Pipeline Technology Conference 2008
Topic
CARBON DIOXIDE TRANSPORT IN PIPELINES - UNDER
SPECIAL CONSIDERATION OF SAFETY-RELATED ASPECTS
Author
Dr. -Ing. Klaus-Dieter Kaufmann
Presenter
Dr. -Ing. Klaus-Dieter Kaufmann
Organization
ILF Beratende Ingenieure GmbH, Werner-Eckert-Strae 7, D-81829
Munich
Country:
Germany
Abstract:
The capture of CO2 from power plants, its transportation through pipeline systems
and its long term deposition in suitable storage reservoirs both on and offshore
appears to be an effective method for preventing CO2 from entering the atmosphere,
allowing to mitigate adverse greenhouse gas effects due to anthropogenic activities.
Depending on the process or power plant application, three main approaches to
capturing the CO2 generated from a primary fossil fuel (coal, natural gas or oil), biomass, or mixtures of these fuels are presently considered: Post-combustion capture
systems, Pre-combustion capture systems and Oxyfuel combustion capture systems.
The production of SO2 or H2S as unintended and potentially dangerous by-products
is also considered.
While existing CO2 pipelines in U.S.A. with predominantly relatively pure CO2
streams are running generally through sparsely populated areas, CO2 pipelines from
projected fossil-fuelled power plants containing H2S and SO2 as impurities may
need to cross densely populated areas, for instance in Western Europe.
The availability of reliable calculation methods for the most relevant CO2 properties
(density and viscosity), for the influence estimation of impurity concentrations on
phase behavior and calculation methods for determination of the optimum technoeconomic pipeline diameter, are therefore a pre-requirement for a safe, environmentfriendly and economic pipeline design.
A potential new pipeline route has to be examined within the frame of a risk analysis
to identify hypothetical hazard scenarios and to estimate potential consequences with
regard to severity and estimated frequency. The examination covers the hypothetical
case of leakage, evaluates the time-dependent CO2 leak rate (source term),
estimates the CO2 outflow / jet formation in the immediate vicinity of the leak, and
estimates the dispersion of cold CO2 clouds depending on atmospheric and
topographic conditions like hilly terrain, depressions and big buildings.
The integrated approach which is partially an iterative process starts with the
investigation of a suitable route avoiding exposed areas and close proximity to
inhabited areas, under consideration of the special CO2 and impurity related
properties and local conditions. The study continues with selection of appropriate
P.7 Kaufmann Contact + Abstract + Paper
1/28
Pipeline Technology Conference 2008
pipe material, wall thickness (design factor), burial depth and optimized number and
locations of valve stations, increased quality control during pipe manufacturing, weld
control and supervision during construction works. The study additionally refers to
measures for leak detection and fast pipeline shutdown including quick closure of
sectionalizing valves, as well as implementation of an integrated leak response plan
in order to minimize the potential consequences of a hypothetical CO2 leakage to
people, environment, assets and project reputation. These measures will be
completed by integrated procedures and training measures for the pipeline operators
and by pipeline maintenance measures including running of so-called intelligent pigs.
Nevertheless, there remain some challenging engineering tasks like comparison of
dispersion models versus CO2 test release results, definition of measures to avoid /
reduce potential rupture propagation, investigations referring to potential influence of
H2S and H2 impurities on stress cracking promotion, potential interaction of coabsorbed H2S and SO2 (S2 generation) and investigations (based on reported
transportation system failures) to minimize the remaining pipeline risk to a level as
low as reasonably practicable (ALARP).
P.7 Kaufmann Contact + Abstract + Paper
2/28
Pipeline Technology Conference 2008
TABLE OF CONTENTS
1
INTRODUCTION
CO2 SEPARATION FROM POWER PLANTS AND MAX. LEVEL OF
IMPURITIES
PHYSICAL PROPERTIES OF THE CO2 TRANSPORTED
10
TECHNO-ECONOMICAL DESIGN OF CO2 PIPELINES
12
DEHYDRATION AND PIPELINE MATERIAL SELECTION
15
5.1 Dehydration
15
5.2 Pipe Material Selection
15
HEALTH RELATED ASPECTS OF CO2, H2S AND SO2
16
SAFETY RELATED ASPECTS OF CO2 TRANSPORT IN PIPELINES
17
7.1 Overview
17
7.2 Main Influences on Pipeline Leak Frequencies
18
7.3 Most Probable Hazard Scenario Expected
18
7.4 Basic Data and Correlations for the Source Term
19
7.5 Dispersion Calculation for CO2 Considering Heavy Gas Behavior
20
7.6 General Safety Aspects to be Considered for CO2 Pipelines
21
7.7 Risk Calculation for CO2 Pipelines
23
7.8 Measures for Risk Minimization
25
FURTHER INVESTIGATION DEMAND
27
REFERENCES
27
P.7 Kaufmann Contact + Abstract + Paper
3/28
Pipeline Technology Conference 2008
INTRODUCTION
Approximately 60% of the global carbon dioxide emissions (about 23.5 Gt in
year 2000) is attributed to large stationary sources (Power Plants). The main
clusters of emissions are located in North America, Europe (northwest
region), East Asia (east coast of China) and South Asia (Indian subcontinent) /1/.
The capture of CO2 from power plants, its transportation through pipeline
systems and its long term deposition in suitable storage reservoirs both on
and offshore appears to be an effective method for preventing CO2 from
entering the atmosphere, allowing to mitigate adverse greenhouse gas
effects due to anthropogenic activities. The distances between CO2 sources
and sinks can vary considerably and in some cases will amount to several
hundreds of kilometres.
