copy distribution for Sample

December, 2010
Vol.9, No.4
Journal of
Pipeline Engineering
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
incorporating
The Journal of Pipeline Integrity
Great Southern Press
Clarion Technical Publishers
4th Quarter, 2010
209
The Journal of
Pipeline Engineering
incorporating
The Journal of Pipeline Integrity
Volume 9, No 4 • Fourth Quarter, 2011
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Contents
James Watt ................................................................................................................................................................213
Carbon dioxide transport infrastructure key learning and critical issues
Saulat Lone, Dr Tim Cockerill , and Prof. Sandro Macchietto . ......................................................................... 223
The techno-economics of a phased approach to developing a UK carbon dioxide pipeline network
Dr Brian N Leis, Dr James H Saunders, Ted B Clark, and Dr Xian-Kui Zhu .................................................... 235
Transporting anthropogenic CO2 in contrast to pipelines supporting early EOR
Graeme G King and Satish Kumar ....................................................................................................................... 253
How to select wall thickness, steel toughness, and operating pressure for long CO2 pipelines
Prof. Haroun Mahgrefteh, Solomon Brown, and Peng Zhang ........................................................................... 265
A dynamic boundary ductile-fracture-propagation model for CO2 pipelines
Dr Robert Andrews, Dr Jane Haswell, and Russell Cooper ................................................................................ 275
Will fractures propagate in a leaking CO2 pipeline?
Dr Tim Cockerill, Dr Naser Odeh, and Scott Laczay . ......................................................................................... 285
Greenhouse gas emissions from electricity generating CCS upstream and downstream transport processes
❖❖❖
Our cover photo shows construction work under way on Denbury Resources’ 24-in diameter, 512-km long
Green Pipeline for both natural and man-made CO2. The new pipeline will be one of the first designed to transport
anthropogenic CO2 in the Gulf Coast area of the US. Photograph courtesy of Denbury Resources Inc., www.denbury.com
210
The Journal of Pipeline Engineering
T
HE Journal of Pipeline Engineering (incorporating the Journal of Pipeline Integrity) is an independent, international,
quarterly journal, devoted to the subject of promoting the science of pipeline engineering – and maintaining and
improving pipeline integrity – for oil, gas, and products pipelines. The editorial content is original papers on all aspects
of the subject. Papers sent to the Journal should not be submitted elsewhere while under editorial consideration.
Authors wishing to submit papers should send them to the Editor, The Journal of Pipeline Engineering, PO Box 21,
Beaconsfield, HP9 1NS, UK or to Clarion Technical Publishers, 3401 Louisiana, Suite 255, Houston, TX 77002, USA.
Instructions for authors are available on request: please contact the Editor at the address given below. All contributions
will be reviewed for technical content and general presentation.
The Journal of Pipeline Engineering aims to publish papers of quality within six months of manuscript acceptance.
Notes
4. Back issues: Single issues from current and past volumes
are available for US$87.50 per copy.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
1. Disclaimer: While every effort is made to check the
accuracy of the contributions published in The Journal of
Pipeline Engineering, Great Southern Press Ltd and Clarion
Technical Publishers do not accept responsibility for the
views expressed which, although made in good faith, are
those of the authors alone.
5. Publisher: The Journal of Pipeline Engineering is
published by Great Southern Press Ltd (UK and Australia)
and Clarion Technical Publishers (USA):
2. Copyright and photocopying: © 2010 Great Southern
Press Ltd and Clarion Technical Publishers. All rights
reserved. No part of this publication may be reproduced,
stored or transmitted in any form or by any means without
the prior permission in writing from the copyright holder.
Authorization to photocopy items for internal and personal
use is granted by the copyright holder for libraries and
other users registered with their local reproduction rights
organization. This consent does not extend to other kinds
of copying such as copying for general distribution, for
advertising and promotional purposes, for creating new
collective works, or for resale. Special requests should
be addressed to Great Southern Press Ltd, PO Box 21,
Beaconsfield HP9 1NS, UK, or to the editor.
3. Information for subscribers: The Journal of Pipeline
Engineering (incorporating the Journal of Pipeline Integrity)
is published four times each year. The subscription price for
2010 is US$350 per year (inc. airmail postage). Members of
the Professional Institute of Pipeline Engineers can subscribe
for the special rate of US$175/year (inc. airmail postage).
Subscribers receive free on-line access to all issues of the
Journal during the period of their subscription.
v
Great Southern Press, PO Box 21, Beaconsfield
HP9 1NS, UK
tel: +44 (0)1494 675139
fax: +44 (0)1494 670155
email: [email protected]
web: www.j-pipe-eng.com
www.pipelinesinternational.com
Editor: John Tiratsoo
email: [email protected]
Clarion Technical Publishers, 3401 Louisiana,
Suite 255, Houston TX 77002, USA
tel:
+1 713 521 5929
fax: +1 713 521 9255
web: www.clarion.org
Associate publisher: BJ Lowe
email: [email protected]
6. ISSN 1753 2116
v
v
www.j-pipe-eng.com
is available for subscribers
4th Quarter, 2010
211
Editorial
Guest editorial: CO2 transportation by pipeline – a special issue
T
operational in the next five years. Much of the research has
therefore to be conducted in parallel with FEED studies and
it is vital that there is rapid information exchange between
academia and industry.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
HIS SPECIAL ISSUE of the Journal of Pipeline
Engineering is dedicated to the topic of carbon
dioxide transportation by pipeline. The subject of pipeline
transportation of CO2 is becoming increasingly important
as carbon capture and storage (CCS) schemes world-wide
are moving from the pilot to the demonstration phases.
In recognition of the increased interest in CO2 transport,
Newcastle University – in association with Tiratsoo Technical
and Clarion Technical Conferences – organized the first
Forum on the transportation of CO2 by pipeline in July
this year. The aims of the Forum were to highlight the key
issues relating to CO2 pipelines and bring together leading
academics and industry experts to discuss current and future
CCS ventures and the international research activities being
undertaken to support these projects.
In this issue, five papers from the technical sessions of the
Forum are published, together with two further papers from
international authors working in the field. The conference
opened with a session of keynote and scene-setting papers
including the paper published in this volume by James
Watt on ‘Carbon dioxide transport infrastructure: key
learning and critical issues’. In this paper, the world-wide
operational experience in the large-scale transportation
of high pressure CO2 is reviewed in order that the issues
relevant to the development of networks of pipelines for
CCS schemes can be understood. It is estimated that there
are approximately 6000km of CO2 pipelines globally,
predominantly in North America; the majority of these
pipelines are transporting naturally sourced CO2 for
the purpose of enhanced oil recovery (EOR). Whilst it is
concluded that the knowledge gained from this experience
is vital, there remain areas pertinent to the development of
CCS transport networks which still need to be addressed.
However, one of the key messages of this paper is that, due
to the urgent requirement to reduce CO2 emissions, the
industry is not in the position of being able to wait until
this research is complete before designing, constructing, and
operating pipeline networks. If CCS is to have an impact
in the reduction of CO2 emissions, then plants have to be
One of the areas highlighted in James Watt’s paper for further
research is that of materials’ selection. This important topic
is the subject of the paper given at the Forum by Dr Paul and
co-authors, who conclude that the main issues for the pipeline
material are corrosion in the CO2 process stream, resistance
to brittle and ductile fracture propagation, and degradation
of polymers in supercritical CO2. With respect to corrosion,
it is recognized that typical carbon steel materials used for
pipelines are not corrosive in pure, dry, supercritical CO2,
and the paper highlights that there is considerable experience
and research in this area relevant to the currently operating
pipelines. However, one of the significant differences that
will be encountered in pipelines transporting CO2 in CCS
schemes is that the product stream will contain impurities
in combinations not currently transported in pipelines for
EOR. There is very little research work on the effect of these
impurities in the event of a process upset which could allow
free water to enter the pipeline.
In terms of fracture propagation, Dr Paul and his co-authors
conclude that although there is extensive experience with the
specification of material properties to prevent long running
ductile and brittle fractures from propagating in natural gas
pipelines, this knowledge cannot be directly applied to the
design and fracture control of dense phase CO2 pipelines.
This view is shared by Leis and co-authors in their paper
published here on ‘Transporting anthropogenic CO2 in
contrast to pipelines supporting early EOR’. In particular,
they conclude that the main approach used for specifying
ductile fracture control requirements in natural gas pipelines,
the Battelle Two-Curve Method, has not been validated
for CO2 containing impurities from a capture plant, and
neither has the equation of state which forms the basis of this
model. One of the key conclusions from this paper is that
the decades of experience with the transportation of CO2
212
The Journal of Pipeline Engineering
for EOR should be considered carefully when applying this
knowledge to the design of pipelines for CCS, particularly
in the area of fracture control. Indeed it is highlighted
that many of the early CO2 pipelines were retrofitted with
crack arrestors to manage concerns with fracture arrest.
In a complementary paper, King and Kumar illustrate the
issues to be considered in the design of a high-pressure CO2
pipeline for ductile fracture arrest using the example of the
propsoed CO2 Masdar pipeline in Abu Dhabi. In this paper
it is demonstrated that ductile fracture arrest is possible in
high-pressure CO2 pipelines using the pipe wall thickness
alone without the need for crack arrestors.
Finally, Dr Tim Cockerill and co-authors present a lifecycle
analysis to determine the CO2 emissions which are generated
by the pipeline transportation of CO2 and the impact that
these might have on the CCS process. The analysis indicates
that the greenhouse gas emissions from the transport phase
are almost negligible and therefore the optimal location of
power plant in the design of networks should be driven by
the requirement to minimize fuel processing and transport
rather than CO2 transportation.
As you read through the papers in this issue, one of the
recurring conclusions drawn by the authors is that there is a
still a significant amount of research required in the area of
CO2 pipeline transportation. In particular, the experimental
database on which so much of the knowledge relating to
natural gas pipelines is built has not been established for
CO2. There is therefore an urgent need for researchers to
work in collaboration in order that the required research can
be completed within timescales which allow CO2 pipelines
for CCS to be designed and operated safely and efficiently.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
As well as concerns over ductile fracture propagation, as
discussed in the previous papers, one of the issues with CO2
is that there is a significant temperature drop around the leak
site as the escaping fluid expands, due to the Joule-Thompson
effect. It is postulated that this drop in temperature could
cause continuous brittle initiation of a crack in a CO2
pipeline as the steel is cooled below the ductile-brittle
transition temperature. This effect is investigated in the
paper by Andrews, Haswell, and Cooper ‘Will fractures
propagate in a leaking CO2 pipeline?’.
infrastructure is set up to which smaller CO2 sources could
then be added. The analysis conducted indicates that there
comes a point in the network development where addingin smaller sources, particularly those that are remote from
clusters of CO2 sources, considerably increases the marginal
costs and diminishes the returns.
An important additional conclusion drawn by King and
Kumar is that, when designing a pipeline or network, there
are a number of potential options that could be selected
and cost optimization techniques, in combination with the
technical requirements, should be implemented to derive the
final solution. This theme is taken up by Saulat Lone and
his co-authors in their paper ‘The techno-economics of a
phased approach to developing a UK CO2 pipeline network’.
This paper investigates the establishment of CO2 networks
in the UK using idealized scenarios in which a backbone
Dr Julia Race
Senior Lecturer in Pipeline Engineering
School of Marine Science and Technology,
Newcastle University,
Newcastle-upon-Tyne, UK
[email protected]
4th Quarter, 2010
213
Carbon dioxide transport
infrastructure: key learning and
critical issues
by James Watt
AMEC Power and Process, Europe, Darlington, UK
C
ARBON CAPTURE AND STORAGE is acknowledged as one of the key technologies in carbon
dioxide abatement. Whilst not a permanent solution, it can enable the continued use of hydrocarbon
based power generation and reduce emissions from industrial processes. This is critical in decarbonizing
whilst renewable and cleaner energy sources come online to resolve the issues around energy security
and security of supply.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Transportation of carbon dioxide in a CCS scheme is critically important but has not been well addressed.
While increasing attention is paid to storage and capture technology, transportation issues still lag behind. In
particular, storage assessments and research increases and demonstration projects drive capture technology
development. Shipping is rising as a potential solution and more consideration is being given to the impact
of clusters and networks.
The amount of carbon dioxide pipelines is approximately 6000km globally, the majority of which are in North
America. Compared to other pipeline distances for natural or hydrocarbon pipelines, this is a relatively
small experience base. Systems that do exist are also different. The majority of pipelines are installed for
the purposes of enhanced oil recovery, often using natural sources. There are anthropogenic sources, but
not many. Whilst pipeline design is common practice, the concern – if any – is the fluid being transferred
and the dynamics of the system. In such a new field set for rapid growth industry needs to understand and
make use of what experience is available and transferable and, more importantly, identify the gaps.
Maturity of transport systems
When considering the use of pipelines it is important first
to take stock of the existing facilities. Pipelines are good
references, but the knowledge embedded in them comes
with the supporting research, engineering, and learning
that comes with each design. The majority of carbon
dioxide pipelines are in the USA and Canada, along with
substantial in-field pipework for EOR schemes. There are
other pipelines, including 90km in Turkey, and pipelines
in Algeria and Hungary, but there are relatively few outside
North America. Only two projects have offshore pipelines:
Snovhit and Sleipner in the North Sea.
This paper was presented at the First International Forum on Transportation of CO2
by Pipeline, organized in Newcastle upon Tyne in July, 2010, by Tiratsoo Technical
and Clarion Technical Conferences, and with the support of the University of
Newcastle and the Carbon Capture and Storage Association.
Author’s contact details
tel: +44 (0)1740 646100
email: [email protected]
In North America since 1972 carbon dioxide has been used
for enhanced oil recovery from some man-made sources, but
the majority of transported carbon dioxide is from naturally
occurring gasfields along the mid-continental mountain
ranges and Mississippi Basin, as shown in Fig.1 and Table
1; the gas is transmitted above critical pressure, Fig.2.
The maturity of the systems is still limited, but the experience
that is there has formed a nucleus upon which to build CCS.
The pipelines have been developed at the required scale; for
example, the Cortez pipeline delivers 19.3 million tonnes
per year, roughly equivalent to one of Europe’s largest power
station at Drax, UK. So the pipelines have been proved at
scale in terms of flow. That only 6000km of pipelines are
in place is often pointed to as a weakness, not providing
an extensive knowledge base, when compared for instance
to the 490,000km of gas pipelines or the 278,000km of
hazardous liquid pipelines in the USA alone. Whilst specific
knowledge may be limited to 2% or less of the hazardous
liquids’ experience in the US, there is still a knowledge
base to consider: over 600,000km [1] of pipeline designed
to the same codes of practice.
The Journal of Pipeline Engineering
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
214
Fig.1. North American CO2 pipelines
Location
Capacity
(Mt CO2/y)
Length
(km)
Pressure
(bar)
Year
Complete
USA
19.3
808
186
1984
Sheep Mountain
USA
9.5
660
132
Bravo
USA
7.3
350
165
1984
Bravo Dome
Canyon reef Carriers
(SACROC)
USA
5.2
225
175
1972
Gasification
Val Verde
USA
2.5
130
-
1998
Val Verde Gas Plants
Turkey
1.1
90
170
1983
Dodan field
USA &
Canada
5
328
Up to 204
2000
Gasification
Pipeline
Cortez
Bati Raman
Weyburn
Origin of CO2
McElmo Dome
Sheep Mountain
Table 1: Major carbon dioxide pipelines [26].
Within the US and Canada, regulatory frameworks that
govern carbon dioxide pipelines have been developed
and deployed for a number of years. The design of such
pipelines is essentially uses the same standards as for any
hazardous liquid pipeline such as ethylene, crude oil, or
petroleum products. These codes are, for the US 49 CFR
195, Transportation of hazardous liquids by pipeline, and for
Canada Z662-07, Oil and gas pipeline systems. For gaseous
carbon dioxide, 49 CFR 192 applies rather than the liquidspecific 49 CFR 195 [2].
The approach within the US is that the code or regulation
contains a number of established standards from the API,
AMSE, ASTM and others. Those referenced in the regulation
or quoted, such as ASME B31.4, are the required minimum
and therefore become part of the regulation, in theory a
hard standard on which to base design. The core standard
is ASME B31.4 – the code for liquid pipelines – see Fig.3.
However, evidence suggests that ASME B31.8 is also applied.
This gas-specific code is used to evaluate the safety issues
around a gas pipeline, applying these rules to carbon dioxide
liquid lines as the fluid transitions to gas on release. The
process of design for pipelines in hazardous liquid service
is therefore robust and well understood.
European experience
Current European experience rests with Statoil at Sleipner
and Snovhit, although there is some carbon dioxide
experience on the Continent. Predominantly, current
transportation is either by ship or road tanker. The
expectation is that European CCS will evolve to target
215
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
4th Quarter, 2010
Fig.2. Carbon dixoide phase diagram and North American pipeline operating envelope.
Fig.3. Prescribed standards and codes
under 49 CFR 195.
Fig.4. ISO 13623 petroleum and
natural gas industries – pipeline
transportation systems.
216
The Journal of Pipeline Engineering
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Fig.5. BS PD 8010 steel pipelines
on land.
offshore storage sites, and due to the potential for CCS
clusters, develop in a network format. In some estimates the
infrastructure for European CCS is estimated at between
30,000 and 150,000km [3]. This brings new challenges that
have not been addressed in the current experience base,
both in terms of scale and the need for offshore pipelines.
As in the US and Canada, European and international
codes are also mature: ISO 13623 (BS EN 14161) and
DNV OS-F101 are equivalent to the US codes, although
neither specifically consider dense-phase or supercritical
carbon dioxide. However, both are adequate codes, but
should be supported by industry best practice. In the UK
the HSE recommends that BS PD 8010 is used as a wider
appreciation of good practice in pipeline design. In each
case the standards recommend the same or similar sub-codes
and standards, as shown in Figs 4 and 5.
Requirements in CCS systems
The infrastructure needs for CCS are different from those
of simpler enhanced oil recovery schemes. In the existing
schemes, the carbon dioxide is a by-product or natural
emission that is used to drive crude oil out of the ground.
The economic driver is therefore the demand for the oil,
and when the demand is not there, the extraction of natural
carbon dioxide need not occur.
Future CCS schemes will not have the same relationship.
Whilst the sensitivity of the storage is an issue, the
requirements here of flow, pressure, and temperature set
the downstream conditions. The emitter in CCS provides
another set of upstream conditions in terms of flow rates,
ramp rates, composition, temperature, and pressure, and
each of these systems imposes conditions on the transport
infrastructure. This is multiplied when considering networks.
Transportation systems have little ability to respond to
variances: essentially, even the flow rate is dictated by
the power station or storage. The dynamics of the whole
chain system, with drivers at both ends, is therefore more
complex. There will be an economic driver here to maintain
transportation to the storage and maintain a storage solution
that is flexible and allows accommodation of the emitter’s
operating regime.
What has the US learned?
The experience in the US has highlighted the critical issues
that must be considered for carbon dioxide service [4, 5].
These include
• process conditions
• properties
• operating conditions at entry and exit
• flow calculation method
• transient (surge) modelling
• flow characteristics
• typical carbon dioxide compositions
• piping design
• fracture propagation
• blowdown assembly design
• blowdown rate basis and calculation
• line break controls
• pig trap
• depth of cover
• routeing topography
• safety and environmental
• ambient/ground temperature
• blowdown rate basis and calculation
4th Quarter, 2010
217
Fig.6. CO2 pipeline incidents
1986-2008.
dispersion pattern
frequency and position of block valves
leak-detection systems
line inventory
For inspection there have been problems with dense-phase
or supercritical pipeline inspection and cleaning using
pigs. Simple scraper pigs used in cleaning pipelines need a
lubricating fluid such as diesel, and they are adversely affected
by the carbon dioxide which damages some non-metallic
materials [6, 7]. For inspection, the use of intelligent pigs
is routine; however, inspection experience in the US has
highlighted a key problem. The carbon dioxide penetrates
the non-metallic components and, as the tool is depressurized
in the receiver, the systems are often subject to rapid gas
decompression, destroying the unit. In fact Oosterkamp [8]
indicates that it has been reported that by 2008 only two
intelligent pig operations in North America have resulted
in the expensive tool surviving.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
•
•
•
•
• material selection
• pipeline materials
• carbon equivalent
• hardness value
• fracture strength
• valve, fitting and trim types
• seal, packing materials of construction
• valve actuators
• cleaning and strength testing
• cleaning
• hydrostatic testing/drying/dewatering
• construction techniques
• corrosion monitoring
• external corrosion
• fracture propagation
• special construction and welding
• stress relief
• pipeline operation
• refrigeration effects during start-up/blowdown
• start-up/shutdown methodology
• line pressuring
• requirement for blowdown noise control
• environmental considerations
• operational problems
• operational safety
• measurement
• custody transfer methods
• moisture analysis
The list is extensive and typical to the design of pipelines no
matter the fluid, but has particular relevance here. Particular
learning can be drawn from the experience in the areas of
inspection, corrosion, material specification, operational
safety, and thermodynamics.
Corrosion of carbon dioxide/water systems has been studied
extensively, not just for the carbon dioxide EOR industry
but also for processes involving the production of ammonia,
urea, steam reforming, and the sweetening of acid gas. The
key example to be considered is the research and practical
observation that was undertaken for the SACROC project1.
The research into corrosion by Schremp and Roberson [9]
tested a number of compositions against three common
weld types. The samples were full-size X60 12-in and 16-in
pipe sections that were exposed to a carbon dioxide mix
consistent with real SACROC pipeline conditions. Tests were
conducted at two design temperatures and pressures, and at
chemical compositions including 600-800ppm of H2S and
800-1000ppm of water. As a result, it is now accepted that
corrosion in carbon dioxide pipelines will not occur if the
water content is kept less than 60% of the saturation value
[10]. The original test used a water content 20 times higher
than specified for the SACROC pipeline, and concluded
that corrosion did not occur, even at this elevated water
content. Operational experience showed that after 12 years
the SACROC pipeline, with a 50-ppm water content limit,
had a corrosion rate of 0.25-2.5μm/y [11, 12].
1 The so-called SACROC (Scurry Area Canyon Reef Operators Committee) unit
in West Texas near the town of Snyder, one of the US’ largest oil fields, initiated
a 350-km long carbon dioxide injection pipeline network on 26 January, 1972.
218
The Journal of Pipeline Engineering
Fig.7.Typical orifice meter.
The operational problems most reported are safety incidents
reported in the US to PHMSA for all pipelines. In the
period from 1986 to March 2008 there were 42 [16] reported
incidents to PHMSA, Fig.6. This relates to approximately
0.36 incidents/1000km/year, (assuming US pipeline
distance average of 5000km over the period), compared
to 2447 incidents on the US gas transmission network of
488,000km [17] or 0.22 incidents/1000km/year. In the
period 1990-2001, the incident rate for natural gas was 0.17,
while for hazardous liquids it was 0.82 [18], so the incident
rate is still comparable and within the expected bounds for
a hazardous liquid.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Care always has to be taken with regard to corrosion; however,
it is clear from continued operation that carbon steel is
acceptable as material for carbon dioxide pipelines. There
have been incidents, however, with corrosion: SACROC
reports some issues [13], notably a corrosion incident in a
spur line from a main pipeline. The area had free-standing
water remaining from a hydrostatic test, and so corrosion set
in. This experience shows that care must be taken not only
to ensure a water content limit in the entry specification, but
also for the water introduced during testing, commissioning,
and maintenance activities.
In specifying water content the industry-accepted level are
conservatively specified as between 288-480mg/m 3 [14].
In addition the presence of other additional ‘acid gases’
such as H2S, SOx, and NOx needs to be considered. In
the case of CCS, the control of SOx and NOx compounds
will to be subject to earlier restrictions in the capture
process and are unlikely to present themselves in the
pipeline in significant volume. Hydrogen sulphide may
be an issue for some processes but should be kept below
the 200-ppm limit from Dynamis2 [15], Table 2; this
restriction and the water content specification should
prevent corrosion occurring.
Non-metallic components such as seals, valve seats, O-rings,
and even greases have also shown to be affected by carbon
dioxide. Petroleum-based seals can become saturated with
the high-pressure fluid and rapidly decompress when the
pressure is reduced or structurally weakened. Some greases
are also known to become hard and no longer effective.
Inorganic materials and greases are therefore more often
recommended.
Operational safety
Operational problems are not necessarily reported in the
public domain unless a release occurs, so public information
refers to where an incident has occurred, and has not
necessarily been prevented. In effect, only the failure is
recorded, so this limits the learning from operational sources.
2 The DYNAMIS project is investigating viable routes to large-scale cost-effective
hydrogen production with integrated CO2 management. The project is an element
of the HYPOGEN initiative, and forms a part of the European Commission’s QuickStart Programme for the Initiative for Growth. HYPOGEN has the goal of providing
Europe with a viable route to a hydrogen economy and includes, as an interim step,
the construction of a large-scale test facility for the production of hydrogen and
electricity from decarbonized fossil fuels with permanent CO2 storage.
Another reference source for pipelines (although not
specifically carbon dioxide) is CONCAWE. The data held
by CONCAWE indicates a significant number of incidents
relate to corrosion and third-party intervention. Hence, for
carbon dioxide systems, drying is a critical aspect of design
and operation in addition to the usual protection applied
to pipelines against third-party incidents.
There are two other operational issues to be considered.
The first is valve operation: here it is recommended that
all valves are slow opening, as this avoids damage to the
valves and surging in the pipeline, and this is particularly
important for blowdown valves. Where segments are above
ground, thermal relief should be provided and valves
must be capable of seating under high carbon dioxide
pressures. Current practice is not to work on pressurized
pipelines at all and, where necessary, sections and valves
bodies are blown down before removal of a valve or other
equipment item.
The thermodynamics of the fluid are generally based on
the determination of its properties by using equations of
state. The correct selection of the equations is therefore
critical. Experience has also shown that for key operations
such as start-up and blowdown, the thermodynamic
characteristics require much longer periods to avoid the
very low temperatures that are possible with carbon dioxide.
There are other key lessons to consider, including transient
fluid effects, leak proving (for which the use of air or nitrogen
is not sufficient), spacing of block valves, the high level of
sensitivity to temperature and pressure, and the attendant
effect on pipeline operations. All of these need to be
carefully considered.
4th Quarter, 2010
Component
219
Post Combustion
IGCC
Oxyfuel
Weyburn
Dynamis
CO2
>95%
>95%
N2/Ar
<4%
<4% (for noncondensable
gases)
0.01
0.03 –
0.06%
4.1%
<4%
EOR 1001000ppm
O2
Hydrocarbons
0
0.01%
H2
0
0.8 – 2%
H2O
0
H2S
0
Hg
Sox
Nox
Glycol
0
0
<100ppm
<500ppm
<1450ppmv
<200ppmv
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
CO
0.01-0.6%
<5%
Saline
Formation <4%
EOR <2%
<4% for all noncondensable
gases and
Hydrocarbons
0
0.03-0.4%
0
<2000ppmv
<0.01%,
<100ppm
<0.01%,
<100ppm
0.5%
<100 ppm
<0.01%,
<100ppm
<0.01%,
<100ppm
0.01%
<100 ppm
0
Table 2. Examples of carbon dioxide stream composition.
Major differences
There are major technical and economic differences to be
considered in CCS schemes. The demand-led EOR schemes
in which the need for oil production places a requirement
on the provider and the provision of carbon dioxide. The
compressor operating regime is dictated by the production
rate required and the geological configuration of the
reservoir. In CCS schemes, the same geological factors
dictate to the transport system the operating conditions and
flow rates, but there is also the influence of the upstream
technology. The power plant or industrial process does not
fit the same operating profile as a geologic storage. Carbon
dioxide from CCS-enabled plant must be accommodated,
or the emitter will have to free vent, incurring penalties and
making the idea of a CCS scheme redundant.
This has an impact on the design of the infrastructure,
particularly the downstream configuration at the storage site.
Whilst the influences from the storage sites are common
with EOR, the addition of deep saline formations to the
mix adds another level of complexity and another series
of unknowns.
The upstream processes also vary and whilst not relevant
for a bespoke source-to-sink design, they are for network
considerations. There are a number of issues that need to
be addressed in the system design, and key amongst these
has to be the composition of the carbon dioxide stream
entering any transport system. Two considerations have
to be made here: firstly safety, and secondly technical. In
terms of safety, the impact of contaminants needs to be
considered alongside carbon dioxide. It is not enough
to model the dispersion of a carbon dioxide stream, but
also its constituent parts must be modelled. This has been
done by the Dynamis [19] project, which recommended
the specification in Table 2.
Typically entry into a US scheme is similar; the Canyon
Reef project advises [20] the following specification for
carbon dioxide:
• 95% mol carbon dioxide minimum
• 0.489g/m3 (250ppm wt) water in the vapour phase,
no free water
• 1500 ppm (w/w) hydrogen sulphide
• 1450 ppm (w/w) total sulphur
• 4% mole nitrogen
220
The Journal of Pipeline Engineering
Fig.8.Typical vortex meter and
contaminant analysis.
• 5% mole, < -28.9°C dewpoint for hydrocarbons
• 10ppm (w/w) oxygen
The Dynamis project considered not only the technical
issues, deviation of properties, density-phase envelopes,
but also the safety aspects. As a result the criteria for
H2S, CO, SOx, and NOx are established on health and
safety grounds.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
In terms of design safety, the methods are almost global,
although this is one of the major issues in the industry
today. Whilst we can draw methods and experience from
the US and Canadian pipelines, there is an underlying
issue. It can be argued that the EOR-based systems are
more tolerant, more conservative in their approach, and
with lower population densities, and income-generating oil
revenue from EOR can afford this. The example to consider
here is the approach to dispersion modelling adopted in
the US. In the example there is a single assumption that a
percentage of the carbon dioxide immediately upon release
forms a solid which falls to the ground [21]. The remaining
vapour is then modelled using the EPA’s Aloha programme
or the models DEGADIS or SLAB. The rate of flow from
the pipeline is also simplified to a common assessment.
This methodology is conservative and the approach in
Europe is much more precautionary. Hence a more defined
understanding is desired in understanding key risks, such as
fracture propagation and dispersion modelling. It must be
stressed that the US approach is not wrong, but the driver
in CCS appears to be much more of a considered approach,
more accurate and more economic.
The dynamics of a CCS system become more complex when
considering networks. Large networks of CCS infrastructure
are already proposed in Scotland, as well as in the Humber,
Mersey and Dee, Thames, Teesside, and Rotterdam regions.
These clusters range from 20 million to 90 million tonnes
per year of carbon dioxide, and link diverse emitters in
industry and power generation. How these networks behave
and cope in different operating scenarios is critical in terms
of both the design of the system and also, more importantly,
the economics.
In terms of system dynamics and the interactions of all
the processing elements, CCS has a series of different
operating modes to consider at both ends. The preference
for storages is to be a generally constant flow, whereas
power station emitters are cyclic and diurnal, and here is
little scope for change between the two. The driver here is
economics and affects the capture plant as well: the emitter
can vent if a storage closes or requires a reduce inlet rate or
pressure, but there will be an associated cost penalty. The
simpler systems in the US with their demand-led supply
can only provide some of the operational dynamic models
that CCS requires. It is likely that the storage will have to
be flexible enough in terms of dispersed entry points or
multiple storage options.
Flow measurement in carbon dioxide is also acknowledged
as a possible challenge. Previous experience shows that two
measurements are made: contaminants and flow. Typically
either orifice or vortex flowmeters (Figs 7 and 8) in pipeline
systems are insulated to limit temperature-induced density
changes, and all meters are fitted with flow computers. The
major difference in CCS is that the metering scheme will
have to be compliant with the EU ETS requirements, and
the meters will have to be of a fiscal standard. Whether this
is possible is the subject of continuing research [22], and
careful consideration must be given to the requirements of
EU ETS when metering flow.
To gain funding for the provision of networks, building them
to cope with future additions, or ‘right sizing’ as it has become
known, mandates that the economics and tariffs that could
be expected need to be clear. This is not the funding model
elsewhere, and for investment to move forward in CCS, the
issue needs to be clearly understood. The need for networks
are fundamental: CCS may not succeed if the reliance is on
multiple source-to-sink solutions. Cluster networks enable
significant savings over the collected costs of A to B solutions.
In these areas, whilst current learning can be applied, CCS is
fundamentally different. The learning required here to give
confidence to investors, industry, and other stakeholders
cannot come from the current experience or knowledge
alone. Some of the lessons industry needs to learn will
only come from the first CCS schemes and a high rate of
knowledge sharing.
Issues and approaches
There are a number of issues that need to be considered in
transportation and pipelines. These issues do not prevent
4th Quarter, 2010
221
development, but their resolution would aid both technical
and economic decisions. The technical challenges are already
known, but an understanding of how to address them needs
to be developed. Some good practice-based guidelines by
the Energy Institute and DNV aim to resolve some of these
issues, whilst research programmes aim to address gaps in
the knowledge. Typically existing experience has either
highlighted the issue, or it is specifically CCS-related.
At the other end of the scale, the storage companies need to
understand that the power utilities and emitters may dictate
the system parameters, not the storage. The dynamics of these
systems are different and the future operators and owners
need to understand this. There is a cultural shift within
corporations, business models, and individual sites that
needs to be considered and addressed. When deployment of
CCS comes, it must be received by an intelligent, informed,
and correctly-resourced workforce with training complete.
When considering resources, there needs to be an
understanding of the market size and rate of deployment.
Any evaluation of the CCS market is made with caution due
to the number of dependencies involved. The IEA roadmap
[25] provides one such analysis, stating the expectation that
by 2020 we need 100 projects, by 2030 we need 850, and
by 2050 we need 3400 projects, in order to meet the BLUE
map scenario. Projects rely on three issues: the supply chain,
engineers to design, and skilled technicians to construct. How
many people does it take to design and build 100 projects in
10 years? The engineers in the marketplace now are needed
in existing areas. New and emerging areas like CCS and
biofuels, or even new nuclear construction, will draw on
the same resource pool and supply chain. The CCS industry
and academia must address these issues rapidly and build
capacity now, ready for delivery to the market in a decade.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Flow-assurance guidelines for carbon dioxide, and an
understanding of the fluid behaviour, needs to be clearer.
Experience in the US [23] indicates that there is an
issue with surge and transient pressure. Because the US
experience is restricted to onshore pipelines, and with
offshore pipelines anticipated for CCS, there is a concern.
Few mitigation measures can be added to offshore pipelines,
particularly those that terminate at subsea completions.
In this discipline there are also concerns that, even at
low water content in the fluid, carbon dioxide clathrates
(hydrates) may form. The definition of clathrate formation
behaviour, and this area of vapour-liquid equilibrium,
clearly require further research.
emitters, particularly power generators, face in the addition
of a capture plant, pipeline, and storage are considerable.
The current staff will have new processes to control and
monitor, and chemical stocks to maintain and dispose of.
Whilst these processes are common to the energy industry,
they are not common to the power-generation companies.
Importantly, the resolution of physical properties of possible
fluid streams remains an issue. Equation-of-state selection
is covered in multiple academic papers, but empirical data
to support the assumptions and outputs of the predictive
methods are needed. Whilst engineering design can be
achieved using the predictions, the design margins applied
may prove to be excessive. This is particularly of interest
in phase-envelope predictions to protect systems against
multi-phase flow. The data must cover the range of possible
contaminants and process conditions for a pipeline system.
