Date post: | 13-Nov-2023 |
Category: |
Documents |
Upload: | independent |
View: | 0 times |
Download: | 0 times |
Author's Accepted Manuscript
Engineering evaluation of CO2 separation bymembrane gas separation systems
Adele Brunetti, Enrico Drioli, Young Moo Lee,Giuseppe Barbieri
PII: S0376-7388(13)00990-3DOI: http://dx.doi.org/10.1016/j.memsci.2013.12.037Reference: MEMSCI12610
To appear in: Journal of Membrane Science
Received date: 24 September 2013Revised date: 24 November 2013Accepted date: 16 December 2013
Cite this article as: Adele Brunetti, Enrico Drioli, Young Moo Lee, GiuseppeBarbieri, Engineering evaluation of CO2 separation by membrane gasseparation systems, Journal of Membrane Science, http://dx.doi.org/10.1016/j.memsci.2013.12.037
This is a PDF file of an unedited manuscript that has been accepted forpublication. As a service to our customers we are providing this early version ofthe manuscript. The manuscript will undergo copyediting, typesetting, andreview of the resulting galley proof before it is published in its final citable form.Please note that during the production process errors may be discovered whichcould affect the content, and all legal disclaimers that apply to the journalpertain.
www.elsevier.com/locate/memsci
1
Engineering evaluation of CO2 separation by membrane
gas separation systems
Adele Brunetti1, Enrico Drioli1,2, Young Moo Lee2, Giuseppe Barbieri1,*
1 Institute on Membrane Technology (ITM-CNR), National Research Council, c/o The University of Calabria, Cubo 17C, Via Pietro Bucci, 87036 Rende CS, Italy
2 WCU Department of Energy Engineering, College of Engineering, Hanyang University, Seongdong-gu, Seoul 133-791, S. Korea
Author to whom correspondence should be addressed:
Tel. +39 0984 492029; Fax +39 0984 402103; E-mail: [email protected]
Abstract
The possible application of membranes for CO2 separation in the treatment of non‐valuable streams
(e.g., flue gas of a power plant or cement industry) or valuable streams (e.g., biogas) has been analyzed.
Some selection criteria useful in the choice of membrane gas separation for CO2 capture are discussed
to evaluate the advantages potentially offered by membrane systems . Membrane selectivity ranging
from 30‐50 (values of commercial membranes) to 100‐500 (values of most promising laboratory
membranes) and different feed/permeate pressure ratios were considered for the various cases. The
composition and recovery of carbon dioxide in the membrane‐treated stream were the target
parameters taken into account as guidelines in the evaluation of the separation technology
performance. General “maps” of CO2 permeate concentration versus CO2 recovery have been
developed by means of a simple tool that takes into account the influence of the most important
parameters affecting the membrane system performance (that is, membrane selectivity and
2
permeation driving force). The analyses indicated that the separation depends on various interrelated
factors: the membrane material (selectivity and flux), the operating conditions (pressure ratio), and
the final requirements (CO2 recovery and composition). Also, the operational limit and the
potentialities of the membrane gas separation technology were analyzed under these conditions. The
“maps” proposed and utilized for CO2 separation are valid and can be utilized for other gas separations
in which the membrane shows selectivities similar to those taken into account here.
Keywords: Separation maps; CO2 separation; membranes; maps; flue gas; biogas
1. Introduction
Power and hydrogen production, heating systems (for example, in steel and cement industries),
natural gas and biogas purification, etc. are examples of circumstances in which carbon dioxide is
produced in huge (thousands of tons) streams (Table 1). Carbon dioxide separation from hydrogen and
methane streams has long been used since the high value of these streams [1, 2, 3, 4, 5]. Recent
constrains and regulations on CO2 emissions from power plants have focused on the separation of CO2
from flue gas streams [6, 7, 8, 9, 10, 11, 12, 13, 14]. Although technologies such as adsorption,
absorption, and cryogenic distillation [15] were the first to be considered suitable for this purpose,
membrane technology is a valid alternative for carbon dioxide separation from the various
aforementioned streams. Conventional liquid solvent‐based technologies for separating CO2 are used
commercially in the chemical industry [16]. The U.S. Department of Energy (DOE) estimates that post‐
combustion capture using conventional solvents will increase the cost of electricity by about 80% and
incur a $68/ton avoided cost for CO2 [17]. Corresponding projections for pre‐combustion capture are a
30‐40% increase in cost of electricity and $32‐42/ton avoided cost [16,17]. Considering that streams
containing carbon dioxide coming from power plants or heating systems are waste streams with no
economic value, no “profit” margin is involved in their treatment. A significant separation cost (no less
than 20/25 US$/ton) would significantly affect the final cost (e.g., of the electricity) [6]. Therefore, the
3
separation/capture process must consider these aspects in terms of energy demand and limit them to
be as low as possible.
Table 1 – Typical sources of CO2 emissions
Source Separation Feed composition Temperature and pressure
Ref.
Flue gas streams
Power plants Coal gasification plants Steel factory Cement factory Transportation
CO2/N2
5‐25% CO2 65‐80% N2
3‐5% O2 Rest N2, SOx, H2S, H2O
35‐100°C and 1 bar [14, 18]
Natural Gas Natural gas pipes Sweetening of natural gas, etc.
CO2/CH4
1‐8% CO270‐90% CH4
0‐20%C2H6, C3H8, C4H10 Rest O2, N2, H2S, Ar, Xe,
He
25‐30°C and 1.2 bar
[19, 20, 21]
Biogas Various 34‐40% CO250‐70 %CH4
Rest N2, O2, H2S, H2O
25‐35°C and 1 bar
[22]
Because of their fundamental engineering and economic advantages over competing separation
technologies, membrane operations (used in the past for the separation of CO2 from natural gas or
hydrogen streams separations) [1,2,3,4,5, 23] are now being explored for CO2 capture from power
plant emissions and other fossil‐fuel‐based flue gas streams. For example, the great interest in this
new capturing technology has recently been confirmed by the EU‐FP6 project “Nanostructured
Materials against Global Warming” (Nanoglowa) [14,24], where the field application of membrane
modules with large membrane area has been tested in pilot plants in power stations, and the
membrane unit performed well in terms of both separation and durability [25].
