United States /
              Environmental Protection
              Agency
              Industrial Environmental
              Research Laboratory
              Cincinnati OH 45268
EPA-600/2-78-069
April 1978
              Research and Development
£EPA
Oil/Water Separation:
State of the Art

Environmental Protection
Technology Series

-------
                RESEARCH REPORTING SERIES

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

      1.   Environmental  Health Effects Research
      2.   Environmental  Protection Technology
      3.   Ecological Research
      4.   Environmental  Monitoring
      5.   Socioeconomic Environmental Studies
      6.   Scientific and Technical Assessment Reports (STAR)
      7.   Interagency  Energy-Environment Research and Development
      8.   "Special"  Reports
      9.   Miscellaneous Reports

This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
 NOLOGY series. This series describes research performed to develop and dem-
 onstrate instrumentation, equipment, and methodology to repair or prevent en-
 vironmental degradation from point and non-point sources of pollution. This work
 provides the new or improved technology required for the control and treatment
 of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia  22161.

-------
                                                 EPA-600/2-78-069
                                                 April 1978
                  OIL/WATER SEPARATION
                    STATE-OF-THE-ART
                           by

                   Fidelis A. Osamor
                Robert C.  Ahlert, Ph.D.
   Department of Chemical and Biochemical Engineering
      Rutgers, The State University of New Jersey
            New Brunswick, New Jersey 08903
                   Grant No. R803978
                    Project Officer

                  Leo T. McCarthy, Jr.
        Oil & Hazardous Materials Spills Branch
Industrial Environmental Research Laboratory-Cincinnati
                Edison, New Jersey 08817
      INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
           OFFICE OF RESEARCH AND DEVELOPMENT
          U.S. ENVIRONMENTAL PROTECTION AGENCY
                 CINCINNATI, OHIO 45268

-------
                                DISCLAIMER


     This report has been reviewed by the Industrial Environmental
Research Laboratory, U.S. Environmental Protection Agency,  and approved for
publication.  Approval does not signify that the contents necessarily
reflect the views and policies of the U.S. Environmental Protection Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
                                    ii

-------
                                 FOREWORD
     When energy and material resources, are extracted, processed, con-
verted, and used, the related pollutional impacts on our environment and
even on our health often require that new and increasingly more efficient
pollution control methods be used.  The Industrial Environmental Research
Laboratory - Cincinnati (IERL - Ci) assists in developing and demonstrating
new and improved methodologies that will meet these needs both efficiently
and economically.

     Effluent discharge guidelines for oil have been established for
existing onshore and offshore industries, and performance standards have
been stipulated for new point sources.

     This report identifies, organizes, and interprets technical and
commercial literature resources on oil/water separation.  As such, this
state-of-the-art report will be most useful to regulatory personnel of
Federal and State agencies in assessing the capabilities of existing
technology to meet standards established for the control of oil discharges.
                                David G. Stephan
                                    Director
                  Industrial Environmental Research Laboratory
                                   Cincinnati
                                    iii

-------
                                ABSTRACT

     This report reviews the state-of-the art for oil/water separating
devices and processes.  Devices and processes are classified according to
the primary mechanism that induces separation of oil/water mixtures.  The
basic concepts, specific design features, operational conditions, and
limitations of each category are discussed.

     Literature on test evaluation of a variety of devices is critiqued on
the basis of actual or potential success in treating various oil/water
systems states.  No single technique can separate all oil/water system
states efficiently.  Specific deficiencies in existing technology have
been identified.

     Reverse osmosis, ultrafiltration, and carbon adsorption possess great
potential, but high equipment and operational costs will continue to limit
their use for treating oily wastestreams.  A combination of devices in a
process chain is therefore necessary for production of effluents with
desired discharge quality (<10 ppm of oil).  The specific characteristics
of an oily wastewater determine the combination of devices that will yield
required effluent concentrations.

     This report was submitted in fulfillment of Research Grant No.
R803978 by Rutgers University under the sponsorship of the U.S.
Environmental Protection Agency.  The report covers the period July 1,
1975, to June 30, 1977, and work was completed as of July 31, 1977.
                                      IV

-------
                                 CONTENTS

Foreword	ill
Abstract	iv
Figures	vi
Tables	vi
Acknowledgments	vii

   1.  Introduction	    1
            Background 	    1
            Objective of report  	    2
            Approach	    2
   2.  Conclusions	    4
   3.  Recommendations 	    8
   4.  System Characterization 	   10
            Oil/water systems  	   10
            Free oil	10
            Dispersed oil (emulsions)   	10
            Chemically stabilized dispersions  	  .......   11
            Molecularly dissolved oil   	   11
            Solubilized oil	12
            Oil-coated suspended solids  	   12
   5.  Characterization of Oily Wastewaters	13
            Ballast water	   13
            Tank-cleaning water  	   14
            Bilge water	14
            Oil-field production water 	   14
            Summary	14
   6.  Devices and Processes	16
            Technology	16
   7.  Critical Review of Selected Literature  	   54
   8.  Selected Manufacturers of Oil/Water Separating Equipment  ....   81
   9.  References	84
  10-  Bibliography	93

-------
                                 FIGURES


Number                                                              Page

  1    API Oil/Water Separator	20

  2    Shell Parallel-Plate Interceptor  	 22

  3    Shell Corrugated-Plate Interceptor  	 23

  4    Total Pressurlzation System 	 31

  5    Partial Pressurization System 	 32

  6    Recycle Pressurlzation System 	 33

  7    A Complete Liquid/Liquid Coalescing System  	 44

  8    Coanda-Effect Separator 	 50

  9    Orbiting Buttress Threaded Device 	 52

 10    Vortex-Flow Separator	57


                                 TABLES


Number                                                              Page

  1    Potential of Separation Techniques to Separate Various Oil/
         Water System States	5

  2    Oil Removal in API Separators	25

  3    Oil and Suspended-Solids Removal in Gravity-Type Separators  . 26

  4    Estimated Effluent Quality from Primary Oil/Water Separation
         Processes	26
                                     VI

-------
                            ACKNOWLEDGEMENTS
     The authors are indebted to Frank Freestone and Leo McCarthy of the
U.S. Environmental Protection Agency, Edison, New Jersey, for their
suggestions and assistance.
                                      vii

-------
                                SECTION 1

                               INTRODUCTION
BACKGROUND

     The rising demand for energy by a growing population has led to con-
tinued increases in world production of crude oil.  As an immediate conse-
quence, there is a worsening problem of oil pollution of waterways and the
marine environment'by discharges of oily wastewaters.

     There are various sources of oily substances, and the following key
sources have been identified:

     a) natural seeps;
     b) petroleum mining and refining operations;
     c) accidental oil spills;
     d) discharges from transporting vessels;
     e) discharges from chemical and industrial plants; and
     f) stormwater runoff.

The magnitude of the oily waste problem has been described in the literature
(Boesch et al., 1974; SCEP, 1970), but data on the amount of the discharge
from these key sources are at best speculative.

     In response to public demand for a clean environment, laws and regula-
tions have been promulgated.  Effluent discharge guidelines have been
established for existing onshore and offshore industries, and performance
standards have been stipulated for new point sources (U.S. EPA, 1974, 1975).

     Industries are approaching oil pollution abatement by process modifica-
tions and the use of advanced waste treatment technology for end-of-pipe
treatment.  Secondary and tertiary recovery methods for crude oil are
minimizing the volume of production water generated during production
operations.  New ship designs, incorporating separate holding tanks for bilge
and ballast water, and use of the load-on-top procedure have cut down on the
volume of oily wastewater discharged at sea.  Separation devices for oil/
water mixtures are being installed onboard ships, and there is an increasing
number of dockside treatment facilities for bilge and ballast waters.
However, large volumes of oily wastewaters must be treated by even more
advanced methods to meet effluent standards before discharge into United
States waters.

-------
OBJECTIVE OF REPORT

     Strict pollution control laws place heavy emphasis on the development
of economical and reliable devices for separation of oil/water mixtures.
The objective of this report is to review the state-of-the-art for oil/water
separation devices and processes.   The scope of this investigation will be
the technology available for shipboard and shoreside treatment of bilge and
ballast waters and the oily wastewaters associated with crude oil produc-
tion.  Separation devices will be classified according to the primary mech-
anism that induces separation of oil/water mixtures.  The basic concepts of
each class of equipment, specific design features, operational conditions,
and limitations of each device will be discussed.  An evaluation of each
category of equipment will be based on the overall performance standard.

     Finally, a list of some major manufacturers of commercially available
equipment, presently used worldwide for oil/water separation, will be
compiled.
APPROACH

     This project will identify, organize, and interpret technical and com-
mercial literature resources on oil/water separation.

     The first effort will be a description of the various states in which
oil can co-exist with water after intimate contact of  both liquids.   Each
of the states will be characterized in operational terms,  as well as in the
physical-chemical sense.  Next, the process streams to be  considered—oily
wastewaters generated onboard ship, oil in process water,  and formation
water from crude oil production—will be characterized based on oil/water
system states.  Other parameters to be considered for  an efficient separa-
tion will be outlined.

     The second phase will be devoted to a review of the state-of-the-art
of separation devices and processes.  A distinction will be made between
device performance and the performance of a chain of several devices or a
process.   Since the performance standard of each device is of utmost
importance, the review of a device will be reported using the following
format:

     a) title of report;
     b) report number and date of report;
     c) author(s);
     d) manufacturer of equipment;
     e) design features;
     f) characteristics of wastewater on which device was tested-
     g) method(s) of analysis;
     h) results (performance of standard based on published data)•
     i) approximate purchase price of equipment (if specified); and
     j) critical comments.

-------
The overall effectiveness of each class of devices for separation of oil/
water system states will be noted.

     Finally, a list of manufacturers for each equipment category will be
compiled.  Material to be reviewed is generated through searches in available
commercial literature and unpublished reports of industries and laboratories
involved in the manufacture or testing of equipment and new techniques.
Since existing literature is vast, the search and review process will focus
primarily on utility, rather than exhaustiveness.

-------
                                SECTION 2

                               CONCLUSIONS

      This  report  is  a  review of the state-of-the-art for oil/water separating
 devices and processes.  Techniques proposed as feasible but not yet evalu-
 ated as candidates for oil/water separation are discussed also.  Discussion
 does not include  coagulation-flocculation processes with chemical addition,
 nor is  the list of devices exhaustive.

      The problems associated with the removal of oil from water are complex;
 to meet effluent  guidelines or discharge limits, many oil/water treatment
 facilities must be upgraded.  Before a choice of separating equipment can be
 made, the  specific nature of the oil/water separation problem has to be
 examined thoroughly.   Sources of oily wastewaters are diverse; oily waste-
 waters  produced by different processes have characteristics that differ from
 each other depending on the type of oil(s), oil(s) and suspended-solids
 concentrations, physical and chemical properties of the aqueous phase, the
 process originating  the oily wastes, salinity, temperature, etc.  Moreover,
 the characteristics  of oily wastewaters may change with time.  The problem
 is complicated further by the different states in which oil can exist in
 water.   The. thermodynamic states in which oil can exist in water have not
 been identified completely; only the major states have been described.
 These states can  co-exist in wastewater depending on oil type, degree of
 mixing  of  the oil and water phases, concentration of surfactants, and other
 factors.   Also, it is  difficult to estimate oil concentrations in given oily
 wastewaters, because of inadequate analytical methods.  Concentrations of
 states  depend on  the chemical interactions between oil and water, number of
 days of equilibration, and the dissolution processes (physical, bacterial,
 or chemical oxidation) occurring during the equilibration period.

      Each  state has  a  certain degree of difficulty associated with its re-
 moval.   There are many devices available commercially capable of removing
 one  or  a few states, but a single device capable of separating all states is
 still lacking.  With progress in membrane technology, reverse osmosis and
 ultrafiltration may qualify eventually as the best technologies available
 for  separating oil/water mixtures.  For economic reasons, carbon adsorption
 should  continue to be  used in polishing states.  The potential and
 limitations  of devices and processes are presented in Table 1.

      The degree of difficulty of separating oil from wastewaters depends
 largely on  the number of states present.  Because of the inability of devices
 to effectively separate several states, combination of separation techniques
 into  process  trains is necessary to produce effluents that will meet dis-
 charge standards.  A modern oil wastewater treatment system may include an
API  gravity  separator and dissolved air flotation for removing free oil

-------
            TABLE 1.  POTENTIAL OF SEPARATION TECHNIQUES TO SEPARATE VARIOUS OIL/WATER SYSTEM STATES
Ul
Stabilized
Oil-coated solids Unstabilized dispersions Molecularly
Free Settle-
Technique oil able
A. Gravity
Differential
API XXX *
Hydro gard XXX
Circular XXX
PPI XXX
CPI XXX
Fram-Akers XXX
Curved-plate
finger XXX
Gravi-Pak XXX
Centrifuges XXX
Hydro cyclones XX
Vortex XX
Dispersed air
flotation XXX
Dissolved air
flotation XXX
Vacuum desorp-
tion XXX
Electro-
chemical
B. Filtration
Granular media
Multimedia
C. Coalescence/
Filtration
Fibrous media XXX
Centrifuge XXX
Bimetallic


XX
XX
XX
XX
XX
XXX

XXX
XXX
XXX
XXX
XX

XXX

XXX

XXX



XXX
XXX





Neutrally dispersions Surface Solubilized dissolved
buoyant Primary


X
X
X
XX
XX
XXX XXX

XX XX
XXX
XXX



X XX

XX XXX

XX XXX

X XX

XXX XXX
XXX XXX


XXX
XXX

Secondary Chemically charge oil oil






X
XX

X







XX

XX

XX XXX

XX XX
XXX XX XX


XXX
XXX
XX XXX
       *X, poor separation; XX, average separation; XXX, excellent separation.

-------
   TABLE 1.  POTENTIAL OF SEPARATION TECHNIQUES TO SEPARATE VARIOUS OIL/WATER SYSTEM STATES  (cont.)
Oil-coated solids
Free Settle- Neutrally
Technique oil able buoyant
D . Memb rane
Electro-
dialysis XX *
Reverse
osmosis XXX
Ultrafil-
tration XXX
E. Adsorption
Carbon XXX
F. Electric
& Magnetic
Electro-
phoretic
Magnetic
G. Thermal
H. Coanda
Effect X
I. Viscosity-
Actuated X
J. Chroma-
tography
K. Sonic &
Ultrasonic
Stabilized
Unstabilized dispersions Molecularly
dispersions Surface Solubilized dissolved
Primary Secondary Chemically charge oil oil
XX XX XX XX X X
XXX XXX XXX XXX XXX XXX
XXX XXX XXX XXX XXX XX
XXX XXX XXX XXX XXX
X XX
XX
XX X
XXX XXX
XX X
*X,  poor  separation; XX, average  separation;  XXX,  excellent separation.

-------
oil-coated suspended solids, and unemulsified primary dispersions;  a co-
alescer equipped with a prefilter for removing neutrally buoyant,  oil-coated
solids and remaining unemulsified primary and secondary dispersions; reverse
osmosis for removing emulsified, solubilized, and dissolved oil; and carbon
adsorption for removing the last traces of dissolved oil.  A combination of
other separation methods can be used; however, the general trend is gross
separation followed by finer separation and, finally, a polishing state.
This trend should prove to be most economical; desired effluent quality can
be achieved, the life of the polishing stage is extended, and throughput is
reasonable.  Therefore, utilization of several separation techniques is an
efficient means of separating oil/water mixtures.

     An attempt to review patent literature was unsuccessful, because of the
limited information usually available in patents and absence of performance
data.

     It is hoped the information developed in this report will be useful to
both  manufacturers and users of oil/water separating equipment.

-------
                                SECTION 3

                             RECOMMENDATIONS
 1.  The thermodynamic states in which oil can co-exist with water are not
    defined  completely.  Identification of oil/water system states present
    in a  given wastewater is necessary to a good choice of separating
    devices.

 2.  Dissolution rates of a variety of oils, petroleum products,  and other
    toxic organic pollutants are in need of measurement.

 3.  Until recently, it was thought that the concentration of dissolved oil
    present  in effluents could not be higher than the solubility of the oil.
    The phenomenon of solubilization of oil, in the presence of  surface-
    active agents and dissolved organic matter, increases oil concentrations
    considerably.  Therefore, removal of dissolved oil is necessary if
    effluents are to meet discharge limits.  Solubilization of oil should
    be investigated.

 4.  Estimating oil content with on-line oil/water monitors, turbidity mea-
    surements, visual observation, and other dubious analytical  techniques
    should be avoided.  Oil/water monitors and turbidity meters  are highly
    variable.  Since regulations are aimed at the total oil content of
    effluents, total organic carbon analysis should be preferred over
    extraction-gravimetric or extraction-infrared-spectrophotometric tech-
    niques for measuring oil content of oil/distilled water samples.

 5.  The efficiencies of different organic solvents—heptane, hexane, chloro-
    form, carbon tetrachloride, petroleum ether, pentane, and methylene
    chloride—commonly used in extracting oil for analysis, must be investi-
    gated.   Efficiencies of solvents for different hydrocarbon groups
    (paraffins, aromatics, etc.) merit investigation.  Depending on the
    solvent  and the number of extractions, results of oil concentrations in
    effluents can be in great error.

 6.  Since oil dispersions are formed through turbulent mixing during pumping
    operations, the efficiency of separation can be enhanced by  gravity flow
    or use of low-shear pumps having limited emulsification tendencies

 7.  A method to characterize oily wastewaters in terms of separation
    requirements is needed.

8.  For any oily-waste problem, segregation of wastes containing detergents
    proper water management techniques, maintenance of devices in good     '

-------
     operating condition,  and adequate operator training are  useful.

 9.   Presently, there is no format for evaluating the  performance  of  oil/
     water separation devices.   Oil/water mixtures used in  test  evaluations
     should be characteristic of the oily wastewaters  that  the equipment will
     treat upon installation.  Tests using oil/water'emulsions formed by
     passage through a centrifugal pump do not demonstrate  equipment  capa-
     bility to separate other emulsions without data on comparative emulsion
     stability.  Factors that affect the efficiency of the  separation process
     should be varied systematically.

10.   Many performance claims are substantiated with limited test results,
     using inexact analytical methods.  Performance data are  not stated, often
     because data are considered proprietary.   Often,  devices have not been
     tested adequately.  Therefore, evaluation can be  made  only  by comparing
     the principles of separation, instead of  design variations  and/or special
     design features.

11.   For comparison of promising techniques and adequate evaluation of
     existing technology,  a test facility similar to OHMSETT  is  needed.

-------
                                 SECTION 4

                         SYSTEM CHARACTERIZATION
OIL/WATER SYSTEMS

     Before a separation device or process  can be  selected,  there  Is  need
for an understanding of the type of oily wastewater to be  treated. As  such,
a characterization of oil/water systems is  necessary.  In  this  discussion,
"oil" will be used in a generic sense to  refer to  the non-aqueous  phase and
"water" will refer to the aqueous phase.

     After intimate contact of oil and water,  oil  can contaminate  the water
by existing in the aqueous phase in various forms.   These  states have already
been identified and reported in the literature as  free oil,  dispersed oil,
chemically emulsified oil, molecularly dissolved and solubilized oil, and
oil-coated suspended solids.

FREE OIL

     When a mixture of oil and water is left undisturbed for a  short  time,  a
continuous layer of oil forms at the surface of  the aqueous  phase  if  the oil
is less dense than water.  Separation of  the mixture into  two separate  layers
is due to the action of buoyant forces on  the large oil  masses  initially
present in the water body.  The oil in the  top layer has the essential  physi-
cal and chemical properties of the source  (parent)  oil,  unless  it  has been
modified by extrusion or reaction with chemicals present in  the aqueous
environment.  Modification may be due, also, to  the action of any  or  all of
the following mechanisms:  evaporation of  the volatile components, atmo-
spheric oxidation, microbial activity, and  dissolution of  soluble  fractions
of the oil.  These processes occur if there is prolonged contact of the
liquids, e.g. ballast and bilge waters retained  onboard  vessels for several
days or weeks.

DISPERSED OIL (EMULSIONS)

     Due to turbulent mixing, oil droplets  may be  dispersed  in  the aqueous
phase to form an oil-in-water emulsion,  depending  on phase volume  ratios and
other factors.  The particle size of the  dispersed oil in  an emulsion is
important in characterizing the type of  dispersion. Depending  on  the inten-
sity of mixing, primary or secondary dispersions result.  Both  dispersions
usually account for only a very small volume fraction of oil, in  the  order
of 50 to 1,000 ppm.
                                    10

-------
     Primary dispersions are formed from macroscopic oil droplets that range
in diameter from 1,000 to 10,000 A and remain in suspension due to Brownian
motion.  These droplets are generally unstable thermodynamically and agglom-
erate or coalesce into larger droplets, if allowed to remain undisturbed for
periods of approximately 24 hours.  Primary dispersions are produced by low
shear pumps, e.g. diaphragm and vane pumps, as well as low-speed centrifugal
pumps.

     Secondary dispersions are formed from very fine, microscopic oil drop-
lets that have mean diameters between 50 and 600 A and do not separate from
water if left undisturbed for a very long time; they are stable thermody-
namically.  Highly turbulent flow conditions are favorable to the formation
of secondary dispersions.

     Interfacial tension is an effective stabilization mechanism for disper-
sions.  Additional stability of dispersions arises if electrical charges are
present on the surfaces of oil droplets.  A detailed discussion on this
electrokinetic phenomenon can be found in the literature (Kruyt, 1952; Over-
beek, 1952; Adamson, 1967).  A double layer of charges is formed at the oil/
water interface of each droplet, and coalescence of adjacent droplets is
prevented by mutual repulsion.  The potential difference in the diffuse
double layer is called the zeta potential.  Stable emulsions exhibit zeta
potentials in excess of 25 mV (Churchill and Kaufman, 1973; Orr and Kang,
1974).  Measurement of this potential is a useful tool in estimating the
stability of emulsions (Orr and Kang, 1974).

CHEMICALLY STABILIZED DISPERSIONS

     The presence of surfactants favors formation of chemically stabilized
dispersions.  These have the same particle sizes as the primary and secondary
dispersions discussed above.  The oil droplets have additional stability
because of the presence of a third component in either the oily or the
aqueous phase.  The third component is variously referred to as a surfactant,
surface-active agent, detergent, soap, stabilizing agent, emulsifier, etc.
Small concentrations of this agent are enough to chemically stabilize oil
droplets.  Extensive literature is available on surfactants and their effects
on the stability of emulsions (Jefferson and Boulavare, 1973; Churchill and
Kaufman, 1973; Gloyna and Ford, 1974).  The chemical nature of the surfactant
is important.  A theory has been advanced for surfactant modification of an
oil/water interface.  It states that surfactants are molecules composed of
lipophilic and hydrophilic end groups that orient themselves in an emulsion
such that their lipophilic ends project into the non-aqueous phase while
their hydrophilic ends are anchored in the aqueous phase.  A protective
"film" is formed around each droplet, as a result of surface interaction.
Reduction in interfacial tension leads to a low free energy, which is
unfavorable to coalescence of the oil droplets.

MOLECULARLY DISSOLVED OIL

     Generally, hydrocarbons exhibit limited solubilities in water, with
aromatic hydrocarbons somewhat more soluble than aliphatic hydrocarbons
                                     11

-------
 (McAuliffe, 1969a,b).  Molecularly dissolved oil is  oil that is  in  true
 chemical solution in the aqueous phase.   Dissolved oil is  generally classi-
 fied as submicroscopic oil droplets, below 50-A diameter.   Because  of limited
 solubility of oils in water, the concentration of molecularly dissolved  oil
 is probably less than 20 ppm.  However,  after prolonged equilibration, oil
 concentrations can be higher than 200 ppm.  Dissolution of petroleum-derived
products has been studied extensively, but is not completely understood
 (McAuliffe, 1969a; Boehm and Quinn, 1974; Lysyj and  Russell, 1974).  One
 reason for the inadequacy of knowledge in this subject is  the lack  of tech-
nology for measuring soluble oil without interference from emulsion droplets.
Lai and Adams (1974) developed a method  for determining the molecular solu-
bility of Navy oils in water, using an osmometric device.

SOLUBILIZED OIL

     Studies indicate that, in the presence of dissolved organic matter  (DOM)
in the aqueous phase, the solubility of  oils increases, particularly oils
containing large fractions of aliphatic  hydrocarbons (Boehm, 1973;  Boehm and
Quinn, 1974).  Acceleration of organics  transfer into the  aquous phase is
due to chemical modification of water-insoluble petroleum  fractions.  The
presence of surface-active agents in petroleum products can cause solubiliza-
tion, also.  Solubilized oil droplets are less than  0.5 p  in size.   Solubili-
zation of hydrocarbons in seawater may not be appreciable  if the DOM concen-
tration is so low that the concentration of dissolved surfactants does not
exceed the critical micelle concentration (Elworthy  et al., 1968).

OIL-COATED SUSPENDED SOLIDS

     Solids suspended in the aqueous phase become coated with oil.   These
solids have a wide range of origins, densities, compositions, and sizes
 (Finger and Tabakin, 1973; Freestone and Tabakin, 1975) .  These  solids are
mainly clays, silica, drill muds, corrosion products, asphaltenes,  heavy
metals or alkaline-earth salts, and the  fine sediments that are  abundant in
natural waters and oil field brine formation water.   Finely divided solids
play a major role in crude oil emulsion  stabilization.  Solids that are
neutrally buoyant require special treatment before they can be separated
effectively.  Oil adsorbed on the surfaces of solids enhances solubilization.
                                    12

-------
                                SECTION 5

                   CHARACTERIZATION OF OILY WASTEWATERS

     Since this review focuses primarily on technology available for separat-
ing oily wastewaters from ships and crude oil production operations, a brief
characterization of these wastewaters in parallel with the characterization
of oil/water systems follows.

     Oily wastewaters generated onboard ships and vessels are ballast water,
tank cleaning water, and bilge water.