Dispersion modelling is already acknowledged as an issue,
and some experimental work has been done by BP [24] and
Scottish and Southern Electricity for the now-cancelled DF1
project. Further modelling work is also planned by DNV
in the Pipetrans research programme. The issues around
dispersion are complex. Clear dispersion modelling and the
behaviour of a depressuring pipeline are critical elements
in both determining the major accident response and
also in defining safe distances. The behaviour of carbon
dioxide at the release point, the source terms, needs to be
defined, and the computational models validated against it.
Computational-fluid-dynamic models can be used instead
of the simpler commercial programmes. However, the
efficient and accurate modelling which is at an early stage
for the purposes of safety cases and route definition is key
to efficient, practical, and safe design.
Human factor
While the critical technological issues discussed here aid
design, safety, and integrity of the pipeline, the human
factor should not be ignored. Important in the learning
from the US is that the schemes there are operated in an
oil and gas production environment. The changes that the
Perhaps the biggest potential issue for the industry to face
is the public. There is a growing need to address public
education and perception. Poor responses in Continental
Europe and opposition in the USA are already showing
public resistance to storage and pipelines. These issues
already exist in all fields, and experience can be drawn
from the UK’s position on gas storage. The reserve capacity
in the UK gas supply is approximately 4% of the annual
consumption, which compares poorly to that of France at
24% or Germany at 21%. The number of potential projects
to rectify this is significant, but the majority of projects suffer
in planning, face local opposition, or enter public enquiry,
lengthening the development time and cost. Despite the
urgent need for capacity, the poor educational message and
engagement has prevented a number of projects proceeding
in a timely fashion. This challenge now faces CCS: failure
to engage with the public will have devastating effect on
project development. There needs to be project-specific
information, but also a wider education programme.
Conclusions
In conclusion, it is apparent that the knowledge gained from
the current carbon dioxide pipelines will prove vital. It is
222
The Journal of Pipeline Engineering
not the only the experience and solutions, but also those
issues that have not been fully addressed. There are critical
areas to address, including flow assurance, dispersion,
properties, and engagement. These need to be resolved to
enable deployment.
However the body of evidence, experience, and proven
design methodology, codes, and regulation all enable
pipelines and infrastructure to be designed. While these
parameters may prove to be conservative and therefore
more costly, it does not prevent pipelines from moving
forward. Nor can CCS afford to wait: design for the
transport infrastructure is underway now. The industry
must maximize the transfer of knowledge from carbon
dioxide and hazardous liquid pipeline design, while parallel
research must address the issues discussed here and generate
resources and tools that can meet the challenge to bring
to wide-scale deployment.
References
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
1. www.bts.gov/publications/national_transportation_statistics/
html/table_01_10.html
2. CFR – Code of Federal Regulation.
3. Building the cost curves for CO2 storage: European sector.
Report 2005/2 IEA GHG, UK.
4. M.Mohitpour, 2007. Pipeline design and construction: a
practical approach. ASME Press, New York.
5. J.Barrie. Carbon dioxide pipelines: a preliminary review
of design and risks. 7th Int. Conf. on Greenhouse Gas
Technologies. http://uregina.ca/ghgt7/PDF/papers/
peer/126.pdf
6. M.Mohitpour, 2008. A generalized overview of requirements
for the design, construction, and operation of new pipelines
for CO2 sequestration. The Journal of Pipeline Engineering.
7. A.Oosterkamp and J.Ramsen, 2008. State-of-the-art overview
of CO2 pipeline transport with relevance to offshore pipelines.
Polytec, Norway.
8. Idem, ibid.
9. F.Schremp and G.Roberson, 1978. Effect of supercritical
carbon dioxide on construction materials. Society of Petroleum
Engineers.
10. G.Najera, 1986. Maintenance techniques proven on CO2 line.
Oil and Gas Journal.
11. T.Gill, 1985. Canyon Reef Carriers Inc, CO2 pipeline
description and 12 years of operation. ASME.
12. M.Seiersten, 2001. Material selection for transportation and
disposal for CO2. Corrosion 2001.
13. L.Newton, 1977. Corrosion and operational problems, CO2
project, SACROC Unit. Society of Petroleum Engineers.
14. M.Mohitpour, 2007. Pipeline design & construction: a practical
approach. ASME Press, New York.
15. de Visser et al., 2007. Towards hydrogen and electricity
production with carbon dioxide capture and storage. Dynamis
Consortium.
16. http://www.phmsa.dot.gov/pipeline
17. http://www.eia.doe.gov/pub/oil_gas/natural_gas/analysis_
publications/ngpipeline/
18. J.Gale and J.Davison, 2004. Transmission of CO2 – safety and
economic considerations. Energ,29.
19. de Visser et al., 2007. Towards hydrogen and electricity
production with carbon dioxide capture and storage. Dynamis
Consortium.
20. IPCC, 2005. Special report on carbon capture and storage.
21. www.energy.ca.gov/sitingcases/hydrogen_energy/documents/
applicant/revised_afc/Volume_II/Appendix%20E.pdf
22. A study of measurement issues for carbon capture and storage
(CCS). Report 2009/54, April 2009, TUVNEL.
23. M.Mohitpour, 2007. Pipeline design & construction: a practical
approach. ASME Press, New York.
24. http://archivos.labcontrol.cl/wcce8/offline/techsched/
manuscripts%5C8mnkk4.pdf
25. www.iea.org/papers/2009/CCS_Roadmap.pdf
26. IPCC, 2005. Special report on carbon capture and storage.
4th Quarter, 2010
223
The techno-economics of a
phased approach to developing
a UK carbon dioxide pipeline
network
by Saulat Lone1, Dr Tim Cockerill*2, and Prof. Sandro Macchietto3
1 Sui Northern Gas Pipelines Ltd, Pakistan
2 ICEPT, Imperial College London, UK
3 Department of Chemical Engineering, Imperial College London, UK
A
T
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
PHASED APPROACH to developing a CCS pipeline network would see an initial ‘backbone’ system
constructed to collect carbon dioxide from the very largest sources. Once the backbone was in place,
it might be possible to add a large number of smaller sources for relatively little additional cost.This paper
analyses the techno-economics of a phased approach to rolling-out a comprehensive UK CO2 onshore
pipeline network. We have developed a series of idealized scenarios where, initially, a new UK network is
established to carry emissions from large-scale producers of carbon dioxide, defined here as more than
3Mtonnes per annum. In a second phase of development, medium-scale emitters are added to the network.
A final third phase incorporates small producers with emissions in the range 0.5-1 Mtonnes per annum. For
all scenarios, two different approaches to network construction have been compared, one using intermediate
re-pressurization stations and one relying only on initial pressurization. Our results compare the construction
and transportation costs of the different network configurations in each scenario, indicating the cost per
tonne of CO2 transport. While there are some benefits offered to smaller sources by a phased approach,
a rule of diminishing returns operates, with each tier experiencing an increase in marginal transport costs.
The sensitivity of the costs to changes in the network configuration and design assumptions is investigated.
HERE IS CURRENTLY interest in the technoeconomics of constructing a carbon-dioxide pipeline
network for the UK to support the future deployment
of CCS. It is likely that early development will focus on
transporting CO2 from a small number of large carboncapture-fitted power stations to a limited number of offshore
storage sites. As the number of carbon-capture-fitted point
sources increases there may be some benefit from developing
a common pipeline infrastructure. Once a ‘backbone’
network starts to emerge, an appealing possibility is that a
large number of smaller CO2 producers could be connected
to the network for relatively little cost.
To investigate the feasibility of this idea, we have analysed
the techno-economics of a phased approach to rolling-out a
This paper was presented at the First International Forum on Transportation of CO2
by Pipeline, organized in Newcastle upon Tyne in July, 2010, by Tiratsoo Technical
and Clarion Technical Conferences, and with the support of the University of
Newcastle and the Carbon Capture and Storage Association.
*Author’s contact details
email: [email protected]
comprehensive UK CO2 onshore pipeline network. We have
developed a series of idealized scenarios where, initially, a new
UK network is established to carry emissions from large-scale
producers of carbon dioxide. In subsequent phases, smaller
CO2 sources are progressively added to the growing network.
The aims of this paper are to:
• Develop an idealized, but representative, spatial
data set detailing the locations of carbon dioxide
sources and the quantities likely to be produced.
The locations in the data set are disaggregated by
size, such that they may be connected to facilitate
study of a phased development of a carbon dioxide
pipeline network.
• Establish a method for modelling the phased
development of a CO2 network.
• Examine the technical requirements and performance
of the network at each phase, and estimate the capital
costs associated with network roll-out.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
REGISTE
R TODA
Y
Visit the
Clarion
website
to
register
now
www.cla
rion.org
Marriott Westchase Hotel
Houston, Texas, USA
Courses
Conference
Exhibition
Now entering it 23rd year, the PPIM
Conference is recognized as the
foremost international forum for
sharing and learning about best
practices in lifetime maintenance
and condition-monitoring technology
for natural gas, crude oil and product
pipelines.
Plan to be there: www.clarion.org
or call us at +1 713 521 5929
PLATINUM SPONSOR
SILVER SPONSOR
The international gathering of the global pigging industry!
Conference Organizers
4th Quarter, 2010
225
Areas considered
Approximate
Year
References
North West UK
2006
[1]
Scotland, Yorkshire
2008-2010
[2]
East of England
2007
[3]
Scottish Regional Carbon Capture and Storage
(CCS) Study
Scotland and Northern
England
2009
[4]
Yorkshire Forward Study
Yorkshire and Humber
2008
[5]
Europe Wide
2008
[6]
Study
Feasibility Study on the Transmission of CO2
CO2 Aquifer Storage Site Evaluation and Monitoring
(CASSEM)
North Sea Basin Task Force Study
Ramp-up of large-scale CCS infrastructure in Europe
Table 1. Selected studies of UK CCS pipeline infrastructure roll out.
Brief UK literature overview
For the tier of the largest producers, the GIS was used to
identify corridors along which a backbone network of carbon
dioxide pipelines could be constructed. The pipeline network
was then designed using hydraulic analysis techniques. Finally
a simple unit cost per in-km approach was used to estimate
construction costs, assuming all pipelines are newly built.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Most work on the development of CO2 infrastructure for
the UK has been performed by commercial consortia, with
a focus on a regional development approaches. Figure 1
summarizes the areas considered by the first five major
studies listed in Table 1.
of a phased pipeline roll out, the sources were classified
into three tiers according to their annual CO2 production.
The literature suggests that links between regions have
not been considered for developing a country-wide central
CO2 transportation network. One drawback of regional
approaches is the potential underutilization of storage
capacity. Similarly, local CO2 transmission and storage
capacity constraints could limit the extent to which CCS
could be deployed within any region. This is particularly true
in the North West of the country, where storage capacity is
limited, but there are a large number of CO2 sources. Finally,
a series of optimizations at local level is likely to deliver
suboptimal solutions for the UK as a whole. Integration of
regions with a central CO2 transmission system will help
maximize utilization of both capture and storage potential.
A similar approach was used to add all sources in each of
the smaller production tiers to the backbone network. This
allowed the implications of a phased approach to network
development to be assessed, and in particular the marginal
cost of adding smaller sources to be estimated. Calculations
were carried out for networks with and without intermediate
recompression stations.
Identification of export terminals
and sources
Analytical approach
Fig.1. Areas considered by UK pipeline studies.
Figure 2 illustrates the methodology adopted for this study,
which only considers the development of onshore pipelines
connecting point carbon dioxide sources to a limited number
of export terminals located on the coast. Further offshore
pipelines will carry the CO2 to offshore storage locations,
but the development of these components of the network are
not examined here. In common with several other studies,
it is assumed that onshore/offshore connections will only
be permitted at existing pipeline terminals. Hence the first
stage of the analysis was to identity existing UK pipeline
terminals and their locations.
Secondly, a spatial database of existing and prospective CO2
producers was developed within a geographical-information
system (GIS) drawing on data from several sources. Due to
the large number of potential sources, the database only
included those above a threshold value. To facilitate study
Identification of CO2 export terminals
The UK’s existing oil and gas terminals and the nearest
offshore oil and gas sedimentary basins with CO2 storage
potential are summarized in Table 2. These CO2 export
terminals would be equipped with compressor or pumping
units for export of CO2 to the offshore storage. The storage
potentials mentioned in the Table represent the realistic
storage potential for each basin suggested by the British
Geological Survey (BGS), which has been divided against
each CO2 export terminal. At present, it is difficult to
establish when individual fields or transport pipelines will
become available due to commercial confidentiality of
information. Therefore, for this study, it is assumed that
UK’s offshore oil and gas fields will be available for CO2
storage when required.
226
The Journal of Pipeline Engineering
Name of terminal
St Fergus Gas terminal
Nearest UK offshore Oil & Gas
sedimentary basin
CO2 Storage Capacity
Northern & Central North Sea basin
1,346
Southern North Sea Basin
3,886
East Irish Sea Basin
1,043
Teesside Terminal
Easington/Dimlington gas terminal
Theddlethorpe gas terminal
Bacton gas terminal
Point of Ayr terminal
Barrow-in-Furness gas terminal
Table 2. UK existing onshore oil and gas terminals.
Types of emitter
Tier-0
3 million and above
Coal & CCGT Power stations, Refineries, Steel industry
Tier-1
1 million – 3 million
CCGT & Oil Power stations, Refineries, Cement factories, CHP
Tier-2
0.5 million – 1 million
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
CO2 Emissions Range
Tonnes per annum
Cement factories, CCGT Power stations, fertilizer, petrochemical complexes
Table 3. Classification of emitters according to emissions.
Design Approach
Pipeline based
Pipeline
+
Compression based
Scenario name
Captured CO2
volumes
References
S1A
156
Design “pipeline based” transmission network for
Tier-0 emitters
S2A
156 + 57 = 213
Design “pipeline based” transmission network for
Tier-0 and Tier-1 emitters
S3A
213 + 15 = 228
Design “pipeline based” transmission network for
Tier-0, Tier-1 and Tier-2 emitters
S1B
156
Design “pipeline + compression based” transmission network for Tier-0 and Tier-1 emitters
S2B
156 + 57 = 213
Design “pipeline + compression based” transmission network for Tier-0 and Tier-1 emitters
S3C
213 + 15 = 228
Design “pipeline + compression based” transmission network for Tier-0, Tier-1 and Tier-2 emitters
Table 4. Definition of pipeline design scenarios.
Selection of CO2 sources
For this study, all current and planned (to 2015) industrial
and power station CO2 emitting sources in the UK with CO2
emissions greater than 500,000t/a have been considered. In
total, these sources account for 228Mt CO2 – approximately
50% of UK current CO2 emissions.
The cut-off threshold is broadly consistent with the UK
Government policy that any new combustion power station
at or over 300MWe should be built ‘carbon-capture ready’
(CCR), as existing coal-fired power stations in UK with the
threshold capacity would be expected to emit approximately
500,000t/a of CO2. The sources considered here therefore
include larger power stations and large industrial sources.
Non-power plant industrial sources of this scale comprise
refineries, steel manufacturers, petrochemical complexes,
fertilizers and cement factories. CO2 emissions data for
the selection process were taken from a variety of sources,
including the Environment Agency, published data under
the EUETS programme, and the annual reports of each
emitter. Figure 3 illustrates the approach.
227
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
4th Quarter, 2010
Fig.2. Analytic approach used in
this study.
Division of emitters by CO2 emissions range
Emitters above the threshold value have been categorized
into three tiers as set out in Table 3. The division points
were selected based on engineering judgment, with each
tier distinguishing decreasing importance of the individual
emitter for UK greenhouse gas emissions.
• Tier-0 emitters: Tier-0 comprises emitters producing
CO2 emissions of 3Mt/a and above, typically large
coal-fired power stations with generation capacity
from 1.5-4.0GWe. The largest emitter in this category
is Drax power station, with the largest industrial
emitter being Corus’ steel works. In total there are 20
existing power stations, five industrial installations,
and five new power projects, totalling 156Mt/a (68%
of the total UK emissions considered).
and oil power stations with generation capacities in
the range of 300–500MWe. Across the tier there are
13 existing power stations, 16 industrial installations,
and three new power projects, totalling 15Mt/a (7%
of the total UK emissions considered).
Pipeline development scenarios
The onshore CO2 pipeline transmission network for UK
is designed in this study by considering following capacity
phased approach:
• Scenario 1 (S1): Evaluate the design and topology
of a CO2 onshore ‘backbone’ transmission network
to collect CO2 emissions from Tier-0 emitters only.
• Tier-1 emitters: Tier-1 emitters are typically CCGT
power stations and refineries, producing CO2
emissions in the range of 1–3Mt/a. The total inventory
includes 19 existing power stations, 10 industrial
installations, and eight new power projects, totalling
57Mt/a (25% of the total UK emissions considered).
• Scenario 2 (S2): Re-evaluate the design of an onshore
CO2 transmission network where Tier-1 emitters are
added to the already laid transmission system for
Tier-0 emitters. Determine the additional changes
required in the transmission network to handle
combined flows of Tier-0 and Tier-1 emitters to the
CO2 export terminals.
• Tier-2 emitters: Most Tier-2 emitters are industrial
sources, typically cement and petrochemical plant,
with emissions ranging from 0.5-1Mt/a. Power stations
included in the Tier-2 band are predominantly CCGT
• Scenario 3 (S3): Complete the design of the onshore
CO 2 transmission system by considering the
connection of Tier-2 emitters to the transmission
system already in place for Tier-0 and Tier-1 emitters.
228
The Journal of Pipeline Engineering
Fig.3. Approach to developing CO2
sources spatial database.
flange of the CO2 compressor installed at export terminals
to inject gas in offshore pipelines for subsequent storage in
offshore oil and gas fields.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Two separate types of transmission models have been
developed for each scenario, one based entirely on
pipelines without re-pressurization (A) and the other using
combinations of pipeline and intermediate compression
stations (B). The full set of scenarios is summarized in Table 4.
Data and assumptions
Pipeline design assumptions
Many factors have to be considered in the design of the
new pipelines (assumed buried underground), including
the properties and quantities of the fluid to be transmitted,
underground conditions, and safety requirements. Some key
data are set out in Table 5, with further discussion following.
It is assumed here that CO2 transportation systems will
be based on new onshore carbon steel pipelines, with a
maximum diameter of 1,067mm. No re-use of existing
infrastructure is accommodated within the study. The
quantities of CO2 provided to the network by each source
have been estimated by taking a carbon dioxide capture
efficiency of 90% for power stations and 60% for industrial
installations.
The networks are designed for carbon dioxide in a
supercritical state. Keeping in view the critical point of
pure CO2, i.e. 74 bar, the minimum pressure at which CO2
would leave each source is taken to be 95 bar. This allows
for a pipeline network pressure drop of 20 bar above the
CO2 critical pressure. The operating pressure of the pipeline
transmission network is specified at 100 bar due to the
limitation of the operating pressures of the flanges and
fittings at intermediate facilities. For intermediate booster
stations, where used, it is assumed that recompression would
be required when the pressure drops to 85 bar after which
the pressure will be boosted again to 100 bar. It is assumed
that arrival pressure at each CO2 export terminal would be
85 bar. This arrival pressure is the pressure at the upstream
Another assumption is that CO2 streams from all the
sources will be dehydrated to -5°C dewpoint, representing
the temperature in the CO2 mixture at which water will start
condensing. It is assumed that for Tier-0 and Tier-1 emitters,
CO2 drying facilities would be installed at the premises as a
part of the CO2 capture plant. The CO2 streams from Tier-2
emitters would be collected together at different locations
for centralized drying. For deciding the -5°C water dew
point of CO2 streams from each emitter, the available UK
soil temperature data at 30cm depth for peak winter month
(January) has been used. Since the majority of emitters are
located in England, an average subsurface soil temperature
of 3–5°C is taken for this study.
Network topology
A crucial factor in the design of any pipeline transmission
network is the determination of pipeline route corridors.
In this study, it is assumed that wherever feasible, the CO2
transmission network will follow the existing route corridors
of onshore oil and gas pipelines in the country. However
some new pipeline route corridors have been proposed for
the emitters which are located away from existing oil and
gas onshore pipeline infrastructure.
Cost data
Detailed construction costings for CO2 pipelines are difficult
to obtain thanks to the limited number of CO2 pipelines
operating worldwide. Instead, we assume that new carbon
steel pipelines will be laid, and rely on approximate cost data
for natural gas pipelines construction drawn from the IPCC
special report [7] and the Oil and Gas Journal 2005-2008
[8]. In this study, the base unit pipeline total construction
cost (including materials, labour, and all works) is taken to
be US$ 30,000/in–km, including pipeline material costs,
4th Quarter, 2010
229
Fluid
CO2 purity
100%
Phase of CO2
Supercritical
Critical Temperature
31°C
Critical Pressure
74 bar
Pipeline
PN100 (100 bar nominal operating pressure)
Standard used for pipeline fittings and equipment
DIN2512
Pipeline Material
A105-Carbon Steel
Standard used for pipeline design criteria
BS EN 14161 / BS EN 1594
Maximum allowable operating pressure of pipeline network
110 bar
Pipeline internal design pressure
100 bar
CO2 pressure leaving emitter’s premises
95 bar
CO2 temperature leaving emitter’s premises
35 oC
CO2 arrival pressure at export terminals
85 bar
Minimum pipeline diameter
323.9 mm
Maximum pipeline diameter
1,067 mm
Onshore pipeline buried depth
1.2 - 1.8 m
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Pressure rating of valves & fittings
CO2 capture plants
Efficiency of CO2 capture plant at power stations
90%
Efficiency of CO2 capture plant at industrial installations
60%
Table 5. Summary of pipeline design assumptions.
labour costs, and costs related to intermediate facilities, but
excluding additional compression. The impact of pipeline
construction cost increases over this base level is later
evaluated by performing sensitivity analysis.
The cost of intermediate booster stations, where employed,
has been estimated based on literature data. A review of
several studies [9, 10, 11, 12] suggests that capital costs for
installed compressor stations would range from $1500-4800/
kW. This wide variation can be attributed to differences in
the year of installation, type of compression, geographical
location, and installation of intermediate heating or
cooling facilities. For the results reported here, a value
for compression station costs towards the middle of the
literature range, specifically $2500/kW, is adopted. This
encompasses the cost of buildings, compressor units, prime
mover, pipeline fittings, and drying equipment.
Only the costs of intermediate booster stations are accounted
for, with compression at sources and export terminals excluded
from the calculations. However, for the booster stations, it is
assumed that 50% of the compression requirement would
be available as stand-by in order to make up in case of any
emergency or during maintenance of main unit.
All cost data have been converted to UK Pounds Sterling
using 2008 exchange rates.
Hydraulic modelling
Introduction
For designing the CO2 pipeline transmission system, a stateof-the-art hydraulic simulator, PipelineStudio version 3.0
has been used to simulate the transmission pipeline system
for steady-state operation. For all scenarios, the pipeline
network model has been constructed using:
pipeline lengths and elevations from the GIS database,
CO2 volumes from each emitter and location of CO2
export terminals,
subsoil temperature data depending upon the region.
Simulation methodology
For the hydraulic simulation of the CO2 transmission
network, the arrival pressure at each CO2 export terminal
has been fixed at 85 bar whereas the captured CO2 volumes
are fed at the source nodes. The pipeline diameters are then
230
The Journal of Pipeline Engineering
Fig.4. CO2 transmission network for (A) Tier-0 emitters, (B) Tier 0+1 emitters, (C) Tier 0+1+2 emitters. Numbers 1,2,3…
denotes CO2 export terminals as mentioned in modelling results table.
where
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
optimized to keep the calculated pressures of each source
within the pipeline operating limits and above the critical
pressure of CO2.
For estimating the thermodynamic properties of the CO2
mixtures, the Peng–Robinson equation of state has been
used. While keeping the input flows constant, the pressures
at each emitter are then back-calculated by the software by
using the ‘general flow equation’ after Menon [11]:
Q = gas flow rate, measured at standard conditions, ft3/day (SCFD)
f = friction factor, dimensionless
Pb = base pressure, psia
Tb = base temperature, R (460 + F)
P1 = upstream pressure, psia
P2 = downstream pressure, psia
G = gas gravity (air =1.00)
Tf = average gas flowing temperature, R
L = pipe segment length, mi
Z = gas compressibility factor at the flowing temperature, dimensionless
D = pipe inside diameter, in
For compression-based scenarios, pipeline diameters of the
transmission model have been optimized while considering
intermediate booster stations. The intermediate booster
stations are placed where the pipeline pressure drops to 85
bar after which the pressure is boosted again to 100 bar. The
power requirement at each booster station are then calculated
by the software using Eqn 2, again after Menon [11]:
where
HP = compressor horsepower
γ = ratio of specific heats of gas, dimensionless
Q = gas flow rate, MMSCFD
T1 = suction temperature of gas, oR
P1 = suction pressure of gas, psia
P2 = discharge pressure of gas, psia
Z1 = compressibility of gas at suction conditions, dimensionless
Z2 = compressibility of gas at discharge conditions, dimensionless
ηa = compressor adiabatic (isentropic) efficiency, decimal value
Mass balance
The pipeline transmission model drawn in the network
has been verified by balancing the inputs and outputs of
the transmission network. For steady-state calculations it is
assumed that there will be no accumulation of CO2 volumes
in the pipeline transmission network.
Results
Main network characteristics
Table 6 summarizes the main technical results of the
simulation and optimization study, and the network layouts
are shown in Fig.4. It is clear that the pipeline-only scenarios
(A) have pipeline diameters larger than the scenarios where
re-compression is used (B). Without re-compression, larger
4th Quarter, 2010
231
Emission Sources:
Tier-0
Scenario
S1A
Tier- 0 + 1
S1B
S2A
Tier- 0 + 1 + 2
S2B
S3A
S3B
Basis: Power station CO2 capture Eff = 90% , Industrial CO2 capture Eff = 60%
CO2 volume reaching export terminals
Mtpa
34
34
71
58
75
75
2-Easington Gas Terminal
Mtpa
46
46
55
55
56
56
3-Point of Ayr Terminal
Mtpa
10
10
16
16
20
20
4-Theddlethorpe Gas Terminal
Mtpa
44
44
40
53
42
42
5-Barrow-In-Furness Terminal
Mtpa
-
-
-
-
1
1
6-Ireland Platform
Mtpa
-
-
3
3
4
4
7-Teesside Gas Terminal
Mtpa
19
19
25
25
27
27
8-St Fergus Gas Terminal
Mtpa
3
3
3
3
3
3
Sub Total
Mtpa
156
156
213
213
228
228
Pipeline Network Length
Miles
883
883
1359
1359
1559
1559
Network Equivalent Diameter*
mm
3,689
3,200
4,301
3,897
4,632
4,376
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
1- Bacton Gas Terminal
Compression Requirements at compressor stations
Comp-1
MW
2.0
6.6
6.5
Comp-2
MW
4.4
3.3
5.0
Comp-3
MW
5.3
3.5
4.8
Comp-4
MW
3.3
4.2
6.0
Sub Total
MW
0
15.1
0
17.6
0
22.3
Table 6. Summary of main technical results for pipeline networks.
* The term “network equivalent diameter” is used in this study only for indicative purpose in order to make an equal
comparison of results of different scenarios. For each scenario it characterizes the diameter of the entire pipeline network as
single diameter for ease of comparison of the results.
pipeline diameters must be used to meet the pressure
requirements at the CO2 export terminals.
Transmission network capital costs
A summary of total costs required to establish the CO2
transmission networks is presented in Table 7. All the
pipeline-only scenarios (A) have higher up-front capital costs
when compared to the total costs of the corresponding recompression based scenarios (B). The higher up-front costs of
scenarios (A) are due to the larger pipeline diameters employed.
Considering the overall CO2 network costs (Table 7
and Fig.5), the results suggest that systems relying on recompression offer more attractive techno-economics than
those without. However, re-compression based networks will
have increased operation and maintenance (O&M) costs in
comparison to pipeline-only systems: these are not accounted
Fig.5. Summary of CO2 costs for all base scenarios.
for here, and more-detailed work is required to determine
the optimum combination of compressor requirements and
related O&M costs.
232
The Journal of Pipeline Engineering
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Fig.6. Marginal tier costs, defined as
the incremental costs divided by the
increase in quantity of carbon dioxide
transported, for all scenarios.
Fig.7. Range of transmission costs
identified across all cases considered.
For each set of scenarios, the cost of building a pipeline
transmission network increases with the addition of smaller
emitters. This trend is reflected in transportation costs,
because the extra pipe length required increases more quickly
than the additional CO2 provided by each tier. Adding
all Tier-1 and 2 sources to the network would produce
an increase in per tonne transportation capital costs of
approximately 15% if no re-compression is used and 22%
in a system that relies on re-compression.
to a network with compression is 44% higher than that for
the Tier-0 sources. The comparable marginal cost increase
for a network without compression is 21%. Incorporating
increasingly small sources demonstrates poorer returns, as
further adding Tier-2 sources to networks with and without
compression provides further marginal cost increases of 94%
and 109% respectively. The value of adding such smaller
sources to a network must be questioned, particularly if they
are remote from natural clusters of carbon dioxide sources.
Analysis
Sensitivity analysis
Marginal tier costs
The marginal costs for each scenario have been calculated
to determine the extra investment required for the addition
of emissions of lower tiers by dividing the incremental
capital costs of each scenario by the incremental increase
in emissions. The results are shown in Fig.6.
Marginal costs increase with the addition of emissions from
the lower tiers. The marginal cost of adding Tier-1 sources
For this study, sensitivities have been calculated to determine
the impact of different parameters on CO2 transportations
costs. Results of each sensitivity analysis have been compared
with a base case.
Change in pipeline network length
This sensitivity analysis has been performed to determine
the impact of 15% and 20% increases in pipeline length on
CO2 transportation cost, perhaps due to the need to avoid
4th Quarter, 2010
233
Pipeline only based scenarios
Basis: Power station CO2 capture Eff = 90% , Industrial CO2 capture Eff = 60%
Capital Cost
CO2 Transportation Cost
Captured
Emissions
Pipeline
Compression
Total
Pipeline
Compression
Total
Mtpa
M£
M£
M£
£ / Tonne
£ / Tonne
£ / Tonne
S1A
156
952
0
952
6.12
0.00
6.12
S2A
213
1,375
0
1,375
6.46
0.00
6.46
S3A
228
1,608
0
1,608
7.05
0.00
7.05
Scenario
Pipeline + compression based scenarios
Basis: Power station CO2 capture Eff = 90% , Industrial CO2 capture Eff = 60%
CO2 Transportation Cost
Pipeline
Compression
Total
Pipeline
Compression
Total
Mtpa
M£
M£
M£
£ / Tonne
£ / Tonne
£ / Tonne
S1B
156
822
38
860
5.28
0.24
5.53
S2B
213
1,270
44
1,314
5.97
0.21
6.18
S3B
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Capital Cost
Captured
Emissions
228
1,490
56
1,546
6.53
0.24
6.77
Scenario
Table 7. Summary of CO2 transmission network costs for pipeline and compression-based scenarios.
geographical obstacles not accounted for in the relatively
high-level study. For the base-case scenario, the pipeline
route was evaluated using a GIS database. It was assumed
that pipeline diameters, CO2 emissions, and pipeline unit
construction costs remained constant in all cases. However,
in scenarios involving re-compression, compression powers
were calculated again.
Further hydraulic simulations determined the operating
pressures of the transmission network and pressures required
at sources. For a 15% increase in pipeline length, pressure
at sources increased by an average of 2% as compared to the
base-case pressures. Similarly for a 20% increase in pipeline
length, the pressure at sources increased by an average of
3%. In all cases, the calculated pressures remained below
the maximum allowable pipeline operating pressures as well
as adequately above the CO2 critical pressure.
Costs were then re-evaluated for the increased in pipeline
lengths, unsurprisingly showing an increase in transmission
costs in all cases (Fig.7).
Change in CO2 capture plant efficiency
This sensitivity analysis has been performed to determine
the changes in CO2 transmission network resulting from
changes in the CO2 removal efficiencies of the capture plants
installed at each point source. The base-case assumption
in this work is that the CO2 removal efficiency for power
stations is 90% and 60% for the industrial installations. Any
change in capture efficiencies will increase or decrease the
volume of carbon dioxide flowing though the network, and
hence the design and infrastructure requirements.
Two perturbations of the base-case were examined, changing
the capture efficiency at the power station sites to 85% and
95%. It was assumed that CO2 capture efficiency of the
industrial units remained constant at the base-case value.
Using the new flow volumes, new hydraulic simulations were
carried out, assuming that pipeline lengths and route corridors
will remain unchanged. Only changes in pipeline diameters
and compression requirements have been re-calculated.
In all cases the network pipeline diameters increased
or decreased in accordance with the change in carbon
dioxide flow volumes. For the compression-based cases,
the compression energy requirements increased for the
95% capture case and decreased for 85% capture case.
However for the 95% capture scenario, CO2 transportation
cost decreased compared to the base-case whereas for 85%
capture scenario, the unit cost increased. The decrease in CO2
transportation cost for 95% case scenarios is countered by
increased capital cost due to using larger-diameter pipelines.
Similarly for 85% scenario, pipeline diameters have been
reduced due to a reduction in emissions volume. The change
in CO2 transportation costs in all cases is comparatively
small, at around 2%.
Increase in pipeline construction costs
This sensitivity analysis investigates the impact of increased
pipeline unit construction costs, representing the effect
234
The Journal of Pipeline Engineering
of increased steel prices, difficult construction terrain, or
other project-management-related factors. As part of this
sensitivity analysis, a sub-analysis has been conducted to
check the impact of increased pipeline construction costs
on the previously-calculated sensitivities to CO2 capture
efficiencies and increased pipeline lengths.
The results of the combined sensitivity analysis shown in
Fig.7 indicate the range of CO2 transportation costs across
all the cases considered. In all cases the transportation
costs of compression-based scenarios are less than for
non-compression scenarios. The average costs of the three
scenarios – shown as small, light-coloured squares in Fig.7
– range from 8.24 £/tonne for scenario S1 to 9.2 £/tone
for scenario S3; the cost variability within each scenario is
rather larger, approximately 7-8 £/tonne.
Conclusions
These general trends are hardly influenced by changes in the
major underlying assumptions and design choices, although
of course the absolute value of the CO2 transportation cost
does vary substantially.
References
1. GASTEC at CRE Ltd (GaC), 2006. Feasibility study on the
transmission of CO2. GaC Report 3484, October [Available
at www.gastecuk.com/case-studies-detail.php?id=3 ].
2. G.Pickup, 2009. CASSEM Overview, November, www.geos.
ed.ac.uk/ccs/UKCCSC/Pickup.pdf [Consulted May 2010].
3. UK DBERR. Development of a CO2 transport and storage
network in the North Sea,
[Available at www.nsbtf.org/documents/file42476.pdf ].
4. Scottish Centre for Carbon Storage & Scottish Government,
2009. Opportunities for CO2 storage around Scotland. [www.
geos.ed.ac.uk/sccs/regional-study/CO2-JointStudy-Full.pdf].