Demonstration projects are underway worldwide to investigate the feasibility
of commercial scale carbon capture and storage (CCS) technology. Some
major carbon capture facilities already exist for example in Sleipner
(Norway), Weyburn (Canada), In Salah (Algeria) and two projects in
Germany (RWE IGCC & Ketzin).
Pipeline systems for the safe transportation of oil, gas and speciality products
are already widely accepted around the world as a means for the long
distance transfer of large quantities of these products. It is also noted that
there exists commercial scale pipeline systems for the transportation of CO2
mainly from enhanced oil recovery (EOR) projects in U.S.A. (approx. 2,600
km in 2002) which have been operated for about 30 years /1/.
Despite this, there are two significant differences between existing CO2
pipelines installed in U.S.A. and those foreseen for projected carbon capture
and storage projects elsewhere:
a)
Existing CO2 pipelines in U.S.A. are running generally through sparsely
populated areas, while in Western Europe, potentially densely
populated areas may need to be crossed
b)
CO2 streams from natural CO2 storage fields (within the frame of EOR
projects) in U.S.A. are predominantly relatively pure CO2 streams, while
CO2 streams from projected fossil-fuelled power plants will contain
small amounts of co-absorbed components like H2S and SO2.
P.7 Kaufmann Contact + Abstract + Paper
4/28
Pipeline Technology Conference 2008
Safety related aspects are therefore of high importance and represent, in
addition to the techno-economical aspects, a special focus of this
presentation.
Figure 1-1 shows the basic scheme for CO2 capture, pipeline transportation
and storage options.
Figure 1-1: Basic Scheme of CO2 Capture, Pipeline Transportation and
Storage Options
CO2 Capture Facilities
Post Combustion
Pre Combustion
Oxy-Fuel Systems
CO2 Pipeline System
CO2 Storage
Head Station
Pipeline Sections
Line Valve Stations
Intermediate Transport
Station(s)
Terminal Station
Enhanced Oil Recovery
Depleted Oil and Gas
Reservoirs
Saline Aquifers
Enhanced Coal Bed Methane
CO2 SEPARATION FROM POWER PLANTS AND MAX. LEVEL OF
IMPURITIES
Figure 2-1 shows the basic scheme of a typical conventional power plant.
Heat and power is generated in a boiler plant with steam turbines, optional
selective catalytic reduction (SRC) for nitrogen oxides (NOx), particulate
collection and cooling, and optionally, flue gas desulphurization (FGD).
Carbon dioxide (CO2) leaves the power plant via the stack into the
environment at near atmospheric pressure with large proportions of diluents
and contaminants. The capture of CO2 in this condition (a concentration of
around 8%) is both very expensive and requires substantial, additional
expenditure of energy.
Figure 2-1: Basic Scheme of a Typical Conventional Power Plant
N2, O2, H2O,
Typical Conventional Power Plant
CO2*)
to Stack
Coal
Gas
Biomass
Power and Heat
Generation
Selective Catalytic
Reduction
(NOx Removal)
Particulate
Collection and
Cooling
Flue Gas
Desulphurization
Ash
CaSO4
to Disposal
Air
Elaborated / Simplified from IEA GHG Report (8/2004)
P.7 Kaufmann Contact + Abstract + Paper
ILF
5/28
Pipeline Technology Conference 2008
Depending on the process or power plant application, there are three main
approaches to capturing the CO2 generated from a primary fossil fuel (coal,
natural gas or oil), bio-mass, or mixtures of these fuels:
a)
Post-combustion systems
b)
Pre-combustion systems
c)
Oxyfuel combustion systems.
Figure 2-2 shows the basic scheme of a typical post-combustion CO2 capture
process with co-absorption of sulphur dioxide (SO2) in a common liquid
solvent process. CO2 and SO2 impurities are compressed to pipeline
transportation pressure, dehydrated e.g. by triethylene glycol (TEG) in an
intermediate compression stage, and injected into the CO2 pipeline. The CO2
stream contains SO2 and O2 /12/ as major impurities.
Figure 2-2: Basic Scheme of a Typical Post-Combustion CO2 Capture
Process with Co-Absorption of Sulphur Dioxide (SO2)
N2, O2, H2O
to Stack
Typical Post-Combustion Process
CO2 Pipeline
(co-absorbed SO2)
Alternative 2:
Without SO2 Removal Upstream the
CO2 Absorption
CO2-Absorber &
Regenerator,
(SO2 co-absorbed)
CO2 Compression
(SO2 co-absorbed)
TEGDehydration
Flue Gas
Coal
Gas
Biomass
Power and Heat
Generation
Selective Catalytic
Reduction
(NOx Removal)
Particulate
Collection and
Cooling
(probably required)
Air
Ash
ILF
Elaborated / Simplified from IEA GHG Report (8/2004)
Figure 2-3 shows the basic scheme of a typical pre-combustion CO2 capture
process with co-absorption of hydrogen sulphide (H2S) in a common liquid
solvent process.
Primary fuel is processed to generate synthesis gas (H2, CO2, CO), which is
reformed to mainly CO2 and H2; while H2 is used for power and heat
generation, the CO2 including H2S impurities are compressed to pipeline
transportation pressure, dehydrated, and injected into the CO2 pipeline. The
CO2 stream contains H2S as major impurity (also H2 has to be considered in
terms of how the properties of CO2 are affected /12/). The CO2 stream leaves
the process at above atmospheric pressures which saves on compression
investment costs and energy expenditure.
