Petroleum 2 (2016) 296e300

Contents lists available at ScienceDirect

Petroleum
journal homepage: www.keaipublishing.com/en/journals/petlm

Original article

A prediction method of natural gas hydrate formation

in deepwater gas well and its application
Yanli Guo a, *, Baojiang Sun a, Keke Zhao b, Hongkun Zhang a
a
b

School of Petroleum Engineering, China University of Petroleum (East China), Qingdao, China
China Petroleum Engineering Co., Ltd., Beijing Company, Beijing, China

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 20 April 2016
Received in revised form
16 May 2016
Accepted 16 June 2016

To prevent the deposition of natural gas hydrate in deepwater gas well, the hydrate formation area
in wellbore must be predicted. Herein, by comparing four prediction methods of temperature in
pipe with eld data and comparing ve prediction methods of hydrate formation with experiment
data, a method based on OLGA & PVTsim for predicting the hydrate formation area in wellbore was
proposed. Meanwhile, The hydrate formation under the conditions of steady production, throttling
and shut-in was predicted by using this method based on a well data in the South China Sea. The
results indicate that the hydrate formation area decreases with the increase of gas production,
inhibitor concentrations and the thickness of insulation materials and increases with the increase of
thermal conductivity of insulation materials and shutdown time. Throttling effect causes a plunge
in temperature and pressure in wellbore, thus leading to an increase of hydrate formation area.
Copyright 2016, Southwest Petroleum University. Production and hosting by Elsevier B.V. on
behalf of KeAi Communications Co., Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords:
Deepwater gas well
Hydrate prediction
Steady production
Downhole throttling
Well shut-in

1. Introduction
The low temperature and high pressure makes it easy for the
formations and depositions of natural gas hydrate in deepwater
oil and gas eld, thus the ow assurance has drawn more attentions [1e4]. To prevent the blockages caused by hydrate in
deepwater gas well, the prediction of hydrate formation conditions must be performed. Currently, many domestic and foreign
scholars proposed respective prediction methods of natural gas
hydrate formation [5e8]. However, these methods were established based on some assumptions, the accuracy and applicable
conditions tend to be limited. OLGA is the industry standard tool
for transient simulation of multiphase petroleum production [9].
Its models cover from bottom hole into the production system,
involving the wells, owlines, pipelines and process equipment.
It is being used for design and engineering, mapping of

operational limits and safety analysis. The uid physical property

parameters are calculated using PVTsim for the simulation with
OLGA. PVTsim is provided by Calsep company, with the functions
of the PVT simulation, hydrate formation prediction and wax
scale, multiphase ash calculation, regression analysis, unit operations, and so on. Hence, according to the characteristics of
deepwater gas well test and exploration, a method based on
OLGA & PVTsim for predicting the hydrate formation area in
wellbore was proposed by comparing four prediction methods of
temperature in pipe with eld data and comparing ve prediction methods of hydrate formation with experiment data.
Moreover, the hydrate formation under the conditions of steady
production, throttling and shut-in was predicted by using this
method based on a well data in the South China Sea. The results
can guide the ow assurance and gas well test operation.

2. Methods
* Corresponding author.
E-mail address: guoynli@126.com (Y. Guo).
Peer review under responsibility of Southwest Petroleum University.

Production and Hosting by Elsevier on behalf of KeAi

2.1. Temperature and pressure model

The main ow pattern in the deepwater gas well is gaseliquid
annular ow that a layer of liquid lm slowly rises along a pipe
wall dragged by a column of high speed gas ow containing lots
of liquid droplets [7]. The extended two-uid model applied by

http://dx.doi.org/10.1016/j.petlm.2016.06.004
2405-6561/Copyright 2016, Southwest Petroleum University. Production and hosting by Elsevier B.V. on behalf of KeAi Communications Co., Ltd. This is an open
access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Y. Guo et al. / Petroleum 2 (2016) 296e300

OLGA includes three separate continuity equations, two momentum equations and one mixture energy-conservation equation [9].
(1) Continuity equations.
For the gas phase,



v

1 v

Vg rg 
AVg rg vg jg Gg
vt
A vz

2.2. Natural gas hydrate equilibrium model

(1b)

For the liquid droplets,

v
1 v
Vd
V r 
AVd rl vd  jg
jl  jd Gd
vt d l
A vz
Vl Vd

(1c)

(2) Momentum equations.

