Step-by-step hand

calculation for
three-phase
separator sizing
(vertical and
horizontal)—Part 1
S. AVAJI, ODCC Co., Tehran, Iran; Z.
KHEZRI, PETROFORCE Co., Tehran,
Iran; S. AVAJI, FALCON Co., Tehran, Iran;
and Y. R. KHODAVERDILOU, ODCC
Co., Tehran, Iran

Separators play a crucial role in separation

processes, particularly in chemical
engineering, and the sizing of this equipment
is equally vital. Improper sizing can result in
economic losses and create unfavorable
process conditions.1 The separation process
in this equipment relies on the differences in
density among the various phases, and the
design of the separator must promote
efficient separation.2

In the oil and gas industry, various streams,

including hydrocarbons, water and gas, are
separated under specific temperature and
pressure conditions. This separation process
is primarily conducted using equipment such
as slug catchers and separators. Separators,
as fundamental pieces of process
equipment, must be meticulously designed
to ensure that downstream devices operate
to ensure that downstream devices operate
under optimal conditions.3 Process
engineers recognize that determining the
size of separators—particularly in three-
phase flow situations—presents an intricate
challenge due to the many variables that
must be considered in the design.4

Separators can be configured either

vertically or horizontally. Each design case
should be evaluated individually; however,
based on industrial usage and experience,
vertical separators are recommended for
applications such as compressor knockout
(KO) drums, degassing boots, fuel gas KO
drums, absorber feed KO drums and high-
pressure production separators. Conversely,
horizontal separators are typically used for
reflux drums, three-phase separation and
flare KO drums.5

The primary design principle of separators is

the balance between gravitational and drag
forces acting on the liquid droplet. According
to gravitational theory, a liquid droplet can
settle and separate from the accompanying
phase when the net effect of gravitational
and drag forces is directed downward and
the force of gravity can overcome the drag
force acting on a droplet.6 The two theories
proposed for the sizing of separators are the
droplet-settling theory and retention-time.
The first theory is related to the balance of
drag and gravity forces applied to the
droplet. The second theory states that to
reach the equilibrium of the liquid and gas
phases inside the separator, the liquid must
have a certain residence time. This time
varies across various sizing projects and is
contingent on the process.3 The application
of software like computational fluid
dynamics (CFD), in combination with neural
networks and genetic algorithms for
separator design, is progressing quickly.
Nevertheless, there is a lack of studies that
employ numerical solution techniques for
sizing separators.7–11

Over the years, researchers have consistently

sought to develop methods for sizing
separators. Monnery and Svrcek were
among the first to provide a comprehensive
solution for sizing three-phase separators.12
Their method continues to be widely used by
companies and process engineers. They
outlined all necessary equations for sizing
boot, weir and bucket-type separators, but it
is important to note that their article lacks
references regarding the accuracy of the
equations and the final solution. Additionally,
they did not compare their examples with
real samples or experimental data, nor did
they mention this issue. Boukadi, et al.,13
developed the method proposed by Stewart
and Arnold,3 acknowledging that the
changing specifications of the inlet flow to
the separators over time may result in errors
in sizing. This can lead to either over-design
or under-design in the long term.

It was stated that fluid viscosity does not

affect the relationships described by Stewart
and Arnold; therefore, they introduced a new
retention time that accounts for variations in
viscosity over time. In their research, they
utilized a diagram illustrating emulsion
viscosity as a function of time. They
identified the peak point on the diagram,
which indicated the highest viscosity of the
fluid within a specific time frame, as the
viscosity of the fluid under investigation.
Selecting this specific viscosity of the fluid
resulted in a longer retention time and,
consequently, a larger separator.13

Deng, et al.,14 introduced an equation to

calculate the levels of light and heavy liquids
by analyzing the flow dynamics in a three-
phase separator equipped with a weir and
bucket. Their equation is derived from the
energy equation and incorporates the
resistance of the bucket. The results of their
study indicated that the height of the heavy
liquid within the separator increases with an
increase in flowrate and the ratio of heavy
liquid density to light liquid density.
Conversely, the liquid level decreases as the
difference in height between the light and
heavy liquids in the weir overflow increases.
This indicates that the height disparity
between the light and heavy liquids in weir
overflows can serve as an adjustable input to
regulate the liquid level. Using their proposed
equation, the distinct levels of light and
heavy liquids can be identified, allowing for
the precise determination of the minimum
length required to separate the phases, in
comparison to laboratory findings.14

