Optimized Process Design of 3-Phase Production

Separators on Offshore Oil and Gas Facilities

John T. Small

John T. Small

Process Engineer

Published Feb 21, 2024

TABLE OF CONTENTS

1. INTRODUCTION

2. SEPARATOR CONFIGURATION AND DESIGN INPUTS

2.1 Separator Inlet Zone Lɪɴ

2.1.1 Dimensioning of Feed Pipe, Inlet Nozzle and Inlet Device

2.1.2 Separator Flow Straightening Baffle

2.2 Separator Gravity Settling Zone Lᴇꜰꜰ

2.2.1 Size of Oil Droplets to be Removed from Gas

2.2.2 Sizes of Water and Oil Droplets to be Removed from Liquid

2.2.3 Axial Length of Gravity Settling Zone

2.2.4 Liquid Axial Velocity Considerations

2.3 Dimensioning of Gas Outlet Nozzle

2.3.1 Velocity Head Criterion

2.3.2 Erosion Rate Criterion

2.4 Dimensioning of Water Outlet Nozzle and Water Outlet Zone

2.4.1 Water Outlet Nozzle

2.4.2 Water Outlet Zone Lᴡ

2.5 Dimensioning of Oil Outlet Nozzle and Oil Outlet Zone

2.5.1 Oil Outlet Nozzle

2.5.2 Oil Outlet Zone Lᴏ

2.6 Oil and Water Retention Times

2.7 Re-entrainment of Oil Droplets into Gas Phase

2.8 Slenderness Ratio

2.9 Level Controls and Trips

2.9.1 Level Control Intervals

2.9.2 Unsteady State Operations – Slugging and Surging

2.9.3 Level Trips

2.10 User Guide for the Calculation Spreadsheet

2.11 Navigating the Results of the Separator Design Calculations

2.12 Architecture of Spreadsheet Calculations

3. CASE STUDY:

3-PHASE SEPARATOR ADEQUATE FOR EARLY AND LATE FIELD LIFE

FLUIDS

3.1 Summary

3.2 Problem Description

3.3 Performance Criteria and Constraints

3.4 Results

4. NOMENCLATURE

5. REFERENCES
1. INTRODUCTION

The 3- phase separator configuration which forms the basis for this work is
illustrated above.

The separator is the flooded-weir type, which is commonly employed in oil

and gas production facilities.

These large vessels generally contain the most substantial hydrocarbon

inventory on the facility, and they utilize costly deck space and loading
capacity. The associated pipework and valving are often bulky and
challenging to accommodate in restricted module spaces. Modification
projects especially may place hard limits on these factors.

At the same time, separator performance and control are often critical to
successful operation of downstream equipment such as de-oiling, de-
watering and gas compression units.

It follows that there is a strong incentive to minimize the size of the vessel
and associated piping whilst achieving satisfactory performance in the
separation, process control and safeguarding functions.

This optimization task requires resolution of key design aspects of the

separator unit:

· Gravity settling zone dimensions for successful phase separation.

Opposing schools of thought favor computation via droplet settling theory

versus liquid retention times.

· Provision of adequate liquid hold up times and height differences

between contiguous level control and trip settings.

· Type of inlet device, and its impact upon the size of inlet pipework and
the dimensions of the separator vessel.

· Nozzle sizing according to erosional, velocity head and velocity criteria,

and the impact on vessel dimensions.

· Maintaining axial velocities of oil and water phases within acceptable

limits.

· Prevention of re-entrainment of oil droplets into the gas phase.

· Prevention of transient gas blowby or gross carryover of liquid during

unplanned shutdown.
· Provision of adequate liquid and gas holdup volumes for slugging and
surging.

· Prevention of spurious trips due to foaming, where applicable.

This article, which is adapted from Small (2022), describes a Microsoft®

Excel spreadsheet that models numerically the factors listed above and
optimizes the vessel dimensions using the Solver® add-in, based on the
engineer´s design input parameters and constraints. Macros are not
employed, to maximize flexibility of use.

The principal calculation method for dimensioning the gravity separation

zone of the vessel is based on settling theory, according to which droplets /
bubbles of certain diameters settle to their respective continuous phases,
however the spreadsheet allows the engineer to concurrently specify
minimum allowable oil and / or water retention times if desired.

The objective function of the Solver algorithm may be set to minimize the
volume of a separator vessel for the selected design parameters and
constraints, or the engineer may choose other targets for optimization such
as length of the gravity settling zone.