Membranes play an important role in the treatment of these waste streams, mainly because of their
greater flexibility with respect to other separation technologies. For instance, the temperature and
heat requirements in the solvent recovery stage of absorption cannot be changed much. On the
contrary, the separation driving force (and hence the compression costs) can change over a wide range
and can be compensated by the appropriate membrane surface area in membrane processes. The
separation driving force (and hence the compression costs) can change over a wide range and can be
4
compensated by the appropriate membrane surface area in membrane processes. On the contrary, for
instance, heat requirements are set from the chemical bond between the gas and solvent as well as the
operating temperature in solvent recovery stage of absorption; thus they cannot be changed much.
Many materials can be considered suitable for the separation of CO2 from flue gas or methane streams
[26, 27, 28, 29, 30, 31, 32, 33], and many advances have been made in the maximization of their mass
transport properties. However, questions remain about the scalability and durability of these
materials under real conditions. Generally, there is a debate among the scientific community about the
fact that it could be more convenient to have a membrane with high permeance and low selectivity, or
vice versa (the two limit conditions of Robeson diagram) [34]. In this regard, an interesting parametric
study on the impact of membrane materials and process operating conditions on carbon capture from
humidified flue gas was performed [35]. For all cases, high membrane CO2 permeance minimizes
membrane area requirements, while high CO2/N2 selectivity improves the CO2 permeate concentration
and reduces the energy needed for CO2 purification. The benefits of higher selectivity are accentuated
at higher feed‐to‐permeate pressure ratios, at the expense of increased energy cost. The advantages of
higher permeance are most pronounced at lower pressure ratios. The engineering design of a
membrane separation unit for the recovery of CO2 must thus take into account various factors. In our
previous work [6], we developed a simple tool that uses “maps” to enable analysis of performance and
the perspectives of membranes in CO2 capture. That study focused on the the treatment of a stream
with the typical composition of the flue gas coming from the power plant, thus containing a low
percentage of CO2 (ca. 13 mole%). However, from various considerations, it appears evident that the
suitability of the membrane gas separation with respect to other separation technologies is strictly
related to various parameters, particularly the CO2 feed composition, the feed conditions (pressure
and temperature), the product permeate concentration, and the final destination. While some of these
variables are intrinsic, and in some cases, limited by the specific process (such as the feed composition
and conditions), the others are strictly connected to the design of the separation process and
significantly to the characteristics of the membranes chosen for that defined separation. In most cases,
the final design of the process depends on the synergistic evaluation of the effect that these variables
exercise on the separation.
5
This work more widely analyzes the application of membrane gas separation in CO2 processing
with a general approach considering the effect and, eventually, the limitations offered by the main
variables that affect the separation performance: the pressure ratio, the feed composition, and the
mass transport properties (permeance and selectivity) of the membrane considered in the installation.
As performed, the study represents a useful guide for readers interested in CO2 separation
independent from the other gases present in the feed stream. It is thus suitable for flue gas separation
as well as natural gas and biogas. The model results are also valid for other separation characterized
by similar membrane selectivity. The cost analysis is not part of the study that is focused on the
evaluation of the feasibility of this type of technology.
In addition, the analysis is valid for the preliminary design of the performance of a single‐stage or
multi‐stage separation system, since the maps can be used to evaluate the performance and the
requirements of each membrane unit depending on the input parameters.
2. Methods
As mentioned previously, the simulations described in this work have been carried out using a simple
tool developed in our previous work [6], taking into account the influence of the most important
parameters affecting the membrane system performance. This tool consists of a dimensionless 1D
mathematical model for the multi‐species steady‐state permeation in no sweep mode and co‐current
configuration. For convenience, the tool is explained in the Appendix. The results achieved by these
simulations are described in terms of general maps of CO2 permeate concentration versus CO2
recovery. These are useful for analyzing different design solutions in terms of membrane area and
pressure ratio to be installed and for comparing different systems having the same performance. More
details on this tool can be found in [6], where the validation of the model via experimental literature
results is also presented.
In the dimensionless form of the equations, the terms Θi and φ can be distinguished as the
permeation number and the feed to permeate pressure ratio, respectively.
6
2
2
2
Membrane FeedCO
CO Feed FeedCO
A PΘ
x QΠ
= (1)
Feed
Permeate
PP
ϕ = (2)
2
2
2CO
CO
CO permeating fluxP
Π =Δ
(3)
2 2 2
/ReFeed tentate PermeateCO CO COP P PΔ = − (4)
The permeation number (1) expresses a comparison between the two main transport mechanisms
involved. A high permeation number corresponds to a high membrane area and/or permeance (3) for
the stream and to a high permeation through the membrane with respect to the total flux along the
module. The pressure ratio (2) represents one of the most important and determinant operating
parameters affecting the performance of the membrane unit and is the driving force for the separation.
For a given feed composition, membrane properties (species permeance and selectivity), module
geometry (total installed membrane area and module length), and fixed operating conditions (feed
flow rate and pressures), the solution of the equation system provides the species dimensionless flow
rate profiles and composition along the module length for both the membrane sides. The overall
membrane module performance in terms of, for instance, final species concentration and total CO2
recovery in the permeate (5) stream are easily calculated considering the value of state variables
(species composition) at the module exit.
22
2
cov *100CO permeate flowrateCO re eryCO feed flowrate
=
(5)
In particular, the performance of a membrane separation is measured as CO2 concentration in the
permeate stream and CO2 recovery as function of several parameters such as, membrane selectivity,
7
pressure ratio and feed composition. Hence, permeance, membrane area and feed flow rate are taken
into account through the permeation number. Most of the proposed plots, namely maps, show CO2
concentration in the permeate stream versus CO2 recovery.
