Because inhibitors inhibitor concentrations.

are important in reduc- Required model parameters for one

ing formation of hy- of the shortcut methods are reported
drates in gas streams, for the first time.
we performed a study to In general, engineers should check
investigate and evaluate the process simulator results against
, the accuracy of available experimental data before using the

- tools that predict hydrate simulator for an actual process. All three
formation conditions in the presence of shortcut methods studied were accurate
inhibitors. down to 200 F; below that temperature,
We the Moshfeghian-Maddox method is
evaluated most accurate but requires more calcu-
two com- lations.
studytestsaccuracyof methods mercial
process- Hydrate inhibitors
that estimatehydrateformation simula-
tion
Inhibitor injection is one of the
practical means for preventing hydrates
programs from forming in process equipment
and three shortcut methods. The study and natural-gas transportation pipe-
covered wide ranges of pressures and lines. Accurate knowledge of hydrate
Mahmood Moshfeghian formation conditions in the presence
John C. Bourdon Based on a presentation to the 22nd European of inhibitors is therefore very useful for
JohnM. Campbell & Co. Symposium on Applied Thermodynamics, June 28- safety and economic reasons.
Norman, Okla. July 1,2006, Elsinore, Denmark. Many materials when added to water
R.N. Maddox will depress
Oklahoma State University PROCESSFLOWDIAGRAM Fig.1 the hydrate
Stillwater, Okla.
Feed gas Gas and freezing
temperatures.
For many
practical
reasons,
Heater feed . Separator feed Liquid
hydrocarbon
alcohol or a
glycol-usu-
Heater/cooler
'0";"'"' I Separator
ally methanol,
diethylene

-=.J Aqueous phase

glycol (DEG),
or monoeth-
ylene glycol
(MEG)-is

EFFECT OF C02* Fig.2

5.5
5.0
c: 4.5
.g~
'" . 4.0
E ~
~ ::> 3.5
.E1O
Q) ~
~ Q)
3.0
'" Co
-t; E 2.5
>-Q)
J:~ 2.0
1.5
1.0 .

o 2 4 6 8 10 12
C02 in feed gas, mole %
"Gas E at 66.021 MPa with 40 wt % methanol.

44 Oil & Cas Journal/Jan. 8, 2007

used as an inhibitor. All can be recov- EFFECTOFN-HEXANECOMPOSITION* Fig.3
ered and recirculated in a process, but -44.8
recovering methanol may not be eco-
-45.0 ~ ProMax
nomic in many cases. c:
The total injection rate is the amount .2
_0U -45.2
'" .
of inhibitor needed in the liquid water E ~ -45.4
0_'"
plus inhibitor that enters the vapor and !'" Q.
:;;-45.6
hydrocarbon liquid phases. Any inhibi- -oE
tor in the vapor or liquid hydrocarbon ~! -45.8
-46.0
phase has little effect on hydrate forma-
tion conditions. -46.2 .

o 2 3 4 5
Determining the amount and con- nC6H14in feed gas, mole %
centration of inhibitors and their distri- "Gas D at 19.703 MPa with 85 wt % methanol.