Section 2 discusses development of the salient features of the 3-phase

separator model, citing the technical references and criteria that have been
considered. Key input design parameters and constraints are identified for
solution of the dimensioning calculations. Operation of the spreadsheet
calculator and navigation of the results panel are described. The architecture
of the spreadsheet is mapped on a flowsheet.

In Section 3 a case study is presented which addresses the design of a

separator that is required to accommodate a major change in water cut
during field life. The impact upon the key separator design aspects of
installing a half-pipe versus a diffuser type inlet device is also evaluated. The
case study is founded on published field data from a North Sea separator
which reportedly operated successfully. This allows a degree of validation of
the calculation method and results.
2. SEPARATOR CONFIGURATION AND DESIGN INPUTS

This section discusses the technical references and criteria that were
considered in development of the salient features of the 3-phase separator,
the input design parameters and constraints that are required for execution
of the separator dimensioning calculations, how to navigate the results
panel, and lastly, the architecture of the spreadsheet.

The topics are discussed under the following sequence of headings:

· Separator Inlet Zone Lɪɴ

· Separator Gravity Settling Zone Lᴇꜰꜰ

· Gas Outlet Nozzle

· Water Outlet Nozzle and Water Outlet Zone Lᴡ

· Oil Outlet Nozzle and Oil Outlet Zone Lᴏ

· Oil and Water Retention Times

· Re-entrainment of Oil Droplets into Gas Phase

· Slenderness Ratio

· Level Controls and Trips

· Summary of Required Inputs to the Calculation Modules

· Navigating the Results of the Separator Design Calculations

· Architecture of Spreadsheet Calculations

2.1 Separator Inlet Zone Lɪɴ

The Inlet Zone components of the 3-phase separator are generally defined in
accordance with GPSA (2012) to include the feed pipe, inlet nozzle and inlet
device. In addition, a flow straightening baffle is included downstream of the
inlet device.

In compliance with GPSA and API 12J (2008), bulk phase separation is
deemed to take place in the Inlet Zone during steady state operation.

2.1.1 Dimensioning of Feed Pipe, Inlet Nozzle and Inlet Device

The sizing calculations for these components require input to Design

Parameters in respect of flowrate, density and dynamic viscosity for each
flowing phase, labelled Qɢ: Qᴏ: Qᴡ: ϱɢ: ϱᴏ: ϱᴡ: µɢ: µᴏ: µᴡ.
GPSA (2012) recommends that the fluid velocity head J from a point 10 pipe
diameters upstream of the inlet should match the requirements of the inlet
device, to ensure good distribution of the phases and to minimize shattering
of liquid droplets.

The feed pipe, inlet nozzle and inlet device are therefore considered to have
the same internal diameter.

The engineer calculates component diameters which meet velocity head and
erosional criteria, as described below. The spreadsheet selects from a look-up
table the pipe diameter which satisfies the more stringent sizing criterion.

Velocity Head Considerations

The allowable velocity head J at the inlet device, and by extension in the feed
pipe and inlet nozzle, depends on its design configuration.

Having selected a particular type of inlet device, the engineer inputs to the
calculation panel for the inlet nozzle a value for allowable velocity head J.

GPSA (2012) cites typical allowable velocity head ranges for various types of
inlet device:

J = 1500 – 3700 Pa for half-pipe, v-baffle or elbow type.

J = 6000 – 9000 Pa for diffuser type.

Erosion Rate Considerations

The size of the feed pipe, inlet nozzle and inlet device must also be great
enough to satisfy erosional constraints during the service life of the
separator.

In this work erosion rate is calculated in accordance with DNV-GL-RP-501

(2015).

DNV has developed models for prediction of erosive wear rates in the
standard steel piping components found in oil and gas facilities. Grounded in
the results of experimental investigations, the models enable sizing of
pipework components based on the physical properties of the sand particles,
the process fluid, and the pipework, taking account also of piping geometry.

It is assumed that from the perspective of erosion rate, the minimum

diameter of the feed pipe / inlet nozzle / inlet device is governed by a pipe
bend component in the feed pipe.
The engineer calculates the nozzle diameter required to meet the erosion
target using Goal Seek function as per step number 6 in Section 2.10 below.

Upper Edge of Inlet Nozzle Relative to Top of Vessel

The upper edge of the inlet nozzle at (hɪɴ + Di) should be located as close as
permissible to the top of the vessel head, since this tends to minimize vessel
diameter Di. The proximity is however constrained by mechanical
considerations to a certain proportion of Di.