In their research, Igbagara, et al.,15

conducted a study on the sizing of a three-
phase separator using the method developed
by Monnery and Svrcek.12 This method was
employed to determine the fundamental
dimensions of the separator, including its
diameter and length, as well as the various
liquid levels within the separator (oil, water
and gas). The sizing calculations in their
study were computerized, and the results
obtained exhibited a minimal margin of error
when compared to actual data.15

This article seeks to define the concepts and

principles of design for engineers by
p p g g y

numerically solving various industrial and

real-world examples related to the sizing of
separators. Based on the results obtained
and their comparison with the industrial data
sheets provided by vendors, the numerical
solution method presented in this article can
serve as a reliable initial estimate. In this
context, three examples of three-phase
separators are presented: vertical
separators, horizontal separators with weirs,
and horizontal separators with boots.

VERTICAL THREE-PHASE
SEPARATOR
FIG. 1 illustrates a vertical three-phase
separator. All necessary data have been
extracted from the actual project. The
upcoming sections will detail the
procedures/steps to manually calculate the
dimensions of three-phase vertical
separators (TABLE 1). Some input data, such
as hold-up time, surge time, operating
pressure and operating temperature, are
subject to change, and the designer must
determine them based on project conditions
and criteria.

FIG. 1. Vertical separator

scheme used for three-phase
separation.
Step 1. Calculate or select the K value
(ft/sec). Both the York16 and Gaussian
process spatial alignment (GPSA) methods
are utilized for this step, calculating K using
each of the specified procedures. Ultimately,
the minimum value obtained was chosen.
This value may be specified by the owner
and project process criteria and should be
verified.

For 400 < P (Psi) < 5,500, the York relation16

is calculated using Eq. 1:

0.43 – 0.023 × lnP = 0.43 – 0.023 ×

ln(377) = 0.2935 (1)

and K from the GPSA design practice [0 < P

(Psi) < 1,500] is calculated using Eq. 2:

0.35 – 0.0001 × (P – 100) = 0.35 –

0.0001 × (377 – 100) = 0.3223 -->
selected K = 0.2935 (2)

Step 2. The calculation of vertical terminal

velocity is next (Eq. 3). Terminal velocity is
the highest speed that an object can reach
while descending through a fluid.

Vt = K √[(ρL.L – ρv ) / ρv ] = 0.2935 ×

√[(40.578 – 1.81) / 1.81] = 1.36 (3)

Step 3. The vapor velocity is about 0.75–

0.95 of the terminal velocity (Eq. 4).12

Vv = 0.75 × Vt = 0.75 × 1.36 = 1.02 (4)

Step 4. The vapor disengagement diameter
is calculated using Eqs. 5 and 6:

Dvd = √[4Qv / (π × Vv )] = √[(4 ×

(72,607.02 / 3,600) / (π × 1.02)] =

5.02 (5)

Dvd = 5.02 × 12 = 60.24 (6)

Dvd must be rounded up to the nearest

multiple of 6 in., so 63 in. (or 5.25 ft) is

selected.

Dvd = 5.25

Step 5. The settling velocity of heavy liquid

droplets in light liquid is calculated using Eq.
7:

VHL = [ks × (ρH.L – ρL.L)] / μL.L (7)

ks is Stokes’ law of the terminal velocity

constant ([Link]./min.ft3). This constant

value is typically 0.163, except in cases
where hydrocarbons are present as light
liquids with a specific gravity (SG) of < 0.85
at 60°F and water or caustic is present as a
heavy liquid. In this example, this constant of
0.163 will be maintained. Eq. 8 shows the
numerical values for Eq. 7:

VHL = [0.163 × (65.86 – 40.578)] / 0.1 =

41.21 (8)

Step 6. The rising velocity of light liquid

droplets in heavy liquid is calculated using
Eq. 9:

VLH = [(ks × (ρH.L – ρL.L )) / μH.L ] =

[(0.163 × (65.86 – 40.578)) / 1.8] =

2.29 (9)
Step 7. Calculation of the required time for
heavy liquid droplets to settle through HL,

which represents the length of the light liquid

up to the nozzle. The minimum value for this
distance is 1 ft—for this example, 1 ft is
assumed. This value should be verified
against project criteria (Eq. 10).

