Post-Combustion CO2 Capture Using

Chemical Absorption
Minimizing Energy Requirement

Henning Leifsen

Master of Science in Energy and Environment

Submission date: December 2007
Supervisor:
Olav Bolland, EPT

Norwegian University of Science and Technology

Department of Energy and Process Engineering

Problem Description
There is an increasing interest in CO2 capture and storage as a measure to reduce man-made
emissions of the greenhouse gas CO2. Several methods have been proposed for how to do CO2
capture from power plants. It may seem that the post-combustion method including amine
absorption is the most likely choice of technology in the short term. However, there is a significant
efficiency penalty. When capturing the CO2, the use of fuel energy increases 15-25% per kWh of
electricity produced.
The main focus of this master thesis will be modelling and simulation of a process for chemical
absorption of CO2 from atmospheric exhaust gas coming from a gas turbine plant. In an
absorption plant for gas separation there are many parameters which can be varied as well as
there are many practical limitations. It is very important to have a good model representation of
the solvent to be used. The impact on overall plant (including power plant) efficiency shall be
quantified.
The objective of the work is to develop a computational model which with sufficient accuracy can
predict performance with respect to energy requirement, and further to find optimal and realistic
parameter values for low energy use.
The modelling and simulation shall be performed in Hysis, and will be built on the student’
s previous project work in spring 2007. Aker Kvrner Engineering & Technology will provide data
for solvent characteristics beyond those in Hysis, if necessary. Issues to be considered are
performance, safety and cost (CAPEX/OPEX) Aker Kvrner Engineering & Technology can assist in
obtaining cost data.
The following tasks should be considered in the project work:
1)
Make a literature review of work done for modelling and simulation of absorption/
desorption systems focussing on plant performance and the energy use.
2)
A simulation model is to be made, on a level in line with the above stated objective, which
is capable of a mass and heat balances of the process.
3)
Parameters and configurations (as split stripper) with importance for the energy use in
the process shall be identified, and realistic boundaries for parameter values shall be discussed.
4)
Using simulations with energy use as guidance, an optimal combination of parameters as
well as configuration choices shall be sought an found.

-- ” -Within 14 days of receiving the written text on the diploma thesis, the candidate shall submit a
research plan for his project to the supervisors. A planning with milestones should also be clearly

Master thesis for Henning Leifsen, Autumn 2007

Master thesis for Henning Leifsen, Autumn 2007

ii

Master thesis for Henning Leifsen, Autumn 2007

iii

Master thesis for Henning Leifsen, Autumn 2007

iv

Master thesis for Henning Leifsen, Autumn 2007

Report no:

Classification:

The Norwegian University of Science and Technology

Norges Teknisk-Naturvitenskapelige Universitet
TELEPHONE

ADDRESS:

NTNU
DEPARTMENT OF ENERGY AND
PROCESS Engineering
Kolbjrn Hejes vei 1A
N-7491 Trondheim - NTNU

Switchboard NTNU:
Department office:
Hydropower section:

TELEFAX
73 59 40 00
73 59 27 00
73 59 38 57

Department office:
Hydropower section:

Title of report
Post-combustion CO2 capture using chemical absorption minimizing energy
requirement

73 59 83 90
73 59 38 54

Date
21 December 2007
No. of
pages/appendices
71/8
Project manager

Author
Henning Leifsen
Division
Faculty of Engineering Science and technology
Department of Energy and Process Engineering
ISBN no.

Project no.
Price group

Client/Sponsor of project
Aker Kvrner / NTNU

Clients ref.

Abstract
Capture and storage from fossil fuel fired power plants is drawing increasing interest as a potential
method for the control of greenhouse gas emissions. An optimization and technical parameter study for a
CO2 capture process of the flue gas of a commercial gas power plant, based on absorption/desorption
process with MEA solutions, using HYSYS with the Amine Property Package fluid package, has been
performed.
The optimization has aimed to reduce the energy requirement for solvent regeneration, by investigating
the effects of circulation rate, cross-flow heat exchanger minimum approach, desorber operating pressure
and the absorber diameter. In addition, an economic evaluation including investment cost has been
performed for the first three parameters.
Major energy savings can be realized by optimizing the desorber pressure and the solvent circulation rate.
The circulation rate will have a clearly defined optimal point, while for the desorber pressure the
temperature will be a limiting factor. A too high temperature may lead to amine degradation and corrosion
problems. The cross-flow heat exchanger minimum temperature approach will not affect the energy
consumption significantly.
An optimum absorber column diameter was not found, but the column should be designed with a diameter
large enough to prevent flooding through the column. A too large diameter will not favour the energy
consumption very much, and other factors will be more decisive when the column diameter is chosen.

Group 1

Indexing Terms: English

CO2 ,capture, MEA, optimization

Norwegian
CO2, innfanging, MEA, optimalisering

Group 2
Selected
by author

Master thesis for Henning Leifsen, Autumn 2007

vi

Master thesis for Henning Leifsen, Autumn 2007

Preface
CO2-capture has a high priority on the agenda of the Norwegian government today. A
collaboration between the government, Gassnova and several major industrial companies with
Aker Kvaerner as one of the leading members has resulted in the project called Just Catch.
The objective for Just Catch is to reduce the operating costs and improve the efficiency of
CO2-capturing plants. The aim is to offer a competitive technology for the Norwegian market
as well as the international market.
This assignment was written in Trondheim for the Department of Energy and Process
Engineering at the Norwegian University of Science and Technology (NTNU) autumn 2007.
Aker Kvaerner supported the necessary cost data and helped with the making of a base model
for simulation.
I would like to thank my supervisor professor Olav Bolland for his help and advices.
I would also like to thank Bjrn Magnus Berg at Aker Kvaerner for many helpful phonemeetings and for always being available for questions.

Trondheim , 21 day of December 2007

Henning Leifsen

vii

Master thesis for Henning Leifsen, Autumn 2007

viii

Master thesis for Henning Leifsen, Autumn 2007

Summary
Capture and storage from fossil fuel fired power plants is drawing increasing interest as a
potential method for the control of greenhouse gas emissions. An optimization and technical
parameter study for a CO2 capture process of the flue gas of a commercial gas power plant,
based on absorption/desorption process with MEA solutions, using HYSYS with the Amine
Property Package fluid package, has been performed.
The optimization has aimed to reduce the energy requirement for solvent regeneration, by
investigating the effects of circulation rate, cross-flow heat exchanger minimum approach,
desorber operating pressure and the absorber diameter. In addition, an economic evaluation
including investment cost has been performed for the first three parameters.
Major energy savings can be realized by optimizing the desorber pressure and the solvent
circulation rate. The circulation rate will have a clearly defined optimal point, while for the
desorber pressure the temperature will be a limiting factor. A too high temperature may lead
to amine degradation and corrosion problems. The cross-flow heat exchanger minimum
temperature approach will not affect the energy consumption significantly.
An optimum absorber column diameter was not found, but the column should be designed
with a diameter large enough to prevent flooding through the column. A too large diameter
will not favour the energy consumption very much, and other factors will be more decisive
when the column diameter is chosen.

ix

Master thesis for Henning Leifsen, Autumn 2007

Master thesis for Henning Leifsen, Autumn 2007

Sammendrag
Innfanging og lagring av CO2 fra kraftverk basert p fossilt brensel har i kende grad vekket
interesse som en mulig metode for redusere drivhuseffekten. En optimalisering og teknisk
parameter studie for en innfangingsprosess av CO2 fra eksosen til et kommersielt
gasskraftverk er gjennomfrt. Studien er basert p absorpsjon/desorpsjons-prosesser med
MEA-lsninger, der HYSYS og medflgene amine fluid-pakke er blitt benyttet.
Optimaliseringen har hatt som forml redusere energiforbruket til regenereringen av amin
ved underske betydningen av sirkulasjonsrate, kryss-strms varmevekslers minste
temperaturdifferanse, driftstrykket til desorber og absorber diameter. I tillegg har det blitt
gjennomfrt en konomisk analyse av de tre frstnevnte parametrene hvor ogs
invisteringskostnadene ble tatt hensyn til.
Ved optimalisere desorber trykk og sirkulasjonsrate av aminer kan energiforbruket
reduserers betraktelig. Sirkulasjonsraten vil ha et klart definert optimalt punkt, mens desorbertrykket vil vre begrenset av temperaturen i kolonnen. Hye temperaturer kan fre til
degradering av aminer og problemer med korrosjon. Minste temperaturdifferanse for kryssstrms varmeveksler vil ha liten betydning for energiforbruket.
En optimal absorber diameter ble ikke funnet, men kolonnen br designes med en diameter
som er stor nok til forebygge flooding gjennom kolonnen. En for stor diameter vil ikke
pvirke energiforbruket betydelig og andre faktorer vil ha strre betydning nr kolonnediameter skal velges.

xi

Master thesis for Henning Leifsen, Autumn 2007

xii

Master thesis for Henning Leifsen, Autumn 2007

Table of Contents
1

INTRODUCTION ................................................................................................. 1

DESCRIPTION OF A TYPICAL AMINE-BASED ABSORPTION PROCESS..... 3

HEAT REQUIREMENTS FOR THE STRIPPER COLUMN................................. 5

3.1

Heat of Absorption ................................................................................................................................. 5

3.2

Sensible Heat ........................................................................................................................................... 5

3.3

Dilution Heat ........................................................................................................................................... 6

3.4

Stripper- Heat Pressure and Flow Dependency ................................................................................... 7

3.5

Primary Amines versus Tertiary Amines ............................................................................................. 8

3.6

Corrosion and Amine Degradation due to High Stripper Temperature ........................................... 9

ABSORBER MODELLING ............................................................................... 11

4.1

Design Principles................................................................................................................................... 11

4.2

Contactor Selection............................................................................................................................... 11

4.3

Material and Energy Balance .............................................................................................................. 14

4.4
Column Height ...................................................................................................................................... 16
4.4.1
Absorption Coefficient....................................................................................................................... 16
4.4.2
The Gas-Liquid Interfacial Area ........................................................................................................ 17
4.4.3
Height of Column .............................................................................................................................. 17
4.4.4
Volumetric Overall Mass-transfer Coefficient................................................................................... 18
4.4.5
Height of Column Using Overall Mass-Transfer Coefficient ............................................................ 19
4.4.6
Selecting the Volumetric Overall Mass-Transfer Coefficient............................................................ 20
4.5

Column Diameter.................................................................................................................................. 22

4.6

Flooding ................................................................................................................................................. 23

4.7

Number of Equilibrium Stages ............................................................................................................ 24

4.8

Tray Efficiency...................................................................................................................................... 24

4.9

Residence Time Requirement .............................................................................................................. 24

AMINES PROPERTY PACKAGE IN HYSYS ................................................... 25

xiii

Master thesis for Henning Leifsen, Autumn 2007

6 MINIMIZING THE ENERGY CONSUMPTION IN AMINE-BASED ABSORPTION
PLANTS ................................................................................................................... 27
6.1

Effect of Condenser Temperature ....................................................................................................... 27

6.2

Effect of Cross-Flow Heat Exchanger Performance.......................................................................... 28

6.3

Effect of CO2 Capture Efficiency ........................................................................................................ 29

6.4
Effect of Rich and Lean Amine Level ................................................................................................. 32
6.4.1
Effect of Different Lean Solvent Loading Including the Effect of the CO2-Removal Percentage..... 33
6.4.2
Effect of Different Lean Solvent Loading Including the Effect of the MEA Weight%..................... 35
6.5

Effect of Reboiler Pressure .................................................................................................................. 37

6.6

Effect of Varying Reboiler Equilibrium Efficiency ........................................................................... 37

6.7

Importance of the Accuracy of the Equilibrium Model .................................................................... 38

MODELLING THEORY ..................................................................................... 41

7.1

Selection of Process Simulation Programme ...................................................................................... 41

7.2

Selection of Thermodynamic Package ................................................................................................ 41

7.3

Heat versus Mechanical Work............................................................................................................. 42

METHODOLOGY FOR ECONOMIC EVALUATION......................................... 43

8.1

Investment Expenses (CAPEX) ........................................................................................................... 43

8.2

Operating Expenses (OPEX) ............................................................................................................... 44

8.3

Profitability Analysis ............................................................................................................................ 45

SIMULATION CASES....................................................................................... 47

9.1

Base Case Model ................................................................................................................................... 48

9.2

Circulation Rate.................................................................................................................................... 51

9.3

Minimum Temperature Approach for Rich/Lean Heat Exchanger................................................. 54

9.4

Stripper Pressure .................................................................................................................................. 56

9.5

Economic Evaluation............................................................................................................................ 59

9.6

Absorber Column Diameter................................................................................................................. 64

10

CONCLUSION............................................................................................... 69

11

RECOMMENDATIONS FOR FURTHER WORK........................................... 71

12

REFERENCES .............................................................................................. 73

xiv

Master thesis for Henning Leifsen, Autumn 2007

Nomenclature
Latin Letters
A
a
E
Emv
C
c
cp
Fc
G
GM`
H
h
He
Ie
i
K
KGa
KLa
kG
kL
L
LM`
m&
M
Mw
N
n&
n
P
p
Pcol
Pel
Q
T
U
V
V&col
W
wG
X
x
Y
y
Z
z

= area
= interfacial area
= total mechanical work
= Murphree vapour efficiency
= cost
= concentration
= specific heat
= total consumption of cooling water
= gas flux
= CO2-free gas molar flow rate
= enthalpy
= enthalpy
= Henry`s law constant
= enhancement factor
= interest rate
= equilibrium ratio
= overall mass-transfer coefficient for gas
= overall mass-transfer coefficient for liquid
= gas absorption coefficient
= liquid absorption coefficient
= liquid flux
= CO2-free liquid molar flow rate
= mass stream
= maintenance cost
= molar weight
= flux
= molar stream
= project life
= pressure
= partial pressure
= price of cooling water
= price of electricity
= heat duty
= temperature
= overall heat transfer coefficient
= volume
= volume stream of cooling water
= mechanical work
= superficial gas velocity
= mole ratio component in liquid phase
= mole fraction in liquid phase
= mole ratio of component in gas phase
= mole fraction in vapour phase
= mole fraction in feed stream
= height of column

xv

m2
m2/m3
MJ/kg CO2
NOK
kmol/m3
kJ/kg C
m3/year
kg/m2h
kmol/sm
kJ/kg
kJ/kg
pi/ci
%
kmol/smPa
kmol/smkmol/m3
kmol/m2sPa
kmol/m2skmol/m3
kg/hm2Pa
kmol/sm
kg/h
NOK/year
kg/kmol
kmol/m2s
kmol/h
years
bara
bara
NOK/ m3
NOK/kWh
MJ
C
kJ/Chm2
m3
M3/h
MJ
m/s
m