P.7 Kaufmann Contact + Abstract + Paper
6/28
Pipeline Technology Conference 2008
Figure 2-3: Basic Scheme of a Typical Pre-Combustion CO2 Capture
Process with Co-Absorption of Hydrogen Sulphide (H2S) in a
common liquid solvent process
Typical Pre-Combustion Process
CO2 Pipeline
Alternative 2:
Without H2S Removal
(co-absorbed H2S)
CO2 Compression
(H2S co-absorbed)
N2 , O 2 , H2 O
to Stack
TEG-Dehydration
CO2 + H2S
Coal
Biomass
Gasification
(H2+CO2+CO)
Reformer + Acid
Gas Separation
(CO2 & H2S)
H2
Power and Heat
Generation
O2
Steam
Gas, Oil
Air
ILF
Elaborated / Simplified from IEA GHG Report (8/2004) and mod.: only O2 blown into process (A. Brown)
Finally, Figure 2-4 shows the basic scheme of a typical oxyfuel process with
CO2 capture with co-absorption of sulphur oxides (SOx) and of nitrogen
oxides (NOx).
Primary fuel is combusted in oxygen-rich atmosphere, flue gas is recycled
from down-stream the particulate collection to combustion inlet. The CO2 with
SOx and NOx impurities at above atmospheric pressures, which saves on
compression costs and energy expenditure, is compressed to pipeline
transportation pressure, dehydrated, and injected into the CO2 pipeline.
P.7 Kaufmann Contact + Abstract + Paper
7/28
Pipeline Technology Conference 2008
Figure 2-4: Basic Scheme of a Typical Oxyfuel Process with CO2 Capture
with Co-Absorption of Sulphur Oxides (SOx) and of Nitrogen
Oxides (NOx).
Typical Oxyfuel Combustion Process
Alternatives:
Alternatives with NOx/SOx separation to be evaluated
95-99% O2
CO2 Compression
Cryogenic Air
Separation
Unit (ASU)
CO2 Compression
(SOx, NOx coabsorbed)
CO2 Pipeline
(co-absorbed SOx,
NOx)
TEGDehydration
CO2 (+ SOx + NOx)
Flue Gas Recycle
Coal
Power and Heat
Generation
Particulate
Collection
Cooling and
Water KO
Ash
Elaborated / Simplified from IEA GHG Report (8/2004)
ILF
If not anyway required, all the carbon capture processes mentioned above
can have additional steps to adsorb the impurities like H2S and SO2,
however, from the economic point of electrical energy production, it is
advantageous to co-absorb these components into the CO2 stream /2/.
The following Figure 2-5 shows the maximum level of impurities that migh be
produced in captured CO2 streams.
P.7 Kaufmann Contact + Abstract + Paper
8/28
Pipeline Technology Conference 2008
Figure 2-5: Maximum Level of Impurities Potentially Produced in Captured
CO2 Streams
Maximum Level of Relevant Impurities in Captured CO2
Values from IEA GHG Report No. PH4/32 August 2004 (for Coal Fired Power Plants only)
Component
Hydrogen Sulfide
Sulphur Dioxide
Oxygen1)
Hydrogen
Nitrogen
Concentration
mole-%
3.4
H2 S
SO2
2.9
O2
1.9
H2
1.8
N2
0.6
Component
Water
Carbon Monoxide
Nitrogen Oxides
Argon
Methane2)
ILF
Concentration
mole-%
H2 O
0.3
CO
0.2
NOx
0.14
Ar
0.05
CH4
traces
1) Value for Membrane Process (Much lower for Other Priocesses)
2) Value might be considerably higher for gas-fired power plants
The levels shown in Figure 2-5 can be used to define a worst-combination
envelope for impurities potentially produced in CO2-streams from power
plants equipped with carbon capture technology. The most relevant impurity
expected in co-captured CO2 is sulphur either as H2S from integrated
gasification combined cycle plants (IGCC) which use pre-combustion
capture, or as SO2 from conventional steam plants which use postcombustion capture.
There are additional issues to be considered from:
a)
additional requirements referring to the maximum admissible water
content for internal corrosion protection
b)
maximum H2S and SO2 concentrations from safety related aspects
(hypothetical leakage case)
c)
maximum allowable oxygen content from reservoir mechanical aspects
(e.g. bacteriological aspects); the answer is to have a very tight control
on the admissible oxygen content /12/
d)
maximum allowable H2 and H2S concentrations with respect to potential
influences on pipe material properties
e)
potential interaction of H2S and SO2 impurities in CO2 streams which
might undergo catalyzed Claus reaction and potentially could cause
equipment blockages or plug reservoirs /2/.
P.7 Kaufmann Contact + Abstract + Paper
9/28
Pipeline Technology Conference 2008
PHYSICAL PROPERTIES OF THE CO2 TRANSPORTED
The following figures show the vapor pressure curve (Fig. 3-1), and further
the density (Fig. 3-2) and the kinematic viscosity (Fig. 3-3) of pure CO2 as
function of pressure and temperature, calculated with high-accurate property
calculation routines /3,10/. Additionally, the diagrams show the typical
process condition variations (black curves) of CO2 transport in a 24 land
based pipeline system of assumed length of 300 km, able to transport 1,200
t/h CO2 originating e.g. from a 1,200 MWel power plant equipped with carbon
capture technology.
As long as the pressure inside the CO2 pipeline system remains above the
so-called critical point of CO2 (30.98 C, 73.77 bar), the CO2 can be
transported as single phase in the pipeline system which is a pre-condition
for continuous operation at quasi-steady operating conditions. Transportation
in the liquid or in the supercritical state (physically only a question of
definition); is also known as transport in the dense-phase; in UK, from
legislative aspects, transportation in the supercritical state has to be avoided
/12/). The typical transport density amounts to approximately 740-800 kg/m
(see Fig. 3-2) and the typical kinematic viscosity (see Fig. 3-3) to
approximately 0.08-0.09 cSt (mm/s).
