United States
                          Environmental Protection
                          Agency
 x°/EPA
 Office of
 Research and Development
 Cincinnati, OH 45268
EPA/540/R-94/525a
December 1994
Introduction

    In   1980,   the   U.S.   Congress   passed  the
Comprehensive   Environmental   Response,
Compensation, and Liability Act (CERCLA), also known
as Superfund,  committed to protecting human  health
and  the  environment from  uncontrolled  hazardous
waste sites.  CERCLA was amended by the Superfund
Amendments and Reauthorization Act (SARA) in 1986.
SARA  mandates the  implementation  of  permanent
solutions  and  the   use  of  alternative   treatment
technologies or resource recovery technologies, to the
maximum extent possible, to clean up hazardous waste
sites.

    State and Federal agencies, as well  as private
parties,  are now exploring  a  growing number  of
innovative technologies for treating hazardous wastes.
The sites on the National Priorities List total over 1,200
and comprise a broad  spectrum of physical, chemical,
and environmental conditions requiring varying types of
remediation. The U.S. Environmental Protection Agency
(EPA)  has  focused   on  policy,  technical,  and
informational issues  related to exploring and applying
new remediation technologies to Superfund sites. One
such   initiative  is   EPA's  Superfund  Innovative
Technology Evaluation (SITE)  Program, which was
established to accelerate development, demonstration,
and use  of innovative  technologies for site cleanups.
EPA SITE Technology Capsules summarize the latest
information available on selected innovative treatment
and site  remediation technologies and  related issues.
These  capsules are designed to help EPA remedial
project   managers,   EPA   on-scene   coordinators,
contractors,  and   other  site  cleanup   managers
understand the types of data and  site  characteristics
needed   to  effectively  evaluate  a   technology's
applicability for cleaning up Superfund sites.

    This Capsule provides information on the InPlant
Systems, Inc.  (InPlant)  Oleofiltration technology,  a
technology   developed   to   separate   suspended,
emulsified,  and a  portion of  dissolved  hydrocarbons
from  water.    The  Oleofiltration  technology was
evaluated under EPA's SITE Demonstration Program at
a former oil reprocessing facility in June 1994. This
Capsule presents the following information:
        Abstract
        Technology Description
        Technology Applicability
        Technology Limitations
        Process Residuals
        Site Requirements
        Performance Data
        Technology Status
        Disclaimer
        Sources of Further Information
        References
Abstract

    Oleofiltration is an innovative hydrocarbon recovery
technology  that  utilizes  amine-coated,   oleophilic
granules  to separate suspended  and  mechanically
emulsified hydrocarbons from aqueous solutions.  The
granules are also reported to separate several types of
chemical  emulsions  and to reduce concentrations of
dissolved   hydrocarbons.    The   technology   was
developed  by  Exxon Research   and Engineering
Company and manufactured under exclusive license
and  patent  by InPlant of Houston,  Texas.   North
                                       SUPERFUND INNOVATIVE
                                       TECHNOLOGY EVALUATION
                                                                                      Printed on Recycled Paper

-------
 American Technologies Group, Inc. (NATGI) is the sole
 marketer of the technology.

     The InPlant SFC System combines an innovative,
 vertical-fin, coalescing separator and a patented, amine-
 coated, ceramic granule filtration system (the Oleofilter)
 into one unit, capable, according to InPlant, of treating
 virtually  any insoluble  hydrocarbon/water  mixture.
 When  the  hydrocarbon/water  mixture  entering the
 granules contains less than 500 milligrams/liter (mg/L)
 of total recoverable petroleum hydrocarbons (TRPH)
 and less  than 50 mg/L of suspended solids, InPlant
 claims that the SFC System will produce a treated water
 effluent that contains 15 rng/L or less of TRPH.  SFC
 Systems  operate at atmospheric  pressure  and are
 available  in sizes capable of treating 2.2 to 44 gallons
 per minute (gpm).   For treatment  of larger flow  rates
 (up to 1,000 gpm), the coalescing unit is manufactured
 as  a  separate stand-alone  component  from  the
 Oleofilter.  The Oleofilters designed to treat larger flow
 rates operate under low pressure [less than 30 pounds
 per square inch (psi)].   The  units can  be operated
 independently or installed in series on a single  skid.
 The latter configuration provides the same treatment
 capabilities as the SFC System.

    The SFC 0.5 System was evaluated under the EPA
 SITE  Demonstration   Program   at  a  former  oil
 reprocessing facility in Pembroke Park,  Florida.  This
 Superfund site has a layer of free  product (waste oil)
 floating on  groundwater that is contaminated with  a
 variety  of   organic  and  inorganic    constituents.
 Demonstration activities were initiated on June 2, 1994
 and were  concluded on June 18, 1994.  The  SFC 0.5
 System has  a treatment  capacity  of 2.2 gpm.  The
 waste   oil   recovered  for the  demonstration  was
 significantly more viscous  than the oil collected for the
 pre-demonstration treatability study.  Consequently, the
 feed stream to the SFC System was thinned with virgin,
 lighter weight motor oil  and then emulsified with site
 groundwater using an air-powered inline blender.  The
 unit was evaluated over five separate operating cycles
 ("runs").  The feed  stream was the same for  all runs
 except Run 4. The feed  stream for  Run 4  was a 3-to-1
 mixture of thinned oil to kerosene emulsified in  ground-
 water.  The TRPH concentration in the feed stream for
 Run 4 was two to five times higher than the concentra-
 tions for the other runs.   These differences in Run 4
 were implemented in an attempt to  resolve filter back-
 flushing difficulties  associated  with  treating  a very
 viscous oil.

