design of pipelines for the simultaneous flow of oil and gas

8/10/2019 Design of Pipelines for the Simultaneous Flow of Oil and Gas

1/16


2/16

2

DESIGN

OF

PIPELINES

FOR

THE SIMULTANEOUS FUM

OF

OIL

323-G

tenfold greater than for smooth pipes.

The major factor causing high pressure drop

i s

the

energy

required

to

move

the

l iquid

through

the

l ine. For most cases

this i s

energy supplied by

the

gas.

Additional energy is used in the violent

r i s -

ing and

fe.lling

of the l iquid in

the

l ine.

This

energy

must

come from a

reduction

in

pressure.

When

th is

factor

i s combined with

the

effects

of decreas

ed

pipe

diameter

and

roughness t becomes more

ap

parent why

large pressure drops

occur.

FLOW

PATTERNS

Several

flow p a t t e r n ~

have been

recognized

in

two-phase

flow. Sketches of

the various

types

are

shown

in

Figure

2. Alves

3

has described

these

as

follows

I

Assume a horizontal pipe with l iquid flow

ing

so as to f i l l

the pipe and consider

the

types of

flow that occur

as

gas i s

added

in increasing amounts

a.

Bubble

Flows Flow

in which

bubbles of

gas

move along

the

upper par t

of the pipe a t approximate

ly the

same

velocity as the l iquid. This

type i s

similar

to

Froth

Flow where

the

entire

pipe

is

f i l led

wi th a froth similar to an emulsion.

b. Plug

Flows Flow

in which alternate

plugs

of

liquid and gas

move along the upper part of

the

pipe.

c. Strat i f ied Flow: Flow in

which the

l iquid

flows along the bottom of the pipe and the gas flows

above, over a smooth l iquid-gas interface.

d.

Wavy

Flow:

Flow which

i s

similar to

s t ra t i

fied flow except that

the

gas moves a t a higher ve

loci ty and

the interface

i s

disturbed by

waves trav

eling in

the

direction

of

flow.

flow pattern regions as functions of G,

the

mass ve

loc i tyof the

gas phase,

and L/

G

, the

rat io

of mass

veloci i es of the

liquid

and

gas phase.

Since most of

the

ave.ilable date. were for

the

air -

water system

a t

atmospheric pressure, correction fac

tors have been introduced to gad

just

for other l iquids

and gases. Holmes suggested these terms

for

corre

lat ing the flooding point

in

wetted wall dist i l la t ion

columns.

The gas mass velocity i s

divided by

A

=

~ G / o . 0 7 S ) (62o.3/f L) and the L/G

rat io i s

multiplied

by'l \ p where ~ 1 / . 3

0/

=

~ ~

+

~ t J

0

P and

are

the gas and l iquid densit ies a t

flow

ing cond1tions in pounds per cubic

foot.

The

surface

tension of the l iquid 'Y',

i s in

dynes per centimeter

and the l iquid viscosi t y f i L, i s

in

centipoise.

Although

the

borders of the

various

flow pattern

regions

in

Figure .3 are shown

as

l ines , in real i 7

these

borders

are rather broad

transition

zones.

Each observer

probablY,selected the

t ransi t ion

at

sl ightly

different

points.

Not

a l l observers

have

used the

same nomenclature

and t

was

necessary

to

equate

terms in

some

cases. Figure.3

i s

based on

data

from

one,

two

and four inch pipe.

LITERATURE

ON

PRESSURE DROP

Much of

the early

work on

two-phase pressure

drop

was published by

workers a t the

University of

California s tar t ing

in

19.39.

9

,10,11,12,1.3 These

studies resulted in the

correlation

presented by

Lockhart

and

Martinelli

in

1949.

12

Their

method

i s

as

follows: Calculate

.the

pressure drop of the l iq-

uid phase

assuming

that i t is the

only fluid flow

ing in the

pipeline.

A

similar calculation

i s

made


3/16

323-0

OVID BAKER

this paper. Lockhart

and

Martinelli

did not

con

sider

the

effect

of the various flow patterns in

their correlation although their basic assumptions

tended

to

l imi

tit to

annular

flow.

W h ~ n l t h e i r paper

was presented the data

of

Jenkins

4

, 2 cited d lU ing

the discussion,

indicated

that additional terms would be required to accurate

17 predict ~ e s s u r e

drop.

Later

in

1949 Gazle7

and

Bergelin a t

the

UniverSity of Delaware pre

.ented

data

on st rat i f ied and wave flow

in

a two

inch pipe. They

obtained

pressure

drops

consider

ab17 lower

th n those

predicted by the Lockhart and

Martinelli Clll Ve.

lbeir

results suggested that

the

Lockhart ana178is

was not valid for st rat i f ied flow

or

that

two

inch pipe had

a

different

relationship

between

G

and

X.

