SPE-173598-MS

A Simplified Approach to Sizing 2 and 3 Phase Separators for low GOR

and low Pressure Onshore Production Batteries
William P. Dokianos, Alto Engineering and Consulting Inc.

Copyright 2015, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Production and Operations Symposium held in Oklahoma City, Oklahoma, USA, 1–5 March 2015.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents
of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written
consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may
not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract
This paper provides a simplified approach to sizing 2 and 3 phase separators in low GOR (⬍2000
scf/STB) applications. The paper will provide the facility engineer with a straight forward method for
beginning the sizing of the separation system. Once the first pass sizing has been executed, the sizing can
be refined and adjusted to meet any specific requirements. The paper will also describe a simplified Stokes
equation to confirm that the retention time approach has a high probability of meeting process require-
ments. Finally, the methodology lends itself to building an excel spreadsheet for multiple vessel sizing
requirements.
The oil should be ⱖ 25 API crude, although the application can be applied to heavier crudes providing
that additional safeguards are applied. There are no limits on the water cut, but the paper assumes that the
water cut is greater than 50%. The approach uses a retention time method which sizes the liquid section
of the separator. The recommended retention time will be approximately 2 times the API 12J values for
2 reasons:
1. To account for hydraulic short circuiting in the separator
2. To ensure that the separation process is in the Stokes or Laminar regime.
The paper will also recommend using baffle plates in the 3 phase separation to facilitate plug flow. The
paper will recommend the use of either a mesh pad or vane pack in the gas section to minimize liquid
aerosols carried out with the gas.
The sizing methodology is summarized as follows:
1. Establish the actual oil, water and gas properties and rates.
2. Select a retention time for water and oil.
3. Select a Length to Diameter ratio or Height to Diameter ratio.
4. Solve for the vessel diameter and length.
5. The gas section is then sized.
6. The liquid control volume and alarm heights are established.
7. Inlet and outlet nozzles are sized.
8. Stokes calculation is executed to determine separable droplet size, oil in water, water in oil and oil
2 SPE-173598-MS

in gas.
The paper will discuss the merits of vertical and horizontal separators based on production environment
and environmental conditions.

Introduction
Domestic onshore oil production has dramatically increased over the past decade. Small and midsized
E&P companies rather than the majors have driven this development. The developments have been in
unconventional and or tight formations. The technical emphasis has been associated with downhole
processes. The surface facilities, separators and treaters, have in general been purchased as if thru a
catalogue whereas offshore systems are commonly designed based on production issues, risk and profiles.
The typical purchase of a separator in this environment is based on normal operating pressure and
temperature, an oil rate, a water rate and possibly a gas rate. The vessel design is left to a fabricator who
has been building vessels based on previous vessels which may have originally been sized on the basis
API 12J or another companies design which has been scaled up or down to match stated rates.
This approach creates a number of number of problems for the end user which includes but are not
limited to inadequate retention times, design not based on worst case, poor separation, inadequate relief,
making bad oil, rerunning oil, and high oil concentrations in water. The remainder of the paper will review
the basic principles associated with the process design of 2 and 3 phase liquid dominated separators and
a sizing approach to generate desired separation results.

Statement of Theory and Definitions

A separator whether 2 or 3 phase, vertical or horizontal is a system of components which much function
collaboratively in order to achieve the desired results. In this paper a 3 phase separator and a Free Water
Knockout (FWKO) have the same meaning. The components are:
1. The inlet nozzle and deflector
2. The gas gravity section
3. The gas outlet device
4. The gas outlet nozzle
5. The liquid section
a. Degassing the oil
b. Gravity based water from oil separation and oil from water separation for 3 phase separators
6. Liquid outlet piping
7. Process controls
8. Overpressure protection
In order to size the above components, a process table should be developed. The table is used to identify
the key process variables which are then applied to the separator design. Table 1 summarizes the minimum
process data required.
The following steps develop the sizing calculations and their derivations for completeness. As a result
the steps appear more complex than they are in execution. The example will use the steps in their
simplified form as one would in sizing a new vessel or troubleshooting an existing vessel.

Application
Step 1
Obtain the data required for Table 1. The pressure and temperature data should be available from existing
field data and from production data. The temperature information must also consider the heat loss between
the well head and the production battery. As an example, many producers use polyethylene flow lines laid
SPE-173598-MS 3

Table 1—Summary of process description

Design variable Operating Range Units

Pressure PSIG Minimum Normal Maximum Pounds per Square Inch Gage
Temperature °F Minimum Normal Maximum Degrees Fahrenheit
Gas Rate SCFD Minimum Normal Maximum Standard Feet Cubed per Day
Actual Gas Rate AFt3/S Minimum Normal Maximum Actual Feet Cubed per Second
Actual Gas Density lbm/Ft3 Nominal Pounds per Foot Cubed
Oil Rate BPD Normal Maximum Barrels per Day
Oil Rate CFM Maximum Cubic Feet per Minute
Oil Viscosity (cP/1488) Minimum Normal Maximum Pounds per Foot-Second
Produced Water Rate BPD Maximum Barrel per Day
Produced Water Rate CFM Maximum Cubic Feet Per Minute
Produced Water Viscosity (cP/1488) Nominal Pounds per Foot -Second
Oil Density lbm/Ft3 Nominal Pounds per Foot Cubed
Water Density #/Ft3 Nominal Pounds Per Foot Cubed

