RECD

UNSOLICITED

AUG 13 i$85

IY(2Q36PUBLICATJOP4S
sp,
-MOLECULAR WEIGHTS IN THE CHARACTERIZATION
DJUSTMENT F C7+
OF PETROLEUM MIXTURES CONTAINING HEAVY HYDROCARBONS

Per Thomassen, STATOIL

*Karen

Schou

Pedersen,
DK-2800 Lyngby, Denmark.
Aage

as.,

N-4001 Stavanger

CALSEP A/S,

Lyngby

Norway.

Hovedgade

Fredenslund,

Instituttet
for Kemiteknik,
Tekniske Hrajskole, DK-2800 Lyngby, Denmark.

29,

Danmarks

ABSTRACT.
Deviations between measured and calculated phase equilibrium
results for oil and gas mixtures are usually

accounted for by

tuning the equation of state parameters.

In this work it is
shown that major improvements in the calculations of the phase
behavior of gas condensate mixtures can be obtained by

taking
into consideration the inaccuracies in the measurements of the
molecular weights of the plus-fractions.
*Author to whom correspondence should be addressed.

Copyright

1987 Society of Petroleum

Engineers

This manuscript was provided to the Society of Petroleum Engineers for distribution
and possible publication in an SPE journal. The materialis subject to correction
by the author(s). Permission tocopy is restricted toan abstract of not more tt,an
300 words. Write SPE Publications Dept., P.O. Box 833836, Richardson, TX
75083-3836 U.S.A. Telex 730989 SPEDAL.

SPE 1,936

INTRODUCTION .
An increasing part of the North Sea petroleum comes from gas
condensate fields.
pressures

are

simulations.

Accurate
needed

Pedersen

i11
et

predictions
reservoir
al. have

predictive C,+ -

characterization

is

of

the dew point

and
process
plant
presented a fully

procedure (1) based on the

Soave-Redlich- Kwong equation of state (2). The composition of
plus-fraction

estimated

dependence of the molar

fraction
mixtures,

slightly
the

better

oil mixtures.

variations

results

logarithmic

carbon

number
In case of oil

are

obtained

true boiling point distillation

100% distilled off.

mixtures.

content

assuming
a given

on the corresponding carbon number.

extrapolating

4%. Larger

by
of

Bubble

Very accurate

results

are

curve

by
to

obtained for

point pressures are predicted to within

deviations are however found for gas condensate

.23.
~ dew point predictions are very sensitive to
in

the

gas/liquici ratios are

C,+

analyses.

underestimated

In
at

most

cases

pressures

the

slightly

below the dew point pressure.

Tuning of equation of state parameters is commonly used in
an attempt to account for deficiencies in
the phase behavior.

the

predictions

As shown by Pedersen et al.

be dangerous, in particular when

erroneous analy+ ~al i.ita.

the

tuning

of

(3) this can

is based on

WE
The composi t ional ana lyses used for oil and ga,s mixtures are
most often given in wei.ght uni,ts. Conversion to molar
composi

each

tions requires knowled,ge of the mo 1 ecular weight of

component and carbon number fraction. The experimental

inaccuracy on the determi,nation of the molecu,lar weight s of

the plus-fractions is of the order 5
lo%.
Deviations in that
order of magni t ude can influence a cal ,cula
ted dew point
pressure of a gas conde nsa te mixture as much as
bar. An
50
obviou,s application of this fact is to use the molecular
weight of the plus fraction as an adjustable pa,raneter.
Adjustment s within the experimenta 1 uncertainty can e.
be
used to match a measured saturat
point.

DATA,.
In this work simulation results are presented for 10
gas
condensate mixtures and 2 oil mixture s. The c
-fracti.ons
7+
have been split into carbon number fractions
For the gas
condensate mixtures, the ana,lyses stops between
and c
c1 o+
20+
For the oi1 mixtures a na.lytical data are available up to c
30+
The simulated results are compared with reSults of con,stant

mass expansion, constant

liberation experiment s.
in

(1b

Volume
The se

depletion

and

differential

PVT-experi,ments are described

).

