- Email: [email protected]
Membrane selective exchange process for dilute methane recovery
Journal of Membrane Science 469 (2014) 11–18
Contents lists available at ScienceDirect
Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci
Membrane selective exchange process for dilute methane recovery Haiqing Lin n, Ramin Daniels, Scott M. Thompson, Karl D. Amo, Zhenjie He, Timothy C. Merkel, J.G. Wijmans Membrane Technology & Research, Inc., 39630 Eureka Drive, Newark, CA 94560, USA
art ic l e i nf o
a b s t r a c t
Article history: Received 18 April 2014 Received in revised form 13 June 2014 Accepted 13 June 2014 Available online 20 June 2014
Methane gas is a valuable energy resource. Methane emissions are the second largest human-caused contributor to global warming. Significant quantities of the methane emissions are vented as carbon dioxide/methane mixtures from landfills and natural gas processing plants, since they are too dilute to burn or flare. This paper describes a novel membrane selective exchange process to upgrade dilute methane emission streams into useful fuel by removing CO2 and introducing oxygen into the methane product stream. This is achieved in a single step without the need for compression. Flat sheet thin film composite membranes based on perfluoropolymers and a spiral-wound module were prepared and tested with model dilute methane mixtures. A technical and economic analysis shows that the selective exchange process is cost-effective, improves methane utilization and reduces global warming emissions. & 2014 Elsevier B.V. All rights reserved.
Keywords: CO2/CH4 separation Dilute methane Membrane exchange process
1. Introduction Methane emissions are the second largest contributor to global warming, after carbon dioxide (CO2), accounting for 10% of U.S. greenhouse gas emissions [1,2]. Due to increasing public awareness during the past two decades, various technologies and process improvements have been developed to curb methane emissions. However, dilute methane emissions containing 5–40 vol% methane are problematic because they do not burn [3]. Such streams often contain CO2; for example, 1. older, closed landfill sites have methane emission streams with methane contents as low as 20%, with the remainder of the stream being comprised of CO2 and nitrogen [4]; 2. carbon dioxide removal units (amine absorption or membrane systems) in natural gas processing systems have vent streams containing 5-20% methane, with the balance composed of CO2 [5,6]. Streams containing less than 40% methane have an energy content of less than 400 Btu/scf, which is too low for the streams to be burned or flared economically [7]. As a result, they are often vented, increasing global warming. The current unrecovered methane emissions from these sources are estimated at 1.0 million n Corresponding author. Present address: Department of Chemical and Biological Engineering, University at Buffalo, State University of New York, Buffalo, NY 14260, USA. Tel.: þ1 716 6451856. E-mail address: [email protected] (H. Lin).
http://dx.doi.org/10.1016/j.memsci.2014.06.025 0376-7388/& 2014 Elsevier B.V. All rights reserved.
metric tons or 50 billion scf (standard cubic feet) per year in the U. S. alone [4,8]. With 7.6 times the global warming effect of an equivalent volume of CO2, these methane emissions are equivalent to emissions of 21 Tg CO2 per year. For comparison, a 600 MWe coal-fired power plant emits about 4.0 Tg CO2 per year [9]. Therefore, if the methane emissions from the above sources can be recovered and used, it would be equivalent to capturing and sequestering CO2 from five 600 MWe coal-fired power plants. Table 1 compares treatment options and their global warming potential (GWP) for a typical dilute methane stream containing 20% methane. GWP is a measure of the heat that a gas traps over a certain period of time (usually 100 years), compared to CO2 [2]. The larger the GWP value, the more warming the gas causes. CO2 and methane have GWP values of 1 and 7.6, respectively. If the stream is simply vented without treatment, the methane content is responsible for 66% of the GWP of the whole stream, even though the methane content is only 20%. The dilute methane streams can be mixed with supplemental fuel to bring them to 500 Btu/scf so they can be flared, which reduces the GWP by 31% (see Table 1). However, this approach requires additional methane, adding significantly to the cost. The best solution would be to burn the stream as fuel, without adding any additional methane. If this is possible, the GWP could be reduced by 66% and the cost of separation could be compensated for or even covered by the value of the fuel produced. Conventional CO2 separation technologies (such as amine absorption, pressure swing adsorption and membrane technology) cannot recover methane economically from these dilute streams [10]. The CO2 content (60–90%) of the gas is too high for economic
12
H. Lin et al. / Journal of Membrane Science 469 (2014) 11–18
Table 1 Options for dealing with 1 volume of dilute methane emission containing 20% methane and 80% CO2, and the resulting global warming potential (GWP). Options
Emission composition
Volume
GWP of emission
GWP reduction
Vent Flare requiring additional 0.6 volume of CH4 Upgrade to fuel without additional CH4
20% CH4/80% CO2 100% CO2 100% CO2
1 1.6 0.8b
2.32a 1.6 0.8b
– 31% 66%
a b
GWP for untreated emission vent is calculated as 1 0.2 7.6 þ1 0.8 1.0 ¼ 2.32. Credit is taken for the 0.2 volume of methane usefully combusted.
Fig. 1. Ternary composition diagram showing the flammable region and the preferred combustion range for methane/oxygen/inert gas mixtures [12,13]. Methane/inert gas mixtures with more than 70% methane can be used directly as fuel. Dilute methane streams with less than 70% methane cannot be easily burned with ordinary air and need methane upgrading to be used as fuel [7].
use of amine absorption or pressure swing adsorption, because the cost of the sorption system is usually proportional to the amount of CO2 to be removed [6,10]. The low feed pressures (less than 50 psig) mean that conventional membrane systems are not an attractive option either, because large membrane areas will be required [11]. In this paper, we describe a novel membrane selective exchange process to upgrade the quality of the dilute methane/ CO2 streams. Using air as a sweep on the permeate side, the membrane removes CO2 from the dilute methane stream, while introducing oxygen to the upgraded methane stream. The upgraded methane gas is thus combustible when mixed with ordinary air, and can be used as fuel for an engine, heater or boiler. The process reduces global warming emissions, and converts what was previously a waste stream into useful fuel in a potentially economical manner.
2. Background 2.1. Combustion of methane/inert mixtures Combustion of methane-containing gas mixtures can be described using a ternary diagram of the type shown in Fig. 1 [3,12,13]. In this figure, the three components are methane, oxygen, and an inert gas, where the inert gas in this case is the summed concentrations of nitrogen and CO2. The apexes of the
triangle represent pure methane, pure oxygen and pure inert gas. Binary mixtures are represented by positions along the sides of the triangle. Mixtures of all three components are represented by points within the triangle; each mixture is represented by a single point. The diagram is divided into two regions – the flammable region, where combustion is possible, and the non-flammable region, where it is not. The flammable region has a low content of inerts, and balanced concentrations of oxygen and methane. A small preferred combustion region within the flammable region is also shown. The preferred region is away from the edges of the flammable region, to ensure that minor concentration fluctuations do not take the mixture outside the flammable region and stall combustion. The preferred region is also relatively lean in oxygen (below 28%) to avoid excessive combustion temperatures, which lead to NOx formation and create materials‐of‐construction issues [14,15]. In a ternary diagram, all gas compositions that can be obtained by combining two gas mixtures of different compositions are represented by a straight line connecting the two gas mixture compositions. Following this rule, as shown in Fig. 1, any binary methane/inert gas mixture containing more than 70% methane can be mixed with ordinary air to form a mixture that is within the preferred combustion range. The lines connecting binary mixtures containing less than 70% methane-in-inerts and air do not pass comfortably through the preferred combustion region. Therefore, these mixtures cannot be easily burned in air. Moreover, the presence of high content inerts often decreases the
H. Lin et al. / Journal of Membrane Science 469 (2014) 11–18
combustion efficiency [15,16]. To bring these methane/inert mixtures into the combustion range, the gas needs to be upgraded to increase its methane content and/or oxygen enriched air (containing more than 21% oxygen) needs to be used for combustion. The membrane selective exchange process described in this paper effectively integrates both approaches to convert methane/inert vent gas mixtures with low methane concentrations into compositions that can be directly burned with regular air. 2.2. Equivalent methane concentration (EMC)
yCH4 yO2 þ ¼1 70% 21%
ð1Þ
where yCH4 is the methane concentration (%) and yO2 is the oxygen concentration (%) of the ternary mixture. In a more general sense, the equivalent methane concentration (EMC) of any ternary mixture can be calculated from yCH4 1 ðyO2 =21%Þ
ð2Þ
For example, Eq. (2) indicates that a mixture containing 50% methane is as easily combusted as a 70% methane mixture, if the 50% mixture contains 6% oxygen. However, the energy content of the 50% methane mixture (500 Btu/scf) is lower than the energy content of the 70% methane mixture (700 Btu/scf). 2.3. Membrane permeation The gas flux, NA [cm3(STP)/s], through a membrane is expressed as follows [17]: NA ¼