Figure 3-1: CO2 Vapor Pressure Curve and Phase / State Definition (also
included are typical operating conditions of a CO2 pipeline
system)
Vapor Pressure Curve and CO2 Phases
(Vapor pressure curve calculated w ith FLUIDCAL Routines)
200
180
Pressure (bar)
160
solid
liquid
140
supercritical
Pipeline 24" (DN 600),
300 km,1200t/h
120
100
80
Triple Point
-56.6 C
5.18 bar
60
40
Critical Point
30.98 C, 73.77 bar
20
gas
0
-80
-50
-20
10
Temperature (C)
P.7 Kaufmann Contact + Abstract + Paper
40
70
100
ILF
10/28
Pipeline Technology Conference 2008
Figure 3-2: Density of CO2 as Function of Pressure and Temperature (typical
operating conditions ca. 740-800 kg/m)
CO2 Density as Function of Pressure and Temperature
(Vapor pressure curve calculated w ith FLUIDCAL Routines)
0 C
1,000
800
10 C
400
20 C
30.98 C
600
0 C
Density (kg/m)
1,200
40 C
60 C
50 C
Pipeline 24" (DN 600),
300 km,1200t/h
200
0
20
50
80
110
140
170
200
Pressure (bar)
0 C
10 C
20 C
30.98 C
40 C
50 C
ILF
60 C
DEW
B UB
P IP E
Figure 3-3: Kinematic Viscosity of CO2 as Function of Pressure and
Temperature (typical operating conditions ca. 0.08-0.09 cSt
(mm/s))
CO2 Viscosity as Function of Pressure and Temperature
(Vapor pressure curve calculated w ith FLUIDCAL Routines)
Pipeline 24" (DN 600),
300 km,1200t/h
20 C
0.05
10 C
0.10
0 C
Viscosity (mPas)
0.15
30.98 C
40 C
50 C
60 C
0.00
20
50
80
110
140
170
Pressure (bar)
0 C
10 C
20 C
30.98 C
40 C
50 C
60 C
200
ILF
DEW
B UB
P IP E
In order to estimate the influence of impurities in real CO2 streams on the
transport properties, available literature /4/ has been evaluated. Figure 3-4
shows the increase of the critical pressure of CO2 streams with impurity
concentration. As long as the amount of impurities remains moderate (e.g.
2.5 %), the pressure increase remains also moderate (< 5 bar). By
corresponding increase of the operating pressure, single-phase
transportation of the CO2 stream can thus be assured.
P.7 Kaufmann Contact + Abstract + Paper
11/28
Pipeline Technology Conference 2008
Figure 3-4: Increase of the Critical Pressure of CO2 Streams with Impurity
Concentration
Increase of Critical Pressure of CO2 Streams with
Impurity Concentration
(Basis: Seevam/Race/Dow nie: J. Pipel. Engng., 3rd. Qu. 2007, pp. 140-141)
Pressure Increase
(bar)
25
Impurity
20
N2
15
H2
10
NO2
5%N2+5%NO2
5%N2+5%CH4
0
0
2.5
7.5
10
Impurity Concentration (%)
ILF
TECHNO-ECONOMICAL DESIGN OF CO2 PIPELINES
The optimized design of a CO2 pipeline system requires the determination of
the optimum pipeline diameter. For this purpose, the specific transportation
costs are determined as a function of assumed CO2 throughput and pipeline
diameter values. The specific transportation costs are hereby calculated
considering the annuity of the capital investment and the annual operating
cost (mainly energy cost for transportation, and maintenance cost).
The techno-economically optimum diameter can then be determined from an
optimization diagram as shown in Figure 4-1. The diagram shows e.g. that for
transportation of annually 10 million tons (MTA) of CO2, a 24 (DN 600)
pipeline system would represent the optimum techno-economic solution. It
has however to be considered that Fig. 4-1 shows only the comparable
transportation cost comprising the main diameter-dependent cost factors. In
case CO2 must be compressed at the head station from approx. atmospheric
conditions to approx. 80 bar (theoretical intermediate reference pressure
level), the related specific compression cost would by far dominate the
specific transportation cost; 80 bar is hereby considered to be a realistic
minimum transportation pressure in order to keep the CO2 steadily in its
dense phase.
P.7 Kaufmann Contact + Abstract + Paper
12/28
Pipeline Technology Conference 2008
Figure 4-1: Pipeline Diameter Optimization Diagram
OPTIMISATION OF CO2 PIPELINE TRANSPORTATION SYSTEM
Comparable Specific Transportation Cost
(Euro/ton/100 km)
Compression in Head Station (1 to 80 bar)
excluded
Remark: For compression of CO2 from 1 bar to approx. 80 bar (ca. 275 kJ/kg),
approx. 11.5 /t energy cost and approx. 2 /t annuity cost are to be added
Pipeline
Diameter
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0 MTA
5 MTA
10 MTA
15 MTA
20 MTA
25 MTA
Throughput (MTA)
16 Inch
20 Inch
24 Inch
28 Inch
32 Inch
30 MTA
ILF
36 Inch
40 Inch
The following Figures 4-2 and 4-3 show the pressure and temperature
profiles as well as the density and elevation profiles of a hypothetical landbased CO2 pipeline system able to transport 1,200 tons per hour of CO2 over
a distance of 300 km.
