Proceedings of the 2nd IFAC Workshop on Automatic Control in

Proceedings
Offshore of the
Oil and Gas2nd IFAC Workshop on Automatic Control in
Production
Proceedings
Proceedings of the
of the 2nd
2nd IFAC Workshop
IFAC Workshop onon Automatic
Automatic Control
Control in
in
Offshore
May Oil2015.
27-29, and Gas Production
Florianópolis, BrazilAvailable online at www.sciencedirect.com
Offshore
Offshore Oil and
Oil2015. Gas
and Gas Production
Production
May 27-29, Florianópolis, Brazil
May 27-29,
May 27-29, 2015.
2015. Florianópolis,
Florianópolis, Brazil
Brazil

ScienceDirect
IFAC-PapersOnLine 48-6 (2015) 089–096
Review
Review of
of Slug
Slug Detection,
Detection, Modeling
Modeling and
and
Review
Control of Slug Detection,
Techniques for Modeling
Offshore Oil &and
Gas
Control Techniques
Control Techniques for
for Offshore
Offshore ⋆Oil
Oil &
& Gas
Gas
Production Processes
Production Processes ⋆
⋆
Production Processes
Simon Pedersen ∗∗ Petar Durdevic ∗∗ and Zhenyu Yang ∗∗
Simon
Simon Pedersen ∗∗ Petar Durdevic ∗∗ and Zhenyu Yang ∗
Simon Pedersen
Pedersen Petar Petar Durdevic
Durdevic and and Zhenyu
Zhenyu Yang Yang ∗
∗
∗ Department of Energy Technology, Aalborg University, Esbjerg
∗ Department
Department
∗ Campus,
of
of Energy
Energy Technology,
Technology, Aalborg University,
Aalborg Denmark
University, Esbjerg
Esbjerg
Department Nielsof Bohrs Vej
Energy 8, 6700 Esbjerg,
Technology, Aalborg University, (E-mail:
Esbjerg
Campus,
Campus, Niels
Niels Bohrs
Bohrs Vej
Vej 8,
8, 6700
6700 Esbjerg,
Esbjerg, Denmark
Denmark (E-mail:
(E-mail:
Campus, spe@et.aau.dk,
Niels Bohrs pdl@et.aau.dk,
Vej 8, 6700 yang@et.aau.dk)
Esbjerg, Denmark (E-mail:
spe@et.aau.dk,
spe@et.aau.dk, pdl@et.aau.dk, yang@et.aau.dk)
spe@et.aau.dk, pdl@et.aau.dk,
pdl@et.aau.dk, yang@et.aau.dk)
yang@et.aau.dk)
Abstract: The current offshore oil & gas multi-phase production and transportation installa-
Abstract:
Abstract: The
The current
current offshore oil & gas multi-phase production and transportation installa-
tions
Abstract:
tions
have big
have big current offshore
Thechallenges
challenges
related oil
offshore
related
with
oil
with
&
& gasthe multi-phase
gas
the
slugging flow:
multi-phase
slugging flow:
production
An unstable
production
An unstable
and transportation
and multi-phase
transportation
multi-phase
flowinstalla-
flow
regime
installa-
regime
tions
where have
tions have the big
flow
big challenges
rates,
challenges related
pressures with
and
related and the slugging
temperatures
withtemperatures flow:
oscillate
the slugging oscillate An
flow: An in unstable
in the
unstable multi-phase
considered
multi-phase flow
flow regime
processes. Slug
regime
where
where the
the flow
flow by rates,
rates, pressures
pressures and temperatures
temperatures oscillate the
in the considered
thestructures.
consideredThe processes.
processes. Slug
Slug
can
where be the
caused
flow different
rates, operating
pressures and conditions and installation
oscillate in considered most severe
processes. Slug
can
can be
be caused
caused by different operating conditions and installation structures. The most severe
slugs
can beare often by
caused by different
induced
different in operating
long vertical
operating conditions
risers orand
conditions and installation
production
installation structures.
wells, where liquid
structures. The
The most
blocks
most severe
gas
severe
slugs
slugs are
are often
often induced
induced in
in long
long vertical
vertical risers
risers or
or production
production wells,
wells, where
where liquid
liquid blocks
blocks gas
gas
at
slugsthe areriser/well
often inducedbase and correspondingly
in correspondingly
long vertical risers it causes the
or production pressure to accumulate
wells,towhere and
liquid blocks hence
gas
at
at the
the riser/well
riser/well base
base and
and correspondingly it
it causes
causes the
the pressure
pressure to accumulate
accumulate and
and hence
hence
originates the
at the riser/well oscillating
base and performance.
correspondingly There are
itare many
causes severe
the consequences
pressure to
to accumulate the production
and hence
originates the
originates because oscillating
the oscillating
oscillating performance.
performance. There
There are many
many severe
severesomeconsequences
consequences to
to the production
the production
production
processes
originates the of the performance.
slugging flow.There This arepaper manyreviews
severe observed to
consequences latest
the status and
processes
processes because
because of
of the
the slugging flow. This paper reviews some observed latest status and
key challenges
processes because aboutof slugslugging
the detection,
slugging flow.
flow. This
dynamical
This paper
paper reviews
modeling
reviews andsome
some observed
elimination
observed latest
of status
slugging
latest statusflows.and
and
key
key challenges
challenges about
about slug
slug detection,
detection, dynamical
dynamical modeling
modeling and
and elimination
elimination of
of slugging
slugging flows.
flows.
Mathematical
key challenges modeling
about slugof slug has
detection, been used
dynamical to investigate
modeling the
and slug mechanism
elimination of and
slugging anti-slug
flows.
Mathematical
Mathematical modeling
modeling of slug
slug has
ofmodels has been
been used
used to
to investigate
investigate the the slug
slug mechanism
mechanism and
and anti-slug
anti-slug
control.
Mathematical Most of availableof
modeling slug has arebeen
basedusedonto mass-balance
investigate formulations,
the slug which and
mechanism oftenanti-slug
require
control.
control. MostMost
Most of available
offoravailable
available models
models are based
are based
based on mass-balance
on mass-balance
mass-balanceSlug formulations,
formulations, which
which often
often require
require
sufficient
control. dataof reliable models
parameter are tuning/identification.
on elimination
formulations, whichand often
control have
require
sufficient
sufficient data
data for
for reliable
reliable parameter tuning/identification. Slug elimination and control have
been
sufficientinvestigated
data for for many parameter
reliable years and there
parameter tuning/identification.
exist many solutions
tuning/identification. Slug
Slug toelimination
eliminate the
elimination and
and control
slug, have
but some
control have
been
been investigated
investigated for
for many
many years
years and
and there
there exist
exist many
many solutions
solutions to
to eliminate
eliminate the
the slug,
slug, but
but some
some
of these
been methods
investigated can simultaneously
for simultaneously
many years andreduce reduce the
there exist oil &
many gas production,
solutions which
to eliminate is a
the very
slug,bigbutconcern
some
of these
of these
these methods
methods can can
can simultaneously the
reduce parameter oil
the oil &
oil & gas
& gas
gas production,
production, which
which is is a
is aaWevery
very big concern
big concern
concern
as
of the production
methods ratesimultaneously
is the key evaluationreduce the forproduction,
offshore production.
which very conclude
big that
as
as the
the production
production rate
rate is
is the
the key
key evaluation
evaluation parameter
parameter for offshore production. We conclude that
the
as slugging
the production flow is a well-defined
rate is the key phenomenon,
evaluation even for
parameter for offshore
though
offshorethisproduction.
subject hasWe
production. Webeen conclude
conclude that
extensively
that
the
the slugging
slugging flow
flow is
is a
a well-defined
well-defined phenomenon,
phenomenon, even
even though
though this
this subject
subject has
has been
been extensively
extensively
investigated
the slugging in
flow theis past
a decades,
well-defined the cost-effective
phenomenon, evenand optimal
though thisslug modeling
subject has and
been control
extensively are
investigated
investigated in
in the
the past
past decades,
decades, the
the cost-effective
cost-effective and optimal slug modeling and control are
still open