Data fer crude oi l well streams a t various gas

oi l ratio. were reported by Van Wingen

in

1949.

1

He

measured two-phase pressure drops

in

the gatherilll


4/16


5/16

323 G

OVID BAKER

5

s t rat i f ied

flow

are shown in Figure 12.

4

,14

ore

experimental work i s

needed

for th is

flow

pattern.

Indications are

that

the pressure drop calculated

by

the lockhart and

Martinelli curve

l118 y

be 105 to

205

times

greater

than experimental values.

How-

ever, the data points for one inch and

ten

inch

pipe f a l l very close to

the lockhart

and

Martinelli

correlation.

0'.l'H R

TYPES

OF FWd

In

th is

paper a l l

the

experimental

data con

sidered

were

for both gas

and

l iquid

phases

in tur-

bulent flow. This i s the

usual

case in indus t r ia l

applications.

For

data

oovering

the ojher cases

referenoes

should

be consulted. Alves

and Gasley-U+

give helpful

data for

one and

two inoh pipe

sizes.

DESIGN SmGFSTIONS

The modifications of the

basic Lockhart

and

.....

nel l i correlation proposed in

this

paper are proD-

)ly

fai r ly rel iable,

but

the similarity

of

new de

signs to

our experimental condi tions

should

be con

sidered carefully in each oase. The l imits

of

error

for the various

equations

should be studied

before

a safety

factor or

load factor for

the

flowing quan

t i t i e s

is

seleoted.

I t will

be

noted that

the

measured pressure

drops

were

always

small

fractions of

the

to tal pres

sure. t

i s

suggested that i the caloulated pres

sure drop i s

more

than 10 per cent of the

downstream

absolute pressure the pipeline be calculated

in two

or

more

sections.

The magnitude

of the

errors in -

volved

for

_gas

flow

are

discussed

by

Poettmann

23

and

Clinedinst0Z4.

t should

be

emphasized

that

the quanti

t ies

and

physical properties of the fluids used in the cal-

Perhaps

pipe diameters may

be selected

that

~

keep

uphill flow

in a

safe

pattern. Kosterin s tates

that

in a one

inch

pipe

a t

a

72

0

angle

with

the

hori

zontal, the flow patterns were

not

affected by

the

inclination

a t

velocit ies

exceeding

ten

feet

per

second.

For

his air-water system a t

atmospheric

pressure th is

velocity

corresponds

to the

s tar t

of

froth

flow. He

found the

greatest pressure surges

in s t rat i f ied

flow a t

velocit ies

in

the range

of

1.6

to 6.6 feet per second.

Data

for

vertical

two-phase flow

have

been pre

sented

by

various

authors

8

,27,28,29,30,31

Perhaps

their

data

would be

helpful in evaluating

inclined

two-phase

flow.

ACKNOWLEDGMENTS

The author wishes

to

thank the ~ g n o l i a Petrole

um

Company

ana the Uontinental Oil

Company

for per

miSSion

to

publish th is information.

So

many per

sons

have contributed

to

th is project that

t

would

be impractical to l i s t

them

al l . Magnolia

personnel

included

the following people.

The

equiIDent

was

in -

stalled and operated

under the

direction

of J. Eo

Shannon,

Superintendent,

C. A.

Nevels,

Do M. Ball,

and

C. C.

Baird. B.

C.

Stone Was construction

en

gineer for the

project . W

H. Speaker

supervised

the

t es t

and planned the project . M. R.

Hindes

made

many of the calculations. G. A.

Lundberg, J .

F.

Wright, Fred Wilson,

C.

O. Childress and Will GUl.ett

of Magnolia and

C. E. Iamb

of

Continental made the

tests

with the help of

those l i s ted

above.

J. C. Van

daveer

and

R.S. Garvie

prepared the

figures. n-

alyses and viSCOSity measurements were

made by

the

Magnolia

Natural

Gas and

Field

Research

laboratories.

NOMENCLATURE

D

= nside diameter of pipe,

inches


6/16


7/16

323-0

OVID BAKER

7

a Vertical Pipe I : . AIME, (1940), ~ 79.

29. Poettmann, F. H. and Carpenter, P. G.a liThe

Multiphase Flow

of

Gas, Oil and Water Through

Vertical

Flow Strings I (March

21, 1952) Paper

presented

a t

Wichita, Kansas,

API

Paper

No.

851-26-1, available from Dlvision of Produc

tion

American Petroleum Institute Dallas

Tems.

.30. Kra1bill

R. R.

and Williams,

Brymer

a

"Two

Phase

Fluid

Flow, ridging

Velocities

in

Wetted-Wall Columna,. Paper presented

at

the

San

Francisco meeting

of

the American Insti-

tute of Chemical Engineers, September 14, 195.3.