The proportion (hɪɴ + Di) / Di is defined by the engineer and is input to the
Design Parameters as a value labelled Rɪɴ.

The value of Rɪɴ should ideally be obtained from a vessel specialist.

Alternatively, a rule of thumb estimate of 0.8 may be used.

2.1.2 Separator Flow Straightening Baffle

The purpose of the flow straightening baffle downstream of the inlet device is
to induce plug flow conditions in the Gravity Settling Zone, which is
discussed further in Section 2.2.3 below. The location of the baffle relative to
the vessel seam defines the axial length of the Inlet Zone and is labelled Lɪɴ.

The engineer inputs to the Design Parameters a fixed value for Lɪɴ.

2.2 Separator Gravity Settling Zone Lᴇꜰꜰ

In this work, the principal calculation method for the phase separation aspect
of separator sizing is based on settling theory, according to which droplets /
bubbles of certain diameters settle to their respective continuous phases
within the Gravity Settling Zone.

(Note: The calculation program also allows the engineer to concurrently

specify minimum allowable oil and / or water retention times; refer to Section
2.6).

It is assumed that by designing the separator using settling theory for certain
droplet sizes, all droplets of that size and larger settle out at terminal
velocity to their respective continuous phase.

The gas phase holdup time in the Settling Zone needs to be sufficient for
liquid droplets to settle out to the surface of the oil. Similarly, the holdup
times of the oil and water phases need to be sufficient for settle out of gas
bubbles and liquid droplets from the respective dispersed phase.
Noting that the terminal velocity for a given droplet size increases according
to the density difference between the continuous and dispersed phases,
Grødal and Realff (1999) highlight for each continuous phase the settling
process which governs holdup time requirements in a 3-phase separator:

· In the continuous gas phase, settling out of oil droplets governs

when oil density is closer to gas than is water density.

· In the continuous oil layer, settling of water droplets governs, since

water density is closer to oil than is gas density.

· In the continuous water layer, flotation of the oil droplets governs,

since oil density is closer to water than is gas density.

The engineer inputs to Design Parameters for each fluid phase the minimum
diameter droplet that is to settle out in the Gravity Settling Zone:

dᴏɢ :dᴡᴏ :dᴏᴡ.

The engineer calculates for each continuous phase the terminal settling
velocity of dispersed droplets/bubbles using Goal Seek function, as per step
number 4 in Section 2.10 below.

2.2.1 Size of Oil Droplets to be Removed from Gas

Arnold and Stewart (2008) cite field experience which suggests that selecting
an oil droplet size dᴏɢ = 140 microns results in adequate oil removal from
the gas phase.

2.2.2 Sizes of Water and Oil Droplets to be Removed from Liquid

The relative densities of the oil and water constituents being similar, the
settling velocity of dispersed phase droplets is strongly influenced by the
viscosity of the continuous phase. Oil viscosity is typically the order of 10
times that of water, therefore it is more often the allowable water droplet
diameter that determines the dimension Lᴇꜰꜰ for oil/water separation.

Arnold and Koszela (1990), writing from the perspective of design engineers,
have reported good results sizing separators based on water droplet
diameters dᴡᴏ = 500 – 1000 microns.

Arnold and Stewart (2008) provide estimated water-in-oil concentrations

corresponding with this range of design water droplet diameter, for separator
oil discharge streams.
This author analyzed production data from a paper by Laleh et al (2013)
pertaining to an actual 3-phase separator operating satisfactorily on an
offshore installation; oil from this separator is further processed sequentially
in two downstream separators according to Hansen et al (1993).

The separator was found to be oil capacity limited, with a minimum water
droplet diameter dᴡᴏ = 960 microns.

Arnold and Stewart (2008) suggest allowable oil droplet size dᴏᴡ = 200
microns for flotation of oil from the water phase.

2.2.3 Axial Length of Gravity Settling Zone

The axial length of the Gravity Settling Zone is labelled Lᴇꜰꜰ

The balance of forces upon which settling theory is predicated applies strictly
to horizontal plug flow of the continuous phase, as discussed in GPSA (2012),
therefore it is necessary to define the serviceable length of the Gravity
Settling Zone.

In a 3-phase separator, it is considered that the downwards flow component

of liquid in the Water Outlet Zone would impede flotation of oil droplets and
gas bubbles rising in the continuous liquid layers. Thus, Lᴇꜰꜰ occupies the
zone between the inlet flow straightening baffle and the inboard edge of the
Water Outlet Zone.

