Download as pdf or txt
Download as pdf or txt
You are on page 1of 21

OTC-27681-MS

Experimental Investigation of Severe Slugging Control by Surfactant


Injection

Carolina V. Barreto, Amanda Pimenta, Hamidreza Karami, Eduardo Pereyra, and Cem Sarica, The University of
Tulsa

Copyright 2017, Offshore Technology Conference

This paper was prepared for presentation at the Offshore Technology Conference held in Houston, Texas, USA, 1–4 May 2017.

This paper was selected for presentation by an OTC program committee following review of information contained in an abstract submitted by the author(s). Contents of
the paper have not been reviewed by the Offshore Technology Conference and are subject to correction by the author(s). The material does not necessarily reflect any
position of the Offshore Technology Conference, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written
consent of the Offshore Technology Conference is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may
not be copied. The abstract must contain conspicuous acknowledgment of OTC copyright.

Abstract
Recently, with the ever-increasing development of offshore fields, the use of long risers fed by lateral
flow lines has significantly increased. This type of configuration exposes the production system to severe
slugging, particularly with decline in reservoir pressure and consequently flow rates. Severe slugging can
cause huge pressure fluctuations and may result in decreased recovery and separator flooding. This study is
an experimental and theoretical effort to analyze how the application of surfactants can suppress or eliminate
severe slugging.
The experimental study is performed in a 2-in ID facility, consisting of a 64-ft flow line with 1° downward
inclination, followed by a 41-ft riser. Water and compressed air are the liquid and gas phases, and an anionic
surfactant is applied continuously. First, effects of surfactant delivery location are investigated by injecting
at three points, namely, riser base and flowline inlet without or with a static mixer, used as an external source
of agitation. Pressure and holdup fluctuations with time, slugging frequency and surfactant efficiency are
monitored. Moreover, a high-speed camera is used to monitor blowout stage and analyze gas front thickness
and velocity. Pressure fluctuations and surfactant efficiency are experimentally investigated by changing
the operating conditions. Tests include without and with surfactant at different liquid and gas flow rates.
Firstly, the experiments are conducted with lowest superficial gas velocity (vSg) and highest superficial
liquid velocity (vSL) values. At these conditions, severe slugging is partially eliminated only with high
surfactant concentration and the use of static mixer. The use of surfactant in other locations results in
elimination of slug production stage and reduced cycle duration, but negligible changes in pressure profile.
Then, a new test matrix is designed to conduct tests at higher vSg and lower vSL values, moving towards the
boundary of Bøe's criterion for severe slugging. Under these conditions, severe slugging cycle durations
and pressure fluctuations reduce and elimination efficiency of surfactant increases significantly. In addition,
severe slugging severity and elimination efficiency are correlated with Boe's criterion coefficient.
This study is an attempt to characterize severe slugging magnitude based on operating conditions,
and investigate surfactant efficiency for different delivery locations. Considering the limited number of
studies on this subject, the results can provide an experimental and theoretical source to optimize surfactant
application in production systems facing severe slugging.
2 OTC-27681-MS

Introduction
Severe slugging is a phenomenon occurring in two-phase flow through a downward inclined flow line
followed by a vertical riser at low gas flow rates. In this phenomenon, the liquid accumulates in the riser
and curvature section of the flow line blocking the passage of gas in the lowest point of the system. As
a result, the gas front penetrates the liquid blockage intermittently, causing extremely large slugs, severe
fluctuations, and flooding of downstream equipment.
The first stage of severe slugging is the slug formation, in which, a large liquid slug is formed as the
liquid blocks the passage of the gas and increases its level into the riser section. Consequently, the gas
phase accumulates in the flow line and compresses. The liquid slug reaches the top of the riser initiating the
second stage, slug production. In this stage, liquid is produced, while the gas pocket is being pressurized
for an eventual penetration into the liquid. The third stage or blowout occurs once the gas pressure in the
downward inclined flow line overcomes the hydrostatic pressure of the liquid column. In this stage, the gas
pushes the liquid column violently out of the riser. As the pressure declines in the pipeline, the fourth and
last stage of the severe slugging phenomenon, liquid fallback, occurs. In this stage, the remaining liquids
fall back and accumulate at the riser base and curvature sections. Figure 1 shows a schematic of the severe
slugging phenomenon.

Figure 1—Severe slugging schematic (Malekzadeh et al. 2012)

Based on the balance of pipeline and riser forces, Bøe (1981) developed a criterion to determine the
occurrence of severe slugging phenomenon. The ratio between the rate of gas pressure increase in the
flowline and the rate of hydrostatic pressure increase in the riser leads to the following equation:

(1)
OTC-27681-MS 3

where Wg and WL are the mass rates of gas and liquid, and HL is the liquid holdup in the pipeline upstream
of the liquid blockage. For the cases that ΠSS value is less than 1, severe slugging is expected to occur based
on this criterion. In this study, Bøe criterion is used for comparison purposes.
Severe slugging elimination methods have been proposed based on the backpressure increase and gas
lift (Tengesdal et al. 2002). The first method eliminates severe slugging by increasing the system pressure
and, consequently reducing the production capacity. Gas lift consists of injecting gas into the riser to reduce
hydrostatic head in the riser and increase the gas flow rate in the pipeline. This method requires large
amounts of gas to accomplish the elimination and has high operational cost.
Tengesdal et al. (2002) tested a new approach, named self-gas lifting, to attenuate severe slugging in
pipeline-riser systems, by transferring the pipeline gas (in-situ gas) to the riser at a point above the riser
base. The transferring process reduces both the hydrostatic head in the riser and the pressure in the pipeline
with no additional gas injection required from the platform. The proposed method is proven very effective
for severe slugging attenuation in pipeline-riser system.
Another method to mitigate severe slugging is foaming as mentioned by Hassanein and Fairhurst (1998).
This method requires foaming agents (surfactants) and a foam generation method. Surfactants, or surface
active agents, are amphipathic molecules that consist of a non-polar hydrophobic portion, attached to a
polar hydrophilic portion (see Figure 2). They are characterized by the tendency to absorb at surfaces and
interfaces, or in other words, the boundary between two immiscible phases. The driving force for surfactant
absorption lowers the free energy of the interface boundary. This interfacial free energy is referred to as
surface tension (γ).

Figure 2—Surfactant structure (Holmberg et al. 2003)

Increasing the surfactant concentration decreases the liquid surface tension to a minimum value, which
is reached when micelles start to form. This concentration is called critical micelle concentration (CMC).
Figure 3 shows a generic behavior of surface tension (γ) as a function of surfactant concentration.

Figure 3—Critical micelle concentration (CMC)

Foam is a dispersed system consisting of gas bubbles separated by liquid films. Pure water or oil cannot
form foam unless a surfactant is present. In addition to surfactant presence, the process of gas-liquid mixing
and generation of interfacial area is required to generate foam. In other words, bubbles need to be generated
to allow the surfactants to stabilize this colloidal system or foam. These mechanisms are controlled primarily
by the power dissipated in the fluids and the gas volume fraction (Paul et al. 2004). In this work, the
4 OTC-27681-MS

aforementioned power is referred to as agitation, and can be provided mechanically by the use of a static
mixer or by the existing flow patterns. Together with the fluid properties, the agitation influences the bubble
size and consequently the foam stability.
Zukoski (1954) studied the influence of viscosity, surface tension, and inclination angle on motion of
long bubbles in closed tubes. Results of this study show that a decrease of the surface tension leads to an
increase in bubble penetration velocity. This result can be achieved by surfactant injection, and is believed
to impact the gas penetration rate during blowout stage of severe slugging.
Sarica et al. (2014) showed that the use of surfactant to an air-water system reduced severe slugging.
After the injection of surfactant, foam was generated by the self-agitation of the phases. The severe slugging
is mitigated by changing the flow pattern occurring in the system. They conducted baseline tests without
surfactant injection and tests with surfactant injection. The tests were conducted in a flow loop consisting
in a 65.6 ft pipeline with -3° inclination, followed by a 49.2 ft riser. A new elimination performance index
(EPI) was proposed to quantify the severe slugging mitigation efficiency. Equation 2 shows this parameter.

(2)

where Δpmin and Δpmax are respectively the minimum and maximum values of pressure drop along the flow
loop during the severe slugging cycle. The elimination performance was classified into three categories,
namely, complete elimination (CE) for 50<EPI<100, partial elimination (PE) for 0<EPI<50, and no
elimination (NE) for EPI≤0. It was shown that surfactant injection in the riser base can mitigate severe
slugging for transition region at relatively high gas flow rates within the severe slugging envelope.
Sarica et al. (2015) investigated the feasibility of surfactant and gas lift combination as a severe slugging
suppression method. They proposed a new modified elimination performance index (MEPI) to include the
backpressure effects, as follows:

(3)

This study is an effort to analyze the effects of different parameters on surfactant efficiency in severe
slugging mitigation. These parameters include surfactant injection location, concentration, and gas and
liquid rates. Compared to Sarica et al. (2014) the tests are conducted with a smaller pipe ID (2 in.) and for
lower liquid and gas operating rates, and a more detailed analysis is carried out on severe slugging process.

Experimental Program
Experimental work of this study was conducted in a 2-in. ID facility. This facility operates with a maximum
operating pressure of 25 psig at the test section. The flow loop consists of a 62.6 ft long flow line followed
by a 41 ft long riser. The the inclination angle of the flow line can be varied ±1° from horizontal. The pipe
material is acrylic, which allows visual observation of the flow behavior.
The facility is equipped with multiple pressure, temperature, and differential pressure transmitters. The
pressure transmitters (PT) and temperature transmitters (TT) are placed in the gas and liquid inlets, flow line
inlet, riser base and in the middle of the riser. The measurements are used to calculate the in-situ superficial
gas velocity and monitor the changes in fluid properties. Differential pressure transmitters (DPT) are located
at the inlet and the end of the flow line, at the curvature, and in the middle of the riser. The DPTs are used
to measure the pressure losses along each section, and consequently evaluate the effects of the surfactant
presence.
Four quick closing valves (QCVs) are installed along the flow line to provide two sections for liquid
holdup measurement. In addition, four QCVs are installed along the riser providing three consecutive
OTC-27681-MS 5

trapping sections. The average liquid holdup in the riser is obtained by calculating the ratio between the
height of the trapped liquid (hL-n) and the total height of each section (hn). Then, the final holdup is obtained
by averaging the results from the three sections. In the lateral section, liquid holdup is measured by draining
the trapped liquid and calculating the ratio between the drained liquid volume (VL) and the total volume of
the section (VT). The facility schematic is shown in Figure 4.