BALLAST WATER

     After discharging fuel oil cargo, a ship pumps water or seawater into
storage tanks to maintain stability.  This water may contain large amounts
of silt or fine solids if picked up in a river or estuary, or offshore from
a large seaward river flow.  Oil that adheres to tank walls mixes with the
water as a result of ship motion.  During deballasting operations, the oily
wastewater is pumped out, usually by a high-speed centrifugal or reciprocat-
ing pump, in order to minimize turnaround time.  The phase-forming effects of
various pumps on mixtures of oil and water have been reviewed (Shackleton et
al., 1960; Fruman and Sundaram, 1974).  Many factors affect the dispersion
of oil/water mixtures as they pass through pumps and piping (Shackleton et
al., 1960):

     a) excessive velocities and accelerations;
     b) restricted ducts;
     c) rapid changes in fluid direction; and
     d) varying speeds and discharge pressures.

     It is likely that debaliasted water contains all the oil/water systems
discussed in Section 4, depending on the following factors:

     a) type of fuel cargo;
     b) characteristics of water used as ballast water;
     c) duration of voyage between ballasting and deballasting operations;
        and
     d) the emulsification characteristics of the pump used in deballasting.

Ballast water from cargo tanks usually contains oily residues from prior
loads.  Occasionally, washwater from tank cleaning is added to ballast
water.
                                      13

-------
 TANK-CLEANING WATER

     Before a change in fuel cargo is made, cargo tanks are washed.  Oily
 wastewater generated by tank cleaning is similar to ballast water, except
 detergent cleaning of cargo tanks may be necessary.  Cleaning with deter-
 gents  leads to formation of chemically emulsified oil.  The amount of oil
 that clings to tank walls may be up to 1% of the total cargo, depending on
 the viscosity of the oil.  This portion of the load ends up in ballast and
 tank-cleaning waters.  Dirt and scale are produced by tank cleaning, also.

 BILGE  WATER

     Bilge water comprises leakages of lubricating oil, fuel oils, and hy-
 draulic fluids, and water resulting from drains and drippings in the engine
 room.  It contains solids and rust scale, also.  Oils present in bilge water
 contain many additives.  Bilge water must be pumped out of the ship, and
 chemically emulsified oil will be the dominant of the two types of emul-
 sions  discussed previously.  Becuase of the complexity of bilge waters,
 there  is, as yet, no meaningful characterization scheme (Budininkas and
 Remus, 1974).  In undiluted bilge water, the ratio of oil-to-water is higher
 than in ballast water.  Therefore, multiple emulsions are probably present.
 The rate at which a ship generates bilge depends on the age, condition, and
 maintenance history.

 OIL-FIELD PRODUCTION WATER

     There is increasing interest in development of offshore oil production
 facilities.  During crude oil exploration, drill cuttings and mud chemicals
 are the main pollutants.  In the production phase, the wastewater generated
 is oily brine formation water (production water).  The composition of forma-
 tion water differs from well to well.  However, these wastewaters are char-
 acterized by a high content of dissolved and suspended solids.  Therefore,
 the potential for formation of oil-coated solids and stable emulsions is
 high.

     Crude oils are complex mixtures; they differ in characteristics,
 according to geologic age, chemical constitution, and associated impurities.
 They contain many natural emulsifiers, usually naphthenic and other organic
 acids, resinous substances, and asphaltenes (Reisberg and Doscher, 1956).
 Formation brines for different wells have different compositions  (USEPA,
 1975).  Consequently, emulsions resulting from crude oil production are
 stabilized by a variety of mechanisms, depending on origin.

     Discharge pressures during crude oil production are usually high, and
 entrainment of fine gas bubbles in the oil/water mixture is likely to occur.

 SUMMARY

     The various states in which oil and water can co-exist have been dis-
cussed.  Oily wastewaters from vessels and crude oil production have been
characterized crudely.  Technology available for the separation of these
oily wastewaters will be reviewed.

                                    14

-------
     In addition to the oil/water system states already presented, other
parameters influence the performance of a separation device or process:

     a) oil concentration of the influent stream;
     b) flow rate;
     c) physio-chemical properties of the wastewater, including
          1) temperature,
          2) pH,
          3) salinity,
          4) ionic strength, and
          5) dielectric constant;
     d) density ratio of the oily and aqueous phases; and
     e) mechanical motions during separation.
                                    15

-------
                                SECTION 6

                          DEVICES AND PROCESSES

     Available methods for separating oil/water mixtures include:  physical,
chemical, mechanical, electrical, magnetic,  and thermal treatments, and
combinations of these.  Because manufacturers'  trade names are often mislead-
ing, devices are classified according to the primary mechanism which induces
separation in wastewaters.  The basic principle of each group of; devices is
stated.  Variations in devices and processes, including the different modes
of operation, are specified.  Pertinent literature on a device, particularly
literature on text evaluation of equipment,  is  summarized and reviewed
critically.

     Since the performance standard for each class of devices is of utmost
importance, the workability of each group relative to what is known about
oil/water system states is discussed.  The specific oil/water system(s)
which each group of devices (if adequately designed and operated) is capable
of treating is noted.

     Applicability of devices relative to some  additional constraints is
stated.  These include:  limiting space requirements, low weight, and sensi-
tivity to motion during processing.  These are  imposed by shipboard or off-
shore platform operations, but are not necessarily important for shoreside
facilities.  Finally, it should be noted that this survey does not purport to
include all devices presently available.

TECHNOLOGY

     Commercial and experimental oil/water separation devices are listed
below.  Devices operating on principles proposed as feasible but which are,
as yet, in the developmental stage are included.

Gravity Differential Separation

     API Oil/Water Separators
     Circular Separators
     Plate Separators
          Shell parallel-plate interceptors  (PPI)
          Shell corrugated-plate interceptors (CPI)
     Curved-Plate Finger Separators
     Rotational Separation
          Centrifuges
          Hydrocyclones
          Vortex flow
                                    16

-------
     Gas Flotation
          Dispersed  air
          Dissolved  air
          Vacuum desdrption
          Electrochemical

 Filtration

     Layer Filtration
          Granular media
          Multimedia
     Membrane
          Electrodialysis
          Reverse osmosis
          Ultrafiltration

 Coalescence/Filtration

     Fibrous-Media
     Membrane
     Centrifuge
     Bimetallic
     Granular-Media
     Other Porous Materials

 Adsorption and Absorption

 Electric and Magnetic Separation

     Electrophoretic
     Magnetic

 Thermal Separation

     Heating
     Evaporation and Distillation
     Freezing and Crystallization

 Sonic and Ultrasonic Separation

 Coanda-Effect Separation

 Viscosity-Actuated Phase Separation

 Chromatographic Separation


Gravity Differential Separation

     Gravity differential separation is the oldest and most common method for
separating oil/water mixtures.  It is usually the first step in the treatment
of oily wastewaters and provides coarse separation of oil and water.  In

                                     17

-------
general, oil/water mixtures will separate naturally into two distinct layers
of oil and water, if allowed to stand undisturbed for a sufficient period of
time.  Ease of separation depends on the magnitude of the difference in den-
sities of the two immiscible liquids; the basic principle governing this
technique is Stokes 's Law, which is applicable to the rate of rise of oil
globules in water.                 9
                             v
where v = rate of rise of an oil globule

      g = acceleration due to gravity

      D = diameter of an oil globule
  6  ,6  = densities of the aqueous phase and oil, respectively
   0)  O
      ]i = absolute viscosity of the aqueous phase

Stokesfs Law applies to solids suspended in water (oil-coated), also.  From
this equation, oil globules and/or suspended particles will rise to the sur-
face or fall to the bottom, depending on the sign of the density differen-
tial.  The difference in densities between most contaminant oils and water is
usually small, and the viscosity of the aqueous phase is temperature-depen-
dent but is essentially constant.  Therefore, the rate of rise of an oil
globule is dependent to a large extent on the particle size.  Hence, for an
appreciable separation to occur, within reasonable residence times, the oil
droplets and suspended solids must be large.  As oil globules rise to the
surface, collisions occur, coalescence takes place, and a floating oil film
forms at the surface.  Coalesced oil is subsequently skimmed off.

     Gravity separation is inefficient when the density difference is small,
viscosity of the aqueous phase is low, and oil droplets are small.  As parti-
cle size becomes smaller, residence times and space requirements increase.
Because of these limitations, gravity separation methods are used only to
separate free oil, primary dispersions, and large oil-coated solids.  Devices
operating on the gravity principle will not separate dissolved oil or emul-
sions (API, 1969).  Neutrally buoyant solids, coated with oil, are not sep-
arated.  If gravity separation is used in conjunction with chemical addition,
stable emulsions can be broken and separated.

     The most economical state-of-the-art methods in oil/water separation are
of the gravity type.  Devices can handle large flow rates, have low power re-
quirements, and need minimum operator attention; but processes are slow,
necessitating large equipment.  Gravity separation is basic to almost all
oil/water separators.  Several methods have been devised for accelerating the
process.  These include provisions for heating the influent to reduce the
viscosity of the aqueous phase, extended plate surfaces to increase the hori-
zontal distance traveled by oil globules, rotational forces instead of gravi-
tational force, and air flotation.  In attempts to increase oil/water separa-
tion efficiency, there have been modifications of existing designs.  As a
result, devices in this category are the most abundant.  The slight varia-
tions and modifications in designs have already been reviewed excellently by
Harris  (1973).   Of major concern are the improvement of the hydraulic

                                     18

-------
characteristics of the devices and the reduction of turbulence in separators.
Almost all gravity-type separators produce effluents that must be treated fur-
ther, in subsequent separating devices.  However, they are reliable, simple,
and inexpensive to operate, and serve to attentuate fluctuations in flow and
oil concentration in lieu of more sophisticated oil/water separators.

     API Oil/Water Separators—The design of gravity-type separators has been
studied extensively by several investigators and particularly by the American
Petroleum Institute (Ingersoll, 1951; Rohlich, 1951; University of Wisconsin,
1949, 1950, 1951).  As a result of these efforts, the API has set forth de-
sign recommendations for oil/water separators in the Manual on Disposal of
Refinery Wastes;  Volume on Liquid Wastes (API, 1969).  Construction details
of API separators are found in Chapter 6 of the manual.  Important design
considerations are minimum horizontal area, minimum vertical cross-sectional
area, and minimum depth-to-width ratio of 0.3-0.5.  The API design is based
on Stokes's Law and hydraulic overflow rates.  An oil-droplet rise rate of
0.2 ft/min, with a forward wastewater flow of 3.00 ft/min, is used.  Deten-
tion time is about one-half hour.  Separators are designed to remove non-
emulsified oil particles of 130-150 u and larger for oils having typical
specific gravity.  Design nomographs are presented in the manual, with cor-
rections for turbulence, short-circuiting, and wastewater temperature.  API
separators are usually rectangular in shape and multichanneled; expansion is
possible, and single channels can be cleaned without interrupting operation.
The API oil/water separator consists of inlet and outlet sections, a pre-
treatment stage, separating stages, baffles, skimming devices, and flight
scrapers.  Some separators are equipped with covers or floating roofs.  In
operation, oily wastewater enters the separator at the inlet, flow is slowed
down, and turbulence is minimized by the inlet structure and baffles.  Oil
globules larger than 150 y rise to the surface and settleable solids (oil-
coated and non-oil-coated) sink to the bottom.  Provision is made for skim-
ming the oil and removing the sludge.  A schematic diagram of an API
oil/water separator is given in Figure 1.

     Hydrogard separators are prepackaged oil/water separator units, manu-
factured by Inland Environmental Corporation.  These separators are designed
according to API guidelines.  Units with flow rates up to 210 gpm are avail-
able.  Cleaning and maintenance are carried out easily.  Inland Environmental
Corporation claims that effluents from these separators can contain as little
as 5 ppm of oil.

     Circular Separators—Oil/water separators designed according to the con-
ventional arrangement of a circular clarifier are used in some oil refineries
with satisfactory results, but a rational design procedure for circular sep-
arators has not been developed.  An advantage of circular units is the ease
of installing oil-skimming and sludge-scraping devices.  The capacity of
these devices can be varied by adjusting the height of the oil skimmer.  Oily
wastewater is fed through a central inlet; effluent outlets are located in
the peripheral wall.  Circular units are more compact than API separators.

     Plate Separators—Concern over the large space requirements of API oil/
water separators led to studies on methods of reducing equipment size without
decreasing oil-removal efficiency.  Different methods were used to reduce

                                     19

-------
            h-
                                     SEPARATOR CHANNEL
                                                             	   J
      DIFFUSION DEVICE
    (VERTICAL-SLOT BAFFLE)
GATEWAY PIER
       \
      -A	
        \
     ±
        JJ.
     /    /
     /     /
FOREBAY   i
       SLOT FOR
    CHANNEL GATE
                                 FLIGHT SCRAPER
                                 CHAIN SPROCKET
           ROTATABLE OIL
           SKIMMING PIPE
                  \
                                    WOOD FLIGHTS
FLIGHT
SCRAPER
 CHAIN
    \
    \
                                                                WATER   \
                                                                LEVEL   .
                                                                   *     \
                                        =yt
                                                          \
                                                  1    I    I
                                             FLOW
                                                                               OIL-RETENTION
                                                                                   BAFFLE
                                "--SLUDGE-COLLECTING HOPPER
                                   DISCHARGE WITH LEAD PIPE
                                                                                            EFFLUENT
                                                                                          /WEIR AND
                                                                                            WALL
                                                                         EFFLUENT FLUME
                                  EFFLUENT SEWER
      SLUDGE-COLLECTING
            HOPPER     '
                    SLUDGE  PUMP
                   SUCTION PIPE
                                  Figure 1.  API Oil/Water Separator

-------
settling length, because the time required for oil particles to rise to the
surface depends on the depth of the separator.  Flat plates, plate packs,
convoluted plates, perforated conical plates, and perforated plates of other
geometry have been used to subdivide the settling chamber into a number of
sections.  Plates increase the surface area and reduce the maximum rise
height of oil globules.  It is observed that coalescence of oil globules
occurs on these plates, also,  Therefore, the maximum distance oil droplets
have to travel, before coalescence occurs, is dependent on plate spacing in-
stead of depth as in the API design.  To improve coalescence, plates are
manufactured from oleophilic materials and inclined at an angle to the incom-
ing flow.  As oil globules rise to the surface, they coalesce on the under-
side of the plates, creep up plate surfaces and break loose as large particles
that rise rapidly to the top.  Because of these design modifications, plate
separators are approximately one-fifth to one-half the size of API separa-
tors.  Oil skimming and sludge removal are easier; the main problem with
these units is plugging of the spaces between plates with solids, biological
growth, or highly viscous oils.  Several types of plate separators are avail-
able commercially.

          Shell parallel-plate interceptor.  The Shell parallel-plate inter-
ceptor (PPI) oil/water separator has parallel plates spaced 100 mm (approxi-
mately 4 inches) apart.  Two such systems are available.  One system consists
of one or more sets of plates inclined at an angle of 45° to the long axis of
the separator.  The other system consists of one set of plates parallel to
the long axis and inclined at an angle of 45° to the horizontal.  The spacing
between the plates can be varied.  Because of the inclination, the effective
surface area for coalescence is increased and the net path oil globules trav-
el before reaching the surface is decreased (Kirby, 1964).  These added fea-
tures make the Shell PPI oil/water separator capable of separating oil drop-
lets of 60 y in diameter or larger.  In spite of improvement in performance,
stable emulsions and dissolved oil cannot be separated in these devices.

          Shell corrugated-plate interceptor (CPI).  An improvement on the
Shell PPI oil/water separator design is the Shell CPI separator.  It features
plates arranged at an angle of 45° to the horizontal in the direction of
wastewater flow, similar to the Shell PPI unit; the major difference is that
the plates are corrugated and the spacing between plates is smaller, only
approximately 20 to 40 mm.  Plates are made of fiber glass-reinforced poly-
ester.  Because of closer spacing, CPI units are more compact than PPI units
and oil-removal efficiency is greater.  Reduction in space requirement can be
as high as two-thirds, but plugging from solids is a major problem.  Newer
designs feature accesses for easy cleaning.  Schematic diagrams of the Shell
PPI and CPI are given in Figures  2  and  3, respectively.

     The Fram Corporation manufactures an oil/water separating system fea-
turing a combination of two separation processes.  The separator has three
stages:  the first stage contains two preconditioners (filter-cartridge type,
having 75-y pore openings) for suspended-solids removal; the second stage
utilizes parallel-plate-type gravity separation; and the third stage contains
a cartridge-type coalescer.  Because of the presence of a cartridge-type
coalescer, this system is capable of breaking and separating emulsions.
pact design makes the system suitable for use on board ship.  Oil-removal
Com-
                                     21

-------
   OVERFLOW PIPE FOR
     TREATED WATER
VENT AND OIL
OVERFLOW PIPE  TREATED
                WATER
ISJ
l-o
  TRASH
CONTAINER
                                                                           SCREEN
                         PACKS OF PARALLEL
                          INCLINED PLATES
                      SLUDGE SUMP
         SUCTION HOSE
          FOR SLUDGE
                                                                                         SAND SUMP
                                Figure  2,   Shell Parallel-Plate  Interceptor

-------
U>
            ADJUSTABLE WEIR
                    \
                     \
  OIL
SKIMMER
           _    ,
           LAYER
           /      OIL GLOBULES
                     v\\\\\\\\\\\\\\\\\
INLET
                                                              \\\ \Y\\A\
                                                                          SEDIMENT TRAP
                                                            PACKS OF CORRUGATED
                                                             PARALLEL PLATES
       TREATED WATER
      OUTLET CHANNEL
                                                  — SLUDGE PIT
                           Figure  3.   Shell Corrugated-Plate Interceptor

-------
 efficiency  is  reasonable, but effluents require some further treatment before
 discharge.  Media plugging and low liquid flow rates are disadvantages.
 System backwashing, when head loss reaches a specified level, is automatic.

      Curved-Plate Finger Separators—The basis for the design of the curved-
 plate finger separator is accelerated gravity settling of oil from oily
 wastewaters.   This separation technique is a combination of gas flotation and
 parallel  plate-type gravity separation.  Incoming wastewater feed is first
 mixed with  air,  as in dispersed gas flotation; the separation of oil and
 water takes place between curved steel plates.  The plates are inclined hori-
 zontally  and arranged in the direction of flow-  Oil films form at the under-
 sides of  the plates, rise gradually toward the upper ends of the plates, and
 leave as  "fingers" at the plate tips.  Oil that collects at the surface is
 removed by  an  oil-skimming device.  Because of the gas flotation character-
 istics of this device, small oil globules can be separated.  Space require-
 ments are less than for API oil/water separators, for the same throughput.

      The  Gravi-Pak oil/water separator is manufactured by Keene Corporation.
 Units are multistaged, and consist of a primary separation chamber of the
 parallel-plate type and a secondary separation stage in which natural gravity
 settling  occurs.  Devices are suitable for gross separation of oil and water
 and are more compact than the API design, but larger than Fram-Akers units.
 Manufacturers  claim effluents from these units can contain less than 30 ppm
 of  oil, with wastewater having up to 20,000 ppm of oil.

      Design—Stokes's Law is the fundamental principle governing design of
 gravity-type oil/water separators.  Design considerations are flow rate,
 rise rate of oil globules, turbulence correction factors, type and concentra-
 tion of oil, characteristics of wastewater, and geometry of basin.

      Performance—Gravity-type oil/water separators are used primarily to
 remove free oil, large oil globules (unstable primary dispersions), and oil-
 and/or non-oil-coated solids in suspension.  Stable emulsions (surface- or
 chemically  stabilized), solubilized oil, and dissolved oil are not separated
 by  devices  operating on the gravity differential principle.  Devices can
 handle varying influent oil concentration and are normally used for gross
 separation.  Effluents from these devices usually require some further treat-
 ment before discharge.  Factors affecting oil-removal efficiency are flow
 rate,  oil particle size, density of oil, characteristics of wastewater,
 temperature, and separator design.  For good separation, it is essential
 that feed velocity distribution is as uniform as possible at the inlet
 section of  the device.

      The  API oil/water separator is an effective and versatile device for
 removing  oil and suspended solids.  Because of design limitations, API sep-
 arators are only effective in removing oil globules having a lower limit of
 130  y  in  diameter.  Neutrally buoyant particles may pass out with the efflu-
 ent.  Beychok (1973)  has summarized the available performance data on API
 separators.   His data are presented in Table 2.  Data indicate that oil con-
 tent of effluents can be as high as 120 ppm, when the oil content of influ-
ents is small.   Operation of devices at flow rates grossly below design
                                    24

-------
rates yields lower efficiences.  In experiments with a cylindrical-type sep-
arator having a residence time of about 20 minutes, at a flow rate of 10 gpm,
Finger and Tabakin (1973) report that more than 80% of the oil was removed
from oily wastewaters containing up to 4% oil.  The performance data of
several gravity-type separators, published by Gloyna and Ford (1974), are
presented in Table 3.  Oil-removal efficiencies, for refinery wastes, range
from 50 to 90%.  Data indicate that oil-removal efficiencies in parallel-
plate separators may be higher than achievable in API separators.   A compari-
son of results of API and PPI separators by Kirby (1964)  indicates reductions
in oil content of up to 67% more in the PPI than in the API.  Data presented
by Brunsmann et al. (1962) substantiate the belief that PPI separators
achieve higher oil removal than comparable API separators.  Harris (1973)
presents estimated effluent quality from primary oil/water separation proces-
ses, as shown in Table 4.  The oil content of effluents from devices operat-
ing on the gravity differential principle can be much higher.  Effluents con-
taining more than 500 ppm of oil should be expected, if the influent contains
high concentrations of emulsified and dissolved oil.  Average oil  removal,of
25 to 65% should be expected; oil-removal efficiency can be increased to more
than 80%, if the process is used in conjunction with chemical addition.  Sus-
pended-solids (oil-coated and/or non-oil-coated) removal of up to  65% can be
accomplished.  Gravity-type separators will continue to be basic to treatment
of oily wastes; and operation of devices, at design flow rates by  trained
personnel, will yield reasonable separation.


                 TABLE 2.  OIL REMOVAL IN API SEPARATORS*
            Oil content (ppm)
           Influent   Effluent      % Removal         Shape
-
-
-
-
50-100
90-98
42
108
20-70
20
80-115
75
20-40
40-44
20
20
-
-
-
-
60
55
52
54
Rectangular
ir
it
ii
it
it
it
Circular
          * Reproduced with permission, from Aqueous Wastes
            from Petroleum and Petrochemical Plants, by
            Milton R. Beychok.
            Copyright ©1967, John Wiley & Sons Limited
                                    25

-------
  TABLE 3.  OIL AND SUSPENDED-SOLIDS REMOVAL IN GRAVITY-TYPE SEPAEATORS*
Oil content (ppm)
Influent
300
220
108
108
98
100
42
2000
1250
1400
Effluent
40
49
20
50
44
40
20
746
170
270
% Oil % Suspended-
removal solids removal
87
78
82
54
55
60
52
63 33
87 68
81 35
Type
PPI
API
circular
"
API
n
"
n
ti
n
  * Gloyna and Ford (1974).
       TABLE 4.  ESTIMATED EFFLUENT QUALITY FROM PRIMARY OIL/WATER
                          SEPARATION PROCESSES*
       Separators commercially
             available
Effluent oil
concentration
   (fflg/D
      API rectangular
      Circular
      Inland Steel—Hydrogard
      Shell PPI
      Shell CPI
      Finger-plate separator
      Fram-Akers plate separator
      Keene—Gravi-Pak
    50-75
    50-75
    50-75
    35-50
    35-50
    35-50
    50-100
      20
     * Harris (1973, p. 85).
     Rotational Separation—Successful separation of oil from water by uti-
lizing rotational motion has been reported in the literature (Guzdar et al.
1975; Yu, 1969; Sinkin and Olney, 1956; Sheng and Welker, 1969).  However,"
this separation method is practical and economical only when the concentra-
tions of oil and suspended solids are high.  Rotational separation is an
accelerated gravity differential separation method and Stokes's Law applies-
gravitational force is replaced by centrifugal force.  The centrifugal force
                                    26

-------
can be 1000 to 5000 times the gravity force, and the rate of separation is
faster than in ordinary gravity separation.  There are three basic types of
rotational separators:  (1) centrifuges, (2) hydrocyclones, and (3) vortex
flow.

          Centrifuges.  Centrifuges have been used extensively in sludge-
dewatering applications and for removal of solid contaminants from waste-
waters.  In recent applications, centrifuges have been used to separate oil
from oil/water mixtures.  In centrifugation, the oil/water mixture is moved
along a circular path by the rotational motion of the device.  The centri-
fugal force developed by rapid rotation of the system enhances the separation
of the two phases.  The phase with the higher density has a larger momentum
and moves toward the outer periphery of the centrifuge; the less dense phase
concentrates at the center line of the centrifuge.  These phenomena occur as
a result of the centrifugal force field and the difference in densities of
the two phases.  The greater the difference in the densities of the two
phases, the faster the separation and the less the energy requirement.  A
minimum density difference of 5% is enough for separation.  The location of
the boundary between the two liquid phases can be predicted in theory, and it
is possible to determine the critical oil-droplet size that will be present
in the separated water phase.

          Hydrocyclones.  The basic principle of separation by a hydrocyclone
is similar to that of a centrifuge.  In a hydrocyclone, the liquid is forced
into circular motion due to tangential injection of the oil/water mixture
against the circular configuration of the hydrocyclone (Yu and Ventriglio,
1969) .  The advantages of a hydrocyclone over a centrifuge are low initial
cost, ease of maintenance, and absence of moving parts.  But hydrocyclones
require considerable pumping power to achieve the centrifugal force needed
for separation.  Separation by a hydrocyclone is similar to vortex flow sep-
aration, and problems associated with the turbulence created during operation
have made both methods inefficient for treating oil/water mixtures.