P.7 Kaufmann Contact + Abstract + Paper
13/28
Pipeline Technology Conference 2008
Figure 4-2: Pressure and Temperature Profile of a 24 (DN 600) CO2
Pipeline System
CO2 Pipeline Pressure and Temperature Profile
140
120
100
80
60
40
20
0
60
50
40
30
20
50
100
150
200
10
0
300
250
Temperature (C)
Pressure (bar)
Pipeline 24" (DN 600), 1200 t/h
Length (km)
ILF
Figure 4-3: Density and Elevation Profile of a 24 (DN 600) CO2 Pipeline
System
CO2 Pipeline Density and Elevation Profile
Density (kg/m)
800
600
400
200
0
0
50
100
150
Length (km)
200
250
800
700
600
500
400
300
200
100
0
300
Elevation (m aSL)
Pipeline 24" (DN 600), 1200 t/h
1000
ILF
Pipeline inlet conditions were assumed to be 130 bar and 40 C; if the CO2
had to be compressed from atmospheric conditions to the pipeline inlet
pressure, the power requirement for a four-stage compression plant would
amount to approximately 130 MW shaft rated power; the optimum number of
compression stages (4-7) has to be determined during the more detailed
design. The maximum allowable inlet temperature into the pipeline system
has to be reviewed project-specifically with respect to environmental and
authority requirements.
P.7 Kaufmann Contact + Abstract + Paper
14/28
Pipeline Technology Conference 2008
DEHYDRATION AND PIPELINE MATERIAL SELECTION
5.1
Dehydration
Due to the high corrosion potential of CO2, it is not considered to be
practicable to transport wet CO2 in low-alloy carbon steel pipeline systems.
However, once dehydration has been performed, corrosion is not considered
a major issue for transport of CO2 streams /2/.
The review of different literature showed maximum admissible water
concentrations in the CO2 stream between 0.039 g/m /5/ and 0.48 g/m /2/,
but these dehydration requirements for CO2 streams can be met by glycol
dehydration which is typically performed at an intermediate stage of
compression (approx. 50 bar) /2/. However, potential hydrate formation at low
temperature must be considered /12/.
5.2
Pipe Material Selection
Experiences with carbon steels (e.g. with API X-60 and X-65) for
transportation of CO2 streams are already available since more than 30 years
e.g. from EOR projects /5/. For pipeline systems operated with dry
supercritical CO2 streams, the corrosion rate is low (approx. 0.01 mm/year at
temperatures of 160C-180C). Field experience confirms this; for a carbonsteel pipeline system operated with high-pressure CO2 during 12 years, a
corrosion rate of only 0.25-2.5 micrometer per year was reported /1/.
For wet service prior to dehydration, 304L stainless steel has been found a
suitable material /5/.
For carbon dioxide service at high pressures in valves, control seals and
packing special CO2 resistant materials like nylon or viton are considered
appropriate /5, 12/.
When specifying the requirements for dry supercritically operated carbon
steel pipelines, in addition to the potential corrosion aspect (specifying a
corrosion allowance), especially the following must be respected:
Low temperatures down to approx. -20C (normal operation /6/) or even
to approx. -78.5C (during depressurization) may occur; material
specification and pipeline operation must respect this.
Susceptibility to long running brittle fracture should be avoided /
prevented by a series of measures, e.g. minimization of defects
introduced during pipe manufacture, construction supervision and
specification of appropriate toughness requirements (Charpy V-notch
(CVN) energy, drop-weight tear test (DWTT) shear area /7/).
P.7 Kaufmann Contact + Abstract + Paper
15/28
Pipeline Technology Conference 2008
The installation of mechanical crack arrestors is not the only option for
fracture-propagation control, considering the capability of modern pipe
mills to produce high-toughness linepipe steel /7/.
The potential influence of the presence of H2S and H2 potentially
promoting stress cracking of the pipeline shall be considered; this
concern is usually lessened by selection of softer pipe steel /2/ or by
careful attention to the amount of H2S in the CO2 /12/.
HEALTH RELATED ASPECTS OF CO2, H2S AND SO2
Carbon Dioxide
CO2 (relative density to air: 1.52) is a colorless, odorless, noncombustible gas, non-toxic at moderate concentrations
Natural ambient concentration ca. 0.04%; Human expired air contains
ca. 4% CO2, physical discomfort at 5%, life-threatening at
concentrations > 20% CO2 /2/ (to be critically reviewed)
Hydrogen Sulphide
H2S (relative density to air: 1.18) is a flammable, extremely toxic gas
with strong odor at low concentrations
Maximum exposure level 10 ppm /11/, 50 ppm in air causes headaches
and nausea, 100 ppm can cause unconsciousness within a few minutes
/2/
Sulphur Dioxide
SO2 (relative density to air: 2.21) is a non-combustible gas with strong
odor, key component in air pollution
Levels of 100 ppm in air can be considered life-threatening /2/
Comparable Life-Threatening Impact of H2S and SO2 in CO2 Streams
Based on the Life-Threatening Impact of H2S and SO2 in CO2 Streams as
reported above, Table 6-1 below shows the calculated concentrations of
equal life-threatening impact in CO2 streams with impurities H2S and SO2.
Table 6-1: Concentrations of Equal Life-Threatening Impact of H2S and SO2
in CO2 Streams
Component
Unit
Concentration
CO2
H2S
SO2
vol-%
ppm
ppm
100
500
500
P.7 Kaufmann Contact + Abstract + Paper
16/28
Pipeline Technology Conference 2008
From the results shown in Table 6-1, the following can be concluded:
In order to avoid, that the impurity components H2S and SO2 dominate
the health-related impact in case of a leakage, the maximum allowable
concentrations of these components in the CO2 stream must be limited
to certain values which were preliminary estimated to be in the order of
magnitude of 500 ppm for H2S as well as for SO2.
This requirement may have considerable consequences on the design
of the carbon capture processes / technologies foreseen / required for
installation, especially if the pipeline route runs through densely
populated areas.
Because CO2 itself has no smell, a small leak can remain undetected.