In general, the approach proposed with the maps use is a top‐down one: it starts from the target of
separation, that is, recovery and concentration of the permeate stream. Each curve is plotted in order
to directly read this information as function of a membrane property (selectivity) and operating
condition (feed/permeate pressure ratio). Other information, such as, permeance and the related
membrane area (or other variables) are related with one‐another by means of the permeation number
which is also reported on the maps. The complexity of the variable named permeation number seems
as a limitation but a same curve (maps) can be intended valid. For instance, the same product between
membrane area and permeance provides the same value of permeation number. In other word, the
same curve is valid for any permeance value, very high as well as medium or very low. The permeation
number red in the plot will be used, for example, to set the membrane area of a chosen membrane,
once the operating point is identified on the maps. This is a clear advantage of using the permeation
number.
Investigation of the whole range of operating conditions (feed pressure and flow rate) produces
global maps showing all the possible solutions for the considered gas separation membrane system,
which are expressed as parametric curves for the pressure ratio. In addition, curves at a set
permeation number cross these maps. Each point in a map corresponds to the total membrane module
performance for a given set of operating conditions. The map can supply two important pieces of
information: the maximum performance achievable by a membrane unit once such operating
conditions have been fixed; or the operating conditions, membrane area, or membrane type required
to produce a stream with a certain recovery and permeate concentration.
Table 2 shows the stream conditions, membrane selectivity, and operating parameters used as a
case study in the following calculations. These have been chosen with consideration to membrane
materials properties currently available in the literature, as indicated in previous sections of the paper.
Throughout the text, the composition of the gas mixtures is expressed in mole%.
8
Table 2 – Membrane properties and operating conditions used in the calculations
Membrane properties
Selectivity, CO2/i 30; 50; 100; 250; 500
Operating conditions
CO2 concentration in the feed, mole % 10; 25; 35; 50
Pressure ratio, ‐ 1.5; 2.5; 5 and 50
Permeation number, ‐ 1; 10; 20
Any selectivity value can be set in the simulation for using the predictive property of the model.
Therefore, the range of selectivity considered is chosen to explore a wide spectrum of selectivity
values actually present in the literature, including selectivities typical of membranes already available
in the market (30‐50) [11, 25] and values only available at lab‐scale (100‐500). Most polymeric
membranes exhibit selectivity in a mixture (or so‐called separation factor) that differs from the one
measured in ideal conditions with pure gas tests. In most cases, the consideration of this value with
respect to the ideal one allows more proper evaluation of the membrane separation unit performance.
In the maps, the term indicated as “selectivity” can be independently intended as the one measured
with pure gases or the one measured in mixture, when available.
The CO2 concentration cases chosen for calculation explore the various conditions that can be
found in real applications; from the low value of concentrations typical of flue gas from power plants
(10 %) to flue gas exiting a cement factory (25‐35 %) to the typical concentration of CO2 in biogas (35‐
50 %).
In the separation of CO2 from flue gas, the main limitation is the pressure of the stream coming
from the chimney of the plant, which in most cases is atmospheric. The large flows tend to be treated,
and the absence of added value in the product to be recovered (CO2) limits the availability of extra‐
pressure supplied to the stream by means of compressors. Thus, pressure ratios considered in these
calculations are limited (ranging from 1.5‐5) when discussing flue gas separation. Circumstances are
different for biogas separation, where not only is the separation product concentrated CO2 in the
9
permeate, but purified methane can be retained in the retentate of the already compressed membrane
stream. In this case, the high added value methane, the lower flow rates, and the fact that the purified
methane has to be pumped in the pipeline at 40‐50 bar, make it affordable to operate with high
pressure ratios in the membrane unit. For these reasons, this work focuses on the study of membrane
unit performance for low pressure ratios, typical case of flue gas streams, and also includes the case of
a high pressure ratio of 50 to evidence the potentialities achievable when a high driving force is
available, as in the case of biogas and natural gas.
3. Results and Discussion
As mentioned previously, the possibility of using a membrane unit in the separation of a gaseous
stream is strictly connected to three main factors: the composition of the feed, the available operating
conditions, and the separation properties of the membrane chosen for the specific application. This
work discusses the effect of each of these parameters and their relative interaction on the
performance of a typical membrane unit, before discussing the performance maps and their practical
use in a preliminary design of a membrane separation system.
In the following, CO2/N2 is used as an example to show the application of the tool. However, since
other membranes show similar selectivity for different gases, the same graphs can be used as
reference for other separations, identifying from time to time the correct selectivity and permeation
number for the desired operating condition (e.g., pressure ratio).
3.1 Effect of selectivity at different feed molar compositions
Figures 1‐3 show the permeate concentration versus recovery for different values of CO2
concentration in the feed at different pressure ratios. In all the cases, the permeation number has
unitary value and using eq.(1) and the membrane area can be calculated setting the permeance and
the other process parameters. Figure 1 shows the case at a pressure ratio of 1.5. At first observation, it
10
appears that, for all the conditions considered, a very low recovery is achieved because the low
pressure ratio strongly limits the driving force required for permeation.
Figure 1. CO2 permeate concentration as function of CO2 recovery at various values of selectivity and CO2 feed concentration. Pressure ratio φ=1.5.
The permeate concentration is instead much more dependent on the feed composition, which
defines the driving force together with the pressure ratio. At a pressure ratio of 1.5, a feed stream
containing a low percentage of CO2 cannot reach high values of permeate concentration due to the
insignificant driving force (Figure 1). As the fraction of CO2 in the feed increases, the achievable
permeate concentration also increases. The permeate concentration cannot exceed 16 % if the feed
stream contains only 10 % CO2, whereas a value of ca. 75 % can be achieved when the feed contains
around 50 % CO2. The strong limitation on the driving force does not allow increased effect of the
membrane selectivity on the performance of the membrane unit. The low driving force represents the
rate determining step of the permeation, and the use of a membrane with a very low value of
selectivity (α=30) or a membrane with extraordinary separation properties (α=500) is equivalent
each‐other for the final performance of the unit. The situation changes when a higher pressure ratio is
utilized (Figure 2, Figure 3).