investigated
still open
topics
in the
topics
with
with
many
past
many
related
decades,
related the cost-effective and optimal slug modeling and control are
challenges.
challenges.
and optimal slug modeling and control are
still open
still open topics
topics withwith many
many related
related challenges.
challenges.
© 2015, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.
Keywords: Slugging flow, multi-phase, detection, modeling, anti-slug control, Offshore, oil &
Keywords:
Keywords: Slugging flow, flow, multi-phase, detection,detection, modeling, anti-slug anti-slug control, Offshore,Offshore, oil & &
Keywords: Slugging
gas
gas Slugging flow, multi-phase,
multi-phase, detection, modeling, modeling, anti-slug control, control, Offshore, oil oil &
gas
gas
1. INTRODUCTION
1. INTRODUCTION
1.
1. INTRODUCTION
INTRODUCTION
Slug is a common flow pattern in multiple-phase flow
Slug is a common flow pattern in multiple-phase flow
Slug
system,
Slug is
is aasuch
common
commonas in flow the oil pattern
& gasin multiple-phase
upstream productionflow
system, such as in flow the oil pattern
& gasinupstream
multiple-phase
productionflow
system,
process.
system, such
The gas
such as
as in
andthe
in the oil
liquid
oil &
& gas
(water
gas upstream
and oil) may
upstream production
not be
production
process. The gas and liquid (water and oil) may not be
process.
evenly The
The gas
process.distributed gas and liquid
liquid (water
throughout
and and
and oil)
the production
(water may
may not
oil) wells, be
trans-
not be
evenly distributed throughout the production wells, trans-
evenly
port
evenly distributed
pipelines
distributed and throughout
risers due tothe
throughout the production
specific wells,
configuration
production wells, trans-
and
trans-
port pipelines and risers due to specific configuration and
port pipelines
operating
port and
and risers
condition,
pipelines suchdue
risers thatto specific
the liquidconfiguration
and gas traveland as
operating condition, suchdue thattothe specific
liquidconfiguration
and gas traveland as
operating
plugged condition,
aoperating train
condition, with such
a
such that
large
that the
plug
the liquid
of
liquidoneand
and gas
phase
gas travel
medium
travel as
as
a plugged train with a large plug of one phase medium
afollowed
a plugged
pluggedby train
the with
otheraa phase
large
large plugmedium of
of oneplugphase medium
through the
followed bytrain
the with
other phase plug
medium one
plugphase medium
through the
followed
pipeline.
followed by
As the
the other
by shown phase
in Figure
other phase1, medium
these large plug through
plugs the
are often
pipeline. As shown in Figure 1,medium
these large plug through
plugs the
are often
pipeline.
referred
pipeline. As
to
As shown
as
shownslugsin
in Figure
(Schmidt
Figure 1,
1, these
et al.large
these plugs
(1980);
large plugs are
Taitel
are often
and
often
referred to as slugs (Schmidt et al. (1980); Taitel and
referred
Dukler
referred to
to as
(1976)).as slugs
This (Schmidt
slugs et
et al.
type of irregular al. (1980);
flows can Taitel
resultandin
Dukler (1976)). This (Schmidt
type of irregular (1980);
flows can Taitel
resultandin
Dukler
very
Dukler (1976)).
poor oil
(1976)). andThis
This type
water of irregular
separation,
typeseparation, flows
reduced
of irregularreducedflows cancan result
result in
production in
very poor oil and water production
very
very poor
capability, oil
oil and
poor extra and water
fatigue
waterloadsseparation, reduced
to installations
separation, reducedand production
facilities,
production
capability, extra fatigue loads to installations and facilities,
capability,
shortening
capability, extra
devicefatigue loads
life-times, to
to installations
loads accelerating and
and facilities,
component cor-
shortening extradevicefatigue
life-times, installations
accelerating component facilities,
cor-
shortening
rosion,
shorteningand device
even life-times,
emergent accelerating
shut-off of component
production cor-
(Aamo
rosion, and device life-times,shut-off
even emergent accelerating component
of production cor- Fig. 1. Typical flow patterns in a vertical pipeline (from
(Aamo
rosion,
et
rosion, and
and even
al. (2004); even emergent
Eikrem
emergent shut-off
(2006);
shut-off of
of production
Hassanein (Aamo
and Fairhurst
production (Aamo Fig. 1. Typical flow patterns in aa vertical pipeline (from
et al. (2004); Eikrem (2006); Hassanein and Fairhurst
et al.
(1998);
et al. (2004);
Havre and
(2004); Eikrem
Dalsmo
Eikrem (2006);
(2001);Hassanein
(2006); Di-Meglioand
Hassanein (2012a); Fig.
et al.Fairhurst
and Fairhurst
1.
1. Typical
Fig. left
left
to right):
Typical
to
flow
flow
right):
patterns
Bubble
patterns
Bubble
flow,in
flow, a vertical
inslug
slug
flow, churn
vertical
flow,
pipeline
churn
flow,(from
pipeline
flow,
and
(from
and
(1998); Havre and Dalsmo (2001); Di-Meglio et al. (2012a); left
annual
left to
to right): Bubble
flow (Taitel
right): Bubble et flow, slug
al. (1980))
flow, slug flow,
flow, churn
churn flow,
flow, and
and
(1998);
Storkaas
(1998); Havre
(2005);
Havre and
and Dalsmo
Taitel
Dalsmo (2001);
et al.
(2001); Di-Meglio
(1990);
Di-Meglio et
Tengesdal
et al.
al. (2012a);
et
(2012a);al. annual flow (Taitel et al. (1980))
Storkaas (2005); Taitel et al. (1990); Tengesdal et al. annual
annual flow
flow (Taitel
(Taitel et
et al.
al. (1980))
(1980))
Storkaas
(2002)).
Storkaas (2005); Taitel et al. (1990); Tengesdal et al.
(2002)). (2005); Taitel et al. (1990); Tengesdal et al. geometric locations within the offshore upstream produc-
(2002)).
(2002)). to specific operating condition and system con- geometric locations within the offshore upstream produc-
Subject geometric locations within the
Subject to specific operating condition and system con- tion tion
process.
geometric
process.
As shown
locations
As shown in the offshore
in Figure
within Figure offshore
2,
upstream
2, a slugging
upstream
aa slugging
flow
flow
produc-
could
produc-
could
Subject
figuration,
Subject to
to specific
the slugging
specific operating
flow can
operating condition
occur at
condition and
andmanysystem
system con-
different
con- tion
appear process.
in the As shown
gas-lifting in Figure
production 2, slugging
wells due flow
to a could
casing-
figuration, the slugging flow can occur at many different appear tion process.
in the As shown inproduction
gas-lifting Figure 2, wells a slugging
due flow
to a could
casing-
figuration,
⋆ figuration, the
the slugging
slugging flow
flow can
can occur
occur at
at many
many different
different appear
heading in the gas-lifting
mechanism production
(Eikrem (2006); wells
Hu due to
(2004)), a casing-
terrain
Supported by the Danish National Advanced
⋆ Supported by the Danish National Advanced Technology Founda- Technology Founda- appear
heading in the gas-lifting
mechanism production
(Eikrem (2006); wells
Hu due to
(2004)), aa casing-
terrain
⋆ heading mechanism
could occur(Eikrem (2006);
tion through by
Supported
⋆ Supported
tion through by
PDPWAC
the Danish
the
PDPWAC
DanishProject (J.nr.Advanced
National
National
Project
95-2012-3).
Advanced
(J.nr. 95-2012-3).
Technology Founda-
Technology Founda- slugging
heading
slugging mechanism
could occur (Eikrem
in (2006); Hu
in transportation
transportation Hu (2004)),
pipelinesaa terrain
(2004)),
pipelines
due to
terrain
due to
tion through PDPWAC Project (J.nr.
tion through PDPWAC Project (J.nr. 95-2012-3). 95-2012-3). slugging
slugging could occur in transportation pipelines due
could occur in transportation pipelines due to
to
Copyright
2405-8963 © © 2015,
2015 IFAC 89 Hosting by Elsevier Ltd. All rights reserved.
IFAC (International Federation of Automatic Control)
Copyright © 2015 IFAC 89
Peer review©
Copyright
Copyright ©under
2015 responsibility
2015 IFAC
IFAC of International Federation of Automatic
89 Control.
89
10.1016/j.ifacol.2015.08.015
IFAC Oilfield 2015
90
May 27-29, 2015. Florianópolis, Brazil Simon Pedersen et al. / IFAC-PapersOnLine 48-6 (2015) 089–096

