# Design and Analysis of a Three-Phase Horizontal Separator

VerifiedAdded on 2019/10/18

|22

|6063

|360

Report

AI Summary

The assignment content is about designing a three-phase horizontal separator with a specific set of dimensions and operating conditions. The design process involves selecting materials, calculating thicknesses for the component parts, and determining the most economical option for the separator's head. The chosen material for construction is carbon-manganese steel, which can withstand internal pressures and corrosion. The minimum thickness required for the cylindrical shell, torispherical-head, and ellipsoidal-head are calculated using various equations. The results indicate that an ellipsoidal-head with a wall thickness of 81.85mm is the most economical option.

## Contribute Materials

Your contribution can guide someone’s learning journey. Share your
documents today.

CHAPTER ONE

1.0 Background

Hydrocarbons which are produced from wells whether oil or gas are not in the form of simple mixtures but of very complex

form of liquid or gas mixture of hundreds of different composites of chemically bonded compounds. A well stream typically

is turbulent of mixed substances such as gas, oil, water including minerals of dissolved form containing a substantial

amount of salt, corrosive gasses such as Carbon dioxide (CO2) and Hydrogen sulfide (H2S), and different solids e.g. sands of

unconsolidated forms producing intervals on a reservoir from which particles of sands may become detached at the

slightest flow rate of production. Therefore, the need for separation processes to align these components to their export

standards and bring about the flow of branched-out but simultaneous components is carried out by the separator.

Nonetheless, the types of separators used in the oil and gas industry are three which vary in terms of both form and

function such as Horizontal separator, vertical separator and spherical separator. The Horizontal type can be subdivided

into even further sections such as single tube and double tubes. Both horizontal and vertical separators may be classed as

two phase and or three phase separators, this depends on their specifications and purposes. Depending on the objective

targeted, a two phase separator will be used for the separation of a two phase only (oil-gas) or (gas-water) separation and a

three phase separator is used for the separation of a gas from water and gas equally. Among the entire field processing

operations, the separation of a gas from free-liquid has shown to be the most critical stage. The desired size of a particular

separation unit is equally (directly) proportional to that of a composition of fluid mixture and pressure. Hence, this project

aims to demonstrate for the design of a three-phase horizontal separator, whilst considering factors that influences the

design of a separator including; oil and gas capacity constraints, bending moments and stress, design materials, cost for

design amongst others.

However, Guo, B. et al, (2007) argues that separators function on the basis of gravity segregation/settling or centrifugal

segregation. The separator construction normally depends on the:

Inlet device, where the primary separation of gas-liquid bulk occurs.

A large settling section of adequate height or length to allow for liquid droplets.

Equipped with a mist extractor near the gas outlet to coalesce any entrained liquid droplets that do not settle due to

gravity.

adequate controls, such as: liquid level controller, placed between oil-water level and in the oil section, liquid dump valves,

gas back-pressure valve, pressure gauges, instrument gas regulator and piping.

1.0 Background

Hydrocarbons which are produced from wells whether oil or gas are not in the form of simple mixtures but of very complex

form of liquid or gas mixture of hundreds of different composites of chemically bonded compounds. A well stream typically

is turbulent of mixed substances such as gas, oil, water including minerals of dissolved form containing a substantial

amount of salt, corrosive gasses such as Carbon dioxide (CO2) and Hydrogen sulfide (H2S), and different solids e.g. sands of

unconsolidated forms producing intervals on a reservoir from which particles of sands may become detached at the

slightest flow rate of production. Therefore, the need for separation processes to align these components to their export

standards and bring about the flow of branched-out but simultaneous components is carried out by the separator.

Nonetheless, the types of separators used in the oil and gas industry are three which vary in terms of both form and

function such as Horizontal separator, vertical separator and spherical separator. The Horizontal type can be subdivided

into even further sections such as single tube and double tubes. Both horizontal and vertical separators may be classed as

two phase and or three phase separators, this depends on their specifications and purposes. Depending on the objective

targeted, a two phase separator will be used for the separation of a two phase only (oil-gas) or (gas-water) separation and a

three phase separator is used for the separation of a gas from water and gas equally. Among the entire field processing

operations, the separation of a gas from free-liquid has shown to be the most critical stage. The desired size of a particular

separation unit is equally (directly) proportional to that of a composition of fluid mixture and pressure. Hence, this project

aims to demonstrate for the design of a three-phase horizontal separator, whilst considering factors that influences the

design of a separator including; oil and gas capacity constraints, bending moments and stress, design materials, cost for

design amongst others.

However, Guo, B. et al, (2007) argues that separators function on the basis of gravity segregation/settling or centrifugal

segregation. The separator construction normally depends on the:

Inlet device, where the primary separation of gas-liquid bulk occurs.

A large settling section of adequate height or length to allow for liquid droplets.

Equipped with a mist extractor near the gas outlet to coalesce any entrained liquid droplets that do not settle due to

gravity.

adequate controls, such as: liquid level controller, placed between oil-water level and in the oil section, liquid dump valves,

gas back-pressure valve, pressure gauges, instrument gas regulator and piping.

## Secure Best Marks with AI Grader

Need help grading? Try our AI Grader for instant feedback on your assignments.

2

CHAPTER TWO

2.0 Methodology

Since the separator that works on the basis of gravity segregation or settling, the multi-phase fluid that enters the separator

is separated as a result of gravity and density differences. Gas phase occupies the upper part of the vessel, while the oil and

water fill the lower section of the vessel and a layer of two immiscible phases, oil and water, occurs on the top of the water

layer-emulsion layer. The layer above the emulsion is the oil layer, as shown in Fig.1. It is essential to understand emulsions,

as it limits the liquid/liquid separation. If the emulsion layer is too big an emulsion-breaker or demulsifier, is used to reduce

the thickness of that layer.

Figure 1: A schematic of three-phase horizontal separator(Stewart, M. and Arnold, K.,2008).

Depending on the flow rate of the incoming mixture, the reaction force from the wall of the vessel generates a centripetal

acceleration of up to 500 times the gravitational acceleration. This centripetal action forces the liquid droplets to coalesce

to bigger droplets and fall to the bottom of the separator section and the gas to leave from the top due to density

differences. Guo, B. et al (2007). However, in a momentum breaker-inlet device, a sudden change in momentum occurs

when the fluid stream hits it. This sudden change in momentum generates the primary gas-liquid separation in which

then the liquid splashes into the liquid pool, the lower part of the vessel, and flow as dispersion with low horizontal

velocity, while the gas together with entrained liquid droplets flow horizontally in the upper part of the separator. Inlet

diverters or momentum breakers normally designed with a liquid down comer that directs the liquid flow below the oil-

water interface.

According to Arnold, K et., al (2008) the function of the down comer is to force the oil-water mixture (emulsion) to mix

with the water being the continuous phase at the bottom of the vessel and rise through the oil-water interface. This

process is called “water washing” and it promotes the coalescence of water droplets, which are entrained in the oil phase.

The inlet diverter assures that little gas is carried with the liquid, and the water wash assures that the liquid does not fall

on top of the gas-oil or oil-water interface.

3

Weir

2.0 Methodology

Since the separator that works on the basis of gravity segregation or settling, the multi-phase fluid that enters the separator

is separated as a result of gravity and density differences. Gas phase occupies the upper part of the vessel, while the oil and

water fill the lower section of the vessel and a layer of two immiscible phases, oil and water, occurs on the top of the water

layer-emulsion layer. The layer above the emulsion is the oil layer, as shown in Fig.1. It is essential to understand emulsions,

as it limits the liquid/liquid separation. If the emulsion layer is too big an emulsion-breaker or demulsifier, is used to reduce

the thickness of that layer.

Figure 1: A schematic of three-phase horizontal separator(Stewart, M. and Arnold, K.,2008).