5. Yorkshire Forward, 2008. A carbon capture and storage network
for Yorkshire and Humber.
[Available at: www.yorkshireforward.com/sites/default/files/
documents/Yorkshire%20%20Humber%20Carbon%20
Capture%20%20Storage%20Network.pdf].
6. J.Kjärstad and F.Johnsson, 2008. Ramp-up of large-scale CCS
infrastructure in Europe. Int. J. of Greenhouse Gas Control, 2,
4, pp417-438, October.
7.E.S.Menon, 2005. Gas pipelines hydraulics. CRC Press, May.
8. Ramgen Compressors, 2010. Ramgen’s low-cost, highefficiency CO2 compressor technology. www.ramgen.com/
apps_comp_unique.html [Consulted May 2010].
9. D.Simbeck and E.Chang, 2002. Hydrogen supply cost estimate
for hydrogen pathways scoping analysis. US National Renewable
Energy Laboratory (NREL) Report/SR-540-32525, November.
10. US Department of Energy, 2007. Conceptual engineering/
socioeconomic impact study – Alaska spur pipeline : Appendix
3-5: compressor cost estimate. Report on DOE-NTL Contract
No DE-AM26-05NT42653, January. [Available at www.
jpo.doi.gov/SPCO/DOE%20Spurline%20Documents/
Appendix%203-5%20Compressor%20Cost%20Estimate.pdf]
11. B.Metz, O.Davidson, H.de Coninck, M.Loos, and L.Meyer,
(Eds), 2005. IPCC special report on carbon dioxide capture
and storage, IPCC.
12. Oil and Gas Journal (various issues, 2005-2008). PennWell
Petroleum Group.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Approximately 50% of UK industrial and energy CO2
emissions are produced by emitters that generate more than
500,000Mt/a of CO2. Providing a CO2 pipeline transport
network for each of these sources has the potential, assuming
they are also retro-fitted with carbon capture equipment, to
facilitate a major reduction in UK CO2 emissions.
without re-compression. The CO2 transport cost per tonne
are overall smaller, as the expense of the compression stations
is outweighed by the reduced cost of smaller-diameter pipes.
Our simplified estimates take no account of operations
and maintenance costs, meaning this difference is certainly
within the range of uncertainty for the results. More-detailed
work is required to determine the optimum combination of
compressor requirements and related O&M costs.
The conceptual design and techno-economics of a phased
approach to rolling out such a network have been investigated,
subject to a number of simplifying assumptions. If a
‘backbone’ network connecting the very largest sources is
constructed first, smaller sources could be later added to the
network with a relatively small impact on the transportation
cost per tonne of carbon dioxide. Typically, adding all Tier-1
and-2 sources to the network would produce an increase
in per unit transportation costs of approximately 15% if
no re-compression is used and 22% in a system that relies
on re-compression.
Considering the marginal cost of making these additions,
however, tells a different story, as increasingly large expense
is required to add sources of rapidly decreasing size. The
marginal per tonne cost of adding Tier-1 sources to a
network with compression is 44% higher than that for
the Tier-0 sources. The comparable marginal cost increase
for a network without compression is 21%. Incorporating
increasingly small sources demonstrates diminishing returns,
as further adding Tier-2 sources to networks with and
without compression provides further marginal per tonne
cost increases of 94% and 109% respectively. The value of
adding such smaller sources to a CCS network must be
questioned, particularly if they are remote from natural
clusters of carbon dioxide sources.
The relatively simple cost analysis carried out for this work
suggests that a system relying on re-compression perhaps
offers a 10% capital cost (about £100m) advantage over one
4th Quarter, 2010
235
Transporting anthropogenic
CO2 in contrast to pipelines
supporting early EOR
by Dr Brian N Leis*, Dr James H Saunders, Ted B Clark, and
Dr Xian-Kui Zhu
Energy Systems and Carbon Management, Battelle Columbus Laboratory, Columbus, OH, USA
T
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
HIS PAPER EXAMINES aspects in quantifying arrest-toughness requirements to control running fracture
in CO2 pipelines. Four key risk and safety discriminators were used to contrast transport of nearly
pure CO2 for EOR to that of CO2 for CCS applications, including: the retrofitting of early EOR pipelines to
provide for fracture control; differing impurities for CCS service that can increase required arrest toughness
as compared to EOR applications; differing transported volume and routeing that lead to increased risk
exposure for CCS pipelines; and technological uncertainties in assessing fracture-control requirements
that develop due to impurity effects. Against this background key elements such as the equation of state
and critical assumptions are evaluated as the basis for establishing practical direction for fracture control.
Finally, the historic design space for many CO2 pipelines supporting EOR is contrasted to that for pipelines
in CCS service.
Whereas some widely recognized reports indicate that technology to design CO2 pipelines is mature,
some significant gaps were identified for CCS applications. The results indicated that arrest toughness is a
very strong function of the minimum CO2 level, with an order of magnitude swing in the arrest toughness
required for a 10% swing in minimum CO2 content as compared to pure CO2, with subtle differences in
the constituents present being a major driver. On-line monitoring of injected streams was suggested to
help manage the related risk. Finally, the often-used Battelle two-curve model adapted to CO2 applications
– while validated in regard to near-pure CO2 applications and cases involving rich (dense-phase) natural
gas – remains unvalidated in application to typical CCS product streams. Such was also the case for many
supporting elements like the equation of state, with an expanded empirical database being key to ensuring
viable fracture-arrest predictions.
T
HE LITERATURE ON schemes to capture and store
anthropogenic carbon dioxide (CO2) reflects the
increase in concern for the effects of greenhouse gas (GHG)
and the need for GHG management. The extent of concern
is evident when ‘GHG + greenhouse gases’ is entered into
a web browser, which recently pointed to 15,300,000 hits.
Tracking the early history of GHG management leads to
sites associated with the intergovernmental panel on climate
change (IPCC, whose formation traces to the late 1980s),
and other agencies such as the International Energy Agency
(IEA), and its focus on GHG (IEAGHG).
Early schemes to capture and store anthropogenic CO2
were based on its injection into the earth at sites local to
its generation, which precludes the need for transport,
and potentially underlies the commonly used acronym
CCS 1 (carbon capture and storage). Over time it became
evident that, in many cases, local CO2 storage sites might not
provide adequate long-term storage integrity, resulting in CO2
seeping back to the surface. This could have consequences such
as litigation, and/or other issues, like verification monitoring
or economic drivers, which are beyond the present scope. As
a result, transport became a necessary aspect of CCS, such
that moving anthropogenic CO2 to sites or schemes better
suited to its retention or mitigation is now a consideration
in this process. The need for transport in conjunction with
CCS implies that CCST might be the appropriate acronym, in
lieu of CCS. However, searching CCST (as ‘CCST + carbon
capture’) does not quickly lead to transport, but rather points
to terms like CCS technology or training, possibly implying
that transport (aside from economic considerations) does
not pose concerns in parallel to the other elements of CCS.
*Author’s contact details
tel: +1 614 424 4421
email: [email protected]
1. With CCS defined in reference to managing GHG implies that re-injecting
naturally sourced CO2 in support of EOR is not CCS – but simply a commercial
operation that returns the CO2 below ground. In the same vein, EOR supported
by CO2 from GHG is CCS, as it helps manage GHG emissions.
Are you up to speed?
0 0 0 2 0 1 0
Training courses – 2011
TRAINING
FEB 2011
January 25–28
Subsea Production Systems Engineering (Aberdeen)
January 31–
February 4
Subsea Pipeline Engineering Course (Amsterdam)
January 31–
February 4
In-line Inspection of Pipelines (Amsterdam)
February 1–
February 3
Defect Assessment in Pipelines (Amsterdam)
February 14–15
Defect Assessment in Pipelines (Houston)
February 14–15
DOT Pipeline Safety Regulations – Overview and Guidelines for Compliance
(Houston)
February 14–15
Pigging & In–line Inspection (Houston)
February 14–15
Pipeline Repair Methods / In–Service Welding (Houston)
February 14–15
Introduction to Excavation Inspection & Applied NDE for Pipeline Integrity
Assessment (Houston)
February 14–15
Performing Pipeline Rehabilitation (Houston)
February 14–15
Stress Corrosion Cracking in Pipelines (Houston)
February 14–15
Advanced Pipeline Risk Management (Houston)
March 17–18
Microbiological Corrosion in Pipelines (Houston)
March 21–25
Onshore Pipeline Engineering (Houston)
March 30–31
Unpiggable pipeline solutions forum (Houston)
April 25–29
Deepwater Riser Engineering Course (Houston)
April 25–29
Subsea Pipeline Engineering Course (Houston)
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
JAN 2011
2011
MAR 2011
APR 2011
Working with a faculty of some 38 leading industry experts, Clarion and Tiratsoo Technical are
privileged to provide some of the best available industry based technical training courses for
those working in the oil and gas pipeline industry, both onshore and offshore.
Complete syllabus and registration details for each course are available at:
www.clarion.org
4th Quarter, 2010
237
Examination of some major governmental reports tends to
underscore the view that the transport of anthropogenic CO2
regardless of its end use, is straightforward. For example,
in 2005 the IPCC [1] stated that “many analysts consider
CO2 pipeline technology to be mature.” Likewise, a US
Congressional report in 2007[2] stated “pipeline transport
of CO2 operates as a mature market technology,” a view that
remained unchanged when the report was updated a year
later to address jurisdictional issues. A quick check of the
facts indicates that both statements are accurate in reference
to transporting naturally occurring (relatively pure) CO2, in
volumes needed to support enhanced oil recovery (EOR), as
dome-sourced CO2 has been transported for decades. This
opens to the question: why would this or other papers be
written concerning CO2 transport?
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
This question is simply answered in terms of three important
safety and risk discriminators that emerge if pipelines
transporting anthropogenic CO2 service are contrasted to
those in service moving relatively pure CO2 in support of
EOR. These discriminators include:
• the retrofitting of some early CO2 pipelines to provide
for fracture control;
• the role and significance of trace impurities; and
• the volume transported that drives the use of largerdiameter pipelines, which tend to run longer distances
with some traversing high-consequence areas.
Fig.1. Photo after a running ductile fracture along a full-scale
test section (circa 1970s).
While not recognized in the IPCC and US Congressional
reporting circa 2005 to 2007, these discriminators are
significant, as elaborated later in this paper.
After demonstrating the significance of the three safety and
risk discriminators, this paper presents the technological
background to ensure pipeline fracture control, considering
important aspects like the equation of state (EoS) and the
assumptions embedded in the analyses of fracture-arrest
requirements to offset fracture concerns. The role of
impurities is illustrated in regard to arrest requirements,
and select design scenarios are considered as the basis for
evaluating the practical implications of both fracture control
and hydraulics in the light of various platforms available to
quantify required arrest toughness. This discussion considers
a range of fluid properties determined relative to past and
proposed anthropogenic service conditions, along with are
a range of potential pipeline designs. Finally, by reference
to the results of such analysis, the historic design space for
many CO2 pipelines is contrasted to the design space for
pipelines in anthropogenic service. The paper closes with
some important conclusions regarding safety and risk.
What is running fracture, when is
it an issue, and why?
While consideration of and concern for running fracture
are second nature to specialists in this technology, many in
Fig.2. Retrofit fracture arrestors (courtesy of Clock Spring): (a top) successful fracture arrest on a CO2 pipeline; (b - bottom)
installation of one retrofit arrestor scheme.
238
The Journal of Pipeline Engineering
the design community seem unaware of the details and/or
the safety implications, while others seem unaware that the
decompression of a supercritical CO2 pipeline can involve
multi-phase response that cannot be represented by the
broadly available analysis for single-phase gas behaviour.
Such scenarios are known first-hand to the authors within
the last two years in the context of CO2 pipeline design, so
such considerations are neither hypothetical nor are they
relegated to the distant past.
In addition to considerations such as loss of service and/or
the cost to replace potentially significant lengths of pipeline,
the consequences of running fracture for some transported
products and pipeline locations require the certainty that, if
initiated, such fractures would be quickly arrested. Considering
the cost of lost service, product make-up for uninterrupted
delivery contracts, and pipeline replacement, the lost value
can easily amount to millions of dollars per mile through
which the fracture propagates, depending on the location,
product, and length of the propagation. However, far greater
losses can accrue due to litigation consequent to such a
failure. It follows that there is a need for rational schemes
to assess and quantify fracture propagation and arrest in any
scenario where a running ductile fracture is plausible. Models
to quantify fracture propagation and arrest will be presented
following consideration of the key differences between pipelines
supporting EOR versus those involved in CCS service.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Fracture propagation occurs following the unlikely event of
fracture initiation that leads to a rupture, because the stored
energy in certain transported fluids, such as supercritical
CO2, is sufficient in a high-pressure transmission pipeline to
sustain the unstable axial extension of that rupture. Where
the decompression front that develops as expansion waves
propagate back into the pipeline runs axially at a speed
greater than the fracture speed, the pipeline depressurizes
faster than the fracture propagates, leading to fracture
arrest when the pressure decays to a level that no longer
will support axial extension. In contrast, so long as the
crack-tip sustains pressure above that level, unstable axial
propagation continues. Figure 1 (from Battelle’s archives)
shows the outcome of propagating fracture following a fullscale test to quantify this behaviour, which is illustrative of
ductile propagation that has caused failure on pipelines
while in revenue-service.
of the propagating crack. This opening is evident in Fig.1,
where the upper quadrants of the pipe have opened. The size
of the flaps reflects the energy associated with the passage
of the crack and the extent of the energy stored prior to
its passing. Flap formation and the extent of the opening
develop longitudinal and circumferential stresses ahead of
the crack. These stresses cause thinning of the wall thickness
and induce significant ovality in the pipe’s cross-section.
Figure 1 shows the effects of the longitudinal yielding ahead
of propagation and adjacent to the crack, which is evident
in the wavy response apparent on either side of the crack
over the upper half of the pipeline. Some fracture-arrestor
concepts rely on constraint of flap formation, on the
presumption that the stresses due to the flap inertia and the
related dynamics contribute to the crack’s advance. Other
running ductile fracture-arrestor concepts act to reduce
the wall stress, or dissipate the fracture energy, or provide
enhanced fracture resistance. In addition, arrestors can act
to ‘ring-off’ the cracking, although violent ring-off should be
avoided due to the chance of fracture re-initiation.
Factors that can contribute to arrest include:
• dissipation of energy through the inherent toughness
of the linepipe steel and its strain-hardening
characteristics;
• reduction in local stress, due to increased wall thickness
at a road crossing or comparable condition; and
• the effects of external factors that act to restrain
crack advance or lower the net load carried by the
pipecwall, such as due to a fracture arrestor.
Early in-service failures involving propagating or running
fractures occurred via ‘brittle’ fracture, which showed
limited dissipation in terms of either deformation or
crack-tip response. As changes to the pipe steels, coupled
with appropriate steel specifications, managed concern for
running brittle fracture, it became apparent that a running
‘ductile’ fracture was also possible. In contrast to the brittle
scenario, a running ductile fracture exhibits significant plastic
deformation and involves cracking mechanisms associated
with locally ductile stretching, leading to void nucleation,
growth, and coalescence.
As fractures propagate axially along a pipe, the fracture tends
to open in the wake of the crack, creating what have been
termed ‘flaps’. For a running brittle fracture, the extent of
flap-opening can be very limited, with some distribution
pipe materials like polyethylene remaining tight in the
wake of the crack due to residual stresses induced in pipe
manufacture. In contrast, with a running ductile fracture
the flaps open and can fully flatten the pipe in the wake
Early pipelines supporting EOR in
contrast to CCS applications
This section elaborates on:
• the need to retrofit some of the early CO2 pipelines
to provide for fracture control;
• the role and significance of impurities (including
trace levels in some cases); and
• the volume transported that motivates use of
larger-diameter pipelines running longer distances
and traversing high-consequence areas to identify
differences in the safety and risk aspects for domesourced pipelines that support EOR versus those
designed for CCS applications.
2. Some other practical concerns are provided for in design, such as hydraulics,
compression/pumping, repair, start-up/shutdown, in addition to fracture control.
Some of these impact efficiency and cost-drivers, while others impact safety and risk,
with the focus here being fracture control because of its safety and risk implications.
4th Quarter, 2010
239
The focus here is aspects in which the design basis for such lines
supporting EOR will underestimate what must be specified
to ensure safe design in CCS applications, underscoring the
fallacy of CO2 pipeline design as a mature technology.
The need to retrofit fracture arrestors on CO2
pipelines
Role of impurities
The second key discriminator is the presence of impurities
and their impact on the fluid’s properties, which becomes
clear in contrasting results for near-pure naturally occurring
‘dome’ CO2 to that for some EOR service, and the flue gas
anthropogenic CO2 mix emitted in fossil power generation4,5.
While knowledge of and concern for this aspect also traces
to the 1980s [17], it too became topical circa 2005 [14], and
since has been emphasized in the broadly available literature
[18-20], particularly in regard to CO2 compression and
pipeline transport.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
The view that pipeline transport of CO2 is a mature
technology can be considered flawed unless there is a
mature basis to quantify fracture-arrest requirements for
such applications2. This view reflects the observation that
some early-design CO2 pipelines supporting EOR had to be
retrofitted to manage concern for running fracture3 [3, 4],
a need that was realized after the first of the CO2 pipelines
went into service. Likewise, it reflects the observation that
while regulations covering CO2 pipeline design require
that fracture arrest be considered since 1988 [7], it is only
in 2010 that a recommended practice to address this aspect
became available [8]. Finally, this view follows from the
observation that the guidance available in Reference 8 is
performance-based rather than prescriptive, with the tools
needed to implement that guidance are still in development
or in are use only by specialists, and generally lack full-scale
proof of their utility/applicability, except for pure CO2.
Accordingly, the design basis for CO2 pipelines is well
short of mature, and use of the design basis of pipelines
built to move nearly pure naturally occurring CO2 can be
misleading when employed where the transported product
is typical anthropogenic CO2.
Concern for running fracture has been recognized for
decades for any pipeline transporting compressed gases or
supercritical fluids, with the latter being the common state
used for CO2 transport due to related efficiencies. Yet the
threat it poses is not evident in what was termed mature
technology. The need to address fracture control also was
emphasized in regard to CO2 pipelines in textbooks on
pipeline design [12], which were published well before the
IPCC and US Congressional reporting. This need has
been emphasized by some regulators [13], and included in
some pipeline regulations for decades [7], with the need
for fracture control becoming topical again, circa 2005
[14, 15] into 2007 [16]. As becomes clear later, confidence
in future design cannot be gained from such testing, nor
can the design process for the pressure boundary of such
pipelines – however mature – be taken as the benchmark
for other pipelines, unless all relevant design parameters
are comparable.
Several resources that trace to the 1980s [9-11] provide
perspective for the running-fracture concerns in CO2
pipelines designed then to support EOR. While such
resources underlay this retrofitting, such is not evident in the
post-2000 IPCC and US Congressional reporting. Reference
9 couples two similar papers that reflect on the unique traits
of supercritical CO2, while Reference 10 outlines the threat
posed by long-running ductile fracture, as does Reference
11, so the concern for complexity well beyond mature
technology was in print and readily accessible for decades.
In regard to retrofitting as discussed in Reference 3, a series
of full-scale fracture-arrest tests was conducted to assess the
utility of wrap-on fracture arrestors for such applications, but
not broadly published. Figure 2a shows a view of a successful
arrest (in a full-scale test) from that work. It is apparent from
this photograph that while the fracture ran up to and below
the arrestor, the pressure was reduced and the pipe restrained
sufficiently to limit its propagation beyond the arrestor. Figure
2b illustrates the installation of such arrestors, although this
is not specific to a CO2 pipeline application. References 9to
11 are notable in regard to the knowledgebase concerning
the threat posed running fracture in CO2 pipelines circa the
1980s, with other resources also available then.
3. Pipelines already in EOR service were retrofitted after the concern for running
fracture became apparent. Others for which pipe had been ordered, or at various
stages beyond that into construction, addressed the concern for running fracture by
applying fracture arrestors during construction [5, 6].
It is apparent from the work reported circa 2005 [14] that
the saturation pressure changes significantly due to the
presence of modest impurity levels, and it can be inferred
that the critical temperature will show a comparable
dependence. While such outcomes are specific to the
EoS that was used (the Peng-Robinson [21] (P-R) EOS was
adopted for that work), nevertheless the concern for the
role of impurities was clear then, which given the impact
of saturation pressure on fracture propagation and arrest
must be addressed in CO2 pipeline design. Recent work
indicates the P-R EoS can significantly underestimate the
saturation pressure [22] in contrast to other EoSs, and
the density dependence of fluid on pressure, depending
on the initial conditions. In turn this means that fracture
control of pipelines in CCS service based on the P-R EoS
could be less conservative than required if referenced to
the inappropriate mix of CO2 and impurities, and their
relative percentages. This becomes an even more acute
issue if their design basis was benchmarked to outcomes
for nearly pure CO2 (i.e., typical EOR cases).
4. Depending on the fossil generation plant (coal versus gas-fired) and the combustion
process (pre versus post-combustion versus oxy-fuel) the impurities present, their
absolute or relative levels, and their effects differ significantly – for example, see
Sass et al. [14].
5. Continuing work indicates that nominal impurity levels reported for a given
process can vary significantly from what is considered nominal due to fuel and
process variations.
240
The Journal of Pipeline Engineering
Fig.3.The BTCM.
Transport volume, distances, and routeing
also be addressed. This particularly true given that some
early-use EOR pipelines required retrofitting.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
The third key discriminator involves the volume of CO2
to be transported, the distance moved, and the locations
traversed. While the above two aspects affect differences in
the threat posed by running fractures, parameters including
transported volume affect pipe size and pressure, which
like routeing and distance, impact public exposure and
the consequences associated with the unlikely occurrence
of fracture initiation. While there is no simple avenue to
determine the eventual volume, it has been stated [1] that
“pipelines can be expected to play a significant role in the
required transportation infrastructure.” This indicates it is
likely that many CO2 pipelines will be built for CCS-related
service, which will develop in addition to the existing network
of more than 30 pipelines supporting EOR6 [23].
Reference 23 designates three types of CO2 pipelines in
EOR service circa 2007, with the discriminator being the
transported product specification. What was termed Type I
covered special, single-use pipelines with case-by-case (rather
open) specifications, with Type II being multiple source /
user lines and ‘strict’ specifications (this type was noted as
typical of most of the North American network), and Type III
represented hybrid lines, with relaxed but ‘controlled’ CO2
composition. These three scenarios differ somewhat from
the compositions moved by pipelines supporting early EOR
– say prior to 1985. Of the six or so CO2 pipelines operating
circa 1985, about half moved nearly pure dome-sourced
CO2, while the others moved CO2 with small amounts of
hydrocarbons whose levels depended on the source field and
its variability. If history is to serve as a benchmark for current
practices, the role of impurities must be considered. If that
same history is to be a benchmark for CCS applications –
where CCS involves anthropogenic sources given its GHG
roots – then differences in impurities and their levels must
6. Of the 30 pipelines reported in EOR service, about one-third transport CO2 whose
source is gas plants or other processing facilities that results in CO2 separated by
man from a process stream that originated naturally – as such it is not anthropogenic
as a consequence of CCS, nor is it naturally occurring in the form it is transported.
The sizes of the two CO2 pipeline segments commissioned
in the US in support of EOR prior to 1980 involved a
diameter of 16in (406mm) or less, with the pipe wall made
of Grade X65 (448MPa) or lower. The sizes and grades
of the CO2 pipelines built in the interval thereafter, but
before CCS transport became a major consideration, had
diameters that ranged up to 30in, which were typically
built in Grade X70 or below. Pipelines supporting EOR
tended to run through quite remote areas, with source sites
selected, in general, as close as possible to the reservoir
being worked to minimize the length of the pipeline. For
this reason, well over half of these pipelines have a length
less than 100miles (160km), with almost all having lengths
less than twice that distance. In general, the diameter
of such lines is also small in comparison to natural-gas
transmission pipelines, with well over half having a diameter
of 12in (305mm) or less, with almost all having diameters
less than twice that size.
In contrast, to the authors’ knowledge, the diameter of lines
currently now built in the US for CCS service run up to
24in7 (610mm), and are made in Grades up to X80 (551MPa).
Trunklines are under consideration for anthropogenic CO2
that will run from the north to the south of the US, leading
to distances the order of five to ten times that just cited.
Currently constructed pipelines for CCS service, such as
the Green Pipeline, run through high-consequence areas
(HCAs) that pose a public concern given the asphyxiative
properties of gaseous CO2, and its terrain-tracking transport
as a dense (heavier than air) vapour for onshore pipelines,
but are much less an issue in an offshore context [25]. Its
properties as a supercritical fluid lead to unique concerns,
which increase nonlinearly with diameter given the
7. As indicated above there are some quite large-diameter segments supporting EOR,
notably the Cortez pipeline [5].
4th Quarter, 2010
241
illustrates this graphical scheme circa the 1970s. The curve
labelled ‘fracture’ reflects the relationship between fracture
speed and pressure as a function of toughness, which was
plotted by trial and error until the fracture speed for a given
toughness was tangent to the curve labelled ‘gas’ that reflects
speed of decompression in the wake of the expansion waves
propagating back into the pipeline. Arrest is ensured by use
pipe steel that is specified with that or greater toughness.
As becomes is evident later, arrest occurs rapidly once the
stress at the crack-tip drops below its critical value.
Each of the above three considerations points to the need
to address running fracture in regard to CO2 pipelines in
CCS-related service. Likewise, it was evident that differences
between pipelines designed to support EOR versus transport
mixtures of CO2 and the trace impurities common to CCS
service that precludes use of existing pipelines as design
benchmarks. On this basis, the next section considers
approaches to quantify fracture propagation and arrest in
CCS applications. As noted above, such considerations are
not new – they have been with us since the 1970s for natural
gas transmission [26], the 1980s in reference to dome-sourced
near-pure CO2 pipelines [4-6], and more recently in regard
to CO2 pipelines in CCS service [15, 18-20]. Consequently,
what follows is a brief introduction and review, which suffices
as background to assess the viability of the assumptions and
the viability of the model and its predictions.
By the late 1970s analytic approaches [27] and semi-analytical
schemes [28, 29] appeared, as did empiricism in the form
of curve-fits to full-scale test data [30]. The mid-1980s saw
continued attempts to analytically capture this phenomenon
[31-33], including a comprehensive energy-balance
formulation, and a large-scale numerical formulation that
made use of crack-tip opening angle (CTOA) as its measure of
cracking resistance. After significant work to refine the metric
for fracture resistance, the formulation solidified [34], but its
blind application to predict arrest for the Alliance Pipeline
full-scale tests led to values of CTOA the order of 25º for
arrest [35], well beyond the moderate toughness levels actually
required. This gave rise to redefinition of CTOA and some
reformulation of the model, which resulted in a more correct
outcome the order of 11-12º [36]. Subsequently, definition
of CTOA was further updated [37], and then revised again
in the late 1990s along with additional ‘tweaks’ [32], which
contributed to its improved case-specific predictability. While
this work continues, primarily in Italy [34-38] and the UK
[39], and holds much promise, day to day arrest-toughness
predictions still make use of the BTCM, except as modified
to address higher-toughness steels [40, 41]. Subsequent work
has addressed other issues, but relies on the same concepts
[42, 43]. In this context there are three approaches to make
predictions for CO2 applications:
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
functional dependence of running fracture on diameter [26].
As such, the requirements for fracture control of pipelines
transporting nearly pure CO2 in the volumes associated with
EOR are much less demanding than for the typically largerdiameter systems anticipated for service transporting CO2
containing impurities typically associated with CCS. Tables
2 and 3 in Reference 14 identify possible trace impurities by
source type and summarize their potential effects on capture,
compression, pipeline transmission, and injection, and so
are useful assessing related safety implications.
Modelling to quantify fracture
propagation and arrest
Technology to quantify running fracture spans from raw
empiricism into numerical formulations which, while
potentially elegant in concept, still embed empirical
calibration(s). While the rapid evolution of fracturemechanics’ theory and technology development that began in
the 1960s and has continued since provides the foundation,
running ductile fracture is a complex phenomenon. Running
ductile fracture couples fluid and solid mechanics with
fracture mechanics, gas dynamics, and thermodynamics.
The coupled nonlinearities of these disciplines involve soilstructure interaction for buried onshore pipelines, and its
parallel offshore, where the fact that pipelines can operate
at significant depths adds further complexity.
Work through the mid 1970s led to what has been referred
to as the Battelle Two-Curve Model [25] (BTCM), which
reflects the significant efforts of Maxey [26]. This formulation
capitalized on the basic fracture concepts, and coupled that
with gas dynamics, and thermodynamics to characterize
driving force and the inherent fracture resistance. The BTCM
is referred to as such because it quantifies the driving force
reflecting the gas’ decompression speed versus pressureinduced wall stress response and the fracture speed versus
pressure (stress) response through use of two curves, whose
iterative solution for computational reasons in the 1970s was
done by plotting their speed versus pressure trends. Figure 3
• the BTCM as is [16, 44];
• adaptations of GASDECOM with other elements
of the BTCM largely intact [45, 46]; and
• use of the concepts that underlie the BTCM with a
return to first principles, as needed [22].
As for dense-phase (rich) natural gas, the expansions waves
leading to decompression for CO2 propagate through a one
or two/multi-phase medium. The gas dynamics’ aspects of the
BTCM were formulated under the assumption that the flow is
a homogeneous isentropic process. The model was formulated
using the Benedict-Webb-Rubin EoS as modified by Starling
(BWRS) [47], which was packaged as software that has become
known as GASDECOM. Because this formulation was
developed for typical natural gas producing fields, its scope
included binary interactions for a range of NGLs, as well
as for CO2. Use of this technology in applications to quite
rich gases decades later showed it to be quite robust, well
beyond expectations [41]. The basic GASDECOM model
has seen only modest changes since, which focused on the
solver and related algorithms early in 2000.
242
The Journal of Pipeline Engineering
in a pipeline that suddenly suffers a guillotine break. At
the time of the break, the flow is zero at the exit plane
and a sonic wave propagates upstream at the local speed of
sound. The moving expansion wave converts fluid at rest
into fluid in motion. As the exit plane velocity increases
from zero, another expansion wave propagates into the fluid,
which is now moving at a speed, denoted u, at a slightly
reduced pressure and temperature. In reality, this process
is continuous with expansion waves propagating upstream,
each moving a little slower than the preceding one, with
the exit plane velocity continuously increasing until the
exit plane velocity equals the expansion wave speed and
the flow is considered choked. The exit plane speed can
be described by the following [51]:
(1)
where ρ is the density, a is the local speed of sound, and t is
the time after the rupture. The speed of sound is a property
of the fluid and can be found in the literature for a wide
variety of fluids. For pure CO2, the speed of sound is available
from a number of studies, with the results based upon the
Span and Wagner [49] EoS, which is widely considered as
highly accurate.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
While adaptation of the scheme developed for methanedominated mixes was found necessary in Battelle’s 1980s
work involving nearly pure CO2 [10, 48], others have used
GASDECOM for gases dominated by CO2 at least in their
initial work [16, 44]. As Battelle continued to experience
issues in adapting the source-code for GASDECOM to the
CO2-impurity mixes typical of CCS, we have since adopted
alternative EoSs, depending on the application, and returned
to a first-principles approach. The Span and Wagner EoS
[49], which was developed and calibrated specifically for pure
CO2 is used for this benchmark scenario. Both the P-R EoS
[21] and the GERG EoS [50] were initially considered when
impurities effects were involved, however, an adaptation
of the GERG EoS is now preferred, with care still taken to
assess the practical viability of the outcomes. The speed of
decompression is then inferred from gas dynamics referenced
to the speed of sound as a function of composition and
density in the manner due to Liepmann and Rosko [51]. The
fracture velocity is determined as a function of pressure and
toughness using variations of historically proven practices,
with the required arrest toughness being that needed to
slow the speed of propagation to equal or less than that
for decompression.
Key assumptions and their
implications
Fundamental to predicting fracture arrest is the assumption
that arrest occurs when the wall stress falls below some critical
level due to decompression, because the rate of propagation
has slowed such that the pressure-induced wall stress
decreases. The historic view with the BTCM is that arrest
ensues within a diameter or so once toughness suffices to
cause tangency between the gas and fracture curves in Fig.3.
A second key assumption is that the decompression response
can be calculated independent of the fracture response,
with these being brought together through an empirically
determined function that couples their response, that was
termed the backfill coefficient. Yet another key assumption
is that the flow during decompression is a one-dimensional,
homogeneous, isentropic process that develops in response
to expansion waves that propagate back into the pipeline,
the implication of which is that it is an isentropic, reversible
process, with relative motion of liquid and vapour neglected.
Finally, it is assumed that the speed of the expansion waves
can be determined relative to the instantaneous local density
of the fluid. These assumptions are considered next, more
or less in the reverse order they have been cited.
Speed of expansion waves and flow response
Expansion waves of single-component, single-phase fluids
have a long history of study, are generally considered well
understood and thus provide a solid basis for understanding
the more complex case of expansion-wave propagation in
CO2 mixtures with impurities and phase change. Such
analysis is formulated in regard to a fluid at high pressure
In single-phase fluids, the process is considered isentropic
and the density path can be found by tracing an isentropic
path on a thermodynamic state diagram from the initial
pressure to a final pressure. Under this assumption, and using
the necessary data, Equn 1 can be readily integrated and
the exit velocity obtained. The focus here is the expansion
wave, which propagates at the local speed of sound in a
frame moving with the exit velocity. In a frame at rest, the
propagation velocity is then a + u (where the direction is
positive in the upstream direction, so u < 0, a > 0). With this,
the propagation velocity can be tracked and compared to
the crack propagation velocity in the manner of the BTCM.
For a mixture of CO2 and impurities as dictated by the
product stream, the situation is more complex. First, the
properties must be determined by an appropriate EoS. As
noted above, these range from relatively simple expressions,
such as the P-R EoS, to more-complex and realistic expressions
that are grounded in thermodynamics and fit with extensive
data sets. One of the most accurate sets is based upon the
GERG database, as modified by Lemmon [52] of the US
National Institute of Standards and Technology. These
expressions are widely used for hydrocarbon mixtures, but
do not have extensive data for mixtures with high fractions
of CO2 and impurities, indicating that there is a need for
additional data to support CO2 mixture studies.
As the pressure drops in the pipeline, the fluid may undergo a
change in phase. Figure 4a illustrates an isentropic expansion
that starts at 2160psia (149bar) and 90°F (32°C) for a CO2
mixture specific to one client. Such plots develop for pure
243
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
4th Quarter, 2010
Fig.4. Aspects of the thermodynamic
state and expansion waves for CO2
pipelines: (a - top) thermodynamic state;
(b - left) traits of process timeline.
