SPE-201509-MS

A Stability Criterion for Gas-Liquid Flow in a Well-Flowline-Separator

System

Oscar Moran Ochoa, Sergio Vázquez Nolasco, and Ignacio Díaz-Soto Pérez, PEMEX Exploration & Production;
Victor Fairuzov and Yuri Fairuzov, Multiphase Energy Corporation

Copyright 2020, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Annual Technical Conference & Exhibition originally scheduled to be held in Denver, Colorado, USA, 5 – 7
October 2020. Due to COVID-19 the physical event was postponed until 26 – 29 October 2020 and was changed to a virtual event. The official proceedings were
published online on 21 October 2020.

This paper was selected for presentation by an SPE 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 Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written
consent of the Society of Petroleum Engineers 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 SPE copyright.

Abstract
Oil and gas gathering systems can experience flow instabilities characterized by cyclic pressure variations
and liquid or free water surges out of the flowlines that typically cause liquid or gas handling problems
in receiving facilities and dynamic external loads in elbows and T-junctions. This study describes a new
mechanism of instability of gas-liquid flow in a well-flowline-receiving facility system referred to in
other industrial applications as the density or liquid holdup wave oscillations (DWO). Unlike instability
phenomena related to hydrodynamic slugs, terrain-induced slugs or riser-induced severe slugging, this
type of instability is a result of multiple feedback effects between the flow rate, pressure drop, and liquid
inventory (the total volume) in the flowline. The specific DWO instability mechanism proposed in the
present study is related to generation and propagation of a liquid holdup wave with a large amplitude
and wavelength several times greater than the length of a hydrodynamic or terrain-induced liquid slug.
A new stability criterion to predict the DWO onset is proposed based on Ultra-High Definition (UHD)
simulations of three-phase gas-oil-water flow to accurately predict the liquid inventory. Field data showing
a transition from unstable to stable operation in an offshore oil production system are presented. The system
is used to transport produced fluids from a satellite platform with naturally flowing and gas-lifted wells to
a central platform with processing facilities, via a 30-in. diameter flowline. The stability criterion was used
to identify the root cause of large-amplitude oscillations of pressure and flow rate in the flowline. Thus,
proposed dynamic two-phase flow instability mechanism and criterion are a cost-effective and practical
method for predicting instability onset conditions and developing mitigation strategies in existing and new-
built production facilities.

Introduction
Oil and gas gathering systems can experience flow instabilities characterized by cyclic pressure variations
and liquid or free water surges out of the flowlines that typically cause liquid or gas handling problems
in receiving facilities and dynamic external loads in elbows and T-junctions. Several mechanisms, such as
hydrodynamic (normal) slugs, terrain-induced slugs, and riser-induced severe slugging, were proposed in
2 SPE-201509-MS

previous studies to provide a reasonable explanation for the unstable flow in both onshore and offshore
gathering systems. Pedersen et al. (2015) provide a review of slug detection, modeling and control
techniques for offshore oil and gas production processes. The traditional method of eliminating slugs is
manually choking flow at the riser outlet until stable operation is achieved, Campos et al (2015). The
application of this method is straightforward in the existing facilities. However, predicting the maximum
valve opening that enables a stable operation is a challenging task in the design of new production facilities.
We observed large flow rate oscillations in many offshore oil gathering systems with large diameter
flowlines (24-in., 30-in., and 36-in.) connected to the first stage separator, with liquid and subsequently gas
entering the separator throughout a period of a few tens of seconds and minutes, respectively. This instability
cannot be attributed to severe slugging since the gas superficial velocity in the flowlines was much greater
than 1 m/s (Fig.1). Neither numerical simulations of transient gas-liquid flow, Bendiksen (1982) and nor
the modeling approach proposed in a previous study for long hydrodynamic slugs, Guzman and Fairuzov
(2009), enabled us to identify the root cause of the observed oscillations. The predicted oscillation frequency
was too high or too low. Also, the flow pattern in the flowline was slug flow before and after choking the
flow, therefore the achieved stable operation was not associated to a flow pattern transition.

