction in the amount of waste
generated in the future, this
trend bringing the cost of land
disposal to the level of its true
costs to society should continue.

     An important impediment to im-
plementing low-cost waste reduction
practices is lack of access to in-
formation about them.  Developing
means to exchange and disseminate
information about successful waste
reduction projects is an essential
first step toward reducing future
     Approaches other than the direct
regulation of manufacturing processes
are needed.  Within the regulatory
framework, regulations would be bene-
ficial that are administered consistently
and predictably and are flexible enough
to encourage the use of methods that
reduce the generation of hazardous
waste.

     In the long term, as implementation
of newer, more capital-intensive tech-
nology becomes necessary to reduce waste
generation further, public policies
will need to adapt to the different
considerations.  Industry may require
assistance — in the form of incentives
or subsidies, for example — to help
defray R&D as well as capital costs.
Risk assessment and risk management
studies will be needed for assessments
of more sophisticated waste reduction
options.  Research on these topics should
begin now.
REFERENCES

     Booz-Allen and Hamilton, Inc.
     (1982) Review of Activities of
     Major Firms in the Commercial
     Hazardous Waste Management
     Industry:  1981 update.
     Prepared for EPA (SW-894.1).
     17 pp.

     Bower, B.T., C.N. Ehler, and
     A.V. Kneese (1977) Incentives
     for managing the environment.
     Environmental Science and
     Technology 11:250-254.

     Campbell, M.E., and W.M. Glenn
     (1982) Profit from Pollution
     Prevention.  Toronto, Ontario:
     Pollution Probe Foundation.

     Chemical Manufacturer's Associa-
     tion (1983) The CMA Hazardous
     Waste Survey for 1981 and 1982.
                                         507
-------
Final Report.  Washington, DC.

Dow, E.W. and M.T. McAdams
(1982) Waste Exchanges:  An
Informational Tool for Linking
Waste Generators with Users.
pp. 86-94 in Huisingh, D. and
V. Bailey, eds. (1982) Making
Pollution Prevention Pay
Ecology with Economy as Policy.
New York:  Pergamon Press, 156
PP-

Economic Commission for Europe
(1981) Compedium on Low- and
Non-Waste Technology, Publica-
tion No. ECE/ENV/36, 2 vol.
Geneva:
United Nations.

Gaines, L.L. (1982) Industrial
Waste Exchange:  A Mechanism
for Saving Energy and Money.
Contract No. W31-109-ENG-38.
Agronne National Laboratory,
49 pp.

Governor's Waste Management
Board (1983) Annual Report.
Raleigh: State of North
Carolina.

Hall, R.M. Jr. (1983) The Pro-
blem of Unending Liability of
Hazardous Waste Management.
The Business Lawyer:  Vol.
38.  Feb. 83.

Herdon, R.C., ed (1983) Proce-
edings of the National Confer-
ence on Waste Exchange.
March 8-9, 1983.  Florida
Chamber of Commerce, Florida
State University.  U.S. Envir-
onmental Protection Agency.

Hirschhorn, J.S. (1983) Hazar-
dous waste source reduction
and a wasteend superfund tax.
Paper presented at Massachusets
Hazardous Waste Source Reduction
Conference.  Oct. 13, 1983.
Massachusetts Depart ment of
Environmental Management.

Huisingh, D., and V. Bailey, eds.
(1982) Making Pollution Prevention
Pay:  Ecology with Economy Policy.
New York:  Pergamon Press.  156 pp.

Ministere de L'Environment (1981)
Les Techniques Propres dans
L'Industrie Francaise.  Neuilly-
sur—Seine, France.

National Research Council (1983)
Management of Hazardous Industrial
Wastes:  Research and Development
Needs.  National Materials Advisory
Board.  Washington, DC:  National
Academy Press.

New York Environmental Facilities
Corp. (1983) Industrial Materials
Recycling act, Second Annual Report.

Office of Technology Assessment
(1983) Technologies and Management
Strategies for Hazardous Waste
Control.  Washington, DC.

Partlngton, B.P., J. Kohl, and E.
Dorn (1983) Making Pollution Pre-
vention Pay in the Electroplating
and Metal Finishing Industries:
Summary of a Workshop.  April 13-14,
1983.  Raleigh:  Water Resources
Institute of the University of
North Carolina.

Petulla, J.M. (1984) Environ mental
Management:  The Black Box Syndrome.
Environment: Vol. 26, No. 1.

Royston, M.G. (1979) Pollution Pre-
vention Pays.  New York: Pergamon
Press.  197 pp.

Sarokin, D. (1983) Source reduction
of hazardous and toxic wastes:
obstacles and incentives.  Paper
presented at Massachusetts Hazardous
Waste Source Reduction Conference,
Oct. 13, 1983.  Massachusetts
                                     508
-------
Department of Environmental
Management.

U.S. Environmental Protection
Agency (1981) Identification
and Listing of Hazardous
Waste, Sections 261.31 and
261.32 Background Document,
EPA Office of Solid Waste.

U.S. General Accounting Office,
EPA's Efforts to Clean Up Three
Hazardous Waste Sites, Report
to the Chairman, Subcommittee
on Commerce, Transportation
and Tourism, House Committee
on Energy and Commerce, GA01RCEO-
84-91, June 7, 1984, 1-6 pp.

U.S. General Accounting Office,
State Experiences with Taxes
on Generators or Disposers of
Hazardous Waste, Report . . .
GA01RCFO-84-146, May 4, 1984

Westat, Inc. (1984) National
Survey of Hazardous Waste
Generators and Treatment,
Storage, and Disposal Facili
ties Regulated under RCEA in
1981.  Prepared for U.S.
Environmental Protection
Agency, Office of Solid Waste.
                                    Disclaimer

                      The work described in this paper was
                      not funded by the U.S. Environmental
                      Protection Agency.  The contents do
                      not necessarily reflect the views of
                      the Agency and no official endorse-
                      ment should be inferred.

                                     509
-------
             HAZARDOUS WASTE MANAGEMENT STRATEGY IN ILLINOIS:
                             GOVERNMENT'S ROLE  '

             Michael  J.  Barcelona  and Stanley A. Changnon, Jr.
  Hazardous  Waste  Research  and  Information Center - Water Survey Division
            Illinois  Department of Energy and Natural Resources
                         P.O. Box 5050, Station A
                            Champaign,  IL 61820
                                 ABSTRACT

     Government can play a valuable advocacy role in the development of a
reasonable hazardous waste management strategy within its jurisdiction.
State government bodies, in particular, are in a position to stimulate
the adoption of waste reduction, resource recovery and alternative treat-
ment practices, as well as to discourage the continued use of land
disposal waste management options.

     The experience in Illinois clearly demonstrates that the Department
of Energy and Natural Resources, as a nonregulatory agency, is a good
venue for the promotion of safer waste management options.  A balanced
program of research, information transfer and industrial assistance has
been established to coordinate the development of a statewide hazardous
waste management policy.  Industrial, academic and public interest groups
are involved as contributors to this effort which supplements ongoing
activities on hazardous waste problem solving in other states and at
federal institutions.  The states must recognize that they can assume a
waste treatment information transfer role which federal agencies have
deeraphasized in the past ten years.
INTRODUCTION AND PURPOSE

The development of a comprehen-
sive hazardous waste (HW) manage-
ment strategy for an entire state
is an ambitious undertaking.
There are numerous groups with de-
cidedly parochial interests in-
volved in the research, regulatory
and enforcement aspects of stra-
tegy development which must be
consulted at various stages.
Complexities involved in the
issues of equitable waste regula-
tion, adoption of alternative
technologies to land disposal,
information transfer and problem
assessment research are all
potential obstacles to the
development of sound policies for
waste management.  Regardless of
the approach taken by various
                                   510
-------
governments, hazardous waste man-
agement strategy must be
approached from a basis of sound
information on waste management
practices.

     Illinois, as a major indus-
trial state with a significant
agricultural economic component
may provide a suitable model for
the development of reasoned waste
management strategy for other
states.  There are definite advan-
tages which result from allowing a
comprehensive and consistent tech-
nical approach to drive policy
development in specific areas of
hazardous waste management as
contrasted with a confining regu-
latory approach. A degree of
flexibility must be incorporated
into the development of an initial
strategy to allow for the discov-
ery of new data or waste manage-
ment market reactions.  Clearly, a
nonregulatory agency with neutral
technically sound response capa-
bilities has an edge in bringing
together the diverse societal
elements of waste management solu-
tions.  Public interest groups,
industry and various levels of
government share both the cause
and solution to hazardous waste
management problems if they can be
brought together and work coopera-
tively.

     In July 1981*, the Governor
and the legislature took action to
create a group which would provide
technical support towards the
development of a comprehensive
statewide hazardous waste manage-
ment strategy. The purpose of the
program was to provide a balanced
program of research, information
transfer and technical assistance
to industry in Illinois in support
of a statewide HW management
strategy.  This paper describes
the early development of the
Illinois Hazardous Waste Research
and Information Center (HWRIC). It
emphasizes the technical obstacles
to hazardous waste problem
assessment activities in a major
industrial state which may be
useful to other government bodies.
Industrial and public interest
groups may also find the results
useful as they face similar chal-
lenges in other states.
APPROACH

     The initial approach to the
development of the HWRIC concept
was the transformation of a vari-
ety of research and technical
assistance activities into a bal-
anced program which would meet the
long-term hazardous waste manage-
ment needs of the State.  The
experiences of a number of scien-
tists and engineers involved in:
ground-water, water and waste-
water, geological and toxicologi-
cal research (in support of regu-
latory programs) could be summed
up as being fragmented and lacking
a constituency for necessary
long-term support and product
implementation.  It was clear that
legislative and regulatory
time-frames were shrinking in
response to Federal actions and
the public recognition of the
seriousness of waste management
issues.  Further, these
time-frames were not synchronized
with research agendas dictated by
external funding arrangements or
priorities.  Problems of scale and
the basic geographic and climatic
differences between the states,
which were provided essentially
the same Federal fiscal and tech-
nical support, placed additional
pressure on the state's regulatory
agencies as they worked to respond
to a number of waste management
planning needs.  In summary, the
                                   511
-------
principal outcomes of scientific
and administrative task forces
(6,7,8,12) were the need for care-
fully directed research, informa-
tion transfer and industrial
assistance activities in support
of a comprehensive hazardous waste
management strategy (5).  It is
reasonable to state that if the
technical solutions to various
waste management problems could be
identified, necessary administra-
tive and regulatory adjustments
could be made for effective policy
implementation.
RESEARCH

     Research activities are fre-
quently undertaken to provide
answers to deceptively simple
questions of a basic nature.
Hazardous waste management issues
entail the study of a diverse
array of interrelated questions,
many of which have no simple,
obvious answers.  It was obvious
that waste problems relate to the
properties of the materials
involved, where they were gener-
ated, transported, treated or
disposed and the opportunities for
their release or exposure In the
environment.  It was also obvious
that no single scientific disci-
pline could approach unilaterally
the solution to the problems of a
major industrial state and that
Federal research priorities were
not necessarily those most criti-
cal to our unique situation.  In
order to focus the limited techni-
cal resources available to inves-
tigate the symptoms of improper
waste management, the research
activities of HWRIC were planned
in two main areas:  problem
assessment and problem solving.

     Problem assessment research
activities were needed to provide
a clearer picture of the extent
and magnitude of hazardous waste
problems in the state.  The need
was complicated by the overlapping
definitions of hazardous sub-
stances covered by various regula-
tions or by hazardous characteris-
tics (e.g. ignitability,
corrosivity, toxicity and reac-
tivity) .  The dimensions of these
activities had to be broadened In
view of the fact that hazardous
wastes had been managed by a
variety of means for over 50 years
in an environment which may take
decades to display symptoms of
waste-related impacts.  Available
regulatory information was limited
in waste composition data and the
short (i.e. 3-5 year) annual
reporting period required that the
initial problem assessment efforts
were planned as pilot projects.
There had been at least three
efforts to assess the statewide HW
problem in Illinois prior to 1982"
(8,9,10,11). These studies had
established several important
points;

  © There was sufficient overlap
    between the reporting in the
    special and hazardous waste
    categories to discourage gen-
    eralizations as to major HW
    generating industries, problem
    regions or major disposal
    practices;

  ® Deep-well injection and dis-
    posal, in or on the land, were
    major disposal practices for
    both treated and untreated
    waste streams; and

  © Estimates of both the total
    quantity of HW generated and
    that accounted for by treat-
    ment, storage or disposal
    facilities showed large
    discrepancies owing to one or
    more of the following factors:
                                    512
-------
    unspecified on-site practices,
    accounting difficulties with
    existing manifest systems and
    changing economic conditions.

     Figure 1  depicts the envelope
of estimates for the annual
amounts of hazardous waste gener-
ated in Illinois from 1920 to 1981
from a variety of sources (2).
The estimated quantities were
derived from state information,  as
well as from extrapolations on
industrial growth and Illinois'
contribution to total U.S. indus-
trial output.  Recent generation
estimates varied widely and at
least half of the total waste
generated prior to 1975 was
managed in the absence of regula-
tory controls.

     The need for pilot projects
for problem assessment was recog-
nized in a number of areas,
including:  estimation of the
potential health risk to human
populations via ground water,
development of a methodology for
searching historical records of
industrial and commercial activity
and identification of major liquid
HW streams, disposal practices and
options for their treatment and
disposal other than landfill ing.
The regulatory data has improved
in the last two years and an
expanded research program is cur-
rently underway.  The following
discussion provides a capsule
summary of selected initial
research efforts.
Potential Riskvia Ground Water

     Ground-water contamination
may occur by a number of mecha-
nisms.  Inappropriate waste man-
agement practices have been
identified as major potential
sources.  A study of the
assessment of past sources of
hazardous waste related activity
in both a rural and urban county
provided some very useful insights
into the magnitude of the problems
which may affect shallow sand and
gravel aquifers in areas of high
and moderate HW related activity.
The counties selected were among
the top ten in generation and
disposal of manifested RCRA waste
streams.  A simple HW activity
ranking'scheme was developed
(Fig. 2) and.applied ,in an attempt
to screen potentially serious
situations towards prioritizing
more in-depth investigations.
The susceptibility of aquifers to
surface contamination based on
geologic considerations was
supplemented by the development of
time-related capture zones for
public water supply wells.  This
approach acknowledges the'dynamics
of ground-water systems and has
been applied to the zoning of in-
dustrial/commercial sites in
Europe.  An example of a
time-related capture zone for a
well finished in shallow sand and
gravel is shown in Figure 3-  The
18-month study led to several
useful conclusions.

  © The accuracy and completeness
    of regulatory data on the gen-
    erators, waste composition and
    quantities was frequently
    suspect because of critical'
    data base management consid-
    erations.

  o Past HW handling activities
    prevent an equal and perhaps
    more serious threat to
    ground-water quality than the
    more recent data would
    indicate.

  © Historical documentation of
    past waste generation and
    management practices is
                                   513
-------
    contained in a diverse array
    of chamber of commerce, county
    census, property records and
    trade publications. The most
    reliable searching methodolo-
    gies require the expertise of
    an experienced industrial
    geographer and considerable
    manpower.

  © The ranking scheme had to be
    limited to planning applica-
    tions since much of the
    available data to support its
    use was incomplete.  Risk
    assessment is a very complex,
    information intensive disci-
    pline (1).

  © Illinois counties with more
    than 20,000 gallons (83 tons)
    of RCRA manifested wastes need
    more detailed study.  This set
    would include H2 of 102
    counties and account for >99I
    of the total quantity mani-
    fested in 1982.

     Figures H and 5 provide a
graphic illustration of the areas
which were judged to have degrees
of potential risk after applica-
tion of the ranking scheme in
county and metropolitan areas,
respectively.
Liquid Waste Characteristicsand
Management Options
     Legislative bans on the land-
filling of HW streams have been
established in a number of states.
In 1982 several related actions
were being discussed in Illinois
which ultimately lead to a ban on
untreated liquid waste landfilling
in 198H.  The final action encour-
aged pretreatment but retained an
emphasis on free liquids and par-
ticularly those associated with
organic solvents.  Anticipating
the enactment of the ban,  an
assessment of major liquid waste
generation and off-site management
alternatives was initiated in a
three-county metropolitan  area
(4).  The major findings of this
study were:

  © In 1982, industries in Cook,
    DuPage,  and Lake Counties in
    the Chicago area generated an
    estimated total of over 31*
    million gallons of liquid
    hazardous wastes that  were
    managed off-site.  Cook County
    generated nearly 90 percent of
    the total (over 30 million
    gallons) whereas DuPage County
    generated the least, at less
    than 2 percent (about  550,000
    gallons). The three counties
    contributed nearly half of the
    total state generation of
    liquid hazardous wastes mani-
    fested for off-site manage-
    ment.

  © Land disposed liquid hazardous
    wastes managed off-site in the
    three-county area in 1982 were
    estimated at nearly 22 million
    gallons.  12.5 million gallons
    of this total are generated in
    the three-county area. Most of
    the remainder is from  industry
    outside of the area (some from
    outside Illinois).

  © Acidic wastes and alkaline
    wastes were far greater than
    other waste categories gener-
    ated.  Other major waste gen-
    eration categories include;
    paint, solvents, distillation
    residues, plating, and metal
    wastes,  hydrocarbons,  and
    wastes within a miscellaneous
    category.  The most prominent
    land disposed waste category
    is alkaline wastes—with an
    even greater quantity  land
    disposed than generated.
                                   514
-------
  ® The results of the waste
    stream evaluations suggested
    that feasible waste management
    alternatives exist for large
    quantities of liquid hazardous
    wastes currently being land
    disposed.  Estimated quanti-
    ties of off-site managed
    liquid hazardous wastes which
    could be managed alterna-
    tively, were estimated at 12.5
    million gallons.

  e The time required to effect
    the potentially feasible
    reductions in land disposed
    hazardous waste is unknown.
    Some reductions from the 1982
    quantities have already been
    noted. In addition, the cost
    for land disposal of hazardous
    wastes is expected to continue
    increasing above inflationary
    levels.  The cost effective-
    ness of alternatives to land
    disposal was expected to
    become more evident. •

     The results of these and
several ongoing projects have
established a basis for more
detailed, statewide HW problem
assessment research.  We certainly
have a long way to go towards a
comprehensive waste management
strategy.

