Available online at www.sciencedirect.

com

Energy Procedia 37 (2013) 3018 – 3030

GHGT-11

Flow Assurance Study

Wolfgang Bösera,b*, Stefan Belfroidc
a
ROAD Maasvlakte CCS Project C.V., NL 3112 NA Schiedam, The Netherlands
b
E.ON Ruhrgas AG Brüsseler Platz 1, 45131 Essen, Germany
, cTNO, NL-2600 AD Delft, The Netherlands

Abstract

Generally large scale carbon capture projects require pipeline systems for the transporting of the CO2
from its point of capture to the storage site. The article will give information on the proposed operational
management system. This has to work for all process situations, ranging from steady flow at varying
injection conditions and flow rates, to start-up and shutdown procedures and also for emergency
shutdown at the platform. In all these operational situations the phase behaviour of CO2 and the process
conditions will be affected by differential pressures and the temperature losses in the pipeline. The study
was carried out on the specifics of the ROAD project and has to find operational procedures on the
intended conditions determined by the operation of a power plant and a storage side with changing
conditions at the storage side during the project lifetime.

©
© 2013 The
The Authors.
Authors.Published
PublishedbybyElsevier
Elsevier Ltd.
Ltd.
Selection and/or peer-review
Selection and/or peer-reviewunder
underresponsibility
responsibility
of of GHGT
GHGT

Keywords: CO2 transport, Pipeline, Flow Assurance, Operation, Phase Transition

1. Introduction

The Rotterdam Afvang en Opslag Demonstratie Project (ROAD) (www.road2020.nl) is a 250MWe

equivalent carbon capture and storage (CCS) demonstration project being developed by E.ON Benelux
and GDF SUEZ Energy Nederland. They initiated the joint venture Maasvlakte CCS Project C.V. (MCP).
Intended partners of MCP are GDF Suez E&P Nederland for the CO2 transport and TAQA Energy for the
CO2 injection and permanent storage under the seabed of the North Sea. The ROAD project is co-financed
by the European Commission within the framework of the European Energy Programme for Recovery
(EEPR), the Government of the Netherlands and the Global CCS Institute. The capture plant will be

* Corresponding author. Tel.: +49-209-6013232; fax: +49-201-184248178.

E-mail address: Wolfgang.Boeser@eon-ruhrgas.com.

1876-6102 © 2013 The Authors. Published by Elsevier Ltd.

Selection and/or peer-review under responsibility of GHGT
doi:10.1016/j.egypro.2013.06.188
 Wolfgang Böser and Stefan Belfroid / Energy Procedia 37 (2013) 3018 – 3030 3019

and capturing 90% of the CO2 from that gas. (Figure 1) This CO2 will be compressed, transported by
pipeline and permanently stored under the seabed just off the Dutch coastline in a depleted gas reservoir
operated by TAQA. Contracts needed for the project include those for the capture facility, the pipeline and
the storage of CO2.
The capture facility will be built under an agreed procurement and construction contract. Transport will
be via a new build on- and offs
TAQA will carry out all the work on the platform and well. All of the main permits, including those for
the handling of long-term storage liabilities, have been approved by the Dutch Government but public
consultation and consultation with the EU for the storage license is not yet complete.
Compression
and Cooling
dehydration, P18 A
CO2 O2--removal,
O2 Platform
Capture Metering
Maasvlakte
Maasvlakte Plant
Power Plant 3
Power Onshore
Transport
Pipeline Offshore
Transport
Subsea
Pipeline
p

5 km 20 km
3,5 km
3

Start of CCS CO2 Storage into depleted

d
scope gas reservoirs

(reservoir pressure 20 300 bar)

Figure 1: Schematic of process design.

Purpose of the Flow Assurance Study (FAS)

The goal of the FAS was to derive operational guidelines. This was done by dynamic modeling of the
network from compressor outlet to injection point. The interactions between Capture, Transport and
Storage had been investigated. Results are recommendations how to operate the system, what are
restrictions and how to handle them and what are the impacts of this information for the design of the
process. Operating risks were also identified.