RESULTS.
Dew point
condensate

ca,lculations
mixtures.

pre:sented in

able

have

Calculated
1.

In

been
and

ord,er to

made

for

measu red
improve

the

10

results

gas
are

the dew point

calculations I it was att,empted to treat the molecu lar weigh

of the plus fraction as an adjUstable parameter, ma,intaining
the measu
weight compos ition
Table 1 also shows the
adjustmen,ts needed in the mo,lecular weights of the plus
fractions to get agre:ement with the measu red dew point.

-3-

SPE 1403(3
The maximum adjustment required is 12 % (gas condensate 10).
The molar compositions before and

after

this

adjustment

is

shown in the Appendix.

Constant

mass expansion and constant volume depletion

simulations were performed

The Peneloux modification
used to calculate

the

(lb)

for the gas condensate mixtures.

(4) of the SRK equation of state is

phase densities.

The relation between

che SRK-volume, ~, and the Peneloux volume, V, is given by:

n
(1)

i=1
where n is the
molar fraction
component i.

number
and

of

the

components and x i and Ci are the

volume

correction,

respectively of

For non-hydrocarbons

and hydrocarbons < C, the

expression for c recommended in (4)-is used:

c=
R is

0.40768(RTc/Pc)(0.29441

the

gas

-ZRA)

(2)

constant, Tc is the critical temperature, Pc

the critical pressure

and

is the Rackett compressibility

RA
factor (5). For C7+ - components Peneloux et al. recommend to
calculate c from a 5th degree polynomial in the carbon
number.

It is seen in (lb) that very inaccurate oil densities

are calculated using this pr~ced.ure. In this work the c-v~lue

of c,+ - components is calculated from the measured density at
atmospheric pressure and

15

C.

-4-

In (1]

the

correlations of Kesler and Lee

are used for

(6)

the acentric factor, u, of the C,+ - fractions.

The following

expression is used for

lighter

the

C7+
approximately C15 (TBr < 0.8):

components

than

(ln P
+ a2/TBr +a~ in TBr
~Br + al
+ a4 Br ) / (a5 + a6/TBr +
a7 ln Br + a8 Br 6,
(II=

Br is the reduced pressure PB/Pc, where PB is the pressure

at which the boiling point
B was measured. The values
recommended in (6) for the coefficients of eq. (3)are listed
in

table 2. A parameter estimation was performed to determine

the optimum values of

these coefficients.

coefficients

of

hydrocarbon

interactions.

zero

were

used
The

for

Binary interaction
all

optimum

hydrocarbon

values

of

the
coefficients a, in eq. (3) are given in tdble 2. Results of
1
dew point calculations on the gas condensate mixtlnes 1 - 10
using this modified expression for

the

acentric

factor

are

presented in table 3.
Figs. 1
curves.

- 5 present measured and calculated liquid

dropout
Calculated results are shown which are based on:

1.

The characterization procedure of (1)

2.

The characterization procedure of (1) where the

molecular weights of the plus-fractions have been
adjusted to get agreement with the measured dew points

3. The characterization procedure of (1) where the modified

w-expression for TBr < 0.8 is used and the molecular
weights of the plus-fractions adjusted to get agreement
with the measured dew points

In figs
depletion

and 7

measured

and calculated
results for gas phases are shown.

the molar fractions of methane

constant

volume

Fig 6 presents
the liberated gas from gas

in

condensate 8 and fig 7 the gas phase

compressibility factors
of the gas from gas condensate 10. These results indicate that
the adjustments of the molecular weights of the plus-fractions
and the modified w-expression

have

very

little influence on

the calculated gas phase properties.

