CO2 Properties and EoS
for Pipeline Engineering Tuesday 11 November 2014, York
www.ukccsrc.ac.uk
Agenda.................................................................................................................................................................................................... 2Delegate List.................................................................................................................................................................................................... 3Speaker biographies.................................................................................................................................................................................................... 4Roland Span - Thermodynamic Property Models for Transport and Storage of CO2.................................................................................................................................................................................................... 8Javier Rodriguez - gSAFT: advanced physical properties for CCS system modelling.................................................................................................................................................................................................... 44Martin Trussler - Phase Behabiour and EoS Modelling of the CO2-H2 System.................................................................................................................................................................................................... 85Richard Graham - Understanding and predicting CO2 properties.................................................................................................................................................................................................... 108Solomon Brown - Impact of EoS on Simulating CO2 Pipeline Decompression.................................................................................................................................................................................................... 171Xiaobo Luo - Simulation-based Techno-economic Evalusation for Optical Design.................................................................................................................................................................................................... 188Chris Wareing - Numerical modelling of trans-triple point temperature....................................................................................................................................................................................................... 205Jie Ke - Phase equilibrium studies of impure CO2 systems....................................................................................................................................................................................................... 216
This meeting will bring together researchers and practitioners in the field of CO2 properties and Equations of
State relevant to the pipeline transportation of CO2. The aim of the meeting will be to share information
relating to both research needs in this area and the status of existing research projects, with a view to
developing a more co-ordinated research effort within the UK CCS research community.
The format of the meeting will include:
- Invited presentations from industrial and academic participants aimed at highlighting research needs
- Short (5 minute) presentations from researchers aimed at illuminating the topics and status of current
research efforts
- Round-table discussions aimed at improved co-ordination of existing research and identification of
opportunities for new research initiative and funding.
Convened by: Dr Richard Graham (University of Nottingham), Dr Julia Race (University of Strathclyde), Prof
Martin Trusler (Imperial College London)
AGENDA
10.00-10.30 Arrivals and registration
10.30-10.45 Welcome/introduction
10.45-11.30 KEYNOTE: Professor Roland Span, Ruhr-Universität Bochum, Germany “Thermodynamic
Property Models for Transport and Storage of CO2”
11:30-11:55 Russell Cooper, National Grid "The Certainties and Uncertainties of CO2 Transport and Storage"
11:55-12:20 Javier Rodriguez, Process Systems Enterprise Ltd "gSAFT: Advanced Physical Properties for
Carbon Capture and Storage System Modelling"
12:20-12:45 Martin Trusler, Imperial College London "Phase Behaviour and EoS Modelling of the Carbon
Dioxide-Hydrogen System"
12:45-13:10 Richard Graham, University of Nottingham "Understanding and Predicting CO2Properties for CCS
Transport”
13:15-14:15 Lunch
14.15-15.45 Group discussions with academic presentations
Solomon Brown, University College London “Impact of Equation of State on Simulating
CO2Pipeline Decompression”
Xiaobo Luo, University of Hull “Study of the Pipeline Network Planned in the Humber Region of
the UK”
Chris Wareing, University of Leeds “Numerical Modelling of Trans-Triple Point Temperature
Near-Field Sonic Dispersion of CO2 from High Pressure Dense Phase Pipelines”
15.45-16.00 Break
16.00-16.30 Discussion and wrap-up
Hamed Aghajani Newcastle University
Saif Al Ghafri Imperial College London
David Allen Doosan Babcock Limited
Emilie Brady UKCCSRC
Solomon Brown University College London
Russell Cooper National Grid
Andrew Cox Energy Intelligence & Marketing Research
Richard Graham University of Nottingham
Yong Hua Leeds University
Bilaal Hussain Birmingham University
Andrew Laughton DNV GL (Oil & Gas UK)
Chih-Wei Lin Heriot-Watt University
Xiaobo Luo University of Hull
Roger Macdonald Pipeline Industry Guild
Haroun Mahgerefteh University College London
Victor Emeka Onyebuchi Cranfield University
Javier Rodriguez Process Systems Enterprise Ltd
Roland Span Ruhr-Universität Bochum
Michael Thomson University of Nottingham
Martin Trusler Imperial College London
Meihong Wang University of Hull
Chris Wareing University of Leeds
Ben Wetenhall Newcastle University
DELEGATE LIST
SPEAKERS
Solomon Brown
University College London
Solomon Brown received his Ph.D. from University College London in 2011 and is currently a research associate
and teaching fellow in the same institution. His expertise belong to the field of mathematical and numerical
modelling of transient multiphase flows. Dr Brown has published on the use of CFD in various safety-related
aspects of CCS and is the recipient of the IChemE Frank Lees Medal for his collaborative work with HSE on CO2
pipelines safety. He has participated in several projects sponsored by the EC, EPSRC, UKCCSRC and industry. He
is a co-investigator on the UKCCSRC project “The Development and Demonstration of Best Practice Guidelines
for the Safe Start-up Injection of CO2 into Depleted Gas Fields” and EC FP7 project “CO2QUEST - Techno-
economic Assessment of CO2 Quality Effect on its Storage and Transport”.
Russell Cooper
National Grid
Russell has been working on CCS transport for over 6-years now. This started off as a side-line to his main job of
managing the design of the gas National Transmission system on behalf of National Grid.
Russell managed the transportation feed for the Longannet CCS proposal that was part of the DECC-1
competition. Since then he has managed the concept selection and consenting process for the proposed
Yorkshire and Humberside CCS transportation and Storage system. During this time Russell has also managed
the research programme that has been developed to address the safety and environmental questions relating to
the transport of CO2.
Richard Graham
University of Nottingham
Richard Graham is an Associate Professor in the School of Mathematical Sciences at the University of
Nottingham. His research broadly encompasses molecular modelling of flow dynamics and phase transitions. His
CCS interests centre on models for impure CO2, for pipeline transport and rupture. To his problem, he has
applied multi-scale techniques, including equations of state, non-parametric modelling and molecular
simulation. This work has been funded by EPSRC, UKCCSRC, E.ON and RWE npower. Previously, Richard worked
at the School of Physics and Astronomy, University of Leeds, and the Department of Chemical Engineering,
University of Michigan.
Xiaobo Luo
University of Hull
Xiaobo received his BEng in Chemical Engineering and Technology in 2001. Following this he was a process
engineer and project manager in National Engineering Research Centre of Distillation Technology of China
(based in Tianjin University) for 7 years. He obtained my MSc degree in Process System Engineering in Cranfield
University in 2008 and from 2009 to 2012 he worked for BP as a lead process optimisation engineer in BP Zhuhai
chemical plant. He them moved to the University of Hull where he is currently a research associate on
process\energy system engineering and CCS. His main research directions include:
(1) modelling, simulation and optimization of power plants integrated with post-combustion carbon capture
process;
(2) techno-economic evaluation of CO2 transport pipeline network;
(3) developing key technologies for energy saving and emission reduction for large scale chemical and
metallurgy processes;
Javier Rodriguez
Process Systems Enterprise Ltd
Javier Rodríguez is a Senior Consultant at Process Systems Enterprise (PSE). He is the lead developer of physical
property models in gCCS, a modelling and simulation package for CCS systems. He played a key role in the
execution of the ETI CCS System Modelling Tool-kit Project, and greatly contributed to the design, development
and testing of gCCS model libraries and their productisation. Before joining PSE, Javier held a position as
research associate at Imperial College and worked on the development of SAFT models for amine-based
solvents relevant to carbon capture. Javier has a PhD from Imperial College London, where he worked on
developing a hybrid strategy to enhance state estimation for nonlinear lumped and distributed parameter
systems. His research focused on model-based state and parameter estimation techniques under a stochastic
setting. Prior to his PhD, Javier obtained his Masters in Chemical Engineering from the University of Oviedo,
Spain, in 2004.
Prof Roland Span
Ruhr-Universität Bochum
2006 - Chair of Thermodynamics, Faculty Mechanical Engineering, Ruhr-Universität Bochum, Germany
2002 - 2006 Chair of Thermodynamics and Energy Technologies, Faculty Mechanical Engineering, University of
Paderborn, Germany
2001 - 2002 Project and group leader in gas-turbine research, ALSTOM Power Technology Ltd., Baden,
Switzerland
1993 - 2000 Group leader “multiparameter equations of state”, Thermodynamics (chair: Prof. Dr.-Ing. W.
Wagner), Faculty Mechanical Engineering, Ruhr-Universität Bochum, Germany
1988 - 1993 Doctoral researcher, Thermodynamics (chair: Prof. Dr.-Ing. W. Wagner), Faculty Mechanical
Engineering, Ruhr-Universität Bochum, Germany
Prof J. P. Martin Trusler
Imperial College London
Martin Trusler is Professor of Thermophysics in the Department of Chemical Engineering at Imperial College
London. He gained his bachelor’s and PhD degrees in Chemistry from UCL, and held Lindermann Trust and
Ramsay Memorial Fellowships before joining the academic staff at Imperial in 1988. Martin’s research interests
focus on measurement and modelling of the thermophysical properties and phase behaviour of fluids, especially
under extreme conditions of temperature and pressure, with applications in CCS and oil/gas exploration and
production. He is a former editor of the Journal of Chemical Thermodynamics and the author of over 120 papers
in peer-reviewed journals.
Christopher John Wareing
University of Leeds
Dr Chris Wareing is a post-doctoral research fellow in the Department of Applied Mathematics at the University
of Leeds. He holds a cross-departmental post where he works with colleagues in the School of Chemical and
Process Engineering performing research into the transport methods that will be used in CO2 Capture and
Storage – a short-term way to stop the release of carbon dioxide into the atmosphere from emitters such as
power stations and mitigate man-made climate change. In his other role, he also supports academic researchers
through one-to-one guidance and by developing and delivering training courses on how to efficiently use the
University’s ARC supercomputing clusters to solve massively complicated problems that would on a single
computer take years to execute, but on the supercomputer can be performed in parallel in a matter of weeks or
days.
UK Carbon Capture and Storage Research Centre (UKCCSRC)
The UKCCSRC brings together over 1000 members including over 200 of the UK’s world-class CCS academics to
provide a national focal point for CCS research and development. The Centre is a virtual network where
academics, industry, regulators and others in the sector collaborate to analyse problems devise and carry out
world-leading research and share delivery, thus maximising impact. A key priority is supporting the UK economy
by driving an integrated research programme and building research capacity that is focused on maximising the
contribution of CCS to a low-carbon energy system for the UK.
The UKCCSRC is supported by the Engineering and Physical Sciences Research Council (EPSRC) www.epsrc.ac.uk
as part of the Research Councils UK Energy Programme, with additional funding from the Department of Energy
and Climate Change (DECC) www.decc.gov.uk for the UKCCSRC PACT Facilities www.pact.ac.uk
www.ukccsrc.ac.uk
RUHR-UNIVERSITÄT BOCHUM
Thermodynamic Property Models for
Transport and Storage of CO2
Roland Span
UKCCS Research Centre Meeting, York 2014
2 Span | UKCCS Research Centre Meeting | York 2014
Options for CCS
Pre-Combustion Process
(IGCC / NGRCC)
Category
Lignite
Fuel
Hard Coal
Cryogenic
O2 Supply
OTM / ITM
Physical Absorpt.
CO2 Separation
Chemical Absorpt.