bution in different phases is important

for practical purposes. Determining the
required amount and concentration of GAS COMPOSITIONS Table1
these inhibitors is possible with several
Stream A B C D E F
thermodynamic models for hand and
rigorous calculations that have been de- Component, mole %
N 7.00 5.96
veloped and incorporated into software cb, 14.19 3.00
CH 84.13 71.60 93.51 89.99 84.50 74.130
programs. c,i46 4.67 4.73 4.58 6.31 8.70 7.210
CH 2.34 1.94 1.31 2.40 3.80 4.500
This study evaluated the accuracy of 0.10 0.30 0.900
iS~\O 0.93 0.79 0.20 0.50 1.810
two commercial process simulators- nl.~':'0 0.10 0.10 0.870
ProMax1and HYSYS.2Threeshortcut iC..:;\I"
m"'SH12 0.93 0.79 0.10 0.10 0.890
0.10 0.30 9.690
methods can be used to calculate the nC6H'4
Inhibitor MeOH MeOH MeOH MeOH MeOH MEG
required concentration of inhibitor and Concentration
range, wt % 10-20 10-20 65-85 65-85 20-40 25-50
the injection rate for dewpoint correc- Pressure, psia 131-2,729 151-2,778 95-2,955 110-2,926 5.366-13.474 105-2.879
10
Data points 13 18 15 14 13
tion, NGL recovery, or pipeline trans- Reference 7 7 8 8 9 10
portation of natural gas.The calculation
procedure is shown in Chapter 6 of
"Gas Conditioning and Processing," Vol. experimental data include pressure, In some reports, overall composition
1.3The three methods we evaluated are temperature, dry gas composition, and can be approximately inferred by the
those developed by Hammerschmidt,4 concentration of inhibitor in water phase equilibrium data presented, but
Nielsen-Bucklin,5 and Moshfeghian- solution mixed with the hydrate former. McIntyre10disputes that these data do
Maddox (OGJ,Aug. 30,1993, p. 78). Overall compositions are not provided. not extend to the higher inhibitor con-
The study
examinedmulti- PROCESS
SIMULATOR
ESTIMATES* Fig.4
component natural
15
gas mixtures over Methanol added Iwt %),
a wide pressure data source
13
range, up to 100 _ 0,Experimental
MPa, and identi- '" 11 _ 0,ProMax
'iij
fied the strengths Q.
o _ 0,HYSYS
and limitations of o
q
9
~ _ 20,Experimental
these methods. ~
::J 7 _ 20,ProMax
II)
II)
Q) _ 20,HYSYS
Computer Q: 5
simulators _ 40,Experimental
Various litera- 3 -0- 40,ProMax
ture sources report _ 40,HYSYS
hydrate formation
pressure and tem- 30 40 50 60 70 80 90 100
perature condi- "GasE. Hydrate formation temperature, oF.

tions.6-9Reported

Oil & Gas Journal/Jan. 8, 2007 45

 PROCESS SIMULATOR ACCURACY FOR METHANOL* Fig.5 EOUATIONS
100 .",

_ (MW)(T.- T)
More than 65 wt % methanol x.. - A + (MW)(T.- T) (1)

1 -T
'"
a
50 x. =1- exp (~ ) (2)

E
.., IIII
.,a.
E X.
_- 1 - 1
~,Oexp n!1[T (Mifll - To]) (3)
!
c: 0
...., i I .9
10 ~H - 2,063
II E fiR = a+ a. 10-3 + II. InP (4)
.E
.,
10 -50
I I II In"l",o= (x.)'(B+ 2(A - B)(l-x.)] (5)
-0
>-
I ""
"C a,T
I ! A= a. + T;OOO (6)
.,

c;;
-100
·~ a,T
u
. ~;~.~~" I I B = ~ + 1,000 (7)
Nomenclature
-150 A = Moshfeghian-Maddox constant
-150 -100 -50 0 50 100 AH = Hammerschmidt constant,
1,297 in 51 units or 2,335 in
Measured hydrate formation temperature, oF. English units
"GasesA-Ewith 10-85WI % methanol. AN. = Nielsen-Bucklin constant, -72 in
51 units or -129.6 in
English units
B = Moshfeghian-Maddoxconstant
of inhibitor in the MW = Molecular weight of inhibitor
PROCESSSIMULATORACCURACYFORMEG* Fig.6 P = Pressure, psia
aqueous phase. xR = Mole fraction of inhibitor in
100 aqueous phase
Mixing effects XR = Weight fraction of inhibitor in
aqueous phase
=
-
Q)
"Cm 50
:3
will change the
temperature; we
T Minimum flowing temperature,
°C. or oF.for Equations 1 and 2,
oR. for Equation 3
>-.,
""a. therefore added To = Hydrate forming temperature,
"C E °C. or oF.for Equations 1 and 2,
--
., .,
.!!! c:
a heater-cooler oR, for Equation 3
:3 0
.2 0 with no pressure " = Constant
., ';::
., ", = Constant
U E drop to readjust ", = Constant
.E the mixed separa- 13 = Constant
13, = Constant
-50 _ tor feed stream 13, = Constant
-50 o 50 100 & = Constant
Measured hydrate formation temperature, oF. temperature to 'YH,o= Water activitycoefficient
"Gas F with 25 WI % and 50 WI % MEG.
the experimental
hydrate formation
temperature value. Waterrate
centrations presented, the region where This process flow diagram ensures For 100 mole/hr of Gas E in the
they are most needed. that the condition of hydrate-formation presence of 40 wt % methanol, we
Fig. 1 shows the process flow dia- temperature in the three-phase separa- varied the water rate from 1-10 mole/
gram used in this study to simulate the tor is close to those measured experi- hr and predicted hydrate-formation
experimental measurements. Feed gas, mentally.We therefore used it in both temperatures using the two simulators.
inhibitor, and water are mixed at the commercial process simulators. For each water flow rate, we adjusted
mixer. All three streams are at the same the inhibitor stream rate to produce the
pressure and temperature. Overall effects specified inhibitor concentration of 40
Table 1 shows different feed-gas wt % methanol.
compositions. The inhibitor stream is To investigate the impact of overall. Hydrate-formation temperature is
either pure methanol or MEG,and the composition (which was unknown) independent of the water rate in both
11 water stream is pure water.The water- on hydrate formation temperature, we simulators. In subsequent runs, we set
stream flow rate was set to a small flow adjusted the water rate, C02 concentra- the water-stream rate to 10 mole/hr for
rate for the resulting aqueous phase. We tion, and heavy ends. 100 mole/hr of gas.
adjusted the inhibitor-stream flow rate
to produce the specified weight percent