0 20 40 60 80 100
CO2 recovery, %
0
20
40
60
80
100 1.5
30
500
x CO2
Feed=10%mol
x CO2
Feed=25%mol
x CO2
Feed=50%mol
x CO2
Feed=35%mol
11
Figure 2. CO2 permeate concentration as function of CO2 recovery at various values of selectivity and CO2 feed concentration. Pressure ratio φ=2.5
When the pressure ratio is 2.5 (Figure 2), the performance of the system improves, particularly
when the feed is not too diluted. For 10 % CO2 in the feed, a permeate concentration no higher than 25
% can be achieved. Even though this performance is better than that obtained for a pressure ratio of
1.5 (and 16 % CO2 in the permeate), no significant improvements can be seen in terms of CO2 recovery.
When the CO2 feed concentration increases, the performance of the unit improves significantly due to
the favored driving force, and the effect of selectivity becomes more evident.
For instance, at 35 % CO2 in the feed, the membrane unit reaches a permeate concentration in the
range of 60‐88 % depending on the selectivity of the membrane. In general, a higher permeate
concentration and a lower recovery correspond to a higher selectivity, owing to the lower driving
force of the more permeating component along the module axis, and vice versa, since the calculations
are made considering the same permeance and area of membrane. Therefore, a permeate
concentration of 60‐72 % can be achieved with a membrane selectivity of 30, whereas these values
increase to 70‐80 % and 80‐87 % for selectivity values of 250 and 500, respectively. These trends are
much more evident at a pressure ratio of 5 (Figure 3). In this case (also at a low CO2 concentration in
the feed), both recovery and permeate concentration can be significantly improved since the limiting
effect of the diluted stream on the driving force is compensated by the higher pressure ratio.
0 20 40 60 80 100
CO2 recovery, %
0
20
40
60
80
100
CO2
perm
eate
con
cent
ratio
n, %
2.5
30
50100
250500
x CO2
Feed=10%mol
x CO2
Feed=25%mol
x CO2
Feed=35%mol
x CO2
Feed=50%mol
12
Therefore, a stream with 10 % CO2 in the feed can be concentrated at ca. 20 %, recovering ca. 75 % of
the initial CO2 with a membrane having a selectivity of just 30. Alternately, by using a membrane with
a higher selectivity (250‐500), a permeate concentration of 43‐48 % can be achieved, at which
correspond a recovery of less than 20 %. The improvement offered by the higher pressure ratio in this
case is even more evident at a higher CO2 feed concentration.
Figure 3. CO2 permeate concentration as function of CO2 recovery at various selectivities and CO2 feed concentrations. Pressure ratio φ=5.
On the basis of the above considerations, the selectivity of the membrane is fundamental for the
final permeate concentration of the permeate stream. Its effect is limited when the permeation driving
force is not sufficient; therefore, streams with low CO2 concentration in the feed and/or low pressure
ratio cannot achieve high permeate concentration level in one stage, even using membranes with
extraordinary selectivity. Figure 5 can be utilized to analyze a multistage separation process, but a
description of the figure is required first. Each curve shows the relationship between the feed
concentrations of the CO2 and permeate stream for different values of membrane selectivity and
pressure ratio. Therefore, for a selected CO2 feed composition at any membrane stage (for instance, the
first or second), the permeate composition of the same stage can be read on the y‐axis.
0 20 40 60 80 100
CO2 recovery, %
0
20
40
60
80
100
CO2
perm
eate
con
cent
ratio
n, %
5
3050
100
250500
x CO2
Feed=10%mol
x CO2
Feed=25%mol
x CO2
Feed=50%mol
13
Considering a stream with a CO2 feed concentration of 15 % (typical of flue gas from a power plant)
and a pressure ratio of 1.5 (Figure 5A), the use of a first stage that has a highly permeable membrane
with a selectivity typical of commercial membranes (α=30) allows a stream concentration of ca. 25 %
CO2. When this latter stream at 25 % CO2 enters the second stage, a permeate concentration of 60 %
can be achieved by using a membrane with a higher selectivity (ca. 100) at a pressure ratio of 2.5
(Figure 5 and Figure 4A case I). The permeate concentration can be higher than the 80 % value
imposed by the International Energy Agency [36] by using a pressure ratio equal to 5 (Figure 5 and
Figure 4A case II) in the second stage. A permeate concentration greater than 80 % can also be
reached by using the solution proposed in Figure 5 (right side), where both membrane stages (The
case III considers the same selectivity but a different pressure ratio in the first stage) are operated at a
pressure ratio of 2.5 (Figure 4B).
(A) ‐ Case I
(A) ‐ Case II
(B)
14
Figure 4. Schemes of multistage configurations
Figure 5. CO2 permeate concentration as function of CO2 feed concentration at different selectivities for a multistage configuration
15
3.2 Effect of pressure ratio at different permeation numbers and selectivity values
The examples described above illustrate the fundamental role of the pressure ratio in the final
performance of the membrane module. The effect of this parameter is emphasized depending on the
other variables affecting the separation, such as the CO2 concentration in the feed, membrane
selectivity, and permeation number. In principle, a membrane module with set membrane properties
(permeance and selectivity) operated at a higher pressure ratio does not necessarily indicate a higher
permeate concentration but does mean a greater recovery of the desired product at the higher CO2
feed composition due to the effect on the overall driving force promoting the CO2 permeation. Figure 6
(left side) shows the CO2 permeate concentration and recovery as functions of the pressure ratio at
various values of CO2 concentration in the feed for a membrane with a selectivity of 50. The permeate
concentration initially increases with the pressure ratio up to a maximum and then decreases. An
increase in the pressure ratio first promotes the permeation of the most permeable specie; however,
as it increases, the partial pressure difference of the other less permeable species present in the feed
stream also increases, causing their permeating fluxes to increase and the consequent dilution of the
permeate stream.
