close

Вход

Забыли?

вход по аккаунту

?

j.jngse.2018.08.010

код для вставкиСкачать
Accepted Manuscript
A Transient Simulation Model to Predict Hydrate Formation Rate in both Oil- and
Water-Dominated Systems in Pipelines
Yan Wang, Carolyn A. Koh, J. Alejandro Dapena, Luis E. Zerpa
PII:
S1875-5100(18)30354-8
DOI:
10.1016/j.jngse.2018.08.010
Reference:
JNGSE 2682
To appear in:
Journal of Natural Gas Science and Engineering
Received Date: 25 May 2018
Revised Date:
11 July 2018
Accepted Date: 16 August 2018
Please cite this article as: Wang, Y., Koh, C.A., Dapena, J.A., Zerpa, L.E., A Transient Simulation Model
to Predict Hydrate Formation Rate in both Oil- and Water-Dominated Systems in Pipelines, Journal of
Natural Gas Science & Engineering (2018), doi: 10.1016/j.jngse.2018.08.010.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Pipeline with irregular geometry
RI
PT
Oil-dominated Section
TE
D
M
AN
U
SC
Water-dominated
Section
Gas Hydrate Forms
at Interfaces
Gas
Entrainment
AC
C
EP
Liquid
Entrainment
Oil
Gas
Water
Water-dominated
Section
ACCEPTED MANUSCRIPT
2
A Transient Simulation Model to Predict Hydrate Formation Rate
in both Oil- and Water-Dominated Systems in Pipelines
3
Yan Wang a, Carolyn A. Koh a, J. Alejandro Dapena a, Luis E. Zerpa a, b,*
1
4
5
a
6
7
b,*
8
Keywords: Flow assurance; gas hydrates; multiphase flow; high water cut; modeling
9
Abstract
RI
PT
Center for Hydrate Research, Department of Chemical & Biological Engineering, Colorado
School of Mines, 1600 Illinois St., Golden, CO 80401, USA
SC
(Corresponding author) Department of Petroleum Engineering, Colorado School of Mines,
1600 Arapahoe St., Golden, CO 80401, USA. Email: lzerpa@mines.edu
The high pressure and low temperature operational conditions of deepwater subsea facilities
often result in the formation of gas hydrates, which is one of the most challenging flow assurance
issues. The variant water cuts and flowline geometries could result in complicated flow patterns
leading to different hydrate formation mechanisms. In previous work, we have developed
hydrate formation models both for oil- and water-dominated systems. The model for oildominated systems considers the water phase being dispersed as droplets in an oil continuous
layer, and uses either a kinetics or a transport model to calculate hydrate formation at the
interface of water droplets. On the other hand, water-dominated systems contain small amounts
of oil, and a mass transfer-based hydrate growth model is used. This paper presents a new
transient hydrate formation model that predicts hydrate growth and transportability in oil- and
water-dominated environments, which might be present simultaneously in oil and gas pipelines
due to changes in fluid distribution in complex multiphase flow systems. Simulations of highpressure pilot-scale flowloop experiments at different water cuts are used to verify this modeling
approach. Furthermore, the proposed hydrate simulation tool is applied to analyze the flow
dynamics of a subsea tieback involving hydrate formation at varying water cuts after flowloop
verification. This simulation model is targeted to be widely used to simulate oil-dominated
pipelines, as well as high water cut systems leading to phase inversions, and represents a step
further towards the development of more comprehensive hydrate formation models.
28
1 Introduction
29
30
31
32
33
34
35
36
37
38
Gas hydrates are ice-like crystalline compounds where small gas molecules are surrounded by
hydrogen-bonded water cages, which usually form under high pressure and low temperature
conditions typically encountered in oil and gas offshore pipelines (Sloan and Koh, 2007). In deep
to ultra-deep waters, if subsea tieback facilities exchange sufficient heat to the ocean
environment, the produced reservoir fluids might cool down while the pressure remains high,
making gas hydrates the thermodynamically preferred phase for natural gas and water. Once gas
hydrates form, the hydrate particles may cohere upon contact and form large aggregates,
increasing the slurry viscosity. Hydrate plugs can form from bulk hydrate aggregates and large
hydrate slurry viscosity increase, and may cause flowline blockage, posing severe safety,
economic and environmental concerns (Koh et al., 2011; Liu et al., 2016). Although “hydrate
AC
C
EP
TE
D
M
AN
U
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
1
ACCEPTED MANUSCRIPT
avoidance” methods, such as pipeline insulation, electrical heating and injection of
thermodynamic hydrate inhibitors (THI), as well as “hydrate management” methods, which
include injection of kinetic hydrate inhibitors (KHIs) and anti-agglomerants (AAs), have been
developed to limit the severity of hydrate plugs, transient modelling tools that predict hydrate
formation are highly desirable for the design of safe production operations considering the
complicated multiphase flow scenarios in the oil/gas field (Sloan, 2005; Zerpa et al., 2012a).
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
In oil/gas production, the liquid holdup, void fraction and water cut could vary dramatically
between different fields. Accordingly, the oil and gas transportation in pipelines could be
classified as oil-dominated, water-dominated and gas-dominated systems (Zerpa et al., 2012a).
High gas to liquid ratios might result in gas-dominated systems (the discussion of gas-dominated
systems is out of the scope of this paper, and it is not considered in the proposed hydrate
prediction model detailed here). On the other hand, low gas to liquid ratios could result in oildominated systems, in which all the liquid water phase is dispersed in an oil continuous layer,
forming water-in-oil (W/O) emulsions. However, at high water cuts, only part of the water
might be dispersed in the oil phase, leading to a free-water layer during flow and the formation
of oil-in-water (O/W) emulsions, which characterize water-dominated systems (Zerpa et al.,
2012b). In addition, there might be phase inversions from oil- to water-dominated systems and
vice versa in subsea pipelines, especially in the presence of low spots where water could
accumulate. The investigation of emulsion phase inversions is of great significance since they
not only affect the hydrate dispersion properties, but also might change the fluid slurry viscosity,
influencing the production pressure drop (Høiland et al., 2005; Salager and Forgiarini, 2012).