46 Oil & Gas Journal! Jan. 8, 2007

-
 ~-~. ._-~---~ ~._--------_._- ~-~ ~_.--.--

tion from 90.17 to

ESTIMATIONMETHODS* Fig.7
85.95 mole %.The
15
Methanol added Iwt %1, reported composi-
13 estimation method tion of nC6HI4was
. 0, Experimental 0.3 mole % and
'" 11 - 0, Parish-Prausnitz that of methane
"iii
c. Q) 20, Experimental was 89.99 mole
0
0
q
9 _ 20,Hammerschmidt %.
!I
i _ 20, Nielsen-Bucklin
7
Changing the
::>

. I II)
II)
Q)
-0- 20, Moshfeghian-Maddox nC6HI4composi-
It 5 e 40, Experimental tion by a factor of
_ 40, Hammerschmidt 50, and conse-

1LJ
3
__ 40,Nielsen-Bucklin
40, Moshfeghian-Maddox
quently changing
the overall compo-
sition, the hydrate
30 40 50 60 70 80 90 100 formation tem-
'GasE. Hydrate formation temperature, OF.
perature changed
only 1.2° C. with
ProMax. HYSYS
METHOD ACCURACYFORMETHANOL* Fig.8 could not converge for this test run.
100 Fig. 3 shows these results.
Even though overall composition
. Hammerschmidt is important, Figs. 2 and 3 show that
if . Nielsen-Bucklin variation of overall composition has
50
a'" .6.Moshfeghian-Maddox only a slight effect on the predicted
:;; hydrate temperature and the error is
c.
E within 12° C.This magnitude of error
I: o is normally within the experimental
.g'" error and is not in agreement with
E McIntyre's remark that "predictions
.E
-50 of hydrate formation conditions in
f!
"C process simulators is difficult to verify
>-
.l:
Ii "C
Q)
in cases where the mutual solubility
1;; .
-----!. of the hydrate former and inhibitor is
"3
<.> -100
c;;
.. . relatively high."lo
U

... .
-150
-150 -100
Measured hydrate formation temperature, OF.
'Gases A-E with 10-85 wt % methanol.
C02 composition
The reported composition of C02 in of 40 wt % methanol is less than 2° C.
Gas E was 3 mole %; we therefore var-
ied the C02 concentration in our study
from 0.1 to more than 10 mole %. For
each mole % of C02' the gas mixture
was normalized, but other conditions
remained the same as the reported
experimental values.
Fig. 2 shows that, even though C02
concentration varied more than 100-
. I
The authors also warn that, "with-
out knowledge of the overall compo-
sition, a significantly wide range of
-50 0 50 100
predicted conditions is possible. More
significantly, depending on the nature
of the system, therefore they suggest
that caution must be exercised when
fold, the variation of predicted hydrate- using these experimental data for
formation temperatures in the presence design or other use since the reported
conditions are for an unknown overall
composition."
Heavy ends
We conducted a similar study for Gas Evaluation results
D in the presence of 85 wt % methanol Table 1 shows the composition,
in which the nC6H14composition in the inhibitor range, pressure range, number
feed gas was varied from 0.1-4.8 mole of data points, and the reference of the
%.This change in nC6HI4corresponds experimental data for the gas mixtures
to a change in the methane composi- studied. Fig. 4 shows the ability of Pro-
Max and HYSYSto predict the hydrate-