     Problem solving research
activities have yet to begin but
we are working towards the ele-
ments of a ground-water protection
plan with the fifteen or more
agencies with jurisdiction in this
area.
Information_ Transfer

     As a government agency with
no regulatory or enforcement
responsibilities,  ENR has adopted
a neutral, technically based
advocacy role in promoting HWRIC's
main goals.  The reception from
the public, state agencies and
industry has been encouraging.
These diverse interests agree'that
the emphasis of HWRIC programs
should be on the development of
high quality technical products
which deal with the most critical
problems which face the state.  As
the information base and our un-
derstanding improve, we expect to
be an active contributor to the
development of reasonable waste
reduction and management alterna-
tives.  We are committed to the
free exchange of our research
accomplishments (as well as recog-
nized difficulties) with all
related efforts.
Industrial and Technical Assis-
tance
     The ultimate usefulness of HW
research activities can only be
realized if the products and
understanding gained are eventu-
ally applied.  A number of states
(i.e. California, New York) have
initiated industrial assistance
efforts towards waste reduction as
exemplified by North Carolina's
"Pollution Prevention Pays" pro-
gram (3)« Alternative technologies
to land-based disposal and
end-of-pipe treatments are being
developed around the country for
application to HW problems.
HWRIC's industrial and technical
assistance efforts have been
planned to supplement these
efforts for specific applications
to high-risk, high volume waste
streams and to support site reme-
diation activities.   Cooperative
relationships with"other state and
industry programs have been initi-
ated to insure that economic and
environmentally sound practices
are considered in future
                                   515
-------
regulatory and administrative
actions.
RESULTS

     The beginning efforts of
Illinois government towards the
development of a comprehensive
waste management strategy hold
promise for the future.  The
dimensions of hazardous waste
problems are imposing and involve
the cooperation of all sectors of
society.  We intend to share our
results with all interested par-
ties in the hope that tractable
solutions to these problems can be
identified and implemented.
ACKNOWLEDGEMENTS

     The authors would like to
acknowledge the time, effort and
cooperation of ENR staff, their
contractors and the staff of the
Illinois Environmental Protection
Agency.  The valued advice and
consultation of representatives of
industry, state government and the
public have been most helpful in
the development of these activi-
ties.
REFERENCES

 1.  American Chemical Society,
     1984, Chemical Risk: A
     Primerj  Department of
     Government Relations and
     Science Policy, American
     Chemical Society, Washington,
     D.C., 12 pp.

 2.  Gibb, J. P., Barcelona, M.
     J., S. C. Schock and M. W.
     Hampton, 1983, Hazardous
     Waste in Ogle and Winnebago
     Counties: Potential Risk Via
6.
Present Activities; Illinois
Department of Energy and
Natural Resources, Document
No. 83/26, State Water Survey
Contract Report 336,
September, 1983, 73 pp.

Huisingh, D., 1984, The
Experience and Promise of a
Statewide' Pollution Preven-.
tion Program; Presented at
Pollution to Profit: Reducing
Industrial Waste in Illinois;
Proceedings of the Waste
Reduction Conference, April
16-17, 1984, Chicago, IL, ENR
Document No. HW84/02, 180 pp.

Hunt, R., N. Artz, J.
Sellers, R. Young, R; Welch,
T. Ferguson, 'T. Lapp, R.
Kakarlapudi, F. Hopkins, W.
Frerichs and K. R. Reddy,
1984, The Development of
Detailed Characterization of
Liquid Waste Streams
Generated by Illinois
Industries; Department of
Energy and Natural Resources
(ENR) Document No. 84/01;
Prepared by Franklin and
Assoc./ Midwest Research
Institute, Kansas and ENR
staff; March 1984, 86 pp and
Appendices.

Illinois Department of Energy.
and Natural Resources, 1983,
Hazardous Wastes in Illinois:
An Overview, Illinois Depart-
ment of Energy and Natural
Resources, Document No.
83/17, June, 1983, 116'ppand
Appendices.

Illinois Department of Energy
and Natural Resources, 1984,
Report of the Agency Task
Force on Hazardous Wastes
"Solving the Problem with
Information - A Plan for
                                  516
-------
     Action,  January,
     117 PP.
     Illinois Hazardous Waste Task,
     Force, 1983,  Committee Recom-
     mendations from the Final
     Conference of the Task Force,
     Attorney General Neil F.
     Hartigan and Senate President
     Philip J. Rock; Summary of
     the Activities of ten
     committees.
    State Water Plan Task Force;
    Water Resources Center
    UILU-WRG-83-0013,  University
    of Illinois,  March,  1983,
    91  pp.
 8.   Illinois Legislative Investi-
     gating Commission,  1981,
     Lamifilling of Special and
     Hazardous Waste in  Illinois;
     A Report to the General
     Assembly, August 1981,
     287 pp.

 9.   Klemm, H. A,  and M. A.
     Chillingworth, 1980,
     Industrial Waste Survey for
     the State of Illinois;
     Prepared by OCA Corporation
     for the Illinois Environment-
     al Protection Agency,
     4 Volumes (var. page),
     November 1980.

10.   Patterson, J. W., 1979,
     Hazardous Wastes Management
     in Illinois;  Illinois
     Institute of Natural
     Resources, Document No.
     79/32, October, 1979.

11.   Patterson, J. W. and
     C. N.  Haas, 1982, Management
     of Hazardous Wastes; An
     Illinois Perspective Report
     to the Illinois Department of
     Energy and Natural  Resources
     by Patterson  Assoc.,  Inc.,
     Chicago, IL,  250 pp.

12,   State  Water Plan Task  Force,
     1985,  Illinois Water Research
     Needs  and a Catalog of Water
     Research in Illinois,  Special
     Report No. 5 of the Illinois
              Disciafraer

The work described in this paper was
not funded by the U.S. Environmental
Protection Agency.  The contents do
not necessarily reflect the views of
the Agency and no official endorse-
ment should be inferred.
                                  517
-------
                        IEPA fISSQ) Documented Estimate
                        Estimate from 1960-1980 National Total
                        (ref. 1)
                     O  Estimate from ACS figures lot Total Solid
                        Wanes Allotting 50% to industry and 10%
                        of Industrial as Hazardous Iref. 7,8)
                        Estimated from USEPA 1973 figures for
                        Non-radioactive Hazardous Wastes (ref. 9.10)
                        Assorted Estimates (1-7 from ref. B)
                  1910
                            1920
                        Figure 1.
                                      1930
                                               1940  .   1950
                                                   YEAR
                                                                   1960
                                                                             1970
                                                                                      1980
Estimated hazardous waste  generation
  In  Illinois  1920-1980
(UHXINC MCTOKS ,
H«:m 1*
uolvvmt
SnfUlKa
                    2.   Schematic diagram  of ranking factors,
                           values,  and Input data
                                             518
-------
CJI
                                                                                                                             WINWE3AOO COUNTY
                                                                                                                             SCAIE 01= MILES
                              Figurt 3.  Diagp»» of a  oaptur* «re« with horizontal
                               trtv«l flM« (bracketed number) and vertical travel
                                        tlaw* tdded (unbraoketed nunbar)
                                                                                                                    •      10      IS      SO      2S      3D
                                                                                                                            SCALE OT KKOMETEflS
Figure 1.  Potential hazardous waste problem areas
                In Wlnnabago County
-------
:
                   POTBNTIAL RISK RATINGS


                     Secondary  Moderatt    Cpi*
                0   I   J   3   «   S   »
                     SCALE OF KILOM£TEAS
Figure  5.   potential  hazardous waste problea areas
         In tha  Rookford notropolltan area  •
                          520
-------
                         EMERGING TOXIC ISSUES FOR THE
                           ELECTRIC UTILITY INDUSTRY

                               Dr.  Ralph  Y.  Komai
                       Electric Power Research Institute
                          Palo Alto, California 94303


                                    ABSTRACT

     The  issue  of toxic  materials has  been  the  driving  force  from the  very
beginning of  environmental regulations,  because  of public concern about  wastes
in the  air,  the  water,  or  the land.   While  regulations  still  tie  public  and
environmental  health  to  those  hazardous  substances  found in  air,  water,  and
land, there has  been  an evolution  in the  basic  rules  of regulations,  application
of control, and  monitoring.

     Observation  of  contaminants   is  no  longer  strictly  a  matter  of  seeing,
tasting,  or  smelling a  problem.   The  tools  of  science  have  moved us  orders
of magnitude  lower  in what  man  can perceive.  As  technology progresses,  the
targets  for  measurement  become  smaller  and  smaller.  Some  regulations  seem
to have  shrinking targets,  even  as  companies  are  trying to  comply,  such  as
the target for PCB spill  clean-up.  While 50 ppm was  once  considered reasonable
for  regulating  PCBs, numbers  such as  0.05 ppm  or  whatever  is  decided to  be
the  detection  limit  are  considered  now  for  clean-up  standards.    Sometimes
that  number  is  vague because  of the  "no-threshold" level  for  cancer,  where
any molecule may potentially  lead  to an increase  in mortality rate.

     When  the  target was   larger,  it  was  easier  to  detect  compliance  or
non-compliance  and,  therefore, easier to  attribute  effects  to  the  substance
being  controlled.   Likewise  a  regulatory standard   was  something  which,  if
met, absolved one of  consequences  concommitant  with the standard.

     That  is  another  change  to  which  companies  must  adjust.  You can  meet
every  permit  requirement and  be  consistent with  every  legal  limitation,  but
still be  liable  for  the  consequences and remedial action  if it is  subsequently
observed  that  something  in  your  waste  is  causing harm  or  even  if it  is  one
of a  number  of  potential causes of  a  victim's  ailments.   This means that  past
practices  are not things  to  be   forgotten  and  everyone  does not   start  with
a  clean  slate as new  regulations  are  promulgated.  It means that today's  best
technology may  not be good  enough when  human  health  and the environment  are
at stake.

     The  importance   of   monitoring  is   not   just  in  proving  compliance  to
regulators, but of increasing  importance,  as  a warning to you  that  your  system
may  have  a  flaw.   The  sooner the environmental  problem is  discovered,  the
cheaper  the   remedial  effort.   The responsibility  and liability  continue  as
long as the waste has a potential  to cause harm.

     The  regulatory  picture is changing  also  with  such concepts  as  workers
right-to-know,  toxic  torts,  and victim's  compensation.   More  information  must
be collected  for  companies about  the chemicals in use, in both large and small
quantities, and about chemicals formerly used.

                                       521
-------
     In  come cases  the  data  is  sparse or  non-existent about  the  toxicology,
health  effects,  modes of  transport,  or rates  of migration.  As  processes  are
developed,  the  question  must  be  asked, is  there  a variation which will  yield
less of a toxic waste or  chemicals in less  mobile forms.   The data  base must
continually be expanded since the responsibility lasts forever.

     What are  some of the  problems which  the utility industry  must face? Some
that  I  foresee  are  the   PCB-contaminated   sites,  town  gas  disposal  sites,
regulated underground storage tanks, and pesticides/herbicides usage.

     PCB  or polychlorinated  Mphenyl   has  been  a  leader in  the  toxics  area.
It  is  the  only  compound which was specifically called  out for  regulation  in
the  Toxic Substances Control  Act.   It has  been  driven  by  public fear despite
toxicological   data.    Special   handling,   disposal,    transportation,   and
recordkeeping are required for this family of compounds.

     Utilities,  are  not  discovering new  problems  from the  residues of  past
practices,  which  are  unsatisfactory  by  today's  standards.  Wastehaulers  and
metal  salvagers  who  took  askarel  transformers  and PCB fluids and mismanaged
them have created  some contaminated  sites.   Utilities  because  of past  practices
may have created some themselves.   In both  cases, as these sites are  discovered,
and  finding them  even  if  you  are  seeking  them is  not easy without records,
utilities  can  find  themselves  in the  midst  of costly clean-up efforts.  As
stated  previously,  liability continues, and  the  financial responsibility finds
its way back to the deepest pockets — the  utilities.

     Another residual  of  past  practices is  the problem of town  gas disposal
sites.   From the  late  1800's  to  about  1950,  a major  source of  residential
gas  was pyrolysis/combustion  of  coal.  Coal  tar was a  by-product  which  was
landfilled as a waste product when  it  was not a salable  commodity.  Polynuclear
aromatic  hydrocarbons  (PAHs)  from the  coal  tars are on  the  list of  troublesome
organic compounds.   Cyanide  evolved from the organic nitrogen and was scrubbed
out  by  iron  oxide,  producing ferri-  and  ferrocyanides.   As  residuals  from
pre-1950,  it is  not only  difficult  to ascertain  what  was buried  where,  but
also  to find  people in  the companies who  know anything  about  the  operation
and  disposal  procedures.  This  is  another  expensive remedial  activity.   EPA
has  performed  an  historical  investigation  and turned up a  list of about 1500
companies  who  made  town  gas.   Most  are  extinct  companies.    Remedial  action
options  are few  and  not  without  drawback.   Excavation and   landfilling  may
be establishing next year's clean-up problem.

      Regulated  underground  storage tanks   (RUST)  is not   something  which one
would  necessarily  associate with  the utility  industry, but  there  are  a lot
of tanks, mainly for storing gasoline.  Again, this is  a major  clean-up  problem
with few alternatives.

      Herbicides/pesticides  occasionally receive  headlines   when tied  to  foods.
Utilities have used these chemicals  in  and around their  facilities  and along
rights-of-way.   Results  of  storage and  utilization potentially can  lead to
another toxic  clean-up problem.

      There  are  new  rules  for  regulation   and  compliance,   and  these  lead to
new problems  from  old  practices.   Utilities  could  find  themselves  involved
in significant toxic clean-up activities in  the  coming decade.

                                        522
-------
         THE EPA HAZARDOUS WASTE ENGINEERING RESEARCH LABORATORY'S
                  RESEARCH PROGRAM IN SUPPORT OF SUPERFUND

                                     by

                               Ronald D. Hill
                    U.S. Environmental  Protection Agency
              Hazardous Waste Engineering Research Laboratory
                           Cincinnati,  Ohio 45268

                                  ABSTRACT

In anticipation of the passage of the Comprehensive Environmental  Response,
Compensation, and Liability Act of 1980 (CERCLA or Superfund),  the Office
of Research and Development of the USEPA began a program in 1980 to support
the Agency's activities concerned with  uncontrolled hazardous waste sites.
In the area of environmental engineering and technology, the Agency looked
to the ongoing and established programs in land disposal  of hazardous  waste
and hazardous spill cleanup.  Thus a research program was established  to
develop and evaluate removal and remedial  cleanup technology.  This program
is now housed in the Hazardous Waste Engineering Research Laboratory in
Cincinnati, Ohio.

The Superfund program has been organized into four research areas:  (1)
containment technologies for remedial actions; (2) on-site cleanup equip-
ment; (3) in-situ treatment of hazardous waste/contaminated soils; and (4)
personnel protection.

The containment technologies area includes projects dealing with barriers
to mi "hi mi ze surface water Trif 11 trati on, limit or divert groundwater contact
with waste, and control leachate migration; waste storage, either temporary
or long-term; waste and leachate treatment; remedial  action technical
evaluations and modeling; and post closure evaluations.

An analysis of In-situ methods for cleaning contaminated soils  has identified
three approaches to the problem.  The first approach is the extraction of
contaminants from the soil with ultimate treatment of the extracted material
in conventional  wastewater treatment plants either on or off site.  Methods
such as surfactant washing and freezing would fall into this category. A
second approach is the immobilization of the contamination within the  soil
column and thus eliminating its transport.  Included in this approach  would
be methods such as grouting, pH adjustment, and thermal  fusion.   The third
approach would be the degradation/detoxification of the contaminants in
place, for example, the application of  microbes or chemicals to  the soil
that chemically change the contaminants to a nontoxic or less toxic form.
                                     523
-------
Crosscutting these three approaches is the need for methods  to deliver  the
chemicals, grouts, microbes, etc., to the soil  and methods to  recover
material where applicable.

On-slte separation, concentration, and detoxification technologies  are  an
alternative to containment and off-site disposal.   On-site technologies
being evaluated include a mobile incinerator, soils washer,  and carbon  re-
generator.

The objectives under the personnel protection area are:   (1) to develop and
evaluate chemical protectiveclothing and equipment; and (2) to develop and
evaluate procedures that will result in reduced chemical  exposures,  safer
working conditions, and more economical and efficient working  conditions
without a reduction in safety.

In order to transfer the state-of-the-art being evaluated within this pro-
gram to the user community, a series of Superfund remedial/removal  criteria
handbooks has been prepared.  Following is a listing of  the  twenty-five
handbooks that have been or will be published through FY'86.

1.   Guidance Manual for Minimizing Pollutions from Waste Disposal  Sites,
     EPA 600/2-78-142

2.   Compatibility of Grouts with Hazardous Wastes,
     EPA 600/2-84-015

3.   Review of In-Place Treatment Techniques for Contaminated  Surface Soils,
       Volume l--Technical Evaluation, EPA 540/2-84-003a
       Volume 2—Background Information for In-Situ Treatment,
                 EPA 540/2-84-003b

4.   Handbook for Remedial Action at Waste Disposal Sites,
     EPA 625/6-82-006

5.   Handbook for Evaluating Remedial Action Technology  Plans,
     EPA 600/2-83-076

6.   Slurry Trench Construction for Pollution Migration  Control,
     EPA 540/2-84-001

7.   Remedial Response at Hazardous Waste Sites,
       Summary Report—EPA 540/2-84-002a
       Case Studies—EPA 540/2-84-002b

8.   Design and Development of a Hazardous Waste Reactivity  Testing  Protocol,
     EPA 600/2-84-057

9,   Leachate Plume Management, (to be published in FY'85)

10.  Modeling Remedial Actions at Uncontrolled Hazardous Waste Sites,
     EPA 540/2-85-001
                                   524
-------
11.  Costs of Remedial  Actions at Uncontrolled Hazardous Waste Sites:
     Worker Health and  Safety Considerations,  (to  be  published in FY'85)

12.  Guide for Decontaminating Buildings,  Structures  and Equipment at
     Superfund Sites,
     EPA 600/2-85-028

13.  Evaluation of Systems  to Accelerate the Stabilization of Waste Piles
     and Deposits, (to  be published in  FY'85)

14.  Covers for Uncontrolled Hazardous  Waste Sites,  (to be published in FY'85)

15.  Stabilization/Solidification Alternatives for Remedial Action at
     Uncontrolled Waste Sites, (to be published in FY'85)

16.  Fugitive Dust Control  at Hazardous Waste  Sites,  (to be published in
     FY'85)

17.  Field Manual for Plunging Water Jet Use in Oil Spill Cleanup,
     EPA 600/2-84-045

18.  Guidance Manual  on Overtopping Control Techniques for Hazardous Waste
     Impoundments, (to  be published in  FY'85)

19.  A Guidance Manual  for  the Selection and Use of Sorbents for Liquid
     Hazardous Substance Releases, (to  be  published in FY'85)

20.  Drum Handling Practices at Hazardous  Waste Sites, (to be published in
     FY'85)

21.  Manual of Practice for Marine Safety  Officers and On-Scene Coordina-
     tors Involved in  Chemically-and/or Biologically-Contaminated Under-
     water Operations,  (to  be published in FY'86)

22.  Manual for Preventing  Spills of Hazardous Substances at Fixed Facili-
     ties, (to be published in FY'86)

23.  Reference Manual  of Countermeasures for Hazardous Substance Releases,
     (to be published in FY'86)

24.  Environmental Emergency Control Handbook  for  Hazardous Substances, (to
     be published in  FY'86)

25.  Flammable Hazardous Substances Emergency  Response Handbook: Control
     and Safety Procedures, (to  be published in FY'86)

Ordering information  for published reports is  available from the USEPA,
Center for Environmental  Research Information,  Cincinnati, OH 45268,
(513) 684-7562 or (FTS) 684-7562.