Numerous studies have been conducted to investigate the emulsion phase inversion in the
presence of hydrates. Høiland et al. (2005) evaluated the phase inversion transitions of 12 crude
oil-brine systems containing Freon hydrates to investigate the effect of these solids on the
dispersion properties. The results showed that phase inversions were highly dependent on both
the crude oil properties and the brine volume fractions; moreover, the presence of hydrate
particles might either favor or hinder such phase inversions. Moradpour et al. (2011) indicated
that phase inversions in systems involving hydrates could change the slurry viscosity
significantly. Dapena et al. (2017) inferred emulsion phase inversions from oil-continuous to
water-continuous systems upon hydrate formation during experiments using a high-pressure
pilot-scale flowloop based on changes in the measured pressure drop behavior.
32
33
34
35
36
37
38
39
40
41
42
However, to date no simulation tool has been reported that captures the emulsion phase
inversions related to hydrate formation in oil/gas systems. Independent hydrate formation models
for both oil- and water-dominated systems have been developed at the Center for Hydrate
Research in the Colorado School of Mines (Davies et al., 2009; Joshi et al., 2013; Turner et al.,
2009). In oil-dominated systems, the hydrate formation rate is calculated by either a kinetics or a
transport model, while the increase in the slurry viscosity is a function of the calculated hydrate
agglomerate size and corresponding effective volume fraction. In water-dominated systems, a
mass transfer-based hydrate growth model is used, with a hydrate plugging criterion based on
fluid velocity, hydrate volume fraction and liquid holdup (Zerpa et al., 2012b). Further studies on
hydrate formation indicate that the growth rate, viscosity increase and plugging mechanisms in
these two systems are different (Zerpa et al., 2012a). In oil/gas production, the water content
AC
C
EP
TE
D
M
AN
U
SC
RI
PT
1
2
3
4
5
6
2
ACCEPTED MANUSCRIPT
increases as the field matures, leading to conditions where a transition from oil- to watercontinuous systems could occur. Considering the complex multiphase flow scenarios and the
dynamic water cut that can occur in oil and gas pipelines, a new transient hydrate simulation tool
has been developed as a plug-in module coupled with a commercial multiphase flow simulator.
This transient hydrate simulation tool has the ability to predict the hydrate formation in pipelines
exhibiting both water- and oil-dominated environments in different sections of the pipe, as well
as modeling the transition from one system to another in a single simulation run. This paper first
describes the models and assumptions of this comprehensive hydrate simulation tool. Then, the
hydrate formation models for oil- and water-dominated systems, as well as systems with phase
inversion are validated against high-pressure pilot-scale flowloop experimental results. Finally,
this transient hydrate simulation tool is applied to a subsea tieback to analyze the effect of water
cut and geometry on flow dynamics.
13
2 Model Development
14
15
2.1 Integration of the transient hydrate simulation models into a commercial multiphase
flow simulator
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
This transient hydrate simulation tool is based on the algorithm of a one-dimensional commercial
multiphase flow simulator (Bendiksen et al., 1991). This commercial flow simulator spatially
discretizes the pipeline into multiple control volumes, and within each control volume, it will
calculate and update the following: fluid properties, phase masses, phase equilibria, momentum
balance, velocity, volume and pressure, along with energy and temperature. After initialization,
an iterative algorithm is executed for every control volume at each time step. The hydrate
simulation tool includes four parts: a user defined PVT (UDPVT) module, a flash module, a
rheology module for oil-dominated systems and a PVT property update module. In the UDPVT
module, both the hydrate equilibrium curves and the hydrate former guest concentration tables
are checked in preparation for hydrate growth calculations. In the flash module, the hydrate
formation onset is calculated based on the hydrate equilibrium curves and the subcooling; the
hydrate growth or dissociation rates are computed using either the hydrate formation or
dissociation models; and masses for each phase are updated accordingly. In the rheology module
for oil-dominated systems, the dynamic hydrate agglomerate size and hydrate slurry viscosity are
calculated. For water-dominated systems, measured cohesive forces between hydrate particles
are significantly lower than those observed in oil-dominated systems, and the hydrate
agglomeration mechanisms with water as the continuous phase are still not fully understood,
hence an agglomeration model for water-dominated systems is not proposed (Aman et al., 2012;
Joshi et al., 2013). This paper focuses on the hydrate formation models, and the description of
hydrate rheology and dissociation models are not presented here, but references are provided
(Zerpa et al., 2012b). Finally, the hydrate PVT properties, including density, heat capacity and
thermal conductivity, are updated. An overview of this transient hydrate simulation tool and the
coupling with a commercial multiphase flow simulator is shown in Figure 1 (Davies et al., 2010).
Figure 2 (a) and (b) illustrate the phase distributions in the pipe cross sections for oil- and waterdominated systems, respectively. This multiphase flow simulator predefines water droplets and
gas bubbles dispersed in the oil layer, or oil droplets and gas bubbles dispersed in the water layer.
AC
C
EP
TE
D
M
AN
U
SC
RI
PT
1
2
3
4
5
6
7
8
9
10
11
12
3
ACCEPTED MANUSCRIPT
The transient hydrate simulation tool defines the hydrate particles dispersed in either the oil or
the water layer for oil- and water-dominated systems, respectively. The hydrate formation in the
gas phase is not considered in this hydrate simulation tool.