    The first critical objective of the demonstration was
to evaluate whether the SFC System could remove at
 least 90 percent of the TRPH from the emulsified oil/
 water influent stream.  Data  indicate that the  SFC
 System met this goal for all runs except Run 4.

     The second critical objective was to determine
 whether the  SFC  System  could  reduce  TRPH
 concentrations in the treated water exiting the system
 to 15 mg/L or less.  When data are combined and
 evaluated for the runs where the system operated within
 normal design parameters (Runs 1 and 5), this goal was
 met.  For the other runs, the 15 mg/L threshold was
 exceeded.

     The third  critical  objective was to evaluate the
 effectiveness of the oleophilic granules by comparing
 the TRPH  concentration  in the  oil/water  emulsion
 before  and  after   passing through  the  granules.
 Combined data for the runs with similar feed streams
 (Runs 1, 2, 3, and 5) show the granules achieved a 95
 percent  reduction in TRPH concentration. A 65 percent
 reduction in TRPH was obtained in Run 4.

     Several noncritical objectives were evaluated for the
 demonstration. One of these objectives was evaluation
 of  the  relative effectiveness  of the  SFC  System
 hydrocarbon-capturing components. Results indicate
 that the  coalescing separator accounted for 45 to 62
 percent  of  the  total TRPH  removed;  the oleophilic
 granules removed the corresponding 55 to 38 percent.

     Another noncritical objective was to evaluate the
 ability of the SFC System to remove suspended solids
 (measured as  non-filterable  residue, NFR) from the
 oil/water influent. NFR removal ranged from 27 percent
 to 58 percent; NFR values in the oil/water influent were
 generally below 50 mg/L

     The  ability of the SFC System to remove selected
 semlvolatile organic compounds (SVOCs) was another
 noncritical objective.    SVOC concentrations  in  the
 oil/water influent for Runs 1, 2, 3, and 5 were too low
 to support any conclusions about removal effectiveness.
 Run 4 had higher SVOC concentrations in the oil/water
 influent.    For  this  run,  75  percent removal  of
 naphthalene   and   81  percent   removal  of   2-
 methylnaphthalene were achieved.

     During the demonstration, the SFC System did not
achieve steady-state operating conditions.  The lack of
steady-state conditions apparently resulted from treating
the  unexpectedly high-viscosity oil  during a  short-
duration  evaluation  of the  technology.  This situation
precluded the evaluation of two noncritical  objectives.
An  evaluation of the effectiveness of the  coalescing
separator at  segregating oil from water, as determined
by the percent water in the concentrated  oil effluent,

-------
could not be made since the increased agitation that
occurred during backflushing resulted in overflowing of
backflushing water into the  concentrated oil  effluent
stream.  An acceptable materials mass balance closure
could not be achieved since the amount of oil retained
in the unit was not constant across the runs.

Technology Description

    SFC systems, which  contain only one  internal
moving part (a liquid-level control float), are designed to
be explosion-proof and are  operated at atmospheric
pressure.  Figure 1 shows the configuration and cross-
sectional view of the liquid flow through the SFC 0.5
System.   The hydrocarbon/water mixture  (oil/water
influent stream) feeds into the top of the unit through
Port A,  moves downward  inside the outer shell, and
flows upward past the vertical-fin coalescing separator.
Free-floating and emulsified hydrocarbons passing over
the  coalescing fins  combine  with  droplets  already
adsorbed on the fins' surface.

    The hydrocarbon droplets continue to increase in
size  until the buoyancy of the droplets overcomes the
adsorptive forces.  The droplets then release from the
fins,  float toward the top of the unit, and are discharged
from the system through Port B as the concentrated oil
effluent stream. Final hydrocarbon filtration occurs as
the remaining emulsified and dissolved hydrocarbons
flow upward through  the center of the unit and gravity
flow through  the bed of  amine-coated,  oleophilic
granules.   The majority of  remaining  hydrocarbons
attach to the granules,  and the treated water  (treated
water effluent stream) exits the system through Port C.

     When  the Oleofilter  becomes saturated  with
hydrocarbons and suspended solids (InPlant states that
15 to 20 liters of hydrocarbons can be retained by 100
liters of   oleophilic   granules),  the  granule  bed
regenerates  itself  automatically  by  backflushing.
Backflushing  is activated when  the system reaches a
set pressure differential across the bed. The pressure
drop that initiates  backflushing can be adjusted by the
operator to optimize filtration time, while preventing filter
breakthrough.

     The  backflush cycle takes 20 minutes.  Water for
backflushing is pumped into the bottom of the system
through  Port C.   During  the  first 4 minutes of the
backflush cycle, only water is introduced. During the
next 8  minutes,  both  air  (supplied by an  external
compressor)  and  water are flushed  through the filter.
The  air increases the agitation that physically strips the
hydrocarbons  from the granules.  During the last 8
minutes of the backflush cycle, a water only rinse is
performed. The backflush water flow rate is equal to
the nominal throughput of the filter (2.2 gpm for the
SFC 0.5 System), and the air flow rate is 0.3 standard
cubic feet  per minute  (scfm) per gpm of water flow
(0.66 scfm for the SFC 0.5  System).  Therefore, the
amount of air exiting the  SFC  0.5  System during
backflushing is approximately 5 scf. The hydrocarbon/
water mixture generated during backflushing (backflush
water effluent stream) gravity flows from Port D near the
top of the  unit (not shown in Figure 1) to a sump or
holding tank. The coalesced hydrocarbons within the
mixture typically separate within 10 to 30 minutes and
can  be reprocessed through the SFC System, leaving
only the concentrated hydrocarbons to be recycled or
disposed of.