on the surface between the well head and the battery. In the winter the production fluids arrive at the
battery at or close to ambient temperature. The critical aspect of the temperature is its effect on viscosity.
The viscosity of 35 API gravity oil at 100°F at the well head is approximately 4cP. Based on heat loss
the well head fluid can easily arrive at the production battery at 40°F which changes the viscosity from
4cP to 16cP.1 The oil and water rates in many cases have a correlation based on how the well comes in.
In many cases facilities are designed on the basis of best case oil and historical water rates. The worst case
oil rate is also based on historical rates in the producing area. The producer ultimately has to decide on
whether the design basis is the optimistic or the realistic rate. This paper assumes that the optimistic rate
be defined as the maximum and the realistic rate is defined as normal. Most of the data required for the
table is easily obtained internally with the exception of viscosity. There are a number of correlations which
can be used to obtain reasonable values such as the Glaso correlation. Given the fact that these systems
operate at low pressure, less than 200 PSIG, the dead oil correlation is adequate since Gas in Solution
values are very low at these pressures. For example 35 API oil at 200 PSIA and 100°F has a live oil
viscosity value of 3.4cP versus a dead oil viscosity value of 4cP. Using the dead oil viscosity at low
pressure can be thought of as a safety factor when a separable droplet is calculated.
Step 2
Decide on whether the separator should be vertical or horizontal. As with most design decisions there is
not a single correct answer. General statements can be made:
1. Horizontal is better for 3 phase separation because for the same design parameters the horizontal
has better retention time and less steel.
2. Horizontal is better for foamy crudes because increasing retention in the horizontal vessel still
results in less steel than an equivalent vertical vessel.
3. Sand and Wax control and removal is more easily addressed in a vertical vessel, especially one
with a conical bottom.
4. Liquid dominated 2 phase vessels may be either vertical or horizontal. In general, the horizontal
vessel will result in less steel.
Numerous texts such as Arnold and Stewart’s Surface Production Operations Volume 1 (2008) and
Manning and Thompson’s Oilfield Processing Volume 2 (1995) discuss the merits of vertical versus
horizontal separators.
If the service is 2 phase degassing of a liquid load, the decision could be either a vertical or horizontal
and the deciding factors may be cost and other process issues. If the liquid load is substantial, greater than
5000 BPD then the most common decision is a horizontal separator. This tends to be an economic decision
4 SPE-173598-MS

Figure 1—Drag Coefficient versus Reynolds Number from Chemical Engineers’ Handbook, McGraw Hill, New York, NY

If the service is 3 phase separation, liquid dominated then the most common selection is a horizontal
separator. The decision is based on economics, less steel.
Step 3
The design criteria will be based on liquids in the laminar flow regime. This means that the Reynolds
number for the separable droplet of interest, i.e. a gas bubble out of the liquid phase, will be less than or
equal to 1. See Figure 1 from the Chemical Engineers’ Handbook
The boundary between the Stokes (Laminar) flow regime and the Intermediate flow regime is a
Reynolds Number value between 1 and 2. This paper will use a Reynolds value of 1 to define that
boundary.
The Reynolds Number Re ⫽ (DpV␳)/␮ ⬍ 1 for Laminar or Stokes flow
Where:

Dp Diameter of the droplet in the dispersed phase ft

V Terminal velocity of the dispersed droplet falling or rising in the continuous phase ft/sec
␳ Density of the continuous phase lbm/ft3
␮ Absolute viscosity of the continuous phase lbm/ft-sec

The vessel sizing approach assumes laminar or Stokes flow in the liquid section. The separable droplet
will be calculated on the basis of laminar flow. The Reynolds Number is then calculated on the calculated
droplet and terminal velocity which are determined by the vessel size. If the calculated Reynolds number
is less than 1 then the vessel is in the laminar or Stoke regime and the separable droplet is valid. This
calculation is used to confirm that the assumptions for the residence time sizing criteria are valid.
The following steps will be segregated to address the sizing for the 2 most common applications:
1. 3 phase oil water separator
2. 2 phase gas liquid separator
SPE-173598-MS 5

Figure 2—Schematic of a 3-phase separator

Step 4: 3 Phase Horizontal Separator

1. Select a retention time. API 12J (1989) establishes a set of minimum criteria. For example, crude
oil greater than 35 API gravity should have a retention time of 3 to 5 minutes. Since most onshore
production batteries utilize a 3 phase separator following by a heater treater, retention time is a
critical decision. The short retention times can lead to high water cut in the oils and higher than
desired oil in water in the produced water. The paper suggests that retention time for:

a. greater than 35 API gravity crudes, Retention Time is a minimum of 10 minutes,

b. 25 to 35 API gravity crudes at a minimum operating temperature of 100°F, Retention Time
(Tro) is a minimum of 20 minutes
c. 25 API to 35 API gravity crudes and above at a minimum temperature of 80°F, Retention
Time (Tro) is a minimum of 30 minutes
d. 25 API to 35 API gravity crudes and above at a minimum temperature of 60°F, Retention
Time (Tro) is a minimum of 60 minutes
2. Select a retention time for the water. API 12J (1989) recommends a retention equal to oil. API 12L
(1987) uses a nominal retention time of one half of the oil retention time. The paper suggests that
the water retention time is 50% to 100% of the oil retention time. The longer retention times will
ensure that the separator is operating in the laminar regime and it will also account for known short
circuiting which will occur in the separator.
3. Select a length to diameter ratio. For the 3 phase separator a simple selection is an L/D ratio of
4 or 5 is a good first choice. Lengths to Diameter Ratios from 3 to 6 are acceptable. Arnold and
Stewart (2008) suggest that reentrainment should be calculated when the ratio is greater than 5. For
low GOR and low pressure systems reentrainment tends not to be an issue. Because the retention
time numbers are very conservative, the paper does not use the concept of effective length.
a. L/D ⫽ C1 where C1 is a numerical constant ) L ⫽ C1D for example C1 ⫽ 5
4. Select the Normal Liquid Level (NLL) as percentage of Vessel Diameter. For example, a low GOR
vessel with minimal surging a NLL of 0.7 to 0.8D is a reasonable choice. It is important to note
that the NLL/D ratio would have to be reduced if conventional High Level Alarms and High level
Shutdowns are employed. If a tuning fork style level sensor is used a high level shutdown can still
be employed.
a. HI ⫽ C2 ⫻ D where C2 is a fractional value of the diameter, for example 0.5 and Hl is the
height of Liquid
6 SPE-173598-MS

Figure 3—Fraction of area for gas, oil and water flow as a function of interface height

5. Calculate the volume for oil, volume for water and the total liquid volume using Retention Times
chosen in step 1:
a. Vo ⫽ Qo ⫻ Tro
b. Vw ⫽ Qw ⫻ Trw
c. Vl ⫽ Vo ⫹ Vw
Where

Vo, Vw, Vl Feet Cubed Ft3

Tro, Trw Retention Time Oil, Water Minutes

6. Use a graph or table to obtain the percentage of vessel area available for Liquid. The 1987 GPSA
Volume 1 Figure 6-22 has a complete table of partial volume of horizontal cylinders. Figure 3
below describes the area available for Gas, Oil and Water. The M values used in the 3 phase
horizontal vessel are obtained from the GPSA.
a. The Area Available for Liquid is Ml which is equal to Mo ⫹ Mw
b. The Area Available for Oil is Mo
c. The Area Available for Water is Mw
d. The Area Available for Gas is (1- Ml)
7. The vessel dimension can now be calculated. The Volume for liquid is a function of the diameter
cubed because the length is a function of the diameter from step 3 and the percentage of vessel area
for liquid is defined by the height of liquid to diameter ratio from step 6.
a.
b. where L ⫽ C1D and Ml is the percentage of area for liquid
c.
d. where Diameter and Lenqth are in feet