For the oil mixtures of this study
which deviate

by

-3.5%

bubble point pressures

and 0.7% from the measured ones were

calculated using the characterization procedure of (1 ). If the

7+ - molecular weights were adjusted by 19.3% and -2.1%,
respectively, agreement was obtained between the measured and
calculated bubble points.
c

The results of table 3 reflect

point results of a

reduction

in

the
the

influence
amount

on
of

the

dew

analytical

information supplied to

the program. Results are presented

where only the molar fraction, the specific gravity and the
molecular weight of the total C7+ - fraction are used. For
the

two

oil

mixtures

this reduction in the amount

of
analytical information gives bubble point pressures deviati::g
by

-0.93 and 4.2% from the measured ones. This means that
neither for the gas condensate mixtures nor for the oil
mixtures it is necessary to have analytical data beyond C7+ in
order to describe the properties of the mixtures.

-6-

DISCUSSION+
The inaccuracy

in the determination of the molecular weight

of plus-fractions is of the order 5-10%. The results presented
in

tables

and 3 show that deviations of that order of

influeace the
calculated
dew
points of gas
condensate mixtures significantly. There is no systematic
deviation between the calculated and measured dew points of

magnitude

table

(BIAS

-0.2]. We therefore believe that the

are
mainly caused by inaccuracies
in
the
determination of the molecular weights of the plus-fractions
and we find it justified to allow an adjustment of these
molecular weights by up to lo%,
maintaining constant weight
deviations

compositions.

The
dropout curves are

results of figs 1-5 show that the liquid

markedly improved when agreement with the

measured dew point is secured by adjusting the molecular

weights of the plus-fractions. As seen in the Appendix the
molar composition of gas condensate 10 still looks very
similar to the original one after an adjustment of the
molucular weight of the plus -fraction of 12%
Improved liquid dropout curves are obtained by modifying the
correlation of Kesler and Lee

for calculating the acentric

factors of components with a reduced boiling point below 0.8.

This modification

has only little influence on the calculated

saturation pressures (table 3 and figs

recommended for general use.

-7-

6 and 7) and it can be

It

is

seen that

it

has

only

little

influence on the
calculation results if the amount of analytical information
for the C.1+ -components is reduced to a minimum. This is a
strong indication that the
assumption of a logarithmic
dependence
of molar fraction against carbon number
is
reaso~lable.

CONCLUSION.
The

results

condensate

of

phase

mixtures

inaccuracies in the

equilibrium

can

be

molecular

calculations

markedly

on

improved

gas

if

the
weights of the plus -fractions

are taken into consideration by treating these molecular

weights as adjustable parameters, maintaining constant weight
compositions.
points

as

Deviations

large

as

between measured and calculated dew

50 bar

can be accounted for

adjustment of this molecular weight of around 10%, i.

by
e.

an
an

adjustment of the same order of magnitude as the

experimental
inaccuracy on the determination of the molecular weight of a

plus-fraction.
The liquid dropout

calculations for gas condensates can be

further improved by a modification of the expression of Kesler
and

Lee

for

-components

calculating

with

modification has only

the

reduced
little

other properties and is

Accurate phase equilibrium

acentric

boiling

factors

point below

of

C7+
0.8. This

influence on the predictions of

therefore

generally

calculations

applicable.

do not require very

detailed analytical data. An accurate analysis up to C7+ is
for most purposes sufficient.

-8-

NOMENCLATURE .

al

a8

Coefficients of eq. (3)

Peneloux parameter of eq.(1 )

Binary interaction coefficient

Number of components
Molecular weight

Pressure

Gas constant

SG

Specific gravity

Temperature

Molar volume

Molar fraction

Compressibility factor

Greek symbol

(l)

Acentric factor

Subscripts
B

Boiling point

Critical property

Component index

j
r

Component index

RA

Rackett

Reduced property

SPE 1036

REFERENCES .

1.