CO2, through mixtures of CO2 and impurities – the key
differences being the locations of the phase boundaries and
associated properties. The y-axis in this figure is pressure
on a logarithmic scale, while the x-axis is density. The
light layered trends are isotherms, while the heavy curved
trend that runs from the upper right down to the lower
left across the layered isotherms is the isentrope from the
initial condition. The region above the critical point at the
top of the two-phase dome comprises supercritical fluid,
while the liquid phase is to the right side of the figure, and
the vapour phase is to the left side, with the region within
the dome being a mix of these two subcritical phases. For
a supercritical CO2 pipeline, there is a spectrum of initial
temperatures and pressures ranging from the inlet to the
exit of the pipeline, so there can be a spectrum of responses
depending on position along the pipeline considered in
regard to the guillotine rupture, and its local pressure and
temperature. For the particular initial conditions considered
in Fig.4a, the CO2 starts as a supercritical fluid, passes into
the liquid region, and then encounters the liquid-vapour
saturation boundary, and remains therein through the
remainder of the process.
244
The Journal of Pipeline Engineering
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Fig.5. Speed of sound of CO2 and CO2
mixtures referenced to three EoSs.
In reference to Fig.4b, the x-axis is density as it is in Fig.4a,
whereas the y-axis is speed in regard to each of the three
curves shown in the figure. Each of these curves represent
a process that begins at 70°F (21°C) and 2200psia (152bar),
which in this figure originates from the upper right corner
of the figure, as the local fluid density corresponds to that
prior to rupture. The upper curve is the speed of sound,
relative to the moving liquid. As time passes, the local density
decreases as the fluid increases in velocity. As evident, the
sonic velocity is strongly nonlinear with density initially,
with this dependence diminishing with time. The sonic
velocity is a property of the liquid and vapour determined
from the equation of state, with multi-phase sonic speeds
determined from the method outlined in Wallis [53]. The
lower curve represents the outflow fluid velocity from
Fig.6. Pressure versus time for a
location 12ins (~30cm) upstream of
the exit plane.
Equn 1, starting from rest (highest density) and increasing
to choked flow. The middle curve is the wave speed in a
stationary reference frame and is simply the vector addition
of the upper and lower curves. The wave speed starts at the
sonic speed (highest density) and then decreases as the fluid
speed increases, finally reaching zero, which is the choked
flow condition.
Each of Figs 4a and 4b are unique to the initial conditions, to
the fluid mix in terms of the fraction of CO2, and to the mix
of impurities and their relative levels. As becomes apparent
next, they are also dependent on the EoS embedded in the
process description, so a compendium of such behaviour
as a function of circumstances along a pipeline, much less
for differing product streams is implausible.
4th Quarter, 2010
245
Fig.7. Effect of impurities on
decompression wave velocity.
flow is a strong but common assumption; among others it
assumes equality of liquid and vapour velocities, which at best
is true only for low bubble volume fractions. For the results in
Fig.5, the speed of sound of the liquid and vapour phases has
been estimated in the context of REFPROP [52], while the
speed of sound in the multi-phase region has been estimated
using an expression from Wallis [53].
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Figure 5 further illustrates trends in the speed of sound for
CO2 as a function of density and fluid composition, to better
quantify the range of responses that can occur depending
on pressure and composition, and also the manner these
are related via the EoS that underlies the analysis. While
speed was the dependent parameter in regard to density in
Fig.4b, in Fig.5 the sonic speed is a function of pressure
that is shown on the x-axis. It is clear from this figure that
for any trend considered, the sonic speed is high when the
transported fluid exists as a liquid, and then drops sharply
as the isentrope crosses the liquid-vapour phase boundary,
reaching a much lower value as a gas-liquid mixture, and
even lower as a gas. This response is central to controlling
running fracture, because this behaviour underlies changes
that lead to decompression local to the rupture plane, such
that arrest can occur, with sufficient toughness supplied
such that the crack propagation speed is equal or less than
the decompression speed. Consequently, understanding the
speed of sound as a function of conditions is important to
the designer, and especially understanding the boundary of
the phase change regions for CO2 mixtures with impurities.
Figure 5 provides insight into the role of the EoS that underlies
the locations of the phase boundaries and how these move
as a function of impurities, which taken together dictate the
breakpoints in the trends shown in this figure. The speed of
sound is shown for three fluids: pure CO2, a 95% CO2 and
impurities mixture, and a 98% CO2 and impurities mixture.
The mixes in both cases represent client estimates of what might
be transported in their pipeline, which might be considered
proprietary and so are not elaborated. Results are presented
for these mixtures characterized by the GERG EoS and the
P-R EoS, both of which were introduced previously along with
the Span and Wagner EoS for pure CO2. These results reflect
the assumption of homogenous one-dimensional flow in the
two-phase region: that is, the properties are a weighted average
based upon the liquid and vapour fractions. Homogeneous
The expansion process for the mixtures in Fig.5 has been
assumed isentropic – which is reasonable for the singlephase part of the expansion but is not generally valid for
the multi-phase part. Essentially then it is the transition
region between the liquid and vapour speed of sound that
is open to question. However, it should be noted that the
phase transition itself may not necessarily occur at the phase
boundary, due to nucleation delays.
Realizing that the underlying formulation can be used
to quantify pressure response within the pipeline, it is
instructive to consider that response and the rate of that
process in contrast to crack speeds approaching arrest.
Consider Fig.6 in this regard, which shows the pressure
response 12ins (30cm) upstream of the exit (rupture) plane
as a function of time. This result is generated for pure CO2,
for isentropic expansion beginning at 2200psia (152bar)
and 70°F (21°C). It is apparent from this figure that there
is a strong drop in pressure until the liquid-vapour phase
boundary is reached, and that the response time is the
order of milliseconds. While this outcome reflects the same
assumptions as noted above, it is apparent that when such
events occur they do so at very high speeds. Because the
timeframe is short the occurrence of subtleties arising due
to three-dimensional versus one-dimensional response are
indicated to have only modest influence on where along
the pipeline – relative to its diameter – the decompression
velocity is found to match the propagation speed, which in
the BTCM framework defines arrest. As such, the minimum
246
The Journal of Pipeline Engineering
Fig.8. Isentropic processes plausible in
design vary significantly.
practical applications involving rich (dense-phase) gases
[41]. Furthermore, the lack of equality between liquid and
vapour velocities generally means that these flows are not
isentropic, since the relative phase motion results in interphase drag, heat transfer, and mixing. In addition, viscous
and inertial effects will become important as the flow speed
increases, which are not included here.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
arrest toughness that results from this analysis is a reasonable
indicator of the toughness needed for fracture control.
Figure 7 develops in the context of analyses for the conditions
that underlie Fig.5, presenting a plot of pressure shown
on the y-axis and the decompression wave speed on the
x-axis in regard to the pure CO2 and the 95% CO2 plus
impurities cases. Comparing these trends shows a practically
quite significant shift in response due to the presence of
impurities. As can be seen, this shift for the impurities mix
is associated with a higher pressure – which occurs due to
the increased saturation pressure, and the shifting phase
boundary in the pressure-density space. At pressures above
the phase boundary, the decompression speed is slower for
CO2 with impurities compared to pure CO2 at the same
pressure. At pressures below the phase boundary, these trends
slow as expansion occurs within the two-phase boundary
and the temperature drops. Because expansions waves form
continuously and there is feedback in the pressure-volumetime space consistent with the EoS, the decompression
velocity gradually decreases, with that decrease generally
stronger for the mixtures than for pure CO2.
Some results indicate that a plateau develops once the
phase boundary is crossed regardless of the EoS used,
which is evident (for example) in References 45 and 46.
Clearly this is in conflict with the view shown in Fig.7, but
as yet the reason for such has not been identified. In this
regard the need for data to support the reliable prediction
of phase boundaries, equations of state, and the transport
properties for these CO2 mixtures, is emphasized, as such
data are generally lacking. In addition, the multi-phase flow
regime is uncertain. The homogenous assumption has not
been verified for these fluids; essentially it assumes that
the liquid and vapour phases move at the same velocity
and these velocities will depart significantly as the bubble
volume increases. However, we note that the homogenous
flow assumption has been successfully applied in many
Clearly a compendium of pressure-volume outcomes would
be useful from a design perspective, but is virtually impossible
to generate because of the range of product compositions and
initial conditions. That view follows from our experience in
working with the product streams anticipated to be carried
under conditions akin to common-carrier / commodity-type
service, which indicates the range of possible mixes could
be very large. It also follows from the observation that the
results of analyses as those in Figs 5 and 7 are highly sensitive
to both the impurities present and to subtle differences in
this mix or the relative percentage of a given constituent.
As such, the focus here is on cause-effect aspects, and the
assumptions involved, and their implications, in lieu of
attempts to trend such outcomes.
The last point to make in this context involves multi-phase
effects, which are as important to hydraulics and other design
aspects (which are not considered herein) as they are to running
fractures. Suffice it to note that multi-phase wave propagation
has been extensively studied for supercritical water-blowdown
events related to a loss of coolant accident scenario in the
nuclear research community. While CO2 mixture flows are
more complex, the existing related knowledge could provide
a strong base for understanding such flow response.
Discussion and design implications
Prior sections have identified the dependence of the
outcomes on the initial conditions used in the analysis
(pressure and temperature), the scope of the impurities, and
4th Quarter, 2010
247
the assumptions made. The need for data to assess/establish
the viability of the predictive schemes also exists, as illustrated,
for example, by the apparent disparity in a plateau forming
beyond the phase boundary in other work [45, 46] versus
the steady decay in the pressure-velocity response evident in
Fig.7. These are considered next in regard to decompression
and arrest-toughness prediction, and then illustrated in
regard to historic CO2 pipeline designs in contrast to the
spectrum of potential EOR-CCS applications.
Viability of decompression and arrest toughness
predictions
The design space for historic EOR versus plausible
CCS applications
Perhaps the most instructive contrast between the outcome
of the design approach that underlies the early CO2 pipelines
and has continued since – and is still considered by some to
be mature technology [1, 2] – is to compare the required arrest
toughness for a range of designs for pipelines transporting
CO2 across a range of pipeline capacities and CO2 quality
specifications. To simplify this comparison, the outcomes
are presented in a normalized format – which avoids making
this comparison specific to details such as inlet pressure
and temperature. Suffice it to say that parameters typical
of supercritical CO2 transport have been considered in
regard to inlet pressure and temperature. To manage the
scope of this comparison and focus the outcomes, a simple
binary CO2 product stream is considered, with the second
constituent being one common to both EOR and CCS
product streams whose effect on arrest toughness is neither
worst-case nor trivial. The CO2 content ranges from pure,
which is well above what is cited as the usual Permian Basin
EOR minimum CO2 content of 95% [54], down to 90%,
which is equally below that minimum. Finally, to avoid the
nonlinear influence of diameter on the outcome of this
analysis, a constant diameter pipeline is considered.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
As identified above, a valid concern exists in regard to the
EoS used, which in turn impacts the saturation pressure
and temperature, and the phase boundaries. It also impacts
aspects such as compression / pumping decisions, and
hydraulics issues that control the recompression distance
for longer CO2 pipeline systems. Finally, such concerns
impact predictions of the decompression behaviour of CO2
and CO2 mixtures, and carry through to the viability of the
concepts that underlie the BTCM and its reliance on the
Charpy V-notch (CVN) energy as the toughness metric.
These are considered next in regard to decompression and
arrest toughness prediction, and then illustrated in regard
to historic CO2 pipeline designs in contrast to the spectrum
of potential EOR-CCS applications.
diagram it is readily apparent that substantially different
behaviour can develop over the length of a given pipeline.
This is evident, for example, in Fig.8, which illustrates the
thermodynamic state of a 95% CO2 plus impurities mixture
for isentropic processes that start at 40°F up to 160°F (4
to 71°C), all at 2160psia (104bar). Depending upon the
initial temperature of the fluid, the expansion process can
descend on either the liquid or vapour side of the critical
point, or possibly through the critical point. Reality for a
pipeline is that the pressure changes along its length, which
also plays into the need for the designer to consider the full
range of conditions that might develop across the full range
of possible compositions. On this basis there is need to
identify worst-case scenarios for purposes of decompression
assessment and arrest-toughness prediction.
Developing a viable database to address these concerns with
scope adequate to address the range of potential EOR-CCS
applications is a bit like a Christmas list where there is no
Santa Claus. While work is planned or continuing to address
these gaps, the scope is limited, so the outcomes tend to be
marginal in contrast to what might be needed given the scope
of concerns involved. More critical is the observation that
some EoS and/or predictive schemes are comparable under
certain circumstances but quite different for others, which
complicates defining the empirical basis to discriminate
between them. This inconsistent disparity is perhaps the
biggest concern, as empirical understanding in regard to:
• the EoS (saturation pressure and temperature, and
the phase boundaries);
• Multi-phase flow (compression/pumping decisions,
hydraulics issues, recompression distance,
decompression behaviour); and
• the viability of the concepts that underlie the BTCM
for CO2 applications could require a significant
testing matrix.
Design implications for CO2 pipelines
An important consideration to the pipeline designer is
the temperature, pressure, and composition of the CO2
mixture across the range of potential rupture locations.
Given the random nature of threats such as third-party
damage, fracture initiation leading to possible running
fracture could occur most anywhere along the length of some
pipelines. By following the isentropic processes on the state
The reference pipeline used to normalize the outcomes
in this comparison is an EOR-service pipeline that for
the sake of simplicity transports pure CO2. To keep the
normalization benchmark representative, this benchmark
is sized slightly larger than ‘average’ relative to the outcomes
in earlier discussion in the section titled Transport volume,
distances, and routeing. On that basis, the benchmark has
been chosen as a 20-in (508-mm) diameter pipeline made of
Grade X65 (448MPa), whose wall was sized using a design
factor of 0.72 for operation at a given inlet pressure and
temperature8. To address the apparent shift from the earlier
EOR designs through what could emerge if CCS service
8. While this benchmark design has been chosen to represent typical onshore CO2
pipelines, its dimensions and pipe grade are comparable to one of the early-design
EOR pipelines that moved dome-sourced CO2, parts of which were retrofitted with
fracture arrestors in the 1980s.
248
The Journal of Pipeline Engineering
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Fig.9. Design space for CO2 pipelines
in terms of arrest toughness and
capacity.
Fig.10. Distribution of CVN energies
for a major order of large-diameter
linepipe.
dominates CO2 pipelines, the need to increase pipeline
capacity (transported volume) is inferred by considering a
range of linepipe grades (diameter is fixed as noted above)
that permit relatively higher stresses. The resulting wall
thickness for a given grade is determined by design factor,
pressure, and diameter, which are constant for all scenarios.
Based on earlier discussion, the range of grades in addition
to the X65 benchmark includes X70 (482MPa) and X80
(551MPa), which in current use in CCS service.
Figure 9 presents the normalized results of this comparison
in terms of required arrest toughness shown on the y-axis
as a function of pipeline capacity shown on the x-axis. The
normalized outcomes are trended as a function the impurity
level, which as noted above involves the variation of one
constituent in a binary fluid, which has been selected from
among the candidate impurities because it is less demanding
than most from a fracture-control perspective for usual EOR
scenarios, and much less demanding than the constituents
in typical CCS-related streams.
Because the benchmark pipeline has been used to normalize
the outcomes, that result is evident in the lower left corner
of this figure at (1,1), which is emphasized by the solid
circular symbol. Pipelines that require lower arrest toughness
and are relatively smaller than the benchmark exist, as do
others that are somewhat larger and require somewhat
higher toughness, the scope of which is typical of many
pipelines supporting EOR. This group of designs tends to
fall within the dashed box that is roughly centred at the
benchmark pipeline. As can be seen from the x-axis labels,
only a 50% relative increase in capacity has been inferred
for this analysis, which is less than some might anticipate.
The scale along the y-axis runs up to a value of 10, with the
worst-case considered in these analyses regarding capacity
and impurities leading to a required arrest toughness that
4th Quarter, 2010
249
approaches ten-times the toughness required for arrest in
the benchmark pipeline. Archival records at Battelle located
for one of the EOR pipelines that had arrestors installed
back in the 1980s indicates toughness levels (hereafter CVN
full-size equivalent energy at service temperature) in excess
of 60ft-lb (81J) in the sections where arrestors were installed.
While this historic design scenario is comparable to the
benchmark pipeline used to normalize the results shown
in Fig.9, this actual toughness level is slightly larger than
the required arrest toughness for the benchmark, which was
the order of 40ft-lb (54J).
This paper has examined aspects determining arresttoughness requirements to control running fracture in CO2
pipelines, and addressed the validity of the view that the
design of these systems deploys mature technology, as has
been suggested in some widely distributed reports on CCS
as the means to manage GHG.[1, 2] While this might be
inferred from decades of experience in transporting CO2 to
support EOR, this view is open to question when moving
large volumes of anthropogenic CO2 from four perspectives:
• some of the early CO2 pipelines were retrofitted to
provide for fracture control;
• trace impurities in CCS and EOR product streams
can significantly increase the arrest toughness
required for fracture control;
• the volume transported for CCS and its routeing
implies risks that can be much greater than for many
EOR applications; and
• the technology used to quantify fracture control
requirements is neither mature nor is the empirical
database supporting it complete relative to current
expectations.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
It can be seen from Fig.9 that the presence of impurities
drives the required arrest toughness much more so than
pipeline capacity. This is evident in regard to the benchmark
capacity (x-axis value equal to one), where a ten-fold increase
in toughness is needed to deal with impurity levels the order
of 10%, as compared to the modest increase in toughness
due to capacity. Realizing that the minimum purity is often
set at 95%, the decision to establish this level as compared
to a lower level helps to offset concern for running fracture.
While there is a desire to control the minimum CO2 content
(apparently due to miscibility for the benefit of EOR, not
safety), there is a clear need to monitor and/or control the
product stream in regard to this parameter given its strong
impact on required arrest toughness.
Summary and conclusions
Equally, there is a need to better understand sources of
variability in toughness for linepipe, as such variability can
diminish the all-heat average (AHA) toughness and opens the
door to longer propagation, which undermines specification
of toughness as the control for running fracture. The data
in Fig.10 help to illustrate this point, where the results show
the distribution of toughnesses for a recent large order of
transmission linepipe. The toughness coordinates in this
figure run from 50 to 400ft-lb, equally 68 to 542J. Whereas
the AHA for this large pipe order exceeds 200ft-lb (271J),
the lowest toughness measured was 56ft-lb (76J), with a
significant fraction of the heats with toughness less than one
standard deviation below the mean. A tighter population,
or a population skewed to above the mean, would greatly
improve fracture control, and could reduce the AHA – that
might simplify making specifications where high AHA
levels are needed.
Central to fracture control is a fracture control plan,
which should be developed during the FEED (front-end
engineering and design), and thereafter serves as the basis
to specify steel for the pipeline. Such planning offsets the
eventual need to retrofit arrestors – which opens to costs
and maintenance concerns that otherwise can be avoided.
Without such planning, where the inherent toughness is
found to be inadequate after the line is commissioned and
the operational parameters are fixed, requires the use of
retrofit arrestors. Their use and placement is motivated by
risk assessment and other considerations, just as was done
in the 1980s, which is about where this paper began.
While not evaluated directly in regard to theses perspectives,
some have considered running fracture a greater risk for
CO2 pipelines than for hydrocarbon pipelines [55]. This
paper has considered the technological background to
ensure pipeline fracture control, considering important
aspects like the EoS and the assumptions embedded in the
analyses of fracture-arrest requirements to offset fracture
concerns. That technology was used to illustrate the role
of trace impurities after which the select design scenarios
have been considered as the basis for evaluating practical
implications in regard to both fracture control and related
aspects. This discussion has considered a range of fluid
properties for naturally occurring and anthropogenic CO2
sources, and the service conditions anticipated for a range
of pipeline designs. Finally, the historic design space for
many CO2 pipelines supporting EOR has been contrasted
to that for pipelines in CCS service.
Important conclusions that can be drawn from this work
follow:
• the minimum CO2 level can cause significant
swings in the arrest toughness, such that toughness
requirements for pipelines in use for dome-sourced
CO2 underestimate that for CCS service by a factor of
two – possibly more in contrast to relatively pure CO2;
• minor out-of-specification swings below the usual
95% minimum can further increase the required
arrest toughness – with a decrease in the minimum
below that level of 2.5% increasing the required
arrest toughness about two-fold;
Held under the Patronage of His Excellency Dr. Abdul Hussain bin Ali Mirza,
Minister of Oil & Gas Affairs and Chairman of the National Oil & Gas Authority, Kingdom of Bahrain
PLATINUM SPONSOR
15–17 May 2011, Bahrain
GULF CONVENTION CENTRE, BAHRAIN
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
ORGANIZERS
Join leaders in the international pipeline industry as they converge for the Best Practice
in Pipeline Operations and Integrity Management Conference and Exhibition in Bahrain.
CONFERENCE
EXHIBITION
Six technical streams covering a wide range of
subjects will run over the two and a half day event
(and presented by industry leaders).
• Planning, design, construction and materials
• Operations and maintenance
• Asset integrity management
• Inspection and cathodic protection
• Repair and rehabilitation
• Automation and control
• Leak detection
Paper abstracts are now being accepted.
A comprehensive exhibition will be part of the
event, allowing companies from around the world
to showcase their products and services. Contact
us today to book your space.
NETWORKING
Throughout the event there will be ample
opportunities to network with participants to
further your business relationships. Meet with
industry leaders from around the world.
vent.
ndmark e
pen
ons will o
ti
a
tr
is
g
e
R
this la
u attend
o
y
e
r
u
s
e
11 – mak
0
in early 2
www.pipelineconf.com
4th Quarter, 2010
251
• minor differences in the relative proportions of a
trace impurity can have a significant influence on the
required arrest toughness, with subtle differences in
the constituents present also a major driver – highly
volatile constituents can complicate the analysis and
significantly drive fracture arrest requirements;
• on-line monitoring of the transported stream and/
or the injected streams on a trunkline should be
considered to help manage risk;
• the BTCM adapted to CO2 applications has been
validated in regard to near-pure CO2 applications
and cases involving rich (dense-phase) natural gas:
however, like most other aspects involved in fracture
control, like the EoS, it remains unvalidated in
applications to typical CCS product streams; and
• an expanded empirical database is essential to ensure
viable fracture arrest predictions.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Acknowledgements
Useful discussions with Dr Bruce Sass of Battelle’s Energy
Systems and Carbon Management product line are gratefully
acknowledged, as is support from Battelle’s Science and
Technology fund in modelling the EoS and the expansionwave response.
References
by regulating saturation arrest pressures. Oil and Gas Journal,
pp44-46.
11. A.B.Rothwell, 1988. Fracture control in natural gas and CO2
pipelines. In: Microalloyed HSLA Steels. ASM International,
pp95-108.
12. M.Mohitpour et al., 2003. Pipeline design and construction:
a practical approach. 2nd edition, ASME Press, (Note: the 3rd
edition is now available but is not used here, to emphasize the
timeline of this 2nd edition).
13. anon., 2008. Interim guidance on conveying CO2 in pipelines
in connection with carbon capture, storage and sequestration
projects. UK-HSE.
14. B.Sass, B.Monzyk, S.Ricci, A.Gupta, B.Hindin, and N.Gupta,
2005. Impact of SOx and NOx in flue gas on CO2 separation,
compression, and pipeline transmission. Carbon Dioxide
Capture for Storage in Deep Geologic Formations, Vol.2,
D.C.Thomas and S.M.Benson (Eds.), Elsevier.
15. J.Barrie, K.Brown, P.R.Hatcher, and H.U.Schellhase, 2005.
Carbon dioxide pipelines: a preliminary review of design and risks.
Proc. 7th Int. Conf. on Greenhouse Gas Control Technologies.
16. A.Cosham and R.Eiber, 2007. Fracture control in carbon
dioxide pipelines. J. Pipeline Engineering, 6, 3, pp150–158.
17. C.B.Farris, 183. Unusual design factors for supercritical CO2
pipelines. Energy Prog. 3, 3.
18. P.N.Seevam, J.M.Race, M.J.Downie, and P.Hopkins, 2008.
Transporting the next generation of CO2 for carbon, capture and
storage: the impact of impurities on supercritical CO2 pipelines.
Proc.7th Int. Pipeline Conf., IPC2008-64063, Calgary, October.
19. A.Cosham and R.J.Eiber, 2008. Fracture control in carbon
dioxide pipelines: the effect of impurities. Idem, IPC2008-64346.
20. H.Li and J.Yan, 2009. Impacts of equations of state (EOS) and
impurities on the volume calculations of CO2 mixtures in the
applications of CO2 capture and storage (CCS) processes. J.
Applied Energy, Elsevier.
21. D.-Y.Peng and D.B.Robinson, 1976. A new two-constant
equation of state. Ind. Eng. Chem., Fundam., 15, 59–64.
22. B.N.Leis, J.H.Saunders, and E.B.Clark, 2010. Issues in
quantifying the expansion wave response in CO2 pipelines.
9th Annual Conf. on Carbon Capture and Sequestration,
DOE/NETL, Pittsburgh.
23. L.S.Meltzer, 2007. CO2 transport – building on the current
framework to meet demands of widely deployed commercial
scale CCS systems. 6th Annual Conf. on Carbon Capture
and Sequestration, DOE/NETL, Pittsburgh.
24. www.statoil.com/en/ouroperations/explorationprod/ncs/
snoehvit/pages/default.aspx.
25. B.N.Leis and R.J.Eiber, 2010. Fracture control technology for
transmission pipelines. PRCI Catalog L51846, 2010: updates
Eiber, R. J., Bubenik, T. A. and Maxey, W. A., Fracture control
technology for natural gas pipelines, NG-18 Report No. 208,
Pipeline Research Council International, Project PR-3-9113,
Battelle, 1993.
26. W.A.Maxey, 1974. Fracture, initiation, propagation, and arrest. 5th
Symposium on Line Pipe Research, American Gas Association.
27. H.C.van Elst, 1974. Criteria for steady state crack extension in
gas pipelines. Int. Conf. on Prospects of Fracture Mechanics,
Netherlands, pp299–318, June 24-28.
28. P.A.McGuire, S.G.Sampath, C.Popelar, and M.F.Kanninen,
1978. A theoretical model for crack propagation and arrest in
1. Carbon dioxide capture and storage, IPCC Special Report,
2005. B.Metz, O.Davidson, H.de Coninck, M.Loos, and
L.Meyer, (Editors).
2. P.W.Parformak and P.Folger, 2007. Carbon dioxide (CO2)
pipelines for carbon sequestration: emerging policy issues. CRS
Report for Congress, 2007: Updated in 2008 as A.Vann and
P.W.Parfomak, ‘CRS report for Congress: regulation of carbon
dioxide (CO2) sequestration pipelines: jurisdictional issues,’
which notes pipeline safety in a US jurisdictional perspective,
but does not otherwise consider the topic.
3. D.I.Marsili and G.R.Stevick, 1990. Reducing the risk of ductile
fracture on the Canyon Reef Carriers CO2 pipeline. SPE 65th
Annual Technical Conference, New Orleans, LA, Proceedings
311-20, September.
4. J.Watts, 1983. Sheep Mountain CO2 pipeline to boost West
Texas production. Pipeline & Gas J. Vol/Issue 210: 6, May 1.
5. W.R.Quarles, 1983. Willbros nears completion on Cortez
CO2 trunkline. Pipe Line Industry, Aug.
6. C.Horner, 1985. Choctaw carbon dioxide line laid in Mississippi.
Pipeline & Underground Utilities Construction, 40, 4, pp4-6, April.
7. anon. §195.111 Transportation of hazardous liquids by pipeline.
US CFR Part 195.
8. anon., 2010. Design and operation of CO2 pipelines.
Recommended Practice, DNV-RP-J202, April.
9. G.G.King, 1981. Design of carbon dioxide pipelines. EnergySources Technology Conference and Exhibition, Houston,
January: see also G.G.King, Design considerations for carbon
dioxide pipelines. Pipe Line Industry, 1981, pp125–132.
10. W.A.Maxey, 1986. Long shear fractures in CO2 lines controlled
252
The Journal of Pipeline Engineering
Eng. Frac. Mech., 71, pp1997-2013, 2004.
40. B.N.Leis, R.J.Eiber, L.E.Carlson, and A.Gilroy-Scott, 1998.
Relationship between apparent Charpy V-Notch toughness and
the corresponding dynamic crack-propagation resistance. Int.
Pipeline Conf., ASME, Calgary, pp723-732: see also Leis, B.
N., Relationship between apparent Charpy V-Notch toughness
and the corresponding dynamic crack-propagation resistance,
1997, Battelle Report to R J. Eiber, Consultant, Inc. Exhibit
B-82, Proceeding GH 3-97, National Energy Board of Canada,
1997-1998.
41.R.J.Eiber, B.N.Leis, L.E.Carlson, N.Horner, and A.GilroyScott, 1999. Full-scale tests confirm pipe toughness. Oil &
Gas Journal, Nov.8, pp48-54.
42. D.Rudland, D.-J.Shim, H.Xu, D.Rider, P.Mincer, D.Shoemaker,
and G.Wilkowski, 2007. First major improvements to the two
curve ductile fracture model. DOT-PHMSA No. DTRS5603-T-0007, May.
43. B.N.Leis, and T.P.Forte, 2007. New approach to assess running
fracture in transmission pipelines. DOT/PHMSA DTRS5605-T-0003, February.
44. anon., 2010. Material requirements for CO2 line pipe. JFE
Steel Corporation / Marubeni-Itochu Steel Inc Presentation,
Battelle, February.
45. A.Cosham, 2009. CO2: it's a gas, Jim, but not as we know
it. 5th Pipeline Technology Conference, Ostend, Belgium,
October.
46. A.Cosham, R.J.Eiber, and E.B.Clark, 2010. GASDECOM:
carbon dioxide and other components. 8th Int. Pipeline Conf.,
IPC2010-31572, October.
47. K.E.Starling, 1973. Fluid thermodynamic properties for
light petroleum systems. Gulf Publishing Co., Houston: see
also Hopke, S. W. and Lin, C. J., Applications of the BWRS
equation to natural gas systems., paper presented at the 75th
National AIChE Meeting, Denver, March 1974.
48. W.A.Maxey, 1983. Gas expansion studies. AGA NG-18 Report
133.
49. R.Span, and W.Wagner, 1996. A new equation of state for
carbon dioxide covering the fluid region from the triple point
temperature to 1100 K at pressures up to 800 MPa. J. Phys.
Chem. Ref. Data, 25, 6.
50. O.Kunz, R.Klimeck, W.Wagner, and M.Jaeschke, 2007. The
GERG-2004 wide-range equation of state for natural gases and
other mixtures. GERG Technical Monograph 15. Fortschr.-Ber.
VDI, VDI-Verlag, Düsseldorf.
51. H.W.Liepmann and A.Roshko, 1957. Elements of gas dynamics.
John Wiley and Sons.
52. E.W.Lemmon, M.L.Huber, and M.O.McLinden, 2010.
NIST Standard Reference Database 23: Reference Fluid
Thermodynamic and Transport Properties-REFPROP,”
Version 9.0, National Institute of Standards and Technology,
Standard Reference Data Program, Gaithersburg.
53. G.B.Wallis, 1969. One-dimensional two-phase flow. McGrawHill.
54. anon., 2009. Table 2 in Guidelines and regulations for oxy-fuel
carbon dioxide capture, transport and storage. IEA Oxy-Fuel
Working Group.
55.J.Barrie, K.Brown, P.R.Hatcher, and H.U.Schellhase, 2004.
Carbon dioxide pipelines: a preliminary review of design and
risks. 7th Int. Conf. in GHG Control Tech., Vancouver.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
pressurized pipelines. American Gas Association, Catalogue
No. L00033, November.
29. W.A.Poynton, 1974. A theoretical analysis of shear fracture
propagation in backfilled gas pipelines. Crack Propagation
in Pipelines, Paper No. 14, published by the Institute of Gas
Engineers, Newcastle upon Tyne, UK, March.
30.G.D.Fearnehough and D.G.Jones, 1980. Toughness
specification for shear fracture arrest in pipelines. Int. Conf.
On Analytical and Experimental Fracture Mechanics, Rome,
June 23-27. See also Bonomo, F. et al., Survey and tentative
revision of ductile fracture arrest criteria in pipelines for gas
transmission, Ibid. See also Vogt, G. H., et al., Toughness for
crack arrest in gas pipelines. EPRG Report, 3R International,
22, 1983, pp 98-105, and others.
31. F.Abbassian, 1985. Long-running ductile fracture of high
pressure gas pipelines. Dissertation for PhD, University of
Cambridge, November.
32. L.B.Fruend and D.M.Parks, 1980. Analytical interpretation
of running ductile fracture experiments in gas-pressurized
linepipe. Crack Arrest Methodology and Applications, ASTM
STP 711, G. T. Hahn and M. F. Kanninen, Eds., pp359-378.
33. G.Buzzichelli, F.Nicolazzo, G.Demofonti, G.Re, S.Venzi,
M.F.Kanninen, J.W.Cardinal, E.Z.Polch, T.B.Morrow,
S.T.Green, and C.H.Popelar, 1987. Second annual report on
the development of a ductile pipe fracture model. Southwest
Research Institute report to PRCI, AGA/PRC Contract Nos.
PR 182-527 and PR 15-527, May 1987. See also Kanninen, M.
F., O’Donoghue, P. E., Cardinal, J. W., Leung, C. P., Morrow,
T. B., Green, S. T., Popelar, C. F., Buzzichelli, G., Demofonti,
G., Rizzi, L., and Venzi, S., Dynamic fracture mechanics
analysis and experimentation for the arrest of ductile fracture
propagation in gas transmission pipelines, Pipeline Technology
Conference, Belgium, October.
34. G.Demofonti and I.Hadley, 1992. Review of fracture parameters for
laboratory measurement of resistance to ductile crack propagation
in line pipe steels. CANMET Pipeline Conference, Calgary.
35. S WRI, 1997. Written private communication to Von
Rosenburg, E. L., March.
36. S WRI, 1997. Written private communication to Von
Rosenburg, E. L., May.
37. G.Demofonti, S.Venzi, and M.Kanninen, 1995. Step by step
procedure for the two specimen CTOA Test. 9th EPRG/PRCI
Symposium, pp18-1 through 18-10.
38. G.Berardo, P.Salvini, G.Mannucci, and G.Demofonti, 2000.
On longitudinal propagation of a ductile fracture in a buried
gas pipeline: numerical and experimental analysis. Proc. 2000
Int. Pipeline Conf., 1, New York, ASME, pp287 –294. See
also Salvini P., Fonzo A., and Mannucci G., Identification of
CTOA and fracture process parameters by drop weight test
and finite element simulation, Engineering Fracture Mechanics,
70, 3-4, 553-566, 2003. See also Mannucci, G., Buzzichelli, G.,
Salvini, P., Eiber, R. and Carlson, L., Ductile fracture arrest
assessment in a gas transmission pipeline using CTOA. 3rd
Int. Pipeline Conf., Calgary, Alberta, Canada, 1, pp315-320.
39. S.H.Hashimi, et al., 2004. A specimen for studying the resistance
to ductile crack propagation in pipes. 5th Int. Pipeline Conf.,
IPC04-0610, Calgary, 2004: see also Shterenlikht, A., Hashemi, S.