Figure 4—Experimental facility schematic

Tap water is pumped by a centrifugal pump from a 2000 gallon water tank to the flow loop. Compressed
air is supplied by a rotary screw air compressor with a maximum outlet pressure of 100 psig. Flow rates of
the two fluids are controlled by coriolis type flow meters. A control valve is used to adjust air flow rate,
which is complemented by a bypass valve for the cases of higher air flow rate. A more detailed description
of the facility is provided in Barreto (2016).
An anionic surfactant with a specific gravity varying from 0.94-1.14, is used for the experiments. The
reported critical micelle concentration for this surfactant is 2400 ppm, according to Ajani (2014). Figure
5 shows surface tension of the water-surfactant mixture as a function of surfactant concentration for this
surfactant.
6 OTC-27681-MS

Figure 5—Surface tension for the anionic surfactant

Surfactant is pumped using a digital pump, from a 5 gallon surfactant solution tank at the desired injection
point. The injection points include the riser base (IP3), flow line inlet (IP2), and flow line inlet with a static
mixer used as an external source of agitation (IP1). Pressure and holdup fluctuations with time, slugging
frequency and gas front penetration rate are monitored. After flowing through the test section, the mixture
of foam, water, and gas is disposed into two tanks, the drain tank and broken foam tank.
In order to determine whether an external source of agitation is essential to generate foam, a static mixer
is installed at the beginning of the flow line. This static mixer is used for the tests applying IP1, and can be
bypassed for the tests without an external source of agitation. The mixing section consists of a 2-ft long and
4-in. ID acrylic pipe with flanges at the bottom and the top. Air flows from the bottom of the static mixer,
while a solution containing surfactant and water enters from the side. Inside the acrylic pipe, a wire mesh
is placed. The wire mesh spreads air around the mixing section, and creates sufficient agitation for foam
generation (bubble formation). Finally, the foam flows out homogeneously through the top of the static
mixer, into the flow line. Figure 6 shows a schematic of the static mixer.

Figure 6—Static mixer schematic

In order to visualize flow patterns in the riser, a surveilance camera is installed on a rope and spool device.
This allows visualization and recording the flowing conditions at different points along the vertical section,
as shown in Figure 7. In addition, the height of the trapped liquid in each section is obtained using the basler
camera, and hence the liquid holdup is calculated. In the flow line, the flow patterns were investigated by
the recordings obtained with a full HD resolution (1920 × 1080 pixels) camera.
OTC-27681-MS 7

Figure 7—Surveilance camera

Experimental Results
The experimental campaign of this study was conducted in two stages. First, experiments were conducted to
investigate the effects of surfactant concentration and injection location on its efficiency in severe slugging
mitigation. These tests were completed at a fixed vSL and vSg value. In the next stage, surfactant concentration
and injection location were fixed, and the effects of change in vSL and vSg on surfactant efficiency were
evaluated. In this section, the test matrices and the acquired experimental results for these two stages are
presented.

Effects of Foamer Delivery Location


A test matrix was designed and the effects of varying dosages of surfactant on severe slugging are
investigated. In addition, transient pressure and liquid holdup measurements are made, and high-speed
camera captures were used to investigate the effects of surfactant on the gas front penetration rate at
the blowout phase. The changes in different parameters with respect to surfactant injection location and
concentration are discussed.
Test Matrix. Two different surfactant concentrations (500 and 3000 ppm) were tested for all the injection
points (IP1, IP2 and IP3). Fixed vSL, vSg and qinj values of 0.25 ft/s, 0.33 ft/s, and 10.72 gal/day were
considered. In order to reach the desired surfactant concentration, 1 L of surfactant solution included
50.5 mL of pure surfactant for 500 ppm concentration cases and 303 mL of pure surfactant at 3000 ppm
concentration tests. Table 1 presents the test matrix used in the severe slugging tests.

Table 1—Severe slugging experimental test matrix


8 OTC-27681-MS

Gas Penetration Rate Analysis. Initial observations with the vSL of 0.25 ft/s and vSg of 0.33 ft/s showed
that the severe slugging cycle reduced significantly for all the injection points. In addition, the blowout
phase started immediately after the riser was completely filled by liquid. The hydrostatic pressure of the
liquid slug was no longer increasing and the gas was penetrating through the liquid to reach the riser and
commence the blowout stage.
A possible explanation for shorter cycle duration and elimination of slug production phase with surfactant
application is the drop in surface tension of the liquid phase. The surfactant acts by reducing the surface
tension of the liquid, and as a result, reducing the forces that keep the surface intact. Therefore, the
penetration of the gas phase through the liquid blockage is facilitated (Lea et al. 2008).
Gas front penetration velocity (vbubble) and thickness (hbubble) were measured for the severe slugging cases.
In order to obtain vbubble and hbubble, a high-speed video system was used. Videos were recorded at 250 fps at
the riser base to capture the moment when the gas front penetrates through the liquid. This moment is just
before the blowout. A 12-in measuring tape was placed below the flow line to obtain the length travelled
by the gas (Lbubble). The gas front penetration velocity, vbubble, was then calculated using Eqn. 4:

(4)

where Framei and Framef are the frame numbers used to record the gas front location at initial and final
stages. Thickness of the gas front, hbubble, was determined using a software. Initially, the number of pixels
representing the entire pipe diameter (Pixp) was measured and associated to a value of 2 in. Then, the number
of pixels equivalent to the gas front thickness was measured (Pixg), and converted to inches as shown in
Eqn. 5:

(5)

Figure 8 shows an example of gas front penetration frames, for surfactant concentration of 500 ppm
injected at IP2. Table 2 shows the obtained vbubble and hbubble values for air-water flow base case (AW) and
the cases of surfactant injection at IP2 and IP3. Since a foam layer flows on top of the liquid layer for IP1,
the gas penetration cannot be quantified.

Table 2—Gas front penetration rate (vbubble) and thickness (hbubble)


OTC-27681-MS 9

Figure 8—Gas penetration rate (vSg = 0.328 ft/s and vSL = 0.25 ft/s, IP2, 500 ppm)

As observed from Table 2, the use of surfactant facilitates the gas flow through the liquid blockage and
increases gas front thickness and penetration rate. The effects are more pronounced for the cases with higher
surfactant concentration. Using IP2, vbubble increases 250% and 442% for 500 and 3000 ppm, respectively,
while for IP3, vbubble increases 64% for 500 ppm case and 221% with 3000 ppm surfactant concentration.
The injection point location also influences the results. When injected at the riser base (IP3), the surfactant
cannot completely diffuse through the liquid present in the flow line, while in the case of IP2, surfactant
diffusion is better, resulting in a uniform water/surfactant solution in the flow line.
Pressure Analysis and Slug Frequency. From the conducted experiments, inlet pressure data were
recorded and analyzed. Air-water base case (AW) shows severe slugging pressure behavior (see Figure 9).
The inlet pressure increases due to the slug accumulation stage until it reaches a plateau indicating slug
production phase. This plateau is followed by a sudden pressure drop associated with blowout and liquid
fallback phases. All the surfactant injection points eliminate the slug production stage, and as soon as the
liquid column reaches the top of the vertical section, foam blowout is observed. The severe slugging cycle
duration is reduced with the use of surfactant independently of the surfactant injection location. Visual
observations confirm that, using surfactant, gas penetrates through the liquid blockage at the curvature as
soon as the liquid column reaches the top of the riser leading to the blowout stage. This is in agreement of
the bubble penetration analisys presented before.
10 OTC-27681-MS

Figure 9—Inlet pressure fluctuations for vSL = 0.25 ft/s and vSg = 0.328 ft/s, using 500 ppm of surfactant

For the surfactant concentration of 500 ppm, all injection points show a reduction in cycle duration.
However, the case of static mixer (IP1) presents the largest reduction in slug duration. For IP2 and IP3,
the amplitude of pressure fluctuations is similar to the air-water base case, while for IP1 this amplitude
is slightly reduced. This is due to the presence of a small amount of foam generated by the static mixer
agitation. This foam layer avoids complete liquid blockage at the curvature, allowing the passage of gas in
the shape of small bubbles. Bubble flow has lower mixture density than single-phase liquid flow, leading to
a smaller hydrostatic pressure and consequently lower pressure fluctuation amplitude than the air-water base
case. Figure 9 shows the inlet pressure fluctuations, given by the pressure transducer PT-3, as a function of
time for vSL of 0.25 ft/s and vSg of 0.33 ft/s, with surfactant concentration of 500 ppm.
Figure 10 shows the inlet pressure fluctuations, given by the pressure transducer PT-3, as a function of
time for vSL of 0.25 ft/s and vSg of 0.33 ft/s, with surfactant concentration of 3000 ppm. For this surfactant
concentration, IP1 is the injection point with the highest reduction in pressure fluctuations. Once again,
amplitude of pressure fluctuations is similar to the air-water base case for IP2 and IP3, while a significant
reduction in pressure fluctuations is observed for IP1. Due to the larger surfactant concentration, more foam
is generated by the static mixer. The foam, flowing on top of the liquid, avoids complete liquid blockage
at the curvature. Larger bubbles, in theshape of Taylor bubbles, climb through the vertical section and help
unload some of the liquid. The blowout in this case occurs before the liquid column is entirely filled with
liquid, leading to a partial severe slugging phenomenon and significantly lower pressure fluctuations. Thus,
an increase in foam quality causes a reduction in maximum pressure and promotes partial blowout.
OTC-27681-MS 11

Figure 10—Inlet pressure fluctuations for vSL = 0.25 ft/s and vSg = 0.328 ft/s using 3000 ppm of surfactant

Overall, the number of severe slugging cycles increases with the use of surfactant, indicating a higher
gas production potential from a gas well. Compared to the air-water base case, IP1 presents 53% and 79%
increase in severe slugging cycle frequency, for the cases with 500 and 3000 ppm surfactant concentration,
respectively. However, IP2 presents relative increase of 22% and 12% in cycle frequency for 500 and 3000
ppm. The relative frequency increases for the case of IP3 are 18% and 23%, corresponding to 500 and
3000 ppm, respectively. Figure 11 shows the severe slugging cycle frequency for air-water flow and all the
surfactant injection tests.