                                    Di sclaimer

   This paper has been  reviewed  in accordance  with the U.S. Environmental
   Protection Agency  peer and administrative review policies and approved for
   presentation and publication.
                                    525
-------
     SURROGATE  COMPOUNDS AS  INDICATORS OF HAZARDOUS WASTE INCINERATOR PERFORMANCE

                       Robert E. Mournighan, Robert A. Olexsey
                              Thermal Destruction Branch
                          Alternative Technologies Division
                   Hazardous Waste Engineering Research Laboratory
                         U.S. Environmental Protection Agency
                               Cincinnati, Ohio  45268


                                       ABSTRACT

     Evaluation of hazardous waste incinerators for Resource Conservation and Recovery
Act permits is a difficult and sometimes costly task.  The Environmental  Protection
Agency is evaluating, in its research program, several chemical compounds which may be
used to monitor incinerator destruction and removal efficiency and combustion conditions
more effectively.  This paper discusses the research aimed at developing reliable and
cost-effective methods for trial burns, and permit compliance monitoring.
INTRODUCTION

     Regulations promulgated under the
Resource Conservation and Recovery Act
(RCRA) have identified nearly 400 com-
pounds [17] which, if present in a waste
material, can cause it to be designated
as a hazardous waste and be subject to
the rigorous disposal requirements re-
quired by law.  When wastes containing any
of these compounds are incinerated, the
regulations require that the major waste
components (called principal organic haz-
ardous constituents or POHCs) be destroyed
and/or removed to at least 99.99* (de-
struction and removal efficiency or ORE).
Hazardous waste streams are often extreme-
ly complex mixtures containing hundreds of
compounds of various chemical classes.  An
analysis for POHC's and other major waste
components on a continuing basis is a dif-
ficult, time consuming and expensive task.

     Since 1979, the Hazardous Waste Engi-
neering Research Laboratory of the USEPA
has been intensively involved in research
to assess the capabilities of incineration
facilities to meet the RCRA standards.
This research is currently directed toward
Identification of easily-monitored facil-
ity operating parameters which correlate
with, or are surrogates for, system per-
formance.  From these efforts, we antic-
ipate establishing a sound scientific
basis for the selection of monitoring
methods which would simplify and reduce
costs of incinerator facility validations
and permit compliance monitoring.

BACKGROUND
     Much of the attention in thermal
destruction research has centered on the
incineration of chlorinated organic com-
pounds.  Of all  the compounds listed as
components of hazardous waste, chlorinated
organics were given high research priority
because they are the most common compounds
found in waste and represent a large por-
tion of incinerable wastes generated [11].

     Because of the large number of these
compounds efforts were Initiated early in
the research program to reduce the number
under study to a manageable level.  By
selecting POHC's representative of the
compounds most prevalent in priority waste
categories and which were generated in
largest amounts, twenty-eight compounds
were initially selected for the research
program [11].

     A major element in HWERL's thermal
destruction research program was to com-
                                         526
-------
 are the relative performance of different
 types  of incinerators,  industrial  boilers
 and kilns by means of conducting research
 trial  burns  to evaluate ORE.  It readily
 became apparent, in the beginning of this
 program, that waste analysis and stack
 sampling was more complex and costly than
 expected. To simplify  research trial
 burns  and to keep sampling and analysis
 costs  in line with the  research program
 budgets, nine of the twenty-eight com-
 pounds were  chosen for  full  scale evalua-
 tion tests in industrial  boilers and proc-
 esses  and pilot  scale research.  These
 compounds are listed in Table 1.  To be
 selected, the compounds had  to meet  the
 following criteria:

     0  easily detectable in stack
         emissions with  existing Sampling
         and  Analysis methods
     0  have high thermal  stability  for
         its  chemical  class
     °  representative  of physical and
         chemical  properties  of priority
         wastes.
           Table  I

Compounds  Selected  for EPA Research
Program
      carbon tetrachloride
      methylene chloride
      trichloroethylene
      perch!oroethyl ene
      chlorobenzene
      hexachlorobenzene
      pyridine
      acetonitrile
      toluene
SURROGATES FOR HYDROCARBONS AND
HALOCARBQNS

     After initial field tests, it
became apparent that (1) these compounds
were not always present in wastes at the
sites that were available for testing;
(2) there were analytical problems using
existing methods with methylene chloride,
toluene, acetonitrile and pyridine.  The
solution to the first problem, was to
"spike" or add a known amount of the
desired compounds to the waste in order to
measure destruction and removal efficiency
(ORE).  This approach, for the most part,
worked well, except that other problems
 appeared, such as handling  problems  due
 to toxicity, lack of  commercial  avail-
 ability and high cost of the spiked
 compounds.  Because of the  need  for
 spiking compounds into the  waste,  and the
 problems just discussed,  it  was  decided
 to choose compounds that  minimized these
 problems.  Additional  selection  criteria
 were added and were made  more restrict-
 ive by requiring the  compounds be  [11,
 18]:

     0  non-inflammable
     0  non-toxic
     0  commercially available at relative
        low cost
     0  high thermal stability
     0  not likely to  be  formed as  a  PIC
        (Product of Incomplete Combustion)

     A group of compounds  that fit  these
 criteria are the FREONS,  halogenated
 organic compounds which are  commercially
 available.  Table 2 lists some of  the more
 common ones which fit  the selection  cri-
 teria and would simulate  a range of  POHC
 physical  properties.   At  a workshop  on
 surrogate's sponsored  by  the  California
 Air Resources Board [12], Freon  113  was
 selected as the most promising of  those
 on  the list.   Its volatility  is  low, it'is
 nriscible with most of  the wastes under
 study, and is low in toxicity.   Its  chief
 use is as a degreasing solvent.

     The incineration of FREON 113  has
 also been  studied  in Sweden, both  in
 incinerators  [2]  and cement kilns  [1].
 The Swedish  studies reported it to be
 easy to detect  but difficult to destroy.
 Because  of the  ease of detection, de-
 struction  efficiencies could be deter-
 mined  to  a  very high degree.

             Table 2

         Trial Burn Surrogates

 Compound              - Boiling Point, °C
Freon 11
Freon 113
Freon 114
Freon 22B1
 24
 48
  4
150
Since 1983, the USEPA has conducted five
research trial burn tests to evaluate
the feasibility of using Freon 113 as a
surrogate for wastes; three  of the tests
were at cement plants, one at a light-
weight aggregate production  plant  and the
                                         527
-------
 fifth was  at  a  facility processing
 montmorillonite clay.   Table 3  summarizes
 the ORE  results for  each of the processes
 studied.

      The ORE's  of  Freon 113 and the POHC's
 in  the cement kiln and  aggregate plants
 exceeded the  incinerator standards of
 99.99S ORE.   In the  test of the plant
 processing montraorillonite,  Freon 113 was
 the only halogenated compound present at a
 significant concentration (>1000 ppm),
 Even though the process  temperature was
 low,  the ORE  of Freon 113 was greater than
 99.99S throughout  the test.  The results
 demonstrate that Freon  113  is as  difficult
 to  incinerate as other  POHC's in the tests
 and that it could  feasibly  be used as a
 reliable surrogate for most  wastes with
 volatile components.

      The other  Freons listed in  Table 2
 should be evaluated  as potential  surro-
 gates for less  volatile  and  even  solid
 hazardous wastes.  The physical  state of
 the surrogate compound  should be  the same
 as  the POHC under evaluation; e.g. Freon
 11  and Freon  113 should  be  used  for
 volatile POHC's and  Freon 22B1 should be
 used  for solid waste.  Evaluation of these
 compounds and further testing of the Freon
 compounds will be conducted  in EPA's pilot
 scale and laboratory research program to
 further document the reliability of this
 technique.

A DIFFERENT APPROACH
     The use of Freons and other halo-
genated orfanics, used as spiked material
1n hazardous wastes, may solve many of the
problems associated with conducting trial
burns.  Further Improvements, such as a
reduction 1n complexity of analysis re-
quirements, test time, and the associated
costs, would be benefits to both industry
and government.

     An approach which could conceivably
accomplish the objectives of shorter test
                time  and  lower cost  is  to  use an alto-
                gether  different  compound  such as sulfur
                hexafluoride  {SFg) [12,18].  SF6 is a gas
                which has  been widely employed as a tracer
                in  experiments studying atmospheric
                transport  of  pollutants [16], meteorology
                [13], and  atmospheric chemistry £5, 7, 8,
                10].  Its  high thermal  and chemical
                stability, which  made it very useful in
                the above  mentioned  studies, make it
                attractive as a potential  surrogate for
                use in  hazardous  waste  incineration trial
                burns [12, 18].   In  addition to these
                properites, it  is non-toxic [9, 14] and is
                commercially  available  at  relatively low
                cost [16].

                   Being  a gas,  SFs cannot be easily
                or accurately added as  a "spike" to liquid
                or solid hazardous waste.  It is most
                easily  introduced into  an incinerator or
                other device  by adding  it continuously to
                the combustion  air.  A  potential  drawback
                to this approach  is that it may not
                "simulate" the  combustion dynamics, matrix
                effects or waste  form that a POHC experi-
                ences.  However, it is only a presumption
                and a theoretical assumption that this
                method of addition may not be preferred
                [12].  The practicality and relative low
                cost of using SFg may override any con-
                census about  its suitability as a sur-
                rogate.

                    As part  of continuing evaluation of
                surrogates, the California Air Resources
                Board and the USEPA sponsored three tests
                where the feasibility of using SFg  as a
                surrogate was investigated.  The  tests
               were conducted at a circulating bed corti-
               bustor (CBC)  [4], a dry-process  cement
                kiln and an asphalt plant Table 4 lists
               the SFg destruction and  removal effi-
               ciencies (ORE) from each test,  and  the
               temperature for each  process.  DRE's for
               the POHC's in these tests  are not yet
               available. Further studies, specifically
               devoted  to developing the  relationship
               between POHC and SFg  destruction, are
               needed before final  conclusions can  be
                                         Table 3

                            Freon 113 Feasibility Test Results
             Process

             Cement
             Aggregate
             Montmorillonite
Freon 113
   ORE

   >99.999
   >99.9999
   >99.99
POHC
ORE	

>99.996
99.99992
   Process
Temperature,  °C

     1500
     1150
     650
                                           528
-------
                                        Table  4

                              SFg  Feasibility  Test  Results
                    Facility

                    Cement  Kiln
                    CBC*
                    Asphalt Plant
   ORE

 >99.999%
90 - 99.99*
99.3 - 99.84
Process Temp °C

      1500
    790 - 850
       190
                    Circulating Bed Combustor

drawn on the  suitability of SFg as a
surrogate.

    The SFs destruction efficiency in the
circulating bed combustor varied con-
siderably and appeared to be a function of
temperature.  In the cement kiln test how-
ever, destruction was so complete that SPg
could not be  detected in the stack gas
from the kiln.  The destruction efficiency
was calculated based on the detection
limit of the 6C/ECD used for the analysis.
Freon 113 was also used as a surrogate for
this test and its ORE was >99.99£.

    The asphalt plant test was quite dif-
ferent from the other tests in that the
SFs ORE was not very high and the values
were confined to a very narrow range.

    The significance of these results is
that for a range of process temperatures,
there is a range of SFg ORE.  As these
were feasibility tests for the evaluation
of the concept, further work is needed to
establish the relationship between haz-
ardous waste and POHC destruction with SFs
ORE.

    A driving force which might spur
further research is the ease with which
SFs data was collected.  By the third
test, at the asphalt plant, the experi-
mental  protocol  had been developed to the
point where all  of the necessary data
could be collected in 4 hours of testing.
More than 15 stack samples per hour were
taken and analyzed on a semi-continuous
basis.

    Recently, analyzers developed for
continuously monitoring SFg concentra-
tions in air, have been used for atmos-
pheric studies [3, 6, 15] and could
potentially be adapted for use in stack
gas analysis.  This would enable a
facility operator to have a continuous
readout of SFs destruction and removal
        efficiency. Continuous monitors  for SFs
        could also be used to monitor  permit
        compliance on a  real-time  basis, a tool
        which environmental  agencies do  not yet
        have.

        CONCLUSIONS AND  RECOMMENDATIONS

             Based on feasibility  and  proof of
        concept  tests, certain Freon compounds and
        sulfur hexafluoride  may prove  to be useful
        surrogates for hazardous waste incin-
        eration  and be used  for permit compliance
        monitoring.  Experiments showed  that these
        compounds  are easily introduced  into the
        incineration systems, require  relatively
        simple and low cost  equipment  such as GC/
        ECO  and  are amenable to on-site analysis.

             Further work  on developing  and demon-
        strating the relationships between sur-
        rogate ORE and hazardous waste destruction
        is recommended.  Laboratory and pilot
        plant research is  needed to further
        develop  sampling and analysis protocols,
        and  to make direct comparisons of POHC
        DRE's with  those of  surrogates.
                                         529
-------
REFERENCES
 1.  Ahling, B., "Destruction of Chlo-
    rinated Hydrocarbons in a Cement
    K1ln," Environmental Science and
    Tehnology, 13(11):1377-1379,
    November 1979.

 2.  Ahling, B. and Wiberger, K.,
    "Incineration of Freon TF,"  Swedish
    Institute for Water and Air  Pollution
    Research, Report K 648, Stockholm,
    Sweden, January 1982.

 3.  Baxter, R» A., Pankratz, D.  and
    Tombach, I., "An Advanced Continuous
    SFg Analyzer for Real-Time Tracer Gas
    Dispersion Measurements from Moving
    Platforms," AV-TP 83/504, Aero-
    Vironment, Inc., Pasadena, CA, 1983.

 4.  Chang, 0. P. Y., and Sorbo,  N.
    "Evaluation of a Pilot Scale Circul-
    ating Bed Combustor with Surrogate
    Hazardous Waste Mixtures," USEPA
    Eleventh Annual Research Symposium on
    Land Disposal  and Remedial Action,
    Incineration and Treatment of Haz-
    ardous Wastes, Cincinnati, Ohio,
    April 1985.

 5.  Dletz, R. N. and Cote,  E. A.,
    "Tracing Atmospheric Pollutants by
    Gas Chromatographic Determination of
    Sulfur Hexafluoride," Environ. Sci.
    Technology, 7:338,  1973.

6.  Dietz, R. N. and Goodrich, R. W.,
    "The Continuously Operating  Per-
    fluorocarbon Sniffer,"  BNL 28114,
    Brookhaven National  Laboratory,
    Upton, NY, 1980.

7.  Dietz, R. N. and Oabberdt, W. F.,
    "Gaseous Tracer Technology and Appli-
    cations," BNL 33585, Brookhaven
    National  Laboratory, Upton,  NY, 1983.

8.  Dietz, R. N. and Senum,  6. I.,
    "Capabilities, Needs and Applications
    of Gaseous Tracers," BNL 35108,
    Brookhaven National  Laboratory,
    Upton, NY, presented at  Atmospheric
    Tracers Workshop, Sante  Fe,  N.M.,
    May 1984.

9.  Johnson,  W. B., "Meteorological
    Tracer Techniques for Parameterizing
    Atmospheric Dispersion,"  J.  Appl.
    Meteorol. 22:  931,  1983.
 10.  Lester, D. and Greenberg,  L. A.,
      "The Toxicity of Sulfur Hexa-
      fluoride,1' Arch.  Ind.  Hyg.
      Occupational Med. 2:348-349, 1950.

 11.  Mournighan, R. E., "Preliminary POHC
      Ranking and Test Matrix,"  EPA
      internal planning document, June 1981.

 12.  Report on Workshop on  Incineration
      Surrogate Compounds, University of
      California-Davis to the California
      Air Resources Board, November 1983.

 13.  Saltzman, B. E., Coleman,  A. I. and
      demons, C., "Halogenated  Compounds
      as Gaseous Meteorological  Tracers,"
      Anal. Chem. 38:753, 1966.

 14.  Sax, N. I., "Dangerous  Properties of
      Industrial  Materials",Van  Nostrand
      Reinhold, New York, 1984.

 15.  Simmonds, P. G., Lovelock, A. J. and
      Lovelock, J. E., "Continuous and
      Ultrasensitive Apparatus for the
      Measurement of Airborne Tracer Sub
      stances," J. Chromatogr. 126:3, 1976.

 16.  Swartz, S.  E., Leahy, D. F. and Fink,
      S., "Aircraft Release of Sulfur
      Hexafluoride as an Atmospheric
      Tracer," Jour. APCA 35:555, May 1985.

 17.  Title 40, Code of Federal  Reg-
      ulations, Part 261, Appendix VIII.

 18.  Tsang, W.,  and Schaub,  W.  M.,
      "Surrogates as Substitutes for
      Principal Organic Hazardous Con-
      stituents of Incinerator Operations,"
      Second Conference on Municipal, Haz-
      ardous and  Coal Wastes  Management,
      Miami Beach, Florida, December 1983.
                  Disclaimer

This paper has been reviewed  in
accordance with the U.S.  Environ-
mental  Protection  Agency  peer and
administrative review  policies and
approved  for presentation  and publi-
cation.
                                       530
-------
                     EVALUATION OF ON-SITE INCINERATION
                FOR CLEANUP OF DIOXIN-CONTAMINATED MATERIALS

                                     by

                                F. Freestone
                    U.S. Environmental  Protection Agency
              Hazardous Waste Engineering Research Laboratory
                          Edison, New Jersey 08837

                                    and

                         R. Miller and C. Pfrommer
                               IT Corporation
                          Edison, New Jersey 08837


                                  ABSTRACT

The EPA Mobile Incineration System (MIS), a kiln-plus-secondary-combustion-
chamber-type system, had previously (1982-1983) been tested using PCB-contami-
nated liquids, and destruction and removal efficiencies of 99.9999% were
achieved.  Based upon this performance data, a project was initiated to
evaluate the technical, economic and administrative viability of on-site
incineration for the cleanup of dioxin-contaminated materials.  During
1984, the MIS was extensively modified to prepare for field use on dioxin-
contami nated soils, and these modifications were tested at Edison, NJ.
Additionally, permit documentation, operating and safety protocols, site
modification planning documentation, and a risk assessment for the planned
trial burn operating conditions were prepared.  An aggressive public
information activity was conducted by EPA's Region VII office to provide
an interested public with planning and permit-related information on the
project.