H., Howard, I. C., Yates, J. R. and Andrews, R. M., A specimen
for studying the resistance to ductile crack propagation in pipes,
4th Quarter, 2010
253
How to select wall thickness,
steel toughness, and operating
pressure for long CO2 pipelines
by Graeme G King*1 and Satish Kumar2
1 Tensor Engineering Ltd, Calgary, Alberta, Canada
2 Masdar Carbon, Abu Dhabi, UAE
M
ASDAR IS PLANNING to capture CO2 from power plants, smelters, steel works, industrial facilities,
and oil and gas processing plants in Abu Dhabi in a phased series of projects. Captured CO2 will be
transported in a new national CO2 pipeline network with a nominal capacity of 20 x 106 t/a to oil reservoirs
where it will be injected for reservoir management and sequestration.
I
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
The design of the wall thickness,pipe toughness,and operating pressure of the network considered fundamental
thermodynamic properties of CO2, code requirements, toughness needed to control long ductile fractures,
and cost optimization to resolve contention between the different technical requirements and arrive at a
safe and economical pipeline design.The work selected a design pressure of 24.5MPa, well above the critical
point for CO2 and much higher than is normally seen in conventional oil and gas pipelines. Despite its high
operating pressure, the proposed network will be one of the safest pipeline systems in the world today.
n June, 2007, Masdar announced the Abu Dhabi
Carbon Capture and Storage (CCS) project. Front-end
engineering design (FEED) for the first phase of the project
started in November, 2008, and has now been completed:
Fig.1 shows the proposed route, which will form part of the
basic infrastructure that will significantly reduce greenhouse
gas emissions in the UAE from 2020 onwards.
Captured CO2 will be transported in a new national CO2
pipeline network to onshore oil reservoirs throughout Abu
Dhabi where it will be injected for reservoir management
and sequestration. Masdar is working closely with the Abu
Dhabi National Oil Co (ADNOC) and the Abu Dhabi Co
for Onshore Oil Operations (ADCO).
The project will have a threefold benefit: it will reduce
greenhouse gas emissions in the UAE, make CO2 available
for enhanced oil recovery (EOR), and free-up natural gas
that is currently being injected to maintain pressure in
some of the fields.
Three different and sometimes conflicting sets of
requirements are needed to select wall thickness and
toughness of CO2 pipelines. The first set of requirements
limits stresses in the pipe wall under steady and transient
*Author’s contact details
tel: +1 403 398 3858
email: [email protected]
operating pressures and temperatures, and is embodied in
the governing pipeline codes and regulations. The first set
of requirements is used to select the minimum allowable
wall thickness of the pipe.
The second set of requirements is based on the need to
prevent longitudinal ductile fractures and has been developed
by the gas pipeline industry using empirical data obtained
from full-scale pipe-burst tests with natural gas. This set of
requirements is used to establish pipe toughness and, if the
required toughness is unachievable, it dictates the minimum
required wall thickness or the use of crack arrestors.
The final set of requirements is related to the optimization of
total project owning and operating costs. Cost optimization
provides a rational methodology for choosing between
the apparently conflicting results of the first two sets of
requirements, and leads to the selection of a pipeline that
is both safe and economic.
Nomenclature
A = area beneath Charpy notch, m2
CS = velocity of sound in CO2, m/s
CV = Charpy notch toughness, J
D = pipe outside diameter, m
E = modulus of elasticity, Pa
EN = normalized toughness parameter, (-)
F = hoop stress design factor, (-)
The Journal of Pipeline Engineering
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
254
Fig.1. Route map for Abu Dhabi CO2 pipeline network.
FJ = weld joint factor (ASME B31.4 Table 402.4.3)
Pd = decompressed pressure at phase boundary, Pa
SMYS = specified minimum yield strength
T1 = temperature at time of installation, °C
T2 = maximum operating temperature, °C
t = pipe wall thickness, m
α = linear coefficient of thermal expansion, 1/°C
∆v = change in velocity of the fluid, m/s
∆P = surge pressure, Pa
ν = Poisson’s ratio for steel, (-)
ρ = density of the fluid, kg/m3
σd = hoop stress due to internal pressure Pd, Pa
σf = pipe steel flow stress, Pa
σh = hoop stress, Pa
σL = longitudinal compressive stress, Pa
Process design
CO2 can be transported most efficiently over long distances
in the dense phase. The Cortez pipeline, for example,
transmits CO2 from Wyoming to Texas and operates at
pressures up to 18MPa (2600psi), and the Weyburn-Souris
pipeline in Montana and Saskatchewan has a maximum
operating pressure of almost 21MPa (3000psi).
These high operating pressures are due in part to the need for
a high minimum pressure to maintain single-phase operation,
and in part to the optimum frictional pressure loss along CO2
pipelines. As a result the optimum operating pressure can be
higher than 21 MPa. For any specific project, the optimum
operating pressure is influenced by the composition of the
CO2 mixture it carries and specific details of cost and economic
models used for the project.
The dense phase was originally defined as a single phase
separating the gas and liquid phases immediately above the
two-phase region, where fluid properties transition between
those of a gas and a liquid without any change of phase, and
where fluid properties can be distinctly different from those
of either a gas or a liquid [1]. Figure 2 shows the dense phase
region on a pressure-enthalpy chart developed using the BWRS
equation of state for a design mixture of 95% pure CO2.
Specific heat is one of the properties of dense-phase fluids
that is different from either a gas or a liquid. Figure 2 shows
specific heat for gas, dense, and liquid phases (single-phase
region): it shows that the specific heat of CO2 in both gas
and liquid phases is less than 2.5kJ/kg-K but in the dense
phase the specific heat is higher than 2.5kJ/kg-K and can
reach values as high as 10kJ/kg-K.
Another unusual property of dense-phase fluids is their
volumetric sensitivity to changes in temperature. For
example, the density of CO2 at 10MPa and 40°C is 400kg/
m3 (see Point 1 on Fig.2) and the density of carbon dioxide
at 15°C and 10MPa is 800kg/m3 (see Point 2 on Fig.2). If the
fluid followed the gas law, this 25°C change in temperature
from 40°C (313K) to 15°C (288K) would cause the density
to change only 8% from 500 to 540kg/m3. But dense-phase
CO2 is an order of magnitude more sensitive to changes
4th Quarter, 2010
255
Location
Dwellings per Mile
General Description
Design Factor
Class 1
<10
Sparsely populated wasteland, wilderness, grazing and
farmland
0.72
Class 2
10 to 46
Intermediate fringe and areas around cities and towns
0.60
Class 3
>46
Suburban residential and industrial areas
0.50
Class 4
>46
Multi-story urban areas with high traffic and buried utilities
0.40
Tier-2
0.5 million – 1 million
Cement factories, CCGT Power stations, fertilizer, petrochemical complexes
Table 1. Basic design factor (F).
in temperature, and in this particular case a temperature
change of only 25°C causes the density to change more than
100% from 400 to 800kg/m3.
parameters subject to all the relevant technical restraints
imposed by the requirements of hydraulic performance,
governing codes, and good engineering practice.
The UAE has long summers with air temperatures rising
to about 48°C between May and September, and short
moderate winters between December and March with air
temperatures rarely dropping below 6°C. The maximum
design temperature of CO2 from aerial coolers after
compression has therefore been set at 55°C. The ground
temperature at pipeline depth fluctuates between 13°C in
late winter and 38°C in late summer.
Allowable stress
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Requirements for CO2 pipelines are included in ASME B31.4
Pipeline transportation systems for liquid hydrocarbons and
other liquids [2], in which paragraph 402.3. uses a design
factor (F) of 0.72 to limit the allowable hoop stress (σh)
under steady flow conditions:
The volumetric sensitivity of dense-phase CO2 to changes in
temperature means that during summer when the ground
temperature at pipeline depth is close to 40°C, the density of
carbon dioxide at the inlet to intermediate booster stations
and down-hole injection facilities would be approximately
400kg/m3, but at low flow during winter when the ground
temperature at pipeline depth is around 15°C, the density
of CO2 at pump suction would be approximately 800kg/
m3. Wide swings in density between summer and winter,
and between high and low flows, can create challenges to
the smooth operation of the pipeline unless the magnitude
of the swings is identified early in the design and solutions
are found and implemented.
If centrifugal equipment is used to pump the dense-phase
fluid, wide swings in fluid density cause proportionally wide
swings in differential pressure between suction and discharge.
Speed control can be used to manage the swings by allowing
the equipment to be run faster during summer when density
is low and slower in winter when density is high. Variablefrequency and variable-speed drives (VFDs and VSDs) are
now widely used throughout the pipeline industry, so that
once variations in density have been properly evaluated it is
not difficult to size and configure variable-speed centrifugal
equipment to handle the full range of operating flows,
pressures, and temperatures throughout the year.
The overall goal of pipeline design work therefore is to
configure the proposed system to handle a nominal flow of
20 × 106t/a of CO2 with a maximum design temperature of
55°C, ground temperatures between of 17°C and 38°C, and a
maximum design pressure of around 21MPa, and to optimize
σh ≤ 0.72 FJ SMYS
(1)
Metallurgical investigations conducted as part of the
optimization work looked at a range of steel strengths and
selected API 5L X65 (L450) linepipe as the most suitable
for the project. Table 402.4.3 of ASME B31.4 sets the joint
factor (FJ) in Equn 1 for pipe made from this steel equal
to unity so that the allowable hoop stress for the pipeline
network under steady flow conditions is 72% SMYS. In
addition, paragraphs 402.3.2(c) and 419.6.4(b) of ASME
B31.4 limit the combined stress of buried pipelines:
σh + σL ≤ 0.90 SMYS (2)
In Equn 2, σL is the net longitudinal compressive stress
due to the combined effects of temperature rise and fluid
pressure computed from:
σL = Eα(T2 – T1) – νσh
(3)
Paragraph 419.6.4(b) of ASME B31.4 makes it clear that
bending stresses only need to be included in combined
stress calculations when designing above-grade portions of
restrained lines and do not need to be considered when
designing buried portions providing, of course, that the
pipeline is built using proven pipeline construction practices,
good engineering, and other more-specific rules included
in Chapter V of ASME B31.4.
If a design factor of 0.72 is used, Equns 2 and 3 allow
a maximum temperature differential of 160°C between
construction and operating conditions for buried pipelines
before extra wall thickness is required to keep the combined
256
The Journal of Pipeline Engineering
Fig.2. Pressure-enthalpy diagram for
95% pure CO2 showing specific heat in
liquid, dense, and gas phases.
density over the normal range of operating conditions is
approximately 800kg/m3.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
stress below 0.90 SMYS. The maximum temperature
differential between construction and operation for the
proposed pipeline is less than 50°C, and the combined
stress limitations of ASME B31.4 therefore do not require
additional wall thickness.
In order to enhance safety, the design factor of 0.72 has been
modified depending on its location by additional restrictions
of ASME B31.8 Gas transmission and distribution piping
systems [3]. Table 1 summarizes the design factors for the
four location classes specified in Table 841.114A of ASME
B31.8. These additional requirements over and above the
requirements of ASME B31.4 increase wall thickness of the
CO2 pipeline network in populated areas.
Surge pressures
Surge pressures are produced by changes in velocity of the
moving fluid that result from shutting-down pumps or
pump stations, closing valves, or otherwise blocking the
flow. Internal viscous effects and inelastic properties of the
backfill and pipe wall attenuate surge pressure waves as they
move away from the point of origin.
Paragraph 402.2.4 of ASME B31.4 requires pipeline designers
to make surge calculations and to provide adequate controls
and protective equipment to prevent pressure rise due to
surges and other variations from normal operations from
exceeding the internal steady state design pressure anywhere
in the pipeline system and equipment by more than 10%.
The maximum surge pressure (∆P) can be precisely calculated
from the fundamental equation:
∆P = ρ CS ∆v (4)
Figure 3 shows the density of dense-phase CO2 over a broad
range of operating pressures and temperatures. Typical fluid
Figure 4 shows the velocity of sound in CO2, the typical value
of which over the normal range of operating conditions is
approximately 500m/s. Both density and velocity of sound
in CO2 are slightly lower than for oil, so that the magnitude
of surge pressure waves in CO2 pipelines can be expected
to be less than in oil pipelines.
Figure 5 shows the unit surge pressure defined as the change
in pressure for a change in flow velocity of 1m/s. The
typical value for unit surge pressure over the normal range
of operating conditions is approximately 0.4MPa/(m/s).
Surge pressure is proportional to change in flow velocity,
and maximum surge pressure therefore depends on the
maximum flow velocity. The optimum flow for CO2 systems
is less than 4m/s and therefore, for cost efficiency, it is best
to design and build CO2 systems so that they operate at
velocities less than 4m/s. The maximum surge pressure in
a properly designed CO2 pipeline, if a valve suddenly closes
or a pump station suddenly stops working, is therefore less
than 1.6MPa (that is, 4m/s x 0.4MPa/(m/s)).
Since the design pressure of the proposed pipeline network is
24.5MPa, the maximum increase in pressure due to sudden
valve closure or station outage is no more than 6.5% above
the design pressure. Because the maximum surge pressure
is less than allowed by the code (that is, less than 10%
above the design pressure), no increase in wall thickness is
required. The next section will comment on the effect of
surge pressure on ductile fracture arrest in CO2 pipelines.
Ductile fractures
Paragraph 402.5.1 of ASME B31.4 requires the designer of
CO2 pipeline systems to consider the possibility of ductile
4th Quarter, 2010
257
Fig.3. Density of CO2 with highlighting
showing the normal range of operating
conditions.
Decompression and crack velocity
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
fractures and provide reasonable protection to limit their
occurrence and length throughout the pipeline, with special
consideration at crossings and other appropriate locations.
More specifically, Paragraph 402.5.3 requires the designer
to minimize ductile fracture propagation by the selection
of pipe steel with appropriate fracture toughness and/or
by the installation of suitable fracture arrestors. Design
considerations include pipe diameter, wall thickness, fracture
toughness, yield strength, operating pressure, operating
temperature, and the decompression characteristics of CO2
with its associated impurities.
The possibility that very long ductile fractures can occur
in CO2 pipelines was first identified in the late 1970s [4]
almost 10 years after the first CO2 transmission pipelines
had been built in North America. Crack arrestors were
subsequently retrofitted to existing CO2 pipelines to help
reduce risk. Since that time new CO2 pipelines have been
built with crack arrestors.
Thicker walls and tougher steels can be used instead of crack
arrestors effectively to control long ductile fractures. If a
longitudinal subcritical crack in the pipe wall grows during
operation, it can initiate a longitudinal tear in the pipe wall.
If the pipe wall is not strong enough or tough enough to resist
the force of the decompressed pressure pushing against the
unrestrained walls of the pipe on either side of the crack, a
ductile fracture will form and run along the pipe.
The solution is to increase either wall thickness or fracture
toughness of the pipe steel, or both. Increasing wall thickness
helps by reducing the stress at the tip of the fracture, and
increasing toughness helps by enabling the steel to absorb
more energy when it tears. Alternatively, crack arrestors can
be installed to limit the length of the fracture.
In the unlikely event that a dense-phase CO2 pipeline
bursts and a ductile fracture starts to run, the sudden loss
of gas causes the CO2 to decompress isentropically into the
two-phase region. The decompression is rapid and highly
turbulent, and the gas and liquid components do not have
time to separate. The two phases behave like an homogenous
single phase with a very low sonic velocity (less than 100m/s).
The low sonic velocity causes a sustained pressure at the
moving tip of the fracture equal to the pressure at which
the decompression crosses the phase boundary into the
two-phase region. This high sustained pressure acts on the
unrestrained flaps of the fractured pipe just behind the
moving fracture tip, and concentrates stresses at the tip of
the fracture large enough to tear the steel wall and drive the
fracture along the length of the pipe. Pipe steels need to be
both strong and tough to resist the forces and prevent the
crack from propagating.
In order to calculate the pressure at the tip of the moving
fracture it is necessary to understand how CO2 decompresses
and the effect of phase behaviour. The actual composition
of CO2 in the proposed pipeline is important in this regard.
CO2 from identified sources is relatively pure but during
design it is not possible to identify all the future sources of
CO2. Therefore a worst-case estimate of composition from
present and future sources needs to be used during design
to protect the pipeline system against long ductile fractures
over its full operating life.
Of the possible impurities, hydrogen has the largest effect
on the phase boundary and decompression behaviour
of CO2. Nitrogen has a lesser but still large effect while
258
The Journal of Pipeline Engineering
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Fig.4.Velocity of sound in CO2 with
highlighting showing the normal range
of conditions.
Fig.5. Unit pressure surge for a change
in flow velocity of 1m/s.
other possible impurities have relatively minor effects.
The effect of hydrogen on the properties of CO2 has been
determined accurately by experimental and theoretical work
of Prausnitz and Gunn [5] and others [6, 7, 8]. Comparison
of experimental results with predictions of equations of
state using parameters developed by Prausnitz and Gunn
[5] have shown that the PR equation of state is sufficiently
accurate for design work over a wide range of pressures and
temperatures. and that the BWRS equation of state was
sufficiently accurate for temperatures greater than -20°C.
After studying a range of possible compositions from a
variety of sources that might need to be transported in the
proposed network, a design-case mixture of 95% CO2 with
1% hydrogen and 4% nitrogen by volume was chosen as
the basis for design work to prevent long ductile fractures
over the operating life of the project.
Figures 6, 7, and 8 illustrate how the pipe toughness that is
needed to arrest ductile fractures is found using the Battelle
two-curve method [9]. Each figure shows two sets of curves:
the first is a set of decompression pressure-wave velocity curves
for the design-case mixture for a range of temperatures from
15 to 55°C (59 to 131°F) when the pipeline decompresses
following a rupture. The decompression pressure wave
curves define the speed that each pressure level propagates
back along the pipe from the initial fracture site as the
pipeline decompresses. Figures 6, 7, and 8 are for sudden
decompression from operating pressures of 17.5, 21, and
24.5MPa (2538, 3046, and 3553 psi) respectively, and were
calculated using the BWRS equation of state.
The second set of curves in Figs 6, 7, and 8 is a set of Battelle
fracture-velocity J-curves for a range of toughness values. For
illustrative purposes the J-curves were calculated for NPS 18
(18-in diameter) Grade L450 (X65) pipe with wall thickness
for the three different operating pressures calculated using
a design factor of 72%. The fracture-velocity J-curves define
the relationship between the pressure driving the fracture
along the pipe and the velocity of the fracture. The toughness
4th Quarter, 2010
259
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Fig.6. Decompression pressure wave velocity curves for design-case CO2 decompressing from 17.5MPa intersecting
toughness J-curves for NPS18 L450 pipe with 12.4-mm wall.
required to arrest a ductile fracture is found where the two
sets of curves are tangential (just touching).
Figures 6, 7 and 8 show that higher operating pressures require
lower-toughness pipe steel. Figure 6 for decompression from
17.5MPa shows a required toughness of 120J; Fig.7 for
decompression from 21MPa shows a toughness of 60J; and
Fig.8, for decompression from 24.5MPa, shows a toughness
requirement of 40J.
An interesting observation from these figures is that,
when the initial operating temperature is kept constant,
the sustained pressure plateau is slightly lower for higher
initial operating pressures. This means that to prevent long
ductile fractures in CO2 pipelines, unlike gas pipelines,
they need to be designed to handle low operating pressures
rather than the maximum allowable operating pressure. If
CO2 pipelines are designed to prevent ductile fractures at
low operating pressures they are generally safer from the
point of view of controlling ductile fractures when they
are operated at higher pressures, up to the maximum hoop
stress allowed by the code. Another conclusion that can be
drawn from this counterintuitive result is that increases
in operating pressure due to surge waves do not need to
be considered when selecting pipe toughness to arrest
ductile fractures.
This counterintuitive behaviour of dense-phase pipelines is
distinctly different from gas pipelines where the sustained
pressure plateau increases markedly as operating pressure
increases. Figure 9 illustrates the reason for the difference,
and shows decompression paths from four different
operating points on a pressure-enthalpy diagram developed
using the BWRS equation of state for a mixture of 95%
pure CO2. Decompression paths from Points 1 and 2 in
the dense phase intersect the bubble-point line, whereas
decompression paths from Points 3 and 4 in the gas phase
intersect the dew-point line. Because bubble-point and dewpoint lines have different slopes, dense-phase decompression
from high-pressure Point 1 intersects the phase boundary
at a lower pressure than decompression from low-pressure
Point 2. On the other hand, gas-phase decompression from
high-pressure Point 3 intersects the phase boundary at a
higher pressure than decompression from low-pressure Point
4. This difference explains why it is easy to control ductile
fractures in dense-phase pipelines by increasing the design
pressure, but not so easy in gas-phase pipelines.
Toughness calculations
The decompression path from 24.5MPa and 40°C for
design-case CO2 is shown in Fig.8. It shows a sustained
pressure plateau extending between 60m/s and 330m/s at
a pressure of 8.4MPa. The toughness requirement where
the decompression path just touches the fracture velocity
curve is determined from the low-velocity end of the pressure
plateau. at a velocity of 60m/s.
At velocities as low as 60m/s, the fracture-velocity J-curve is
horizontal; it can be seen from Fig.8 that if the decompressed
pressure is kept constant at 8.4MPa there is no practical
difference between the toughness required for a fracture
speed of 60m/s and the toughness required for a fracture
speed of 0m/s.
For CO2 pipelines, this observation allows a simplification,
with no loss of accuracy for steel toughness up to
approximately 200J, by setting fracture velocity equal to zero
and fracture arrest pressure equal to the highest pressure at
which the decompression enters the two-phase region [4].
260
The Journal of Pipeline Engineering
Wall @ 175 bar & 72%
Wall @ 210 bar & 72%
Wall @ 245 bar & 72%
Diameter
Steel
Grade
Thickness
Toughness
Thickness
Toughness
Thickness
Toughness
inches
MPa
mm
J
mm
J
mm
J
8.625
450
5.9
∞(1)
7.1
35.6
8.3
22.7
10.75
450
7.4
∞(1)
8.8
44.4
10.3
28.3
12.75
450
8.7
∞(1)
10.5
52.6
12.2
33.6
16
450
11.0
∞(1)
13.2
66.0
15.4
42.1
20
450
13.7
∞(1)
16.5
82.5
19.2
52.7
24
450
16.5
∞(1)
19.8
99.0
23.0
63.2
Table 2.Wall thickness to satisfy the requirements of ASME B31.4 and notch toughness to prevent long ductile fractures in
pipelines carrying the design-case mixture. Note (1): use crack arrestor or increase wall thickness when the decompressed
stress ratio (σd/σf) exceeds 0.28
(5)
where
There are approximations in the development of the
Battelle fracture-arrest equations and uncertainties in their
application to CO2 pipelines that need to be considered.
The underlying theory was originally calibrated using
results from full-scale pipe-burst tests to advance the design
of pipelines carrying natural gas. The theory has been
extended without additional experimental validation to
CO2 pipelines operating at much higher pressures than
natural gas pipelines. Full-scale burst tests with proposed
pipe, proposed operating pressure, and proposed pipelinequality CO2, are recommended to validate the applicability
of the theory to new CO2 pipelines.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
For many CO2 pipelines the fracture arrest pressure will
be equal to the cricondenbar of the mixture selected for
the design work. If the fracture velocity is set equal to zero
and the arrest pressure is set equal to the highest pressure
at which the decompression path intersects the phase
boundary (Pd), the Battelle fracture-arrest equations [9],
after rearranging terms, become:
(6)
(7)
As the ratio of hoop stress at the fracture tip to flow stress
(σd/σf) approaches 0.30, the normalized toughness (EN) in
Equn 6 approaches infinity and becomes highly sensitive
to small errors in the estimation of decompressed pressure
and flow stress. When the ratio exceeds 0.28, wall thickness
rather than toughness should be increased to control
ductile fractures, or crack arrestors should be used. As a
result, the practical limit of applicability of Equns 5, 6,
and 7 is given by:
(8)
Equations 5 to 8 can be used to calculate the toughness
(CV) required for the arrest of ductile fractures using
wall thickness (t), pipe diameter (D), decompressed
plateau pressure (Pd), and flow stress (σf). Alternatively,
the equations can be solved iteratively to find the wall
thickness (t) needed to arrest ductile fractures for any
given toughness (CV).
Future variations in product quality have been accounted
for in this work by selecting an appropriate design-case
mixture but, until full-scale burst tests are conducted to
validate the Battelle fracture-arrest equations for CO2
pipelines, an additional margin of safety is required in
order to account for uncertainties in the theory itself and
its extension to dense-phase CO2. This has been done by
adding 0.4MPa to the arrest pressure (cricondenbar). The
cricondenbar of the design-case mixture is 8.4MPa, so that
the arrest pressure selected for this project for use in Equn
7 was therefore 8.8MPa.
Table 2 shows the minimum wall thickness required to
satisfy the hoop-stress requirements of ASME B31.4 with a
design factor of 0.72 as well as the Charpy V-notch toughness
required to prevent long ductile fractures using fracture
arrest Equns 5, 6, and 7 when the pipeline is transporting
the design-case mixture. Crack arrestors are indicated in
Table 2 for cases where the decompressed stress ratio (σd/
σf) exceeds 0.28.
Optimization
Table 2 defines three long CO2 transmission pipelines with
design pressures of 17.5, 21.0, and 24.5MPa using pipe sizes
from NPS 8 (8.625ih) to NPS 24 (24in). The three different
systems can be characterized as follows:
4th Quarter, 2010
261
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Fig.7. Decompression pressure wave velocity curves for design-case CO2 decompressing from 21MPa intersecting toughness
J-curves for NPS18 L450 pipe with 14.9-mm wall.
Fig.8. Decompression pressure wave velocity curves for design-case CO2 decompressing from 24.5MPa intersecting
toughness J-curves for NPS18 L450 pipe with 17.3-mm wall.
• pipelines with design pressure of 17.5MPa and the
thinnest wall of the three alternatives using crack
arrestors instead of toughness to arrest ductile
fractures;
• pipelines with design pressure of 21.0MPa using
relatively high toughness steel to arrest ductile
fractures;
• pipelines with design pressure of 24.5MPa and the
thickest wall of the three alternatives using relatively
low-toughness steel to arrest ductile fractures.
Optimization on a cost basis can be used to choose between
these alternatives. To make a fair comparison of systems with
different maximum design pressures, compressor facilities
for all the systems were designed to take CO2 from source at
the minimum pipeline operating pressure and deliver it to
injection facilities at the same minimum operating pressure.
Initial and intermediate compressor stations were located
as required to boost pressure from the minimum operating
pressure to the system-design pressure.
The pipe costs were based on current prices for API 5L
L450 PSL2 pipe and do not include a premium for highertoughness steels. It was assumed that crack arrestors could
be installed at no additional cost so that the lower-pressure
systems would not be unfairly penalized. Figure 10 shows
262
The Journal of Pipeline Engineering
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Fig.9. Decompression paths for 95%
pure CO2 showing why burst pressure
inside gas pipelines is higher when
initial pressure is higher, but in densephase pipelines it is lower when initial
pressure is higher.
Fig.10. Comparative unit cost of
transportation curves for dense phase
CO2 pipelines showing optimum
capacities for each pipe size.
comparative unit cost of transportation curves based on total
lifetime owning and operating costs for the three systems: it
can be seen that the lowest unit transportation cost (lowest
lifetime owning and operating cost) is obtained with a
design pressure of 24.5MPa for all pipe sizes. This means it
is more cost-effective for long CO2 transmission pipelines
to use pipe with thick walls, high operating pressures, and
low toughness steels, rather than thin walls, low operating
pressures, and high-toughness steels or crack arrestors.
A further increase in operating pressure is not expected
to further reduce cost of transportation because 24.5MPa
is close to the pressure rating of 1500 Class flanges at the
maximum pipeline operating temperature, and the high
cost of heavier flanges and fittings (2500 Class) would put
a step in the optimization curves.
Conclusions
ASME B31.4 sets out the equations required for calculating
minimum wall thickness of CO2 pipelines for any given
operating pressure and the Battelle fracture arrest Equns 5
to 7 in combination with Equn 8 can be used to calculate
toughness to arrest ductile fractures, as well as the need
for extra wall thickness or crack arrestors. Finally, cost
optimization provides a rational methodology for choosing
between the different alternatives produced by the first two
sets of requirements, leading to the selection of a pipeline
system that is both safe and economic. For this project a
design pressure of 24.5MPa was selected as the most costeffective alternative resulting in the adoption of thick walls
and low-toughness steels for the arrest of ductile fractures.
Although it is more cost effective to use the pipe wall rather
4th Quarter, 2010
263
than crack arrestors to arrest ductile fractures, the main
benefit is the increase in safety that comes from limiting the
length of ductile fractures to much shorter lengths than is
practical with crack arrestors. A secondary benefit of using
the pipe wall rather than crack arrestors to control ductile
fractures is that the pipeline is built with a thicker wall
making it more resistant to third-party damage.
References
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
1. D.L.Katz and G.G.King, 1973. Dense phase transmission of
natural gas. Energy Processing Canada, November and December.
2. American Society of Mechanical Engineers, 2006. ASME B31.42006 Pipeline transportation systems for liquid hydrocarbons
and other liquids.
3. Idem, 2007. ASME B31.8-2007 Gas transmission and
distribution piping systems.
G.G.King, 1981. Design considerations for CO2 pipe line.
Pipe Line Industry, November, pp125-132.
4. J.M.Prausnitz and R.D.Gunn, 1958. Volumetric properties of
nonpolar gaseous mixtures. A.I.Ch.E Journal, 4, 4, December,
pp430-435.
5. C.Yokoyama, K.Arai, S.Saito, and H.Mori, 1988. Bubble-point
pressures of the H2-CO-CO2 system. Fluid Phase Equilibria, 39,
101-110.
6. J.O.Spano, C.K.Heck, and P.L.Brick, 1968. Liquid-vapor
equilibria of the hydrogen-CO2 system. J. Chem. Eng. Data,
13, 168-171.
7. C.Y.Tsang and W.B.Streett, 1981. Phase equilibria in the H2/
CO2 system at temperatures from 220 to 290 K and pressures
to 172 MPa. Chem. Eng. Sci., 36, 993-1000.
8. R.J.Eiber, T.A.Bubenik, and W.A.Maxey, 1993. Final report
on fracture control technology for natural gas pipelines.
Project PR-3-9113, NG-18 Report No. 208, Pipeline Research
Committee, American Gas Association, December.
UPSF
This two-day forum will address capabilities and guidance concerning
tools for corrosion and mechanical damage inspection of ‘unpiggable’
oil, gas and hazardous liquids pipelines.
The focus will be on existing, new, and developing in-line tools as well as
those in research and development.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Advances and best practices in guided-wave technologies will also
comprise a significant part of the program. Live-line access techniques,
combined with low-flow restriction robotic and wire-line-powered
internal inspection tools, will also be a special focus, especially for
pipeline segments that cannot be taken out of service when these tools
suggest a need for further integrity inspections.
Contact us today for information on registration or about sponsorship
& exhibiting options.
Program Chairman
Dr. Keith Leewis, P-PIC
Program advisory Committee
Mark Andraka, PECO Energy
Drew Hevle, El Paso Corp.
Richard Kania, TransCanada Pipelines
Garry Matocha, Spectra Energy
Bryan Melan, Marathon Oil Co.
Andrew Pulsifer, CenterPoint Energy
Albert Van Roodselaar, Chevron Energy Technology
houston marriott Westchase hotel
ConferenCe organizers
B.J. Lowe,
Clarion Technical Conferences
John Tiratsoo,
Tiratsoo Technical
Call +1 713 521 5929 or visit www.clarion.org
4th Quarter, 2010
265
A dynamic boundary ductilefracture-propagation model for
CO2 pipelines
by Prof. Haroun Mahgrefteh*, Solomon Brown, and Peng Zhang
Department of Chemical Engineering, University College London, UK
T
T
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
HE DEVELOPMENT and testing of a dynamic boundary ductile-fracture-propagation model for
pressurized CO2 pipelines is presented. The model accounts for all the important fluid-structure
interactions governing the fracture process.These include expansion-wave propagation, real fluid behaviour,
pipe/wall fiction, and heat transfer, as well as the rapidly diminishing dynamic loading effects as the crack tip
opens. The resistance to crack-tip propagation is determined based on the drop-weight tear test energy
approach. The performance of the fracture model is tested by comparison of its predictions of the crackpropagation velocity versus crack length against real data. The latter include the High-Strength Line Pipe
Committee, ECSC X100 and Alliance full-scale burst tests conducted for pipes containing either air or rich
gas mixtures. In all cases good agreement is obtained between the model predictions and the real data.The
validated model is used to test the propensity of a hypothetical but realistic pressurised CO2 pipeline to
ductile fracture propagation failure. The simulations indicate the remarkably significant role of the starting
line temperature on fracture propagation in CO2 pipelines.
HOUSANDS OF KILOMETRES of pressurized pipelines
are used to transport large amounts of hydrocarbons
across the world. Although this method of transportation
is generally considered to be safe, pipeline failures do occur
with some leading to catastrophic consequences (see for
example Refs 1, 2). In the US alone, despite having one of
the most stringent safety requirements across the globe, over
202 pipeline incidents were reported during 2005 – 2009 [3].
These resulted in an estimated $2 billion of damage leading
to 69 deaths and 254 serious injuries.
Propagating factures are considered as by far the most
catastrophic type of pipeline failure. Such failures
involve the rapid axial splitting or tearing of the pipeline,
sometimes running over distances of several hundred
meters resulting in massive loss of inventory in a very
short time. Deservedly, understanding and modelling of
the mechanisms responsible for such type of failure has
led to a large number of studies (see for example Refs 4,
5). Such interest has intensified recently [6-8] given the
prospect of using pressurised pipelines for transporting
captured CO2 from fossil plants for subsequent storage.
This paper is based on one presented at the First International Forum on Transportation
of CO2 by Pipeline, organized in Newcastle upon Tyne in July, 2010, by Tiratsoo
Technical and Clarion Technical Conferences, and with the support of the University
of Newcastle and the Carbon Capture and Storage Association.
*Author’s contact details
tel: +44 (0)20 7679 3835
email: [email protected]
Given that CO2 at concentration of >10% v/v is likely to
be instantly fatal [9], the rupture of a CO2 pipeline near
a populated area can lead to catastrophic consequences.
Fractures can initiate from defects introduced into the
pipe by outside forces such as mechanical damage, soil
movement, corrosion, material defects, or adverse operating
conditions. Fractures propagate when the stresses acting
on the defect overcome the fracture initiation tolerance
of the pipe, reaching a critical size based on the pipeline
material properties and operating condition. As such it is
highly desirable to design pipelines such that when a defect
reaches a critical size and fails, the result is a leak rather than
a long running facture.
The above requires a two-tiered design approach involving:
• providing sufficient fracture initiation resistance,
mainly via specifying the required pipe toughness,
wall thickness and operating conditions
• ensuring sufficient fracture propagation resistance
such that if a running fracture occurs its length is
limited to a short distance
Notwithstanding cost implications, fracture initiation can
be largely controlled a priori by specifying the required
fracture initiation toughness, minimum wall thickness and
the maximum stresses acting upon the defect.
