WATER POLLUTION CONTROL RESEARCH SERIES •15080EUU 10/70
            VORTEX SEPARATION FOR
          OIL SPILL RECOVERY SYSTEMS
ENVIRONMENTAL PROTECTION AGENCY • WATER QUALITY OFFICE

-------
         WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Reports describe the
results and progress in the control and abatement of
pollution in our Nation's waters.  They provide a central
source of information on the research, development, and
demonstration activities in the Water Quality Office of
the Environmental Protection Agency, through inhouse re-
search and grants and contracts with Federal, State and
local agencies, research institutions, and industrial
organizations.

Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Head, Project Reports
System, Planning and Resources Office, Office of Research
and Development, Water Quality Office, Environmental Pro-
tection Agency, Room 1108, Washington, D. C.  20242.

-------
                  VORTEX SEPARATION PROCESS FOR
                   OIL SPILL RECOVERY SYSTEMS
            AMERICAN PROCESS EQUIPMENT CORPORATION
                       2015  Lisenby Avenue
                   Panama  City, Florida  324-01
                              for the
                      WATER QUALITY OFFICE
                 ENVIRONMENTAL PROTECTION AGENCY
                      Program No. 15080 EUU
                           October, 1970
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price 60 cents

-------
                EPA Review Notice
This report has been reviewed by the Environmental
Protection Agency and approved for publication.
Approval does not signify that the contents necessarily
reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or
commercial products constitute endorsement or recommenda-
tion for use.

-------
                          ABSTRACT
A two-phase program to design, fabricate,  and test a 1 MGD proto-
type oil/water separator  for use in oil spill recovery systems was
completed successfully,  In Phase I,  an experimental 50  GPM vortex
separator was built and tested in a variety of operating modes, lead-
ing to definition of a preferred vortex separation process.

The Phase II full-scale prototype was tested at a Standard Oil Company
refinery,,  Oil/water mixtures containing from 1 - 10% oil were suc-
cessfully separated to yield water  effluents containing 65 - 235 ppm
total oil.  Recovered oil contained 1 - 5% water. Recovery of influent
oil from the mixture exceeded 99.5% under reproducible  operating
conditions.

Operation and performance of the full-scale prototype duplicated those
of the 50 GPM experimental model.   The total influent capacity of the
prototype was  about 50% of the project objective of 1  MGD,  due to use of
theoretically derived scaling laws  which were  subsequently shown to be
inapplicable to the preferred process defined.

The preferred vortex separation process uses a combination  of
centrifugal force and the force of gravity to facilitate oil  recovery by
causing floating oil to collect in a "submerged oil vortex" from which
it  can be removed by pumping.   The submerged oil vortex acts as an
oil surge tank, eliminating the need for precise control of rate of oil
removal in response to variable influent oil rates.   The process
efficiency,  in common with that of other proposed oil separation
processes,  varied inversely with the extent of  emulsification or
dispersion of oil in water, and directly with the concentration of oil
in the influent mixture.

This report was  submitted in fulfillment of Contract  No.  14-12-825
between the Federal Water Quality Administration and American
Process Equipment Corporation,

Key Words:  Vortex separation,  submerged oil  vortex, separator
efficiency.

-------
                          CONTENTS


Section                                                     Page

         Abstract                                             i

         Contents                                             ii

         List of Figures                                      iii

         List of Tables                                        iv

I        Conclusions                                          1

 II       Recommendations                                     3

 III      Introduction                                          4

IV       Phase I.  Vortex Separation Process Development     6

V        Phase IL  Design and Construction                   24

VI       Phase II.  Full-Scale Prototype Demonstration
         and Evaluation                                       30

VII      Discussion                                          33

VIII     Acknowledgements                                   36

IX       References                                          37

X        Patents and Publications                             38

XI       Glossary                                            39

XII      Appendices                                          40
                                11

-------
                            FIGURES


                                                           Page

I        Deep Vortex Skimming Process                        7

II        Shallow Vortex Overflow Process                     10

III       Formation of "Submerged Oil Vortex"                 13

IV       Gravity Flow Oil Recovery from Submerged
         Oil Vortex                                           15

V        Upward Vertical Suction Oil Recovery from
         Submerged Oil Vortex                                17

VI       Phase I (50 GPM) Vortex Separator                   18

VII      Phase I Experimental Vortex Separator System        19

VIII     Phase II Vortex Separator Test Installation at
         Refinery                                             26

IX       Phase II Full-Scale Prototype Vortex Separator
         Top View                                            28

X        Phase II Full-Scale Prototype Vortex Separator
         Side View                                           29
                               111

-------
                            TABLES


No.                                                         Page

1     Selected Vortex Separator Process Development
      Tests                                                 11

2     Effect of Tilting on Submerged Oil Vortex               21

3     Effect of Oil Dispersion on Vortex Separator            23

4     Phase II Vortex Separator Oil Recovery Efficiency      32
                                IV

-------
                           SECTION I
                          CONCLUSIONS
A  simple vortex separation process has been demonstrated to be an
efficient method for separating floating oils from mixtures with water.
Recovery of better than 99.5% of influent oil was demonstrated from
influents containing 1  - 10% oil.  Recovery of  90% was obtained with
influents containing only 0.1% oil.  Effluent water containing as  little
as 10 ppm  oil was obtained, and effluent quality in the range  of 100-
200 ppm was reproducibly achieved.  Recovered oil contains 0   5%
water.

The efficiency and simplicity of the vortex separator process and
its operability over a wide range of processing variables make it
favorably suited for incorporation into a total oil recovery system
comprising sweeping-skimming-separation-storage.

Definition of preferred conditions for operation of the vortex sepa-
ration process was achieved. The "submerged oil vortex" which
forms under those conditons is a key element in the performance
characteristics described herein.  Critical design parameters and
scaling laws for the vortex separation process equipment were
defined and verified.

The vortex separation process was equally efficient 'with representa-
tive crude oils of specific gravity 0.82  and 0.90. Additionally,
water-in-oil emulsions  containing up to  30% water were effectively
separated.   These thixotropic emulsions with a viscosity similar to
that of mayonnaise had a specific gravity of 0. 93.

The vortex separation process was shown to be operable while the
separator tank was being tilted up to 15 degrees  from vertical.
Because efficiencies and controllability  are impaired by tilting,  however,
it  is concluded that operation aboard ship should be isolated  from rolling
and pitching motion by gimbal mounting.  A functional gimbal stand and
flexible hose connections were used in the full-scale prototype demon-
strations .

The submerged oil vortex feature of this separator  system provides
an integral oil "surge tank", obviating the need for continuous and
precise matching of oil  influx and withdrawal rates.

A  simple pneumatic level control closed loop system was very
effective in maintaining optimum operating level.  Control point
was an automatic water effluent  valve0

                                  1

-------
The efficiency of the vortex separation process varies inversely
with the degree of dispersion of oil in the water.  True emulsions
cannot be separated.  Careful attention to minimize mixing in
influent pumping and transport systems is,  herefore, requisite  in
the overall process.

For a vortex separator of given diameter and height, separation
efficiency varies inversely as the water influent flow rate, other
factors being  constant.  This relationship determines the practical
upper limit of flow capacity for a given unit.

The flow capacity of the Phase II unit is approximately 50% of the
project objective of 1 MGD.  Subsequent to final design of the unit,
clarification of scaling laws for the submerged vortex separation
process predicted this result.

-------
                            SECTION II
                       RECOMMENDATIONS
The  simple,  reliable, and efficient vortex separation process defined
under Contract 14 - 12 -  825  should be the subject of continuing engineer
ing studies designed to facilitate its  application in oil spill recovery
and other oily pollution abatement problems.  Recommended programs
are:

      Testing and evaluation  of the Phase II separator under
      at-sea field conditions.

      Integration of the vortex separator into a  total oil spill
      recovery system comprising sweeping,  skimming,
      separation,  and storage.

      Definition of the vortex separation process  efficiencies
      with a  wider variety of oily pollutants,  including
      petroleum products, vegetable and animal oils, fatty
      acids,  solvents, and greases.   Process limitations with
      respect to oil compostions,  densities, and viscosities
      should be determined, to permit reliable  appraisal of
      process applicability to a wide variety of  oily waste treat-
      ment problems.

      Development of an  automatic control loop to sense the
      volume of the submerged oil vortex and control oil
      recovery flow rate accordingly.   This would assure
      maximum efficiency of operation and reduce required
      operator skill.