Fig. 2. A schematic illustration of a well-pipeline-riser Fig. 3. A illustration of different flow patterns in a flow
system in the upstream offshore oil & gas production map (Taitel (1986))
the seafloor elevations (Jansen (1990); Ogazi (2011)), and source consumption could be extremely high and time-
a severe slug could appear at the riser part due to the consuming.
significant gravity influence (Jahanshahi et al. (2013b); Di-
Meglio et al. (2012a)). Some slugging flow could be sim- 2.2 Bøe Criterion
ply induced due to some intermittent system/operational
performances, such as pigging, start-up, blow-down, or Slug flow can be estimated by using some simple criteria
changes of production references etc. (Sivertsen et al. than an empirical flow map subject to some specific
(2010)). constraints. Such as, Taitel and Dukler (1976) gave a
criterion for stratified flow in horizontal pipelines, and
This paper intends to give a brief review of the latest this criterion has been adopted in oil & gas industry in
status and key techniques about slug detection, dynam- combination with the studies from Schmidt et al. (1980),
ical modeling and slug elimination for offshore oil & gas which proved that the occurrence of stratified flow in a
production processes. The rest of the paper is organized horizontal pipeline is a pre-condition for severe slug. Later
as: Section 2 focuses on discussion of different slug de- on, an extension of this criterion is presented in Bøe (1981)
tection criteria; Section 3 introduces several typical slug for a pipeline-riser system, and it is now referred to as Bøe
dynamic models for the purpose of supporting anti-slug Criterion and it is [described as:
control design and analysis; Section 4 illustrates different ]
Pp
slug elimination methods which are classified into passive ULS ≥ US, (1)
and active approaches; Finally, we conclude the paper in ρL g(1 − ϵL )Lsin(α) G
Section 5 with the opinion that even though the slugging where ULS is the (pipeline) injected superficial velocity of
flow is a well defined concept and this subject has been S
liquid, UG is the (pipeline) injected superficial velocity of
extensively studied in several decades, the cost-effective gas, α is the inclination angle of the riser, ϵL is the ratio
and optimal slug modeling and control are still open topics between the liquid and the combined gas and liquid in the
with many related challenges. riser, and Pp is the pressure at the riser base (bottom of
the riser). The interpretation of this criterion is based on
2. SLUG DETECTION the observation (Schmidt et al. (1980)) that: The rate of
gas pressure accumulation at the riser base must be greater
The occurrence of slug can be determined according to than that at the pipeline section, in order to have a severe
relevant flow dynamic theory by checking the system con- slug to be formed in the riser. However, the Bøe Criterion
figuration, parameters and operating conditions (Hamathy is based on the assumptions of constant inlet flow rates,
(1960); Taitel and Dukler (1976)). a pressure balance over the riser as well as the gas mass
balance in the pipeline.
2.1 Flow Pattern Map As Bøe criterion assumes that if the liquid column is
stable, a constant steady state should exist. However, some
As the slug is defined as a steady-state flow pattern, some exceptions have been discovered in Taitel et al. (1990) with
flow pattern map as shown in Figure 3 can be exper- a fact that a tendency of a cyclic process still can exist even
imentally obtained for all possible operating conditions when the liquid column is stable. This work concluded that
(Taitel et al. (1980); Taitel (1986)). These flow patterns the Bøe criterion is good at differing from steady to cyclic
in Figure 3 correlate with superficial velocities of both flow with a few exceptions, especially with the cases of
liquid and gas phases. If there are more operational and high liquid flow rates, where a predicted severe slugging
manipulated variables/parameters, this type of flow map region with Bøe Criterion can actually be stable. Jansen
needs to be extended to cover all possible operating ranges. et al. (1996) further noticed that the Bøe Criterion is only
Correspondingly, the required experimental work and re- valid when no slug elimination methods are applied.

90
IFAC Oilfield 2015
May 27-29, 2015. Florianópolis, Brazil Simon Pedersen et al. / IFAC-PapersOnLine 48-6 (2015) 089–096 91

2.3 Taitel Criterion

An alternative slug criterion is proposed by Taitel et al.