Master thesis for Henning Leifsen, Autumn 2007

Greek Letters

= CO2-load
= energy quality-factor
= size exponent
= capture coefficient
= stage efficiency
= density
= kinematic liquid viscosity

mole/mole
kg/m3
cst

Indexes (superscript)
sat
vap

= saturation
= vapour

Indexes (subscript)
bc
bp
c
co
diff
e
g
i
i
in
j
l
lm
p

= base case
= boiling point
= condenser
= compression
= difference
= equilibrium
= gas
= interface
= component number
= inlet
= stage number
= liquid
= log mean
= constant pressure

xvi

Master thesis for Henning Leifsen, Autumn 2007

Abbreviations
CAPEX
HETP
LMTD
MDEA
MEA
NPV
NVE
NTNU
OPEX
PR
VLE

= Capital expenditures
= Heat equivalent to a theoretical plate
= Logarithmic mean temperature difference
= Methyl-diethanolamine
= Monoethanolamine
= Net present value
= Norges vassdrags- og energidirektorat
= Norwegian University of Science and Technology
= Operational expenditure
= Peng-Robinson
= Vapour-liquid equilibrium representation

xvii

Master thesis for Henning Leifsen, Autumn 2007

xviii

Master thesis for Henning Leifsen, Autumn 2007

List of Figures
Figure 2-1 Schematic diagram for an amine absorption plant _______________________________________ 4
Figure 3-1 Streams in and out of the absorption tower ____________________________________________ 7
Figure 3-2 Heat demand for a regression function at a 40C absorption temperature[2]__________________ 8
Figure 4-1 Random tower packing ___________________________________________________________ 12
Figure 4-2 Examples of random packing elements _______________________________________________ 13
Figure 4-3 Typical structured packing (intalox)_________________________________________________ 13
Figure 4-4 Material balance diagram for countercurrent contactor _________________________________ 14
Figure 4-5 Operating line-equilibrium curve diagram for absorption column _________________________ 15
Figure 4-6 Effect of liquid flow rate on overall mass-transfer coefficient _____________________________ 20
Figure 4-7 Overall mass transfer coefficient of different structured packings at various liquid flow rates ____ 21
Figure 4-8 Overall mass-transfer coefficient at various liquid temperatures___________________________ 21
Figure 4-9 Generalized pressure drop correlation for packed towers ________________________________ 23
Figure 5-1 Stage j in a column simulated in HYSYS ______________________________________________ 25
Figure 6-1 Condenser temperature versus reboiler duty[10] _______________________________________ 27
Figure 6-2 Sorbent regeneration heat requirement (kJ/kg CO2), and the equivalent loss in power generation
(MWh/tonne CO2) as a function of CO2 capture efficiency [11]. ____________________________________ 29
Figure 6-3 Electrical energy requirement for capture and compression (MWh/tonne CO2), and comparison to
the equivalent thermal loss in power generation (MWh/tonne CO2) [11]. _____________________________ 30
Figure 6-4 Total energy requirement (MWh/tonne CO2) of the amine-based CO2 capture system, as a function of
CO2 capture efficiency. ____________________________________________________________________ 30
Figure 6-5 Net power output of a power plant with amine-based CO2 capture system as a function of CO2
capture efficiency (MWg = gross power plant size) [11].__________________________________________ 31
Figure 6-6 CO2 capture system capital cost as a function of the CO2 capture efficiency for two plant sizes. __ 31
Figure 6-7 CO2 capture system capital cost normalized on the CO2 removal capacity, as a function of the CO2
capture efficiency for two plant sizes. _________________________________________________________ 32
Figure 6-8 Thermal energy requirement at various CO2/amine lean loadings for different CO2 removal (CF)
[12] ___________________________________________________________________________________ 34
Figure 6-9 Solvent flow rate requirement at various CO2/amine lean solvent loadings for different CO2 removal
[12] ___________________________________________________________________________________ 34
Figure 6-10 Cooling water consumption at various CO2/amine lean solvent loadings for different CO2 removal
[12] ___________________________________________________________________________________ 35
Figure 6-11 Cost of electricity as a function of lean solvent loading for different CO2 removal [13]________ 35
Figure 6-12 Thermal energy requirement at various CO2/amine lean loadings for different MEA wt%. (for a
bituminous coal fired plant) [12] ____________________________________________________________ 36
Figure 6-13 Solvent flow rate requirement at various CO2/amine lean solvent loadings for different MEA (wt.%)
______________________________________________________________________________________ 36
Figure 6-14 Cooling water consumption at various CO2/amine lean solvent loadings for different MEA (wt.%)36
Figure 6-15 Tuned vapour liquid equilibrium (VLE) data for 30 wt% MEA at stripper conditions. a. normal plot,
b. log-log plot [10] _______________________________________________________________________ 38
Figure 7-1 Ratio () between reduced electricity production and heat outlet from a power plant [14]. ______ 42
Figure 9-1 Base case for simulation __________________________________________________________ 48
Figure 9-2 Intersection point where was found ________________________________________________ 49
Figure 9-3 Base case, CO2 compression and drying _____________________________________________ 49
Figure 9-4 Total mechanical work as a function of circulation rate _________________________________ 52
Figure 9-5 Reboiler duty as function of circulation rate __________________________________________ 52
Figure 9-6 Total mechanical work as a function of lean solvent loading ______________________________ 53
Figure 9-7 Temperature of solvent entering the absorber tower as a function of circulation rate___________ 53
Figure 9-8 Cooling water as a function of solvent sirculation rate __________________________________ 54
Figure 9-9 Total mechanical work as a function of rich/lean minimum temperature approach_____________ 55
Figure 9-10 Reboiler duty as a function of rich/lean minimum temperature approach ___________________ 55
Figure 9-11 Total mechanical work as a function of stripper column pressure _________________________ 57
Figure 9-12 Reboiler duty as a function of stripper column pressure ________________________________ 57
Figure 9-13 Cooling water consumption as a function of stripper column pressure _____________________ 58
Figure 9-14 Temperature of reboiler reflux as a function of stripper column pressure ___________________ 58
Figure 9-15 Total mechanical work as a function of pressure drop in absorber column __________________ 66
Figure 9-16 Superficial gas velocity as a function of absorber column diameter _______________________ 67
Figure 9-17 Height and diameter as a function of pressure drop in absorber __________________________ 67
Figure 9-18 Total mechanical work as a function of absorber column diameter ________________________ 68

xix

Master thesis for Henning Leifsen, Autumn 2007

xx

Master thesis for Henning Leifsen, Autumn 2007

List of Tables
Table 6-1 Varying the cross-flow heat exchanger pinch temperature approach[10] _____________________
Table 6-2 Effect of: a. Increasing circulation rate and lean loading level and b. Increasing the rich and lean
levels simultaneously[10] __________________________________________________________________
Table 6-3 Effect of varying reboiler pressure[10] _______________________________________________
Table 6-4 Effect of varying reboiler equilibrium efficiency[10] _____________________________________
Table 6-5 Effect of varying tuning factor in equilibrium model[10]__________________________________
Table 9-1 Difference in investment cost and operational cost for different circulation rates compared to base
case of 2358 ton/h ________________________________________________________________________
Table 9-2 Difference in net present value (NPV) for different circulation rates compared to base case of 2358
ton/h __________________________________________________________________________________
Table 9-3 Difference in investment cost and operational cost for different minimum temperature approaches
compared to base case at 8.5C _____________________________________________________________
Table 9-4 Difference in net present value (NPV) for different minimum temperature approaches compared to
base case at 8.5C________________________________________________________________________
Table 9-5 Difference in investment cost and operational cost for different stripper column bottom pressures
compared to base case of 1.862 bara _________________________________________________________
Table 9-6 Difference in net present value (NPV) for different stripper column bottom pressures compared to
base case of 1.862 bara, given for interest rates of 3%, 5% and 7%._________________________________
Table 9-7 Difference in investment cost for a sub-optimal operating point compared to base case__________

xxi

28
33
37
37
39
60
61
61
61
62
63
63

Master thesis for Henning Leifsen, Autumn 2007

xxii

Master thesis for Henning Leifsen, Autumn 2007

Introduction

Background
The climate change is the biggest environmental challenge the world face today. To avoid
man-made climate changes it is necessary to reduce the CO2 emissions drastically. Given the
advantages inherent in fossil fuels, such as their availability, relatively low cost, and the
existing infrastructure for delivery and distribution, they are likely to play an important role in
the world wide energy production for at least the next 75 years. Therefore there is also an
increasing interest in CO2 capture and storage.
Several different methods have been proposed for how to perform CO2 capture from power
plants, but the maturity of the technologies differs. The most mature technology today is the
post-combustion method including amine absorption. The major problem using this
technology has been the high operational costs, mainly due to the regeneration of the amines.
The operational costs are closely connected to the reduced efficiency of the power plant when
capturing the CO2.
Scope and Objective
The main area of investigation for this master thesis will be modelling and simulation of a
chemical absorption process for CO2 capture from atmospheric exhaust gas coming from a
gas turbine plant. A simulation model which is capable of calculating mass and heat balances
with sufficient accuracy will be presented and discussed.
In the CO2 capture process there are many different parameters which can be varied. There
are also many limitations. The objective of this thesis will be to find optimal and realistic
parameter values for low energy use.
Four parameters will be presented and varied. The first parameter variation introduced is the
solvent circulation rate, secondly the rich/lean heat exchanger minimum temperature approach
and then the stripper column pressure. All these parameters will be presented with respect to
both performance and cost. The fourth parameter is the absorber column diameter, where
pressure loss, superficial gas velocity, diameter and height will be varied and discussed.
The structure of the thesis is listed below:

Chapter two gives a short introduction to chemical absorption of CO2.

Chapter three explains the stripper heat demand for desorption.
Chapter four describes the absorber column design.
Chapter five is a short description of the Amine Property Package in HYSYS.
Chapter six is a literature review of work done for modelling and simulation of
absorption/desorption systems focussing on plant performance and the energy use.
Chapter seven and eight describes how the different parameter variations were
measured and compared in terms of energy use and cost.
Chapter nine introduces the parameters examined in this thesis and includes
assumptions, results and discussion.
Chapter ten and eleven give the conclusion and recommendations for further work.

Master thesis for Henning Leifsen, Autumn 2007

Master thesis for Henning Leifsen, Autumn 2007

Description of a Typical Amine-based Absorption Process

Figure 2-1 depicts the operation of a typical amine based absorption plant for recovery of CO2
from flue gas. Before the flue gas enters the absorption column, it is cooled and treated for
particles and other impurities such as SOX and NOX to tolerable levels. The flue gas is also
compressed to a higher pressure in order to meet the absorber operational pressure.
In the absorber the rising flue gas reacts with the lean amine solution flowing downwards.
The amines selectively absorb CO2 from the flue gas by chemically reacting with it. The
absorber column typically operates within a temperature range of 40-45C at the top and 5060C at the bottom[1], depending on column pressure and composition.
The treated flue gas exits from the top of the absorber, while the rich amine solution (rich on
CO2) from the bottom of the column is pumped to the cross-flow heat exchanger. Here the
rich amine solution is preheated to about 105 C by heat integration with the lean amine
solution[1]..
The heated CO2-rich amine then enters the upper section of the stripper column, which
normally operates at a temperature of about 110 C at the top and 120 C at the bottom. The
operation pressure in the bottom of the column and in the reboiler is typically 2 bara with a
pressure drop across the column of about 0.2 bar[1].
In the stripper column the CO2 is desorbed from the amine solution and exits from the top of
the column along with water vapour and some amines. The CO2-rich vapour stream is passed
through a reflux condenser where it is partially condensed. The bulk CO2 is separated from
the condensate, dried and further compressed to the required pressure for CO2 sequestration,
e.g. about 100 bara for injection into the ocean or a geological formation[1]. The condensate,
which is mainly water, is fed back to the stripper.
Since the CO2 desorption is an endoterm process, additional heat is required. This heat is
normally provided by saturated steam at 3 bara or higher. Heating of the amine solution drives
off some water which helps desorbing CO2 from the rich amine solution[1].
The lean amine solution that leaves the bottom of the stripper is pumped up to the absorber
pressure and cooled down to the absorber temperature before entering the absorber.

Master thesis for Henning Leifsen, Autumn 2007

CO2 to compression

Treated flue gas

Flash tank

Condenser

Absorber

Cooler

Stripper
Flue gas

Reboiler
Lean/rich heat exchanger

Pump

Pump

Figure 2-1 Schematic diagram for an amine absorption plant

Master thesis for Henning Leifsen, Autumn 2007

Heat Requirements for the Stripper Column

The heat requirement for the stripper column could be divided into three;

heat of absorption
sensible heat
dilution heat

By dividing the heat into three different types, it is easier to explain the pressure and flow
dependency as well a load dependency for the desorption process. All numbers and equations
in this chapter rest on work performed by Pernille Sire Seljom[2].

3.1 Heat of Absorption

Heat of absorption is the energy needed to break the chemical bonds between the CO2 and the
amines. It is called the heat of absorption because it is the same heat that is released in the
absorber when the amines react with CO2. The heat of absorption is higher for primary amines
(MEA) than for tertiary amines (MDEA).

3.2 Sensible Heat

The CO2-rich amine solution that enters the stripper has a temperature below the boiling
temperature of the solution. The heat that is required to heat the solution up to its boiling point
is called the sensible heat. Mathematically it can be written as:

Qsensible (Tin , rich )c p (Tin , rich )

=
mCO
(rich lean)CmineMw, CO
2

(3.1)

where
Qsensible = energy needed to heat solution to boiling point (kJ)
mCO = mass CO2 (kg)
c p = specific heat for rich amine solution (kJ/kgC))
= density of rich amine solution (kg/m3)
T = Tin-Tbp = difference between inlet temperature and boiling
temperature of CO2 rich feed (C)
(rich-lean) = molefraction CO2 sent to compression per mole amine
Camine = concentration of amine in rich amine solution (kmol/m3)
M w,CO = moleweight CO2 (kg/kmol)
2

Master thesis for Henning Leifsen, Autumn 2007

3.3 Dilution Heat

In the stripper it is desirable that the CO2 changes from liquid phase to gas phase. For this to
happen it is necessary for the CO2 saturation pressure in the amine solution flowing
downwards to be greater than the CO2 partial pressure of the gas solution rising.
The gas solution that rises in the stripper mainly consists of CO2 and water. According to
Dalton`s Law the partial pressure of CO2 is a function of the total pressure and the molar
portion of water, as shown in equation 3.2.

pCO = P pH O = P(1 yH O )
2

(3.2)

where

P = total pressure (bar)

pCO = partial pressure of CO2 (bar)
pH O = partial pressure of water (bar)
y H O = molar fraction of water
2

The purpose of the dilution heat is to increase the portion of water in the gas flowing upwards
in the stripper column by evaporating some of the water in the amine solution. This will
decrease the partial pressure of the CO2.
As the gas stream rises in the column the partial pressure of CO2 will increase. There are
mainly two reasons for this; the CO2 phase change from liquid to gas and the condensation of
water caused by a decrease in temperature.
When the partial pressure exceeds the saturation pressure, CO2 changes from gas to liquid.
The dilution heat is the minimum heat requirement to keep the saturation pressure greater than
the partial pressure, and make CO2 change from liquid to gas, in the whole column.
Mathematically it can be written:

Qstrip
PHsatO (Tin ) xH O
=
H HvapO (Tin )
mCO
pCO (Tin, rich ) Mw , CO
2

where
Qstrip = heat required to keep PCOsat PCO (kJ)
mCO = mass CO2 (kg)
H HvapO = energy required to evaporate water (kJ/kmol)
xH O = fraction of liquid water
PHsatO = saturation pressure water (bar)
pCO = partial pressure of CO2 (bar)
M w,CO = moleweight for CO2 (kg/kmol)
2

(3.3)

Master thesis for Henning Leifsen, Autumn 2007

3.4 Stripper- Heat Pressure and Flow Dependency

The load, , of the streams in the absorption process, is defined as mole CO2 per mole amine.
The rich load is the mole CO2 per mole amine in the stream leaving the bottom of the
absorber while the lean load is the mole CO2 per amine in the stream leaving the bottom of
the stripper. The figure below shows the streams in and out of the absorber.

Figure 3-1 Streams in and out of the absorption tower

namine =
where

CO nCO
rich lean
2

(3.4)

CO = capture coefficient for CO2

nCO = moles of CO2 entering the absorption tower
namine = moles of amines used for absorption
rich = rich load
lean = lean load
2

The rich load increases as the partial pressure of CO2 entering the absorber increases,
assuming a constant temperature and capture coefficient in the absorption tower. Equation 3-4
shows that, with a given CO2 molar flow, capture coefficient and lean load, this lead to a
smaller amount of amines needed. With less amines to be heated up to boiling temperature,
the sensitive heat decreases. An increase in the rich load also increases the CO2 saturation
pressure in the stream entering the stripper. A higher saturation pressure leads to a greater
driving force between the CO2 partial pressure and saturation pressure, and less dilution heat
is needed. Equation 3-3 shows that the dilution heat decreases with an increasing rich load.

Master thesis for Henning Leifsen, Autumn 2007

By increasing the lean load the dilution heat will decrease. The higher the lean load is, the
higher the CO2 saturation pressure is at the bottom of the column. This means that less water
vapour is needed to keep the saturation pressure higher than the partial pressure. However,
equation 3-4 shows that by keeping the molar flow of CO2, the capture coefficient and the rich
load constant, the amount of amines necessary will increase when the lean load increases.
Therefore the sensible heat will increase with an increasing lean load.
Figure 3-2 depicts the distribution between absorption heat, sensitive heat and dilution heat.
The absorption heat is the largest and is independent of the partial pressure. Sensitive heat is
larger than the dilution heat and they both decrease with an increase in partial pressure.

Figure 3-2 Heat demand for a regression function at a 40C absorption temperature[2]

3.5 Primary Amines versus Tertiary Amines

The absorption heat is larger for primary amines than tertiary amines. In addition tertiary
amines have got a higher load at low temperatures and a lower load at higher temperatures.
Therefore tertiary amines have got a smaller heat demand in the stripper than primary amines.
This difference in heat demand between the different amines increases as the partial pressure
into the absorber increases.

Master thesis for Henning Leifsen, Autumn 2007

3.6 Corrosion and Amine Degradation due to High Stripper Temperature

Amine solution corrosion is the corrosion of carbon steel in the precense of aqueous amine.
Amine solution corrosion is most significant in the hot bottom section of the desorber column
and is influenced by a number of factors. The most pronounced are high operating
temperatures, high loadings, amine solution concentration, amine type and amine solution
contaminants including amine degration products and heat-stable salts[3].
Alkanolamines are subject to degradation by contact with free oxygen. Degradation products
can reduce amine solution absorption capacity, increase solution viscosity, increase solution
foaming tendency, and in some cases contribute to amine plant corrosion. MEA reacts with
CO2 to form substituted imidazolidone that later hydrolyzes to produce a diamine and release
the CO2. Both imidazolidone and the diamine degration products can be removed from MEA
solutions by thermal reclaiming. The amine-CO2 degradation reactions are relatively slow, but
do occur at a significant rate under the conditions prevailing in the regeneration section of an
absorption plant. The extent of these reactions may be limited by avoiding elevated
temperatures. Reboiler heat flux should be limited, and amine circulation through the reboiler
should be kept high. Desorber column operating temperatures should be limited in order to
minimize the rate of amine degradation.