Depending on the flow rate of the incoming mixture, the reaction force from the wall of the vessel generates a centripetal

acceleration of up to 500 times the gravitational acceleration. This centripetal action forces the liquid droplets to coalesce

to bigger droplets and fall to the bottom of the separator section and the gas to leave from the top due to density

differences. Guo, B. et al (2007). However, in a momentum breaker-inlet device, a sudden change in momentum occurs

when the fluid stream hits it. This sudden change in momentum generates the primary gas-liquid separation in which

then the liquid splashes into the liquid pool, the lower part of the vessel, and flow as dispersion with low horizontal

velocity, while the gas together with entrained liquid droplets flow horizontally in the upper part of the separator. Inlet

diverters or momentum breakers normally designed with a liquid down comer that directs the liquid flow below the oil-

water interface.

According to Arnold, K et., al (2008) the function of the down comer is to force the oil-water mixture (emulsion) to mix

with the water being the continuous phase at the bottom of the vessel and rise through the oil-water interface. This

process is called “water washing” and it promotes the coalescence of water droplets, which are entrained in the oil phase.

The inlet diverter assures that little gas is carried with the liquid, and the water wash assures that the liquid does not fall

on top of the gas-oil or oil-water interface.

3

Weir

2.1 Factors Affecting Separation Processes

There are some factors that affect the separation efficiency in a separator. The following factors must be determined

before separation design:

Flow Rate of Gas and liquid.

Operating and design pressures and temperature

Physical properties of the fluid such as , Z, , Sg, etc.,

Surging or slugging tendencies of feed streams.

Size of droplet to be removed.

Presence of impurities e.g: Paraffin, sand, scale, etc.,

Foaming tendencies of the crude oil,

Corrosive tendencies of the liquid or gas.

Arnold, K. 2007, reported challenges associated with separation processes such as:

Foaming: Pressure reduction in certain types of crude oils may produce foaming. It occurs when small bubbles of gases

that are dissolved in the oil start to come out of solution, and are incased in a thin film of oil.

Emulsion: Materials become emulsified due to the presence of oil and water at the same time. The liquid/liquid

separation will be limited, if a large emulsion layer exists.

Corrosion: Produced well fluids can be very corrosive and may cause the primary failure of process equipment. The two

most corrosive elements are hydrogen sulfide and carbon dioxide. These two gases may present in the well fluids in

quantities from a race up to 40% to 50% of the gas by volume.

Paraffin: The deposition of paraffin in separators could reduce their efficiency through the build-up action inside the

vessel. Paraffin build-up may block the fluid passages and the mist extractor, and consequently leads to costly operation

shut downs.

Sand: When dealing with separators, sand could have a worrying effect; as it is capable of causing metal erosion,

especially in valves and chocks. In addition, it can accumulate at the bottom of the separator leading to blockage or

clogging of the separator internals.

4

There are some factors that affect the separation efficiency in a separator. The following factors must be determined

before separation design:

Flow Rate of Gas and liquid.

Operating and design pressures and temperature

Physical properties of the fluid such as , Z, , Sg, etc.,

Surging or slugging tendencies of feed streams.

Size of droplet to be removed.

Presence of impurities e.g: Paraffin, sand, scale, etc.,

Foaming tendencies of the crude oil,

Corrosive tendencies of the liquid or gas.

Arnold, K. 2007, reported challenges associated with separation processes such as:

Foaming: Pressure reduction in certain types of crude oils may produce foaming. It occurs when small bubbles of gases

that are dissolved in the oil start to come out of solution, and are incased in a thin film of oil.

Emulsion: Materials become emulsified due to the presence of oil and water at the same time. The liquid/liquid

separation will be limited, if a large emulsion layer exists.

Corrosion: Produced well fluids can be very corrosive and may cause the primary failure of process equipment. The two

most corrosive elements are hydrogen sulfide and carbon dioxide. These two gases may present in the well fluids in

quantities from a race up to 40% to 50% of the gas by volume.

Paraffin: The deposition of paraffin in separators could reduce their efficiency through the build-up action inside the

vessel. Paraffin build-up may block the fluid passages and the mist extractor, and consequently leads to costly operation

shut downs.

Sand: When dealing with separators, sand could have a worrying effect; as it is capable of causing metal erosion,

especially in valves and chocks. In addition, it can accumulate at the bottom of the separator leading to blockage or

clogging of the separator internals.

4

## Paraphrase This Document

Need a fresh take? Get an instant paraphrase of this document with our AI Paraphraser

2.2 THEORY OF SEPARATION

2.3 Settling velocity

In a three-phase separator, the upper part represents the gravity settling section of the gas-phase in which oil droplets are

entrained with the gas. The relative motion exists between oil droplets and the surrounding gas. An oil droplet, being

denser than the surrounding gas tends to move vertically downwards due to gravitational or buoyant force, F g, while the

surrounding gas exerts a drag force on the opposite direction of that oil droplet, Fd. Consequently, the oil droplet will

accelerate downward until the drag force balances the gravitational or buoyant force; and therefore, the droplet will fall at

a uniform velocity, known as terminal or settling velocity.

The low half of the three-phase separator is occupied by the liquid phase (oil and water). The retention time constraint of

the liquid in a three-phase separator must be considered for both oil droplets in water and water droplets in oil. Similar to

gas settling section, in separating oil droplets from water or water droplets from oil, a relative motion exists between the

droplet and the surrounding continuous phase.

The following equation shows how to determine terminal velocity:

Terminal velocity of oil droplet from gas:

V t = ( 0 . 01186 ) [ ( ρo −ρg

ρg )x dm

Cd ]0 .5

[1]

Terminal velocity of oil:

V t = ( 1 .787 x 10−6 ) x ( △ γ ) dm

2

μ [2]

Where:

V t = Terminal velocity of the droplet (ft/s).

△ γ =(γ ¿¿ w−γo )¿= the difference in specific gravities of water and oil.

μ=Viscosity of the continuous phase (cP.)

dm= diameter of the droplet (microns).

ρo = Density oil oil (Ib/ft3)

ρg= Density of gas (Ib/ft3)

5

2.3 Settling velocity

In a three-phase separator, the upper part represents the gravity settling section of the gas-phase in which oil droplets are

entrained with the gas. The relative motion exists between oil droplets and the surrounding gas. An oil droplet, being

denser than the surrounding gas tends to move vertically downwards due to gravitational or buoyant force, F g, while the

surrounding gas exerts a drag force on the opposite direction of that oil droplet, Fd. Consequently, the oil droplet will

accelerate downward until the drag force balances the gravitational or buoyant force; and therefore, the droplet will fall at

a uniform velocity, known as terminal or settling velocity.

The low half of the three-phase separator is occupied by the liquid phase (oil and water). The retention time constraint of

the liquid in a three-phase separator must be considered for both oil droplets in water and water droplets in oil. Similar to

gas settling section, in separating oil droplets from water or water droplets from oil, a relative motion exists between the

droplet and the surrounding continuous phase.

The following equation shows how to determine terminal velocity:

Terminal velocity of oil droplet from gas:

V t = ( 0 . 01186 ) [ ( ρo −ρg

ρg )x dm

Cd ]0 .5

[1]

Terminal velocity of oil:

V t = ( 1 .787 x 10−6 ) x ( △ γ ) dm

2

μ [2]

Where:

V t = Terminal velocity of the droplet (ft/s).

△ γ =(γ ¿¿ w−γo )¿= the difference in specific gravities of water and oil.

μ=Viscosity of the continuous phase (cP.)

dm= diameter of the droplet (microns).

ρo = Density oil oil (Ib/ft3)

ρg= Density of gas (Ib/ft3)

5

CD = Drag coefficient.

In considering equation 2, the settling velocity of a droplet is inversely proportional to the viscosity of the continuous

phase. Knowing that the viscosity of oil is several times higher than that of the water, therefore the terminal velocity of oil

droplets in water is greater than the terminal velocity of water droplet in the oil. Since, the required time for a droplet to

settle out of one continuous phase and reach the oil-water interface depends on the terminal velocity and the distance

traveled by the droplet, therefore the separation of water droplets from the continuous oil phase would always be taken as

the design criterion for three-phase separators (Abdel-Aal, H., et al, 2003).

2.4 Drag coefficient ( CD)

The drag coefficient (CD) concept, mentioned previously, is very important in separator designs and it is a function of

Reynolds number and the shape of particle of the flowing gas. For the purpose of design, the shape of the particle is

considered to be a solid, rigid sphere. The value of the drag coefficient can be determined using trial-and-error solution,

since both particle size (dm) and terminal velocity (Vt) are involved.