Pedersen,

K.S.,

P. Thomassen and Aa. Fredenslund

a. IEC Process Des. Dev. 23, 163 (1984)

b. ibid 23, 566 (1984)
c. ibid 24, 948 (1985)
2. Soave, G.,Chem. Eng. Sci. 27, 1197 (1972)
3.

Pedersen, K.S., P. Thomassen and Aa. Fredenslund,On the

Dangers of Tuning Equation of State Parameters, Presented
at the ACS National Meeting Miami Beach, Florida, April 1985

4.

I?eneloux,A., E. Rauzy and R. Freze, Fluid Phase Equilibria 8, 7

(1982)

5.

Spencer, C.F. and R.P. Danner, J. Chem. Eng. Data 18, 230 (1973)

6.

Kesler, ~4.G. and B. 1. Lee, Hydrocarbon Processing 55, 153 (1976)

SPE 14056

TABLE 1.

Experimental and Calculated Dew Point Results for 10 Gas

Condensate mixtures. The calculations are based
on the characterization procedure of (l).

.---No

Temp

Exp Dew Pt

Calc Dew Pt

(c)

(bar)

(bar)

- - - -- - -- -----

-- - -- ---

96.6

- - - -- ------

---

Adjustment

Plus-

of MW

fraction

(1)

- -- - -- -- --.

----

- - - - ---

311.6

282.0
-- -- ---

%Dev

-_----

10.5

- -- - - . -- - - - -- -- - ---_--

(2)

----

-- -- - - - . - --

-8.8

11+
-- - -- -- __ .- _- - - - . _ - - - -

-- . -_----

118.9

398.0

377.1

-5.3

8.3

119.7

394.0

369.7

-6.2

9.6

150.3

381.5

5
- ---

- -- -- - - -- - - - - ---

154.1

- - - - - - - - - ---

7
8

396.3
- - - -- - - ------

385.5
---

- - - - ---

3.9

---

---

410.2
------

- - ----

---

----

367.0

378.7

3.0

129.0

464.0

454.2

-2.1

- - - - ---

----

- - - - ----

-----

136.1

386.4

375.6

10

148.9

542.5

493.5

_____
-9.6

- - -_ -- ------

131.5

- - - -- - -- - -- ---

-5.4
------

6.4
- ---

- - -- - -----2.8

. -- - .

--

- ---

----

---

lo+

C20+
- _- - - - __ _ - -- - - .
c
10+
- _ ~__ - - _ _ ___ _ - _ .

-9.0

5.6

1 o+

20+

20+
- __ - ___ - - - _ - _ __ _

8.4

12.1

20+

-9.0
c20t
---- ------_ ----- ---- --___ -___ _-- --- ---- -------- -- --- -_ -___ _-__ --______ _

%AAD
%BIAS

4.9

-0.2

7.7

1.1

(1) All the analytical data are Used

(2) % adjustment needed in the molecular wei$hts ~f the plus
fraction to get agreement with the measured dew points
maintaining the measured weight compositions

SPE 15036

TABLE 2.
Coefficients of eq
-----

-----

-----

___

(3)

-------

----

----

----

----

----

----

----

----

____

----

____

Coefficient
Recommended in (7)
This Work
--------------------------------------------------------------________
5.92714
5.77881
al
------------------------------------------------------------______ ____
6.09648
6.05615
a2
----------------------------------------------------_______________ ___
I .28862
1.37882
a3
--------------------------- -------------------------_____________-----

0.169347
0.173914
34
-------------------------_________________----------___________ ----- ___
15.2518
15.5523
a5
--------------- ------------------- -------------------------_______ ____
-

-----

6
-----

-----

_____

-----

7
-----

-----

------

-----

a8
--------

15.6875

-----

________

_______

-----

------

13.4721
_____

_____

-----

-----

_____

-----

0.43577
-----

------

__________

15.7915
-----

-----

________

12.7855
-----

________

-----

0.43487
-----

-----

------

------

------

_____

___

SPE 1$036

TABLE 3.
Experimental and Calculated Dew Point Results for the 10 Gas
Condensate mixtures of Appendix A. The coefficients of table 2
I

are used in the omega-correlation of eq.