H2 Membrane
CO2 Membrane
Integrated Process
(Oxy-Fuel) Cryogenic
OTM / ITM
Condensation
Chemical Looping
Post-Combustion Process
(exhaust gas cleaning) Chemical Absorpt.
Chilled Ammonia
Solid Adsorbents
Conditioning
Natural Gas
Lignite
Hard Coal
Natural Gas
Lignite
Hard Coal
Natural Gas
Membrane Reactor
Compression, Transport and Storage of CO2
3 Span | UKCCS Research Centre Meeting | York 2014
Multi-Disciplinary Property Research
Pre-Combustion Process
(IGCC / NGRCC)
Category
Lignite
Fuel
Hard Coal
Cryogenic
O2 Supply
OTM / ITM
Physical Absorpt.
CO2 Separation
Chemical Absorpt.
H2 Membrane
CO2 Membrane
Integrated Process
(Oxy-Fuel) Cryogenic
OTM / ITM
Condensation
Chemical Looping
Post-Combustion Process
(exhaust gas cleaning) Chemical Absorpt.
Chilled Ammonia
Solid Adsorbents
Natural Gas
Lignite
Hard Coal
Natural Gas
Lignite
Hard Coal
Natural Gas
Membrane Reactor
Conditioning Compression, Transport and Storage of CO2
Property models typical for chemical engineering
4 Span | UKCCS Research Centre Meeting | York 2014
Pre-Combustion Process
(IGCC / NGRCC)
Category
Lignite
Fuel
Hard Coal
Cryogenic
O2 Supply
OTM / ITM
Physical Absorpt.
CO2 Separation
Chemical Absorpt.
H2 Membrane
CO2 Membrane
Integrated Process
(Oxy-Fuel) Cryogenic
OTM / ITM
Condensation
Chemical Looping
Post-Combustion Process
(exhaust gas cleaning) Chemical Absorpt.
Chilled Ammonia
Solid Adsorbents
Natural Gas
Lignite
Hard Coal
Natural Gas
Lignite
Hard Coal
Natural Gas
Membrane Reactor
Conditioning Compression, Transport and Storage of CO2
Property models typical for geology / geo sciences
Multi-Disciplinary Property Research
5 Span | UKCCS Research Centre Meeting | York 2014
Pre-Combustion Process
(IGCC / NGRCC)
Category
Lignite
Fuel
Hard Coal
Cryogenic
O2 Supply
OTM / ITM
Physical Absorpt.
CO2 Separation
Chemical Absorpt.
H2 Membrane
CO2 Membrane
Integrated Process
(Oxy-Fuel) Cryogenic
OTM / ITM
Condensation
Chemical Looping
Post-Combustion Process
(exhaust gas cleaning) Chemical Absorpt.
Chilled Ammonia
Solid Adsorbents
Natural Gas
Lignite
Hard Coal
Natural Gas
Lignite
Hard Coal
Natural Gas
Membrane Reactor
Conditioning Compression, Transport and Storage of CO2
Property models typical for energy technologies
Multi-Disciplinary Property Research
6 Span | UKCCS Research Centre Meeting | York 2014
Pre-Combustion Process
(IGCC / NGRCC)
Category
Lignite
Fuel
Hard Coal
Cryogenic
O2 Supply
OTM / ITM
Physical Absorpt.
CO2 Separation
Chemical Absorpt.
H2 Membrane
CO2 Membrane
Integrated Process
(Oxy-Fuel) Cryogenic
OTM / ITM
Condensation
Chemical Looping
Post-Combustion Process
(exhaust gas cleaning) Chemical Absorpt.
Chilled Ammonia
Solid Adsorbents
Natural Gas
Lignite
Hard Coal
Natural Gas
Lignite
Hard Coal
Natural Gas
Membrane Reactor
Conditioning Compression, Transport and Storage of CO2
Multi-Disciplinary Property Research
Oxyflame – DFG Collaborative Research Centre
RUB with RWTH Aachen and TU Darmstadt
7 Span | UKCCS Research Centre Meeting | York 2014
Pre-Combustion Process
(IGCC / NGRCC)
Category
Lignite
Fuel
Hard Coal
Cryogenic
O2 Supply
OTM / ITM
Physical Absorpt.
CO2 Separation
Chemical Absorpt.
H2 Membrane
CO2 Membrane
Integrated Process
(Oxy-Fuel) Cryogenic
OTM / ITM
Condensation
Chemical Looping
Post-Combustion Process
(exhaust gas cleaning) Chemical Absorpt.
Chilled Ammonia
Solid Adsorbents
Natural Gas
Lignite
Hard Coal
Natural Gas
Lignite
Hard Coal
Natural Gas
Membrane Reactor
Conditioning Compression, Transport and Storage of CO2
Multi-Disciplinary Property Research
Develop a model / a set of models which …
• describes homogeneous states with high
(reference) accuracy
• consistently describes VLE / LLE equilibria
• consistently describes equilibria with solid phases
(ice, dry ice, hydrates)
8 Span | UKCCS Research Centre Meeting | York 2014
Thermodynamic Properties of Pure CO2
Span and Wagner (2003), fundamental EOS with 12 fitted coefficients (high technical quality)
Span and Wagner (1996), fundamental EOS with 42 fitted coefficients (reference quality)
9 Span | UKCCS Research Centre Meeting | York 2014
Helmholtz-model for mixtures (fundamental equation of state!)
Introduced independently by Lemmon & Tillner-Roth in mid 90’s
1
0 r r
m m m m
1 1 1 1
, , , ln , ,
N N N N
i oi i i oi i j ij ij
i i i j i
x x T x x x x F
The GERG-2008 Model by Kunz and Wagner
10 Span | UKCCS Research Centre Meeting | York 2014
Helmholtz-model for mixtures (fundamental equation of state!)
Introduced independently by Lemmon & Tillner-Roth in mid 90’s
Pure fluid equations of state (EOS)
1
0 r r
m m m m
1 1 1 1
, , , ln , ,
N N N N
i oi i i oi i j ij ij
i i i j i
x x T x x x x F
The GERG-2008 Model by Kunz and Wagner
11 Span | UKCCS Research Centre Meeting | York 2014
Helmholtz-model for mixtures (fundamental equation of state!)
Introduced independently by Lemmon & Tillner-Roth in mid 90’s
Pure fluid equations of state (EOS)
Mixing rules for reduced input parameters m and m
Two options:
• Mixing rules with four adjustable, binary specific parameters
• Combination rules
1
0 r r
m m m m
1 1 1 1
, , , ln , ,
N N N N
i oi i i oi i j ij ij
i i i j i
x x T x x x x F
The GERG-2008 Model by Kunz and Wagner
0.5
rm r , , , ,2
1 1 ,
( )with ( )
N Ni j
i j T ij T ij c i c j
i j T ij i j
x xTT x x T T
T x x
xx
3
m , , 2 1 3 1 31 1r r , , ,
1 1 1 1with
( ) ( ) 8
N Ni j
i j ij ij
i j ij i j c i c j
x xx x
x x
x x
12 Span | UKCCS Research Centre Meeting | York 2014
Helmholtz-model for mixtures (fundamental equation of state!)
Introduced independently by Lemmon & Tillner-Roth in mid 90’s
Pure fluid equations of state (EOS)
Mixing rules for reduced input parameters m and m
Binary excess functions
Two options:
• Binary specific excess function – Fij = 1, parameter in ij fitted
• Generalized excess function – Fij fitted, parameter in ij generalized
• Combination rules
1
0 r r
m m m m
1 1 1 1
, , , ln , ,
N N N N
i oi i i oi i j ij ij
i i i j i
x x T x x x x F
The GERG-2008 Model by Kunz and Wagner
, , ,
,
2
m m m m m m m m
1 1
( , ) expPol ij Pol ij Exp ij
k k k k
Pol ij
K K K
d t d t
ij k k k k k k
k k K
n n
13 Span | UKCCS Research Centre Meeting | York 2014
Helmholtz-model for mixtures (fundamental equation of state!)
Introduced independently by Lemmon & Tillner-Roth in mid 90’s
Pure fluid equations of state (EOS)
Mixing rules for reduced input parameters m and m
Four levels of accuracy, depending on available experimental data and relevance of the binary subsystem
• Only combination rules for Tr, r
• Mixing rules with four adjustable parameters for Tr, r
• Adjusted mixing rules & generalized excess function
• Adjusted mixing rules & binary specific excess function
• Combination rules
1
0 r r
m m m m
1 1 1 1
, , , ln , ,
N N N N
i oi i i oi i j ij ij
i i i j i
x x T x x x x F
The GERG-2008 Model by Kunz and Wagner
UNIQUE!
14 Span | UKCCS Research Centre Meeting | York 2014
Helmholtz-model for mixtures (fundamental equation of state!)
Introduced independently by Lemmon & Tillner-Roth in mid 90’s
Pure fluid equations of state (EOS)
Mixing rules for reduced input parameters m and m
1
0 r r
m m m m
1 1 1 1
, , , ln , ,
N N N N
i oi i i oi i j ij ij
i i i j i
x x T x x x x F
The GERG-2008 Model by Kunz and Wagner
Methane (CH4) n-Pentane (n-C5H12) Hydrogen (H2)
Nitrogen (N2) Isopentan (i-C5H12) Carbon monoxide (CO)
Carbon dioxide (CO2) n-Hexane (n-C6H14) Hydrogen sulphide (H2S)
Ethane (C2H6) n-Heptane (n-C7H16) Water (H2O)
Propane (C3H8) n-Octane (n-C8H18) Oxygen (O2)
n-Butane (n-C4H10) n-Nonane (n-C9H20) Argon (Ar)
Isobutane (i-C4H10) n-Decane (n-C10H22) Helium (He)
15 Span | UKCCS Research Centre Meeting | York 2014
5 mixtures: new excess functions
5 mixtures: new reducing parameters
EOS-CG – Improving GERG-2008 for CO2-Rich Mixtures
O 2
N 2
CO 2
CO
Ar
Binary specific excess function
r ij
Adjusted reducing functions for r and Tr
Lorentz-Berthelot combining rules for r and Tr
16 Span | UKCCS Research Centre Meeting | York 2014
Property Models for CO2-Rich Mixtures – EOS-CG
Example CO2 – Ar: Phase boundaries
Improvements compared to GERG-2008 at high pressure
17 Span | UKCCS Research Centre Meeting | York 2014
Example H2O – CO2: Phase Boundaries
Property Models for CO2-Rich Mixtures – EOS-CG
18 Span | UKCCS Research Centre Meeting | York 2014
Example H2O – CO2: Densities
Property Models for CO2-Rich Mixtures – EOS-CG
19 Span | UKCCS Research Centre Meeting | York 2014
Numerically stable (phase-equilibrium) algorithms available
Property Models for CO2-Rich Mixtures – EOS-CG
20 Span | UKCCS Research Centre Meeting | York 2014
Much Room for Improvement: IMPACTS
Binary systems for experimental work within IMPACTS
EOS
Ch
lori
ne
Hyd
roge
n C
hlo
rid
e
Die
than
ola
min
e
Mo
no
eth
ano
lam
ine
Me
than
ol
Am
mo
nia
Sulp
hu
r Tr
ioxi
de
Sulp
hu
r D
ioxi
de
Nit
roge
n D
ioxi
de
Nit
roge
n O
xid
e
Hyd
roge
n S
ulf
ide
Me
than
e
Hyd
roge
n
Car
bo
n M
on
oxi
de
Arg
on
Oxy
gen
Nit
roge
n
Wat
er
Carbon Dioxide
Water
Nitrogen
Oxygen
Argon
Carbon Monoxide
Hydrogen
Methane
Hydrogen Sulfide
Nitrogen Oxide
Nitrogen Dioxide
Sulphur Dioxide
Sulphur Trioxide
Ammonia
Methanol
Monoethanolamine Probably covered quite well by existing Helmholtz models
Diethanolamine Helmholtz models available, but accuracy unclear
Hydrogen Chloride Current work at RUB
Chlorine
Maj
or
Co
mp
on
en
tsM
ino
r C
om
po
ne
nts
NIST & WSU
21 Span | UKCCS Research Centre Meeting | York 2014
• CO2 in pipelines is liquid
• Pressure loss (leaks or flooding of
evacuated pipeline segments)
results in liquid / vapor system
• Further expansion leads to
formation of a dry ice / vapor
system at about 195 K
• For a description of the process
(enthalpy) flash calculation with
solid phase
• Similar effects for low tempera-
ture transport / capture
Consistent fundamental equation
for dry ice required!