48 Oil & Gas Journal/Jan. 8, 2007

 formation temperature for METHOD
ACCURACY
FORMEG*
GasE. Figs. 5 and 6 show the 100
accuracy of these software ~ Fig. 9 shows that all three
products for the mixtures in Q) oj
~ ~
Table 1. '" :::J
-g. ~ 50
Figs. 5 and 6 indicate that ..c: Q)
Q.
for methanol inhibition, the S~
"'~
lower limits of hydrate forma- g o
(ij ',j:;
a . Hammerschmidt
tion temperatures are -75° E u '"
E
. Nielsen-Bucklin
.a. Moshfeghian-Maddox
.,
(-60° C.) for ProMax (maxi- .E!
mum of 70 wt % methanol) -50 .
-50 o 50
and -25° E (-32° C.) for HY-
SYS(maximum of 50 wt %
methanol). HYSYScould not
Measured hydrate formation temperature, oF.
"Gas F with 25 WI % and 50 WI % MEG.
converge for casesof 85 wt %
methanol. For MEG, both simulators give model were not previously published.
accurate results down to 0°E (-18° C.) In this study,Table 2 shows the required
corresponding to 50 wt % MEG.
parameters for Equations 4-7; these are of hydrate formation temperatures are
released for the first time.
Shortcutmethods The required hydrate-formation
Three shortcut methods were evalu- temperatures in the absence of inhibitor
ated in this study (see equation box)- (pure water) were predicted by the Par-
those developed by Hammerschmidt, ish ahd Prausnitz11model for all three in-
Nielsen-Bucklin, and Moshfeghian-
Maddox.
hibition shortcut methods. This ensured
the same basis and accurate results.
Hammerschmidt4 was the first'to Fig. 7 shows that the Parish-Prausnitz
present a correlation for calculation of
the required mass fraction of inhibitor
method predicts the hydrate- formation
temperature for Gas E in the absence of
in the aqueous phase. Equation 1 shows inhibitor (0 methanol) accurately.This
the widely used correlation of Ham- method was used to add its prediction
merschmidt. temperature to the depression tempera-
Nielsen and Bucklin proposed a ture predicted by the three shortcut
more accurate correlation,S shown in methods for the sake of easy compari-
Equation 2. son with the experimental data.
To predict inhibition effects accurate- Fig. 7 also indicates that all three
ly for pressures as high as 14,500 psia methods give good results for 20 wt %
(100 MPa) and higher concentrations methanol; however,for 40 wt % metha-
of methanol and MEG,Moshfeghian
and Maddox suggested a correlation,
nol, the Hammerschmidt results deviate
from the experimental data considerably.
shown in Equation 3. Fig. 8 shows that all three methods
They proposed correlations for the give accurate results for temperatures
enthalpy of formation-hydrate number as low as 200 E (-6P C.) equivalent
(Equation 4) and used a temperature- to maximum of 25 wt % methanol.
dependent Margules model for water At lower temperatures (or higher
activity coefficient (Equation 5). methanol concentrations) the Hammer-
Equations 6 and 7 define the tem- schmidt method deviates from the ex-
perature-dependent parameters. perimental data considerably. For lower
Unfortunately, parameters for this temperatures, the Moshfeghian-Maddox
MOSHFEGHIAN-MADDOX MODEL PARAMETERS
'"
Inhibitor "', "', 13, 13, 13
MEG -2.32455 -2.17985 0.66598 -2742068 2.24035 0.03563
MeOH -5.84663 15.04912 -2.82261 10.91332 0.76499 -0.03124
50
5 -
Fig.9
gives better results than the
Nielsen-Bucklin method.
methods give accurate results
for Gas F up to a concentra-
tion of 50 wt % MEG (as low
as 00 E) for which experi-
mental data were available.
Findings
Based on our study, we
100
concluded that:
. The accuracy of a
process simulator should be
checked before its application
for an actual process.
· For methanol, the lower limits
-75° E (-60° C.) for ProMax (cor-
responding to a maximum of 70 wt %
methanol) and -25° E (_32° C.) for
HYSYS(corresponding to a maximum
of so wt % methanol).
· For MEG, both ProMax and HYSYS
give accurate results down to 0° E,
which corresponds to SOwt % MEG.
· All three shortcut methods give
accurate results for temperatures as
low as 20° E (-6.7° C.), which cor-
responds to a maximum of 25 wt %
methanol. At lower temperatures (or
higher methanol concentrations) the
Hammerschmidt method deviates from
the experimental data considerably. For
lower temperatures, the Moshfeghian-
Maddox method gives better results
than the Nielsen-Bucklin method.
. All three shortcut methods give ac-
curate results for Gas F up to a concen-
tration of SOwt % MEG (as low as 0°
E), for which experimental data were
available. ..
References
1. ProMax, version 1.2, Bryan Re-
search & Engineering Inc, Bryan,Tex.,
2005.
2. HYSYS,version 3.2, Build 5029,
Aspen Technology Inc., Cambridge,
Table
2 Mass., 2003.
3. Campbell, J.M., "Gas Condition-
I)
ing and Processing, Vol. 1, the Basic
-0.07746
0.10006 Principles," 8th Ed., Norman, Okla.:
J.M. Campbell & Co., 2001.
Oil &CasJournal/Jan. 8, 2007
 4. Hammerschmidt, E.G.,"Forma-
tion of gas hydrates in natural gas
transmission lines," Ind. & Eng. Chern,
Vol.26 (I 934), p. 85 I.
5. Nielsen, R.B.,and Bucklin, R.W,
"Why not use methanol for hydrate
control," Hydrocarbon Processing, Vol.
8. Blanc, c., and Tournier-Lasserve,
J., "Controlling hydrates in high-pres-
sure flowlines," World Oil, November
1990.
9. Ng, H.-]., Chen, q., and Rob-
inson, D.B.,"Research Report RR-92,"
Tulsa: Gas Processors Association, 1985.
Methanol-A Review and New Concerns
over Experimental Data Presentation,"
presented at the 83rd Annual GPACon-
vention, New Orleans, Mar. 14-17, 2004.
11. Parrish, WR., and Prausnitz, ].M,
"Dissociation Pressures of Gas Hydrates
Formed by Gas Mixtures," Ind. Eng. I II