No

Exp Dew Pt

Calc Dew Pt

(c)

(bar)

(bar)

-. - - ----

%Dev

- ------

- - ---

---

(bar)

(2)

296.2

6.8

-- ----

Adjustment

Calc Dew Pt %Dev

(1)

301.2

282.0

96.6

1
---

Temp

(3)

- - ___ -_ - - - ---

-5.7

5.0

-- - -- _ - - ______

. - --

__ _ _ _ _ . .
-13.5

93.8

235.0

225.7

-4.0

237.7

1.1

118.9

398.0

374.7

-5.9

369.3

7.2

8.3

119.7

-7.4

370.2

6.1

3.8

- - --

----

394.0
----

364.8

- - ----

--

----

- -- -- - -- --

---

- -- -----

----

- - - - _ - -- - - __ . _ - _ _ _ _ - .

150.3

381.5

392.5

2.9

390.5

2.4

-3.5

154.1

385.5

409.7

6.3

406.5

5.4

-7.4

131.5

367.0

373.8

1.9

388.8

6.0

-4.1

129.0

464.0

460.6

444.3

4.2

-- - - - - -- - -- _- - -- - _- ______
9

136.1

----

----

10

148.9

-----

-----

----

-------

_______

-----

-- - -- -- - - - ___ -_ - - - - --1.7

----

----

497.6

542.5
-----

- - _____
379.7

386.4
----

-0.7

------

-----

-----

5.7

----

----

----

491.9
-----

-----

-----

-----

-----

-----

-----

------

-----

-----

-----

-----

------

-----

-----

(1) All the analytical data are used

-----

----

-----

-----

--------

-----

-----

-----

____

----

_____
6.9

----1.3

-----

_ _ _ _ .,

9.5

5.2

-1.0
-----

----

_____
5.1

9.3

4.6

%BIAS
-----

-----

- - _____

364.5

-8.3

%AAD
-----

----

1.7

_____

____

0.0
-----

_____

----

(2)

The C7+ - characterization is based on the analytical

information available for the total C7+ - fraction only

(3)

adjustment needed in the molecular weight of the plus

fraction to get agreement with the measured dew points .
%

maintaining the measured weight compositions. All the

available analytical data are used i.e. the plus-fractions
are those listed in table 1

APPENDIX

MEASURED

COMPONENT

MOLAR

COMPOSITION

MOLE%

OF

MOLECULAR
WEIGHT

N2
C02
cl

GAS

CONDENSATE

DENSITY

AT 15

(G/cM3)

.39
3.47
80.17

C2

6.28

C3

2.75

IC4

. 43

C4

.88

IC5

.31

C5

.35

C6

.54

C7

.72

C8

.88

C9

cl o

95.
106.
118.

.7432
.7612

.33

:32.

.7869

cl 1

.24

149.

.7883

cl 2

.20

163.

.7980

cl 3

.21

.8195

cl 4

.18

175.
194.

cl 5

.15

.8332

Cl 6

.11

203.
217.

c1 7

. 12

235.

.3383
.8360

Cl 8

.03

247.

.8415

cl 9

.06

255.

.8520

C20+

.57

396.

.8673

55

10

.7786

.8307

COMPONENT

MOLE%

MOLECU LAR

DENSITY

WEIGHT

N2
co 2
cl

G / CM 3)

*39
3.47
80.22

C2

3.28

C3

2.75

IC4

.43

C4

.88

IC5

.31

C5

.35

C6

.54

C7

. 72

C8

.88

106.

.761

C9

.55

116.

.779

cl o

.33

132.

.787

cl 1

.24

149.

.788

C12

.20

163.

.799

cl 3

.21

175.

.820

cl 4

.18

194.