Low-Temperature Phase Equilibria for CO2
22 Span | UKCCS Research Centre Meeting | York 2014
Phase Equilibria with Solid Phases – Dry Ice
Fundamental equations for solid CO2 (dry ice) developed
Jäger and Span (2010, 2012): Gibbs enthalpy as a function of pressure and temperature
Approach by Tillner-Roth (1998) adapted
Fitted only to data for solid CO2
Trusler (2011): Helmholtz energy as a function of molare volume and temperature
Fitted also phase equilibrium with adjacent fluid-phase
0 0 0
0
0 0 0
( , )( , ) ( , )d d ( , )d
pT Tp
p
T T p
c T pg p T h Ts c T p T T T v p T p
T
23 Span | UKCCS Research Centre Meeting | York 2014
Phase Equilibria with Solid Phases – Dry Ice
Intersection with Gibbs enthalpies from EOS for fluid states yields consistent SVE / SLE data (sublimation / melting pressure)
Allows for flash calculations into the melting / sublimation region
The property model used for the fluid phases has a significant impact!
Intersection with
Span and Wagner (1996)
Intersection with
Ely (1987)
24 Span | UKCCS Research Centre Meeting | York 2014
Phase Equilibria in the System CO2 / H2O
Solid H2O: Feistel and Wagner (2006)
Solid CO2: Trusler (2011) / Jäger and Span (2010, 2012)
Hydrates: Jäger, Vinš, Hrubý, and Span, R. (2013)
Fluid region:
H2O: Wagner and Pruss (2002)
CO2: Span and Wagner (1996)
Mixing rules: Gernert and Span (2013)
25 Span | UKCCS Research Centre Meeting | York 2014
• The model of Ballard und Sloan (2002) was chosen and
slightly modified:
,, , , ln 1 ,H
w J w i i J J
i J
T p f g T p RT v C T p f
Phase Equilibria with Solid Phases – Hydrates
26 Span | UKCCS Research Centre Meeting | York 2014
• The model of Ballard und Sloan (2002) was chosen and
slightly modified:
• Adjustable parameters of the model:
,, , , ln 1 ,H
w J w i i J J
i J
T p f g T p RT v C T p f
1 2, ,0 ,0,w wg h ,
Reference state Pressure dependence
of the molar volume
Potential-
parameters
Phase Equilibria with Solid Phases – Hydrates
fitted but almost
unchanged fitted to (T,p)-data generated from
experimental data & fluid model
consistency to reference
state of fluid model
27 Span | UKCCS Research Centre Meeting | York 2014
Phase Equilibria with Solid Phases – Hydrates
Accurate description of CO2 / H2O hydrate formation
Consistent to accurate VLE / LLE / homogeneous phase model
Intersection with EOS-CG yields
equilibrium temperatures with
deviations < 1 K to exp. data
LcH
VH
VIw
HIc
VLc
VLw
28 Span | UKCCS Research Centre Meeting | York 2014
Allowable Water Content in CO2
29 Span | UKCCS Research Centre Meeting | York 2014
Allowable Water Content in CO2
30 Span | UKCCS Research Centre Meeting | York 2014
TREND – Software Made Available
• TREND 1.1 was made available early in 2014
• Launch of TREND 2.0 is expected by the end of 2014
Preview
TREND 2.0
31 Span | UKCCS Research Centre Meeting | York 2014
Other Hydrates
• Hydrate models (consistent to accurate multiparam. mixture
models for fluid phase) will soon be published for water with
- Nitrogen
- Oxygen
- Argon
- Carbon monoxide
- Methane
- Ethane
- Propane
• A corresponding model for mixed hydrates is still pending
32 Span | UKCCS Research Centre Meeting | York 2014
Avoiding Inconsistencies – Injection of CO2
• Engineers working on pipelining of CO2 use GERG-2008 / EOS-CG
• Reservoir engineers use cubic EOS (+ hydrate / electrolyte models)
• Severe inconsistencies, e.g., for injection of CO2 (-rich mixtures)
Pipeline engineer delivers
500 to/h 577 m3/h at 19.6 MPa
Reservoir engineer receives
577 m3/h at 18.1 MPa 440 to/h
33 Span | UKCCS Research Centre Meeting | York 2014
Avoiding Inconsistencies
• CO2 in geological storage
Mixtures with brines instead of water
• CO2 capture from natural gas / two phase gas pipelines
Mixed hydrates
Mixtures with hydrate inhibitors
• CO2 scrubbing from flue gas / natural gas
Systems containing scrubbing agents
• CO2 ….
In the long run the main challenge will
be to ensure that property models used
in adjacent process steps are consistent
to accurate models applied for CO2 transport
34 Span | UKCCS Research Centre Meeting | York 2014
Thank you for your attention!
The author is grateful to all organizations that
contributed funding to the presented work, namely to
- E.ON for awarding an E.ON Research Award
- E.ON Ruhrgas for the contract "Calculation of Complex
Phase Equilibria"
- the federal government of Nordrhein Westfalen in
conjunction with EFRE for funding under contract
315-43-02/2-005-WFBO-011Z
- the European Commission for the contract
"Seventh Framework Program, Nr. 308809, IMPACTS“
- the DFG for the framework of the collaborative research
centre ”Oxyflame“
35 Span | UKCCS Research Centre Meeting | York 2014
The „Killer Application“
• If a CO2 pipeline leaks, the
mantle cools down drastically, the
material becomes brittle
• The crack propagates in both
directions
• The pressure loss propagates
with about speed of sound
• If the crack propagates faster
than speed of sound, small
cracks result in a disaster
The issue is a safety issue,
accurate properties required for
homogeneous, VLE, VLSE states
• Accurate properties are certainly
more important for other
applications, but who cares …
liquid CO2 in the pipeline
cold CO2 escapes
Pipeline wall
SINTEF
36 Span | UKCCS Research Centre Meeting | York 2014
International Cooperation
Focus on CO2-rich mixtures
• Measurement of thermodynamic properties (pvT, w)
• Measurement of phase equilibria (VLE, LLE, VLSE)
• Measurement of transport properties (viscosity)
• Accurate property models for CO2-rich mixtures
• Description of phase equilibria including solid phases
• Improvement of phase equilibrium algorithms
• Test of new property models for various applications
© 2014 Process Systems Enterprise Limited
CO2 Properties and EoS for pipeline engineering York 11 November 2014
J. Rodriguez, M. Calado, E. Dias, A. Lawal, N. Samsatli, A. Ramos, T. Lafitte, J. Fuentes, C. Pantelides
gSAFT: advanced physical properties for carbon capture and storage system modelling
THE ADVANCED PROCESS MODELING COMPANY
© 2014 Process Systems Enterprise Limited
Overview
gCCS whole-chain system modelling environment
ETI’s CCS system modelling tool-kit project
Challenges in providing physical properties for the systems downstream of the capture plant
gSAFT technology
Based on a predictive molecular equation of state
gSAFT for the compression, transmission and injection subsystems within gCCS
Application to typical CCS flowsheets
Using gCCS libraries
Conclusions
© 2014 Process Systems Enterprise Limited
gCCS whole-chain system modelling environment
ETI’s CCS system modelling tool-kit project
© 2014 Process Systems Enterprise Limited
The CCS System Modelling Tool-kit Project 2011-2014
Energy Technologies Institute (ETI)
gPROMS modelling platform & expertise
Project Management
~$5m project commissioned & co-funded by the ETI
Objective: “end-to-end” CCS modelling tool
© 2014 Process Systems Enterprise Limited
gCCS initial scope (2014/Q2)
Process models
Power generation
Conventional: pulverised-coal, CCGT
Non-conventional: oxy-fuelled, IGCC
Solvent-based CO2 capture
CO2 compression & liquefaction
CO2 transportation
CO2 injection in sub-sea storage
Materials models
cubic EoS (PR 78)
flue gas in power plant
Corresponding States Model
water/steam streams
SAFT-VR SW/ SAFT- Mie
amine-containing streams in CO2 capture
SAFT- Mie
near-pure post-capture CO2 streams
Open architecture allows incorporation of 3rd party models 5
© 2014 Process Systems Enterprise Limited
Physical properties for subsystems downstream of the capture plant
Challenges
© 2014 Process Systems Enterprise Limited
Physical properties for downstream of the capture plant
CO2 phase diagram
Physical properties for pure CO2 predicted very accurately by Span & Wagner EoS Span, Wagner. "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." Journal of physical and chemical reference data 25 (1996): 1509.
Best choices for CO2 transmission liquid-like density gas-like viscosity
© 2014 Process Systems Enterprise Limited
Physical properties for downstream of the capture plant
Challenges I: Impurities
From post-combustion (dry basis):
CO2 (>99%), N2 (<0.17%), O2 (<0.01%), SOx (10 ppmv), traces of Ar
From pre-combustion (dry basis):
CO2 (>95.6%), H2S (<3.4%), H2 (<3%), N2 (<0.6%), CO (<0.4%), Ar (<0.05%), CH4 (350 ppmv)
From oxyfuel (dry basis):
CO2 (>74.66%), N2 (<15%), Ar (<2.5%), O2 (<6.15%), SOx (<2.5%), traces of CO
…plus H2O
The presence of impurities significantly affects physical properties (densities, phase envelope, critical temperature and pressure,…)
impact on compressor/pump power, pipeline capacity, potential for hydrate formation & two phase flow, distance between booster stations…
© 2014 Process Systems Enterprise Limited
Physical properties for downstream of the capture plant
Challenges II : Wide range of conditions
Compression subsystem
Pressures Inlet 0.5 to 5 bara
Outlet 10 to 200 bara
Temperatures Inlet 20-41 °C
Outlet 40-130 °C
Transmission subsystem
Pressures 50-200 bara
Temperatures -5-40 °C
Compression Transmission
© 2014 Process Systems Enterprise Limited
Physical properties for downstream of the capture plant
Challenges III: Limited experimental data
Recent literature review of experimental data Li, Hailong, et al.
"PVTxy properties of CO2 mixtures relevant for CO2 capture, transport and storage: Review of available experimental data and theoretical models." Applied Energy 88.11 (2011): 3567-3579.