Figure 6. CO2 permeate concentration (left side) and CO2 recovery (right side) as a function of (both) pressure ratio at different values of CO2 feed concentration. Theta=1, selectivity=50
16
On the contrary, the recovery of CO2 (Figure 6 right side) increases at all the CO2 feed
concentrations until reaching a plateau where all the CO2 is permeated through the membrane. Higher
concentrations of CO2 in the feed stream correspond to lower pressure ratios once the plateau is
reached.
The promotion of recovery can also be obtained by operating with a greater permeation number,
which means increasing, for example, the membrane area or using a membrane with a greater
permeance. As shown in Figure 7 (where a permeation number of 10 is used), the gain in terms of
recovery is also evident when a diluted CO2 stream is fed to the membrane module. At
a CO2 feed composition of 25 % and a pressure ratio of 2.5, less than 25 % is recovered, but this value
increases up to around 60 % at a pressure ratio of 5. The higher recovery is counterbalanced by a loss
in CO2 permeate concentration, particularly at high pressure ratios. The combination of a high
permeation number and pressure ratio implies the maximization of the recovery, since the module
offers both sufficient membrane area and driving force to allow permeate into the more permeable
component, which in this case is CO2. Meanwhile, the molar fraction of the less permeable components
also increases, and consequently, their permeation is promoted, depleting the final concentration of
the permeate stream. On the basis of the feed conditions and the final stream target, a compromise
between these two parameters must be found to obtain the desired levels of permeate concentration
and recovery.
In general, the trend is similar for all the selectivities investigated, even though (as discussed in
Section 3.1) higher selectivity corresponds to higher permeate concentration and lower recovery once
the permeation number and pressure ratio are fixed.
17
Figure 7. CO2 permeate concentration and CO2 recovery as a function of pressure ratio at different values of CO2 feed concentration. Permeation number Θ=10, selectivity α=50
3.3 Effect of permeation number at different pressure ratios and selectivities
The proper design of a membrane unit consists of a comprehensive evaluation of the feed conditions,
the final targets, and the available operating condition ranges that are plausible for the specific stream
to be treated, leading to the choice of the membrane material and thus to the definition of the
geometric characteristics (membrane area, number of modules, etc.) of the separation unit. The
permeation number contains these process variables in its definition; therefore, it can be used in
various ways during the design of the membrane separation unit. Permeation number:
• indicates the membrane area necessary to carry out the separation of a specified feed stream
with its own characteristics (under certain operating conditions) once the membrane material
that constitutes the membrane unit has been chosen on the basis of its mass transport
properties (permeance and selectivity);
• can be used to identify the proper membrane material (specifically referring to the permeance
of the latter) once a certain geometry of the module (and thus the membrane area) have been
18
identified on the basis of different design considerations, such as the footprint occupied by the
installation, the dimension of the single membrane modules, etc.;
• identifies the feed flow rate of a stream with a certain composition that can be treated by that
module under set operating conditions, e.g., the feed pressure when module characteristics
such as membrane material and membrane area are already defined.
In most cases, the first listed application/option is the preferred one, since the membrane material
can be chosen on the basis of the final target (leading to the choice of a membrane with certain mass
transport properties) or on the basis of limitations dictated by some process conditions like high
temperature, presence of contaminants, etc., restricting the choice of material suitable for that use.
The other two approaches are valid if coupled with the evaluation of other parameters such as the
pressure ratio and membrane selectivity needed to obtain a certain permeate concentration, etc.
As mentioned previously, the permeation number influences the final permeate concentration once
the pressure ratio and the membrane selectivity are set. In these conditions, a high permeation
number corresponds to a high residence time for the stream to be separated and then to a high
permeation through the membrane with respect to the total flux along the module. In terms of final
product conditions (permeate concentration and recovery), an increase in the permeation number
promotes the permeation firstly of the most permeable species (the desired product), then also of the
less permeable components. Their concentrations increase along the membrane module due to the
permeation of CO2 consequently increasing their driving force (Figure 8). If a greater membrane area
is available, then these other species permeate, and the final permeate concentration of the permeate
stream is depleted. However, the greater permeating flux allows a higher recovery that is thus an
increasing function of the permeation number up to a plateau where all the CO2 is permeated. The
trend is analogous to when a membrane with a very high selectivity is used (Figure 9). In this case, the
reductive effect of the permeation number on the final concentration of the permeate stream is less
evident due to the better separation properties of the membrane, but a lower recovery can be
achieved. This latter property can be improved by increasing the pressure ratio (Figure 10), which can
double the recovery when it increases from 2.5 to 5. As shown in Section 3.2, a significant increase in
19
the pressure ratio leads to a reduction in the final permeate concentration. This pressure ratio limit
(beyond which the permeate concentration is depleted) shifts toward higher values as the membrane
selectivity is greatly increased.
Figure 8. CO2 permeate concentration and CO2 recovery index as a function of permeation number at different values of CO2 feed concentration. Pressure ratio=2.5, selectivity=50
Figure 9. CO2 permeate concentration and CO2 recovery index as a function of CO2 feed concentration at different values of permeation number. Pressure ratio=2.5, selectivity=500
20
Figure 10. CO2 permeate concentration and CO2 recovery index as a function of CO2 feed concentration at different values of permeation number. Pressure ratio=5, selectivity=500
3.4 Maps
The potential of the tool developed in our previous work [6] and utilized here is its possibility for use
in a preliminary design of a membrane unit for gas separation. Global economic considerations of the
final electricity cost and CO2 storage technology allow the optimal performance (that is, a point on the
plot of CO2 permeate concentration vs. CO2 recovery) to be univocally individuated on the maps; the
parametric curves crossing this optimal point provide the corresponding pressure ratio and
permeation number. This leads to the identification of the operating conditions, membrane
characteristics (permeance, area, etc.), or feed conditions required to obtain the final product with
certain characteristics. The maps are singularly calculated for a set value of selectivity (30‐500) and
CO2 feed concentration; the curves of permeate concentration versus recovery are parametric with
regard to pressure ratio and permeation number (Figure 11).