      Design, fabrication, and  evaluation of a highly portable
      skimmer/vortex separator system for rapid recovery of
      small spills in rivers,  harbors, terminal docks,  and
      other  restricted waters.  Such a system could be  trailer-
      mounted for shore-based operation or adapted for use on
      small, maneuverable vessels.

-------
                           SECTION III
                         INTRODUCTION
Oil spills of major proportions in recent years have resulted in
widespread recognition of their ecological impact, of the urgent
need to prevent such spills or to recover  them rapidly  if they do
occur.  A serious lack of existing technology and equipment for
their  effective recovery is also recognized.  The publicity attend-
ant to a disaster such as  in the cases of the Torrey Canyon or  the
Santa Barbara Channelhas resulted in public demand for quick,
effective solutions to such problems.  Industry and government
officials recognize,  also, that thousands of lesser oil contamination
incidents occur in many natural waters annually.   The  summation
of the damaging effects of these minor spills may indeed exceed that
for the  less frequent major accidents.

This Contract is a part of the Federal Water Quality Administration's
extensive program to develop and provide technology and equipment
for the  safe,  fast, and efficient recovery  of spilled oils from inland
and coastal waters.

A  total  oil spill recovery system may involve at  least four basic,
separate but interdependent,  unit operations. They are:  sweeping
and/or  containment;  skimming; oil/water separation; and  recovered
oil storage.  A variety of sweeping and skimming methods have been
tested with  varying degrees  of success.  Environmental conditions
are always  a major  factor in the effectiveness of these operations „

Particularly in choppy or heavy seas,  the efficiency of skimming
operations may be relatively low,  so that in attempting to remove
oil from the water's  surface,  in fact a mixture of oil and water is
obtained. Such mixtures often times contain only a few per cent  of
oil. These dilute mixtures must then be rapidly and efficiently
separated into a water fraction suitable for return to sea, and an
oil fraction containing a minimum of water in order to  reduce  interim
storage requirements for recovered oil at the recovery site.

This project had as  its objective the demonstration of a full-scale
prototype oil/water  separator, potentially capable of filling the need
outlined above: a continuous, rapid,  efficient oil/water separator;
operable in choppy to moderately heavy seas, and capable of separa-
ting low concentration oil/water mixtures into recovered oil nearly
free of  water  and clear water suitable for return to sea.

-------
American Process Equipment Corporation had conceived such an oil/
water separator, based on a modified hydrocyclone process.   Model
testing and a small prototype successfully demonstrated that the device
•was,  qualitatively, a simple, rapid, continuous oil/water separator.
This project was then authorized by Contract No.  14-12-825 "to design,
fabricate, and test a 1  MGD prototype  oil/water separator system".
The effort was divided into two phases.  Phase I comprised design,
building, and testing of a 50 GPM system and a theoretical study to
assist in the design of the 1 MGD prototype.  Phase II comprised
design, fabrication, and testing of the  1 MGD system.  The project
schedule and consideration of equipment and material procurement
timing required that certain key features of the Phase II design be
frozen prior to completion of the Phase I activity.

-------
                            SECTION IV

  PHASE I.  VORTEX SEPARATION PROCESS DEVELOPMENT
The Statement of Work in Contract No.  14-12-825 was limited to
"design, fabricate,  and test a 1 MGD prototype oil/water separator
system",,  The Period of Performance and the Estimated Cost
reflected this limitation,  and no process development effort was
anticipated.  Nonetheless, initial tests of the Phase I, 50 GPM model
demonstrated that the proposed modified hydrocyclonic concept was
severely limited in applicability as an oil/water separator in oil spill
recovery systems.  Accordingly, the Contractor undertook  develop-
ment studies which  successfully defined a simple, effective  oil/water
separation process  of broad applicability.  This section describes
the process development studies conducted on the Phase I system.
Evaluation of the process is traced through four phases:

      Deep vortex skimming process (original concept).

      Shallow vortex overflow process.

      Gravity flow oil recovery from submerged oil vortex.

      Upward vertical suction oil recovery from submerged
      oil vortex.

The deep vortex skimming process reproduces the 2.5 GPM and
50 GPM prototype test models developed by Contractor at his expense
prior to Contract No» 14-12-825.  The process  operates as  follows
(refer to Figure I.):

A mixed oil/water influent is introduced tangentially through inlet
ports (1), at the bottom of an influent annular trough  (2).  The tangen-
tial velocity imparts a rotational motion to the mixture in the trough,
filling the inner tank of the  separator.  Rotational speed is increased
by tangential jets of recirculated water introduced through vortex
generator jets (7)0   The rotational motion and the flow of water out
the bottom through effluent  pipe  (5) cause  formation of a deep vortex.
The apex of the vortex stabilizes at a concentrically  positioned vortex
finder tube (3). Centrifugal force and gravity, acting on the density
differential between the water and oil,  cause the lighter oil  (8) to
seek the surface of the vortex.  The  oil remains at the vortex surface
while flowing  down to the vortex finder tube (3).  The vortex finder
tube then skims the  oil from the vortex  surface,  permitting  the sepa-
rated oil to flow from and be recovered at (4).   Water bypasses the
vortex finder  tube and exits  through gate valve (6) at (5).

-------
 l) Influent Ports
 ?**^k
 2) Influent Annular Trough
^•v
 3 ) Vortex Finder Tube

fA-) Recovered Oil
   Water Effluent

   Gate Valve

T) Vortex Generator Jets

Q) Floating Oil
         Figure I.  Deep Vortex Skimming Process


                              7

-------
Test runs of this deep vortex skimming process demonstrated the
following limitations:

      Recovery of a high percentage of influent oil could be
      achieved only at the expense of high water content (35-90%)
      in the recovered oil.

      If skimming was controlled to '.achieve low water content in
      recovered oil,  excessive oil by-passed the vortex finder tube
      and was not recovered.

      Control of flow  through the vortex finder tube was excessively
      erratic.   This flow is  dependent on liquid level, rotational
      velocity,  vertical  position of the vortex finder tube relative
      to the vortex apex, ripples and other perturbations in the
      vortex surface,  and the deviation of the tank axis  from
      vertical.  These variables are partially interdependent and
      in turn are functions of total flow rate  and vortex  generator
      jet flow rate.

      Influent oil concentrations higher than  about 1% resulted in
      excessive oil loss with the effluent water.

      The process -was extremely sensitive to tilting.  When the
      tank was slightly tilted,  the vortex left the vortex finder
      tube and followed a vertical path into the water effluent end-
      pipe,  resulting in  total loss of separation.  On more extreme
      tilting (5 degrees or more),  the vortex collapsed completely.

These process characteristics were deemed so restrictive of applica-
bility in oil/water separations that  no utility in oil spill recovery
operations could be predicted.  Selected test data illustrating some  of
the above conclusions are summarized in Table 1,  runs  1-5.

Experimental effort was then directed to process modifications which
would ameliorate the most serious  limitations of the deep vortex skim-
ming process.  The major independent variable available  for study was
the vortex generator jet flow rate.  In tests with water only,  it was
shown that the depth of the vortex varied directly with that flow rate,
and that a shallow vortex (depth of 2   4 inches compared  to 18" deep
vortex) formed with no vortex generator jet  flow.   This  shallow vortex
was  little more than a wide, concentric depression in the  rotating
fluid  surface.  Rotation  was caused by the  influent flow, and thus
rotational velocity was proportional to influent flow rate.

This  last relationship was investigated in more detail.  The Phase I
unit  (Figure I) has two influent ports,  each of three square inch cross
sectional  area.  It was observed that the linear tangential velocity at

-------
the surface of the rotating liquid approached the calculated linear
velocity of flow through the influent ports, over a range of total
flow rates from 30 to 60 GPM.  Furthermore, when one of the two
influent ports was blocked, directing the entire flow through one port,
the surface tangential velocity approximately doubled at constant
total flow.  As is true of such rotating fluids,  linear tangential velocity
was essentially constant at all points of  the surface, with the result
that rotational velocity (rpm) varied indirectly with the radius.  In
this type  of rotational flow , a given particle this moves in response
to its imparted tangential velocity and its radial distance from the
center of rotation.  This holds independently of the  size of the tank.