Limited range of conditions
Gaps for several binary mixtures
some mixtures (e.g. CO2-SO2) are very corrosive experimentation problematic
Very scarce data for ternaries and beyond
Working on solving this • Release of experimental data from several projects • Experimental plan at University of Nottingham for VLE measurements of near-
pure CO2 mixtures
© 2014 Process Systems Enterprise Limited
applied to mixtures of CO2, CO, H2O, Ar…..
small molecules single group each
Physical properties for downstream of the capture plant
Challenges
Impurities
Wide range of conditions
Limited experimental data
A predictive equation of state
is required
© 2014 Process Systems Enterprise Limited
gSAFT
A commercial implementation of the SAFT-γ Mie equation of state
© 2014 Process Systems Enterprise Limited
gSAFT
The Statistical Association Fluid Theory I
Molecular-based EOS are a very appealing alternative to more classical approaches, such as cubic EOS
The Statistical Association Fluid Theory (SAFT) is especially relevant for its ability to deal with complex fluids
SAFT-based EOS are rooted on statistical mechanics, so
they involve a limited number of parameters with a clear physical meaning
can be fitted to a limited amount of experimental data can predict phase behaviour and physical properties for a wide range of
conditions, including those far from the ones employed for parameter estimation
© 2014 Process Systems Enterprise Limited
gSAFT
The Statistical Association Fluid Theory II
PSE’s gSAFT is a commercial implementation of one of the most advanced SAFT-based EOS
SAFT- Mie, developed by Imperial College London
SAFT: Chapman, Gubbins, Jackson, Radosz, Ind. Eng. Chem. Res., 29, 1709 (1990)
SAFT-VR: Gil-Villegas, Galindo, Whitehead, Mills, Jackson, Burgess, J. Chem. Phys., 106, 4168 (1997)
SAFT-γ: Lymperiadis, Adjiman, Jackson, Galindo, Fluid Phase Equilib., 274, 85 (2008)
SAFT-γ Mie: Papaioannou, Lafitte, Avendaño, Adjiman, Jackson, Muller, Galindo, in preparation (2014)
© 2014 Process Systems Enterprise Limited
gSAFT
SAFT-γ Mie molecular model
Molecules are modelled as chains of spheres
Interactions
dispersion/repulsion (van der Waals) forces
hydrogen bonding via off-centre electron donor/acceptor (“association”) sites
ionic (coulombic) forces
Mie potential
( )R A
U r Cr r
Incr
eas
ing
stre
ngt
h
© 2014 Process Systems Enterprise Limited
gSAFT
Transferability of parameter values
The values of the interaction parameters
are assumed to be constant across
different molecules and mixtures
in different phases
under different temperatures, pressures and compositions
An approximation based on SAFT- Mie’s fundamental molecular basis supported by practical evidence
© 2014 Process Systems Enterprise Limited
gSAFT for near-pure CO2 streams in gCCS
© 2014 Process Systems Enterprise Limited
gSAFT for compression/transmission in CCS
The gSAFT Databank
H2O
H2S
CO2
CH3OH
CH4
Ar H2
SO2
O2
N2 CO
© 2014 Process Systems Enterprise Limited
gSAFT for compression/transmission in CCS
Comparisons: Pure CO2
© 2014 Process Systems Enterprise Limited
gSAFT for compression/transmission in CCS
Comparisons: Binary mixture H2O + CO2
Isotherms:
T=323.2 K (red)
T=333.2 K (yellow)
T=353.1 K (green)
CPA: Cubic+Association EoS
CO2 rich phase
Bamberger et al., 2000
© 2014 Process Systems Enterprise Limited
gSAFT for compression/transmission in CCS
Comparisons: Binary mixture H2O + CO2
CO2 rich phase – low temperatures
King et al., 1992
© 2014 Process Systems Enterprise Limited
gSAFT for compression/transmission in CCS
Comparisons: Binary CO2 + impurities
CO2+CH4
CO2+H2S
CO2+O2
© 2014 Process Systems Enterprise Limited
gSAFT for compression/transmission in CCS
Predictions: Bubble point of CO2+H2
Chapoy et al., 2011
© 2014 Process Systems Enterprise Limited
gSAFT for compression/transmission in CCS
Predictions: CO2+N2 densities
Brugge et al., 1997
© 2014 Process Systems Enterprise Limited
gSAFT for compression/transmission in CCS
Predictions: CO2+N2+Ar densities
University of Nottingham
COZOC project, University of Nottingham
T=303.15 T=313.15
T=313.15
© 2014 Process Systems Enterprise Limited
University of Nottingham measurements
Experimental plan
Mixture Name Component 1 Component 2 Component 3 x1 x2 x3
E1 CO2 N2 Ar 0.90 0.05 0.05
E2 CO2 N2 Ar 0.98 0.01 0.01
E3 CO2 Ar H2 0.95 0.02 0.03
gSAFT predictive accuracy being tested Specifically two-body interaction assumption
gSAFT model parameters will be readjusted if necessary
Dew-point and bubble-point lines for the following mixtures
© 2014 Process Systems Enterprise Limited
University of Nottingham measurements
VLE predictions
Pure CO2
CO2 + N2 (x=0.05 )+ Ar (x=0.05)
Dew-point line
© 2014 Process Systems Enterprise Limited
University of Nottingham measurements
VLE predictions
Pure CO2
Dew-point line
CO2 + N2 (x=0.01 )+ Ar (x=0.01)
Pure CO2
© 2014 Process Systems Enterprise Limited
University of Nottingham measurements
VLE predictions
Dew-point line
Pure CO2
CO2 + H2 (x=0.03)+ Ar (x=0.02)
© 2014 Process Systems Enterprise Limited
Application to typical CCS compression, transmission and injection flowsheets
Using gCCS libraries
© 2014 Process Systems Enterprise Limited
Compression
© 2014 Process Systems Enterprise Limited
gCCS model libraries
Compression
SourceCO2
ElectricDrive
CompressorSection
CoolerKODrum Dehydrator
SinkCO2
© 2014 Process Systems Enterprise Limited
Transmission & injection
© 2014 Process Systems Enterprise Limited
gCCS model libraries
Transmission and Injection
Well
PipeSegment
Emergency shutdown valve (ESD)
Gate Valve
Vertical Riser
CO2 Flowmeter
Distribution header
Choke Valve
Reservoir
Wellhead connection
© 2014 Process Systems Enterprise Limited
Case Study
Line-packing operation
• Assumed constant inlet flowrate at CO2Source (275tonnes CO2 per day)
• Gas phase injection with discharge pressure in CO2 sink ~ 21bara
• Total pipeline length – 132.2km • Pipeline is located offshore (in water)
System dynamics
Simulating line-packing operation: Sudden valve closure
© 2014 Process Systems Enterprise Limited
System dynamics
Simulating line-packing operation: Sudden valve closure
Case Study
Line-packing operation
Warning: Phase change identified!
© 2014 Process Systems Enterprise Limited
Conclusions
© 2014 Process Systems Enterprise Limited
Conclusions
Providing physical properties for a modelling tool for the systems downstream of the capture plant is challenging
Experimental data are limited
gSAFT is an implementation of a SAFT equation of state, perfectly suited to address these challenges
a parameter databank for the relevant components has been developed
excellent correlations and predictions have been demonstrated
gSAFT physical properties are already available within gCCS, an “end-to-end” modelling tool for CCS
for the simulation of compression/transmission/injection flowsheets
© 2014 Process Systems Enterprise Limited
Acknowledgements
ETI Tool-kit development consortium
Energy Technologies Institute
E.On
EdF
Rolls-Royce
CO2DeepStore
E4Tech
© 2014 Process Systems Enterprise Limited
PSE’s CCS Technology Team
Gerardo Sanchis Power plant
Mário Calado Compression Systems
Capture processes
Dr Adekola Lawal Capture processes
Transmission & injection
Dr Javier Rodríguez Capture processes
Physical properties (gSAFT)
Dr Tom Laffite
Physical properties (gSAFT)
Dr Nouri Samsatli Power plant
Product development
Dr Javier Fuentes Software development
Alfredo Ramos Technology Manager
Mark Matzopoulos Marketing & Business
Development
Prof Costas Pantelides Chief Technologist
© 2014 Process Systems Enterprise Limited
Advanced Process Modelling
World leaders in …
Software & services
Phase Behaviour and EoS Modelling
of the Carbon Dioxide-Hydrogen
System
J P Martin Trusler
Department of Chemical Engineering
Imperial College London, UK
11 November 2014
1
Acknowledgements
Researchers
David Vega Maza (now at University of Aberdeen, Scotland)
Olivia Fernandez Torres (now at University of Gelph, Canada)
Sponsors
Costain Energy & Process
Energy Technologies Institute
2
CO2 + H2 CO2 + N2
Background to Project
Original motivation for study: generation of VLE data to support pre-
combustion decarbonisation of fuel gases.
Examples include:
• processing of high-CO2 natural gases
• hydrogen production from synthesis gas
Technologies include:
• traditional solvent processes (e.g. MEA process)
• membrane separations
• cryogenic flash or distillation processes
• hybrids of the above
3
Costain’s Next Generation Capture Technology (NGCT)
• Process for electricity production from coal with 95% carbon capture
• Based on synthesis gas production and CO2 separation to yield H2
• Combustion/electricity production in a combined a cycle process
• NGCT achieves primary CO2 removal by low-temperature flash processes
Sour Water
Gas Shift Gasifier
Air Separation
Unit
CO2
Comp.
Combined
Cycle
O2
Coal
Claus Plant Sulphur CO2
12 MPa
Power
Syn
Gas
Syn
Gas
H2
Steam
CO2
H2S
Low Temp.