Master thesis for Henning Leifsen, Autumn 2007

10

Master thesis for Henning Leifsen, Autumn 2007

Absorber Modelling
4.1 Design Principles

In the design of counter-current contactors, there are normally five major steps [3, 4]:

Selection of contactor, including types of trays or packing based on process

requirements and expected service conditions.
Calculation of heat and material balances.
Estimation of required column height (number of trays or height of packing) based on
mass transfer analysis.
Calculation of required column diameter and tray or packing parameters based on gas
and liquid flow rates and hydraulic considerations.
Mechanical design of the hardware.

4.2 Contactor Selection

The primary function for the contactor is to provide sufficiently large area for liquid surface in
contact with the gas phase under the conditions favouring mass transfer. The three most
commonly employed mechanisms for absorption/desorption are: bubble cap trays, spray
chambers and packed columns. Although specific requirements and conditions may favour
one over the other, they are interchangeable to a considerable extent. Packed columns are
generally preferred over tray columns for small installations, corrosive service, liquids with a
tendency to foam, very high liquid/gas ratios and applications in which a low pressure drop is
desired [3]. Their use in larger sizes appears to be increasing, and there is also a growing use
of packing to replace trays where an improvement in column performance is required. As
packed column was found to be the best choice for the purpose of this thesis, only the packed
column will be described in detail.
Packed columns are vertical columns which have been filled with packing or devices of large
surface. One example of a randomly packed column is shown in figure 4-1.

11

Master thesis for Henning Leifsen, Autumn 2007

Figure 4-1 Random tower packing

The tower packing should offer the following characteristics [4]:

Provide a large interfacial surface between liquid and gas.
Process desirable fluid-flow characteristics. This ordinarily means that the fractional
void volume, or fraction of empty space, in the packed bed should be large. The
packing must permit passage of large volumes of fluid through small tower cross
sections without loading and flooding (see below) and with low pressure drop for the
gas.
Be chemically inert to fluids being processed.
Have structural strength to permit easy handling and installation.
Represent low cost.
Packed columns could be packed randomly or structured. Random packings are simply
dumped into the tower during installation and allowed to fall at random. The tower may first
be filled with water to reduce the velocity of fall. Random packings most frequently used at
present are manufactured, and some of the common types are shown in figure 4-2.

12

Master thesis for Henning Leifsen, Autumn 2007

Figure 4-2 Examples of random packing elements

While random packing has been the most employed packing, the use of structured packing is
expanding. These are of great variety. The structured packing offer the advantages of low
pressure drop for the gas and greater possible fluid flow rates, usually at the expense of more
costly installation than random packing. The use of packing in larger installations appears to
becoming more common.

Figure 4-3 Typical structured packing (intalox)

13

Master thesis for Henning Leifsen, Autumn 2007

4.3 Material and Energy Balance

Figure 4-4 shows a simple model of an absorption column containing either trays or packing.

Figure 4-4 Material balance diagram for countercurrent contactor

Equation 4.1 gives the material balance for the contactor column [3]:

GM (Y Y2 ) = LM ( X X 2 )
Where

(4.1)

GM' = CO2-free gas molar flow rate (kmol /( s m))

LM' = CO2-free liquid molar flow rate (kmol /( s m))
x
X = mole ratio CO2 in the liquid phase =
, where x = mole fraction
1 x
y
Y = mole ratio CO2 in the gas phase =
, where y = mole fraction
1 y

The equation of the operating line is given by rearranging equation 4.1:

Y = LM' / GM'(X-X2) + Y2

(4.2)

On rectangular coordinate paper this equation will be a straight line with a slope of LM'/GM'.
The coordinates at the ends of the operating line represent conditions at the ends of the
column. The known parameters are normally the feed gas flow rate (GM'), the mole ratio of
CO2 in the feed gas (Y1), the mole ratio of CO2 in the lean solvent (X2) and the required mole
ratio of CO2 in the product gas (Y2). The objective is to estimate the required liquid flow rate
and, finally, the dimensions of the column.

14

Master thesis for Henning Leifsen, Autumn 2007

The performance of counter-current absorbers can best be visualised with a diagram. In figure
4-5 two possible operating lines has been drawn. Line A represents a typical design, while B
represents the theoretical minimum liquid flow rate. The distance between the operating line
and the equilibrium curve represents the driving force for mass transfer at any point in the
column. Since line B actually touches the equilibrium curve at the bottom of the column, it
would require an infinitely tall column, and therefore represents the limiting liquid flow rate.

Figure 4-5 Operating line-equilibrium curve diagram for absorption column

Absorption increases the temperature within the column. It is therefore necessary to modify
the equilibrium curve so that it corresponds to the actual conditions at each point in the
column.
The amount of heat released between the liquid and gas streams is primarily determined by
the ratio of the total heat capacities of the two streams, LMCq/GMCp, where Cq and Cp are the
heat capacities of the liquid and gas respectively. With a high ratio (higher than 2), the liquid
carries the heat of reaction down the column, the product gas leaves at approximately the
temperature of the liquid feed, and the product liquid leaves at an elevated temperature
determined by the overall heat balance. The outgoing liquid is somewhat cooled by the feed
gas which leads to a temperature bulge within the column. With a low ratio (lower than 0.5)
the product gas carries most of the heat of reaction of the column. A ratio close to 1.0 splits
the heat of reaction between the gas and liquid products and might increase the temperature
well above that of the incoming streams [3].

15

Master thesis for Henning Leifsen, Autumn 2007

4.4

Column Height

Counter-current packed columns operate in a different manner compared to trayed columns.

In packed columns the fluids are in contact continuously in their path trough the tower, rather
than intermittently. The liquid and gas compositions therefore change continuously with
height of packing. Every point on an operating line represents conditions found somewhere in
the column. The height equivalent to a theoretical plate (HETP) is a frequently used
approach to describe the mass transfer efficiency of packing. HETP is defined as height of
packed zone divided by the number of theoretical plates achieved in packed zone. To estimate
the packing height, the calculated number of theoretical plates required is simply multiplied
by the HETP. If the efficiency of packing for a reason is not available, or a more detailed
examination of the column height is necessary, the column height may be calculated with the
use of absorption coefficients.

4.4.1 Absorption Coefficient

The absorption coefficient is based upon a two-film theory originally proposed by W. G.
Whitman [3] where it is assumed that the gas and liquid are in equilibrium at the interface and
that thin films separate the interface from the main bodies of the two phases. The liquid
absorption coefficient, kL, is defined as the quantity of material transferred through the liquid
film per unit time, per unit area, per unit of driving force in terms of liquid concentration. The
gas absorption coefficient, kG, is defined as the quantity transferred through the gas film per
unit time, per unit area, per unit of driving force in terms of pressure. The quantity of material
transferred from the body of the gas to the interface must equal the quantity transferred from
the interface to the body of the liquid. This balance is given as:
N CO = kG ( p pi ) = kG P( y yi ) = k L (ci c)
2

(4.3)

where
N CO = quantity of CO2 transferred per unit time, per unit area ( kmol /(m 2 s ) )
kG = gas absorption coefficient ( kmol /(m 2 s Pa) )
kL = liquid absorption coefficient ( kmol /(m 2 s kmol / m3 ) )
p = partial pressure of CO2 in main body of gas (Pa)
pi = partial pressure of CO2 in gas at interface (Pa)
P = total system pressure (Pa)
y = mole fraction of CO2 in main body of gas
yi = mole fraction of CO2 in gas at interface
c = concentration of CO2 in main body of liquid (kmol / m3 )
ci = concentration of CO2 in liquid at interface (kmol / m3 )
2

The partial pressure of CO2 at the interface is generated by the absorbed CO2 gas in the liquid
solution that still has not reacted with the amines. It can be seen from equation 4.3 that the
driving force for CO2 absorption and hence the mass flux of CO2 is greatest when the CO2
back pressure from the absorption liquid approaches zero and the partial pressure of CO2 in
the gas stream is at a maximum.

16

Master thesis for Henning Leifsen, Autumn 2007

4.4.2 The Gas-Liquid Interfacial Area

The gas-liquid interfacial area is an important factor for mass transfer in the gas absorption
process. A larger interfacial area is required to achieve greater mass transfer efficiency.
Determination of the mass-transfer area is considered to be the most extensive task in column
design because it involves both experiments and rather complicated calculation steps. The
calculations require a large number of fluid dynamic parameters and solvent physical
properties, as well as the geometry of column internals that constantly vary from location
within the column. In the case of CO2 absorption into aqueous solutions these calculations can
become even more difficult as the mass-transfer process involves exothermic chemical
reactions, causing a significant variation in the column temperature. The gas-liquid interfacial
area is defined as the transfer area divided by the volume:
a=

Where

dA
Ac dz

(4.4)

dA = transfer area (m2)

Ac = cross-sectional area of column (m2)

dz = h

= height of column (m)

4.4.3 Height of Column

The absorption coefficient data are used to express the height of the column. It is then
necessary to consider the changes in liquid and gas compositions that occur over the length of
the column. The quantity of material transferred to the quantity indicated to be transferred on
the basis of the absorption coefficient and driving forces has to be equated and integrated over
the length of the column as shown in equation 4.5 [3]:
p1

h = GM

p2

c1

plm dp
L
dc
=
2

k G a ( P(1 y )) P ( y ye ) L c k L a(ce c)

(4.5)

Where
h = height of packed zone (m)
G'M = superficial molar mass flux of inert gas (kmol /(h m 2 ))
a = interfacial area per unit volume of absorber (m2/m3)
ye = mole fraction of CO2 in equilibrium with a solution having the composition
of main body of liquid (Pa)
ce = concentration of CO2 in a solution in equilibrium with main body of gas
(kmol / m3 )
plm = log mean of inert gas pressures
k'G = kG (pBM/p) = special mass-transfer coefficient which is independent of gas
composition (kmol /(h m 2 Pa))
L = liquid flux ( kg /( h m 2 ))
L = liquid density (assumed constant) (kg/m3)
The subscript 1 refers to the bottom of the column, while the subscript 2 refers to the top of
the column.

17

Master thesis for Henning Leifsen, Autumn 2007

4.4.4 Volumetric Overall Mass-transfer Coefficient

The use of equation 4.5 for design requires knowledge of both kG and kL as well as the
equilibrium relationship and the interfacial area per unit volume of absorber. Usually it is
more practical to use overall coefficients which are based on the total driving force from the
main body of the gas to the main body of the liquid. These overall coefficients relate directly
to the contactor volume rather than to the interfacial area:
N CO adV = K G aP( y ye )dV = K L a (ce c)dV
2

Where

(4.6)

KGa = overall mass-transfer coefficient for gas ( kmol /( s m Pa) )

KLa = overall mass-transfer coefficient for liquid ( kmol /( s m (kmol / m3 )) )
V = volume of packing (m3)

The relationship between the individual film coefficients and the overall coefficients is given
as:
1
1
He
(4.7a)
=
+
K G a kG a k L a

1
1
1
=
+
K L a k L a HekG a

(4.7b)

He is the Henry`s law constant and is commonly defined as:

He =

pi
ci

(4.8)

In cases where Henry`s law does not apply, (pi-pe)/(ci-ce) is used instead. The overall
coefficients are only valid where the equilibrium line is straight over the operating region but,
nevertheless, they are widely used in reporting test data due to their convenience.
When chemical absorption takes place, as in the case for CO2 absorption by amines, the liquid
side mass transfer is improved. To describe the influence of this chemical reaction an
enhancement factor, Ie, can be added to equation 4.7[5]:
He
1
1
=
+
K G a kG a I e k L ,ref a

(4.9a)

1
1
1
=
+
K L a I e k L ,ref a HekG a

(4.9b)

Where
kL,ref = liquid side mass transfer coefficient without chemical reaction
( kmol /( m 2 s kmol / m3 ) )
Ie = enhancement factor

18

Master thesis for Henning Leifsen, Autumn 2007

4.4.5 Height of Column Using Overall Mass-Transfer Coefficient

More simplified forms of equation 4.5 are developed and are more commonly employed to
design the height of columns. These forms are sufficiently accurate for most engineeringdesign calculations, particularly for gas purification with low gas and liquid concentrations.
For these equations to apply, the following assumptions must be valid [3]:

The equilibrium curve is linear over the range of concentrations encountered, which
means that overall coefficients can be applyed.
The partial pressure of the inert gas is essentially constant over the length of the
column.
The CO2 content of gaseous and liquid phases are sufficiently low so that the partial
pressure and liquid concentration values may be assumed proportional to the
corresponding values when expressed in terms of mole of CO2 per mole of inert gas.

The column height can then be estimated, in terms of the overall gas coefficient and gas-phase
composition, by equation 4.10a, or by equation 4.10b if the overall liquid absorption
coefficient is available:
y1

h=

GM
dy

K G aP y y ye

(4.10a)

x1

h=

LM
dx

M K L a x xe x

(4.10b)

Where
x = mole fraction of solute in the liquid streams
xe = mole fraction of solute in the liquid streams at equilibrium
GM = molar flow rate value of gas per cross-sectional area (kmol / s m 2 )
LM = molar flow rate of liquid per cross-sectional area (kmol / s m 2 )
If a dilute mixture (<10% CO2) and a constant GM along the column is assumed, a simpler
expression for the mass balance can be given as:

d (GM y )
dy
GM
= K G aP( y yeq )
dz
dz

(4.11)

The height of the column can then be obtained by rearranging equation 4.11:
h

y2

y1

dz = h = K

GM
y1 yeq
GM
ln
dy =
K G aP y2 yeq
G aP ( y yeq )

19

(4.12)

Master thesis for Henning Leifsen, Autumn 2007

4.4.6 Selecting the Volumetric Overall Mass-Transfer Coefficient

The volumetric overall mass-transfer coefficient, KGa, is a combination of three contributions
associated with mass transfer, i.e., thermodynamics, kinetics, and hydrodynamics of the CO2
absorption system.
The value of the volumetric mass transfer coefficient is unaffected by the gas flow rate
through the absorber over a range of 30-100 kmol/(h m2))[6]. A constant mass flow rate of
CO2 is absorbed despite the increased mass flow rate of CO2 molecules at a higher gas flow
rate. Therefore the liquid phase mass transfer to the gas-liquid interface, which depends upon
diffusion of reagent molecules within the liquid phase, is considered to be the major
controlling factor.
If the CO2 concentration of the gas stream is increased to values up to 15%, it is found to
reduce the value of KGa for the system. However, the increase in CO2 partial pressure causes
the mass flux of CO2 absorbed in the system to remain relatively constant [6].
The increase in value of KGa on the basis of the increase of liquid flow rate is the result of two
different effects. Firstly, increasing the liquid flow rate increases the liquid phase mass
transfer rate to the gas-liquid interface, which directly increases availability of the amine at
this location and the liquid side mass transfer coefficient component of KGa. Secondly,
increasing the liquid flow rate can improve the liquid distribution on the packing surface
leading to an increase in the effective interfacial area component of KGa. The figure below
shows the effect of liquid flow rate on overall mass-transfer coefficient for the structured
packing Mellapak 500Y [7].

Figure 4-6 Effect of liquid flow rate on overall mass-transfer coefficient

It is important to note that the impact of liquid flow rate varies considerably from packing to
packing. This can be observed in figure 4-7 [7], which gives KGa values for three different
packings under the range of liquid flow rate.

20

Master thesis for Henning Leifsen, Autumn 2007

Figure 4-7 Overall mass transfer coefficient of different structured packings at various liquid flow rates

The temperature of the absorbing reagent solvent is found to have a complex effect on the
value of the mass transfer coefficient[6]. Figure 4-8 shows that an increase in liquid
temperature from 20C to 30C causes an increase in KGa. However, as the temperature of the
solution is increased from 40C to 65C the value of KGa decreases. It is proposed that this
effect is caused by an increase in the Henry`s Law coefficient with temperature which then
limits the capability of CO2 to transfer from the gas phase to the solution. If the liquid
temperature is further increased towards temperatures close to 120C the CO2 absorption
reaction will reverse as in the regenerator. The maximum KGa value in a 3 kmol/m3 MEA
system with a CO2 loading of 0.36 mol/mol is found to occur at approximately 36C[6].

Figure 4-8 Overall mass-transfer coefficient at various liquid temperatures

The amine concentration within the absorbing liquid has a straightforward effect on the CO2
absorption performance. An increase in the amine concentration provides a higher number of
amine molecules per unit volume and therefore a greater ability to absorb the gas-liquid
interface. This will increase the value of KGa and thereby improve the CO2 removal
performance of the system. However, it is important to balance factors such as corrosion and
the operating cost with this improvement in CO2 removal performance.

21

Master thesis for Henning Leifsen, Autumn 2007

The CO2 loading of the system is expressed as the number of moles of CO2 per moles of
amines in solution. The CO2 loading is therefore directly linked to the amine concentration
and yields an equally straightforward relationship between the CO2 loading and the mass
transfer coefficient of the system. An increase in the CO2 loading reduces the concentration of
active reagent molecules in the liquid and hence the KGa value decreases.