M
AN
U
SC
RI
PT
1
2
3
4
Figure 1: Integration of the transient hydrate simulation models into a commercial multiphase
flow simulator.
8
9
10
(a)
(b)
Figure 2: Illustration of phase distributions in (a) pipe cross section with oil-dominated systems,
and (b) pipe cross section with water-dominated systems.
AC
C
7
EP
TE
D
5
6
11
2.2 Description of the hydrate prediction model
12
2.2.1 Oil-dominated systems
13
14
15
16
17
This transient hydrate prediction tool includes both oil- and water-dominated models. As shown
in Figure 3, hydrate formation is divided into four main stages in oil-dominated systems: (i)
water entrainment considering the emulsification of water droplets in a continuous oil phase
(W/O emulsion); (ii) hydrate growth at the interface of water droplets when a certain subcooling
has been reached; (iii) hydrate agglomeration caused by the interaction between particles in the
4
ACCEPTED MANUSCRIPT
form of cohesive forces; and (iv) plugging due to an increase in hydrate slurry viscosity that
could lead to pipeline blockage (Zerpa et al., 2012b).
RI
PT
1
2
SC
3
Figure 3: Conceptual picture of hydrate formation and plugging in oil-dominated systems,
adapted from Turner (2005), with input from J. Abrahamson (U. Canterbury, Christchurch, NZ).
6
7
8
9
10
The size of the dispersed water droplets determines the interfacial area available for hydrate
formation, as well as the emulsion rheology and stability. The model developed by Boxall et al.
(2012) is used to estimate the water droplet size based on the oil phase density, viscosity,
velocity, and water-oil interfacial tension. In this model, the flow is categorized into two
subranges: inertial and viscous. The mean droplet size in the inertial subrange follows Eq. 1:
M
AN
U
4
5
̅ 0.063 /
The equation for the mean droplet size in the viscous subrange follows Eq. 2:
̅ 0.016 / TE
D
11
Eq. 1
Eq. 2
2
where We is the Weber number (We = ρoU Dp/σ), and Re is the Reynolds number (Re =
ρoUDp/µo). In Eqs. 1 and 2, the prefactors 0.063 and 0.016 both come from experimental
regression through analyzing the water droplet size dispersed in different types of oils (Boxall et
al., 2012).
16
17
18
19
The transition from the inertial to the viscous subrange occurs when We > 0.0674Re5/4. Upon
hydrate formation, the hydrate particle size is assumed to be the same as the water droplet size.
After reaching a certain subcooling (∆Tsub = Thyd_eq -Tsys), the hydrate onset takes place, and the
intrinsic kinetics hydrate growth rate is calculated using a first order equation according to Eq. 3:
20
21
22
Eq. 3
∆"#$
In Eq. 3, the gas consumption rate is a function of the intrinsic kinetics rate constants (k1 and k2),
which were regressed from data presented by Vysniauskas and Bishnoi (1983), the surface area
between water droplets and the oil phase (As), and the subcooling (∆Tsub) (Boxall et al., 2009).
23
24
25
As an alternative to the first order hydrate formation kinetics equation (Eq. 3), a shrinking core
model, which considers heat and mass transfer limitations, can be used for hydrate growth rate
calculations. After formation of hydrate shells on the surface of water droplets, further hydrate
AC
C
EP
12
13
14
15
5
ACCEPTED MANUSCRIPT
growth occurs due to gas diffusion through the hydrate shells and upon contact with the water
core. The gas consumption rate considering both the external and internal mass and heat transfer
limitations is described in Eq. 4 and Eq. 5 (Davies et al., 2010; Zerpa et al., 2012b). Finally, the
hydrate growth rate would be limited by the slowest gas diffusion or heat transfer through either
the oil layer surrounding the hydrate particle (external resistance), or through the hydrate shell
(internal resistance).
RI
PT
1
2
3
4
5
6
(&$#'
&() )
- ∙ /
1
+ ∙/
+
0 1
Eq. 4
2.2.2 Water-dominated Systems
M
AN
U
7
8
Eq. 5
SC
>_345 ∙ ?
2"345_() "4 7
(
)
- ∙ /
>1@(A_345 ∙ ?
1@(A + >_345 ∙ ?
1
8 +
< (∆= )
ℎ :;+ ∙ /1
The amount of produced water is expected to increase during oil and gas field maturation.
Eventually, at high water cuts, the excess water forms a free water phase, dispersing oil and gas
into this free water layer by shear forces, known as a water-dominated system. However, the
hydrate formation model for water-dominated systems (Figure 4) considers a less complicated
phase distribution by assuming only the presence of an aqueous and a hydrocarbon gaseous
phase. The contribution from the dispersed oil phase to hydrate formation is neglected. This
model splits the hydrate formation process into the following three steps (Joshi et al., 2013;
Zerpa, 2013):
17
18
(1) Gas entrainment: Flow shear disperses gas bubbles in the continuous water layer and
multiphase flow correlations are used to estimate the surface area of water and hydrocarbon.
19
20
21
22
(2) Hydrate growth: After reaching a specified subcooling, hydrate onset is assumed to occur
immediately. At this point, the interface of the hydrocarbon gas bubbles contacting surrounding
water forms a hydrate shell. The mass transfer model (Eq. 6) developed by Skovborg and
Rasmussen (1994) is used to calculate the formation rate:
dm gas
23
24
25
AC
C
EP
TE
D
9
10
11
12
13
14
15
16
Eq. 6
= − k mass As _ w ( C bulk − C eq )
dt
(3) Plugging: The significant aggregation of hydrate particles can be recognized by a large
increase in the slurry viscosity and sizeable fluctuations in pressure drop. The plugging criterion
can be estimated as a function of hydrate volume fraction, fluid velocity, and liquid holdup.