(1990), which was an extension of the result from Taitel
(1986). This Taitel Criterion sets up the correlation of the
gas holdup pressure at the riser base and the riser (top-
side) back-pressure as:
ϵl+L
Ps ϵ′ − h
> P0
, (2)
P0 ρL gΦ
where Ps is the back-pressure that the riser needs to
overcome in order to generate the production flow, and it
is often correlated with the downstream separator pressure
located on the separation platform. P0 is the atmospheric
pressure, ρL is the (combined oil and water) liquid density,
ϵ is the ratio of volume of the liquid over the volume of Fig. 4. Flow map comparison of Bøe Criterion and other
combined liquid and gas, ϵ′ is the void fraction for a Taylor criteria (Jansen et al. (1996))
bubble that penetrates into the riser. A Taylor bubble
(also called gas slug) is the large asymmetric bullet-shaped
bubble within a gas-liquid multi-phase flow. The Taylor developed a criterion for S-shaped risers. Tchambak (2004)
bubble can occupy almost the entire cross-section of the presented a prediction criterion which also focused on
pipeline and has a length of several times of the pipeline S-shaped risers, but with three different gas injection
diameter (Liao and Zhao (2003)). l is the pipeline length locations: the gas injection at the pipeline inlet, the
before the riser, h is the height of the riser, L is the length downstream, and the upstream of the riser base.
before the liquid and gas is being combined into two phase,
g is the gravitational acceleration, and Φ is an index of
local liquid holdup in the riser, and it can be calculated
by Φ = 1− uuGS , where ut is the Taylor bubble velocity, and 2.6 Slug Detection Methods
t
uGS is the gas superficial velocity. In Taitel et al. (1990)
ϵ′ is assumed to be a constant (0.9) in any vertical flow, All the above mentioned slug criteria can provide some
and ϵ has only one value based on the separator pressure. guidelines for real-life applications, such as providing pa-
rameter limitations for physical system/process design,
2.4 Jansen Criterion committing condition monitoring for operations, as well
as supporting anti-slug control design and analysis (Yang
An extension of Taitel Criterion was carried out in Jansen et al. (2013, 2014)). Of course, the applicability of any
et al. (1996), considering that the gas from the artificial criterion depends on (i) whether the corresponding crite-
lifting is the only gas flowing through the riser. However, rion’s assumptions are fully valid; (ii) the evaluation vari-
this criterion assumed a constant steady-state gas injection able/parameters are available or measurable. We observed
into the riser base. The Jansen Criterion can be expressed that almost all slug criteria require either the pressure
in the following: information at the riser base, which unfortunately is not
ϵGR L
available (no installed transmitters at all) in most real-
Ps ϵ′ − hR life offshore constructions, or the liquid and gas injection
> G P0 , (3) information as well as very detailed physical construction
P0 ϵ gρ
S
GR L
description. An alternative approach for slug detection is
UGR to purely use measured data and conduct some signal
where ϵGR = 1 − UBubble and UBubble = C0 Us + UD .
S
UBubble is the Taylor bubble’s superficial velocity, UGR is processing analysis on it. For instance, the riser topside
the superficial velocity of gas in the riser, and Us is the pressure and its changing rate are employed in our previous
combined superficial velocity of gas and liquid. Here C0 is work (Pedersen et al. (2014a)) to detect whether a severe
the drift parameter, and UD is the bubble drift velocity. slug is happening or not based on a lab facility developed
This study concluded that for complete Taylor bubbles by Jepsen et al. (2013), so that a supervisor will decide
these two parameter values are constant: C0 = 1.2 and switches between an anti-slug controller and a production
UD = 0.35; For complete bubble flow these values are controller.
C0 = 1.0 and UD (Hamathy (1960)). Figure 4 shows the In practice, the alignment of these theoretical oriented
flow map comparison of the Bøe Criterion and the Jansen criteria with real-time data analysis can lead to more
criterion. These experiments were committed with no slug reliable and accurate slug detection. Besides that, where
elimination methods applied and only constant liquid and in the considered system/process to install what type of
gas injection rates. transmitter(s) to collect slug-relevant signals, as well as
how to efficiently and reliably process the obtained data,
2.5 Other Criteria turn to be also very important issues as well. Sometimes,
some model-based estimation approaches may also need
Even though the Jansen Criterion is still widely used, to be coordinated in order to retrieve unavailable key
other criteria have also been proposed and used for specific slug-relevant parameters (Helgesen (2010),Sivertsen and
pipeline/riser constructions, such as Montgomery (2002) Skogestad (2005)).

91
IFAC Oilfield 2015
92
May 27-29, 2015. Florianópolis, Brazil Simon Pedersen et al. / IFAC-PapersOnLine 48-6 (2015) 089–096

3. SLUG MODELING or time-consuming to make a relevant model fit to the real

data. Some tradeoff between model precision and complex-
Modeling multi-phase flow dynamics in a process is always ity needs to be managed. It is also noticed that a 6-state
a challenging and active topic. Hereby we do not focus model is proposed in Jahanshahi (2013) for considering
on sophisticated CFD-based modeling approaches, with a well-pipeline-riser system. This model however requests
respect to fact that we are interested in reviewing some the average mass ratio of gas and liquid to the well from
dynamic models which can be potentially employed for reservoir known, and this can be an open issue since this
the purpose of anti-slugging control design and analysis. information is hardly available in reality. We noticed that
the model developed and used in Di-Meglio et al. (2009,
3.1 Early-Stage models 2012b) is based on a virtual valve model locating at the
bottom of the riser, thereby this model does not depend
Taitel et al. (1980) is one of the early studies to de- much on the physical appearance as the model proposed
scribe the transition relationships among flow patterns by Jahanshahi and Skogestad (2011) does. Consequently
for two-phase gas and liquid flows in vertical pipelines. the Di-Meglio model can be easily adapted to handle both
The achieved models mainly focus however, on the steady- pipeline-riser and (gas-lifting) well facilities. Our previous
state flows. A similar study was observed in Viggiani et al. Biltoft et al. (2013) used Di-Meglio model and described
(1988). Sarica and Shoham (1991) presented a dynamic the model tuning and Pedersen et al. (2014a) designed a
model for pipeline-riser systems based on a 1-D gravity- hybrid switching controller on top of the tuned model.
dominant flow in both the pipeline and riser. The experi-
mental verification showed that this model can satisfyingly 3.3 PDE-Based Models
predict pipeline pressure transients, liquid accumulation,
slug length, and cycle time for all flow conditions tested. For other types of relevant models, we would mention that
The detailed physical size and dimensions of the considered Sinègre et al. (2006) developed a PDE model to predict
system are required using this model, and it is also noticed slugs in gas-lifting wells where the stability analysis was
that the pipeline inclination angle is a very sensitive pa- performed through small gain theorem. Di-Meglio et al.
rameter. This model was compared to a model developed (2011) proposed a low-dimensional PDE model which
by Jansen (1990). It is concluded that this model devel- comprises the gas mass fraction, the pressure, and gas
oped in Sarica and Shoham (1991) can successfully predict velocity as states. Compared with numerical simulations
severe slug outside the Bøe (1981) region. However, when the model proves to be accurate according to oscillation
non-slugging flow occurs, this model does not converge to frequencies and shapes. Another PDE model was devel-
the same validated model from Jansen (1990). oped by Nemoto and Bali no (2012), which is based on two
switchable states: One indicates the status where the gas is
3.2 Mass-Balance Models able to penetrate into the riser (steady flow), and the other
state in which there is a liquid accumulation preventing
Storkaas et al. (2003) presented an ODE model for severe the gas from penetrating into the riser (severe slugging).
slugging in a pipeline-riser based on mass balance equa- The model considers the liquid penetration length and the
tions with 3 states: liquid height in the riser, thus the model can distinguish
ṁL = wL,in − wL,out different sizes of slugs.
ṁG1 = wG,in − wG1 (4)
ṁG2 = wG1 − wG,out 3.4 Challenges with Slug Models
along with an algebraic model of the choke valve locat-
The main issue of slug modeling is that most models
ing at the topside of the riser. The developed model
are based on mass-balance principle, thus the liquids and
was compared with OLGA simulation and tested at a
gasses injected into the system have to be known, which is
scaled medium-sized testing facility. Both results showed
often not the case in reality. For this reason several studies
good consistencies of the models and data. Several slug
have focused on developing observers to estimate flow
elimination controllers were designed based on this model
rates. For instance, Grimstad and Foss (2014) developed
(Sivertsen and Skogestad (2005),Storkaas (2005)).
an adaptive extension to the observer developed by Aamo
An improved model of (Storkaas et al. (2003)) with 4 states et al. (2004) which estimates the well flow rate and
for a pipeline-riser facility is proposed in Jahanshahi and downhole pressure from topside measurements on gas-lift
Skogestad (2011). The considered system is divided into wells. However, this adaptive observer only works with a
two subsystems: one describing the horizontal pipeline, slowly varying reservoir pressure. Mansoori et al. (2014)
and the other one describing the vertical riser. Each studied different transients of the bottom-hole pressure in
subsystem are modeled with two mass equations (gas and the well using system identification techniques to estimate
liquid). Similar 3-state or 4-state models can also be found a reservoir model. Another issue arises as the well-pipeline-
in Eikrem (2006, 2008); Kaasa and V. Alstad (2008); Di- riser models heavily depend on the initial conditions of the
Meglio et al. (2009); Silva and Nydal (2010). masses of all phases in the pipelines, which also can be hard
to estimate (Jepsen et al. (2013); Pedersen et al. (2014b)).
It has been noticed that no matter which model formu-
lation people intend to use, there are always a number Not all wells/risers are limited by few measurements. Some
of (model) tuning parameters need to be handled. For new ones have more transmitters and actuators integrated,
example, Kaasa and V. Alstad (2008) with 7, and Di- and they are commonly referred to as ”Smart wells/risers”
Meglio et al. (2009) with 5, both Eikrem (2008) and Silva or ”Intelligent wells/risers” (Johal and Cousins (1999)).
and Nydal (2010) with 3. This indicates that it can be hard Within these smart wells, the downhole transmitters can