4.5 Column Diameter

The diameter of packed columns is usually based on flooding correlations (see below). The
Eckert correlation [3] below is a widely used correlation for estimating pressure drop in
structured packing. The Y axis is called the Flow Capacity Factor and the X axis the Relative
Flow Capacity. The flow capacity factor includes a packing factor, F, which is a characteristic
of the packing configuration. For most packings acceptable packing factor values are made
available by the packing vendor or the open literature. The equations for the Flow Capacity
Factor and Relative Flow Capacity are given below as:
G2 F 0.1
32.2 G ( L G )

Flow Capacity Factor

Relative Flow Capacity

L G
=
G L

Where

0.5

(4.15)

F = packing factor
1
kg /(m 2 h))
3
1.356 10
1
kg /(m 2 s ))
= gas flux ((lb / ft 2 s ) =
1.356 103
1
kg /(m 2 h))
= liquid flux ((lb / ft 2 h) =
1.356 103
= kinematic liquid viscosity (cst)
1
= gas density ((lb / ft 3 ) =
kg / m 3 )
16.019
1
= liquid density ((lb / ft 3 ) =
kg / m 3 )
16.019

G = gas flux ((lb / ft 2 h) =

G
L

G
L

(4.14)

22

Master thesis for Henning Leifsen, Autumn 2007

Figure 4-9 Generalized pressure drop correlation for packed towers

It is normally considered good practice to design for a gas rate that gives a pressure drop of
less than about 3.3 mbar per meter of packing. Many gas purification absorbers have a high
L/G ratio. The pressure drop may then exceed the above value, but the gas rate should not
exceed 85% of the rate that results in a pressure drop of 12.5 mbar per meter of packing.
Systems that tend to foam should be operated to give a low pressure drop and vacuum
systems may require an even lower pressure drop to minimize overall column pressure drop.

4.6 Flooding
When the inlet gas flow rate is so high that it interferes with the downward flow of the solvent
liquid, it may cause an upward flow of the liquid through the tower. This is known as flooding
[8]. Most absorbers are designed to operate at no more than 70% of the maximum gas velocity
that can cause flooding. Besides a high inlet gas flow rate, low circulation rates and small
diameter towers could also lead to flooding. Flooding conditions in random packings depend
on the method of packing (wet or dry) and settling of packing. Flooding velocities for
structured packings will generally be considerably greater than for random packing[4], i.e.
structured packing is more resistant to flooding.

23

Master thesis for Henning Leifsen, Autumn 2007

4.7 Number of Equilibrium Stages

In order to determine the number of equilibrium stages required in a cascade, the equilibrium
characteristics of the system and material-balance calculations are required. The design of
absorbers and strippers typically involves a computer-assisted, tray-by-tray (or section-by
section of packing) heat- and material-balance calculation to determine the required number
of equilibrium stages. The required number of equilibrium stages is related to the required
number of actual trays based on an estimated tray efficiency. Tray efficiencies are available
for various services in literature.

4.8 Tray Efficiency

The number of actual trays in an absorber is related to the number of theoretical trays by a
factor known as the tray efficiency. The overall tray efficiency is defined as the ratio of
theoretical to actual trays required for a given separation[3]. The overall tray efficiency can
only be used after the total number of theoretical trays has been calculated using a graphical
or analytical technique. To relate the actual performance to the theoretical trays for individual
trays, the Murphree vapour efficiency is used. This is defined as:
Emv =

y p y p+1
y pe y p+1

(4.16)

Where
yp = average mole fraction of solute in gas leaving tray
yp+1 = average mole fraction of solute in gas entering tray (leaving tray below)
ype = mole fraction of solute in gas in equilibrium with liquid leaving tray
Murphree tray efficiency values can be used to correct the individual steps in graphical
analysis of the number of trays required.

4.9 Residence Time Requirement

The time of contact is closely connected to the stage efficiency, whereas for continuouscontact equipment the time leads ultimately to the volume or length of the required device.
There are several factors which help establish the time necessary. Material balances permit
calculation of the relative quantities required of the various phases. The equilibrium
characteristics of the system establish the ultimate concentrations possible, and the rate of
transfer of material between phases depends upon the departure from equilibrium which is
maintained. The rate of transfer is also depending on upon the physical properties of the
phases as well as the flow regime within the equipment.

24

Master thesis for Henning Leifsen, Autumn 2007

5 Amines Property Package in HYSYS

The Amines Property Package in HYSYS is a property package designed especially for
modelling of alkanolamine treating units in which H2S and CO2 are removed from gas
streams. The package contains data to model the absorption/desorption process where
aqueous solutions of single amines and aqueous solutions of blended amines are used. The
Amine Property Package uses technology developed by D. B. Robinson and Associates Ltd.
[9] to model the equilibrium solubility of acid gases in aqueous solutions. A non-equilibrium
stage model which is based on the tray efficiency concept is used to simulate the performance
of contactors and regenerators.
The generalized stage model as shown in figure 5-1 gives the flow geometry and
nomenclature for an individual stage in a column. The fundamental concept used is that the
rate of absorption/desorption of acid gases to/from the amine solution must be considered as a
mass-transfer rate process.

Figure 5-1 Stage j in a column simulated in HYSYS

where
F = stage feed stream (kmole/h)
L = liquid stream travelling to stage below (kmole/h)
V = vapour stream travelling to stage above (kmole/h)
SL = liquid side draw from stage (kmole/h)
SV = vapour side draw from stage (kmole/h)
Q = energy stream entering stage (kJ/h)
H = enthalpy (kJ/h)
h = enthalpy (kJ/h)
Z = mole fraction in feed stream
x = mole fraction in liquid phase
y = mole fraction in vapour phase
j = stage number
i = component number

25

Master thesis for Henning Leifsen, Autumn 2007

The rate process depends on the equilibrium and kinetic parameters that describe the acid
gas/amine system. This model uses a modified Murphree-type vapour efficiency to account
for the varying mass-transfer rates of individual acid gas components. The acid gas stage
efficiencies are, in turn, functions of mass-transfer coefficients and the mechanical design of
the tray.
A modified Newton-Raphson method is used to solve the rigorous non-linear stage equations
simultaneously for temperature, composition and phase rates on each stage in a column.
The Amines property package defines the stage efficiency as:

(V j + SV j ) y j V j +1 yij +1
(V j + SV j ) K1 j xij V j +1 yij +1

Where
= stage efficiency
K = equilibrium ratio
The stage efficiency is a function of the kinetic rate constants for the reactions between each
acid gas and the amine, the physico-chemical properties of the amine solution, the pressure,
the temperature and the mechanical tray design variables such as tray diameter, weir height
and weir length. The stage efficiencies could be specified or calculated by HYSYS.
The range of applicability for the amines in the Amine Property Package is given as 25126.7C (77 -260F). To avoid physical amine degradation into corrosive products, HYSYS
recommends a maximum reboiler temperature of 137.8C (280F).
The most important equations for the non-equilibrium stage model in the Amine Property
Package are shown in appendix B.

26

Master thesis for Henning Leifsen, Autumn 2007

6 Minimizing the Energy Consumption in Amine-Based

Absorption Plants
The operational cost of the overall gas sweetening process is mainly determined by the
regeneration of the loaded liquid in the desorber. Several parameters have been identified and
varied over a given span in order to see the performance effects of the regeneration system.
This chapter gives the results of previous work in order to minimize the energy consumption
in amine-based absorption plants.

6.1 Effect of Condenser Temperature

The condenser temperature determines the amount of steam that follows along with CO2 to
the compression unit as well as the temperature of the liquid reflux returning to the column.
The higher this temperature is, the more amines and water follows the CO2 leaving the
condenser. On the other hand, less steam is required for heating the liquid flowing down the
column. Figure 6-1 shows the effect of varying the condenser temperature between 45-75C,
with respect to reboiler duty and the total amount of steam that follows at the CO2 out of the
reboiler with a fixed CO2 pickup of 0.32[10].

Figure 6-1 Condenser temperature versus reboiler duty[10]

The figure shows that an increase in condenser temperature from 45C to 75C only reduces
the reboiler duty marginally. The amount of steam following the exit CO2 is on the other hand
increased from about 50 to 200kmol/h. This occurs because only a small amount of the total
liquid that flows down the main packing is refluxed liquid. A 30C change in condenser
temperature will only change the temperature of the liquid meeting the main packing after
being mixed with the reflux with about one degree. As the temperature increases, the amount
of MEA following the CO2 will become more significant. At 60C, the amine loss is
calculated to about 0.4 kg MEA/ton CO2 recovered. At 45C the loss of MEA is reduced to
the half, but the energy consumption is higher. An economic optimum between these two
factors is possible to calculate. It is also important to take the environmental concerns
regarding MEA loss into account as well as the steam content of the exit CO2 steam.

27

Master thesis for Henning Leifsen, Autumn 2007

6.2 Effect of Cross-Flow Heat Exchanger Performance

Varying Temperature Approach
The rich amine inlet temperature to the desorber is determined by the performance of the
rich/lean cross-flow heat exchanger. This performance can be defined by the temperature on
the pinch side, which in this case is the cold side. The minimum acceptable temperature
approach is normally set to 10-15C, but more modern heat exchangers can bring this
minimum approach down to less than 5C[10].The results for all three temperature
approaches are given in table 6-1.
Table 6-1 Varying the cross-flow heat exchanger pinch temperature approach[10]

When the pinch is reduced from 15C to 5C, less steam is needed for desorption in the tower
at the same pressure and less water is condensed in the tower. The loading is also reduced
from 0.45 to 0.38, while the inlet rich feed temperature to the desorber increases with about
one degree. Furthermore, the overall steam consumption parameter is lowered from 4.53 to
4.44 GJ/t CO2 and both the liquid and vapour flow in the desorber are lower.
Interaction Between Cross-Flow Heat Exchanger and Desorber Column
As the heat exchanger duty is increased the vapour fraction will increase and the use of a flash
separator will reduce the loading significantly. When the feed, which is at its equilibrium,
meets the water wash reflux, the solvent is cooled and diluted. This could result in a drop in
the equilibrium partial pressure below the partial pressure of the vapour phase in the tower.
To keep desorption occurring throughout the column, the cross-flow heat exchanger should
only have a limited duty. The duty should be chosen so that the inlet rich feed to the desorber
column, after being mixed with the reflux, is close to the desorber exit gas phase equilibrium.
In other words, the more efficient the heat exchanger is, the higher the chances are for the
desorber to be performing below optimum. The size of the heat exchanger should therefore be
chosen depending on the desorber size and efficiency[10].

If a larger and more efficient heat exchanger is used, the temperature of the rich stream to the
desorber will not increase as much. This is because the bubble point is exceeded. Instead CO2
and other vapours will start desorbing. By promoting turbulence and increasing local heat
transfer this might serve the heat exchanger favourably. If the vapour fraction becomes too
large, however, it can decrease the heat transfer coefficient and give a reduced total heat
transfer. This can be prevented by boosting the rich amine pressure to above the bubble point
before the cross-flow heat exchanger and followed by a pressure reduction and flash.

28

Master thesis for Henning Leifsen, Autumn 2007

6.3 Effect of CO2 Capture Efficiency

Thermal Energy
The thermal energy required for the regeneration of the amines is a function of the design and
operating parameters of the CO2 capture system, including amine concentration, CO2 mole
fraction, lean loading and CO2 capture efficiency. The heat in power-plant applications is
usually supplied in the form of low-pressure steam extracted from the steam turbine. The
steam extraction leads to a loss in power generation and the thermal energy requirement can
therefore be expressed in terms of an equivalent electrical energy penalty. Figure 6-2 shows
the amine regeneration energy requirement per unit of CO2 captured, and the equivalent loss
in power generation[11]. The numbers are based on detailed studies of LP steam extraction
points for modern steam turbines. Higher capture efficiency requires more efficient use of the
loading capacity of the amines and, hence, more regeneration energy. The specific thermal
energy requirement increases therefore as the CO2 removal efficiency increases.

Figure 6-2 Sorbent regeneration heat requirement (kJ/kg CO2), and the equivalent loss in power
generation (MWh/tonne CO2) as a function of CO2 capture efficiency [11].

Electrical Energy
The gain in CO2 capture efficiency by building a higher absorber column is found to be quite
low [11]. However, by keeping the column height and increasing the liquid/gas ratio a higher
CO2 capture efficiency is achieved. This requires a larger column diameter. Since there is a
practical limit on column diameter size (also in terms of available column vendors), the flue
gas flow rate must be reduced to increase the liquid/gas ratio further. Thus, it may be required
to increase the number of trains to achieve higher CO2 removal for a given flue gas flow rate.

The flue gas blower power requirement will remain constant if the column height is
unchanged. However, a higher capture efficiency will increase the sorbent pumping power
requirement. Overall, the total electrical energy requirement per unit of CO2 captured
decreases as the capture efficiency increases, as shown in figure 6-3.

29

Master thesis for Henning Leifsen, Autumn 2007

Figure 6-3 Electrical energy requirement for capture and compression (MWh/tonne CO2), and
comparison to the equivalent thermal loss in power generation (MWh/tonne CO2) [11].

Figure 6-4 combines the two results in figure 6-3 to show the total energy penalty of the
amine-based CO2 capture system in terms of MWh/tonne CO2. This graph includes both the
electrical energy requirement and the equivalent power loss that is due to the solvent
regeneration heat requirement. At a 90% capture efficiency 53% of the total energy comes
from sorbent regeneration. The electrical energy requirements for capture and compression
are 37% and 10% of the total respectively [11]. The minimum overall energy requirement is
observed to be at a capture efficiency of about 86%.

Figure 6-4 Total energy requirement (MWh/tonne CO2) of the amine-based CO2 capture system, as a
function of CO2 capture efficiency.

30

Master thesis for Henning Leifsen, Autumn 2007

The net power output of a power plant with CO2 capture as a function of CO2 capture
efficiency is shown in figure 6-5. Although the curves may appear linear at the scales
presented, the slope of the line gradually increases with higher capture efficiency.

Figure 6-5 Net power output of a power plant with amine-based CO2 capture system as a function of CO2
capture efficiency (MWg = gross power plant size) [11].

Capital Cost of the CO2 Capture System

Figure 6-6 and 6-7 show how the total capital cost varies with an increasing CO2 capture
efficiency. The minimum capital cost in both figures is different from the 86% capture
efficiency at which the total energy requirement of the system was determined to be minimum
given in figure 6-4. Although the energy penalty strongly affects the net capital cost via its
impact on net power output, other factors (especially the number of operating trains for CO2
capture and compression) also influence the location of the minimum capital cost per unit of
CO2 removed.

Figure 6-6 CO2 capture system capital cost as a function of the CO2 capture efficiency for two plant sizes.

31

Master thesis for Henning Leifsen, Autumn 2007

Figure 6-7 CO2 capture system capital cost normalized on the CO2 removal capacity, as a function of the
CO2 capture efficiency for two plant sizes.

6.4 Effect of Rich and Lean Amine Level

The amount of steam needed for the desorber is highly dependent on the lean amine loading.
A higher lean amine loading results in higher desorber driving forces, which reduces the
steam consumption per ton CO2. The lean amine load can be increased in two ways, either by
keeping the rich amine loading and increasing the liquid circulation rate, or by increasing both
the rich and lean amine loadings such that the cyclic capacity, and circulation rate, remains
the same.
Increasing Circulation Rate and Lean Loading Level
Table 6-2 gives the results of varying the lean loadings from 0.16 to 0.20 and 0.24, while
increasing the circulation rate so that the CO2 production rate is kept constant. The table
shows that the steam consumption reduces radically as the circulation rate and lean loading is
increased. Even though this will have a negative effect on the absorber operation, the driving
force at the top of the absorber will still be very high. The additional cost associated with the
increase in pump energy required is very small compared to the overall energy requirement
for steam production. An increase in the liquid flow rate of 30% will give a negligible
efficiency penalty compared to the energy required for steam production.

32

Master thesis for Henning Leifsen, Autumn 2007

Table 6-2 Effect of: a. Increasing circulation rate and lean loading level and b. Increasing the rich and
lean levels simultaneously[10]

Increasing the Rich and Lean Levels Simultaneously

Part b of table 6-2 shows that the steam consumption is drastically reduced as the rich and
lean loads are increased.

Table 6-2 shows that an increase in the amine loading by either one of the two methods
reduces the desorber energy consumption very effectively, and should be considered when
optimizing a removal plant. However, it is also important to take the effect on the absorber
into account.

6.4.1 Effect of Different Lean Solvent Loading Including the Effect of the CO2Removal Percentage
Figure 6-8 shows that the thermal energy requirement decreases with increasing lean loading
until a minimum is reached. If the lean load is further increased, the thermal energy
requirement will increase as well. Figure 6-8 also shows that decreasing the degree of
regeneration will decrease the thermal energy requirement. Furthermore, at very low degrees
of regeneration the amount of solvent needed increases considerably and this would require
more regeneration energy. Therefore, an optimal lean loading will exist. The figures of this
chapter are based on a bituminous coal fired plant (CF), and the numbers would differ from a
natural gas combined circle (NGCC).

33

Master thesis for Henning Leifsen, Autumn 2007

Figure 6-8 Thermal energy requirement at various CO2/amine lean loadings for different CO2 removal
(CF) [12]

Since the lean load is varied by varying the solvent circulation rate, a high lean load also
represents a high solvent circulation rate as shown below.