PA ΔpA Am l
ð3Þ
where PA [cm3(STP) cm/cm2 s cmHg] is the permeability coefficient of gas component A, l is the thickness of the membrane selective layer (cm), ΔpA is the partial pressure difference across the membrane (cmHg), and Am is the membrane area (cm2). Because the thickness of the selective layer often is not known accurately, thin film composite membranes are usually characterized by the gas permeance, PA/l, which is expressed in gas permeation units (gpu), where 1 gpu¼10 6 cm3(STP)/cm2 s cmHg¼7.5 10 12 m3(STP)/m2 s Pa. Based on the solution-diffusion mechanism governing gas transport in nonporous polymers, permeability of gas A can also be written as [17] P A ¼ DA S A
ð4Þ 2
Table 2 Physical properties of relevant gas molecules [19,20]. Gas
Kinetic diameter (Å)
Critical temperature (K)
CH4 CO2 N2 O2
3.8 3.3 3.64 3.46
190 304 126 154
αA=B , defined as the ratio of their permeances [17]
The ternary diagram presented in Fig. 1 shows that a 70% methane-in-inerts mixture can be combusted when mixed with regular air. However, mixtures with less than 70% methane also can be combusted with air if the methane mixture contains a sufficient amount of oxygen. All compositions that lie on the line connecting the 70% methane/30% inert point with the air point (21% oxygen/79% nitrogen) have a methane concentration below 70%, but can be combusted just as easily as the 70% methane/30% inert mixture. These compositions are defined as having an equivalent methane concentration (EMC) of 70% and are characterized by the following equation which follows from the mass balance:
EMC ¼
13
3
where DA (cm /s) is the gas diffusivity and SA [cm (STP)/ cm3 cmHg] the gas solubility. Diffusion coefficients increase with decreasing penetrant size, and solubility coefficients increase with penetrant condensability [18]. A measure of a membrane's ability to separate two components, A and B, is the membrane selectivity of gas A over gas B,
αA=B ¼
P A =l SA DA ¼ P B =l SB DB
ð5Þ
Diffusivity selectivity, DA/DB, is influenced by the relative molecular sizes, while solubility selectivity, SA/SB, is determined by the relative condensabilities of the penetrants [18]. Table 2 shows physical properties of the relevant gases [19,20]. A CO2 molecule is smaller than a methane molecule, and therefore diffusivity selectivity always favors CO2 over methane. CO2 is also more condensable than methane (as indicated by higher critical temperature for CO2), and therefore solubility selectivity favors CO2 over methane as well. Consequently, polymeric membranes are always selective for CO2 over methane to varying degrees [21]. The oxygen/methane separation is complicated by the favorable diffusivity selectivity and unfavorable solubility selectivity, since an oxygen molecule is smaller than methane and less condensable than methane (as shown in Table 2). Therefore, membranes with high oxygen/methane selectivity should have a strong size-sieving ability, to maximize the favorable diffusivity selectivity. 2.4. Selective exchange process Fig. 2 shows an example of a membrane selective exchange process to recover methane from waste gas streams [3]. The process simulation is performed using a commercial process simulator (ChemCad 6.3, ChemStations, Inc., Houston, TX), to which we have added a code for a countercurrent membrane process. The membrane simulation uses finite differential element analysis by separating the membrane process into many small steps in series. The gas stream conditions are computed in each membrane step using a thermodynamic package with the Soave– Redlich–Kwong equation of state. Similar methods for membrane simulations have been described elsewhere [22–25]. In Fig. 2, a waste methane stream feed (stream ①) contains 20% methane and 80% CO2; this is the composition of a landfill gas stream at a later stage of gas recovery operation, or a typical gas stream composition for the off-gas from an acid gas removal unit in a natural gas processing plant. The gas stream is introduced on the feed side of the CO2-permeable membrane. Air at atmospheric pressure (③) is circulated countercurrently on the permeate side of the membrane using a blower. Because of the partial pressure differences across the membrane, CO2 permeates into the air sweep and oxygen from the air permeates into the methane stream. The result is a fuel gas product stream (②) with a methane concentration of 53.6%, to which 4.9% oxygen has been added. The resulting stream has an EMC equal to 70%. The methane residue stream can be easily used as engine, turbine or boiler fuel, even though the gas contains only 53.6% methane with a Btu value of 530 Btu/scf. This is because the gas also contains 4.9% oxygen and thus requires less additional air to be mixed with the gas to bring the oxygen concentration to the level of approximately 18% required for good combustion. The diluting effect of nitrogen in the air is thus reduced. The phase diagram positions of the low methane feed gas (①), the methane concentrated gas (②), and the methane concentrated
14
H. Lin et al. / Journal of Membrane Science 469 (2014) 11–18
Fig. 2. Process simulation of a membrane selective exchanger used to upgrade a dilute methane stream into a more concentrated methane stream that can be directly used as fuel. The process recovers 80% methane. The process simulation assumes membrane permeances of 320 gpu for CO2, 200 gpu for O2, 64 gpu for N2 and 32 gpu for CH4.
Fig. 3. (a) Phase diagram showing how a membrane contactor upgrades low methane feed gas to produce a fuel gas that readily burns when mixed with air. Streams ① and ② refer to the feed and product stream compositions in Fig. 2, respectively; stream ➄ refers to the combustion mixture. (b) Simplified schematic of the process shown in Fig. 2.
3. Experimental
minimal resistance to gas flow. The paper and microporous layers provide mechanical strength to the membrane composite structure. The smooth surface of the gutter layer allows a very thin, dense selective layer of perfluoropolymers to be deposited [27,28]. The flat sheet thin film composite membranes were fabricated into four-port spiral‐wound modules that operate in countercurrent/ sweep mode [15,26,29–32]. The feed gas passes down the module in the channel created by the feed spacer. A portion of the feed gas permeates the membrane and then enters the permeate channel. The permeate channel can be swept with a sweep gas, allowing the module to operate in a countercurrent mode [26,29]. More details on the module structure have been shown in our earlier work [15,26] and in the work of others [29–32].
3.1. Membrane and module preparation
3.2. Sweep experiments
Flat sheet thin film composite membranes for gas separation were prepared in this study [5,15]. The separation properties are governed by the thin selective layer, and the other three layers (paper, microporous and gutter layers) are designed to have
Composite membranes and membrane modules were tested under countercurrent/sweep conditions using a system shown in Fig. 4. The CO2/CH4 mixture flowed on the feed side of the membrane, and low pressure air flowed on the permeate side.
gas after mixing with an appropriate amount of air (⑤) are shown in Fig. 3. The concentrated methane gas is still comfortably above the upper flammability limit, but when mixed with air it can be brought into the preferred combustion range and used as fuel. The sweep air leaving the contactor (stream ④ in Fig. 2) contains 26% CO2 and only 1.4% methane and can be safely discharged to the atmosphere without further treatment. The use of air sweep in the membrane exchanger dilutes the selectively permeated CO2 and increases the CO2 pressure driving force across the membrane, thus improving CO2 removal efficiency [5,11,26].
H. Lin et al. / Journal of Membrane Science 469 (2014) 11–18
15
Table 3 Range of pure-gas permeances of composite membranes at 50 psig and 20 1C.
Fig. 4. Schematic of countercurrent/sweep test system for flat sheet composite membranes and spiral‐wound modules. A rectangular countercurrent test cell (with a membrane area of 150 cm2) was used for membrane testing, and one module housing was used to test countercurrent/sweep spiral‐wound modules (with a membrane area of 0.5–1 m2). MFM: mass flow meter. GC: gas chromatography unit.
The test system was equipped with an Agilent MicroGC 3000 portable gas chromatography (GC) unit, to monitor the gas composition of all the streams (including the feed, residue, sweep and ‘sweep þpermeate’ stream). The flow rates were measured using mass flow meters (Sierra Instruments, Monterey, CA). The experimental data were only considered valid when the mass of each component entering and leaving the permeation cell or module housing was within 715%. The mixed-gas permeance in countercurrent/sweep operation can be estimated using the following equation [26, 33]: P A N A lnðΔpA;L =ΔpA;O Þ ¼ Am ðΔpA;L =ΔpA;O Þ l
ð6Þ
ΔpA;O ¼ pA;F pA;S þ P and ΔpA;L ¼ pA;R pA;S
ð7Þ
where pA;F , pA;R , pA;S and pA;S þ P are the partial pressures of component A in the feed, residue, sweep and ‘sweepþ permeate’ stream, respectively. NA is the flow rate difference of component A between the sweep and ‘sweepþpermeate’ stream. Eq. (6) is similar to Eq. (3), except that the partial pressure difference is expressed as the logarithmic mean of ΔpA;L and ΔpA;O . The composite membrane sheets were mounted in a modified Osmonics SEPAs CF II Med/High Foulant System cell (GE Osmonics Labstore, Minnetonka, MN). The test cell was modified to allow a near-perfect countercurrent flow test [34]. The countercurrent/ sweep spiral‐wound modules were installed in a custom-built module housing, with separate ports to accommodate the four gas streams. Chemical-grade nitrogen, oxygen and air, each with a purity of 99%, were used as received from Praxair Inc. (Hayward, CA). An air compressor was also used to provide high-pressure air, the composition of which was determined using the GC.