Figure 1—Comparison of Bøe criterion and other criteria for severe slugging, Jansen et al. (1996)

Two-phase flow instabilities in boiling water reactors (BWR) were extensively studied in the nuclear
industry. A comprehensive review of previous research done in this field is provided by Ruspini (2013). We
are citing here several examples and important ideas from this review; these will be referred to in subsequent
sections of this paper. Firstly, two-phase flow instabilities are classified into two types: microscopic and
macroscopic instabilities. The term "microscopic instabilities" describes "the phenomena occurring locally
at the liquid-gas interface"; for example, the Kelvin-Helmholtz instability causing the flow regime transition
from stratified to slug flow. The term "macroscopic instabilities" is used to describe instability mechanisms
involving "the entire two-phase flow system". Secondly, the macroscopic instabilities can be static or
dynamic, according to the most widely accepted classification proposed by Bouré et al. (1973). The
onset of static instability can be predicted "using the steady-state conservation laws." To predict dynamic
SPE-201509-MS 3

instabilities, in contrast, it is necessary to consider "different dynamic effects, such as the propagation time,
the inertia, compressibility, etc."
Ledinegg instability is a well-known example of a static instability in a boiling channel. This phenomenon
is associated with two types of hydraulic characteristic curves, the system’s curve or "internal curve" and the
relationship between the pressure drop and flow rate of the external device (e.g., a pump or a combination
of inlet and outlet manifolds) or "external curve", as shown in Fig.2. The internal curve of a boiling channel
has an S-shape and several external curves are also plotted in this figure. The points of intersection represent
the steady-state operation points, which can be stable or unstable. By definition, "an operation point is called
stable when the slope of the internal characteristic is smaller than the slope of the external characteristic
curve." As can be seen, the external curves of the case 1 (constant flow, e.g. choke) and case 2 (positive
displacement pump) intersect the internal curve in only one point (1). As the internal curve slope is larger
than the external curve slope, then this point is stable. The external curves of case 3 (centrifugal pump) and
case 4 (inlet and outlet manifolds with constant pressures) intersect the internal curve at the same point (1).
In these two cases the point (1) becomes unstable. The static methodology is a convenient and practical
method for identifying flow instabilities in two-phase systems.

Figure 2—Internal pressure drop vs. flow rate characteristic curve for a boiling
system and five different external characteristics curves (cases), Ruspini (2013)

(1) stability
condition
Density Wave Oscillations (DWO) are the most representative type of dynamic instabilities in the boiling
systems. This instability mode results from multiple feedback effects between the flow rate, the void fraction
(or liquid holdup) and the pressure drops in the boiling channel. According to the classical interpretation
of DWO proposed by Yadigaroglu & Bergles (1972) the origin of the instability is attributed to waves of
heavier and lighter fluids, and respective delays through the channel. More specifically, "the difference in
density between the fluid entering the heated channel (subcooled liquid) and the fluid exiting (low density
two-phase mixture) triggers delays in the transient distribution of pressure drops along the tube, which may
induce self-sustained oscillations. A constant pressure drop (or better, the same, not necessarily constant
4 SPE-201509-MS

with time, pressure drop for the multiple parallel channels) is the proper boundary condition that can excite
those dynamic feedbacks which are at the source of the instability mechanism", Papini et al. (2011).
In this paper, the methods developed in the nuclear industry for the analysis of macroscopic instabilities
in boiling water reactors (BWR) are extended to flow instability problems in offshore oil gathering systems
with large-diameter flowlines and naturally flowing and gas-lifted wells.

Description of a Well-Flowline-Separator System

The gathering system of an offshore oil field comprises several large diameter flowlines that are used to
transport produced fluids (Table 1) from satellite platforms (export platforms) to central platforms with
processing facilities (import platforms). Most of wells are naturally flowing and gas lift is used to optimize
production in some wells. The main elements of the offshore oil production system (in Figure 3) are

• Wells connected to a production manifold

• Flowline

• Separator with a manifold at the inlet

• Topside choke valve (manual)

Figure 3—A schematic diagram of a well-flowline-separator-system

The elevation profile of a 30-in diameter flowline of the oil gathering system is shown in Fig.4.
SPE-201509-MS 5

Figure 4—Elevation profile of a 30-in diameter flowline of the oil gathering system

Fig. 5 shows pressure trends upstream and downstream of the choke valve from a real-time data report
for a period during which the choke valve opening (area) was reduced from a larger value (Position 1) to
a smaller value (Position 2). Setting the choke valve at Position 1 triggered severe instability characterized
by very large amplitude, 8 kg/cm2, pressure oscillations that are attributed to variations of liquid flow rate
through the valve. When the fluid flowing through the valve comprises mostly liquid, the upstream pressure
is high, since the liquid density is much greater than the gas density, and vice versa, the upstream pressure
is low when mostly gas flows through the valve. The downstream pressure exhibits a similar trend because
there is an automatic control valve at the separator inlet (not shown in Fig.3).