Dimensioning of the Water Outlet Zone is discussed in Section 2.4 below.

The Gravity Settling Zone also contributes to other functions such as liquid
holdup to meet Design Constraints on level control, surge / slug volumes,
and where specified, liquid retention times. These functions may force
extending Lᴇꜰꜰ beyond what is required to accommodate the settling
process.

The variable Lᴇꜰꜰᵀᴼᵀᴬᴸ is introduced to enable a solution for Lᴇꜰꜰ which

encompasses the multiple functions of the Gravity Settling Zone.

The engineer inputs to the calculation panel a trial value for Lᴇꜰꜰᵀᴼᵀᴬᴸ.

2.2.4 Liquid Axial Velocity Considerations

The engineer inputs to Design Constraints maximum allowable values for

water and oil axial velocity Uᴡᵐᵃˣ and Uᴏᵐᵃˣ within the Gravity Settling Zone.

These values affect the thickness of the respective layers and therefore
influence Lᴇꜰꜰ.
The engineer may wish to consider the following factors when setting values
for Uᴡᵐᵃˣ and Uᴏᵐᵃˣ:

· GPSA (2012) mentions using a maximum axial velocity of 0.015 m/s for
the oil and water continuous phases, in a horizontal separator design with
target droplet diameters of 150 microns.

· Analysis by this author of data pertaining to an actual offshore

installation in a paper by Laleh et al (2013), indicates a 3-phase production
separator operating satisfactorily with oil and water axial velocities of 0.2
and 0.1 m/s respectively. The water droplet cut-off diameter was 960
microns.

Arntzen (2016) writing from the PACO perspective advises consideration of

liquid velocity upon the accuracy of level control instrumentation.

2.3 Dimensioning of Gas Outlet Nozzle

The gas outlet pipe and nozzle are assumed to have the same internal
diameter. They are sized to meet the more stringent criterion of allowable
velocity head and erosion rate. The engineer calculates component
diameters which meet velocity head and erosional criteria, as described
below. The spreadsheet selects from a look-up table the pipe diameter which
satisfies the more stringent sizing criterion.

2.3.1 Velocity Head Criterion

The engineer inputs to the calculation panel for the gas outlet nozzle, a value
for allowable velocity head J .

GPSA (2012) quotes industry guidelines of J = 4500 - 5400 Pa for gas outlet
nozzles.

2.3.2 Erosion Rate Criterion

From the perspective of erosion rate, it is assumed that the size of the gas
outlet nozzle and outlet pipe is limited by a pipe bend component in the
outlet pipe.

The engineer calculates the nozzle diameter required to meet the erosion
target using Goal Seek function as per step number 6 in Section 2.10 below.

2.4 Dimensioning of Water Outlet Nozzle and Water Outlet Zone

2.4.1 Water Outlet Nozzle

The water outlet pipe and nozzle are assumed to have the same internal
diameter. They are sized to meet the more stringent criterion of allowable
velocity and erosion rate. The engineer calculates nozzle diameters which
meet velocity and erosional criteria, as described below. The spreadsheet
selects from a look-up table the pipe diameter which satisfies the more
stringent sizing criterion.

Velocity Criterion

The engineer inputs to the calculation panel for the water outlet nozzle, a
value for allowable velocity Uɴ,ᴡᵐᵃˣ.

The value may be selected from the ranges presented in API RP 14E (1991):

0.9 ≤ Uᵐᵃˣ ≤ 4.6 m/s for pipe between vessels.

Uᵐᵃˣ = 0.9 m/s for pipe into pump suction.

The lower limit on Uᵐᵃˣ is to prevent sand deposition.

Recommended by LinkedIn

Different Types of Impellers Used in Centrifugal Pumps…

Mechanical Engineering World 1 year ago

Erosion Rate Criterion

From the perspective of erosion rate, it is assumed that the size of the water
outlet nozzle and outlet pipe is limited by a pipe bend component in the
outlet pipe.

The engineer calculates the nozzle diameter required to meet the erosion
target using Goal Seek function as per step number 6 in Section 2.10 below.

2.4.2 Water Outlet Zone Lᴡ

Water Outlet Zone Lᴡ is deemed to extend vertically up through the phases

and restricts the serviceable length of the Gravity Settling Zone Lᴇꜰꜰ, as
referred to above in Section 2.2.