The shallow vortex formed at 50 GPM total flow had a rotational
velocity of about 22 rpm, and a  tangential velocity of 1500 inches/
minutes at radius 10 inches.  At this  relatively low velocity, the surface
was smooth and free of turbulence.  At 60 GPM flow rate,  turbulence
increased and the vortex became deeper (about lOinches), and showed a
tendency to form a  stable air core extending to the end pipe at the
bottom of the tank  (similar to Figure I).

When oil was introduced into a quiescent shallow vortex at 50 GPM flow-
through the tank,  it collected and remained in the vortex at the center.
The vortex finder tube was then elevated so that its upper, open end
•was even with the height of the center of the vortex (Figure II).  A con-
tinuous flow of mixed oil/water  influent  was then introduced.  With
level controlled to provide a  constant overflow of oil into the vortex
finder tube,  effective separation was achieved.  This was called the
shallow vortex overflow process.

The shallow vortex overflow  process is  an improvement over the
deep vortex skimming process,  and has  applicability for separation
of oil/water  mixtures under ideal conditions (constant flow rates and
influent composition).  However, it also has limitations which would
limit its applicability in oil spill recovery systems  where ideal con-
ditions cannot be anticipated.  The key characteristics are:

      Oil recovery  rate is determined by the overflow into
      the vortex finder tube.   The overflow in turn is con-
      trolled  by the height of liquid at the center of the vortex
      relative to the height of the vortex finder tube.   For a
      fixed tube position, the controlling variables are total
      flow rate and liquid level  in the tank.

      Assuming constant total flow rate  and precise liquid
      level control, constant overflow oil recovery is  possible.
      On  the other,  if oil influent rate is variable (as it would
      be under field conditions), recovered oil has a variable
      and uncontrollable water content,  and/or at high oil flows,
      only partial recovery is achieved.

-------
(T)   Vortex Finder Tube




(2-J   Recovered Oil





(3)   Water Effluent





(4)   Floating Oil




      Influent Ports
 Figure II.  Shallow Vortex Overflow Process




                     10

-------
TABLE 1.  SELECTED VORTEX SEPARATOR PROCESS DEVELOPMENT TESTS
INFLUENT, GPM
RUN NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
WATER
50
40
40
45
45
50
37
50
50
50
40
45
45
30
40
40
40
50
50
OIL
0.
1.
1.
0.
0.
4.
3.
2.
1.
2.
0.
0.
0.
0.
3.
0.
0.
3.
6.
25
0
0
7
7
2
7
0
0
0
. 1
,5
,5
,5
0
05
5
0
0
% OIL
0.
2,
2.
1.
1.
8.
10.
4.
2.
4.
0.
1.
1.
1.
7.
0.
1.
6.
12.
5
5
5
5
5
5
0
0
0
0
25
1
1
7
5
1
1
0
0
RECOVERED OIL
FLOW, GPM
2.
1.
1.
0.
1.
4.
3.
2.
1.
2.
0.
0.
0.
0,
3.
0.
0.
3.
6.
0
0
5
25
5
0
5
.0
,0
0
. 15
.5
.5
,6 +
, 0+
,05
.5
,5
0
RECOVERED OIL,
% WATER
90
5
35
0
55
4
20
4
30
20
35
2-
2-
10
5
10
2-
5
2-
% OF INFLUENT
OIL RECOVERED COMMENT
80
95
97
36
96
91
75
96
70
80
97
98+
98+
98+
98+
90
98+
98+
98+
10 GPM Recirculation, deep vortex skimming.
35
35
15
15
No recirculation. Shallow overflow vortex.





Submerged vortex tube, gravity recovery.



intermittent recovery.
Vertical upward suction oil recovery.



-------
      A control system to sense influent composition and total
      flow rate and adjust vortex finder tube position continuously
      to provide variable oil recovery rate in  response to variable
      oil influent rate is feasible.  However,  such a system would
      be complex and expensive.

Selected data illustrating these characteristics of the shallow vortex
overflow process are shown in Table  1,  runs 6 -  10.

In the course of experimentation with the shallow vortex overflow
process,  it was observed that  at times a pool  of oil formed and
remained around the vortex finder tube.  At such times,  the com-
position of recovered oil approached 100% oil.  The pool could be
eliminated by raising the liquid level  to increase the oil recovery
flow rate, so that recovery rate exceeded influent rate.   This
observation led to a hypothesis,  which was confirmed by experiment.

The hypothesis assumed that in a rotating oil/water mixture, floating
oil would migrate to and remain in a "pocket" at  the surface and
center of the liquid.  A simple laboratory test was  run to test the
hypothesis.  About 750 ml of a 5% oil/water mixture was placed in a
one liter glass beaker, and allowed to stand at rest.  Of course, the
oil floated in a uniform layer on top of the water.  A magnetic stirring
bar -was added,  and the mixture was  stirred slowly.  The oil immediately
moved to the center of rotation,  and formed a largely submerged pocket
of oil.  The photographs in Figure III show clearly what happened „   It
was further demonstrated that the diameter of the "submerged oil
vortex" decreased and the depth accordingly increased as the rotational
speed was increased.  The submerged oil vortex forms as a result  of
the minor centrifugal force present acting on the density differential
between the oil and the water.

It was reasoned that the submerged oil vortex could be utilized as a
reservoir from which to recover nearly water-free oil.   The Phase I
vortex separator was further modified to test this concept.

The modified setup is shown schematically in  Figure IV. The vortex-
finder tube was lowered so that its open end was  three inches below the
surface of the shallow vortex.   An oil recovery flow control valve was
installed at (3). A mixed oil /water  influent was pumped to  the separator
at 50 GPM,  and liquid level control was established by manual  setting
of the valve in the water effluent line  (5).  No  oil was removed until a sub
merged oil vortex of about  3-5 gallons was accumulated.  Then the oil
flow control valve (3) was  regulated to maintain a submerged oil vortex
of about that volume.  Recovered oil flowing out at  (4) contained con-
sistently low quantities of water.  The size  of the submerged oil vortex
was not critical as long as  it was large  enough to surround the open

                                12

-------
Figure III.  Formation of "Submerged Oil Vortex"
                    13

-------
end of the vortex finder tube with a pocket of oil.  In fact, oil recovery
could be stopped to permit the submerged oil volume to increase,  and
then oil could be withdrawn rapidly until the pocket was so small that
entrained water appeared in the recovered oil.

Obviously, this process eliminates some of the severe control  limitations
noted for the deep vortex skimming process and the shallow vortex over-
flow process.  The improvement  essentially stems  from the fact that
the two earlier processes require plug flow of oil through the system,
in turn requiring constant control of oil recovery  flow in response to
variable oil influx rate.  The submerged  oil vortex comprises an oil
"surge chamber", which eliminates plug  flow of oil through the system
and, therefore, eliminates the need for constant oil recovery flow rate
control.

Critically of all process variables is  reduced by this essential  change.
Because the physical size of the submerged oil vortex is not critical
as long as the  oil intake is surrounded with oil, level control is less
critical,,  Rotational speed, which varies  with total influent flow rate,
is not  critical.  This system is also markedly  less sensitive to tilting
(to be  discussed in more detail later). Data  illustrating the perform-
ance capability of the gravity flow oil recovery from submerged oil
vortex process are shown in Table I,  runs 12 - 16.

Additional tests with this process  setup (Figure IV)  demonstrated that
oil concentrations as low as 0.1% in the influent could be recovered,
by the technique of intermittent recovered oil flow.   The very low rate
of oil influx (0.05 GPM) was allowed to accumulate in the  submerged
oil vortex between withdrawals of 0.5 gallons every 10  minutes  (Table
I,  run 16).

Attempts to demonstrate recovery of oil from mixtures containing
10% oil showed a limitation of this gravity flow process.  While intro-
ducing an influent flow  of 45 GPM water and 5 GPM  oil, removal of
recovered oil at 5 GPM was not possible.  The reason  for this  limitation
is that the flow of oil into the vortex tube  at such high rates resulted
in formation of a normal,  air-cored "bathtub"  vortex,  feeding into
the vortex tube.  The air core increases  in size with increasing flow
rate,  and effectively starves the flow into the tube.   Tests showed
the upper limit of recoverable oil flow on the Phase I equipment to be
about 3 GPM, or 6% oil in a total influent flow  of 50 GPM.