Figure 11—Surfactant effects on severe slugging cycle frequency

Elimination Performance Index. Table 3 shows the severe slugging suppression performance rated by
means of EPI and MEPI terms for all injection points using the anionic surfactant at concentrations of 500
ppm and 3000 ppm for vSL of 0.25 ft/s and vSg of 0.328 ft/s. Green, yellow and red show CE, PE and NE,
respectively.
12 OTC-27681-MS

Table 3—Severe slugging suppression efficiency, rated by EPI and MEPI terms

As observed, severe slugging phenomenon is partially eliminated only for IP1 at the high surfactant
concentration of 3000 ppm. All other cases show that severe slugging is not eliminated, based on both EPI
and MEPI analyses.
Liquid Holdup. Throughout the severe slugging cycle, several HL measurements were made by draining the
flow line and by recording the liquid height in the riser. The measurements were made after several cycles,
when the system was stabilized to assure that the acquired values are representative of the phenomenon.
The measurements were taken right after blowout, at the mid-point of liquid accumulation phase, at the end
of liquid accumulation phase, and at the end of slug production phase (for air-water flow).
From these measurements, it is possible to obtain the overall system HL during the severe slugging cycle
for air-water base case and all the surfactant injection points. For this purpose, the HL values in the flow line
and riser sections are multiplied by the respective volumes of the sections. The resulting liquid volumes
of the flow line and riser sections are then summed to obtain the total system liquid inventory. Finally,
dividing the liquid inventory by the total volume of the test section, average system HL profile is acquired.
The effects of liquid accumulation at the curvature are neglected in this estimation. Figure 12 shows the HL
profile during a severe slugging cycle with a surfactant concentration of 500 ppm (a) and 3000 ppm (b).

Figure 12—Liquid holdup (HL) profile. With Anionic Surfactant,


at 500 ppm (a) and 3000 ppm (b) at vSg = 0.33 ft/s and vSL =0.25 ft/s

The minimum liquid holdups correspond to the measurements immediately after blowout, while the
maximum holdups correspond to the beginning of slug production for air-water flow and before blowout
for all injection points. The average HL over a severe slugging cycle is obtained from the plots above by
determining the area under each plot over the corresponding cycle time. For this purpose, a linear changing
OTC-27681-MS 13

trend of HL with time is assumed, and the areas under the curves are estimated for different cases using
simple geometrical relationships. Finally, dividing the result by the cycle duration, average HL over a cycle
is estimated.
The observed time-averaged HL values in flow line and riser sections do not change significantly between
air-water case and surfactant application cases for all the injection points. This means that, although the
reduction in slug frequency has a positive effect on gas production for gas wells, it is not very effective
in unloading the liquids. One exception is for the case of IP1 at 3000 ppm concentration, in which severe
slugging is partially eliminated and liquid holdup drops 43.1% compared to air-water case. From the
combined HL values of flow line and riser sections, it is possible to calculate the average liquid inventory in
the system. Figure 13 shows the liquid inventory in the system for every case tested. It is again confirmed
that IP1 at 3000 ppm is the only injection point presenting significant liquid removal compared to the air-
water base case. Based on these results, for the tested values of vSL and vSg and given surfactant type, it is
concluded that surfactant use does not provide a considerable improvement in severe slugging mitigation,
unless an external source of agitation is used at a high surfactant concentration.

Figure 13—Liquid Inventory with Anionic Surfactant at Different Injection Points

Effects of Operating Conditions


Above, surfactant efficiency to remediate severe slugging at different injection points was presented.
However, the experiments were conducted only at one operating condition deep inside the severe slugging
envelope based on Boe criterion. This section provides the results for a second experimental campaign,
conducted to look at surfactant efficiency, while moving towards the boundary of severe slugging envelope
by increasing vSg and decreasing the vSL values.
Test Matrix. The test section configuration, experimental fluids, and the surfactant type are the same
as in the previous section. However, the injection point is fixed at IP2 (flow line section inlet) and the
lower surfactant concentration of 500 ppm is considered in the test matrix along with the base case of air-
water flow. The recorded parameters for this test matrix include video observations, and transient pressure
measurements along the test section using the available pressure and differential pressure transducers. Table
4 shows the designed test matrix for this section.
14 OTC-27681-MS

Table 4—Operating Conditions Investigation Test Matrix

Video Observations. Three flow patterns were observed in the conducted test matrix, including slug flow,
partial severe slugging and severe slugging. Partial severe slugging is similar to what is described in the
previous section with high concentration surfactant injection at IP1. The blockage at the riser base for the
gas flowthrough, and consequently, bubble flow is observed in the riser during the slug accumulation phase.
Both severe slugging and partial severe slugging cases exhibit a blowout phase. However, the blowout phase
is not observed for the tests with slug flow pattern, where intermittent flow of relatively shorter slug bodies
and Taylor bubbles is recorded. Figure 14 shows the screenshots of the flow for tests with slug flow (a),
partial severe slugging (b) and severe slugging (c).