According to (Stewart, M. and Arnold, K., 2008), for production facility design, Stokes’ law does not govern, and the

following formula for drag coefficient must be used instead.

CD =0 . 34+ 24

Re

+ 3

¿ ¿¿ ¿ [3]

The terminal velocity is given by equations (1 and 4), when the drag force acting on the droplet is equal to the buoyant

force:

V t =0 . 01186 x ¿ ¿ [4]

Re=(0 . 0049) x ρg x dm x V t

μ [5]

Where:

Re = Reynolds number (dimensionless).

2.5 Retention time

According to (Stewart, M. and Arnold, K.,2008), retention time is one of the important aspects that need to be taken into

consideration when designing a separator. It can be defined as the following “the average time a molecule of liquid is

retained in the vessel, assuming plug flow”. It determines the required liquid volumes within the separator, so in other

words, it is the volume of liquid storage in the vessel divided by liquid flow rate. In a three-phase separator, a sufficient

6

In considering equation 2, the settling velocity of a droplet is inversely proportional to the viscosity of the continuous

phase. Knowing that the viscosity of oil is several times higher than that of the water, therefore the terminal velocity of oil

droplets in water is greater than the terminal velocity of water droplet in the oil. Since, the required time for a droplet to

settle out of one continuous phase and reach the oil-water interface depends on the terminal velocity and the distance

traveled by the droplet, therefore the separation of water droplets from the continuous oil phase would always be taken as

the design criterion for three-phase separators (Abdel-Aal, H., et al, 2003).

2.4 Drag coefficient ( CD)

The drag coefficient (CD) concept, mentioned previously, is very important in separator designs and it is a function of

Reynolds number and the shape of particle of the flowing gas. For the purpose of design, the shape of the particle is

considered to be a solid, rigid sphere. The value of the drag coefficient can be determined using trial-and-error solution,

since both particle size (dm) and terminal velocity (Vt) are involved.

According to (Stewart, M. and Arnold, K., 2008), for production facility design, Stokes’ law does not govern, and the

following formula for drag coefficient must be used instead.

CD =0 . 34+ 24

Re

+ 3

¿ ¿¿ ¿ [3]

The terminal velocity is given by equations (1 and 4), when the drag force acting on the droplet is equal to the buoyant

force:

V t =0 . 01186 x ¿ ¿ [4]

Re=(0 . 0049) x ρg x dm x V t

μ [5]

Where:

Re = Reynolds number (dimensionless).

2.5 Retention time

According to (Stewart, M. and Arnold, K.,2008), retention time is one of the important aspects that need to be taken into

consideration when designing a separator. It can be defined as the following “the average time a molecule of liquid is

retained in the vessel, assuming plug flow”. It determines the required liquid volumes within the separator, so in other

words, it is the volume of liquid storage in the vessel divided by liquid flow rate. In a three-phase separator, a sufficient

6

period of time has to be given for the oil to reach equilibrium for the dissolved gas to be liberated. Retention time has to be

long enough for water droplets, entrained in the oil, to coalesce and settle to water zone. Simultaneously, water phase

needs to be retained for sufficient period of time to allow for the coalescence of oil droplets, suspended in water.

2.6 Gas capacity

According to Arnold, K., and Steward, M. (2008), the principles of liquid droplets settling from a gas bulk to liquid-gas

interface can be used to develop an equation to size a separator for a gas flow rate. The gas capacity constraint equations

are based on setting the gas retention time equal to the time required for a droplet to settle to the liquid interface. Fora

half-full horizontal vessel of liquid, and separation of 100 micron liquid droplets from the gas, the following equation may

be used:

d Leff =240 x [ TZ Qg

P ] x [ [ ρg

ρl −ρg ][ Cd

dm ] ]1/ 2

[1]

Where:

d = vessel internal diameter (inches).

Leff = the effective length of the vessel where separation occurs (ft).

T = operating temperature (° R ¿.

Qg= gas flow rate (MMscf/day).

P = operating pressure (psia).

Z = gas compressibility.

CD = drag coefficient.

dm = liquid droplet to be separated (micron).

ρg= density of gas (Ib/ft3).

ρl = density of liquid (Ib/ft3).

2.7 Liquid Capacity Retention Time Constraint

The separator size must provide sufficient space for the oil and water such that each phase is retained adequately within

the separator. The assumption used is that the liquid will occupy half of the separator volume, as Stewart, M. and Arnold,

K., (2008) suggested. However, in the present case, both oil and water occupy that volume. Therefore, the volume occupied

by the liquid-phase (both oil and water), VL, is the sum of the volume occupied by oil (Vo) and water(Vw) in a separator

having a diameter, D (in.), and effective length, L (ft). So the combination of vessel diameter (d) and effective length (L eff)

can be calculated using the following equation:

D2Leff = 1.429¿ [7]

Where:

7

long enough for water droplets, entrained in the oil, to coalesce and settle to water zone. Simultaneously, water phase

needs to be retained for sufficient period of time to allow for the coalescence of oil droplets, suspended in water.

2.6 Gas capacity

According to Arnold, K., and Steward, M. (2008), the principles of liquid droplets settling from a gas bulk to liquid-gas

interface can be used to develop an equation to size a separator for a gas flow rate. The gas capacity constraint equations

are based on setting the gas retention time equal to the time required for a droplet to settle to the liquid interface. Fora

half-full horizontal vessel of liquid, and separation of 100 micron liquid droplets from the gas, the following equation may

be used:

d Leff =240 x [ TZ Qg

P ] x [ [ ρg

ρl −ρg ][ Cd

dm ] ]1/ 2

[1]

Where:

d = vessel internal diameter (inches).

Leff = the effective length of the vessel where separation occurs (ft).

T = operating temperature (° R ¿.

Qg= gas flow rate (MMscf/day).

P = operating pressure (psia).

Z = gas compressibility.

CD = drag coefficient.

dm = liquid droplet to be separated (micron).

ρg= density of gas (Ib/ft3).

ρl = density of liquid (Ib/ft3).

2.7 Liquid Capacity Retention Time Constraint

The separator size must provide sufficient space for the oil and water such that each phase is retained adequately within

the separator. The assumption used is that the liquid will occupy half of the separator volume, as Stewart, M. and Arnold,

K., (2008) suggested. However, in the present case, both oil and water occupy that volume. Therefore, the volume occupied

by the liquid-phase (both oil and water), VL, is the sum of the volume occupied by oil (Vo) and water(Vw) in a separator

having a diameter, D (in.), and effective length, L (ft). So the combination of vessel diameter (d) and effective length (L eff)

can be calculated using the following equation:

D2Leff = 1.429¿ [7]

Where:

7

## Secure Best Marks with AI Grader

Need help grading? Try our AI Grader for instant feedback on your assignments.

Qw = water flow rate (bbl/day).

¿ = water retention time (min.).

( Qo )= oil flow rate (bbl/day).

¿ = oil retention time (min.).

d = vessel internal diameter (inches).

Leff = the effective length of the vessel where separation occurs (ft)

2.8 Seam-to-Seam length

The effective length where the phase separation occurs can be determined, if the liquid retention time constraint governs,

otherwise equation (6) can be used to determine the effective length. From this, a vessel seam-to-seam length may be

determined. The actual seam-to-seam length depends on the physical design of the internals in a vessel Fig.3.

Some vessels are sized based on a gas capacity constraint; therefore, some portion of the vessel length is required to

distribute the flow evenly near the inlet diverter.

Another portion of the vessel length is required for the mist extractor. The length of the vessel between the inlet diverter

and the mist extractor with evenly distributed flow is known as the effective vessel length ( Leff ¿. This effective length is

directly proportional to the diameter of the vessel, so as a vessel’s diameter increases more length is required to distribute

the flow of the gas evenly after the inlet diverter.

The seam-to-seam length can be determined from the following equation, if the gas capacity constraint governs:

Lss= Leff

12 [2]

Figure 1: a schematic of a horizontal separator showing the

effective and seam-to-seam lengths (Arnold, K., and Steward, M., 2008).