Simultaneously, laboratory and pilot plant studies were carried out to
establish optimum kiln conditions for decontaminating soils.  These studies
indicated that conditions necessary to decontaminate soils thermally
could be achieved in the kiln of the MIS.  Previous laboratory efforts by
others have established that dioxins could be destroyed using time and
temperature conditions achievable in the secondary combustion chamber of
the EPA system.  Therefore, the MIS was judged to have adequate operating
conditions for feeding dioxin-contaroinated solids and liquids into the
kiln, and would have a reasonable chance of success during a dioxin trial
burn.  The MIS was transported to the Denney Farm site in McDowell, MO,
in December 1984.
                                     531
-------
During 1985, extensive field shakedown activities were conducted during
January-March, and a trial burn on dioxin-contaminated liquids and solids
was conducted in April.  Results from that trial  burn indicate that
destruction and removal efficiencies exceeding 99.9999% were achieved
for 2,3,7,8 TCDD,  Additionally, the kiln ash and process wastewater byproducts
have been shown to meet "delisting" guidelines proposed by EPA's Office
of Solid Waste.

A 30-day field demonstration and a further trial  burn on RCRA- and
TSCA-designated materials were conducted during the Summer of 1985.  These
results will also be presented.
                                  Disclaimer

                    This  paper has been  reviewed  in
                    accordance with the  U.S. Environ-
                    mental  Protection Agency peer and
                    administrative review policies and
                    approved for presentation and publi-
                    cation.
                                     532
-------
         HAZARDOUS  WASTE  INCINERATION IN  INDUSTRIAL  PROCESSES:  CEMENT AND LIME KILNS

                           Robert  E,  Mournighan,  Harry  Freeman
                                Thermal Destruction  Branch
                     Hazardous Waste  Engineering  Research Laboratory
                           U.S.  Environmental  Protection Agency
                                Cincinnati, Ohio 45268

                                     James A.  Peters
                             Environmental Sciences Center
                                     Monsanto  Company
                                     800  N. Lindbergh
                                St. Louis, Missouri   63167

                                  Marvin R. Branscome
                              Research Triangle  Institute
                                     P.O. Box 12194
                     Research Triangle Park,  North  Carolina  27709


                                        ABSTRACT

    With more liquid wastes due to be banned  from land disposal facilities, expanding
hazardous waste incineration capacity becomes increasingly important.  At the same time,
industrial plants are increasingly seeking to find  new sources of lower cost fuel, spec-
ifically from the disposal of hazardous wastes with  heating value.

    The Hazardous Waste Engineering  Research Laboratory (HWERL) is currently evaluating
the disposal of hazardous wastes  in  a wide range of  industrial  processes.  The effort
includes sampling stack emissions at cement, lime and aggregate plants, asphalt plants
and blast furnaces, which use waste  as a supplemental fuel.  This research program is an
essential part of EPA's determination of the overall environmental impact of various
disposal options available to industry.  This paper  summarizes  the results of the HWERL
program of monitoring emissions from cement and  lime kilns burning hazardous wastes as
fuel.
INTRODUCTION

         With the passage of the Hazard-
ous and Solid Waste Disposal Act of 1984,
more categories of liquid hazardous
wastes will be banned from land disposal
facilities.  At the same time, energy
intensive industries are increasingly
seeking to find new sources of less ex-
pensive fuel.  Because many industrial
waste products can be readily used as
fuels and some hazardous wastes can be
economically processed and made into
fuels, a market based on hazardous wastes
has been developing in the U.S. [8],  [22].
If reprocessed waste liquids  do not con-
ain significant quantities of toxic metals,
halogenated materials, and PCBs, and  have
a high heating value, they can be econom-
ically substituted for coal,  coke, oil  or
natural gas in many industrial processes.
There are many examples of high temperature
industrial furnaces and processes which
already burn hazardous waste  as supplement-
al fuel: cement kilns (both wet and dry
proceses), lime and dolomite  kilns, clay
processing kilns, steel  blast furnaces,
phosphate rock calciners  and  dryers,  iron
                                              533
-------
 ore  dryers,  brick and tile tunnel  kilns,
 mineral  wool  furnaces and  glass melt
 furnaces.

     In particular,  there has  been  a great
 deal  of  interest  in the use of cement
 kilns for the disposal of  industrial
 wastes as supplemental fuel for several
 reasons: 1)  the production  process is
 highly energy intensive; fuel savings may
 translate into into a competitive  advan-
 tage; 2) kiln temperatures  are higher
 (270Q°F) and gas  residence  times are long-
 er (6-10 seconds) than those  encountered
 in most hazardous waste incinerators; 3)
 cement product quality is  relatively in-
 sensitive to addition of most waste trace
 impurities.

    The Alternative Technologies Division
                   of EPA's Hazardous Haste Engineering
                   Research Laboratory-Cincinnati  has been
                   investigating hazardous waste incineration
                   in cement kilns since 1977.  A number of
                   test programs on waste incineration by
                   other organizations at cement plants have
                   been conducted since 1974 in Canada, the
                   U.S., and Western Europe and are listed
                   in Table I.

                       The purpose of the tests, both in the
                   U.S. and abroad, was to determine the
                   effectiveness of the process in destroying
                   hazardous wastes, while providing signifi-
                   cant heat input into the process.  The
                   primary objectives of the test  program
                   were to document the performance of each
                   facility and compare it to performance
                   standards promulgated for incinerators.
                   Secondary objectives were to document any
               TABLE  I.  SUMMARY OF CEMENT AND LIME KILN TESTS
     Plant
Date
Process
APCD
Fuel
Wastes
1.
2.
3.
4.
St. Lawrence Cement
Hlssissauga, ON
C3, 12]
Stora Vika
Sweden [1, 2]
Marquette Cement
Oglesby, IL [9]
San Juan Cement
Puerto Rico [173
1974
197S-76
1978
1981
1981-82
Dry
Wet
Wet
Dry
Dry
ESP
ESP
ESP
ESP
ESP
Fuel oil
Fuel oil
Coal
Coal
Coal
Waste oil .
Chlorinated
organics
Chlorinated
organics,
PCB's,
Freon 113
Hydrocarbon
solvents
(<5% chlorine)
Chlorinated
organics
5.  Alpha Cement            1982       Wet          ESP      Coal
    Cementon, NY [15]

6.  General Portland        1982       Dry          Baghouse Coal
    Lebec, CA  [10]

7.  General Portland        1983       yet          ESP      Coal
    Paulding, OH  [4]
                                           Solvents
                                           Hydrocarbon
                                           solvents

                                           Hydrocarbon
                                           solvents
                                           Freon 113
8.  Lone Star Industries    1983       Dry          ESP
    Oglesby, IL  [5]
                                 Coal/Coke  Hydrocarbon
                                           solvents
                                           Freon 113
9.  Rockwell Lime           1983       Lime         Baghouse Coke       Hydrocarbon
    Hanltowoc, WI  [7]                                                  solvents
                                           534
-------
 changes in conventional pollutant emis-
 sions that could be attributed to burning
 hazardous waste and identify new problems.

     This paper summarizes the data from
 these tests on: waste destruction and
 removal  efficiency (ORE), hydrogen
 chloride (HC1) emissions, and changes in
 conventional  pollutant emissions from
 kilns when burning hazardous wastes as
 supplemental  fuel.  From the RCRA stand-
 point, most important of these concerns
 are particulate matter, (ORE), and HC1
 results.  Carbon monoxide, NOX, total
 hydocarbons (THC), and SOX will also be
 discussed.  For reference, a generalized
 diagram of a  cement kiln burning waste as
 a  fuel  is shown in Figure 1.

 PARTICULATE MATTER
    Most  of  the  tests  conducted  at
 kilns  using  electrostatic precip-
 itators exhibited  small  or little
 change in particulate  emissions  when
 burning hazardous  wastes.  A  summary
 of  the average data  for  each  test Is
 listed in Table  II.  The major ex-
 ceptions  are tests during which  there
 were either  process  equipment mal-
 functions [5, 12]  or high amounts of
 chlorine  being fed to  the kiln (1, 2,
 12).   The latter tests have led  to the
 conclusion that  the  higher chlorine
 input  to  the kiln  will lead to greater
 particulate  emissions.

    In a  review  article,  Weitzman [21]
 discussed particulate emissions  from
 cement kilns burning chlorinated
 wastes as fuel.  Data from the St.
 Lawrence  Cement  and  Stora Vika tests
 were re-examined,  compared, and were
 plotted as a function of  chlorine feed
 rate.  The resulting graph  showed a
 linear increase  in particulate emis-
 sions  as  a function  of chlorine input
 to the process.  In  the  article, he
 hypothesized that  the hydrogen
 chloride  combustion  product reacts
with the  potassium,  sodium, and cal-
cium compounds in the hot kiln to
 form the  respective  chlorides and the
volatilized chloride salts were
carried by the hot gas to the cooler
end of the kiln where they condensed
at about 800°C to  form a  fine partic-
ulate.  The higher chloride content  of
the particulate,  with concomitant
changes in dust resistivity, lowered
the electrostatic precipitator (ESP)
 collection  efficiency.  This  resulted
 in  significantly  increased dust emis-
 sions  during  the  two chlorinated or-
 ganic  waste firing tests.  Based on
 the results of  these tests and the
 hypothesis  of very fine particles,
 Weitzman  concluded that a kiln using a
 fabric filter for particulate removal
 would  encounter a similar deteriora-
 tion in efficiency.

     The particulate emissions from
 the kiln  tested at San Juan Cement
 [17],  equipped  with a fabric filter
 dust collecting system, did not ex-
 hibit  an  increase with progressively
 larger chlorine input as shown in
 Figure 2.   For  comparison, the linear
 regression  plot developed from the St.
 Lawrence  Cement and Stora Vika data
 are also  displayed.  It can be seen
 that,  not only was more organic chlo-
 rine fed  to the kiln at San Juan
 Cement than more  of the previous
 tests, the  fabric filter system was
 able to maintain a high collection
 efficiency  regardless of the amount of
 chlorine  fed to the kiln.

     Thus,  the addition of substantial
 amounts of organic chlorine in waste •
 fuel may contribute to lowering ESP
 collection efficiency, but has no
 effect on control  efficiency of fabric
 filters. [14]

 DESTRUCTION AND REMOVAL EFFICIENCIES
 FORPRINCIPAL ORGANIC HAZARDOUS
 CONSTITUENTS

     Incinerators  and,  in the near
 future, other facilities  burning  haz-
 ardous wastes, must perform at a  mini-
mum to the standards  described in  the
 Federal regulations [20].   Specifi-
cally,  with  regard to the  Principal
Organic Hazardous  Constituents
 (POHC's),  the  facility  must achieve a
destruction  and  removal efficiency
 (ORE) of 99.99%  for each designated
POHC.  The ORE is  determined  as
follows [20]:
           ORE =
           where:
                       -  wout
            X 100S
                      M
                       in
W-jn = mass feed
rate of a specific
POHC in waste feed
stream.
                                        535
-------
Hot Cost. Mma*Y *
"     secondary AU
     Waste Fuel, Air
                                      Ground Feed (Dry)
                                                                tttt
KHM560W     Soiias  \   Gas
                         r Clinker 11.750TPD)

                                       Process Schematic
Hot Coal
•ndAlf
                                                 WhlrilngVano
                    "Compf eased Air
                    -Waste Fuel
                                                  Coal and Air
                                                  Waste Fuel
                                                    Nozzle
                                                            Burner Cross Section
               Burner Side View
          Figure 1,   Generalized Diaqran of a Cement Kiln Burning
                     Hazardous Haste
                                          536
-------
  ST    2-°"
  UJ
  z
t/J Q
-------
                                                    TABLE II.   SUMMARY  OF PARTICULATE  EMISSION DATA
                         Plant
Test Condition
en
OJ
CD
    Particulate Emissions
gr/scf      Ib/hr      Ib/ton
Chloride Input to Kiln
       (Icg/Mg)
St. Lawrence





Rockwell Lime

Stora Vika








Marquette

Alpha Cement

San Juan

General Portland
Paulding
Lone Star

Chlorinated aliphatlcs1
Chlorinated aromatics
PCB's
Baseline
Lubricating oil
Baseline
Waste
Baseline
Aliphatics
Baseline
PCB's
Baseline
Chlorophenols
and Phenoxyacids
Basel ine
Freon 113
Baseline
Waste solvents
Baseline
Solvents
Baseline
Wastes
Basel ine
Wastes
Baseline
Waste*
Baseline
0.21
0.286
0.078
0.038
0.064
0.107
0.016
0.013
0.039
0.009
0.024
0.011
0.058

0.014
0.062
0.022
0.014
0.093
0.041
0.060
0.043
0.041
0.030
0.030

0.17
123
45
44
21
83
139
2.2
2.0
21
4.7
12.7
5.9
30.9

7.7
33.3
11.7
58
80
44
53
22.4
21.7
18.9
19.6

116
3
1.1
1.1
0.5
0.7
1.1
0.26
0.24
0.88
0.21
0.53
0.25
1.36

0.34
1.39
0.49
21
20.8
0.8
1.1
0.66
0.64
0.65
0.64

2.0
4.0
5.5
2.5
-
-
„
2.7
-
4.4
0
3.6
0
0.95

0
1.7
0
1.1
-
-
-
5.5
-
2.2
0.2
1.2
0.2
                         1Process upset during test
                         2  ESP malfunctioned during test
-------
                       Wout = mass flow
                       rate of the same
                       POHC in exhaust
                       emissions to the
                       atmosphere.

 Table  III  summarizes  the  results  of the
 tests  which  revealed  reliable ORE data.

    One of  the  first tests  to examine the
 ORE of hazardous  waste in  cement  kilns was
 conducted  at a Canadian plant,  St.
 Lawrence Cement.   The wet-process kiln was
 normally fired with fuel oil  and  the
 exhaust gas was cleaned by an ESP.   Three
 tests  were conducted  with  batches of waste
 solvents labeled  as chlorinated ali-
 phatics, chlorinated  aromatics, and PCBs.
 The chlorinated aliphatics were composed
 primarily  of chloropropane, chloropropene,
 1,2-dlchloroethane, 1,1,2tr1 chloroethane,
 and chlorobenzene. The chlorinated
 aromatics  batch was primarily chloro-
 toluene, and the  waste  labeled PCBs was
 about  50 percent  PCB, 28.5 percent  chloro-
 toluene, and a mixture of  chlorinated
 aliphatics.  Destruction efficiencies were
 calculated conservatively  by not  sub-
 tracting or correcting for the  background
 levels from the baseline test or  inter-
 ferences (contamination) on the control
 blanks.  In addition, a conservative col-
 lection efficiency of 80 percent  was  used.
 The  result was  a  maximum estimated  con-
 centration of  50  ppb  for all  of the waste
 fuel tests, and "because of the high  and
 uncertain  background  levels,  the  estimate
 of  emissions is higher than the actual
 levels  in  the emissions" [12].  The DREs
 were calculated on total chlorinated  or-
 ganics  basis rather than on  the actual
 analysis of specific  compounds.   The
 reported OREs  for  wastes with mostly
 chlorinated organics  was >99.99 pecent,
 >99.989 percent for chlorinated aromatics,
 and >99.986 percent for PCB mixtures  [12].

    A test similar to the one conducted at
 St. Lawrence was  conducted  in Sweden  at a
wet process kiln  in Stora Vika.   Short-
 term tests were conducted with  chlorinated
 aliphatics (100 hr),  PCBs  (24 hr), chloro-
 phenols and phenoxy acids  (12 hr) and
trlchlorotrifluoroethane (Freon 113}  (3
 hr).  Long-term tests were  conducted with
the chlorinated organics (1.5 months) and
 PCBs (about 1 month).

    Results are listed in Table III.
During both the short- and  long-term tests
with chlorinated aliphatics, none of the
 waste fuel's major components  were detec-
 ted in the stack gas.   Based on  the detec-
 tion limit, the ORE of  methylene chloride
 exceeded 99.995 percent and the  ORE of
 trichloroethylene exceeded 99.9998 percent
 Cl], C2].

      No PCBs were detected in  the stack
 gas during either the short- or  long-term
 test and yielded a ORE  exceeding 99.99998
 percent for PCB,  Similarly, no  chloro-
 phenols or phenoxy acids were  detected in
 the stack gas;  DREs were calculated to be
 greater than 99.99999 and 99.99998 per-
 cent, respectively. Detectable  levels of
 Freon 113 were  found during one  experiment
 with this waste fuel and yielded a ORE of
 99.99986 percent [1], [2].

      The test at the General  Portland's
 Los Robles dry  process  kiln included dich-
 loromethane, 1,1,1-trichloroethane, 1,3,
 5-trimethylbenzene, and xylene as POHC's
 in  the waste fuel.  The results  show a ORE
 of  99.99 percent  or greater for  3 of the 4
 compounds and again, the calculations are
 conservative.   No  corrections were made
 for baseline levels or  for the con-
 tribution from  ambient  air.  No trim-
 ethyl benzene was  detected; therefore, the
 ORE for this  compound is likely to have
 exceeded 99.95  percent  [10].

    ,  In  the  tests  at General  Portland's
 Paulding,  Ohio  plant, ORE's for the most
 part  were greater  than 99.99  percent and
 in  some  cases higher than 99.999 percent.
 DRE's  less than 99.99 percent for methyl
 ethyl  ketone and methylene chloride were
 related  to sample  contamination problems
 and do  not represent poor incineration
 performance  of  the  kiln.  No  problems  with
 contaminants were seen with the 1,1,1-
 trichloroethane and Freon 113 results,
which demonstrated ORE's of 99.999  percent
 or  greater.