4. Results and discussion 4.1. Composite membrane preparation and characterization Thin film composite membranes comprised of a perfluoropolymer as the selective layer were prepared. Table 3 shows typical pure-gas permeance and selectivity at feed pressure of 50 psig and 20 1C. The gas selectivity values are similar to those of dense films, indicating that the composite films are defect-free. The membranes show good CO2/CH4 selectivity to selectively remove CO2 from the methane stream, and superior oxygen/methane selectivity to selectively introduce oxygen into the methane product stream. The high CO2/CH4 and O2/CH4 selectivities are ascribed to the unfavorable interactions between perfluoropolymers and
Gas
Pure-gas permeance (gpu)a
Gas/CH4 selectivity
CH4 CO2 N2 O2
10–30 240–720 25–75 75–225
– 24 2.5 7.5
a
1 gpu ¼10 6 cm3(STP)/cm2 s cmHg ¼7.5 10 12 m3(STP)/m2 s Pa.
hydrocarbons such as CH4, leading to high solubility selectivity of CO2/CH4 and O2/CH4 [7,35].
4.2. Evidence of sweep effectiveness The separation performance of a membrane sheet was tested in a countercurrent test cell (150 cm2 area) with and without application of a permeate side sweep. The feed contained 47% methane in CO2 and was fed into the cell at 50 psig and a flow rate of 16 standard liters per minute per square meter of membrane area (slpm/m2). The permeate side was kept at three pressure values (0 psig, 20 psig and 40 psig), and swept with air at various flow rates. The selective layer of the membrane faced the highpressure feed gas. Fig. 5(a) shows that the CO2 permeation rate can be tripled by sweeping the permeate side of the membrane with air. The air sweep reduces the partial pressure of CO2 directly under the selective layer, resulting in an increased driving force for permeation. The enhancement of the CO2 removal is more pronounced at higher permeate pressure. For example, as the air sweep flow rate increases from 0 to 8 slpm/m2, the increase in CO2 removal is 13%, 110% and 250% for the tests at sweep pressures of 0 psig, 20 psig and 40 psig, respectively. This behavior is expected and also has been observed in natural gas dehydration field tests [20]. Simultaneously, as shown in Fig. 5(b), oxygen permeates from the air sweep stream to the CO2/methane stream, in spite of the net permeation flow rate flowing in the opposite direction. A higher sweep pressure means a higher driving force for oxygen to permeate across the membrane, increasing oxygen flux. The data presented in Fig. 5 show that the use of an air sweep significantly increases the CO2 removal rate, thereby increasing the actual methane concentration, at the same time that it introduces oxygen to the methane product stream, which increases the equivalent methane concentration beyond the actual methane concentration. Fig. 6 summarizes these results for a larger set of experiments. The results are shown as the equivalent methane concentration (EMC) of the product gas versus the methane lost to the air sweep exiting the system (methane loss is defined as the fraction of the methane in the original feed stream that is lost with the air sweep exiting the system). As shown in Fig. 6, the use of the air sweep increases the EMC, while the loss of methane is significantly reduced. The oxygen concentration in the methane stream in these experiments ranges from 1% to 5%, which means that oxygen permeation is responsible for up to 30% of the increase in the EMC. The intrinsic pure-gas CO2 permeance of the membrane used in these experiments is 240 gpu. The effective permeance observed in the countercurrent/sweep experiments is in the range of 80 to 200 gpu. This reduction in permeance is presumably caused by concentration polarization on the permeate side (such as in the support layers of the membrane) [5,36,37].
16
H. Lin et al. / Journal of Membrane Science 469 (2014) 11–18
Sweep : feed flow ratio 1.0
Sweep : feed flow ratio
2.0
3.0
0
1.0
2.0
3.0
1.5
Sweep pressure:
8
O2 permeation into 2 methane stream (slpm/m )
CO2 removal from 2 methane stream (slpm/m )
10
0
0 psig 20 psig 40 psig
6
4
2
1.2
Sweep pressure: 40 psig
0.9 20 psig 0.6
0.3 0 psig
0
0
10
20
30
40
50
0
0
10
2
20
30
40
50
2
Air sweep flow rate (slpm/m )
Air sweep flow rate (slpm/m )
Fig. 5. Amount of (a) CO2 and (b) oxygen permeating through the perfluoropolymer-based membrane as a function of air sweep flow rate at various air sweep pressures (0 psig, 20 psig and 40 psig). The feed contains 47% CH4 at about 16 slpm/m2 and 50 psig.
100
With air sweep
90
EMC of methane product (%)
Equivalent methane concentration (%)
100
80 70 No air sweep 60 50
Module with sweep
80
60
40
Module without sweep 20
No separation
40
0
0
5
10 15 Methane loss (%)
20
25
0
5
10
15
20
25
Feed methane concentration (%)
Fig. 6. Equivalent methane concentration (EMC, defined in Eq. (2)) versus methane loss for a set of experiments using perfluoropolymer-based composite membranes. The feed gas contains 47% methane at 50 psig and feed flow rates are in the range of 5.7–31 slpm/m2. The air sweep is at pressures of 0–40 psig and sweep flow rates of 0–34 slpm/m2. The process recovers 80–95% of the methane that is otherwise lost as a dilute methane vent stream.
Fig. 7. Equivalent methane concentration (EMC) achieved by the countercurrent/ sweep test module, as a function of the feed methane concentration. The module operates significantly more efficiently in sweep mode than without sweep.
4.3. Performance of a countercurrent/sweep module
4.4. Economic analysis
A spiral‐wound countercurrent/sweep module was fabricated using the same membrane tested in the permeation cell. The module has one membrane envelope with an effective area of 0.6 m2, about 40 times the area of the permeation cell. The module has a pure gas CO2 permeance of 200 gpu and CO2/CH4 selectivity of 24, suggesting that the module is defect-free. The module was tested with a CO2/methane feed stream at 50 psig and 15 slpm/m2, and an air sweep stream at 0 psig and 9.0 slpm/m2. Fig. 7 shows the EMC values achieved by the module in a single pass, as a function of the methane concentration in the feed stream. The module effectively upgrades the dilute methane streams in the sweep mode because the module both removes more CO2 than the
To explore the economic feasibility of the membrane-based process for utilization of dilute methane streams, we performed process simulations for streams with methane concentrations ranging from 20% to 60%. The goal was to create a methane product stream with an EMC value of 70%, so the stream could be easily used as fuel. Table 4 lists the membrane areas required, the methane recoveries achieved, the value of the recovered methane, and the capital cost of the membrane system to recover a combustible product at five different dilute methane starting compositions. Air sweep flow rate was 1 MMscfd. Assumed membrane permeances were 320 gpu for CO2, 200 gpu for O2, 64 gpu for N2 and 32 gpu for CH4, higher than those achieved during the membrane and module
operation without sweep and introduces oxygen into the produced fuel stream.
H. Lin et al. / Journal of Membrane Science 469 (2014) 11–18
Table 4 Process parameters and capital costs for a membrane selective exchange unit treating 1 MMscfd of dilute methane gas at various methane concentrations and producing a fuel product stream with 70% equivalent methane concentration. Feed methane content (%)
Membrane area (m2)
Methane recovery (%)
Fuel product value ($1000/ year)a
Membrane system capital cost ($1000)b
60 50 40 30 20
130 225 295 345 380
95.8 92.1 88.5 84.8 80.4
840 670 520 370 230
130 225 295 345 380
a b
Based on fuel value of $4/1000 scf methane. Based on $1000 per square meter membrane area.
17
in CO2 partial pressure difference across the membrane and the increase in the CO2 mass transfer coefficient in the boundary layer on the permeate side. The process is cost-effective and reduces methane emissions to the atmosphere. For example, the membrane exchanger can upgrade a dilute methane stream containing as little as 20% methane, to a fuel quality methane stream with equivalent methane content of 70%. The value of the methane recovery is $230,000 per year (assuming $4/1000 scf methane), while the operating cost of the membrane exchange system is $133,000 per year. The economics of the membrane selective exchanger become better if the waste streams have higher methane content and the credit for the avoidance of methane emissions is considered.
Acknowledgments Table 5 Estimate of the operating cost for a membrane system converting 1 MMscfd of 20% methane gas into useful fuel with an EMC value of 70%. Cost category Installed system cost Depreciation Annual module replacement costs Total annual operating cost
Unit cost
Annual value of recovered methane
$4/1000 scf CH4
20% of capital cost $150/m2
Value ($1000s) $380 $76 $57 $133 $230
experiments, but very realistic based on the most recent membranes produced at Membrane Technology & Research, Inc. The cost of the membrane system installed in the field is estimated to be $1000/m2 membrane area, which is realistic for the relatively small flow rates considered here. This cost is based on our experience in building commercial membrane systems for the natural gas industry. As expected, the capital cost of the membrane system decreases with increasing methane content in the dilute methane streams. Table 5 gives a breakdown of the capital and operating costs for the 20% methane case and shows how promising the process economics are for this least attractive case. The membrane system is assumed to be depreciated over a five-year period. The membrane modules are assumed to have a three-year operating lifetime, and the module replacement cost is estimated at $450/m2 membrane area. Therefore, the annual cost of membrane modules is $150/m2 membrane area. As shown in Table 5, the operating cost of the membrane system ($133,000 per year) is easily offset by the methane value recovered ($230,000 per year assuming a value of $4/1000 scf methane). The economics of the membrane selective exchanger are even more attractive, if the dilute methane streams have higher methane content and the credit for the avoidance of methane emissions is considered (as shown in Table 1).