Figure 14—Screenshots of the Observed Flow Patterns, Slug Flow (a) Partial Severe Slugging (b), Severe Slugging (c)

Pressure Analysis and Slug Frequency. The fluctuations in inlet pressure are used to investigate the
changes in flow pattern for the cases without and with surfactant injection. As described earlier, severe
slugging is identified by slug lengths as large as the riser length, and consequently, pressure fluctuations
as large as the hydrostatic liquid column in the riser. However, for partial severe slugging, bubbles of gas
can penetrate through the liquid blockage during the slug accumulation phase and flow through the riser.
As a result, the pressure fluctuations decrease depending on the liquid holdup of the riser at the moment of
blowout corresponding to the point with maximum riser base pressure. Finally, for slug flow pattern, the
fluctuation magnitute decreases sharply due to the elimination of blowout stage.
Figure 15 shows the sample transient data of the test section inlet pressure for tests conducted with the
highest vSL of 0.25 ft/s, and vSg values of 0.33 (a) and 0.99 ft/s (b). For the case of lower vSg (a), severe
slugging is observed with and without surfactant, and magnitude of pressure fluctuations are the same.
However, the slug production phase is eliminated with surfactant and the slug cycle duration is reduced.
OTC-27681-MS 15

For the case of higher vSg (b), partial severe slugging is observed with and without surfactant. However,
the surfactant helps reduce the pressure fluctuations due to the earlier blowout of the gas into the riser and
the slug cycle duration.

Figure 15—Inlet Pressure Fluctuations with Time with vSL = 0.25 ft/s, and vSg = 0.33 (a) and 0.99 ft/s (b)

Figure 16 shows another sample of transient data of the test section inlet pressure for tests conducted with
the lowest vSL of 0.08 ft/s, and vSg values of 0.33 (a) and 0.99 ft/s (b). For the case of lower vSg (a), severe
slugging is observed with and without surfactant with no slug production phase, relatively lower pressure
fluctuations and higher cycle duration compared to the case of Figure 15 (a). In addition, the surfactant
helps reduce the pressure fluctuations due to the earlier gas blowout the slug cycle duration. For the case
of higher vSg (b), partial severe slugging is observed without surfactant. However, severe slugging seems
to be eliminated with surfactant, and slug flow pattern is observed with significantly lower cycle durations
and pressure fluctuations.

Figure 16—Inlet Pressure Fluctuations with Time with vSL = 0.08 ft/s, and vSg = 0.33 (a) and 0.99 ft/s (b)
16 OTC-27681-MS

Figure 17 shows a summary of the measured severe slugging cycle frequency in cycles/hour for all the
tested cases of vSL and vSg, with and without surfactant injection. A general trend that can be observed for all
cases is that increasing vSg causes an increase in cycle frequency or a decrease in cycle duration. Comparing
the three plots for water-air flow, it is also observed that increasing vSL has a similar effect and increases the
cycle frequency. It is observed that the surfactant has the highest impact in the tests with higher vSg and lower
vSL values. For these cases, the surfactant makes a substantial increase in severe slugging cycle frequency
or decrease in cycle duration. This effect can be easily observed for the case of tests with vSL = 0.08 ft/s.
However, as seen from the plot of high vSL tests with vSL = 0.25 ft/s, the surfactant effects on increasing the
cycle frequency are not as significant.

Figure 17—Effects of Surfactant on Severe Slugging Frequency for the tested vSL and vSg Values

In addition to surfactant effects on reducing the severe slugging cycle duration, another desirable impact
is to help reduce the pressure fluctuations in the system. Since severe slugging is a phenomenon occuring
in pipeline-riser systems with extremely low operating rates, the frictional pressure drops in the system
are negligible in comparison to gravitational drops. Therefore, the pressure losses in the system are mainly
governed by the hydrostatic column of liquid in the riser and riser base. The pressure fluctuations can be
used to estimate the fluctuation in the holdup of liquid in the riser. For this purpose, liquid buildup (ΔHL,vertical)
is defined as the difference between the maximum liquid holdup (right before blowout) and the minimum
liquid holdup at the riser (after the blowout phase). Equation 6 shows the formulation to calculate liquid
buildup based on the fluctuations in hydrostatic pressure of the liquid column.

(6)

Liquid buildup values close to unity show that the liquid completely fills the riser before the blowout, and
then it gets all blown out during blowout phase. Therefore, these values are indicator of severe slugging in
the system. Mid-range values of liquid buildup show a partial liquid filling of the riser before blowout, and
therefore can be an indicator of partial severe slugging. Liquid buildup values close to zero show minimum
changes in the liquid inventory throughout the slugging cycle. Therefore, they are indicators of a slug flow
pattern and no severe slugging.
Figure 18 shows the changes in the estimated liquid buildup for tests with different vSL and vSg values, and
with or without surfactant injection. As observed, increasing vSg results in a decrease in liquid buildup and
OTC-27681-MS 17

slugging severity for all cases. On the other hand, increasing vSL causes an increase in liquid accumulation in
the system during the cycle, and therefore, results in higher liquid buildup values for all cases. The surfactant
causes an improvement by decreasing the liquid buildup in most cases, especially for the tests with partial
severe slugging or slug flow.