     The  results of the test  burn at  Lone
 Star showed  similar contamination problems
with rnethylene chloride.  However,  the
ORE's  for  other compounds were  all  greater
than 99.99 percent on conservative  basis;
 i.e., no correction for blanks.  In
addition to the POHC's,  several other  com-
 pounds were tracked for ORE.  Styrene,
ethyl benzene, orthoxylene, and benzald-
ehyde all  showed DRE's  greater  than
99.998%, which indicates that excellent
destruction conditions  existed  in the  kiln
[14].
                                          539
-------
                        TABLE III.   SUMMARY  OF  ORE  DATA
   Plant
Waste  Component
Destruction Efficiency
St. Lawrence  Cement
Stora Vika
San Juan Cement
Los Robles
(General Portland)
Paul ding
(General Portland)
Qglesby
(Lone Star)
Rockwell Lime
 Chlorinated  aliphatics
 Chlorinated  aromatics
 PCB's

 Methylene chloride
 Tri chloroethylene
 All chlorinated hydrocarbons
 PCB
 Chlorinated  phenols
 Phenoxy acids  •
 Freon 113

 Methylene chloride
 Carbon tetrachloride

 Methylene chloride
 1,1,1-Trichloroethane
 1,3,5-Trimethylbenzene
 Xylene

 Methylene chloride
 Freon 113
 Methyl ethyl ketone
 1,1,1-Trichloroethane
 To 1uene

Methylene chloride
 Freon 113
Methyl ethyl ketone
 1,1,1-Trichloroethane
 Toluene

Methylene chloride
Methyl ethyl ketone
 1,1,1-Trichloroethane
Tri chloroethylene
Tetrachloroethylene
Toluene
     >99.990
     >99.989
     >99.986

     >99.995
     >99.9998
     >99.988
     >99,99998
     >99.99999
     >99.99998
     >99.99986

   93.292-99.997
   91.043-99.996

     >99.99
      99.99
     >99.95
     >99.99

   99.956-99.998
     >99.999
   99.978-99.997
   99.991-99.999
   99.940-99.988

   99.94-99.99
     99.999
   99.997-99.999
     >99.999
   99.986-99.998

   99.9947-99.9995
   99.9992-99.9997
   99.9955-99.9982
   99.997-99.9999
   99.997-99.9999
   99.995-99.998
                                    540
-------
     Waste fuel  containing methylene
 chloride, methyl  ethyl  ketone, 1,1,1-
 trichloroethane,  trlchloroethylene,
 tetrachloroethylene,  and toluene was
 burned during the test  at Rockwell  Lime.
 The  average  ORE results are listed  in
 Table  III and show an average of 99.99
 percent  or greater.   The reported DREs are
 again  conservative because no blank cor-
 rections were applied.   DREs  exceeding
 99.99  percent were consistently  obtained
 for  all  POHCs.

     The  test  at San Juan used methylene
 chloride, chloroform, and carbon tetra-
 chloride as  the designated POHCs in a
 waste  fuel with very  high chlorine  con-
 tents  (6.5-35.1 percent).  This  test
 showed a phenomenon seen at other cement
 kiln tests:   measurable emission rates of
 the  POHCs during  the  baseline test  when no
 waste was burned.  Blank samples showed no
 contamination problems;  therefore,  the
 source of the POHCs during the baseline
 was  unexplainable.  The reported DREs for
 this test (Table  III) are low compared
 with other test results.  The test  report
 concluded that  lack of  air atomization of
 the waste fuel  and the  difficulty of
 incinerating  highly chlorinated  monocarbon
 compounds  contributed to the  low DREs.
 Above normal  free  lime  content of the
 clinker  product and removal of chloride in
 the  clinker  instead of  the waste dust also
 suggest  that operating  difficulties  were
 experienced  in  the kiln.   However,  the
 detection  of POHCs during the  baseline
 make the ORE results difficult to inter-
 pret.  If  the measured  POHCs  originated
 from sources other than  burning  waste
 fuel, the  actual DREs may have been  higher
 than those measured [17].

    The  San Juan test, when viewed with
 results  from other cement kiln tests,
 suggests  that burner design, waste com-
 position,  and the kiln's  operation can
 have a significant effect  on ORE  perform-
 ance.  The extent of their effect is  in-
 conclusive because of baseline POHC
 emissions.

 CARBON MONOXIDE AND TOTAL  HYDROCARBONS

     CO  emissions are not  affected by  the
 burning  of waste fuels.   Weitzman noted
 that changes in CO can be  indicative of
 flame quenching, improper burner  adjust-
ment, or other  imbalances  in the  flame
 [21].  Peters and Mournighan [14] found
 that any process change can create a  sign-
 ificant CO excursion.  Examples of process
 changes include  alteration of:  the
 primary air/fuel  ratio, secondary air
 flow,  solid fuel  feed rates, and product
 cooler exhaust  flow.  These effects were
 observed during  the Paulding test during
 a  kiln upset  with coal as the only fuel.
 Over the course  of the event, several
 process parameters were changed and large
 swings  in CO  concentrations (as well  as
 other  monitored  gas concentrations) were
 observed.   The CO results for Stora Vika
 also show a wide  range of 50-1500 ppm for
 both the baseline and waste fuel burns.
 Table  IV lists CO and total hydrocarbons
 for each  of the tests.

    The CO results for Lone Star appear
 to be  the  most consistently low.  The kiln
 was operated  with a higher Oj input (to
 aid in  drying wet coal),  which apparently
 resulted in consistently low levels of
 THC, CO and SOg with increased NOX con-
 centrations.  The Los Robles kiln was also
 very stable during three waste firing
 tests  in which the maximum CO was 100 ppm.

     The CO results indicate that the
 feeding  of  waste fuel  to  the kiln does not
 significantly affect CO concentrations.
 Process  stability apparently has a greater
 effect  on  CO  concentrations.  CO monitors
 on kilns provide the operator with a
 measure  of  this stability.  The CO meas-
 uring equipment in general use by the
 industry  is not as sensitive as the equip-
 ment used by the test  crews and is not
 capable  of  seeing CO concentrations much
 below 1000 ppm.   Installation of more
 sensitive CO monitors  would help operators
 control combustion conditions more effect-
 ively.

     Total hydrocarbons  (THC)  increased
 during waste fuel combustion at 3 tests,
decreased at 2 tests,  and  remained the
 same at another.  The  consistently low
 THC concentrations for several  different
 types of compounds at  Stora Vika suggest
that kiln operation  may affect THC con-
 centrations more than  the  waste type  does.
The levels at Rockwell Lime,  which  attain-
ed consistently high DRE's,  averaged  3.5
 ppm.  Similarly, the Lone  Star kiln main-
tained consistently  low levels  of THC
 between 2.5 and 5 ppm.  These data  for
different types of kilns and  wastes fuels
 indicate that a well-controlled kiln  can
operate with average THC concentrations
below 10 ppm.
                                           541
-------
              TABLE  IV.  SUMMARY OF THE THC AND CO CONCENTRATIONS
Site
San Juan
Los Rabies
General Portland
Paul ding
General Portland
Lone Star
Rockwel 1 Lime
Stora Vika


Test condition
Waste burn
Baseline
Waste burn
Baseline
Waste burn
Baseline
Waste burn
Baseline
Waste burn
Baseline
Chlorinated aliphatics
Baseline
PCBs
Baseline
Chlorophenols/phenoxy acids
Baseline
Freon
Basel i ne
THC (ppm)a
12.7
8.3
c
4
21
10
5
2.5
3.5
8.2
<10
<10
<10
<10
10
10
<10
<10
CO (ppm)
24 - 738°
25 - 349&
25 - 100
10 - 618
190
212
24 - 49b
35 - 40&
599
477
300 - 1500
1,500
100 - 1500
100
50 - 500
50
100 - 500
100
a  Expressed as ppm methane unless otherwise  noted.
b  Range of test averages.
c  Not measured.
                                  542
-------
 NITROGEN OXIDES
 ACID GASES
      In  cement  and  lime  kilns,  there  is  a
 direct  functional  relationship  between NOX
 emissions  and secondary  combustion  air
 flow  and temperature.  There  is  usually
 little  or  no  NOX  formed  from  nitrogen in
 fuel  relative to  NQX  from combustion  air,
 Kilns are  generally operated  as  "poor"
 combustors.   When compared to boilers and
 incinerators, kilns have a reduced  primary
 air/fuel mix, poorer  secondary  air  mixing,
 and often  require a confined, narrow  flame
 configuration for good product  quality.
 Typical  NOX concentrations in the stack
 gas can  range from  200 to 1500  ppm, de-
 pending  on combustion conditions. NOX
 levels are very operator-dependent; con-
 tinuous  monitoring  has shown  that at  some
 kilns NOX  levels vary widely  from hour to
 hour, while others  can maintain  steady NOX
 concentrations  for  days  at a  time.  These
 and other tests conducted  under  the USEPA
 NOX Reduction Program [18] indicate that
 waste firing  has little  to do with NOX
 variations.   The amount  of excess air,
 secondary air temperature,  and primary
 air/fuel ratios are the  main  determinants
 in kiln  NOX emissions [14].

    Table V present a summary of the  tests
 for which NOX emissions  data  is  available.
 At both Marquette Cement  and  San Juan
 Cement a significant decrease was observed
 during waste  firing.  This  may have been
 the result of lowered oxygen  input more
 than  waste firing.  At the  General Port-
 land  - Los Robles Plant, the  NOX emissions
 decreased slightly  the change was not
 statistically significant.  A significant
 increase was  observed at Rockwell Lime,
 but the change was  intended so that the
 NOX emissions increased during waste
 firing.

     A definite relationship  between 02,
 NOX,  and S02  in calcination kilns has been
 shown [18].   NOX emissions  decrease with
decreasing Q£ input and lower secondary
 air temperatures, while SOg emissions in-
crease.   For  instance, a 38%  reduction in
NOX due  to Q£ lowering can cause a 47%
 increase in S02.  Continuous monitoring at
Rockwell Lime and General Portland-
Paul ding illustrated that NOX and $02
change inversely, as shown  in Figure 3.
Moreover, the NOX/S02  change  is followed
within 5-15 minutes by a CO spike, the
 kiln's indicator of a  momentary process
 instability.  As shown in Figure 3, this
event recurred at uneven intervals.
     The process  materials in cement kilns,
 by their alkaline  nature, provide  built-in
 control for acid gases  such as  S02 and
 HC1.  In a  well  operated cement kiln with
 medium-to high  feed  alkalinity, S02 will
 be scrubbed in  the kiln environment and be
 adsorbed by the  alkaline dust of the air
 pollution control  device to control effi-
 ciencies of greater  than 90%.   For HC1
 emissions,  a removal of efficiency greater
 than 99% was observed at San Juan Cement,
 where high  amounts of chlorine  were added
 to a kiln processing relatively low
 alkaline materials [17],  Tables VI and
 VII  present  the  emission results for S02
 and  HC1, respectively, for all tests.

      As shown in Table VII, HC1 results
 are  quite variable from test to test.  For
 the  most part, however, HC1  emissions are
 low  and less that the standard set for
 incinerators [19].  Notable exceptions are
 the  Marquette and Lone Star tests.  The
 results of  the Lone Star tests are suspect
 because of ESP malfunction.

      Haste  fuels typically do not contain
 much sulfur  and  thus the partial sub-
 stitution of waste fuel  for  fuel oil  or
 coal  will reduce sulfur input to the kiln,
 resulting in less S02 to being formed.
 During  the Marquette Cement  test,  where a
 13%  energy substitution was  fired, S02
 stack gas concentrations were reduced from
 93 ppm  to 18 ppm.  At the Los Robles
 cement  plant, no significant difference
 in SC>2  emissions was noted.   S02 emissions
 increased significantly  at San Juan
 Cement,  from 280 ppm to 450  ppm, probably
 for  two  reasons:  the NOX emissions were
 reduced—indicative of lower 02 input—
 and  the  high  rate of chlorine input may
 have caused  a competitive acid gas
 "scrubbing"  situation-HCl  was  preferent-
 ially absorbed over S02-  The kiln
 operator's choice of Og  input will  have a
 greater  effect on SC>2 emissions than
 addition of  a waste fuel.  The test results
 indicate that it is possible,  and  probably
preferable to operate a  cement  kiln with
 less  than 50 ppm S02  in  the  stack  gas,
when  burning hazardous waste.

     At Rockwell  Linie, however,  a differ-
 ent  mode of  operation was  required;  S02 is
 not  intentionally absorbed because the
presence of  sulfur  in their  lime product
 is undesirable.   No significant difference
was  observed for SOg  emissions,  which
                                          543
-------
1250
             1030
1100
1130      1200      1300

     TIME OF DAY
1230
1330
1400
1430
                       Figure 3.   Illustration of Interrelationship Between
                                 NOX, S02, and CO Emissions in a Lime  Kiln
-------
TABLE V.  SUMMARY OF NOX  EMISSIONS DATA
Test
Marquette
San Juan
Los Robles
en
en
Paulding
Lone Star
Rockwell Lime
Condition
Waste
Baseline
Waste
Baseline
Waste
Baseline
Waste
Baseline
Waste
Baseline
Waste
Baseline
(Ib/hr)
275
404
31,3
60.4
304
444
174
140
472
371
134
N0x
(Ib/ton)
4.6
6.7
0.9
1.8
5.3
8.2
6.0
4.6
8.6
6.9
15.8
(ppm)
544
920
68
136
486
680
478
371
814
620
446
386
Comments
Significant decrease 1n NOX
Significant decrease in NOX
No significant decrease in NOX
Significant increase in NOX
Significant increase in Nox
Significant increase in NOX
-------
 remained between 500-600 ppm regardless of
 waste firing.

     The data indicate that effects of
 chlorinated waste combustion on KC1  emis-
 sions are quite variable from kiln to
 kiln.  Variations in HC1 emissions with
 time can also exist within the same kiln,
 depending on operating conditions.
 Sporadic variations in HC1  can be expected
 because of cyclic temperature variations,
 feed and fuel changes that affect the kiln
 temperature profile, and chloride loading,
 and variations in recycle dust  rate  and
 composition.

     The continuous monitoring of SC>2,  NOX,
 D£ and CO indicated that the concentration
 of these compounds in the stack gas  are
 interdependent and combustion operation
 can be optimized.  Use of continuous mon-
 itors proved to be an excellent process
 monitoring tool because of the increased
 Information feedback on combustion con-
 ditions.  Although continuous monitoring
 equipment requires constant  attention  and
 maintenance, their worth as  a process
 control  tool  may prove  quite valuable.

 CONCLUSIONS

      The foregoing examples  of  burning
 liquid  hazardous wastes as a fuel  supple-
 ment  in cement and lime kilns  in  Canada,
 Sweden,  Illinois,  Puerto Rico,  California,
 and Wisconsin have demonstrated that there
 is  little or no significant  change in
 emissions of conventional  pollutants and
 that  process  operation  changes  have  a
 greater effect on  changes  in emissions
 than  the  partial  substitution of  hazardous
 waste fuel.

      The  conventional pollutants  which
 show  little or no  significant change and
 are independent of hazardous-waste-as-fuel
 burning are CO,  THC,  NOX and S02.  Previ-
 ous studies and  evaluation of continuous
monitoring demonstrated  that emission
 rates of  these  compounds are more a
 function  of combustion  air flows  and fuel
 rate  control than  any of the other para-
meters.

      Particulate emissions from kilns
equipped with electrostatic precipitators
may be expected  to  increase with  in-
creasing chlorine  loading in the  kiln.
Those kilns equipped with fabric  filters,
or baghouses, will not experience any
significant change in emissions.
      HC1 emissions from cement and lime
kilns meet emission limits set for incin-
erators.  However, variations in test
results warrant further investigation into
this issue.

     Destruction and Removal  Efficiencies
(ORE) of POHC's were, for the most part,
greater than the incinerator  regulatory
requirement of 99.99%.  It appears that
there are enough variations in kiln
operation from plant to plant, that var-
iations in ORE would be expected,  but a
well operated facility should have little
trouble in demonstrating compliance with
the ORE requirement.
                                           546
-------
                                                          TABLE VI.   SUMMARY OF S02 EMISSION DATA
en
*»
•v)
Test
Alpha Cement [19]
Marquette
San Juan
Los Robles
General Portland
Paulding
General Portland
Lone Star
Rockwell Lime
Condition
Waste
Basel ine
Waste
Baseline
Waste
Baseline
Waste
Basel ine
Waste
Baseline
Waste
Baseline
Waste
Baseline
(Ib/hr)
58.5
138
11.5
57.1
264
170
21.7
23.7
297
526
6.7
5.6
97
149
SO? Emissions
(lb/ton)
1.1
2.7
0.19
0.95
8
5
0.36
0.38
6.8
17.2
.12
.10
11.4
17.5
(ppm)
33
78
18
93
450
279
27
27
265
636
19
7
596
553
Comments
S in coal = 2.*6%; S in waste = 0.2%
S during waste burn = 2.0%
S in waste = 0.08%
Significant decrease
S in fuel oil = 2.15%
Significant increase
S in coal = 0.43%
No significant difference
S in coal = 4.3%
Significant decrease
S in coal /coke = 2.7%
Significant increase
S in coke = 4.15
No significant difference
-------
                           TABLE VII.  SUMMARY OF HC1  EMISSION DATA
        Test
Condition
   HC1  Emissions
(Ib/hr)     (Ib/ton)
Alpha Cement
Marquette
San Juan
Los Robles
(General Portland)
Paul ding
(General Portland)
Lone Star
St. Lawrence
Rockwell L1me
Waste
Baseline
Haste
Baseline
Waste
Baseline
Waste
Baseline
Waste
Baseline
Waste
Baseline
Waste
Baseline
Waste
Baseline
5.8
2.4
115
190
0.79
0.19
1.03
0.55
4.62
1.25
25.3
2.9
1
1
0.44
0.20
0.11
0.05
1.9
3.2
0.02
0.006
0.015
0.007
0.16
0.04
0.46
0.054
0.02
0.02
0.05
0.02
REFERENCES
1.  Ahling, B.  "Combustion Test with
    Chlorinated Hydrocarbons in a Cement
    Kiln at Stora Vika Test Center."
    Swedish Water and Air Pollution
    Research Institute, Stockholm,
    Sweden.  March 1978.

2.  Ahling, B.  "Destruction of Chlo-
    rinated Hydrocarbons in a Cement
    Kiln." Environmental Science and
    Technology, 13 (11):  1377-1379,
    November 1979.

3.  Berry, E. E., L. P. MacOonald, and
    0. J. Skinner.  "Experimental Burning
    of Waste Oil as a Fuel  in Cement Man-
    ufacturing."  Environment Canada,
    EPS 4-WP-75-1.  June 1976.

4.  Branscome, M. R. et al.  "Evaluation
    of Waste Combustion in Cement Kilns
    at General Portland, Inc., Paulding,
    Ohio."  Research Triangle Institute,
    draft report on EPA Contract 68-02-
    3149, Work Assignment 11-1.  March
    1984.
                   Branscome, M. R. et al.  "Evaluation
                   of Waste Combustion in a Dry-Process
                   Cement Kiln at Lone Star Industries,
                   Oglesby, Illinois," Research Triangle
                   Institute draft report on EPA Con-
                   tract 68-02-3149, Work Assignment
                   11-1, December 1984.