The study was carried out with regard to the proposed depleted gas field in the North Sea 20 km from
Rotterdam harbor. Studies concluded operational situations are mainly determined by the storage side and
the well and tubing design.

Characteristics of the depleted gas field

The gas field compartments in block P18-A (P18-2, P18-4 and P18-6) are situated at approximately
3500 m depth below sea level. The reservoir rock is part of the Triassic Main Buntsandstein Subgroup and
the primary seal for the gas fields consists of disconformably overlaying siltstones, claystones, evaporites
and dolostones. The P18 gas fields including P18-4 are located in a heavily faulted area and consist
mainly of fault bounded compartments, which are (at least on production time scales) hydraulically
isolated from their surroundings.
The use of compartments P18-2 and P18-6 was rejected early in the project, The reason for 18-2 were
the higher cost and uncertainty due to the extra wells. Moreover, the compartment was probably not
3020 Wolfgang Böser and Stefan Belfroid / Energy Procedia 37 (2013) 3018 – 3030

available early enough due to ongoing gas production. For P18-6 the reasons were limited capacity
(approximately 1 Mton) and limited injectivity.

2. Design conditions

Below (Table 1) a short summary on the relevant design data of the project is given.

Table 1: Basic design of the transport and storage chain.

Max operational pressure 129 bar

Operating pressure range 40 85 129 bar
Pipeline
Design Pressure for calculation DP 175
Length 25 km
min. load ( 0 kg/s ) 18.8 kg/s
max load 47 kg/s
Installation
onshore Subsurface
coverage > 1m
Industrial area
offshore below sea bottom
coverage >1m
shallow water
Insulation 1.5 W/(m²K)
Platform Choke valve
Min temperature -10°C
Well
Min temperature > -10°C
Length 4200m
Depth 3300m
Reservoir P 18 4
Bottom hole
pressure 20 300 bar
temperature > 15°C

Based on the results some general guidelines for the design could be provided.
o Additional venting at the platform to reduce start-up times does not have any reasonable
advantages.
o To store CO2 at the envisaged storage side some basic requirements have to be fulfilled:
- Temperature of CO2 at the platform has to be above threshold values to ensure stable phase
conditions of CO2 in the pipeline at the platform.
- Guarantee that the temperature of CO2 in the well section is not be below thes specification
values of material and installations.
- To avoid hydrate forming the temperature of CO2 at the bottom hole of the storage side must
be above 15°C.
o To enable the system to meet these requirements installation of an additional heater at the
 Wolfgang Böser and Stefan Belfroid / Energy Procedia 37 (2013) 3018 – 3030 3021

platform was discussed.

o Besides the requirements in respect of energy, deck space, weight of the fuel in the case of an oil
driven heater needed deck space at the platform and mechanical stability, simulation showed the
limited benefit for operation. Instead of installation a heater at the platform it became obvious
that operational procedures should be ramped up. To limit start up times at low reservoir pressure
and to avoid restrictions by hydrate formation in the storage field, strategies using a heater
installed at the platform were analyzed. Based on restrictions determined by the platform the
heating capacity was limited to 4 MW. It was found that a 4 MW heater installed at the platform
would not have enough capacity to ensure temperatures adequately high during start-up in order
to reduce start-up time to less than a day. Only if the heater capacity were much higher it would
be beneficial. Looking at the increase of the reservoir pressure due to the injection rate, a heater
might be useful until the reservoir pressure is above 50 bar. Based on the proposed injection rate,
an initial pressure for the project of 20 bar and injectivity of the storage side this period would be
roughly less than 2 months. By developing start-up procedures which take longer (3.5 days at a
most) installation of a heater was discarded.
o The valve section at the platform should be suitable for low temperatures to withstand effects
caused by local pressure loses during opening of the valves and the Joule Thompson effect. By
using valves of an Arctic design operational flexibility was gained.