5. Conclusions This paper describes a membrane selective exchange process to upgrade dilute methane streams (containing inerts such as CO2) to fuel grade products. The membrane uses an air sweep on the permeate side, which improves the CO2 removal from the dilute methane stream and introduces oxygen to the methane product, thus improving the quality of the methane stream as a fuel. Laboratory experiments with membrane stamps and a prototype membrane module have demonstrated the feasibility of the selective exchange process that generates useful fuel streams from dilute methane vent streams. The efficiency improves with increasing air sweep flow rate, presumably due to the increase
We gratefully acknowledge the partial financial support of this work by the U.S. Department of Energy Small Business Innovation Program (SBIR) Phase I, under Contract number DE-FG02-06ER84609, and the U.S. Environmental Protection Agency SBIR Phase I under Contract number EP-D-10-036 and Phase II under EP-D-11071. One of the authors, H. Lin, also acknowledges the funds provided by the University at Buffalo, State University of New York, during the preparation of the manuscript.
Nomenclature membrane area (cm2) British thermal unit diffusivity of gas A (cm2/s) diffusivity selectivity of gas A over gas B equivalent methane content, defined in Eq. (2) gas permeation unit [10 6 cm3(STP)/cm2 s cmHg] thickness of the membrane selective layer (cm) million British thermal units steady state flux of gas A across the membrane [cm3(STP)/s] PA permeability of membrane selective layer to gas A (Barrer) PA/l permeance of composite membrane to gas A (gpu) Δp A partial pressure difference across the membrane for gas A (cmHg) ΔpA;0 partial pressure difference across the membrane for gas A between feed and ‘sweepþpermeate’ streams (cmHg) ΔpA;L partial pressure difference across the membrane for gas A between residue and sweep streams (cmHg) SA solubility of gas A in a polymer [cm3(STP)/(cm3 polymer atm)] SA/SB solubility selectivity of gas A over gas B scf standard cubic foot slpm/m2 unit of gas flow rate, standard litter per minute per m2 membrane area T temperature (K) yCH4 methane concentration in the ternary mixture (vol %) yO2 oxygen concentration in the ternary mixture (vol%) Am Btu DA DA/DB EMC gpu l MMBtu NA
Greek letter
αA=B
permeability or permeance selectivity of gas A over gas B
18
H. Lin et al. / Journal of Membrane Science 469 (2014) 11–18
Subscript A B F R S S þP
gas component A gas component B feed stream residue stream sweep stream the exiting stream in the permeate including the permeate and sweep gas
References [1] Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990–2006, U.S. Environmental Protection Agency, 2008. [2] I. Karakurt, G. Aydin, K. Aydiner, Mine ventilation air methane as a sustainable energy source, Renew. Sustain. Energy Rev. 15 (2011) 1042–1049. [3] J.G. Wijmans, T.C. Merkel, H. Lin, S.M. Thompson, R. Daniels, Efficient gasseparation process to upgrade dilute methane stream for use as fuel, U.S. Patent 8124733, 2012. [4] N.P. Cheremisinoff, Handbook of Solid Waste Management and Waste Minimization Technologies, Butterworth-Heinemann, Burlington, MA (2003). [5] R.W. Baker, Membrane Technology and Applications, 3rd ed., John Wiley and Sons Ltd., Chichester, UK, 2012. [6] R.W. Baker, K. Lokhandwala, Natural gas processing with membranes: an overview, Ind. Eng. Chem. Res. 47 (2008) 2109–2121. [7] T.C. Merkel, I. Pinnau, R.S. Prabhakar, B.D. Freeman, Gas and vapor transport properties of perfluoropolymers, in: B.D. Freeman, Y.P. Yampolskii, I. Pinnau (Eds.), Materials Science of Membranes for Gas and Vapor Separation, John Wiley & Sons Ltd, Chichester, UK, 2006, pp. 251–270. [8] EPA's Landfill Methane Outreach Program, 〈http://www.epa.gov/lmop/〉, 2014. [9] Cost and Performance Baseline for Fossil Energy Plants (V 1): Bituminous Coal and Natural Gas to Electricity, Report No. DOE/NETL-2010-1397, September 2013. [10] S. Su, A. Beath, H. Guo, C. Mallett, An assessment of mine methane mitigation and utilisation technologies, Prog. Energy Combust. 31 (2005) 123–170. [11] T.C. Merkel, H. Lin, X. Wei, R.W. Baker, Power plant post-combustion carbon dioxide capture: an opportunity for membranes, J. Membr. Sci. 359 (2010) 126–139. [12] H.F. Coward, G.W. Jones, Limits of Flammability of Gases and Vapors, U.S. Bureau of Mines, 1952. [13] C.V. Mashuga, D.A. Crowl, Application of the flammability diagram for evaluation of fire and explosion hazards of flammable vapors, Process Saf. Prog. 17 (1998) 176–183. [14] C.E. Baukal, Oxygen-Enhanced Combustion, CRC Press, Boca Raton, 1998. [15] H. Lin, M. Zhou, J. Ly, J. Vu, J.G. Wijmans, T.C. Merkel, J. Jin, A. Haldeman, E. H. Wagener, D. Rue, Membrane-based oxygen-enriched combustion, Ind. Eng. Chem. Res. 52 (2013) 10820–10834. [16] S.G. Kimura, W.R. Browall, Membrane oxygen enrichment I: Demonstration of membrane oxygen enrichment for natural gas combustion, J. Membr. Sci. 29 (1986) 69–77.
[17] J.G. Wijmans, R.W. Baker, The solution-diffusion model: a review, J. Membr. Sci. 107 (1995) 1–21. [18] B.D. Freeman, Basis of permeability/selectivity tradeoff relations in polymeric gas separation membranes, Macromolecules 32 (1999) 375–380. [19] R.C. Reid, J.M. Prausnitz, B.E. Poling, The Properties of Gases and Liquids, McGraw-Hill, New York, 1987. [20] D.W. Breck, Zeolite Molecular Sieves, Krieger Publishing Company, Florida, 1974. [21] D.E. Sanders, Z.P. Smith, R.L. Guo, L.M. Robeson, J.E. McGrath, D.R. Paul, B. D. Freeman, Energy-efficient polymeric gas separation membranes for a sustainable future: a review, Polymer 54 (2013) 4729–4761. [22] C.Y. Pan, H.W. Hapgood, Gas separation by permeation, Can. J. Chem. Eng. 56 (1978) 197. [23] D.T. Coker, T. Allen, B.D. Freeman, G.K. Fleming, Nonisothermal model for gas separation hollow-fiber membranes, AIChE J. 45 (1999) 1451–1468. [24] B. Belaissaoui, Y.L. Moullec, D. Willson, E. Favre, Hybrid membrane cryogenic process for post-combustion CO2 capture, J. Membr. Sci. 415-416 (2012) 424–434. [25] S. Kelman, H. Lin, E.S. Sanders, B.D. Freeman, CO2/C2H6 separation using solubility selective membranes, J. Membr. Sci. 305 (2007) 57–68. [26] H. Lin, S.M. Thompson, A. Serbanescu-Martin, J.G. Wijmans, K.D. Amo, K. A. Lokhandwala, B. Low, T.C. Merkel, Dehydration of natural gas using membranes. Part II: Sweep/countercurrent design and field test, J. Membr. Sci. 432 (2013) 106–114. [27] K.A. Lundy, I. Cabasso, Analysis and construction of multilayer composite membranes for the separation of gas mixtures, Ind. Eng. Chem. Res. 28 (1989) 742–756. [28] W. Yave, A. Car, J. Wind, K. Peinemann, Nanometric thin film membranes manufactured on square meter scale: ultra-thin films for CO2 capture, Nanotechnology 21 (2010) 1–7. [29] D. Reddy, T.Y. Moon, C.E. Reineke, Counter current dual-flow spiral wound dual-pipe membrane separation, U.S. Patent 5,034,126, 1991. [30] T.Y. Cath, A.E. Childress, M. Elimelech, Forward osmosis: principles, applications, and recent developments, J. Membr. Sci. 261 (2006) 70–87. [31] G.D. Mehta, Further results on the performance of present-day osmotic membranes in various osmotic regions, J. Membr. Sci. 10 (1982) 3–19. [32] C. Bayer, M. Follmann, H. Breisig, I.M. Wienk, F.P. Cuperus, M. Wessling, T. Melin, On the design of a 4-end spiral-wound L/L extraction membrane module, Ind. Eng. Chem. Res. 52 (2013) 1004–1014. [33] 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. Membr. Sci. 367 (2011) 233–239. [34] H. Lin, S.M. Thompson, A. Serbanescu-Martin, H.G. Wijmans, K.D. Amo, K. Lokhandwala, T.C. Merkel, Dehydration of natural gas using membranes. Part I: Composite membranes, J. Membr. Sci. 413-414 (2012) 70–81. [35] R.S. Prabhakar, B.D. Freeman, I. Roman, Gas and vapor sorption and permeation in poly(2,2,4-trifluoro-5-trifluoromethoxy-1,3-dioxole-co-tetrafluoroethylene), Macromolecules 37 (2004) 7688–7697. [36] J.G. Wijmans, A.L. Athayde, R. Daniels, J.H. Ly, H.D. Kamaruddin, I. Pinnau, The role of boundary layers in the removal of volatile organic compounds from water by pervaporation, J. Membr. Sci. 109 (1996) 135–146. [37] J.R. McCutcheon, M. Elimelech, Influence of concentrative and dilutive internal concentration polarization on flux behavior in forward osmosis, J. Membr. Sci. 284 (2006) 237–247.