A combined momentum equation for the gas phase and
possible liquid droplets,

 



 vp
v

Vg rg vg Vd rl vd  Vg Vd
vt
vz

1 v 

AVg rg v2g AVd rl v2d
A vz
1   Sg
1
S
 li rg jvr jvr i
 lg rg vg vg
2
2
4A
4A



Vg rg Vd rl g cos a
jg

entrainment and deposition rates, kg/m3$s: Gf is the possible

mass source of Phase f , assumed to enter at a 90 angle to the
pipe wall, carrying no net momentum, kg/m3 s; p is pressure, Pa;
Sg ; Sl ; Si is the wetted perimeter of the gas, liquid and interface,
respectively, m; a is the pipe obliquity, rad; HS is the enthalpy
from mass source, J/kg; E is the internal energy per unit mass,
J/kg; h is elevation, m; U is the heat transfer from pipe walls,
J/m3; g is gravitational acceleration, m/s2.

(1a)

For the liquid lm,

v
1 v
Vl
V r 
AVl rl vl  jg
 jl jd Gl
vt l l
A vz
Vl Vd

297

The multiphase ows always develop towards the minimized

state of chemical potential. Therefore, the hydrate phase equilibrium model applied by PVTsim is established using the
chemical potential theory [10e12].
The difference between the chemical potential of water in the
hydrate state and in the pure water state can be expressed as:





mH  ma mH  mb mb  ma

where, a; b and H and used to identify each of the three states

considered. mH  mb is regarded as the stabilizing effect on the
hydrate lattice caused by the adsorption of gas molecules,
calculated as follows:

mH  mb RT

NCAV
X

vi ln 1 

(2a)
A separate equation for the liquid lm,

 

v
vp
1 v 
1
S
Vl rl vl Vl

AVl rl v2l  ll rl jvl jvl l
vt
vz
A vz
2
4A
1
S
Vl
li rg jvr jvr i Vl rl g cos a  jg
va
2
4A
Vl Vd


 vV
 je vi jd vd  Vl d rl  rg g l sin a
vz

(3) Energy equation.

A mixture energy-conservation equation,





v
1
1
mg Eg v2g gh ml El v2l gh
vt
2
2






1 2
v
1
mg vg Hg v2g gh

md Ed vd gh
2
vz
2





1 2
1 2
ml vl Hl vl gh md vd Hd vd gh
2
2
Hs U
(3)
where, V is volume fraction; r is density, kg/m3; v is velocity, m/s;
A is the pipe cross-sectional area, m2; Subscripts g; l; i; d indicate
the gas, liquid, interface and droplets, respectively; jg is the mass
transfer rate between the phases, kg/m3 s; jl , jd is the

!
YKi

 . 
RT ln fw fw0

(5)

where, vi is the number of cavities of type i; YKi denotes the

possibility that a cavity i is occupied by a gas molecule of type K;
NCAV is the number of cavities per unit cell in the hydrate lattice;
N is the number of components present, which may enter into a
cavity in the hydrate lattice; R is the gas constant;
lnfw =fw0 ln xw , if injecting the inhibitors into the drilling
uids, lnfw =fw0 lnyw xw ; yw ; xw is the activity coefcient and
the molar fraction of the water in the water-rich phase, respectively. mb  ma is the difference between the chemical potentials
of water in the empty hydrate lattice and water in the form of
liquid or ice, calculated as:

mb  ma DmT0 ; P0


RTo
RT
(2b)

N
X
K1

i1

Vl
va je vi  jd vd
Vd Vl

(4)