                   Compilation of Air Pollutant Emission
                   Factors, Third Edition.  U.S. Envi-
                   ron Environmental Protection Agency,
                   Research Triangle Park, N.C.  Pub-
                   lication No. AP-42, July 1979.

                   Day, D. R., and L. A. Cox.  "Eval-
                   uation of Hazardous Waste Incin-
                   eration in Lime Kilns at Rockwell
                   Lime Company."  EPA 600/S2-84-132;
                   U.S. Environmental Protection Agency,
                   Cincinnati, Ohio.

                   Hazelwood, D. L., F. J. Smith and
                   E. M. Gartner.  "Assessment of Waste
                   Fuel Use in Cement Kilns."  EPA-600
                   /S2-82-013, U.S. Environmental
                   Protection Agency, Cincinnati, Ohio,
                   October 1982.
                                          548
-------
 9.  Higgins, G. M,, and A. J.
     Helmstetter.  "Evaluation of Hazard-
     ous Waste  Incineration in a Dry
     Process Cement Kiln."  Proceedings of
     the Eight  Annual Reserach Symposium:
     Incineration and Treatment of Hazard-
     ous Waste.  EPA-600/9-83-003. April
     1983.  pp. 243-252.

10.  Jenkins, H. C., 6. Murchinson, R. C.
     Adrian, R. D. Fletcher, and 0. C.
     Simerotb.  "Emissions from Los Robles
     Kiln."  State of California Air Re-
     sources Board, Engineering Evaluation
     Report C-82-080.  August 1983.

11.  Lauber, J. D.  "Burning Chemical
     Wastes as  Fuels in Cement Kilns,"
     JAPCA 32(7),771-776, July 1982.

12.  MacDonald, L. P., D. J. Skinner,
     F.  J. Hopton, and G. H. Thomas.
    ."Burning Waste Chlorinated Hydro-
     carbons in a Cement Kiln,"  Fisheries
     and Environment Canada, EPS 4-WP-77-2
     (also EPA/530/SW-147C).

13.  Mix, T. W. and B. L. Murphy.   "Risks
     Associated with Waste-Fuel Use in
     Cement Kilns."  Environmental Pro-
     gress, 3 (1):  64-70, February 1974.

14.  Mournighan, R. E., Peters, J, A.,
     and Branscome, M. R., "Effects of
     Burning Hazardous Waste in Cement
     Kilns on Conventional Pollutant Emis-
     sions," 78th APCA Meeting, Detroit,
     MI, June 1985.

15,  Nesselbeck, E. R., "Baseline and
     Solvent Fuels Stack Emission Tests,"
     Prepared for Energy and Resources
     Recovery Corporation, Albany, NY,
     January 1982.

16.  Peters, J. A., T. W, Hughes,  and
     R.  E. Mournighan.  "Evaluation of
     Hazardous Waste Incineration in a
     Cement Kiln at San Juan Cement."
     Presented  at the Ninth Annual
     Research Symposium on Land Disposal,
     Incineration and Treatment of Hazard-
     ous Waste, May 2-4, 1983, Ft.
     Mitchell,  Kentucky.
 17.   Peters, J. A., T. W. Hughes,  J.  R.
      McKendree, L. A. Cox, and B.  M,
      Hughes.  "Industrial Kilns Processing
      Hazardous Wastes-San Juan Cement
      Company Demonstration Program."
      EPA-600/S2-84-129, U.S. Environmental
      Protection Agency, Cincinnati, Ohio.

 18.   Tidona, R. J., W. A. Carter,  H. J.
      Buening, S. S. Cherry,  and M. N.
      Mansour.  "Evaluation of Combustion
      Variable Effects on NOX Emissions
      from Mineral  Kilns."  EPA-600/S7-
      83-045, U.S.  Environmental  Protection
      Agency, Research Triangle Park, N.C.,
      November 1983.

 19.   Title 40 Code of Federal  Regulations,
     Part 261.

20.  Title 40 Code of Federal  Regulations,
     Part 264,343.

21.  Weitzman,  L.   "Cement Kilns as Haz-
     ardous Waste  Incinerators."  Environ-
     mental Progress,  2  (1):   10-14,
     February 1983.

22.  Worthy, W.  "Hazardous Waste:   Treat-
     ment Technology Grows."  Chemical  and
     Engineering News, 60 (10):  10-16,
     March 1982.
                                                               Disclaimer

                                              This paper  has  been reviewed In
                                              accordance  with the U.S. Environ-
                                              mental  Protection Agency peer  and
                                              administrative  review  policies and
                                              approved for presentation and  publi-
                                              cation.
                                          549
-------
       COMBUSTION FUNDAMENTAL STUDIES
                         INCINERATION
   FOR HAZARDOUS WASTE
         Kun-chieh  Lee,  Wei-yeong  Wang,  and  Joe  E.  Neff
          Engineering  and  Technology  Services  Division
              c  1985  Union Carbide Corporation
                  South Charleston, WV  25303
                            ABSTRACT

     This paper discusses Union Carbide's fundamental research
program being conducted to better understand the spray com-
bustion of liquid chemical wastes.  The program focuses on the
study of turbulent mixing phenomena in the flame zone and con-
stitutes a fuel atomization study and a 3-million Btu/hour
research combustor swirl combustion study.  The installation
and initial start-up troubleshooting were completed for both
facilities in December, 1984.  The actual test program is
on-going.
INTRODUCTION

     In the past, the design
and operation of a chemical
waste incinerator was mostly an
art and not a science.  The
performance of an incinerator
was basically determined by
visual observation of the flame
pattern and color.  Trial and
error, and each individual
company's past experience were
the sources of information.
Although this approach has been
satisfactory, better scientific
understanding of the design and
operating principles is needed
in order to be sure that the
stringent regulatory requirements
for the combustion of hazardous
waste can be met.

     From our many years ex-
perience in operating various
chemical waste incinerators, we
have learned that the per-
formance of an incinerator is
determined basically by the
design and the operation of the
system and not by kinetics.  In-
cineration can be a very reliable
operation if the system is de-
signed and operated properly.
This observation is also sup-
ported by several trial burn
test related programs (4, 13)
conducted in the past few years.
Those tests also found that the
incinerability of each indivi-
dual organic compound is really
not an important factor to be
considered in evaluating the
performance of an incinerator
under flame conditions.  If an
incinerator is designed and
operated with the capability to
destroy one compound, which is
reasonably stable (thermally),
the chances are that it can
destroy other compounds.  If one
compound fails to meet a 99.99%
destruction and removal (ORE)
for a good incinerator, the
chances are that either the
                                550
-------
compound is a combustion inter-
mediate of other compound(s), or
the compound has a very low
concentration (between TOO to
1000 ppm) in the waste feed.

    Union Carbide has contri-
buted to the advancement of
knowledge in flameless thermal
oxidation (11).   However, in the
real world, the  factors which
really impact on organic de-
struction in a liquid injection
incinerator are  those things
which happen in  the flame zone
and not what happens in the
post-flame zone.

     In view of  this, Union
Carbide has constructed a fuel
(waste) atomization facility and
a research combustor facility to
advance the knowledge of liquid
injection chemical waste combus-
tion.

PURPOSE

     In a liquid waste inciner-
ator, waste is normally atomized
into a mist.  The organics have
to be vaporized  before they can
be burned.  The  waste mist is
mixed with combustion air in the
burner zone.  As will be shown
later, the most  critical factor
affecting the destruction ef-
ficiency of liquid wastes often
is the performance of the burner
nozzle.  The pattern of turbu-
lent mixing of the combustion
air with the mist is also cri-
tical; the right amount of
oxygen has to be supplied to the
burning mist droplets at the
right time and right location.
For a we!1-designed and
properly-operated incinerator,
most of the organic destruction
occurs near the  flame zone
through flame oxidation pro-
cesses.  What takes place in the
combustion chamber after the
flame zone is relatively less
critical because the combustion
chamber (or what is normally
called incineration) temperature
is much lower than the flame
zone temperature.  The gas flow
in the rest of the combustion
chamber is normally somewhat
stratified.  The combustion
chamber just provides additional
time and temperature to destroy
the trace organics left over
from the flame zone.  This
occurs through much slower
flameless thermal oxidation
processes.

     If the three T's of com-
bustion for liquid injection
waste incinerators are ranked,
"Turbulence" is the most impor-
tant, followed by "Temperature"
and "Time".  Time is the least
important factor of the three if
a certain minimum operating
temperature (between 1,500 to
1,600°F) is met and good tur-
bulence has been achieved.  The
temperature and time consider-
ations are related to the reac-
tion kinetics or the chemical
processes of the combustion.
The turbulence consideration is
related to the very complex and
difficult to handle physical
processes of the combustion.
Factors which have impact on the
turbulence include:  fuel
atomization, droplet's size
distribution, droplet's eva-
poration or gasification rate,
air/fuel mixing pattern and
rate, stages of air addition,
etc.  Just about any design an.d-'"
operating parameters have
impacts on turbulent mixing.
                                551
-------
     In the past few years,
laboratory studies (7, 11) were
conducted to study the destruc-
tion of organics under non-flame
conditions.  Results from those
studies have indicated that above
1500°F (820°C), the reaction
rate is so fast that, theoreti-
cally, less than 0.1 second is
required to achieve the 99.99%
ORE requirement.  At tempera-
tures higher than 1600°F
(870°C), any organic compound
will be destroyed in no mea-
surable time (again, theoreti-
cally).  We all know that in the
real world case, this does not
happen due to the fact that the
reaction rate is mixing- (or
mass transfer-) limited and not
kinetics-1imited.
                       Calculated results shown
                   in Table I  demonstrate those
                   points discussed.   Non-flame
                   kinetics data for  vinyl
                   chloride (which is known  to be
                   thermally stable)  are used  in
                   the calculation.   The detailed
                   data and equations used  can be
                   found in Reference 11.

                        Of the many  physical  pro-
                   cesses going on in a flame
                   zone, fuel  atomization is  the
                   most critical step.   Many
                   theoretical studies  have  been
                   conducted on either  single
                   droplet combustion phenomena,
                   or the combustion  of vapor
                   cloud (6, 8, 10).   All those
                   studies tell us that the  li-
                   quid fuel combustion is  very
                             TABLE I
  T h e or e ti c a1  Vinyl  C h1o r i de  DestructionEfficiency  Calculation

(Note:  The  number shown  below is "the fraction remaining"
which  is  one minus the destruction removal efficiency (ORE).)
Residence time

Temperature

1200°F


1300°F


1400°F


1500°F


1600°F

1650°F
   0.1 sec
9.66 x 10-1
(3.4% ORE)

7.8 x 10-"1
(22.% ORE)

2.4 x 10-"1
(76.% ORE)

9.5 x lO'4
(99.905% ORE)

3.0 x 10-13

5.8 x 10-25
 0.5 sec
8.4 x 10-1
(16.% ORE)

2.9 x ID'1
(71.% ORE)

7.4 x lO-4
(99.926% ORE)

7.6 x lO-1^
2.4 x 10-63

1 .0 x 10-99
 1.0 sec
7.0 x 10-1
(30.% ORE)

8.2 x lO-2
(92.% ORE)

5.4 x 10-7
(99.999946% ORE)

5.7 x 10-31
1.0 x 10-99

1.0 x 10-99
                               552
-------
complex.  In the near future,
no satisfactory theoretical or
mathematical solutions can be
resolved to fully understand
the fuel atomization and the
associated combustion pheno-
mena.  The addition of the
swirl mixing with air makes
the problem even more com-
plex.  In our opinion, the
best way to resolve those
problems is a statistically
designed experimental program
followed with statistical data
analyses to obtain some kind
of empirical correlations.

     Several experimental fuel
atomization studies are on-
going (3, 5, 9) to address the
importance of spray combus-
tion.  This also indicates the
lack of fundamental knowledge
in this area.  Typical fuel
oils have to meet certain ASTM
standards and have defined
fuel properties.  This is not
the case for chemical wastes.
A variety of physical and
chemical properties are pos-
sible for chemical wastes.
Hence, the study of atomi-
zation for chemical wastes is
even more difficult.

     Information on spray
nozzle performance alone is of
little use without an actual
performance evaluation under
flame combustion conditions.
The smallest, practical spray
nozzle for chemical wastes is
around three million Btu/hr
due to plugging considera-
tions.  Hence, we have built a
multiple purpose, pilot-scale
research combustor to further
evaluate the nozzle per-
formance which has been deter-
mined to be satisfactory from
the cold-fuel atomization
study, under flame conditions.

     The majority of the RCRA
Appendix VIII compounds are
chlorinated chemicals.  The
others are basically nitrogen-
and sulfur-containing
chemicals.  Hence, one goal of
our flame combustion study is
to assess the design and
operating conditions which
favor the destruction of
chlorinated compounds.  At a
later time, we would also like
to evaluate the combustion of
nitrogen- and sulfur-containing
chemicals.

     Emission of Products of
Incomplete Combustion (PICs)
is becoming of concern to the
general public.  Although the
emission of incomplete com-
bustion products is a pheno-
menon of all combustion
devices and the trace level
emitted from an incinerator
may not present any more
threat to public health and
environment than the other
combustion devices, we would
like to understand this pro-
blem better with our research
combustor.

     The objectives of the
Union Carbide combustion fun-
damental research can be sum-
marized as:

o  To develop a better under-
standing of what is going on
in the flame zone when chemi-
cal wastes of different chem-
ical and physical properties
are burned.

o  To develop design and
operating criteria to maximize
the organic destruction effi-
                               553
-------
ciency and minimize the pro-
duction of incomplete com-
bustion intermediates.

APPROACH

     Two test facilities have
been constructed at Union
Carbide's South Charleston
Technical Center to carry out
the fundamental chemical waste
combustion studies.  One is a
cold, non-flame fuel atomiza-
tion pilot unit and the other
is a flame combustion
multiple-purpose pilot test
unit.

Fuel Atomization Pilot
Faci1 ity

     It is very difficult to
conduct atomization tests on
hazardous wastes due to the
many concerns of safety,
health, and environment.  The
three major considerations in
the design of the test facility
are: the safe handling of the
waste material, the explosive
nature of the fuel mist and
air mixture,  and the recovery
of all spray droplets and
vapor from the exhaust gas.

     A schematic diagram of
the fuel atomization pilot
facility is shown in Figure
1.  The facility can be func-
tionally separated into four
parts: (a) fuel handling
system; (b) spray vessel; (c)
effluent gas organic emission
control system; and (d)
emergency shutdown system.

     The fuel handling system
allows the test liquid to be
transferred in a closed system
between a fuel feed drum, a
spray test vessel, and a waste
storage drum.  The fuel lines
are designed to allow maximum
flexibility in operation
during the test.  The waste
liquid (including the line
purging solvent) is contained
in the waste storage drum
after the test.  Drum heaters
and heating tapes are used to
control the test fluid vis-
cosity.

     The two-foot diameter
spray vessel is designed as a
pressure vessel able to con-
tain an explosion.  An explo-
sion is very unlikely to occur
since all potential ignition
sources are eliminated.  Every
part of the unit is properly
grounded to avoid accumulation
of static charges.  The vessel
is equipped with two, 6 x
10-inch viewing windows and
two, 6 x 6-inch windows for
optical diagnosis.  In one of
the test setups, a honeycomb
metal plate can be inserted
above the window to stratify
the screening gas to avoid
fogging of the windows by the
fuel mist.  The vessel is
operated under slight vacuum
to avoid fugitive organic
emi ss i ons.

     An exhaust fan is used to
remove the gas and droplet
mixture from the test vessel
and create a slight negative
pressure.  The desired pres-
sure level can be adjusted
through a damper.  Following a
droplet knock-out device, a
wiremesh filter and a fabric
filter are currently used as a
two-stage demister to remove
                               554
-------
all liquid droplets from the
exhaust gas before it is dis-
charged into the atmosphere.
This demister has performed
satisfactorily in the initial
tests.

     The safe operation of the
test facility is further en-
sured by an emergency shutdown
system.  Both the fuel and air
influent flows will be shut
off automatically in case any
one of the following condi-
tions occur:

o  Overheating of the test
liquid in the fuel management
system.
o  Pressure build-up in the
spray vessel.
o  Overfilling of the test
liquid in the spray vessel.
o  Loss of the exhaust fan.

     Two techniques are cur-
rently used for spray diag-
nosis: visual observation and
photography.  The visual ob-
servation of the spray pro-
vides the preliminary opera-
tional information of the test
waste atomizer.  Two still
picture camera systems are
used to study the sprays.  A
35-mm camera system is used
for spray structure study and
a 4 x 5-in. format camera
system is used for in-depth
spray quality analysis.  A
single-flash shadowgraphic
technique was used (Figure 2)
for picture taking.  These
camera systems are similar to
those used by researchers at
Carnegie Mellon University
(5).
Multiple-Purpose Research Corn-
bus tor

     The research combustor
system consists of several
skid-mounted modules.  Maximum
flexibility was the key idea in
the design of the research com-
bustor.  The system was de-
signed to be able to test dif-
ferent burners, nozzles, and
fuels with different proper-
ties.  In the initial stage, a
research burner developed by
the International Flame
Research Foundation is being
used extensively to study the
impact of various air mixing
patterns on the flame zone.
Other burners will be studied
at a later time.  The research
combustor is designed to have a
long residence time (about 4
seconds total) with enhanced
turbulent mixing to make sure
that everything is burned.

     In the initial phase of
study, only fuel oils or fuel
oil spiked with low concentra-
tions of selected non-toxic
chemicals will be burned in the
research combustor.  No hazar-
dous waste will be handled
until we have received a RCRA
Part B permit.

     Figure 3 shows the top
view for the combustion chamber
section.  The combustion cham-
ber is designed for a normal
load of three million Btu/ hour
and a three second residence
time at 30% excess air.  An ad-
ditional one second residence
time is provided by the after-
burner section.  The maximum
load is five million Btu/hr.
The normal operating
                               555
-------
temperature is between 1,600 to
2,200°F and the maximum opera-
ting temperature is 2,500°F.  A
series of nine thermocouples is
permanently installed along the
combustion chamber to monitor
the transverse temperature
profile continuously.

     The combustion chamber of
the research combustor consists
of five sections.  The first
section is a burner support
section which can be changed
easily to adapt to different
size or design burners.  There
are three combustion sections
so that special test set-ups
can be installed.  For example,
simulated boiler tubes may be
installed in the third combus-
tion section for fouling and
slagging studies.  Corrosion
coupons may be installed for
material of construction
studies.  The fifth section is
the end cover with a viewing
port.  The combustion chamber
is lined with Durablanket (a
low-density firebrick) for fast
heat up so that minimum time is
wasted in start-up period and
minimum thermal shock damage is
experienced by the firebrick.