CO2 Removal
Syn
Gas H2S Removal
4
Phase Behaviour Project Plan
• Development of new VLE apparatus for studying CO2-rich mixtures at
low temperatures and high pressures
• Measurements of VLE (and also SVLE) for the CO2 + H2 binary system
• Pressures up to 16 MPa
• Temperatures approx. triple-point to critical point of CO2
• Fully analytical approach
• Modelling of VLE data in a form suitable for process design
• Measurement range covers two areas of interest:
• T < 270 K (cryogenic separations)
• T > 270 K (pipeline engineering)
5
Apparatus for VLE and SVLE Measurements
6
Emphasis on:
• uncertainty
• reliability
• automation
• safety
Working ranges:
• pressure to 20 MPa
• temperature 193 to 474 K
H2
N2
Gas 4
Gas 5
CO2
V-4
E-2
P-3RD-2
V-3
V-5
E-1
V-8 V-9
P-2 P-1
V-1
I-1
E-3V-7
Burner
V-2
RD-1
CG
T-1
E-4
Phase-Equilibrium Cell
• pmax = 200 bar
• V = 0.15 L
• Stainless steel
• Gold-plated, N2-filled
stainless-steel o-ring
• Magnetic stirrer
• Rolsi electro-
magnetic sampling
valves
7
Phase-Equilibrium Cell
8
GC Calibration for H2 and CO2
2,1 Component / 2 iAbAaρVn iiii
0.00
0.02
0.04
0.06
0.08
0.10
0 50000 100000 150000 200000
ρ/(
mol·dm
-3)
A / (25 μVs)
H2
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0 5000 10000 15000 20000
ρ/(
mol·dm
-3)
A / (25 μVs)
CO2
-0.002
-0.001
0.000
0.001
0.002
0 50000 100000 150000 200000
Δρ/(
mol·dm
-3)
A / (25 μVs)
H2
-0.002
-0.001
0.000
0.001
0.002
0 5000 10000 15000 20000
Δρ/(
mol·dm
-3)
A / (25 μVs)
CO2
Parameters: ai and bi for each
gas
9
Verification: Vapour Pressure of Pure CO2
Vapour pressure Deviations from Span-Wagner EoS
10
0
2
4
6
8
220 240 260 280 300
p/M
Pa
T/K
Experiment
Span-Wagner EoS
0.000
0.002
0.004
0.006
0.008
220 240 260 280 300
Δp/M
Pa
T/K
Experiment
Uncertainty
Overall standard uncertainties: u(T) = 0.01 K; u(p) = 0.003 MPa; u(x) = 0.011x(1 - x); u(y) = 0.011y(1 - y)
VLE Results for (CO2 + H2): Low Temperatures
11
0
2
4
6
8
10
12
14
16
0.0 0.2 0.4 0.6 0.8 1.0
p/M
Pa
x2, y2
0
2
4
6
8
10
12
14
16
0.0 0.2 0.4 0.6 0.8 1.0
p/M
Pa
x2, y2
0
2
4
6
8
10
12
14
16
0.0 0.2 0.4 0.6 0.8 1.0
p/M
Pa
x2, y2
0
2
4
6
8
10
12
14
16
0.0 0.2 0.4 0.6 0.8
p/M
Pa
x2, y2
T = 218.15 K T = 233.15 K
T = 243.15 K T = 258.15 K
VLE Results for (CO2 + H2): Pipeline Regime
12
0
2
4
6
8
10
12
14
16
0.0 0.2 0.4 0.6
p/M
Pa
x2, y2
0
2
4
6
8
10
12
14
16
0.0 0.1 0.2 0.3 0.4 0.5
p/M
Pa
x2, y2
0
2
4
6
8
10
12
14
16
0.0 0.1 0.2 0.3 0.4
p/M
Pa
x2, y2
0
2
4
6
8
10
12
14
16
0.00 0.05 0.10 0.15 0.20 0.25
p/M
Pa
x2, y2
(a)
T = 273.15 K T = 280.65 K
T = 288.15 K
T = 295.65 K
VLE Results for (CO2 + H2): ... & close to CO2 critical point
13
7.0
7.2
7.4
7.6
7.8
8.0
0.00 0.01 0.02
p/M
Pa
x2, y2
Modelling
• Standard Peng-Robinson Equation of State
• Quadratic mixing rules for a and b parameters:
• lij = constant
• kij = kij,1 + kij,2/T
• Up to three parameters to fit nine isotherms
• ai, bi from critical (or effective critical) constants and acentric factor
14
jiiji jj i aakxxa )1(
2/))(1( jiiji jj i bblxxb
Quantum Effects for H2
• Typical equations of state (including cubic EoS and SAFT models)
regress pure-component parameters against pure-component data:
• e.g. critical constants, vapour pressure, saturated liquid density
• Hydrogen is a quantum gas with large quantum-mechanical effects
below its critical temperature (Tc = 33 K)
• Quantum effects diminish at higher temperatures, so that H2 behaves
essentially classically at the present experimental temperatures
• Errors arise when EoS parameters are fitted to pure-H2 VLE data
because of the quantum effects that prevail under those conditions
• Thus for cubic EoS models (which base parameters on Tc, pc and ω)
effective critical constants fitted to virial coefficients are used
15
Fit for Virial Coefficients of H2
16
0
200
400
600
800
1000
-20
-10
0
10
20
50 150 250 350 450
C/(
cm
6 m
ol-2
)
B/(
cm
3 m
ol-1
)
T/K
Parameters Tc/K pc/MPa ω
True 33.15 1.296 -0.219
Effective 31.76 1.276 -0.063
Critical Constants of Normal Hydrogen
Modelling (CO2 + H2) with Peng-Robinson Equation
• Objective function based on composition deviations at given T and p:
• Up to three parameters: kij,1, kij,2 and lij
• Common approach is to set lij = 0
• ... but this leads to a relatively poor fit for CO2 + H2
• Fits with lij ≠ 0 lower objective function by a factor of 4
• Global fit to all isotherms (excluding a few near-critical states):
S = 0.004
17
N
i
i,ii,i yyxxN
S1
2calc,,22
2calc,,22
2 )()(1
Experiment vs Model for (CO2 + H2) at Low Temperatures
18
0
2
4
6
8
10
12
14
16
0.0 0.2 0.4 0.6 0.8 1.0
p/M
Pa
x2, y2
0
2
4
6
8
10
12
14
16
0.0 0.2 0.4 0.6 0.8 1.0
p/M
Pa
x2, y2
0
2
4
6
8
10
12
14
16
0.0 0.2 0.4 0.6 0.8 1.0
p/M
Pa
x2, y2
0
2
4
6
8
10
12
14
16
0.0 0.2 0.4 0.6 0.8
p/M
Pa
x2, y2
T = 218.15 K T = 233.15 K
T = 243.15 K T = 258.15 K
Experiment vs Model for (CO2 + H2) at PipelineTemperatures
19
0
2
4
6
8
10
12
14
16
0.0 0.2 0.4 0.6
p/M
Pa
x2, y2
0
2
4
6
8
10
12
14
16
0.0 0.1 0.2 0.3 0.4 0.5
p/M
Pa
x2, y2
0
2
4
6
8
10
12
14
16
0.0 0.1 0.2 0.3 0.4
p/M
Pa
x2, y2
0
2
4
6
8
10
12
14
16
0.00 0.05 0.10 0.15 0.20 0.25
p/M
Pa
x2, y2
(a)
T = 273.15 K T = 280.65 K
T = 288.15 K
T = 295.65 K
Experiment vs Model for (CO2 + H2)
20
7.0
7.2
7.4
7.6
7.8
8.0
0.00 0.01 0.02
p/M
Pa
x2, y2
General conclusions:
• Model with 3-parameters provides a reasonable global fit
• Fails near to the critical locus (as expected)
• Not within experimental uncertainty – especially bubble curves at higher temperatures
• Improved fits for restricted regions can be obtained
Limited Model for Pipeline Applications
• Same Peng-Robinson model but parameters fitted:
• T ≥ 273 K
• xH2 ≤ 0.06 with coexisting values of yH2
• Much improved representation with S = 0.002
21
Phase Envelopes
(0.95 CO2 + 0.05 H2): ,
(0.98 CO2 + 0.02 H2): ,
PR model:
Critical locus:
CO2 VP curve:
Summary
Experiment:
• New equipment constructed and validated
• High-quality calibrations and low overall uncertainties
• (CO2 + H2) measured at temperatures between the triple point and
critical points of CO2
Modelling:
• ‘Standard’ Peng-Robinson equation used
• ‘Effective’ critical constants for H2
• Quadratic mixing rules for both a and b parameters
• Provides a fair global fit
• Local fit for pipeline regime gives a good representation of the data
22
PE EoS
23
)()(
)(
mmmm bVbbVV
Ta
bV
RTp
c2
c /)()(457235.0 pTαRTa
2c2 /126992.054226.137464.01)( TTωωTα
cc /077796.0 pRTb
jiiji jj i aakxxa )1( 2/))(1( jiiji jj i bblxxb
Single substance: Mixture:
Understanding and predicting CO2 properties
Richard Graham Tom Demetriades, Alex Cresswell, Martin Nelson,
Richard Wilkinson and Simon Preston School of Mathematical Sciences, University of Nottingham.
Overview!
•Parametric equations of state (pressure explicit)
•Non-parametric EoS (pressure explicit or free energy formulation).
•Molecular simulations
Overview!
•Parametric equations of state (pressure explicit)
•Non-parametric EoS (pressure explicit or free energy formulation).
•Molecular simulations!
•Uncertainty quantification
Potential applications: Avoiding pipeline issues
Two-phase flow
Vapour-solid mix
JetSolid entrainment
Warm air entrainment Solar heating
Solid sublimation
Snow/ dry ice Ground heat flux
Pipe rupture
Coexistence - Impurities
Coexistence - Impurities
Liquid
Gas
xG: Gas composition vG: Gas volume
xL: Liquid composition vL: Liquid volume
Coexistence - Impurities
CO2+N2 dataLi
quid
Gas
Molar Volume (litres/mol)
Pres
sure
(M
Pa)
Liqu
id
Gas
Pres
sure
(M
Pa)
Mole fraction of impurity
Liquid
Gas
xG: Gas composition vG: Gas volume
xL: Liquid composition vL: Liquid volume
Uncertainty quantification
Uncertainty quantification
Uncertainty quantification
Economic recovery!
Uncertainty quantification
Huge uncertainty!
Uncertainty quantification
Must account for uncertainty due to: •Incomplete data •Measurements errors •Model imperfection
Huge uncertainty!
R
A generalised equation of state
Peng-Robinson
This work
R
A generalised equation of state
Peng-Robinson
This work
Higher order terms enable a longer plateau and improved critical volume
R
A generalised equation of state
Peng-Robinson
This work
Higher order singularity provides a sharper ‘liquid’ region
Higher order terms enable a longer plateau and improved critical volume
10-4 10-3
Molar volume [m^3/mol]
2
4
6
8
10
12
Pres
sure
[MPa
]
294K
Fitting methodFitting criterion
10-4 10-3
Molar volume [m^3/mol]
2
4
6
8
10
12
Pres
sure
[MPa
]
294K
Fitting methodFitting criterion
10-4 10-3
Molar volume [m^3/mol]
2
4
6
8
10
12
Pres
sure
[MPa
]
294K
Fitting methodFitting criterion
10-4 10-3
Molar volume [m^3/mol]
2
4
6
8
10
12
Pres
sure
[MPa
]
294K
Fitting methodFitting criterion
10-4 10-3
Molar volume [m^3/mol]
2
4
6
8
10
12
Pres
sure
[MPa
]
294K
Fitting method
Numerically minimise the sum of these 4 quantities over the parameters a...g
Fitting criterion
MCMC: an example
θ1
θ2
2 4 6 8 10 12 14
4
6
8
10
12
14
16
18
The search algorithm explores the fitting criterion, spending more time in regions of good fit.
Markov-Chain Monte-Carlo: an example
MCMC: an example
θ1
θ2
2 4 6 8 10 12 14
4
6
8
10
12
14
16
18
The search algorithm explores the fitting criterion, spending more time in regions of good fit.
Markov-Chain Monte-Carlo: an example
MCMC: an example
θ1
θ2
2 4 6 8 10 12 14
4
6
8
10
12
14
16
18
The search algorithm explores the fitting criterion, spending more time in regions of good fit.
Markov-Chain Monte-Carlo: an example
MCMC: an example
θ1
θ2
2 4 6 8 10 12 14
4
6
8
10
12
14
16
18
The search algorithm explores the fitting criterion, spending more time in regions of good fit.
Markov-Chain Monte-Carlo: an example
MCMC: an example
θ1
θ2
2 4 6 8 10 12 14
4
6
8
10
12
14
16
18
The search algorithm explores the fitting criterion, spending more time in regions of good fit.