Figure 6-9 Solvent flow rate requirement at various CO2/amine lean solvent loadings for different CO2
removal [12]

The amount of thermal cooling water decreases with increasing lean solvent loadings as
shown in figure 6-4. The reason that the amount of cooling water remains constant for a
loading between 0.26 and 0.33 mol CO2/mol MEA in this graph, is that the temperature for
high lean solvent loadings was allowed to increase to meet the requirement of a closed water
balance[12]. The consequence of this is that the absorber operates at a higher temperature
which allows evaporation of water at the top of the absorber to maintain a closed water
balance in the complete process. If the lean solvent temperature was kept constant at high
solvent flow rates, this would have led to excessive condensation in the absorber. This water
would have to be removed in the stripper.

34

Master thesis for Henning Leifsen, Autumn 2007

Figure 6-10 Cooling water consumption at various CO2/amine lean solvent loadings for different CO2
removal [12]

The cost of electricity (CoE) could be described as a function of lean loading for different
degrees of CO2 removal as shown in figure 6-4.The range where the cost of electricity is at its
minimum, is the same range where the thermal energy is at its minimum. This indicates that
there is a close link between the thermal energy requirements and increased electricity costs.

Figure 6-11 Cost of electricity as a function of lean solvent loading for different CO2 removal [13]

6.4.2 Effect of Different Lean Solvent Loading Including the Effect of the MEA
Weight%
The thermal energy requirement decreases substantially with increasing MEA concentration
in the absorption solution. The increasing MEA concentration will however have a corrosive
effect on the process equipment. This problem could be solved by adding corrosion inhibitors
to the process. Figure 6-12 shows that there are clear benefits by increasing the MEA
concentration in the solvent. By increasing the MEA concentration from 30 to 40 wt% the
thermal energy requirement is decreased with 5-8% [12]. The cooling water and solvent
consumption are also decreased as a result of a higher MEA concentration as shown in figure
6-13 and figure 6-14.

35

Master thesis for Henning Leifsen, Autumn 2007

Figure 6-12 Thermal energy requirement at various CO2/amine lean loadings for different MEA wt%.
(for a bituminous coal fired plant) [12]

Figure 6-13 Solvent flow rate requirement at various CO2/amine lean solvent loadings for different MEA
(wt.%)

Figure 6-14 Cooling water consumption at various CO2/amine lean solvent loadings for different MEA
(wt.%)

36

Master thesis for Henning Leifsen, Autumn 2007

6.5 Effect of Reboiler Pressure

By increasing the pressure in the reboiler, one will increase both temperature and pressure
profiles of the column. The effect of varying the reboiler pressure between 180 and 220 kPa is
shown in table 6-3.
Table 6-3 Effect of varying reboiler pressure[10]
Simulation
Reboiler Pressure
Reboiler heat duty
kPa
kJ/h
Press_180kPa
180
157.5
Base_200kPa
200
143.5
Press_220kPa
220
132.8
Rich loading 0.48, lean loading 0.16

Specific Steam consumption

GJ/ton CO2
4.93
4.49
4.15

T reboiler
C
118.2
121.2
123.9

The table shows that an increase in pressure will enhance stripping and reduce CO2
compression cost, if compression is required. By increasing the pressure, more steam
condenses and the CO2 mass transfer throughout the column will increase. As shown in table
6-3 an increase in pressure from 180 to 220 kPa will lead to a reduction of 15% in the reboiler
duty. The disadvantage of increasing the pressure is that the boiling point of the amine
solution will increase also, which may lead to higher amine degradation rates and to increased
corrosion problems.

6.6 Effect of Varying Reboiler Equilibrium Efficiency

Larger reboiler units may operate at different equilibrium stages. This has an effect on the
performance on the regenerator. Equilibrium efficiency is defined as the approach to phaseequilibrium of CO2 that can be achieved in the reboiler by reducing the equilibrium value of
YCO 2 by a certain percentage. Table 6-4 shows the effect of varying the reboiler equilibrium
efficiency with a constant rich and lean loading of 0.48 and 0.16 respectively.
Table 6-4 Effect of varying reboiler equilibrium efficiency[10]

By reducing the reboiler efficiency, the CO2 gas concentration out of the reboiler will reduce
proportionally. The total CO2 leaving the desorber does however not decrease by the same
magnitude. The table shows that with a 70% approach to equilibrium the amount CO2 stripped
in the column packing is 391.7 kmol/h, while its only 314.3 kmol/h at 100%. Thus, there are
two counteracting effects: a more efficient reboiler indicated by a high outgoing CO2 partial
pressure will be counteracted by the effect of a less efficient CO2 mass transfer in the desorber
column and vice versa.

37

Master thesis for Henning Leifsen, Autumn 2007

The steam consumption is lower with a reboiler that approaches equilibrium, and the
efficiency is therefore an important element in the optimization process. Reducing the
efficiency from 100 to 70% will however only increase the steam consumption by 10% and it
is thus not the most important factor.

6.7 Importance of the Accuracy of the Equilibrium Model

The accuracy of the vapour-liquid equilibrium representation (VLE) is very important for the
results of the simulation as these models always depend on experimental data to which they
have been fitted. At high CO2-loading the equilibrium partial pressure increases very rapidly
with loading, especially at high temperatures. The deviations in this area might therefore be
significant. Figure 6-15 gives an example of this.

Figure 6-15 Tuned vapour liquid equilibrium (VLE) data for 30 wt% MEA at stripper conditions. a.
normal plot, b. log-log plot [10]

38

Master thesis for Henning Leifsen, Autumn 2007

Table 6-5 shows how the result of the simulation is affected by the equilibrium model when
the equilibrium curves are adjusted at high temperatures. The curve Eq1_3 gives higher
equilibrium partial pressures for CO2 as the temperature is increased at a given loading
compared to the base case Eq1_5. This yields a higher driving force for desorption, which
again yields higher desorption fluxes. This is the case in both the flash drum (if a flash drum
is used), reboiler and in the column since all these units operate at a high temperature. To
reach a desired lean loading of 0.16, however, a higher temperature in the reboiler is needed.
This leads to a higher lean temperature in the cross-flow heat exchanger, which gives a higher
inlet rich temperature or a higher vapour phase fraction in the flash unit. An accurate VLE
description at desorber temperatures is therefore critical for the models performance. A 2%
shift in the equilibrium line at 120C changes the overall steam consumption in the
simulations with more than 15%.
Table 6-5 Effect of varying tuning factor in equilibrium model[10]

39

Master thesis for Henning Leifsen, Autumn 2007

40

Master thesis for Henning Leifsen, Autumn 2007

Modelling Theory

In simulation modelling there are a few important things to be aware of. There are several
different simulation programmes, but not all the programmes are designed for the same
purpose. The first step will be to find the right programme.
All process simulation programmes uses preinstalled thermodynamic models that include a
predefined set of unit operational modules. To be able to choose the correct package it is
necessary to know the process behind the simulation. Sometimes more than one fluid package
can be used, and the engineer has to decide which one to choose.
For two cases to be comparable it is necessary that they use the same basis. If the subject is to
find a way of reducing the energy consumption, the quality of the energy used needs to be
taken into consideration. Just as important is it that the product from the process is the same
for all cases. If not a comparison of the energy consumption would be useless.

7.1 Selection of Process Simulation Programme

A chemical process simulator calculates a thermodynamic consistent solution for a process
flow diagram based on input from the user. The simulator contains an input and an output
section, a library of unit operation models, a database for physical properties, a solver library
and an administrative section. The simulator performs calculations based on the instructions
of an engineer and helps the engineer making decisions by providing the necessary
information. The three most well-known commercial process simulators are:

ASPEN Plus
PRO-II/Provision
HYSYS

ASPEN and PRO-II are steady state process simulators with process optimization features,
while HYSYS provides both steady state and dynamic simulation features. ASPEN has the
largest number of users followed by PRO-II and then HYSYS. They all have an extensive
database of compounds for physical properties and a large collection of unit operation models
and thermodynamic models. All the three simulators have Windows-based PC-versions.

7.2 Selection of Thermodynamic Package

Fluid packages contain all necessary information regarding pure component flash and
physical property calculations. By using a fluid package it is possible to define all the required
information in a single location, making the information easy to modify. The fluid packages
can be imported and exported as completely defined packages for use in any simulation. It is
also possible to use multiple fluid packages within one simulation. The choice of fluid
package is depending on the fluid or mixture of fluids as well as pressure regime used in the
process.

41

Master thesis for Henning Leifsen, Autumn 2007

7.3 Heat versus Mechanical Work

The total mechanical energy used per kg CO2 sent to compression is a valuable benchmark for
evaluating the efficiency of a capturing process.
Heat at a certain temperature can only be partially converted into work/electricity. Generally
heat is considered as having a lower energy quality than mechanical work. To be able to
calculate the total mechanical work for the capturing process, the heat duty from the reboiler
has to be corrected for the difference in energy quality. This factor can be found from the
graph below, which is based on a combined power plant with a condenser pressure of 0.04
bara. The graph is calculated for heat, in the form of saturated steam, which is taken from the
power plant steam turbine and returned at a temperature of 70C [14].

Figure 7-1 Ratio () between reduced electricity production and heat outlet from a power plant [14].

The value of depends on the temperature of the steam used. The steam from the power
production is however not used directly. First it has to be transported from the power plant to
the sweetening plant. Even though the distance usually is short, this will lead to an additional
pressure drop. It is necessary to account for this when the interaction with the graph is made.
A general equation for the mechanical work in a sweetening process is given below [5].
E=
where

Q& + W&co + W&blower + W& pumps

m&

(7.2)

E = total mechanical work (MJ/kg CO2)

Q& = heat from saturated steam used for separation (MJ/h)
= factor that describes the different energy quality between electricity
and heat
&
Wco = CO2 compression work (MJ/h)
&
Wblower = blower compression work (MJ/h)
W& pumps = pump work (MJ/h)
m& = mass of CO2 sent to compression (kg/h)

42

Master thesis for Henning Leifsen, Autumn 2007

Methodology for Economic Evaluation

Even though a parameter variation could improve the energy efficiency of a process, it might
still not be the optimum solution. As the main objective for a company is to increase the
profit, optimum technical solutions have to be weighted against the extra investment expenses
(CAPEX) and operational expenses (OPEX).
There are several different methods for making an economic evaluation of a project. The
choice of method depends on which information is available and what the project is compared
to. Usually more than one method is chosen.

8.1 Investment Expenses (CAPEX)

The investment expenses depend on the size and type of equipment needed, as well as the
amount of equipment needed. The choice of equipment is based on several different
parameters. The choice of equipment material is usually the cheapest material that is good
enough. In other words, it is not practical to build a process plant that would last for 150 years
if it is meant to be in operation for 15 years. It is however important that the equipment fulfils
the safety requirements and performs satisfactory.
The main investment items for the absorption system are the two towers, the inlet cooler, and
the lean/rich cross-flow heat exchanger.
Estimating Cost by Scaling
The different equipment is measured differently when a cost calculation is performed. During
screening of process alternatives it is for example normal to scale the heat exchangers and
columns by area, while the weight of the separators is more interesting. For scale-up purposes,
the following equation can be used to determine the scaled base unit cost.

C2 = C1 (size2 / size1)

(8.1)

where
C2 = scaled base unit cost (MNOK)
C1 = base unit cost (MNOK)
size1 = base unit size
size2 = scaled base unit size
= size exponent

43

Master thesis for Henning Leifsen, Autumn 2007

Simulation programs like HYSYS can help calculating many of the necessary parameters, but
sometimes some additional calculations are necessary. Equation 8.2 shows how the area of
heat exchangers can be calculated using Q and LMTD found in HYSYS and a U found from
tables [15].
A=
where

Q
U LMTD

(8.2)

A = area (m2)
Q = heat duty (kJ/h)
U = overall heat transfer coefficient (kJ/Chm2)
LMTD = logarithmic mean temperature difference (C)

The necessary column area can be calculated based on a maximum superficial gas velocity.
This maximum velocity is depending on type of packing and this information is available
from the vendor for most packings. The column area is calculated as a function of the largest
volume flow of gas in the column and a specified maximum gas velocity as shown below.
Acolumn =
where

8.2

V&
wG

(8.3)

Acolumn = column area (m2)

V& = maximum volume flow through column (m3/s)
wG = maximum superficial gas velocity (m/s)

Operating Expenses (OPEX)

The main operating expenses for a CO2 sweetening plant are connected to the consumption of
steam and electricity. Electricity only amounts to about 10-20% of the total energy required,
and the largest consumers are the product CO2 compressors, the inlet flue gas blower, and to a
minor extent, the two pumps for amine recycling.
The main thermal energy for separation is added to the process as low pressured steam to the
stripper reboiler. The steam might be extracted from the gas power plant, or from a separate
steam generating boiler as a part of a utility system. This is by far the major energy sink in the
process and amounts to about 80-90% of the total energy requirement. The steam can be
converted to equivalent mechanical energy by multiplying it with a factor found from figure
7-1. The operating expenses for the steam can then be calculated by multiplying the steam
equivalent mechanical energy with the price of electricity.

44

Master thesis for Henning Leifsen, Autumn 2007

The amount of cooling water used is also an important parameter when the operating cost is
being calculated. The largest sources are the stripper condenser, the compressor intercoolers,
the lean liquid cooler, and the flue gas cooler (if a flue gas cooler is included). The total
consumption of cooling water is found by summing all the cooling water in all heat
exchangers used for cooling, and the cost of the cooling water is calculated by equation 8.4:
Pc = Fc Pc ol

(8.4)

Where
Pc = price of cooling water (NOK/year)
Fc = total consumption of cooling water (m3/year)
Pcol = price of cooling water per volume (NOK/m3)
The cost of operating an absorption plant varies with the complexity of the technology used.
A more complex technology might involve more maintenance and thereby the need of more
operators on the plant. If there is not an extensive change in technology applied, the
maintenance can be calculated as a percentage of the investment cost.

8.3

Profitability Analysis

Net Present Value

A standard method for the financial appraisal of long-time projects is the net present value
(NPV). I this method, a discount interest rate i is specified. This may be based on the average
rate of return that the company is currently obtaining on its capital, or the bank prime lending
rate. The NPV can be calculated using equation 8.5[16].
j =n

NPV =
j =0

(cash flow) j
(1 + i ) j

(8.5)

where
NPV = net present value (NOK)
cash flow = annual cash flow (NOK/year)
i = interest rate (%)
n = project life (year)

45

Master thesis for Henning Leifsen, Autumn 2007

46

Master thesis for Henning Leifsen, Autumn 2007

9 Simulation Cases
In all aspects of industrial activity, one must continuously try to improve the performance and
it is common to aim at a target for ultimate or best operating point. In almost all engineering
problems, the functions and variables can be converted into NOK. Hence, most engineering
optimization problems can be reduced to a minimization of cost or maximization of profit.
The basis for this master thesis was a post-combustion capture plant using amine absorption.
The parameters and dimensions were similar to a commercial plant intended to recover CO2
from flue gas from a gas power plant at atmospheric pressure. The simulations were
performed in HYSYS. The choice of process simulation programme was stated in the scope of
work. HYSYS is also widely used for amine process simulation in industry and it is known as
one of the best programmes for simulation with MEA amines[15].
Since gas based power plants exhaust flue gas at atmospheric pressure, MEA was chosen as
the preferred amine. MEA amines have a high reaction rate and are therefore the most used
amines for absorption at atmospheric pressure. The amine fluid package was used for all
streams that contained amines, while the Peng-Robinson fluid package was used for cooling
water.
The first and most time-consuming part of this master thesis was the making of a base case
simulation model. It was desired to make a model with realistic parameters and simple
solutions.
The scope of work left the choice of parameters to be optimized open. As the major energy
consumption in a capture plant is in the desorption process, the main focus was set on
parameters that affected the desorber column directly. The aim was to find an optimum
operating point in terms of energy consumption and economy. In order to make an
economical evaluation both the difference in operational costs and investment costs were
included.
Three parameters were varied with respect to achieve an optimum operating point;
the solvent circulation rate,
the rich/lean cross-flow heat exchanger minimum temperature approach and
the desorber pressure.
All these parameters have a significant effect on the desorber operation and an optimization
of these parameters can reduce the expenses of a capture plant greatly. For each parameter
variation the other parameters were kept constant at the base case value. In the end the
optimal point for each varied parameter were included in the same model. This will not
necessarily be the optimized solution since these parameters depend on each other. It will
however give an indication of the possible improvement in energy consumption and economy.
A fourth parameter, the absorber diameter, was varied to get a better understanding of column
design. For this purpose HYSYS has its limitations and an additional programme and
calculations by hand were necessary. It was found difficult to include these results in the
optimization of the other parameters. The variation of the absorber diameter was therefore
presented as a separate part of the results.

47

Master thesis for Henning Leifsen, Autumn 2007

9.1 Base Case Model

The base case for the simulation in HYSYS is shown in figure 9-1 and figure 9-3. Since some
of the water and amines followed the flue gas out of the absorber and the CO2 to compression,
two additional streams were added to the simulation, keeping the circulation rate of water and
amines constant. The streams were added right before the absorber inlet. To simplify the
simulation model, a water scrubber was not included in the absorber. Neither was an amine
filter or a reclaimer. This lead to a relatively high consumption of amines. The consumption
of amines did however not vary significantly for the different cases.