62, No.4, April 1983, p. 7 I. 10. McIntyre, G., Hlavinka,M., and Chern. Proc. Dev.,Vol. 11 (1972), No. I,
6. Ng, H.J., and Robinson, D.B., Hernandez, v., "Hydrate Inhibition with pp. 26-35.
"Research Report RR-66,"Tulsa: Gas
Processors Association, 1983.
7. Ng, H.J., Chen, c.]., and Robin-
son, D.B.,"Research Report RR-l 06,"
Tulsa: Gas Processors Association, 1987.

The authors
Mahmood Moshfeghian '-'
(mahmoodm@jmcampbell.
com) is a senior research
engineer at John M. Campbell
& Co., Norman, Okla. Before "
joining JMC, he was profes-
sor of chemical engineering at
Shiraz University, Iran, where
he served as department head
and associate dean of research in the college of en-
gineering. He was previously professor of chemical
engineering at the University of Qatar and a senior
research scientist at the Kuwait Institute for Scien-
tific Research. Moshfeghian holds BS, MS, and PhD
degrees in chemical engineering from Oklahoma
State University. He is a member of AICHE. THEENERGY
INDUSTRY'S
MOSTPOWERFUL
JOBBOARD
John C. Bourdon is vice-presi-
dent of the John M. Campbell
& Co. He has more than 30
years' experience in hydro-
carbon processing. Bourdon
Post. Search. Work!
is a Campbell gas instructor
for "Gas Conditioning and
Thousands of new industry jobs (Apply for free!)
Processing" and consultant spe-
cializing in sour gas processing. Confidential resume posting available
He has been involved in the development of several
sulfur-related technologies and mechanical innova- E-mail job alerts for instant notification of the latest postings
tions. Bourdon holds a BS in chemical engineer-
ing from the Georgia Institute of Technology and Weekly career-oriented newsletter
advanced degrees in other fields. He is a registered
Salary Wizards (Are you getting paid enough?)
Professional Engineer and member of AIChE, SPE,
and Chi Epsilon Sigma Honor Society.

RobertN. Maddox (rmaddox

@okstate.edu)is the Sheerar
Chair emeritusprofesso.r Post your profile today: www.PennEnergyJOBS.com
of chemicalengineeringat
OklahomaStateUniversity.He I
servedasheadof the chemical
engineeringdepartmentfrom
1958 to 1978. He is a fellow
of AIChE anda memberof
GPA.
~ TurningInformationinto innovationI ServingStrategicMarketsWorldwidesince1910

Oil & Gas Journal/Jan. 8, 2007 57

I
.I