3. Simulations model

Due to changes on pressures temperatures and well design during basic engineering, the study had to
be constantly adapted. Based on the typical expected impurities for ROAD (Table 2) an analysis was
carried out to investigate effects of these impurities on the condensation line compared to 100% of CO2.
The effects found were minor and within the range of accuracy. It was decided to go on with simulation
by using pure CO2.

Table 2: Impurities of CO2 fluid for ROAD.

N2 02 H2O Acetaldehyde Ar
350 ppmv 40 ppmv 40 ppmv 10 ppmv 7 ppmv

Simulations of the entire system were performed for a range of mass flow rates, reservoir pressures
and inlet temperatures, starting from the inlet of the transport pipeline (onshore) up to the reservoir and
down in the well. To understand the dynamics and the pipeline flow the transient multiphase flow
simulator OLGA (SPT Group versions 7 and higher with different modules) was taken as the main
simulation tool for FAS for the project. Transportation of large volumes of CO 2 is preferably done in a
dense phase condition. Transmission at a gaseous phase is not economical as the case with two-phase
flow, in which high pressure losses particularly in hilly subsea terrain can occur.
OLGA is the worldwide leader and is the standard in dynamic simulation tools for multiphase
simulation software. It is widely used in the industry for simulating multiphase pipe flows for the oil and
gas industry. It also incorporates a CO2 module for accuracy in computing CO2 flows. The equation of
state model used within OLGA for the CO2 single component module is the Span and Wagner Equation.
The Span Wagner equation is a generalized corresponding states equation of state (EOS) that supersedes
the earlier equations and now is generally recognised by industry as the most accurate representation of
the available experimental pressure, volume and temperature data for CO2 and its mixtures.
The basis is a one dimensional, three-phase fully dynamic simulator including heat transfer. Fluid
properties are based on external programs, such as PVTSIM, and are used in OLGA in the form of a
3022 Wolfgang Böser and Stefan Belfroid / Energy Procedia 37 (2013) 3018 – 3030

matrix in which all properties are tabulated. For use with CO2 systems, a single component module was
added. The main difference between the single component and normal modules is the evaluation
concerning phase transitions. In the base OLGA, the pressure, volume and temperature data is supplied to
OLGA in the form of tab files. In general for oil and gas applications, the gas liquid fraction only
gradually changes and forms a 2-phase envelope that is taken into account in OLGA. For a single
component system, the tab file is generated in OLGA using the Wagner equation of state, which was
setup specifically for pure CO2. For a single component, the phase transition region is not an envelope but
a single line at which there is a discontinuity in the fluid enthalpy. Furthermore, as the critical point is in
the normal operating range, a smoothing by numerical means needs to be done around this point. Around
the critical point the fluid properties are extrapolated from the region boundaries. Like this singularities at
the critical point could be avoided. This treatment of the phase envelope results in additional uncertainty
as now a relaxation time in the condensation and boiling must be used to avoid unphysical results and to
avoid numerical instabilities.
Except for the treatment of the phase diagram, there are no further modifications to the main OLGA
code. This means that all normal multiphase correlations regarding slug initiation, entrainment rate, slip
velocities etc. are maintained. With respect to the single phase flow, Kim et. al. made an inventory of
available data on the pressure drop and heat transfer behaviour of CO2 including for the supercritical state
[1]. Their conclusion was that currently insufficient data or reasons exist to believe that the dependence of
heat transfer and pressure drop on fluid properties are not captured adequately by conventional single-
phase turbulent flow correlations. Of course near the critical point large variation can occur. This means
that all normal design tools can be used to design and dimension CO2 transport networks. Whether this is
correct, with respect to multiphase is an unknown and too little open literature date is available at present.
Due to the low surface tension and low viscosity of liquid CO2 compared to water and most oils, the
multiphase flow behaviour is different and classic flow maps cannot be used to predict the actual flow
regime. A lot of work is being done on the multiphase flow behaviour for CO 2 boiling but mostly in
smaller diameter tubes. Although for tubes larger than 1 mm in diameter they can be considered large for
a two-phase CO2 flow. The results were that the flow maps deviated from the classic flow maps. Statoil
recently presented some good prediction results using the Friedel relation recently [7]. With respect to the
multiphase flow behaviour and especially the flow regime identification further research is required.
Groups such as Sintef , SPT group/IFE and CATO2 are currently working on that topic. While there
is still uncertainty about the 2-phase aspects of a multiphase CO2 fluid flow, as long as no other validated
dynamic tools are available, the use of OLGA for CO2 flow assurance is as good as one can get.
However, the main uncertainties in the dynamic simulations using OLGA remain:
o Validity of two-phase models tuned and validated for oil-gas flow, for a CO2 stream.
o Choice of thermodynamic non-equilibrium parameter. OLGA takes this phase transition
relaxation time by a parameter which determines the time delay with respect to the
thermodynamic non-equilibrium. A low relaxation time means all phase transfer happens
extremely fast. For simulation the default value of 1 sec is used for boiling and condensation.
This is the value recommended by the SPT Group.
At the moment there is too limited a set of field data on the open literature to remove these
uncertainties. For validation, field data or test data should be gathered or obtained.