Contents lists available at ScienceDirect
Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci
Membrane selective exchange process for dilute methane recovery Haiqing Lin n, Ramin Daniels, Scott M. Thompson, Karl D. Amo, Zhenjie He, Timothy C. Merkel, J.G. Wijmans Membrane Technology & Research, Inc., 39630 Eureka Drive, Newark, CA 94560, USA
art ic l e i nf o
a b s t r a c t
Article history: Received 18 April 2014 Received in revised form 13 June 2014 Accepted 13 June 2014 Available online 20 June 2014
Methane gas is a valuable energy resource. Methane emissions are the second largest human-caused contributor to global warming. Significant quantities of the methane emissions are vented as carbon dioxide/methane mixtures from landfills and natural gas processing plants, since they are too dilute to burn or flare. This paper describes a novel membrane selective exchange process to upgrade dilute methane emission streams into useful fuel by removing CO2 and introducing oxygen into the methane product stream. This is achieved in a single step without the need for compression. Flat sheet thin film composite membranes based on perfluoropolymers and a spiral-wound module were prepared and tested with model dilute methane mixtures. A technical and economic analysis shows that the selective exchange process is cost-effective, improves methane utilization and reduces global warming emissions. & 2014 Elsevier B.V. All rights reserved.
Keywords: CO2/CH4 separation Dilute methane Membrane exchange process
1. Introduction Methane emissions are the second largest contributor to global warming, after carbon dioxide (CO2), accounting for 10% of U.S. greenhouse gas emissions [1,2]. Due to increasing public awareness during the past two decades, various technologies and process improvements have been developed to curb methane emissions. However, dilute methane emissions containing 5–40 vol% methane are problematic because they do not burn [3]. Such streams often contain CO2; for example, 1. older, closed landfill sites have methane emission streams with methane contents as low as 20%, with the remainder of the stream being comprised of CO2 and nitrogen [4]; 2. carbon dioxide removal units (amine absorption or membrane systems) in natural gas processing systems have vent streams containing 5-20% methane, with the balance composed of CO2 [5,6]. Streams containing less than 40% methane have an energy content of less than 400 Btu/scf, which is too low for the streams to be burned or flared economically [7]. As a result, they are often vented, increasing global warming. The current unrecovered methane emissions from these sources are estimated at 1.0 million n Corresponding author. Present address: Department of Chemical and Biological Engineering, University at Buffalo, State University of New York, Buffalo, NY 14260, USA. Tel.: þ1 716 6451856. E-mail address: [email protected] (H. Lin).
http://dx.doi.org/10.1016/j.memsci.2014.06.025 0376-7388/& 2014 Elsevier B.V. All rights reserved.
metric tons or 50 billion scf (standard cubic feet) per year in the U. S. alone [4,8]. With 7.6 times the global warming effect of an equivalent volume of CO2, these methane emissions are equivalent to emissions of 21 Tg CO2 per year. For comparison, a 600 MWe coal-fired power plant emits about 4.0 Tg CO2 per year [9]. Therefore, if the methane emissions from the above sources can be recovered and used, it would be equivalent to capturing and sequestering CO2 from five 600 MWe coal-fired power plants. Table 1 compares treatment options and their global warming potential (GWP) for a typical dilute methane stream containing 20% methane. GWP is a measure of the heat that a gas traps over a certain period of time (usually 100 years), compared to CO2 [2]. The larger the GWP value, the more warming the gas causes. CO2 and methane have GWP values of 1 and 7.6, respectively. If the stream is simply vented without treatment, the methane content is responsible for 66% of the GWP of the whole stream, even though the methane content is only 20%. The dilute methane streams can be mixed with supplemental fuel to bring them to 500 Btu/scf so they can be flared, which reduces the GWP by 31% (see Table 1). However, this approach requires additional methane, adding significantly to the cost. The best solution would be to burn the stream as fuel, without adding any additional methane. If this is possible, the GWP could be reduced by 66% and the cost of separation could be compensated for or even covered by the value of the fuel produced. Conventional CO2 separation technologies (such as amine absorption, pressure swing adsorption and membrane technology) cannot recover methane economically from these dilute streams [10]. The CO2 content (60–90%) of the gas is too high for economic
12
H. Lin et al. / Journal of Membrane Science 469 (2014) 11–18
Table 1 Options for dealing with 1 volume of dilute methane emission containing 20% methane and 80% CO2, and the resulting global warming potential (GWP). Options
Emission composition
Volume
GWP of emission
GWP reduction
Vent Flare requiring additional 0.6 volume of CH4 Upgrade to fuel without additional CH4
20% CH4/80% CO2 100% CO2 100% CO2
1 1.6 0.8b
2.32a 1.6 0.8b
– 31% 66%
a b
GWP for untreated emission vent is calculated as 1 0.2 7.6 þ1 0.8 1.0 ¼ 2.32. Credit is taken for the 0.2 volume of methane usefully combusted.
Fig. 1. Ternary composition diagram showing the flammable region and the preferred combustion range for methane/oxygen/inert gas mixtures [12,13]. Methane/inert gas mixtures with more than 70% methane can be used directly as fuel. Dilute methane streams with less than 70% methane cannot be easily burned with ordinary air and need methane upgrading to be used as fuel [7].
use of amine absorption or pressure swing adsorption, because the cost of the sorption system is usually proportional to the amount of CO2 to be removed [6,10]. The low feed pressures (less than 50 psig) mean that conventional membrane systems are not an attractive option either, because large membrane areas will be required [11]. In this paper, we describe a novel membrane selective exchange process to upgrade the quality of the dilute methane/ CO2 streams. Using air as a sweep on the permeate side, the membrane removes CO2 from the dilute methane stream, while introducing oxygen to the upgraded methane stream. The upgraded methane gas is thus combustible when mixed with ordinary air, and can be used as fuel for an engine, heater or boiler. The process reduces global warming emissions, and converts what was previously a waste stream into useful fuel in a potentially economical manner.
2. Background 2.1. Combustion of methane/inert mixtures Combustion of methane-containing gas mixtures can be described using a ternary diagram of the type shown in Fig. 1 [3,12,13]. In this figure, the three components are methane, oxygen, and an inert gas, where the inert gas in this case is the summed concentrations of nitrogen and CO2. The apexes of the
triangle represent pure methane, pure oxygen and pure inert gas. Binary mixtures are represented by positions along the sides of the triangle. Mixtures of all three components are represented by points within the triangle; each mixture is represented by a single point. The diagram is divided into two regions – the flammable region, where combustion is possible, and the non-flammable region, where it is not. The flammable region has a low content of inerts, and balanced concentrations of oxygen and methane. A small preferred combustion region within the flammable region is also shown. The preferred region is away from the edges of the flammable region, to ensure that minor concentration fluctuations do not take the mixture outside the flammable region and stall combustion. The preferred region is also relatively lean in oxygen (below 28%) to avoid excessive combustion temperatures, which lead to NOx formation and create materials‐of‐construction issues [14,15]. In a ternary diagram, all gas compositions that can be obtained by combining two gas mixtures of different compositions are represented by a straight line connecting the two gas mixture compositions. Following this rule, as shown in Fig. 1, any binary methane/inert gas mixture containing more than 70% methane can be mixed with ordinary air to form a mixture that is within the preferred combustion range. The lines connecting binary mixtures containing less than 70% methane-in-inerts and air do not pass comfortably through the preferred combustion region. Therefore, these mixtures cannot be easily burned in air. Moreover, the presence of high content inerts often decreases the
H. Lin et al. / Journal of Membrane Science 469 (2014) 11–18
combustion efficiency [15,16]. To bring these methane/inert mixtures into the combustion range, the gas needs to be upgraded to increase its methane content and/or oxygen enriched air (containing more than 21% oxygen) needs to be used for combustion. The membrane selective exchange process described in this paper effectively integrates both approaches to convert methane/inert vent gas mixtures with low methane concentrations into compositions that can be directly burned with regular air. 2.2. Equivalent methane concentration (EMC)
yCH4 yO2 þ ¼1 70% 21%
ð1Þ
where yCH4 is the methane concentration (%) and yO2 is the oxygen concentration (%) of the ternary mixture. In a more general sense, the equivalent methane concentration (EMC) of any ternary mixture can be calculated from yCH4 1 ðyO2 =21%Þ
ð2Þ
For example, Eq. (2) indicates that a mixture containing 50% methane is as easily combusted as a 70% methane mixture, if the 50% mixture contains 6% oxygen. However, the energy content of the 50% methane mixture (500 Btu/scf) is lower than the energy content of the 70% methane mixture (700 Btu/scf). 2.3. Membrane permeation The gas flux, NA [cm3(STP)/s], through a membrane is expressed as follows [17]: NA ¼
PA ΔpA Am l
ð3Þ