However, sizing a vesselwhere the liquid capacity governs, some portion of the vessel’s length must be considered for inlet

diverter flow distribution and liquid outlet. The seam-to-seam length should not exceed the following:

8

¿ = water retention time (min.).

( Qo )= oil flow rate (bbl/day).

¿ = oil retention time (min.).

d = vessel internal diameter (inches).

Leff = the effective length of the vessel where separation occurs (ft)

2.8 Seam-to-Seam length

The effective length where the phase separation occurs can be determined, if the liquid retention time constraint governs,

otherwise equation (6) can be used to determine the effective length. From this, a vessel seam-to-seam length may be

determined. The actual seam-to-seam length depends on the physical design of the internals in a vessel Fig.3.

Some vessels are sized based on a gas capacity constraint; therefore, some portion of the vessel length is required to

distribute the flow evenly near the inlet diverter.

Another portion of the vessel length is required for the mist extractor. The length of the vessel between the inlet diverter

and the mist extractor with evenly distributed flow is known as the effective vessel length ( Leff ¿. This effective length is

directly proportional to the diameter of the vessel, so as a vessel’s diameter increases more length is required to distribute

the flow of the gas evenly after the inlet diverter.

The seam-to-seam length can be determined from the following equation, if the gas capacity constraint governs:

Lss= Leff

12 [2]

Figure 1: a schematic of a horizontal separator showing the

effective and seam-to-seam lengths (Arnold, K., and Steward, M., 2008).

However, sizing a vesselwhere the liquid capacity governs, some portion of the vessel’s length must be considered for inlet

diverter flow distribution and liquid outlet. The seam-to-seam length should not exceed the following:

8

Lss =( 4

3 ) Leff [3]

2.9 Slenderness ratio

In using equations (6,7,8 and9), depending on which constraint governs, various choices of vessel diameter and length are

permitted. For each vessel design, there is a combination of Leff and d that will minimize the cost of the vessel. It can be

shown that the smaller the diameter, the less the vessel will weigh and thus the lower its cost. However, there is a limit or a

point, where decreasing the vessel diameter too much will increase the possibility of high velocity in the gas flow rate and

create waves, which have high momentum that can break liquid droplets into smaller ones. This action will re-entrain liquid

droplets at the gas-liquid interface.

AsStewart, M. and Arnold, K., (2008) stated, that the ratio of the seam-to- seam length divided by the vessel’s diameter

should be between 3 and 5.

This ratio is known as the slenderness ratio (SR) of a vessel. Slenderness ratios between the 3 to 5 range is recommended.

On the other hand, slenderness ratios outside that range may be used, but the design should be checked to assure that re-

entrainment of the liquid droplets would not occur.

2.1.1 Materials

According to Coulson&Richardson’s, (2005), separators or pressure vessels, regardless of the configuration, are normally

constructed from plain carbon steels, low and high steel grades, other alloys, and reinforced plastics. So, when deciding the

suitable material of construction, it must be taken into consideration the compatibility of the selected material with the

process environment, and the suitability of the material for fabrication, especially welding.

2.1.2 Design Stress (nominal design strength)

In any design, it is necessary to choose a value of the maximum allowable stress or nominal design strength, which is

acceptable for material construction. This can be achieved through applying a suitable design stress factor or safety factor

to the maximum stress that the material could be expected to resist without failure. Safety factors are essential in any

design, as they allow for any uncertainties in the design methods used for determining; for instance, the loading, quality of

the chosen material, and the workmanship. Materials that are not subjected to high temperatures, however, the design

stress is, normally, based on the yield stress (or proof stress), or the tensile strength (ultimate tensile stress) of the material

at the design temperature. Typical design stress values for some of the common materials that can be used for preliminary

designs are shown in Table 6 under mechanical design section below.

2.1.3 Corrosion Allowance

Any material is likely to be subjected to corrosion if it was used is processes that deal with corrosive gasses, such as

hydrogen sulfide and carbon dioxide, or liquids. Therefore, in any major unit design, an additional thickness of metal should

be added to allow for material lost by corrosion, erosion, or scaling processes. This is known as “corrosion allowance”, and

there is no specific rule for the estimation of the corrosion allowance required for all circumstances, however, it is mainly

9

3 ) Leff [3]

2.9 Slenderness ratio

In using equations (6,7,8 and9), depending on which constraint governs, various choices of vessel diameter and length are

permitted. For each vessel design, there is a combination of Leff and d that will minimize the cost of the vessel. It can be

shown that the smaller the diameter, the less the vessel will weigh and thus the lower its cost. However, there is a limit or a

point, where decreasing the vessel diameter too much will increase the possibility of high velocity in the gas flow rate and

create waves, which have high momentum that can break liquid droplets into smaller ones. This action will re-entrain liquid

droplets at the gas-liquid interface.

AsStewart, M. and Arnold, K., (2008) stated, that the ratio of the seam-to- seam length divided by the vessel’s diameter

should be between 3 and 5.

This ratio is known as the slenderness ratio (SR) of a vessel. Slenderness ratios between the 3 to 5 range is recommended.

On the other hand, slenderness ratios outside that range may be used, but the design should be checked to assure that re-

entrainment of the liquid droplets would not occur.

2.1.1 Materials

According to Coulson&Richardson’s, (2005), separators or pressure vessels, regardless of the configuration, are normally

constructed from plain carbon steels, low and high steel grades, other alloys, and reinforced plastics. So, when deciding the

suitable material of construction, it must be taken into consideration the compatibility of the selected material with the

process environment, and the suitability of the material for fabrication, especially welding.

2.1.2 Design Stress (nominal design strength)

In any design, it is necessary to choose a value of the maximum allowable stress or nominal design strength, which is

acceptable for material construction. This can be achieved through applying a suitable design stress factor or safety factor

to the maximum stress that the material could be expected to resist without failure. Safety factors are essential in any

design, as they allow for any uncertainties in the design methods used for determining; for instance, the loading, quality of

the chosen material, and the workmanship. Materials that are not subjected to high temperatures, however, the design

stress is, normally, based on the yield stress (or proof stress), or the tensile strength (ultimate tensile stress) of the material

at the design temperature. Typical design stress values for some of the common materials that can be used for preliminary

designs are shown in Table 6 under mechanical design section below.

2.1.3 Corrosion Allowance

Any material is likely to be subjected to corrosion if it was used is processes that deal with corrosive gasses, such as

hydrogen sulfide and carbon dioxide, or liquids. Therefore, in any major unit design, an additional thickness of metal should

be added to allow for material lost by corrosion, erosion, or scaling processes. This is known as “corrosion allowance”, and

there is no specific rule for the estimation of the corrosion allowance required for all circumstances, however, it is mainly

9

based on the experience with the material of construction under similar service conditions to those proposed for the new

design. As Coulson& Richardson’s, (2005) suggestions in the book, for carbon and low-alloy steels, where severe corrosion is

not expected, a minimum allowance of 2.0 mm should be used; where more severe conditions are expected minimum

thickness this should be increased to 4.0 mm. A minimum corrosion allowance of 1.0 mm is the standard in most of design

codes. Process design:

A three-phase inlet separator unit is designed acceding to the data given in table 1( E.W.M. Hansen and William Y., et al,

2012) were taken from the first stage inlet separator. Some assumptions were made in order to carry out the design.

Parameters Gas Oil Water Field units

Flow rate (Q) 51.96

(MMscf/d)

92586 (bbl/d) 14441.33 (bbl/d) -

Density (ρ) 3.103 51.91 64.3 Ib/ft3

Viscosity (μ) 0.013 5.25 0.43 Cp

Specific gravity (Sg) 0.683 0.8403 1.04 Ib / ft 3

Ib / ft 3

GOR 561.21 - Scf/STB

API of oil 36.9

Compressibility factor (Z) 0.99 - Dimensionless

Droplet removal size

(assumed)

- 100 500 Microns (μm ¿

Retention times

(Assumed)

- 5 5 minutes

Operating conditions

Pressure 996.41 Psia

Temperature 592 °R

Table 1: The physical properties of the fluid and production rates at operation conditions.(E.W.M.