26
27
6
ACCEPTED MANUSCRIPT
RI
PT
1
2
3
Figure 4: Conceptual picture for hydrate formation and plugging in water-dominated systems
consisting of gas and water phases (Zerpa et al., 2013).
4
2.2.3 Phase inversion
The developed tool accounts for dispersion phase inversions by coupling the models for oil- and
water-dominated systems, and by defining a phase inversion region. Figure 5 shows a general bidimensional map considering the emulsion formulation variables versus the water/oil
composition. Figure 5 generally divides the water-oil mixture as oil- or water-continuous
systems based on the oil/water ratio. As indicated in Figure 5, if the water content in the wateroil mixture increases and passes the inversion region, the system will invert from oil- to watercontinuous. During inversion, the W/O emulsion destabilize and break, forming an O/W
emulsion. On the contrary, a water content decrease may generate the transition from water- to
oil-continuous system when passing the inversion region (Høiland et al., 2005; Salager and
Forgiarini, 2012). The water-oil emulsion inversion not only depends on the compositional
variable change (i.e., oil/water ratio), but also depends on the change of formulation variables
(e.g., salinity, temperature, surfactant structure and type) often expressed as hydrophiliclipophilic deviation (HLD), associated to the relative affinity of emulsifying agents to oil and
water (Salager and Forgiarini, 2012).
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Similar to the emulsion inversion map in Figure 5, if the water cut is below the inversion range,
the oil-dominated model is applied for hydrate predictions assuming the hydrate particles are
dispersed in the continuous oil layer. If the water cut increases to above the inversion range, the
system inverts from an oil-continuous to a water-continuous dispersion, with the oil droplets and
the hydrate particles dispersed in the water layer. In this case, the water-dominated model is
applied for hydrate formation calculations. Partially dispersed systems, where the water is
present both as a free water phase and partially dispersed as W/O emulsions in the oil phase, are
not considered in the current model but will be studied further in the future (Vijayamohan et al.,
2014). This new transient hydrate simulation tool is able to capture the catastrophic phase
inversion which is caused by a compositional variable change, but does not capture the
translational phase inversion which is caused by the change of formulation variables (GalindoAlvarez et al., 2011). Since the inversion region differs with respect to oil types and
compositions, hydrate formation, AA and/or salt addition, the inversion point is a user input in
the simulations in terms of the water cut (Moradpour et al., 2011).
AC
C
EP
TE
D
M
AN
U
SC
5
6
7
8
9
10
11
12
13
14
15
16
17
18
7
SC
RI
PT
ACCEPTED MANUSCRIPT
1
M
AN
U
The hydrate formation model is validated against experimental results obtained using a highpressure pilot-scale flowloop. The flowloop testing section has an inner diameter of 9.7 cm and
is 96 m long. A sliding-vane pump with an operating speed from 300 - 1500 rpm, which
corresponds to a mixture velocity of 0.75 - 3.75 m/s, is installed in order to drive the fluids
through the flowloop. The entire system, with an exception of a piston-driven high-pressure gas
accumulator, is located in a temperature-controlled test chamber. The gas accumulator allows
maintaining a constant pressure in the system as gas is consumed due to hydrate formation. The
temperature control within the chamber relies on a forced air-convection system. A Coriolis
multiphase flowmeter provides measurements of the flowrate throughout the tests. The pressure
drop across the pump is recorded. Multiple pressure and temperature sensors are installed along
the flowloop (Joshi et al., 2013). Figure 6 presents a schematic diagram of the experimental setup.
In all the simulations, the liquid holdup is defined as the sum of water, oil, and hydrate volume
fractions in the pipe. The water cut is defined as the water volume fraction divided by the liquid
holdup.
AC
C
4
5
6
7
8
9
10
11
12
13
14
15
16
17
3 Model Validation
TE
D
3
Figure 5: Bi-dimensional map of emulsion properties versus water/oil composition.
EP
2
18
8
ACCEPTED MANUSCRIPT
Figure 6: Schematic diagram of the high-pressure pilot-scale flowloop (Joshi et al., 2013).
2
3
4
5
6
7
8
9
10
11
12
13
14
15
The hydrate formation model has been shown to successfully predict the hydrate formation rate
in the pilot-scale flowloop for both oil- and water-dominated systems, as well as for systems with
phase inversion. Figure 7 shows the simulation results in oil-dominated systems corresponding to
a flowloop test conducted at a constant pressure (i.e., 69 bar) using 2014 Conroe crude oil, 60
vol.% initial liquid holdup, 30% water cut, and pure methane as the gas phase. The pump speed
was set to 1200 rpm, corresponding to a mixture velocity of 3 m/s (Srivastava et al., 2017). The
aqueous phase contains 5.0 wt.% Instant Ocean sea salt (Grasso et al., 2014). From an initial
temperature of 29 ºC, the flowloop is cool down to a set point temperature of 4 ºC. In Figure 7(a),
hydrate starts to form at a time of about 1.8 hrs, and exhibits a very fast hydrate formation rate
during the initial 5 hrs. The periodic hydrate dissociation and formation periods at 9 hrs, 17.5 hrs
and 25 hrs are caused by small external ambient temperature fluctuations (note that ambient
temperature is an input to the model). From Figure 7(b), the liquid holdup goes from 60 to 65
vol.% due to the formation of hydrates in the bulk liquid phase. The water is not fully converted
into hydrate due to salt inhibition and mass transfer limitations.
TE
D
M
AN
U
SC
RI
PT
1
AC
C
EP
16
17
18
(a)
(b)
9
ACCEPTED MANUSCRIPT
Figure 7: (a) Experimental (Srivastava et al., 2017) and simulated hydrate volume fractions from
flowloop tests in an oil-dominated system. (b) Calculated phase volume fractions using the oildominated hydrate formation model.