Figure 5—Pressure trends upstream and downstream of the choke valve from a real-time data report

The frequency of large amplitude oscillations is approximately 2 times per hour. A stable operation
condition was achieved after the average upstream pressure was increased from 18 kg/cm2 to 22 kg/cm2 by
setting the choke valve at Position 2. The average downstream pressure did not change significantly after
chocking the flow and was about 12 kg/cm2. The low amplitude and high frequency pressure oscillations
did not represent a problem for controlling the liquid level at the separator.
6 SPE-201509-MS

System Model
A model of the well-flowline-separator system (Fig.3, Table 1) was constructed using a software
application that contains the broad, physical modeling capabilities needed to simulate, visualize, and analyze
simultaneous flow of all types of fluids and solids (gas, all types of liquid hydrocarbons, water, sand,
proppant, sludge, black powder, elemental sulfur, drilling mud, and control fluids) encountered in oil and
gas production operations Fairuzov and Fairuzov (2019). It is based on an advanced mechanistic modeling
technique that has the following two features important for gas-liquid flow stability studies:

• allows simulations using an accurate (real, not simplified) and complete elevation profile,

• predicts simultaneously a gas-liquid flow pattern and an oil-water flow pattern in the liquid at each
location.
The construction of a pipeline elevation profile is performed considering the elastic deformation of the
pipeline in the longitudinal direction. This is achieved by ensuring that the first derivative of the elevation
with respect to the distance is a continuous function, Fairuzov and Fairuzov (2020). The average mesh cell
size smaller than the pipe diameter is required in the pipeline numerical model to meet this requirement in
most cases. We use the term Ultra-High Definition Simulation (UHD) to refer to this modeling technique.

Table 1—Key parameters of the model of the well-flowline-separator system

Transported fluids Gas, oil, water

Flow patterns predicted Five gas-liquid flow patterns (Stratified/Drop Flow; Elongated Bubble Flow; Slug Flow; Annular/
Pulsating Wavy Flow; Dispersed Bubble Flow) and five liquid-liquid flow patterns in the liquid phase
(stratified flow, dispersion of oil-in-water and water, oil-in-water emulsion, dispersions of water-in-oil and
oil-in-water, water-in-oil emulsion)

Number of sections in the provided 21506

elevation profile

Sum of elevation changes in uphill sections 43.103 m

of pipeline laid on the seabed or below it
inside the trench

Minimum cell size 0.36 m

Maximum cell size 0.72 m

Total number of cells 43302

Oil API gravity at 60°F 32.5

Gas Specific Gravity 1.07

Production GOR 258.15 m3/m3

TDS 200000 mg/l

Water cut 3%

Well model IPR+VLP

Choke valve model Δp=K ρ m v m 2 2

Separator model Sink with a constant pressure

Comparisons of other software used in the industry, Bendiksen (1982), with field measurements show that
there may be both large under-and over-predictions of the total liquid volumes (inventories) in flowlines,
Flaten (2008). To resolve this problem, the modeling technique used in the present work was validated using
six groups of different data types, including full-scale experiments, historical production data, historical
coupon data, corrosion coupon installation and retrieval reports, thermography (Fig.6), scale-model tests,
and ILI-survey data.
SPE-201509-MS 7

Figure 6—Comparison of predicted liquid holdup to thermography results for a 20-in. diameter flowline on the export platform

Fig.7 shows a 3D visualization and predicted phase holdup profiles in a 20-in.diamter flowline of the
field.

Figure 7—UHD simulation of a 20-in. diameter flowline

8 SPE-201509-MS

Results and Discussion

Static stability analysis
The static methodology used to analyze the Ledinegg instability was applied to identify root cause of
large amplitude pressure (and the liquid flow rate) oscillations shown in Fig. 5. The internal characteristics
(curves) were predicted using the system model for two choke valve openings (Fig.8): completely open
(K=0) and the valve opening at which the stable operation condition was achieved (K=752, Position 2 in
Fig.5). The external curve is a combination of IPR and VLP (outflow curve) of the wells connected to the
production manifold.