4. Steady state operating conditions

4.1. Well behaviour

For injection into a well, the mass flow rate is determined by the upstream compressor or pump
system, which means that at steady state the mass flow rate is given. As the reservoir pressure is also
given at a given time, all other pressures throughout the system can be determined from these inputs,
 Wolfgang Böser and Stefan Belfroid / Energy Procedia 37 (2013) 3018 – 3030 3023

including the bottom hole pressure, even though small variations in bottom hole pressure may occur due
to variations in bottom hole temperatures. Furthermore, the wellhead temperature is determined by the
pipeline heat losses and the pipeline inlet temperature. This means that for an injection scenario, the
wellhead pressure is a result rather than a clear control parameter.
The pressure drop through the well is made up of two components: the gravitational and the frictional
pressure drops:
p 1
p gravity p friction g 1
2 u2 , (1)
z D
where p/ z and p are the pressure gradients along the tubing [Pa/m], is the fluid density [kg/m3], g
the gravitational acceleration [m/s2], u the fluid velocity [m/s], D the tube diameter [m] and the friction
coefficient [-]. The gravitational pressure is the only component when a static column of fluid is present
and corresponds to a hydrostatic head: it therefore tends to decrease the wellhead pressure compared to
the bottom hole pressure. The frictional pressure drop corresponds to a pressure drop in the direction of
the flow, due to friction. Since the flow is downward, its effect is opposite to the gravitational pressure
drop: it tends to increase the wellhead pressure. In a production well, both of these components act in
concert, resulting in the well known Tubing Performance Curve (TPC). In an injection well, the TPC
takes a different shape, as illustrated in Figure 2, where the left part (negative mass flow rate) corresponds
to a fluid injection. It should be noted that generally, the tubing curves are generated for a fixed wellhead
pressure. This is directly relevant and representative for hydrocarbon production, where the wellhead
pressure is being fixed by process equipment at the production site. However, during (CO 2) injection,
such an approach to looking at the pressure drop is less directly applicable as the wellhead pressure will
not be fixed but rather is the result of an imposed mass flow rate (at the outlet of the compressor) and the
reservoir pressure.
Wellhead temperature = 40 degC
350

300
Bottomhole pressure [bar]

250

200

150

100 Gravity
dominated
50 Wellhead pressure = 85 bar
Wellhead pressure = 60 bar
Wellhead pressure = 40 bar
0
-50 -40 -30 -20 -10 0 10 20 30 40 50
Mass flow rate (negative is injection) [kg/s]
Figure 2: Tubing performance curve of a CO2 well during production (positive mass flow rates) and injection at 3 different reservoir
pressures.