One additional modification was made to eliminate this limit on the
oil concentration range.  The vortex tube was removed, and in its
place,  a vertical upward suction oil recovery tube was installed.
Refer  to Figure V.  The oil suction intake (3) comprised a bell
immersed in the submerged oil vortex (2).  The bell was connected

                                14

-------
                Floating Oil





                Submerged Oil Vortex





          ()   Oil Flow Control Valve




           4    Recovered Oil




                Water Effluent




                Influent Ports
Figure IV.  Gravity Flow Oil Recovery From Submerged Oil Vortex




                           15

-------
by piping (4) to an oil recovery pump.  This configuration eliminated
the formation of an air-cored vortex  at high oil flow rates,  while
retaining all the favorable attributes of the submerged oil vortex
process „  Recovery of as high as 7.5 GPM of oil from a 50 GPM,
15% oil influent -was achieved continuously, with very low  (less than
2%) water in the recovered oil.  Additional data from selected runs
in this configuration are shown in Table I,  runs 17  - 19.

The foregoing discussion of the development  of the submerged oil
vortex separation process is largely qualitative,  to permit a clear
understanding of the principles involved (See also Section  VII).   The
objectives  sought in the development studies  were:

      A process which rapidly separates floating oil/water
      mixtures, using simple equipment and  characterized
      by simple,  non-critical process  control.  These
      attributes were deemed prerequisite for applicability
      in  oil spill recovery systems  for use under field con-
      ditions.

      A recovered oil characterized by low water content
      (less than 5%) to minimize required at  sea  storage
      capacity.

      A separated -water effluent containing minimum residual
      oil, suitable for direct return to sea without  further
      treatment or storage.

Progress toward and achievement of the first two objectives are
detailed  in the  preceeding discussions.  Quantitative measurement
of performance with regard to the third objective was deferred until
the first two were achieved, for two reasons.  First, the  large
number of test runs required (in excess of 150 runs were made)  could
not have been accomplished in the time available if quantitative analyses
had been performed in each case.  Secondly, it was a practical nec-
essity with available facilities to recycle water and oil through the
system from and to storage reservoirs.  A constant supply of fresh
water at 50 GPM was not available, and no means to dispose of large
quantities of oily waste water existed.   The Phase  I system  is shown
schematically in Figure VII.  An inescapable consequence of the continual
recycle of water  was a build-up of emulsified oil in the water supply.
Consequently,  quantitiative data on effluent water  quality could not be
obtained up to this point of the development.  After the process was
optimized, the entire  system was cleaned and prepared for a typical
run in the preferred configuration with "once through" water and oil
flows. A pumper truck of clean water was provided by the Panama City
Fire Department, permitting a continuous  run of 30 minutes at 40 GPM.

                                16

-------
       Floating Oil

       Submerged Oil Vortex

       Oil Suction Intake

       Oil to Recovered Oil Pump

       Water Effluent

       Influent Ports
Figure V.  Upward Vertical Suction Oil Recovery
           From Submerged Oil Vortex
                    17

-------
Figure VI.  Phase I (50 GPM) Vortex Separator
                  18

-------
    Water Reservoir
(2) Moyno Influent Pump
(3) Pump Bypass
(4-) Vortex Separator
Oil Suction Intake
Oil Recovery Pump
    Reservoir
Oil Pump
     Oil Flowmeter
     Water Effluent
     Water Return Pump
J2)  Water Flowmeter
     Gate Valves
               Figure VII.  Phase I Experimental Vortex Separator System

-------
An oil feed rate of 0.5 GPM (1.25%) was selected for the test, and
oil was continuously recovered at about 0.5 GPM.  Four water samples
were taken during the test, and analyzed for oil content. The samples
contained 0-10 ppm,  10 -50 ppm,  50 - 100 ppm, and 0-10 ppm, respec-
tively,  showing that water effluent of excellent quality was achieved.
Control samples collected before oil addition and after recovery of all
the oil had 0-10 ppm oil by analysis.  Analysis was by colorimetric
matching  of trichloroethylene extracts to  standard  comparison samples
(See Appendix E)0  Oil samples taken during the  run were consistently
less than  5% in water content.  Oil recovery exceeded 99.6% of oil
influent.

The effect of tilting the separator on process efficiency was studied,
to permit prediction of operability at sea.  The separator had been
designed to  permit rocking back and forth on trunnions,  so that the
effect of wave motion on a shipboard separator could be estimated.
It was quickly shown that even slight tilting had a major negative
effect on the deep vortex skimming and shallow vortex overflow
processes.  Complete loss of control resulted, with oil losses to
the water effluent,  and excessive water in the recovered oil fraction.

Tests •with the  submerged oil vortex process (gravity recovery through
a submerged tube)  showed that this preferred method is much less
sensitive  to tilting.  The data in Table 2 show that  tilting back and
forth 7.5  degrees from vertical every four seconds does affect sepa-
ration efficiency.  Run #1 and #2 were identical,  except  that the volume of oil
in the submerged oil vortex was increased from two gallons to five gallons,
largely overcoming the effect of tilting.  At higher  oil flow and with a sub-
merged oil vortex volume of three gallons in run #3,  slight effects  of
tilting 7.5 degrees were seen.  In all these and other tests,  tilting  15
degrees from vertical did have a serious negative effect on efficiency.
Accordingly, the Phase  II system was designed with a simple  gimbal sup-
port stand to permit the separator to  remain essentially vertical despite
ship's motion.

The  crude oil used in the Phase  I studies was supplied by the Chevron
Oil Company.  Two grades, representative of a medium low and medium
high density crude, were used.  They were:

      Sample #1:  Reedy Creek Field,  Laurel,  Mississippi.
                 Well depth 11,400 feet sand.  Water cut 0.3%.
                  Gravity 25 degrees API (specific gravity 0.905).

      Sample #2:  Raleigh Field, Raleigh, Mississippi,
                  Well depth 11, 750  feet sand.   Water cut 0.4%.
                  Gravity 41.1 degrees API (specific  gravity 0. 820).

                                20

-------
      TABLE 2.  EFFECT OF TILTING ON
SUBMERGED OIL VORTEX SEPARATION PROCESS
                              %  Water in
 Water Influent  Oil Influent  Oil Effluent   Tilting
1. a.
b.
c.
2. a.
b.
c .
3. a.
b.
c .
45 GPM
45
45
45
45
45
40
40
40
0.
0.
0.
0.
0.
0.
3.
3.
3.
5 GPM
5
5
5
5
5
0
0
0
2
59
81
2
2
80
5
7
42
None
7.5° back and
15° one way
None
7. 5° back and
i r O
1 5 one way
None
7.5° back and
i r O
15 one way

forth


forth


forth

                          21

-------
Performance  of the vortex separator over the enitre range of opera-
ting conditions investigated was not affected by the crude oil sample
used0  Thus,  at least over the range  of specific gravity 0.82 to 0.905,
oil density and viscosity are not critical variables in this recovery
process.

Both crude oil samples showed some propensity for formation of
water-in-oil emulsions „  As oil was recycled through the system
during the course of experimental investigations,  it increased in
viscosity due  to loss of volatile components and due to formation of
such water-in-oil emulsions.  Recovery of floating "water-in-oil
emulsions was repeatedly demonstrated with the submerged oil
vortex process, again over the full operating range discussed in this
section.   Recovery efficiency of a thixotropic, viscous emulsion of
composition 70% Sample #1 crude oil  and 30% water was identical to
that of fresh crude oil.  The emulsion had the consistency of mayon-
naise and a specific gravity of 00930  The  single operating variable
affected was,  of c.ourse,  power required to pump  the recovered
emulsion. For this reason, the oil recovery pump should be a
positive displacement pump adequately powered to handle a  wide range
of viscosities „

The vortex separator process  is designed  to and does recover  float-
ing oils from  water/oil mixtures.  It will not efficiently separate oil
from permanent oil-in-water emulsions,,  Efficiency of recovery from
temporary disperions will vary inversely with the degree of mixing
or the  extent of oil dispersion present in very fine droplets.  Some
investigations were carried out to provide semi-quantitative information
on this relationship.