where PA [cm3(STP) cm/cm2 s cmHg] is the permeability coefficient of gas component A, l is the thickness of the membrane selective layer (cm), ΔpA is the partial pressure difference across the membrane (cmHg), and Am is the membrane area (cm2). Because the thickness of the selective layer often is not known accurately, thin film composite membranes are usually characterized by the gas permeance, PA/l, which is expressed in gas permeation units (gpu), where 1 gpu¼10 6 cm3(STP)/cm2 s cmHg¼7.5 10 12 m3(STP)/m2 s Pa. Based on the solution-diffusion mechanism governing gas transport in nonporous polymers, permeability of gas A can also be written as [17] P A ¼ DA S A
ð4Þ 2
Table 2 Physical properties of relevant gas molecules [19,20]. Gas
Kinetic diameter (Å)
Critical temperature (K)
CH4 CO2 N2 O2
3.8 3.3 3.64 3.46
190 304 126 154
αA=B , defined as the ratio of their permeances [17]
The ternary diagram presented in Fig. 1 shows that a 70% methane-in-inerts mixture can be combusted when mixed with regular air. However, mixtures with less than 70% methane also can be combusted with air if the methane mixture contains a sufficient amount of oxygen. All compositions that lie on the line connecting the 70% methane/30% inert point with the air point (21% oxygen/79% nitrogen) have a methane concentration below 70%, but can be combusted just as easily as the 70% methane/30% inert mixture. These compositions are defined as having an equivalent methane concentration (EMC) of 70% and are characterized by the following equation which follows from the mass balance:
EMC ¼
13
3
where DA (cm /s) is the gas diffusivity and SA [cm (STP)/ cm3 cmHg] the gas solubility. Diffusion coefficients increase with decreasing penetrant size, and solubility coefficients increase with penetrant condensability [18]. A measure of a membrane's ability to separate two components, A and B, is the membrane selectivity of gas A over gas B,
αA=B ¼
P A =l SA DA ¼ P B =l SB DB
ð5Þ
Diffusivity selectivity, DA/DB, is influenced by the relative molecular sizes, while solubility selectivity, SA/SB, is determined by the relative condensabilities of the penetrants [18]. Table 2 shows physical properties of the relevant gases [19,20]. A CO2 molecule is smaller than a methane molecule, and therefore diffusivity selectivity always favors CO2 over methane. CO2 is also more condensable than methane (as indicated by higher critical temperature for CO2), and therefore solubility selectivity favors CO2 over methane as well. Consequently, polymeric membranes are always selective for CO2 over methane to varying degrees [21]. The oxygen/methane separation is complicated by the favorable diffusivity selectivity and unfavorable solubility selectivity, since an oxygen molecule is smaller than methane and less condensable than methane (as shown in Table 2). Therefore, membranes with high oxygen/methane selectivity should have a strong size-sieving ability, to maximize the favorable diffusivity selectivity. 2.4. Selective exchange process Fig. 2 shows an example of a membrane selective exchange process to recover methane from waste gas streams [3]. The process simulation is performed using a commercial process simulator (ChemCad 6.3, ChemStations, Inc., Houston, TX), to which we have added a code for a countercurrent membrane process. The membrane simulation uses finite differential element analysis by separating the membrane process into many small steps in series. The gas stream conditions are computed in each membrane step using a thermodynamic package with the Soave– Redlich–Kwong equation of state. Similar methods for membrane simulations have been described elsewhere [22–25]. In Fig. 2, a waste methane stream feed (stream ①) contains 20% methane and 80% CO2; this is the composition of a landfill gas stream at a later stage of gas recovery operation, or a typical gas stream composition for the off-gas from an acid gas removal unit in a natural gas processing plant. The gas stream is introduced on the feed side of the CO2-permeable membrane. Air at atmospheric pressure (③) is circulated countercurrently on the permeate side of the membrane using a blower. Because of the partial pressure differences across the membrane, CO2 permeates into the air sweep and oxygen from the air permeates into the methane stream. The result is a fuel gas product stream (②) with a methane concentration of 53.6%, to which 4.9% oxygen has been added. The resulting stream has an EMC equal to 70%. The methane residue stream can be easily used as engine, turbine or boiler fuel, even though the gas contains only 53.6% methane with a Btu value of 530 Btu/scf. This is because the gas also contains 4.9% oxygen and thus requires less additional air to be mixed with the gas to bring the oxygen concentration to the level of approximately 18% required for good combustion. The diluting effect of nitrogen in the air is thus reduced. The phase diagram positions of the low methane feed gas (①), the methane concentrated gas (②), and the methane concentrated
14
H. Lin et al. / Journal of Membrane Science 469 (2014) 11–18
Fig. 2. Process simulation of a membrane selective exchanger used to upgrade a dilute methane stream into a more concentrated methane stream that can be directly used as fuel. The process recovers 80% methane. The process simulation assumes membrane permeances of 320 gpu for CO2, 200 gpu for O2, 64 gpu for N2 and 32 gpu for CH4.
Fig. 3. (a) Phase diagram showing how a membrane contactor upgrades low methane feed gas to produce a fuel gas that readily burns when mixed with air. Streams ① and ② refer to the feed and product stream compositions in Fig. 2, respectively; stream ➄ refers to the combustion mixture. (b) Simplified schematic of the process shown in Fig. 2.
3. Experimental
minimal resistance to gas flow. The paper and microporous layers provide mechanical strength to the membrane composite structure. The smooth surface of the gutter layer allows a very thin, dense selective layer of perfluoropolymers to be deposited [27,28]. The flat sheet thin film composite membranes were fabricated into four-port spiral‐wound modules that operate in countercurrent/ sweep mode [15,26,29–32]. The feed gas passes down the module in the channel created by the feed spacer. A portion of the feed gas permeates the membrane and then enters the permeate channel. The permeate channel can be swept with a sweep gas, allowing the module to operate in a countercurrent mode [26,29]. More details on the module structure have been shown in our earlier work [15,26] and in the work of others [29–32].
3.1. Membrane and module preparation
3.2. Sweep experiments
Flat sheet thin film composite membranes for gas separation were prepared in this study [5,15]. The separation properties are governed by the thin selective layer, and the other three layers (paper, microporous and gutter layers) are designed to have
Composite membranes and membrane modules were tested under countercurrent/sweep conditions using a system shown in Fig. 4. The CO2/CH4 mixture flowed on the feed side of the membrane, and low pressure air flowed on the permeate side.
gas after mixing with an appropriate amount of air (⑤) are shown in Fig. 3. The concentrated methane gas is still comfortably above the upper flammability limit, but when mixed with air it can be brought into the preferred combustion range and used as fuel. The sweep air leaving the contactor (stream ④ in Fig. 2) contains 26% CO2 and only 1.4% methane and can be safely discharged to the atmosphere without further treatment. The use of air sweep in the membrane exchanger dilutes the selectively permeated CO2 and increases the CO2 pressure driving force across the membrane, thus improving CO2 removal efficiency [5,11,26].
H. Lin et al. / Journal of Membrane Science 469 (2014) 11–18
15
Table 3 Range of pure-gas permeances of composite membranes at 50 psig and 20 1C.
Fig. 4. Schematic of countercurrent/sweep test system for flat sheet composite membranes and spiral‐wound modules. A rectangular countercurrent test cell (with a membrane area of 150 cm2) was used for membrane testing, and one module housing was used to test countercurrent/sweep spiral‐wound modules (with a membrane area of 0.5–1 m2). MFM: mass flow meter. GC: gas chromatography unit.
The test system was equipped with an Agilent MicroGC 3000 portable gas chromatography (GC) unit, to monitor the gas composition of all the streams (including the feed, residue, sweep and ‘sweep þpermeate’ stream). The flow rates were measured using mass flow meters (Sierra Instruments, Monterey, CA). The experimental data were only considered valid when the mass of each component entering and leaving the permeation cell or module housing was within 715%. The mixed-gas permeance in countercurrent/sweep operation can be estimated using the following equation [26, 33]: P A N A lnðΔpA;L =ΔpA;O Þ ¼ Am ðΔpA;L =ΔpA;O Þ l
ð6Þ
ΔpA;O ¼ pA;F pA;S þ P and ΔpA;L ¼ pA;R pA;S
ð7Þ
where pA;F , pA;R , pA;S and pA;S þ P are the partial pressures of component A in the feed, residue, sweep and ‘sweepþ permeate’ stream, respectively. NA is the flow rate difference of component A between the sweep and ‘sweepþpermeate’ stream. Eq. (6) is similar to Eq. (3), except that the partial pressure difference is expressed as the logarithmic mean of ΔpA;L and ΔpA;O . The composite membrane sheets were mounted in a modified Osmonics SEPAs CF II Med/High Foulant System cell (GE Osmonics Labstore, Minnetonka, MN). The test cell was modified to allow a near-perfect countercurrent flow test [34]. The countercurrent/ sweep spiral‐wound modules were installed in a custom-built module housing, with separate ports to accommodate the four gas streams. Chemical-grade nitrogen, oxygen and air, each with a purity of 99%, were used as received from Praxair Inc. (Hayward, CA). An air compressor was also used to provide high-pressure air, the composition of which was determined using the GC.
