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
------- |