     Design—The design of centrifuges has not changed much since their first
introduction.  Emphasis is in providing centrifuges capable of achieving
greater throughput, at low speed.  The three types of centrifuges most often
used in marine oil/water clarification applications are the barrel, tubular,
and disc or plate types (Harris, 1973).  Centrifuges can be designed for
radial or axial flow.  The Navy, Coast Guard, and Maritime Administration are
sponsoring jointly a contract to develop a centrifuge.  It is to incorporate
axial flow, allowing for,comparatively high throughput and low speed.  Axial
flow design allows sufficient residence time for interdrop coalescence to
occur, increasing the effectiveness of separation (Finger and Tabakin, 1973).

     Performance—Oil/water separation by centrifugation is practical only
when the oil and/or solid particles are present in relatively high concentra-
tion and the densities of the oily and liquid phases are not close.  Centri-
fugation is effective in removing oil- and non-oil-coated suspended solids,
free oil, and primary dispersions, if the average oil globule size is greater
than the critical drop size.  It is ineffective in removing stable emulsions
and solubilized and dissolved oil.  With chemical addition, this method has
been used to destabilize and coalesce stable emulsions.  Hence this method is


                                     27

-------
used mainly to concentrate and recover fuel oil from oily wastewaters with
little water content.  Finger and Tabakin (1973) report, in laboratory tests,
effluent concentrations in the range of 10 ppm were obtained with feeds con-
taining 1000 ppm of #2, #4, and Nigerian crude oil.  At higher concentra-
tions, ranging up to about 59,000 ppm of #2 oil and 19,000 ppm of #4 oil,
effluent concentrations were higher, but did not exceed 100 ppm.  The oil
content of effluents reported by these authors is unusually low, and higher
values should be expected.  Efficiences of centrifuges are usually in the
range 60-80%, depending on the specific nature of the oily wastes concerned,
oil and suspended-solids concentrations, temperature, etc.  The presence of
detergents has an adverse effect on device performance.

     Efficiencies ranging from 77 to 91% have been reported for hydrocy-
clones, but the oil content of effluents was persistently high and the sep-
arated oil phase contained water droplets (Sinkin and Olney, 1956) .  Sheng
and Welker (1969) suggest the use of several hydrocyclone units, in a cascade
operation, to produce effluents containing less than 100 ppm of oil; informa-
tion on this process is currently not available.  Centrifuges and hydrocy-
clones can be useful for gross separation, in lieu of more advanced and
better separation methods.

          Vortex flow.  Separation of liquid/liquid mixtures, utilizing the
fluid phenomenon known as confined vortex flow, can be achieved if the
liquids differ in density.  Vortex flow is similar to centrifugation; as
such, it is an accelerated gravity settling technique.  The separation of
oil/water mixtures is accomplished by imparting relatively large rotational
motion to the mixture, in a cylindrical vessel.  As a result of the confined
vortex flow, the lighter fluid (in most instances the oil) is accelerated
radially inward faster than water.  Therefore, oil accumulates and forms a
central core, where it is removed by extraction tubes (or a perforated core)
located at the center line of the cylinder.

     Two types of vortex-flow separators are available commercially.  In one
type, the influent or recycle stream is injected tangentially into the cylin-
der through various inlet ports.  Injection points are located at the circum-
ference on one end of the cylinder.  In the second type, a shrouded axial-
flow pump impeller is used to rotate the fluid; this type is very similar to
a centrifuge.  However, vortex separation is different from centrifugation
because the volume of fluid rotated at any time is smaller, and reinjection
of some effluent water is necessary to maintain the vortex formed within
the tube.

     Design—Vortex separators consist primarily of a vortex tube, a cylin--
drical vessel with a perforated tube or extractor located at the center line
and have no moving parts.  Important design parameters are:  operating pres-
sure, location of injection and exit ports, and length-to-width ratio of
vortex tube.  Several geometric variations are possible.

     Performance—Several factors affect the performance of vortex-flow sep-
arators.   These include:  separator geometry, feed-oil concentration, oil
type,  oil-droplet-size distribution, debris, external motion, oil-collection
rate,  and temperature.  In tests conducted with the United Aircraft vortex

                                     28

-------
separator, separation was poor; slightly emulsified oil could not be separ-
ated.  Average oil content of effluents was above 50 ppm, and separated oil
contained as much as 90% free water.  Better separation was obtained when the
unit was followed by a gravity separator.  Because of physical limitations,
vortex separation is capable of separating only free oil or oils with large
droplet sizes.  Even then, oil-removal efficiency can be lower than that of
ordinary API gravity separators.  Furthermore, injection of oil/water mix-
tures creates turbulence, which reduces oil-droplet sizes and hinders
separation.

     Gas Flotation—The success of gas flotation in the mineral industry led
to its use for separating oil from wastewater (Gaudin, 1957).  Gas flotation
is an accelerated gravitational separation technique in which flotation of
oil dispersions, wax and grease, and suspended solids in wastewater is accom-
plished by numerous microscopic gas bubbles.  The process is a composite of
the following steps (Vrablik, 1957):

     a) introduction of gas bubbles into wastewater;
     b) collisions between gas bubbles and suspended matter;
     c) attachment of fine bubbles to the surfaces of suspended matter;
     d) collisions between gas-attached suspended particles forming
        agglomerates;
     e) entrapment of more gas bubbles into agglomerates; and
     f) upward rise of floe structures in a "sweeping" action ("sweep
        flocculation").

A froth layer is formed at the surface of the wastewater and is removed by
an appropriate skimming device.

     The rise rate of the floe structures is expressed by a modified
Stokes's Law:
                            v
                             o        18y

where v  = rise rate of oil-air-particle agglomerates
       o
       g = gravitational constant
      D  = effective diameter of oil-air-particle agglomerates
       o
      S  = density of aqueous phase
       to
      6  = density of oil-air-particle agglomerates

       y = absolute viscosity of aqueous phase.

The attachment of gas bubbles to suspended matter in the flotation process
affects density and diameter in the Stokes's Law equation.  The result is a
net increase in rise rate.  Two- to tenfold increases in rise rate are
encountered, depending on other factors, e.g.,

     a) gas-input rate and volume of gas released per unit volume of liquid;
     b) bubble-size distribution and degree of dispersion;
     c) surface properties of suspended matter;

                                    29

-------
     d) hydraulic design of flotation chamber;
     e) concentration and type of dissolved material;
     f) concentration and type of suspended matter; and
     g) temperature, pH, etc.

     Different gases have been used for the flotation process (Vrablik,  1957;
Berry and Engel, 1969), but flotation by air is most common.  The essential
property of the gas is limited solubility in water.  Different flotation
methods are practiced:  dispersed air flotation, dissolved air flotation
(DAF), vacuum desorption flotation, and electrochemical flotation.  The  first
two methods differ in the way air is introduced to the wastewater; in the
other two, the air bubbles are generated from the wastewater itself.

          Dispersed air.  In dispersed (diffused) air flotation, air is  in-
troduced through a special type of disperser (e.g., diffuser, sparger, re-
volving impellers, perforated tubes, etc.) to the influent of a flotation
tank.  The use of diffusers poses problems, particularly in oily wastewaters
having high concentrations of suspended solids, waxes, and greases, because
diffusers are susceptible to plugging.  Air bubbles generated in dispersed
air  flotation have diameters of approximately 1000 p and higher.

          Dissolved air (DAF) .  In DAF, wastewater is saturated with air at
an elevated pressure  (usually 15 to 60 psig) in a retention tank for 1 to 5
minutes.  Then, the pressure on the air-saturated wastewater is reduced to
atmospheric, at the inlet to a flotation chamber.  Reduction in pressure
leads  to the release of tiny air bubbles from solution.  These have diameters
ranging upwards from 50 y.  Retention time in the flotation chamber is about
15 minutes.  Dissolving air in wastewater provides the maximum possible con-
tact that can be established, and oil droplets or suspended particles may act
as nucleation sites for bubble precipitation (Churchill, 1973) .  Three meth-
ods  are used in operating DAF units:  full-stream (total) pressurization,
split-stream (partial) pressurization, and recycle-stream (recycle) pressuri-
zation.  These three methods are shown in Figures 4 to 6.  Each mode has its
advantages and disadvantages (Churchill, 1973), but recycle-stream pressuri-
zation is considered superior to others (Rohlich, 1954; Simonsen, 1962).
Typical recycle rates are about one-third of influent flow.  Recycle units
are  the most common of all four flotation methods.

          Vacuum desorption.  Vacuum desorption flotation is accomplished in
three steps (Rohlich, 1954):

     a) a preaeration period to saturate wastewater with air at atmospheric
        pressure;
     b) release of larger air bubbles; and
     c) application of vacuum to the wastewater.

Depending on the vacuum applied, air bubbles have sizes similar to those in
dissolved air flotation but, because of limited solubility, the desorption
process may require higher energy than the dissolved flotation process
However, there is a reduction in turbulence relative to that which occurs in
DAF units (flotation chamber).  This turbulence is a deterrent to eff  f
particle/bubble collisions.                                          ective

                                     30

-------
                      AIR
OIL/WATER
 MIXTURE
                                                                                   SCUM
FLOTATION
 CHAMBER
                                                                                      CLARIFIED
                                                                                      EFFLUENT
                                  PRESSURE
                                 RETENTION
                                    TANK
                               Figure 4.  Total Pressurization System

-------
     OIL/WATER
      MIXTURE
N3
       FLOCCULATING
          AGENT
FLOCCULATION
  CHAMBER
                                                  PRESSURE
                                                 RETENTION-
                                                    TANK
FLOTATION
 CHAMBER
                                                                                            OILY SCUM
                                                                                               CLARIFIED
                                                                                               EFFLUENT
                                    Figure 5.  Partial Pressurization  System

-------
                                                                     OILY SCUM
   OIL/WATER
    MIXTURE
to
LO
    FLOCCULATING
       AGENT
FLOCCULATION
  CHAMBER
                                                                PRESSURE
                                                               RETENTION
                                                                  TANK
                                                                                        CLARIFIED
                                                                                        EFFLUENT
                                                  RECYCLE
                                                   PUMP
                                   Figure 6.  Recycle Pressurization System

-------
          Electrochemical.  Electrochemical flotation is a recent concept.
In this process, microgas bubbles are produced in'wastewater by electrolysis.
Bubble-size distribution is in the colloidal range,  0.001 to 0.1 u; it is
estimated that about 10 times as many bubbles are generated by this flotation
process than are produced by air pressurization systems.  Furthermore, gas
bubbles produced electrolytically possess surface charges of either polarity.
These surface charges can be effective in neutralizing oil dispersions and
suspended solids stabilized by surface charges.

     Design—Commercially available DAF units are rectangular or circular in
shape and are constructed of steel.  Important design parameters are:  pres-
sure, recycle ratio, inlet oil concentration, and residence time.  DAF units
generally consist of horizontal flow chambers, with  inside configurations
required for operation as gravity separators.  The hydraulic design of DAF
units is critical.  Length-to-width ratios vary from 2/1 to 5/1.  API has
made recommendations on the design of DAF units (API Manual, 1969).  Multi-
cell DAF units are manufactured and are claimed to be more efficient than
single-cell units (Ellis and Fisher, 1970).  DAF units are generally more
compact than API gravity separators.

     Performance—Air flotation is used for removal  of large oil globules
and suspended solids.  Flotation units are normally  preceded by API gravity
separators.  The efficiency of air flotation depends, to a large extent, on
air-bubble size.  The gas bubbles are in a state of  motion; collisions with
suspended matter must occur before attachment takes  place.  If a suspended
particle is much smaller than a bubble, the former may follow the viscous
streamlines of the bubble and not make contact.

     Because the bubble sizes expected in dispersed  air flotation are of the
order of 1000 y, only unstable primary dispersions and large, oil-coated
suspended solids will be separated.  Consequently, oil-removal efficiencies
lower than 25% and suspended-solids removal of about 20% can be achieved.
Despite the smaller bubble-size distribution in DAF, oil-removal efficiencies
are in the range of 60 to 90% (Stormont, 1956; Quigley and Hoffman, 1966).
Suspended-solids removal as high as 95% have been reported.  Particles exhib-
iting neutral buoyancy may be removed by flotation,  if they are not stabi-
lized by surface charges.  D'Arcy (1951) reported an unusually high final
effluent quality (between 5 and 7 ppm).  If properly designed and operated,
oil concentrations of effluents from DAF units can be kept below 50 ppm.

     The efficiencies of vacuum desorption flotation units are generally
lower than those of DAF units, even when high vacuums are used.  Electro-
chemical flotation may be capable of separating oil  droplets and suspended
solids stabilized by surface charges.

     Motion during the separation process has significant effect on device
performance; motion interferes with skimming of the  froth layer.  Floated
oil and solids may redisperse in the aqueous phase.   For an efficient separa-
tion, oil concentration in influent should not vary  significantly.
                                    34

-------
Filtration

     Filtration through granular materials is one of the oldest methods for
separating oil/water mixtures.  This technique is useful for removing sus-
pended matter and associated materials, particularly oil, from oily waste-
waters.  It is best suited for removing oil-coated solids that are a major
cause of fouling in coalescing devices.  Neutrally buoyant oil-coated solids
that may require infinite settling times can be removed, also.  As a result,
this technique is becoming very important in the petroleum industry because
of its capacity to reduce the concentratiori of oil and level of suspended
solids in production waters that are passed, subsequently, to secondary and
tertiary recovery operations.

     The mechanisms involved in removal of suspended solids by filtration are
very complex and little understood.  With deep granular filters of coarse
material, removal is primarily within  the filter bed (commonly referred to as
depth filtration) .  Some solids may be removed by a process of interstitial
straining, and oil may be removed by adsorption on the bed material.

     There are three main filtration techniques:  layer, membrane, and
fibrous-media.  Membrane filtration is discussed in the section on membrane
processes, and filtration by fibrous media is essentially a coalescence/
filtration technique which has been discussed elsewhere.  The present discus-
sion focuses on layer filtration only.  Layer filtration can be divided into
two types:  deep granular and multimedia.

     Layer Filtration—Granular media.  Several materials have been used as
granular media; these include sand, crushed anthracite coal, perlite, diato-
maceous earth, garnet or ilmenite sand, and powdered or granular activated
carbon.  A granular-media filter normally utilizes a bed of these materials
to remove the contaminant.  The most common and economical filter material is
fine, graded sand (slow and rapid sand filtration).  The effective size of
the filter media may vary from 0.35 to 1.0 mm.  In the backwash cycle, hy-
draulic grading of the sand occurs, with the finest sand forming the top of
the bed.  Therefore, if a suspended solid present in the process stream is
not trapped in the top layer, it is most likely to be found in the effluent
because larger voids occur in the direction of liquid flow.

          Multimedia.  To overcome the shortcomings of layered systems,
multimedia filters were introduced.  The principle of multimedia filtration
is that coarser grains of lower specific gravity will settle more slowly
than heavier but finer grains, during  the backwash cycle, provided the size
ratio between different materials is properly selected (Hsiung et al.,1974).
With dual media, combinations of two materials are used, e.g. crushed an-
thracite coal and silica sand.  The dual media have the largest grains
(coal) at the top and the smaller sand grains below, but the average grain
size at the bottom is still relatively large.

     To decrease average grain size, mixed-media filters are used.  A mixed-
media filter uses three or more materials, each of different size and den-
sity, that intermix, not stratify, to  form'a filter.grading from coarse at
the top to fine at the bottom in the direction of flow.  Typically,

                                     35

-------
anthracite coal, silica sand, and garnet are used (Evers,  1975).   The main
advantage of using mixed-media filters  is  pressure-drop  reduction at'high  "
flow rates, without deterioration of effluent quality.   Longer filter  runs
can be achieved, also.  Increasing the  number of filter  materials (different
densities and sizes)  approaches an ideal filter, one  in  which  the size of
media particles decreases uniformly. However, the additional  advantages  of
the ideal filter, for three-media or four-media cases, do not  offset  the
extra cost.

     Design—Deep granular filters consist generally  of  18 to  30 inches of
filter medium.  The filter material is  supported on an under-drain system
inside a pressure or gravity-flow vessel.   Concrete systems are available,
also.

     The design of mixed-media filters  depends on several factors:   the de-
sired effluent quality, the oil and suspended-solids  content of feed,  desired
flow rate, and maximum allowable head loss.  An excellent filter can be made
from 18 inches of 1.0-mm "effective"-size  coal (sp. gr.  1.6),  0.5-mm
"effective"-size sand (sp. gr. 2.6), and 0.3-mm "effective"-size garnet (sp.
gr. 4.0).  The size and quantity of each filter material can be varied inde-
pendently to meet specific requirements.  The average oil-droplet and
suspended-solids sizes are the primary  determinants of the size of the vari-
ous materials used.  After the first backwash, the materials become inter-
mixed but the filtration efficiency increases in the  direction of liquid
flow.  Typical filtration rates are 5 to 10 gal/min/sq ft.  The preferred
surface wash system is the rotating-arm type; during  backwash, the bed is
expanded at least 15% for efficient cleaning.  Air wash  is sometimes  used for
cleaning mixed media, but it is not as  effective as backwashing unless suffi-
cient water is provided for fluidization of the filter,  after  the air scour.
Multistage filters are being designed,  so  there is no interruption of  opera-
tion during backwash of one unit.  Devices are equipped  with automatic back-
wash systems which initiate backwash cycles on a time cycle or on the  basis
of filter head loss.

     Performance—Oil-removal efficiency and storage  capacity  of a filter
depend upon the media design.  Filters  have a limited oil-retention capacity,
and when effluent quality deteriorates  and becomes unacceptable, or when
pressure drop through the filter reaches a predetermined value, operation is
discontinued and filters must be washed to restore capacity.  Literature  on
the subject of backwashing filters used for oil removal  shows  that fouling
of media is a common problem.  Fouling  results from the  accumulation of sus-
pended solids within the bed and/or from biological growth in  the bed.  Cak-
ing or mud-balling in filters is common; filters are  not well  suited for
intermittent flows.  Failure of automatic  backwashing devices  is a major
cause of poor effluent quality.  Scheduling, or backwash frequency, consti-
tutes a problem, also.  Properly designed surface-wash and backwash facili-
ties can keep filter media clean over prolonged periods  of use.

     Tests using multimedia filtration  to precondition production waters for
reinjection during secondary and tertiary recovery operations  indicate the
absence of visible oil in effluents. Pilot tests conducted at the University
of Houston indicate that mixed-media filters are very effective in separating

                                     36

-------
unstable oil/water emulsions formed by pumping light Iranian crude oil into
the suction side of a fresh-water feed pump.  This filter was operated at 10
to 12 gal/min/sq ft, with an influent containing 750 ppm of oil; COD measure-
ments indicated a 95 to 100% oil-removal efficiency, in runs of up to 6 hours.
With 100 ppm of oil in the feed, a run lasted more than 30 hours and effluent
contained about 10 to 20 ppm of oil (COD measurements) .  Tests of a deep-bed
sand filter at the same loading rate indicated an oil-removal efficiency of
85 to 90%, when the feed contained 250 ppm of oil.  A filter run lasted about
16 hours (Hooper and Myrick, 1972).  It should be noted, however, that the
very long operating times and high oil-retaining capability of the bed were
probably due to the absence of suspended solids in the feed.

     Results of these tests show that filtration using granular materials is
a candidate for separating oil/water mixtures containing unstable emulsions,
oil-coated solids in suspension, and some free oil.  However, the process
generates secondary wastes (backwash) which have to be treated before dis-
charge.  The volumes of such wastes are generally small.  When used in con-
junction with chemical treatment, filtration can separate chemically stabi-
lized oil emulsions but, because of physical limitations, dissolved oil will
still be present in effluents from devices operating on the filtration
principle.

     Membrane—During the last decade, there has been rapid progress in the
development of artificial membranes.  As a result, membrane processes are
becoming increasingly important in wastewater treatment.  Hydrophobic and
hydrophilic membranes have been used for oil/water separation, particularly
for polishing purposes, because membrane fouling is a major problem.  For
efficient separation, the membrane has to be matched to the oil/water system
to be separated, and the feed must be free from suspended solids.  Pore size
and structure of the membrane determine the quality of effluent.  Membrane
processes for separating oil/water mixtures are still in the developmental
state; some of their disadvantages are limited rates and throughputs achiev-
able and high cost.  Processes are slow, but effluents from membrane devices
usually have oil content below acceptable discharge limit.

     Membrane processes are different from conventional microporous filtra-
tion.  Membranes are semipermeable, extremely fine in porosity, and easily
fouled.  Depending on a membrane's surface characteristics, it will pass oil
but not water, or vice versa.  If the viscosity of the oily phase is too high
for passage through a membrane at a practical rate at ambient temperature, a
method has to be devised for heating the feed mixture.  The energy require-
ment for processing large volumes of wastewater is large.  If the oily phase
has a low viscosity, a large surface area is required to process a volume of
wastewater.  Therefore, membrane processes are prohibitively expensive and
can be justified only for handling small-volume wastewaters or for removing
the last traces of oil, notably soluble oils and chemically stabilized emul-
sions.  In membrane processes, little pretreatment of feed is necessary.
New membranes are being developed that can process untreated feeds.  As the
research activity currently in progress yields good results, and if membranes
can be made cheaply, membrane processes will become more attractive and may
eventually replace other treatment and separation methods because of the
high-quality effluents that are possible.

                                     37

-------
     The major membrane processes are electrodialysis,  reverse osmosis,  and
ultrafiltration.

          Electrodialysis.   Electrodialysis has been used extensively for de-
salting applications.  The driving force for this technique is electrical
potential gradient; separation is based on selective ion transport across a
membrane.  The flow of electrical current in electrodialysis causes water to
flow (electro-osmosis) and to be separated from the oil phase by a water-
selective membrane across the current path.  Desired membrane properties are
high water-transport number, high electrical conductivity, good chemical re-
sistivity, and high mechanical strength.  Electrodialysis has not been inves-
tigated extensively as a possible separation technique, but it may be useful
for separating wastewaters containing emulsions stabilized by surface char-
ges .  This technique may be practical for small-volume  wastewaters; however,
pretreatment of feed will be necessary to diminish membrane plugging.  Some
of the disadvantages of this technique are high energy  requirements, dissoci-
ation of water at high current densities, imperfectly selective membranes,
low liquid flow rates, and concentration polarization.   These phenomena re-
duce the overall efficiency of this technique.

          Reverse osmosis.  Like electrodialysis, reverse osmosis has found
application to desalting operations for brackish water.  The principal driv-
ing force for this technique is pressure; reverse osmosis is essentially a
membrane filtration technique, in which pressures are greater than the os-
motic pressure exerted by species in solution.  Separation is on the basis of
molecular size.  Applied pressure forces water, and species smaller in size
than the rejection level of the membrane, through the membrane.  Oil and
larger species will be rejected at the membrane surface.  The smaller the
size of the species present in the oily wastewater, the greater the osmotic
pressure generated; therefore, reverse osmosis has an added advantage of
being able to remove species of atomic dimensions by using tight membranes
and operating pressures higher than the osmotic pressure.  Experiments have
shown that conventional dialysis membranes of the semipermeable, molecular
diffusion type are not appropriate for use in reverse osmosis.  Cellulose
acetate membranes are presently used most widely.  New  casting techniques
have led to production of ordinary and modified cellulose acetate membranes
which appear adequate for reverse osmosis processes on  the bases of high
water flux and good rejection of organic molecules.  A  wide variety of mem-
branes and backing systems is being developed.  One important commercial
system is based on a polyamide membrane.  Fluxes greater than 25 gal/day/sq
ft of membrane are possible, at applied pressures up to 1500 psig.  One
problem encountered is the slow decrease, with time, of oil rejection because
of membrane hydrolysis.  As a result, operation is applicable to a narrow pH
range, to increase membrane life.  Fouling is not a serious problem because
rejected matter is usually passed downstream by the flowing process stream,
but periodic cleaning is necessary.

          Ultrafiltration.   Ultrafiltration is similar  to reverse osmosis,
differing because it is not impeded by osmotic pressure.  Reverse osmosis'
systems operate at elevated pressures of 250 to 1500 psig, while ultrafiltra-
tion systems are operated at lower pressures of only 50 to 200 psig.  Because
of lower operating pressures, ultrafiltration membranes are more open, and

                                     38

-------
rejection of only colloidal or suspended matter and other macromolecules is
possible.  The predominant mechanism in ultrafiltration is selective sieving
through pores.  In the ultrafiltration process, feed is pumped through the
center of a porous tube on which a membrane has been cast integrally.  Hy-
draulic pumping pressure causes water and some dissolved, low molecular
weight materials to pass; however, emulsified oil, free oil, and oil-coated
solids are retained and concentrated in the tube.  To be useful for ultrafil-
tration, membranes must have a narrow molecular weight cutoff and a high
solvent flux at low-pressure differentials.  Early work in ultrafiltration
was done using cellophane or porous cellulose nitrate membranes, but repro-
ducibility was poor; adsorption on pore walls and plugging were common.  Be-
cause of breakthroughs in membrane technology, a variety of synthetic poly-
meric membranes is being tested in the ultrafiltration process.  The most
promising, thus far, are non-cellulosic in nature and allow operation over a
wide range of pH and temperature.  The limitations of this process are flux
decline with time and the phenomenon of concentration polarization.  Flux
decline is the sum of membrane compaction and fouling.  Economic studies have
shown that the most fruitful areas for significant gains are higher water
fluxes and longer membrane life.

     Design—Ultrafiltration and reverse osmosis plants consist of a series-
parallel arrangement of modules.  About four different module designs are
possible:  plate and frame, tubular, spiral-wound, and hollow-fiber configu-
ration.  Fibers are housed in cartridges that allow for expansion.  A de-
tailed comparison of module designs has been investigated and summarized by
Schatzberg et al. (1975) .  Ease of cleaning modules varies and backflushing
systems, with or without detergent cleaning, are incorporated into some of
the newer units.  A typical membrane has an asymmetric structure, consisting
of a thin, dense skin on a porous support.  It is approximately 100 y thick,
with a surface skin of approximately 0.2 y that acts as the rejecting sur-
face.  The pore size is of the order of 5 to 50 A and, thus, is approximately
1000 times smaller than emulsified oil droplets (Nordstrom, 1974).