HL = 1

tH.L – res = (12HL / VHL ) = [(12 × 1) /

41.21] = 0.29 (10)

Step 7. The calculation of the required time

for settling light liquid droplets through HH ,

which is the height from the bottom of the

bottom tangent line (T.L) up to the heavy
liquid interface (holdup for heavy liquid). The
minimum value for this height is typically set
at 1 ft; however, in this example, it is fixed at
2 ft. This value may vary based on project
criteria, and process engineers should be
mindful of this adjustment (Eq. 11).

HH = 2

tL.L – res = (12HH / VLH ) = [(12 × 2) / 2.29]

= 10.48 (11)

Step 8. The calculations related to the

baffle plate. If a baffle plate is present in the
separator, the area of the baffle plate must
be calculated to determine the area occupied
by heavy and light liquids in each phase.13–
16 Subsequently, the residence time for each

phase will be calculated (Eq. 12) and

compared with the previously calculated
residence time for each phase (FIG. 2).

∆ρ = ρL.L + ρv = 40.578 – 1.81 =

38.76 (12)
 FIG. 2. Baffle
plate and
downcomer
area.

HR indicates the height of light liquid above

the holdup height. Assume a height of 1 ft or

12 in. This value will be checked by the
selected holdup time (TH = 15 min) (Eq. 13).

HR + HL = 1 + 1 = 2 ft = 24

AD = 7.48 × [(QH.L + QL.L ) / G] (13)

G represents the volume of liquid that can be

loaded onto the baffle, considering the
density difference, as illustrated in FIG. 3 and
Eq. 14. The numerical values have been
inserted here into Eq. 14:

G = 9,800

AD = 7.48 × [(38.85 + 155.385) / 9,800] =

0.148 (14)

FIG. 3. Downcomer allowable

flow.

Step 9. In this step, we will recalculate the

value using WD and compare the two

calculated amounts, selecting the larger one.

For this purpose, assume WD is 4 in. (this

value should also be verified against the

separator vendor), and calculate AD using

Eqs. 15 and 16 and TABLE 2.

It can be seen from FIG. 2 that AL = At – AD =

21.64 – 0.577 = 21.07 ft2. The new residence

time for heavy and light liquids can be
calculated using Eqs. 17 and 18:

Now, τL.L–res and τH.L–res should be

compared to tL.L–res and tH.L–res,

respectively. If these values exceed τL.L–res

and τH.L–res, the calculations are correct;

otherwise, the diameter should be increased.

In this example, the condition is met.
Additionally, HR must be compared to the

previously selected HR (Eq. 19).

It is evident that [Link] is greater than the

old value; therefore, rounding up [Link] has

been chosen as the final value ([Link] = 2 ft).

Step 10. The next step is to calculate the

height of the liquid under surge conditions.
Surge conditions refer to the scenario in
which the height of the liquid rises from the
normal liquid level (NLL) to the high liquid
level (HLL) (Eq. 20).

Round up the Hs to the nearest integer value

(Hs = 2 ft).

Step 11. The calculation of the total liquid

volume fraction, the density of the liquid
mixtures, the density of the entire mixture
(including liquid, liquid-vapor) and the total
volumetric flowrate are calculated using Eqs.
21–24, respectively.

Step 12. The nozzle diameter can be

calculated using Eq. 25:
Round up dN to 10 in.

dN = 10 in.

The vapor disengagement height can be

calculated as 0.5D; however, the presence of
a mist eliminator should be carefully studied.
If there is a mist eliminator, Eq. 26 is used.

HD = max [0.5D & ((24 + 0.5dN ) /

12)] (26)

Without a mist eliminator, Eq. 27 is used.

HD = max [0.5D & ((36 + 0.5dN ) /

12)] (27)

In this example, a mist eliminator is

considered (Eq. 28).