Markov-Chain Monte-Carlo: an example
MCMC: an example
2 4 6 8 10 12 14
4
6
8
10
12
14
16
18
θ1
θ2
The result is samples of the probability distribution of the parameters
Markov-Chain Monte-Carlo: an example
Predictions (pure CO2)R
10-4 10-3
Molar volume [m^3/mol]
2
4
6
8
10
12
Pre
ssur
e [M
Pa]
304.3K (Tc)294K285K
Predictions (pure CO2)R
10-4 10-3
Molar volume [m^3/mol]
2
4
6
8
10
12
Pre
ssur
e [M
Pa]
304.3K (Tc)294K285K
Predictions (pure CO2)
10-4
Molar volume [m^3/mol]
4
6
8
Pre
ssur
e [M
Pa]
Coexisting liquidCoexisting vapour
R
Mixture modelling CO2+N2
Introduction to non-parametric methods
[6]
Introduction to non-parametric methods
•Model for pressure against volume, as with an equation of state. •However, no need to specify terms or parameters •Model ‘learns’ the P(v) functional form from the measurements [6]
0 0.2 0.4 0.6 0.8 1x0
0.2
0.4
0.6
0.8
1
f(x)
Introduction to non-parametric methods
•Model for pressure against volume, as with an equation of state. •However, no need to specify terms or parameters •Model ‘learns’ the P(v) functional form from the measurements
•Basic examples include splines and other interpolation techniques•Modern implementations are significantly more sophisticated
[6]
0 0.2 0.4 0.6 0.8 1x0
0.2
0.4
0.6
0.8
1
f(x)
Gaussian processes
a) Generate random functions from a distribution that favours smooth functions
0 0.2 0.4 0.6 0.8 1x0
0.2
0.4
0.6
0.8
1
f(x)
Gaussian processes
a) Generate random functions from a distribution that favours smooth functions
b) Keep only the functions that pass through the data points
Mean of accepted functions = Model Variance of accepted functions = Uncertainty quantification
0 0.2 0.4 0.6 0.8 1x0
0.2
0.4
0.6
0.8
1
f(x)
Data Mean
Variance
0.2 0.4 0.6 0.8 1.0
−2
−1
01
2
volume
pre
ssu
re
0.2 0.4 0.6 0.8 1.0
−2
−1
01
2
volume
pre
ssu
re
A Gaussian process for pure CO2P
ress
ure/
(Crit
ical
Pre
ssur
e)
Molar volume/(Ideal gas volume)
Temperature=290K
CO2 data Gaussian Process mean. 95% confidence interval Individual Gaussian Processes
0.2 0.4 0.6 0.8 1.0
−2
−1
01
2
volume
pre
ssu
re
0.2 0.4 0.6 0.8 1.0
−2
−1
01
2
volume
pre
ssu
re
A Gaussian process for pure CO2P
ress
ure/
(Crit
ical
Pre
ssur
e)
Molar volume/(Ideal gas volume)
Temperature=290K
CO2 data Gaussian Process mean. 95% confidence interval Individual Gaussian Processes
Gaussian Process accurately captures the data
0.2 0.4 0.6 0.8 1.0
−2
−1
01
2
volume
pre
ssu
re
0.2 0.4 0.6 0.8 1.0
−2
−1
01
2
volume
pre
ssu
re
A Gaussian process for pure CO2P
ress
ure/
(Crit
ical
Pre
ssur
e)
Molar volume/(Ideal gas volume)
Temperature=290K
CO2 data Gaussian Process mean. 95% confidence interval Individual Gaussian Processes
Gaussian Process accurately captures the data
Uncertainty is only significant in the coexistence region
0.2 0.4 0.6 0.8 1.0
−2
−1
01
2
volume
pre
ssu
re
0.2 0.4 0.6 0.8 1.0
−2
−1
01
2
volume
pre
ssu
re
A Gaussian process for pure CO2P
ress
ure/
(Crit
ical
Pre
ssur
e)
Molar volume/(Ideal gas volume)
Temperature=290K
CO2 data Gaussian Process mean. 95% confidence interval Individual Gaussian Processes
Gaussian Process accurately captures the data
Uncertainty is only significant in the coexistence region
Generalisation to mixtures is ongoing
Molecular simulationComputer model of individual molecules within a small box of fluid.
Can predict: •Pressure-‐volume •Coexistence •Effect of impurity •Most other quanBBes of interest
[7]
Molecular simulationComputer model of individual molecules within a small box of fluid.
Can predict: •Pressure-‐volume •Coexistence •Effect of impurity •Most other quanBBes of interest
Can be used where experiments are unavailable?
[7]
Molecular simulationComputer model of individual molecules within a small box of fluid.
Can predict: •Pressure-‐volume •Coexistence •Effect of impurity •Most other quanBBes of interest
Can be used where experiments are unavailable?
Can be used to derive an EquaBon of State?
[7]
Gibbs ensemble simulaBonsGas
Liquid
Gibbs ensemble simulaBonsTwo simulaBon boxes, represenBng coexisBng phases
Gas
Liquid
Gibbs ensemble simulaBonsTwo simulaBon boxes, represenBng coexisBng phases
The system approaches equilibrium by making a series of moves, consistent with staBsBcal mechanics
Once in equilibrium, the system predicts the coexistence properBes !
Gas
Liquid
ParBcle displacement Volume change
ParBcle transfer
�23
Molecular force-fields
�23
Molecular force-fields•All physical proper-es are ulBmately determined by interac-ons between molecules•Force-‐fields that describe these interacBons are a key input to simula-ons
�23
Molecular force-fields•All physical proper-es are ulBmately determined by interac-ons between molecules•Force-‐fields that describe these interacBons are a key input to simula-ons•InteracBons of CO2 with itself and with impuri-es must be specified
!
Semi-empirical forcefields CO2+N2
Semi-empirical forcefields CO2+N2
Simulations using literature force fields
Semi-empirical forcefields CO2+N2
Simulations after optimising the force field
Simulations using literature force fields
Bubble point comparison CO2 + 5%H2
Phase boundary measurements by Jie Ke, Mike
George et al
Two phase re
gion
Simulation aids EoS development
Liqu
id
Gas
Molar Volume (litres/mol)
Pres
sure
(M
Pa)
Liqu
id
Gas
Pres
sure
(M
Pa)
Mole fraction of impurity
Liquid
Gas
xG: Gas composition vG: Gas volume
xL: Liquid composition vL: Liquid volume
Two phase re
gion
Simulation aids EoS development
Liqu
id
Gas
Molar Volume (litres/mol)
Pres
sure
(M
Pa)
Liqu
id
Gas
Pres
sure
(M
Pa)
Mole fraction of impurity
Liquid
Gas
xG: Gas composition vG: Gas volume
xL: Liquid composition vL: Liquid volume
Two phase re
gion
Simulation aids EoS development
Liqu
id
Gas
Molar Volume (litres/mol)
Pres
sure
(M
Pa)
Liqu
id
Gas
Pres
sure
(M
Pa)
Mole fraction of impurity
Liquid
Gas
xG: Gas composition vG: Gas volume
xL: Liquid composition vL: Liquid volume
Ab initio force fields CO2+H2
Quantum Chemistry calculations of CO2-H2 interaction
Gaussian Process fit for use in simulations
+
Ab initio force fields CO2+H2
Quantum Chemistry calculations of CO2-H2 interaction
Force field computed from first principles
Potential for accurate predictions without data fitting⇒
Gaussian Process fit for use in simulations
+
Making it all work together!
•Parametric equations of state
•Non-parametric EoS
•Semi-empirical molecular simulation
•Ab-initio molecular simulation
Making it all work together!
•Parametric equations of state
•Non-parametric EoS
•Semi-empirical molecular simulation
•Ab-initio molecular simulation
Making it all work together!
•Parametric equations of state
•Non-parametric EoS
•Semi-empirical molecular simulation
•Ab-initio molecular simulation
• Fast, flexible models for computational studies • Fit to experiments, simulation data more advanced
EoS
Making it all work together!
•Parametric equations of state
•Non-parametric EoS
•Semi-empirical molecular simulation
•Ab-initio molecular simulation
• Fast, flexible models for computational studies • Fit to experiments, simulation data more advanced
EoS
• Rigorous uncertainty quantification - optimise choice of experiments
• (Somewhat) expensive but very accurate EoS
Making it all work together!
•Parametric equations of state
•Non-parametric EoS
•Semi-empirical molecular simulation
•Ab-initio molecular simulation
• Fast, flexible models for computational studies • Fit to experiments, simulation data more advanced
EoS
• Rigorous uncertainty quantification - optimise choice of experiments
• (Somewhat) expensive but very accurate EoS
• Accurate treatment of temperature variation • Completes coexistence measurements to help EoS fitting
Making it all work together!
•Parametric equations of state
•Non-parametric EoS
•Semi-empirical molecular simulation
•Ab-initio molecular simulation
• Fast, flexible models for computational studies • Fit to experiments, simulation data more advanced
EoS
• Rigorous uncertainty quantification - optimise choice of experiments
• (Somewhat) expensive but very accurate EoS
• Accurate treatment of temperature variation • Completes coexistence measurements to help EoS fitting
• Most physically realistic but also most expensive. • Can augment or replace experiments
1
Impact of Equation of State on
Simulating CO2 Pipeline Decompression
Dr Solomon Brown
UCL
CO2 Properties and EoS for Pipeline Engineering
11 August 2013, Athens, Greece
Structure
2
1. Background
2. Equations of State
3. Impact on simulation of pipeline decompression
4. Conclusions
4
A ductile fracture will come
to rest when the fluid
pressure at the crack tip,
Pt falls below the Crack
Arrest Pressure, Pa.
Running fractures
represent a threat to CO2
pipelines
Pipeline ductile fracture
5
Pressurised CO2
Rupture
plane: 1 atm
• At the rupture plane the fluid is exposed to ambient air
• Following the rupture, the rarefaction wave starts propagating along the
pipe at the speed of sound
• The vapour phase emerges in the expansion wave reducing the mixture
speed of sound
• Due to rapid cooling of the fluid in the decompression wave, the solid
phase may also be released from the pipe
Pipeline decompression
7
Generalized cubic EoS
3 * * 2 * *2 * *2 * * *2 *3
* *
2
(1 ) ( ) 0
,
z B uB z A B uB uB z A B B B
aP bPA B
RT RT
EoS
RK
u=1, w=0
SRK
u=1, w=0
PR
u=2, w=-1
PR/G
u=2, w=-1
Cubic Equations of State
8
( , ) ( , ) ( , ) ( , ) ( , ) ( , ) ( , ) ( , ) ( , )res ideal ref disp hs chain assoc dispa T a T a T a T a T a T a T a T a T
RT RT RT RT RT RT RT RT RT
The SAFT (Statistical Associating Fluid Theory) equation of state is written as a summation of residual Helmholtz free energy terms that occur due to different types of molecular interactions in the system under study.
Hard sphere term: Chain term: Association term: Dispersion term:
22
1
34
n
nn
RT
Ahs
4
1
9
1i j
ji
ij
disp n
kT
uD
RT
A
31
5.01ln1
n
nm
RT
Achain
M
A
AA
assoc
MX
XRT
A
1 2
1
2ln
Carnahan-Starling EoS for hard spheres based on Wertheim’s TPT1 based on TPT1
Alder equation from molecular dynamics
SAFT and PC-SAFT Equations of State
Structure
9
1. Background
2. Equations of State
3. Impact on simulation of pipeline decompression
10
Saturated density predictions
Saturated liquid and vapour density predictions
using the various Equations of State
11
SAFT PC-SAFT RK PR SRK Yokozeki
Density 1.81 1.14 6.25 4.89 5.97 3.35
Cv 6.37 3.77 9.96 3.94 8.80 58.75
Cp 11.09 3.53 10.35 4.54 3.35 28.76
Sound Speed 6.73 3.26 16.65 13.53 11.86 15.39
Joule-Thomson 113.11 44.20 75.70 102.08 66.65 85.62
Accuracy of derivative properties
Comparison of the predictive accuracy of EoS (AAD%) for
important derivative properties.