In each test, the Phase I separator system was thoroughly cleaned,  and
placed in operation with a clear water influent. Oil was then intro-
duced by various methods which caused varying degrees of oil disper-
sion in the influent.  The degree of dispersion was estimated by sampling
mixed  influent as  it entered the separator.  Droplet size and time
required for the oil in the mixture to float completely to the  surface of
the sample were observed.  Effluent  water samples were collected and
analyzed for oil content.   The results are  summarized in Table 3.   They
illustrate the  importance of an ifluent transport system which •will not
increase the degree of dispersion of oil in water encountered at the  site
of a spill. Low shear, nonemulsifying pumps  are required,  and overall
process efficiency can be  improved by providing even momentary pre-
settling so that minimum dispersed oil enters  the separator.  The data
also  show that for a given separator and influent system,  efficiency
varies inversely with water flow rate.  This is true for all gravity sepa-
rators and comprises one practical limit on the capacity of  any given
size vortex separator.
                                22

-------
            TABLE 3.  EFFECT OF OIL DISPERSION ON
                VORTEX SEPARATOR EFFICIENCY
    Water Influent     Oil Droplet     Oil Flotation       Water  Effluent
    	GPM         Diameter,  in.     Time, Min.          ppm Oil


1.   Water and oil mixed by pumping through 1L10H Moyno pump.

          50           1   to  J_        50%  -  1  min.        1000 +
                      128     64~        80%  -  2  min.
                                       100%    10  min.

2.   Oil added through tee in water influent hose to tank.

          50          —   t0  —       100%     1  min.        500 - 1000
                      64      32

          40          J_   to  J_       100o7c  _  l_min,        250
                      64      16                3

3.   Oil/water influent to tank by gravity flow in a  trough.

          50          —               100%  -  1  min.        500 - 1000
                      64

4.   Oil added directly to water  surface in influent trough.

          50          I  to I          100%  -  I min.       35
                      84                   5

          40          1  to 1          100%     Imin.        10   50
                      84                   5
                                   23

-------
                           SECTION V

          PHASE II FULL SCALE PROTOTYPE DESIGN,
            CONSTRUCTION, AND TEST FACILITIES
Design of the full-scale prototype -was based on the initial Phase I
experimental results and on a theoretical analysis of the deep vortex
skimming process by the Contractor's sonsultant, Dr.  A. B.  Troesch
of the University of Southern California.  Unfortunately,  the Contract
Schedule and procurement delivery times required that key elements
of the Phase II system  design be frozen before process development
studies under  Phase I were well underway.  The main consequence
of this situation was that  scaling factors derived in the theoretical
study were used to determine major dimersions  of the Phase II vortex
separator.  Later studies which elucidated a different and preferred
vortex separation process also showed that  different scaling factors
apply (see p0 43)0  As a result, the practical influent flow capacity
of the Phase II system  is about 50% of the Contract objective  of 1 MGD.

The Phase II vortex separator inner tank was built with  diameter 60
inches and height 31 inches.  Our present knowledge of the submerged oil
vortex separation process suggests that the corresponding dimensions
for a 1 MGD separator  should be 75 inches and 65 inches,  respectively.
Influent degree of mixing was recognized as a critical factor  in overall
separator efficiency.  An effort to reduce the amount of dispersed oil
entering the tank was made by introducing the influent flow into a
plenum  chamber underlying the separator tank itself.  Flow was then
directed tangentially out  of the  plenum and into the influent annular
trough surrounding the separation tank.  This design provided both
time (10   20 sec) and metal surface contact to permit coalescence of
fine droplets into larger  drops  which would  float and be  separated more
readily.

The Phase I tests showed that the vortex separator does function even
when tilted from vertical.  However, reliability, control, and effi-
ciency are reduced under such  conditions.   Accordingly, the  Phase  II
unit was  designed for support on a gimbal stand.  It is thus isolated
from pitching  and rolling motion and can be  tested at sea on its
present mounting.  Flexible hose connections were also provided to
permit relative motion between the tank and fixed pumps on shipboard.

The Phase II tank was of welded construction, fabricated from 6061-TG
aluminum.  The tank and the gimbal support stand were fabricated for
the Contractor by Metric Systems,  Inc., Fort Walton Beach,   Florida.

The influent pump for the Phase II system was an SWG88 Moyno pump
manufactured  by Robbins and Myers. It was powered by a 25 hp,

                                 24

-------
440 V T.  E.  F. C. motor.  Contactors and start-stop controls were
Class I,  Division 2, Group D, for use in hazardous environment.  The
oil recovery pump was a 1 L10 Moyno pump with a 5 hp,  440 V T. E.
F. C. motor.

Operation of the submerged oil vortex separation process requires
maintenance of a  stable liquid level in the separator tank, so that the
position  of the oil suction intake relative to the submerged oil vortex
will remain constant.   A level control loop  was designed for the
Phase II system,  comprising a Foxboro Type 13A differential pres-
sure cell sensor, a Foxboro Type 43A A4 reset controller, and a
Foxboro Model M3L pneumatically actuated 6 inches butterfly valve.
The pressure tap for the d/p cell was placed at the bottom of the
separator tank.  The butterfly valve controlled effluent flow to main-
tain tank level at controller set-point.  Control at any point over a
ZO inch range was provided.

A  test site for Phase II testing and evaluation was provided by the
Standard Oil Company  Refinery at Pascagoula, Mississippi.  A
sketch of the test equipment arrangement is shown in Figure  VIIL
In addition to providing the site, Standard Oil assisted by providing
•water piping and hose,  electrical power, rigging service for  place-
ment and removal of equipment, crude oil,  recovered oil receiving
tanks, and a pump for  influent  oil.  The Refinery  Laboratory also
performed effluent analyses for oil-in-water and water-in-oil.

The Phase II system was  set up so that there was no risk of contami-
nating natural waters  as a result of the test program.
                                 25

-------
IV
                      See key,  opposite page.
                           Figure VIII.  Phase II Vortex Separator Test Installation at Refinery

-------
                       Key to Figure VIII








(1)    Tidal Salt Water Canal





(2)    Pond





(3)    Strainer Intake





(4)    Moyno SWG88 Influent Pump





(5)    Pump Bypass





(6)    Oil Supply (by Refinery)





(7)    Oil Pump (by Refinery)





(8)    Gate Valve Control





(9)    Oil Flowmeter





(10)   Vortex Separator





(11)   Oil Recovery Pump





(12)   Pump Bypass





(13)   Sample Tap





(14)   Recovered Oil  Tank (by Refinery)





(15)   Level Control System





(16)   Automatic Butterfly Valve





(17)   Water Effluent Trough





(18)   A.P.I.  Separator (Refinery)





(19)   Flexible Hose
                                27

-------
Figure IX.  Phase II Full Scale Prototype Vortex Separator
                     Top View
                    28

-------
Figure X.  Phase II Full Scale Prototype Vortex Separator
                Side View
                    29

-------
                           SECTION VI

       PHASE II.  FULL-SCALE PROTOTYPE VORTEX
 SEPARATOR PROCESS DEMONSTRATION AND EVALUATION
The Test installation is  shown schematically in Figure VIII.

Salt water influent from a tidal canal was supplied to the SWG88
Moyno influent pump through a  strainer intake.  A manually controlled
butterfly valve in the bypass piping from pump discharge to intake was
used to control net discharge flow to the separator.   The separator
was located about 200 feet from the influent pump, at the end of an
A.  P.  I. separator.

Oil was injected  into the influent water stream through a tee in the
line 15 feet ahead of the  separator. Oil flow was controlled by a manual
gate valve and measured by a Foxboro d/p cell and orifice  plate flow-
meter.   The crude oil was  supplied by tank truck by the Refinery,  and
pumped by a centrifugal oil pump furnished by the Refinery.

Recovered oil was pumped out of the submerged oil vortex by the 1L10
Moyno pump. Net oil recovery rate was controlled by a manually
operated plug valve  in the bypass piping on the pump.

After all equipment  had been installed and checked out for proper
functionality, a period of qualitative testing was undertaken to estab-
lish optimum operating conditions.

At full influent pump output of 700 GPM (1 MGD), excessive rotational
velocity in the separator  tank resulted in  formation of a deep vortex,
and no recovery  of oil was possible (see discussion of scaling factors,
Section VII,  p. 43)0  The influent flow was reduced stepwise until a
controllable  and  stable submerged oil vortex could be generated.  This
occurred at a flow of 300   325 GPM.  Application of the proper scaling
relationship  between diameters and optimum flows in phase I and
Phase II equipment gave a calculated capacity of 360 GPM.   The experi-
mental optimum was less because the tank depth was  also improperly
scaled.  It is the Contractor's opinion that the Phase II unit capacity
can be increased to  360 GPM or slightly more, by increasing the tank
depth by about 30 inches.