HD = max [0.5 × 5.25 & ((24 + 0.5 × 10) /

12)] = max (2.625 & 2.417) = 2.625

(28)

HBN refers to the height of the liquid

measured from the top of the baffle to the

feed nozzle (Eq. 29).

HBN = 0.5 × dN + max(2 & (Hs + 0.5)) =

0.5 × 0.77 + max(2 & (2 + 0.5)) =

2.88 (29)

Round up HBN to 3 ft.

The minimum height of the liquid on the

baffle plate is typically considered to be 0.5
ft. The thickness of the mist eliminator (Hmist

) is at least 0.5 ft, and the height between the

top of the tangent line (TL ) and the mist

eliminator (HU ) has a minimum value of 1 ft

(Eq. 30).

HA = 0.5 & Hmist = 0.5 & HU = 1 (30)

The calculation has been completed, and the

final result is reported in Eq. 31.

HT = HA + HBN + Hmist + HU + HD + HS +

HR + HL + HL (31)

HT = 0.5 + 3 + 0.5 + 1 + 2.625 + 2 + 2 + 1

+ 2 = 14.63 ft = 14.63 × 0.3048 = 4.46 m

D = 5.25 ft = 5.25 × 0.3048 = 1.6 m

Additionally, HT / D should be monitored and

must fall within the range of 1.5 to 6 (Eq. 32).

(HT / D) = (14.63 / 5.25) = 2.78 (32)

HORIZONTAL THREE-PHASE
SEPARATOR (WITH WEIR)
FIG. 4 is a diagram of a real three-phase
separator. The specifications for any outlet
stream are demonstrated here. The diameter
and length of this separator will be
calculated numerically, without relying on
computers, to enhance understanding. This
example serves as a useful guide for initial
calculations; however, sizing must be
checked case by case and may vary
depending on project criteria (TABLE 3). For
instance, residence time is a crucial
parameter that must be discussed and
specified by the client. In this example, a
residence time of 5 min is assumed.
Additionally, a particle diameter of 150
microns is selected for this vessel sizing,
which may differ in other projects.

FIG. 4. Horizontal separator scheme used for

three-phase separation.

FIG. 5. Drag coefficient vs. Reynolds number.17

Step 1. Calculate the C(Re)2 and drag

coefficient from FIG. 5 (Eq. 33).

Step 2. Calculate the settling velocity using

Eq. 34:
Step 3. Standard length-to-diameter (L/D)
ratios for horizontal separators generally
range from 2.5–6. A significant liquid surge
volume is needed to achieve a longer
retention time, which facilitates a more
thorough release of dissolved gas and, if
needed, provides surge volume for the
circulation system (Eq. 35).

L/D = 3.3 (35)

Step 4. The velocity coefficient is typically

regarded as ranging from 0.5 to 0.6. In this
instance, 0.6 is used for this parameter;
however, a designer has the option to modify
and choose different values for their project.

F = 0.6

Step 5. The required vapor velocity is

calculated using Eq. 36:

Vm = F × (L / D) × Vs = 0.6 × 3.3 × 0.197 =

0.39 (36)

Step 6. The light liquid velocity in heavy

liquid is calculated using Eq. 37:
Based on the total design criteria, the
maximum value of UL,L is 250 mm/min;

therefore, UL.L/H.L = 250 is selected for this

velocity.

Step 7. The heavy liquid velocity in light

liquid is calculated using Eq. 38:

Step 8. The required vapor cross sectional

area is calculated using Eq. 39:

Step 9. The liquid height (HLL) to vessel

diameter, which is shown as h1 / D, should be

specified. The value of 0.7 is recommended

for this height and is selected for this
purpose (Eq. 40).

h1 / D = 0.7 (40)

Step 10. Calculate the ratio of the surface

area occupied by liquids to the total surface
area of ​the separator by using Eq. 41:17
Step 11. Calculate the ratio of the surface
area occupied by vapor to the total surface
using Eq. 42:

Av / AT = 1 – (Al / AT ) = 1 – 0.75 =

0.25 (42)

Step 12. Calculate the total area using Eq.