Figure 18—Surfactant Effects on Liquid Buildup in the Riser for the tested vSL and vSg Values

Table 5 shows the summary of the liquid buildup data for all the tests with or without surfactant. The
tests with liquid buildup values higher than 0.9 are considered to be in severe slugging region, shown in
red. The tests with liquid buildup values between 0.5 and 0.9 are considered to have partial severe slugging,
shown in yellow. Finally, the tests with liquid buildup values less than 0.5 are in the slug flow pattern, and
are shown in green. The effects of surfactant on shrinking the severe slugging region and eliminating severe
slugging at two cases can be observed from the table.

Table 5—Flow Pattern and Surfactant Efficiency Identification with Liquid Buildup

In addition, the values of modified elimination performance index (MEPI) are presented in Table 5, to
evaluate the surfactant efficiency in eliminating severe slugging. The results also show a similar trend by
using EPI to analyze the effects. As expected, complete severe slugging elimination (CE) is obtained for
tests with lower vSL and higher vSg values, shown in green. At lower vSg and higher vSL values, shown in
red, no severe slugging elimination (NE) is obtained. For the tests conducted at the mid-range operating
conditions, shown in yellow, partial severe slugging elimination (PE) is obtained.

Evaluation of Bøe Criterion


In this section, the experimental results of this study on the existence of severe slugs are compared to the
predictions of Bøe's criterion, the most well known predictive method in the literature. The comparisons are
18 OTC-27681-MS

done for both sets of tests with and without surfactant. For this purpose, Eqn. 1 is used in the calculations,
and two-phase flow model of Taitel and Dukler (1976) is used to estimate liquid holdup in the pipeline with
the assumption of stratified gas-liquid flow.
Figure 19 shows the comparison between Bøe's predictions and experimental severe slugging data for
the case of air-water flow. It is commonly believed that Bøe's criterion is a conservative estimate of severe
slugging occurance. However, it is observed here that even some tests falling outside the Bøe's envelope
experience full severe slugging. In addition, all the tests with partial severe slugging fall outside this
envelope.

Figure 19—Predictions of Bøe Criterion on Severe Slugging Existence Compared to Experimental Data

Surfactant can help reduce severe slugging by foam generation and reduction of hydrostatic pressure in
the riser, and also surface tension reduction and facilitating gas blowout. Bøe's criterion does not consider
the effects of surface forces on severe slugging process, and is merely a balance of hydrostatic pressure in the
riser and gas compression in the flow line. In this study, the effects of foam presence in the riser were applied
on Bøe's criterion by replacing liquid density with foam density. However, no foam quality measurements
were conducted at this study. Barreto (2016) completed foam quality measurements at the operating range
of this study, and observed a wide range of foam qualities, ranging between 0.4 and 0.8. Figure 20 shows
the results of this study on severe slugging existence with 500 ppm of surfactant plotted against Bøe's
predictions for air-water and air-water-surfactant with foam qualities of 40%, 60%, and 80%. Increasing the
foam quality decreases the hydrostatic pressure at the riser, and therefore shrinks Boe's criterion towards
lower vSg and higher vSL values. Bøe's criterion again predicts the onset of severe slugging, and the tests with
partial severe slugging are not included in the predicted envelope.
OTC-27681-MS 19

Figure 20—Predictions of Bøe Criterion, Considering Different Foam Qualities, Compared to Experimental Data

It was observed earlier that by moving deeper into the severe slugging envelope the value of liquid buildup
(ΔHL,vertical), calculated using Eqn. 6, increases and approaches unity. At the same time, the Bøe's criterion
value (ΠSS) gets lower than unity and decreases. Figure 21 shows the changing trend of liquid buildup
with ΠSS, for tests conducted with and without surfactant injection. For both cases, at high values of ΠSS,
liquid buildup values approach zero and severe slugging is eliminated. On the other hand, when ΠSS values
decrease, liquid buildup values approach unity. In addition, the surfactant efficiency in eliminating severe
svere slugging is presented by plotting MEPI with respect to ΠSS. As observed, the surfactant efficiency
significantly increases by increasing ΠSS.