     The combustion chambers
are mounted on three skids.
The section on the first skid
has twelve, 10-inch sampling
ports.  At the center of the
10-inch port is a smaller
3-inch sampling port.  The
sections on the second and
third skids have twelve, 4-inch
sampling ports each.  The
sampling ports can be used for
either non-intrusive laser
diagnostic studies or for in-
trusive probe sampling studies,
The effects of residence time
can be studied by obtaining
samples at different ports
along the combustion chamber.
In this mode of operation,
long overall residence time
and high destruction effi-
ciency can be assured at the
stack at all times.

     Figure 4 shows  the block
diagram for the complete
system.  Fuel tanks, a com-
bustion air blower (1,500 SCFM
and 35-inch water pressure),
and a burner management system
are skid-mounted for maximum
flexibility.  All process
fluid (air, steam, fuel) flow
rates and pressures  are
measured in order to have
total control of the system
operati on.

     The research burner which
we are using for the initial
phase of the research is
called a "movable-block swirl
generator" (2).  The swirl
generator is designed to be
able to deliver combustion air
at any ratio of radial and
tangential flows.  Advantages
of this device are that it
makes change of swirl inten-
sity to any desired  level
during burner operation
possi b1e.

     The research combustor
has a complicated burner
management system with several
interlock loops to assure safe
operation.  It also  has an
emergency shutdown system
which will shut down all fuel
feeds automatically  in case
                                556
-------
any one of the following
conditions occurs:

o  Loss of flame.
o  Low combustion air flow.
o  High or low fuel gas
   pressure.
o  High thermal oxidizer
   pressure (loss of ID fan).
o  High or low thermal
   oxidizer temperature.
o  Low liquid waste feed
   pressure.
o  Low steam or air
   atomization pressure.
o  High stack CO level (for
   hazardous  waste feed).

     A special water-quenched
and water-cooled probe (14) is
used for hot  flue gas sampling
in the combustion chamber.  A
special suction pyrometer probe
(1) with water cooling is used
for radial temperature profile
measurement in the combustion
chamber.

RESULTS

     This research program is
designed to obtain some em-
pirical correlation from
statistical analysis of
statistically designed test
results.  With the knowledge
generated from the studies, we
can provide scientifically-
based recommendations to the
designers and operators to
help them identify problems
for an operating incinerator
and provide design basis for
new incinerators.  The tech-
nology is also applicable in
burning waste in boilers or
process furnaces.

     We expect that we can
provide the following in-
formation from research to be
conducted with the liquid-
injection combustion research
faci1ities.

1.  Spray nozzle selection.
2.  Best fuel mist and com-
bustion air mixing pattern for
a particular fuel (chemical
waste).
3.  Burner selection and per-
formance evaluation.
4 .  Selection of optimum
operating conditions.
5.  Combustion tests to assess
special combustion problems
associated with a particular
waste: such as boiler tube
corrosion and fouling, re-
fractory attacking, and
special pollutant emissions.

     An experimental program
is being conducted now and we
expect to have some prelimin-
ary data in the near future.

REFERENCES

1.  Beer, J. M. and Thring, M.
W., "Measurements in Flames",
Chapter 2, Edward Arnold,
London; Crane, Russak, New
York,  (1972)

2.  Beer, J. M. and Chigier,
N. A., "Combustion
Aerodynamics", Chapter 5,
Robert E. Krieger Publishing
Company, Malabar, Florida,
(1983)

3.  Carmij S., Ghassemzadeh,
M. R., " Viscosity and Spray
Formation Studies of Coal-Oil
Mixtures ",  Fuel, Vol. 60, pp
529-533, June, 1981.
                              557
-------
4.  Castaldini,  C.,  et  a 1 .,
"Full-scale  Boiler  Emissions
Testing of Hazardous  Waste
Coflring," Proceedings  of  the
Ninth EPA Annual  Research
Symposium,    pp.  180-193.

5.  Chigier,  N.  A.,  Meyer,  P.
L., " Vaporization  and
Oevolati1izatIon  of  Coal  Water
Sprays ", Carnegie-Mellon
Univ., Progress  Report,
DOE-RA-50266,  1983-1984.

6.  Chiu, H.  H.,  Kim, H.  Y.,
and Croker,  E. 0.,  "  Internal
Group Combustion  of  Liquid
Droplets  ",  19th  Symposium
(International)  on  Combustion,
The Combustion Institute,
1982.

7.  Del 1inger, B.,  et al. ,
"Determination of the Thermal
Decomposition  Properties  of 20
Selected Hazardous Organic
Compounds",  EPA-60Q/2-84-138,
NTIS PB84-232487, August,
1984.

8.  Faeth, G.  M., "Evaporation
and Combustion of Sprays",
Progress  in  Energy and
Combu s t i on Science, Vo~l 9,  pp.
1-76, 1983.
9.  Jackson, T. A., and
Samuelsen, G.  S., "An Evalua-
tion of Fuel Spray Performance
in a Swirl Stabilized Com-
bustor Using Optical Methods
for Drop Sizing", Paper pre-
sented at AIAA.SAE/ASME 20th
Joint Propulsion Conference,
June 11-13,  1984, Cincinnati.

10. Law, C. K.,  " Recent Ad-
vances in Droplet Vaporization
and Combustion," Progress in
Energy and Combustion Sciencfe,
Vol. 8, 1982.
 11 . Lee,  K.  C.,  et  al.,  "Re-
 vised Model  for  the  Prediction
 of  the  Time  Temperature  Re-
 quirements for Thermal
 Destruction  of Dilute  Organic
 Vapors  and Its Usage for
 Predicting Compound
 Destructibi1ity," APCA New
 Orleans,  June, 1982.

 12. Seeker,  W. R.,  Heap,  M.
 P., and Samuelsen,  S.  S.,  "
 Characterization Performance,
 and Dynamics of  Fuel Oil  Spray
 Atomization ," Paper  presented
 at  the  EPA Fundamental
 Combustion Workshop, Newport
 Beach,  California,  February,
 1980.

 13. Trenholm, A. et  al.,
 "Emission Test Results for a
 Hazardous Waste  Incineration
 RIA," Proceeding of  the  EPA
 Ninth Annual Research
 Symposium, Ft. Mitchell,
 Kentucky, May 2-4,  1983,  pp.
 160-170.

 14. Wendt, J. 0. L., et  al.,
 "Advanced Staged Combustion
Configurations for Pulverized
Coal", University of Arizona,
pp. 17-19, DOE contract
DE-AC01-79ET-15184, August,
1983.
            Disclaimer

 The work described in this paper was
 not funded by the U.S. Environmental
 Protection Agency. The contents do
 not necessarily reflect the views of
 the Agency and no official endorse-
 ment should be inferred.
                               558
-------
                          ATOMIZING GAS

                              -SCREEN GAS
                                                 EXHAUST
                                                   FAN
                                                                               SPRAY
                                                                                        ATOMIZER
                                                                        DIFFUSER
                                                                             *\
                                                                    FLASH LIGHT
                                                                     SOURCE a
                                                                    DEFLECTOR
                                                                                                             FILM
                                                                                                             PLANE
         FIGURE 1.   SCHEMATICS OF THE FUEL
                     ATOMIZATION PILOT FACILITY
                                                FIGURE 2.  OPTICAL SYSTEM  FOR FLASH SHADOW
                                                            PHOTOGRAPHY
CJl
tn
<£>
        VARIABLE
         SWIRL
         BURNER
r-VIEWING PORT
                           SAMPLING PORTS
»

» I t

o o o o
T i 1
lit
O O O O
1 I i
                          -THERMOCOUPLES
      COMBUSTION
          AIR
                                              VIEWING
                                               PORT
                                         FLUE GAS
                                                                               STEAM
                                                                                             -AIR

NATURAL GAS
CONTROL
SYSTEM

COMBUSTION
AIR CONTROL
SYSTEM

PILOT
CONTROL
SYSTEM

r
I
• 	 to*
FUEL
ANAGEMENT
SYSTEM

RESEARC
COMBUSTC
^ 1

EMERQENCY
SHUTDOWN "
SYSTEM
*
H
)R
1
AFTER-
BURNER
_J
HEATER
& PUMP -
STATION




QUENCH
•.& SCRUB-
BER

J1EATEO
QUID FUEL
 TANK
         FIGURE 3.  UNION CARBIDE RESEARCH
                    COMBUSTOR - TOP  VIEW
                                            FIGURE  4.   BLOCK  DIAGRAM FOR THE RESEARCH  COMBUSTOR
-------
                   FATE OF POLYNUCLEAR AROMATIC  COMPOUNDS
                     DURING SEWAGE  SLUDGE  INCINERATION

              T.R. Bridle*, M.J. Bumbaco^  and  P.O.  Crescuolo*
                            iEnvironment Canada
                      Environmental Protection Service
                       Wastewater Technology Centare
                        Burlington, Ontario, Canada

                            ^Environment Canada
                      Environmental Protection Service
                 River Road Environmental  Technology  Centre
                          Ottawa, Ontario, Canada
                                  ABSTRACT
     Sewage sludge accumulates biorefractory compounds  such  as  heavy  metals
and toxic organics,  which  may pose problems if subsequent disposal options
are not well managed.   While the fate of heavy metals  during  sludge  incin-
eration has been well  defined,  little data  is  available regarding the  fate
of toxic organics.   The prime objective of this study  was to  determine the
fate  of seven  selected  polynuclear  aromatics  during  full-scale   sludge
Incineration.     The  results   indicated   that,   with  the  exception   of
dibenzofuran, destruction efficiencies for  the target  compounds  ranged  from
95.03%  to  99.98%  at a maximum furnace temperature  of 786°C.    In general,
the destruction  efficiency  of  a  compound  increased as  its boiling point
increased and as the number of benzene rings is contained increased.   This
suggests  that   in  multiple  hearth   furnaces,  operated  countercurrently,
compounds  with  low  boiling  points  are volatilized  and  swept  out  of  the
incinerator  without   significant  destruction  occurring.    The  data  also
indicates  that  atmospheric  emissions  are more  significant  than aqueous
emissions.  The ratio  of atmospheric to aqueous  emissions  varied from  2:1
for carbazole to greater  than 20:1 for dibenzofuran.   Generally,  about 0.5
to  5%  of  the  polynuclear  aromatics  in   the  sludge  feed  were released
atmospherically with 0 to  1.5% discharged aqueously.
INTRODUCTION

     Incineration   is  widely   used
for the ultimate disposal  of sewage
sludges, with  approximately 40%  of
Canadian  sludges  (Schmidtke,  1981)
and  about   20%  of   U.S  sludges
(Bunch,  1981)  being  disposed  in
this manner.   Sewage sludge accumu-
lates  biorefractory compounds  such
as heavy metals and toxic  organics,
which may pose problems if subsequent
disposal  options  are  not well  man-
aged.  While the fate of heavy metals
during  sludge  incineration  has  been
well defined, (Campbell et a!., 1982;
Dew!ing et  al.,  1980)  little data is
available regarding the fate of toxic
organics.   Canadian (IJC,  1983)  and
U.S. (U.S. EPA, 1982) surveys of sew-
age sludges have indicated that poly-
nuclear   aromatic   hydrocarbons   are
                                    560
-------
frequently  found  to be  present at
relatively   high   concentrations.
For  this  reason  polynuclear  aro-
matics  (PNA's  )are  receiving  con-
siderable  attention from  a sludge
management   viewpofnt.      Sewage
sludge  from  Hamilton,  Ontario, was
sampled  and  analyzed  as   part  of
Environment    Canada's     "Toxics
Screening Study of Municipal Sewage
Treatment  Plants"  conducted during
1981/82.   This survey  revealed the
sludge  contained  relatively  large
concentrations  of   organic   con-
taminants,  especially  PNA's.   Con-
sequently, Hamilton sludge  is being
used  extensively  for  Environment
Canada's  ongoing  studies  aimed at
determining  the  effect  of various
sludge management options  on  toxic
organics1  fate  and  mobility.   The
assessment   of  organic   emissions
from the  Hamilton STP incinerators
is  thus   just  one  component  of
Environment  Canada's overall toxics
assessment program.

STUDY OBJECTIVES

     The  prime objective   of  this
study was  to determine the fate of
seven  selected  priority   organics
during    multiple-hearth    sewage
sludge   incineration.     A  second
objective  was  to  assess  the  suit-
ability  of two stack  gas   sampling
techniques   for  the   seven  target
compounds,   which   included  acena-
phthylene,  fluorene,   fluoranthene,
pyrene, benzo-a-pyrene,  (B-a-P),di-
benzofuran and carbazole.

DESCRIPTION  OF HAMILTON'S  SLUDGE
PROCESSING TRAIN

     The City of Hamilton  is served
by  a  single sewage  treatment plant
with a  design  capacity of  approxi-
mately   400   000   m3/d  (90  mgd).
Sewage  sludge is   currently  being
produced  at  a  rate  of about  60 dry
tonnes per day.

     Primary   and  waste   activated
sludge   are   anaerobically  digested
prior    to    mechanical    dewatering
(vacuum  filters)  and  subsequent in-
cineration.     Two   Envirotech  BSP
multiple-hearth furnaces are used for
sludge incineration.   Each unit com-
prises    nine    alternating    in/out
hearths, six metres in diameter.  The
units  are  rated  at  9 000  kg  of wet
sludge  per  hour,  at  a  nominal  20%
total    solids   and   75%   volatile
content.  Both  units  are  natural gas
fired  on  hearths  3,  5 and 7.   Sludge
is fed via a  weigh  belt  from the top
of the  furnace (hearth #1) and moves
downward through  the  drying, burning
and cooling zones prior to being dis-
charged  as  ash at the bottom  of the
furnace.  Rabble  arms move the  sludge
from hearth to hearth.  Off gases are
quenched and  scrubbed prior to being
discharged,  via  an   ID  fan  to  the
stack. Quench/scrubber water (68 m3/h
to each  unit), which  is secondary ef-
fluent   from  the  STP,  is  used  to
slurry the  bottom ash  and transport
it  to  the  impoundment   lagoon.     A
schematic of  the  Hamilton  STP  incin-
erators  is shown  in Figure 1.

SAMPLING AND ANALYTICAL METHODS

     Feed  and effluent  streams from
the  incinerators  were sampled  during
a four day  period in  September  1982.
Samples  of  feed  sludge,  bottom ash,
scrubber  water and stack  gases were
collected  during  the  4-hour steady-
state  run  on  each  of the  four con-
secutive sampling days (September 15-
18).   Sampling  locations  are   shown
on Figure 1.  The sludge feed,  common
to both  incinerators, was  sampled via
grab  methods  each   hour   during  the
steady-state  period.    These samples
were   composited  to   represent  the
                                    561
-------
sludge  fed   during  the  sampling
period.   Grab  samples  of  ash and
scrubber water  were also collected
each  hour  and  composited.    Stack
gas  was  sampled   using  two  stack
sampling trains  using  both wet and
dry  collection  trains.    For each
system,  isokinetic  sampling  was
carried out along a multipoint tra-
verse  of the stack.   The wet col-
lection train used  ethylene  glycol
as  the primary  collection medium,
while   the  dry   train   used  an
Amberlite  XAD-2  sorbent  cartridge.
Details of the  stack  gas sampling
are    presented   elsewhere    (EPS,
1983).

     Composited  samples  of sludge,
bottom  ash and  scrubber water were
analysed for the target organics by
the  Analytical  Services  Section of
the  Wastewater   Technology Centre,
while  the   stack  gas  samples were
analysed by  the  River Road Labora-
tory  in  Ottawa.   Sludge  and  bottom
ash  were  extracted with acetone-
hexane  in  a  soxhlet apparatus  while
the  scrubber water was  extracted
with   methylene  chloride  in  a   1
litre  separatory   funnel.     Solid
samples from the stack gas monitor-
ing  were  also  soxhlet  extracted
while  liquid samples were  extracted
using   separatory   funnels.     All
extracts   were  analysed   by   GC/MS
techniques.

RESULTS

     Complete   sampling   of  all
streams  was  successfully effected
on  three of  the four sampling days
(September  15,  16  and   17), and
results  for  these  three  days are
detailed below.   On September 18th
only  stack  gas  sampling  was con-
ducted and  results  for  the  amber-
lite train are  not  available.  The
results are  subdivided to show
operating,  analytical  and  emission
data.

Operating data for the sampling
period

     The operating conditions of each
incinerator were monitored and logged
on  a  hourly  basis  during  each  four
hour   steady-state   run,   with   the
exception  of  wet  sludge  feed  rate,
which   was   monitored  continuously.
The average operating  conditions  for
the  two  incinerators  for  the  three
test periods is shown in Table 1,   On
September 15th, a ruptured ash trans-
port line forced the  run  to be term-
inated  after  1  1/2  hours. Also, dur-
ing  this   run   only  84%   of  iso-
kineticity  was  achieved,  which  is
outside the  normal  accepted range of
100  +   10%.   Nonetheless,  it  can  be
seen that operating  conditions  were
similar  during  all  three  sampling
periods.   The two  exceptions  appear
to  .be   a  higher  furnace  temperature
and   stack   gas   particulates   on
September 15th.  The high particulate
loading on September 15th is a direct
result    of   not    sampling    iso-
kinetically.  Sludge  loading to both
units ranged from 61 to 75% of design
(based  on wet sludge feed rate)  dur-
ing  the three  day  test  period.   It
should  be  noted  that  the  feed  was
lower   in  both  total  and  volatile
solids  than the  values  assumed  for
design  capacity calculations.  Hearth
temperature  profiles  are  shown  in
Figure  2.   The  profiles  for  the   3
test periods  also show that for some
unknown  reason,  hearth 6 temperature
on  September 15 appears abnormal.