43:

Av / AT = 0.25 --> AT = Av / 0.25 = 0.09 /

0.25 = 0.36 (43)

Step 13. Calculate the liquid area using

Eq. 44:

Al / AT = 0.75 --> Al = 0.75 × 0.36 =

0.27 (44)

Step 14. Calculate the total vessel

diameter using Eq. 45:

D = √4AT / π = √(4 × 0.36) / π = 0.6807 m

= 680.07 (45)

Step 15. According to the calculated value

for the diameter, a rounded-up value is
selected. The next appropriate choice for the
diameter of the separator is 800 mm.
Although 700 mm was initially considered, it
was excluded to simplify the calculations.
However, the reader is encouraged to
perform the calculations using the 700 mm
value if desired.
 D = 800

Step 16. The flow path length is calculated

using Eq. 46:

L/D = 3.3 --> L = 3.3 × 800 = 2,640

(46)

Step 17. The TL – TL length is calculated

using Eq. 47:

L’ = L + 25.4 × (1.5 × (d1 + d2 )) = 2,640 +

25.4 × (1.5 × (4 + 4)) = 2,944.8 (47)

Step 18. The high liquid level (h1 = HLL) is

calculated using Eq. 48:

h1 / D = HLL / D = 0.7 --> h1 = HLL = 0.7 ×

800 = 560 (48)

Step 19. The volume of the HLL is

calculated using Eq. 49:17

Step 20. Select the low liquid level (h2 =

LLL). This value should be specified by the
owner or project criteria (Eq. 50).

h2 = LLL = 700 mm (50)

Step 21. The volume of the LLL is

calculated using Eq. 51:
Step 22. Calculate the surge volume using
Eq. 52:

Vsurge = VHLL – VLLL = 1.1 – 1.37 = –

0.27 (52)

Due to the surge volume being negative, the

diameter must be increased.

Step 23. Select the new diameter and

repeat Steps 16 to 22.

Dnew = 1,550

L = 5,115 mm
l’ = 5,419.8 mm ≈ 5,500 mm
h1 = 1,085 mm

VHLL = 7.65 m3

h2 = 700 mm

VLLL = 4.48 m3

Vsurge = 3.16 m3

Step 24. Calculate the residence time of

light liquid. The calculated residence time
must be ≤ to the residence time selected (5
min) (Eq. 53).
Step 25. Calculate the NLL (hN.L) and
liquid volume at NLL using Eqs. 54 and 55:

At this time, the heavy liquid falling and light

liquid rising should be checked. Calculations
are first performed for the heavy liquid
phase.

Heavy liquid calculation (falling)

Step 26. Calculate the baffle (weir) space

using Eq. 56:

B = (2 / 3)l’ = (2 / 3) × 5,419 =
3,613.2 (56)

Step 27. 3,700 mm is chosen as the first

trial for baffle space height (Eq. 57).

B = 3,700 (57)

Step 28. The total liquid volumetric

flowrate can be calculated using Eq. 58:

Qt = QH.L + QL.L = [mH.L / (60 × ρH.L )] +

[mL.L / (60 × ρL.L )] =

[9,955 / (60 × 988)] + [31,000 / (60 ×

728.4)] = 0.88 (58)

Step 29. Calculate the high liquid level

area using Eq. 59:

AHLL = (VHLL / l’ ) = [7.65 / (5,419 × 10–3)]

= 1.41 (59)
Step 30. Calculate the horizontal velocity
at the HLL area using Eq. 60:

UhHLL = 1,000 × (Qt / AHLL ) = [(1,000 ×

0.88) / 1.41] = 621.8 (60)

Step 31. Calculate the required vertical

distance from HLL for suitable settling using
Eq. 61:

ZHLL = [(B(from step 2) × UH.L / L.L) / UhHLL ]

= [(3,700 × 250) / 621.8] = 1,487.6

(61)

Step 32. Calculate the HLL (or h1 ) – ZHLL

height. This value should be negative until
liquid droplets can reach the bottom of the
vessel. If the value is positive, L/D or HLL/D
must be changed (Eq. 62).

h1 – ZHLL = 1,085 – 1,487.6 = –

402.6 (62)

Step 33. Calculate the LLL area using Eq.