Hansen and William Y., et al, 2012)

10

design. As Coulson& Richardson’s, (2005) suggestions in the book, for carbon and low-alloy steels, where severe corrosion is

not expected, a minimum allowance of 2.0 mm should be used; where more severe conditions are expected minimum

thickness this should be increased to 4.0 mm. A minimum corrosion allowance of 1.0 mm is the standard in most of design

codes. Process design:

A three-phase inlet separator unit is designed acceding to the data given in table 1( E.W.M. Hansen and William Y., et al,

2012) were taken from the first stage inlet separator. Some assumptions were made in order to carry out the design.

Parameters Gas Oil Water Field units

Flow rate (Q) 51.96

(MMscf/d)

92586 (bbl/d) 14441.33 (bbl/d) -

Density (ρ) 3.103 51.91 64.3 Ib/ft3

Viscosity (μ) 0.013 5.25 0.43 Cp

Specific gravity (Sg) 0.683 0.8403 1.04 Ib / ft 3

Ib / ft 3

GOR 561.21 - Scf/STB

API of oil 36.9

Compressibility factor (Z) 0.99 - Dimensionless

Droplet removal size

(assumed)

- 100 500 Microns (μm ¿

Retention times

(Assumed)

- 5 5 minutes

Operating conditions

Pressure 996.41 Psia

Temperature 592 °R

Table 1: The physical properties of the fluid and production rates at operation conditions.(E.W.M.

Hansen and William Y., et al, 2012)

10

## Paraphrase This Document

Need a fresh take? Get an instant paraphrase of this document with our AI Paraphraser

Assumptions: Some assumptions were made for the purpose of carrying on the design of the three-phase separator, as

there was insufficient information about some of the data needed. Some of the required data for the design were taken

from Stewart, M. and Arnold, K. (2008)as assumed values. Assumptions were made: The specific gravity of the gas ( γg) is

assumed to be 0.683, and he compressibility factor of the gas Z = 0.99.

The gas flow rate is calculated based on the GOR = 100 sm3/sm3 (gas to oil ratio) and oil flow rate (Qo = 277758 bbl/day)

given in the following document (E.W.M. Hansen). The conversion of the GOR from sm3/sm3 to scf/STB is shown below:

∵ 1 (sm3) gas = 35.30 (sft3)

∵ 1 (sm3) oil = 6.29 (bbl)

∴ 1 sm3/sm3 = 35.30(ft3)

6.29(bbl)

∴ 100 sm3/sm3 = 561.21 (scf/STB)

∴gas flow rate (Qg) = flow rate of oil (Qo) x GOR

Qg = 277758 (bbl/d) x 561.21 (scf/STB) = 155.88 (MMscf/d)

Then, this gas flow rate is divided by 3 for the purpose of design (Qg = 51.96 MMscf/d)

In order to calculate the drag coefficient using equations (3,4 and 5), an assumed value of drag coefficient of (0.34) is used.

Using the assumed value of the drag coefficient, equation (4) is solved for Vt, then using the calculated value of Vt in

equation (5) to solve for Reynolds number (Re). After that, the calculated value of Reynolds number in equation (3)was

used to solve for Cd. If the calculated value of Cd equals the assumed value, then the solution has been reached. If not, then

the same process is repeated, using the new calculated value of the drag coefficient as the new assumption. According to

(Stewart, M. and Arnold, K.,2008), the assumed value of 0.34 for the drag coefficient was used, as this value id the limiting

value for large Reynolds number.

Sample calculations are shown below in how to determine the value of Cd, and then the iterated values are tabulated in

Table 2.

Step 1: An assumed the value of Cd = 0.34 is used and equation (4) is solved for Vt for liquid droplet diameter of (100

microns).

V t =0.01186 x [ [ 51.9−3.1

3.1 ] x [ 100

0.34 ] ]0.5

= 0.807 (ft/s)

Step 2: By using equation (5), Reynolds number is calculated using the calculated Vt:

11

there was insufficient information about some of the data needed. Some of the required data for the design were taken

from Stewart, M. and Arnold, K. (2008)as assumed values. Assumptions were made: The specific gravity of the gas ( γg) is

assumed to be 0.683, and he compressibility factor of the gas Z = 0.99.

The gas flow rate is calculated based on the GOR = 100 sm3/sm3 (gas to oil ratio) and oil flow rate (Qo = 277758 bbl/day)

given in the following document (E.W.M. Hansen). The conversion of the GOR from sm3/sm3 to scf/STB is shown below:

∵ 1 (sm3) gas = 35.30 (sft3)

∵ 1 (sm3) oil = 6.29 (bbl)

∴ 1 sm3/sm3 = 35.30(ft3)

6.29(bbl)

∴ 100 sm3/sm3 = 561.21 (scf/STB)

∴gas flow rate (Qg) = flow rate of oil (Qo) x GOR

Qg = 277758 (bbl/d) x 561.21 (scf/STB) = 155.88 (MMscf/d)

Then, this gas flow rate is divided by 3 for the purpose of design (Qg = 51.96 MMscf/d)

In order to calculate the drag coefficient using equations (3,4 and 5), an assumed value of drag coefficient of (0.34) is used.

Using the assumed value of the drag coefficient, equation (4) is solved for Vt, then using the calculated value of Vt in

equation (5) to solve for Reynolds number (Re). After that, the calculated value of Reynolds number in equation (3)was

used to solve for Cd. If the calculated value of Cd equals the assumed value, then the solution has been reached. If not, then

the same process is repeated, using the new calculated value of the drag coefficient as the new assumption. According to

(Stewart, M. and Arnold, K.,2008), the assumed value of 0.34 for the drag coefficient was used, as this value id the limiting

value for large Reynolds number.

Sample calculations are shown below in how to determine the value of Cd, and then the iterated values are tabulated in

Table 2.

Step 1: An assumed the value of Cd = 0.34 is used and equation (4) is solved for Vt for liquid droplet diameter of (100

microns).

V t =0.01186 x [ [ 51.9−3.1

3.1 ] x [ 100

0.34 ] ]0.5

= 0.807 (ft/s)

Step 2: By using equation (5), Reynolds number is calculated using the calculated Vt:

11

Re=(0.0049)x 3.1 x 100 x 0.807

0.013 = 94.29

Step 3: By using equation (3)and the previously calculated Reynolds number, the value of Cd can be determined as shown

below:

Cd=0.34+ 24

(94.29) + 3

¿ ¿ ¿ = 0.9035

As it can be seen, the value of the calculated drag coefficient is nowhere close to the assumed value of 0.34, hence steps 1

to 3 must be repeated until the calculated Cdisequal or very close to the assumed value. Table2 shows the result of the

iteration process.

Table 2: Drag coefficient calculation using iterative method. the value of the drag coefficient(Cd = 1.257)

Vt Re Cd

(ft/s) Dimensionless Dimensionless

0.807 94.29 0.90348

0.4951 57.8505 1.1493

0.4389 51.2838 1.2269

0.4248 49.636 1.249

0.421 49.192 1.2556

0.414994 49.0684 1.257

2.1.4 Three-Phase Horizontal Separator Sizing:

It is important to choose a seam-to-seam vessel length and diameter, when sizing a horizontal separator. The choice must

not only satisfy the conditions for gas capacity that allow the liquid droplet, entrained in the gas, to coalesce and fall from

the gas section to the liquid-gas interface as the gas crosses the effective length of the vessel, but also it must provide

sufficient hold-up or retention time to allow for the liquid to reach equilibrium. Fig. 3 shows a half-full vessel of liquid,

which is the case used to develop sizing equations for a horizontal separator in this project. Oil retention times (Stewart, M.

and Arnold, K.,2008) suggested if no information is available for water retention time, recommends retention times for

water from 5 to 10 minutes. assumed retention times for both oil and water is equal to (5 min.)

12

0.013 = 94.29

Step 3: By using equation (3)and the previously calculated Reynolds number, the value of Cd can be determined as shown

below:

Cd=0.34+ 24

(94.29) + 3

¿ ¿ ¿ = 0.9035

As it can be seen, the value of the calculated drag coefficient is nowhere close to the assumed value of 0.34, hence steps 1

to 3 must be repeated until the calculated Cdisequal or very close to the assumed value. Table2 shows the result of the

iteration process.