When two-phase flow occurs, the lowering of the wellhead pressure due to gravity can be limited by
the phase line. At high bottom hole pressure, the well can be fully filled with liquid CO2, resulting in
wellhead conditions beyond the critical point. At lower bottom hole pressures, the pressure decreases to
the wellhead up to the point that the phase line is reached, at which point CO2 is in gaseous form and the
3024 Wolfgang Böser and Stefan Belfroid / Energy Procedia 37 (2013) 3018 – 3030

gravitational pressure reduces sharply. In Figure 3, the pressure drop and the individual parts of the
frictional and gravitational pressure drop are plotted for a case at a reservoir pressure of 20 bar and for a
case of a reservoir pressure of 300 bar. For the 20 bar case, the friction is dominant, especially in the
smaller ID sections. However, for the 300 bar case, the pressure drop is almost completely determined by
the gravity term.

Figure 3: Pressure drop, frictional and gravitational pressure drop along wellbore for a reservoir pressure of 20 bar and a reservoir
pressure of 300 bar.

For two-phase conditions, the wellhead pressure is fully determined by the temperature. This is valid
as long as the pressure drop is nearly dominated by gravity. At high mass flow rates, the frictional
pressure drop becomes important and the wellhead pressure increases such that the phase line can be
avoided and the entire well is filled with supercritical fluid.
As discussed earlier, the pressure profile in the well is constrained by the occurrence of two phase
conditions in the well. This is visualized in Figure 5. In this figure, the wellhead pressure is plotted as
function of the wellhead temperature for different reservoir pressures and for two mass flow rates. At high
inlet temperatures, the wellhead is in the dense phase region and the remainder of the well is also filled
with supercritical fluid. At lower mass flow rates, there are three regimes, at extremely low rate; the well
is filled with gas. At medium flow rates, the well is partly filled with two-phase flow and at extremely
high flow rates, the complete well is filled with liquid or supercritical fluid.
At lower inlet temperatures, roughly up to the critical point, the wellhead pressure is determined by
the wellhead temperature. That is, there are almost always two-phase conditions at the wellhead at low to
medium wellhead temperatures. A direct observation from this plot is also that lower wellhead
temperatures lead to lower required injection pressures for the same mass flow rate. This is primarily due
to the fact that the density of the fluid is higher at lower temperature, leading to (a) a higher gravitational
component in the well (a larger hydrostatic head), and (b) slower velocities, hence lower friction
throughout the well. The effect is very significant: at a reservoir pressure of 100bar, the required injection
pressure for an inlet temperature of 20 C is 57 bar, whereas at 40 C it is already 76 bar. This means that
by lowering the wellhead temperature, a significant saving can be made on the required compressor
power, although at the cost of added operational uncertainty due to the operation at two-phase conditions.
In the current project an alternative has been chosen. At low reservoir pressure, the bottom hole
temperature is the limiting factor for the wellhead temperature. At too low wellhead temperature, the
bottom hole is limited by the occurrence of two-phase flow conditions (Figure 6). That means that, with a
critical temperature of 15 °C, as long as the bottom hole pressure is below 50 bar, the temperature can fall
below the critical temperature, but as for higher bottom hole pressures, the wellhead temperature is no
longer the limiting factor. In that case, the limiting factor is the occurrence of two-phase flow in the
pipeline.
 Wolfgang Böser and Stefan Belfroid / Energy Procedia 37 (2013) 3018 – 3030 3025

In the project some gain was achieved with respect to the compressor outlet pressure, by using a
platform valve which maintains the pipeline at 85 bar but due to the cooling over the valve, the resulting
wellhead pressure is lower than in the scenario where no valve is used. It must be noted that this can lead
to instabilities as the pressure drop across the valve, the temperature drop across the valve and the
pressure drop in the well must match.