92
IFAC Oilfield 2015
May 27-29, 2015. Florianópolis, Brazil Simon Pedersen et al. / IFAC-PapersOnLine 48-6 (2015) 089–096 93

monitor the well and reservoir conditions, and some of

them even equipped with control valves to control the
inflow of fluids from the reservoir to the well. Johal and
Cousins (1999) patented an intelligent riser for deep-
water oil & gas fields where gas-lifting, slug catching,
and measurements are combined into one big riser sys-
tem. These Smart wells are also being used to improve
the performance of artificial reservoir flooding, generally
water-flooding (van Essen et al. (2006, 2009); Zandvliet
et al. (2006)). As the industry now is considering more ad-
Fig. 5. Two different flow conditioners: Wavy Pipe (left)
vanced model-based control methods to the water-flooding
and Helix Pipe (right) developed at Cranfield Univer-
technique, some focus has been put into the integration of
sity (Xing et al. (2013); Adedigba (2007))
reservoir modeling and slug modeling (Doren et al. (2011)).
the movement of the gas in the pipeline to the riser base
4. SLUG CONTROL is accelerated compared with the liquid accumulation.
Another type of flow conditioner using a helix-shaped
Due to it negative influences, the slug flow, especially pipeline is reported in Adedigba (2007) and it is illustrated
a severe slug, is not expected in any normal operation. in Figure 5 as well.
Thereby slug elimination, or we call slug control in the A venturi-shaped device is patented by Almeida and
following, need to be carefully dealt. In the following, from Gonçalves (1999) as one type of flow conditioner, which
the control engineering point of view, we classify most consists of a convergent nozzle section followed by a
methods/approaches into two categories: divergent diffuser section. This device is supposed to be
• Passive approaches: The slug elimination is conducted located as part of the horizontal pipeline near to the
by some proper and dedicated system/process design, riser base. Venturi-shaped devices can give a pressure drop
instead of using feedback control strategy; and causing a mixing effect and converting the stratified flow to
• Active approaches: The slug elimination is realized by a non-stratified flow temporarily. A similar functional flow
some automatic feedback control strategy based on a conditioner can be found in Makogan (2007). It should
given system/process. be noticed that the flow conditioner approach is similar
as the permanent choking approach proposed by Jansen
In some cases, some feedback control strategy is applied et al. (1996), thereby they both may have a payoff with
along with some dedicated change in the system/process. a reduced production rate (Ogazi et al. (2009); Pedersen
To avoid confusion, we put them into the active approach et al. (2014a)).
category.
Slug Catchers Using a slug catcher after the riser or
4.1 Passive Slug Control topside of well is the most commonly used passive slug
elimination approach in the actual production systems.
Elimination of severe riser slug by creating a change in The slug catchers can be classified as vessel-type, finger-
the process has been investigated for a long time. Early type and parking-loop according to their different config-
studies, such as Yocum (1973), identified several different urations. However, it should be aware that this type of
solutions for process changes, which still are being used in approach is very effective but with a big price (McGuiness
practice today to handle the slug. These solutions can be and Cooke (1993)).
categorized into three groups:
Other Alternatives Hassanein and Fairhurst (1998) pre-
(1) Reducing the incoming line diameter near the riser to sented a method to avoid slug formation by attenuating
establish a stable flow regime; the non-homogeneous liquid and gas into one homogeneous
(2) Using dual multiple risers, instead of a single riser; fluid. The idea was to reduce the surface tension of the fluid
(3) Using fluid remix device, which purposely mixes fluids by injecting a surfactant which could change the fluid into
at the riser base to avoid liquid accumulation, hence foam, hence making the fluid homogeneous. However, this
to prevent a stratified flow to progress into a severe approach will definitely increase the difficulty of separation
slugging. at the downstream separation process, thereby ultimately
These three kinds of solutions form the fundamental basis affect the product quality and capability. A self-gas lifting
for all passive slug control methods explained in the approach is proposed in Sarica and Tengesdal (2000). The
following. basic idea is to reduce the static head (weight of liquid)
in the riser by an extra pipeline to pass the gas directly
Flow conditioners A flow conditioner is referred to a into the riser (bypassing the riser base). Tengesdal et al.
specific device that is installed in the pipeline with the (2002) investigated the possibility of compressing gas in
objective to affect the original flow regime. A typical the pipeline and then separating it from liquid at the
example of this is a Wavy Pipe developed by Xing et al. upstream of the riser base.
(2013) at Cranfield University (UK). A 7-bend Wavy Pipe
is illustrated in Figure 5 and it is placed close to the 4.2 Active Slug Control
riser base. The basic idea here is to artificially introduce
a number of small slugs through the wavy pipe, so that Active slug elimination approach involves some automatic
a severe riser slug can be avoided due to the fact that feedback control mechanism, which manipulates some ac-