Analytical Results

     The  structure  and  some  of  the
physical   properties   of  the  seven
target  PNA's are shown in Table 2.
                                     562
-------


® SLUDGE FEED ',





INCINERATOR
*1



1
2

4
A.
?
7
a
e




—



I J
STACK
ID FAN
-Q-

©






	 l r" ! i ™" •*££

!
Jvl (SCRUeBEl?) fA.
"gJ !*•»«•;
SCRUBBiR/
«>1 I



f~ i ®
COMBUSTION 1
ER-*1 j*f
\ /QUENCH/
>SCfeJB8£R

©








(2


INCINERATOR



© ^ f 	
1 COMBUSTION
AIR 1 ASH 1 *«
I SUMP. J
    BAMPILINQ UXafflONS

      Ot| SLUDGE FEED
      ffi DRV BOTTOM ASH
      6 SCRUBBER WATER
        STACK OAS
      ASH SLURRY TO
      IMPOUNDMENT LAGOON
Figure 1.   Schematic  of Hamilton  Incinerators
      800-


      700-


   .  600-


   €5ooH

   5  400-
   5

   1  300-

      200-

      100-
       0
NATURAL GAS FIRED

   i     I    4
                  i  i   i   i  i   l   I   I  l   i
            EXHAUST 1X3456789
                  HEARTH NUMBER

    Figure  2.   Hearth Temperature Profiles
                         563
-------
                TABLE  1.   INCINERATOR  OPERATING  CONDITIONS
Parameter
                 Sept.15/82   Sept.16/82  Sept.17/82
Total wet sludge feed rate (kg/h)
Sludge solids (%)
Volatile content (%)
Dry sludge feed rate (kg/h)
Ash production (kg/h)
Scrubber water flow rate (m3/h)
Max. hearth temperature (°C)
Stack gas data
- Oz concentration (vol. %)
- flow rate (mVmin)
- velocity (m/s)
- temperature (°C)
- particulates (kg/h)
13530
15.8
46.0
2138
1154
134.1
786

12.7
450.7
3.8
116
4.05
11102
15.8
46.0
1754
947
134.1
732

12.8
473.5
3.2
66
1.42
11243
15.8
46.0
1833
990
134.1
722

12.7
426.9
2.9
63
0.95
   TABLE 2.  STRUCTURE AND SELECTED  PHYSICAL  PROPERTIES  OF  TARGET  PNA'S
Compound Name
Structure
Boiling    Aqueous       lonization
Point      Solubility    Energy
(°C)       (mg/L)        (ev)
Acenaphthylene



Dibenzofuran



Fluorene



Carbazole



Fluoranthene



Pyrene



Benzo-a-pyrene
   WH
                   266
                   287
                   294
                   355
                   375
                   393
                   495
             3.93
                              **
             1.98
                              **
             0.26
             0.14
             0.0038
7.76
                            7.90
7.89
                            7.57
7.80
7.41
7.12
*Notavailable.
                                    564
-------
      As  indicated, an  objective of
 the  study  was  to assess the collec-
 tion efficiency of two gas sampling
 trains,  namely  a  mt  collection
 system using ethyl ene-glycol  and a
 dry   train  using  Amber!ite  XAD-2
 cartridges.   Results  from these two
 collection  trains  are   shown   in
 Table 3,    Neglecting the  results
 from  September   15th   (not   i so-
 kinetic)   it  is  apparent that  the
 Amberlite  train  generally collects
 more PNA's than the ethylene glycol
--train.  With the  exception  of car-
 bazole and  B-a-P, the  PNA  concen-
 tration     measured    using    the
 Amberlite  train  were  several  times
 higher than the  same  species mea-
 sured  with  the  ethylene  glycol
 train.  In  addition,  the distribu-
 tion  of   PNA  in  the   various  com-
 ponents  of  each  train  showed  the
 Amberlite  to  be  superior  to eth-
Vylene glycol.    For these reasons,
 'subsequent stack gas data shown are
 those generated using the Amberlite
 train.

      The  results  of   the analyses
 conducted   on   the various  samples
 collected   during  the  three  test
 peHods  are shown in  Table 4.  All
 of;the bottom ash samples contained
 noil-detectable  (NO)  levels  of  the
 target  PNA's   (detection  level  0.1
 yg.'i'g)-  The data  in  Table 4 indi-
 cate that sludge quality (with re-
 sp4ct to the target PNA's) was uni-
 fo.rm during the three test periods.
 Tho  only  exception  is  that  B-a-P
 wa:i  not detected in  the  sludge on
 September  16th.   Since  B-a-P  was
 detjected  in  the  stack   gas,  this
 sludge  analysis  is suspect.   This
 conclusion  is   supported  by  MOE
 anchyses  of  sludge  samples  taken
 concurrently   (Rush   and  Taylor,
 19£i3).   These  sludge  analyses are
 in  general  agreement  with the ones
 generated in this  study.
     Scrubber water analyses indicate
low  levels  of  dibenzofuran,  fluor-
anthene,  pyrene,  carbazole  and  flu-
orene were present.   Analysis  of the
STP final  effluent,  which is used as
raw  scrubber  water  indicates  only
trace  quantities  of  these compounds
are present  (Rush and Taylor, 1983).

     QA/QC  on  the  sludge,  ash  and
scrubber   water   samples  indicated
recoveries of the target  PNA's ranged
from  49% to  214%.    The  average re-
covery data  for these three matrices
are shown in Table 5.

The  replicated  sludge  samples  were
spiked at 50 pg/g, the ash samples at
25 pg/g  and  the scrubber water at 60
yg/L.   It  can  be  seen that while re-
coveries did vary, the values report-
ed are  considered  acceptable for the
levels  of  contamination  encountered
in these  matrices  and are represent-
ative  of recoveries  reported  in the
literature.    Note  however  that the
sludge,  ash  and  scrubber  water an-
alyses   reported  are not  corrected
based on this  recovery data.

Mass emission data

     Based on the  incinerator operat-
ing  conditions  (Table 1)  and the an-
alytical  data  (Table 4),  mass   bal-
ances  around  the  incinerator system
were  determined.    These results for
the  three test  periods  are shown  in
Table  6.

     The  destruction efficiency  (DE)
for  each compound  for each sampling
day  is  also  shown  in  Table 6.  The  DE
is defined  as:

      (contaminant   feed   rate-scrub-
DE =  ber  emission  rate-stack emis-
      sion  rate)xlOO

          contaminant feed rate
                                      565
-------
                 TABLE  3.  STACK GAS  CONCENTRATION  DATA
September 15
Compound
Acenaphthylene
Dibenzofuran
Fluorene
Carbazole
Fluoranthene
Pyrene
Benzo-a-pyrene
Ethyl ene
Slycol
23.1
166.9
49.1
32.5
102.9
29.3
2.9
Amberlite
8.4
53.6
16.0
14.9
43.1
17.8
2.0
September 16
Ethyl ene
Glycol
2.9
0.7
7.1
17,5
4.2
1.2
1.9
Amberlite
9.6
67.3
25.2
12.7
49.9
10.6
0.1
September 17
Ethyl ene
Glycol
9.3
22.9
8.5
26.2
11.4
3.4
0
Amberlite
8.5
95.4
27.2
6.
58.0
17.6
0.5
                                 TABLE 4.  ANALYTICAL RESULTS
September 15
Compound
Aconaphthy 1 one
Dlbanzofuran
F I uorena
Carbazole
Fluoranthene
Pyrena
Banzo-8-pyrene
Sludge*
4.9
4.4
14.4
6.7
77.0
60.0
36.6
Scrubber
Water*
NO
1
<1
trace
3
2
NO
Stack
Gas*
8.4
53.6
16.0
14.9
43.1
17.8
2.0
September 16
Sludge
2.1
2.5
8.3
3.4
52.2
38.9
ND
Scrubber
Water
NO
1
<1
trace
2
<1
ND
Stack
Gas
9.6
67.3
25.2
12.7
49.9
10.6
0.1
September 17
Sludge
5.6
5.3
16.4
10.0
77.1
57.3
57.2
Scrubber
Water
NO
1
1
trace
3
1
NO
Stack
Gas
8.5
95.4
27.4
6.6
58.0
17.6
0.5
•Concentration expressed as jjg/g
*ConcentratIon expressed as ug/L
.Concentration expressed as ugAn3
  TABLE  5.  RECOVERY  DATA FROM SLUDGE, ASH  AND SCRUBBER
             HATER SAMPLES
Compound
Acenaphthylene
Fluorene
Fluoranthene
Pyrene
Benzo-a-pyrene
Percent Recovery
Sludge
62
83
114
97
214
Bottom Asn
49
76
75
97
148
Scrubber Water
66
81
71
80
140
                                      566
-------
     It can be seen that Individual
PHA feed  rates to  the incinerators
vc.ried from  a low of  3.68  g/h for
acenaphthylene  to  a   maximum  of
lf.4,63 g/h for  fluoranthene.   By
contrast,   stack   emission   rates
vo'ried from  0.003  g/h  for benzo-a-
p^'rene to 2.444  g/h   for  dibenzo-
fu.ran.    Destruction   efficiencies
fcr the PNA's varied from 91.74% to
9S;.98%,  with  the exception  of di-
benzofuran,   which  exhibited  de-
struction  efficiencies  as   low  as
53;.39%.   The  results  achieved dur-
ing the three test periods are very
consistent, especially  taking into
account  the  non-isokineticity  of
the stack gas analyses generated on
September 15th.

     Average  destruction efficien-
cies for  the  seven  PNA's are plot-
ted as a function  of  their  boiling
points  (Figure  3)  and  ionization
energies  (Figure 4).    It   appears
that   destruction   efficiency  in-
crpases  non-1inearly  with increas-
ing boiling  point.   This  is to be
exbected   for a  multiple   hearth
furnace  operated  in  countercurrent
mo'de with sludge addition on  hearth
one.    As  the  sludge   is   heated,
volatile  compounds  are  vaporized
an{l swept to  the  stack  with com-
bustion air.  Thus, a compound like
benzo-a-pyrene,   with   a    boiling
point of  495°C, would progress much
farther down the incinerator  before
being  vaporized.    Once vaporized,
it would  then  be exposed to a more
aggressive  thermal  regime   (higher
tenperature,  longer residence time)
than would, say, dibenzofuran.  The
correlation  of DE  with ionization
energy indicates  that thermal sta-
bility   increases  with  increasing
ionization  energy,  which  is to be
expected.

     The   average  PNA   destruction
data,  and  the  percentage discharged
via the stack and scrubber water, are
shown  in  Table  7.   Since  the stack
data generated  on  September 15th was
not  collected   isokinetically,  the
averages shown  in Table  7 do not in-
clude the September 15th  values.

     The  data  in  Table  7   highlight
the fact  that with the  exception of
dibenzofuran, DE's for all PNA's were
greater than 95%.  There  are no emis-
sion standards  for PNA's  from combus-
tion sources  in Canada,  however EPA
standards for hazardous wastes, which
excludes sewage sludges,  do  require a
minimum  DE  of  99.99%  for "principal
organic    hazardous    constituents"
(Federal Register,  1982).  The emis-
sion data  in   Table  7  also confirm
that air  emissions  are far more  sig-
nificant than aqueous emissions.

     There  are  very   few documented
cases of PNA  fate during  incineration
with which  to  compare  this study's
results.   A  cement  kiln operated in
Sweden,  using  tar  as  an auxilliary
fuel,  was  able to  assess  PNA fate.
At  an   operating  gas  temperature of
>1400°C,   99.998%   destruction   of
total  PNA's   was  effected   (Trovaag,
1981).     The   University of   Dayton
Research  Institute  (UDRI)   has  con-
ducted  laboratory-scale  thermal  de-
composition studies on selected PNA's
in  free flowing  air  (Rubey,   1982).
Their  results are shown  in  Table 8.
This   data   indicates   that   PNA's
exhibit  considerable  thermal stabil-
ity.    The  data also  indicates  that
naphthalene   and   chlorinated  naph-
thalenes  are  more  stable  than  the
other  PNA's   studied.    Surprisingly,
as  the number  of  aromatic  rings in-
crease,  the  thermal  stability  of the
compound  appears to  decrease, i.e.,
the order of  stability decreases  from
naphthalene   to anthracene  to  tri-
phenylene to  to benzo-a-pyrene.
                                    567
-------
99.99tr
 99.9
  99
  so
                                                    * Benzo-o-pyrane
             Fluoreno »
               liconaphthylnnB
   •Flworanthanfl

fCorbozole
        D i bonioFuron *>
   0'  .... i  .... i  .•.•)•.,.! ,  ...  ! ,  ...  i ....  i ...
   ZOO     250     3QO     350     400     450     SOD      S50      600


                          COMPOUND BOILING POINT CD


       Figure 3.    PNA Destruction versus Boiling Point
69.99
 99. S
  99
  GO
                                                               Fluorene
                                                     Oibenzo? uran *
   7.0    7.1    7.2    7.3    7.4    7.5    7.6    7.7    7.8    7.9    8.0

                             HMIZATION B£RG¥ <»v!




    Figure 4.   PNA  Destruction  versus  lonization Energy
                                568
-------
                          TABLE 6.  INCINERATOR MASS BALANCE DATA
September 15
Mass f low rate
Compound
Acenaphthy I ene
Oi benzofuran
F 1 yorene
Carbazole
Fluoranthene
Pyrene
Benzo-a-pyrane
SI udge
10.48
9.41
30.79
14.33
164.63
128.28
78.25
Stack
0.227
1.449
0.433
0.403
1.166
0.482
0.054
(q/h >
Scrubber
Water
ND
0.134

Sludge
10.16
9.72
30.06
18.33
141.32
105.03
104.85
Stack
0.218
2.444
0.697
0.169
1.486
0.451
0.013
Scrubber
Water
ND
0.134
0.134
<0.134
0.402
0.134
ND
DE
97.85
74.71
97.24
98.35
98.66
99.44
99.98
    TABLE  7.   AVERAGE PNA DESTRUCTION/DISTRIBUTION DATA*
Compound
Acenaphthylene
Dibenzofuran
Fluorene
Carbazole
Fluoranthene
Pyrene
Benzo-a-pyrene
Average
DE (%)
95.22
64.05
95.69
95.03
98.41
99.40
99.98
Percent in
Stack Gas
4.78
34.35
3.63
3.49
1.30
0.44
0.02
Percent in
Scrubber Water
0
1.60
0.68
1.48
0.29
0.16
0
*Excluding  September 15th data
    TABLE  8.   LAB SCALE THERMAL DESTRUCTION  PROFILES FOR SELECTED  PNA'S
               (RUBEY, 1982)
Percent destruction at 2 second gas retention time
Compound 400°C 600°C 650°C 700°C 725°C
Benzene
Naphthalene
Monochl oronaphthal ene
1,2 Dichl oronaphthal ene
Anthracene
1,2,3,4 Tetrachl oro-
naphthal ene
Triphenylene
Benzo-a-pyrene
0
0
0
0
0
0
0
0
0
0
0
0
47
0
27
36
15
NA
16
29
71
30
36
49
89.8
77
69
76
NA
67
94
96.5
98.5
85.0
81.0
83.0
99.83
90.0
99.50
99.79
                                      569
-------
This  trend  Is  supported  by  the
results  generated  in  the  present
study,  which  found benzo-a-pyrene
to  be the  least  thermally  stable
compound.  The UDRI data does indi-
cate   thermal   stability   as   the
PNA's,  in  their  vapour  state,  are
exposed  to  the  high  temperature
zone.   It  is therefore significant
that  vapour-phased thermal  stabi-
lity  appears  to  decrease  as  the
number of rings is  increased.

SUMMARY AND CONCLUSIONS

     The  results  from this  study
indicated that,  with  the exception
of  dibenzofuran,   destruction  ef-
ficiencies  for  the  target  PNA's
ranged  from  95.03% to 99,98%  at  a
maximum   furnace   temperature   of
786°G.    In   general»  the  thermal
stability  of  a  compound  decreases
as  the number of  benzene  rings it
contains increases.

     The  data also  indicates   that
under   the   conditions   evaluated
atmospheric   emissions   are    more
significant   than  aqueous   emis-
sions.  The  ratio of atmospheric to
aqueous emissions varied from about
2:1  for  carbazole to   20:1 for di-
benzofuran.    Generally  about  0.5
to  5% of the  PNA's present in the
sludge  feed   are   released  atmos-
pherically,  with 0-1.5% discharged
aqueously.

    Stack gas  sampling has indicat-
ed  that for  the  target  compounds
the  dry  collection  train,  using
Amber!ite  XAD-2  cartridges,   was
superior  to  the  wet  train  using
ethylene glycol as  the absorbent.

     The  results from  this survey
suggest  that  atmospheric  emission
of  PNA's  from the  Hamilton sludge
incinerators   amount   to   roughly
60  kg/yr.   Current  modifications to
the incinerators are, however, likely
to reduce these emissions.

REFERENCES

 1.  Schmidtke, N.W.,  "Sludge Gener-
     ation, Handling  and Disposal at
     Phosphorus Control Facilities in
     Ontario",      Characterization,
     Treatment  and   Use   of  Sewage
     Sludge,  Comm,  of the  European
     Communities,     Editors     P.L.
     L'Hermite and H. Ott, (1981).

 2.  Bunch,   R.L.,   "Sewage  Sludge
     Dilemma  of the  Eighties",   pre-
     sented  at  3rd   Int.   Conf.  on
     Physiochemical Methods  for Water
     and     Wastewater    Treatment,
     Poland, (1981).

 3.  Campbell, H.W., et a!.,  "Fate of
     Heavy  Metals  and  Potential  for
     Clinker  Formation  during Pilot
     Scale  Incineration  of  Municipal
     Sludge",  Wat.  Sci.  Tech.  14,
     (1982).

 4.  Dewling,  R.T.  et  a!.,  "Fate and
     Behaviour   of   Selected  Heavy
     Metals  in  Incinerated  Sludge",
     JWPCF  52,  10,  (1980).

 5.  IJC,  "A  Review of the  Municipal
     Pollution  Abatement  Programs in
     the  Great  Lakes  Basin".  Report
     to   GLWQ   Board  of  the   IJC,
     November  (1983).

 6.  U.S. EPS,  "Fate of Priority  Pol-
     lutants   in   Publically  Owned
     Treatment  Works", EPA  440/1-82/
     103  (1982).

 7.  EPS  "Report  on a Stack  Sampling
     Program to Measure the  Emissions
     of  Selected  Trace  Organic   Con-
     taminants   from  the   Hamilton/
     Wentworth      Sewage     Sludge
                                     570
-------
     Incinerator",  Technical  Ser-
     vices  Branch,   EPS,   October
     (1983).

8.   Rush,  R.J.   and  L.J.  Taylor,
     "Removal   of  Hazardous  Cont-
     aminants   in  an  Ontario  Water
     Pollution    Control    Plant".
     Proceedings  of  the MOE  Tech-
     nology Transfer  Conf.  No.  4,
     Toronto,  November 1983.

9.   U.S.  Federal  Register,  Vol.
     47, No. 122, June 24,  (1982).

10.  Trovaag,  K.,  "Hazardous  Waste
     Incineration   in   a    Cement
     Kiln", Proceedings,  NATO/CCMS
     Symposium  on Hazardous  Waste
     Disposal,   Washington,   D.C.,
     (1981).

11.  Rubey,   W.A.  et   al,   "The
     Thermal  Decomposition  Behav-
     iour  of  Selected  Polynuclear
     Aromatic   Hydrocarbons",   Un-
     published  Report,   University
     of Dayton, (1982).
                                 Disclaimer

                   The work described in this paper was
                   not funded by the U.S. Environmental
                   Protection Agency.  The contents do
                   not necessarily reflect the views of
                   the Agency and no official endorse-
                   ment should be inferred.
                                   571
-------
-------