Figure 4: Operational corridor.

In the final design this resulted in the fact that at low reservoir pressure the pipeline inlet temperature
must be set to 80 °C whereas at high reservoir pressure the pipeline temperature could be lowered down
to 40 °C, (Figure 4) which was the minimum temperature due to the cooler constraints (Figure 4).
Mass flow rate = 47 kg/s Mass flow rate = 23.5 kg/s
140 140
phase line phase line
Reservoir pressure 20 bar Reservoir pressure 20 bar
120 120
Reservoir pressure 40 bar Reservoir pressure 40 bar
Reservoir pressure 60 bar Reservoir pressure 60 bar
100 Reservoir pressure 100 bar 100 Reservoir pressure 100 bar
Inlet pressure [bar]
Inlet pressure [bar]

Reservoir pressure 200 bar Reservoir pressure 200 bar

Reservoir pressure 300 bar Reservoir pressure 300 bar
80 80

60 60

40 40

20 20

0 0
-60 -40 -20 0 20 40 60 -60 -40 -20 0 20 40 60
Inlet temperature [degC] Inlet temperature [degC]
Figure 5: Wellhead pressure as function of wellhead temperature for different reservoir pressures for a mass flow rate of 47 kg/s
(left) and 23.5 kg/s (right).
3026 Wolfgang Böser and Stefan Belfroid / Energy Procedia 37 (2013) 3018 – 3030

70
Mass flow rate 47 kg/s
60 Mass flow rate 23.5 kg/s
phase line downhole pressure 47 kg/s

Downhole temperature [degC]

50 phase line downhole pressure 23.5 kg/s

40

30

20

10

-10

-20
-60 -40 -20 0 20 40 60
Inlet temperature [degC]
Figure 6: Down hole temperature as function of wellhead temperature for a reservoir pressure of 20bar and a mass flow rate of 47
and 23.5 kg/s.

5. Start-up and shut-in

Transients such as start-up, shut-in and ESD scenarios are critical in the overall design. Especially in
the first period of operations frequent starts and stops are expected and the required start-up time is
critical. In contrast to most fluids, the pure CO2 results in sharp phase behaviour and non-standard
behaviour when the system is in two-phase conditions. This means that the cool-down time becomes
important whether the pipeline is in gas, liquid or two-phase conditions.

In principle the base condition is that the pipeline is maintained in single phase conditions. However,
as indicated in section 4.1, the pipeline operating temperature varies from 80 40 °C and a typical
operating pressure around 85 bar. For long shut-in conditions, with a seabed temperature of 4 °C, the
resulting condition is at two-phase conditions. The liquid fraction varies, depending on the operating
conditions, between 13-75 %. The cool-down to reach two-phase conditions is between 4- 105 hrs (Figure
7). The cool-down time increases with increasing operating temperature and increases marginally with
increasing operating pressure. The scenario of pressurizing the pipeline before completely shutting it in
does extend the single phase period but not drastically. The two-phase flow might be prevented by
emptying the pipeline in the well. i.e., keep the well open up to the point of potential back-flow with a
closed in pipeline at the compressor inlet. This scenario works very well for the low reservoir scenario.
Due to the high operating temperature, the pipeline goes from supercritical conditions to gaseous
conditions. However, at higher reservoir pressures, the operating conditions are too close to the phase line
and two-phase flow or back-flow from the reservoir occurs too fast and the pipeline still ends up in two-
phase conditions. This means that often, the two-phase conditions cannot be avoided. This has direct
consequences for the start-up.
 Wolfgang Böser and Stefan Belfroid / Energy Procedia 37 (2013) 3018 – 3030 3027

31 hr
48 hr
18 hr
84 hr
4 hr
42 hr
105 hr

4 hr, 29 hr

Figure 7: Cool-down Temperature-Pressure profiles plotted in the phase diagram. Included are the cool-down times.