     Generally, the membrane must be maintained wet at all times, although
some of the newer membranes can be handled dry.  More recently, there has
been a trend toward development of thin-channel hollow fibers (Messinger,
1974).  Hollow-fiber reverse osmosis systems are smaller in size than ultra-
filtration systems, but require greater power inputs (Finger and Tabakin,
1973).  Ultrafiltration systems can be operated in batch or continuous modes;
throughput rates are usually small in both cases.  Permeate rates in ultra-
filtration are higher than obtainable in reverse osmosis systems.

     Performance—Applications of reverse osmosis and ultrafiltration systems
to the separation of oily wastewaters have been investigated (Desai, 1971;
Goldsmith and Hossain, 1973; Nordstrom, 1974; Schatzberg et al., .1975).  Re-
sults of these tests indicate good separation, even though membrane processes
are still in the developmental stage.  Studies at the Naval Ship Research and
Development Center, Annapolis, with reverse osmosis systems, reveal that
treated oily wastewater, containing 500 ppm of oil, gave an effluent contain-
ing approximately 10 ppm of oil.  Permeate fluxes ranged from 10 to 25 gal/
day/sq ft, but periodic cleaning was necessary to minimize decline in per-
meate rates as a result of fouling (Finger and Tabakin, 1973).  Results of

                                     39

-------
experiments with ultrafiltration systems tested at the Center showed 90% oil-
removal efficiency, with a feed containing 100 ppm of oil.  Permeate was free
of emulsified oil and contained less than 10 ppm of dissolved oil.  Membrane
fouling was a problem, and only a detergent wash restored membrane flux
capacity.  Finger and Tabakin (1973) note that chemically emulsified oil
wastes from bilge cleaning would actually improve ultrafiltration, by mini-
mizing fouling, but detergents are known to cause membrane disintegration.
Membrane processes are very useful for separating stable emulsions, oil-
coated solids, and free oil.  Processes can achieve close to 100% oil-removal
efficiency, if adequately maintained and operated.  Removal of dissolved oils
cannot be achieved in these systems but, with the advent of tighter membranes
capable of withstanding very large operating pressures, reverse osmosis will
become a candidate for separating dissolved oil.  Ultrafiltration and reverse
osmosis are expensive processes, throughputs are low, and membrane replace-
ment is necessary as a result of fouling and other aging factors.  It is pos-
sible to minimize concentration polarization in ultrafiltration by operating
at high feed velocities parallel to the membrane surface and/or utilizing
thin-channel hollow fibers.

Coalescence/Filtration

     Coalescence of dispersed oils in aqueous suspension occurs in almost all
liquid/liquid separation processes.  Originally, the coalescence process was
used to remove water from oil in aerosol filtration and, because it was suc-
cessful, it was adapted for removing oil from oil/water mixtures.  The term,
as used in this report, refers to the coalescence process induced by flow
through porous media.  Coalescence is a complicated operation; it has been
studied extensively, but the phenomenon is not completely understood (Jordan,
1953; Voyutskii et al., 1955, 1958; Redmon, 1963; Farley and Valentin, 1965;
Sareen et al., 1966; Spielman, 1968; Splelman and Goren, 1972a,b) .  Several
theories have been advanced to explain the mechanism of coalescence (Vinson,
1965; Hazlett, 1969a, Jeffreys and Davies, 1971).

     The following major steps were proposed by  Voyutskii et al. (1955):

     a) collisions of emulsion globules with the ends of filter fibers;
     b) adhesion of droplets to the fiber;
     c) coalescence of the microdrops;
     d) adhesion of microdrops to the surfaces of fibers; and
     e) trickling of coalesced drops down the fibers.

Hazlett (1969a) proposed dividing the coalescence process into three main
steps:  approach of a droplet to a fiber or to a droplet attached to a fiber
attachment of a droplet to a fiber or to a droplet already attached to a     '
fiber, and release of an enlarged droplet from the fiber surface.  Each step
involves considerable complexity or alternative mechanisms.  The  fibers and
coalesced  liquid matrix may capture droplets by a number of mechanisms-
interception, Brownian diffusion, inertial impaction, gravity settling  'and
long-range attractive forces.  Except for-small particles, interception is
considered the dominant mechanism (Spielman and Goren, 1970).  Therefore  in

                                     40

-------
coalescence, oil dispersions  are  retained in  the porous media where the drops
grow larger, until  they  are large enough to be swept away by fluid-flow
forces.  Coalesced  oil snakes  through  the porous media until it reaches the
downstream  face of  the media, where it  is released in individual droplets
large enough to separate from the aqueous medium by gravity.

     Jordan (1965)  classified liquid/liquid coalescing into two general
types:

     a)  depth-type:   the coalescing operation occurs within the porous
         material and both phases  of the liquid system pass through the porous
         material; and
     b)  surface-type:  coalescence occurs on  the surface of the porous mate-
         rial and only one liquid  phase  passes through the porous material.

Surface-type coalescence and  ordinary  filtration are similar.  In fact, some
filtration  occurs in  any coalescence process. Therefore, devices operating
on  the  coalescence  principle  are  often called filter-coalescers.

     A wide variety of materials  has been used as coalescing media:  natural
fibers  (cotton and  wool), synthetic materials (fiber glass, viscose, nylon,
orlon,  and  felt), reticulated (open-pore) foams, membranes (hydrolyzed and
cellulose acetate), screens,  mats, and granular materials (pebbles, sand,
and diatomaceous earth). Because of the complex nature of coalescence, var-
ious approaches to  coalescence  processes have been identified.  The major
differences in these  modes are  almost  entirely the type of porous media used.

     Fibrous-Media—Coalescence using  fibrous media is the most important of
the coalescence types mentioned above.  Fibrous materials with different
internal, geometric,  and surface  properties have been used.  Voyutskii et al.
(1958)  used viscose and  wool  fibers.   Cotton  and glass wool were used by
Gudesen  (1964) .  The  most common  fibrous material in use is fiber glass
(Burtis  and Kirkbride, 1946;  Hayes et  al., 1949; Graham, 1962; Rose, 1963;
Sweeney, 1964).  Coalescence  performance of fiber glass can be enhanced by
coating  the fiber surfaces with certain synthetic resins, e.g. phenolic
resins,  to  render them hydrophobic or hydrophilic.  The use of fibrous media
implies  depth-type  coalescence.

     Membrane—Membrane  coalescence belongs to the surface-type discussed
above"Several separatory membranes are available.  Membrane materials are
usually  such that they can be treated  to render them hydrophobic or hydro-
philic.  Liquid flow  rates are  small.   Membrane coalescence devices are
normally preceded by  fibrous-bed  coalescers,  i.e. they are used as polishing
stages.  This mode  is similar to  membrane filtration and is covered in that
discussion.

     Centrifuge—As mentioned,  gravity  separation is an integral part of the
coalescence process.  In centrifugal coalescence, fibrous material is used
as  the coalescing medium, but  the final gravity separation of the oil and
water phases is enhanced by centrifugation.   Literature on this technique
has been reviewed.

                                     41

-------
     Bimetallic—Coalescence of dilute o/w emulsions by passage of the mix-
 ture through a bed containing a granular mixture of dissimilar metals (iron
 and aluminum) or carbon and a metal is a new concept proposed by Fowkes et
 al.  (1970).  Various bimetal and carbon-metal beds have been tested, but no
 complete  engineering assessment of this technique's potential has been re-
 ported.   The process is essentially an electrokinetic phenomenon (Koelmans
 and Overbeek, 1954), similar to electrophoresis and electrodeposition, but it
 has coalescence/filtration characteristics (Ghosh and Brown, 1975).

     Granular-Media—Beds of gravel, pebble or sand have been investigated as
 coalescing media for oil/water separation (Douglas and Elliot, 1962;
 Shackleton et al., 1960).  Coalescence using diatomaceous earth and other
 granular  materials has been reported.  This type of coalescence process is
 discussed further in the section on layer filtration.

     Other Porous Materials—Fine-mesh screens, non^woven mattings (polyester
 felt,  polypropylene felt, and glass mats) and reticulated foams have been
 used as coalescence media (Vinson, 1965; The Permutit Company, 1966; Chieu et
 al., 1975).  In this coalescence process and that using granular media, oil
 droplets  are adsorbed until the bed is saturated or the breakthrough oil con-
 centration in effluent is reached.  Then, operation is discontinued so that
 adsorbed  oil can be removed by backwashing or squeezing the media.

     Design of Fibrous-Media Coalescers—Since the mechanisms which lead to
 coalescence are not fully understood, design of filter-coalescers is largely
 empirical.  The most common configuration for fibrous-media coalescers is a
 cartridge.  Fibers are arranged and bonded together to provide a tortuous
 path for  fine oil droplets in order to achieve reliable coalescence (Finger
 and Tabakin, 1973; Freestone and Tabakin, 1975).  Desired properties of car-
 tridge elements are:  uniform structure, sufficient pore openings, and ade-
 quate  mechanical strength to withstand operating pressures of 25 to 75 psig.
 The direction of liquid flow through the coalescing element is usually radi-
 ally outward.  Cartridges (filter elements)  are housed in cylindrical vessels
 that can  be mounted vertically or horizontally.

     Coalescing elements are available in different sizes depending on unit
 capacity.  Coalescing devices normally contain more than one cartridge, each
 cartridge can be removed independently, and the entire unit can be assembled
 easily.   Fiber sizes vary from less than 5-p to 25-y diameter.  Failures of
 cartridge-type coalesce^ occur as a result of improper end-cap sealing (poor
 end cap-to-element sealing and/or poor gasket-to-end cap sealing) and poor
 quality control (defective elements and voids in filter media) .  Because of
 the problems associated with solids, most devices are equipped with a screen
 or prefilter ahead of the coalescer elements.  The main purpose of the pre-
 filter is to remove solid particles that may plug the pores of the coalescing
 filter elements.  However, the prefilter also acts as a coalescing filter and
 preconditions the feed.  Coarse filter cartridges, with approximately 25- to
 100-y-diameter fibers,  are used as prefilters.

     Since gravity separation is an integral part of coalescence, commercial-
ly available coalescers are multistaged.  The first stage is used as a gross
gravity separator or may contain a prefilter.  Succeeding stages alternate

                                    42

-------
coalescence and gravity separation.  Important design parameters are:  amount
of fibrous material (voidage), arrangement of fibers, and hydrodynamic
factors.

     Coalescers are compact and easily assembled, and operation is flexible.
The? are usually not regenerated through fluidization or backwashing; there-
fore, replacement of coalescing elements is necessary.  Coalescers equipped
with automatic cartridge-cleaning devices are available commercially. Figure
7 is a schematic diagram of a cartridge filter/coalescer.

     Performance of Fibrous-Media Coalescers—In coalescence, a principal
driving force is interfacial  tension (Yu and Ventriglio, 1969).  Lowering the
interfacial tension of the oil and water phases promotes coalescence of the
dispersed oil phase.  But coalescence can also occur in systems having high
interfacial tension, if the density difference between the two phases is high
and  coalescing surfaces have  a high degree of roughness (Jeffreys and Davies,
1971).  Attempts have been made to analyze the performance of fibrous-bed
coalescers using filter coefficients.  Theoretical expressions for filter
coefficients have been developed by several investigators using various co-
alescence models (Spielman and Goren, 1970, 1972a,b; Sherony and Kientner,
1971; Rosenfeld and Wasan, 1974).  Important coalescing parameters are:

     a) physical properties of coalescing media-preferential wettability of
        bed, bed-spreading characteristics, fiber size, packing, pore-size
        distribution, uniformity of structure, surface roughness of fibers,
        and two-phase permeability of bed;
     b) interfacial tension and contact angle;
     c) flow velocity; and
     d) physicochemical properties of fluids (viscosity, density,
        temperature, etc.).

     Combinations of small- and large-size fibers improve oil-removal effi-
ciency, if the larger-size fibers are located at the downstream end of the
cartridge (Voyutskii, 1958; Hazlett, 1969b; Jeffreys and Davies, 1971).
Resins used in bonding fibers together usually give an intermediate contact
angle (Hazlett, 1969a).  Preferentially oil-wetted mats are believed to be
less efficient than aqueous-wetted mats for separation of o/w emulsions
(Spielman and Goren, 1972b).  Voyutskii (1955) found a critical flow veloc-
ity, less than 1 cm/sec, below which separation was possible.  Gloyna and
Ford (1974) found that oil-removal efficiency of coalescers is related to
two major factors:

     a) variation in type of  oil, degree of emulsion, droplet size, and
        suspended-solids concentration; and
     b) fluctuations in flow  rates, influent oil concentration, and equipment
        upsets.

     For successful operation of filter beds, it is important that solid par-
ticles are removed from the oil/water mixtures before passage through the
bed.  Solid matter deposited  in the bed will not only change the voidage and
local fluid velocities in the bed but, more important, the surface properties
of the bed will be changed.   Particulate matter (suspended solids and

                                     43

-------
                                 TREATED WATER  EFFLUENT
OIL/WATER
MIXTURE
\\v\\\\\\\\\\\\\ >
              \\\\ \\v\\\\\\\\\\
                 CARTRIDGE
                 COALESCING
                  ELEMENT

                                           t
                                                 V

                                           J

                                    \
                                                      \
                                                        \
                                                    SEPARATORY
                                                    MEMBRANES
                                     SEPARATED OIL
    Figure 7.  A Complete  Liquid/Liquid Coalescing System

                              44

-------
gelatinous materials)  interferes with effective coalescing.  Plugging of
coalescing media often is  the  cause of coalescer failure.  Plugging results
in, an increase in head loss and coalescer elements require elaborate cleaning
to restore efficient separation or replacement of filter elements.

     The effects of surfactants on coalescence have not been studied exten-
sively.  Detergents concentrate at the oil/water interface during coalescence
(Lindenhofen and Shertzer, 1967), limit droplet growth prior to detachment,
and hinder droplet release from the downstream face of the filter (Hazlett,
1969b) .  It has been suggested that surfactants degrade coalescer performance
by absorption on the fibers, resulting in an increase in the contact angle
and decrease in wettability (Hazlett, 1969b).  This alteration of the surface
properties of the coalescing fibers leads to poisoning of the coalescer and
eventual failure.

     Coalescers are particularly suited for removal of oil dispersions, but
segregating phases after separation is difficult and expensive.  Small quan-
tities of water may be present in the separated oily effluent of coalescers.
Oil content of influents to coalescers should be limited to 200 ppm and
should fluctuate only  within a narrow range.  However, if coalescers are well
operated and maintained, oil-removal efficiencies can be as high as 99%.
Typical values are higher  than 90%.  Results of studies at the Annapolis
evaluation facility indicate that 10 ppm of oil in effluent water can be at-
tained with some commercially  available coalescers (Finger and Tabakin,
1973).  Coalescers are not designed to remove dissolved and non-colloidal
oils (solubilized oil); therefore, oil/water mixtures containing these sys-
tems cannot be treated effectively.  Oil dispersions stabilized by surface
charges cannot be treated  adequately, although bimetallic coalescers have a
potential for treating this system.  Solids, stabilized oil emulsions, and
suspended slimy materials, if  present in influents to coalescers, may reduce
efficiencies below those obtained with gravity separators (Gloyna and Ford,
1974).

Adsorption and Absorption

     Sorption on solids, particularly activated carbon, has become a widely
used operation for purification of waters and wastewaters.  Activated carbon
has an affinity for organic matter present in petrochemical and refinery
wastewaters and, thus, is an effective means of removing dissolved oil, solu-
bilized oil, and chemically stabilized emulsions that cannot be destabilized
by chemical addition and other methods.  These oil/water systems pose serious
problems in various oil/water  separating devices.

     Activated carbon has demonstrated large adsorptive capacity and desired
surface properties, making it  adequate for adsorption processes; it is the
material most widely used.  Adsorption is an interface phenomenon; oil is
selectively adsorbed on the surface of the carbon, but the adsorbed film is
only a few molecules thick.  Adsorption of oil molecules from the aqueous
phase to the carbon surface occurs as a result of a combination of various
forces:   adhesive, cohesive, electrical, surface tension, and van der Waals.
Contact between carbon surfaces and oil wastewater is achieved through
fixed-bed or expanded-bed carbon columns.  As adsorption of oil and other

                                    45

-------
 organic matter present in the waste stream is accomplished, the carbon loses
 its  adsorptive capacity and breakthrough occurs.  Spent carbon must be re-
 placed or reactivated.  The process is uneconomical if the spent carbon is
 wasted; therefore, it must be reactivated and reused with new carbon added to
 make up the losses of regeneration.  Regeneration is an extremely important
 consideration in the use of activated carbon for treatment of oily waste-
 waters.  Currently, regenerative methods are mainly thermal.  The regenera-
 tive process requires a large capital investment and has high operational
 costs.  It is presently feasible to regenerate carbon, by conventional ther-
 mal  techniques, for several cycles of successful saturation and regeneration.
 The  feasibility of other regenerative techniques is being investigated.
 These are (1) the use of steam to drive off the adsorbed oil and (2) repeated
 washings of the spent carbon with solvents.  The results of these attempts
 have been disappointing because of high cost and the small efficiency
 achieved.

      Design—The basic design and operating concepts of gravity flow, pres-
 sure, and expanded-bed-type-flow adsorbers are essentially similar to those
 for  granular-bed filters of the corresponding types.  Fixed-bed units are
 usually vertical pressure vessels containing activated carbon supported on an
 under-drain system.  Important design considerations are flow rate, contact
 time, depth of carbon, influent oil and suspended-solids concentrations, and
 desired effluent quality.  Multiple adsorption columns are usually provided
 in series or in parallel so that a unit can be taken out for regeneration.
 Moving-bed carbon adsorption columns are also being used to eliminate the
 spare columns required for regeneration.  Chiyoda, Japan, uses a new method
 of contacting carbon and wastewater.  Unlike conventional fixed-bed systems,
 the  Chiyoda multistage, fluidized-bed system is based on a unique process in
 which activated carbon is circulated in continuous, countercurrent contact
 with the wastewater and, when spent, carbon  is reactivated for re circulation.
 The  advantage of this system is the absence of the removal of spent carbon
 and  replenishment with fresh carbon.  The system can be operated without
 interruption.  Design is compact, space requirements are reduced, and there
 is a flexibility of future expansion (Chiyoda Company, 1974) .

      Performance—If properly operated, carbon adsorption columns provide
 effluents that will meet practically all discharge limits for oil.  Pilot
 tests conducted by the Ben Holt Company, California, using carbon adsorption
 columns in lieu of coalescing devices, indicated oil and surfactant present
 in the influent were removed efficiently.  Feed to the combined system was
 typical Navy oily waste containing seawater, oil, sludge, dirt, and chemi-
 cally stabilized emulsion.  Influent to the carbon columns had an oil content
 varying from 35 to 530 ppm.  The highest effluent oil content was 2.5 ppm,
 measured by chloroform extraction and IR spectrophotometry.  Surfactant  con-
 centration at the carbon inlet varied from 63 to 630 ppm and was removed to
 a  level not detectable by IR analysis (Ben Holt Company, 1974) .  Feasibility
 studies of variations of the adsorption process by Calspan Corporation,  New
 York, showed efficient separations.  In the latter studies, carbon adsorption
was  used in conjunction with plain sedimentation, dissolved air flotation,
 mixing and filtration, and polyelectrolyte coagulation.  The average oil
 content of effluents was 1 ppm, with contact times of less than 3 minutes
 and  influents containing up to 2000 ppm of oil.  All samples were analyzed

                                     46

-------
for oil content by percent transmittance measurement.   Though results  showed
efficient separation, the carbon dosages required were quite large:  1 g of
activated carbon per liter of wastewater (Wang et al., 1973).

     Oil-removal efficiencies in carbon adsorption are usually very high: in
the range 95 to 100%; a typical value is 98%.  An obvious disadvantage of the
simple adsorption process is low capacity, necessitating huge surface  areas.
Therefore, carbon adsorption columns are used as polishing stages in lieu of
other separators.  Removal of suspended solids is accomplished, also,  because
of the filtration characteristics of carbon columns.  However, in spite of
the high efficiencies achievable with this process, it has not found wide-
spread use because of the large expenditure involved.   Attempts to use coke,
instead of activated carbon, as the absorption medium and to regenerate a
saturated adsorption surface by coking have not provided any encouragement
(Freestone and Tabakin, 1975).  Nonetheless, carbon adsorption has proven to
be the answer where other physical and chemical treatment techniques have
failed and, as wastewater control regulations become more stringent, the
process has become indispensable.

Electric and Magnetic Separation

     Separation of oil from water by electric and magnetic means has not been
fully investigated because of the high cost involved and the low liquid flow
rates achievable in such devices.  These disadvantages, together with  the
technological problems encountered, have made these methods unattractive.
The two methods in this category are electrophoresis and magnetization.

     Electrophoretic—Electrophoretic separation has been used in oil  refin-
eries for recovering oil from oil-rich wastewaters, but its application has
been limited to processing small volumes of wastes in which w/o emulsions are
the major contaminant.  Electrophoresis, an electrokinetic phenomenon, is the
principle of separation in this technique.  A strong electrical field  is es-
tablished in the wastewater; dispersed particles move along the lines  of
force and become separated from the continuous medium because of the net
electrical charge on each particle.  The electrophoretic mobility of a water
particle is greater than that of an oil globule of the same size.  Therefore,
this method is practical for separating w/o emulsions which are produced
during drilling operations.  This method becomes more effective as the sizes
of the water droplets become smaller.  When the water droplets are larger
than 10 y in size, a considerable electrical energy has to be used for an
appreciable separation to occur.  An increase in the energy requirement can
lead to hydrolysis of water; hydrogen and/or chlorine gas may be generated
at the electrodes.  Gas generation makes the wastewater turbulent and further
degrades separation efficiency.  Separation of oil-in-seawater emulsions is
difficult because of the high electrical conductivity of seawater and the
possibility of concentration polarization.  In general, electrophoretic
separation is a slow, inefficient process and is not suitable for the
treatment of large volumes of oily wastewater.

     Magnetic—The feasibility of recovering oil from fine, stable, o/w
emulsions by magnetic means has been investigated by Kaiser et al. (1971).  In
operation, a ferrofluid is added to the wastewater to make the dispersed oily

                                     47

-------
phase magnetically responsive.  The wastewater is  passed through  a packed bed
placed in a magnetic field.  Packed beds are composed of magnetic particles
or screens.  Results of experiments indicate virtually complete removal of
oil particles, as small as 1 y in diameter,  using  air-gap fields  of several
thousand oersteds, bed packing of several inches in length,  and oil-phase
magnetization of about 2 to 10 gauss.   Residence times of less than a minute
were used.  Magnetic separation is practical for separating  small volumes of
oily wastewater in which the major oil/water system present  is secondary
dispersions; energy requirements are large,  and operational  costs are high.

Thermal Separation

     Separation of oil/water mixtures  using  thermal treatments is a feasible
concept, but impractical if large volumes of wastewater are  handled.  How-
ever, thermal methods have been used to resolve emulsions present in oily
wastewaters from crude oil production.  After the  emulsions  are destabilized,
oil can be separated from wastewater through use of other separation tech-
niques .  These methods have found only limited application and are adapted
specifically for those wastewaters which contain mainly w/o  emulsions or when
water is present in small amounts.  In systems such as these, the wastewater
can be demulsified using thermal treatment.   The process is  economically
attractive, if oil recovered from the  wastewater has a high  heating value or
can be reprocessed and used as fuel oil.  The major thermal  treatment methods
are heating, evaporation and distillation, and freezing or crystallization.

     Heating—Heating has been used extensively to resolve crude  oil emul-
sions.  It has been used in conjunction with chemical addition to destabilize
chemically stabilized emulsions.  The  process also increases the  amount of
dissolved oil in the wastewater.  It is simple: the basic principle of this
technique is alteration of the vapor pressure difference between  oil and
water.  Energy requirements for large  volumes of wastewater  make  this process
uneconomical.  Large equipment sizes are required, also.

     Evaporation and Distillation—These separation techniques are similar to
heating and suffer from the same disadvantages, even when the distillation
process is carried out at reduced pressure.   Complete separation  of waste-
water into distinct oil and water phases is  impossible because, during opera-
tion, the oil fractionates and some fractions will be present in  both con-
densate and distillate.  Even when the distillation process  is carried out in
stages, there is a large energy demand for heating and providing  reduced
pressure before oil can be removed completely.

     Freezing and Crystallization—In these  methods, the difference in the
freezing points of the two liquids is  used to effect separation.   Oil/water
separation by freezing and crystallization is generally considered as
economically infeasible when large volumes of wastewater are handled.  Little
literature on these techniques for oil/water separation is available.  How-
ever, there is obvious need for complex refrigeration equipment  and
relatively large inputs of energy (Yu and Ventriglio, 1969).

     Thus, thermal separation, though  feasible, is not practical  because of
the high cost involved.

                                    48

-------
Sonic and Ultrasonic Separation

     The use of sonic and ultrasonic devices to separate liquid/liquid mix-
tures is prevalent in the dairy industries, but the capability of these tech-
niques to separate oil from oil/water mixtures has not been investigated
fully.  The lack of engineering assessment of these separation methods is a
consequence of the use of sonic and ultrasonic devices, originally, in emul-
sification processes; they were deemed unsuitable for the reverse operation
(demulsification).  There are suggestions that these techniques may be fea-
sible, if the appropriate wave frequency is used, for destabilizing emul-
sions.  Destabilized emulsions can be coalesced and separated, using other
separation techniques.