Additional tests were run to establish the preferred liquid level and
positioning of the oil recovery suction intake.   Liquid level was estab-
lished at 0.5 inches below the top edge of the  inner tank. The preferred
position of the bottom  of the suction intake was 4 inches below the liquid
level,

                                30

-------
The flow in the influent trough -was of substantially lower linear
velocity than in the Phase I system (approx.  200 inches per min com-
pared to 1ZOO  - 1500 inches per min).  This resulted from a combina-
tion of two factors.  First, inlets were designed to provide the desired
velocity at 700 GPM, and the system was operable at only half that rate.
Secondly, the design of the baffles directing  influent flow from the
plenum tangentially into the influent trough was inadequate.  Because
these baffles were too short, the flow was actually about 30 degrees
from the desired tangential direction.   This  resulted in dissipation of
tangential velocity in the influent trough.

With the exception of the total capacity and influent baffles  discussed
above,  the entire  vortex separator  system performed according to
designed objectives.  The operation was exactly as expected from that
in the smaller Phase I unit.

A  series of quantitiative tests was then undertaken to provide  reliable
data on the  efficiency of the Phase II vortex  separator,   operating at
the preferred conditions  established by the qualitative  tests.   As each
set of test flows was set up, the system achieved stable level  control
and performance in 1 - 2 minutes.  Stable operation was maintained for
20   30 minutes before samples for analysis were taken.  Frequent
sampling during this period permitted qualitative assurance that
operation was stable and that the analytical samples were representative
of the separator's performance.  Data are summarized in Table 4.
Data describing the crude oil and the salt water used in these  tests are
included in the Appendix G,  and were furnished by Standard Oil.  Test
Numbers 3, 4, and 5 were witnessed by the  Project Officer.
                                 31

-------
     TABLE 4.  PHASE II VORTEX SEPARATOR OIL RECOVERY
                             EFFICIENCY
TEST NO.
% Oil Recovered(1)               97.4    96.9   99.7   99.8   99.9
Water Influent,  GPM             325      250    280     280    280


Oil Influent,  GPM                   3      7.5      10      20     30


Total Influent,  GPM              328     257.5   290     300    310


% Oil in Influent                 0.9       3     3.4     6.7    9.7


                            (2)
Water Effluent, ppm  total oil     235      980    106     111     63


Recovered oil, % Water           5(3)     2(3)   0.7(2)  1.0(2)    1(3)
(1)  Calculated by the materials balance equation;
                                                  ppm oil in
          % Recovered = 100   influent % water  x water  effluent
                               influent % oil            j Q^

(2)  Analysis by Refinery Laboratory

(3)  Approximate value
                                   32

-------
                           SECTION VII
                           DISCUSSION
Analysis of all observations and tests performed in Phase I and
Phase II of the project effort leads to the following general conclusions
regarding the Contractor's submerged  oil vortex oil/water  separation
process:

      The oil/water vortex separation process described in this
      report is reproducible,  simple in equipment construction
      and operation,  and is applicable as an itegral part of sweeping-
      skimming-separation-storage oil recovery systems for treating
      oil spills in natural waters.  It offers additional applicability
      in oil pollution control applications requiring removal of gross
      floating oils from large  volumes  of water.  Such applications
      may include refinery waste or process waters, oily animal or
      vegetable waste waters, and  storm runoff from  such oil-laden
      areas as airports.  Treatment of marine  ballast and bilge -waters
      may also be effective.

      Work under this Contract has demonstrated that the vortex
      separation process can provide continuous recovery of 99.5%
      or more  of the floating oil from an  influent stream  containing
      5 - 15% of oil.  More dilute influents down to 1% oil content
      can be treated  to recover 99% of the oil.  Efficiency as
      measured by percent recovery falls off with  still more dilute
      influents, to about 90% for a  0.1% mixture of oil in water
      (see  Appendix F)0

      Water  effluent  from the  vortex separator  treating an
      influent of floating oil and water will contain well under
      200 ppm  oil with proper operation.  Emulsified oils or
      extensively  dispersed oils will not  be efficiently separated.
      Oil recovered can be controlled in composition to the range
      of 0  - 5% -water content.

      Process  control comprising a simple pneumatic level control
      loop  was demonstrated.   For simplest field operations,
      additional instrumentation to provide automatic control of
      rate  of oil removal from the  separator is recommended.

      Many of the operating characteristics of this vortex separa-
      tion process are shared by other  "instantaneous" oil separa
      tion process.   Efficiency of such  processes is highly dependent
      on degree of emulsification of oil in the influent and on oil
      concentration in the influent.  This process involves simple
      equipment and  controls, provides high space/time processing
     capacity,  and is not significantly affected by oil densities or
                                33

-------
      viscosities over ranges most commonly encountered.

      A simple influent plenum chamber to facilitate oil droplet
      coalescence and thus improve efficiency by reducing oil
      dispersion was effective.  This  result can logically be
      extended to larger pre-coalescence tanks to provide further
      improved efficiencies.

      A gimbal support stand for the vortex separator tank serves
      to isolate the separator process from pitching and rolling
      motion at sea,  thus overcoming undesirable effects of
      excessive tilting on separator performance.

      The submerged oil vortex separation process has been
      adequately defined to permit reliable design and operation
      of units ranging from 50 GPM to 1000 GPM in nominal
      influent capacity.

      The vortex separation process equipment has no moving
      parts  and no power requirements  except the pumps for
      supplying influent and removing oil to storage.  Water
      effluent may be discharged by pumps or by gravity flow.
      This leads to favorably low initial investment requirements
      and operating costs.

      The simplicity  of the vortex separation process equipment
      will facilitate development of air-transportable or air-
      deployable models for rapid applicability in spills anywhere
      within a large area.   This can reduce the number of units
      necessary to protect the extensive inland and coastal waters
      of the United States.

A summary of the principles of operation and key design parameters
of the submerged oil  vortex separation process, as elucidated by the
studies so far discussed, follows to provide a clear  understanding of
the  simplicity and applicability of the  process„

The vortex  separator utilizes the force  of gravity and centrifugal
force operating on the density differential between oil and -water.
The resultant of these forces causes floating oil in an oil/water
mixture to collect in  a "submerged oil vortex",  which forms  at and
below the surface of the liquid rotating in a cylindrical tank,  and is
coaxial with the tank. The submerged oil vortex provides a convenient
pocket of oil from which oil is recovered by pumping through a vertical
upward suction intake with discharge  to storage.  Clear water effluent
                                34

-------
passes out a bottom pipe,  coaxial with the tank, and may be returned
to sea or otherwise disposed of by gravity flow or pumping.

Rotary motion of the oil/water mixture in the separator tank is generated
by introducing the influent in a tangential direction through two or
more inlet ports into an annular influent trough surrounding the
separator tank.  The rotational velocity is primarily a function of
tangential velocity of the influent flow,  which in turn is determined
by the flow rate and the cross-sectional area of inlet ports.  Flow
velocities should be in the range  of 1000 - ZOOO inches per minute.

Flotation (gravity) provides the main separation force, while the
small centrifugal force is all that is required to cause formation of
the submerged oil vortex.  The submerged oil vortex comprises an
effective surge tank, which  obviates the need for precise continuous
control of oil recovery rate in response to varying oil influent rates.

Separation efficiency is dependent on rise rate of the oil droplets
present in the influent mixture as compared to the vertical downflow
rate of separated water through the tank.  Accordingly,  mixing or
dispersion of oil into the water should be minimized in the influent
handling system.  Also, because water downflow rate is a function
of the cross-sectional area  of the tank, the scaling  law to be used
in designing  tanks of various diameters,  d, for various capacities,
Q,  is:
The average -water downflow rate should be about 3 feet per minute to
provide efficient separation of oil droplets 1/40 inch in diameter or
larger.  The annular influent trough increases separation efficiency
because all elements of the oil/water mixture are required to closely
approach the liquid surface when overflowing from the trough into the
separation tank.  Hence,  the rise distance required for each oil
droplet is minimal.
                                 35

-------
                           SECTION VIII
                      ACKNOWLEDGEMENTS
The cooperation and support of the Standard Oil Company Refinery
at Pascagoula, Mississippi, Mr.  Thron Riggs, Manager, was vital
to our  successful Phase II evaluation program.  They provided a
site for testing without risk of polluting natural waters, and facili-
tated the entire effort  with  support services, power, crude oil supply,
and analytical services.  Their assistance is gratefully acknowledged.

Metric Systems,  Inc.,  Fort Walton Beach, Florida, fabricated both
Phase  I and Phase II vortex separator tanks.  In addition to delivering
first-class workmanship against a very demanding time schedule,
their interest and constructive suggestions were of value to the Project
and are sincerely acknowledged.