Table 2: Drag coefficient calculation using iterative method. the value of the drag coefficient(Cd = 1.257)

Vt Re Cd

(ft/s) Dimensionless Dimensionless

0.807 94.29 0.90348

0.4951 57.8505 1.1493

0.4389 51.2838 1.2269

0.4248 49.636 1.249

0.421 49.192 1.2556

0.414994 49.0684 1.257

2.1.4 Three-Phase Horizontal Separator Sizing:

It is important to choose a seam-to-seam vessel length and diameter, when sizing a horizontal separator. The choice must

not only satisfy the conditions for gas capacity that allow the liquid droplet, entrained in the gas, to coalesce and fall from

the gas section to the liquid-gas interface as the gas crosses the effective length of the vessel, but also it must provide

sufficient hold-up or retention time to allow for the liquid to reach equilibrium. Fig. 3 shows a half-full vessel of liquid,

which is the case used to develop sizing equations for a horizontal separator in this project. Oil retention times (Stewart, M.

and Arnold, K.,2008) suggested if no information is available for water retention time, recommends retention times for

water from 5 to 10 minutes. assumed retention times for both oil and water is equal to (5 min.)

12

CHAPTER THREE

3.0 Procedure to Size Three-Phase Horizontal Separator

The design procedure is adopted from Stewart, M. and Arnold, K. (2008).

Step 1: Specific gravities ( Δ SG ¿:

API = 141. 5

γ o

−131 .5 [4]

Re-arranging equation (10) and solve for γogives the following:

Since the oil gravity is (36.9) and water specific gravity equals (1.04) as shown in Table 3, therefore:

γo= 141.5

36.9+131.5 = 0.8403

Δ SG = 1.04 – 0.8403 = 0.1997

Step 2: the maximum oil pad thickness ( ho ¿max can be calculated using the following equation, with an assumed water

droplet size of (500 microns) and oil retention time of (5 min.) as suggested by Stewart, M. and Arnold, K.,2008:

(ho ¿max = ( 1 .28 E−3 ) ¿ ¿ [5]

= ( 1.28 E−3 ) ( 5 ) ( 0.1997 ) ¿ ¿ = 60.8 in.

Step 3: the fraction of the vessel cross-sectional area occupied by water phase can be calculated using the following

equation:

Since the flow rates of oil and water are (Qo = 92586 bbl/d and Qw=14441.33), and the retention time of oil is assumed

to be the same as that of water ¿= ¿ = 5 min.

Therefore,

Aw

A = 0 . 5 ( Qw ) ¿ ¿ [6]

Aw

A = 0.5 ( 14441.33 ) (5)

( 5 ) ( 92586 ) + ( 5 ) (14441.33) = 0.067

Step 4: The ratio of oil pad thickness to vessel diameter ( ho

d = β) is determined using Figure 2: Graph showing the

relationship between the fraction of vessel cross-sectional area occupied by the water phase and the ratio of oil pad

thickness to vessel diameter (modified from by Stewart, M. and Arnold, K.,2008)

13

3.0 Procedure to Size Three-Phase Horizontal Separator

The design procedure is adopted from Stewart, M. and Arnold, K. (2008).

Step 1: Specific gravities ( Δ SG ¿:

API = 141. 5

γ o

−131 .5 [4]

Re-arranging equation (10) and solve for γogives the following:

Since the oil gravity is (36.9) and water specific gravity equals (1.04) as shown in Table 3, therefore:

γo= 141.5

36.9+131.5 = 0.8403

Δ SG = 1.04 – 0.8403 = 0.1997

Step 2: the maximum oil pad thickness ( ho ¿max can be calculated using the following equation, with an assumed water

droplet size of (500 microns) and oil retention time of (5 min.) as suggested by Stewart, M. and Arnold, K.,2008:

(ho ¿max = ( 1 .28 E−3 ) ¿ ¿ [5]

= ( 1.28 E−3 ) ( 5 ) ( 0.1997 ) ¿ ¿ = 60.8 in.

Step 3: the fraction of the vessel cross-sectional area occupied by water phase can be calculated using the following

equation:

Since the flow rates of oil and water are (Qo = 92586 bbl/d and Qw=14441.33), and the retention time of oil is assumed

to be the same as that of water ¿= ¿ = 5 min.

Therefore,

Aw

A = 0 . 5 ( Qw ) ¿ ¿ [6]

Aw

A = 0.5 ( 14441.33 ) (5)

( 5 ) ( 92586 ) + ( 5 ) (14441.33) = 0.067

Step 4: The ratio of oil pad thickness to vessel diameter ( ho

d = β) is determined using Figure 2: Graph showing the

relationship between the fraction of vessel cross-sectional area occupied by the water phase and the ratio of oil pad

thickness to vessel diameter (modified from by Stewart, M. and Arnold, K.,2008)

13

## Secure Best Marks with AI Grader

Need help grading? Try our AI Grader for instant feedback on your assignments.

Therefore from Figure 4, β = 0.381

Step 5: The maximum allowable diameter of the vessel (Dmax ¿ can be calculated from the following relationship:

Dmax = ¿ ¿ = 159.58 in. (4.06 m)

Step 6: The gas capacity constraint can be calculating using equation (6) and Table 3, as the following:

The combinations of d and Leff for diameters less than Dmax that satisfy the gas capacity constraint is calculated.

d Leff =240 x [ TZ Qg

P ] x [ [ ρg

ρl −ρg ][ Cd

dm ] ]1/ 2

d Leff =240 x [ 592 x 0.99 x 51.96

996.41 ] x [ [ 3.105

51.91−3.105 ][ 1.257

100 ] ]1/ 2

= 362.997 (in.ft)

∴ d Leff = 362.997 (in.ft)

14

Step 5: The maximum allowable diameter of the vessel (Dmax ¿ can be calculated from the following relationship:

Dmax = ¿ ¿ = 159.58 in. (4.06 m)

Step 6: The gas capacity constraint can be calculating using equation (6) and Table 3, as the following:

The combinations of d and Leff for diameters less than Dmax that satisfy the gas capacity constraint is calculated.

d Leff =240 x [ TZ Qg

P ] x [ [ ρg

ρl −ρg ][ Cd

dm ] ]1/ 2

d Leff =240 x [ 592 x 0.99 x 51.96

996.41 ] x [ [ 3.105

51.91−3.105 ][ 1.257

100 ] ]1/ 2

= 362.997 (in.ft)

∴ d Leff = 362.997 (in.ft)

14

For assumed vessel diameter (d) less than Dmax =159.58 (inches) values of Leff are tabulated using excel spreadsheet, as

shown below:

D (in) Leff (ft)

148.0000 2.4527

140.0000 2.5928

130.0000 2.7923

120.0000 3.0250

110.0000 3.3000

100.0000 3.6300

90.0000 4.0333

80.0000 4.5375

70.0000 5.1857

66.0000 5.5000

60.0000 6.0500

Table 3: Gas capacity constraint calculations for diameters less than that of Dmax

Since the values of Leff are low, then the gas capacity does not govern the vessel sizing.

Sample calculation for determining Leff in Table3 is shown below:

Since, d Leff = 362.997 (in.ft)

Therefore,

Leff = 362.997

148 = 2.4527 ft.

Step 7: The combinations of d and Leff for diameters less than Dmax that satisfy the oil and water retention

timeconstraints are calculated, using equation (7) and Table3 as the following:

D2Leff = 1.429¿

D2Leff = 1.429[ ( 14441.33 ) (5)+ ( 92586 ) ( 5) ]= 764710.297 (in2.ft)

∴ D2Leff = 764710.297(in2.ft).

15

shown below:

D (in) Leff (ft)

148.0000 2.4527

140.0000 2.5928

130.0000 2.7923

120.0000 3.0250

110.0000 3.3000

100.0000 3.6300

90.0000 4.0333

80.0000 4.5375

70.0000 5.1857

66.0000 5.5000

60.0000 6.0500

Table 3: Gas capacity constraint calculations for diameters less than that of Dmax

Since the values of Leff are low, then the gas capacity does not govern the vessel sizing.