93
IFAC Oilfield 2015
94
May 27-29, 2015. Florianópolis, Brazil Simon Pedersen et al. / IFAC-PapersOnLine 48-6 (2015) 089–096

tuators, which installed in the process system, subject to MIMO or MISO control system problem. For the well case,
some sensor feedback signals. These signals can be from Pagano et al. (2008) developed a model-free PI-controller
pressure, temperature and/or flow transmitters, depend- where the injection valve of the gas-lifting is controlled to
ing on which specific system is studied. The selections of stabilize the gas flow injected into the production tube,
actuators and sensors can be guided by some fundamental meanwhile the topside choke valve is used to stabilize the
system property analysis, e.g., following the input-output topside pressure. Abardeh (2013) investigated anti-slug
controllability analysis (Jahanshahi et al. (2012); Skoges- control in S-shaped risers, where a robust control solution
tad and Postlethwaite (2005)). was proposed using the topside choke valve and artificial
gas-lifting. Jahanshahi et al. (2013b) used feedback lin-
Choke Valve Control Choking some controllable valve(s) earization to design a nonlinear model-based control for
in a considered process is the most investigated active a pipeline-riser system using both the riser-base pressure
slug elimination approach. The anti-slug control using the and the topside pressure. Nevertheless, the stability of the
(riser) topside choke valve has been studied for many concerned system needs to be guaranteed (Asheim (1988)).
years, and typical work can be found in Havre and Dalsmo
(2001); Di-Meglio et al. (2012a); Storkaas and Skogestad Slug Compression A slug compression system is reported
(2008); Jahanshahi et al. (2012). Ogazi (2011) also consid- in Kovalev (2003). This solution is a combination of
ered the control valve located at the separator gas outlet process change and active feedback control of choke valves.
as an alternative anti-slug control actuator. Eikrem et al. A topside mini-separator is installed to separate the liquid
(2004); Jahanshahi et al. (2013a); Scibilia et al. (2008) from the gas upstream the first stage separator. Between
focused on the estimation of the seabed/downhole pressure the two separators there are two choke valves: One for
from a topside pressure transmitter for the purpose of the gas pipeline and one for the liquid pipeline. This
regulating the topside choke valve. Ogazi et al. (2009) also way the liquid injection into the first stage separator is
investigated the possibility of using large valve openings controlled to stabilize the height of the liquid, while the gas
to maximize the oil production rate while also eliminating injection is used to compensate for the possible slugging.
the slug. Enricone Stasiak et al. (2012) also developed a This is an advantage as the gas pipe is much easier to
topside control design for minimized either the flow or control when no liquid is in. This slug suppression system
pressure oscillations while keeping the choke valve opening was implemented and experimentally verified that this
higher than the opening which characterizes the beginning solution can successfully eliminate all types of slug and
of the limit cycle. Pedersen et al. (2014a) developed a self- improve the production rate of both oil and gas as well.
learning controller which consists of a supervisor and two However, the investment for extra equipment as well as the
baseline PID controllers, in order to automatically find corresponding extra maintenance will lead to the increase
out the best (choke valve’s) operating position with the of costs for running production.
maximal production rate subject to no-slug. Sinègre and
Petit (2006) proved that the density-wave instability in 5. CONCLUSIONS
gas-lifted wells, which is different from the casing-heading
phenomenon (Sinègre et al. (2005)), can also be controlled This paper examines the historical and current status and
by manipulating the choke valve only based on the well some key techniques related with the slugging flow in
head pressure measurement. Most of practical anti-slug offshore oil & gas production processes, specially focusing
control structures are PID types, thereby how to effec- on the riser-induced slug occurring in production wells or
tively tune these controllers have been focused, such as pipeline-riser installations.
Godhavn et al. (2005) proposed three PI tuning methods For detection and modeling of the slug flow, if all oper-
for eliminating slug. Jahanshahi et al. (2014) developed a ating conditions and physical structure and parameters of
new IMC-PIDF controller as an extension to the tuning considered well-pipeline-riser systems are known or mea-
methods proposed by Godhavn et al. (2005). surable, the corresponding slug criteria or dynamic models
can provide precise slug detection and prediction. However
Gas-Lifting Control It has been proved that using ar- in many practical cases, these detail information is very
tificial gas-lifting is also an effective approach in elimina- limited.
tion severe slugs (Asheim (1988); Plucenio et al. (2012)),
through a huge amount of gas might be needed to generate Some key slug elimination approaches have been analyzed
an actual effect on the flow pattern. Hu (2004) mentioned and they are classified into passive and active slug con-
two methods to obtain stable flow in the production well, trol categories. The two most popular active slug control
i.e., using the gas-lifting approach and using the water- approaches are feedback control of a riser/well topside
flooding to increase reservoir pressure. Krima et al. (2012) choke valve and feedback control of artificial gas-lifting
developed several PI controller design strategies for the in a riser/well. Two main objectives of slug control are
gas-lifting focusing on mitigating hydrodynamic slug in (i) eliminating the (severe) slugging flow; meanwhile (ii)
OLGA simulations. It is concluded that a good control optimizing the production rates. Model-free (developed)
design for the topside control choke valve can reduce the controllers have proven to be effective as well as they are
required amount of injection gas, and thus a combined data driven and hence only depending on relevant mea-
MIMO control design for gas lifting and topside choke surements. However, besides requiring ad hoc parameter
valve can be a more optimal solution. tunings, the model-free developed solutions often face to
the time-delay challenge due to the fact that the slug has
MIMO Slug Control The two most available actuators to occur before any feedback controller can react, while the
are the topside choke valve and the external gas lift- model-based control solution can enjoy the big advantage
ing. They have the possibility of being combined in a of slug prediction.

94
IFAC Oilfield 2015
May 27-29, 2015. Florianópolis, Brazil Simon Pedersen et al. / IFAC-PapersOnLine 48-6 (2015) 089–096 95