ZT
To

DH
dT
RT 2

Zp
po

DV
dp
RT

(6)

where, DH, DV are the changes in molar enthalpy and molar

volume associated with the transition. T0 is the temperature
under the standard condition,  C; p0 is the pressure under the
standard condition, Pa.
mH  ma now can be calculated using the Equations (4)e(6)
above. mH  ma 0 indicates that the hydrate state and the
liquid or solid water states are equally favorable; mH  ma < 0
indicates that hydrate can exist and some of the water will at
equilibrium be in a hydrate form. mH  ma > 0 indicates that no
hydrate can exist and the water will be in the form of either
liquid or ice.
2.3. Prediction method of natural gas hydrate formation area in
wellbore
Coupling optimum temperature and pressure model with
natural gas hydrate equilibrium model, the method based on
OLGA & PVTsim for predicting the natural gas hydrate formation
area in wellbore was obtained. The prediction procedure is

298

Y. Guo et al. / Petroleum 2 (2016) 296e300

shown in Fig. 1, as follows: (1) Drawing the temperature prole

curve in wellbore using the temperature model. (2) Drawing the
pressure prole curve in wellbore using the pressure model. (3)
Drawing the temperature prole curve of natural gas hydrate
equilibrium in wellbore based on the pressure prole and the gas
hydrate equilibrium model. (4) Comparing the temperature
prole curve and the natural gas hydrate equilibrium curve in
wellbore [6]. The natural gas hydrate formation area in wellbore
is that enclosed by the temperature prole curve and the natural
gas hydrate equilibrium curve. The longer the area's vertical
length is, the broader the hydrate formation's scope is. The wider
the area's horizontal width is, the more easily the hydrate forms.
The method can directly and accurately reect the hydrate formation conditions in wellbore.
2.4. Model validation
2.4.1. Temperature and pressure model's validation
Yu et al. (2000) [13] proposed three temperature drop formulas for multiphase pipeline. The effects of gas throttling and
liquid friction are neglected in Method 1. Rather, these effects are
considered in Method 2. Furthermore, given the effects of pipe
undulation, the mass quality is replaced by liquid holdup in
Method 3. Based on the data in Ref. [13], the temperature drop of
a multiphase pipeline was calculated using the above three
methods and the method in this paper. The results are shown in
Table 1. It can be seen from the comparison that the relative error
of method in this paper is minimum. Moreover, the relative error
of predicted pressure drop calculated using method in this paper
was 8.41%. Hence, the model in this paper can be used to predict
the temperature and pressure in wellbore.
2.4.2. Natural gas hydrate equilibrium model's validation
The prediction methods of natural gas hydrate equilibrium
conditions mainly include empirical chart method, Polo Maleev
method, Stergaard method, Hammerschmid method and so on
[14e16]. Based on the experimental data [7,17], Methane hydrate
equilibrium conditions in pure water were calculated using the
above four methods and the method in this paper, as shown in
Fig. 2. The relative error of method in this paper is 0.057%, and that
of empirical chart method, Polo Maleev method, Stergaard method
and Hammerschmid method is 0.314%, 5.027%, 4.516% and 2.978%,
respectively. It can be seen from the comparison that this model
shows a better agreement with the experimental data. Hence, it can
be used to predict the natural gas hydrate equilibrium conditions.

Table 1
Comparison of methods for predicting pipe temperature drop with the measured
values.
Method