     Determination of the appropriate wave frequencies necessary to destabi-
lize emulsified oil/water systems is a costly operation.  The characteristics
of untreated wastewaters change and volumes of oily wastes to be processed
are usually large.  Using the wrong frequencies may break stable emulsions
into finer dispersions or shatter already coalesced globules.

Coanda-Effeet Separation

     Separation of oil/water mixtures by utilizing the fluid-dynamic phenom-
enon called the "Coanda effect" is a new concept in liquid/liquid separation.
The basic principle of this process was proposed by Henry Coanda, and the
Navy has designed a separator capable of treating bilge and ballast waste-
waters by the wall-attachment (Coanda-effeet) phenomenon (Navy, 1974; Paszyc
et al., 1975).  In this process, an oil/water mixture is injected into an
unbounded region.  The jet splits into two subjets at the apex of concave
walls; each subjet is deflected toward an adjacent wall, becomes attached,and
flows along the wall enclosing a separation-bubble zone.  The curved flow of
each subjet creates a centrifugal force and the separation-bubble zone is
formed.  An oil droplet released at the jet nozzle experiences an inward
motion due to the centrifugal force and undergoes radial displacement toward
the bubble zone; it can coalesce with other oil droplets trapped there.  Oil
which accumulates in the bubble zone can be drawn off by suction.  A
schematic diagram of the process is shown in Figure 8.

     Design—Important design parameters are:  length of the attachment walls,
radii of curvature of the splitting walls, velocity of fluid at the nozzle,
and the size of the nozzle.  Basically, a Coanda-effeet separator comprises:
inlet and outlet ports, an oil-collection chamber, oil/water interface detec-
tion probes, and an oil suction device.  Multistage units are under design.

     Performance—Results of test evaluations indicate that devices cannot
achieve greater oil-removal efficiency than an API gravity separator or other
primary separation technique.  A major difficulty encountered with these
devices is the turbulence that arises during processing, as a result of jet
flow.  Turbulence breaks oil dispersions into smaller droplets that are more
difficult to separate.  However, in the absence of turbulence, devices will
be capable of separating only free oil and unstable primary dispersions.
Oils with the same density as the aqueous phase will not be separated.
                                     49

-------
             CLARIFIED
             EFFLUENTS.
        REATTACHED JETS
          . VELOCITY
           PROFILES
 NOZZLE
                         OIL/WATER
                          MIXTURE
                           INLET
                                              SEPARATION
                                              BUBBLE
                                              ZONE
                                              MIXTURE
                                              JETS
Figure 8.   Coanda-Effeet Separator
                50

-------
Viscosity-Actuated Phase Separation

     A new concept for separating a mixture of two immiscible liquids has
been proposed and tested by Union Carbide Corporation.  The principal driving
force for the separation is the difference in viscosity of the liquids.  The
basic principle of this separation technique was proposed after laboratory
observations of flow fields induced when a mixture of two immiscible liquids
was subjected to pressure gradients.  These laboratory tests revealed that if
a drop of a mixture of oil and water is placed in the V-notch formed by two
flat plates that can be squeezed together, separation of the liquids occurs
because the low-viscosity water flows from the pressurized area more rapidly
than the higher-viscosity oil (Union Carbide Corporation, 1973).

     An apparatus was constructed to subject oil/water mixtures to the proper
type of flow field in order to achieve separation of the mixture.  The appa-
ratus consists of a screw capable of rotating inside a threaded hollow cylin-
der.  The rotor and stator are fitted so there is maximum eccentricity be-
tween the two parts during motion of the rotor.  In operation, the changing
clearance between the rotor and stator induces a squeezing action on oil/
water mixtures and separation occurs.

     Design—During a test evaluation, two types of viscosity-actuated phase
separation devices were used.  One unit (Orbiting Buttress Threaded Device)
consists of an orbiting-screw rotor and a threaded hollow cylinder; both
rotor and stator are constructed of rigid material.  The second unit (Rotat-
ing Buttress Threaded Device) is similar to the first unit; however, the
rotor is threaded and the stator is smooth.  Later in the test evaluation,
both units were constructed of an elastomeric material.

     Important design parameters include the degree of eccentricity between
the rotor and stator and the clearance between the two parts.  Figure 9 is a
diagram of an Orbiting Buttress Threaded Device.

     Performance—Factors that influence device performance are separator
geometry, the viscosity differential between the two liquids, and the balance
between feed and separation rates.  Results of test evaluations were not
encouraging, and the poor performance reported was attributed to excessive
clearances between the rotor and stator.  For an effective separation, the
oil-droplet size must be larger than the clearances between the stator and
rotor.  Because of geometric design limitations and inaccuracies in machin-
ing, the lower limit of the clearances could not be achieved.  Results indi-
cate that the devices increased emulsification instead of effecting separa-
tion.  Some success was reported during separation of oil/water mixtures
with oil as the continuous phase.  These devices cannot qualify as oil/water
separators, at present.

Chromatographic Separation

     Chromatographic techniques are widely used in liquid/liquid separation
and purification applications.  Gas, liquid, and gel chromatographic methods
have been suggested as processes capable of separating oil/water mixtures.
However, there is no literature available on the use of these methods for

                                     51

-------
               OIL/WATER MIXTURE INLET
OIL EFFLUENT
ORBITING
 SHAFT
                                                      ECCENTRIC
                                                      DRIVE  SHAFT
 WATER
EFFLUENT
    Figure 9.   Orbiting  Buttress  Threaded Device
                          52

-------
bulk oil/water separation (Yu and Ventriglio, 1969).  Chromatography has use
in several applications, but the technique suffers from the following disad-
vantages:  small throughput, high head loss, and very low rates.  Selection
of column packings capable of extracting the oil phase from oil/water mix-
tures is difficult and expensive.  Disposal of spent column materials is a
problem and, if material is regenerated using an appropriate solvent, a
secondary waste is generated.  Therefore, the high cost of these methods has
discouraged investigation as candidate oil/water separation processes.

     Such sophisticated and costly separation methods may prove useful for
small volumes of oily wastes containing little oil as regulations and
discharge limits become more stringent.
                                     53

-------
                                  SECTION  7

                    CRITICAL  REVIEW OF SELECTED LITERATURE

 Title:   Coalescing Plates  and  Packs  for Oil/Water Separation in Various
         Shipboard Applications

 Report  No.  and Date:   CG-724305.2/6, January 1973

 Authors:   J.  G. Merryman and E.  R. Osterstock

 Manufacturer of Equipment:   General  Electric Company, Philadelphia, Pa. 19101

         Design Features:   Device is  a gravity-differential, parallel-plate
 separator.   Plates are convoluted and made of polypropylene.  Plates are per-
 forated,  also, to enhance  oil-globule coalescence.  Equipment is divided into
 four compartments; each compartment  contains a plate pack.

         Wastewater Characteristics;  Four oils were used in the tests:  (1)
 Navy Special Fuel, (2) Navy  Distillate Fuel, (3) Venezuelan Crude, and (4)
 a mixture of hydraulic fluid and lubricating oil.  Aqueous test fluids were:
 (1)  tap water and (2)  seawater.   Mixing of oil and water was accomplished by
 three different methods, yielding different levels of emulsification.  In the
 first two methods, a Lightning mixer was operated at two speeds.  In the
 third method,  oil and  water  were mixed and re circulated through a centrifugal
 pump.  Bilge water was simulated by  adding detergent and sand to the oil/
 water mixture containing test  oil (4).

         Methods of Analysis;   Oil content of samples was measured by (1)
 chloroform extraction  and  transmittance measurement in a B & L spectrophotom-
 eter and (2)  gravimetric analysis.  All samples were acidified before analy-
 sis. Results  of both  analytical techniques were comparable.  In most tests,
 the  colorimetric method was  used, because of accuracy in the low concentra-
,tion range.

         Results:   Several  parameters that affect separation efficiency were
 investigated.   These parameters  were plate length, oil concentration, temper-
 ature,  and  flow rate.   Results showed the oil content of effluents was always
 less than 50 ppm,  for  influent containing 500 ppm of oil.  When influent oil
 concentration  was  increased  to 5000 ppm, effluent oil content was higher than
 100  ppm.  Effluents in tests using the centrifugal pump had more oil than
 those produced in tests  using  the Lightning mixer.  Addition of detergents
 degraded  device performance, also.  Poor performance was observed when ship
motions were simulated by  rocking the separator.
                                      54

-------
        Critical Comments;  Plate- type, gravity-differential separators can-
not separate emulsions effectively.  They are adequate for gross separation
only.  Effluents from these devices have to be processed further to achieve
desired discharge quality.


Title:  Feasibility Test Program of Application of Coalescing Phase Oil/Water
        Separators to Self-compensating Fuel Tanks in Surface Ships

Report No. and Date:  CG-D-88-74, May 1974

Authors:  J. B. Arnaiz and E. Batutis

Manufacturer of Equipment:  General Electric Company, Philadelphia, Pa. 19101

        Design Features;  Equipment is a commercial, parallel-plate-type,
gravity separator, capable of flow rates up to 1150 gpm.  Length of the sep-
arator is 14 feet; plate length is 8 feet.  This device has provision for
connecting the inlet to a ship fuel tank discharge port.

        Wastewater Characteristics;  Actual ballast water discharged from a
Navy oiler (U.S.S. Koelsh) during routine fueling operations.  Ship had been
docked a few days .

        Method of Analysis;  During the test evaluation, on-line measurements
of entrained particle size and number were obtained for influent and efflu-
ent, using an HI-AC particle counter.  On-line turbidity measurements were
made on the effluent, using a Keene turbidity meter.  All samples were as-
sayed for oil content by carbon tetrachloride extraction and spectrophototn-
etry.  Prior to extraction, samples were acidified with a mixture of sul-
furic and hydrochloric acids in a one-to-one ratio.

        Results ;  Data reported by the performing agency were generally lower
than the results of chemical analyses of samples carried out by U.S. Navy
Laboratories.  However, both sets of results indicate the oil content of
effluents was less than 8 ppm, in all tests .  Influent oil concentration
varied from 4 to about 1000 ppm.

        Critical Comments^  Device functioned satisfactorily and separating
efficiency was unusually high.  This may be the result of the absence of
emulsions in the oil/water mixture.  Ship had been docked some time before
deballasting; it is conceivable that considerable separation had occurred
before the test evaluation was begun.
Title:  Oil/Water Separator Evaluation

Report No. and Date:  NCSL 252-75, July 1975

Author:  John Mittleman

Manufacturer of Equipment:  Assembled by author

                                     55

-------
         Design  Features:  Device is a gravity-differential, plate-type sep-
 arator.   Plates are made of oleophilic material and arranged in stacks verti-
 cally in one  chamber and horizontally in others.  Equipment contains auto-
 matic valves, sensors, and other hardware items necessary for maintaining
 control  of the  movement of fluids in the system.

         Wastewater Characteristics;  In preliminary tests, synthetic waste-
 water was produced by emulsifying Navy Distillate Fuel Oil in tap water.
 Later, ballast  and bilge water pumped from a naval vessel were used as feed.

         Methods of Analysis:  Two methods were used to quantify oil concen-
 trations in samples:  (1) visual quantification for those effluent samples in
 which there was a persistent sheen and (2) extractions with carbon tetrachlo-
 ride followed by light-transmittance measurements at 420 nanometers.
                                                           ',
         Results;  Data were presented graphically, and indicate poor perfor-
 mance throughout the test period.  The oil content of all effluent samples
 averaged more than 50 ppm.  Separated oil contained a high concentration of
 water, also.  Device could not separate oil-coated solids.

         Critical Comments:  Gravity-differential separators are useful only
 for gross separations.  Oily wastewater containing high oil concentration, as
 emulsions,  cannot be separated in such devices.  Results of this test evalu-
 ation are indicative of the limitations of plate-type oil/water separators.
 Title:  Vortex Concept for Separating Oil from Water

 Report No.  and Date:  4105.2/1, January 1973

 Authors:  R. C. Stoeffler and C. E. Jones

 Manufacturer of Equipment:  United Aircraft Corporation, East Hartford,
                            Conn. 06108

        Design Features:  Separators tested consist  of 6-inch and 9.5-inch
 vortex tubes having four injection points; tubes were made of Lucite to allow
 visual inspection of operation; end walls were plain discs fastened to the
 vortex tube.  A schematic diagram of the device is shown in Figure 10.

        Wastewater Characteristics;  Six types of oil (different densities
 and viscosities) were injected separately into the water feed line using a
 pressurized injection probe.  In some tests, screens of different mesh sizes
 were inserted in the input line to emulsify the oil; in other tests, a
 centrifugal pump was used.

        Method of Analysis;   Oil content of samples was measured by carbon
 tetrachloride extraction and infrared analysis; values of the actual and
measured oil concentrations  were close.

        Results;   Overall, separator performance was poor.  Oil content in


                                    56

-------
                                                                                         r
01
=n
EXIT END-x
WALL \
\
N
WATER EXIT-v
DUCT
r
"






__\J
	 Ij
=r==:v





•L^^


£ 1
1



-,
1h-












^=3

1
s
r-OIL-COLLECTOR
DUCT
D ,
f ¥ °



r
- -\

t



^~ — SUPPORT STRUTS)
i SECONDARY /
NJECTOR (TYPICAL)j
/
i

;

c



r-LUCITE/ 1
b \
1 -


r"
A—a —

/
r


/



0



/
o
1 0 '
1 Or

j=>


y- INLET E^
,/ WALL



^-MIXTURE
INJECTOR
(TYPICAL)



                                                                     TYPICAL
                                                                     INJECTOR
                                                                       BI-
                                  SECTION B-B
                                                                                            B
                                                                                    SECTION A-A
NOMINAL SEPARATOR
CAPACITY, QM- GPM
10
50
DIMENSIONS
L - IN.
18, 36
28.5, 57
D - IN.
6
9.5
d* - IN.
0.6, 1.2
0.95, 1.9
do-
0.18
0.305 .
Dj - IN.
0.209, 0.272, 0.323, 0.453
0.609, 0.718, 0.922

-------
 effluents was  generally greater than 50 ppm; separated oil contained up to
 95% water.

        Critical Comments;  Test data indicate the vortex concept is not
 feasible as an oil/water separation technique; the devices fail to meet
 stringent discharge limits.  Separators were unable to separate emulsified
 oil;  separated oil contained excessive free water, making reprocessing
 necessary.  Slightly better separation was achieved when the process was
 followed by gravity separation.

        Cost:  The following prices were reported:
                   100 gpm unit:   $10,000
                 1,000  "    " :   $94,000
                10,000  "    " :  $940,000
 Title:   Investigation of the Use of a Vortex Flow to Separate Oil from an
         Oil/Water Mixture

 Report No. and Date:  714103/A/001, November 1970

 Authors:  A. E. Mensing, R. C. Stoeffler, W. R. Davison, and T. E. Hoover

 Manufacturer of Equipment:  United Aircraft Corporation, East Hartford,
                            Conn. 06108

         Design Features;  Device is a 10-inch-diameter by 29.25-inch-long
 vortex tube; tube was made of Lucite to allow visual inspection; end walls
 were made of plain discs fastened directly to the tube.  Injection points
 were located on the peripheral wall of the vortex tube.

        Wastewater Characteristics;  Influent oil/water mixtures.were pre-
 pared by metering oil into a water line; four types of oil were used but
 none of  the oils were emulsified.

        Method of Analysis;  Samples were left quiescent for several hours
 until the oil and water phases separated; then, the respective volumes of
 the two liquid phases were measured.

        Results:  Authors claim that it is possible to "capture" up to 85%
 of the injected oil and the separated oil can contain less than 15% water.

        Critical Comments:  The analytical method is unacceptable, because
natural gravity separation is possible as the influent did not contain emul-
 sified oil.  Data presented graphically indicate overall performance was
poor even when the process was followed by gravity separation.  The authors'
claims cannot be verified; thorough testing of the device is necessary and
test data need to be reported in a better way before meaningful conclusions
can be drawn.
                                    58

-------
Title:  Vacuum Desorption Concept  for Removing Oil from Water

Report No. and Data:  USCG  734305.2/8, March 1973

Author:  George M. Pomonik

Manufacturer of Equipment:  Mechanics Research, Los Angeles, Calif. 90045

        Design Features;  System consists basically of:  collection tank,
vacuum separation  tank, vacuum pump, revolving drum skimmer, and flow-control
devices.  A modification included  polypropylene coalescing plates added to
the vacuum tank.

        Wastewater Characteristics;  Mechanically emulsified oil/water mix-
tures were prepared with tap water and various oils.  Occasionally, salt was
added to simulate  bilge and ballast water.  A few tests were conducted with
detergent and fine sand added to the influent.  Mixing of oil and water was
done by pump.

        Method of  Analysis;  Oil content of samples was determined by extrac-
tion with petroleum ether,  evaporation to dryness, and weighing of the
residue.

        Results;   Oil concentration in effluents varied from 5 to 1000 ppm.
Oil injected ahead of the main centrifugal pump could not be separated.  The
best effluents (5  and 7 ppm of oil) resulted when oil was added to tap water
effluent from the  main pump.

        Critical Comments;  This system is not suitable as a final oil/water
separation device  because of the poor performance.  Improved design may en-
hance system capability for treating oil dispersions and suspended solids
free of surface charges.


Title:  Electrochemical Flotation  Concept for Removing Oil from Water

Report No. and Date:  USCG  734305.2/4, January 1973

Authors:  Q. H. McKenna, H. Helber, L. M. Carrell, and R. F. Tobias

Manufacturer of Equipment:  Lockheed Aircraft Service Company, Ontario,
                            Calif. 91761

        Design Features;  The system consists of a rectangular flotation
cell, approximately 4 feet  long, constructed from glass-reinforced polyester.
The cathode is stainless steel mesh.  The anode is made from fine platinum-
10% iridium alloy wire, spot-welded to a Columbian substrate.

        Wastewater Characteristics^  Simulated bilge and ballast waters using
a combination of tap water, sea salts, and different oils were tested.  Emul-
sification involved blending the mixture with a Lightning mixer for 5 minutes
before feeding, by gravity, to a high shear pump operating at 500 psig.


                                    59

-------
 Stability  tests indicated that the mixtures contained unstable emulsions
 after  they were prepared.  Oil concentrations in the aqueous phases dropped
 to  about 100 ppm from higher initial concentrations after a few hours.

        Methods of Analysis;  Two methods were used for analyzing samples.
 One method involved extraction with solvent, followed by UV analysis.  Sol-
 vent was a mixture of the following components:  60% 2-propahol, 10% petro-
 leum ether, 20% seawater, and 10% tap water.  In the second method, pure
 ether  was  used as the extracting solvent.  Oil concentrations determined by
 both methods were similar.

        Results;  Experimental data indicate the oil concentration of all
 effluents  was lower than 20 ppm, for influents containing oil concentrations
 of  3000 to 4000 ppm.

        Critical Comments:  If developed adequately, the electrochemical
 flotation  concept can separate oil from water effectively.  The oil/water
 systems most susceptible to separation will be oil dispersions and suspended
 solids stabilized by surface charges.  Consistent effluent quality of less
 than 10 ppm of oil can be achieved.  The system was not tested thoroughly for
 the effects of variable oil concentration in the influent.  Simulated ship
 motion did not degrade system performance.  Cell volume is large and the cost
 of  chemicals and equipment is very high.  Problems encountered in the process
 are:   (1)  production of chlorine gas which dissolves in the effluent, to a
 concentration of 250 ppm—concentration is beyond the allowable discharge
 level  for  chlorine, making further treatment a necessity; (2) production of
 hydrogen gas, causing a fire hazard; and (3) temperature elevation of efflu-
 ent, due to ohmic heating of the electrodes, leading to thermal pollution and
 energy wastage.

        Cost;      10 gpm unit:   $10,000
                  100  "   "  :   $80,500
                 1000  "   "  :  $600,000


 Title:  Separation of Oil in Bilge Water by Semipermeable Membrane

 Report No. and Date:   AD-A023-289, May 1971

 Authors:   W. L. Adamson and M. W. Titus

 Manufacturer of Equipment:  Bench-scale apparatus assembled by authors

        Design Features;  Device consists of a 3-inch-inside-diameter, stain-
 less steel cylinder in which was mounted the cellulose-acetate membrane
 (Eastman Chemical type HF) .  The membrane was placed on a porous stainless
 steel  disk supported on a perforated metal disk.  Operating pressures ranged
 from 550 to 675 psig.

        Wastewater Characteristics:  The feed solution was prepared  from  dis-
 tilled water and 2190-TEP lubricating oil.  The oil/water solution was
stirred mechanically for 1 hour, in a mixing tank, and gravity-fed to a

                                     60

-------
reciprocating pump.   A magnetic stirrer located at the feed inlet  provided
further emulsification of the feed.

        Method of Analysis;  All samples were analyzed for oil content using
carbon tetrachloride extraction and infrared absorbance measurement.   Two
extractions with carbon tetrachloride were made:  25 ml of CCl^ in  the first,
and 15 ml of CC14 in the second.

        Results;  The average oil content of all effluent samples was  below
25 ppm of oil with feed containing up to 10,000 ppm of oil.  The longest run
lasted for about 15 hours.

        Critical Comments:  This test evaluation shows the feasibility of
using cellulose-acetate membranes for separating oil/water mixtures.   The
device was not tested in detail, because the feed wastewater characteristics
were different from those of bilge and ballast water or petrochemical  waste-
waters.  There was a sharp decline in permeate flow rate with increasing
operating time, indicating that fouling of the membrane may have occurred.
Repeated cleaning with trichloroethylene partially restored permeate flow
rate.
Title:  Study of Hydrophilic Membranes for Oil-Water Separation

Report No. and Date:  4305.2/7, January 1973

Authors:  C. E. Milstead and J. F. Loos

Manufacturer of Equipment:  Gulf Environmental Systems Company, P.  0.  Box
                            81608, San Diego, Calif. 81608

        Design Features:  Twenty membrane materials were evaluated  as  candi-
dates for use in ultrafiltration.  Hydrolyzed asymmetric cellulose-acetate
was selected, and tests were conducted with this membrane in a spiral-wound
configuration.

        Wastewater Characteristics;  Four different oils were used: Gulf
Harmony lubricating oil, diesel fuel, a California crude oil, and actual
bilge water from a U.S. Navy ship (U.S.S. Monti cello) that had been in harbor
for 2 weeks.  Oil/water mixtures were prepared in a feed reservoir  by  a mix-
ing pump, with tap water and oil.

        Methods of Analysis;  Oil content of samples was determined using an
extraction-gravimetric technique and Total Carbon Analysis.  The precision
of the extraction-gravimetric technique, based on data, was better  than +5%
for lubricating oil/water mixtures but large errors were encountered with
crude oil/water mixtures.

        Results;  Test results indicate the following:

        a) Oil content of effluents averaged 1.3 ppm with feeds containing
up to 50,000 ppm of lube oil, 2.4, 5 and 1.5 ppm with feeds containing 2,500


                                     61

-------
ppm of crude oil, 2,500 ppm of diesel oil, and actual bilge water,  respective-
ly.
        b) Oil content of effluent reached a maximum of 18 ppm with feeds
containing 10,000 ppm of diesel oil.

        Critical Comments;  The surface-hydrolyzed cellulose-acetate membrane
performed satisfactorily in all tests and can produce effluents that contain
less than 10 ppm of oil.  Variations of operating conditions did not affect
device performance adversely.  Membrane fouling is a problem, and an adequate
cleaning method to restore product flux is lacking.  For consistent perfor-
mance, feed should be free of suspended solids.

        Cost;  The estimated costs of two units are as follows:  100 gpm unit:
$27,000; 1000 gpm unit:  $245,000.

Title:  Ultrafiltration Concept for Separating Oil from Water

Report No. and Date:  734305.2/2, January 1973

Authors:  R. L. Goldsmith and S. Hossain

Manufacturer of Equipment:  Abcor, Inc., 341 Vassar Street, Cambridge,
                            Mass. 02139

        Design Features;  Ultrafiltration membranes tested were (1) moderate-
ly hydrophilic cellulose-acetate and (2) highly hydrophilic cellulose-acetate.
Both were studied in a tubular configuration; each tube had an internal
diameter of 1 inch and membrane pore sizes were less than 0.01 v~.  Ultrafil-
tration rates were high, generally in the range 25-150 gal/day/sq ft.

        Wastewater Characteristics:  The oils tested were No. 6 fuel oil, a
Venezuelan crude, lubricating oil, and kerosene.  Very unstable emulsions
were obtained by gravity feeding these oils into recirculated tap water.

        Methods of Analysis:  The analytical methods used were:

        a) Gravimetric—samples were acidified with sulfuric acid and oil in
           the samples was extracted with petroleum ether; extract was dried
           and the residue from the drying step was weighed.
        b) Infrared spectrophotometry following extraction with carbon.
           tetrachloride.
        c) UV spectrophotometry following extraction with carbon tetrachlo-
           ride.  Method (b) was the least sensitive.  Samples were kept
           refrigerated (35°F) for 1 to 5 days before analysis.

        Results;  The authors claim that, at oil input ratios of up to 90%,
effluents were uniformly free of visible oil and had less than 10 ppm of oil.
All effluents were reported to be completely free of turbidity and  floating
oil sheen, and were crystal-clear.  However, a very faint oil odor  and taste
were generally detectable.

        Critical Comments;  Test results from the gravimetric and infrared
spectrophotometric techniques were quite dissimilar.  Most analyses were by

                                     62

-------
the gravimetric  technique, which  is very sensitive to drying temperature.
Test results  cannot be  entirely correct.  The presence of oil odor and taste
indicate that  dissolved oil passed the membrane, and the concentration of
dissolved oil  could be  far greater than the  10 ppm reported for all efflu-
ents.  Test data were not properly reported, and further testing is necessary
before meaningful  conclusions  can be made.  Membrane fouling is a problem.