266
The Journal of Pipeline Engineering
Parameter
HLP
ECSC
Alliance
Rich Gas (see
table 2)
Air
Rich Gas (see
table 2)
1.182
1.182
1.4223
0.8856
0.0183
0.0183
0.0183
0.0191
0.0142
Initial pressure (bara)
116
116
104
126
120.2
Initial temperature (oC)
12
6
-5
20
23.9
Ambient pressure (bara)
1.01
1.01
1.01
1.01
1.01
Ambient temperature (oC)
20
20
20
20
20
Pipe length (m)
35
35
35
35
100
Tensile stress (MPa)
505
505
505
807
505
Yield stress (MPa)
482
482
482
728
482
Pipe grade
X70
X70
X70
X100
X70
B1
C2
Inventory
Air
Air
Internal diameter (m)
1.182
Pipe thickness (m)
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
A1
Table 1. Pipeline characteristics and prevailing conditions used for the full-scale burst tests.
However, controlling fracture propagation once a leak has
formed is more complex, presenting a unique set of challenges.
As well as the fracture toughness of the steel and the backfill
conditions, the fracture-propagation velocity and arrest length
depend on the depressurization rate, the thermal stresses, and
the minimum pipe wall temperature relative to its ductileto-brittle transition temperature. To model the above and
hence develop methodologies for overcoming such a type of
failure, we need to understand the nature of the processes
taking place once a fracture has been initiated.
The onset of a leak in the pressurized pipeline results in a
series of expansion waves that propagate from the rupture
plane towards the intact end of the pipeline at the speed of
sound [10]. As the main driving force for crack propagation
is the crack tip pressure [11], the precise tracking of the
expansion waves, and their effect on the pressure profile
along the pipeline, is essential for the proper modelling of
fracture propagation.
Fig.1. Schematic representation of the
experimental setup used in the HLP
full-scale pipe burst tests [20].
Additionally, given the significant drop in the speed of
sound and hence the depressurization rate during the
transition from the gaseous to the two-phase region [12],
such analysis must also account for real fluid behaviour
through the use of an appropriate equation of state. Also,
non-isentropic effects such as the fluid/pipe wall friction
and heat transfer must be accounted for as these also directly
affect the depressurization rate. Finally the temperature
drop as a result of the Joule-Thomson expansion cooling
[13] of the fluid within the pipeline during discharge can be
significant. In the case of CO2, depending on the starting
conditions, such temperatures can reach as low as -70°C
resulting in very significant localised cooling of the pipe
wall in contact with the escaping fluid.
The minimum pipe wall temperature reached relative to
its ductile to brittle transition temperature will dictate
whether the pipeline will fail in the ductile or brittle
fracture manner. The modelling of brittle fractures in
267
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
4th Quarter, 2010
Fig.2.Variation of crack velocity with
crack length for test A1 south-running
crack.
Inventory:
air, initial pressure = 116bara, initial
temperature = 12°C.
Curve A: experimental data [20].
Curve B. DBFM prediction.
pressurized pipelines has been presented in the authors’
previous publication [13]. Ductile fractures are the focus
of our attention in this work.
The so called Battelle Two-Curve (BTC) approach by
Maxey [5] was the first used to express the criterion for the
propagation of a ductile fracture in terms of the relation
between the fluid decompression-wave velocity and the
crack-propagation velocity. If the fluid decompressionwave velocity is larger than the crack velocity, the crack tip
stress will decrease, eventually dropping below the arrest
stress and causing the crack to arrest. Conversely, if the
decompression-wave velocity remains smaller than the crack
velocity, the crack tip pressure will remain constant resulting
in indefinite propagation.
Several studies have since been conducted for modelling
ductile fractures based on the BTC approach (see for
example Refs14, 15). Some employ sophisticated finiteelement methods for simulating material deformation
but use over-simplistic transient fluid flow models for
predicting the rupture plane pressure and hence the crack
driving force (see for example Refs 16, 17). Others, on
the other hand, although accounting for the transient
depressurization profile within the pipeline, do not
deal with the impact of pipe wall heat transfer and
friction effects on the fluid decompression behaviour
(see for example Refs 18, 19). Additionally a reliable
decompression model must also incorporate a suitable
equation of state. This is especially important in the
case of CO2 pipelines given the unique depressurization
thermodynamic trajectory of CO2 [8].
Crucially none of the studies reviewed simulate the dynamic
interaction between the rapidly changing crack tip opening
area and the pressure loading as the crack propagates.
In this paper, we report the development and validation of
a rigorous dynamic boundary ductile-fracture-propagation
model which takes into account all of the important transient
fluid/structure interactions governing the fracture process.
The performance of the model in terms of predicting the
crack-propagation velocity and arrest length is tested by
comparison against real data. These full-scale burst tests
conducted by the High-Strength Line Pipe Committee[20],
ECSC X100 [21] and Alliance[22] for pipes containing either
air or rich gas mixtures.
The validated model is used to test the propensity of a
hypothetical but realistic pressurized CO2 pipeline to ductile
fracture propagation failure, paying particulate attention
to the impact of the starting line temperature. The latter
investigation was prompted by the findings of Cosham
and Eiber [7] indicating the significance of the starting
temperature on the CO2 depressurization trajectory relative
to its phase transition boundary.
268
The Journal of Pipeline Engineering
Theory
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Fig.3.Variation of crack velocity with
crack length for test B1 south-running
crack.
Inventory:
air, initial pressure = 116bara, initial
temperature = 6°C.
Curve A: experimental data [20].
Curve B: DBFM prediction.
The full background theory of the fluid flow model employed
in this study to predict the fluid decompression velocity
and the crack tip pressure for a given opening area is given
elsewhere [23-26], and hence only a brief account of its
main features is given here. Based on the homogeneous flow
assumption, in the case of unsteady, one-dimensional flow
the mass, momentum and energy conservation equations
are respectively are given by:
where ρ, u, P and h are the density, velocity, pressure and
specific enthalpy of the homogeneous fluid as function of
time, t, and space, x; qh is the heat transferred through the
pipe wall to the fluid and βy is the friction force term given by:
where, fw is the Fanning friction factor and D the pipeline
diameter.
Also,
where θ is the angle of inclination of the pipeline to the
horizontal.
Equations 1-3 are quasi-linear and must be solved numerically.
In this study, the Method of Characteristics (MOC) [27] is
used as the numerical solution method, as opposed to other
numerical techniques such as finite-element [28, 29] and
finite-difference methods [30-32] as both have difficulty in
handling the choking condition at the rupture plane. The
MOC handles the choked flow intrinsically via the Mach
line characteristics. Moreover, MOC is considered to be more
accurate than the finite-difference method as it is based on
the characteristics of wave propagation. Hence, numerical
diffusion associated with a finite-difference approximation
of partial derivatives is reduced.
The key step in the BTC method is the derivation of two sets
of curves: one set describing the crack velocity, and the other
the velocity of fluid decompression wave. The resistance to
crack propagation is indicated by the Charpy V-Notch (CVN)
269
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
4th Quarter, 2010
Fig.4.Variation of crack velocity with
crack length for test C2 south-running
crack.
Inventory:
rich gas (Table 2), initial pressure =
104bara, initial temperature = -5°C.
Curve A: experimental data [20].
Curve B: DBFM prediction.
energy [5]. However, in the full-scale pipe bust tests conducted
by the High-Strength Line Pipe Committee (HLP) [33], the
BTC theory is used in conjunction with the drop-weight
tear test (DWTT) energy, as this is shown to provide a more
accurate indication of the pipeline resistance to fracture.
Consequently this is the model applied in this work.
The two-curve model for the crack propagation velocity, vc,
and crack arrest pressure, Pa, are respectively given by [33]:
Component
HLP C2
Alliance Test 1
CH4
89.57
80.665
C2H6
4.7
15.409
C3H8
3.47
3.090
iC4H10
0.24
0.232
nC4H10
0.56
0.527
iC5H12
0.106
0.021
nC5H12
0.075
0.014
nC6H14
0.033
0.003
nC7H16
0.017
0
nC8H18
0.008
0
nC9H20
0.001
0
N2
0.5
0.039
CO2
0.72
0
Table 2. Rich gas feed compositions.
where σflow, Dp, and Ap are respectively the flow stress (the
mean value of the tensile and yield stresses), pre-cracked
DWTT energy and ligament area of a pre-cracked DWTT
specimen. On the other hand Pt and tw are the crack tip
pressure and pipe wall thickness respectively. The crack
tip pressure Pt is taken to be the choked pressure at the
pipeline release plane.
For brevity, full details of the coupling of the fracture and the
fluid decompression models will be reported in a separate
The Journal of Pipeline Engineering
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
270
Fig.5.Variation of crack velocity with
crack length for test ECSC X100
south-running crack.
Inventory:
air, initial pressure = 126bara, initial
temperature = 20°C.
Curve A: experimental data [33].
Curve B: DBFM prediction.
study. The calculation algorithm automatically corrects
for the effective pipeline length and hence the crack tip
pressure as crack opens. The required fluid decompression
velocity and the crack tip pressure are determined from
the numerical solution of the conservation equations (1-3)
using our CFD computational package, PipeTech [34]. The
Peng-Robinson [35] equation of state (PR EoS) is used for
the prediction of the pertinent fluid phase equilibrium data
for both air and rich gas mixtures. In the case of CO2, the
Modified Peng-Robinson [36] EoS is used. As compared
to the PR EoS, this equation has been shown to produce
better predictions of the phase equilibrium data during the
most part of the depressurization process [37].
Results and discussion
Validation
The following shows the results relating to the validation of the
dynamic boundary ductile-fracture model presented above,
hereby referred to as DBFM, by comparison of its predictions
against the following published experimental data:
• HLP full-scale burst test [20]
• ECSC X100 pipe full-scale burst test [21]
• Alliance full-scale burst tests [22]
Table 1 shows the pertinent conditions relating to each
test. Table 2 on the other hand shows the rich gas feed
compositions for HLP C2 and Alliance tests.
The full-burst-test pipelines used comprised several sections
of differing toughness for which the corresponding DWTT
energy may be calculated. In all simulations, the pipe wall
roughness and heat-transfer coefficient are taken as 0.05 mm
and 5 W/(m2 K) respectively. The latter correspond to the
uninsulated pipeline exposed to still air in all simulations.
An equidistant grid system comprising 100 nodal points is
employed for the fluid dynamic simulations using PipeTech.
The corresponding discretisation time element is determined
using 90% of the Courant, Friedrichs and Lewy value [38].
The HLP full-scale experiments involved three series of
burst tests, referred to as test series A, B, and C using X70
API grade pipelines containing air and a rich gas mixture.
Pipeline fracture was initiated using an explosive charge.
Figure 1 shows a schematic representation of the pipe setup.
Figures 2 to 4 show the variation of the crack velocity with
crack length for the south-running A1, B1 and C2 tests
respectively for the HLP full-scale experiments. Curves A
show the measured crack length; Curves B, on the other
hand are the simulation predictions. In all cases, the
271
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
4th Quarter, 2010
Fig.6.Variation of crack velocity with
crack length for test Alliance Test 1.
Inventory:
rich gas (Table 2), initial pressure
= 120.2bara, initial temperature =
23.9°C.
Curve A: experimental data [22].
Curve B: DBFM prediction.
corresponding Charpy Energy, Cv, for each pipe section is
given in the figures.
Figures 5 and 6 show the corresponding data for ECSC
X100 [21] and Alliance full-scale burst tests [22], respectively.
Returning to Figs 2-6, as it may be observed, the crack
velocity significantly decreases with increase in crack length.
This is due to the significant rapid decrease in the crack
tip pressure as the pipeline depressurizes. As an example,
such behaviour expressed in terms of the variation of the
cark tip pressure with time is shown in Fig.7 for the HLP
A1 south-running crack.
Also as expected, the crack velocity decreases as the crack
propagates into the pipeline section with the higher
toughness, eventually coming to rest in all cases. As expected,
the data in Fig.4 show the smallest crack length as compared
to the other tests due to the combination of the much higher
fracture toughness pipe material employed together with
the lowest initial pressure.
The initial rapid increase in the crack velocity observed in many
of the test data is due to the finite time taken for the initial
notch to fully develop into an open flap following detonation.
This time domain is ignored in the present simulations.
Returning to the simulation data (curves A), given the
experimental uncertainties, it is clear that in all cases the
DBFM predictions produce reasonably good agreement
with the test data.
CO2 pipeline ductile fracture investigation
The following describes the results of the application of
the validated DBFM to the rupture of a hypothetical CO2
pipeline. To ensure practical relevance, the respective
pipeline internal diameter and wall thickness of 590.7mm
and 9.45mm are employed in the simulations. Cosham and
Eiber [7] suggest that such dimensions are the most likely for
CO2 pipelines to be employed in CCS. The same authors
also propose that a Cv of 50 J would be sufficient to arrest
a fracture for typical operating conditions of 100barg and
10oC. The same pipeline operating conditions are chosen
in the proceeding simulations.
Figure 8, curve A, shows the predicted variation of the crack
velocity versus crack length based on the above conditions
for the CO2 pipeline. For the sake of comparison, the
analogous data for methane (curve B) and natural gas
(85% methane-15% ethane, curve C) inventories are also
presented.
The Journal of Pipeline Engineering
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
272
Fig.7. Simulated variation of crack tip
pressure versus time for test HLP A1
south-running crack.
As may be observed following an initial drop, the natural gas
pipeline (curve C) exhibits a relatively constant high-velocity
crack which propagates through almost the entire length of the
pipeline before coming to rest at approximately 89m. Similar
trends in behaviour but of a smaller magnitude is observed in
the case of the methane pipeline (curve B) where the fracture
comes to rest at a distance of approximately 18m. Of the
three cases examined, the CO2 pipeline (curve A) offers the
best resistance to ductile fracture. Here the fracture almost
instantaneously comes to rest at a distance of only 6m.
Impact of line temperature
Based on an analysis of the CO2 depressurization trajectory,
Cosham and Eiber [7] postulate that the initial temperature
of the CO2 pipeline may have a significant impact on its
resistance to ductile fracture failure.
Figure 9 shows impact of the line temperature on the
variation of fracture velocity versus fracture length for the
CO2 pipeline at four selected temperatures of 30°C (curve
A), 20°C (curve B), 10°C (curve C), and 0°C (curve D).
As it may be observed, in the range 0-20°C, an increase in
temperature results in a relatively modest increase in the
fracture velocity and fracture arrest length. The data at 30°C
(curve A) is an exception to this rule. Remarkably only a
10°C rise in the line temperature results in a fast-running
propagating fracture which covers the entire length of the
pipeline.
To explain the above, Fig.10 shows the variation of the
crack tip pressure with temperature for the starting line
temperatures of 10, 20, and 30°C relative to the CO2
saturation curve. The latter is generated using the Span
and Wagner [39] equation of state for CO2. The calculated
crack arrest pressure of 43bara is also indicated in the same
figure. In the case of the highest line temperature of 30°C
(curve A), soon after pipeline failure, the dense-phase CO2
inventory crosses the saturation curve at the maximum
pressure of ca. 60bara, some 17bara higher than the crack
arrest pressure of 43bara, thus resulting in a propagating
fracture. The fracture comes to rest once the crack tip pressure
is equal to the crack arrest pressure. By the time this occurs
in the case of the 30°C pipeline, the crack will have already
propagated through the entire length of the pipe.
Conclusion
The development and validation of a dynamic boundary
ductile-fracture-propagation model for pressurized pipelines
was presented. The model, based on the coupling of a
semi-empirical fracture model with the transient real fluid
273
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
4th Quarter, 2010
Fig.8.Variation of crack velocity with
crack length for a 100-m long pipe at
100barg and 10°C containing various
inventories.
flow simulator, PipeTech, takes into account all of the
important fluid/structure interactions governing the fracture
propagation and arrest process.
A particularly important feature is accounting for the change
in the effective pipeline length as the pipeline unzips and
its impact on the crack tip pressure. Following its successful
validation against real pipe burst data reported for air and
rich-gas mixtures, the model is used to test the propensity
of a hypothetical but realistic pressurized CO2 pipeline to
ductile-fracture-propagation failure. Such investigations are
particularly timely given the real prospects of using CO2
pipelines as part of the carbon capture and sequestration
(CCS) chain.
propagation. A relatively modest increase in the line
temperature from 20 to 30°C resulted in a running fracture
which propagated through the entire length of the 100m
pipe. Analysis of the data revealed that upon crack initiation,
the pipeline inventory rapidly transforms from the dense
phase into the saturated state, thereafter following a relatively
prolonged depressurization trajectory along the saturation
curve. The crack will propagate for as long as the crack tip
pressure remains higher than the crack arrest pressure. In
the case of the CO2 pipeline at 30°C this cross over will
not happen before the fracture has propagated through the
entire length of the pipe. Obviously such phenomenon will
have significant practical implications when transporting
CO2 at different ambient temperatures as part of the CCS.
Simulations conducted using a 100-m long pipeline
containing methane, natural gas or CO2 at 100barg and
10°C revealed that whereas for the natural gas and methane
inventories the fracture propagated through most of the pipe
length, in the case CO2 pipe, the crack length was limited
to only a short distance.
The study assumes that the adopted rather simplistic but
nevertheless effective drop-weight tear test (DWTT) energy
approach validated for air and rich gas mixtures is also
applicable to CO2. This assumption is justified given the fact
that the only fluid parameter introduced into the DWTT
approach is the crack tip pressure.
The change in the temperature of CO2 however was found
to have a remarkable impact on the resistance to fracture
Furthermore the fluid-flow model employed is based on
the plausible homogenous flow assumption in which the
The Journal of Pipeline Engineering
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
274
Fig.9.Variation of crack velocity with
crack length for the 100-m long, 100barg CO2 pipe at different starting line
temperatures.
constituent fluid phases remain at thermal and mechanical
equilibrium during the fracture-propagation process.
Addressing the non-equilibrium phase behaviour, together
with the impact of the typical impurities present within the
CO2 stream for the various capture technologies on the
ductile fracture behaviour of CO2 pipelines, is currently
being investigated by the authors.
References
1. A.Bartenev, 1996. Statistical analysis of accidents on the Middle
Asia-Centre gas pipelines. J. Haz. Mat., 46, 1, pp57-69.
2. G.Papadakis, 1999. Major hazard pipelines: a comparative
study of onshore transmission accidents. J. Los. Prev. Proc. Ind.,
12, 1, pp91-107.
3. PHMSA, 2010. Pipeline incidents and mileage reports. Retrieved
from primis.phmsa.dot.gov/comm/reports/safety/PSI.html.
4. B.N.Leis, X.Zhu, and T.Forte, 2005. Modelling running fracture
in pipelines: past, present and plausible future directions.
Proc. 11th Int. Conf. Frac., Italy.
5. W.A.Maxey, 1974. Fracture initiation, propagation and arrest.
In: Proc. 5th Symposium in Line Pressure Research. Houston.
6. H.Mahgerefteh, G.Denton, and Y.Rykov, 2008. CO2 pipeline
rupture. IChemE Symposium Series: HAZARDS XX Process
Safety and Environmental Protection, 154, pp869 - 879.
Manchester: IChemE.
7. A.Cosham and R.J.Eiber, 2008. Fracture propagation in CO2
pipelines. J. Pipe. Eng., 7, pp115-124.
8. M.Bilio, S.Brown, M.Fairwheather, and H.Mahgerefteh, 2009. CO2
pipelines material and safety considerations. IChemE Symposium
Series: HAZARDS XXI Process Safety and Environmental
Protection, 155, pp423–429. Manchester: IChemE.
9. H.Kruse and M.Tekiela, 1996. Calculating the consequences of
a CO2-pipeline rupture. En. Conv. Man., 37, 95, pp1013-1018.
10. D.J.Picard and P.R.Bishnoi, 1988. The importance of realfluid behaviour and non-isentropic effects in modeling
decompression characteristics of pipeline fluids for application
in ductile fracture propagation analysis. Cana. J Chem. Eng.,
66, 1, pp3-12.
11. G.Fearnehough, 1974. Fracture propagation control in gas
pipelines: a survey of relevant studies. Int. J. Pres. Ves. Piping,
2, 4, pp257-282.
12. Idem., 1987. Calculation of the thermodynamic sound velocity
in two-phase multicomponent fluids. Int. J. Multip. Flow, 13,
3, pp295–308.
13. H.Mahgerefteh and O.Atti, 2006. Modelling low-temperatureinduced failure of pressurized pipelines. AIChE J., 52, 3, pp1248-1256.
14. B.N.Leis, 1997. Relationship between apparent (total Charpy
V-Notch toughness) and the corresponding dynamic crack-
275
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
4th Quarter, 2010
Fig.10.The variation of crack tip
pressure with temperature for the 100m long, 100-barg CO2 pipe at different
starting line temperatures.
propagation resistance. Proc. Int. Pipe. Conf., Calgary: ASME.
15. H.Makino, I.Takeuchi, and R.Higuchi, 2008. Fracture
propagation and arrest in high-pressure gas transmission
pipeline by ultra strength line pipes. In: 7th Int. Pipeline
Conf.. Calgary.
16. P.E.O’Donoghue, M.F.Kanninen, C.P.Leung, G.Demofonti,
and S.Venzi, 1997. The development and validation of a
dynamic fracture propagation model for gas transmission
pipelines. Int. J. of Pres. Ves. Piping, 70, 1, pp11-25.
17. Z.Zhuang, 1999. Analysis of dynamic fracture mechanisms in
gas pipelines. Eng. Frac. Mech., 64, 3, pp271-289.
18. H.Makino, T.Sugie, H.Watanabe, T.Kubo, T.Shiwaku, S.Endo,
et al., 2001. Natural gas decompression behaviour in high
pressure pipelines. ISIJ Int., 41, 4, pp389-395.
19. A.Terenzi, 2005. Influence of real-fluid properties in modelling
decompression wave interacting with ductile fracture
propagation. Oil Gas Sci. Tech., 60, 4, pp711–719.
20. T.Inoue, H.Makino, S.Endo, T.Kubo, and T.Matsumoto, 2003.
Simulation method for shear fracture propagation in natural
gas transmission pipelines. In: International Offshore and
Polar Engineering Conf., 5, pp121-128. Honolulu.
21. I.Takeuchi, H.Makino, S.Okaguchi, N.Takahashi, and
A.Yamamoto, 2006. Crack arrestability of high-pressure gas
pipelines by X100 or X120. In: 23rd World Gas Conference.
Amsterdam.
22. D.M.Johnson, N.Horner, L.Carlson, and R.J.Eiber, 2000. Full
scale validation of the fracture control of a pipeline designed
to transport rich natural gas. Pipeline Technology, 1, 331.
23. H.Mahgerefteh, O.Atti, and G.Denton, 2007. An interpolation
technique for rapid CFD simulation of turbulent two-phase
flows. Pro. Safe. Env.
24. H.Mahgerefteh, A.Oke, and Y.Rykov, 2006. Efficient numerical
solution for highly transient flows. Chem. Eng. Sci., 61, 15,
pp5049-5056.
25. H.Mahgerefteh, P.Saha, and I.G.Economou, 1999. Fast
numerical simulation for full bore rupture of pressurized
pipelines. AIChE J., 45, 6, pp1191–1201.
26. A.Oke, H.Mahgerefteh, I.Economou, and Y.Rykov, 2003. A
transient outflow model for pipeline puncture. Chem. Eng.
Sci., 58, pp4591-4694.
27. M.J.Zucrow and J.D.Hoffman, 1975. Gas dynamics. Wiley,
New York.
28. C.Bisgaard and H.Sørensen, 1987. A finite element method
for transient compressible flow in pipelines. Int. J. Numer.
Methods Fluids, 7, pp291-303.
29. E.Lang, 1991. Gas flow in pipelines following a rupture
computed by a spectral method. J. App. Math. Phy., 42, March.
30. K.Bendiksen, D.Maines, and R.Moe, 1991. The dynamic twofluid model OLGA: theory and application. SPE Production,
6, 6, pp171–180. Society of Petroleum Engineers.
276
The Journal of Pipeline Engineering
36. D.Wu and S.Chen, 1997. A modified Peng-Robinson equation
of state. Chem. Eng. Comm., 156, 1, pp215-225.
37. H.Mahgerefteh, G.Denton, and Y.Rykov, 2008. A hybrid
multiphase flow model. AIChE J., 54, 9, pp2261–2268.
38. H.Mahgerefteh, Y.Rykov, and G.Denton, 2009. Courant,
Friedrichs and Lewy (CFL) impact on numerical convergence
of highly transient flows. Chem. Eng. Sci., 64, 23, pp4969-4975.
39. R.Span and W.Wagner, 1996. A new equation of state for
carbon dioxide covering the fluid region from the triple-point
temperature to 1100 K at pressures up to 800 MPa. J. Phys.
Chem. Ref. Data, 25, 6.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
31. J.R.Chen, S.M.Richardson, and G.Saville, 1995. Modelling
of two-phase blowdown from pipelines – 1: a hyperbolic
model based on variational principles. Chem. Eng. Sci., 50, 4,
pp695–713.
32. idem, 1995. Modelling of two-phase blowdown from pipelines
– 2: a simplified numerical method for multi-component
mixtures.Ibid., 50, 13, pp2173-2187.
33. H.Makino, I.Takeuchi, M.Tsukamoto, and Y.Kawaguchi, 2001.
Study on the propagating shear fracture in high strength line
pipes by partial-gas burst test. ISIJ Int., 41, 7, pp788-794.
34. PipeTech. Pipeline rupture simulation software: www.
pipetechsoftware.com.
35. D.Peng and D.B.Robinson, 1976. A new two-constant equation
of state. Ind. Eng. Chem. Fund., 15, 1, pp59-64.
4th Quarter, 2010
277
Will fractures propagate in a
leaking CO2 pipeline?
by Dr Robert Andrews*1, Dr Jane Haswell2, and Russell Cooper3
1 BMT Fleet Technology, Loughborough, UK
2 Pipeline Integrity Engineers, Newcastle upon Tyne, UK
3 National Grid Gas Transmission, Warwick, UK
A
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
HYPOTHETICAL CONCERN has been raised that leaks in a CO2 pipeline could escalate to a propagating
fracture.This is due to the potentially large temperature drop associated with the expansion of either
gaseous or dense-phase CO2 to ambient conditions. It is suggested this local cooling would lower the pipe
wall temperature to an extent that a brittle fracture would initiate followed by a transition to a propagating
fracture. Although such a mechanism could theoretically occur in natural gas pipelines, there is increased
concern for CO2 transport because of the different thermodynamic behaviour of the contents, particularly
for dense-phase transport.
This paper critically reviews the literature associated with this postulated failure mechanism and other
studies on the cooling of cracks and holes by escaping fluid. It is concluded that pipelines constructed to
modern standards are not at risk. Limited crack extension may occur when the leak is through a ‘tight’
crack in a material of low toughness. However, the crack will arrest as it enters warmer material remote
from the leak. Escalation to a propagating fracture can be controlled using methods which are widely used
and understood in the pipeline industry.
Introduction
Background
With the current drive towards reducing emissions of carbon
dioxide (CO2) gas to the atmosphere, pipeline operators
and designers are investigating options for the transport of
bulk quantities of CO2. This will involve moving CO2 from
sources such as power stations equipped with carbon capture
to storage sites such as aquifers or depleted oil or gas fields.
This will require either the construction of new pipelines,
or the re-use of existing pipelines originally constructed to
carry natural gas or other fluids.
A concern has been raised that leaks in a CO2 pipeline
could escalate to a propagating fracture. This is due to
the potentially large temperature drop associated with the
expansion of either gaseous or dense-phase CO2 to ambient
conditions. It is suggested this cooling would lower the
pipe wall temperature to an extent that a brittle fracture
This paper was presented at the First International Forum on Transportation of CO2
by Pipeline, organized in Newcastle upon Tyne in July, 2010, by Tiratsoo Technical
and Clarion Technical Conferences, and with the support of the University of
Newcastle and the Carbon Capture and Storage Association.
*Author’s contact details
tel: +44 (0)1509 621814
email: [email protected]
would occur followed by a propagating fracture. Although
such a mechanism could theoretically occur in natural gas
pipelines, there is increased concern for CO2 pipelines
because of the different thermodynamic behaviour of the
contents, particularly for dense-phase transport. There is
no public domain evidence that such a failure mechanism
has occurred either in natural gas pipelines in service or in
laboratory experiments.
The postulated failure mechanism involves a complex
interaction of the thermo-fluid mechanics of a leaking
pipeline, heat transfer driven by cold fluid escaping through
either a crack or a hole, crack initiation and propagation,
and crack arrest. These issues are discussed separately, after
a summary of the postulated failure mechanism in the next
section. The issues considered are grouped as fluid flow
and heat transfer at a leak, fracture initiation, immediate
arrest after initiation, and full-bore fracture propagation,
and these are considered successively in the subsequent
sections. This arrangement of the material does involve
some repetition, but it was considered the best approach
to ensure that all issues are covered. A general discussion
and conclusions then follow.
It should be noted that this paper has been prepared in the
context of UK pipeline design practices, although most of
the analysis and the conclusions should be appropriate for
other locations.
278
The Journal of Pipeline Engineering
Terminology
The following terminology has been used to distinguish
between different possible types of defect. It is important
that these distinctions are maintained, as there can easily
be confusion between the different possible types of loss of
containment in the pipe wall.
• Leak – a break in the pipe wall through which fluid
can escape. This is used as a generic term. A leak
is stable at the current length, applied stress and
metal temperature.
• Crack – a sharp-ended feature which can be analysed
by fracture mechanics methods which characterize
the crack tip stresses as a singularity.
• Hole – a rounded feature, typically (but not
necessarily) circular. Fracture mechanics’ methods
are not applicable to holes.
• Although not completely clear in the published
papers, it appears to be assumed in [2] that once
the condition in Eqn (1) is satisfied, a propagating
fracture occurs immediately. In [2] this is described
as “a secondary more catastrophic running brittle
fracture”; in [1] both ductile and brittle propagating
fractures are described.
No experimental evidence is given to support this
mechanism. In [3] it is suggested that this mechanism was the
cause of the gas pipeline failure at Ghislenghien, Belgium,
in July 2004. At the time of writing this incident is still the
subject of litigation and the official Belgian government
enquiry has not issued a report, so it is not clear if this
speculation is correct.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
• Rupture – a break in the pipeline which has an
opening equivalent to at least the full-bore area.
A rupture may remain stable or may escalate to a
propagating fracture.
• were Kc is described as “the critical fracture toughness
below which a fracture propagates”, Y is a “shape
factor depending on the crack length and geometry”,
a is the crack half length, and σ is “the sum of the
pressure and thermal stresses”.
• Propagating fracture – a fracture moving continuously
along the pipeline at a velocity of around 100 m/s
or higher.
Postulated failure due to local
cooling at a leak
The failure mechanism postulated in [1, 2] can be summarized
as follows:
• Fluid (usually a gas but possibly a dense-phase fluid)
escapes through a leak in the pipeline wall, expanding
and cooling.
• The low temperature fluid cools the pipe wall at
the leak.
• Cooling a carbon steel reduces its toughness as the
material becomes brittle at lower temperatures.
Additionally, the temperature differential between
cold material at the leak and warmer areas remote
from the leak induces thermal stresses.
• The combined effect of reduced toughness and
thermal stress initiates a fracture. The condition for
fracture is set in reference [2] by a pure linear elastic
fracture mechanics (LEFM) approach, with fracture
assumed to occur when:
(1)
Heat transfer and
fluid flow at a leak
As noted above, heat transfer between the leaking fluid and
the pipe wall is essential if the material around the leak is to
be cooled and possibly reach the lower shelf of the ductilebrittle transition curve. This section considers these issues.
Heat transfer and flow at a leak in a pipeline
The model in [2] assumes that flow both within the pipeline
and through the leak in the wall is turbulent and that heat
transfer is by forced convection. The leak is considered to
be a 5-mm diameter hole, with a crack extending from it
for 50mm. It is not clear if leakage is assumed through only
the hole, or through both hole and crack. Heat transfer
coefficients are estimated using coefficients from the
literature. Although not clearly stated, it appears that the
flow through the leak is treated in the same way as flow
within the pipe. The model also does not allow for heat
recovery from the surrounding soil as [2] refers to natural
and forced convective heat transfer “to ambient”. Whilst in
the immediate vicinity of the leak the pipe may be exposed,
even here it will be surrounded by escaped gas rather than the
ambient. Remote from the leak the pipeline will be buried
and heat transfer would be expected to be by conduction
from the soil, not by a convective mechanism.
An analysis has been carried out for leaking ethylene pipelines
in [4]. This analysed flow and heat transfer at circular holes
in the wall of ethylene pipelines, with contents in both
the gas and dense phases. In most cases the flow through
the hole was choked and the bulk of the fluid expansion
actually occurred in a shock wave outside the pipe wall.
4th Quarter, 2010
279
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
The fluid flowing through the hole was thus at a similar
temperature to the bulk fluid and so there was little local
cooling of the material around the hole. The exterior of
the pipe was surrounded by a cloud of cooled gas, but the
heat transfer from this gas to the pipe wall was relatively
low. Other points from this study were:
Fig.1. Predicted local wall temperatures
from [4] during blowdown of a
supercritical ethylene pipeline through
a 50-mm diameter hole at either the
top or bottom of the pipeline.
greatest cooling occurs when the pipeline is shut-in so that
the bulk fluid pressure and temperature fall. For the more
likely case of a small undetected leak, where flow continues,
the temperatures return to ambient over shorter distances as
the pipe wall is heated by the warm flowing bulk contents.
• Steady-state pipe metal temperatures were reached
within 1 minute, rather than taking up to 10 minutes
(the timescale predicted in [2]) to reach a steady state.
• Figure 1 shows predicted wall temperatures from
[4] as a function of pressure during blowdown
of a supercritical ethylene pipeline. The lowest
predicted temperature is -58°C but this occurs at
only approximately 25% of the initial pressure,
which would considerably reduce the driving force
for fracture initiation.
• For supercritical conditions, Fig.1 shows that the
location of the hole affects the temperatures and
behaviour once the pressure falls to the boiling point
and two-phase flow is established. This is because
– with a hole at the bottom of the pipe – relatively
warm liquid is forced through the hole; with a hole
at the top, colder gas escapes from the hole.
• The model in [2] appears to assume the pipe is above
ground, as there are references to convective heat
transfer. In contrast, Saville [4] explicitly models a
buried pipeline and includes heat recovery from the
soil. Crater formation at the leak due to displacement
of the soil by the escaping fluid was also considered.
Although there are differences in the treatment of the local
fluid flow and heat transfer, both models show that the metal
temperature recovers to the bulk fluid temperature over relatively
short distances. Saville’s model predicts a more rapid recovery
as there is less local cooling. Both models also predict that the
Experimental studies on leakage from full-scale pipelines
are rare, as they are difficult and expensive to perform. One
relevant study was carried out by SZMF [5] on a 1067-mm
diameter, 33.5-mm wall thickness grade L555 (X80) vessel
simulating a storage vessel for compressed natural gas (CNG)
transport applications. Leakage of lean natural gas through
a fatigue crack produced a lowest measured temperature (on
the outer surface near the crack tip) of -70°C, in agreement
with predictions made by an unknown method, and these
predictions are reproduced in Fig.2. They show a significant
through-wall temperature gradient, with the crack at inner
surface of the vessel at around -10°C. The cooling is greater
than [2] predicts for natural gas, albeit starting from a higher
pressure of 180 bar. This result suggests that leakage through
cracks may be more severe than leakage through holes.