4. Results and discussion 4.1. Composite membrane preparation and characterization Thin film composite membranes comprised of a perfluoropolymer as the selective layer were prepared. Table 3 shows typical pure-gas permeance and selectivity at feed pressure of 50 psig and 20 1C. The gas selectivity values are similar to those of dense films, indicating that the composite films are defect-free. The membranes show good CO2/CH4 selectivity to selectively remove CO2 from the methane stream, and superior oxygen/methane selectivity to selectively introduce oxygen into the methane product stream. The high CO2/CH4 and O2/CH4 selectivities are ascribed to the unfavorable interactions between perfluoropolymers and
Gas
Pure-gas permeance (gpu)a
Gas/CH4 selectivity
CH4 CO2 N2 O2
10–30 240–720 25–75 75–225
– 24 2.5 7.5
a
1 gpu ¼10 6 cm3(STP)/cm2 s cmHg ¼7.5 10 12 m3(STP)/m2 s Pa.
hydrocarbons such as CH4, leading to high solubility selectivity of CO2/CH4 and O2/CH4 [7,35].
4.2. Evidence of sweep effectiveness The separation performance of a membrane sheet was tested in a countercurrent test cell (150 cm2 area) with and without application of a permeate side sweep. The feed contained 47% methane in CO2 and was fed into the cell at 50 psig and a flow rate of 16 standard liters per minute per square meter of membrane area (slpm/m2). The permeate side was kept at three pressure values (0 psig, 20 psig and 40 psig), and swept with air at various flow rates. The selective layer of the membrane faced the highpressure feed gas. Fig. 5(a) shows that the CO2 permeation rate can be tripled by sweeping the permeate side of the membrane with air. The air sweep reduces the partial pressure of CO2 directly under the selective layer, resulting in an increased driving force for permeation. The enhancement of the CO2 removal is more pronounced at higher permeate pressure. For example, as the air sweep flow rate increases from 0 to 8 slpm/m2, the increase in CO2 removal is 13%, 110% and 250% for the tests at sweep pressures of 0 psig, 20 psig and 40 psig, respectively. This behavior is expected and also has been observed in natural gas dehydration field tests [20]. Simultaneously, as shown in Fig. 5(b), oxygen permeates from the air sweep stream to the CO2/methane stream, in spite of the net permeation flow rate flowing in the opposite direction. A higher sweep pressure means a higher driving force for oxygen to permeate across the membrane, increasing oxygen flux. The data presented in Fig. 5 show that the use of an air sweep significantly increases the CO2 removal rate, thereby increasing the actual methane concentration, at the same time that it introduces oxygen to the methane product stream, which increases the equivalent methane concentration beyond the actual methane concentration. Fig. 6 summarizes these results for a larger set of experiments. The results are shown as the equivalent methane concentration (EMC) of the product gas versus the methane lost to the air sweep exiting the system (methane loss is defined as the fraction of the methane in the original feed stream that is lost with the air sweep exiting the system). As shown in Fig. 6, the use of the air sweep increases the EMC, while the loss of methane is significantly reduced. The oxygen concentration in the methane stream in these experiments ranges from 1% to 5%, which means that oxygen permeation is responsible for up to 30% of the increase in the EMC. The intrinsic pure-gas CO2 permeance of the membrane used in these experiments is 240 gpu. The effective permeance observed in the countercurrent/sweep experiments is in the range of 80 to 200 gpu. This reduction in permeance is presumably caused by concentration polarization on the permeate side (such as in the support layers of the membrane) [5,36,37].
16
H. Lin et al. / Journal of Membrane Science 469 (2014) 11–18
Sweep : feed flow ratio 1.0
Sweep : feed flow ratio
2.0
3.0
0
1.0
2.0
3.0
1.5
Sweep pressure:
8
O2 permeation into 2 methane stream (slpm/m )
CO2 removal from 2 methane stream (slpm/m )
10
0
0 psig 20 psig 40 psig
6
4
2
1.2
Sweep pressure: 40 psig
0.9 20 psig 0.6
0.3 0 psig
0
0
10
20
30
40
50
0
0
10
2
20
30
40
50
2
Air sweep flow rate (slpm/m )
Air sweep flow rate (slpm/m )
Fig. 5. Amount of (a) CO2 and (b) oxygen permeating through the perfluoropolymer-based membrane as a function of air sweep flow rate at various air sweep pressures (0 psig, 20 psig and 40 psig). The feed contains 47% CH4 at about 16 slpm/m2 and 50 psig.
100
With air sweep
90
EMC of methane product (%)
Equivalent methane concentration (%)
100
80 70 No air sweep 60 50
Module with sweep
80
60
40
Module without sweep 20
No separation
40
0
0
5
10 15 Methane loss (%)
20
25
0
5
10
15
20
25
Feed methane concentration (%)
Fig. 6. Equivalent methane concentration (EMC, defined in Eq. (2)) versus methane loss for a set of experiments using perfluoropolymer-based composite membranes. The feed gas contains 47% methane at 50 psig and feed flow rates are in the range of 5.7–31 slpm/m2. The air sweep is at pressures of 0–40 psig and sweep flow rates of 0–34 slpm/m2. The process recovers 80–95% of the methane that is otherwise lost as a dilute methane vent stream.
Fig. 7. Equivalent methane concentration (EMC) achieved by the countercurrent/ sweep test module, as a function of the feed methane concentration. The module operates significantly more efficiently in sweep mode than without sweep.
4.3. Performance of a countercurrent/sweep module
4.4. Economic analysis
A spiral‐wound countercurrent/sweep module was fabricated using the same membrane tested in the permeation cell. The module has one membrane envelope with an effective area of 0.6 m2, about 40 times the area of the permeation cell. The module has a pure gas CO2 permeance of 200 gpu and CO2/CH4 selectivity of 24, suggesting that the module is defect-free. The module was tested with a CO2/methane feed stream at 50 psig and 15 slpm/m2, and an air sweep stream at 0 psig and 9.0 slpm/m2. Fig. 7 shows the EMC values achieved by the module in a single pass, as a function of the methane concentration in the feed stream. The module effectively upgrades the dilute methane streams in the sweep mode because the module both removes more CO2 than the
To explore the economic feasibility of the membrane-based process for utilization of dilute methane streams, we performed process simulations for streams with methane concentrations ranging from 20% to 60%. The goal was to create a methane product stream with an EMC value of 70%, so the stream could be easily used as fuel. Table 4 lists the membrane areas required, the methane recoveries achieved, the value of the recovered methane, and the capital cost of the membrane system to recover a combustible product at five different dilute methane starting compositions. Air sweep flow rate was 1 MMscfd. Assumed membrane permeances were 320 gpu for CO2, 200 gpu for O2, 64 gpu for N2 and 32 gpu for CH4, higher than those achieved during the membrane and module
operation without sweep and introduces oxygen into the produced fuel stream.
H. Lin et al. / Journal of Membrane Science 469 (2014) 11–18
Table 4 Process parameters and capital costs for a membrane selective exchange unit treating 1 MMscfd of dilute methane gas at various methane concentrations and producing a fuel product stream with 70% equivalent methane concentration. Feed methane content (%)
Membrane area (m2)
Methane recovery (%)
Fuel product value ($1000/ year)a
Membrane system capital cost ($1000)b
60 50 40 30 20
130 225 295 345 380
95.8 92.1 88.5 84.8 80.4
840 670 520 370 230
130 225 295 345 380
a b
Based on fuel value of $4/1000 scf methane. Based on $1000 per square meter membrane area.
17
in CO2 partial pressure difference across the membrane and the increase in the CO2 mass transfer coefficient in the boundary layer on the permeate side. The process is cost-effective and reduces methane emissions to the atmosphere. For example, the membrane exchanger can upgrade a dilute methane stream containing as little as 20% methane, to a fuel quality methane stream with equivalent methane content of 70%. The value of the methane recovery is $230,000 per year (assuming $4/1000 scf methane), while the operating cost of the membrane exchange system is $133,000 per year. The economics of the membrane selective exchanger become better if the waste streams have higher methane content and the credit for the avoidance of methane emissions is considered.
Acknowledgments Table 5 Estimate of the operating cost for a membrane system converting 1 MMscfd of 20% methane gas into useful fuel with an EMC value of 70%. Cost category Installed system cost Depreciation Annual module replacement costs Total annual operating cost
Unit cost
Annual value of recovered methane
$4/1000 scf CH4
20% of capital cost $150/m2
Value ($1000s) $380 $76 $57 $133 $230
experiments, but very realistic based on the most recent membranes produced at Membrane Technology & Research, Inc. The cost of the membrane system installed in the field is estimated to be $1000/m2 membrane area, which is realistic for the relatively small flow rates considered here. This cost is based on our experience in building commercial membrane systems for the natural gas industry. As expected, the capital cost of the membrane system decreases with increasing methane content in the dilute methane streams. Table 5 gives a breakdown of the capital and operating costs for the 20% methane case and shows how promising the process economics are for this least attractive case. The membrane system is assumed to be depreciated over a five-year period. The membrane modules are assumed to have a three-year operating lifetime, and the module replacement cost is estimated at $450/m2 membrane area. Therefore, the annual cost of membrane modules is $150/m2 membrane area. As shown in Table 5, the operating cost of the membrane system ($133,000 per year) is easily offset by the methane value recovered ($230,000 per year assuming a value of $4/1000 scf methane). The economics of the membrane selective exchanger are even more attractive, if the dilute methane streams have higher methane content and the credit for the avoidance of methane emissions is considered (as shown in Table 1).