Title:  Oil-Water  Separation with Noncellulosic Ultrafiltration Systems

Report No. and Date:  Proceedings of Joint Conference on Prevention and
                      Control  of  Oil Spills, 1975, pp. 443-447

Authors:  P.  Schatzberg, L. R. Harris, C. M. Adema, D. F. Jackson, and
          C. M.  Kelly

Manufacturer  of  Equipment:  Laboratory models were assembled by authors

        Design Features;  Four different modules were tested:

        a) A  tubular module consisted of a bundle of porous carbon tubes,
           each  having  an internal diameter of 0.25 inch and a length of 40
           inches; the  membrane had an apparent pore diameter of 20 A.
        b) Hollow-fiber modules,  in two configurations, were used.  One con-
           sisted  of a  bundle  of  hollow fibers having an internal diameter
           of  0.017 inch and a length of 24 inches; this system could be
           backflushed, and total effective membrane surface area was 25 sq
           ft; the membrane had an apparent pore diameter of 38 A and a
           nominal molecular weight cutoff of 10,000.  In the second config-
           uration, the internal  diameter of fibers was 0.020 inch and the
           effective membrane  area of the bundle was 30 sq ft; the nominal
           molecular weight cutoff was 80,000 and the apparent pore diameter
           was 100 A.
        c) The spiral-wound module consisted of a series of membrane sheets,
           separated by corrugated spacers and combined in a spiral-wound
           cylindrical  shape;  total membrane area was 35 sq ft and pore
           diameter was approximately 50 A.  The membrane's nominal molecular
           weight  cutoff was 5,000 to 10,000.
        d) A plate and  frame configuration consisted of a series of membrane
           sheets  separated by fine fiber cloths; the membrane had an effec-
           tive  area gf 5 sq ft and an apparent pore diameter of approxi-
           mately  40 A.  The membrane's nominal molecular weight cutoff was
           100,000.

        Wastewater Characteristics;  Two types of oil-in-water emulsions
were prepared, using a  lubricating oil and fresh water.  Initially a high-
speed blender  was  used  to mix  the oil and water; later, stabler emulsions
were prepared  using an  ultrasonic mixer.

        Method of Analysis;  Oil-in-^water analyses were made by infrared
spectrophotometry, following carbon tetrachloride extraction.


                                     63

-------
         Results;   Each  of the non-celluloslc memoranes investigated demon-
 strated a capacity to separate emulsified and suspended oil from water.  All
 but the tubular modules consistently produced effluents containing less than
 15 ppm of oil.

         Critical  Comments;  All modules tested gave satisfactory performance.
 The major problem encountered was a decline of flux rate as testing pro-
 gressed.   Therefore, permeate flushing and backwashing with detergent were
 necessary.  However, complete recovery tff flux rate could not be obtained
 with any of the non-cellulosic membranes.  Despite the good separation re-
 ported,  it is  doubtful  if these devices are capable of separating dissolved
 oil or emulsified oil much smaller than the membrane pore diameter.


 Title:  Development of  a Coalescing Type Oil/Water Separator for Marine
         Service

 Eeport No. and Date:  Aqua-Chem Technical Report presented at SNAME Meeting,
                      San Diego, Calif., February 18, 1970

 Authors:   Lee  J.  Hartenstein and Thomas E. Lindemuth

 Manufacturer of Equipment:  Aqua-Chem, Inc., Waukesha, Wis.

         Design Features;  Device has three stages.  The first stage is a
 screen;  three  cartridge-type coalescer elements are present in the second
 and third stages.   Removal of oil in all stages is controlled by capacitance-
 type oil/water interface  detectors.  Each chamber is equipped with electrical
 heaters  to reduce oil viscosity and to drive off small amounts of oil
 entrained in the  separated water phase.

        Wastewater Characteristics;  Oil is fed into circulating fresh water
 (occasionally  seawater)  at the suction or discharge port of a centrifugal
 pump to produce differing emulsified oil/water mixtures.

        Method  of Analysis;  Technique(s) used in analyzing samples for oil
 content not stated.

        Results;   Data were presented graphically and indicate effluent oil
 concentrations  were below 80 ppm, with influents containing 10% oil.

        Critical  Compents:  Since analytical methods were not stated, data
 reported  lacked credibility; the operating temperature is high and may alter
 solubility of oil  in the  treated aqueous phase.


 Title:  Oily Water Separator:  Liquid-Liquid Separation by a Commercial
        Self-cleaning Edge Filter

 Report No.  and  Date:  COM-71-01095, January 19, 1971

Author:  J. R.  Hefler


                                    64

-------
Manufacturer of Equipment:  AMF Beaird, Inc., Uncasville, Conn.  06382

        Design Features;  The system consists of a tank with internal baffles
and tangential inlets.  Cartridge-type filter elements having spacings of
0.0015 inch were designed by Cuno Engineering Corporation.  Filter elements
can be cleaned continuously by cartridge rotation.  Automatic oil detection
probes provide recirculation of effluent for reprocessing, if oil concentra-
tion is high.

        Wastewater Characteristics:  Seawater and three grades of Bunker C
oil, of different densities, were used as test fluids.  The oils were injec-
ted upstream and downstream of a 3450-rpm centrifugal pump, into seawater,
to provide the wastewater feed.

        Method of Analysis;  Samples are mixed with "sufficient" chloroform
and evaporated to dryness at 75°C; the weight fraction of oil present in a
sample is determined.

        Results;  Effluents contained as much as 2000 ppm oil.  Separation
efficiency was poor.  Oil with specific gravity close to that of seawater
could not be separated.

        Critical Comments:  Device was poorly designed and separation was not
possible when oil was injected ahead of the pump.  Oil dispersions with drop-
let sizes less than 0.002-inch diameter were not separated.  Test results
cannot be correlated, due to the poor experimental techniques used.
Title:  Experimental Evaluation of Fibrous Bed Coalescers for Separating
        Oil-Water Emulsions

Report No. and Date:  EPA Project No. 12050DRC, November 1971

Authors:  W. M. Langdon and D. T. Wasan

Manufacturer of Equipment:  Illinois Institute of Technology, Chicago,  111.
                            60616

        Design Features;  Device is a 1-sq ft coalescer unit built into a
filter press framework.  The coalescer unit is made of fiberglass filaments,
3.2 y in diameter, coated with isobutyl methacrylate resin for stability.

        Wastewater Characteristics:  Tap water and a mixture of 50% kerosene
and 50% pollutant material from treated hot mill cooling water (similar to
No. 30 lube oil) were agitated and recycled through a 3450-rpm centrifugal
pump for 1 hour.  Primary and secondary dispersions of the oils in water
were produced.

        Methods of Analysis;  Oil concentrations in samples were determined
by light transmission and Hach turbidimeter measurements.  Tabulated data,
on samples analyzed by both methods, were not converted to parts per million
for comparison.

                                     65

-------
        Results:  Oil-separation efficiency varied from 70 to 99%, at influ-
 ent  oil concentrations of 50 to 500 ppm.  Higher efficiences were reported
 for  fibers  coated with resins than for uncoated fibers.

        Critical Comments:  The analytical methods used do not detect dis-
 solved oil, which may have been appreciable.  Therefore, oil concentrations
 in effluents  are higher than reported.  Abrupt increases of oil in effluent
 occurred  if runs were not continuous; feed pretreatment is required to con-
 trol the  large  pressure drops encountered and to prolong fiber life.

        Cost;   Equipment cost was not specified; operating costs are esti-
 mated at  $0.13/103/gal for single-fiber use and $0.01/103/gal if fibers can
 be regenerated.


 Title:  Oily Water Separation System

 Report No.  and  Date:  COM-72-10561, January 1972

 Author:   R. J.  Skocypec

 Manufacturer of Equipment:  Esso Research and Engineering Department,,
                            Linden, N. J. 07036

        Design  Features:  Separator consists of a coalescing element upstream
 of a settling chamber.  There is an AMF-Cuno Super Auto-Klean Filter upstream
 of the device.

        Wastewater Characteristics;  Shoreside tests were performed with
 ballast water discharged from ships.

        Method of Analysis;  All samples were analyzed for oil content using
 an infrared absorption technique.

        Results;  Average oil-removal efficiencies reported range from less
 than 5 to nearly 90%.

        Critical Comments;  The coalescer was easily plugged and device was
 not  tested  sufficiently.  Performance was poor and the test procedure was
 inadequate.


 Title:  Test and Evaluation of a 50-Gallon-per-Minute Oil/Water Separator

 Report No.  and Date:  AD 785-223, July 1972

Author:   E.  C. Russell

Manufacturer of Equipment:   Separations and Recovery Systems, Inc.,
                             Santa Ana, Calif. 92705

        Design Features:   The SRS separator consists of two, skid-mounted,

                                    66

-------
 16-inch-in-diameter by 47-inch-long high-pressure vessels,  connected  in
 series.   Each vessel contains three 6-inch-in-diameter and  22-inch-long
 coalescer elements, mounted in parallel.   The device is equipped with  a supply
 pump  and a capacitance-type probe that controls the automatic  oil-discharge
 cycle of the system.

         Wastewater Characteristics;  Various quantities of  differing  oils
 were  metered into recirculating fresh water, at the suction side of a pump,
 to produce an emulsified oil/water mixture.  In some test runs, dry sand was
 added to the oil/water mixture.

         Methods of Analysis;  Three methods were used for determining the oil
 content  of samples:  (1) turbidity, (2)  visual inspection,  and (3) CC14
 extraction and infrared spectrophotometry.

         Results;  Data from the turbidity meter measurements indicate oil
 concentrations in effluents were below 60 ppm, but values reported for
 infrared analyses were as high as 141 ppm.

         Critical Comments;  The objective of these tests was no visible
 "sheen"  in discharged waters, which device  easily achieved.  Analytical
 results  from the three methods were not similar.
 Title:  Test and Evaluation of Oil-Water Separation Systems

 Report No. and Date:  CR 73.015, November 8,  1972

 Author:  A. V. Sims

 Manufacturer of Equipment:  The devices tested are (1)  Fram  Corporation
                   separator and (2)  Separation and Recovery  Systems separator

         Design Features:  The Fram separator  consists of  a preconditioning
 unit (filter cartridge with 25-y pore size),  a gravity  separation stage, and
 a coalescing stage (5-P pore size filter cartridge).  Device is equipped
 with an interface controller and a manually operated oil  valve.  The S"RS
 separator has a prefilter (25-u pore size) and two stages of coalescing
 filters (5-y pore size).; device is equipped with an interface controller,
 and oil and water control valves.

         Wastewater Characteristics;   Various  quantities of Navy Special Fuel
 Oil were intimately mixed with seawater by a  centrifugal  pump.  Feed mixtures
 contained mechanically emulsified oil.

         Method of Analysis;  Oil in samples was extracted by chloroform
((three separate extractions) and analyzed by  infrared spectrophotometry.
 The report states:  "At concentrations less than 10 ppm,  the average analyti-
 cal accuracy was about 20%."

         Results:  The average oil content of  effluents  from the Fram separa-
 tor was less than 3 ppm, with influents containing 200  ppm oil.  The average

                                      67

-------
oil content of effluents from the SRS separator was less than 1 ppm, 
-------
Authors:  S. M. Finger and T. S. Yu

Manufacturer of Equipment:  Laboratory prototype

        Design Features;  System is a three-stage separation device:  the
first stage is a cylindrical gravity separator; the second stage contains a
prefilter; the third stage is the coalescer filter.  Coalescer elements used
were of the cylindrical cartridge type, made of resin-coated fiberglass
covered with cotton socks.  Four-inch-diameter elements were positioned hori-
zontally in a glass chamber, allowing for visual inspection of element
performance.

        Wastewater Characteristics;  The influent oil/water mixture to the
system was prepared by metering oil and water through a centrifugal pump.
Severn River water was used, and the oils were Navy Distillate Fuel Oil and
MS-2190-TEP lubricating oil; the shearing action of the pump effectively
mixed the oil and water.

        Method of Analysis;  Samples were analyzed for oil content by extrac-
tion with carbon tetrachloride followed by infrared absorbance measurements.

        Results;  Under most conditions studied, effluent water contained
less than 15 ppm oil.  Only 10% of the samples analyzed contained oil more
than this amount.

        Critical Comments;  System was adequately tested for the effects of
several parameters on performance.  Report is well written; however,
stability analyses of feed streams should have been performed.


Title:  Technologies for Shipboard Oil Pollution Abatement:  Effects of
        Operational Parameters on Coalescence

Report No. and Date:  3598, August 1972

Authors:  S. M. Finger and T. S. Yu

Manufacturer of Equipment:  Naval Ship Research and Developmental Center,
                            Bethesda, Md. 20034

        Design Features;  System is a three-stage separator; cartridge-type
coalescer elements were made of resin-coated glass fibers.  The first stage
is empty, acting as a conventional gravity separator; the second stage is a
prefilter; and the third stage contains the coalescer elements.

        Wastewater Characteristics^  Oils were metered into recirculating
river water at the suction side of a centrifugal pump, producing mechanically
emulsified oil/water mixtures.

        Method of Analysis;  An infrared spectrophotometric analysis devel-
oped at the Naval Shipyard Laboratory (methodology was not stated) was used
to detect oil content of samples.

                                     69

-------
         Results:  Authors state that 60 out of 66 samples analyzed for oil
 content contained less  than 15 ppm oil; influent oil concentrations were not
 stated.

         Critical Comments;  Test objective was to determine the effects of
 some operational parameters on a coalescence device; the report states that
 dirt, silt,  and highly  viscous oils can clog filter elements.  The credibil-
 ity of test  results  cannot be confirmed because the analytical technique used
 was not stated in the report.


 Title:  Test and Evaluation of Oil Pollution Abatement Devices for Shipboard
         Use:   Phase  II

 Report No. and Date:  AD 762-499, September 1972

 Author:   L.  B. Norton

 Manufacturer of Equipment:  Not stated

         Design Features;  Three different separators were tested:
         Separator A:  a two-stage vertical filter/coalescer unit, containing
                      automatic oil-discharge control valves; a prefilter was
                      not provided.
         Separator B:  a larger version of separator A, but contains a
                      prefilter.
         Separator C:  a three-stage filter/ coalescer/gravity unit; this
                      device has automatic and manual oil—discharge valves

         Wastewater Characteristics:  Oil and solids were injected in fresh
 water circulating through three different pumps:  centrifugal, vane, and
 double diaphragm.  Oil/water mixtures produced contained emulsified oil.

         Methods of Analysis;  The oil content of samples was  determined by
 two methods:   (1) turbidity and (2) visual observation.

         Results:  Most  effluents had "no visible sheen."  Occasionally sep-
 aration  performance was poor; substitution of vane pumps for the system's
 centrifugal pumps did not improve performance.  Water content of separated
 oils was high.

         Critical Comments;   Analytical methods did not give a true indica-
 tion of  the oil content of samples.  Tests were not complete and the report
 is poorly organized.


Title:  Test and Evaluation of Oil Pollution Abatement Devices for Shipboard
        Use:   Phase III—Final

Report No. and Date:  AD 762-488, January 1973
                                    70

-------
Author:  L. B. Norton

Manufacturer of Equipment:  Omitted

        Design Features;  Device is a two-staged unit, consisting of a pre-
filter made of 10-y-pore-size elements and 10 pleated paper cartridges, and
a second stage containing 5 cartridge filter elements.  Equipment has oil/
water interface probes and discharge valves.

        Wastewater Characteristics:  Device was installed on a U.S. Coast •
Guard cutter to process  feed streams of actual bilge and ballast waters.

        Methods of Analysis;  Samples were analyzed by (1) microbalance and
(2) Total Organic Carbon measurements.

        Results;  Data were not stated.  The average oil content of the ef-
fluent, in one run, was  reported to be 28 ppm; oily effluents generally con-
tained some entrained water.  The author claims that overall separator
performance was acceptable.

        Critical Comments;  The report is badly written.  The device was  not
tested sufficiently; it  is impossible to draw conclusions from reported data.
Analytical techniques are vague.  Device was susceptible to dirt loading.


Title:  Separation of Oil Dispersions from Water by Fibrous Bed Coalescers

Report No. and Date:  Environ. Sci. Technol., 6^ No. 10, 907 (October 1972)

Authors:  W. M. Langdon, P. P. Naik, and D. T. Wasan

Manufacturer of Equipment:  Illinois Institute of Technology, Chicago, 111.
                            60616

        Design Features;  Device is constructed of aluminum and consists  of
two types of fibers (Owens Corning glass fiber having 3.2-u diameter and
Owens Corning Aerocor coarse glass fiber of 10.1-y diameter).  Glass fiber
mats are clamped between perforated plates; 100-mesh Monel screens, coated
with TFE, precede the coalescer/filter elements.

        Wastewater Characteristics;  Standardized oil/water mixtures were
prepared by metering oil into tap water, circulating through a ring pump
operating at 3450 rpm.   Pollutant oil from a skimming tank at Interlake Steel
Corporation and Interlake Steel Corporation No. 30 grade automotive lubrica-
tion oil were used.

        Method of Analysis;  Oil concentrations in influent and effluent
streams were determined by light transmission.

        Results;  Oil content of effluents varied from 7 ppm to generally
non-detectable, with influent oil concentrations of 50 to 500 ppm.


                                     71

-------
         Critical Comments;  It is doubtful that light transmission can detect
 oil  in  the  range 0-50 ppm with accuracy.  Since this method was used in
 analyzing product samples, the reported efficiency of separation is probably
 erroneous.
 Title:   RFC Division, Midland-Ross Corporation 10-Gallon-per-Minute Liquid/
         Liquid  Separator

 Report No. and  Date:  2058, May 1973

 Author:  E. C.  Russell

 Manufacturer of Equipment:  RFC Division, Midland-Moss Corporation, Roxboro,
                            N. C. 27573

         Design  Features;  This is a three-stage coalescence-type device.  A
 cylindrical,  horizontal tank, 70 inches long and 12 inches in diameter
 houses the coalescer elements.  The first-stage element is 12 inches long and
 10  inches in diameter;  in the second and third stages, cartridge elements are
 6 inches long and 10 inches in diameter.

         Wastewater Characteristics;  Quantities of differing oils were
 metered  into circulating, prefiltered fresh water, at the suction side of a
 supply pump, to produce mechanically emulsified oil/water mixtures.

         Methods of Analysis:  Four methods were used for analyzing samples:
 (1)  on-line ultraviolet energy absorption, (2) on-line fluorescence detec-
 tion, (3) turbidity, and (4) Total Organic Carbon (TOC) analysis.  Oil con-
 centrations determined by the different methods were not similar.

         Results;  Data  from TOC analysis indicate oil concentrations in ef-
 fluents  were below 10 ppm, with influents containing up to 32% oil.  Other
 analytical methods gave much higher values (200 ppm), with influents
 containing 3% oil.

         Critical Comments:  Pressure drops in this equipment were high after
 short operating times, indicating the device can be easily plugged. -The
 objective of the test evaluation, i.e. "no visible sheen" in effluents, was
 met  despite the dissimilarity in the data reported for the various analytical
 methods.


 Title:   Fram Corporation Model OWS-23-FCI-USCG Oil/Water Separator

 Report No. and  Date:  2059, May 1973

Author:   E.  C.  Russell

Manufacturer of Equipment:  Fram Corporation, Tulsa, Okla. 74160
                                     72

-------
         Design  Features;   A multistage,  skid-mounted device equipped with a
 double-diaphragm pneumatic supply pump.   The first  stage  is a preconditioner;
 the second stage is  an inclined-plate, gravity separator;  the last stage con-
 tains  a cartridge-type coalescer element.

         Wastewater Characteristics;   Test fluid was prefiltered tap water.
 Various oils were used to  produce oil/water mixtures.  The shearing action of
 a pump emulsified the mixture.

         Methods of Analysis; Three  methods were  used for analyzing samples:
 (1) on-line fluorescence detection,  (2)  turbidity,  and (3) Total Organic
 Carbon (TOG) analyses. Oil concentrations  determined by the different
 methods were not similar.

         Results; TOC analyses  indicate  an  average  oil concentration of 25
 ppm in effluents, with influents having  up  to  9%  oil.  Concentrations
 obtained by on-line  fluorescence detection  were higher.

         Critical Comments;  Device was not  thoroughly  tested; however, per-
 formance was poor.   System is not capable of treating  slightly stable
 emulsions.
 Title:   Separation and Recovery  Systems,  Inc. 100-Gallon-per-Minute Oil/Water
         Separator

 Report No.  and  Date:   2060, May  1973

 Author:  E.  C.  Russell                                                ,

 Manufacturer of Equipment:  Separation and Recovery Systems, Inc., Santa Ana,
                            Calif. 92705

         Design  Features;  Device is comprised of the following:  prefilter;
 two skid-mounted  (20-inch diameter by 60-inch diameter by 60 inches long)
 high-pressure vessels, connected in series; high-shear supply pump; control
 valves,  etc.  Each vessel contains five (6-inch diameter by 22 inches long)
 cartridge-type  filter  elements,  mounted in parallel.

         Wastewater Characteristics;  Various types of contaminant oils were
 mixed with prefiltered fresh water and synthetic seawater to produce oil/
 water mixtures.   Dry fine sand was added  occasionally.  The shear of the
 supply pump  emulsified the mixtures.

         Methods of Analysis;  Two methods were used for analyzing samples:
 (1) on-line  turbidity  measurements and (2) crystal microbalance.

         Results;   Data from turbidity measurements indicated acceptable per-
 formance, but concentrations reported from the crystal microbalance technique
were much higher;  oil  concentrations in effluents averaged 35 ppm.
                                    73

-------
        Critical Comments:  Device performance met the test objective of "no
visible sheen" in discharge waters.  The analytical techniques  used provide
only rough estimates of oil content in effluent samples tested,  and
performance efficiency cannot be determined accurately.


Title:  Oil/Water Pollution Program:   Phase I

Report No. and Date:  NAPTC-PE-27, July 1973

Authors:  A. P. Pontello, F. G. Woessner, and R.  J.  Delfosse

Manufacturer of Equipment:  Velcon Corporation

        Design Features;  System consists of an experimental coalescer/filter
element, approximately 20 inches long and 3.75 inches in diameter (Velcon
TE 3-27), installed vertically in a cylindrical vessel.

        Wastewater Characteristics;  Navy Distillate Fuel Oil was injected
into circulating tap water at the suction end of a vane pump.  Primary and
secondary dispersions (emulsions) were produced by pump shearing forces.

        Methods of Analysis;  Oil concentrations in samples were measured by
two methods:  (1) turbidimetry and (2) carbon tetrachloride extractions  and
infrared spectrophotometry.

        Results;  Data were not tabulated.  The authors claim that 4 ppm of
oil was present in the effluent, when the influent contained 100 ppm oil.

        Critical Comments;  Stability tests were not performed on wastewater;
the oily wastewater used was not representative of bilge or ballast water.
Tests conducted on the device were limited; throughput was small because of
excessive differential pressure on the system.


Title:  Oil/Water Pollution Program:   Phase II

Report No. and Data:  AD-A009-093, October 1974

Authors:  A. P. Pontello, C. J. Collick, J. J. Palmer, and A. J. Rollo

Manufacturer of Equipment:  Several manufacturers of coalescer elements

        Design Features;  Details of the different coalescer elements are as
follows:
        a) FRAM PC-11:  element is 14.5 inches long and 3.5 inches in diam-
           eter; made of fiberglass, pleated paper, and perforated screen
           frames;  flow is inside-out.
        b) Velcon Corporation:   coalescer element is 20 inches long and 1.25
           inches in diameter;  consists of various layers of fiberglass sand-
           wiched between a metal screen and encased in a cylindrical vessel


                                     74

-------
           made of synthetic material.
        c) Keene Corporation:  element is 20 inches long and 1.25 inches in
           diameter; made of variable density fiberglass layers, plastic-
           coated mesh screen, a pleated paper core, and a perforated metal
           screen.
        d) SRS:  element has dimensions similar to the Keene model but
           consists of fiberglass, cloth, and two types of plastic-coated
           mesh.
        e) Bendix:  element has the same dimensions as the Keene coalescer
           but consists of two fiberglass materials of different density.

        Wastewater Characteristics^  Navy Distillate Fuel Oil was injected
at the suction side of a centrifugal pump circulating tap water; the result-
ing emulsified oil/water mixture was used as test fluid.  Detergents were
used, also.

        Methods of Analysis;  Four methods were used to analyze samples:  (1)
samples were rated by visual inspection (method is subjective), (2) turbidity
measurements, (3) CC14 extraction and infrared spectrophotometry, and (4)
estimation of sheen on water surfaces (sheen index).

        Results;  Oil concentration in influents was varied from 100 to
50,000 ppm; the testing process was not systematic, so that data obtained
on various coalescers could not be compared.
        Velcon:  oil concentration in effluents averaged 5 ppm; free water
                 was present in coalesced oil and the quantity of free water
                 increased with throughput; differential pressure in the
                 system reached 28.7 psi after 25 hours of operation.
        Keene:   oil concentration in effluents averaged 4 ppm, without
                 detergents; performance was unsatisfactory when detergent
                 was present in the wastewater.
        Fram:    effluents contained 2 ppm oil, when coalescer was operated
                 with a prefilter; without a prefilter, oil concentration in
                 effluents reached 17 ppm.
        SRS:     in the absence of detergents, effluents contained an average
                 of 2 ppm oil; with detergents, oil concentration in efflu-
                 ents rose to 6.5 ppm.  Free water (20 ppm) was present in
                 coalesced oil.

        Critical Comments;  Stability tests performed on the process streams
indicate that the oil/water mixtures treated were very unstable; separation
of oil (when the mixtures were left undisturbed) was 40 to 80% complete.
Test procedures were not varied, and some of the data obtained were
meaningless.  Analytical methods are suspect.