8. Calculate the vessel length

a. L ⫽ C1D
9. Calculate the height of the liquid (NLL)
SPE-173598-MS 7

a. HI ⫽ C2D where Height of Liquid and Diameter are in feet

10. The vessel diameter, length and height of liquid (NLL) are now calculated. To obtain the height
of water (Hw) and the height of the oil (Ho), the area for water (Mw) must be calculated.
a.
b.
c. Once Mw is known, the Hw can be found in the GPSA table mentioned in step 6.
11. The height for oil is the difference between the height of liquid minus the height for water.
a. Ho ⫽ Hl ⫺ Hw
12. The vessel geometry can now be optimized based on steel, instrumentation requirements for the
interface controller or other process needs.
13. The separable droplets for the water in oil and oil in water are now calculated. Use the vessel
length, diameter, height of oil and height of water decided on in Step 12 if the any if the vessel
parameters were adjusted. Arnold and Stewart (2008) recommend Oil droplets no larger than 500
micron. The author has found that droplet sizes between 300 and 400 micron yields good results
in ensuring that there is less than 10% water in oil entering the heater treater based on imperfect
vessel hydraulics. While the focus of the 3 phase separator design is good oil, Arnold and Stewart
(2008) recommend separable oil from water droplet size of 200 microns or less. This paper agrees
and recommends using the value of 200 microns or less for the separable oil droplet from water.
a. Vt for water out of oil ⫽ Ho/(60 ⫻ Tro) where time is in seconds, height is in feet, and Vt is
in ft/sec.
b. Using Stokes Law in the laminar regime calculate the separable water droplet
i.
where g ⫽ 32.2 ft/ sec2
c. Confirm that the Reynolds Number based on the separable droplet calculated in step c. and the
terminal velocity calculated in Step a. is less than 1.
i. then the flow is laminar.

ii. and if the separable water droplet is ⱕ 400 microns then the retention time sizing is valid
d. Steps a. thru c. are repeated for the separable droplet calculation for oil out of water
e. Vt for oil out of water ⫽ Hw/(60 ⫻ Trw) where time is in seconds
f. Using Stokes Law in the laminar regime calculate the separable oil droplet
i. where g ⫽ 32.2 ft/ sec2

g. Confirm that the Reynolds Number based on the separable droplet calculated in step g. and the
terminal velocity calculated in Step f. is less than 1.
i. If Re ⫽ (Vtoiw ⫻ Doiw ⫻ pw)/␮w ⬍ 1 then the flow is laminar and if the separable oil
droplet is ⬍ 200 micron then the retention time sizing is valid
14. In order to mitigate hydraulic short circuiting, the designer should consider the addition of 2 or 3
baffle plates/flow straighteners (see figure below).The plates should also extend a 4 to 6 inches
into the gas section to break any surging or wave action. The author has good success with cutout
area which slows the fluid velocity thru the cutout area to one half of the vessel fluid velocity. The
cutouts can be circular or elongated circles
15. The gas section is now addressed. This paper assumes that a 6⬙ standard mesh pad is used. The
sizing of the gas space uses API 12J (1987) recommendations for determining the vessel Ksb.
Assuming that the vessel is longer than 10 feet then Ksb will be between 0.7 and 0.9 feet per
8 SPE-173598-MS

second. Because the vessel has been sized for liquid, the gas section may be a compromise. The
vessel Ksb uses the standard Souders Brown equation in API 12J (1987).
a.

b. where the Gas Flow rate is in Actual Feet Cubed per Second
c. Area for Gas ⫽ (␲/4)D2 ⫻ (1 ⫺ Ml)
d. If the vessel Ks is between 0.7 and 0.9 then the mesh pad can be installed across the horizontal
cross section and can be liquid wetted. If the vessel Ks is below 0.7 to 0.9 then the designer
can either use a gas dome sized for the desired Ks or fabricate a box with an internal area inside
the vessel which will satisfy the vessel desired Ks.
16. The final components to consider are the inlet and outlet nozzles. This paper assumes that
conventional impact or diverter plate is used. The inlet nozzle should precondition the flow for
separation and as such it should reduce momentum. Manning (1995) states that erosional velocity
should be the limiting factor. This paper does not agree. Stewart and Arnold (2008) suggest that
inlet and gas outlet nozzles should be sized so that the velocity is less than 60 divided by the square
root of the stream density. This paper suggests a more conservative approach for the inlet nozzle.
The inlet nozzle should be sized on the basis of inlet velocity equal to 30 to 40 divided by the
square root of the inlet mixture density. This allows for the fluid stream to slow down before
entering the vessel and minimizes droplet shattering. The liquid outlet nozzles are sized to
minimize gas carryunder and to minimize restrictions due to scale and other deposits. This paper
recommends a liquid outlet velocity of 4 to 6 feet per second.
a.
b.
c. Voil or V water ⫽ 4 to 6 feet per second
17. The designer must make a decision on what is the worst case overpressure event and the relief
volume associated with the event. In many cases, this reduces to a decision on sizing the relief
valve based on vapor only or a vapor and liquid. The designer is responsible for this decision but
this paper recommends using vapor and liquid as the basis for sizing the relief valve which is the
most conservative approach.
18. Process controls play a critical role in obtaining the design performance. The interface control is
the most critical because it controls the operating residence time of both the oil and water. The
suitability of a control is partially dependent on the how the vessel designer has designed the
vessel. There are 3 possibilities:
a. No weir, the oil simply sits on the water on top and the oil is a drawn from the vessel via an
outlet pipe set at a unique elevation. This paper is not going to discuss this approach because
of the inherent issues and poor control associated with this design.
b. The liquid level in the vessel is set by a weir. This is the most common design. The oil spills
over the weir into an isolated compartment. In this case the oil level in the isolated section is
easily controlled by a float controlled mechanical control valve. The interface level controller
is more problematic. The conventional float control depends on the density difference between
the oil and water. As a result it tends to have a wide operating range. More importantly, when
it actuates, the water level falls dramatically. This means that the relative retention times of the
water and the oil vary in accordance with the water level at the lowest point when the valve has
actuated to the normal interface level just before the valve actuates. The float travel is normally
set to 12 to 18 inch and the level falls in accordance with the float travel. The second issue with
this type of interface controller is how the float is affected by changing temperature. In the
SPE-173598-MS 9

winter or seasonal change, the vessel, the oil and the water cool dramatically. The temperature
change can be a 50°F swing. The oil viscosity increases and the quality or sharpness of the
gravity break at the interface deteriorates. The smaller the difference in gravity and the
thickness of the interface emulsion alters the reliability of the conventional float style interface
controller. The author has good results with the newer generation of controllers using a more
sensitive float (Teflon) and a more sensitive controller.
c. The bucket and weir design uses an oil bucket and a weir which controls the water interface
level. This approach allows the oil to be in its unique compartment and the water to be in its
unique compartment. The interface level is physically controlled by the water weir. This allows
for the satisfactory use of the conventional float style control valves in both the oil bucket and
water compartment.