4
5
6
7
8
9
10
The water-dominated hydrate formation model is validated against a flowloop experiment with
an initial liquid holdup of 50 vol.%, 100% water cut, and methane as the gas phase (Joshi et al.,
2013). The pressure of the flowloop was set constant at 69 bar with a mixture velocity of 2.5 m/s.
Figure 8 (a) and (b) present the flowloop temperature and the hydrate volume fraction in the
experiment and in the simulation, respectively. From Figure 8, it is obvious that with a
subcooling of 2.2 oC, this simulation tool not only can predict hydrate formation rate and amount,
but also can successfully capture the exothermic temperature increase during hydrate formation.
11
12
(a)
TE
D
M
AN
U
SC
RI
PT
1
2
3
(b)
Figure 8: (a) Flowloop temperature comparison between simulation and experiment (Joshi et al.,
2013) of a water-dominated system. (b) Experimental and simulated hydrate volume fraction in
the water-dominated system.
16
17
18
19
20
21
22
23
24
25
26
27
28
Dapena et al. (2017) reported experiments conducted using the same high-pressure pilot-scale
flowloop, with mineral oil as the oil phase and an aqueous phase with 3.5 wt.% NaCl salinity.
The gas phase was a natural gas mixture. A case with an initial liquid holdup of 70 vol.% with 50%
water cut is considered for the validation of the proposed hydrate formation model. In this
experiment, a sudden increase in pressure drop appeared after about 3 vol.% hydrate formation in
a system dosed with 2 vol.% AA. Further water/oil dispersion and rheology tests suggested that
~ 50% water cut is close to the inversion point of the system. The test started with a watercontinuous system and inverted to an oil-continuous system at 3 vol.% of hydrates. Simulations
were run with this new hydrate simulation tool, and the results are shown in Figure 9. From
Figure 9, the inversion point from a water- to an oil-continuous system is observed at 3 vol.% of
hydrates in the simulation, in agreement with the experimental observations. The hydrate
formation rate in the simulation is greater than that in the experiment. This is because this
specific AA showed strong kinetic inhibition effects which delayed the hydrate formation onset
AC
C
EP
13
14
15
10
ACCEPTED MANUSCRIPT
and slowed down the hydrate crystal growth rate, and this is not accounted for in the hydrate
formation model.
M
AN
U
SC
RI
PT
1
2
3
Figure 9: Flowloop simulation of hydrate volume fraction as a function of time in a system
showing a phase inversion from a water- to an oil-dominated system and the comparison to the
experiment. The vertical solid black line indicates the hydrate distribution change from model
prediction; and the vertical dashed green line represents the indicated phase inversion point in the
experiment (Dapena et al., 2017).
9
4 Subsea Tieback Hydrate Simulations
TE
D
4
5
6
7
8
This transient hydrate simulation tool was utilized to simulate an uninsulated subsea flowline and
riser with an increasing water cut from 55% to 75%. The input dispersion phase inversion region
from oil- to water-continuous systems for these fluids is about 70 ± 3% water cut. If the water
cut increases to 73% and above, the water-dominated hydrate formation model is employed. On
the other hand, if the water cut decreases to 67% and below, the oil-dominated hydrate growth
model is used. In order to investigate the influence of water cut and flowline geometry, a
constant well production rate of 7858 kg/h and a constant GOR of 298 Sm3/Sm3 are assumed for
the simulations.
18
19
20
21
22
23
24
25
26
Figure 10 shows the subsea tieback geometry in terms of the pipeline depth from sea level and
the pipeline horizontal length. The wellhead is connected to a flowline which has a rough terrain
slope. The horizontal length of flowline and riser adds up to 23.7 km, and the inner diameter of
the flowline and riser is 12 cm. Figure 11 presents a pressure-temperature diagram showing the
pipeline operating conditions along with the hydrate equilibrium curve. The potential hydrate
formation region lies to the left of the hydrate equilibrium curve. The warm reservoir fluids
flowing from the wellhead would be quickly cooled down to seafloor ambient temperature (4 °C),
and the fluid temperature enters the hydrate formation region at a distance of 0.2 km from the
wellhead. Most of the flowline and riser is located within the hydrate formation region.
AC
C
EP
10
11
12
13
14
15
16
17
11
SC
RI
PT
ACCEPTED MANUSCRIPT
Figure 10: Geometry of the subsea tieback in terms of the pipeline depth from sea level.
3
EP
TE
D
2
M
AN
U
1
Figure 11: Pressure-temperature diagram with the hydrate equilibrium curves and the
flowline/riser operation conditions.
6
7
8
9
10
11
12
13
14
Figure 12 presents the fluid distribution for the case with a wellhead liquid holdup of 25 vol.%
and a water cut of 55%. The results after 3 hrs of simulation time are shown in Figure 12(a). The
fast hydrate growth at a pipeline length of 1 - 5 km is due to low mass transfer limitations, since
a large quantity of water and gas come from the wellhead. The uneven liquid distribution is
mainly caused by the geometry. Compared with Figure 10, when the fluids are transported
downhill at a pipeline length of 0 - 1 km and 11.3 - 15.6 km, the gravity-dominated flow leads to
a higher liquid velocity, and a lower gas velocity, generating a higher void fraction, a lower
liquid holdup, and a shorter residence time for hydrate formation, thus resulting in relatively
lower hydrate formation rates. Conversely, when the liquid is transported uphill, the flow is
AC
C
4
5
12
ACCEPTED MANUSCRIPT
RI
PT
friction-dominated with a lower liquid velocity and a higher gas velocity, leading to a lower void
fraction, a higher liquid holdup, a longer hydrate formation residence time, and a greater hydrate
formation amount. From this point, uphill of the geometry after “low spots” is where the liquids
accumulate, as well as the most likely places where phase inversions from an oil- to a waterdominated system could happen. From Figure 12(a), at a pipeline length of 16.8 - 18.1 km, the
water cut is more than 80% and a catastrophic phase inversion takes place, leading to a waterdominated system in this section. Longer simulations show that the hydrates forming in waterdominated systems are transported and melted up in the riser, and the system will reach steady
state, as presented in Figure 12(b). At steady state, the whole flowline and riser are oil-dominated,
and the highest hydrate accumulation (more than 40 vol.%) is found near the base of the riser.