    Although the design of the vertical-fin coalescing
separator within  the SFC System is novel, the amine-
coated  oleophilic   granules   are  the   innovative
component of the  system.   The  granules  separate
emulsions  not treatable  by conventional  oil/water
separators.     The  oleophilic  granules   use  a
montmorillonite (clay) base  that has been  heated to
800°C [1]*.  The high temperature decomposes the
montmorillonite into an aluminum silicate that assumes
a crystalline, ceramic structure.  The aluminum silicate
is then crushed into granules with diameters between
0.6 and 1 millimeter.

    The granules are subsequently treated to attach the
oleophilic  amine (see Figure 2).  Through a series of
substitution reactions, an amine  molecule bonds to a
silica atom, leaving a long hydrophobic (and oleophilic)
chain (C18H33)  to which hydrocarbons are attracted [2,
pp. 15-16]. As the filtration process continues, hydro-
carbons flowing past the granules agglomerate with the
amine-attracted hydrocarbons, forming droplets.  The
hydrocarbons  remain attached to the amine, while the
separated water exits the system.  The magnitude of
hydrocarbon uptake is inversely proportional to the
compounds' solubility in water and is controlled  by a
partitioning process [3, p. 2054].

      During backflushing, the  hydrocarbon droplets
and  hydrocarbon-laden solids are physically stripped
from the amines and, along with other entrained solids,
exit  the   unit  with  the backflushing  water.   The
hydrocarbons    in   the  backflushing   water   are
predominantly coalesced and now can be removed by
conventional oil/water separation techniques. inPlant
has  installed several systems where the hydrocarbon/
* [Reference Number, Page Number]

-------
  Concentrated Oil
  Outlet (Port B)
       Coalesced Oil
 Treated Water
 Outlet/Backwash
 Water Inlet  (PortC)
  Coalescing Plates
Note: The backwash water outlet
(Port D) is not shown in this view.
Source: Adapted from SFC 0.5x
Operating Manual, 1992.
                                                                                       Oil/Water Inlet
                                                                                       (Port A)
Water for
Oleophilic Filtration
                                                                                        Oleophilic Granules
                            Figure 1. SFC 0.5 Oleofiltration System Configuration.

-------
Only the significant parts of the mineral base,
consisting of silica molecules and hydroxyl groups,
are shown below.
OH
I
SI
/ ^

OH
1
SI
/ ^
Substitution of chlorines (chloride
hydroxyl groups.
Reaction with SO2CL
CL
I
SI
/ X

2
CL
1
SI
/ X

OH
1
SI
/ X
ions) for


OH
1
SI
/ X
Substitution of amines for chlorines.
Reaction with RNH2amine. R=C18
H R
X /
N
II
SI
/ X
H R
X /
N
II
SI
/ X
H33


OH
1
SI
/ X
  Figure 2. Generation of oleophilic amine-coated
              granules.  Source: |2]
water mixture from backflushing  is fed  back into the
system, and the coalesced hydrocarbons are removed
by the vertical-fin coalescing separator. Any emulsified
hydrocarbons are captured by the oleophilic granules.
This approach eliminates the need for disposal of the
hydrocarbon/water mixture resulting from backflushing.

Technology Applicability

    The SFC  System is reportedly capable of treating
virtually an insoluble hydrocarbon/water mixture.  The
stated advantage of this technology over other oil/water
separation  techniques  is its  ability  to  separate
mechanical and several types of  chemical emulsions.
InPlant claims that the SFC System can  remove TRPH
from hydrocarbon/water emulsions to levels below 15
mg/L  when  the  emulsion  reaching  the  granules
contains less  than 500 mg/L TRPH and less than  50
mg/L of suspended solids. According to  InPlant, the
amine-coated granules have been proven effective  on
a wide  variety  of hydrocarbons  including  gasoline;
crude oil;  diesel; benzene, toluene, ethylbenzene, and
xylene (BTEX) compounds; and polynuclear aromatic
hydrocarbons (PAHs).  The granules reportedly also
remove chlorinated hydrocarbons such as pentachloro-
phenol (PCP), polychlorinated  biphenyls (PCBs), and
trichloroethane (TCA), as well as vegetable and animal
oils.

    The ability of the oleophilic granules to separate
hydrocarbon/water emulsions  and reduce dissolved
hydrocarbon concentrations to levels consistent with
other  secondary  treatment systems  indicates  the
potential for the SFC System to be used  in conjunction
with other  treatment technologies.  Site remediation
techniques,  such as steam injection-vapor extraction
and soil  flushing, can generate  hydrocarbon/water
emulsions  that  must be  treated.   Pumps  used  in
transferring oily water also can produce emulsions that
must be separated prior to further treatment. The SFC
System  can  be  employed  in  these  and  other
applications including the remediation of contaminated
groundwater,    in-process  oil/water  separation,
wastewater  filtration,  onsite  waste  reduction and
recovery, and bilge and ballast  water treatment.

    When used as a component of a treatment train,
the technology can significantly reduce hydrocarbon
loading to other downstream treatment equipment such
as air strippers and carbon filtration units. This reduced
loading results in increased on-line time and decreased
operating and maintenance costs for  the treatment
train.  Depending on local pretreatment standards,
treated  water  exiting  the  SFC  System  may  be
acceptable for introduction to the sanitary sewer system
without further treatment.