Flow through cracks
Workers at Sheffield University have studied the flow of
fluids through narrow cracks, rather than through a hole
[6-8]. Whilst they did not study heat transfer effects, and
the pressure differentials were small, they did identify an
effect of surface roughness on the flow regime. At very small
openings, comparable to the surface roughness, the flow
was essentially laminar and “followed” the local surface
roughness. With increasing separation of the crack faces,
the flow could move between the peaks of the surface and
transitioned to turbulent flow at higher velocities.
This work shows that there may be complex effects of the
surface and crack opening; the heat transfer between the
fluid and the crack would be affected by the flow regime
and the fluid velocity.
280
The Journal of Pipeline Engineering
the leak, and so assuming a uniform stress will overestimate
the stress-intensity factor. The analysis also incorrectly treats
the thermal stress as a primary stress, rather than secondary.
Fig.2. Predicted temperatures at a through-wall fatigue crack
due to leakage of natural gas. Figure 3 of [5], Reepmeyer,
Lothe,Valsgaard, Erdelen-Peppler and Knauf: Full-scale gas
leak test at a large-diameter X-80 DSAW pipe. ASME
International Pipelines Conference 2006, IPC2006 10005.
Copyright ASME 2006, used with permission.
Nuclear industry studies
The model in [2] assumes the presence of a sharp crack
in all cases. Work on the failure of volumetric corrosion
defects in low-toughness linepipe [12, 13] has shown a
substantial influence of defect geometry on the failure
behaviour. The most significant factor appears to be
the local defect geometry, in particular the acuity of the
defect. The tests were carried out using material that
was on the lower shelf of the Charpy transition curve at
room temperature, but the behaviour of blunt machined
defects simulating volumetric corrosion at quasi-static
strain rates was predicted by conventional ductile-failure
models. The trend of the results is shown in Fig.3, which
shows the transition temperature as a function of the
stress-concentration factor associated with the defect. The
stress-concentration factor increases as the defect becomes
sharper, showing that the effective transition temperature of
a rounded defect can be well below the Charpy transition
temperature. Thus, for leakage through holes rather than
cracks, it is unlikely that fracture initiation would occur,
even in low-toughness material.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
The nuclear industry has studied the flow of fluids through
cracks as part of ‘leak before break’ (LBB) arguments, which
aim to show that a through-wall crack will leak before
breakage or rupture occurs. A part of these arguments is to
show that a leak can be detected before further sub-critical
crack growth leads to rupture; this requires an estimate of
leakage rates, and much of this work is summarized in [9].
However, the nuclear LBB studies have concentrated on
isothermal leakage, presumably because the containment
is effectively at constant temperature. This is not relevant
to the current issue for CO2 pipelines where heat transfer
to the containment (the pipe) from the fluid expanding in
a non-isothermal manner is the key issue.
The fracture analysis should, in any case, use more modern
methods such as the failure-assessment diagram approach
of [10] or [11]. These can take account of thermal-stress
gradients if these are significant. However, it is true that a
carbon steel pipeline will show transition behaviour and,
depending on the material properties, local cooling may
reduce the toughness.
Fracture initiation
This Section considers issues associated with the analysis
and prediction of fracture initiation from a locally cooled
area of a pipeline.
Fracture analysis methods
The model in [2] assumes LEFM applies, so that the initiation
of fracture is controlled by the stress-intensity factor; the paper
also assumes a step transition between ductile and brittle
behaviour, although this is an oversimplification of the true
behaviour of a pipeline steel. Ignoring the beneficial effects
of warm pre-stressing, discussed below, the behaviour of the
pipeline should reflect the low constraint of a thin-walled
structure loaded in tension. Thus, even on the lower shelf, the
behaviour is likely to be better than would be expected from
simple LEFM considerations. Fracture initiation should be
predicted using an elastic-plastic measure of toughness such as
the crack-tip-opening displacement (CTOD) or the J-integral.
There are other concerns with the fracture-analysis method
in [2], as the authors appear to assume that the thermal
and primary stresses can simply be added to give a higher
uniform hoop tensile stress when calculating the stressintensity factor. The thermal stress will decay remote from
Crack initiation due to local cooling
Fearnehough [14] carried out experiments using liquid
nitrogen vapour to cool steel plates containing a centre crack
of length 600mm and subject to remote axial tension stresses
in the range 15 to 124N/mm2. The aim was to simulate the
effects of spilling LNG onto the outer containment of a
storage tank. The material tested was a grade 43A structural
steel, which is approximately equivalent in strength to a
Grade L290 (X42) pipeline steel. The reported 30-J Charpy
impact temperature was -15°C; as the material thickness
was 6.3mm, this would have passed the typical UK gas
transmission pipeline requirement of 27J at 0°C in a 2/3
specimen.
Local cooling to around 80°C below the bulk metal
temperature was required to initiate cracks, which then
arrested. It was argued that the fracture was initiated by the
additional thermal stress generated by the local cooling and
the applied remote stress level was not a factor. It should
be noted, however, that the stresses used in this study were
relatively low by the standards of many pipelines.
The experimental study on a Grade L555 vessel [5]
discussed above shows that fracture initiation will not
4th Quarter, 2010
281
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Fig.3. Predicted variation of the
transition temperature in a brittle
linepipe for blunt corrosion defects
as a function of the elastic stress
concentration factor.Trend of results
from [12].
Fig.4. British Gas relation between
temperature and stress level for brittle
crack arrest, showing operating point
for case study.
necessarily occur at a sharp crack, even when the material
is subject to significant local cooling. The fatigue crack
in the vessel did not propagate and remained stable
throughout the experiment. This was a modern hightoughness material, with an upper-shelf Charpy energy
exceeding 400J at -10°C, so it is possible that it was not
on the lower shelf at –70°C.
An arrested crack with a larger opening would be closer
to a hole than a crack. If, as the work considered above
suggests, there are differences in the fluid flow and heat
transfer of holes and cracks, a larger defect would behave in
a manner similar to a hole and there would be less cooling.
This would reduce the likelihood of subsequent re-initiation
after an arrest.
Immediate arrest after initiation
The study by Fearnehough [14] shows that cracks initiated by
local cooling will arrest when they run into warmer material
with a higher toughness. The amount of crack extension, or
equivalently the position of the crack arrest, was considered
to be a function of both stress level and temperature, as the
propagation distance was greater in the specimens subjected
to higher remote tension. This observation that arrest was
a function of stress and temperature is consistent with
pipeline experience with propagating brittle fractures [15].
This section considers crack arrest after initiation from a
locally cooled leak. This is important, as the postulated
failure mechanism described above does not clearly
allow for this. Whilst the enlargement of a small leak
to a bigger leak, or a full-bore rupture, is not desirable,
such an escalation is preferable to the formation of a
propagating fracture.
282
The Journal of Pipeline Engineering
The West Jefferson test was developed for brittle crack
arrest tests on pipeline steels and is an example of brittle
crack arrest occurring under a constant applied stress. In
this test a through-wall crack is initiated in a pressurized
vessel (so the stress ahead of the crack is constant at the
hoop stress) and propagates along the length of the pipe.
The fracture appearance and the temperature at the point
of crack arrest correlate well with full-scale test results and
the results of the Battelle drop-weight tear test (DWTT).
The West Jefferson methodology has been used for many
years and is not dependent on the pipe contents – in fact
typically the tests fill the pipe section with 95% of water
and pressurize the remaining space with air or nitrogen to
reduce the stored energy in the vessel.
Fracture propagation and arrest
It is possible that cooling of the remaining contents
during blowdown following a full-bore rupture could
lower the temperature of the pipe wall below the DWTT
transition temperature. However, this cooling would also
be accompanied by a reduction in pressure which would
reduce the hoop stress. Again, the relationship between
arrest temperature and hoop stress shown in Fig.4 could be
used if the pressure–temperature trajectory is known from
simulations of the depressurization.
An alternative approach to predicting brittle fracture arrest in
a blowdown would be that proposed by Battelle [17], which
relates the arrest behaviour to the Charpy transition curve.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
The phenomena of both brittle and ductile fracture
propagation in pipelines are well understood and can be
analysed by accepted methods. The two types of propagating
fracture are considered separately below. It should also be
noted that propagating fractures can originate from pipeline
damage, often due to mechanical interference or corrosion,
that does not involve the postulated cooling and fracture
mechanism. Pipeline designers already design against
propagating fractures for this reason. It may be necessary to
modify the design methods to take account of the specific
decompression behaviour of CO2, but these changes will
be required for a CO2 pipeline in any case.
Thus it is considered that, provided a pipeline is operated
above its DWTT transition temperature, a propagating brittle
fracture should not escalate from a leak, even when operating
at stress levels up to 72% SMYS. If an old pipeline has a
high DWTT transition temperature, it could be operated
at lower stress levels using the relationship between arrest
temperature and hoop stress developed by Fearnehough
[15] and included in Edition 2 of IGE/TD/1 [16]. This
relation is shown in Fig.4.
Propagating brittle fractures
These fractures are characterized by little plastic deformation
and are driven by the elastic stress in the pipe wall. The
fractures travel at axial velocities above the acoustic velocity in
the fluid. As a result, the crack tip is continually propagating
into a region where the pipeline contents are undisturbed.
These cracks cannot be affected by rapid cooling of escaping
fluid as they are moving faster than the fluid can depressurize
and cool. The steel ahead of the crack will remain essentially
at the ambient temperature prior to a leak occurring. Thus a
propagating brittle crack in a CO2 pipeline will behave in the
same manner as a propagating brittle crack in a natural gas
pipeline. The standard fracture-control approach of ensuring
that the pipe steel is operating above its DWTT transition
temperature will prevent propagating brittle fractures.
A DWTT requirement has been routinely specified for gas
transmission pipelines in the UK since the introduction
of Gas Council specifications for steel linepipe in the late
1960s, and most existing UK gas transmission pipelines
will satisfy this requirement. Pipelines pre-dating this with
a high DWTT transition temperature would require special
consideration, but the approach would be the same as has
been used for natural gas pipelines since the publication of
Edition 2 of IGE/TD/1 [16]. For new build, any competent
pipe supplier should be able to produce pipe meeting the
usual DWTT requirement of 85% shear area for design
temperatures around 0°C.
Propagating ductile fractures
The phenomenon of propagating ductile fractures in gas
pipelines has been known for over 30 years, and has been
extensively studied. Rothwell gives a good overview of these
studies [18]; most recent work has concentrated on testing
very high strength linepipe with yield strength over 690N/
mm2 [19] and rich gas mixtures [20]. Propagating ductile
fractures typically run at axial velocities in the range 200300m/s, although in some full-scale tests velocities down to
100m/s have been measured. These velocities are less than
the acoustic velocity in the fluid, and the crack is driven
by the pressure of the escaping fluid acting on the ‘flaps’
developing behind the crack tip as the pipe cracks. At these
velocities there will be negligible heat transfer to or from
the crack tip area, and so the cooling effects described in
Section 2 will not occur.
The steps required to control propagating ductile fractures
in a gas pipeline are also well understood, so that formal
fracture control plans are now explicitly required in some
pipeline design codes such as the Australian code AS 2885
[21]. UK codes such as TD/1 [22] and PD 8010 [23] do
not have such an equivalent explicit requirement, but the
material toughness requirements in the codes will achieve
a high level of resistance to propagating fractures.
The special requirements for CO2 pipelines discussed in [1]
are well known in the pipeline industry. Maxey identified
them in [24] and by 1990 crack arrestors had been installed
on the Canyon Reef Carriers CO2 [25] pipeline in west Texas
because of concerns over ductile crack propagation. Whilst
some of the effects of rapid decompression of dense-phase
CO2 are counter-intuitive, they can be predicted [26]. The
4th Quarter, 2010
283
effects of impurities in the gas have also been considered
[27], but again the fundamental approach to fracture control
remains unchanged. The greatest uncertainty appears to be
in predicting the decompression behaviour of the contents.
If necessary, experimental confirmation of predictions by
shock tube testing may be required.
It is accepted that the control of propagating ductile fractures
in a CO2 pipeline may be more difficult than in a typical UK
onshore natural gas pipeline, particularly for a dense-phase
pipeline. However, the methodology for control is known.
For a new build, these problems should be manageable
with correct specification of the material properties. If an
existing pipeline is being converted to transmit CO2, then
an assessment of the toughness will be required to show
that propagating fractures can be controlled.
General discussion
The available work suggests there may be differences
between the cooling experienced at a hole and that at a
‘tight’ fatigue crack. If [4] is correct and the flow through
a hole is choked so that the bulk of the expansion and
cooling takes place outside the pipe wall, there will be little
cooling at a leak through a hole. If this is the case, the
likelihood of the postulated fracture mechanism occurring
in practice reduces.
In contrast, the test with a fatigue crack reported in [5]
suggests that expansion through a tight crack may generate
a substantial temperature drop in the material. The work
discussed in above also suggests that there are differences
in flow behaviour for tight cracks which would affect heat
transfer. This difference requires further investigation.
Fracture initiation
The available evidence shows that if a fracture should initiate
from a locally cooled crack, it will arrest as it extends into
warmer material. Fearnehough’s tests [14] conclusively
demonstrated this effect, and crack arrest is included in the
R6 procedure [10]. Generally it is assumed that crack arrest
can occur when either the growing crack has a decreasing
stress intensity factor or when the crack grows into an area
of increasing fracture toughness. In the present case, the
fracture toughness will increase as the crack is growing into
warmer material. A longer crack would have an increased
stress-intensity factor simply due to the greater crack length,
but would have grown away from the high stresses associated
with the locally cooled area.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Heat transfer
It was noted above that there were more-realistic fractureassessment methods than the simple LEFM approach used
in [2]. A further consideration is the effect of local constraint.
It is now widely accepted that pipelines are a ‘low-constraint’
structure, as they are thin and defects are loaded predominantly
in tension. This has the effect of increasing the effective
toughness above that measured in a standard fracture-toughness
test. Constraint-based arguments have been used to assess the
behaviour of longitudinal seam-weld defects [29] and pipeline
girth welds [30]. These methods can be used to assess the
behaviour of through-wall cracks in a pipeline subject to local
cooling. Methods of including constraint effects in fracture
assessments are given in Section III.7 of R6 Rev 4 [10].
The available evidence for fracture initiation due to local
cooling is conflicting. Fearnehough’s tests [14] using liquid
nitrogen vapour showed that fracture can be initiated at a
locally cooled crack, and there are other cases known of
failures occurring when components are cooled under load:
for example the failure of a heat exchanger in the Longford
incident [28] was attributed to a combination of thermal
shock and local embrittlement of cold material. However,
the test in [5] showed that fracture will not necessarily occur
under these conditions, and tests on simulated corrosion
defects discussed above have shown that the defect acuity
appears to be a major factor.
A combination of a substantial temperature drop, a sharp
crack, and a material with a high transition temperature
is required to initiate a fracture due to local cooling of
a leaking pipeline. Quantifying these effects will require
understanding of the temperatures generated at a leak,
together with fracture analyses.
It is possible that a crack may grow due to the postulated
local embrittlement mechanism and reach a length exceeding
the stable through-wall crack length for the pipeline. This
would cause the leak to escalate to a rupture. The critical
through-wall crack length can be calculated using the
standard methods for an axial crack in a pipeline.
Fracture propagation
As noted above, it is considered that accepted techniques can
be used to assess fracture propagation in CO2 pipelines, and
indeed these methods will be required to control propagating
fractures originating from other forms of damage such as
external interference. It is understood that work is being
carried out elsewhere to address issues specific to the control
of propagating fractures in CO2 pipelines, in particular the
rapid decompression behaviour of CO2.
Warm pre-stressing
Warm pre-stressing is a phenomenon where the low
temperature fracture toughness of a material is improved by
a prior loading at a higher temperature. Various explanations
exist for the effect, such as crack-tip blunting at the higher
temperature, the generation of beneficial residual stresses at
a crack tip by the prior loading, or metallurgical effects such
as the de-cohesion of brittle second phase particles during
the high temperature loading. Guidance on the application
of warm pre-stressing arguments is given in Annex O of BS
7910 [11] and Section III.10 of R6 Rev 4 [10].
Rio
Pipeline
FP1
2011
Conference & Exposition
September 20-22
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
11/10
Rio de Janeiro • Brazil
Call for Papers:
Submission Deadline
December 17, 2010
Participation
Information:
Phone.: (+55 21) 2112-9000
Fax: (+55 21) 2220-1596
e-mail: [email protected]
www.riopipeline.com.br
Organization / Realization
4th Quarter, 2010
285
For a defect created in an operating pipeline, without
unloading and re-loading, crack tip residual-stress effects
cannot be significant. Thus the operating mechanisms are
likely to be crack-tip blunting and, possibly, metallurgical
effects. There is ample evidence that brittle and transition
fracture toughnesses are strongly influenced by notch acuity:
all the fracture-toughness testing standards require a fatiguesharpened crack, because this gives lower-bound results.
Limits are set on the maximum stress intensity factor Kmax
when fatigue pre-cracking fracture-toughness specimens to
ensure the crack is not artificially blunted before the main
fracture test. The stress acting on a crack before cooling
can be considered to be analogous to a fatigue pre-cracking
cycle with a high Kmax, leading to crack-tip blunting and
an elevation of the fracture toughness.
It can be argued that the initial load due to internal pressure
before the onset of any cooling at a leak is the warm pre-stress,
and the additional local thermal stress generated by cooling
is an additional load, giving a ‘load – cool – fracture’ case.
The applied stress-intensity factor will be increasing during
the cooling due to thermal stresses adding to the driving
force due to pressure loading (which will remain constant if
the leak is small and undetected). As a result the simplified
warm pre-stressing argument, that failure is avoided if the
stress-intensity factor is monotonically falling during cooling,
will not hold, and a more-detailed analysis is required.
Case study
Crack initiation
To assess the possibility of fracture occurring at a sharp crack
under these conditions, the fracture resistance of the material
at -17°C must be estimated. Unless test data exist, it is only
possible to estimate the effects of cooling from the linepipe
specification acceptance criteria. Such an approach should
be conservative, as the actual material properties should
be better than the specification minima. LX/1 compliant
material would have achieved 27J Charpy energy at 0°C
in a 2/3 size specimen, and would have 75% shear area in
the DWTT at 0°C. Unfortunately this information cannot
be used directly to provide the toughness at a different
temperature. The approach developed by Battelle for brittle
fracture arrest predictions [17] can be used, as follows.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
In the authors’ opinion it is not appropriate to claim credit
for the commissioning hydrotest as a prior warm load if
warm pre-stressing arguments are used. This is because most
defects will have been introduced after the hydrotest; for
example mechanical damage will have produced a gouge
or dent-gouge defect at some time after commissioning.
Hence the hydrotest will have stressed defect-free material,
and there can have been no effect of the hydrotest on a
crack-tip region which did not exist at the time of the test.
of 0.56°C per bar, the maximum cooling associated with a
leak would be 21°C. If the operating temperature is +4°C,
the metal temperature could reach as low as -17°C at a leak,
if perfect heat transfer occurred and there is no heat recovery
from the bulk contents or surrounding soil.
This section applies the arguments developed in this paper
to a case where an existing pipeline is being re-used to
transmit CO2. Such re-use is environmentally beneficial as
it avoids the carbon costs associated with the construction
of a new pipeline.
Figure 6 of [17] can be used to estimate that the Charpy
transition temperature in a 2/3 size specimen would be 18°F
(10°C) below the DWTT transition temperature, which is
known to be 0°C or lower for LX/1 material. Hence the
estimated Charpy transition temperature is -10°C, and so
when fully cooled, the material is operating at 7°C below
the transition temperature. Using Fig.5 of the Battelle
report, the shear area in a 2/3 size Charpy specimen tested
at 7°C below the transition temperature is 72%. At this
shear area Fig.7 then shows that the impact energy is 75%
of the maximum upper-shelf energy. Thus the minimum
Charpy toughness of the material at a fully cooled leak
would be predicted to be 0.75 x 27 = 20J. In practice, the
material upper-shelf energy would be above the specification
minimum, and it is likely that the transition temperature
would be lower than that estimated, so the 20-J estimate
is conservative. The effect of reducing the impact energy
from 27J to 20J is to reduce the critical through-wall crack
length calculated using the NG-18 toughness-dependent
model [31] by about 10%, from 350mm to 320mm.
Given the conservative assumptions in the analysis, and
the likely benefits of warm pre-stressing, it is judged that
this difference is not significant and that brittle fracture
initiation is unlikely to occur for this specific case.
Basic data
The pipeline is an existing gas transmission pipeline being
reused to transmit gas phase CO2 at 38bar. This pipeline
is a 36-in (914mm) nominal diameter line with a nominal
wall thickness of 12.7mm, and was constructed from pipe
complying with the former British Gas LX/1 specification.
Thus the material would be expected to have a DWTT
transition temperature of 0°C. The hoop stress under the
proposed operating conditions is 137N/mm2. For an X60
material this is equivalent to 33% SMYS based on the
nominal wall thickness. Assuming Joule-Thomson cooling
The change in toughness predicted by the analysis in the
previous paragraph is a function of the temperature change
and, to a limited extent, the wall thickness. The pipe wall
thickness affects the shift between the Charpy transition
temperature and the assumed DWTT transition temperature
of 0°C for LX/1 compliant material. Smaller shifts will occur
for thinner pipe; other cases will require specific evaluations.
For modern linepipe, the degree of cooling estimated in
this case would not be expected to be a problem, as the
material would be expected to still be on the upper shelf at
temperatures around -20°C.
286
The Journal of Pipeline Engineering
It should also be noted that in Fearnehough’s tests on steel
plates [14] the highest stress was 124N/mm2, comparable
to the nominal stress of 137N/mm2 in this case. A cooling
of over 100°C was required to initiate fracture in this test.
The cracks in these tests were 600mm long, greater than
the tolerable through-wall crack length for this pipeline.
Crack propagation
Using the British Gas relation between hoop stress
and temperature shift below the DWTT transition
temperature, shown in Fig.4, it can be seen that the
operating point at a hoop stress of 33% SMYS and a
temperature shift of 17°C below the DWTT transition
temperature is just above the 95% probability of arrest
line. Thus it is considered that even if a long length
of pipeline were cooled to a low metal temperature, a
propagating brittle fracture is not a credible event under
the proposed operating conditions.
The issues regarding a postulated failure mechanism for
leaking CO2 pipelines have been reviewed. The main
conclusions are:
• It is considered possible that under a restricted range
of circumstances a leak in a CO2 pipeline may enlarge
due to local cooling causing a brittle fracture. Such
enlargement is only likely to occur for a sharp crack
in a material with a high ductile–brittle transition
temperature in a pipeline operating at a high pressure.
• The main uncertainty in assessing if such enlargement
could occur is in the understanding of fluid flow,
heat transfer, and cooling associated with the leak.
This will be strongly influenced by the leak geometry.
• Industry-standard fracture-assessment methods
can be used to determine if fracture will occur at
a leak. These methods can take account of elastic–
plastic fracture and constraint effects to eliminate
unnecessary conservatism.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Further work
Concluding remarks
The greatest unknown in assessing whether it is possible for
a leak to escalate to a rupture due to the postulated local
cooling mechanism is the amount of cooling generated at the
leak. Once the cooling is quantified, given knowledge of the
material’s toughness transition curve, fracture-mechanics’
methods can be used to assess if brittle fracture will occur.
The published work suggests that the amount of cooling
is strongly influenced by the geometry of the leak. Work is
required to address this issue.
It is suggested that the heat transfer and local cooling issue
will require both numerical and experimental work. Whilst
numerical studies can give an indication of trends and the
important parameters, it is considered that experimental
verification is required.
An initial study using computational-fluid-dynamics’
(CFD) methods could be used to estimate leakage rates
and heat-transfer coefficients; this would be combined with
finite-element heat-transfer calculations for the steel pipe
wall and stress analysis to determine the thermal stresses.
The analyses should consider a range of hole sizes and
crack geometries, up to a limiting length of the critical
through-wall crack length for the geometry and material.
Fracture analyses would be carried out using the results
from this study.
The results of the numerical work would be used to design
an experimental programme. This is likely to require tests
on a tough material, where crack growth is not expected,
and a low-toughness material with a high ductile–brittle
transition temperature where brittle crack extension
is predicted. The tests should also include a range of
defect acuity, from a round hole to a tight fatigue crack.
Instrumentation of the tests should include temperature,
strain, and crack growth.
• The beneficial effects of warm pre-stressing should
be quantified and included in future assessments.
• For the specific case study of an existing pipeline
being re-used to carry gas phase CO2 at 38bar, it is
considered that crack extension by the postulated
cooling and embrittlement mechanism will not occur.
Acknowledgements
This study was funded by National Grid Gas Transmission,
which is gratefully acknowledged.
References
1. M.Bilio, S.Brown, M.Fairweather, and H.Mahgerefteh,
20009. CO2 pipelines material and safety considerations. In:
Hazards XXI - Process Safety and Environmental Protection
in a Changing World. IChemE, Manchester, pp423-429.
2. H.Mahgerefteh and O.Atti, 2006. Modelling low-temperatureinduced failure of pressurized pipelines. AIChE Journal, 52, 3,
1248-1256.
3. Ibid., 2006. An analysis of the gas pipeline explosion at
Ghislenghien, Belgium. American Institute of Chemical
Engineers, Orlando, FL, USA.
4. G.Saville, S.M.Richardson, and P.Barker, 2004. Leakage in
ethylene pipelines. Process Safety and Environmental Protection,
82, 1, 61-68.
5. O.Reepmeyer, P.Lothe, S.Valsgard, M.Erdelen-Peppler, and
G.Knauf, 2006. Full scale gas leak test at a large diameter X-80
DSAW pipe. Paper IPC06-10005, vol. 3, Part A, American
Society of Mechanical Engineers, Calgary, AB, Canada, pp1-8.
6. N.M.Bagshaw, S.B.M.Beck, and J.R.Yates, 2000. Identification
of fluid flow regimes in narrow cracks. Proc. Institution of
4th Quarter, 2010
287
20. D.M.Johnson, N.Horner, L.Carlson, and R.J.Eiber, 2000. Full
scale validation of the fracture control of a pipeline designed to
transport rich natural gas. In: Proc 3rd Int. Pipeline Technology
Conference, I, Bruges, Belgium, Elsevier Scientific, Amsterdam,
pp191-209.
21. Standards Australia, 2007. Pipelines - gas and liquid petroleum,
Part 1: design and construction. AS 2885.1: Standards
Australia, Homebush, New South Wales.
22. IGE, 2008. Steel pipelines and associated installations for high
pressure gas transmission. IGE/TD/1 Edition 5. Institution
of Gas Engineers and Managers, Loughborough.
23. BSI, 2004. Code of practice for pipelines - Part 1: Steel pipelines
on land PD 8010-1. British Standards Institution, London.
24. W.A.Maxey 1986. Long shear fractures in CO2 lines controlled
by regulating saturation, arrest pressures. Oil and Gas Journal,
84, 31, 44-46.
25. D.L.Marsili and G.R.Stevick, 1990. Reducing the risk of
ductile fracture on the Canyon Reef Carriers CO2 pipeline.
In: Society of Petroleum Engineers Annual Conference, New
Orleans, Vol. 3. Society of Petroleum Engineers, Richardson,
Texas, pp311-319. 26. A.Cosham, 2009. “It’s a gas, Jim, but not
as we know it”. Pipeline Technology 2009. R.Denys, editor,
Ostend, Belgium, Scientific Surveys Ltd, Beaconsfield, UK.
27. P.N.Seevam, J.M.Race, M.J.Downie, and P.Hopkins, 2008.
Transporting the next generation of CO2 for carbon capture
and storage: the impact of impurities on supercritical CO2
pipelines. IPC2008-64063. Int. Pipeline Conference Calgary.
ASME, New York, pp1-13.
28. D.M.Dawson and B.J.Brooks, 1999. The Esso Longford gas
plant accident report of the Longford Royal Commission.
Melbourne. Government Printer for the State of Victoria.
29. R.M.Andrews, G.C.Morgan, and W.J.Beattie, 2004. The
significance of low toughness areas in the seam weld of linepipe.
IPC04-0422. Proc. International Pipeline Conference. ASME,
New York, pp1-10.
30. Y.Y.Wang and D.J.Horsley, 2003. Tensile strain limits of
pipelines. Paper 35. 14th Joint Technical Meeting on Linepipe
Research, Berlin, May. European Pipeline Research Group,
Duisburg, Germany, pp1-14.
31. J.F.Kiefner, W.A.Maxey, R.J.Eiber, and A.R.Duffy, 1973.
Failure stress levels of flaws in pressurized cylinders. Progress
in flaw growth and fracture toughness testing ASTM STP 536.
American Society for Testing and Materials, Philadelphia,
pp461-481.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Mechanical Engineers, Part C. Journal of Mechanical Engineering
Science, 214, 8, 1099-1106.
7. S.B.M.Beck, N.M.Bagshaw, and J.R.Yates, 2005. Explicit
equations for leak rates through narrow cracks. Int.J. Pressure
Vessels and Piping, 82, 7, 565-570.
8. L.V.Clarke, H.Bainbridge, S.B.M.Beck, and J.R.Yates, 1997.
Measurement of fluid flow rates through cracks. Idem, 71, 1,
71-75.
9. J.P.Taggart and P.J.Budden, 2008. Leak before break: studies in
support of new R6 guidance on leak rate evaluation. J.Pressure
Vessel Technology, Transactions of ASME. 130, 1, 01140210114026.
10. Anon., 2001. Assessment of the integrity of structures
containing defects. R6 Revision 4. Barnwood. British Energy
Generation.
11. BSI, 2005. Guide to methods for assessing the acceptability
of flaws in metallic structures BS 7910:2005 incorporating
Amendment 1, September 2007. British Standards Institution,
London.
12. R.M.Andrews, M. Martin, and V.Chauhan, 2006. Assessment
of corrosion defects in old low toughness pipelines. IPC0610140. Proc.International Pipeline Conference, ASME, New
York, pp1-14.
13. G.Wilkowski, D.Rudland, D.Rider, P.Mincer, and
W.Sloterdijk, 2006. When old line pipes initiate fracture in
a ductile manner. IPC06-10326. Proc. International Pipeline
Conference. Calgary, Canada September, ASME, New York,
pp1-12.
14. G.D.Fearnehough, 1985. Progressive crack extension due to
local cooling of a crack in LNG storage tank material. Int.J.
Pressure Vessels and Piping, 19, 283-292.
15. G.D.Fearnehough, D.W.Jude, and R.T.Weiner, 1971. The
arrest of brittle fracture in pipelines. In: Practical application
of fracture mechanics to pressure vessel technology. Institution
of Mechanical Engineers, London, pp156-162.
16. IGE, 1984. Recommendations on transmission and
distribution practice: steel pipelines for high pressure gat
Transmission. IGE/TD/1 Complete Edition 2. Institution
of Gas Engineers, London.
17. W.A.Maxey, J.F.Kiefner, and R.J.Eiber, 1983. Brittle fracture
arrest in gas pipelines. NG-18 Report 135 PRCI Catalogue
L51436. In: PRCI Report, Washington. Pipeline Research
Committee.
18. A.B.Rothwell, 2007. Evolution and current status of approaches
to fracture control design for gas pipelines. Paper 15. In: 16th
Joint Technical Meeting on Linepipe Research, Canberra,
April. Australian Pipeline Industry Association, pp1-15.
19. R.M.Andrews, N.A.Millwood, A.D.Batte, and B.J.Lowesmith,
2004. The fracture arrest behaviour of 914mm diameter X100
grade steel linepipes. Paper 0596. Proc. International Pipeline
Conference, ASME, New York, pp1-9.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
PIPE
The global organization
for oil and gas
pipeline engineers
• Recognizing your skills and status
• Promoting the highest engineering standards
• Providing a professional network
SIGN UP TODAY!
www.pipeinst.org
4th Quarter, 2010
289
Greenhouse gas emissions from
electricity generating CCS
upstream and downstream
transport processes
by Dr Tim Cockerill*1, Dr Naser Odeh2, and Scott Laczay1
1 ICEPT, Imperial College London, UK
2 AEA PLC (formerly at University of Reading), UK
H
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
EADLINE FIGURES suggest CCS technology will capture 90% or more of the CO2 produced by a
power plant. While this may be true at the stack, on a full lifecycle basis the ‘greenhouse gas’ (GHG)
savings offered are more modest thanks to significant resource consumption in upstream and downstream
processes. Our analysis suggests that lifecycle GHG emissions can be reduced to approximately 170gCO2/
kWh for an integrated gasification combined cycle (IGCC) plant with 90% capture efficiency. This still
represents around an 80% saving compared to conventional coal plant, but is considerably higher than the
better-performing renewables such as wind that produces only 10-30gCO2/kWh in good locations
This paper examines the origin and importance of upstream and downstream CCS GHG emissions, in
particular identifying those associated with transport processes. Sensitivity studies investigate which major
characteristics of a CCS system are likely to have an important impact on transport GHG emissions. The
scope for combining biofuels with CCS in order to improve lifecycle performance is considered. In principle
BioCCS could produce a system with overall negative atmospheric GHG emissions. However that potential
is constrained by emissions arising from the production and transportation of biofuels.
Finally some general conclusions for design approaches for CCS systems aimed at minimizing system GHG
emissions are drawn. Some key areas of uncertainty are also identified for further work.
H
EADLINE FIGURES suggest CCS technology will
capture 90% or more of the CO2 produced by a power
plant. While this may be true at the stack, on a full lifecycle
basis the GHG savings offered are likely to be more modest
thanks to significant resource consumption in upstream and
downstream processes. This paper summarizes results from a
series of lifecycle analysis investigations of hypothetical fossilfuel-based electricity-generating CCS plant, emphasizing
the role, albeit relatively small, that transport systems play
in contributing to the overall emissions. The discussion
encompasses both downstream transport systems (i.e. for
carbon dioxide) and upstream systems (i.e. for fuel and
consumable materials). Much of this discussion draws on
This paper was presented at the First International Forum on Transportation of CO2
by Pipeline, organized in Newcastle upon Tyne in July, 2010, by Tiratsoo Technical
and Clarion Technical Conferences, and with the support of the University of
Newcastle and the Carbon Capture and Storage Association.
*Author’s contact details
email: [email protected]
analysis substantially reported in two publications by two
of the authors [1], [2].
The performance of the fossil-fuel-based CCS systems is
compared to that of several electricity generating renewable
energy technologies. Subsequently the potential offered
by combining biomass and CCS technologies, with the
ultimate objective of producing net carbon dioxide capture
from the atmosphere, is examined. The results from this
section of the paper are necessarily indicative as there are
many uncertainties about the GHG impacts of biomass
production and combustion, let alone the complexities
introduced by combining biomass combustion with carboncapture technologies. In consequence, considerable care
must be exercised in quantitatively comparing the various
results presented throughout the paper. Nevertheless, the
qualitative trends are clear enough to allow some useful
conclusions to be drawn.