Figure 9-1 Base case for simulation

The columns both used packing material. The number of stages in the absorber and the
stripper was set to 11 and 32 respectively, based on results from a simulation model with
similar parameters[15].
The total pressure drop for the absorber was for the base case set to 0.05 bar. The
determination of this pressure drop was based on a similar model which included a washing
section [15]. The stripper pressure drop was set to 0.138 bar based on the same model as the
absorber. The flue gas was blown up to a pressure of 0.01 bar above the absorber bottom
pressure and entered with temperature of 107.9C. To reduce the investment cost, a flue gas
cooler was not included.
To simulate the height of the absorber and the stripper, two valves were added to decrease the
pressure of the streams entering the top of the columns. The pressure drops through the valves
were about 5 bar.
The lean liquid entering the absorber was cooled to 21C by the lean amine cooler, while the
condenser outlet temperature was kept constant at 28C. The condenser and lean amine cooler
both used water at 8C and 4 bara as coolant. The pressure loss for the cooling water was set
to 0.5 bar and the water was specified to exit with a temperature of 18C.
The capture rate was set to 85%, which for the base case was attained by an amine wt% of
30%, a solvent circulation rate of 2358 ton/h and a reboiler duty of 5.22 108 kJ/h. An
accuracy of 0.1% for the capture rate was used for all cases.

48

Master thesis for Henning Leifsen, Autumn 2007

A total mechanical work, which included the equivalent mechanical energy for the reboiler
steam, was calculated to easily compare the different cases in terms of energy consumption
and operational cost. To calculate the total mechanical work equation 7.2 was employed. The
value of was found from figure 7-1. The condensing temperature of the steam used in the
reboiler was 143C. With an additional pressure drop in the steam transportation, was
chosen to be 0.23. The same value was used for all cases since it was assumed that all cases
used the same steam. This resulted in a total mechanical work of 1.441 MJ/kg CO2 for the
base case.

Figure 9-2 Intersection point where was found

The CO2 compression was done in 3 stages with intermediate cooling as shown in figure 9-3.
An equal pressure ratio was used, based on the pressure into the first compressor stage and a
specification of 80 bara out of the third compressor stage. A pump further raised the pressure
to 110 bara. The adiabatic efficiencies were set to 85%, 85% and 80% for compressor stage 1,
2 and 3 respectively. The pump adiabatic efficiency was set to 75%.

Figure 9-3 Base case, CO2 compression and drying

The pressure losses through all 3 heat exchangers were set to 0.3 bar for both the gas stream
and the coolant. As a coolant, water at 8C and 4 bara was chosen. The water was specified to
heat up to 18C, while the temperature of the CO2 was specified to be 20C out of all three
heat exchangers.

49

Master thesis for Henning Leifsen, Autumn 2007

Two separators were included in the model to extract the majority of the remaining water
from the CO2 gas. This water is usually returned to the capturing process so that less water
has to be added.
Appendix A gives the specifications for the model.
Discussion of the Base Case Model

Flue Gas Cooler

The chemical absorption between the CO2 and the amines is favoured by a relatively low
temperature and a pre-cooler at the gas inlet to the absorber is often included in an absorption
plant. Inserting a flue gas cooler does however not a cause significant improvement in
operational cost[17].
The flue gas of a traditional gas power plant will have a water content of approximately
5wt%. At atmospheric pressure this will lead to a saturation temperature of about 43C. The
flue gas leaving a gas power plant will have a temperature of 90-100C. When the flue gas
enters the absorber it will instantly vaporize some of the water in the absorber, and this
vaporization will cool down the flue gas.
If a flue gas cooler is employed the vaporization process occurs before the flue gas enters the
absorber. The flue gas will then be saturated with water and will not experience the cooling
effect of the vaporization in the absorber. Therefore the absorber temperature will not differ
much from the case without pre-cooling. As pre-cooling involves a higher investment cost
and a higher use of cooling water, it was found reasonably to exclude the flue gas cooler.
Impact on Power Plant
The blower duty was unchanged when the circulation rate, the rich/lean minimum temperature
approach and the stripper pressure was varied, and the flue gas composition was kept constant
for all cases. Therefore the only matter that affected the efficiency of the power plant was the
low pressured steam employed in the reboiler, which was assumed to be extracted from the
gas power plant. As the power plant was not included in the simulation model, the impact on
the power plant efficiency can only be discussed on general level.
The steam specifications were kept constant for all cases and the steam mass flow was the
only variable. As less steam was consumed, a higher efficiency was obtained for the power
production.
Total Mechanical Work
The steam employed in the reboiler was converted into an equivalent mechanical energy,
where it was assumed that the condensate was returned to the power plant at 70C. In more
modern power plants, however, it is possible to return the condensate at a lower temperature,
meaning a less conservative factor could be applied to convert the reboiler duty to mechanical
work.

50

Master thesis for Henning Leifsen, Autumn 2007

Water Consumption
The water consumption was not regulated or fixed in the base case model. The only parameter
that had a significant effect on the water consumption was the solvent circulation rate. For this
purpose the temperature out of the lean loading cooler was adjusted so that a more or less
constant mass flow of water followed the cleaned flue gas out of the absorber. Although a
more accurate way of regulating the water consumption will be favourable, this simplification
was found to be sufficient for the purpose of this thesis.
Pressure Loss through Valves
The valves for the streams entering the absorber and the stripper were given rather higher
pressure losses. This will not have a significant effect on the absorber, other than a slightly
increased lean pump duty. For the stripper column, however, a very high pressure loss may
change the phase of the stream entering the column. This will affect the regeneration process
in the stripper column.

9.2 Circulation Rate

Varying the circulation rate also changes the CO2 recovery rate, because the circulation rate
affects the thermal energy requirement in the desorber. To account for this the reboiler duty
was adjusted so that a constant recovery rate of 85% was kept.
There is also a close connection between the circulation rate, the temperature of the lean
amine stream entering the absorber and the water consumption. To keep the water
consumption at a constant level, the temperature of the lean amine stream entering the
absorber was adjusted for the different circulation rates. This resulted in a maximum change
in water consumption of 1.5%.
The lean amine loading varies with the circulation rate and the reboiler duty. The variation in
lean amine loading was therefore also presented as a function of the total mechanical work.
Results
Figure 9-4 and figure 9-5 shows that the optimal circulation rate in terms of total mechanical
work and reboiler duty per kg CO2 with a capture rate of 85% is about 2500 ton/h.

51

Master thesis for Henning Leifsen, Autumn 2007

total work (MJ/kg CO 2)

Total Mechanical Work VS Circulation Rate

1,6
1,58
1,56
1,54
1,52
1,5
1,48
1,46
1,44
1,42
2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

circulation rate (ton/h)

Figure 9-4 Total mechanical work as a function of circulation rate

Reboiler Duty VS Circulation Rate

reboiler duty (MJ/kg CO 2)

4,70
4,60
4,50
4,40
4,30
4,20
4,10
4,00
2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

circulation rate (ton/h)

Figure 9-5 Reboiler duty as function of circulation rate

The lean loading as a function of total mechanical work varies in a similar manner as the
circulation rate, and the minimum for total mechanical work was found for a lean loading of a
about 0.22 as shown in figure 9-6. The rich loading was kept constant at 0.45 since the
number of stages in the absorber column and the recovery rate was unchanged.

52

Master thesis for Henning Leifsen, Autumn 2007

total work (MJ/kg CO 2)

Total Mechanical Work VS Lean Load

1,60
1,58
1,56
1,54
1,52
1,50
1,48
1,46
1,44
1,42
0,18

0,20

0,22

0,24

0,26

0,28

0,30

0,32

lean load

Figure 9-6 Total mechanical work as a function of lean solvent loading

When the circulation rate was increased the necessary cooling of the lean amine stream
decreased with about 1C for every increase in flow rate of 100 ton/h. Figure 9-7 shows that
for a variation in circulation rate of 1800 ton/h, the temperature out of the lean amine cooler
was varied with 17C.
Temperature of Solvent Entering Absorber VS Circulation Rate
40
temperature (C)

35
30
25
20
15
10
5
0
2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

circulation rate (ton/h)

Figure 9-7 Temperature of solvent entering the absorber tower as a function of circulation rate

The consumption of cooling water varied in the same way as the energy consumption, leading
to a minimum consumption for a circulation rate of 2500-2600 ton/h.

53

Master thesis for Henning Leifsen, Autumn 2007

Cooling Water VS Circulation Rate

Cooling water (ton/h)

17000
16500
16000
15500
15000
14500
2200

2400

2600

2800

3000

3200

3400

3600

3800

4000

circulation rate (ton/h)

Figure 9-8 Cooling water as a function of solvent sirculation rate

The composition and flow rate of out of the top of the desorber column varies very little with
the circulation rate, but the flow rate will decrease a little when the circulation rate is
increased.
Discussion of Results
At low circulation rates, the dilution heat is dominant in the thermal energy requirement as a
higher driving force is needed. At high circulation rates more amines has to be heated up
before entering the desorber, and the sensible heat will be dominant in the thermal energy
requirement. Therefore a minimum is expected. Figure 9-4 and figure 9-5 show that the
thermal energy requirement decreases with increasing circulation rate until a minimum is
attained. The point at which the energy requirement is lowest will be defined to be the
optimum lean solvent loading.

The results from figure 9-4 and 9-5 show a lower optimum circulation rate and lean loading
than what was found by Abu-Zahra, Niederer, Feron and Versteeg given in chapter 6.5. The
reason for this is most likely that the results are based on two different feed gas compositions
and a slightly different mass flow.

9.3 Minimum Temperature Approach for Rich/Lean Heat Exchanger

The rich/lean heat exchanger minimum temperature approach was varied between 3C and
15C to observe the effect on the total energy consumption. The composition and flow rate in
and out of the stripper were also observed. The reboiler duty was adjusted so that a constant
recovery rate of 85% was kept.
Results
A variation in the rich/lean minimum temperature approach results in a linear change in total
mechanical work per kg CO2. It is also shown that a change from 15C to 3C changes the
total mechanical work per CO2 with less than 1.5%.

54

Master thesis for Henning Leifsen, Autumn 2007

Total Mechanical Work VS Rich/Lean Min. Temp. Approach

total work (MJ/kg CO 2)

1,60
1,55
1,50
1,45
1,40
1,35
1,30
0

10

12

14

16

minimun temperature approach (C)

Figure 9-9 Total mechanical work as a function of rich/lean minimum temperature approach

The reboiler duty varied with about 2% between 3C and 15C favouring a low temperature
approach.

reboiler duty (MJ/kg CO2)

Reboiler Duty VS Rich/Lean Min. Temp. Approach

4,50
4,25
4,00
3,75
3,50
0

10

12

14

16

minimum temperature approach (C)

Figure 9-10 Reboiler duty as a function of rich/lean minimum temperature approach

With a higher temperature after the rich/lean heat exchanger the vapour fraction in the stripper
inlet stream increases. A change from 15C to 3C almost triples the vapour fraction. A
minimum temperature approach of 3C resulted in a 5.7C higher stripper inlet temperature
than with a minimum temperature approach of 15C. This again resulted in a difference in
stripper top temperature of 5.4C and more water and amines followed the CO2 out of the
stripper. The extra water and amines were however flashed in the overhead receiver and sent
back to the process. The composition and mass flow sent to CO2 compression therefore
remained constant. This resulted however in a higher flow rate in the stripper reflux pump and
the lean amine pump for low temperature approaches. It was also observed that both the
vapour and liquid flow increased throughout the column.

55

Master thesis for Henning Leifsen, Autumn 2007

The total consumption of cooling water decreased with lower minimum approaches. The
difference between 3C and 15C was however less than 2%. The lean amine stream from the
bottom of the stripper, the blower duty and the CO2 compression were not affected by the
temperature approach in the rich/lean heat exchanger.
Discussion of Results
The rich/lean minimum temperature approach was found to have little impact on the energy
consumption for absorption processes. It will therefore not be the most important parameter in
a parameter optimization.

At low temperature approaches the vapour fraction in the rich stream entering the stripper
increases. A flash separator after the rich/lean heat exchanger can reduce the loading
significantly. Although a flash separator was not used in the simulation model of this masters
thesis, the distillation columns of HYSYS have a flash included and the separation will occur
before it interacts with flows of the column.
When the feed, at its equilibrium, meets the water wash reflux, the solvent is cooled and
diluted. As explained in chapter 6.2.2 this could result in a drop in the equilibrium partial
pressure below the partial pressure of the vapour phase in the tower. The cross-flow heat
exchanger should therefore only have a limited duty to keep desorption occurring throughout
the column. The duty should be chosen so that the inlet rich feed to the desorber column, after
being mixed with the reflux, is close to the desorber exit gas phase equilibrium. In other
words, the more efficient the heat exchanger is, the higher the chances are for the desorber to
be performing below optimum. The size of the heat exchanger should therefore be chosen
depending on the desorber size and efficiency.

9.4 Stripper Pressure

The stripper column pressure was varied between 1.462 bara and 3.462 bara at the bottom of
the column with a constant pressure drop of 138 mbar. For every increase in column pressure
an equal increase in outlet pressures were set for the reboiler reflux pump, the rich amine
pump, the condenser reflux pump and the valve in front of the stripper. The reboiler duty was
adjusted so that a constant recovery rate of 85% was kept.
The specifications for the steam used to heat up the reboiler were constant for all pressures
and the reboiler duty was varied by the mass flow of the steam only.
Results
Figure 9-11 and 9-12 show that the energy consumption per kg CO2 can be reduced by
increasing the stripper column pressure. The graphs are steep for an increase in column
bottom pressure up to about 2 bara and stabilizes more to a linear decrease if the pressure is
further increased. Figure 9-13 shows that the consumption of cooling water will vary with the
stripper pressure in the same manner as the energy consumption.

56

Master thesis for Henning Leifsen, Autumn 2007

Total Mechanical Work VS Stripper Pressure

total work (MJ/kg CO2)

1,7
1,65
1,6
1,55
1,5
1,45
1,4
1,35
1,3
1,4

1,6

1,8

2,2

2,4

2,6

2,8

3,2

3,4

stripper btm pressure (bara)

Figure 9-11 Total mechanical work as a function of stripper column pressure

The lean loading decreases and the recovery rate increases with an increased column pressure.
Since a constant recovery rate was desired and the reboiler duty was adjusted for the different
pressures, the lean loading was constant.

reboiler duty (MJ/kg CO2)

Reboiler Duty VS Stripper Pressure

5
4,8
4,6
4,4
4,2
4
3,8
1,4

1,6

1,8

2,2

2,4

2,6

2,8

stripper btm pressure (bara)

Figure 9-12 Reboiler duty as a function of stripper column pressure

57

3,2

3,4

Master thesis for Henning Leifsen, Autumn 2007

Cooling Water VS Stripper Pressure

cooling water (m3/h)

17000
16500
16000
15500
15000
14500
14000
13500
13000
1,4

1,6

1,8

2,2

2,4

2,6

2,8

3,2

3,4

stripper bottom pressure (bara)

Figure 9-13 Cooling water consumption as a function of stripper column pressure

At a bottom pressure of 3.46 bara (3.32 bara in the top) two liquid phases were observed in
the top stage of the stripper column and HYSYS had trouble converging. The solution
temperature after the reboiler will also be relatively high if the pressure is further increased.
With a stripper bottom pressure of 3.46 bara the solution temperature after the reboiler is
140.9C. The variation in temperature after the reboiler is shown below.

Temperature after Reboiler VS Stripper Pressure

temperature (C)

140
135
130
125
120
115
1,4

1,6

1,8

2,2

2,4

2,6

2,8

3,2

3,4

stripper btm pressure (bara)

Figure 9-14 Temperature of reboiler reflux as a function of stripper column pressure

The amount of water following the CO2 out of the stripper tower decreases with a higher
stripper pressure. For bottom pressures up to about 2 bara the amount of water vapour out of
the stripper tower decreases rapidly but decreases with a lesser degree for higher pressures.
The amount of amines following the CO2 and water out of the stripper decreases up to a
pressure of 2.162 bara and start to increase again for higher pressures.

58

Master thesis for Henning Leifsen, Autumn 2007

Discussion of Results
When the stripper pressure is increased more CO2 is stripped. This can be explained based on
the effect of temperature on vapour pressures. The vapour pressure of CO2 over an amine
solution generally increases with temperature more rapidly than does the vapour pressure over
the water/amine mixture over the same solution. As a result, stripping to the same mole
fraction CO2 in the vapour phase means stripping to a lower mole fraction in the liquid if the
pressure (and thereby the temperature) is increased[3]. This also explains why less water
vapour followed the CO2 out of the stripper when the pressure was increased. It should
however be noted that this phenomenon is a characteristic for CO2 in MEA and will not
necessarily occur with other amines. In addition to a decrease in reboiler duty, the CO2
compression work will decrease with higher stripper pressure leading to an even lower total
work per kg CO2.