Although there is still unconverted water and a limited amount of gas inside the pipeline, no
more hydrates form due to strong mass transfer limitations.
13
14
(a)
TE
D
M
AN
U
SC
1
2
3
4
5
6
7
8
9
10
11
12
(b)
Figure 12: Fluid distribution at a wellhead liquid holdup of 25 vol.% and a water cut of 55%: (a)
after 3 hrs of simulation, (b) at steady state.
17
18
19
20
21
22
23
24
25
Figure 13 shows the fluid distribution for a case with a wellhead liquid holdup of 25 vol.% and a
water cut of 65%. Figure 13(a) displays the results after 3 hrs of simulation. A dispersion phase
inversion is predicted for a longer pipeline length compared with 55% water cut system. From
Figure 13(a), at pipeline lengths of 7.8 - 11.3 km and 16.4 - 23.7 km, the system is waterdominated. The steady-state simulation results are shown in Figure 13(b). All the gas at a
pipeline length from 15.8 - 23.7 km is consumed due to hydrate formation, leading to 100 vol.%
liquid holdup. Since the mass production rate and GOR are constant, with an increased water cut
from 55% to 65%, the oil and gas production amount decreases, leading to less hydrate
formation compared to the 55% water cut system.
AC
C
EP
15
16
13
SC
RI
PT
ACCEPTED MANUSCRIPT
1
2
(a)
(b)
Figure 13: Fluid distribution at a wellhead liquid holdup of 25 vol.% and a water cut of 65%: (a)
after 3 hrs of simulation, (b) at steady state.
5
6
7
8
9
10
11
12
13
14
Figure 14 shows the fluid distribution for a case with a wellhead liquid holdup of 25 vol.% and a
water cut of 75%. Figure 14(a) shows the results after 3 hrs of simulation time. Hydrate
dispersion in a water-dominated system is observed for most of the pipeline due to the high water
cut. Hydrate dispersion in oil is only observed downhill of the geometry at pipeline lengths of 0 1 km and 11.8 - 15.8 km, when the liquid holdup and water cut are relatively low. Figure 14(b)
shows a plot of the steady-state simulation results. A dispersion phase inversion is found at a
pipeline length of 5 km. The high wellhead water cut results in hydrate dispersion in the water
layer at 0 - 5 km. The hydrate formation lowers the water cut and at a pipeline length of 5 km,
the water cut decreases below 67% and a catastrophic phase inversion from a water- to an oildominated system takes place.
AC
C
EP
TE
D
M
AN
U
3
4
15
16
(a)
(b)
14
ACCEPTED MANUSCRIPT
1
2
Figure 14: Fluid distribution at a wellhead liquid holdup of 25 vol.% and a water cut of 75%: (a)
after 3 hrs of simulation, (b) at steady state.
3
5 Conclusions
A new transient simulation tool that predicts the hydrate formation rate from both oil- and waterdominated systems has been developed. In oil-dominated systems, the hydrate formation rate is
calculated as a result of intrinsic kinetics as well as mass (gas diffusion) and heat transfer
resistances. In water-dominated systems, a mass transfer-based hydrate growth model is used.
The developed tool can estimate the hydrate formation in pipelines exhibiting both water- and
oil-dominated environments at different locations, as well as model the transition from one
system to another in a single simulation run. This hydrate prediction model is coupled with a
state-of-art transient multiphase flow simulator, and can provide insights for the design and
optimization of subsea transport facilities, as well as the prevention, management and
remediation of hydrates in pipelines.
14
15
16
17
18
19
20
21
This hydrate simulation tool has been validated against high-pressure pilot-scale flowloop tests
with both oil- and water-dominated systems, as well as systems with a phase inversion,
demonstrating the ability to successfully predict hydrate formation in different systems. This tool
has been also applied to estimate the hydrate volume fraction of a subsea tieback with a rough
terrain geometry at different water cuts. It is shown that the void fraction, water cut and
geometry might play a major role in the hydrate formation kinetics. Relatively more hydrates are
formed uphill after the low spots in the geometry than downhill due to water accumulation in the
subsea tieback geometry and longer residence time for liquid phases.