Sample calculation for determining Leff in Table3 is shown below:

Since, d Leff = 362.997 (in.ft)

Therefore,

Leff = 362.997

148 = 2.4527 ft.

Step 7: The combinations of d and Leff for diameters less than Dmax that satisfy the oil and water retention

timeconstraints are calculated, using equation (7) and Table3 as the following:

D2Leff = 1.429¿

D2Leff = 1.429[ ( 14441.33 ) (5)+ ( 92586 ) ( 5) ]= 764710.297 (in2.ft)

∴ D2Leff = 764710.297(in2.ft).

15

For the same assumed vessel diameters (d) less than Dmax =159.58 (inches) values of Leff are tabulated using excel

spreadsheet, as shown below:

D (in) Leff (ft) Lss(ft) SD = 12 Lss

d

148 34.9119 46.5492 3.7743

140 39.0158 52.0211 4.4590

130 45.2491 60.3322 5.5691

120 53.1049 70.8065 7.0807

110 63.1992 84.2656 9.1926

100 76.4710 101.9614 12.2354

90 94.4087 125.8782 16.7838

80 119.4860 159.3146 23.8972

70 156.0633 208.0844 35.6716

66 175.5533 234.0711 42.5584

60 212.4195 283.2260 56.6452

Table 4: Liquid capacity constraint calculations for diameters less than Dmax

Sample calculation for determining Leff , Lss, and the slenderness ratio (SD)

∵D2Leff = 764710.297

∴Leff= 764710.273

¿ ¿ = 34.9119 ft.

The seam-to-seam length of the horizontal vessel ( Lss ¿ is calculated as the following:

∵ The liquid capacity govern the vessel sizing,

∴Lss = ( 4

3 ) Leff = ( 4

3 ) x( 34.9119) = 46.5492 ft.

The slenderness ratio (SD) can be calculated as the following:

SD = 12 x Lss

d = = 12 x 46.5492

148 = 3.7743

There are two possible choices to select the separator dimensions, which are as the following:

1) 148 (in.) diameter by 46.5492 (ft.) seam-to-seam length, and

2) 140 (in.) diameter by 52.0211 (ft.) seam-by-seam length.

16

spreadsheet, as shown below:

D (in) Leff (ft) Lss(ft) SD = 12 Lss

d

148 34.9119 46.5492 3.7743

140 39.0158 52.0211 4.4590

130 45.2491 60.3322 5.5691

120 53.1049 70.8065 7.0807

110 63.1992 84.2656 9.1926

100 76.4710 101.9614 12.2354

90 94.4087 125.8782 16.7838

80 119.4860 159.3146 23.8972

70 156.0633 208.0844 35.6716

66 175.5533 234.0711 42.5584

60 212.4195 283.2260 56.6452

Table 4: Liquid capacity constraint calculations for diameters less than Dmax

Sample calculation for determining Leff , Lss, and the slenderness ratio (SD)

∵D2Leff = 764710.297

∴Leff= 764710.273

¿ ¿ = 34.9119 ft.

The seam-to-seam length of the horizontal vessel ( Lss ¿ is calculated as the following:

∵ The liquid capacity govern the vessel sizing,

∴Lss = ( 4

3 ) Leff = ( 4

3 ) x( 34.9119) = 46.5492 ft.

The slenderness ratio (SD) can be calculated as the following:

SD = 12 x Lss

d = = 12 x 46.5492

148 = 3.7743

There are two possible choices to select the separator dimensions, which are as the following:

1) 148 (in.) diameter by 46.5492 (ft.) seam-to-seam length, and

2) 140 (in.) diameter by 52.0211 (ft.) seam-by-seam length.

16

## Paraphrase This Document

Need a fresh take? Get an instant paraphrase of this document with our AI Paraphraser

The chosen dimensions for the three-phase horizontal separator is (140 in. diameter by 52.0211 ft. seam-by-seam length).

17

17

3.1 Demister Sizing:

Arnold, K. (2007), as explained previously, a knitted mesh is used to enhance the liquid coalescence at the gas outlet while

allowing the gas to bypass those liquid droplets and separate to the outlet gas zone. For general design of knitted mesh, the

following equation can be used:

V m= Kd √ ρl−ρg

ρg

[7]

Where,

V m= Maximum velocity of gas (ft/s).

Kd= Demister capacity factor in (ft/s) and depends on the type of demister.

ρl , ρg= density of liquid and gas phases respectively in (Ib/ft3).

The K value of mesh depends on the type of mesh used, viscosity, surface tension of liquid, liquid loading, as well as the

manufacturer. For knitted mesh, the recommended K value for general sizing is 0.1 (m/s) or 0.328 (ft/s). This typical value

generates a pressure drop only few inches of water, according to Arnold, K. (2007). As in Lyons, C and Plisga, G. 2005 book,

the range of typical pad thicknesses that is sufficient for most separator applications are (4-6 inches).

For horizontal vessels, the required mesh area can be determined using the following equation (Lyons, C and Plisga, G.

2005):

Ad = Qg

V m

and Qg = Rs x Qo

86400 x 14.7

P x TZ

520Where,

Ad = required mesh area in ft3.

Rs = gas to oil ration (GOR) in scf/STB.

The rest are same as before.

Sample calculations:

V m= 0.328 √ 51 . 91−3 . 105

3 .105 = 1.3 (ft/s).(Table 3)

Qg = 561.21 x 92586

86400 x 14.7

996.41 x 592 x 0.99

520 = 9.99(ft3/s). (Table 3)

Ad = 9.99

1.3 = 7.684 (ft2) or 0.714m Therefore, mesh diameter equal to (0.95m).

The weir: Is assumed to has a width of 10 in, height is set at half the diameter of the vessel (70 inor 1.78m).

18

Arnold, K. (2007), as explained previously, a knitted mesh is used to enhance the liquid coalescence at the gas outlet while

allowing the gas to bypass those liquid droplets and separate to the outlet gas zone. For general design of knitted mesh, the

following equation can be used:

V m= Kd √ ρl−ρg

ρg

[7]

Where,

V m= Maximum velocity of gas (ft/s).

Kd= Demister capacity factor in (ft/s) and depends on the type of demister.

ρl , ρg= density of liquid and gas phases respectively in (Ib/ft3).

The K value of mesh depends on the type of mesh used, viscosity, surface tension of liquid, liquid loading, as well as the

manufacturer. For knitted mesh, the recommended K value for general sizing is 0.1 (m/s) or 0.328 (ft/s). This typical value

generates a pressure drop only few inches of water, according to Arnold, K. (2007). As in Lyons, C and Plisga, G. 2005 book,

the range of typical pad thicknesses that is sufficient for most separator applications are (4-6 inches).

For horizontal vessels, the required mesh area can be determined using the following equation (Lyons, C and Plisga, G.

2005):

Ad = Qg

V m

and Qg = Rs x Qo

86400 x 14.7

P x TZ

520Where,

Ad = required mesh area in ft3.

Rs = gas to oil ration (GOR) in scf/STB.

The rest are same as before.

Sample calculations:

V m= 0.328 √ 51 . 91−3 . 105

3 .105 = 1.3 (ft/s).(Table 3)

Qg = 561.21 x 92586

86400 x 14.7

996.41 x 592 x 0.99

520 = 9.99(ft3/s). (Table 3)

Ad = 9.99

1.3 = 7.684 (ft2) or 0.714m Therefore, mesh diameter equal to (0.95m).

The weir: Is assumed to has a width of 10 in, height is set at half the diameter of the vessel (70 inor 1.78m).

18

19

## Secure Best Marks with AI Grader

Need help grading? Try our AI Grader for instant feedback on your assignments.