63:

ALLL = VLLL / l’ = [4.48 / (5,419 × 10–3)] =

0.83 (63)

Step 34. Calculate the horizontal velocity

at the LLL area (Eq. 64).

UhHLL = 1,000 × (Qt / ALLL ) = [(1,000 ×

0.88) / 0.83] = 1,060.26 (64)

Step 35. Calculate the required vertical

distance from LLL for suitable settling (Eq.
65).

ZLLL = [(B(from step 2) × UH.L / L.L ) / UhLLL ]

= [(3,700 × 250) / 1,060.26] = 872.43

(65)
Step 36. Calculate the LLL (or h2) – ZHLL
(Eq. 66).

h2 – ZLLL = 700 – 872.43 = –172.43

(66)

Step 37. Select the baffle height (h3) and

HLL for heavy liquid (h4) (HIL); 75 mm lower

is the suggested value for both heights (Eqs.

67 and 68).

h3 = h2 – 75 = 700 – 75 = 625 (67)

h4 = h3 – 75 = 625 – 75 = 550 (68)

Light liquid calculation (rising)

Step 38. Calculate the vertical rise within

the baffle space distance (Eq. 69).

ZVert = [(B(from step 2) × UH.L / L.L ) / UhHLL

] = [(3,700 × 250) / 1,060.26] =

872.43 (69)

Step 39. Calculate the h4 – ZVert height

(Eq. 70).

h4 – ZVert = 550 – 872.43 = –322.43

(70)

Step 40. Select the LLL of the heavy liquid

(LIL = h5 ). The minimum recommended

value for LLL is 300 mm; however, it may vary

across different projects, and the designer
must follow the project criteria.

h5 = LIL = 300

Step 41. Select the vortex breaker height

(h6). A common value for this height is 100
mm, and this value is considered in this
example.

h6 = 100

Step 42. Calculate the NLL for heavy liquid

(hNIL). It is assumed that the height of hNIL

falls between the heights of h4 and h5 (Eq.

71).

hNIL = [(h4 + h5 ) / 2] = [(550 + 300) / 2] =

425 (71)

Step 43. Calculate the heavy liquid volume

at HIL (Eq. 72).

Step 44. Calculate the heavy liquid volume

at LIL (Eq. 73).

Step 45. Calculate the heavy liquid volume

at NIL (Eq. 74).
Step 46. Calculate the heavy liquid volume
at h6 up to the baffle (Eq. 75).

Step 47. Calculate the surge volume for

heavy liquid (Eq. 76).

VSur/H = VHIL – VLIL = 2.22 – 0.95 =

1.27 (76)

Step 48. Calculate the surge time for

heavy liquid (Eq. 77).

Step 49. Calculate the residence time for

heavy liquid (Eq. 78).
Step 50. Calculate the heavy liquid holding
time. The minimum acceptable value is 4
min–5 min, and this limitation must be
checked vs. project criteria (Eq. 79).

tholding / H = tSur/H + tRes/H = 7.56 + 8.12

= 15.68 (79)

Step 51. Calculate the light liquid volume

at NLL up to the vessel (Eq. 80).

VNLL = [(ANLL (refer to step 25 – section 1) × B

) / 1,000] = [(1.12 × 3,700) / 1,000] =

4.16 (80)

Step 52. Calculate the light liquid

residence time (Eq. 81).

The residence time must be ≥ to the selected

residence time of 5 min. Therefore, the
selected baffle distance (B) must be
increased. Through trial and error, the
minimum selected baffle distance is
determined to be 5,200 mm. Additionally, it
should be less than (l’ ) from Step 23. Due to
3.7 min. being < 5 min., all calculations must
be repeated from Step 28, and 1,000 mm
should be selected for LLL or h2.

Bnew = 5000 mm

Finally, tRes/L = 4.97 ≈ 5 min and the

calculated dimensions are summarized in

TABLE 4 and FIG. 6.

FIG. 6. Horizontal three-phase separator scheme

with weir.