The City of Panama City Fire Department assisted by providing a
tank truck of water for a Phase I experimental run.
                                36

-------
                           SECTION IX
                         REFERENCES
1.  Oceanology International,  March,  1970,  p0  18.  News release
   describing the Seadragon oil spill recovery  system being
   developed by the Garrett Corporation,  System includes a high-
   speed centrifugal oil/water separator.

2.  Offshore Technology, June, 1970,  p. 77, News release reporting
   demonstration by Reynolds Submarine  Services Corporation of a
   "Voraxial  Flow" open-center pump for high-speed liquid-liquid
   separations.

3.  Chemical and Engineering News, July  6, 1970, p. 62.  News
   release describing a mechanically driven vortex  separator  for
   oil spill recovery,  being developed by  Bertin  &: Cie,  Paris,
   France, for a French oil company, Elf-Erap0
                                37

-------
                             SECTION X
                 PATENTS AND PUBLICATIONS
No publication have resulted to date from this project.  Patentability
of the  submerged oil vortex separation process should be investigated
4~ r\ rl o+~ OT* T"n T n o Q T*T\T> r\r\r T a f~^ Ti iTviry ^^f-ir^n
              tj                j.
to determine appropriate filing action
                                 38

-------
                           SECTION XI
                            GLOSSARY
Annular influent trough - an annular trough surrounding the upper
portion of a vortex separator tank.  Mixed influent is  introduced at
appropriate velocity into the trough in a tangential direction, resulting
in a rotational flow in the trough which then overflows into the sepa-
rator tank.

Effluent quality - a)  total oil content of water effluent expressed in
ppm.   b)  total water content of recovered oil expressed in percent.

Moyno pumps - positive displacement,  low-shear, progressive cavity
pumps manufactured by Robbins and Myers,  Springfield,  Ohio.

Oil spill recovery system  a total system for recovery of spilled
oil.  A typical system would include unit operations of sweeping and/
or containment; skimming; oil/water separation; and recovered oil
storage.

Separator efficiency  - a measure of the performance of an oil/water
separator, expressed in  percent Recovery of total influent oil.  Efficiency
is calculated from the materials balance equation:

      Percent Recovery =100 - % water in influent x ppm oil in water effluent
                               % oil in influent             10

Submerged oil vortex - the pocket of oil which forms at and be.low the
surface of a rotating mixture of oil and water, coaxially with the axis
of rotation.

Vortex finder tube - in a  vortex separator, a tube which intercepts
the apex of the vortex and consequently removes components of a
vortexing mixture which  concentrate at the apex.

Vortex separator - a device which utilizes combined forces of gravity
and centrifugal force to effect separation of a mixture into components
of differing  specific gravity.
                                39

-------
                          SECTION XII




                          APPENDICES







                                                            Page





A,    Phase I System Bill Of Materials                        41




B.    Phase II System Bill Of Materials                       42




C.    Phase I Separator Assembly Drawings                   43




D.    Phase II Separator Assembly Drawings 1 thru 4          44




E.    Analytical Methods                                      48




F.    Separator Efficiency Calculations                        52




G.    Phase II:  Specifications of Oil and Salt Water            53
                                40

-------
                            APPENDIX A

                   PHASE I BILL OF MATERIALS


 1.   Welded aluminum vortex separator tank.

 2.   Support stand for vortex separator tank.

 30   Water supply tank,  220 gallon capacity.

 4.   Moyno  1 LI OH pump with 7. 5 hp motor.

 5.   Deming centrifugal water pump with 2 hp motor.

 6.   Oberdorfer Model HRCC-1  oil pump with 0.5 hp motor.

 7.   Electrical power contactors and motor start-stop controls.

 8.   Water flow meter:  Foxboro 13A1 D/P  cell transmitter,
     2 inch threaded orifice flange union #34052, stainless steel
     orifice plate, Model MR receiver gauge.

 9.   Oil flow meter:  Foxboro 13A1 D/P cell transmitter,  1 inch
     threaded orifice flange union #33891, stainless  steel orifice
     plate, Model MR receiver gauge.

10.   Compressed air supply to instruments,  with air pressure
     regulator „

11.   Valves, fittings,  pipe, flexible hose.

12.   Manual oil recovery pump.

13.   Oil drum oil supply tank.
                                 41

-------
                            APPENDIX B

              PHASE II SYSTEM BILL OF MATERIALS


 1.  Welded aluminum vortex separator tank.

 2.  Gimbal support stand for separator tank.

 3.  Moyno SWG88 influent pump with 25 hp,  440 V TEFC motor.

 4.  Moyno 1L10 oil recovery pump with 5 hp, 440 V TEFC motor.

 5.  Class I, Group D,  Div0  2  electrical power contactors and motor
     starters.

 6.  Foxboro orifice plate and  D/P celloil flowmeter (recalibrated
     from Phase I system).

 7.  Foxboro liquid level control loop, including:  13A1 D/P cell
     transmitter,  Model 43A receiver controller,  and Type M3L
     6 inch butterfly control  valve.

 8.  Screen strainer for sea water intake.

 9.  Compressed air for instruments, with air pressure regulators.

10.  Water effluent trough.

11.  Valves,  pipe,  fittings, flexible hose.
                                 42

-------
APPENDIX C.  PHASE I SEPARATOR ASSEMBLY DRAWINGS

-------
PL AC ELS,
                                            H

                                            O

                                            X

                                            ti
                                            *


                                            •ti
                                            en
                                            H
                                            en
                                            H
                                            H
                                            i
                                            en
                                            en
                                            H
                                            O
                                            JO
                                            O

-------
APPENDIX D.  PHASE II SEPARATOR ASSEMBLY DRAWING 2
                           45

-------
APPENDIX D.  PHASE II SEPARATOR ASSEMBLY DRAWING 3
                             46

-------
APPENDIX D.  PHASE II SEPARATOR ASSEMBLY DRAWING 4
 0_©
                                                    00
                                                          (fl
                               47

-------
                          APPENDIX E
                     ANALYTICAL METHODS
Water  effluent analyses in Phase I studies were performed by the
Contractor.  Standard samples of crude oil in trichloroethylene
were prepared at concentrations (by volume) of 0, 10, 50,  100,  250,
500, and 1000 ppm.  Water effluent samples were collected and
extracted with three portions of trichloroethylene.  The extracts
were combined and volume made up to equal the volume of the
original water sample.  Colorimetric matching of extracts to
known  standards was done visually, using equal volumes in test
tubes,  viewed in white fluorescent light against a white background.
Three  to five observers  compared each set of samples, and each
set was compared two or three times.  Reproducibility between
observations and between observers was virtually 100%.   Samples
were rated as being between two standar concentrations; e.g. ,
50-100 ppm.  Fresh  standards were prepared for each analysis,
using crude oil removed from the test in progress to  eliminate any
error due to aging,  evaporation, etc.   Samples -were  analyzed within
two hours, since color changes  occur in standards and extracts
within  48 hours.

Recovered oil samples were analyzed in Phase I  studies by sampling
directly into a 1000  ml graduated cylinder.  The  sample was allowed
to settle for a minimum  of two hours. The volume of  water and total
sample volume were observed,  and the result reported as  percent
water in oil by volume.  These analytical methods were more than
adequate to guide the Phase I process development studies.

Phase  II analyses were performed by  the Standard Oil Refinery Lab-
oratory using Standard's routine analytical methods for total oil  in
water (ppm) and water in oil (percent).

Standard Oil uses the American Petroleum Institute Infrared Spectro-
photometric Method for effluent analyses for total volatile  and non-
volatile oil in water.  The procedure follows:

SCOPE

This method describes a nonreference procedure for  the determination
of oily matter in water.  It is especially suitable  for  routine  monitor-
ing of effluent streams which are known to be  relatively constant as
to the nature of oily matter present.  The method is applicable over a
wide range of concentrations.  Calibration constants  should be determined
                                48

-------
for each type of oily matter encountered,  and periodic recalibration
is recommended.

OUTLINE OF METHOD

The oily matter is extracted from the water with carbon tetrachloride.
The absorbance of the extract is determined with the use of an infrared
spectometer and the oily matter concentration is claculated by means
of Beer's Law.