Figure 8—Internal curves for K=753, K=0 (completely open choke valve) and
external curve (IPR+VLP, outflow) for the well-flowline-separator system

As can be seen, the external curves for K=752 and K=0 intersect the internal curve in point (1) and (2),
respectively. As the internal curve slope is larger than the external curve slope in both cases, then these points
are stable, Eq.(1). Thus, the system is always stable according to the static stability analysis, independently
on the choke valve opening. Therefore, the observed large amplitude pressure oscillations were caused by
another type of two-phase flow instability.

Dynamic stability analysis

In BWR safety analysis, density wave instability (DWO) was found in experiments conducted in upward
vertical systems with a long unheated riser section downstream from the heated section by Fukuda and
Kobori (1979). The well-flowline-separator analyzed in the present work has a similar configuration,
therefore it is reasonable to assume that this phenomenon may occur in oil and gas production systems.
DWO is a result of multiple dynamic feedback effects between flow rate, void fraction (or liquid holdup),
and pressure drop in the entire system. Fig. 9 illustrates the proposed mechanism, which takes into account
dynamic effects related to generation and propagation of density (liquid holdup) waves and the feedback
processes that control the inlet conditions. Initially, the system is at equilibrium, with a liquid holdup profile
corresponding to a certain pressure in the outlet manifold and a flow rate or ΔP, as shown in Fig. 9a.
Then, a positive perturbation of the outlet pressure is generated, for example, by the system that controls
liquid level in the separator. In large diameter flowlines, the difference between the pressures at the riser
bases of the export and import platforms is relatively small (a few kg/cm2), so the generated perturbation
may reduce it several times or make it negative. The speed of sound in methane is 450 m/s, therefore
SPE-201509-MS 9

the pressure perturbation reaches the export platform in less than one minute. This leads to a substantial
reduction in the gas velocity in the flowline. As the result, the liquid accumulated in the flowline flows
back toward the export platform generating one liquid holdup wave with a large amplitude and wavelength
several times greater than the length of a hydrodynamic or terrain-induced liquid slug, as shown in Fig. 9b.
After the positive perturbation of the outlet pressure is removed, the pressure downstream of the formed
liquid holdup wave decreases in less than one minute. The gas is accelerated to the velocity at which it flowed
when the system was at equilibrium, making the liquid holdup wave move toward the import platform.
This holdup wave is not a hydrodynamic or terrain-induced liquid slug, because different flow patterns
simultaneously occur within the wave at different locations (e.g., stratified flow and slug flow), depending
on the pipeline slope and local liquid flow rate. Behind the wave, a quasi-steady liquid holdup profile is
developed similar to that produced when the system is at equilibrium, as shown in Fig. 9c.
Once the holdup wave reaches the separator, the liquid level in it increases due to excessively high
instantaneous liquid-flow rate. The automatic control valve restricts the flow at the separator inlet to control
the liquid level. This action creates the next positive perturbation of the outlet pressure and the process is
repeated producing self-sustained oscillations in the system.
The holdup wave propagation time can be estimated based on the equation of liquid mass conservation

(2)

Assuming that the liquid density variation is negligibly small, Eq.(2) can be rewritten in non-conservative
form as follows

(3)

The holdup wave celerity, c, in Eq.(3) may be several times (an order of magnitude) greater than the
liquid velocity, which is about 1 m/s at the operating conditions analyzed in the present work. Considering
that the length of the flowline is about 20000 m (Fig.4), the estimated holdup wave propagation time is
about 2000 s, which corresponds to the observed frequency of large amplitude oscillations in Fig.5.
It is important to note that the actual shape of the density wave shown in Fig.9 is much more complex
(irregular), as shown in Fairuzov and Fairuzov (2020). The simplification is used for clarity purposes only.
10 SPE-201509-MS