3.2 Mechanical Design of Major Unit

Table 5: Typical design stress for some of the common Materials used for preliminary designs given by

Coulson& Richardson’s, (2005)

Material Design stress at temperature ℃ (N/mm2)

Tensile

strength

(N/mm2)

0-50 100 150 200 250

Carbon steel(semi-

killed or silicon

killed)

360 135 125 115 105 95

Carbon-manganese

steel (semi-killed or

silicon killed)

460 180 170 150 140 130

Carbon-

molybdenum steel

(0.5% Mo)

450 180 170 145 140 130

Stainless steel304

(18Cr/8Ni)

510 165 145 130 115 110

Table 6: design parameters that are used in mechanical design calculations.

Design parameters

Terms Value Unit

Operating pressure 6.87 or 68.7 N/mm2 or bar

Operating temperature 55.4 ℃

Internal Diameter 3550 mm

Shell Length 15.86 or 52.0211 m or ft.

Tensile stress 460 N/mm2

Design Stress at the design temperature

(100℃)

170 N/mm2

Weld factor (J) 1.0 N/mm2

Cv 1.08

Design pressure (10% above operating

pressure)

7.447 N/mm2

Corrosion allowance 2 mm

20

Table 5: Typical design stress for some of the common Materials used for preliminary designs given by

Coulson& Richardson’s, (2005)

Material Design stress at temperature ℃ (N/mm2)

Tensile

strength

(N/mm2)

0-50 100 150 200 250

Carbon steel(semi-

killed or silicon

killed)

360 135 125 115 105 95

Carbon-manganese

steel (semi-killed or

silicon killed)

460 180 170 150 140 130

Carbon-

molybdenum steel

(0.5% Mo)

450 180 170 145 140 130

Stainless steel304

(18Cr/8Ni)

510 165 145 130 115 110

Table 6: design parameters that are used in mechanical design calculations.

Design parameters

Terms Value Unit

Operating pressure 6.87 or 68.7 N/mm2 or bar

Operating temperature 55.4 ℃

Internal Diameter 3550 mm

Shell Length 15.86 or 52.0211 m or ft.

Tensile stress 460 N/mm2

Design Stress at the design temperature

(100℃)

170 N/mm2

Weld factor (J) 1.0 N/mm2

Cv 1.08

Design pressure (10% above operating

pressure)

7.447 N/mm2

Corrosion allowance 2 mm

20

3.3 Material Selection

Selection of a suitable material for the three-phase horizontal separator construction is essential, as well as the

compatibility of the material within the environment and process operating conditions. “Carbone manganese steel” has

been selected for the design taken into consideration the hush environment that the major unit will be exposed to.

Material of

construction

Tensile

strength

0.1 per cent

proof stress

Module of

elasticity

Hardness

Brinell

Specific gravity

Carbon-

manganese

steel

(N/mm2 ) (N/mm2 ) (KN/mm2 )

460 230-460 210 130-160 7.9

Table 7: The chosen material of construction for the three-phase separator, taken from (Coulson&

Richardson’s, (2005).

3.4 Thickness calculation for the component parts of the vessels

The minimum thickness required for a cylindrical shell in order to withstand internal pressure can be calculated using

Table7 and equation (13) with corrosion allowance of 2mm. The following equation is taken from (Economides, M., 2009):

Sample calculation:

e= Pi x Di

2 fJ−Pi

[8]

Where:

Di = vessel diameter (mm).

Pi = internal design pressure (N/mm2).

f = design stress (N/mm2).

e = the minimum thickness required (mm)

J = weld-joint factor (taken as 1.0).

The design pressure is taken as 10% above operating pressure:

(68.7-1) x 1.1 = 74.47 bar

= 7.447 (N/mm2)

Therefore,

e= ( 7.447 ) x (3.55 E+3)

2 x (170) – 1.2(7.447) = 79.85 mm + 2mm = 81.85mm.

21

Selection of a suitable material for the three-phase horizontal separator construction is essential, as well as the

compatibility of the material within the environment and process operating conditions. “Carbone manganese steel” has

been selected for the design taken into consideration the hush environment that the major unit will be exposed to.

Material of

construction

Tensile

strength

0.1 per cent

proof stress

Module of

elasticity

Hardness

Brinell

Specific gravity

Carbon-

manganese

steel

(N/mm2 ) (N/mm2 ) (KN/mm2 )

460 230-460 210 130-160 7.9

Table 7: The chosen material of construction for the three-phase separator, taken from (Coulson&

Richardson’s, (2005).

3.4 Thickness calculation for the component parts of the vessels

The minimum thickness required for a cylindrical shell in order to withstand internal pressure can be calculated using

Table7 and equation (13) with corrosion allowance of 2mm. The following equation is taken from (Economides, M., 2009):

Sample calculation:

e= Pi x Di

2 fJ−Pi

[8]

Where:

Di = vessel diameter (mm).

Pi = internal design pressure (N/mm2).

f = design stress (N/mm2).

e = the minimum thickness required (mm)

J = weld-joint factor (taken as 1.0).

The design pressure is taken as 10% above operating pressure:

(68.7-1) x 1.1 = 74.47 bar

= 7.447 (N/mm2)

Therefore,

e= ( 7.447 ) x (3.55 E+3)

2 x (170) – 1.2(7.447) = 79.85 mm + 2mm = 81.85mm.

21

e= Pi Di

2 Jf −0 .2 Pi

[9]

The stress concentration factor for torispherical-heads can be calculated using equation (15) as shown below:

Cs = 1

4 ( 3+ √ 3.55

0.213 ) = 1.77

1) Therefore, the minimum thickness of a standard dished-head (torisphere) can be calculated using equation (16) as shown

below:

e= Pi Rc CS

2 fJ + Pi (Cs−0 . 2)

= ( 7.447 ) ( 3.55 E+3 ) (1.77)

2 ( 170 ) ( 1.0 ) +7.447( 1.77−0.2) = 133.1mm + 2mm = 135.1mm.

2) Standard ellipsoidal head, ratio of major to minor axes = 2:1

The minimum thickness required for an ellipsoidal head can be calculated using equation(14) as the following:

e= Pi Di

2 f +0.2 Pi

= ( 7.447 ) (3.55 E+3)

2 ( 170 )+ 0.2(7.447) = 77.42mm + 2mm = 79.42mm

Hence: the ellipsoidal head would be the most economical. In addition, its wall thickness is made as the cylindrical shell

thickness of 81.85mm

3) Flat head: A full gasket is used, in which Cp = 0.4, De = bolt circle diameter, and it is taken to be approximately (3.75 m).

The minimum thickness of a flat head end can be calculated as the following:

e=C p De √ Pi

f

C p = A design constant, depending on the edge constraint.

De = nominal plate diameter.

22

2 Jf −0 .2 Pi

[9]

The stress concentration factor for torispherical-heads can be calculated using equation (15) as shown below:

Cs = 1

4 ( 3+ √ 3.55

0.213 ) = 1.77

1) Therefore, the minimum thickness of a standard dished-head (torisphere) can be calculated using equation (16) as shown

below:

e= Pi Rc CS

2 fJ + Pi (Cs−0 . 2)

= ( 7.447 ) ( 3.55 E+3 ) (1.77)

2 ( 170 ) ( 1.0 ) +7.447( 1.77−0.2) = 133.1mm + 2mm = 135.1mm.

2) Standard ellipsoidal head, ratio of major to minor axes = 2:1

The minimum thickness required for an ellipsoidal head can be calculated using equation(14) as the following:

e= Pi Di

2 f +0.2 Pi

= ( 7.447 ) (3.55 E+3)

2 ( 170 )+ 0.2(7.447) = 77.42mm + 2mm = 79.42mm

Hence: the ellipsoidal head would be the most economical. In addition, its wall thickness is made as the cylindrical shell

thickness of 81.85mm

3) Flat head: A full gasket is used, in which Cp = 0.4, De = bolt circle diameter, and it is taken to be approximately (3.75 m).

The minimum thickness of a flat head end can be calculated as the following:

e=C p De √ Pi

f

C p = A design constant, depending on the edge constraint.

De = nominal plate diameter.

22

1 out of 22

### Related Documents

Your All-in-One AI-Powered Toolkit for Academic Success.

##### +13062052269

##### info@desklib.com

Available 24*7 on WhatsApp / Email

Unlock your academic potential

© 2024 | Zucol Services PVT LTD | All rights reserved.