Title:  Development of a Batchwise In-Situ Regeneration-Type Separator to
        Remove Oil from Oil-Water Suspensions

Report No. and Date:  Technical Report 7080-3, December 1974
                                     75

-------
 Author:  D. H. Fruttlan

 Manufacturer  of Equipment:   Hydronautics, Inc., Laurel, Md. 20810

         Design Features;  System  (HOWS Model 0-600) is a completely self-
 contained,  fully  automated module, capable of treating up to 600 gpm of oily
 wastewater.   Device  consists of a three-compartment chamber, made up of an
 upstream header,  a middle filtration section containing the filtering mate-
 rial between  two  perforated plates (one fixed and the other free to move),
 and a downstream  decanter header  containing an inclined plate.  Filter mate-
 rial is  a  thick,  open-reticulated, oleophilic foam which is regenerated by
 squeezing with the movable perforated plate.

         Wastewater Characteristics:  Tests were conducted at the Navy Fuel
 Reclamation Plant in Virginia: influents were taken from the middle chamber
 of a primary  gravity separator and from storage tanks used for ballast and
 bilge waters  discharged  from Navy ships.

         Method of Analysis;  Samples were analyzed by light transmission
 measurements. Some  samples were  analyzed for oil content by the Naval
 Systems  Research  and Development Laboratory, Annapolis; technique was not
 stated.

         Results;   Oil content of  separated water was high; oil-removal effi-
 ciency was  poor.   Oil content of effluents was generally higher than 40 ppm,
 with influents containing up to 300 ppm oil.

         Critical  Comments;  System is suitable for gross oil/water separa-
 tion; efficiency  is  comparable to that of ordinary gravity separators.  Aging
 of foam  leads to  poor performance; high cost of system may discourage use.


 Title:   Coalescence  of Emulsified Oily Wastewater by Fibrous Beds

 Report No.  and Date:  Presented at the 30th Annual Purdue Industrial Waste
                      Conference, Purdue University, Lafayette, Ind.,
                      May 6, 1975

 Authors:  J.-N. Chieu, E. F. Gloyna, and R. S. -Schechter

 Manufacturer  of Equipment:  Laboratory prototype

         Design Features:  System is a 2-foot-long Plexiglas cylindrical
 column, housing coalescing media.  Three types of media were used:  polyester
 felt, polypropylene  felt, and glass mats.

        Wastewater Characteristics;  Influents were prepared by emulsifying
 oil and tap water in a household blender, stabilizing the mixture in an
 ultrasonic disrupter, and stirring continuously before use.  Oils tested
 include refinery slop oil (coker slop oil and API skimmings) and No. 2
heating oil;  the slop oils were filtered before use.
                                     76

-------
        Method  of  Analysis;  Oil  content  of  all  samples was measured using a
Beckman Total Carbon  Analyzer.  Prior to  analysis, each sample was homoge-
nized  ultrasonically  to  insure  representative  sampling.

        Results;   With influent containing 100 ppm oil, the following effi-
ciences were achieved:

        Glass mats          60%
        Polypropylene felt   80%
        Polyester  felt       90%

Higher efficiences were  reached at  low flow  rates.

        Critical Comments:   Fibers  preferentially wetted by the dispersed
phase  favor coalescence  and  exhibit lower head loss.  Oil-removal efficiency
increases  about 10-15% after minimum oil  saturation.  Oil content of
effluents  is higher than the desired limit.


Title:  Centrifuge Coalescer for  Separating  Oil  from Water in Shipboard
        Applications

Report No.  and  Date:  AD-764-006, February 1973

Authors:   A. C. Harvey,  A. R. Guzdar,  V.  K.  Stokes, and A. T. Fisk

Manufacturer of Equipment:   Foster-Miller Associates, Waltham, Mass. 02154

        Design  Features^  System  consists of a three-stage rotor comprising a
primary stage,  a swept vane  separator to  separate large drops; a closely
spaced, axial-plate,  spiral-wrap  coalescer to  coalesce the small drops; and
a secondary-stage  swept  vane separator to collect and separate coalesced
oil.   The  conical  ends of the rotor contain  blades and passageways that act
as  centrifugal  pump and  centripetal turbine  elements, at the inlet and
outlet ends, respectively.

        Wastewater Characteristics:   Oil/water mixtures of No. 2 and No. 4
fuel oils,  Nigerian crude, and  detergents were emulsified by passage through
a centrifugal pump; oil  emulsions had sizes  ranging from 2 to 100 p.  Sta-
bility analyses indicated that  mixtures were quite stable; after 120 hours,
remaining  oil droplets had diameters of 2 to 15  v-

        Methods of Analysis; Samples  were analyzed for oil content by two
methods:   (1) integration of the  drop-size distribution measured by a Coulter
counter and (2) infrared spectrophotometry.  Both measurement techniques gave
about  75 to 80% of the "true" oil concentration  in samples measured; concen-
trations measured  by  the  Coulter  counter  were  approximately 90% of the values
obtained by infrared  analysis.

        Results;   Test data  indicated good separation with an average of
approximately 100  ppm in effluents,  with  influents containing greater than
1%  oil.  Overall oil-removal efficiency was  greater than 90%.  Performance
worsened when influents  contained detergents.

                                     77

-------
         Critical  Comments;  Device was adequately tested.  Test procedure is
 satisfactory,  and if  improvements in design can be made, device is capable
 of  achieving higher oil-removal efficiencies.  Equipment is expensive.
 Effects  of ship motion  during processing were not investigated.
         Cost:  100 gpm  unit $22,900

 Title:   Development of  a Centrifugal System for Separation of Oil and Solids
         from Shipboard  Discharge Water

 Report No.  and Date:  CG-D-118-75, July 1975

 Authors:  A. R. Guzdar, A. C. Harvey, J. Potter, and W. M. Mack

 Manufacturer of Equipment:  Foster-Miller Associates, Inc., 135 Second
                            Avenue, Waltham, Mass. 02154

         Design Features:  Device is an oil/solids/water separating system
 made up  of the following components:  (1) a cleanable bag strainer to retain
 coarse and fibrous solids, (2) a centrifugal pump to separate settleable
 solids,  (3)  a  coalescing centrifuge to separate well-dispersed oil, and (4)
 an  oil/water monitor  to continuously measure and record the oil content of
 effluent.

         Wastewater Characteristics;  Oil/water mixtures were emulsified by
 passage  through a centrifugal pump operating at 30 psi pressure differential.
 Lube oil and No.  2 fuel oil were used.  The oil concentration of influent
 streams  varied from 5 to 100%.

         Method of Analysis;  An on-line oil/water monitor continuously mea-
 sured and recorded the  oil content of effluents.  Analytical results compared
 favorably with results  using EPA solvent extraction-spectrophotometric
 techniques.

         Results;   Laboratory testing of individual components showed favor-
 able performance.  However, the performance of the overall system was poor.
 Oil content  of effluents was much higher than the expected upper limit of
 15  ppm.

         Critical  Comments:  The poor system performance was blamed on "cer-
 tain operational  difficulties" experienced during the testing program.
 Design modifications  may improve performance.
Title:  Bimetallic Coalescers:  Electrophoretic Coalescence of Emulsions in
        Beds of Mixed Metal Granules

Report No. and Date:  Environ. Sci. Technol., 4_, No. 6, 510-514  (1970)

Authors:  F. M. Fowkes, F. W. Anderson, and J. E. Berger

Manufacturer of Equipment:  A laboratory prototype assembled by  authors
                                     78

-------
        Design Features:  Device is a cylindrical column packed with carbon
(4-8-mesh size) and aluminum  (20-mesh size) granules.

        Wastewater Characteristics:  Influents contained oil-in-water emul-
sions.  Methods of preparation of feed streams were not stated.

        Method of Analysis:   Influent and effluent samples were analyzed for
oil content by light transmission.

        Results:  Light transmission data indicate some separation occurred,
with influents containing as  much as 1100 ppm oil; coalescence was faster
as influents became more dilute.

        Critical Comments:  Performance is poor and device is not adequate  as
a polishing stage in a treatment process.  Consumption of metals is a
problem.  Further treatment is a necessity.


Title:  Oil Removal by Carbon-Metal Granular Beds

Report No. and Date:  J. Water Poll. Control Fed., 47, No. 8, 2101-2113
                      (1975)

Authors:  M. M. Ghosh and W.  P. Brown

Manufacturer of Equipment:  Bench-scale coalescer assembled by authors

        Design Features:  Device consists of a glass column packed with a
mixture of carbon and metal granules, supported by a 40-mesh metal screen.
Activated cocoanut charcoal and aluminum or iron were used .as bed materials.

        Wastewater Characteristics;  Stable emulsions of oil droplets in
water were obtained when small volumes of solutions of silicone oil in ace-
tone were jetted into water,  through a small orifice.   In some experiments,
a homogenizer was used, at speeds of 12,000 rpm for 12 minutes, in preparing
oil-in-water emulsions.  Dispersions were less than 1.0 y in diameter and
did not exhibit any self-coalescence over long periods of time.

        Methods of Analysis;  Influent and effluent samples were analyzed
for number and size distribution of oil droplets by a particle counter.
Spectrophotometric analyses of some samples were carried out.

        Results;  The highest oil-removal efficiency achieved, with influents
containing approximately 350  ppm oil, was 82%.

        Critical Comments;  Analytical methods are only accurate to within
50% of actual value.  Throughput is small and head loss gradually increased
during processing.  Formation of metal hydroxides that dissolve in effluent
is a disadvantage.  The carbon-aluminum system performed better than the
carbon-iron system, because of the higher potential difference of the bime-
tallic couple.  Bimetallic coalescers can be used only as polishing devices


                                    79

-------
in oil/water separation, and are useful for separating oil dispersions
stabilized by surface charges.


Title:  The Coanda-Effect Oil-Water Separator:  A Feasibility Study

Report No. and Date:  AD-774-080, February 1974 (NTIS)

Author:  D. Pal

Manufacturer of Equipment:  Experimental model designed by Civil Engineering
                            Laboratory, Port Hueneme, Calif. 93043

        Design Features;  Device consists of inlet and outlet ports, 12-
inch-long attachment wall, oil-collection chambers, and oil/water interface
detection probes.  Housing was made of Plexiglas to allow visual observa-
tion.

        Wastewater Characteristics;  The oil/water mixture used as influent
was prepared by mixing hydraulic oil and tap water.

        Method of Analysis:  Volumes of oil and water present in settled
samples were measured.

        Results;  Test data show that about 50% of the oil present in the
influent was separated; the separated oil contained up to 5% free water.

        Critical Comments;  The Coanda-effeet separator is in a develop-
mental stage; it is useful only for gross separations.  Turbulence created
by jet flow will enhance emulsification of oil in wastewater, making
separation very difficult.
                                   80

-------
                                SECTION 8

         SELECTED MANUFACTURERS OF OIL/WATER SEPARATING EQUIPMENT

Gravity-Differential Separators

        Aerodyne Development Corporation, Cleveland Ohio
        AFL Industries, West Chicago, Illinois
        Aqua-Chem, Inc., Waukesha, Wisconsin
        Butterworth Systems, Inc., Bayonne, New Jersey
        C. E. NATCO, Tulsa, Oklahoma
        Chiyoda Chemical Engineering & Construction Company, Ltd., Tokyo
        De Laval Separator Company, Poughkeepsie,  New York
        Envirex, Inc., Waukesha, Wisconsin
        FMC Corporation, Lansdale, Pennsylvania
        Fram-Akers Corporation, .Tulsa, Oklahoma
        FWI, Pollution Control Division, Tulsa, Oklahoma
        General Electric Corporation, Philadelphia, Pennsylvania
        Heil Process Equipment Corporation, Cleveland, Ohio
        Inland Environmental, Chicago, Illinois
        MAPCO, Inc., Tulsa, Oklahoma
        Midland-Ross, Roxboro, North Carolina
        The Permutit Company, Paramus, New Jersey
        Pielkenroad Separator Company, Houston, Texas
        Separator & Recovery Systems, Inc., Santa  Ana, California
        Smith Industries, Inc., Houston, Texas

Flotation Equipment

        Ecodyne Corporation,  Union, New Jersey
        Envirex, Inc., Waukesha, Wisconsin
        FWI, Pollution Control Division, Tulsa, Oklahoma
        The Galigher Company, Salt Lake City, Utah
        Joy Manufacturing Company, Denver, Colorado
        Lockheed Aircraft Service Company, Ontario, Canada
        Mechanics Research, Inc., Los Angeles, California
        The Permutit Company, Paramus, New Jersey
        Petrolite Corporation, Tretolite Division, St. Louis,  Missouri

Rotational Equipment

        Air Research Manufacturing Company, Torrance, California
        Ametek, Inc., East Moline, Illinois
        Centrico, Inc., Northvale, New Jersey
        De Laval Separator Company, Poughkeepsie,  New York
        Foster-Miller Associates, Waltham, Massachusetts

                                    81

-------
        Pennwalt Corporation, Warminster, Pennsylvania
        Reynolds Submarine Service Company, Richmond, Virginia
        Sharpies Division, Pennwalt Corporation, North White Plains,  New York
        United Aircraft Corporation, East Hartford,  Connecticut

FiltratiQn Equipment (granular media)

        Combustion Engineering Company, East Hartford, Connecticut
        De Laval Separator Company, Poughkeepsie, New York
        Hayward Filter Company, Santa Ana, California
        Neptune Micro-Floe, Inc., Corvallis, Oregon
        Peabody Welles, Roscoe, Illinois
        Smith Industries, Inc., Houston, Texas

Filter/Coales cers

        Aqua-Chem, Inc., Waukesha, Wisconsin
        Fram-Akers Corporation, Tulsa, Oklahoma
        FWI, Tulsa, Oklahoma
        Inland Environmental, Chicago, Illinois
        MAPCO, Inc., Tulsa, Oklahoma
        Midland-Ross Corporation, Roxboro, North Carolina
        Pall Trincor Corporation, Vauxhall, New Jersey
        Selas Flotronics, Spring House, Pennsylvania
        Separator & Recovery Systems, Inc., Santa Ana, California
        Serfilco, Northbrook, Illinois
        Smith Industries, Inc., Houston, Texas
        Velcon Filters, Inc., San Jose, California

Membrane Filtration Equipment

        Abcor, Inc., Cambridge, Massachusetts
        Aqua-Media, Sunnyvale, California
        Fluid Systems Division, UOP, Inc., San Diego, California
        Gulf Environmental Systems, San Diego, California
        Illinois Water Treatment Company, Rockford,  Illinois
        Osmonics, Inc., Hopkins, Minnesota
        Romicon, Inc., Woburn, Massachusetts
        Selas Flotronics, Spring House, Pennsylvania

Adsorption Equipment

        APV Company, Inc., Tonawanda, New York
        Aqua-Media, Sunnyvale, California
        Calgon Corporation, Pittsburgh, Pennsylvania
        Chem-Pro Equipment Corporation, Fairfield, New Jersey
        Chiyoda Chemical Engineering & Construction Company, Ltd., Tokyo
        Diamond Shamrock, Cleveland, Ohio
        Ecodyne Corporation, Union, New Jersey
        Envirex, Inc.,  Conshohocken, Pennsylvania
        General Filter Company, Ames, Iowa
        Hydronautics,  Inc., Laurel, Maryland


                                     82

-------
Illinois Water Treatment Company, Rockford,  Illinois
Liquitech, Inc., Houston, Texas
Met-Pro Systems, Inc., Lansdale, Pennsylvania
Process Equipment Corporation, Bedding, Michigan
Serfilco, Northbrook, Illinois
                             83

-------
                                SECTION 9

                                REFERENCES

Adamson, A. W.  (1967)  Physical Chemistry of Surfaces.   2nd ed.   Inter-
     science, New York.  747 pp.

Adamson, W. L., and M. W. Titus  (1971)  Separation of oil in bilge water by
     semipermeable membrane.  Naval Ship Research and Development Center
     Rept. No. AD-A023-289, NTIS.

API  (1969)  Manual on Disposal of Refinery Wastes;  Volume on Liquid Wastes.
     1st ed.  American Petroleum Institute, Washington,  D.  C.

Arnaiz, J. B., and E. Batutis  (1974)   Feasibility test program of applica-
     tion of coalescing phase oil/water separators to self-compensating fuel
     tanks in surface ships.  U.S. Coast Guard Rept.  No.  CG-D-88-74, NTIS.

Berry, W. L., and R. F. Engel  (1969)   One approach minimizes water pollution
     from offshore platforms.  Pet. Eng., 10:64-66.

Beychok, M. R.  (1973)  Aqueous Wastes from Petroleum and Petrochemical
     Plants.  J. Wiley and Sons, New York.

Boehm, P. D.  (1973)  Solubilization of hydrocarbons  by dissolved organic
     matter in seawater.  M.S. Thesis.  University of Rhode Island, Kingston.

Boehm, P. D., and J. G. Quinn  (1974)   The solubility behavior of No. 2 fuel
     oil in seawater.  Mar. Poll. Bull., 5(7):101-105.

Boesch, D. F., C.  H. Hershner, and J.  H. Milgram  (1974)   Oil Spills  and  the
     Marine Environment.  Ballinger Publishing Co., Cambridge, Mass.

Brunsmann, J. J.,  J. Cornelissen, and H. Eilers  (1962)   Improved oil separa-
     tion in gravity separators.  J. WPCF, 34(l):44-55.

Budininkas, P., and G. A. Remus  (1974)  Development of classification  scale
     for characterizing bilge waters used in evaluating oil removal tech-
     niques .  U.S. Coast Guard Report No. AD-778-929, NTIS.

Burtis, T. A., and C. G. Kirkbride  (1946)  Desalting of petroleum by use of
     fiberglass packing.  Trans. AIChE., 42:413-416.

Chieu,  J.-N., E. F. Gloyna, and R. S.  Schechter  (1975)  Coalescence of emul-
     sified oily wastewater by fibrous beds.  Presented at the 30th Indus-
     trial Waste Conference, Purdue University, Lafayette, Ind.

                                    84

-------
Chiyoda Company  (1974)  Water treatment technologies from Chiyoda.   Tech.
     Kept. No. 209-04, Chiyoda Company, Tokyo.

Churchill, R. J.  (1973)  Air flotation techniques for oily water treatment.
     Tech. Rept., Engineering Science, Inc., Pasadena, Calif.

Churchill, R. J., and W. J. Kaufman   (1973)  Waste processing related surface
     chemistry of oil refinery wastewaters.  SEKL Rept. No. 73-3, University
     of California, Berkeley.

D'Arcy, N. A.  (1951)  Dissolved air flotation separates oil from wastewater.
     Oil Gas J., 6:319-322.

Desai, S. V.  (1971)  An economically attractive application of reverse
     osmosis to refinement of a petrochemical effluent stream.  AIChE Symp.
     Ser., Water,  pp. 379-387.

Douglas, E., and I. G. Elliot  (1962)  Developments in oily-water separator
     design.  Trans. Inst. Mar. Eng., 74:164-168.

Ellis, M. M., and P. W. Fisher  (1970)  Clarifying oil field and refinery
     wastewaters by gas flotation.  J. Pet. Technol. and Soc. Pet. Engrs.
     Rept. No. SPE 3198.

Elworthy, P. H., A. T. Florence, and C. B. Macfarlane  (1968)  Solubilization
     by Surface Active Agents.  Chapman and Hall, Ltd., London,  pp.  11-116.

Evers, R. H.  (1975)  Mixed-media filtration of oily wastewaters. J.  Pet.
     Technol., 2:157-163.

Farley, R., and F. H. H. Valentin  (1965)  Coagulation as a means of  separat-
     ing oil from effluents.  AIChE-IEC Symp. Ser., 1(1):15-20.
    V
Finger, S. M., and R. B. Tabakin  (1973)  Development of shipboard oil/water
     separation systems.  Rept. No. 73-ENAs-38, ASME, New York.

Finger, S. M., and T. S. Yu  (1972)  Technologies for shipboard oil pollution
     abatement:   Effects of operational parameters on coalescence. Naval
     Ship Research and Development Center Rept. No. 3598.

Finger, S. M., and T. S. Yu  (1973)  Analytical examination of oil/water
     separation by coalescence.  USEPA Control Technology Research and
     Development,  pp. 403-408.  Also in Proceedings of the Joint Conference
     on Prevention and Control of Oil Spills.  API, Washington, D. C.
     pp. 407-408.

Fowkes, F. M., F. W. Anderson, and J. E. Berger  (1970)  Bimetallic coales-
     cers:  Electrophoretic coalescence of emulsions in beds of mixed metal
     granules.  Env. Sci. Technol., 4(6):510-514.

Freestone, F. J., and R. B. Tabakin   (1975)  Review of Environmental  Protec-
     tion Agency research in oil/water separation technology.  Proceedings of
     the Joint Conference on Prevention and Control of Oil Spills. API,
     Washington,  D. C.  pp. 437-441.

                                      85

-------
 Fruman, D. H.   (1974)  Development of a batchwise in-situ regeneration-type
     separator  to remove oil from oil-water suspensions.  Tech. Rept.  No.'
     7080-3, Hydronautics, Inc., Maryland.

 Fruman, D. H.,  and T. R. Sundaram  (1974)  Evaluation of pump emulsification
     characteristics.  Naval Ship Engineering Rept. No. 6159-00, NTIS.

 Gaudin, A. M.   (1957)  Flotation.  2nd ed.  McGraw-Hill Book Co., New  York.

 Ghosh, M. M., and W. P. Brown   (1975)  Oil removal by carbon-metal granular
     beds.   J.  WPCF, 47(8):2101-2113.

 Gloyna, E. F.,  and D. L. Ford   (1974)  Control of refinery and petrochemical
     wastewaters and residuals.  Presented at the 1st International Symposium
     on the  Techniques of Liquid-Liquid Separation.  Lamar University,
     Beaumont,  Texas.

 Goldsmith, R. L.,,and S. Hossain  (1973)  Ultrafiltration concept for  separ-
     ating oil  from water.  U.S. Coast Guard Rept. No. 4305.2/2, NTIS
     AD-758-318.

 Graham, R. J.   (1962)  Separation of immiscible liquids by gravity settling
     and induced coalescence.  M.S."Thesis.  University of California,
     Berkeley.

 Gudesen, R.  C.  (1964)  Coalescence of petroleum compounds in mixed fibrous
     beds.   M.S. Thesis.  Illinois Institute of Technology, Chicago.

 Guzdar, A. R.,  A. C. Harvey, J. Potter, and W. M. Mack  (1975)  Development
     of a centrifugal system for separation of oil and solids from shipboard
     discharge  water.  U.S. Coast Guard Rept. No. CG-D-118-75, NTIS.

 Harris, F. R.,  Inc.  (1973)  Port collection and separation facilities for
     oily wastes. II. General technology.  Maritime Administration Rept. No.
     COM-73-11069, NTIS.

 Hartenstein, L. J., and T. E. Lindemuth  (1970)  Development of a coalescing
     type oil/water separator for marine service.  Aqua-Chem Tech. Rept.
     presented  at SNAME Meeting, San Diego, California, February 18.

 Harvey, A. C.,  A. R. Guzdar, V. K. Stokes, and A. T. Fisk  (1973)  Centrifuge
     coalescer  for separating oil from water in shipboard applications.
     U.S. Coast Guard Rept. No. AD-764-006, NTIS.

 Hayes,  J. G., L. A. Hays, and H. S. Wood  (1949)  Commercial desalting unit
     employing  fiberglass as contacting agent.  Chem. Eng. Progr. 45:235.

 Hazlett, R. N.  (1969a)  Fibrous bed coalescence of water:  Steps in the
     coalescence process.  IEC, Fund., 8(4):625-632.

Hazlett, R. N.  (1969b)  Fibrous bed coalescence of water:  Role of sulfo-
     nate surfactant in the coalescence process.  IEC, Fund., 8(4):633-640.

                                    86

-------
Hefler, J. R.   (1971)  Oily water separator:  Liquid-liquid separation by a
     commercial self-cleaning edge filter.  Maritime'Administration Kept. No.
     COM-71-01095, NTIS.

Holt, Ben Company, The   (1974)  Experimental prototype oily wastewater treat-
     ment system.  Civil Engineering Lab. Rept. No. CR-74.008, NTIS.

Hooper, M. W.,  and H. N. Myrick   (1972)  Comparison of multi-media and deep-
     bed sand filter coalescence of oil-water emulsions.  Presented at Con-
     ference on Application of Filtration Technology in Environmental Pollu-
     tion Control and the  Chemical Process Industries.  University of Houston,
     Texas.

Hsiung, K. Y., H. M. Mueller, and W. R. Conley  (1974)  Physical-chemical
     treatment  for oily waste.  Presented at WWEMA Industrial Water and
     Pollution  Conference  and Exposition, Detroit, Mich.

Ingersoll, A. C.  (1951)   Fundamentals and performance of gravity separation.
     Pet. Refiner, 6:9.

Jefferson, T. H., and S. B. Boulavare  (1973)  Surfactants and their effects
     on filter  separators.  U.S. Army Mobility Equipment Research and Devel-
     opment Center Rept. No. 2066, NTIS.
          ^
Jeffreys, G. V., and G. A. Davies  (1971)  Coalescence of liquid droplets and
     liquid dispersion.  In:  Recent Advances in Liquid-Liquid Extraction
     (C. Hanson, Ed.).  Pergamon Press, Oxford,  pp. 561-562.

Jordan, G. V.   (1953)  Separation of immiscible liquids by means of porous
     membranes.  Trans. ASME, 77:393-404.

Jordan, G. V.   (1965)  Coalescence of fluids through porous materials.  Selas
     Corporation Tech. Bull. LFC.  Selas Corporation of America, Flotronics,
     Spring House, Pa.

Kaiser, R., C. K. Colton,  G. Miskolczy, and L. Mir  (1971)  Magnetically
     induced separation of stable emulsions.  AIChE Symp. Ser., Water.
     pp. 115-126.

Kirby, A. W. W.  (1964)  The separation of petroleum oils from aqueous
     effluents.  Trans. Inst. Chem. Eng. (London), 4a(4):76.

Koelmans, H., and J. T. G. Overbeek  (1954)  Stability and electrophoretic
     deposition of suspensions in non-aqueous media.  Dis. Faraday Soc.,
     18:52.