Figure 9—Density wave instability in a well-flowline-separator system

Stability criterion for predicting the DWO onset

The DWO onset can be predicted using UHD numerical modeling, which provides accurate predictions
of transient liquid holdup profiles based on the real (not simplified) elevation profile of the flowline. An
SPE-201509-MS 11

example of the application of this modeling technique to simulate water holdup wave propagation in an oil
pipeline in presented the previous study, Fairuzov and Fairuzov (2020).
An alternative approach is shown in Fig.10, which illustrates the predicted relationship between the total
pressure difference in the system, ΔP, and the liquid inventory in the flowline for the two choke valve
openings K=0 and K=752, Position 2 in Fig.5. The arrow shows how the flow rate changes along the
predicted curves. As can be seen, absolute values of the slopes of the curve for K=0 are great at all its point.
This means that any positive outlet pressure perturbation will lead to the formation of a large density wave
in the flowline, i.e. the system is never stable when the choke valve is open. The curve for K=752 has a
minimum point at ΔP of about 13 kg/cm2. DWO occur on the left from the minimum. The system is stable
on the right from the minimum, where a positive outlet pressure perturbation decreases or does not change
the liquid inventory. The field data presented in Fig.5 confirm this predicted result.

Figure 10—Predicted relationship between the total pressure difference in the system and the liquid inventory

The stability criterion can be formulated as follow

(4) stability
condition
In less conservative form, it can be defined, considering that the separator can handle instabilities in some
degree, as follows

(5)

C is a negative constant depending on design parameters (the turndown ratio) of the separator.
A stability map can be constructed to predict the instability threshold for different outlet pressures, as
shown in Fig.11. In this example, the map was used to develop mitigation strategies for a new separator
to be connected to the 20-in. diameter flowline of the field. As can be seen, this system will operate
within the unstable region beginning from March 2023. Increasing the production gas-oil ratio (GOR) to
a predetermined value using lift-gas injection was proposed to ensure flow stability and avoid a high back
pressure in the production manifold.
12 SPE-201509-MS

Figure 11—DWO instability threshold for a 20-in. diameter flowline of the field

According to Eq.(4), upward inclined flowlines or flowlines having a greater sum of elevation changes
in uphill sections of pipeline laid on the seabed or below it inside the trench (Table 1) are more susceptible
to the DWO instability than those that are downward inclined or horizontal. It is interesting to note that the
proposed criterion explains the experimental results of Fabre et al. (1990) who observed severe slugging in
a relatively short horizontal pipeline-riser system. The existing models and software could not reproduce
these results. In small simple systems, local phenomena become global, therefore the proposed stability
criterion does predict the onset of sever slugging in the pipeline-riser system with a horizontal pipe.

Conclusions
Static and dynamic stability analysis of two-phase flow in a well-flowline-separator system was performed
to identify the root cause of large diameter pressure and flow rate oscillations at the inlet of the separator
observed in the field. The results of the study revealed the following:
1. Static instability does not occur in the system
2. The observed large amplitude oscillations are attributed to the instability mechanism referred to in
other industrial applications as the density wave oscillations (DWO).
3. Existing methods and software used in the oil and gas industry fail to predict this type of instabilities
because hydrodynamic slugs, terrain-induced slugs or riser-induced severe slugging are predicted
based on modeling local phenomena, microscopic instability, while DWO is a macroscopic instability
that involves the entire production system.
4. The specific DWO instability mechanism proposed in this work is related to generation and
propagation of at least one liquid holdup wave with a large amplitude and wavelength several times
greater than the length of a hydrodynamic or terrain-induced liquid slug.
5. A new stability criterion to predict the DWO onset was proposed. It depends on the well and/or
flowline characteristic curve (the system curve), the liquid inventory, and the turndown ratio of the
first stage separator.
SPE-201509-MS 13

The proposed dynamic two-phase flow instability mechanism and criterion are a cost-effective and
practical method for predicting instability onset conditions and developing mitigation strategies in existing
and new-built production facilities.

SI Metric Conversion Factors

To Convert From To Multiply By

barrel (for petroleum, 42 gal) meter3 (m3) 1.589 873 E−0
ft/s m/s 3.048* E−01
foot m 3.048* E−01
lbf/in.2 (psi) pascal (Pa) 6.894 757 E+03
kgf/cm2 pascal (Pa) 9.806 65* E+04
An asterisk (*) after the numbers shown indicates that the conversion factor is exact and. that all subsequent digits (for rounding purposes) are zero.

Acknowledgments
The authors thank PEMEX Exploration & Production (PEP) for permission to publish this work. Sincere
thanks are extended to Julio Yucundo Ruíz Aguilar (Operation Coordinator, PEP) for his very stimulating
and helpful discussions of the topic. The vapor-liquid equilibrium (VLE) calculations used for compositional
models of production streams coming from seven oil fields were carried out by Daniel Rodriguez Montero
(Facilities Design Unit, PEP).