Method 1
Method 2
Method 3
This work

Temperature drop
Measured value/ C

Predicted value/ C

Relative error/%

21.7
21.7
21.7
21.7

20.42
21.04
21.36
21.45

5.90
3.04
1.57
1.15

parameters of the well are as follows: the water depth is 1504 m,

the calculated well depth is 4094 m, the bottom hole pressure is
41.8 MPa, the bottom hole temperature is 80.3  C, the sea surface
temperature is 25  C, the subsea temperature is 2.5  C, the formation specic heat capacity is 1900 J/(kg  C), the formation
thermal conductivity is 1.745 W/(m$ C), the wellbore inner
diameter is 0.12 m, the wellbore specic heat capacity is 500 J/
(kg$ C), the wellbore thermal conductivity is 50 W/(m$ C), the
wellbore density is 7850 kg/m3. The temperature eld of the South
China Sea proposed by Gao (2008) [18] is adopted in this case.
3.1. Gas production
Gas production directly affects the distribution of temperature and pressure in the wellbore. The production optimization
during the proration can avoid the natural gas hydrate and
ensure the ow assurance in wellbore. Fig. 3 shows the predicted
gas hydrate formation area under different gas production rates.
The hydrate area gradually narrows with the increasing gas
production. It can be explained from that the increasing ow rate
can decrease the heat exchange time between wellbore uid
medium and ambient medium, then the wellbore uid temperature will increase with gas production increasing at a certain
depth. Moreover, the gas production change has little effect on
the wellbore pressure, so the hydrate formation temperature
curves changed little. In addition, the temperature difference is
small between wellbore and its outside near wellbore bottom,
the wellbore temperature increases slowly induced by gas production. However, the wellbore temperature difference is large
between wellbore and its outside near the mud line, the wellbore
temperature increases largely induced by gas production. In
other words, in order to avoid the natural gas hydrate formation
in the wellbore, the gas production can be increased properly
while satisfying the production requirements. In this calculation
example, the gas production must be higher than 100  104 m3/
d in order to ensure a safe uid ow in wellbore.

3. Application
The prediction method of natural gas hydrate formation area in
wellbore is applied to a gas well in the South China Sea. The basic

Fig. 1. The computational scheme for calculating the hydrate formation area in
wellbore.

Fig. 2. Comparison of methods for predicting hydrate phase equilibrium with the
experimental data.

Y. Guo et al. / Petroleum 2 (2016) 296e300

299

Fig. 3. Predictions of the natural gas hydrate formation area in wellbore under
different gas production rates.

Fig. 5. Predictions of the natural gas hydrate formation area in wellbore under
different insulation materials.

3.2. Hydrate inhibitors

predicted, as shown in Fig. 5. In this case, thermal conductivity of

PU material is 0.12 W/(m$K), and that of PVC material is 0.22 W/
(m$K). The smaller thermal conductivity reduces the heat loss in
wellbore. Therefore, the natural gas hydrate formation area in
wellbore with a PU insulation layer is smaller than that with a
PVC insulation layer. It also can be seen that the hydrate formation area decreases with the increase of the thickness of
insulation materials. Therefore, the method can be applied to
optimize the type and thickness of insulation materials.

The thermodynamic inhibitors are divided into salt inhibitors

(like sodium chloride (NaCl)) and alcohol inhibitors (like methanol (MeOH)) [19]. Exampling by NaCl and MeOH, the natural gas
hydrate formation area in wellbore under different inhibitor
concentrations was predicted, as shown in Fig. 4. The hydrate
formation in wellbore can be more efciently prevented with the
increase of inhibitor concentrations. NaCl or MeOH injected in
the gaseliquid phase can effectively reduce the activity coefcient of water and change the chemical potential of water or
hydrate, then the original thermodynamic equilibrium conditions between the molecules are damaged. Hence, the hydrate
equilibrium curve moves into the area of low temperature or
high pressure. In addition, the inhibitor can destroy the stability
of hydrate by direct contact and prompt hydrate to resolve. It also
can be seen that the inhibitory effect of NaCl is better than that of
MeOH. Therefore, the method can be applied to optimize the
type and concentration of thermodynamic inhibitors.
3.3. Insulation materials
Due to the low temperature and high pressure of deep water,
the pipe thermal insulation has an important effects on the uid
temperature in wellbore. The multi-foam insulation materials
are used as the insulating layer of pipes in petroleum engineering, and the polyurethane (PU), polyvinyl chloride (PVC), epoxy
resins and rubber products are commonly used [20]. Exampling
by the PU and PVC materials, the natural gas hydrate formation
area in wellbore under different insulation materials was

Fig. 4. Predictions of the natural gas hydrate formation area in wellbore under
different inhibitor concentrations.