The axial length of the Water Outlet Zone is defined by the width of the
vortex breaker on the water outlet, which the spreadsheet calculates to be
twice the nozzle diameter as per Rochelle and Briscoe (2010).

The outboard edge of the Water Outlet Zone is assumed to abut the inboard
side of the weir.

2.5 Dimensioning of Oil Outlet Nozzle and Oil Outlet Zone

2.5.1 Oil Outlet Nozzle

Dimensioning of the oil outlet nozzle is analogous to the water outlet nozzle.

The engineer inputs to the calculation panel for the oil outlet nozzle, a value
for allowable velocity Uɴ,ᴏᵐᵃˣ.

The range of allowable velocity values is similar to the water outlet nozzle.

The engineer then calculates the oil outlet nozzle diameter required to meet
the erosion target rate Eᵐᵃˣ using Goal Seek function as per step number 6 in
Section 2.10 below.

2.5.2 Oil Outlet Zone Lᴏ

The Oil Outlet Zone extends axially from the outboard side of the weir to the
vessel seam line.

The width of the Oil Outlet Zone is deemed to be the same as the oil outlet
vortex breaker, which the spreadsheet calculates at twice the diameter of
the oil outlet nozzle as per Rochelle and Briscoe (2010).

2.6 Oil and Water Retention Times

In this work, retention times for the liquid phases are evaluated in
accordance with NORSOK P-002 (2014), using the respective normal levels
downstream of the flow straightening baffle.

The outboard axial limits of the liquid volumes are demarcated by the weir
plate.

The engineer may specify values in Design Constraints for minimum required
oil and water retention times tʀᴇꜱ,ᴏᴹᴵᴺ and / or tʀᴇꜱ,ᴡᴹᴵᴺ. In this case the
spreadsheet dimensions a separator which meets the more stringent criteria
of retention times or droplet removal.

The engineer may have access to field data for separation based on oil and /
or water retention times or they may wish to follow guidelines for retention
times such as those referred to below.

For 3-phase separators, sufficient oil retention time is required to allow

coalescence of water droplets. Similarly, water retention time is required to
allow coalescence of oil droplets.

Arnold and Stewart (2008) cite oil and water retention times ranging from
180 to 600+ s depending on the density and viscosity of the oil.

This author analyzed data reported by Laleh et al (2013) where a 3-phase

separator operated successfully in the field with ca 60 s oil retention time. Oil
density was 831.5 kg/m³ and viscosity was 5.25 E-3 Pa.s at operating
temperature.

2.7 Re-entrainment of Oil Droplets into Gas Phase

The cost of a vessel increases in line with the diameter, whereas excessive
gas velocity results in re-entrainment of oil droplets into the gas phase. Thus,
the scope for reduction of vessel diameter and cost may be constrained by
the requirement to avoid re-entrainment.

This work uses the correlations which Ishii and Grolmes (1975) developed
based on their experiments, to find the maximum velocity difference
between the gas phase and the oil layer which precludes re-entrainment of
oil droplets into the gas phase.

The spreadsheet design calculations constrain the diameter of the vessel to

meet the maximum velocity difference based on liquid level at hʜʜʟʟ. This is
a more conservative approach than setting the velocity constraint at hʜʟʟ,
since it minimizes the risk of liquid carryover during a level upset.
In practice the gas velocity constraint applies mainly to 2-phase separators.
The diameter of 3-phase separators is more likely constrained by liquid
velocity.

The engineer requires to input a value for oil-gas surface tension in Design
Parameters.

2.8 Slenderness Ratio

In traditional separator sizing methods as per Arnold and Stewart (2008),

customarily a constraint is imposed on the maximum value of

Lꜱ-ꜱ / Di (“slenderness ratio”) to avoid re-entrainment of oil into the gas

phase.

The engineer may input a value for maximum allowable slenderness

ratio Rꜱᵐᵃˣ in Design Constraints when it is desired that the design
calculations employ the slenderness ratio to ensure that the vessel diameter
is sufficient to avoid re-entrainment of oil droplets into the gas.

The engineer may avoid activating the slenderness ratio constraint by

inputting a suitably high value of Rꜱᵐᵃˣ in Design Constraints.

Monnery and Svrcek (1994) provide guidelines for setting the maximum
slenderness ratio as a function of operating pressure.

Viles (1993) provides a more detailed graphical approximation of maximum

allowable slenderness ratio versus oil API gravity, for a range of separator
operating pressures.