In the maps, all the parametric curves at constant pressure ratio have the same trend: higher
recovery corresponds to lower CO2 concentration in the final permeate stream. Moreover, increasing
the pressure ratio causes the curves to shift upward, which means that a higher permeate
21
concentration is achieved at the same CO2 recovery. At a given permeation number (dashed lines), the
CO2 permeate concentration increases with the recovery, as shown in the left part of the plot (low
stage‐cut), whereas a higher permeation number generally leads to a higher recovery but lower
permeate concentration. The main part of the feed flow rate permeates through the membrane, and N2
dilutes the CO2 permeated preferentially in the first part of the module (Section 3.3).
For instance, in a case where the membrane to be used has been chosen, the permeance and
selectivity are known. Moving in the map and for a known feed stream composition, it is possible to
identify the pressure ratio and permeation number required to obtain a permeate stream with certain
characteristics. For a very dilute feed stream (10 % CO2 feed concentration), in most cases, a
multistage membrane system is used for CO2 separation, where the first stage concentrates the stream
to improve the driving force at the second stage. Figure 11A‐D show the maps calculated at different
values of membrane selectivity. Each map is developed at various CO2 feed concentrations for a set
value of selectivity. This means that it can be used to evaluate each membrane gas separation unit
constituting the separation system and in cases of multistage configuration. Therefore, considering the
case of a stream containing the 10 % CO2 fed to a first membrane unit with a selectivity of 50, if a 30‐
35 % concentration of CO2 in the permeate stream is reputedly enough to feed the second membrane
stage, then by entering in the map with a line perpendicular to the y‐axis, it is possible to identify the
pressure ratio and permeation number necessary to achieve that permeate concentration and CO2
recovery in the second stage. In this specific case, a pressure ratio of at least 5 is necessary with a
permeation number of 10, which corresponds to a recovery of ca. 45 %. Since CO2 feed concentration
is fixed, the permeance is known and the feed pressure is a function of the pressure ratio and thus is
also univocally identified in the map; the permeation number ( 2
2
2
Membrane FeedCO
CO Feed FeedCO
Permeance A PΘ
x Q= )
can determine the value of the membrane area required to treat a certain feed flow rate that achieves
those values of permeate concentration and recovery. In cases of defined dimensions of the membrane
module, the permeation number can be used to determine the treatable flow rate value or identify the
value of permeance required to achieve the desired performance.
22
In general, as the CO2 feed concentration increases, the membrane module can be used to attain
highly concentrated permeate streams. This is the job of a second/third stage of a membrane
separation system, for example, in cascade configuration. The multiple maps presented in Figure 11
can also be used to identify the operating conditions for a second/third separation stage. Considering,
for example, that the permeate concentration of the stream exiting the first stage fed with 10 % CO2
mixture is ca. 35 % (condition achievable by operating the first stage at a pressure ratio of 5 and
permeation number of 10) when using a membrane with selectivity of 50, it is possible to enter the
second diagram corresponding to the membrane with a higher selectivity and thus identify the
operating conditions necessary to achieve a defined final permeate concentration. For example,
assuming a membrane unit with a selectivity of 100 (Figure 11B), the permeate concentration of 75 %
in the second stage can be reached at a pressure ratio of 2.5 but with ca. 20 % recovery or at a
pressure ratio of 5 with recovery of more than 60 %. Depending on the choice between these
operating options, the permeation number is univocally defined. The conditions and final recovery
change as expected, thus changing the selectivity. For the same pressure ratios and permeation
numbers, higher recoveries can be achieved or streams with the same recoveries but higher permeate
concentrations can be produced. This is much more evident for high CO2 feed concentrations. The
maps offer the possibility to identify various operating options that can be chosen depending on the
specific requirements and limitations of the process. However, it has to be pointed out that the effect of
the impurities is not considered at this stage of simulation it being a preliminary tool for performance
evaluation.
23
A)
0 20 40 60 80 100
CO2 recovery, %
0
50
100
1.52.5 5
1
2010
0
50
100
1
10
201.5
2.55
0
50
100
1.5 2.5
x CO2
Feed=35%mol
50
50
50
5 50
110
20
x CO2
Feed=10%mol
x CO2
Feed=50%mol
24
B)
0 20 40 60 80 100
CO2 recovery, %
0
50
100
x CO2
Feed=10%mol1.52.5
5
1
2010
0
50
100
x CO2
Feed=50%mol
1
1020
1.5
2.5 5
0
50
1001
1020
1.52.5
5
x CO2
Feed=35%mol
100
50
50
50
26
D)
Figure 11 – Maps of CO2 permeate concentration as a function of CO2 recovery index at various values of CO2 feed concentration. Pressure ratio and permeation number is also reported. A) Selectivity=50; B) Selectivity=100; C) Selectivity=250, D) Selectivity=500
27
3.5 Further remarks
As the major part of the models also the one used for having the presented results prescind from the
chemical nature of membrane material, since the only information taken into account is the selectivity
and permeance. Other membranes will show similar selectivity for different gases. Therefore, the
results discussed here for CO2/N2 separation can be used to analyse other separations, including the
proper selectivity and permeation number for the desired operating condition (e.g., pressure ratio).
For this reason, operating conditions change in a range wider than expected for this separation, such
as a pressure ratio of 50. This latter condition of operation is unusual for the treatment of flue gas
streams since it implies a very high pressure difference between the two membrane sides that can be
concretized by using compressors and/or high performance vacuum pumps. However, with regard to
other streams containing CO2 that must be separated (such as biogas), the possibility to operate with
high pressure on the feed side is not far from the real application. The separation of biogas leads not
only to the recovery and sequestration of CO2, but also to much greater purification and recovery of
value‐added CH4 in order to feed it directly to pipelines for domestic or stationary uses. From this
perspective, since CH4 has to be fed to pipelines at a high pressure, the possibility of installing a
compressor before the membrane system and recovering the methane already concentrated and
compressed as a retentate stream makes this operating option quite realistic.