3.4. Downhole throttling

The downhole throttling method can effectively prevent the
gas hydrate forming for onshore gas wells [21]. Assuming a
throttle is set at 1000 m from the bottom and the effects of
throttling effect on natural gas hydrate formation under different
nozzle orice diameter controlled by the throttle opening are
simulated. Fig. 6 shows the effects of throttle opening on the
natural gas hydrate formation area in wellbore. Both the wellbore temperature and hydrate formation temperature curves
move to the left after the throttle, which indicates that both the
wellbore temperature and pressure are signicantly decreased.
Since the wellbore temperature drop is larger than the hydrate
formation temperature drop, the downhole throttling in deep
water wells are not able to effectively prevent the gas hydrate
forming. There are two possible explanations as follows: (1) The
low temperature and high pressure effect in deep water. Gas
hydrate formation temperature drop under condition of high
pressure drop is less than that under condition of low pressure

Fig. 6. Effects of throttle opening on the natural gas hydrate formation area in
wellbore.

300

Y. Guo et al. / Petroleum 2 (2016) 296e300

drop. The wellbore pressure offshore is higher than that onshore

at the same depth. Hence, hydrate formation temperature curve
decline is small. (b) The gas production effect. The wellbore
temperature after throttling is still higher than the formation
temperature under the higher gas production, so it still loses
heat. The throttling temperature drop could not be restored by
the formation heating.
3.5. Well shut-in
In the gas test or gas reservoir exploitation, the wellbore uid
have complex heat exchange with ambient environment due to
larger temperature difference after well shut-in, so the gas hydrate formation risk in wellbore can be existed. The simulation
conditions are as follows: running 6 h / shut-in 48 h /
restarting and running 6 h. Fig. 7 shows the effects of shut-in
time on the natural gas hydrate formation area in wellbore.
Before shut-in, the wellbore uid velocity is large, so the uid can
have sufcient heat exchange with the environment. The gas
hydrate cannot form due to relatively high temperatures. At the
early period after shut-in, there is a big temperature difference
between the wellbore with its outside since the wellbore uid
gradually stops owing, which induces the efcient heat exchange and the fast wellbore uid temperature drops. The gas
hydrate gradually begins to generate. At the later period after
shut-in, the wellbore temperature decline rate is gradually
slowing due to small temperature difference. The wellbore
temperature approaches the ambient temperature after shut-in
24 h, then the hydrate formation area is maximized. Moreover,
it can easily form near the mud line. The longest hydrate nonformation time after shut-in is dened as the safety shut-in
time or the non-contact time [22], which is an important guarantee for the success of opening wells. Operations in the eld
must be done within the safety shut-in time, otherwise it could
lead to the wellbore blockage. Safety shut-in time is related to
the gas production, inhibitors, piping material thickness and
other factors. The method can provide a basis for the reasonable
design of safety shut-in time.
4. Conclusions
(1) By analysis and evaluation of four prediction methods of
temperature in pipe and ve prediction methods of hydrate formation, coupling the optimum temperature and
pressure model with natural gas hydrate equilibrium
model, a method based on OLGA & PVTsim for predicting

Fig. 7. Effects of shut-in time on the natural gas hydrate formation area in wellbore.