The energy consumption decreases very rapidly for an increase in desorber bottom pressure
up to about 2 bara before it decreases linearly and more gently for higher pressures. A good
explanation for this was not found and should be examined more closely.
A lower energy consumption due to a higher stripper pressure is in accordance with the results
of Tobiesen, Svendsen and Hoff shown in chapter 6.7.
With a column bottom pressure above 3.462 bara two liquid phases were found in the top
stage of the stripper column, and the simulation model had problem converging. This may be
connected to the limitations on the amine temperature in the amine package in HYSYS. For a
column bottom pressure of 3.462 bara, a temperature of almost 141C was observed on the
outlet of the reboiler. The range of applicability for the amines in the Amines Property
Package in HYSYS is given as 25C-126.7C. It is also noted in the simulation basis manual
for HYSYS that the reboiler temperature should not exceed 137.8C to avoid physical
degradation of the amines into corrosive by-products. HYSYS will therefore not be a suitable
programme for simulating high stripper pressures. The optimum operational pressure will be a
trade-off between the gain in decreased operational cost and the increased operational
problems due to degradation and corrosion.

9.5 Economic Evaluation

Investment Expenses (CAPEX)
Only the equipment that differed from the base case was included in the investment
calculations, and all results are given as a cost difference from base case. The investment cost
data were supported by Aker Kvrner. The prices were scaled based on the AKET Just Catch
CO2 capture plant. Due to a request by Aker Kvrner for keeping the economic background
data undisclosed, these numbers were not included in the thesis.

The different equipment was scaled differently. The heat exchangers and columns were scaled
by area, the separators by weight, the compressors by duty and the pumps by volume flow.
The areas, duties and volume flows were found by the student. The separator weight was
calculated based on flow rate and operational pressure, and was calculated in co-operation
with Aker Kvrner.

59

Master thesis for Henning Leifsen, Autumn 2007

Heat exchanger areas were calculated using equation 8.2, while the absorber and stripper
areas were calculated using equation 8.3. A superficial gas velocity of 3 m/s through the
columns were used for all cases. This was the number recommend by Sulzer for the
Mellapack 2X packing [18].
Operational Expenses (OPEX)
The operational expenses were calculated as the sum of mechanical energy cost, cooling
water cost and maintenance. The mechanical energy cost was calculated with an electricity
cost of 0.36 NOK per kWh, while the cost of cooling water was calculated with a price of 0.1
NOK per m3 seawater. The maintenance was calculated as 2.5%/year of investment costs.
These numbers were based on a report from NVE [19] written in December 2006, which
included an economic evaluation of a future CO2 capture plant at Krst, Norway.

To compare the total cost for the different cases from start to end, the net present value (NPV)
was calculated, using equation 8.5. It was assumed that the project life was 25 years, and that
the plant was running for 8000 hours a year[19]. In the report from NVE an interest rate of
5% was used, while interest rates of 3% and 7% were included in the sensitivity analysis. The
reason for applying three different interest rates was that all numbers were calculated on an
early stage and the uncertainties regarding their accuracy were high. The same interest rates
were applied in this thesis, and it was assumed a real interest rate.
Results
Circulation Rate
Table 9-1 and 9-2 compares different circulation rates with the base case of 2358 ton/h in
terms of investment cost, operational cost and net present value. It is found that a circulation
rate of 2500 ton/h will be the best economical option. It has the lowest investment cost and
operational cost and the highest net present value.
Table 9-1 Difference in investment cost and operational cost for different circulation rates compared to
base case of 2358 ton/h

[ton/h]
Cases

[mill. NOK]
Investment Cost

[thousand NOK/year]
Operational Cost

2200
2300
base case (2358)
2500
2600
2700
2800
2900
3000
3250
3500
4000

3,26
0,46
0
-0,40
0,01
0,92
2,80
3,55
5,50
8,70
11,72
19,48

6037
986
0
-809
-515
497
2125
2274
4693
7474
9969
16748

60

Master thesis for Henning Leifsen, Autumn 2007

Table 9-2 Difference in net present value (NPV) for different circulation rates compared to base case of
2358 ton/h

[ton/h]
Cases
2200
2300
base case (2358)
2500
2600
2700
2800
2900
3000
3250
3500
4000

NPV (i=3%)

[mill. NOK]
NPV (i=5%)

NPV (i=7%)

-108,4
-17,6
0
14,5
8,9
-9,6
-39,8
-43,2
-87,2
-138,8
-185,3
-311,1

-88,4
-14,4
0
11,8
7,2
-7,9
-32,7
-35,6
-71,6
-114,0
-152,2
-255,5

-73,6
-12,0
0
9,8
6,0
-6,7
-27,6
-30,1
-60,2
-95,8
-127,9
-214,7

Rich/Lean Minimum Temperature Approach

Table 9-3 and 9-4 compare different minimum temperature approaches in the rich/lean heat
exchanger with the base case temperature of 8.5C in terms of investment cost, operational
cost and net present value. It is found that a minimum temperature approach of 5C will be
the best economical option based on the highest net present value.
Table 9-3 Difference in investment cost and operational cost for different minimum temperature
approaches compared to base case at 8.5C

(C)
Cases

[mill. NOK]
Investment Cost

[thousand NOK/year]
Operational Cost

3
5
7
base case (8,5)
11
13
15

7,70
3,94
1,50
0
-2,07
-3,72
-5,12

-503
-414
-160
0
373
696
1293

Table 9-4 Difference in net present value (NPV) for different minimum temperature approaches
compared to base case at 8.5C

(C)
Cases
3
5
7
base case (8,5)
11
13
15

NPV (i=3%)

[mill. NOK]
NPV (i=5%)

NPV (i=7%)

1,1
3,3
1,3
0,0
-4,4
-8,4
-17,4

-0,6
1,9
0,8
0,0
-3,2
-6,1
-13,1

-1,8
0,9
0,4
0,0
-2,3
-4,4
-10,0

61

Master thesis for Henning Leifsen, Autumn 2007

Stripper Column Pressure
Table 9-5 and 9-6 compare different stripper column pressures with the base case of 1.862
bara in terms of investment cost, operational cost and net present value. It is found that a
bottom pressure of 3.262 bara will give the highest net present value and will therefore be
considered as the best economical option.
Table 9-5 Difference in investment cost and operational cost for different stripper column bottom
pressures compared to base case of 1.862 bara

[bara]
Cases

[mill. NOK]
Investment Cost

[mill. NOK/year]
Operational Cost

1,462
1,562
1,662
1,762
base case (1,862)
1,962
2,062
2,162
2,262
2,362
2,462
2,562
2,662
2,762
2,862
2,962
3,062
3,162
3,262
3,362
3,462

45,45
30,85
18,64
8,48
0
-7,15
-12,32
-17,33
-22,09
-25,94
-28,89
-32,36
-34,55
-37,14
-39,24
-39,62
-40,23
-39,41
-40,22
-33,49
-26,52

23,4
16,0
9,2
4,0
0
-2,7
-4,2
-5,2
-6,2
-7,1
-7,6
-8,3
-8,9
-9,6
-10,3
-10,7
-11,3
-11,8
-12,2
-12,5
-12,8

62

Master thesis for Henning Leifsen, Autumn 2007

Table 9-6 Difference in net present value (NPV) for different stripper column bottom pressures compared
to base case of 1.862 bara, given for interest rates of 3%, 5% and 7%.

[bara]
Cases
1,462
1,562
1,662
1,762
base case (1,862)
1,962
2,062
2,162
2,262
2,362
2,462
2,562
2,662
2,762
2,862
2,962
3,062
3,162
3,262
3,362
3,462

NPV (i=3%)

[mill. NOK]
NPV (i=5%)

NPV (i=7%)

-452,9
-310,0
-179,0
-78,4
0
53,7
84,9
108,0
130,5
149,0
161,2
177,1
190,1
204,2
218,4
225,9
237,3
244,9
253,5
250,5
249,1

-375,2
-256,8
-148,4
-65,1
0
44,8
71,1
90,7
109,8
125,5
136,0
149,5
160,4
172,4
184,2
190,4
199,7
205,8
212,8
209,1
206,7

-318,1
-217,7
-125,9
-55,3
0
38,3
60,9
78,0
94,6
108,3
117,4
129,2
138,6
149,0
159,1
164,3
172,1
177,0
182,9
178,7
175,5

Combining the Optimum Results

An optimal point was found for a circulation rate of 2500 ton/h, for a rich/lean minimum
temperature approach of 5C and for a stripper column pressure of 3.262 bara in the bottom.
These results were combined in a new model. Table 9-7 shows how much the investment cost
is improved compared to base case, including all equipment that changes with the three
parameters above.
Table 9-7 Difference in investment cost for a sub-optimal operating point compared to base case
[mill.NOK]
Unit
optimal
-45,18
CO2-compressor
-0,52
rich lean exchanger
-20,35
stripper
-1,84
stripper condenser
0,50
seperator CO2-compression 1
0,20
seperator CO2-compression 2
0,80
stripper overhead receiver
-0,09
stripper reflux pump
27,82
stripper reboiler
0,66
lean amine pump
-2,06
lean amine cooler
-40,06
total cost

63

Master thesis for Henning Leifsen, Autumn 2007

The investment cost for the three parameters varied, where all equipment that differed is
specified, is shown in appendix C.
Discussion of Results
In many design situations, it is possible to have small increments in capacity; hence the
possibility of small increments in capital investment. By providing more surface area for heat
transfer in a co-current heat exchanger, greater heat transfer may be obtained. But each m2 of
heat transfer area provides a diminishing heat transfer flux. On the other hand, the cost of the
heat exchanger goes up as the area is increased.

The areas of the condenser, the lean amine cooler, the reboiler and the rich/lean cross-flow
heat exchanger vary as a consequence of the difference in minimum temperature approach in
the rich/lean heat exchanger. The reboiler and the lean amine cooler areas both decreased with
a lower minimum approach, while the condenser and rich/lean heat exchanger increased. The
major difference was however for the rich/lean heat exchanger which became very large for a
minimum approach of 3C.
An optimal stripper pressure in terms of net present value (NPV) was found for a bottom
pressure of 3.262 bara. The reason that the net present value was higher at this pressure than
for higher pressures, even though higher pressures resulted in a lower energy consumption, is
connected to the increased investment cost of the reboiler. The specifications for the steam
employed in the simulations of this thesis were kept constant at 4 bara and 270C. For an
increased pressure (and temperature) of the amine solvent, the temperature difference between
the steam and the amine solvent will become smaller and the necessary surface area will
increase. Therefore, if the stripper column pressure is increased, the pressure of the steam
employed should also be increased. This means that the optimal point in terms of net present
value can be higher than 3.262 bara. However, it is very important to be aware of the
increased operational problems due to an increased pressure (and temperature), and most
likely a significantly lower pressure will be more realistic.
Table 9-7 shows that the major reduction in investment costs are found for the stripper
column and the CO2-compressor. The increased investment cost due to a large reboiler at
higher pressures is the main reason that the reduction in investment costs is not greater.

9.6 Absorber Column Diameter

The absorber column diameter was found to be a difficult parameter to vary. Varying the
column diameter changes the gas velocity, and thereby the pressure drop, through the column.
The pressure drop again affects the necessary column height and the energy consumption of
the absorption plant. The value of the pressure drop is also highly dependent of the type of
packing in the column.
To get an understanding of how the pressure drop through the absorber affects the energy
consumption a simple test in HYSYS was performed, where the pressure drop was varied
with constant absorber specifications. The result is shown in figure 9-15.
The connection between the superficial gas velocity and the absorber column diameter can be
described by rearranging equation 8.3:

64

Master thesis for Henning Leifsen, Autumn 2007

wG =

4 V&
D 2

(9.1)

With a constant volume flow the superficial gas velocity will be a function of the diameter
alone. Figure 9-16 gives the result of varying the column diameter in terms of superficial gas
velocity using the maximum volume flow from the base case simulation model.
For column design HYSYS has its limitations. The absorber product depends on the number
of stages, inlet flows, pressure drop, diameter and type of internal material (packing, sieve,
valve etc.). All these parameters have to be given to HYSYS as inputs. Based on these
HYSYS calculates the outlet streams. HYSYS does not calculate the absorber height and it is
not possible to specify the height or type of packing.
It turned out that a variation of the absorber column diameter using HYSYS alone will not
give any interesting results. Therefore a second programme, SULCOL, was used in order to
determine the variation in column pressure as a result of varying the diameter.
SULCOL is a sizing program designed for distillation, stripping and absorption columns with
Sulzer mass transfer products. The sizing procedure leads to estimated values for the product
specific design figures like capacity, pressure drop, holdup etc., based on the specified liquid
and gas flow rates and physical properties at a given column diameter. The column height can
also be specified. However, this programme does not account for the chemical reactions
inside the column.
To be able to determine the necessary height for different diameters the use of an additional
program or calculations by hand was necessary. Other programmes like PROTREAT has the
ability to calculate column height, but due to the time limit it was found better to do some
simplifications and calculate the column height by hand rather than learning to use a third
programme.
The column height was calculated using equation 9.2, which is modification of equation 4-12.
The mole fraction of CO2 in equilibrium, ye, was assumed to be a lot lower than the CO2 mole
fraction at the top and bottom of the column, and was as a simplification neglected. Since the
type of packing, temperature and pressure were kept constant, the overall mass transfer
coefficient, KGa, was also assumed constant. With a constant overall mass transfer coefficient,
a constant column pressure and constant inlet streams, the necessary column height to keep
the CO2 capture rate constant for different diameters could be calculated.
h=

y2
GM
ln
K G aP y1

(9.2)

The overall mass transfer coefficient was found by calculating backwards, using the
specifications of a model with similar specifications[15] where the absorber height was
known.
The diameter was varied between 13 and 30 meters, and for each diameter a new height was
calculated. These specifications were given as inputs to SULCOL along with the inlet and
outlet specifications of the streams entering and leaving the absorber column and the number
of stages in the absorber. The stream specifications and number of stages were found from the

65

Master thesis for Henning Leifsen, Autumn 2007

base case simulation model in HYSYS. The packing type chosen for the absorber column
design was Mellapack 2X from Sulzer. This is the same type of packing recommended to
Aker Kvrner by Sulzer. Based on this information the pressure drop through the column was
calculated using SULCOL. The diameter with the corresponding height as a function of
pressure loss is shown in figure 9-17.
The pressure drop calculated in SULCOL for each diameter was returned to the simulation
model in HYSYS along with the diameter. The pressure drop of 0.05 bar was used for the
base case with a column diameter of 17 meters. The other pressure drops were adjusted to
make the difference in pressure drop as calculated in SULCOL. As the recovery rate increases
with a lower pressure drop, the reboiler duty was varied to keep the recovery rate constant.
The energy consumption as a function of the column diameter is shown in figure 9-18. Since
the overall mass coefficient was applied, the volume of the absorber did not vary and an
economic evaluation was not included.
Results
As shown below the energy consumption varies linearly with the pressure drop through the
absorber column. This mainly due to the change in blower duty.
Total Mechanical Work VS Pressure Drop Absorber

total work (MJ/kg CO2)

1,9
1,8
1,7
1,6
1,5
1,4
1,3
1,2
0

20

40

60

80

100

120

140

160

180

pressure drop absorber (mbara)

Figure 9-15 Total mechanical work as a function of pressure drop in absorber column

Figure 9-16 shows that for a maximum superficial gas velocity of 3 m/s the minimum column
diameter will be about 17 meters.

66

Master thesis for Henning Leifsen, Autumn 2007

absorber diameter (m)

Gas Velocity VS Absorber Diameter

29
27
25
23
21
19
17
15
13
1,00

1,50

2,00

2,50

3,00

3,50

4,00

4,50

5,00

5,50

superficial gas velocity (m/s)

Figure 9-16 Superficial gas velocity as a function of absorber column diameter

The overall mass transfer coefficient, KGa, was calculated to be 0.0155 kmol/(m bar s) for a
superficial gas velocity of 3m/s, a height of 12m and the specifications in the base case
simulation model.
The diameter and the corresponding height as a function of the absorber pressure drop is
shown in figure 9-17. It is shown that increasing the diameter to more than 19 meters will
affect the pressure loss through the column very little. If the diameter is decreased to below 15
meters however, the pressure loss will differ significantly.
Height and Diameter VS Pressure Drop Absorber

diameter and height (m)

35

Diameter

30

Height

25
20
15
10
5
0
0

20

40

60

80

100

pressure drop in absorber (mbara)

Figure 9-17 Height and diameter as a function of pressure drop in absorber

As the pressure drop varied very little for column diameters above 19 meters, the difference in
energy consumption varied very little as well. A diameter below 17 meters will increase the
energy consumption significantly, while increasing the diameter from 17 meters to 19 meters
can give favourable results.

67

Master thesis for Henning Leifsen, Autumn 2007

Total Mechanical Work VS Absorber Column Diameter

total work (MJ/kg CO2)

1,75
1,7
1,65
1,6
1,55
1,5
1,45
1,4
1,35
13

15

17

19

21

23

25

27

29

absorber diameter (m)

Figure 9-18 Total mechanical work as a function of absorber column diameter

Discussion of Results
Figure 9-15 shows that the pressure drop in the absorber will affect the energy consumption
significantly if the pressure loss becomes high. This is mainly due to the increased duty of the
flue gas blower. The pressure loss will however not differ as much as in figure 9-15 for
different diameters, and was only included as an illustration.