Kruyt, H. R.  (1952)  Colloid Science.  Vol. 1.  Elsevier, New York.

Lai, M. G., and C. E. Adams  (1974)  Determination of the molecular solubil-
     ity of Navy oils in water.  Naval Ordnance Lab. Rept. No. NOLTR-74-110,
     NTIS.
                                     87

-------
 Langdon, W. M. , P. P. Naik, and D. T. Wasan  (1972)  Separation of oil dis-
      persions  from water by fibrous bed coalescers.  Env.  Sci.  Technol.,
      6(10):907.

 Langdon, W. M., and D. T. Wasan  (1971)  Experimental evaluation of fibrous
      bed coalescers for separating oil-water emulsions.  USEPA Project No.
      12050 DRC.11/71, NTIS.
                                        ^
 Lindenhofen, H., and R. H. Shertzer  (1967)  Aeronautical Engineering Lab.
      Kept. Nos. NAEC-AEL-1852 and NAEC-AEL-1856.  Aeronautical Engineering
      Lab., Philadelphia, Pa.

 Lysyj,  I., and E. C. Russell  (1974)  Dissolution of petroleum-derived
      products  in water.  Water Res., 8:863-868.

 McAuliffe, C.  (1969a)  Solubility in water of normal Cg and CIQ alkane
      hydrocarbons.  Science, 163:478-479.

 McAuliffe, C.  (1969b)  Determination of dissolved hydrocarbons in subsurface
      brines.   Chem. Geol., 4:225-233.

 McKenna, Q. H., H. Helber, C. M. Carrell,  and R. F. Tobias  (1973)  Electro-
      chemical  flotation concept for removing oil from water.  U.S. Coast
      Guard Rept. No. 734305.2/4, NTIS.

 Mensing, A. E., R. C. Stoeffler, W. R. Davison, and T. E.  Hoover  (1970)
      Investigation of the use of a vortex flow to separate oil from an oil/
      water mixture.  U.S. Coast Guard Rept. No. 714103/A/001, NTIS.

 Merryman, J. G., and E. R. Osterstock  (1973)  Coalescing plates and packs
      for oil/water separation in various shipboard applications.  U.S. Coast
      Guard Rept. No. 724305.2/6, NTIS.

 Messinger, S.  (1974)  Ultrafiltration in water and waste treatment.
      Presented at the 12th Annual Liberty Bell Corrosion Course No. 4.

 Milstead, C. E., and J. F. Loos  (1973)  Study of hydrophilic membranes for
      oil-water separation.  U.S. Coast Guard Rept. No. 4305.2/7, NTIS
      AD-758-321.

 Mittleman, J.  (1975)  Oil/water separator evaluation.  Naval Coastal Systems
      Lab. Rept. No. 252-75, NTIS.

 Navy, Department of the  (1974)  Coanda-effect oil-water separation.  U.S.
      Department of the Navy, Washington, D. C., NTIS AD-D001-437.

Nordstrom, R. P., Jr.  (1974)  Ultrafiltration removal of soluble oil.  Poll.
      Eng., 10:46-47.

Norton, L.  B.  (1972)   Test and evaluation of oil pollution abatement
      devices for shipboard use. Phase II.   U.S. Coast Guard Rept. No.
     AD-762-499,  NTIS.

                                     88

-------
Norton, L. B., and F. Perrj.nl  (1972)  Test and evaluation of oil pollution
     abatement devices for shipboard use. Phase I.  U.S. Coast Guard Rept
     No. AD-&62-498, NTIS.

Norton, L. B.  (1973)  Test and evaluation of oil pollution abatement devices
     for shipboard use. Phase III.  Final U.S. Coast Guard Rept.  No
     AD-762-488, NTIS.

Orr, C., and E. Kang  (1974)  The electrical process in the breaking of
     dilute oil-in-water emulsions.  Office of Water Resources Research  Rept.
     No. PB-235-908, NTIS.

Overbeek, J. T. G.  (1952)  Colloid Science.  Vol. I (H. R. Kruyt, Ed.).
     Elsevier, New York.

Pal, D.  (1974)  The Coanda-effect oil-water separator:  A feasibility study.
     Civil Engineering Lab. Rept. No. AD-774-080, NTIS.

Paszyc, A. J., D. Pal, K. Huang, and J. B. Curry  (1975)  Fluidic oil-water
     separator.  U.S. Department of the Navy, Washington, D. C.,  NTIS
     AD-D001-756.

Pomonik, G. M.  (1973)  Vacuum desorption concept for removing oil from
     water.  U.S. Coast Guard Rept. No. 734305.2/8, NTIS.

Pontello, A. P., F. G. Woessner, and R. J. Delfosse  (1973)  Oil/water pollu-
     tion program. Phase I.  Naval Air Propulsion Test Center Rept. No.
     NAPTC-PE-27, NTIS.

Pontello, A. P., C. J. Collick, J. J. Palmer, and A. J. Rollo  (1974) Oil/
     water pollution program. Phase II.  U.S. Department of the Navy,
     Washington, D. C., NTIS AD-A009-093.

Quigley, R. E., and E. L. Hoffman  (1966)  Flotation of oil wastes.  Pro-
     ceedings of the 21st Industrial Waste Conference.  Purdue University,
     Lafayette, Ind.  pp. 527-533.

Redmon, 0. C.  (1963)  Cartridge type coalescers.  Chem. Eng. Progr.,
     59(9):87.

Reisberg, J., and T. M. Doscher  (1956)  Interfacial phenomena in crude
     oil-water systems.  Producers Monthly, 4:43.

Rohlich, G. A.  (1951)  Hydraulic characteristics of gravity-type oil-water
     separators.  Proc. API, 31M(111):63-85.

Rohlich, G. A.  .(1954)  Application of air flotation to refinery wastewaters.
     Ind. Eng. Chem., 46(2) :304-308.

Rose, P. R.  (1963)  Mechanisms of operation of a fibrous bed coalescer.
     M.S.  Thesis.  Illinois Institute of Technology,  Chicago, 111.
                                    89

-------
Rosenfeld, J. I., and D. T. Wasan  (1974)   Coalescence  of  drops  in  a  liquid-
     liquid dispersion by passage through a fibrous bed.   Can. J. Chem.  Eng.,
     52:26-34.

Russell, E. C.   (1972)  Test and evaluation of a 50-gallon-per-minute oil/
     water separator  (Separation and Recovery Systems,  Inc.).  U.S. Army
     Mobility Equipment Research and Development Center Rept.  No. AD-785-223,
     NTIS.

Russell, E. C.   (1973a)  RPC Division, Midland-Ross Corporation  10-gallon-
     per-minute  liquid/liquid separator.  U.S. Army Mobility Equipment
     Research and Development Center Rept. No. 2058, NTIS.

Russell, E. C.   (1973b)  Fram Corporation Model OWS-23-FCI-USCG  oil/water
     separator.  U.S. Army Mobility Equipment Research and Development
     Center Rept. No. 2059, NTIS.

Russell, E. C.   (1973c)  Separation and Recovery Systems,  Inc. 100-gallon-
     per-minute  oil/water separator.  U.S. Army Mobility Equipment  Research
     and Development  Center Rept. No. 2060, NTIS.

Sareen, S. S., P. M.  Rose, R. C, Gudesen, and R. C. Kintner  (1966)
     Coalescence in fibrous beds.  AIChE J., 12:1045.

SCEP   (1970)  Man's Impact on the Global Environment:  Report of the  Study  of
     Critical Environmental Problems.  MIT Press, Cambridge, Mass.   296  pp.

Schatzberg, P., L. R. Harris, C. M. Adema, D. F. Jackson,  and C. M. Kelly
     (1975)  Oil-water separation with noncellulosic ultrafiltration  systems.
     Proceedings of the Joint Conference on Prevention and Control  of Oil
     Spills.  API, Washington, D. C.  pp. 443-447.

Shackleton, L. R. B., E. Douglas, and T. Walsh  (1960)   Pollution of  the sea
     by oil.  Trans.  Inst. Mar. Eng.

Sheng, H. P., and J.  R. Welker  (1969)  Liquid-liquid separation in a conven-
     tional hydrocyclone.  Proceedings of the AIChE 64th National Meeting,
     New Orleans, La.

Sherony, D. F., and R. C. Kintner  (1971)  Coalescence of an emulsion in a
     fibrous bed. II. Experimental.  Can. J. Chem. Eng., 49:321.
                      r
Simonsen, R. N.  (1962)  Remove oil by air flotation.  Hydrocarbon  Proc.
     and Pet.  Ref.,  41:5.

Sims, A. V.  (1972)   Test and evaluation of oil-water separation systems.
     Naval Civil Engineering Lab. Rept. No. CR-73-015, NTIS.

Sinkin, D. J., and R. B. Olney (1956)  Phase separation and mass transfer in
     a liquid-liquid cyclone.  AIChE J., 2(4):545-551.
                                    90

-------
Skocypec, R. J.  (1972)  Oily water separation system.  U.S. Maritime
     Administration Kept. No. COM-72-10561, NTIS.

Spielman, L. A.  (1968)  Separation of finely dispersed liquid-liquid  suspen-
     sions by flow through fibrous media.  Ph.D. Dissertation.  University  of
     California, Berkeley.

Spielman, L. A., and S. L. Goren   (1970)  Progress in induced coalescence and
     a new theoretical framework for coalescence by porous media.  Ind.  Ene.
     Chem., 62(10)-.10-24.

Spielman, L. A., and S. L. Goren   (1972a)  Theory of coalescence by flow
     through porous media.  Ind. Eng. Chem., Fund., 11(1):66.

Spielman, L. A., and S. L. Goren   (1972b)  Experiments in coalescence  by flow
     through fibrous mats.  Ind. Eng. Chem., Fund., 11(1):73.

Stoeffler, R. C.,  and C. E. Jones  (1973)  Vortex concept for separating oil
     from water.   U.S. Coast Guard Rept. No. 4105.2/1, NTIS.

Stormont, D. H.  (1956)  Air flotation used to separate oil at Richfield's
     new waste water plant.  Oil Gas J., 10:26.

Sweeney, W. F.   (1964)  Some observations on liquid-liquid settling.   M.S.
     Thesis.  University of California, Berkeley.

Union Carbide Corporation  (1973)  Viscosity-actuated phase separating
     tyAPS) for oil-water separations.  Prepared for U.S. Coast Guard,  Rept.
     No. 734305.2/5, NTIS.

USEPA  (1974)  Development document for effluent limitations guidelines  and
     new source  performance standards for petroleum refining:  Point  source
     category.   USEPA No. 440/1-74-014a, NTIS.

USEPA  (1975)  Group II.  Development document for interim final effluent
     limitations guidelines and new source performance standards for the off-
     shore segment of the oil and  gas extraction point source category.
     USEPA No. 440/1-75-055, NTIS.

Vinson, C. G.  (1965)  The coalescence of micro-size drops in liquid-liquid
     dispersions in flow past fine mesh screens.  Ph.D. Dissertation.
     University  of Michigan, Ann Arbor.

Voyutskii, S. S.,  K. A. Kal'yanova, R. Panick, and N. Fodiman  (1955)
     Mechanism of  separation of the dispersed phase of emulsions during
     filtration.   Chem. Abstr., 49:12053d.

Voyutskii, S. S.,  N. M. Fodiman, and R. Panick   (1958)  Filtration of
     emulsions.  Chem. Abstr., 52:19266a.

Vrablik, E. R.   (1957)  An evaluation of circular gravity-type separators  and
     dissolved air flotation for treating oil refinery waste water.

                                     91

-------
     Proceedings of the 12th Industrial Waste  Conference. Purdue University,
     Lafayette, Ind.  p. 73.

Wang, L. K., J. Y. Yang, and D.  B.  Dohm  (1973)  Evaluation and development
     of physical-chemical techniques for the separation of emulsified oil
     from water.  Kept. No. 189, Calspan Corporation, Buffalo, N. Y.

Wisconsin, University of  (1949)  Investigation  of  the behavior of oil-water
     mixtures in separators.  Eng.  Expt. Sta.  Ser.  Kept. No.  CE-78-1.
     University of Wisconsin, Madison.

Wisconsin, University of  (1950)  Investigation  of  the behavior of oil-water
     mixtures in separators.  Eng.  Expt. Sta.  Ser.  Kept. No.  CE-78-2.
     University of Wisconsin, Madison.

Wisconsin, University of  (1951)  Investigation  of  the behavior of oil-water
     mixtures in separators.  Eng.  Expt. Sta.  Ser.  Kept. No.  CE-78-3.
     University of Wisconsin, Madison.

Yu, T. S.  (1969)  A proposed shipboard continuous  oil-pollution control
     process for bilge water.  Naval Ship Research  and Development Lab. Kept.
     No. 3191, NTIS.

Yu, T. S., and D. R. Ventriglio  (1969)  Shipboard  oil-pollution control
     systems for ballast and bilge  waters:   A  state-of-the-art search.
     MATLAB Rept. No. 244, NTIS.
                                    92

-------
                                SECTION 10

                               BIBLIOGRAPHY

Armco Steel Corporation.  1970.  Treatment of Waste Water:   Waste Oil Mix-
     tures.  Federal Water Pollution Control Administration Program No.
     12010 EZV.

Battelle Memorial Laboratory (Pacific Northwest Laboratory).  1967.  Oil
     Spillage Study:  Literature Search and Critical Evaluation for Selection
     of Promising Techniques to Control and Prevent Damage.  NTIS Kept. No.
     AD-666-289.

Beebe, A. H.  1953.  Soluble Oil Wastes Treatment by Pressure Flotation.
     Sewage Ind. Wastes.  25(11):1314-1322.

Blumer, M.  1969.  Oil Pollution of the Ocean.  In:  Oil on the Sea.  Ed.
     D. P. Hoult.  Plenum Press, New York.  pp. 5-13.

Blumer, M.  1970.  Scientific Aspects of the Oil Spill Problem.  Presented at
     the Colloquium on Oil Pollution of the Sea.  North Atlantic Treaty
     Organization, Brussels.

Boyd, J. L., G. L. Shell, and D. A. Dahlstrom.  1972.  Treatment of Oily
     Waste Waters to Meet Regulatory Standards.  AIChE Symp. Ser.
     124:393-401.
                                              *
Breslau, B. R., E. A. Agranat, A. J. Testa, S. Messinger, and R. A. Cross.
     1974.  Hollow Fiber Ultrafiltration:  A Systems Approach for Process
     Water and By-Product Recovery.  Proceedings of the AIChE 79th National
     Meeting, Houston, Texas.

Brown and Root, Inc.  1974.  Determination of Best Practicable Control Tech-
     nology Currently Available to Remove Oil from Water Produced with Oil
     and Gas.  Offshore Operators Committee, Sheen Technical Subcommittee
     Report.

Canadian Plant and Process Engineering, Ltd.  1972.  Polyurethane as  an Oil
     Filter:  A Research Study.  Final Report, Canadian Department of the
     Environment (Fisheries and Forestry), Ontario.

Canevari, G. P.  1969.  Some Basic Concepts Regarding the Separation  of Oily
     Water Mixtures.  Trans. ASCE.  12:190.

Cheesman, D. F., and A. King.  1940.  The Electrical Double Layer in  Relation


                                    93

-------
     to the Stabilization of Emulsions with Electrolytes.  Trans. Faraday
     Soc. 36:241.

Chemical Engineering Deskbook.  1975.  Environmental Engineering Pollution
     Control Equipment.  82(21).  McGraw-Hill Book Co., Hightstown, N. J.

Chieu, J. N., and E. F. Gloyna.  1975.  Treatment of Emulsified Oily Wastes
     by Fibrous Bed Coalescers.  CRWR Report, University of Texas, Austin.

Conley, W. R., and K. Hsiung.  1969.  Design and Application of Multi-media
     Filters.  J. Amer. Water Works Assn. 61:97.

Degler, S. E.  1970.  Oil Pollution:  Problems and Policies.  BNA's Environ-
     mental Management Series.

Edwards, V. H., and R. K. Finn.  1969.  New Separation Techniques.  AIChE,
     Today Series, New Orleans, La.  pp. 21-22.

Engineering Science, Inc. (Texas).  1971.  Preliminary Investigational
     Requirements:  Petrochemical and Refinery Waste Treatment Facilities.
     EPA Project No. 12020 BID, NTIS.

Gloyna, E. F., and D. L. Ford.  1970.  The Characteristics and Pollutional
     Problems Associated with Petrochemical Wastes.  Engineering Science,
     Inc., Texas, Report.

Goldsmith, R. L., and S. Hossain.  1973.  Oil-Water Separation by Ultrafil-
     tration.  Proceedings of the Joint Conference on Prevention and Control
     of Oil Spills.  API, Washington, D. C.  pp. 441-456.

Goldsmith, R. L., D. A. Roberts, and D. L. Burre.  1974.  Ultrafiltration of
     Soluble Oil Wastes.  JWPCF.  46(9):2183-2192.

Goren, S. L.  1974.  Removal of Oil from Aqueous Wastes by Flotation.
     Office of Water Resources Research, Washington, D. C.  Rept. No.
     PB-234-023, NTIS.

Grosz, R. B., and R. A. Kormanck.  1975.  Dissolved Air Flotation.  Unpub-
     lished report.

Gruenfeld, M.  1973.  Extraction of Dispersed Oils from Water  for Quantita-
     tive Analysis by Infrared Spectrophotometry.  Env-  Sci. Technol.
     7(7):636-639.

Hensen, S. P., G. H. Richardson, and K. Hsiung.  1969.   Some Recent Advances
     in Water Treatment Technology.  Chem. Eng. Progr.,  Symp.  Ser.  65.

Harris, R. F.  1975.  Report on Port Collection and Separation Facilities
     for Oily Wastes. Volume IV:  Determination of Impact of the Construc-
     tion of Offshore Terminals on Their Contiguous Ports.  Maritime
     Administration Contract No. 2-36202, NTIS.
                                    94

-------
Harvey, A. C. , A. R. Guzdar,  and  D.  R.  Friswell.   1973.   Laboratory Evalua-
     tion of  the Emulsifying  Characteristics  of  Pumps.   U.S.  Coast Guard
     Rept. No. CG-D-31-74,  NTIS.

Harvey, A. C. , and V. K.  Stokes.   1973.   Evaluation of a Unique Centrifuge
     for Separation of  Oil  from Ship Discharge Waters.   Proceedings of  the
     Joint Conference on  Prevention  and Control  of Oil Spills.  API,
     Washington, D. C.  p.  391.

Hoult, D. P.   1969.  Oil  on the Sea.   Plenum  Press,  New  York.

Hyland, J. R.  1971.  Study of  Oily  Water Marine Treatment Facilities.
     USEPA Office of Water  Programs  Report, NTIS.

Jones, H. R.   1973.  Pollution  Control in the Petroleum  Industry.  Noyes
     Data Corporation,  New  Jersey.

King, A.  1941.  Some Factors Governing the Stability of Oil-in-Water
     Emulsions.  Trans. Faraday Soc.   37:168.

Lederman, P.  B. , and J. S.  Dorrler.   1975.  Development  of Offshore Oil and
     Gas  in  New  England:  Environmental Problems and Solutions.  USEPA
     Report.

Little, R. C.  1974.  Breaking  Emulsions of Water in Navy Fuel Oils.
     Fuel.   53:246-252.

McKay, W. C.   1973.  Evaluation of Concepts for  Separating Oil from Water
     Discharged  from Ships.  U.S. Coast Guard Rept.  No.  CG-D-26-74, NTIS
     734305.2/9.

Milne, D.  1950.  Character of  Waste Oil Emulsions.  Sewage  Ind. Wastes.
     22(3):326.

Nelson-Smith,  A.  1972.   Oil  Pollution and Marine Ecology.   Elek Science,
     London.

Permutit  Company (The).   1966.  Research and  Development for a Shipboard
     Oil  and Water  Separation System.  Maritime  Administration Contract No.
     MA-2722, NTIS.

Poliakoff, M.  Z.  1969.   Oil  Dispersing Chemicals.  Federal  Water  Pollution
     Control Administration Rept. No. 15080FHS-05/69-14/12/549.

Pomeroy,  R.   1953.  Floatability  of  Oil and Grease in  Wastewaters.   Sewage
     Ind. Wastes.   25:1304-1313.

Redlien, W.  H.,  and J.  H. McClintock.  1951.   Reducing Water Pollution for
     Oil  Refineries.  Pet.  Refiner.   30(6):123-126.

Reid, G. W. ,  L.  E.  Streebin,  D. W. Rumfeldt,  and R. Sweazy.   1972.   Evalua-
     tion of Waste  Waters from Petroleum and Coal Processing.  EPA-R2,72-001.

                                      95

-------
Shah, D. 0.  1970.  Molecular Interactions at the Oil/Water Interfaces  and
     the Formation of Microemulsions.   Federal Water Pollution Control  Admin-
     istration Kept. No. 15080EMP-06/70.

Skocypec, R.  1974.  Electrolyte Effects on Flotation of Oily Water.  M.S.
     Thesis.  University of California, Berkeley.

Smookler, A. L., and J. W. Harden.  1973.  Navy Shipboard Investigation of
     Oily Wastes.  Proceedings of the  Joint Conference on Prevention  and
     Control of Oil Spills.  API, Washington, D. C.   pp. 189-193.

Sniegoski, P. J.  1973.  The Determination of Water  Soluble Components  in
     Petroleum Products Used by the Navy.  U.S. Naval Research Laboratory
     2nd Progress Report.

Sollner, S.  1944.  The  Application of Sonic and Ultrasonic Waves  in
     Colloid Chemistry.  Chem. Rev.  34:371.

Sport, M. C.  1969-  Design and Operation of Gas Flotation Equipment  for the
     Treatment of Oilfield Produced Brines.  Presented at the Offshore
     Technology Conference, Houston, Texas.  Preprint No. OTC 1051.
     1:111-145.

Stahl, G. W., D. H. Meyer, and B. H. Rankin.   1972.   Separation of  Oil  in
     Bilge Water by Semipermeable Membranes.   Naval  Systems Engineering
     Rept. No. USNA EW-72-7, NTIS.

Thompson, C. S., J. Stock, and P. L. Mehta.  1972.  Cost and Operating
     Factors for Treatment of Oily Waste Water.  Oil Gas J.  53.

Thompson, D.  1951.  Ultrasoncs.  Chem. Eng.  Progr.  Symp. Series.   47(1).

Vrablik, E. R.  1959.  Fundamental Principles of Dissolved Air Flotation of
     Industrial Wastes.  Proceedings of the 14th Industrial Waste  Conference,
     Purdue University, Lafayette, Ind.  p. 743.

Weston, R. F.  1950.  Separation of Oil Refinery Wastes.  Ind. Eng.  Chem.  4.

Weston, R. F.  1952.  Waste Control at Oil Refineries.  Chem. Eng.  Progr.  9.

Witmer, P. E., and A. Gollan.  1971.  Oil-Water Regenerative Separator—
     Phase I.  Final Report of Phase I:  Development Program of a  Continuous
     Regenerative Moving Bed to Remove Oil from Oil-Water Suspensions.
     Maritime Administration Rept. No. COM-72-11041, NTIS.
                                    96

-------
TECHNICAL REPORT DATA
(Please rcaj Imtmctiom on the rtrcrse before completing)
1. REPORT MO
EPA-600/2-78- 069
i. TITLE AND SUBTITLE OIL/WATER ,
ART
2.
SEPARATION: STATE-OF-THE-

7.AUTHOR(S) Fidelis A. Osamor
Robert C. Ahlert
|9. PERFORMING ORGANIZATION NAME AND ADDRESS
1 Dept. of Chemical & Biochemical Engineering
1 Rutgers, The State University of New Jersey
New Brunswick, New Jersey 08903
112. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Research Laboratory-Gin., OH
Office of Research and Development
1 U.S. Environmental Protection Agency
I Cincinnati, Ohio 45268
3. RECIPIENT'S ACCESSION-NO.
5. REPORT DATE
April 1973 issuinq date
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM CLEMENT NO.
EHE623
11. CONTRACT/GRANT NO.
R803978
13. TYPE OF REPORT AND PERIOD COVERED
Final 7/11/7S - 7/T1/77
t4. SPONSORING AGENCY CODE
EPA/600/12 j
115. SUPPLEMENTARY NOTES
|16. ABSTRACT
           This report  reviews the state-of-the-art for oil/water separating
      devices and processes.   Devices and process are classified according to
      the primary mechanism that induces separation of oil/water mixtures.  The
      basic concepts, specific design features, operational conditions, and
      limitations of each category are discussed.

           Literature on test evaluation of a variety of devices is critiqued
      on the basis of actual or potential success in treating various oil/
      water system states.   No single technique can separate all oil/water
      system states efficiently.  Specific deficiencies in existing technology
      have been identified.

           This report  was  submitted in fulfillment of Research Grant No.
      R803978 by Rutgers University Under the.sponsorship of the U.S.
      Environmental Protection Agency.  This report covers the period July 1,
      1975, to June 30,  1977, and work was completed as of July 31, 1977.
17.
a.
KEY WORDS AND DOCUMENT ANALYSIS ]
DESCRIPTORS
Petroleum industry
Oil/water separation
Oily wastewaters
13.
DISTRIBUTION STATEMENT
TJTTT 17 ACT7 TO PTTRTTP
KEj.L£iAo £t L\J JTUlJJ_iiV>
b.lDENTIFIERS/OPEN ENDED TERMS
Oil pollution
Oil /water treatment
Coalescence
Gravity separation
19. SECURITY CLASS (This Report)
UNCLASSIFIED
20. SECURITY CLASS (This page)
UNCLASSIFIED
. COSATI Field/Group J
68D I
21. NO. OF PAGES j
104 |
22. PRICE
 EPA Form 222O-1 (9-73)
           * u.s. GovwiMBn pimmo OFHCE, i97s-a60-a?o/%

-------