the natural gas hydrate formation area in wellbore was

obtained. It can be applied to the complicated component
and transient working conditions.
(2) The gas production, hydrate inhibitors and insulation
materials have direct impacts on the natural gas hydrate
formation area in wellbore. The more gas production, hydrate inhibitors and thickness of insulation material and
the smaller heat conductivity of insulation material can
increase the risks of hydrate formation.
(3) Downhole throttling will increase the risk of natural gas
hydrate formation under specic conditions in deepwater
gas well. The application of downhole throttle in deepwater gas well should be further discussed.
(4) The uid temperature in wellbore is coming closer to the
ambient temperature after shut-in, which leads to the
increase of natural gas hydrate formation area. This
method can be used to design the safety shut-in time to
save the construction time and economic costs.
References
[1] W.C. Chin, Modern ow assurance methods: clogged pipelines, wax
deposition, hydrate plugs, Offshore 60 (9) (2000) 92e94, 180.
[2] Shaoming Lu, Seismic Characteristic of Two Deep-water Drilling Hazards:
Shallow-water Flow Sands and Gas Hydrate, The University of Texas,
Dallas, 2003.
[3] Baojiang Sun, Shijing Cao, Hao Li, et al., Status and development trend of
deepwater drilling technology and equipment, Pet. Drill. Tech. 39 (2)
(2011) 8e15.
[4] P.M. Dranchuk, R.A. Purvis, D.B. Robinson, Computer calculations of natural
gas compressibility factors using the standing and katz correlation, Inst.
Pet. Tech. Ser. 36 (4) (1974) 76e80.
[5] W.F. Prassl, J.M. Peden, K.W. Wong, et al., Mitigating Gas Hydrate Related
Drilling Risks: A Process-Knowledge Management Approach, 2004. SPE
88529.
[6] Zhiyuan Wang, Baojiang Sun, Haiqing Cheng, et al., Prediction of gas hydrate formation region in the wellbore of deepwater drilling, Pet. Explor.
Dev. 35 (6) (2008) 731e735.
[7] Chenwei Liu, Mingzhong Li, Lei Wang, et al., Prediction of gas hydrate
formation conditions in deepwater oil and gas pipelines and its application,
Nat. Gas. Ind. 30 (10) (2010) 73e77.
[8] J. Rornero, L.E. Touboule, Temperature Prediction for Deepwater Wells: a
Field Validated Methodology, 1998. New Orleans, SPE 49056.
[9] K. Bendiksen, D. Maines, R. Moe, et al., The dynamic two-uid model OLGA:
theory and application, SPE Prod. Eng. 6 (6) (1991) 171e180.
[10] D.D. Erickson, Development of a Natural Gas Hydrate Prediction Computer
Program [M. Sc. thesis], Colorado School of Mines, 1983.
[11] J. Munck, S. Skjold-Jrgensen, P. Rasmussen, Computations of the formation of gas hydrates, Chem. Eng. Sci. 43 (1988) 2661e2672.
[12] J.H. Vander Waals, J.C. Platteeuw, Clathrate solutions, Adv. Chem. Phys. 2
(1) (1959) 1e57.
[13] Xichong Yu, Shuchu Feng, Yuxing Li, The deduction of temperature drop
formula to calculate multiphase pipeline, OGST 19 (4) (2000) 22e25.
[14] A. Konstantin, K.A. Uchadin, J.A. Ripmeester, A complex clathrate hydrate structure showing bimodal guest hydration, Nature 397 (1999)
420e423.
[15] P. Englezos, Z. Huang, P.R. Bishnoi, Prediction of natural gas hydrate formation conditions in the presence of methanol using the Trebble-Bishnoi
equation of state, J. Can. Pet. Tech. 30 (1991) 148e155.
[16] A. Majumadar, E. Mahmoodaghdam, P.R. Boshinoi, Equilibrium hydrate
formation conditions for hydrogen, sulde, carbon dioxide, and ethane in
aqueous solutions of ethylene glycol and sodium chloride, J. Chem. Eng. 45
(2000) 20e22.
[17] Donghai Mei, Jian Liao, Lukun Wang, Jitao Yang, Tianmin Guo, Measurement and prediction of the equilibrium formation conditions for gas hydrates, J. Chem. Eng. Chin. Univ. 11 (3) (1997) 225e230.
[18] Yonghai Gao, Baojiang Sun, Zhiyuan Wang, Shijing Cao, Linsong Song,
Haiqing Cheng, Calculation and analysis of wellbore temperature eld in
deepwater drilling, J. China Univ. Pet. 32 (2) (2008) 58e62.
[19] Xiaodong Bai, Study on the Hydrate Inhibitor in Deepwayer Mud, Southwest Petroleum Institute, Chengdu, 2005.
[20] Yinghui Tian, Theoretical Analysis and Experimental Study of Thermally
Insulated Subsea Pipeline, Tianjin University, Tianjin, 2007.
[21] Yu Wang, Yingchuan Li, Chaoyi She, Prediction of downhole bean performance of gas well, Nat. Gas Ind. 26 (2) (2006) 117e119.
[22] J.P. Dejean, D. Averbuch, M. Gainville, et al., Integrating Flow Assurance
into Risk Management of Deep Offshore Field, OTC, Houston, TX, U.S.A,
2005.