Example-3 Phase Horizontal Separator

Assume that a 3 phase separator is required. The vessel will use a conventional weir. The process data is
in the table 2 below.

Table 2—Process data

Design variable Minimum Normal Maximum Units

Pressure PSIG 80 100 125 Pounds per Square Inch Gage

Temperature °F 40 100 110 Degrees Fahrenheit
Gas Rate SCFD 500000 500000 Standard Feet Cubed per Day
Actual Gas Rate AFt3/S 0.72 0.81 0.84 Actual Feet Cubed per Second
Actual Gas Density lbm/Ft3 0.5 Pounds per Foot Cubed
Gas Viscosity (cP/1488) 0.015/1488 Pounds per Foot-Second
Oil Rate BPD 1000 1000 Barrels per Day
Oil Rate CFM 3.9 3.9 Cubic Feet per Minute
Oil Viscosity (cP/1488) 8/1488 2.5/1488 2/1488 Pounds per Foot-Second
Produced Water Rate BPD 5000 Barrel per Day
Produced Water Rate CFM 19.5 19.5 Cubic Feet Per Minute
Produced Water Viscosity (cP/1488) 1/1488 Pounds per Foot-Second
Oil Density lbm Ft3 (40API) 51 Pounds per Foot Cubed
Water Density lbm/Ft3 (Sg 1.1) 69 Pounds Per Foot Cubed

1. Oil Retention Time Tro ⫽ 10 minutes

2. Water Retention Time Trw ⫽ 5 minutes
3. Length to Diameter Ratio ⫽ 4 so that vessel Length ⫽ 4Diameters, L⫽4D
4. Normal Liquid Level ⫽0.8D, NLL⫽0.8D
5. Volume for oil. Volume for water and volume for liquid.
a. Vo ⫽ Qo ⫻ Tr ⫽ 3.9 ⫻ 10 ⫽ 39ft3
b. Vw ⫽ Qw ⫻ Tr ⫽ 19.5 ⫻ 5 ⫽ 98ft3
c. Vl ⫽ Vo ⫹ Vw ⫽ 39 ⫹ 98 ⫽ 137ft3
6. Based on NLL at 0.8D, the area for liquid is obtained for the GPSA Volume 1 Figure 6-22.
a. Ml ⫽ .857
7. The vessel diameter can now be calculated.
a. where Diameter and Length are in feet
b. ft round up to 4 ft
10 SPE-173598-MS

8. The vessel length

a. L ⫽ C1D ⫽ 4 ⫻ 4 ⫽ 16 feet
9. Calculate the height of the liquid (NLL)
a. Hl ⫽ C2D⫽0.8⫻4⫽3.2feet
10. To obtain the height of water (Hw) and the height of the oil (Ho), the area for water (Mw) must
be calculated.
a. of the total vessel area
b. Once Mw is known, the Hw can be found in the GPSA table mentioned in step 6
c. From the GPSA Vol. 1 page 6-21 Fig. 6-22 the value for Hw ⫽ 0.485 ⫻ D ⫽ 1.94feet
11. The height for oil is the difference between the height of liquid minus the height for water.
a. Ho ⫽ HI ⫺ Hw ⫽ 3.2 ⫺ 1.94 ⫽ 1.26feet
12. Assume that the Diameter, Height and Heights of water and oil are acceptable, then the vessel
geometry is finished.
13. Confirm that both the oil and water are in the Stokes or laminar flow regime based on the vessel
dimensions established using retention time method.
a. note retention time converted to seconds
b. Using Stokes Law in the laminar regime calculate the separable water droplet. Worst case is
the lowest temperature
i.

c. Confirm that the Reynolds Number based on the separable droplet calculate in steps. and
terminal velocity is Step a. is less than 1.
i. 1 confirming that
the flow is laminar and the sepable water droplet is ⬍ 400 micron confirming that the
retention time sizing is valid
d. Steps a. thru c. are repeated for the separable droplet calculation for oil out of water
e. Vt for oil out of water ⫽ Hw/Trw ⫽ 1.94/300 ⫽ .0065ft/sec note retention time converted to
seconds
f. Using Stokes Law in the laminar regime calculate the separable oil droplet
i.

g. Confirm that the Reynolds Number based on the separable droplet calculate in step i and in
Step g. is less than 1.
i. If Re ⫽ (Vtoiw ⫻ Doiw ⫻ ␳w)/␮w ⫽ (0.0065 ⫻ .000368 ⫻ 69)/(1/1488) ⫽ .245 ⬍ 1
confirming the flow is laminar and the separable oil droplet is ⬍ 200 micron confirming that
the retention time sizing is valid
14. Two or three flow conditioners would be designed as discussed above. The decision to use holes
or elongated slots would be a fabrication decision.
15. The gas section is now addressed. A 6⬙ standard mesh pad is used. The vessel is16 feet long then
Ksb should be between 0.52 and 0.65 based on API 12J (1989). Because the vessel has been sized
for liquid, the gas section may be a compromise. The vessel Ksb uses the standard Souders Brown
SPE-173598-MS 11

equation in API 12J (1989).