DEFINITIONS

a)  Absorbance
The logarithm to the base  1 0  of the  ratio of radiant energy in a narrow
spectral range incident on the sample to the radiant energy in the same
range transmitted by the sample.

b)  Absorptivity
An empirical coefficient based on the infrared absorption of the oil
present in the sample per  unit of oil concentration and sample thickness

c)  Oily Matter
Hydrocarbons, hydrocarbon derivatives,  and all substances  containing
CH, CHp, or CH, groups -which show infrared absorption bands at 3.4
and 3.5 microns and are extractable from acidified water with carbon
tetrachloride (CC14).

APPARATUS
a) Infrared Spectrometer
Any infrared spectrometer suitable for measuring absorbance in the
3.5 micron region.

b) Spectrophotometer Cells
Matched,  1 mm cells (NaCl),  perferably with ground-glass stopper.

c) Glassware
Various pipettes and volumetric flasks, graduated cylinders,  and
1 liter separatory funnels.

REAGENTS

Unless otherwise indicated,  it is intended that all reagents conform
to the specifications established by the Committee on Analytical
Reagents of the American Chemical Society when such specifications
are available; otherwise, use best available grade.
                               49

-------
a) Hydrochloric Acid (Specific Gravity,  1.19).
b) Sodium Chloride, Crystals
c) Carbon Tetrachloride, Spectoscopic Grade

Individual bottles should be selected which have a low absorbance in
the critical range.

SAMPLE

The  general precautions governing the collection of all samples for
oil determination should be observed.  Whenever possible, the sample
should be collected in the same calibrated container used for the
extraction operation.  Volume adjustment or transfer of sample is
very likely to result in the loss of oil. Containers should be cleaned
with carbon tetrachloride before use.  If bottles are substituted for
flasks,  the screw caps should be  lined with clean aluminum  foil.

CALIBRATION STANDARDS

Weigh a quantity of the selected oil  of known specific gravity into
carbon  tetrachloride and  dilute to obtain a predetermined concen-
tration,  normally about 300 ppm by volume.  Determine the total
net absorbance as above and claculate the absorptivity constant K.

The  oil used for  calibration should be recovered,  if feasible, from
the water stream which is to be sampled.  If this is  impractical,
an average absorptivity constant may be calculated from the
absorptivities of various  oils likely to be  encountered in the
refinery waste water.

An approximate calibration standard may be synthesized by  blending
37.5 percent isooctane, 37.5 percent cetane,  and 25  percent benzene.
Experience has  indicated that oils found in a typical refinery will not
vary from  the synthetic standard  by more than + 20 percent,  except
in the case of the low boiling aromatics,  such as benzene, xylene,
and cumene, which are essentially undetected.

PROCEDURE

a) Extraction
Transfer the entire  sample (approx. three liters)  into the  extraction
flask.  Add 25 ml of concentrated HC1.  Add 25 ml of C CK to the
sample  bottle,  rinse,  and transfer contents into extraction flask.
Turn on stirrer and extract for 20 minutes.  Let the  phase settle and
repeat extraction with fresh  25 ml portions of CC1 , the designated
number of  times (Note  1). Pass the CC1,  portions through glass wool
into  an appropriate size volumetric  flask  (Note 1)  and make  up to volume,
                                50

-------
Transfer the water sample from the extraction to a graduated cylinder,
measure the volume, and subtract the volume of added HC1 .

b) Infrared Determination
The absorbances of the sample extract and of the carbon tetrachloride
blank are each determined  or in a matched pair of cells.

Care  should be taken to avoid spilling the  sample extract on the cell
while filling.  To avoid leaving a film of oil on the ecterior, wipe the
optical surfaces before each determination, first with lens  tissue
moistened -with carbon tetrachloride, and then with dry lens tissue.

The spectometer is set to scan the spectrum in the region of the
maximum absorption.  These peaks occur in the vicinity of 3.42
microns and at exactly 3.50 microns.  Net absorbances are deter-
mined by subtracting the absorption of the carbon tetrachloride
blank from the sample absorption at these wave lengths.  The sum
of the two net absorbances  is used in the calculations.

If the absorbance at either  wave length of the  original extract is greater
than 1. 0, make aliquot dilutions in carbon tetrachloride until the absorb
ances are between 0.2 and  1.0.

CALCULATION

Calculate the oily matter content of the water samples as follows:


      C =   (A) x (S)
            (K) x (W)

Where:

C = oily matter in parts per million by volume.

A = sum of the two net absorbances measured.

K = absorptivity, a constant developed during the calibration, and
    calculated as:

      K = sum of two measured net absorbances of calibration oil
                 concentration of calibration oil in CCl^. in ppm

      S =  volume of CCl^ extract.

      W = volume of water sample.

      F = dilution of factor, calculated as
                                51

-------
      F =
         Volume of diluted CC14 extract
         volume of original CC1* extract used to prepare the dilution

PRECISION,  ACCURACY, AND SENSITIVITY

The results are reported to the nearest part per million.  Precision
is about +10 percent. Accuracy is about +10 percent if a representa-
tive calibration oil is used in obtaining the calibration constant, or
about + 20  percent with the use of synthetic calibration oil. Sensitivity
is about 1 ppm.

REFERENCE

API Method 733-58
NOTES
Note 1.
Use the following guide fro determining the number of carbon-
tetrachloride extractions and the final volume to which the
CC1 * is to be adjusted.
         Anticipated
         Oil Content
           (ppm)
           20
               Number of 25 -ml
                    Extractions

                      1
                     4
 Final adjusted Volume
	Of CC14 mis

            50
           100
Note 2.  Use 5 to 10 grams of crystalline Na Cl for drying the extract
         CC1 , after having been made up to final volume.

Note 3.  The K factor has  been determined to be  0. 042 for Pascagoula
         Refinery effluents.

The method used by the Standard Oil Refinery Laboratory to determine
water in recovered crude oil was distillation with a water-immiscible
solvent.  The procedure used was ASTM designation D95-62 (Reapproved
1968); API Standard 2560.  The method was  adopted as a joint ASTM-API
standard in 1967.  It is the same as Method 3001  - Federal Test Method
Standard No. 791.
                               52

-------
                          APPENDIX F
           SEPARATOR EFFICIENCY CALCULATIONS
Three parameters are useful in judging the  performance of an oil/
•water  separator.  They are:  1) quality of water effluent,  expressed
in ppm oil content;  2)  quality of recovered oil,  expressed in percent
water  content; and,  3) percent of total influent oil recovered.

A materials balance equation provides the best method of determin-
ation of percent oil recovered,  and also shows the effect of influent com-
position on this  efficiency parameter.  It is:

Percent Oil Recovery=100 - %water in influent  ppm oil in water effluent
                            % oil in influent                  4

For  a  separator producing a water effluent with 100 ppm oil, for
instance,  percent recovery for various influent compositions would be:

         Influent %  oil            % Recovery

             0.1                     90.01
             1.0                     99.01
             5.0                     99.81
            10.0                     99.91
            50.0                     99.99

On the other hand,  a separator treating an influent with 10% oil at
various effluent purity levels would achieve  the following recoveries:

         Ppm Oil in Effluent      % Recovery

                 10                99.991
                 100                99.91
                 500               99.55
                1000               99.1
               10000               91 .0
                                53

-------
                           APPENDIX G

       PHASE II:  SPECIFICATIONS OF OIL AND SALT WATER
                         Feed Inspections

              APEC Separator Test 6/23/70 to 7/8/70


 I.   Specifications for  Crude Used in Test:

     Gravity, °API                     -           31.6
     Vis. @100°,  SSU                   -           53.86
     Sulfur,  Wt.%                      -           0.374
     RVP, ASTM D-323, PSI            -           2.8
     Pour Pt. , ASTM D-97, °F          -          +25
     BS&W,  %                          -           1.0
     Characterization Factor,  UOPK    -           11.8
     Aniline  Point, °F  (estimated)       -           190
     ASTM Distillation, Method          -           D86
     Start (5% on  D-l 160)                -           142/200
     10                                 -           320/448
     30                                 -           506/552
     EP                                -           552
     % Rec.                             -           40.0
     % Res.                             -           60.0


II.   Refinery Effluent Where Test Water Was Drawn:

     pH                                            7.7
     Sulfides, ppm                      -           Nil
     Ammonia,  ppm                     -           8
     Phenols, ppm                      -           0.07
     Dissolved 02, Mg/Liter            -           5.1
     BOD, Mg/Liter                    -           31
     COD, Mg/Liter                    -           102
     Salt Content, %                                0.8%
                                54

-------