For comparison, methane (as natural gas) is also without odour, so in
the UK ethyl mercaptan a foul-smelling odourising compound is added
at a concentration of about 5ppb. Allowing some H2S to remain in the
CO2 will allow early detection of any leaks, as H2S is initially detectable
by smell at levels <1ppm in air /12/
If the level of H2S in CO2 is set as low as possible, then the potential
exists to end up with a plant that is technologically excellent but
commercially unaffordable /12/.
SAFETY RELATED ASPECTS OF CO2 TRANSPORT IN PIPELINES
7.1
Overview
The hypothetical leakage of a CO2 pipeline represents the major risk of CO2
transportation in pipelines.
In case of a leak, the CO2 will be released at a very low temperature (the
therodynamic phase diagrams predict approx. -78.5 C) as mixture of cold
gas and fine solid particles into the environment.
While in case of smaller leaks, the CO2 clouds formed just disperse after
short time, in case of larger leaks, the cold gas/solid mixture forms dense-gas
clouds which move (driven by the wind and by the own weight) slowly over
the terrain and can gather at low points (depressions, low-situated rooms in
buildings), displacing hereby the ambient air (danger of asphyxiation, toxic
effects at higher concentration).
The cooling effect in the CO2 rich plume can cause localized condensation of
atmospheric water into fine aerosol droplets which create regions of very low
visibility that can drift with the prevailing wind. Whilst this is a distinct
advantage in the early identification of a large scale release, it can also
represent e.g. a driving hazard in an extreme scenario.
P.7 Kaufmann Contact + Abstract + Paper
17/28
Pipeline Technology Conference 2008
Cooling down of the pipeline material due to the leakage could result in local
embrittlement of the pipe material potentially forming the starting point of long
running ruptures.
7.2
Main Influences on Pipeline Leak Frequencies
As reported in literature /8/, the following can be concluded referring to
anticipated pipeline leak frequency:
7.3
The external interference activities causing most incidents are those
which involve excavators for digging (39%), drainage works (8%), public
works (8%), and the activities related to agriculture (8%).
The leak frequency decreases dramatically with increasing pipeline
diameters; for all 24 pipelines investigated, the average leak frequency
amounts to approx. 0.02 leakages per 1,000 km * year
The leak frequency decreases with increased pipe wall thickness (> 10
mm) dramatically
Increased soil cover depth has only limited influence on decrease of
leak frequency
Pipeline systems installed since approx. 1980 show considerable lower
leak frequencies than older ones.
Most Probable Hazard Scenario Expected
The most probable hazard scenario expected could potentially be described
as follows:
Pipeline damage by an excavator tooth e.g. during third-party
construction works
typical hole diameter expected: 50 mm
gaseous CO2 with fine dispersed CO2 solid particles is injected as
vertical or inclined jet stream into the atmosphere
dispersed fine CO2 particles in the jet stream de-sublimate
formation of a cold CO2 - air mixture cloud (density higher than air)
movement of the cold CO2 cloud is influenced by meteorologic and
topographic conditions (e.g. hilly terrain, depressions, buildings)
dispersion / dilution of the CO2 - air mixture cloud after some time to
non-dangerous concentration
leak condition will quickly be detected by leak detection system
adjacent line valves close quickly to minimize the CO2 mass loss
P.7 Kaufmann Contact + Abstract + Paper
18/28
Pipeline Technology Conference 2008
7.4
Basic Data and Correlations for the Source Term
In order to perform CO2 dispersion calculations, a suitable source term must
be determined considering
the length and diameter of the pipeline section affected by the leak
the pressure and temperature conditions inside the pipeline
the location of the leak inside the pipe section affected; in worst-case,
the leak would be assumed to occur at the lowest point of the pipe
section considered
a suitable correlation for the description of the (approximately
isenthalpic) chocked leak flow through the orifice as function of
pressure and temperature inside the pipe and of the leak size,
considering gas/liquid and gas/solid phase equilibria during release
a suitable calculation method for the (approximately isentropic)
pressure and temperature change inside the pipeline system,
respecting hereby also heat transfer between pipe and surrounding soil
material.
The following Figure 7.4-1 shows the calculated leak flow rate as function of
time for a hypothetical leak situation at a 24 (DN 600) pipeline section of 3
km length, start conditions 150 bar, 30C with an assumed leak diameter of
50 mm.
Figure 7.4-1: Calculated Leak Flow Rate as Function of Time (Source Term)
CO2 Leak Flow Rate as Function of Time
Pipeline Section 24" (DN 600), Length 3 km, Start Pressure 150 bar, Start Temp. 30 C
Leak Diameter 50 mm
Leak Flow (kg/s)
250
200
150
100
50
0
0
20
40
60
80
100
120
Time (Minutes)
P.7 Kaufmann Contact + Abstract + Paper
140
160
180
200
ILF
19/28
Pipeline Technology Conference 2008
7.5
Dispersion Calculation for CO2 Considering Heavy Gas Behavior
Different models for dispersion calculations are available in literature and on
the market, but these must be reviewed for their potential application. These
models shall respect heavy-gas behavior for CO2 dispersion in the
atmosphere. For our estimations we used a special dense gas model of IBS
/9/, which has been developed for dispersion of heavy gases and vapors in
case of hazardous releases, including estimation of toxicity. Figure 7.5-1
shows a vector graphic of the simulated CO2 jet released from a 50 diameter
hole, calculated with the extended model.
Figure 7.5-1: Vector Graphic of a Simulated CO2 Jet in Case of a Leak
(50 mm), Jet Length ca. 70 m, Width ca. 28 m (3 Dim. Model of
IBS /9/)
P.7 Kaufmann Contact + Abstract + Paper
20/28
Pipeline Technology Conference 2008
Figure 7.5-2: Typical Results of the CO2 Dispersion Calculations, Top View
(3 Dim. Model of IBS /9/)
The results show a maximum CO2 concentration on soil level of approx. 4 %.