a.
b. Vg ⫽ Qg/Area for Gas ⫽ 0.84/2 ⫽ 0.41ft/sec
c. Area for Gas ⫽ (␲/4)D2 ⫻ (1 — Ml) ⫽ (␲/4) ⫻ 42 ⫻ (1 — 0.84) ⫽ 2ft2
d.
e. In this case the gas space is much larger than required. A gas dome could be used where the
diameter of the outlet speeds up the gas to a required velocity. The Ksb for the dome would
be 0.24ft/sec based on vertical flow per API 12J (1989). If a dome were to be used the Gas
Velocity would be 2.4ft/sec and the area would be 0.35ft2. The actual diameter would be either
an 8⬙ or 10⬙ dome.
16. The inlet, gas outlet, water and oil outlet nozzles can be sized.
a. which would require a 6⬙ inlet nozzle
b. which would require a 2⬙ nozzle
c. Voil ⫽ 4 to 6 feet per second ⫽ 0.065/4 ⫽ 0.02ft2 which would require a 2⬙ nozzle? To avoid
issues with scaling, sediment and wax the designer should consider increasing the size to 3⬙
d. Vwater ⫽ 4 to 6 feet per second ⫽ 0.325 ⫼ 4 ⫽ 0.08ft2 which would require a 4⬙ nozzle

Conclusions on the 3 phase separator sizing

1. The author has used this sizing approach on many vessels in the Texas and Oklahoma. The vessels
when installed have been thermally scanned and oil and water quality were analyzed. The
following has been confirmed:
a. Good gravity break between the oil and water.
b. In many cases with lighter oil, emulsion breaking chemical injection has not been required.
c. The oil going to the treater required minimal treating and in some instances was sales grade out
of the treater.
d. The water quality was less than 500 mg/l and in many cases less than 200 mg/l.
2. The interface control was better using the newer generation of controllers over the float style
mechanical system.
3. The vessels, not surprisingly, had better performance when retention time for water was equal to
retention time for oil.
4. In some cases where ambient temperature could be detrimental to the separation quality, direct
fired heater tubes have been added.
5. High and Low level alarms and low level shutdowns were not installed. The high level shutdowns
when installed were tuning fork type devices.
6. Using a simplified form of Stokes Law allows the designer to confirm that the flow is laminar and
that the separable droplet is equal to or less than the design droplet.

The Vertical 2 phase degassing separator

The vertical 2 phase degassing separator is a liquid dominated but its sizing is complicated by a desired
maximum oil droplet in the gas space. The sizing will account for both the liquid domination and the
maximum separable droplet in the gas space. The vertical separator utilizes the first 2 steps described in
the horizontal 3 phase separator. Arnold and Stewart (2008) recommend an oil droplet no larger than 140
micron in the vertical gas space to avoid overloading the mesh pad. API 12J (1989) gives no guidance in
this area. Assuming that the primary process reason for the separator is to degas the oil for a downstream
compressor or wet gas gathering system then the 140 micron size recommended by Arnold and Stewart
12 SPE-173598-MS

Figure 4 —Schematic of a vertical 2-phase separator

(2008) is reasonable. This paper also recommends that the separable gas bubble in the liquid section is less
than or equal to 200 micron for good degassing quality. The liquid section will use a retention time
sufficient to ensure laminar flow using the lowest operating temperature.
Finally the authors experiences suggest that the economics of a vertical separator are limited to 1000
liquid barrels per day assuming there are no mitigating circumstances.
The method for sizing the vertical 2 phase liquid dominated separator will:
a. Use a simplified version of droplet settlement theory to size the vessel diameter (D) based on
separable oil droplet less than or equal to 140 micron. A critical point is that the gas section
operates in the intermediate flow regime because gas viscosity is small.
b. The height (H) of the vessel will be based on retention with flow in the laminar regime. The sizing
will be confirmed by using the simplified Stokes equation in the laminar regime to confirm that
the separable gas bubble out of oil is less than 200 micron.
c. After the vessel Diameter and Height have been calculated. The Vessel Souders Brown coefficient
Ksb is calculated. The Ksb for the vessel will be smaller than the recommended value in API 12J
(1989). A fabrication detail with an area calculated to speed up the velocity which satisfies the
desired Ksb will be calculated
SPE-173598-MS 13

The sizing procedure will use a modified form of the 1987 GPSA Guidelines for Vertical Separator
Dimensions. The sizing procedure also includes High Level Shutdown and a Low Level Shutdown for
safe operating practices. See Figure 4.
Steps 1and 2 are same as the horizontal separator. The sizing method will begin at step 3.

Step 3
1. Arbitrarily select a value of 10 for the Reynolds number in the intermediate flow regime (see
Figure 1 below).

Figure 1—Drag Coefficient versus Reynolds Number from Chemical Engineers Handbook, McGraw Hill, New York, NY

2. Calculate the Terminal Velocity Vt based on 140 micron (460 ⫻ 10⫺6ft). The terminal Velocity
will remain constant thru this process. Viscosity is in units of Pounds per Foot-Second.
a. Re ⫽ (Vt ⫻ (460 ⫻ 10⫺6) ⫻ ␳g)/␮g ⫽ 10 where the regime is intermediate if Re ⬎ 1
b. Vt ⫽ (10 ⫻ ␮g)/((460 ⫻ ⫺6) ⫻ ␳g)
3. Calculate the Drag Coefficient based on the Reynolds Number (Cdre). There are a variety of ways
to evaluate the relationship between the Reynolds number and the drag coefficient (Cd). The drag
coefficient calculation shown on Figure 1. is straight forward. Given that the Reynolds number
will be greater than 1 and the gas section will be in the intermediate regime.
a. Cdre ⫽ (18.5)/Re0.6 the reynolds number raised to the 0.6 power
4. Calculate the Drag Coefficient (Cdge) using the general equation. The general equation will use
the velocity calculated in a.ii. and a droplet size of 460 ⫻ 10⫺6ft. (140 microns).
a. where g is 32.2 ft/sec2 the equation can be
simplified to
b. Cdge ⫽ (43 ⫻ (460 ⫻ 10⫺6)/(Vt)2) ⫻ ((␳o — ␳g)/␳g)
14 SPE-173598-MS

5. If the drag coefficient (Cdre) calculated using the Reynolds number is ⱕ the drag coefficient
(Cdge) using the general equation then the separable droplet will ⱕ 140 microns. If there is desire
to calculate the exact droplets then the 2 drag coefficients are iterated using progressively smaller
separable droplets until agreement is obtained. This method does not iterate a solution. The fact
that the performance is better than required is viewed as conservative design.
a. if Cdre ⱕ Cdge then the vessel will separate oil droplets in the gas space less than 140 microns.
6. The Vessel Diameter is calculated using the Velocity in a.ii.:
a.
Where:

Diameter (D) Feet

Gas Rate (Qg) Actual Feet Cubed per Second
Gas Velocity (Vt) Feet per Second

Step 4
1. Select a liquid retention time. API 12J (1988) recommends minimum retention time for the liquid
to degas. The retention times are:
a. Above 35 °API 1 minute
b. 20 to 30 °API 1 to 2 minutes
c. 10 to 20 °API 2 to 4 minutes
The paper is only concerned with oils greater than 25° API and believes that operating temperature
should be part of the decision making process because it will influence when equilibrium is
reached between the gas and the liquid. Low operating temperature increases the oil viscosity and
therefore increases the size of the separable droplet. This recommends the following retention
times:
i. Above 35 ° API 2minutes
ii. 30 to 35 ° API 3 minutes
iii. 25 to 30 ° API 4 minutes
Arnold and Stewart (2008) make the following points: if foam is present increase retention time
by a factor of 2 to 4 and if high CO2 exists use a minimum retention time of 5 minutes. The author
has seen foam with seven minute collapse time so Arnold and Stewarts comment is well taken. The
author would suggest that unless the foam characteristics are known, a factor of 4 is better than
2 because the need for a chemical defoamant will be less.
2. Calculate the Volume for Liquid. This volume provides space for both degassing and control. The
retention for oil and water is the same.
a. VI ⫽ (Qo ⫹ Qw) ⫻ Tr
Where:

Vl Feet Cubed Ft3

Tr Retention Time for Liquid Control Minutes

3. Calculate the Height of the Volume for Liquid (Hl). This height should be greater than or equal
to18 inches so the level controllers movement is not compromised.
SPE-173598-MS 15

a. HI ⫽ Vl/((␲/4) ⫻ D2)
Where:

Height (Hl) Feet

Volume for Liquid (Vl) Feet Cubed
Diameter (D) Feet

b. If HI ⱖ 1.5 ft then use the value of HI if HI ⬍ 1.5 ft then HI ⫽ 1.5 ft

4. The height required for the high level shutdown and the low level shutdown can be calculated
(Hsd). The height for each will be the same because they both use the same retention time of 1
minute or have a minimum of 1 foot spacing between the normal liquid space and the shutdown
device.
a. Hsd ⫽ (Ql ⫻ 1 minute) ⫼ ((␲/4) ⫻ D2) where Ql is in and Hsd is in feet
b. If Hsd ⱖ 1 ft then use the calculated value of Hsd, if Hsd ⬍ 1 ft then Hsd ⫽ 1ft
5. The Height of the Vessel associated with the gas space (Hg) is the height from the inlet nozzle to
the mesh pad. This space allows for the gas to develop uniform flow. The height is a minimum of
2.5 feet or the vessel diameter.
a. If D ⱖ 2.5ft then Hg ⫽ D, if D ⬍ 2.5 ft then Hg ⫽ 2.5 ft
6. The Height of the Vessel (H) is calculated by adding up the required heights from the bottom of
the vessel. Refer to Figure 4 above. API 12J (1989) uses 2.5 feet increments. The vessel height
should be rounded up to the next standard increment. The designer can decide how to allocate the
additional height. Industry practice uses a Height to Diameter ratio of 3 to 5. Arnold and Stewart
(2008) state that a height to diameter ratio of 3 to 4 is common.
a. H ⫽ 1 ⫹ Hsd Low Level ⫹ HI ⫹ HsdHigh Level ⫹ 1 ⫹ Hg ⫹ 1
7. The Souders Brown Coefficient for the vessel (Ksbv) is obtained by first calculating the gas
velocity in the vessel and then calculating the vessel Ksb. If the Vessel Ksb is within API 12J
(1989) recommendations for Ksb then the diameter is acceptable. The assumption is that the mesh
pad utilizes 100% of the vessel area. If the vessel Ksb is less than the API 12J (1989)
recommendations then the mesh pad area must be reduced. This is accomplished by designing a
fabrication detail which is inserted in the gas space. The mesh pad is installed into the fabrication
such that the open area satisfies the velocity for a correct Ksb for the mesh pad.
a. Vg ⫽ Qg/((␲/4)D2)
b.
c.
Where:

Gas Velocity (Vg) ft/sec Feet per Second

Oil Density (␳o) lbm/ft3 Pound per Foot Cubed
Gs Density (␳g) lbm/ft3 Pound per Foot Cubed
Actual Gas Rate (Qg) ft3/sec Actual Feet Cubed per Second
Diameter (D) ft Feet

8. The inlet nozzle, the gas outlet nozzle and liquid outlet nozzle are sized exactly the same as in the
16 SPE-173598-MS

Horizontal Separator Step16.

9. A simplified Stokes procedure confirms that the liquid section will separate out a gas bubble of
200 micron or less. Although the liquid section is a mix of oil and water, the confirmation process
uses oil only because it is the most conservative approach.
a. Oil Section
i. Calculate the Terminal Velocity for the gas bubble rising by dividing the vessel area into
liquid rate. The subscript gio is gas bubble in oil.
1. Vtgio ⫽ Ql/((␲/4)D2)

ii. Calculate the separable gas bubble in the gas section assuming laminar flow.
1.

iii. Confirm laminar flow in the oil section by calculating the Reynolds number for the oil
section. If the Reynolds number is less 1 then the flow is laminar.
1. If Re ⫽ (Vtgio ⫻ Dgio ⫻ ␳o)/␮o ⬍ 1 and the flow is laminar and if the separable water
droplet is ⱕ 200 microns then the retention time sizing is valid

Example-2 Phase Vertical Separator

Assume that a 2 phase vertical separator is required to degas the wellstream fluids. The vessel will use
a 6⬙ standard mesh pad. The process data is in the table 3 below.

Table 3—Summary process data

Process Variable Minimum Normal Maximum

Pressure PSIG 80 100 125 Pounds per Square Inch Gage

Temperature °F 40 100 110 Degrees Fahrenheit
Gas Rate SCFD 2000000 2000000 Standard Feet Cubed per Day
Actual Gas Rate AFt3/S 2.88 3.32 3.36 Actual Feet Cubed per Second
Actual Gas Density lbm/Ft3 .5 Pounds per Foot Cubed
Gas Viscosity (cP/1488) .013/1488 Pounds per Foot-Second
Oil and Water Rate BPD 1000 1000 Barrels per Day
Oil and Water Rate CFM 3.9 3.9 Cubic Feet per Minute
Oil Viscosity (cP/1488) 8/1488 2.5/1488 2/1488 Pounds per Foot-Second
Produced Water Viscosity (cP/1488) 1 Pounds per Foot-Second
Oil Density lbm/Ft3 (40API) 51 Pounds per Foot Cubed
Water Density lbm/Ft3 (Sg 1.1) 69 Pounds Per Foot Cubed