    Table 1 addresses the performance of the SFC
System based upon the nine evaluation criteria used for
decision-making  in the Superfund  feasibility study (FS)
process.  If  the SFC System is used as a component in
a treatment train,  evaluation of the entire train also
should be performed.

Technology Limitations

    The   Oleofiltration   technology   concentrates
contaminants by separating free, emulsified, and some
dissolved hydrocarbons from  water.   Although  the
toxicity of the water phase decreases, the toxicity and
mobility  of  the  concentrated   hydrocarbons  are
unchanged. The concentrated hydrocarbons must then
be  further  treated or disposed.   Even under ideal
conditions,  the  treated  water  typically will  contain
between  4  and  15  mg/L  of TRPH, requiring further
treatment prior to release at some sites.

    Although the  oleophilic granules  are  relatively
durable, collision between granules during filtration and

-------
Table 1.  Nine Evaluation Criteria for the SFC System
Evaluation Criteria
Overall Protection of Human Health
and the Environment
Compliance with Federal applicable
or relevant and appropriate
requirements (ARARs)
Long-Term Effectiveness and
Performance
Reduction of Toxicity, Mobility, or
Volume through Treatment
Short-Term Effectiveness
Implementability
Cost
State Acceptance
Community Acceptance
Performance
• Provides both short-term and long-term protection by reducing
contaminants in groundwater.
• Prevents further groundwater contamination and offsite migration
caused by emissions during treatment.
• Demonstrated capability of reducing TRPH concentrations in
oil /water mixtures to 15 mg/L
• Concentrates but does not destroy contaminants.
• Effluent needs to be treated further to meet Federal Drinking Water
Standards if it is to be re-injected directly into the ground.
• Effluent may meet pretreatment standards for release to the local
publicly-owned treatment plant (POTW).
• May have to meet substantive requirements of a Resource
Conservation and Recovery Act (RCRA) treatment permit if treating
hazardous wastes.
• May have to meet substantive requirements of a Clean Air Act (CAA)
permit for air discharge during backflushing If volatile organic
compounds (VOCs) are present.
• Concentrated oil effluent may be regulated under the Toxic
Substances Control Act (TSCA) if poiychlorinated biphenyls (PCBs)
are present.
• Residuals treatment or recycling may be required (effluent water,
concentrated oil, oily water from backflushing).
• The technology concentrates contaminants, reducing waste volume,
but does not change the contaminants' mobility or toxiclty.
• Community and workers will be protected because the system is
almost entirely self-contained.
• Most systems are shipped pre-assembled or as modules that are
easily connected.
• Pretreatment of feed stream is typically not required.
• System is explosion-proof.
• If VOCs are present, a release (5 to 106 scfm) of contaminated air
will occur during backflushing.
• Additional treatment options may be needed for residuals.
• Oleophilic granules usage life is shortened if treating solutions with
pH>10.5 (granules become brittle) or chlorinated solvents with
concentrations >100 mg/L (weakens amine bonds).
• Backflush initiation needs to be adjusted, If treating oils of various
viscosities, to prevent breakthrough prior to backflushing.
• The cost to remediate 50 million gallons of contaminated
groundwater (22-gpm system with 95% on-line time) is approximately
$2.57 per thousand gallons.
• Since this system will most often be used as a component in a
treatment train, acceptance is tied to overall treatment acceptability.
• Should be generally acceptable to the public since emissions during
treatment are minimal.

-------
backflushing results in breakage. Broken granules that
are small enough to pass through the retention screen
are discharged from the system during backflushing.
Assuming a backflush  frequency  of  every  10 hours,
InPlant  states that approximately 8 percent of the
granules  must  be  replaced  every  12  months  of
operation.

    InPlant reports that  the  oleophilic granules are
sensitive to two chemical conditions, both of which
shorten the operational life of the granules. Treatment
of solutions having a pH greater than 10.5 for extended
periods of time makes the granules more brittle.  The
increased  breakage  caused  by  this  condition  is
estimated to be an additional 4 percent  every 12
months of operation.   Treatment of solutions  with
chlorinated solvents present in concentrations greater
than 100 mg/L weakens the amine bonds.  A similar
attrition rate (an additional 4 percent every 12 months)
is reportedly caused by prolonged treatment of these
solutions.

    The SFC System is reportedly less  effective  in
treating   chemical   emulsions  containing  anionic
surfactants than other types. Anionic surfactants affect
the ability of the granules' amine coating to attract and
retain hydrocarbons.  InPlant states that  use of SFC
Systems  for  the  treatment   of  hydrocarbon/water
emulsions created by anionic surfactants resulted  in
TRPH concentrations in the treated water of 50 to 80
mg/L.    Although  not  evaluated during  the  SITE
demonstration,  the  granules reportedly  are  more
effective at  removing  hydrocarbons from  chemical
emulsions containing cationic  or nonionic  surfactants.

    Although the SFC  System appears to  effectively
treat  oils of   varying  viscosities  and  densities,
adjustments to the backflushing cycle must be made to
reduce the amount of operator oversight required.  The
pressure at which the  backflushing cycle is  initiated
must  be  adjusted to  maximize filtration time  while
preventing breakthrough of the hydrocarbons prior to
backflushing. During the SITE demonstration, the SFC
System apparently exhibited  breakthrough prior to
backflushing when a kerosene and oil mixture was used
as the feed  oil.  InPlant reportedly has implemented
modifications  to  the   system  that  allow  in-field
adjustment of  the pressure at which backflushing  is
initiated. These modifications, combined with periodic
monitoring of system performance, should eliminate the
difficulties.

