Flow Assurance

서유택
Superficial velocity

• For single phase flow, the mean velocity is defined by the volumetric
flowrate of the single phase divided by the cross sectional area.
• If there is more than one phase, there needs to be a way of describing
the velocity of each phase. The most convenient way is the phase
superficial velocity, which is the velocity the phase would have without
the other phase.
• The superficial velocity is defined as the ratio of the liquid or gas
volumetric flowrate to the total pipeline cross-sectional area.
• Note that superficial velocity is not the same as the velocity at which
the phase itself moves, because superficial velocity reflects the relative
flow rates.
: This can be understood by thinking of the case in which most of the cross
section is liquid, but a few gas bubbles move with the liquid at the same
velocity. The superficial velocity of the gas is then much smaller than the
superficial velocity of the liquid even though the velocities of the two phases
are the same.
• The ratio between the superficial velocities is not the same as the ratio
between the fractions of the cross section occupied by each phase.
: That can be seen by thinking of a pipeline in which half the cross section is
liquid and the other half is gas, but the gas is moving 10 times faster than the
liquid.
Basic flow variables

• The fluid Mixture Velocity is defined as the sum of the superficial gas
and liquid velocities

• Liquid holdup is defined as the ratio of the liquid volume in a pipeline

segment to the whole volume of the pipeline segment. Liquid holdup is
a function of both space and time
• Gas void fraction is defined as the ratio of the gas volume in a pipeline
segment to the whole volume of the pipeline segment.

• From the above two equations, the sum of the liquid holdup and gas
void fraction equals one.
• Average gas and liquid velocities
: If the superficial velocity and liquid holdup are known and the liquid holdup
would not change longitudinally, the average gas and liquid velocities can be
calculated as

• Slip velocity
: Due to the density difference, when gas and liquid flow simultaneously
inside a pipeline, the gas phase tends to flow faster than the liquid phase. The
gas is “slipping” away from the liquid. The Slip Velocity is defined as the
difference of the average gas and liquid velocities
• In homogeneous gas and liquid two phase flow, there is not slippage
between gas and liquid, and the slip velocity equals zero.
: Then the liquid holdup can be calculated as

• Water cut is defined as the ratio of the water volumetric flowrate to the
total water and oil volumetric flowrates,

• The density of gas and liquid homogeneous mixture is expressed as

 Exercise 2.

• Determine actual and superficial velocities

: Determine the superficial liquid and vapor velocities (Vsl and Vsg) and the
actual liquid and vapor velocities (Vl and Vg) for the following liquid holdups,
flow rates, temperatures and pressures.
: Use the gas composition and spreadsheet from exercise 1 to calculate the
vapor volume per MMscf.

Component Mole % In-situ vap Liquid

N2 0.50 Flowline ID Temp. Press. STO Oil Swelling Vsl Vsg Vl Vg
or Holdup
CO2 1.50 (in) (F) (psia) (kbpd) (%) (ft/s) (ft/s) (ft/s) (ft/s)
H2S 0.00 (MMscfd) (%)
He 0.00 8 120 2000 3 30 20 95
C1 88.00 8 80 1200 9 30 12 70
C2 5.00 8 40 300 18 30 8 40
C3 3.00 10 120 2000 5 50 20 96
i-C4 0.40 10 80 1200 15 50 12 72
n-C4 1.00
10 40 300 30 50 8 44
i-C5 0.20
n-C5 0.30
C6 0.10
Beggs and Brill model for two phase flow

• No slip liquid holdup

: λL is defined as the ratio of the volume of the liquid in a pipe segment divided
by the volume of the pipe segment which would exist if the gas and liquid
travelled at the same velocity (no-slippage). It can be calculated directly from
the known gas and liquid volumetric flowrates.
qL
L  ,  g  1  L
qL  q g
• For no slip condition,
U sg U sl U
For No  Slip : u g  u L or   λL  sl
1  λL λL Um
• Froude number of the mixture

um2
Frm 
gD
where, D is pipe ID and g is gravitational constant
• Transition lines for correlation

L1  316 λl0.302 , L2  0.0009252 λl-2.4684 , L3  0.10 λl-1.4516 , L4  0.5 λl-6.738

• Determining flow regimes

Segregated if λl < .01 and Frm < L1 or λl ≥ .01 and Frm < L2

Transition if λl ≥ .01 and L2 < Frm ≤ L3

Intermittent if .01 ≤ λl <0.4 and L3< Frm ≤ L1 or λl ≥ .4 and L3 < Frm ≤ L4

Distributed if λl < .4 and Frm ≥ L1 or λl ≥ .4 and Frm > L4

• For segregated, intermittent and distributed flow regimes used the
following
abl
H l  H l0 , H l0  c
(horizontal liquid holdup)
Frm
• Actual liquid holdup is obtained by multiplying Hl0 by a correction factor


  1  C sin1.8θ   0.333sin3 1.8θ  
C  (1 - λl )ln(d λle N vlf Frmg )
• Liquid velocity number
ρl 0.25
N vl  1.938 u sl ( ) usl: no slip velocity
gσ
• For transitional flow, the liquid holdup is calculated using both the
segregated & intermittent equations and interpolating using the
following
H l  AH l (Segregated)  B H l (Intermittent)
L3  Frm
A , B  1 A
L 3  L2
• General pressure gradient equation
: The pressure gradient equation which is applicable to any fluid flowing in a
pipe inclined at an angle φ from horizontal was derived previously. This
equation is usually adapted for two-phase flow by assuming that the two-
phase flow regime and two-phase properties can be considered
homogeneous over a finite volume of the pipe.