Viles indicates that these values of slenderness ratio are based on negligible
oil velocity relative to gas, leading to a conservative evaluation of separator
diameter.

2.9 Controls and Trips

2.9.1 Level Control Intervals

Resolution of the level control and trip settings is a key part of the vessel
sizing calculation.

Level control points are assigned at LIL, NIL, HIL; LLL, NLL, HLL; process trip
points are assigned at LLIL, HHIL; LLLL, HHLL.

The height differences and holdup times between contiguous level control
points and trips for the respective liquid layers are evaluated in compliance
with NORSOK P-002 (2014).
For the purposes of level control calculations:

· The respective phases are considered to be single phase during

steady state operations.

· When calculating liquid holdup times, credit is taken for volumes

contained in the vessel heads.

· The more stringent requirement of height difference and holdup

time governs the dimension of the control interval.

The engineer enters values in Design Constraints for the minimum

acceptable height differences and holdup times between control levels.

NORSOK P-002 (2014) recommends values of 0.1 m and 30 s respectively.

The engineer inputs to the vessel calculation panel trial values for:

· hᴛᴠ :hʜʜʟʟ :hʜʟʟ :hɴʟʟ :hʟʟʟ :hʟʟʟʟ :hʜʜɪʟ/hᴡᴇɪʀ :hʜɪʟ :hɴɪʟ :hʟɪʟ

2.9.2 Unsteady State Operations – Slugging and Surging

Transient intermittent flow conditions may occur in the feed pipe to the
production separator, resulting in the arrival of alternating slugs of liquid and
essentially dry surges of gas.

The ensuing movements in liquid level are deemed to be contained within

the normal operating range between HLL and LLL, so that such conditions do
not trigger level alarms.

The engineer may enter values in Design Constraints for the minimum
acceptable volumes of slugs Vꜱʟᴜɢᴹᴵᴺ and surges Vꜱᴜʀɢᴇᴹᴵᴺ.

to be accommodated within the control intervals between NLL –

HLL and NLL - LLL respectively.

2.9.3 Level Trips

It is assumed that level trip loops are designed with the appropriate SIL
ratings and when initiated will cause closure of ESD valves in the separator
inlet and outlet lines.

However, separator ESD valves are often large, requiring significant time to
close. This may be considered when setting the trip levels.

Low Oil Level Trip LLLL

Transient gas blow-by via the oil outlet line during closure of the ESD valve
following LLLL trip, would potentially overpressure downstream equipment,
or lead to damage of oil pumps due to low NPSH.

The engineer may enter a value in Design Constraints for the minimum
acceptable holdup time ∆t ꜱᴀꜰ(ʟʟʟʟ)ᴹᴵᴺ between the LLLL trip and the top of
the oil outlet vortex breaker sufficient to prevent gas blow-by during closure
of the ESD valve.

Note that the spreadsheet calculates the top of the vortex breaker to be at a
height equal to the diameter of the oil outlet nozzle, as per Rochelle and
Briscoe (2010).

High Oil Level Trip HHLL

Transient flooding of the separator inlet line during closure of the inlet ESD
valve following HHLL trip, may cause gross carryover of oil into downstream
gas processing equipment, leading to machinery damage.

The engineer may enter a value in Design Constraints for the minimum
height difference ∆hꜱᴀꜰ(ʜʜʟʟ)ᴹᴵᴺ between the HHLL trip and the lower edge of
the inlet nozzle to meet the desired allowance for closure time of the inlet
ESD valve.

Interface Level Trips LLIL AND HHIL

LLIL is set coincident with the top of the water outlet vortex breaker, since
transient breakthrough of oil into the water treatment system while the inlet
ESD valve is closing, would not lead to escalation of the incident.

Note that the spreadsheet calculates the top of the vortex breaker to be at a
height equal to the diameter of the water outlet nozzle, as per Rochelle and
Briscoe (2010).

Transient gas blow-by via the water outlet is not considered a credible
scenario.

HHIL is set coincident with the top of the weir, since transient carryover of
water into the oil compartment would not lead to escalation of the incident.

Foaming Allowance

Foaming oil may potentially cause spurious operation of the HHLL trip.

The engineer may enter a value in Design Constraints for the minimum
height allowance for foaming ∆h ꜰᴏᴀᴍᴹᴵᴺ.
to impose a minimum allowable control interval between HLL and HHLL.