    Treatment  of  high  viscosity  oils may  foul the
granules,  preventing effective backflushing.  InPlant
claims that the use of hot water for backflushing or the
addition of a steam coil attachment to the system will
reduce the viscosity of most retained  oils and allow
normal backflushing.  During the SITE demonstration,
all but one of the runs used a very viscous oil that had
been thinned with virgin motor oil.  Performance of
Runs 1, 2, and 3 resulted in fouling of the granules,
which had to be removed, washed in mineral spirits,
and  reinstalled.   Subsequent  use  of the  hot  water
(approximately 200 °F) increased the effectiveness of
the backflushing. Treatability studies encompassing the
full  range of oil  properties at a site,  along with
provisions for hot  water backflushing,  if indicated,
should resolve backflushing difficulties.

   According to InPlant, when the TRPH concentration
in the pre-granule water exceeds 500 mg/L, the TRPH
concentration in the treated water effluent may exceed
15 mg/L  Run 4 of the demonstration had an average
TRPH concentration in the pre-granule water of 1,242
mg/L. The treated water effluent contained an average
concentration  of  39  mg/L  (these  averages  do  not
include   concentrations   measured    after  filter
breakthrough). This reduction represents a 97 percent
removal of TRPH.  Pilot-scale treatability testing prior to
full-scale implementation should determine the ability of
the unit to meet site-specific performance goals.

Process Residuals

   The SFC System generates three process streams:
treated water, concentrated  contaminants (during the
demonstration this wastestream was a  concentrated
waste  oil),   and   hydrocarbon-laden  water  from
backflushing. Additionally, if the feed stream contains
volatile organic compounds (VOCs), air emissions will
be generated  during  backflushing.   Under optimum
conditions, the treated  water will  reportedly  contain
between  4 and 15  mg/L of TRPH.  Therefore, this
process stream may need to be further treated with a
tertiary process onsite or transported offsite for further
treatment.   The  concentrated  hydrocarbon  effluent
stream can be transported and disposed of offsite. If
the concentrated hydrocarbon is waste oil and meets
the waste oil specifications of 40 CFR 279, it can be
used as fuel.  Two options exist for the  water from
backflushing.  This water can be  fed  back into  the
system where the  coalesced  hydrocarbons  will be
removed  and the water filtered.  Alternately, the water
from  backflushing  can  be  transported  offsite  for
treatment  and disposal.

   Depending on the size  of the  SFC System, air
emissions during backflushing range from 5 to 106 scf.
If the feed  stream contains VOCs, a percentage of them
will become entrained in the backflushing air and exit
through the top of the system. Depending on the types
and concentrations of VOCs and applicable regulations,

-------
 emissions  controls such as  carbon filters  may  be
 required.

 Site Requirements

     Site requirements for the operation of the SFC
 System include a level  area, electricity,  water, and
 compressed air. The SFC System must be operated on
 a level, non-shifting surface.  A 9-square-yard pad of 6-
 inch reinforced concrete will support the largest SFC
 units. Additional space for storage of backflush influent
 and effluent water must  be available.  If  potable water
 is  used for backflushing, water lines or a service  for
 filling the  water tank between backflushes must  be
 available.  A water tank,  with capacrty in  excess of the
 backflush volume, must be provided. Storage capacity
 for the concentrated hydrocarbons  and  treated water
 must be available (if the water is not being treated or
 discharged immediately).  Electrical  power, consisting
 of  4 kilovolt-amp (kVA),  460/230-volt, 3-phase  service
 must be available to operate the largest SFC Systems.
 Smaller systems require 40-amp,  220-volt  service.
 Alternately, electrical power  could be  supplied by  an
 onsite mobile generator.

    Current designs of the SFC System use pneumatic
 controllers,  requiring  approximately  0.5  scfm  of
 compressed air.  Additionally,  the backflushing cycle
 requires compressed air to  increase agitation of the
 granules.  A source of  compressed  air  capable  of
 producing  a volumetric  flow rate of 15 scfm and a
 minimum air pressure of 100 psi will supply sufficient air
 for both purposes on any size SFC System.  InPlant has
 recently begun replacing  pneumatic controllers with
 programmable logic controllers on SFC Systems.

    Depending on the viscosity of the oil, hot water  or
 steam may be required for effective backflushing.  A
 portable hot water washer or steam generator therefore
 may be required.

 Performance Data

    The SFC Oleofiltration System was accepted into
the SITE Demonstration  Program in  December 1992.
The Petroleum  Products Corporation  (PPC)  Superfund
site in  Pembroke Park,  Florida was chosen as the
demonstration  site.  Accidental releases during the
operation of this former oil reprocessing facility resulted
in the deposition of approximately 29,000 gallons of free
product (waste oil)  on  the groundwater surface.  The
groundwater underneath the oil  is contaminated with a
variety of organic and inorganic constituents.

    Prior to the demonstration, samples of oil from the
site were sent to NATGI for treatability studies.  Aliquots
of  the oil were combined with different volumes of
 water, mixed with a  blender, and poured through
 separatory  funnels  containing oleophilic  granules.
 Samples of the water exiting the funnels were analyzed
 for oil and grease by NATGI using EPA Method 413.1
 [4].  Results of the study, presented in Table 2, showed
 the granules to be effective at removing oil and grease
 from the oil/water emulsions.