7.6
General Safety Aspects to be Considered for CO2 Pipelines
General
In U.S.A., CO2 pipelines are mostly routed through low densely populated
areas. Careful routing would especially be required in densely populated
areas with regard to safety distances to be realized there. Dispersion
P.7 Kaufmann Contact + Abstract + Paper
21/28
Pipeline Technology Conference 2008
modeling considering especially topographic (terrain slopes, valleys) and
meteorological conditions would assist to select the optimum pipeline route.
Pipeline Design Codes, Authority Requirements
Safety distances between intended CO2 pipeline routes and populated areas
would have to be designed according to the country-related standards and
regulations as well as considering international standards and rules, Authority
related requirements and expert opinions.
Exemplary Reference Codes
U.S.A.:
49 CFR, Part 195: Law applies for hazardous liquids and CO2;
ASME B31.4 / B31.8
UK:
BS EN 14161:2003: UK recommends combination with PD 18101:2004 (defining the minimum distance for routing purposes
between the pipeline and occupied buildings); CO2 classified as
Cat. C (same as N2, Argon, Air); HSE guidance in UK is that,
until the source terms are better defined (and maybe not even
then) bulk transport of high pressure CO2 such as that in CCS
applications should be class E /12/
Germany: DIN EN 14161 (July 2004): CO2 also classified as Cat. C
substance; this is under review for the same reasons as above
/12/
Remark: In USA, code 49 CFR, Part 195 refers in most parts to hazardous
liquids and carbon dioxide; CO2 is therefore treated legally very
similar a hazardous liquid.
In the European Standard EN 14161, CO2 is only classified as a
Cat C fluid which falls into the category Non-flammable fluids
which are non-toxic gases at ambient temperature and pressure
conditions. Typical examples are nitrogen, carbon dioxide, argon
and air.
According to the Authors opinion, within the frame of future CO2
transportation projects, this divergent classification of CO2
between US codes and EU codes should be subjected to reconsideration, in order to incorporate the relative extensive
P.7 Kaufmann Contact + Abstract + Paper
22/28
Pipeline Technology Conference 2008
experiences especially in USA referring to the hazard potential of
existing CO2 pipelines into the EU legislation.
7.7
Risk Calculation for CO2 Pipelines
General Risk Definition
Risk in general is defined as product of severity of an unintended event and
the expected frequency of this event.
Individual Risk along the Pipeline Route
The individual risk along the pipeline can be calculated as the chance of
fatality of one individual living at a certain distance to the pipeline.
Societal Risk along the Pipeline Route
The societal risk along the pipeline can be calculated as the cumulated
chance of fatality of the individuals living in the influence / impact area of the
pipeline.
Pipeline Route Related Risk Calculation
The following Figure 7.7-1 shows typical results of hypothetical Societal
Risk calculations based on population density and hazard severity along the
pipeline route. The peaks in the cumulated risk curve indicate hereby
especially those locations along the pipeline route where the risk created by
pipeline installation should be further reduced by special measures to a level
as low as reasonably practicable (ALARP principle).
P.7 Kaufmann Contact + Abstract + Paper
23/28
Pipeline Technology Conference 2008
Figure 7.7-1: Typical Results of Hypothetical Societal Risk Calculations
Pop. Density (Pers./m)
Hypothetical Population Density in the Vicinity of the Pipeline
Distance
Range to
Pipeline Axis
Range 1
Range 2
Range 3
Total
Pipeline Location (m)
ILF
Hypothetical Societal Risk in the Vicinity of the Pipeline
Cumulative Risk
(1/(km*yr))
Distance
Range to
Pipeline Axis
Range 1
Range 2
Range 3
Total
Pipeline Location (m)
P.7 Kaufmann Contact + Abstract + Paper
ILF
24/28
Pipeline Technology Conference 2008
7.8
Measures for Risk Minimization
The following shows the main measures considered to reduce the risk from
new CO2 pipeline transportation system to a level as low as reasonably
practicable (ALARP).
a)
b)
General
Minimization of distances between power plants with CC
technology and CO2 storage sites.
Consideration of CC technologies avoiding higher levels of toxic
components (H2S, SO2) in the captured CO2 stream.
CO2 Pre-Processing
c)
d)
e)
Pipe Routing
Selection of pipeline routes avoiding densely populates areas. Site
conditions (e.g. hilly terrain) influencing heavy gas dispersion must
be considered.
Performance of route related risk assessment / minimization
studies.
Pre-Optimization of the Pipeline System
Determination of the technically and economically optimized
pipeline system. The CO2 compression energy demand (especially
at the head station) should be minimized.
Consideration
of
throughput
extension
strategies
by
implementation of additional loops lines and transport stations.
Consideration of blow-down system installation for emergency
cases. Consideration will need to be given to the dispersion of the
CO2 following blow-down to avoid possible impacts on humans
present downwind /12/.
Pipe Material
f)
Dehydration of CO2 streams to the level required to prevent
hydrate formation / internal corrosion.
Specification of carbon steel grades with appropriate material
properties respecting occurrence of low temperatures under
normal operating conditions and in case of line depressurization,
e.g. due to a leak. Steel toughness shall be high enough to
prevent ductile fracture propagation.
Pipe Wall Thickness
Consideration of additional corrosion allowance (to cope for
inadvertent moisture ingress); stringent control and monitoring of
CO2 moisture may be potentially the better strategy /12/.
Increase of pipe resistance against third party impact by increased
pipe wall thickness.
P.7 Kaufmann Contact + Abstract + Paper
25/28
Pipeline Technology Conference 2008
g)
h)
i)
j)
k)
Pipe Manufacturing
Selection of experienced manufacturing companies.