Typical shut-in conditions for the pipeline and well are given in Table 3 and Table 4.

Table 3: Pipeline shut-in conditions. Table 4: Shut-in conditions (near stabilization).

Pipeline fluid Pipeline fluid Reservoir Wellhead Wellhead Well Phase

pressure temperature Pressure Pressure Temperature
[bar] [bar] [° C]
[bar] [°C]
20 12 4 Gas
38.7 4 50 28 3 Gas
50.9 15 100 46 11 Gas/two-phase
12 4 200 41 6 Gas/two-phase
300 72 5 Liquid

At the start-up several procedures are constrained:

- At low reservoir pressure, the bottom hole temperature is a constrained for the injection rate
and/or wellhead temperature.
- A minimum wellhead temperature is a constrained for the pressure drop across the control valve.
- The constraints on the mass flow rate through the tubing due to the limits on the mechanical
vibrations.
- The possible occurrence of slugs from the pipeline to the platform which might impose
constraints due to mechanical vibrations of the riser and/or topside piping.

For the start-up at low reservoir pressure there are three options:
- Start-up from a single phase gas conditions in the pipeline (occurs if the pipeline is shut-in while
keeping the wellhead choke open until back-flow).
- Start-up from two-phase conditions.
- First re-pressurizing of the pipeline.
For the start-up at higher reservoir pressure there are two options:
- Start-up from two-phase conditions.
- First re-pressurizing of the pipeline.
3028 Wolfgang Böser and Stefan Belfroid / Energy Procedia 37 (2013) 3018 – 3030

The start-up procedure is very dependent on the bottom hole pressure. As stated, at low reservoir
pressure, there is a strict limit in the injection temperature to avoid dropping below the critical down hole
temperature of 15 °C. That means that at low reservoir pressure, the injection rate must be limited at the
start. The injection can be increased only if the wellhead temperature has increased enough. Due to the
long pipeline this will take several days. Start-up from two-phase conditions at low reservoir pressure
poses a problem as the pipeline pressure is much higher at a minimum of 39 bar compared to the wellhead
pressure of 11 bar. If the control valve is opened at those conditions the pressure drop is such that the
wellhead temperature drops to approximately -30 °C. This is too low for the design temperature unless
Artic wellhead design constraint is taken. The same does not apply to the higher reservoir pressure
scenarios as the wellhead pressure is in those cases near equal or higher than the pipeline shut-in pressure.
In those cases start-up until stabilization takes in those cases about 20 hours (reservoir pressure of
100bar).
With the pipeline pressurization, several mechanisms play a role. Warm CO2 is injected in a pipeline
consisting of liquid and gaseous CO2, resulting in liquid boiling. At the same time the pipeline is
pressurized. This occurs mainly from the platform side. Liquid is pushed to the platform where the
compression starts and the gas is condensed in liquid due to the pressure.

Time is 0.833 hrs Time is 1.67 hrs Time is 4.58 hrs

1 1 1

0.9 0.9 0.9

0.8 0.8 0.8

0.7 0.7 0.7

Liquid hold-up [-]

Liquid hold-up [-]

Liquid hold-up [-]

0.6 0.6 0.6

0.5 0.5 0.5

0.4 0.4 0.4

0.3 0.3 0.3

0.2 0.2 0.2

0.1 0.1 0.1

0 0 0
0.5 1 1.5 2 2.5 0.5 1 1.5 2 2.5 0.5 1 1.5 2 2.5
along length [m] x 10
4
along length [m] x 10
4 along length [m] x 10
4

Figure 8: Liquid hold-up in the pipeline during pressurization process.