Nomenclature
BWR boiling water reactor
C constant defined in Eq.(5)
c holdup wave celerity
DWO density wave oscillations
G mass flow rate
GOR production gas-oil ratio
IPR inflow performance relationship
Mbbl thousands of barrels
VLP vertical lift performance
ΔP total pressure drop in the system (the pressure difference between the production manifold
and manifold at the separator inlet)
K resistance coefficient of the choke valve
ρl liquid density
Hl liquid holdup
Vl total liquid volume in the system (liquid inventory)
vl liquid velocity
t time
x axial coordinate
Γ liquid generation rate per unit volume
14 SPE-201509-MS

References
Bendiksen, K. H., Maines, D., Moe, R., & Nuland, S. 1991. The Dynamic Two-Fluid Model OLGA: Theory and
Application. Society of Petroleum Engineers. doi: 10.2118/19451-PA.
Bouré J., Bergles A., and Tong L. 1973. Review of two-phase flow instabilities. Nuclear Engineering and Design, 25:
165–192.
Guzman Vazquez, E., & Fairuzov, Y. V. 2009. A Study Of Normal Slug Flow In An Offshore Production Facility With a
Large Diameter Flowline. SPE Production & Operations, Volume 24, Issue 01, doi: 10.2118/108752-PA.
Campos, M. Takahashi, T., Ashikawa, F. et al. 2015. Advanced Anti-slug Control for Offshore Production Plants. Presented
at 2nd IFAC Workshop on Automatic Control in Offshore Oil and Gas Production, Florianopolis, Brazil, May 27-29.
Fabre, J., Peresson, J., Corteville, L. L., Odello, R., and Bourgeols, T. 1990. Severe Slugging in Pipeline/Riser Systems.
SPE Production Eng. J., 5 (3): 299–305.
Fairuzov, Y. and Fairuzov, V., 2019. Diagnosis of Internal Corrosion in Pipelines Based on Mapping Adverse Operational
Conditions, Presented at NACE International Annual Conference & Expo CORROSION 2019 March 25, 2019,
Nashville, Tennessee, Pipeline Integrity, C2019-13250.
Fairuzov, Y. and Fairuzov, V., 2020. Detecting and Preventing Internal Corrosion Damage in Unpiggable, Intermittently-
Operated, Crude-Oil Pipelines, NACE International Annual Conference & Expo CORROSION 2020, March 15-19,
2020, Houston, TX, USA, George R. Brown Convention Center Event Detail: Pipeline Integrity Symposium Doc
Number: C2020-14673 (published in the Digital Conference Proceedings available in the NACE store).
Flaten, G., 2008 Go with the Flow, Offshore Technology. http://www.offshore-technology.com/features/feature41077/.
Fukuda K. and Kobori T. 1979. Classification of two-phase flow instability by density wave oscillation model. Journal
of Nuclear and Technology, 16: 95–108.
Jensen, F. E., Shoham, O., and Taitel, Y., 1996. The Elimination of Severe Slugging-experiments and modeling", Int. J.
Multiphase Flow, 22, 1055–1072.
Papini, D., Cammi, A., Colombo et al. 2011. On Density Wave Instability Phenomena — Modelling and
Experimental Investigation, Two Phase Flow, Phase Change and Numerical Modeling, Dr. Amimul Ahsan (Ed.),
ISBN: 978-953-307-584-6, InTech, http://www.intechopen.com/books/two-phase-flow-phase-change-and-numerical-
modeling/on-density-waveinstability-.
Pedersen S. , Durdevic, P., and Yang. Z. 2015. Review of Slug Detection, Modeling and Control Techniques for Offshore
Oil and Gas Production Processes. Presented at 2nd IFAC Workshop on Automatic Control in Offshore Oil and Gas
Production, Florianopolis, Brazil, May 27-29.
Ruspini C., 2013. Experimental and numerical investigation on two-phase flow instabilities. Doctoral theses at NTNU,
ISBN 978-82-471-3405-4 (electronic version).
Yadigaroglu, G. & Bergles, A. E. 1972. Fundamental and Higher-Mode Density Wave Oscillations in Two-Phase Flow,
Journal of Heat Transfer — Transactions of the ASME, Vol.94, pp. 189–1.