Saturated vapour phase speed of sound
12
Saturated liquid phase speed of sound
Speed of Sound predictions
13
Pressure history during decompression
13.5 m from open end
Pressure history during decompression
18 m from open end
Vapour phase decompression
14
0
2
4
6
8
10
12
14
16
0 0.5 1 1.5 2 2.5 3
Clo
sed E
nd P
ressure
(M
Pa)
Time (s)
SRK
SAFT
PC-SAFT
Experimental
Liquid
Pressure history during decompression 143 m
from open end
Liquid phase decompression
Structure
1
5
1. Background
2. Equations of State
3. Impact on simulation of pipeline decompression
4. Conclusions
16
• Accuracy of speed of sound in liquid predicted by even complex EoS remarkably low
• This greatly affects the modelling of the decompression wave, given that its front moves at the speed of sound
• Very little experimental data for dense phase speed of sound available for development of better EoS
• Results presented relate to pure CO2 only, the little impure CO2 data we have indicates that the same observations are true
17
Acknowledgements and Disclaimer
The research leading to the results described in this
presentation has received funding from the European
Union 7th Framework Programme FP7-ENERGY-2012-1-
2STAGE under grant agreement number 309102.
The presentation reflects only the authors’ views and the
European Union is not liable for any use that may be
made of the information contained therein.
Study of the CO2 Pipeline Network Planned in the Humber Region of the UK:
Simulation-based Techno-economic Evaluation for
Optimal Design
School of Engineering
University of Hull
11th NOV 2014
Xiaobo Luo, Meihong Wang
in collaboration with Ketan Mistry, Russell Cooper
OUTLINE
Pipeline Scheme in Humber Region
Work Package Overview
Techno-economic evaluation
Methodology
Evaluation of compression
Evaluation of trunk pipeline
Whole pipeline system
Findings
Work Package Overview
CO2 transportation pipeline network
Optimal design and operation by using process engineering system techniques
Aspen HYSYS gPROMS APEA
(Aspen Process Economic Analyser)
Dynamic simulation
Steady state simulation
Economic evaluation
Simulation-based techno-economic evaluation for Optimal Design
Methodology-model development
EOS selection in the literature
Span and Wanger ( for pure CO2)
GERG (for CO2 and impurities)
Peng-Robinson (for CO2 and impurities)
SAFT (for CO2 and impurities)
Equation of state (EOS) selection
Table 1. EOS used in published studies
Papers/studies EOS used
Hein et al. 1985 Soave-Redlich-Kwong (SRK) equation
Hein et al. 1986 Peng-Robinson (PR) equation for CO2 mixture
Zhang et al. 2006 Peng-Robinson (PR) equation with Boston -Mathias modification for CO2 mixture
Seevam et al. 2008 Peng-Robinson (PR) equation
Mahgerefteh et al. 2008 Peng-Robinson (PR) equation
E.ON's report , 2010 Span and Wagner EOS for pure CO2
Nimtz et al. 2010 Span and Wagner EOS for pure CO2
Munkejord et al. 2010 Soave–Redlich–Kwong EOS
Liljemark et al.2011 Span and Wagner EOS for pure CO2 and GERG-2004 for the CO2 mixtures
Klinkby et al. 2011 Span and Wagner EOS for pure CO2
Chaczykowski et al. 2012 GERG-2004 for CO2 mixture
An entry specification was agreed to be 96 mole% CO2 and a mixture of nitrogen, oxygen, hydrogen, argon and methane with hydrogen limited to 2.0 mole% and oxygen limited to 10 ppmv.
Methodology-model development
E.ON’s report (2010) show PR EOS is not very accurate in the near-critical region. In this T/P range(4oC - 20oC/ 101 bar -150 bar) in this study, the deviation of pure CO2 density is from -4.8% to 0.1% in E.ON’s report.
Peng-Robinson with calibrated binary interaction parameters
AAD% between experimental data and PR EOS for corresponding kij values
kij Bubble pressure Liquid volume
Reference Temperature (K)
Pressure (Mpa)
AAD% Temperature (K)
Pressure (Mpa)
AAD%
CO2 - N2 -0.007 220-301 1.4-16.7 3.73 209-320 1.4-16.7 1.54 Diamantonis et al. (2013)
Li & Yan (2009)
CO2 - Ar 0.141 288 7.5-9.8 2.32 288 2.4-14.5 1.83 Diamantonis et al. (2013)
CO2 – H2 0.1470 290.2 5.0-20.0 5.6% - - - Foster et al.
Methodology-model development
Model flow sheet in Aspen HYSYS
CO2 from Drax
CO2 from Don Valley
Common onshore pipeline Common offshore pipeline
Pump station
Mid-compressor train for Don Valley
Compressor train
Model validation by comparing the results of PIPEFLO®
no available operating and experimental data
PIPEFLO® is used for the concept design of the project
GERG-2008 EOSwas used in PIPEFLO® for the project
Entry pressure at
White Rose
Entry pressure at
Don Valley
DP of mid-booster for
Don Valley
DP of pump
station
Arrival
pressure
barg barg bar bar barg
Aspen HYSYS® 119.50 34.0 86.92 43.0 125.0
PIPEFLO® 119.20 34.0 86.70 42.4 125.0
Relative difference 0.25% - 0.25% 1.40% -
Methodology-economic evaluation
Economic evaluation using APEA
Aspen Process Economic Analyzer (APEA)
CAPEX • total capital investment cost (capital return factor is 0.15 for annualized capital cost)
Equipments purchase Engineering Construction others during project implement
OPEX •Fixed OPEX
O&M cost (per year) •Available OPEX
Energy and utilities cost (per year)
Diameter calculation in different
correlation methods
Input information
Simulations of trunk pipelines
Economic evaluations of trunk pipelines
Comparison with the literature and analysis
Simulation of whole pipeline network
Compression technology analysis
Simulation of compression train
Economics evaluation of compression train
Comparison with the literature and analysis
Comparison of compression options
to select optimal option
Comparison of pipeline options to
select optimal option
Economics evaluation of whole pipeline
network
Evaluation of CO2 compression
Table 5. Compression technology options and their process definition
Option Unit Base Case C1 C2 C3 C4
Description
Centrifugal
5 stage with
4
intercoolers
Centrifugal 16 stage 4
intercoolers
8 stage
centrifugal
geared with 7
intercoolers
6 stage integrally
geared with 5
intercoolers to 20
oC +pumping
6 stage integrally
geared with 5
intercoolers to 38
oC +pumping
Capacity t/h 245 245 245 245 245
Suction pressure MPa 0.101325 0.101325 0.101325 0.101325 0.101325 Suction temp. oC 20 20 20 20 20
Pumping suck
pressure MPa - - - 8.0 8.0 Pumping suck temp. oC - - - 20 20
Exit pressure MPa 13.5 13.5 13.5 13.5 13.5
Stage - 5 16 8 6 6
Isentropic efficiency % 75 75 75 75 75
Interstage cooler exit
temperature oC 20 38 38 20 38 Last stage exit temp. oC 20 20 20 20 20
Base case Case 1
Case 2 Case 4 Case 3
Compression technology options and their process definition
Evaluation of CO2 compression
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
Base case C1 C2 C3 C4
Brea
k do
wn
annu
al c
osts
(M€2
012/
a)
Options
Annual Energy and utilities cost
Annual O&M cost
Annualized Capital cost
0
2
4
6
8
10
12
Basecase
C1 C2 C3 C4
Ann
ual c
ost (
M€/
a)
McCollum and Ogden (2006)
0
2
4
6
8
10
12
Basecase
C1 C2 C3 C4
Ann
ual c
ost (
M€/
a)
IEA GHG (2002)
0
2
4
6
8
10
12
Basecase
C1 C2 C3 C4
Ann
ual c
ost (
M€/
a)
This study
Annual O&M cost
Annual capital cost
The comparison of levelized cost of different cost model in the literature
Comparison of annual costs of different compression options
Evaluation of trunk pipelines
Hydraulic performance and energy requirement of trunk pipelines in different
diameters
The calculation results of different diameter calculation methods in literature
Common onshore pipeline Common offshore pipeline
Pump station
Pipeline
diameter
Actual initial
velocity
Pressure drop of
onshore pipeline
Pressure drop of
offshore pipeline
Boosting pressure
of pump station
Energy required of
pump station (inch) (m/s) (bar) (bar) (bar) (kWh)
28 1.08 5.9 10.0 5.9 301.5
24 1.49 13.5 20.6 24.1 1243
22 1.81 22.1 32.2 44.3 2305
Diameter calculation method
Calculated
diameter Velocity
Selected diameter
in APEA
Unit (m) (m/s) DN (inch)
Velocity based equation
0.699 1.0 28
0.5713 1.5 24
0.4948 2 20
Hydraulic equation 0.5262 1.77 22
Extensive hydraulic equation 0.6173 1.29 24
McCoy and Rubin model 0.5672 1.52 22
Evaluation of trunk pipelines
0.00
20.00
40.00
60.00
80.00
100.00
120.00
28 in. 24 in. 22 in.
Bre
ak d
ow
n a
nn
ual
co
sts
(M€
20
12
/a)
Pipeline diamter
Annual energy and utilities cost
Annual O&M cost
Annual capital cost
Comparison of capital cost of different cost models in the literature
Annual cost comparison for different diameters of the pipelines
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
22 in. 24 in. 28 in.
Ca
pit
al
cost
(M€
/k
m)
Pipeline diamter
This study
IEA GHG, 2002
MaCollum and Ogden, 2006
McCoy and Rubin, 2008
Piessense et al., 2008
Van Den Broek et al., 2010
Overall cost of whole pipeline network
Comparison of levelized cost of the optimal case and COCATE project (Roussanaly et al., 2013)
Optimal case
COCATE project
Levelized energy and utilities cost €/t-CO2 7.6 8
Levelized CAPEX of the trunk pipeline €/t-CO2 6.0 5.5
Levelized CAPEX of collecting system €/t-CO2 4.4 0.2
Levelized O&M cost €/t-CO2 1.0 2
Levelized total cost €/t-CO2 19.1 15.7
Base case Optimal case
Annual energy and utilities cost M€/a 68.7 68.0
Annual CAPEX of the trunk pipeline M€/a 69.4 53.6
Annual CAPEX of collecting system M€/a 45.2 39.5
Annual O&M cost M€/a 9.2 9.2
Annual total cost M€/a 192.5 170.3
Comparison of annual costs of base case and optimal case
0.00
50.00
100.00
150.00
200.00
250.00
Base case Optimal case
An
nu
al
cost
(M€
/a
)
Annual capital cost of trunk pipelines
Annual capital cost of collecting system
Annual energy and utilities cost
Annual O&M cost
0.00
5.00
10.00
15.00
20.00
25.00
Optimal case COCATE project
Le
ve
lize
d c
ost
(€
/t-
CO
2)
Levelized capital cost of trunk pipelines
Levelized capital cost of collecting system
Levelized energy and utilities cost
Levelized O&M cost
key findings
For CO2 compression, lower intercooler exit temperature (20 oC vs 38oC in this study) contributes lower both energy cost and capital cost.