4. Conclusions
The possibility of using a membrane unit in the separation of a gaseous stream is strictly
connected to three main factors: the composition of the feed, the available operating conditions,
and the separation properties of the membrane chosen for the specific application. Therefore,
together with material science, a crucial role for the real application of membrane technology in
CO2 separation is played by membrane engineering, which involves the integrated scheme design
and optimization of the operating conditions. This work systematically discusses the performance
maps and their practical use in the preliminary design of a membrane separation system, taking
28
into account a wide range of feed and operation conditions, as well as membranes and membrane
module separation properties. The effects of each of these parameters and their relative influences
on the performance of a typical membrane unit are also discussed in detail.
Together with the feed conditions, one variable significantly affecting the performance of the
membrane module is the feed/permeate pressure ratio. The permeation number is a determining
parameter for the module performance. For a set feed flow rate, a set membrane type, and defined
pressure ratio, a low permeation number indicates low recovery and high permeate concentration,
and vice versa. The low CO2 concentration in the feed does not allow high permeate concentration
streams, even when increasing the pressure ratio, and thus more separation stages are necessary.
The effect of the selectivity on the performance of the membrane module is negligible at low
pressure ratios but becomes important as this value is increased. For a high value of selectivity,
doubling of the pressure ratio implies a recovery 2‐3 times greater and improvements in the CO2
permeate concentration.
Acknowledgements
The “Ministero degli Affari Esteri, Direzione Generale per la Promozione e la Cooperazione Culturale”
of Italy is gratefully acknowledged for the financial support of project “New highly innovative
membrane operations for CO2 separation (capture) at medium and high temperature: Experimental
preparation and characterization, theoretical study on elementary transport mechanisms and
separation design” co‐funded in the framework of a bilateral agreement between MAE (Italy) and
MOST (South Korea). YML thanks for the financial support of Korea‐Italy Bilateral Cooperation
Program and Korea CCS R&D Centre, National Research Foundation, the Ministry of IT, Science and
Technology in Korea. The present work was also performed in the framework of the activities of the
“International Joint Laboratory on Membrane Technology” established between ITM‐CNR (Italy) and
Hanyang University (Korea), in Seoul on June 14th, 2011”. G. Barbieri wishes to thank the Korea CCS
29
R&D Center (South Korea) for the inviting to give a talk at the 3rd Korea CCS conference, Jeju (South
Korea), March 14th‐15th, 2013.
List of symbols
α CO2/i selectivity φ Pressure ratio θ Permeation number Π Permeance, m3 m‐2 h‐1 bar ‐1 A Membrane area, m2 P Pressure, bar Q Flow rate, m3 h‐1 x Molar fraction, ‐
Membrane engineering analysis of CO2 separation from flue gas
General “maps” of CO2 permeate concentration versus CO2 recovery
General “maps” with selectivity typical of commercial (30‐50) and latest laboratory generation (100‐
500) membranes
General “maps” with different feed/permeate pressure ratio
References
[1] L. Shao, B. T. Low, T.‐S. Chung, A.R. Grenberg, Polymeric membranes for the hydrogen economy:
Contemporary approaches and prospects for the future, J. Mem. Sci. 327, 1–2, (2009) 18–31
[2 E. Favre, R. Bounaceur, D. Roizard, Biogas, membranes and carbon dioxide capture, J. Mem. Sci. 328 (2009) 11–14
[3] S. Basu, A. L. Khan, A. Cano‐Odena, C. Liu, I. F. J. Vankelecom, Membrane‐based technologies for biogas separations, Chem. Soc. Rev., 39 (2010) 750–768
[4] H. Lin, E. Van Wagner, R. Raharjo, B. D. Freeman, I. Roman, High‐Performance Polymer Membranes for Natural‐Gas Sweetening, Adv. Mater. 18 (2006) 39–44
[5] C. A. Scholes, J. Bacus, G. Q. Chen, W. X. Tao, G. Li, A. Qader, G. W. Stevens, S. E. Kentish, Pilot plant performance of rubbery polymeric membranes for carbon dioxide separation from syngas, J. Mem. Sci., 389 (2012) 470–477
30
[6] A. Brunetti, F. Scura, G. Barbieri, E. Drioli, Membrane technologies for CO2 separation, J. Mem. Sci. 359 (2010)
115–125 [7] J.P. Ciferno, T. E. Fout, A. P. Jones, J.T. Murphy, Capturing carbon from existing coal‐fired power plant, Chem.
Eng. Prog. (2009), 33‐41
[8] H. Herzog, What future for carbon capture and sequestration?, Env. Sci. Tech. 35 (7) (2001) 148‐153
[9] C. M. White, Separation and capture of CO2 from large stationary sources and sequestration in geological formations, J. Air Waste Management Association 53 (2003) 645‐715
[10] E. Favre, Carbon dioxide recovery from post‐combustion processes: Can gas permeation membranes compete with absorption?, J. Mem. Sci. 294 (2007) 50‐59
[11] T. C. Merkel, H. Lin, X. Wei, R. Baker, Power plant post‐combustion carbon dioxide capture: An opportunity for membranes, J. Mem. Sci., 359, 1–2, 1 (2010) 126–139
[12] B. Li, Y. Duan, D. Luebke, B. Morreale, Advances in CO2 capture technology: A patent review, App. En., 102 (2013)1439–1447
[13] L. Peters, A. Hussain, M. Follmann, T. Melin, M. B. Hägg, CO2 removal from natural gas by employing amine absorption and membrane technology—A technical and economical analysis. Chem. Eng. J. 172 (2011) 952–960
[14] L. Daal, L. Claassen, R. Bruns, B. Schallert, G. Barbieri, A. Brunetti, K. Nijmeijer, “Field tests of carbon dioxide removal from flue gasses using polymer membranes”, VGB Powertech, 2013, 6, 78‐84
[15] M.J. Tuinier, H.P. Hamers, M. van Sint Annaland, Techno‐economic evaluation of cryogenic CO2 capture—A comparison with absorption and membrane technology, Int. J. Greenhouse Gas Contr. 5 (2011) 1559–1565
[16] J.M. Klara et al., Cost and performance baseline for fossil Energy plants. Volume 1: Bituminuous coal and natural gas to electricity final report, DOE/NETL‐2007‐1281, (2007) http://www.netl.doe.gov/energy-analyses/pubs/Bituminuous%20Baseline_Final%20Report.pdf.