Through different slugging mechanisms, modeling ap- IFAC Workshop on Automatic Control in Offshore Oil and Gas
proaches and anti-slug control have been extensively in- Production, 122–127.
vestigated over several decades by both academic and Godhavn, J.M., Fard, M.P., and Fuchs, P.H. (2005). New slug
control strategies, tuning rules and experimental results. Journal
industrial societies, these topics are still quite open with
of Process Control, 15, 547557.
many related challenges. Grimstad, B. and Foss, B. (2014). A nonlinear, adaptive observer
for gas-lift wells operating under slowly varying reservoir pressure.
ACKNOWLEDGEMENTS 19th World Congress, The International Federation of Automatic
Control (IFAC’14), Cape Town, South Africa, 2824–2829.
The authors would like to thank the support from the Hamathy, T. (1960). Velocities of large drops and bubbles in media
Danish National Advanced Technology Foundation (via of infinite or restricted extent. AIChE Journal, 6, 281–288.
Hassanein, T. and Fairhurst, P. (1998). Challenges in the mechanical
PDPWAC Project (J.nr. 95-2012-3)). Thanks also go to
and hydraulic aspects of riser design for deep water developments.
our colleagues J.P. Stigkær, A. Aillos, K. G. Nielsen, IBC UK Conf. Ltd. Offshore Pipeline Technology Conference.
M. Haubjerg and P. Molinari from Maersk Oil A/S, our Havre, K. and Dalsmo, M. (2001). Active feedback control as the
colleagues P. Sørensen, A. Andreasen, B. Løhndorf, S. A. solution to severe slugging. SPE Annual Technical Conference &
Meybodi and J. M. Holm from Ramboll Oil & Gas A/S, Exhibition, New Orleans, Louisiana.
and H. Enevoldsen and M. Soltani from AAU, for many Helgesen, A.H. (2010). Anti-slug control of two-phase flow in risers
valuable discussions and technical supports. with: Controllability analysis using alternative measurements.
NTNU.
REFERENCES Hu, B. (2004). Characterizing gas-lift instabilities. Ph.D. thesis,
Norwegian University of Science and Technology (NTNU), De-
Aamo, O., Eikrem, G., Siahaan, H., and Foss, B. (2004). Observer partment of Petroleum Engineering and Applied Geophysics.
design for multiphase flow in vertical pipes with gas-lift - theory Jahanshahi, E., Oliveira, V.D., Grimholt, C., and Skogestad, S.
and experiments. Journal of Process Control. (2014). A comparison between internal model control, optimal
Abardeh, M.E. (2013). Robust control solutions for stabilizing pidf and robust controllers for unstable flow in risers. 19th World
flow from the reservoir: S-Riser experiments. Master’s thesis, Congress, The International Federation of Automatic Control
Norwegian University of Science and Technology, Department of (IFAC’14), 5752–5759.
Chemical Engineering. Jahanshahi, E., Skogestad, S., and Grtli, E.I. (2013a). Anti-slug
Adedigba, A.G. (2007). Two-phase flow of gasliquid mixtures in control experiments using nonlinear observers. American Control
horizontal helical pipes. Ph.D. thesis, Cranfield University. Conference (ACC), Washington DC, USA, 1058–1064.
Almeida, A. and Gonçalves, M. (1999). Device and method for elim- Jahanshahi, E. (2013). Control Solutions for Multiphase Flow - Lin-
inating severe slugging in multiphase-stream flow lines. patent: ear and nonlinear approaches to anti-slug control. Ph.D. thesis,
Us6041803a. Norwegian University of Science and Technology, The Faculty
Asheim, H. (1988). Criteria for gas-lift stability. Journ of Petr.Tech., of Natural Sciences and Technology, Department of Chemical
1452–1456. Engineering.
Biltoft, J., Hansen, L., Pedersen, S., and Yang, Z. (2013). Recreating Jahanshahi, E. and Skogestad, S. (2011). Simplified dynamical
riser slugging flow based on an economic lab-sized setup. IFAC models for control of severe slugging in multiphase risers. 18th
International Workshop on Periodic Control, 5th, 47–52. IFAC World Congress, 1634–1639.
Bøe, A. (1981). Severe slugging characteristics. Course in Two Phase Jahanshahi, E., Skogestad, S., and Grtli, E.I. (2013b). Nonlinear
Flow, NTH, Trondheim, Norway. model-based control of two-phase flow in risers by feedback
Di-Meglio, F., Kaasa, G.O., and Petit, N. (2009). A first principle linearization. IFAC Symposium on Nonlinear Control Systems,
model for multiphase slugging flow in vertical risers. Joint 48th 9th, 301–306.
IEEE Conference on Decision and Control and 28th Chinese Jahanshahi, E., Skogestad, S., and Helgesen, A.H. (2012). Control-
Control Conference, 8244–8251. lability analysis of severe slugging in well-pipeline-riser systems.
Di-Meglio, F., Kaasa, G.O., Petit, N., and Alstad, V. (2011). Slug- IFAC Workshop on Automatic Control in Offshore Oil and Gas
ging in multiphase flow as a mixed initial-boundary value problem Production, NTNU, Trondheim, Norway, 101–108.
for a quasilinear hyperbolic system. American Control Conference Jansen, F.E. (1990). Elimination of severe slugging in a pipeline-riser
on O’Farrell Street, San Francisco, CA, USA. system. M.S. Thesis, University of Tulsa.
Di-Meglio, F., Kaasa, G.O., Petit, N., and Alstad, V. (2012a). Jansen, F.E., Shoham, O., and Taitel, Y. (1996). The elimination of
Model-based control of slugging: Advances and challenges. 2012 severe slugging - experiments and modeling. Int. J. Multiphase
IFAC Workshop on Automatic Control in Offshore Oil and Gas Flow, 22, 1055–1072.
Production, 109–115. Jepsen, K., Hansen, L., Mai, C., and Yang, Z. (2013). Emulation and
Di-Meglio, F., Petit, N., Alstadb, V., and Kaasab, G.O. (2012b). control of slugging flows in a gas-lifted offshore oil production well
Stabilization of slugging in oil production facilities with or without through a lab-sized facility. IEEE Multi-conference on Systems
upstream pressure sensors. Journal of Process Control, 22, 809– and Control.
822. Johal, K.S. and Cousins, A.R. (1999). Intelligent production riser.
Doren, J.F.V., den Hof, P.M.V., Jansen, J.D., and Bosgra, O.H. Kaasa, G.O. and V. Alstad, e.a. (2008). Attenuation of slugging in
(2011). Parameter identification in large-scale models for oil and unstable oil wells. Nonlinear Control. 17th IFAC World Congress,
gas production. 18th IFAC World Congress, Milano, Italy. Seoul, Korea.
Eikrem, G.O. (2006). Stabilization of gas-lift wells by feedback Kovalev, K. (2003). The slug suppression system in operation.
control. Ph.D. thesis, Department of Engineering Cybernetics, Society of Petroleum Engineering (SPE), Offshore Europe 2003,
Norwegian University of Science and Technology. Aberdeen, UK.
Eikrem, G.O. (2008). Eikrem riser model. URL Krima, H., Cao, Y., and Lao, L. (2012). Gas injection for hydro-
http://www.nt.ntnu.no/users dynamic slug control. IFAC Workshop on Automatic Control in
/skoge/diplom/prosjekt08/tuvnes/ Offshore Oil and Gas Production, 1, 116–121.
Eikrem risermodel matlab/. Liao, Q. and Zhao, T. (2003). Modeling of taylor bubble rising in
Eikrem, G.O., Imsland, L., and Foss, B. (2004). Stabilization of gas a vertical mini noncircular channel lled with a stagnant liquid.
lifted wells based on state estimation. IFAC. International Journal of Multiphase Flow, 29, 411434.
Enricone Stasiak, M., Pagano, D.J., and Plucenio, A. (2012). A
new discrete slug-flow controller for production pipeline risers.