2.10 User Guide for the Calculation Spreadsheet

2.11 Navigating the Results of the Separator Design Calculations

The main panel of results for the design calculations shows the level control
heights in semi-graphical format, with final values for the engineer’s design
constraints at the appropriate levels.
The upper part of the main panel shows final values for the dimensions of
the vessel and nozzles.

Below the main panel are final values for the length of the Gravity Separation
Zone, and the mass and volume of each phase of the vessel inventory.

2.12 Architecture of Spreadsheet Calculations

3. CASE STUDY: 3-PHASE PRODUCTION SEPARATOR DIMENSIONS

ADEQUATE FOR EARLY AND LATE FIELD LIFE PRODUCTION RATES
3.1 Summary

This Case Study estimates the minimum size of a 3-phase separator vessel
and associated nozzles required to achieve satisfactory performance in the
separation, process control and safeguarding functions, taking account of the
large variation in water-cut which occurs during its design life.

The peak oil flow case is 100,000 BPD oil, with 25% water-cut: peak water
flow case entails the same total flow, with 75% water-cut. Gas flow is a
constant 40,000 m³/day at separator conditions.

The Case Study uses a Microsoft® Excel spreadsheet to model numerically

the factors listed below and optimizes the vessel dimensions using the
Solver® add-in.

· Gravity settling zone dimensions for successful phase separation.

Droplet settling theory and oil retention time are both considered.

· Axial velocities of oil and water phases maintained within acceptable

limits.

· Adequate liquid hold-up times and height differences between

contiguous level control and trip settings.

· Prevention of transient carryover of liquid during unplanned shutdown.

· Nozzle sizing according to erosional, velocity head and velocity criteria.

The impact of diffuser vs half-pipe inlet-device upon vessel dimensions,

hydrocarbon inventory and the size of the inlet pipework is considered.

Vessel sizing was based upon settling theory, using oil and water droplet cut-
off sizes in the range recommended by Arnold and Stewart (2008), in
conjunction with production fluid density and viscosity data for a North Sea
production separator which was studied by Laleh et al (2013).

Fluid axial velocities and oil residence time were constrained to the values
that were observed in the field for the North Sea separator, which reportedly
operated successfully. This allowed a degree of validation of the Case Study
model and results.

Key Results:
The half-pipe inlet device entails a 24” NB inlet pipe vs 20” NB for the
diffuser device. The half-pipe inlet device results in 539 kg larger
hydrocarbon inventory. This data should be assessed by Mechanical, Piping,
and Technical Safety disciplines to decide on the preferred inlet device.

3.2 Problem Description

A 3-phase separator is to be sized in the FEED study phase of an offshore

oilfield development. The separator shall perform satisfactorily as defined
below, within the operating envelope delimited by peak oil and peak water
production rates.

The volume of the separator is to be minimized, to minimize hydrocarbon

inventory and deck loading.

The impact of installing a diffuser inlet device versus a half-pipe is to be

investigated.

The peak oil production rate is 100,000 BPD during early field life, with a
water cut of 25%; that is to say, coincident water production rate is 33,348
BPD.
The peak water production rate occurs in late field life, when water cut rises
to 75%, while total liquid production rate remains constant; oil and water
production rates are then respectively 33,337 BPD and 100,011 BPD.

Separator gas flowrate throughout field life is 40,000 m³/day measured at

separator operating conditions.

The oil, gas and water feed properties correspond to a North Sea production
separator which was studied by Laleh et al (2013).

3.3 Performance Criteria and Constraints

The separator was designed to deliver water effluent wherein oil droplets are
no larger than 200 microns, and oil effluent wherein water droplets are no
larger than 750 microns. These performance criteria are recommended by
Arnold and Stewart (2008). Note that the production separator studied by
Laleh et al (2013) operated successfully with a less stringent water droplet
cutoff size, of 960 microns.

Maximum oil residence time was constrained to 60s, corresponding to the

observed value in the North Sea production separator mentioned above.

All nozzles were sized adequately for the full range of flowrates, according to
velocity, velocity head and erosion criteria.

Allowable nozzle velocities comply with API RP 14E (1991).

For the diffuser and the half-pipe inlet devices, respective allowable velocity
head values of 7500 Pa and 2600 Pa were selected; these values lie in the
middle of the respective ranges recommended by GPSA (2012).

Erosion rates in nozzles were calculated in accordance with DNV-GL-RP-501

(2015), with Eᵐᵃˣ= 0.02 mm/yr.