31. Calvert Se,mourl

Vertical Upward,

Annular

Two-Alase Flow in

Smooth

Tubea, \I Fh.D. ' heaia,

University

of

Michigan, 1952.

Table 2

Analysis of Fluid Streams from Outlet

Separator (Run No. .3)

Magnolia - Continental Teata

CO e2nent

Mol

Ga.s Mol Liguid

~

0.10

C

95.0.3

22.95

C2

.3.26 4 32

C.3

0.84

2.71

i04

0.22

1.49

n04

0.26 1.71

iC

5

0.06

1.49

C6+

~

63.26*

100.00

100.CO

Engler Distillation

of

Oil

Volume

%

Table 2 (Cont.)

Viscosity

of

Liquid

(Composition Change With Pressure)

PSIG

600

700

800

900

1000

1070

Viscosity

Centipoise @

77

F

0.657

0.635

0.614

0.59.3

0.572

0.557

Table .3

Friction Factors Used for Single

Phase

Flow

Calculations

Reynolds Number,

Be

1,000

2,000

.3,000

10,000

40,000

100,000

150,000

400,000

1,000,000

4,000,000

10,000,000

Friction Factors - CommercialF1pe

1 - 4 inch 6 - 4 inch

0.0157 0.0157

0.01.32 0.0126

0.0119 0.0110

0.0087 0.0078

0.0064 0.0056

0.0054 0.0046

0.0050 0.0042

0.0042 0.00.37

0.00.36 0.00.32

0.0029 0.0027

0.0026 0.0023


8/16


9/16

Table l

(Cont)

Experimental

Data

Experiment No.

8

9

10

11 12

13

14

Run

No

J

2

%

1

:

8

7

MSCFD

25,552 12,050

11,886

6,474

4,348

7,471.6

9,:338.4

Bbls/Day

5,484

4,167

6,592

4,970 5,420

627

721

Length

Line, Ft.

41,333

31,115 41,333

41,333 41,333

3,666

3,666

Inside Diameter,

In.

10.136

10.136

10.136

10.136 10.136 4.026

4.026

Inle t

Pressure,

PSIG

975

962 960

952 930

1,087 1,096

Outlet, Pressure,

PSIG

946

948

936

936

912 1,067

1,075

Gas Gravity Air=l)

0.59 0.59

0.59

0.59

0.59

0.625

0.625

Ga.

Density II/cuft

3.37

3.38 3.32

3.32 3.28

4.31 4.31

Liquid Density II Gal 6 ~ 9 9

6.525 6.499

6.53

6.103 5.11

5.11

Gas

Viscosity,

cp

0.014 0.014 0.014 0.014 0.014 0.0145 0.0145

Liquid Viscosity, cp 0.577 0.58

0.578

0.58

0.589 0.557

0.557

Surface

Tension,

Dynes/em

16.7 16.7 16.7 16.7

16.7

16.7 16.7

4.27

4.27

4.27

4.27

/ 5

4.98 4.98

A

7.62

7.58

7.58

7.58

7.71

9.66 9.66

Line Temperature,

of

80

69

78

66 82

79

80

Two Hla,se L P PSI)

29

14

24

16 18 20

21


10/16

Table 1 Cont)

Experimental

Data

Experiment No 15

16 17

18 19

20

21

Run No 8 7 10 9 7 9 8

MS FD ~ , 9 7 9 . , ~ 9,3 8.4

11,767 6,483.5 4,440.7 6,483.5 6,668.0

B b ~ D a y 627 721 192 136 76 136 107

IMlgtb Line, Ft.

14,790 14,790 11,427 11,427 11,427 10,617 10,617

Inside

Diameter, In. 5.937 5.937 7.750 7.750 7.750 7.750 7.750

Inlet Pressure, PSIG

1,070 1,076

712

705.5 1,075.5 703 1,067

Outlet Pressure, PSIG 1,055 1,060 703 703 1,075.0 701.5 1,065.6

Gas Gravi ty

. A 1 r ~ 1 ) 0.625 0.625 0.62 .60 0.60 0.60 0.60

Gas

DenSity I/cuft

4.28 4.28 2.62 2.62 3.99 2.62 3.99

L1quid

Density I/Oal 5.11 5.11 6.15 6.15 5.68 6.15 5.68

Gas Viscosity, cp .0145 0.0145 .0143 .014 0.014 0.014 0.014

L1quid Viscosity, cp

0.557 0.557 0.63 0.63 0.557 0.63 0.557

Surface

Tension, n,nes/cm

16.7 16.7

4.98

9.66

18

4.25

6.86

65

9

18

15

5.17

8.82

18

4.25

6.86

65

15

5.17

8.82

70

6.86

Line

Temperature,

of

72 73

16

65

69

0.5

wo Phase 6.