Part 2 of this article will appear in the July

issue. HP

LITERATURE CITED
1. Ahmed, T., F. Hamad and P. A. Russell, “The use
of CFD simulations to compare and evaluate
different sizing algorithms for three-phase
separators,” OnePetro, Offshore Technology
Conference Brasil, Rio de Janeiro, Brazil,
October 24–26, 2017.
2. Ling, K., et al., “New method to estimate surface-
separator optimum operating pressures,” Oil and
Gas Facilities, June 2013.
3. Arnold, K. and M. Stewart, Surface production
operations, Volume 2:: Design of gas-handling
systems and facilities, 2nd Ed., Elsevier, 1999.
4. Ahmed, T., et al., “Design and capital cost
optimisation of three-phase gravity separators,”
Heliyon, Vol. 6, Iss. 6, 2020.
 5. Walas, S. M., Chemical process equipment:
Selection and design, Butterworths, January 1,
1988.
6. Wu, F., et al., “Analysis and research on the
automatic control systems of oil-water baffles in
horizontal three-phase separators,” Processes,
June 2022.
7. Kharoua, N., L. Khezzar and H. Saadawi, “CFD
simulation of three-phase separator: Effects of
size distribution,” American Society of
Mechanical Engineers (ASME), Fluids
Engineering Division Summer Meeting, 2013.
8. Khalifat, Z., M. Zivdar and R. Rahimi, “Simulation
of an industrial three-phase boot separator
using computational fluid dynamics,” Journal of
Gas Technology, 2020.
9. Joshy, A., et al., “CFD analysis and optimization
of three phase oil separator,” International
Conference on Systems, Energy and
Environment 2022 (ICSEE 2022).
10. Song, S., et al., “Dynamic simulator for three-
phase gravity separators in oil production
facilities,” ACS Omega, February 2023.
11. Montazeri, V. and A. Ghazi, “Optimization of
three phases horizontal separator using the
genetic algorithm (GA) method to reduce
manufacturing costs and promote the phase
separation process in the petroleum industry,”
Iranian Journal of Oil and Gas Science and
Technology, 2023.
12. Monnery, W. D. and W. Y. Svrcek, “Successfully
specify three-phase separators,” Chemical
Engineering Progress, September 1994.
13. Boukadi, F., et al., “Appropriate separator sizing:
A modified Stewart and Arnold method,”
Modelling and Simulation in Engineering,
November 2012.
14. Deng, C., et al., “Modeling liquid-liquid interface
level in a horizontal three-phase separator with a
bucket and weir,” Journal of Dispersion Science
and Technology, 2018.
15. Igbagara, P., “Design of three phase separator
for 12.5 bpd crude oil atmospheric distillation
unit,” FUPRE Journal of Scientific and Industrial
Research (FJSIR), Vol. 7, No. 1, 2023.
16. York, R. J. and H. L. Pastrick, “Optimal terminal
guidance with constraints at final time,” Journal
of Spacecraft and Rockets, June 1977.
17. company, T., process engineering design
manual. 1991: p. 29.
SINA AVAJI serves as a Process Engineer
in the Process department at Oil Design and
Construction Co. (ODCC) located in Tehran,
Iran. Before that, he spent approximately 3 yr
at Sazeh Co. Previously, Avaji served as
Technical Director at the Shiraz University
Laboratory for 2 yr. The author can be
reached at sinaavaji@[Link].

ZAHRA KHEZRI has been a Process

Engineer in the Process department at
Petroforce Co. for 2 yr. Khezri is a graduate
of the University of Tehran with an MS
degree in chemical engineering.

SAEED AVAJI is a Process Engineer in the

Process department at Falcon Co. He has
been with Falcon Co. for 5 yr. Saeed is a
graduate of the Shiraz University of
Technology.

YASER RASTI KHODAVERDILOU has 2 yr

of professional experience as a Process
Engineer within the Process department at
Oil Design and Construction Company
(ODCC). Prior to this role, he accumulated 2
yr of experience at Oil Industries
Commissioning and Operation Company
(OICO), and an additional 2 yr at Hampa
Engineering Corp. Khodaverdilou is engaged
in doctoral studies in chemical engineering
at Shiraz University, where he is in the final
stages of defending his doctoral thesis.

Scroll down to read