The O&M cost of CO2 compression is found to be low in other published models.
The pipeline diameter models in literature are generally reliable. With optimal diameter of pipelines, the initial velocity of CO2 mixture in dense phase is about 1.7m/s in this study.
The cost range of the pipelines are large for different models. The weight based model (Piessense et al. 2008) has the prediction close to this study.
Simulation-based techno-economics evaluation method offers a powerful tool for optimal designs for the projects, especially for the decision making support about the detailed technical options selection.
Reference Lists
Li H, Yan J. Evaluating cubic equations of state for calculation of vapor–liquid equilibrium of CO2 and CO2-mixtures for CO2 capture and storage processes. Applied Energy 2009;86(6):826-836.
Diamantonis NI, Boulougouris GC, Mansoor E, Tsangaris DM, Economou IG. Evaluation of Cubic, SAFT, and PC-SAFT Equations of State for the Vapor–Liquid Equilibrium Modeling of CO2 Mixtures with Other Gases. Industrial & Engineering Chemistry Research 2013;10.1021/ie303248q:130227083947003.
Foster NR, Bezanehtak K, Dehghani F. Modeling of Phase Equilibria for Binary and Ternary Mixture of Carbon Dioxide, Hydrogen and Methanol. [cited 2014 02/06]. Available from: http://www.isasf.net/fileadmin/files/Docs/Versailles/Papers/PTs15.pdf
IEA GHG, 2002. Pipeline transmission of CO2 and energy. Transmission study report. PH4/6, 1–140.
McCollum, D.L., Ogden, J.M., 2006. Techno-economic models for carbon dioxide compression, transport, and storage & Correlations for estimating carbon dioxide density and viscosity. UCD-ITS-RR-06-14, 1–87.
Piessens, K., Laenen, B., Nijs, W., Mathieu, P., Baele, J.M., et al., 2008. Policy Support System for Carbon Capture and Storage. SD/CP/04A, 1–269.
Roussanaly S, Bureau-Cauchois G, Husebye J. 2013. Costs benchmark of CO2 transport technologies for a group of various size industries. International Journal of Greenhouse Gas Control. 12(0):341-350.
S.T. McCoy, E.S. Rubin, 2008. An engineering-economic model for pipeline transport of CO2 with application to carbon capture and storage. International Journal of Greenhouse Gas Control, 2 , pp. 219–229
M. Van den Broek, A. Ramírez, H. Groenenberg, F. Neele, P. Viebahn, W. Turkenburg, A. Faaij, 2010. Feasibility of storing CO2 in the Utsira formation as part of a long term Dutch CCS strategy: An evaluation based on a GIS/MARKAL toolbox. International Journal of Greenhouse Gas Control, 4 , pp. 351–366
Acknowledgement
Financial support of UK NERC Energy Programme (NE/H013865/2)
Information and discussion from Dr. Russell Cooper Mr. Ketan Mistry
Mr. Julian Field
from National Grid
Related publications: Luo, X., Mistry, K., Okezue, C., Wang, M., Cooper, R., Oko, E., Field, J. (2014), Process simulation and analysis for CO2 transport pipeline design and operation – case study for the Humber Region in the UK, European symposium on computer-aided process engineering (ESCAPE24), Budapest, Hungary. Published in Computer Aided Chemical Engineering, Vol. 33, p1633-1639 Luo X, Wang M, Oko E, Okezue C. Simulation-based techno-economic evaluation for optimal design of CO2 transport pipeline network. Applied Energy 2014;132(0):610-620
Thanks
Thank you for you attention!
Questions are welcome.
Contact us if you are interested in our works.
Meihong Wang
UK office: 01482 466688
Xiaobo Luo
www.co2quest.eu
School of something FACULTY OF OTHER
School of Chemical and Process Engineering FACULTY OF ENGINEERING
Numerical modelling of trans-triple point
temperature near-field sonic dispersion
of CO2 from high pressure
dense phase pipelines
Dr Chris Wareing
UKCCSRC Equation of State Workshop,
11th November 2014, York
C J Wareing, R M Woolley, M Fairweather, S Falle
University of Leeds, Leeds, LS2 9JT, United Kingdom COOLTRANS: The Don Valley CCS Project is co-financed by the European Union’s European Energy Programme for Recovery
COOLTRANS / CO2PIPEHAZ / CO2QUEST: The sole responsibility of this content lies with the author.
The European Union is not responsible for any use that may be made of the information contained herein
Overview
Carbon capture and storage, the short term option for reducing
CO2 emissions, is likely to proceed with transportation from
source to storage along high-pressure dense phase pipelines
• Pipelines fail. Complex CFD simulations can validate pragmatic
approaches used for quantified risk assessment (QRA).
• Leeds: near-field sonic dispersion of carbon dioxide (CO2) from
high pressure pipelines
• Examples
• Thermodynamic model
• Recent developments
• Requirements for impurities
Venting: COOLTRANS dense phase Temperature
Dense phase release from a 150bar reservoir
through a 25mm ventpipe Measurements at:
• 4m (165D)
• 7m (288D)
Near-field shock containing region:
20D x 20D (0.5m x 0.5m)
COOLTRANS punctures and ruptures
Puncture:
Rupture:
Near-field dispersion model
(COOLTRANS & CO2PIPEHAZ)
• Thermodynamic model: (Wareing et al. 2013, AIChE Journal 59 3928-3942)
• Near-field dispersion of pure CO2 in the gas, liquid and solid phases into
dry air.
• Novel composite equation of state for pure CO2 employing:-
• the Peng-Robinson equation of state in the gas phase;
• tabulated data derived from the Span & Wagner equation of state for
the liquid phase and vapour pressure;
• and NIST/DIPPR data for the solid phase and latent heat of fusion.
• Calculations were undertaken using the Helmholtz free energy in terms
of temperature and molar volume, as all other thermodynamic properties
can be readily obtained from it.
• Novel combination of the simple Peng-Robinson equation and tabulated
data on the saturation line allows for crucially a fast implementation
numerically, when updating tens of millions of grid cells per second.
• Internal energy on the saturation line. • Tcrit marks the critical
temperature.
• The triple point can be
identified by the steep
connection between the
liquid and solid phases –
the latent heat of fusion.
• Numerical method, with
unstructured AMR
Near-field dispersion model
(COOLTRANS & CO2PIPEHAZ)
Comparative performance
• Comparison of ideal, Peng-Robinson and composite EoSs
Temperature:- CO2 fraction:-
(a) Ideal
(b) P-R
(c) Composite
Comparative performance
• Comparison of Peng-Robinson and composite EoSs
Condensed
phase fraction:-
(a) P-R
(b) Composite
Recent developments
• Different equations of state
• Solid phase. Now able to run with tables generated from Jager and
Span – Cp and internal energy are very different. Also considered the
Trusler solid phase EoS.
• Gas and liquid phases. Several equations of state under consideration
and comparison:-
• Physical Properties Library (SAFT/pcSAFT) from the National
Research Centre for Physical Sciences, Greece.
• Tables based on Span and Wagner
• Tables based on Richard Graham’s work.
• New considerations of EOS-CG.
• Have previously considered PRSV, PRSV2, Yokozeki, PROPATH &
others and discounted them for our use.
Requirements for impurities
(CO2QUEST)
• EITHER a simple fast equation of state that can be directly embedded
(advantages of speed and solutions over a wide range).
• OR a complex equation that can generate tables (disadvantage – solvers
routinely require solutions outside the tabular ranges).
• Mixing rules with known interaction parameters for e.g. N2, O2, Ar, H2S.
• Pipeline spec.: 96% CO2, 4% impurities (<2% N2, <2% O2 + trace).
Temperature range: 50K – 400K. Pressure range: 0.01 to ~20 MPa.
• Options:-
• A simple option: e.g. Peng-Robinson. But, there are known issues
(incorrect densities, speed of sound, etc. noted again at GHGT12)
• Trust a library function e.g. PPL. Current issues under investigation.
• More complex option: EOS-CG in tabular form. Possible issues with
H2O (presented at GHGT12)?
Thank you for listening. Discussion?
Phase equilibrium studies of impure CO2 systems to underpin
developments of CCS technologies
Jie Ke, Martyn Poliakoff and Michael W. George
School of Chemistry
The University of Nottingham
11 November, 2014, CO2 Properties and EOS for Pipeline Engineering, York
Acknowledgements
National Grid
E.ON
Rolls-Royce
PSE
Our collaborators from the COZOC, MATTRAN and COOLTRANS projects
Dr. Stéphanie Foltran
Dr. Yolanda Sanchez-Vicente
Dr. Andrew J. Parrot
Dr. James Calladine
Dr. Maria-José Tenorio
Dr. Alisdair Wriglesworth
Matthew E. Vosper
Norhidayah Suleiman
Prof. Trevor Drage
UKCCSRC
ETI
EPSRC
TSB
High-pressure facilities in Nottingham
Sensors
t=0 t=0
Sound Wave
t
IR
Holey fibre + GC
ATR Shear-mode quartz Optical fibre
Density meter
J. Phys. Chem. 1996, 100, 9522. J. Phys. Chem. 1997, 101, 5853. Fluid Phase Equilib. 1998, 150, 493. J. Supercrit. Fluids 2004, 30, 259. Phys . Chem. Chem. Phys. 2004, 6, 1258. J. Chem. Eng. Data 2009, 54, 1580.
Vapour-liquid-equilibrium and other thermodynamic properties of CO2 mixtures
Density
CO2 + N2, CO2 + H2, CO2 + N2 + H2 and CO2 + N2 + Ar
VLE Without water With water
o CO2 + N2 (4 mixtures)
o CO2 + H2 (3 mixtures)
o CO2 + N2 + H2 (2 mixtures)
o CO2 + N2 + Ar (in progress, funded by ETI and PSE)
o CO2 + H2 + Ar (in progress, funded by ETI and PSE)
o CO2 + H2O o CO2 + N2 + H2O o CO2 + H2 + H2O
260 270 280 290 300
2
4
6
8
10
12
14
p / M
Pa
T / K
p – T phase boundary of CO2 + N2 and CO2 + H2
260 270 280 290 300
2
4
6
8
10
p / M
Pa
T / K
CO2 + N2 CO2 + H2
Pure CO2
14%
2% 5%
Pure CO2
9.1%
3% 4%
260 270 280 290 300
4
8
12
0.95 CO2 + 0.05H
2
0.95 CO2 + 0.05N
2
p / M
Pa
T / K
0.95 CO2 + 0.02 N
2 + 0.03 H
2
p – T phase boundary of the ternary system of CO2 + N2 + H2
0.93 CO2 + 0.04 N2 + 0.03 H2
Pure CO2
Solubility of H2O in CO2 + N2 (40 oC)
yH2O - mole fraction of H2O
Pure CO2 95% CO2 + 5% N2
90% CO2 + 10% N2
What is next?
• The VLE data of multicomponent mixtures (more than 5
components) of CO2, N2 and O2, H2, Ar, etc.
• The solubility of water in the mixture of CO2 + N2 + H2.
• CO2 and N2 solubilities in sea water.
• Karl-Fischer titration
• Spectroscopic methods
Top Related