.831

cl 5

.15

203.

.833

Cl 6

.11

217.

.833

cl )

.12

235.

.836

Cl 8

.09

247.

.842

cl 9

.08

255.

.852

C20+

.51

444.

.867

96.

AT

743

15

Vol% Liquid of
o
o o

y
0

I?oint

Volume

b
,

m
,

x. \

\\ x

-Al-

o
0
0

o
0

I;x

I i
1

lx

/
/

x /
w y
o
0
0

,/
,/

,/
w
(D

!24

/
/

..

&
o
0

u.
c
m
rt

m
,
0

SPEW376

Vol% Liquid of Dew Point Volume

p
o
, 0
0

N
e
o

b
0

~---

I
I
I
U-I

*o
o

I
-. -

I_
.

m
,
0

,____

____

#
o
*
0

p
0

..

I
II

I
I
I

__J_______
[x .

lx

- . .-. ..__

,\

1-

. . .... .
\

I
.
..

1-

Ul

0
0

L__
I

N
0
0
0

I
--

.-

1
1

5N
w
(D

In,

m
c
n
(D

I
(n

. ..

.__.

.. .

I
I
I

1
1
----

I
... .- ___i._.._..__

%
I
l-.
,

.- . .--

. ----

.,

..-

-,

.-.

vol% ~lqula or Dew Point volume

1-h

po

o
.

D
. . ---- ------

0
..

._.___T_..-

---__-,

-.

.-

I
I

I
I

----

-1
:i
H

CL

u.

c
m

l-r
0

Q
i-r

..-

!- .--. -

QJnJ

PI-h
P.

I
I

\\

.. ----- ..--- ...-

i----.-J.

1OJ
(-F

r-,

I
I

Hl
~

It
lx

c!

Lo

I
L.

_____

.. ----

.--

._

~_..__

,-

. . . . . ------

---l--Y----

;-

/ 1!i

i
1

-----

---

/!

L.-_.__._.
. _-1_._-.
- ,---.L---- ..-_..
....i_

..

. .i

Vol% Liquid in % of Dew Point Volume

p
G

r--

. ....~..

.--------

w
y
0

M
p
0

N
y
0
,-.
.. -1

\
A!

---- -. -

I
I

I
I

-.

________

I
I

k...

P
;
o

. . .. J _. _.__.

ii
-.,----n-._? .1

_x\._

.-

._..-

..-

.-_.

J
I
I
[

z;
YID
x
i-h o

\;

Xj

\\

G
,
c1

.-.

.-._, ._._ .

____

!
.\

- ..-

-x

. -.. .

\
-..-_

_L._-..

.-

---4 .-~=

..

<.-
t

.~_-.-

[1

0
l-h

~- _ ---- ----i/1

-_

-- .

4-

..

..

1
I
I

(n

c1
.

I
L_
.... ._______._ __ --____
._i
____
____ _ i__________.

I
1

I
..-. i

Vol% Liquid of Dew Point Volume

0

p
o

b
*

p
0

.i

I
I
i

lx

0
0

0
a

,1I
I

0
0
0

k0
0

c1

0
0

x
/

/
7/
/

Mel% Methane in the Gas Phase

m
p
o

Ln

0
,0
0

m
(n
s
o

1;

-+$

5s

Q191

@l

>

I-41!21
l--

(-)

fDrt
QIO

I-J.

Ii

m
(Q

Ill

0
p
0

(n

-&_._.

..

_____

!=
1-

rr
m

n
D

In
(lJ

. -______

:
,

(3

.0
a)

0
0
0

_
El

II

/
u-l

0
p
0

I
~

_..__L._-._

-.

tiasPnase
compressibility
Factor

1*
0

o 0

1.
1-

,
N

!1

~1

/7x

d%

SClclla

OS-TX

SLsunJm

N
c1
,-

0
0

u.

k
0

c
UY

Ln

0
0

1----

0-l

0
0