1. Arbitrarily select a value of 10 for the Reynolds number in the intermediate flow regime.
a. Calculate the Terminal Velocity Vt based on 140 micron (460 ⫻ 10⫺6/t). The terminal Velocity
will remain constant thru this process. Viscosity is in units of Pounds per Foot-Second.
i. Re ⫽ (Vt ⫻ (460 ⫻ 10⫺6) ⫻ ␳g)/␮g ⫽ 10 where the regime is intermediate if Re ⬎ 1
ii. Vt ⫽ (10 ⫻ (8.7 ⫻ 10⫺6))/((460 ⫻ 10⫺6) ⫻ 0.5) ⫽ 0.38 ft/sec
b. Calculate the Drag Coefficient based on the Reynolds Number (Cdre). There are a variety of
ways to evaluate the relationship between the Reynolds number and the drag coefficient (Cd).
The drag coefficient shown on Figure 1. is straight forward. Given that the Reynolds number
will be greater than 1 and the gas section will be in the intermediate regime.
i. Cdre ⫽ (18.5)/Re0.6 ⫽ 18.5/100.6 ⫽ 4.6
SPE-173598-MS 17

c. Calculate the Drag Coefficient (Cdge) using the general equation. The general equation will use
the velocity calculated in a.ii. and a droplet size of 460 ⫻ 10⫺6ft. (140 microns).
i. where g is 32.2 ft/sec2 the equation can be
simplified to
ii. Cdge ⫽ (43 ⫻ (460 ⫻ 10⫺6)/(0.38)2) ⫻ ((51 ⫺ 0.5)/.5) ⫽ 13.8
d. If the drag coefficient (Cdre) calculated using the Reynolds number is ⱕ the drag coefficient
(Cdge) using the general equation then the separable droplet will ⱕ 140 microns
i. 4.6 ⱕ 13.8 the vessel will separate oil droplets in the gas space less than 140 microns.
e. The Vessel Diameter is calculated using the Velocity in a.ii.:
i. round up to 3.5 ft
where

Diameter (D) Feet

Actual Gas Rate (Qg) Feet Cubed per Second
Gas Velocity (Vt) Feet per Second

2. Select a liquid retention time.

a. Above 35 °API 2minutes
b. There is no CO2 and assuming there is no foam
c. Retention Time is 2 minutes
3. Calculate the Volume for Liquid. The retention for oil and water is the same.
a. VI ⫽ (Qo ⫹ Qw) ⫻ Trl
b. VI ⫽ 3.9 ⫻ 2 ⫽ 7.8 ft3
4. Calculate the Height of the Volume for Liquid (Hl). This height should be greater than 18 inches
so the level controller’s movement is not compromised.
a. Hl ⫽ Vl/((␲/4)⫻ D2)
b. HI ⫽ 7.8/((␲/4) ⫻ (3.5)2) ⫽ 0.8 ft use HI ⫽ 1.5 ft
where

Height (Hl) Feet

Volume for Liquid (Vl) Feet Cubed
Diameter (D) Feet

5. The height required for the high level shutdown and the low level shutdown can be calculated
(Hsd). The height for each will be the same because they both use the same retention time of 1
minute or have a minimum of 1 foot spacing between the normal liquid space and the shutdown
device.
a.
b. If Hsd ⱖ 1ft then use the calculated value of HSD, if Hsd ⬍ 1 ft then Hsd ⫽ 1 ft
6. The Height of the Vessel associated with the gas space (Hg) is the height from the inlet nozzle to
the mesh pad. This space allows for the gas to develop plug or piston like flow. The height is a
18 SPE-173598-MS

minimum of 2.5 feet or the vessel diameter.

a. If D ⱖ 2.5ft then Hg ⫽ D, if D ⬍ 2.5 ft then Hg ⫽ 2.5 ft
b. Since D ⫽ 3.5ft ⱖ 2.5 then Hg ⫽ 3.5 ft
7. The Height of the Vessel (H) is calculated by adding up the required heights from the bottom of
the vessel. Refer to Figure 4. above API 12J (1989) uses 2.5 feet increments. The vessel height
should be rounded up to the next standard increment. The designer can decide how to allocate the
additional height. Industry practice uses a Height to Diameter ratio of 3 to 5. Arnold and Stewart
(2008) state that a height to diameter ratio of 3 to 4 is common.
a. H ⫽ 1 ⫹ Hsd Low Level ⫹ Hl ⫹ Hsd High Level ⫹ 1 ⫹ Hg ⫹ 1
b. H ⫽ 1 ⫹ 1 ⫹ 1.5 ⫹ 1 ⫹ 1 ⫹ 3.5 ⫹ 1 ⫽ 10 ft
c. Since 10 feet is a standard size H⫽10ft and H/D ⫽ 2.8 depending on the Vessel Ksb
calculation, the height may be increased.
8. The Souders Brown Coefficient for the vessel (Ksbv) is calculated by first calculating the gas
velocity in the vessel and then calculating the vessel Ksb. If the Vessel Ksb is within API 12J
(1989) recommendations for Ksb then the diameter is acceptable. The assumption is that the gas
utilizes 100% of the vessel area. If the vessel Ksb is less than the API 12J (1989) recommendations
then the mist extractor area is reduced by designing a fabrication detail which is inserted in the gas
space. The mesh pas is place into the detail where the open area satisfies the velocity for a correct
Ksb for the mesh pad.
a.

b.

c. The vessel Ksb is much less than the API 12J (1987) recommended values between 0.18 to 0.35
ft/sec. Selecting a value of 0.3 ft/sec, the required gas velocity thru the mesh pad is 3 ft/sec. A
fabrication with an area of 1.12 ft2 would house the mesh pad.

Gas Velocity (Vg) ft/sec Feet per Second

Oil Density (␳o) lbm/Ft3 Pound per Foot Cubed
Gs Density (␳g) lbm/Ft3 Pound per Foot Cubed
Actual Gas Rate (Qg) Ft3/sec Actual Feet Cubed per Second
Diameter (D) ft Feet

9. The vessel can be dimensioned per Figure 4. Figure 5 below illustrates the first pass dimensions.
The vessel diameter and height can be optimized on the basis of API 12J (1987). Standard heights
are in 2.5 foot increments and diameters increase in 0.5 ft. increments.
10. The inlet nozzle, the gas outlet nozzle and liquid outlet nozzle are sized exactly the same as in the
Horizontal Separator Step16.
a. which would require a 6⬙ inlet nozzle
b. which would require a 4⬙ nozzle
c. VI ⫽ 4 to 6 feet per second ⫽ 0.065/4 ⫽ 0.02ft2 which requires a 2⬙ nozzle. To avoid issues
with scaling, sediment and wax the designer should consider increasing the size to 3⬙
11. A simplified Stokes procedure confirms that the liquid section will separate out a gas bubble of
200 micron or less. Although the liquid section is a mix of oil and water, the confirmation process
uses oil only because it is the most conservative approach.
SPE-173598-MS 19

Figure 5—Physical dimensions of the designed separator

a. Oil Section
i. Calculate the Terminal Velocity for the gas bubble rising.
1.

ii. Calculate the separable gas bubble in the gas section assuming laminar flow.
1.