      Table 2.  NATGI Treatability Study Results
  Six one liter samples of groundwater were contaminated with
  20, 100, 300, 500, 2,000, and 10,000 mg/L of oil respectively,
  mechanically emulsified with a high-speed mixer for 1 minute,
  and manually poured into separatory funnels containing
  approximately 0.5 liter of amine-coated ceramic granules.  The
  effluent (output) was analyzed for Oil and Grease using EPA
  Method 413.1. [4]
Input
(mg/L)
20
100
300
500
2,000
10,000
Output
(mg/L)
2.5
4.0
5.0
3.1
3.6
2.5
Percent Removal
(mg/L)
87.5
96.0
98.3
99.4
99.8
99.9
    The SFC 0.5 Oleofiltration System (2.2 gpm) was
 evaluated during the SITE demonstration in June 1994
 at the PPC site. Since the site did not have significant
 amounts of oil  emulsified  in water, an artificial feed,
 consisting  of recovered waste oil  emulsified  in  the
 contaminated groundwater, was formulated to test the
 system.  The contaminated groundwater was obtained
 by diverting a small stream  from the site's  full-scale
 remediation system. This groundwater feed exited the
 bottom of  the full-scale oil/water  separator, passed
 through a flow meter, and entered an air-powered inline
 blender used to create the emulsion to be used in the
 demonstration.

    The feed oil  was collected from a sump where the
 viscous oil had risen to the surface.  Approximately 30
 gallons of highly viscous oil were mixed with 15 gallons
 of 10W-30-weight motor oil to reduce the viscosity of
 the oil. The mixed oil had an average viscosity of 56.3
 centistokes (cs). A peristaltic pump was used to deliver
 the oil through a feed line to the inline blender.  The
 inline blender then created the oil/water emulsion that
 was fed into the system.  After passing through the SFC
 System, the treated water effluent was returned  to the
full-scale oil/water separator. Although the SFC System
was reportedly capable  of  reprocessing the oil/water

-------
mixture resulting from backflushing, the water layer
from  backflushing  was   returned  to  the  full-scale
oil/water separator. The concentrated oil effluent and
the oil layer from backflushing were stored in drums for
offsite disposal.

   Samples were collected from the groundwater feed,
oil feed, emulsified oil/water influent,  water prior to
entering the granules (pre-granule water), treated water
effluent,  backflushing effluent, and concentrated  oil
effluent.

   The demonstration consisted of five separate runs.
Runs 1, 2, 3, and 5 used the mixed feed oil.  Run 4
used a 3-to-1  mixture of  the previously  mixed oil to
kerosene.  The feed oil  for  Run  4 had an average
viscosity of 30.1  cs. Samples were collected for TRPH
analysis using  EPA  Method 418.1  [4].   Additional
samples were collected and analyzed for NFR, SVOCs,
and percent water using  EPA Method  160.2 [4], EPA
Method 8270 [5], and ASTM Method D95-83 [6]. The
average TRPH concentrations for the oil/water influents
ranged from 322 to 2,802 mg/L

   Due to operational difficulties associated with filter
backflushing, only  one complete run  (Run 1) was
accomplished.  Runs 2 and 3  were shortened because
the backflushing cycle preceding each run did not clean
the granules sufficiently to allow the pressure differential
across  the  granule  bed   to   reset  to  InPlant's
specifications of zero inches of mercury (in.  Hg). The
backflush triggering  pressure   of 16  in.  Hg  was
consequently  reached   sooner.   The  operational
difficulties were apparently caused by the high viscosity
and solids content of the feed oil, which were different
from the  oil  provided to InPlant  for  the  treatability
studies.  InPlant claims that adjustments prior to  unit
delivery and the addition  of a steam coil attachment
would have resolved the difficulties.

   Run 4 was terminated when visible oil appeared in
the treated water effluent.  Analytical results confirmed
that  filter breakthrough had  occurred.  Run 5 was
terminated when the level of  pre-granule water in the
unit had risen to the height where it was discharging
through the backflush water outlet.  Additionally, it was
thought that visible oil appeared again in the treated
water  effluent  (analytical  results  indicated  that  this
conclusion was  inaccurate).  Table 3 presents  TRPH
results for the oil/water influent, pre-granule water, and
treated water effluent for all five runs. Table 4 presents
results for NFR, naphthalene, and 2-methylnaphthalene
for the  oil/water influent  and treated  water effluent.
Results from the first sample collected in each run have
not been  presented since the collection time (t =  10
minutes) was less than the calculated residence time of
the unit (i.e., water entering the unit at initiation of the
run had not yet reached the treated water sample port).
Table 5  presents a summary  of project objectives,
results, and conclusions for the demonstration.

    Due to operational differences among some of the
runs,  demonstration data have been  evaluated using
several scenarios.  Since Runs 1, 2,  3, and 5 used the
same feed oil, data from these  runs were pooled and
evaluated together.  Within this group,  only Runs 1 and
5  were  initiated   with  the  granules  backflushed
sufficiently for the initial pressure differential across the
granule bed to approach InPlant's specification of zero
in.  Hg.    Consequently,  evaluation  of demonstration
objectives state a result for the pooled data from Runs
1,  2,  3, and 5 (13  data  points), and  a result for the
pooled data from Runs 1 and 5 only (8 data points).