22
Nomenclature
23
As = surface area between water and oil phase per unit pipeline length (m2/m)
As_w = surface area between water and gas phase per unit pipeline length (m2/m)
Cbulk = hydrate guest concentration in the bulk phase without hydrates (kg/m3)
Ceq = hydrate guest concentration in water phase in the presence of hydrate (kg/m3)
d = mean droplet diameter (m)
26
27
SC
M
AN
U
TE
D
25
EP
24
RI
PT
4
5
6
7
8
9
10
11
12
13
37
MWgas = molecular weight of gas phase (g/mol)
30
31
32
33
34
AC
C
35
36
DA = gas diffusivity through the hydrate shell (m2/s)
Dp = pipe internal diameter (m)
dP = hydrate particle diameter (m)
k1 = intrinsic kinetics rate constant 1 (kg/m2/s/K)
k2 = intrinsic kinetics rate constant 2 (K)
kcomp = heat transfer coefficient of the hydrate shell (J/s/m/K)
kmass = mass transfer coefficient (m/s)
h = convective heat transfer coefficient (J/m2/K/s)
mgas = mass of gas (kg)
28
29
15
ACCEPTED MANUSCRIPT
MWwater = molecular weight of aqueous phase (g/mol)
2
Re = Reynolds number (Re = ρoUDp/µo)
3
rp = hydrate particle radius (m)
rw = water core radius (m)
t = time (s)
Thyd_eq = hydrate equilibrium temperature (K)
Tsys = system temperature (K)
U = fluid velocity (m/s)
We = Weber number (We = ρoU2Dp/σ)
Xgas_hyd = molar fraction of hydrate guest in the hydrate phase
Xwater_hyd = molar fraction of water concentration in the hydrate phase
∆H = enthalpy change during hydrate formation (kJ/mol)
∆Tsub = hydrate formation subcooling (K)
δ = hydrate shell thickness (m)
µo = oil viscosity (cp)
ρo = density of the oil phase (kg/m3)
σ = oil-water interfacial tension (mN/m)
6
7
8
9
10
11
12
13
14
15
16
17
18
SC
5
M
AN
U
4
RI
PT
1
Acknowledgements
20
21
22
23
The authors would like to greatly acknowledge the support from CSM Hydrate Consortium,
which currently include Chevron, Multi-Chem Halliburton, Petrobras, Schlumberger, Statoil,
Total and ExxonMobil. Thanks to ExxonMobil for allowing us to run the flowloop tests. Thanks
to Marshall Pickarts for proofreading.
24
References
25
26
27
Aman, Z.M., Joshi, S.E., Sloan, E.D., Sum, A.K., Koh, C.A., 2012. Micromechanical cohesion
force measurements to determine cyclopentane hydrate interfacial properties. J. Colloid
Interface Sci. 376, 283–288. doi:10.1016/j.jcis.2012.03.019
28
29
Bendiksen, K.H., Maines, D., Moe, R., Nuland, S., 1991. The Dynamic Two-Fluid Model OLGA:
Theory and Application. SPE Prod. Eng. 6, 171–180. doi:10.2118/19451-PA
30
31
Boxall, J.A., Koh, C.A., Sloan, E.D., Sum, A.K., Wu, D.T., 2012. Droplet size scaling of waterin-oil emulsions under turbulent flow. Langmuir 28, 104–110. doi:10.1021/la202293t
32
33
34
Boxall, J., Davies, S., Koh, C., Sloan, E.D., 2009. Predicting When and Where Hydrate Plugs
Form in Oil-Dominated Flowlines. SPE Proj. Facil. Constr. 4, 80–86. doi:10.2118/129538PA
35
36
Dapena, J.A., Majid, A.A., Srivastava, V., Wang, Y., Charlton, T.B., 2017. OTC-27621-MS Gas
Hydrate Management Strategies Using Anti-Agglomerants : Continuous & Transient Large-
AC
C
EP
TE
D
19
16
ACCEPTED MANUSCRIPT
1
Scale Flowloop Studies 1–17.
Dapena, J.A., Majid, A.A., Srivastava, V., Wang, Y., Charlton, T.B., Gardner, A.A., Sloan, E.D.,
Zerpa, L.E., Wu, D.T., Koh, C.A., 2017. Gas Hydrate Management Strategies Using AntiAgglomerants: Continuous &amp;amp; Transient Large-Scale Flowloop Studies, in:
Offshore Technology Conference. doi:10.4043/27621-MS
6
7
8
Davies, S., Boxall, J., Koh, C., Sloan, E., Hemmingsen, P., Kinnari, K., Xu, Z.-G., 2009.
Predicting Hydrate-Plug Formation in a Subsea Tieback. SPE Prod. Oper. 24, 573–578.
doi:10.2118/115763-PA
RI
PT
2
3
4
5
Davies, S.R., Boxall, J.A., Dieker, L.E., Sum, A.K., Koh, C.A., Sloan, E.D., Creek, J.L., Xu,
Z.G., 2010. Predicting hydrate plug formation in oil-dominated flowlines. J. Pet. Sci. Eng.
72, 302–309. doi:10.1016/j.petrol.2010.03.031
12
13
14
Galindo-Alvarez, J., Sadtler, V., Choplin, L., Salager, J.L., 2011. Viscous oil emulsification by
catastrophic phase inversion: Influence of oil viscosity and process conditions. Ind. Eng.
Chem. Res. 50, 5575–5583. doi:10.1021/ie102224k
15
16
Grasso, G. a, Lafond, P.G., Aman, Z.M., Zerpa, L.E., Sloan, E.D., Koh, C. a, Sum, A.K., 2014.
Hydrate Formation Flowloop Experiments. 8th Int. Conf. Gas Hydrates.
17
18
19
Høiland, S., Askvik, K.M., Fotland, P., Alagic, E., Barth, T., Fadnes, F., 2005. Wettability of
Freon hydrates in crude oil/brine emulsions. J. Colloid Interface Sci. 287, 217–225.
doi:10.1016/j.jcis.2005.01.080
20
21
22
Joshi, S. V., Grasso, G.A., Lafond, P.G., Rao, I., Webb, E., Zerpa, L.E., Sloan, E.D., Koh, C.A.,
Sum, A.K., 2013a. Experimental flowloop investigations of gas hydrate formation in high
water cut systems. Chem. Eng. Sci. 97, 198–209. doi:10.1016/j.ces.2013.04.019
23
24
Koh, C.A., Sloan, E., Sum, A., 2011. Natural Gas Hydrates in Flow Assurance, Natural Gas
Hydrates in Flow Assurance. doi:10.1016/C2009-0-62311-4
25
26
27
Liu, Z., Zerpa, L.E., Wang, Y., 2016. Hydrate Problems for Gas Lift Operations for Deepwater
and Arctic Wells and Best Practices for Prevention, in: SPE Western Regional Meeting.
doi:10.2118/180392-MS
28
29
Moradpour, H., Chapoy, A., Tohidi, B., 2011a. Phase Inversion in Water À Oil Emulsions with
and without Gas Hydrates 5736–5745.