In-situ average density Depend on the using correlation

L H L  g H g U sL  U sg

 dP  g f tp ρmum2 ρs dum2
    ρs sin  
 dZ  g c 2g c d 2g dL

ΔPelevation ΔPfriction
Pressure change due Frictional
to hydrostatic head pressure gradient
• The no slip friction factor fn is based on smooth pipe (ε/D=0) and the
Reynolds number, Re.

f tp  f n e S
ln(x)
S 
( 0.0523  3.182ln(x)  0.8725ln(x)  0.01853ln(x)
2 4

λl
x 2
Hl
• Liquid Viscosity (μL): μL may be calculated from the oil and water
viscosities with assumption of no slippage between the oil and water
phases as follows:
 L  o f o   w f w
• Two-Phase Viscosity: Calculation of the two-phase viscosity requires
knowledge of the liquid holdup. Two equations for two-phase viscosity
are used by various investigators in two-phase flow:
μm  μL λL  μ g λg , μs  μ μ
HL Hg
L g

• Liquid Surface Tension (σL):

 L   o fo   w f w
Two phase flow correlations

• Many correlations have been developed for predicting two-phase flow

pressure gradients which differ in the manner used to calculate the
three terms of pressure gradients equation (elevation change, friction
and acceleration terms):

a. No slip, no flow regime considerations: the mixture density is calculated

based on the no slip holdup. No distinction is made for different flow
regimes.

b. Slip considered, no flow regime consideration: The same correlations for

liquid holdup and friction factor are used for all flow regimes.

c. Slip considered, flow regime considered: Usually a different liquid holdup

and friction factor prediction methods are required in each flow regimes.
Exercise 3. Determine pressure drop

• In-situ Vapor MW = 20.6

• Ql =200 bpd, ρl = 49.9 lb/ft3, μl = 2 cp
• Qg=1mmscfd, ρg=2.6 lb/ft3, μg=0.0131 cp

1. Find the flow regime, from Froude number and no-slip liquid holdup

2. Find the pressure drop for elevation change per length (assume θ=90)

3. Find the frictional pressure drop per length

Slug catcher

• What are they?

- Large capacity/volume separators designed to receive unsteady
multiphase flow from multiphase pipelines
- Two main types
: Finger or pipe type
: Vessel type
Causes of slug/intermittent flow

• Operation in the slug flow regimes

: Hydrodynamic slugging
: Gas velocity > Fluid velocity
• Terrain induced slugging
: ± 10% changes
: Start up operation
• Flow rate changes
: Ramping up
• Pigging
: Sweeping liquids
• Flowline/riser geometry: Severe slugging
Flow pattern map of Mandhan

• Horizontal two phase flow in pipes

Hydrodynamic slug size prediction (FPS vs SI)
•

• Design slug length typically taken as 4 ~ 5 times Ls

Pipeline liquid holdup
Simple holdup correlation - Flanigan

• Slug size is based on “Hold Up” difference between flow rate 1 and 2

𝑞𝑔 𝑧 𝑇
where, 𝑉𝑠𝑔 = 𝐴
𝑑2 𝑃
qg in MMm3/d, T in K, d in m, P in kPa, A in 5.19 (SI)
Severe slugging
Finger type slug catcher
Finger type slug catcher

• Kollsnes, Norway
• Woodside slug catcher – 32,500 bbls (5000 m3)
Shell Goldeneye Vessel type slug catcher

• 66’ L * 13.1’ D * 3.6’ wt, 240 ton, 1200 bbls slug capacity
Slug catchers

• Finger type normally used for slug volumes larger than 1500 bbls
• Fingers are normally 36 ~ 48 inch, 300 ~ 800 ft long
• Typical capacities are 3000 ~ 8000 bbls
- Woodside North Rankin (NWS, Australia)
: 32,500 bbls, 14~48” fingers, 1150 ft long
- BP Nam Con Son (Vietnam)
: 25,000 bbls
- Statoil Snohvit (Nroway)
: 17,000 bbls
• Typical installed cost (Finger type): $1500/bbl
Heat transfer on a surface

• If a surface and a moving fluid have a temperature difference,

the convection will occur between the fluid and surface.
• All solid surfaces with a temperature will emit energy in the form
of electromagnetic waves, which is called radiation

• Although these three heat transfer modes occur at all subsea

systems, for typical pipelines, heat transfer from radiation is
relatively insignificant compared with heat transfer from
conduction and convection.
Conduction

• For a one-dimensional plane with a temperature distribution

T(x), the heat conduction is quantified by the following Fourier
equation:

Where,
q” : heat flux, Btu/(hr ft2) or W/m2, heat transfer rate in the x direction per
unit area;
k: thermal conductivity of material, Btu/(ft hr oF) or W/(m K);
dT/dx: temperature gradient in the x direction, oF/ft or oC/m.
• When the thermal conductivity of a material is constant along
the wall thickness, the temperature distribution is linear and the
heat flux becomes:
• For a steady heat transfer, the right side of equation is equal to
zero.
• The total heat flow per unit length of cylinder is calculated by
following equation:

Where,
r1, r2: inner and outer radii of the cylinder medium, ft or m;
T1, T2: temperatures at corresponding points of r1, r2, oF or oC;
qr: heat flow rate per unit length of cylinder, Btu/(hr ft) or W/m.
Convection

• Both internal and external surfaces of a subsea pipeline come

in contact with fluids, so convection heat transfer will occur
when there is a temperature difference between the pipe
surface and the fluid.
• The convection coefficient is also called a film heat transfer
coefficient in the flow assurance field because convection
occurs at a film layer of fluid adjacent to the pipe surface.
U-value

• Figure shows the temperature distribution of a cross section for

a composite subsea pipeline with two insulation layers.
• Radiation between the internal fluid and the pipe wall and the
pipeline outer surface and the environment is ignored because
of the relatively low temperature of subsea systems.
• Convection and conduction occur in an insulated pipeline as
follows,
: Convection from the internal fluid to the pipeline wall;
: Conduction through the pipe wall and exterior coatings, and/or to
the surrounding soil for buried pipelines;
: Convection from flowline outer surface to the external fluid
• For internal convection at the pipeline inner surface, the heat
transfer rate across the surface boundary is given by the
Newton equation:

where
Qi : convection heat transfer rate at internal surface, Btu/hr or W;
hi : internal convection coefficient, Btu/(ft2 hr oF) or W/(m2 K);
ri : internal radius of flowline, ft or m;
L: flowline length, ft or m;
Ai : internal area normal to the heat transfer direction, ft2 or m2;
Ti : internal fluid temperature, oF or oC;
T1: temperature of flowline internal surface, oF or oC.
• Conduction in the radial direction of a cylinder can be described
by Fourier’s equation in radial coordinates:

where
Qr : conduction heat transfer rate in radial direction, Btu/hr or W;
r: radius of cylinder, ft or m;
k: thermal conductivity of cylinder, Btu/(ft hr oF) or W/(m K);
𝜕𝑇/𝜕𝑟: temperature gradient, oF/ft or oC/m.

• Integration gives:
• The temperature distribution in the radial direction can be
calculated for steady-state heat transfer between the internal
fluid and pipe surroundings, where heat transfer rates of
internal convection, external convection, and conduction are the
same.
• The following heat transfer rate equation is obtained:
• The heat transfer rate through a pipe section with length of L,
due to a steady-state heat transfer between the internal fluid
and the pipe surroundings, is also expressed as follows:

where
U: overall heat transfer coefficient based on the surface area A, Btu/ (ft2 hr
oF) or W/ (m2 K);

A: area of heat transfer surface, Ai or Ao, ft2 or m2;

To : ambient temperature of the pipe surroundings, oF or oC;
Ti : average temperature of the flowing fluid in the pipe section, oF or oC.
Calculating U-value for single layer

• For single layer,

where, Di =inside diameter and Do = outside diameter

Multilayer insulation

• The U-value for a multilayer insulation coating system is easily

obtained from an electrical-resistance analogy between heat
transfer and direct current.
• The steady-state heat transfer rate is determined by:

where UA is correspondent with the reverse of the cross section’s

thermal resistivity that comprises three primary resistances: internal
film, external film, and radial material conductance.
• The relationship is written as follows:
• The terms on the right hand side of the above equation
represent the heat transfer resistance due to internal
convection, conduction through steel well of pipe, conduction
through insulation layers and convection at the external surface.
• They can be expressed as follows.
• Therefore, the U-value based on the flowline internal surface
area Ai is:
1
Ui 
1 ri ln(r1 /ri ) ri ln(r2 /r1 ) ri ln(ro /r2 ) r
    i
hi k1 k2 k3 ro ho

and the U-value based on the flowline outer surface area Ao is:
1
Uo 
ro r ln(r /r ) r ln(r /r ) r ln(r /r ) 1
 o 1 i  o 2 1  o o 2 
ri hi k1 k2 k3 ho
• U-value is a function of many factors, including the fluid
properties and fluid flow rates, the convection nature of the
surroundings, and the thickness and properties of the pipe
coatings and insulation.
• Insulation manufacturers typically use a U-value based on the
outside diameter of a pipeline, whereas pipeline designers use
a U-value based on the inside diameter.
• The relationship between these two U-values is:
Exercise 4. U-value calculation

ID (in) WT (in) K-value (W/m/K) U-value (W/m2/K)

6 1 0.20
6 2 0.20
6 4 0.20
6 1 0.01
6 2 0.01
10 1 0.01
10 1 0.20
10 2 0.20
10 4 0.20

Note: U-value is based on the tube ID.

Insulation design for flowlines
Thank you