P PSI) 15 2.5 1.5

1.5


11/16

Experiment No.

Run Noo

MS FD

Bb1s/DaY

Length

Line, Ft.

Inside Diameter, In.

Inlet Pressure, PSIG

Outlet Pressure, PSIG

Gas Gravity (Air=l)

Gas Density /cUt

Liquid Density /Ga1

Gas

Viscosity,

cp

7

6,797.9

102

10,617

7.750

1,07;

1,074

0.60

3.99

;.68

Liquid Viscosity, cp

Surface Tension, pynes/em

0.014

0.557

15

;.17

8.82

Line Temperature,

O

Two Phase

b

P (PSI)

69

1 0

Table l (Cont)

Experimental Data

23

24

2;

8 9

ll,950

9,476.7 6,483.5

236 244

136

ll,313

11,313 22,044

7.750 7.750 7.750

1,064 1,074 705.;

1,062 1,067 701.5

0.60 0.60 0.60

3.99 3.99 2.62

5.68 5.68 6.15

0.014 0.014 0.014

0.557 0.557 0.63

15

5.17

8.82

70

2.0

15

5.17

8.82

69

7.0

18

4.25

6.86

66

26 27

8 7

ll,950

9,476.7

2.36 244

41,317 41,317

10.136 10.136

1,06.3

1,068

1p55.5 1,058

0.60 0.59

3.99 3.88

5.68 5.68

0.014 0.014

0.557 0.557

15

5.17

8.82

70

7.5

16.7

4.65

8.7

69

10


12/16


13/16

glOO

:::> e

o

FIG.

5

I

,-\

..J

6

LIQUID HOLD UP

IN

T E S T s , ~ l ~

al

4

C

lLJ

~

:::>

6

lLJ

~

a::

4

I.L.

o

I -

Z 2

lLJ


14/16

30

FIG.

7

1

SLUG FLOW

/

FOR I

8 SMALLER PIPE USE DASHED

Y d/

INE,

LOCKHART a MARTINELLI CURVE

/

FOR 3

TO 10 PIPE:

,r

YL V

f

1190

XO

8IS

4>GTT=

L

0.5

~ , X

i

000 / _,0 :/ ~ . ,

~ O y

~ ~ . ~ /

;;;7

10

8

c GTT

vY 1)

v

00

W

; f - ~ V

..

/ 1 I /V

/

_

~ -

.'

/

. - - ~

6

4

2

I - - ~

7

~

, ; ~ ~ ~ ~ ~ r ,

1.0

0 1

2

4

6

FIG. 8

ANNULAR

FLOW

~ G T T =(4.8-0.3125 D) X

0.343-0.0210

1.0

1.0.

( f t r

2tl

X=

~ A ~ L

I I

I I I

I I I

8 1 0

2

4

6

8

10 0


15/16

FIG. 9

L

FROTH OR BUBBLE

/-1

FLOW

J7

Lt

>GTT-

14.2

X 0.75

L 0.1

7P

) 0.87"

PIPE

( J

3"

PIPE

I

1,0,

~

O l y . ~ _ /

00 1

I

, 00

I ~ Z ~ O

~ G T T

00

' ~ O O '

J, 0

0

. ~ ~ , O ,

/ ~ /

1/

r---- 7 / - ~ / .

V ~ / LOCKHART MARTINELLI

CURVE

100

, .

I

/

-

i [ ; ~

I

I

I

I

I

I

10

10

o

X ~ A P L

APG

I

JIO

FIG.

10

00

J

PLUG FLOW

I

I

0.855

$GTT=

27.315

X

L

0.17

O.S7"PI PE

I:l

3"

PIPE

i J ? ~ L ' z r

ol

)

/ _ I . -V

\04

d,,1--+-

o \-?-

0 t ~

7

, ,00

10

0

Z

, , -h

' - ,

I ~

GTT

1 ~

84 / / '

1/1 PoO/

~ o o

~ o ~

~ ~ : J -

~ ~

, / ' .

LOCKH RT

Ii MARTINELLI

CURVE

r- -

X - ~ l 1 P L

- l1P

G

I I

10

10

10

o

Io

I

o


16/16

100

FIG. II

8

-

FLOWAVE

-

1.0.

L

REFEIENCE

-

T.02

31,200

-

f

J;;;

> ;;;:;J

1.02

46 100

4

~ ~ = ~ ~ ~

1.02

56,200

4

..- J

I

.....

6

4

0

2.068'

6I,OOO"9I,DOO

14

-

e

115

4,500-7,400

TAILE I

2

I 10.136

4,110

.

UtiLE 1

GTT'

1-.

8

6

./--

/-

4

.l.

design of pipelines for the simultaneous flow of oil and gas

Documents