    Since Run 4 used a different type of feed oil and oil
feed rate, data from this run were evaluated separately.
During this run, the concentration of TRPH  present in
the  pre-granule  water  exceeded   InPlant's  stated
limitation   of   500   mg/L      Consequently,  the
demonstration objective of achieving 15 mg/L or less in
the treated water effluent was not evaluated.  Addition-
ally, InPlant claims that the pressure differential across
the granule bed at which backflushing was triggered (16
in.  Hg)  was  set  to accommodate  the  500  mg/L
maximum TRPH  concentration, and the higher concen-
tration  was   responsible  for  the  apparent  filter
breakthrough. Accordingly, Run 4 was evaluated using
the data for the entire run (5 data points) and also using
only the data prior to filter breakthrough (3 data points).

    The  SFC  System  did not achieve  steady-state
operating conditions during the demonstration.  This
situation precluded the evaluation of two  noncriticai
objectives. An evaluation of the effectiveness  of the
coalescing separator at segregating oil from water, as
determined by the percent water in the concentrated oil
effluent,  could  not be  made since  the  increased
agitation that occurred during backflushing resulted in
overflowing of backflushing water into the concentrated
oil  effluent stream.   An acceptable  materials mass
balance  closure  could  not  be achieved  since  the
amount of oil retained  in the unit was not constant
across the runs.

    InPlant has  provided performance data from a
bench-scale  study   of the  ability  of the  oleophilic
granules to remove TRPH,  BTEX, and PAHs [7]. The
study,   conducted  on tank water bottoms from a
condensate  tank at a  bulk  petroleum storage and
transfer facility in The Netherlands, achieved petroleum
hydro-carbon concentrations in the outlet samples of
1.43 and 2.49 mg/L for times t=10 minutes and t=105

-------
                                 Table 3. Summary of TRPH Analyses
Elasped
Time
Run (min)
1 60
1 120
1 180
1 240
1 300
2 30
2 60
2 90
3 20
3 40
4 45
4 90
4 135
4 180
4 240
5 45
5 75
5 105
5 135
Influent
Concentration
(mg/L)
842
989
1240
1120
1170
366
322
484
988
981
1991
2680
2004
2802
1630
681
NA
565
2448
Pre-granule
Concentration
(mg/L)
691
499
651
445
487
301
227
261
386
137
1260
997
1470
1302
955
456
351
191
189
Effluent
Concentration
(mg/L)
29.4
20.3
13.8
10.9
17.4
16.8
32.2
25.7
25.0
20.7
43.3
26.7
47.5
484
1470
17.2
14.7
7.4
10.1
NA - Not analyzed by laboratory.
                                                10

-------
                        Table 4. Summary of NFR and Specific SVOC Analyses
Run
1
1
1
1
1
2
2
2
3
3
4
4
4
4
4
5
5
5
5
Influent
NFR
(mg/L)
19
11
18
24
20
30
23
26
31
37
32
19
40
34
24
29
22
23
20
Effluent
NFR
(mg/L)
14
6
7
6
11
14
14
13
10
16
17
29
20
21
12
9
9
9
6
Influent
Naphthalene
fc>g/L)
13
***
12
***
6j
12
***
12
13
***
210
***
250
***
100
39 j
***
20
***
Effluent Influent
Naphthalene 2-Methylnapnthalene
G/g/L) 0/g/L)
10 u
***
10 u
***
10 u
10 u
***
10 u
10 u
***
10 u
***
25
***
90
20
***
10 u
***
11
***
10
***
5j
7]
***
7]
7j
***
320
***
410
***
140
38]
***
15
***
Effluent
2-Methylnapnthalene
0/g/L)
10 u
***
10 u
***
10 u
10 u
***
10 U
10 u
***
10 u
***
21
***
130
14
***
10 u
***
u    below detection limits
j     estimated value
***   no sample collected
                                                  11

-------
                         Table 5.  Summary of Project Objectives,  Results, and Conclusions
           Objective
                                                  Results
                                                                 Conclusions
  Critical Objectives

  Evaluate claim of 90%
  minimum removal of TRPH2
  from oil/water emulsion.
  Evaluate claim of 15 mg/L
  maximum TRPH concen-
  tration in effluent. (Test
  hypothesis that sample
  mean is not statistically
  significant from 15 mg/L at
  the 90% confidence interval.)

  Determine TRPH removal
  effectiveness of oleophilic
  granules.
  Noncritical Objectives

  Determine the relative
  contributions to TRPH
  removal of the coalescing
  unit and oleophilic granules.
  (Determine percentage of
  total TRPH removal
  accomplished by the
  coalescing unit and by
  granules.)
  Evaluate the SFC System's
  ability to remove suspended
  solids from the oil/water
  influent.
  Examine the difference in %
  moisture between feed oil
  and oil effluent.

  Evaluate the ability of the
  SFC System to remove
  naphthalene, 2-methyl-
  naphthalene and 1,2-
  dichlorobenzene.
  Determine whether mass
  balance closures of 80 to
  120% can be achieved for
  TRPH and total materials.

  Establish a +50 to -30%
  treatment cost estimate
 The overall TRPH removals were:
  98%-Runs 1,2,3, and 5
    (98% - Runs 1 and 5 only)
  81% - Run 4, all data
    (98% - Run 4, data prior to breakthrough)

 The average effluent concentrations were:
  18.7 mg/L - Runs 1,2,3, and 5
    (15.7 mg/L - Runs 1 and 5)
  414.3  mg/L - Run 4, all data
    (39.2 mg/L - Run  4, data prior to
    breakthrough)


 The TRPH removals of granules were:
  95% - Runs 1,2,3, and 5
    (96% •• Runs 1 and 5)
  65% - Run 4, all data
    (97% Run 4, data  prior to breakthrough)
The TRPH removals for coalescing unit:
  61% - Runs 1,2,3, and 5
   (62% • Runs 1 and 5)
  57% -Run 4, all data
   (45% • Run 4, data prior to breakthrough)

The TRPH removals for granules were:
  39% - Runs 1,2,3, and 5
   (38% - Runs 1 and 5)
  43% - Run 4, all data
   (55% - Run 4, data prior to breakthrough)

The NFRl! removals were:
  57%- Runs 1,2,3, and 5
   (58% - Runs 1 and 5)
  34% - Run 4, all data
   (27% - Run 4, data prior to breakthrough)

Could not be evaluated as system did not reach
steady-state conditions during demonstration.