[17] A.Y. Ku, P. Kulkarni, R. Shisler, W. Wei, Membrane performance requirements for carbon dioxide capture using hydrogen‐selective membranes in integrated gasification combined cycle (IGCC) power plants, J. Mem. Sci., 367 (2011)233‐239
[18] Pidwirny, M. (2006). "Atmospheric Composition". Fundamentals of Physical Geography, 2nd Edition. Date Viewed. http://www.physicalgeography.net/fundamentals/7a.html
[19] http://naturalgas.org/overview/background.asp
[20] Baudot in E. Drioli and G. Barbieri, eds. Membrane Engineering fro vthe treatment of gases, The Royal Society of Chemistry, Cambridge, The United Kingdom, 2011, pages 150‐195, ISBN 978‐1‐84973‐239‐0
[21] Xiao Y., Low B.‐T., Hosseini S. S., Chung T. S., Paul D. R., The strategies of molecular architecture and modification of polyimide‐based membranes for CO2 removal from natural gas—A review”, Progress in Polymer Science 34 (2009) 561–580 [22] Deng L, Ha¨gg M‐B. Techno‐economic evaluation of biogas upgrading process using CO2 facilitated transport membrane. Int. J. Greenhouse Gas Control (2010), 4, 638–646 [23] X. He, M. –B. Hägg, Membranes for Environmentally Friendly Energy Processes, Membranes (2012), 2(4), 706-726
[24] http://www.nanoglowa.com/index.html
[25] L. Daal, L. Claassen, R. Bruns, B. Schallert, G. Barbieri, A. Brunetti, K. Nijmeijer, Field tests of carbon dioxide removal from flue gasses using polymer membranes, VGB Powertech, 2013
[26] C. E. Powell, G. G. Qiao, Polymeric CO2/N2 gas separation membranes for the capture of carbon dioxide from power plant flue gases, J. Mem. Sci. 279 (2006) 1–49
[27] P. Luis, T. Van Gerven, B. Van der Bruggen, Recent developments in membrane‐based technologies for CO2 capture, Progr. En. Com. Sci. 38, 3(2012) 419–448
[28] K. Ramasubramanian, WS W. Ho, Recent developments on membranes for post‐combustion carbon capture, Curr. Op. Chem. Eng. 1, 1 (2011) 47–54
31
[29] H. B. Park, C. H, Jung, Y. M. Lee, A. J. Hill, S. J. Pas, S.T. Mudie, E.Van Wagner, B. D. Freeman, D. J. Cookson,
Polymers with Cavities Tuned for fast and Selective Transport of Small Molecules and Ions, Sci. 318 (2007) 254‐258
[30] C. H. Jung, J. E. Lee, S. H. Han, H. B. Park, Y. M. Lee. Highly permeable and selective poly(benzoxazole‐co‐imide) membranes for gas separation, J. Mem. Sci. 350 (2010) 301‐309
[31] M. Calle, Y. M. Lee, Thermally rearranged (TR) poly(ether‐benzoxazole) membranes for gas separation, Macromol. 44 (2011) 1156‐1165
[32] R.T. Adams, J. S. Lee, T.‐H. Bae, J. K. Ward, J. R. Johnson, C. W. Jones, S. Nair, W. J. Koros, CO2–CH4 permeation in high zeolite 4A loading mixed matrix membranes, J. Mem. Sci. 367 (2011) 197–203.
[33] R Adams, C Carson, J Ward, R Tannenbaum, W Koros, Metal organic framework mixed matrix membranes for gas separations, Micr. Mes. Mat. 131 (1) (2010) 13‐20
[34] L.M. Robeson, A revised upper bound, J. Mem. Sci. 320 (2008) 390–400
[35] B. T. Lowa, L. Zhao, T. C. Merkel, M. Weber, D. Stolten, A parametric study of the impact of membrane materials and process operating conditions on carbon capture from humidified flue gas, J. Mem. Sci. 431 (2013) 139–155
[36] http://www.iea.org/
APPENDIX [6]
A dimensionless 1D mathematical model for the multi‐species steady‐state permeation in no sweep
mode and co‐current configuration was used for the calculations. In the case of binary mixtures (CO2‐
N2), the model consists of a system of two ordinary differentials (for the retentate side) and two
algebraic (for the permeate side) equations (1‐4)
Feed/Retentate side
( )2
2 2 2
RetentateCO Retentate Permeate
CO CO CO
d 1 Θ x xdζ
φϕ
ϕ= − − (1)
( )2 2
2 2 2
2 2 2
Retentate FeedN CO Retentate Permeate
CO N NFeedN CO /N
d x 1 1 Θ x xdζ x
φϕ
α ϕ= − − (2)
Permeate side
32
( ) ( )
2 2 2
Permeate Feed RetentateCO CO COζ ζφ φ φ= − (3)
( ) ( )2 2 2
Permeate Feed RetentateN N Nζ ζφ φ φ= −
(4)
In the equations, ϕCO2 and ϕN2 are the dimensionless molar flow rates for CO2 and N2, respectively, and
ζ is the dimensionless module length.
ii
i
QQFeedφ =
(5)
Lzζ =
(6)
Θi and φ are parameters affecting the performance of a one‐stage membrane system, the permeation
number and the feed to permeate pressures ratio, respectively.
2
2
2
Membrane FeedCO
CO Feed FeedCO
Permeance A PΘ
x Q=
(7)
Permeate
Feed
PP
=φ (8)