Figure 21—Changes of Liquid Buildup in the Riser with Bøe's Criterion Value

Conclusions
Severe slugging phenomenon was observed at superficial gas velocity (vSg) of 0.33 ft/s and superficial liquid
velocity (vSL) of 0.25 ft/s for an air-water base case. The analysis of the transient inlet pressure showed
that the slug production stage is eliminated with surfactant injection. As a result, the slug cycle duration
decreases indicating that more gas is produced at a given time in a gas well. The amplitude of pressure
fluctuations was similar to that of the air-water case for IP2 and IP3, while it was reduced for IP1 due to
the presence of a foam layer generated by the static mixer. The gas front penetration, initiating the blowout
stage, was analyzed using a high-speed camera. By introduction of surfactant, the gas front penetration rate
20 OTC-27681-MS

increased leading to the elimination of slug production stage. In addition, the thickness of the penetrating
gas front increased by surfactant injection. It is suggested that the surface tension reduction has a positive
effect on the penetration rate and gas front thickness. For the case of IP1 with a foam layer present, the
full liquid blockage was avoided and gas front penetration rate and thickness were significantly higher. As
the surfactant concentration increased, the pressure fluctuation amplitude decreased using IP1, and severe
slugging was partially eliminated.
Later, a short test matrix was designed, and severe slugging existence was evaluated at different vSg and vSL
values. Based on visual recordings and transient pressure data, full and partial severe slugging was observed
for air-water. Addition of surfactant helps to completely or partially eliminate severe slugging at higher
vSg (0.66, 0.99 ft/s) and lower vSL (0.08 ft/s) tests. The predictions of Bøe's criterion for severe slugging
existence were compared to the experimental data for the tests with and without surfactant. It was observed
that the tests with partial severe severe slugging are not included in Bøe's severe slugging envelope.

Acknowledgements
The authors would like to thank the Tulsa University Horizontal Well Artificial Lift Projects (TUHWALP)
members for their support of this research. This work was made possible with the support of CAPES,
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil.

Nomenclature
A = Area of the Pipe (in2)
CMC = Critical Micelle Concentration (-)
CE = Complete Elimination (-)
dp/dt = Pressure Buildup Rate (psi/s)
EPI = Elimination Performance Index (-)
hbubble = Gas Front Penetration Thickness (in)
hL = Liquid Film Thickness (in)
hvertical = Riser Length (ft)
HL = Liquid Holdup (-)
L = Pipe length (ft)
Lbubble = Length traveled by gas (in)
MEPI = Modified elimination performance index (-)
Mg = Molecular weight of the gas (lb/lbmol)
NE = No elimination (-)
PE = Partial elimination (-)
R = Gas constant (ft3·psi·°R-1·lb-mol-1)
T = Temperature (K)
vbubble = Gas front penetration velocity (ft/s)
VL = Liquid Volume (L)
vSg = Superficial gas velocity (ft/s)
vSL = Superficial liquid velocity (ft/s)
VT = Total volume of section (L)
Wg = Mass rate of gas (lb/s)
WL = Mass rate of liquid (lb/s)
Z = Gas compressibility factor (-)
ΔHL,vertical = Liquid buildup in the Riser (-)
ρ = Density (lb/ft3)
ΠSS = Bøe criterion (-)
OTC-27681-MS 21

References
1. Ajani, A. A. 2014. Experimental Study and Modeling of Effect of Surfactants on Liquid Loading
in Vertical Pipes (Doctoral dissertation, The University of Tulsa).
2. Barreto, C.V. 2016. Effect of Foamer Delivery Location on Horizontal Wells Deliquification
(Master of Science Thesis, The University of Tulsa)
3. Bøe, A. 1981. Severe Slugging Characteristics; Part 1: Flow Regime for Severe Slugging; Part
2: Point Model Simulation Study. Presented at Selected Topics in Two-Pase Flow, Trondheim,
Norway, March 1981.
4. Hassanein, T., and Fairhurst, P. 1998. Challenges in the Mechanical and Hydraulical Aspects of
Riser Design for Deep Water Developments. IBC UK Conf. Ltd. Offshore Pipeline Technology
Conference, Oslo, Norway.
5. Holmberg, K., Jönsson, B., Kronberg, B., and Lindman, B. 2004. Surfactants and polymers in
aqueous solution. Journal of Synthetic Lubrication, 20(4): 367–370.
6. Malekzadeh, R., and Mudde, R. F. 2012. A Modelling Study of Severe Slugging in Wellbore.
Presented at North Africa Technical Conference and Exhibition. Society of Petroleum Engineers.
7. Paul, E. L., Atiemo-Obeng, V. A. and Kresta, S. M. (Eds.). 2004. Handbook of industrial mixing:
science and practice. John Wiley & Sons.
8. Sarica, C., Begen, P., Pereyra, E. and Kouba, G. 2014. July. Feasibility of surfactants as severe
slugging suppression agents. Presented at 9th North American Conference on Multiphase
Technology. BHR Group.
9. Sarica, C., Yuan, G., Shang, W., Pereyra, E. and Kouba, G. 2015. Feasibility and Evaluation of
Surfactants and Gas Lift in Combination as Severe Slugging Suppression Method. Oil and Gas
Facilities, 4(04): 78–87.
10. Tengesdal, J. Ø., Sarica, C., & Thompson, L. 2002. Severe slugging attenuation for deepwater
multiphase pipeline and riser systems. In SPE Annual Technical Conference and Exhibition.
Society of Petroleum Engineers.
11. Taitel, Y., & Dukler, A. E. 1976. A model for predicting flow regime transitions in horizontal and
near horizontal gas-liquid flow. AIChE Journal, 22(1): 47–55.
12. Zukoski, E. E. 1966. Influence of viscosity, surface tension, and inclination angle on motion of
long bubbles in closed tubes. Journal of Fluid Mechanics, 25(04): 821–837.

You might also like