Careful supervision of pipe manufacturing process.
Use of pre-coated pipes /12/.
Pipeline Construction
Selection of experienced construction companies and welders.
Increase of welding inspection by increased number nondestructive tests.
Increase of soil covering depth; in rural areas below level of
cultivation.
Implementation of measures to mitigate the effects of fracture
propagation (e.g. installation of mechanical crack arrestors,
specification of high-toughness steel).
Running intelligent pig(s) after pressure testing for reference
purposes.
Additional protection for the pipe in critical locations (e.g. at
crossings with roads)
Sectionalizing Line Valve Stations
Increase of number of line valve stations for leak volume
reduction; the additional risks for leakages introduced by the line
valve stations themselves are to be considered as well.
Optimization of line valve station localization considering elevation
profile.
Leak Detection, Localization and Emergency Shut-Down Systems
Installation of high-efficient and reliable leak detection and
localization systems.
Installation of a pipeline-related Geographical Information System
(GIS), also for cross-checking / validation of potential leaks
reported by third party.
Implementation of a controlled operational emergency shut-down
system and of procedures (COESD) for minimization of CO2
releases quantities in case of a leak by means of mainly remotecontrolled operational actions.
Operation and Maintenance
Continuous training of operators to take necessary actions to
avoid / minimize potential releases of CO2.
Implementation of a systematic, comprehensive and integrated
integrity management program to investigate / maintain / improve
the safety of the CO2 pipeline system
P.7 Kaufmann Contact + Abstract + Paper
26/28
Pipeline Technology Conference 2008
l)
Regular pigging to assess e.g. potential pipe wall damage to
external corrosion
Emergency Prevention and Response Plans
Implementing appropriate measures to alert the public on the
existence and routing / location of below-ground pipelines; use of
pipe markers and tape.
Introducing measures to prevent accidental pipeline damage by
third parties carrying out digging and construction activities nearby
(e.g. one-call systems).
Establishing and maintaining liaison with police and organizations
to respond quickly to a hazardous release.
Development of emergency response plan and establishment / performance
of a continuing training program for emergency situations.
FURTHER INVESTIGATION DEMAND
In the opinion of the Author the following main areas within the frame of the
design of CO2 pipeline systems require further detailed investigations:
9
/1/
Comparison of dispersion model calculation results with dispersion test
results for potential larger CO2 releases and, if required, adaptation of
the theoretical models to fit the measurement results
Potential rupture propagation scenarios in CO2 pipeline systems and
counter measures, respecting the actual pipe steel grades and
properties available from the pipe manufacturers
The potential influence of the presence of H2S and H2 in CO2 streams
potentially promoting stress cracking of the pipeline
Potential interaction of H2S and SO2 impurities in CO2 streams which
might undergo a catalyzed Claus reaction and potentially could cause
equipment blockages or plug reservoirs and
Evaluation of reported failures of CO2 transportation systems to
minimize the pipeline risk.
REFERENCES
IPCC 2005: IPCC Special Report on Carbon Dioxide Capture and Storage;
Prepared by Working Group III of the Intergovernmental Panel on Climate
Change [Metz, B., O. Davidson, H. C. de Coninck, M. Loos, and L. A. Meyer
(eds.)]. Cambridge University Press, Cambridge, United Kingdom and New
York, NY, USA, 442 pp.
P.7 Kaufmann Contact + Abstract + Paper
27/28
Pipeline Technology Conference 2008
/2/
Impact of Impurities on CO2 Capture, Transport and Storage; Report No.
PH4/32; IEA Greenhouse Gas R&D Programme, Rev. 2, Date: Aug. 2004
/3/
Overhoff, U. Wagner, W., Software FLUIDCAL, Lehrstuhl fr Thermodynamik,
Ruhr-Universitt Bochum (2005)
/4/
P. N. Seevam, J. M. Race and M. J. Downie: Carbon Dioxide Pipelines for
Sequestration in the UK: An Engineering Gap Analysis; The Journal of
Pipeline Engineering, 3rd Quarter, 2007, pp. 133-146
/5/
James M. West, Chevron Oil Company: Design and Operation of a Supercritical CO2 Pipeline-Compression System, SACROC Unit, Scurry County,
Texas, SPE 4804, Permian Basin Oil and Gas Recovery Conference, March
12, 1974 (11 pages + 4 figures)
/6/
Holloway, S.; Geological Survey: The Underground Disposal of Carbon
Dioxide; Final Report of Joule II Project N. CT920031, 28.Feb. 1995, 355
pp.
/7/
A. Cosham, R. Eiber: Fracture Control in Carbon Dioxide Pipelines, The
Journal of Pipeline Engineering, 3rd Quarter, 2007, pp. 147-158
/8/
6th EGIG-Report 1970-2004, Gas Pipeline Incidents, Doc. Number EGIG
05.R.0002, December 2005
/9/
R. Schenk: Development of a time-dependent calculation method for
dispersion of heavy gases and vapors in case of hazardous releases including
estimation of toxicity. Project sponsored by European Funds for Regional
Development and by country Sachsen-Anhalt (FKZ 76213/07/02-2, 2005)
/10/
Span, R., Wagner, W.: 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 (1996), 1509-1596
/11/
NIOSH (1994), Documentation for Immediately Dangerous to Life or Health
Concentrations, Cincinnati, OH: U.S. Department of Health and Human
Services, Public Health Service, Centers for Disease Control and Prevention,
National Institute for Occupational Safety and Health, NTIS Publication No.
PB-94-195047
/12/
Information by Andy Brown, Engineering Director of Progressive Energy Ltd.
P.7 Kaufmann Contact + Abstract + Paper
28/28