The benefit of first pressurizing the pipeline until single phase conditions are reached is that although
there is a high pressure drop across the choke, the temperature is limited because as soon as the wellhead
is opened, the pressure is high on the downstream side due to the well design and the high flow rates. At
low reservoir pressure, the down hole temperature does drop for a period below the critical temperature.
At high reservoir temperature this scenario is with start-up times of 10 (reservoir pressure 100 bar) - 27
hrs faster than without pressurization.
The downside of first pressurization to single phase conditions is the high mass flow rate initially
injected. This might lead to erosion issues, if the fluid is not completely free of solids, or mechanical
vibrations of the tubing. For the current well design the vibration issues have been handled by setting the
packers at specific locations. Due to the clean fluid erosion is assumed not to be an issue, although
inspection after an initial injection period is planned.
As is clear from the above discussion, the dynamic operation of the injection is not straightforward and
is a delicate balance between different design requirements, such as the design temperature, and
operational requirements, such as the start-up time. The results are very dependent on the well design and
operating conditions such as wellhead pressure and temperature.
 Wolfgang Böser and Stefan Belfroid / Energy Procedia 37 (2013) 3018 – 3030 3029

6. Discussion

Simulations carried out showed that transport and storage of CO2 in the given storage side will work
despite constraints of the storage side and the process. By adding a pressure control valve with an Arctic
design at the platform you get the necessary flexibility to manage different operational situations during
process lifetime. The operational envelope will be from 40 bar up to a maximum pressure of 129 bar and
temperatures in between 40 °C to 80 °C (Figure 9). Increasing the CO2 temperature at shut down will
increase the time interval until phase change of CO2 occurs during shut in. This will facilitate and
accelerate the restart. Good process operation requires a lot of live information on pressure, temperature
and valve positions at different places in the transport chain down to bottom hole. The operational
procedures have to be adapted regularly to the increasing pressure of the storage side.

Figure 9: Operational envelope.

Findings from the study are based on numerical simulations. All the time they have been challenged by
engineering background from power plant projects. But using CO2 as a fluid in this context is still pure
theory. There is a strong need for the results to be validated. Continuation of the project requires further
discussion incorporating experiences from other projects and experimental knowledge.
3030 Wolfgang Böser and Stefan Belfroid / Energy Procedia 37 (2013) 3018 – 3030

7. Acknowledgements

The flow assurance involved a great number of experts from GDF Suez and E.ON. Skills for numerical
software simulation by TNO is acknowledged for the numerical work. Thanks goes to TAQA for sharing
its specific experiences in the architecture of gas production side on a platform. They all provided
valuable input for the ROAD project.

8. References

[1] Kim M-H, Pettersen F., Bullard C.W.,2004 Fundamental process and system design issues in
CO2 vapor compression systems, Progress in Energy and Combustion Science 30 119-174
[2] Snøhvit LNG plant and CO2 transport R&D
Dr. Gelein de Koeijer; Gassnova CO2 transport workshop; 7-11-2008, Porsgrunn
[3] CATO2 WP2.1
-Phase Flow in Carbon D
144847
[5] Havre K., Active Feedback Control as the Solution to Severe Slugging, SPE 71540, 2001
[6] Koeijer G. et al., Experiments and modeling of two phasetransient flow during CO2 pipeline
deprerssurization, GHGT 9, Elsvier, Energy Procedia, 1683-1689, 2009
[7] Paterson L. et al., Numerical Modeling of Pressure and Temperature Profiles including Phase
Transitions in Carbon Dioxide Wells, SPE 115946, 2008
[8] Pekot L. et al., Simulation of two-phase flow in carbon dioxide injection wells, SPE 144847,
2011
[9] Dugstad A. et al., Dense phase CO2 transport when is corrosion a threat, NACE
International 2011
[10] Liljemark S. et al., Dynamic simulation of a carbon dioxide transfer pipeline for analysis of
normal operation and failure modes, GHGT 10, Elsvier, Energy Procedia, 3040 3047, 2011
[11] Chaczykowski M., Osiadacz A., Dynamic simulation of pipeline containing dense
/supercritical CO2-rich mixtures for carbon capture and storage, Elsevier, 446-456, 2012