The maximum acceptable superficial gas velocity is a quality of the packing in terms of at
which velocity flooding can occur. For the Mellapack 2X a maximum velocity of 3m/s was
recommended. This leads to a minimum diameter of about 17 meters with the gas volumetric
flow rate in the model of this thesis. The diameters for Mellapack are given for 0.08m up to
17m, depending on type[18]. If it is desirable to increase the diameter more, a different type
of packing might be more favourable. An increase in diameter to more than 19m did however
not result in much difference in total mechanical work for the absorption plant.
The value of yeq was neglected in equation 9.2. This means that no back pressure was
expected or that the reaction was irreversible. In reality the column will be significantly taller
than the calculated height.
It was found that HYSYS is not a suitable programme for column design. On the basis of the
simplifications and uncertainties regarding the calculations of diameter and height, the degree
of accuracy of the results found for the absorber design has to be taken into consideration. The
column diameter should therefore be examined to a greater extent if an optimal diameter is to
be found.

68

Master thesis for Henning Leifsen, Autumn 2007

10

Conclusion

A simulation model for a CO2 capturing plant using MEA amines has been made, and a base
case with similar parameters as a commercial regeneration plant was determined. Three
parameters were varied in terms of improving the overall energy efficiency of the plant and
thereby reduce operating costs. The parameters varied were the solvent circulation rate, the
rich/lean heat exchanger minimum temperature approach and the desorber pressure. An
economical evaluation has been performed, where investment costs were included in the
calculations. In the end, the optimal values for the three parameters were included in one new
model.
The thermal energy requirement in the desorber is highly dependent on the circulation rate. At
low circulation rates a higher driving force in terms of steam is needed, and for high
circulation rates the thermal energy required for heating of amines, the sensible heat, will be
dominant. An optimum can be found, where the combination of sensible, absorption and
dilution heat is at its lowest. This circulation rate was found to be 2500 ton/h and it was also
this option that gave the highest net present value.
The minimum temperature approach for the rich/lean cross-flow heat exchanger has little
impact on the overall energy consumption of the absorption plant. The ideal minimum
temperature approach is found to be low but not too low. With a lower minimum temperature
approach, the vapour fraction in the rich amine stream will be higher and the saturation
pressure of the CO2 in the rich amine stream will decrease. If the saturation pressure becomes
too low, it will have a bad affect on the regeneration process in the desorber. An optimal
temperature in terms of net present value was found for a temperature of 5C, mainly due to
the increased investment cost with a lower minimum temperature approach.
Increasing the desorber pressure can give favourable results. A higher pressure may reduce
the energy consumption significantly as well as reduce the investment costs, especially for the
stripper column and the CO2-compressor. However, since the temperature will increase with
the pressure, the temperature of the column will be a limiting factor. If the temperature
becomes too high this may lead to operational problems like amine degradation and corrosion.
By including the optimal values for the three parameter variations, a sub-optimal model was
made. It was found that optimizing the reboiler pressure and the circulation rate will result in
a significant reduction in both investment cost and operating cost.
A fourth parameter, the absorber column diameter, was also varied in order to see the
connection between the diameter and height of the column, the pressure loss through the
column and the energy consumption of the absorption plant. It was found that for a packed
column, the pressure loss will be low, and the gain in increased efficiency for the absorption
plant due to an optimization of the absorber diameter is limited. The absorber diameter should
therefore be determined based on flooding correlations rather than pressure drop.

69

Master thesis for Henning Leifsen, Autumn 2007

70

Master thesis for Henning Leifsen, Autumn 2007

11 Recommendations for Further Work

The temperature of the amine solution is the limiting factor for an increase in desorber column
pressure as a high temperature will lead to degradation of amines and corrosion problems.
Since there are many uncertainties regarding the amine behaviour at higher temperatures and
pressures this should be investigated in more detail. Problems to be investigated may be the
cause for the amine degradation, the optimum trade-off between additional operational
problems and the reduced energy consumption, and the possibility of minimizing the
degradation and corrosion problems by changing the basis for the absorption process (amine
type, column material etc.).
A constant pressure for the reboiler steam was applied for all cases and made the reboiler area
very large for high pressures. A similar simulation should be performed where the pressure of
the steam is increased along with the stripper pressure.
The energy consumption decreases very rapidly for an increase in desorber bottom pressure
up to about 2 bara before it decreases linearly and more gently for higher pressures. The
reason for the shape of this graph should be investigated further.

71

Master thesis for Henning Leifsen, Autumn 2007

72

Master thesis for Henning Leifsen, Autumn 2007

12 References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.

13.

14.
15.
16.
17.
18.
19.

Chakravarti, S., Gupta, A. & Hunek, B. . Advanced Technology for the Capture of
Carbon Dioxide from Flue Gases. 2001 May [cited 2007 Jan]; Available from:
http://www.netl.doe.gov.
Sire Seljom, P.M., Termiske kraftprosesser med utskilling av CO2 ved hyt trykk.
2006, NTNU: Trondheim.
Kohl, A.L., Nielsen, R. B., Gas Purification. 5th ed. 1997: Gulf Publishing Company.
Treybal, R.E., Mass-transfer Operations 3rd ed. McGraw-Hill Chemical Engineering
Series. 1980, New York: McGraw-Hill Book Company.
Bolland, O., Professor at the Department of Energy and Process Engineering. 2007.
Booth, N.J., Secondment to the International Test Centre for CO2 Capture (ITC).
2005, University of Regina, Canada: Regina.
Aroonwilas, A., Tontiwachwuthikul, P. and Chakma, A. Effects of Operating and
Design Parameters on CO2 Absorption in Columns with Structured Packings. 2001
[cited 2007 November]; Available from: http://www.sciencedirect.com.
TCEQ. Absorption Units - Technical Disclaimer. 1995 [cited 2007 October];
Available from: http://www.tceq.state.tx.us/.
Hyprotech, Hysys - Simulation Basis. 2003, Aspen Technology.
Tobiesen, F.A., Svendsen, H. F. & Hoff, K. A., Desorber Energy Consumption Amine
Based Absorption Plants. 2005.
Rao, A.B., Rubin, E.S. Identifying Cost-Effective CO2 Control Levels for Amine-Based
CO2 Capture Systems. 2005 [cited 2007 September]; Available from:
http://pubs.acs.org/cgi-bin/article.cgi/iecred/2006/45/i08/html/ie050603p.html.
Abu-Zahra, M.R.M., Niederer, J.P.M., Feron, P.H.M., Versteeg, G.F. CO2 capture
from power plants. Part 1. A parametric study of the technical performance based on
monoethanolamine. 2006 [cited 2007 September]; Available from:
http://www.sciencedirect.com.
Abu-Zahra, M.R.M., Niederer, J.P.M., Feron, P.H.M., Versteeg, G.F. CO2 capture
from power plants Part II. A parametric study of the economical performance based
on mono-ethanolamine. 2006 [cited 2007 September]; Available from:
http://www.sciencedirect.com.
Bolland, O., Energi i Norge - Ressurser, teknologi og milj. 2000.
Berg, B.M., Aker Kvrner. 2007.
K., M.A., ENCH 423 - Chemical Engineering Process Development. 2007: Calgary.
Course notes.
Tobiesen, F.A., Researcher. 2007: Trondheim.
SULZER, Structured Packings for Distillation, Absorption and Reactive Distillation.
2002: Switzerland.
NVE, CO2-hndtering p Krst - fangst, transport, lagring. 2006.

73

Master thesis for Henning Leifsen, Autumn 2007

74

Master thesis for Henning Leifsen, Autumn 2007

Appendix A: Base Case Data

Flue Gas Properties
Temperature:
Pressure:
Mass flow:
Composition:
(mass fraction)

100C
1.013 bara
2.471 106 kg/h
CO2:
0.060941
Nitrogen:
0.743124
Oxygen:
0.143219
H2O:
0.052716

Loading and Recovery

Rich load:
Lean load:
Recovery rate:

0.4559 mol/mol
0.2045 mol/mol
84.94%

Energy Use
Total Mechanical Work:
Reboiler Duty:
Duty CO2 compression:

1.441 MJ/kg CO2

4.086 MJ/kg CO2
0.307 MJ/kg CO2

Compressors
Blower
Inlet pressure:
Outlet pressure:
Duty:
Adiabatic efficiency:

1.013 bara
1.073 bara
5.74 MW
75%

CO2 Compressor
Inlet pressure:
Outlet Pressure:
Stages:
Duty:
Adiabatic efficiency:

1.672 bara
80.00 bara
3
10.89 MW
85%, 85% and 80%

Master thesis for Henning Leifsen, Autumn 2007

Streams
Rich Amine Stream Leaving Absorber
Temperature:
Pressure:
Mass flow:
Composition:
(mass fraction)

52.43C
1.063 bara
2.557 106 kg/h
CO2:
0.090635
Nitrogen:
0.000008
Oxygen:
0.000003
H2O:
0.633410
MEA-amine: 0.275944

Lean Amine Stream Entering Absorber

Temperature:
Pressure
Mass flow:
Composition:
(mass fraction)

20.91C
1.070 bara
2.462 106 kg/h
CO2:
0.042184
H2O:
0.670471
MEA-amine: 0.287345

Condenser Reflux
Temperature:
Pressure:
Mass flow:
Composition:
(mass fraction)

28.15C
6.5 bara
7.773 104 kg/h
CO2:
0.002184
H2O:
0.995596
MEA-amine: 0.002220

Reboiler Reflux
Temperature:
Pressure:
Mass flow:
Composition:
(mass fraction)

122.20C
1.94 bara
1.406 106 kg/h
CO2:
0.042953
H2O:
0.665379
MEA-amine: 0.291668

II

Master thesis for Henning Leifsen, Autumn 2007

Absorber and Stripper Tower

Stripper Tower
Internal type:
Number of stages:
Pressure bottom:
Pressure top:
Rich amine inlet stage:

Packed
32
1.862 bara
1.724 bara
2 (from the top)

Absorber Tower
Internal type:
Number of stages:
Pressure bottom:
Pressure top:
Lean amine inlet stage:

Packed
11
1.063 bara
1.013 bara
1 (from the top)

Heat Exchangers
Condenser
Heat exchanger model:
Pressure loss cooling water:
Pressure loss CO2:
Reflux pump outlet pressure:

End Point
5 bara
3% of inlet pressure
6.5 bara

Reboiler
Heat exchanger model:
Pressure loss Steam:
Pressure loss amine stream
Reboiler duty specification
Steam pressure:
Steam temperature:

Weighted
0. 8 bara
0.6 bara
594000000 kJ/h
4 bara
270C

Rich/Lean Heat Exchanger

Heat exchanger model:
Pressure loss lean stream:
Pressure loss rich stream:
Specified pinch:

Weighted
4 bara
4 bara
8.5C

Lean Amine Cooler

Heat exchanger model:
Pressure loss cooling water:
Pressure loss lean solvent:

End point
4 bara
5 bara

III

Master thesis for Henning Leifsen, Autumn 2007

IV

Master thesis for Henning Leifsen, Autumn 2007

Appendix B: Non-Equilibrium Stage Model in HYSYS

Overall Material Balance
Fj + L 1 ( L j + SL j ) (V j + SV j ) = 0

(b.1)

Component Material Balance

Fj zij + L j 1 xij 1 + V j +1Yij +1 ( L j + SL j ) xij (V j + SV j ) yij = 0

(b.2)

Energy Balance
Fj H Fj + Q j + L j 1h j 1 + V j +1 H j +1 ( L j + SL j )h j (V j + SV j ) H i = 0

(b.3)

Equilibrium Relationship

ij K ij xij (V j + SV j ) (V j + SV j ) yij + (1 ij )V j +1 yij +1 = 0

(b.4)

Master thesis for Henning Leifsen, Autumn 2007

VI

Master thesis for Henning Leifsen, Autumn 2007

Appendix C: Difference in Investment Cost

Table C-1: Difference in investment cost for different circulation rates compared to base case of 2358 ton/h, including all equipment that differs.
Unit
Rich amine pump
rich lean exchanger
stripper
stripper condenser
stripper reflux pump
stripper reboiler
lean amine pump
lean amine cooler
total cost

2200
-0,71
-0,89
1,19
2,06
0,35
0,94
-0,79
1,12
3,26

2300
-0,26
-0,20
0,22
0,55
0,08
0,24
-0,29
0,11
0,46

2500
0,61
0,71
-0,24
-0,48
-0,11
-0,14
0,68
-1,43
-0,40

2600
1,05
1,20
-0,34
-0,64
-0,14
-0,22
1,17
-2,06
0,01

2700
1,48
1,27
-0,26
-0,71
-0,15
-0,17
1,65
-2,19
0,92

[mill.NOK]
2800
1,91
2,07
-0,07
-0,59
-0,13
-0,06
2,14
-2,47
2,80

2900
2,34
2,45
-0,14
-0,76
-0,16
-0,11
2,61
-2,68
3,55

3000
2,75
2,97
0,18
-0,56
-0,12
0,10
3,07
-2,90
5,50

3250
3,79
3,93
0,43
-0,57
-0,13
0,30
4,24
-3,29
8,70

3500
4,82
4,87
0,63
-0,70
-0,15
0,44
5,37
-3,56
11,72

4000
6,82
6,94
1,34
-0,65
-0,14
0,99
7,61
-3,43
19,48

Table C-2: Difference in investment cost for different minimum temperature approaches compared to base case of 8.5C including all equipment that differed.

Unit
rich lean exchanger
stripper condenser
stripper overhead receiver
stripper reflux pump
stripper reboiler
lean amine cooler
total cost

3C
4,76
2,84
0,20
0,52
-0,09
-0,53
7,70

VII

5C
2,04
1,77
0,20
0,33
-0,07
-0,32
3,94

[mill. NOK]
7C
11C
0,68
-0,79
0,75
-1,31
0,10
0,00
0,14
-0,24
-0,03
0,06
-0,14
0,21
1,50
-2,07

13C
-1,25
-2,43
-0,10
-0,44
0,11
0,38
-3,72

15C
-1,63
-3,50
-0,10
-0,63
0,18
0,55
-5,12

Master thesis for Henning Leifsen, Autumn 2007

Table C-3: Difference in investment cost for different reboiler bottom pressures compared to base case of 1,862 bara including all equipment that differed.

Unit
CO2-compressor
rich lean exchanger
stripper
stripper condenser
seperator (CO2-compression) 1
seperator (CO2-compression) 2
stripper overhead receiver
stripper reflux pump
stripper reboiler
lean amine cooler
total cost

1,462
21,52
0,24
18
6,29
-0,4
-0,1
0,1
1,02
-1,16
-0,06
45,45

1,562
14,6
0,16
12,47
4,35
-0,2
-0,2
0
0,71
-1
-0,04
30,85

Unit
CO2-compressor
rich lean exchanger
stripper
stripper condenser
seperator (CO2-compression) 1
seperator (CO2-compression) 2
stripper overhead receiver
stripper reflux pump
stripper reboiler
lean amine cooler
total cost

2,562
-25,85
-0,3
-12,36
-2,42
0,2
0,1
0,4
-0,3
8,07
0,1
-32,36

2,662
-28,43
-0,33
-13,61
-2,63
0,3
0,1
0,4
-0,33
9,86
0,12
-34,55

VIII

1,662
9,36
0,1
7,44
2,5
-0,2
-0,1
0
0,4
-0,82
-0,04
18,64

2,762
-31,89
-0,37
-14,76
-2,81
0,4
0,2
0,4
-0,35
11,91
0,13
-37,14

1,762
4,46
0,05
3,41
1,06
-0,1
-0,1
0
0,17
-0,46
-0,01
8,48

2,862
-35,17
-0,41
-15,84
-2,98
0,4
0,2
0,5
-0,36
14,28
0,14
-39,24

[mill. NOK]
1,962
2,062
-4,38
-7,89
-0,05
-0,09
-2,66
-4,79
-0,76
-1,21
0
0,1
0
0
0,1
0,1
-0,11
-0,17
0,69
1,6
0,02
0,03
-7,15
-12,32

[mill. NOK]
2,962
3,062
-37,23
-40,1
-0,43
-0,46
-16,91
-17,9
-3,12
-3,22
0,4
0,4
0,2
0,2
0,6
0,6
-0,38
-0,38
17,11
20,5
0,14
0,13
-39,62 -40,23

2,162
-11,88
-0,14
-6,55
-1,51
0,1
0
0,2
-0,21
2,62
0,04
-17,33

3,162
-42,44
-0,49
-18,85
-3,34
0,5
0,2
0,6
-0,39
24,66
0,14
-39,41

2,262
-15,84
-0,18
-8,19
-1,79
0,1
0
0,2
-0,24
3,78
0,07
-22,09

3,262
-44,9
-0,52
-19,71
-3,45
0,6
0,3
0,7
-0,4
27,02
0,14
-40,22

2,362
-19,32
-0,22
-9,71
-2,02
0,2
0
0,3
-0,26
5,02
0,07
-25,94

3,362
-47,46
-0,55
-20,52
-3,55
0,5
0,2
0,8
-0,41
37,37
0,13
-33,49

2,462
-22,27
-0,26
-11,08
-2,23
0,3
0,1
0,3
-0,28
6,44
0,09
-28,89

3,462
-50,08
-0,58
-21,33
-3,65
0,6
0,3
0,8
-0,42
47,71
0,13
-26,52