Contiguous control, alarm and shutdown levels were provided at intervals of

0.1 m or 30 s, whichever the greater, measured at production flowrate; this
is in accordance with NORSOK P-002 (2014).

A minimum height difference of 0.1 was imposed between HHLL and the
lower edge of the inlet device, to avoid transient flooding of the inlet
following HHLL trip.

Maximum axial velocities of oil and water were constrained to 0.2 and 0.1
m/s at NLL and NIL respectively. These correspond with velocities observed in
a North Sea production separator which was operating successfully, as
reported by Laleh et al (2013).
3.4 Results

Assuming a diffuser-type inlet device, separator / nozzle sizing calculations

were carried out for the Peak Oil and Peak Water cases, to establish vessel
dimensions and nozzle diameters that cover the full operating envelope.

The key calculation input data and results for these cases are illustrated
below in Figure 3.1 to 3.4.

Comparison of the results in Figure 3.2 vs Figure 3.4 shows that the Peak
Water case requires a much larger separator vessel than the Peak Oil Case.
Water and oil nozzle sizes are larger for the respective cases, as would be
expected. Inlet and gas outlet nozzle sizes are the same for each case.

The more conservative vessel and nozzle dimensions were then combined in
a separator model, to calculate fluid inventories.

The Peak Water case fluids were then used to calculate optimized vessel
dimensions and nozzle sizes based on a half-pipe inlet device. The results are
shown in Figure 3-5.

Key results are tabulated below.

4. NOMENCLATURE
5. REFERENCES

American Petroleum Institute (2008). Specification for Oil and Gas

Separators - API Specification 12J (8th Ed.). Washington, DC: API Publishing
Services

American Petroleum Institute (1991). Recommended Practice for Design and

Installation of Offshore Production Platform Piping Systems - API
Recommended Practice 14E (RP 14E) - (5th Ed.). Washington, DC: API
Publishing Services

Arnold, K and Koszela, P. (1990).: “Droplet-Settling vs. Retention-Time

Theories for Sizing Oil/Water Separator”, SPE Production Engineering,
February 1990

Arnold, K. and Stewart, M. (2008). Surface Production Operations Design of

Oil Handling Systems and Facilities (3rd Ed.). Burlington, MA: Elsevier Inc.

Arntzen, R.: “Level Design and Control in Gravity Separators”, Oil and Gas
Facilities, September 2016

DNV GL AS. Recommended Practice, Managing Sand Production and Erosion -

DNVGL-RP-O501 (Edition August 2015). Oslo

GPSA (2012). Engineering Data Book, (13th Ed.). Tulsa, OK: Gas Processors
Suppliers Association
Grødal, E.O. and Realff, M.J.: “Optimal Design of Two- and Three-Phase
Separators: A Mathematical Programming Formulation”, Paper SPE 56645,
presented at the 1999 SPE Annual Technical Conference and Exhibition,
Houston, TX, USA, Oct. 3-6, 1999

Hansen, E.W.M., Heitmann, H., Laska, B., and Loes, M.: “Numerical

Simulation of Fluid Flow Behavior Inside, and Redesign of a Field

Separator”, Proc., 6th International Conference on Multiphase Production,

Cannes, France, 19–21 June 1993, 117–129.

Ishii, M. and Grolmes, M.A.: “Inception Criteria for Droplet Entrainment in

Two-Phase Concurrent Film Flow”, AIChE J., March 1975

Laleh, A.P., Svrcek, W.Y. and Monnery, W.D.: “Computational Fluid Dynamics-
Based Study of an Oilfield Separator - Part II: An Optimum Design”, Oil and
Gas Facilities, February 2013

Monnery, W.D. and Svrcek, W.Y.: “Successfully Specify Three-Phase

Separators,” Chemical Engineering Progress, September 1994

NORSOK (2014). Process System Design – NORSOK Standard P-002 (1st Ed.).
1326 Lysaker, Norway: Standards Norway

Rochelle, S.G. and Briscoe, M.T.: “Predict and Prevent Air Entrainment in
Draining Tanks”, Chemical Engineering, November 2010

Small, J.T. (2022). Optimal Process Design of 2-Phase and 3-Phase Production
Separators for Offshore Oil and Gas Installations : Amazon Kindle Direct
Publishing

Viles, J.C.: “Predicting Liquid Re-Entrainment in Horizontal Separators,”

Journal of Petroleum Technology, May 1993