30
31
Moradpour, H., Chapoy, A., Tohidi, B., 2011b. Phase Inversion in Water–Oil Emulsions with
and without Gas Hydrates. Energy & Fuels 25, 5736–5745. doi:10.1021/ef201374v
32
33
34
Salager, J.L., Forgiarini, A.M., 2012. Emulsion stabilization, breaking, and inversion depends
upon formulation: Advantage or inconvenience in flow assurance. Energy and Fuels 26,
4027–4033. doi:10.1021/ef3001604
35
36
37
Skovborg, P., Rasmussen, P., 1994. A mass transport limited model for the growth of methane
and ethane gas hydrates. Chem. Eng. Sci. 49, 1131–1143. doi:10.1016/00092509(94)85085-2
38
Sloan, E.D., 2005. A changing hydrate paradigm - From apprehension to avoidance to risk
AC
C
EP
TE
D
M
AN
U
SC
9
10
11
17
ACCEPTED MANUSCRIPT
management, in: Fluid Phase Equilibria. pp. 67–74. doi:10.1016/j.fluid.2004.08.009
2
3
Sloan, E.D., Koh, C., 2007. Clathrate Hydrates of Natural Gases, Third Edition. Clathrate
Hydrates Nat. Gases, Third Ed. 20074156, 1–701. doi:10.1201/9781420008494
4
5
6
7
Srivastava, V., Majid, A.A.A., Warrier, P., Grasso, G., Chaudhari, P., Dendy Sloan, E., Koh,
C.A., Wu, D.T., Zerpa, L.E., 2017. Hydrate Formation and Transportability Investigations
in a High-Pressure Flowloop During Transient Shut-in / Restart Operations, in: Offshore
Technology Conference. doi:10.4043/27849-MS
8
Turner, D.J., 2005. Clathrate hydrate formaton in water-in-oil dispersions.
RI
PT
1
Turner, D.J., Miller, K.T., Dendy Sloan, E., 2009. Methane hydrate formation and an inward
growing shell model in water-in-oil dispersions. Chem. Eng. Sci. 64, 3996–4004.
doi:10.1016/j.ces.2009.05.051
12
13
14
15
Vijayamohan, P., Majid, A., Chaudhari, P., Sloan, E.D., Sum, A.K., Koh, C.A., Dellacase, E.,
Volk, M., 2014. Hydrate Modeling &amp;amp; Flow Loop Experiments for Water
Continuous &amp;amp; Partially Dispersed Systems, in: Offshore Technology Conference.
doi:10.4043/25307-MS
16
17
Vysniauskas, A., Bishnoi, P.R., 1983. A kinetic study of methane hydrate formation. Chem. Eng.
Sci. 38, 1061–1072. doi:10.1016/0009-2509(83)80027-X
18
19
Zerpa, L.E., 2013. A Practical Model to Predict Gas Hydrate Formation, Dissociation and
Transportability in Oil and Gas Flowlines 164.
20
21
22
Zerpa, L.E., Aman, Z.M., Joshi, S., Rao, I., Sloan, E.D., Koh, C. a, Sum, a K., 2012a. Predicting
Hydrate Blockages in Oil , Gas and Water-Dominated Systems. Otc. doi:10.4043/23490MS
23
24
25
Zerpa, L.E., Rao, I., Aman, Z.M., Danielson, T.J., Koh, C.A., Sloan, E.D., Sum, A.K., 2013.
Multiphase flow modeling of gas hydrates with a simple hydrodynamic slug flow model.
Chem. Eng. Sci. 99, 298–304. doi:10.1016/j.ces.2013.06.016
26
27
Zerpa, L.E., Sloan, E.D., Sum, A.K., Koh, C.A., 2012b. Overview of CSMHyK: A transient
hydrate formation model. J. Pet. Sci. Eng. doi:10.1016/j.petrol.2012.08.017
29
M
AN
U
TE
D
EP
AC
C
28
SC
9
10
11
18
ACCEPTED MANUSCRIPT
A Transient Simulation Model to Predict Hydrate Formation Rate
2
in both Oil- and Water-Dominated Systems in Pipelines
3
Yan Wang a, Carolyn A. Koh a, J. Alejandro Dapena a, Luis E. Zerpa a, b,*
RI
PT
1
4
a
5
School of Mines, 1600 Illinois St., Golden, CO 80401, USA
6
b,*
7
1600 Arapahoe St., Golden, CO 80401, USA. Email: lzerpa@mines.edu, +1 (303) 384-2627
8
Highlights
Center for Hydrate Research, Department of Chemical & Biological Engineering, Colorado
M
AN
U
SC
(Corresponding author) Department of Petroleum Engineering, Colorado School of Mines,
•
Hydrates can form simultaneously in oil- and water-dominated sections of a pipeline
10
•
Hydrate formation can trigger dispersion inversion from water to oil-continuous
11
•
Inclination of pipeline affects hydrate accumulation in pipelines
12
•
Proposed hydrate formation model is validated through flowloop simulations
13
•
The validated model is applied to field simulations at high water cuts
AC
C
EP
TE
D
9
1
Документ
Категория
Без категории
Просмотров
0
Размер файла
3 043 Кб
Теги
010, 2018, jngse
1/--страниц
Пожаловаться на содержимое документа