2.

iii. Confirm laminar flow in the oil section by calculating the Reynolds number for the oil
section. If the Reynolds number is less 1 then the flow is laminar.
1. If Re ⫽ (Vtgio ⫻ Dgio ⫻ ␳p)/␮o ⬍ 1 and the flow is laminar and if the separable water
droplet is ⱕ 200 microns then the retention time sizing is valid
20 SPE-173598-MS

2. Re ⫽ (0.007 ⫻ 0.00065 ⫻ 51)/(8/1488) ⫽ 0.04 ⬍ 1, the flow is laminar and if the

separable water droplet is ⱕ 200 microns then the retention time sizing is valid

Conclusions on the 2 phase separator sizing

1. In the author’s experience, there is less emphasis today in sizing 2 phase degassing vessels because
compression is in many cases contracted out. The producer maybe responsible for compressor
failures when the failure is due to liquid carryover. The correct sizing of this type of vessel should
focus on the separable droplet size both in the gas space and the liquid space to ensure minimal
liquid carryover and maximize liquid degassing.
2. Both high and low level shutdowns should be installed to minimize liquid carryover and
subsequent compressor failures and gas carry under which results in excessive emissions or
possible over pressuring low pressure equipment.
3. Attention should be given to both the quality of the level controller and the control valve.
4. Using a simplified form of Stokes Law allows the designer to define where the gas section operates
in the intermediate regime. The diameter is sized on the basis of the 140 micron. The second point
to make about the diameter is that it defines the terminal velocity in the liquid section and by
definition, the diameter determines if the separable droplet is equal to or less than the design
criteria. The retention time ensures that the liquid flow is laminar and the simplified form of Stokes
Equation is valid. The worst case temperature is the lowest temperature. The example illustrates
that what many would say is a more than adequately sized vessel; the separable gas bubble just
meets the sizing criteria due to low oil viscosity.

Nomenclatures
C1 Numerical value of the Length to Diameter ratio for a vertical vessel (dimensionless)
C2 Numerical value of the Height of Liquid to Diameter ratio (dimensionless)
Cdre Drag Coefficient based on the Reynolds number calculation (dimensionless)
Cdge Drag Coefficient based on the Stokes General Equation calculation (dimensionless)
D Vessel Diameter (ft)
Dgio Droplet of Gas in Oil (ft)
Doig Droplet of oil in gas (ft)
Doiw Droplet of Oil in Water (ft)
Dp Diameter of the Droplet in the dispersed phase in the Reynolds number calculation (ft)
Dwio Droplet of water in oil (ft)
g Gravitational Constant (32.2 ft/sec2)
H Heightof Vessel (ft)
Hg Height of the gas space (ft)
Hl Height of Liquid (ft)
Hsd Difference in Height between the High or Low shutdown and height of Liquid (ft)
Ksb Souders Brown Coefficient (ft/sec)
L Vessel Length (ft.)
Ml The percentage of area available for liquid in a horizontal vessel expresses as a decimal
(dimensionless)
Mo The percentage of area available for oil in a horizontal vessel expresses as a decimal
(dimensionless)
Mw The percentage of area available for water in a horizontal vessel expresses as a decimal
(dimensionless)
Qg Gas Volumetric rate (ft3/sec)
SPE-173598-MS 21

Qo Volumetric Rate of Oil (ft3/min)

Qw Volumetric Rate of Water (ft3/min)
Re Reynolds number (dimensionless)
Tro Residence Time of oil (min)
Trw Residence Time for Water (min)
V gas outlet Velocity of the Gas out of the vessel (ft/sec)
V Terminal Velocity of the Dispersed Droplet falling or rising in the continuous phase
(ft/sec)
Vg Gas Velocity (ft/sec)
Vg Velocity of Gas (ft/sec)
Vinlet Velocity of the inlet mixture into vessel (ft/sec)
Vl Volume for Liquid, the sum of Oil and Water, (ft3)
Vo Volume of Oil (ft3)
Vtgio Velocity of the gas droplet in oil (ft/sec)
Vtoig Velocity of the oil droplet in gas (ft/sec)
Vtwio Velocity of the water droplet in oil (ft/sec)
Vtoiw Velocity of the oil droplet in water (ft/sec)
Vw Volume of Water (f3)
␮ Absolute Viscosity of the continuous phase (#/ft-sec)
␮g Absolute Gas Viscosity (lbm/ft-sec)
␮o Absolute Viscosity of oil (lbm/ft-sec)
␮w Absolute Viscosity of water (lbm/ft-sec)
␳ Density of the continuous phase (#/ft3)
␳g Gas Density (lbm/ft3)
␳inlet Composite density of the inlet mixture
␳o Density of oil (lbm/ft3)
␳w Density of water (lbm/ft3)

Acknowledgement
The author would like to thank Mahmood Mosphfegian for his help in both reviewing formatting the
paper.

References
1. Glaso Dead Oil Correlation; PetroWiki published by SPE International
2. Arnold, K. and Stewart, M. 2008, Surface Production Operations Vol 1, third edition, Burlington
Ma. : Gulf Publishing
3. Manning, F and Thompson, R. 1995, Oilfield Processing Vol 2 Crude Oil, Tulsa Ok. : Penwell
Corporation
4. Lapple, C. 1950, Chemical Engineers’ Handbook, Ney York, NY : McGraw Hill
5. Engineering Data Book Vol 1, 1987 tenth edition, Tulsa Ok., Gas processors Association
6. Specifications for Oil and Gas Separators, API Specification 12J, 1989 seventh edition, Wash-
ington DC, American Petroleum Institute
7. Specifications for Vertical an Horizontal Emulsion Treaters, API Specification 12L, 1987 third
edition, Washington DC, American Petroleum Institute