The SVOC4 removals were:
  75% - Naphthalene for Run  4, all data
  81% - 2-Methylnaphthalene  for Run 4,
        afl data
Mass balance closures were not possible due to
lack of steady-state conditions.
Cost for treating 50,000,000 gallons of water
(95% on-line time) is $2.57 per 1,000 gallons
 Runs 1, 2, 3, and 5 met the objective. Runs 1 and 5
 met the objective.  Overall objective not met for
 Run 4 using all data.  Objective met for Run 4 using
 data prior to breakthrough.
 The average of Runs 1, 2, 3, and 5 is statistically
 different from the objective.  The average of Runs 1
 and 5 is not statistically different from the objective.
 For Run 4, the TRPH concentration in the pre-
 granule water exceeded the developer's stated
 limits. Therefore, no conclusions about this
 objective are stated for Run 4.

 The granules were able to significantly reduce TRPH
 concentrations.
 For Runs 1, 2, 3, and 5, the coalescing unit
 removed more TRPH than the granules by a factor
 of 1.56.  For Runs 1 and 5, the coalescing unit
 removed more TRPH than the granules by a factor
 of 1.63.  For Run 4 using all data, the coalescing
 unit removed more TRPH than the granules by a
 factor of 1.32. For Run 4 prior to breakthrough, the
 granules removed more TRPH than the coalescing
 unit by a factor of 1.22.
 For Runs 1, 2, 3, and 5 and Runs 1 and 5, the NFR
 removal was significant. The NFR removal was less
 for Run 4 using all data and for Run 4 using data
 prior to breakthrough.
No conclusions can be made regarding the ability
of the coalescing unit to produce a low-moisture,
concentrated oil stream.

No conclusions can be made regarding the removal
of specific SVOCs for Runs 1, 2, 3, and 5 due to low
influent concentrations.  The  SFC System
significantly removed both naphthalene and 2-
methylnaphthalene during Run 4.  (1,2-
dichlorobenzene was not present above detection
limit.)

No conclusions can be made regarding either TRPH
or total mass balance closure.
Cost estimates are highly dependent on site-specific
factors.  Actual costs may vary significantly.
1   Indicated results obtained by combining data from specified Runs         3
    (e.g., "Runs 1 and 5" Indicates data pooled from those Runs only)         4
2   TRPH Is total recoverable petroleum hydrocarbons (EPA Method 418.1)
                                   NFR is non-filterable residue (a measure of suspended solids)
                                   SVOC Is semlvolatlle organic compound
                                                              12

-------
minutes, respectively. The study also indicated effective
removal of PAHs but less effective removal of BTEX.

Technology Status

   The SFC System is currently being used in industrial
applications including:

   • Treatment of process water at a  laboratory  in
     Oildale, California
   • Treatment of wash water effluent at a car wash -
     the effluent  reportedly  meets the  pretreatment
     water standards for Santa Clara, California
   • Treatment of wash  rack waste water in Ventura,
     California
   • Treatment of storm water runoff in order to meet
     National Pollution Discharge Elimination System
     (NPDES) regulations in Houston, Texas

Disclaimer

   Although the technology conclusions presented  in
this report may not change,  the data have not been
reviewed  by   EPA  Risk   Reduction   Engineering
Laboratory Quality Assurance personnel.

Sources of Further  Information

EPA Contact:
SITE Project Manager
Laurel Staley
U.S. EPA
Risk Reduction Engineering Laboratory
26 West Martin Luther King Drive
Cincinnati, Ohio  45268
(513) 569-7863
Technology Contact:
Cathryn Wimberly
Aprotek
3316Corbin Way
Sacramento, CA  95827
(916) 366-6165

References

1.  Aprotek   Product   Literature,   Aprotek,   Inc.,
   Sacramento, California, 1993.
2.  Subra   Company   Product   Literature,
   Company, New Iberia, Louisiana, 1993.
Subra
3. Smith, J.A. and  P.R.  Jaff6.    Comparison of
   Tetrachloromethane Sorption to an Alkylammonium
   Clay and an Alkyldiammonium Clay.  Environmental
   Science and Technology, Vol.  25, pp. 2054-2058,
   1991.

4. U.S. Environmental Protection Agency. Methods for
   Chemical  Analysis of Water and Wastes.   EPA
   600/4-79-020, 1983.

5. U.S.  Environmental  Protection  Agency.    Test
   Methods for Evaluation of Solid Waste, Third Edition.
   SW-846, December 1987.

6. American Society for Testing and Materials. Annual
   Book of ASTM Standards, 1992.

7. Treatability study performed by HEAD Consultancy
   at the University of Twente, The Netherlands, May
   1993.
                                                  13

-------

-------

-------
United States Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, Ohio 45268

Official Business
Penalty for Private Use
$300
     BULK RATE
POSTAGE & FEE PAID
        EPA
  PERMIT NO. G-35
EPA/540/R-94/525a

-------