FP2
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
The new online information service
that unlocks the secrets of the global
pipeline industry
Pipelines International Premium is the international
oil and gas pipeline industry’s foremost in-depth
source of information, comprising a digest of
high-quality papers covering the latest technology
and reviews of the pipeline industry worldwide, and
a comprehensive project database.
It is comprised of:
Pipelines International Digest which provides a monthly update of papers covering all areas of the
industry – from key projects, and engineering and
construction issues, to environmental, regulatory,
legal and financial issues.
Pipelines International Projects which allows subscribers
to access a searchable database of completed and
current projects.
Subscribe or get a free 14 day trial now at
www.pipelinesinternational.com/premium
291
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
4th Quarter, 2010
Fig.1. Summary of system boundaries for lifecycle analysis.
Overview of the lifecycle analysis
approach
Lifecycle analysis (LCA) is a standardized method for
evaluating the environmental impacts of a given process or
different competing processes. Greenhouse gas emissions,
other air and water emissions, resource consumption,
and energy use are evaluated using energy and material
balances. The evaluation procedure covers all sub-processes
within the lifecycle of the system, starting from raw material
production and ending with product and waste disposal. By
evaluating the environmental impacts of different systems,
recommendations can be made to reduce possible effects.
The work reported here makes use of LCA in studying the
impacts (with emphasis on GHG emissions) of fossil-fuel
power generation with and without CCS. The main objective
is to evaluate the actual reduction in GHG emissions that
can be realized by CCS in various configurations.
Broadly speaking, each of the systems studied consists of
fuel production, its transportation to the power plant,
power plant construction, power plant operation, and
any processes related to power/capture plant operation.
For all CCS technologies, the analysis also includes the
capture plant construction and operation in addition to
CO2 transport and storage.
System specification
System boundaries
Figure 1 summarises the extent of the system considered
by the lifecycle analysis results presented here, with further
detail of individual elements in Table 1. The key element
of course is the power plant itself, and our calculations
include direct emissions from combustion, plant internal
energy consumption, energy used in operating an
monoethanolamine-based capture system, and energy for
plant maintenance activities. Note that the impacts of
electricity transmission beyond the power station are not
included, and thus the results quoted here are for electricity
produced rather than delivered.
Upstream process direct emissions include those arising from
fuel production activities, encompassing the energy used to
operate the machinery and transport systems required. Also
included within upstream processes is the production of
other consumable materials such as limestone, ammonia, and
monoethanolamine (MEA). For gas-cycle systems, leakage
from pipelines is accounted for. Indirect upstream emissions
take account of equipment manufacturing, recognizing that
production facilities are not dedicated to servicing the power
plant and the associated emissions should be shared across
292
The Journal of Pipeline Engineering
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Fig.2. Breakdown of lifecycle GHG
atmospheric emissions from power
plant technologies.
Category
Coal-based systems
Natural gas-based systems
• Power plant construction
• Capture plant construction
• CO2 transport pipeline
•
•
•
•
Fuel combustion
Direct CO2 emissions
Direct CO2 emissions
Fuel production
Mining (Equipment manufacture, Mining Gas extraction (platform construction,
activities, Methane leakage, Coal cleaning, gas sweetening and flaring, methane
Land recovery for surface mining)
leakage)
Other material production
•
•
•
•
•
Limestone/ammonia production
SCR catalyst production
Water treatment and distribution
MEA production
NaOH/activated C production
•
•
•
•
•
•
•
•
•
Boiler/ESP/Gasifier ash
FGD waste
SCR catalyst waste
MEA re-claimer waste
• SCR catalyst waste
• MEA re-claimer waste
Construction
Waste disposal
Transport
• Coal transport
• Local by rail
• International by ship
• Limestone transport by truck
• Chemical transport by rail
• Waste transport by truck
• CO2 compression and injection
Power plant
Gas pipeline
Capture plant
CO2 transport pipeline
Ammonia production
SCR catalyst production
Water treatment and distribution
MEA production
NaOH/activated C production
• Gas transport
• gas compression
• onshore processing
• Methane leakage
• Chemical transport by rail
• Waste transport by truck
• CO2 compression and injection
Table 1. Details of the sub-processes included within the lifecycle analysis described in the paper.
4th Quarter, 2010
293
all uses. Downstream processes consider waste transport and
disposal in relatively nearby locations. As with the upstream
analysis where non-dedicated facilities are used, emissions
are attributed appropriately across all uses.
As the figure shows, emissions arising from the construction
phases across all the supply chain elements are included in
the calculations, taking account of :
• materials
• material production processes
• material transportation by truck over an average
distance of 50km
• on-site energy consumption, comprising 80% diesel
and 20% electricity taken from the UK grid.
Power plant decommissioning is accounted for, but
decommissioning of upstream and downstream equipment
is not included.
Overall results
A breakdown of the results is shown in Fig.2 and Table
7. Unsurprisingly, the supercritical plant without CCS
produces by far the largest atmospheric GHG emissions
per unit of electricity output, dominated by the direct
emissions from combustion. All the CCS-fitted plant
produce considerably less atmospheric emissions, with the
coal-fired IGGC plant generating the smallest quantity of
emissions per unit (kWh) of electricity produced.
Emissions associated with transport processes
Downstream
In all the CCS cases, emissions associated with CO2 transport
are very small, representing between 1% and 1.8% of the
total per unit of electricity produced. This of course only
applies for the system specification set out above. The
component of these emissions arising from powering any
recompression stations could change if energy sources with
differing carbon intensities were employed: the present study
has assumed they are powered by electricity generated with
the UK grid average carbon intensity.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
External factors
As well as the system boundary, another important influence
on the assessment is the location of the plant and the
resources consumed in construction and operation. The
analysis below has been developed on the basis that the
plant is located in North-East England, and that fuel and
other materials are sourced relatively locally so far as possible;
Table 2 provides more comprehensive details.
Life cycle emissions for fossilfuelled CCS plant
Power plant types
of only 37.5% are achieved due to commissioning and
decommissioning activities respectively. Further outline
details of the power plant are contained in Tables 3-6,
but for full information reference should be made to the
authors’ previous work [1, 2].
With longer or less-secure pipelines, the impact of CO2
transport could increase considerably, as investigated below.
Other downstream processes include ash and other solid
waste disposal, which also make only a very small contribution
to overall emissions.
Four types of fossil-fuel plant are considered here:
Upstream
• a supercritical pulverized (SuperPC) coal-fired plant
with selective catalytic reduction (SCR), electrostatic
precipitation (ESP) and flue-gas desulphurization
(FGD) pipe-end clean-up technologies;
• a similar supercritical coal-fired plant, but fitted
additionally with an MEA-based CO2 capture unit
having 90% CO2 capture efficiency;
• a natural gas fired combined-cycle (NGCC) plant
fitted with similar MEA-based capture unit;
• a coal-fired integrated gasification combined-cycle
(IGCC) plant fitted with Selexol-based carbon
dioxide capture, again having 90% capture efficiency.
The calculations assume that the plant has a rated capacity
of 500MWe, an operating lifetime of 30 years, and take
three years to construct. Load factors are taken as 75%
except in the first and last year of operation, where factors
Upstream emissions are included within the ‘operation’
components of Fig.2, which also includes the very small
downstream operational emissions associates with ash and
waste disposal. The upstream calculations assume coal is
produced from a nearby UK deep mine and subsequently
transported to the power plant by rail. This represents
something of an idealized best case, as the limited number
of UK mines means transport will in general be over longer
distances. Limestone is also UK sourced and transported by
truck, with other consumables such as solvents transported
by rail. Natural gas is assumed to be sourced from the UK
North Sea, carried via pipelines with the specification set
out in Table 10.
For the base coal-CCS configurations considered here,
in general just less than 50% of upstream emissions arise
from mining, with a similar contribution coming from the
production and transport of all other consumables. Coal
transport accounts for approximately 1.5% of upstream
294
The Journal of Pipeline Engineering
Category
Coal-based power plants
Gas-based power plants
Power Plant Location
Teesside
Mine location
Surface mine: Maiden’s Hall Extension, Northumberland
Deep mine: Killingley Colliery,
North Yorkshire
–
Limestone Quarry
North Yorkshire: 50 km from power
plant
–
Ammonia production
Billingham, Durham, 20 km from power plant
Concrete manufacturer
Leeds, 100 km from power plant
Steel manufacturers
Teesside
Gas field
–
Southern North Sea
On-shore gas processing
–
Hartlepool
Gas pipeline
–
Offshore: 100 km, on-shore: 50 km
Teesside within 50 km from power
plant
0
CO2 pipeline
50 km on-shore, 150 km offshore
0.039
Bunter Sandstone-Southern North
Sea, Closure
0
CO2 storage
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
CO2 on-shore collection point
Table 2. Location of power plant and key material inputs.
emissions, though as will be seen later, this low value is
largely a reflection of very optimistic assumptions.
For the natural gas CCS system considered, the majority of
the calculated upstream GHG emissions arise because of
escapes in the natural gas supply system, though it should
be noted that we have assumed comparatively high leakage
rate of 1%. Smaller contributions are distributed over the
production and transport of other consumables.
Across all the results for CCS plant, it is clear that the
operational GHG emissions are much large than those from
CO2 transport. Indeed, under the assumptions set out here,
emissions from pipeline CO2 transport are almost negligible
compared to direct emissions, operational emissions, and
emissions associated with capture. In the carbon capture
transport and storage (CCTS) system, the transport element
appears to have a tiny impact on GHG emissions.
Sensitivity study
Figure 3 illustrates how sensitive GHG emissions from each
of the CCS plant are to system changes, with an emphasis on
transport processes. It is immediately clear that the details of the
CO2 transport system have relatively little effect, as increasing
the pipeline network length by 100 km raises lifecycle emissions
by between 0.05% (PC+CCS) and 0.08% (IGCC+CCS). The
details of the other downstream processes, and in particular
plant ash waste disposal, also seem unimportant.
Upstream transport processes are much more important.
Importing coal from Russia, rather than relying on local
production, has a severe influence on the emissions for
the PC and IGCC plant. Much of this is due to emissions
from the transportation processes, though it has also been
assumed here that a poorer quality coal is delivered. Similarly,
supplying the NGCC+CCS plant from a gas network with
two percentage points greater leakage increases lifecycle
GHG emissions by about one-third.
Also of great importance is the effectiveness of CO2
capture. A 5% reduction in the overall proportion of CO2
captured unsurprisingly gives a substantial increase in GHG
emissions in all cases. This result remains qualitatively true
irrespective of whether the reduction is due to less-effective
capture equipment, or increased leakage from a pipeline
transport system.
The key conclusion is that, in terms of GHG emissions of a
CCS system, most of the details of the downstream processes
are relatively unimportant. This presents a stark contrast
to upstream transport processes, which our results suggest
have a much larger impact on emissions. If a design objective
is to minimize lifecycle emissions, CCS systems should in
general be situated to promote ease and effectiveness of fuel
supply, and with little regard to the implications for the CO2
pipeline transport network. However, one important factor
impacting the lifecycle performance is the total CO2 captured,
and the downstream transport system has the potential to
4th Quarter, 2010
295
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Fig.3. Sensitivity of CCS power plant
lifecycle GHG emissions to system
changes.The first columns show the
percentage increase in emissions
for PC+CCS and IGCC+CCS plant
if all coal Is imported from Russia,
rather than locally sourced.The
second column shows the impact for
NGCC+CCS plant if methane leakage
from the supply network increases
by two percentage points.The third
column represents the impact of
recycling 50% of ash and FGD
waste as construction materials.The
fourth column illustrates the result
of lengthening the CO2 transmission
network by 100km.The final column
shows the effect of decreasing
the CCS capture efficiency by 5%
(assuming all other plant parameters
remain the same).
influence this with respect to its resistance to leaks. While
the overall configuration of the CO2 transport system has
little impact on GHG emissions, it is vitally important that
transport is as secure as possible.
Comparison with other lowcarbon energy sources
For comparison purposes, Table 8 shows ranges of values
for GHG emissions from other low-carbon electricityproduction systems taken from the literature. In general it is
clear that, despite producing much lower carbon emissions
than conventional fossil-fuelled plant, CCS cannot produce
electricity that is as low carbon as most renewables.
It should be noted, though, that the dividing line is rather
fuzzy and dependent on the location and system boundary
of the renewable-energy technology. Solar PV in locations
with poor resources can result in emissions per kWh higher
than those calculated here for IGCC systems. Equally, most
of the renewable-energy systems’ assessments do not take
account of the impact of intermittency on the lifecycle
emissions. Including electricity storage facilities within the
system boundary, for example, can dramatically worsen
the environmental performance. This raises a number of
complex issues that are beyond the scope of this paper,
but it should be kept in mind that fossil-fuel-based CCS
systems have the potential to offer supply controllability
and a geographical independence that certain renewables
find difficult to match without additional facilities.
One possible way of further reducing the emissions of CCS
systems is by combining them with biomass fuels. The
potential of this technology is considered in the remainder
of this paper.
Lifecycle analysis of BioCCS
Objectives
CCS with biofuel firing (BioCCS) offers the attractive
potential of producing a net removal of carbon dioxide
from the atmosphere, since the carbon dioxide released
by biomass combustion was originally absorbed from the
atmosphere in photosynthesis. A further high-level study has
examined the lifecycle implications of BioCCS drawing on
the outputs of the UKCCSC supported study, supplemented
with data from the literature as described in detail by Laczay
[3]. Both pure-biofuel and coal co-firing cases have been
examined, with the pure-biofuel cases considering both
miscanthus1 and RC willow as fuels. For the co-firing case,
only miscanthus was analysed.
Approach and assumptions
The study considers a circulating fluidized bed (CFB) power
plant comparable to the 550MWth / 240MWe facility
operated by Alholmens Kraft in Pietarssaari, Finland [4].
This plant operates at a typical thermal efficiency of 38%,
but it was assumed that a 90% effective CO2 capture system
would reduce the power output by 25%, giving an overall
conversion efficiency for the BioCCS system of 28.5%.
1 Miscanthus is a tall perennial grass that has been evaluated in Europe in recent years
as a new bioenergy crop. It is sometimes confused with elephant grass (Pennisetum
purpureum) and has been called both ‘elephant grass’ and ‘E-grass’.
296
The Journal of Pipeline Engineering
Parameter
Value
Parameter
Ambient temperature, ºC
15
Ambient pressure, kPa
101
Steam cycle heating rate, MJ/kWh
7.4
Gasifier temperature, ºC
Excess air, %
20
Gasifier pressure, MPa
Temperature of flue gas exiting boiler, ºC
370
Load factor, %
75
Steam input to gasifier, mol H2O / mol
C
Life time, years
30
ID fan efficiency, %
85
Table 3. Key parameters for the supercritical-PC plant.
Parameter
Number of gas turbines
Value
2
Excess air, %
180
NOx emissions rate, ppm
10
15.7
Compressor efficiency, %
70
Pressure loss across combustor, kPa
28
Temperature into turbine, ºC
1330
Turbine isentropic efficiency, %
85
Mechanical and generator efficiencies, %
98
Table 4. Key parameters for the NGCC plant.
Due to the complexities associated with biomass lifecycle
analysis, this work used a simplified approach. In particular,
emissions associated with power plant and CO2 pipeline
construction processes have been neglected. As the latter
are relatively small this assumption is unlikely to have a
significant influence on the results. The former are more
likely to have an impact on the detail of the calculations,
but not the qualitative conclusions. Emissions associated
with the up-keep of the carbon capture system, for example
solvent replacement, have also been neglected.
It should also be kept in mind that the study takes no
account of the whole system indirect impacts of wider
biomass use, such as induced land use change (ILUC)
which, it is argued, could have a devastating effect on
the lifecycle sustainability of certain biofuels. There has
been much recent debate about the GHG emissions that
should be associated with biofuel production, typified by
the Searching-Wang debate (see for example [5]) with the
Gallagher Review providing an excellent reference [6]. The
calculations reported here account for only the emissions
that arise directly from biomass cultivation, processing and
harvesting operations.
The Biomass Environmental Assessment Tool (BEAT2) [7,
8] was used to calculate the energy yield and combustion
products in all cases, with the following parameters:
GE oxygenblown
1250
6
0.45
Carbon loss, %
1
Oxidant pressure (at outlet of ASU), MPa
4
Oxidant composition, %O2 : % Ar : %
N2
95 : 4: 1
Particulate removal efficiency from
syngas, %
50
COS to H2S conversion efficiency, %
98
H2S removal efficiency, %
98
COS removal efficiency, %
40
CO to CO2 conversion efficiency, %
95
Sulphur recovery efficiency, %
95
Steam added to shift reactor, mol H2O/
mol CO converted
1
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Air compressor ratio
Type of gasifier
Value
Table 5. Key parameters for the IGCC plant with Selexol
capture.Table 5. Key parameters for the IGCC plant with
Selexol capture.
• All power plant operate with an annual load factor
of 90% for miscanthus yields are 18 wet tonnes per
hectare per year, having 30% moisture content. Once
harvested the feedstock is naturally dried in storage for
40 days reducing the moisture content to 10%. 60kg
of nitrogen fertilizer is used when establishing each
hectare, and a production cycle lasts 15 years after which
the plantation must be cleared and re-established.
• For willow yields are 14 wet tones with 50% moisture
content, with the feedstock dried to 10% moisture
content.
• Energy crops are transported by truck 100km from
the plantation to the storage/processing site, with a
further 100-km journey to the power plant.
• Losses of 7% occur, 11% during storage and 3% during
transport of energy crops.
• Coal is imported from South America, USA, Australia
and South Africa
Results
Table 9 shows the calculated lifecycle GHG emissions for the
pure-biomass based CFB power plant with CCS. Both cases
4th Quarter, 2010
297
Fig.4. Lifecycle comparison of fossil
CCS plant with other low carbon
energy systems. Parameter
CO2 removal efficiency, %
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
show strongly negative net GHG emissions, with the 90%
capture rate more than compensating for the emissions that
do reach the atmosphere. The net emissions are sufficiently
negative that the simplifications outlined earlier are very
unlikely to change the qualitative conclusion
More detail is shown in Fig.6, which compares the origins
of the emissions reaching the atmosphere. The upstream
processes for miscanthus and SRC willow show slight
differences in emissions. Miscanthus has more emissions
associated with cultivation/harvest compared to willow. This
is due to differences in the planting and harvesting processes
of the two energy crops. Miscanthus also has much higher
transport emissions than SRC willow because it is bailed
rather than chipped. Chips are more densely transported,
and as a result, miscanthus transport emissions are nearly
twice that of chipped SRC willow.
Results for co-firing with coal are shown in Fig.7, where the
vertical axis represents net atmospheric GHG emissions per
kWh of electricity produced relative to a supercritical coal
power plant without a CO2 capture unit. Unsurprisingly,
net emissions reduce almost in direct proportion to the
proportion by energy value of biomass in the fuel mix. A
useful observation is that miscanthus-based BioCCS appears
to become GHG neutral for a co-firing level of approximately
20%. Higher proportions of biomass produce net capture
from the atmosphere, though some care is necessary in
interpreting the values in the light of the simplifications
outlined earlier.
Discussion
Emission minimization strategies
Presumably a key objective in the design of any electricityproducing CCS system is to generate power with the lowest
achievable GHG emissions per unit. To reach this objective,
Value
90
SO2 removal efficiency in capture plant,
%
99
SO2 removal efficiency in FGD, %
98
SO3 removal efficiency in capture plant,
%
99
HCl removal efficiency in FGD, %
95
NO2 removal efficiency in capture plant,
%
25
Ash removal efficiency in FGD, %
50
MEA concentration, %w/w
30
Lean CO2 loading, mol CO2/mol MEA
0.2
Blower efficiency, %
75
Pressure across blower, kPa
15
Sorbent pump efficiency, %
75
Pressure across pump, kPa
200
Compressor efficiency, %
80
CO2 outlet pressure, MPa
13.5
Table 6. Key parameters for MEA-based capture process.
the results in this paper suggest that there is some value in
adopting an integrated approach to the design of the whole
system, as decisions made in one part of the CCTS chain
can have implications for the GHG emission of another.
Minimizing overall emissions requires that such interactions
are fully accounted for.
In the cases considered here, downstream pipeline-based
CO2 transport does not have a substantial influence on
overall GHG emissions, and thus can largely be designed
independently from the rest of the system in this regard.
298
The Journal of Pipeline Engineering
Source of GHG Emissions (gCO2e/kWh)
Plant Type
Total
Emissions
Construction
Direct
Operation
CO2 Capture
CO2 Transport
Super-Crit Coal
2
788
91
0
0
881
S-C Coal + CCS
3
107
124
22
3
258
NGCC + CCS
3
42
118
25
2
190
IGCC + CCS
3
90
73
1
3
170
Table 7. Summary of lifecycle GHG for representative power plant.
Range of GHG emissions
(gCO2/kWh)
References
Hydro
3-33
[11], [12], [13],
[14]
Geothermal
15-23
[13], [14]
Solar PV
39-217
[15], [14], [16],
[17], [18]
Highly location dependent. Some higher values
included battery storage
Solar thermal (to electricity)
30-120
[19]
Parabolic trough, centralized receiver & parabolic
dish
Wind
9.7-29.5
[20],[14]
Wind with pumped hydro storage
20
[21]
Wind with compressed air storage
109
[21]
6-24.2
[12], [22], [14]
Nuclear fission
Comments
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Technology
Higher values generally for
offshore
Table 8. Representative values for GHG emissions from several low-carbon electricity production technologies. Note that the
large ranges arise partially from incompatible assumptions between the studies considered. Fuel
Miscanthus
Willow
73
57
CO2 from biomass combustion
1291
1449
Combustion CO2 captured
1162
1304
Combustion CO2 to atmosphere (B)
129
145
Net combustion CO2 emissions (C)
-1162
-1304
Other power plant GHG emissions (D)
19
15
Direct emissions to atmosphere (A+B+D)
221
217
-1070
-1232
Upstream Process GHG Emissions (A)
NET GHG EMISSIONS (A+C+D)
Table 9. Indicative lifecycle GHG emissions for CFB biofuel to electricity plant with a 90% capture efficiency carbon dioxide
capture plant, operating on two fuels.The overall conversion efficiency to electricity is taken to be 28.5%. All emissions are
stated in gCO2e/kWh(e).
4th Quarter, 2010
299
Fig.5. Biomass with CCS system
summary. Pipeline CO2 transport can influence system GHG emissions
via leakage, and hence minimizing escapes should be a
primary design objective.
Thickness, mm
On-shore
Off-shore
On-shore
Off-shore
75
100
7.8
9
Table 10. Diameter and wall thickness for natural gas
pipeline.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
For countries of scales similar to the UK, pipeline length
does not have a significant effect on overall GHG emissions.
Hence the CO2 transport distance should not play major
role in CCS plant site selection. Upstream transport
processes, notably fuel transport, have a much stronger
impact on emissions and should have an influence on site
selection. Clearly any whole-system GHG-minimization
strategy should focus on simplifying fuel rather than
CO2 processing and transport, so long as any risk of CO2
leakage is avoided. This is particularly true with biomassbased CCS systems.
Diameter, cm
Areas of uncertainty
While undertaking this work we have identified a number
of areas where limited technical understanding constrains
the usefulness of LCA approaches for the analysis and
optimization of future CCS systems. The two most
important, in the opinion of the authors, are discussed in
this section.
Operational effects
Most LCA analyses assume that the systems they study
operate under steady-state, full-load conditions, and this
is true of almost all CCS studies. Where some account is
taken of variable loading, typically, analysts use only a load
factor approach to account for periods of non-generation.
Where plant will be used predominantly to supply base load,
this is a reasonable assumption, especially as construction
makes a relatively small contribution to overall emissions
even for CCS power plant.
The vagaries of the electricity market mean that base-load
operation is unlikely for all CCS systems, in practice. As a
result, some plant might be subject to substantial numbers
of cold-start and shut-down procedures, as operators try to
optimize their financial return. Operators may also wish
to run plant at part load. From a lifecycle GHG-emissions
perspective, non-steady-state and part-load operations are
likely to exhibit much poorer efficiency than steady-state
full-load operation. In consequence they have the potential
to substantially increase the overall GHG emissions of a
CCS system, particularly if they are frequent events.
New CCS plant will most likely be initially conceived for
base-load operation, and thus it could be argued that their
lifecycle performance will not be impacted by non-steadystate operation. Over their lifetime, though, there will be
substantial changes in energy and electricity markets, and as
they age, CCS plant are likely be moved towards a peaking
role as is common with existing old fossil-fuel plant [9].
Moreover, expected increasing penetration of intermittent
renewables will push even CCS fossil plant towards operating
regimes that are more variable than those experienced by
existing fossil plant.
Evaluating the impact of transient operation on GHG
performance is hindered by poor understanding of both
CCS plant and carbon dioxide transport systems under
such conditions. Further work is required in these areas
in order to fully evaluate the ‘real world’ performance of
future CCS systems.
Impact of biomass combustion products on CCS efficiency
While the results in this paper suggest that biomass-to-power
combined with CCS has the potential to produce negative
lifecycle GHG emissions, it is important to keep in mind
that the underlying calculations assumed there was no
detrimental interaction between the biomass combustion
products and both the capture system together with the CO2
transport system. In general this would seem a reasonable
assumption, it being widely accepted that co-firing reduces the
emission of pollutant elements (including sulphur, nitrogen,
and mercury) in comparison to pure coal. However, biomass
300
The Journal of Pipeline Engineering
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Fig.6. Breakdown of contributions to
direct GHG atmospheric emissions
(i.e. A+B+D in Table 10) for the
biomass with CCS system described in
the text, operating on two fuels
Fig.7. Relative impact of increasing
levels of co-firing with miscanthus
on lifecycle GHG emissions for a
supercritical coal plant, with an
without a carbon dioxide capture
unit with a 90% capture efficiency.
The vertical axis shows the emissions
compared to a representative coal
super-critical plant without CCS.
co-firing can yield increased concentration of hydrochloric
acid in flue gases [10].
Again there is scant data available in the literature regarding
the effect of biomass combustion products on CO2 capture
and transport processes. Further technical data are required
before the LCA can be taken forward.
Conclusions
Transport emissions are a relatively small component of the
GHG emissions from CCS systems, though the quantities
vary considerably with the assumptions underlying the
lifecycle analysis. As a general rule, downstream emissions
associated with pipeline CO2 transport are almost negligible,
certainly with respect to the construction of short pipelines.
Operational emissions depend on the source of the energy
used to power recompressions stations. However any CO2
leakage from the pipeline system to the atmosphere has the
potential to dramatically increase the impact of downstream
transport processes.
Upstream transport emissions, predominantly for fuel,
are more important, typically representing at least 2% of
all GHG emissions. For biofuels, upstream emissions are
considerably more important. This is true even without
taking account of the current uncertainty surrounding the
whole-system sustainability of biofuels, as typified by the
Searchinger-Wang debate.
The results have implications for reducing lifecycle GHG
emissions from CCS plant by optimizing plant location. In
Journal of Pipeline Engineering
Editorial Board - 2010
There are two key areas of uncertainty that have had relatively scant coverage in the literature and require further
work. Firstly, most lifecycle studies of CCS systems make
substantial simplifications with respect to the operational
regime of the plants under consideration. Secondly, the
impact that combustion of biomass fuels might have on the
performance of carbon dioxide capture systems does not
appear to have been extensively considered in LCA studies
of CCS systems. Such an extension can be readily included
within the lifecycle methodology in principle, but there is
a scarcity of data to support such work.
Sa
no m
t f ple
or c
di op
st y
rib
ut
io
n
Obiechina Akpachiogu, Cost Engineering Coordinator, Addax Petroleum
Development Nigeria, Lagos, Nigeria
Mohd Nazmi Ali Napiah, Pipeline Engineer, Petronas Gas, Segamat, Malaysia
Dr Michael Beller, NDT Systems & Services AG, Stutensee, Germany
Jorge Bonnetto, Operations Vice President, TGS, Buenos Aires, Argentina
Mauricio Chequer, Tuboscope Pipeline Services, Mexico City, Mexico
Dr Andrew Cosham, Atkins Boreas, Newcastle upon Tyne, UK
Prof. Rudi Denys, Universiteit Gent – Laboratory Soete, Gent, Belgium
Leigh Fletcher, MIAB Technology Pty Ltd, Bright, Australia
Roger Gomez Boland, Sub-Gerente Control, Transierra SA,
Santa Cruz de la Sierra, Bolivia
Daniel Hamburger, Pipeline Maintenance Manager, El Paso Eastern Pipelines,
Birmingham, AL, USA
Prof. Phil Hopkins, Executive Director, Penspen Ltd, Newcastle upon Tyne, UK
Michael Istre, Engineering Supervisor, Project Consulting Services,
Houston, TX, USA
Dr Shawn Kenny, Memorial University of Newfoundland – Faculty of Engineering
and Applied Science, St John’s, Canada
Dr Gerhard Knauf, Salzgitter Mannesmann Forschung GmbH, Duisburg, Germany
Lino Moreira, General Manager – Development and Technology Innovation,
Petrobras Transporte SA, Rio de Janeiro, Brazil
Prof. Andrew Palmer, Dept of Civil Engineering – National University of Singapore,
Singapore
Prof. Dimitri Pavlou, Professor of Mechanical Engineering,
Technological Institute of Halkida , Halkida, Greece
Dr Julia Race, School of Marine Sciences – University of Newcastle,
Newcastle upon Tyne, UK
Dr John Smart, John Smart & Associates, Houston, TX, USA
Jan Spiekhout, Kema Gas Consulting & Services, Groningen, Netherlands
Dr Nobuhisa Suzuki, JFE R&D Corporation, Kawasaki, Japan
Prof. Sviatoslav Timashev, Russian Academy of Sciences – Science
& Engineering Centre, Ekaterinburg, Russia
Patrick Vieth, Senior Vice President – Integrity & Materials,
CC Technologies, Dublin, OH, USA
Dr Joe Zhou, Technology Leader, TransCanada PipeLines Ltd, Calgary, Canada
Dr Xian-Kui Zhu, Senior Research Scientist, Battelle Pipeline Technology Center,
Columbus, OH, USA
particular, the lifecycle GHG impacts are much more
sensitive to fuel transport processes than downstream
carbon dioxide transport. From the GHG perspective, it is
suggested that optimal plant location strategies for the UK
should focus on minimizing fuel processing and transport,
and not be overly concerned about CO2 transportation
distance. This is especially true for biomass-derived fuels,
even if they are sourced from the UK.
❖❖❖
Acknowledgements
This paper draws on CCS systems’ analysis projects undertaken by the authors over the last eight years. Major
support has been provided by the Tyndall Centre, and
the UK Natural Environment Research Council (NERC)
within the scope of the UK Carbon Capture and Storage
Consortium.
References
1. N.A.Odeh and T.T.Cockerill, 2008. Life cycle analysis of UK
coal fired power plants. Energy Conversion and Management, 49,
212–220. doi:10.1016/j.enconman.2007.06.014.
2. Ibid., 2008. Lifecycle GHG assessment of fossil fuel power
plants with carbon capture and storage. Energy Policy, 36,
367–380.
3. S.A.Laczay, 2009. A comparative analysis of the economics and
GHG emissions of fossil fuel, co-fired and dedicated biomass
electricity generation systems with carbon capture and storage
in the UK. Master’s thesis, Centre for Environmental Policy,
Imperial College London, September.
4. Alholmens Kraft. The history. Website [accessed 15 July 2009]
www.alholmenskraft.com/en/history/index.htm.
5. T.Searchinger, R.Heimlich, R.A.Houghton, F.Dong,
A.Elobeid, J.Fabiosa, S.Tokgoz, D.Hayes, and T.-H.Yu, 2008.
Use of US croplands for biofuels increases greenhouse
gases through emissions from land use change. Science, 319,
1238–1240.
6. EGallagher, ed., 2008. The Gallagher Review of the indirect
effects of biofuels production. Renewable Fuels Agency. Available at www.renewablefuelsagency.gov.uk/reportsandpublications/reviewoftheindirecteffectsofbiofuels.
7. AEA. Biomass Environmental Assessment Tool Version 2 User
Guide., 2008. Available at http://www.biomassenergycentre.
org.uk/pls/portal/url/ITEM/5A8E649760A5B873E04014A
C08044B4D.
8. AEA, 2007. The biomass environmental assessment tool
(BEAT2). www.biomassenergycentre.org.uk/portal/page?_
pageid=74,153193&_dad=portal&_schema=PORTAL.
9. DECC, 2009. Towards carbon capture and storage: Government response to consultation. Technical report, UK Department of Energy and Climate Change.
10. M.F.G.Cremers, 2009. Technical status of biomass co-firing.
Technical Report IEA Bioenergy Task 32 Deliverable 4, International Energy Agency, August.
11. L.Gagnon and J.Van de Vate, 1997. Greenhouse gas emissions
from hydropower: The state of research in 1996. Energy Policy,
25, 1, 7–13.
12. L.Gagnon, C.Belanger, and Y.Ychiyama, 2001. Life-cycle assessment of electricity generation options: The status of research
in year 2001. Idem, 30.
13. Y.Uchiyama, 1996. Life cycle analysis of electricity generation
and supply systems. In Proceedings of a symposium on electricity, health and the environment: Comparative assessment
in support of decision making, pp279–291, Vienna, Austria,
October. International Atomic Energy Agency.
14. H.Hondo, 2005. Life cycle GHG emission analysis of power
generation systems: Japanese case. Energy, 30, 2042–2056.
15. P.Denholm and G.Kulcinski, 2003. Net energy balance and
greenhouse gas emissions from renewable energy storage system. Technical Report 223-1, Energy Centre of Wisconsin.
Available at http://fti.neep.wisc.edu/pdf/fdm1261.pdf.
16. J.L.R.Proops, P.W.Gay, S.Speck, and T.Schroder, 1996. The
lifetime pollution implications of various types of electricity
generation. Energy Policy, 24, 3, 229–237.
17. R.Kannan, K.C.Leong, R.Osman, and H.K.Ho, 2007. Life
cycle energy, emissions and cost inventory of power generation
technologies in Singapore. Renewable and Sustainable Energy
Reviews, 11, 702–715.
18. H.Schaefer and G.Hagedorn, 1992. Hidden energy and correlated environmental characteristics of PV power generation.
Renewable Energy, 2, 2, 159–166.
19. C.Dey and M.Lenzen, 2000. Greenhouse gas analysis of
electricity generation systems. In Proceedings of the ANZSES
Solar 2000 Conference, pages 658–668, Griffith University,
Queensland, Australia, November.
20 L.Schleisner, 2000. Life cycle assessment of a wind farm and
related externalities. Renewable Energy, 20, 279–288.
21. P.Denholm and G.L.Kulcinski, 2004. Life cycle energy requirements and greenhouse gas emissions from large scale
energy storage systems. Energy Conversion and Management, 45,
2153–2172.
22. R.Dones, T.Heck, M.F.Emmenegger, and N.Jungbluth, 2005.
Life cycle inventories for the nuclear and natural gas energy
systems, and examples of uncertainty analysis: EcoInvent:
Energy Supply. Int.J. of Life Cycle Assessment, 10, 1, 10–23.