95
IFAC Oilfield 2015
96
May 27-29, 2015. Florianópolis, Brazil Simon Pedersen et al. / IFAC-PapersOnLine 48-6 (2015) 089–096

Makogan, B. (2007). Patent on device for controlling slugging. Skogestad, S. and Postlethwaite, I. (2005). Multivariable feedback
http://patentscope.wipo.int/ control: Analysis and design. Wiley & Sons, Chichester, West
search/en/WO2007034142. URL http://patentscope.wipo.int/ Sussex, UK.
search/en/WO2007034142. Storkaas, E. (2005). Anti-slug control in pipeline-riser systems.
Mansoori, M., den Hof, P.M.J.V., Jansen, J.D., and Rashtchian, D. Ph.D. thesis, Norwegian University of Science and Technology.
(2014). Pressure transient analysis of bottomhole pressure and Storkaas, E. and Skogestad, S. (2008). Controllability analysis of two-
rate measurements using system identification techniques. URL phase pipeline-riser systems at riser slugging conditions. Control
http://pvandenhof.nl/publications.htm. Engineering Practice, 15, 567–581.
McGuiness, M. and Cooke, D. (1993). Partial stabilization at st. Storkaas, E., Skogestad, S., and Godhavn, J.M. (2003). A low-
joseph. Proceeding of the 3rd International Offshore and Polar dimensional dynamic model of severe slugging for control design
Engineering Conference, 235–241. and analysis.
Montgomery, J. (2002). Severe Slugging and unstable flows in an Taitel, Y., Barnea, D., and Dukler, A.E. (1980). Modeling flow
S-shaped riser. Ph.D. thesis, Cranfield University, England. pattern transitions for steady upward gasliquid flow in vertical
Nemoto, R.H. and Bali no, J.L. (2012). Modeling and simulation of tubes. AIChE Journal, 345–354.
severe slugging with mass transfer effects. International Journal Taitel, Y. and Dukler, A.E. (1976). A model for predicting flow
of Multiphase Flow, 40, 144–157. regime transitions in horizontal and near horizontal gas-liquid
Ogazi, A., Cao, Y., Yeung, H., and Lao, L. (2009). Slug control flow. AIChE Journal, 22, 47–55.
with large valve opening to maximise oil production. Society Taitel, Y., Vierkandt, S., Shoham, O., and Brill, J.P. (1990). Severe
of Petroleum Engineers (SPE), Offshore Europe Oil & Gas slugging in a riser system: experiments and modeling. Int. J.
Conference & Exhibition held in Aberdeen, UK. Multiphase Flow, 16, 57–68.
Ogazi, A.I. (2011). Multiphase Severe Slug Flow Control. Ph.D. Taitel, Y. (1986). Stability of severe slugging. International Journal
thesis, Cranfield University, School of Engineering, Department of Multiphase Flow, 12, 203–217.
of Offshore, Process and Energy Engineering. Tchambak, E. (2004). Mitigation of severe slugging using gas injec-
Pagano, D.J., Plucenio, A., Traple, A., and Gonzaga, C.A. (2008). tion. Interim PhD review, Report No. 2, Cranfield University,
Controlling oscillations and re-starting ooperation in gas-lift wells. England.
XVII Congresso Brasileiro de Automtica, Juiz de Fora, Brasil. Tengesdal, J.O., Sarica, C., and Thompson, L. (2002). Severe
Pedersen, S., Durdevic, P., and Yang, Z. (2014a). Learning control slugging attenuation for deepwater multiphase pipeline and riser
for riser-slug elimination and production-rate optimization for an systems. PE Annual Technical Conference and Exhibition, Paper
offshore oil and gas production process. The 19th World Congress SPE 87089.
of the International Federation of Automatic Control. van Essen, G.M., den Hof, P.M.V., and Jansen, J.D. (2009). Hierar-
Pedersen, S., Stampe, K., Pedersen, S.L., Durdevic, P., and Yang, chical economic optimization of oil production from petroleum
Z. (2014b). Experimental study of stable surfaces for anti- reservoirs. International Symposium on Advanced Control of
slug control in multi-phase flow. International Conference on Chemical Processes (ADCHEM 2009), Istanbul, Turkey.
Automation & Computing, Cranfield University, 20. van Essen, G., Zandvliet, M., and Jansen, J. (2006). Robust opti-
Plucenio, A., Ganzaroli, C.A., and Pagano, D.J. (2012). Stabilizing mization of oil reservoir flooding. Computer Aided Control System
gas-lift well dynamics with free operating point. Proceedings of Design, IEEE International Conference on Control Applications.
the 2012 IFAC Workshop on Automatic Control in Offshore Oil Viggiani, M., Mariani, O., Battarra, V., Annunziato, A., and Bollet-
and Gas Production, 95–100. tini, U. (1988). Pipeline simulation interest group (psig) - a model
Sarica, C. and Shoham, O. (1991). A simplified transient model for to verify the onset of severe slugging. PSIG Annual Meeting.
pipeline-riser systems. Chemical Engineering Science, 46, 2167– Xing, L., Yeung, H., Shen, J., and Cao, Y. (2013). Experimental
2179. study on severe slugging mitigation by applying wavy pipes.
Sarica, C. and Tengesdal, J.. (2000). A new technique to eliminate Chemical Engineering Research and Design, 91, 1828.
severe slugging in pipeline/riser systems. SPE Annual Technical Yang, Z., Pedersen, S., and Durdevic, P. (2014). Cleaning the pro-
Conference & Exhibition, SPE 63185, 633–641. duced water in offshore oil production by using plant-wide optimal
Schmidt, Z., Brill, J.P., and Beggs, H.D. (1980). Experimental study control strategy. IEEE OCEANS’14 ST. JOHN’S Conference.
of severe slugging in a two-phase flow pipeline riser-pipe system. Yang, Z., Stigkær, J.P., and Løhndorf, B. (2013). Plant-wide control
SPE Journal, 20, 407–414. for better de-oiling of produced water in offshore oil & gas
Scibilia, F., Hovd, M., and Bitmead, R. (2008). Stabilization of gas- production. 3rd IFAC International Conference on Intelligent
lift oil wells using topside measurements. IFAC World Congress, Control and Automation Science, 3, 45–50.
17, 13907–13912. Yocum, B. (1973). Offshore riser slug flow avoidance: Mathematical
Silva, C.M.D. and Nydal, O. (2010). Dynamic multiphase flow model for design and optimization. SPE 4312, London, UK.
models for control. BHRG North American Conference on Zandvliet, M., Bosgra, O., van den Hof, P., Jansen, J., and Kraai-
Multiphase Technology. Banff, Canada, 7th. jevanger, J. (2006). Bang-bang control in reservoir flooding.
Sinègre, L. and Petit, N. (2006). Active control strategy for density- 10th European Conference on the Mathematics of Oil Recovery,
wave in gas-lifted wells. Proc. of the International Symposium on Amsterdam, The Netherlands.
Advanced Control of Chemical Processes.
Sinègre, L., Petit, N., and Menegatti, P. (2005). Distributed delay
model for density wave dynamics in gas lifted wells. IEEE
Conference on Decision and Control, and the European Control
Conference, 44.
Sinègre, L., Petit, N., and Menegatti, P. (2006). Predicting instabil-
ities in gas-lifted wells simulation. American Control Conference,
Minnesota, USA, 5530–5537.
Sivertsen, H. and Skogestad, S. (2005). Cascade control experiments
of riser slug flow using topside measurements. Triennial World
Congress, Prague, Czech Republic, 16th, 6.
Sivertsen, H., Storkaas, E., and Skogestad, S. (2010). Small-scale
experiments on stabilizing riser slug ow. Chemical Engineering
Research and Design, 213–228.

96