5. Conclusions This paper describes a membrane selective exchange process to upgrade dilute methane streams (containing inerts such as CO2) to fuel grade products. The membrane uses an air sweep on the permeate side, which improves the CO2 removal from the dilute methane stream and introduces oxygen to the methane product, thus improving the quality of the methane stream as a fuel. Laboratory experiments with membrane stamps and a prototype membrane module have demonstrated the feasibility of the selective exchange process that generates useful fuel streams from dilute methane vent streams. The efficiency improves with increasing air sweep flow rate, presumably due to the increase
We gratefully acknowledge the partial financial support of this work by the U.S. Department of Energy Small Business Innovation Program (SBIR) Phase I, under Contract number DE-FG02-06ER84609, and the U.S. Environmental Protection Agency SBIR Phase I under Contract number EP-D-10-036 and Phase II under EP-D-11071. One of the authors, H. Lin, also acknowledges the funds provided by the University at Buffalo, State University of New York, during the preparation of the manuscript.
Nomenclature membrane area (cm2) British thermal unit diffusivity of gas A (cm2/s) diffusivity selectivity of gas A over gas B equivalent methane content, defined in Eq. (2) gas permeation unit [10 6 cm3(STP)/cm2 s cmHg] thickness of the membrane selective layer (cm) million British thermal units steady state flux of gas A across the membrane [cm3(STP)/s] PA permeability of membrane selective layer to gas A (Barrer) PA/l permeance of composite membrane to gas A (gpu) Δp A partial pressure difference across the membrane for gas A (cmHg) ΔpA;0 partial pressure difference across the membrane for gas A between feed and ‘sweepþpermeate’ streams (cmHg) ΔpA;L partial pressure difference across the membrane for gas A between residue and sweep streams (cmHg) SA solubility of gas A in a polymer [cm3(STP)/(cm3 polymer atm)] SA/SB solubility selectivity of gas A over gas B scf standard cubic foot slpm/m2 unit of gas flow rate, standard litter per minute per m2 membrane area T temperature (K) yCH4 methane concentration in the ternary mixture (vol %) yO2 oxygen concentration in the ternary mixture (vol%) Am Btu DA DA/DB EMC gpu l MMBtu NA
Greek letter
αA=B
permeability or permeance selectivity of gas A over gas B
18
H. Lin et al. / Journal of Membrane Science 469 (2014) 11–18
Subscript A B F R S S þP
gas component A gas component B feed stream residue stream sweep stream the exiting stream in the permeate including the permeate and sweep gas
References [1] Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990–2006, U.S. Environmental Protection Agency, 2008. [2] I. Karakurt, G. Aydin, K. Aydiner, Mine ventilation air methane as a sustainable energy source, Renew. Sustain. Energy Rev. 15 (2011) 1042–1049. [3] J.G. Wijmans, T.C. Merkel, H. Lin, S.M. Thompson, R. Daniels, Efficient gasseparation process to upgrade dilute methane stream for use as fuel, U.S. Patent 8124733, 2012. [4] N.P. Cheremisinoff, Handbook of Solid Waste Management and Waste Minimization Technologies, Butterworth-Heinemann, Burlington, MA (2003). [5] R.W. Baker, Membrane Technology and Applications, 3rd ed., John Wiley and Sons Ltd., Chichester, UK, 2012. [6] R.W. Baker, K. Lokhandwala, Natural gas processing with membranes: an overview, Ind. Eng. Chem. Res. 47 (2008) 2109–2121. [7] T.C. Merkel, I. Pinnau, R.S. Prabhakar, B.D. Freeman, Gas and vapor transport properties of perfluoropolymers, in: B.D. Freeman, Y.P. Yampolskii, I. Pinnau (Eds.), Materials Science of Membranes for Gas and Vapor Separation, John Wiley & Sons Ltd, Chichester, UK, 2006, pp. 251–270. [8] EPA's Landfill Methane Outreach Program, 〈http://www.epa.gov/lmop/〉, 2014. [9] Cost and Performance Baseline for Fossil Energy Plants (V 1): Bituminous Coal and Natural Gas to Electricity, Report No. DOE/NETL-2010-1397, September 2013. [10] S. Su, A. Beath, H. Guo, C. Mallett, An assessment of mine methane mitigation and utilisation technologies, Prog. Energy Combust. 31 (2005) 123–170. [11] T.C. Merkel, H. Lin, X. Wei, R.W. Baker, Power plant post-combustion carbon dioxide capture: an opportunity for membranes, J. Membr. Sci. 359 (2010) 126–139. [12] H.F. Coward, G.W. Jones, Limits of Flammability of Gases and Vapors, U.S. Bureau of Mines, 1952. [13] C.V. Mashuga, D.A. Crowl, Application of the flammability diagram for evaluation of fire and explosion hazards of flammable vapors, Process Saf. Prog. 17 (1998) 176–183. [14] C.E. Baukal, Oxygen-Enhanced Combustion, CRC Press, Boca Raton, 1998. [15] H. Lin, M. Zhou, J. Ly, J. Vu, J.G. Wijmans, T.C. Merkel, J. Jin, A. Haldeman, E. H. Wagener, D. Rue, Membrane-based oxygen-enriched combustion, Ind. Eng. Chem. Res. 52 (2013) 10820–10834. [16] S.G. Kimura, W.R. Browall, Membrane oxygen enrichment I: Demonstration of membrane oxygen enrichment for natural gas combustion, J. Membr. Sci. 29 (1986) 69–77.
[17] J.G. Wijmans, R.W. Baker, The solution-diffusion model: a review, J. Membr. Sci. 107 (1995) 1–21. [18] B.D. Freeman, Basis of permeability/selectivity tradeoff relations in polymeric gas separation membranes, Macromolecules 32 (1999) 375–380. [19] R.C. Reid, J.M. Prausnitz, B.E. Poling, The Properties of Gases and Liquids, McGraw-Hill, New York, 1987. [20] D.W. Breck, Zeolite Molecular Sieves, Krieger Publishing Company, Florida, 1974. [21] D.E. Sanders, Z.P. Smith, R.L. Guo, L.M. Robeson, J.E. McGrath, D.R. Paul, B. D. Freeman, Energy-efficient polymeric gas separation membranes for a sustainable future: a review, Polymer 54 (2013) 4729–4761. [22] C.Y. Pan, H.W. Hapgood, Gas separation by permeation, Can. J. Chem. Eng. 56 (1978) 197. [23] D.T. Coker, T. Allen, B.D. Freeman, G.K. Fleming, Nonisothermal model for gas separation hollow-fiber membranes, AIChE J. 45 (1999) 1451–1468. [24] B. Belaissaoui, Y.L. Moullec, D. Willson, E. Favre, Hybrid membrane cryogenic process for post-combustion CO2 capture, J. Membr. Sci. 415-416 (2012) 424–434. [25] S. Kelman, H. Lin, E.S. Sanders, B.D. Freeman, CO2/C2H6 separation using solubility selective membranes, J. Membr. Sci. 305 (2007) 57–68. [26] H. Lin, S.M. Thompson, A. Serbanescu-Martin, J.G. Wijmans, K.D. Amo, K. A. Lokhandwala, B. Low, T.C. Merkel, Dehydration of natural gas using membranes. Part II: Sweep/countercurrent design and field test, J. Membr. Sci. 432 (2013) 106–114. [27] K.A. Lundy, I. Cabasso, Analysis and construction of multilayer composite membranes for the separation of gas mixtures, Ind. Eng. Chem. Res. 28 (1989) 742–756. [28] W. Yave, A. Car, J. Wind, K. Peinemann, Nanometric thin film membranes manufactured on square meter scale: ultra-thin films for CO2 capture, Nanotechnology 21 (2010) 1–7. [29] D. Reddy, T.Y. Moon, C.E. Reineke, Counter current dual-flow spiral wound dual-pipe membrane separation, U.S. Patent 5,034,126, 1991. [30] T.Y. Cath, A.E. Childress, M. Elimelech, Forward osmosis: principles, applications, and recent developments, J. Membr. Sci. 261 (2006) 70–87. [31] G.D. Mehta, Further results on the performance of present-day osmotic membranes in various osmotic regions, J. Membr. Sci. 10 (1982) 3–19. [32] C. Bayer, M. Follmann, H. Breisig, I.M. Wienk, F.P. Cuperus, M. Wessling, T. Melin, On the design of a 4-end spiral-wound L/L extraction membrane module, Ind. Eng. Chem. Res. 52 (2013) 1004–1014. [33] 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. Membr. Sci. 367 (2011) 233–239. [34] H. Lin, S.M. Thompson, A. Serbanescu-Martin, H.G. Wijmans, K.D. Amo, K. Lokhandwala, T.C. Merkel, Dehydration of natural gas using membranes. Part I: Composite membranes, J. Membr. Sci. 413-414 (2012) 70–81. [35] R.S. Prabhakar, B.D. Freeman, I. Roman, Gas and vapor sorption and permeation in poly(2,2,4-trifluoro-5-trifluoromethoxy-1,3-dioxole-co-tetrafluoroethylene), Macromolecules 37 (2004) 7688–7697. [36] J.G. Wijmans, A.L. Athayde, R. Daniels, J.H. Ly, H.D. Kamaruddin, I. Pinnau, The role of boundary layers in the removal of volatile organic compounds from water by pervaporation, J. Membr. Sci. 109 (1996) 135–146. [37] J.R. McCutcheon, M. Elimelech, Influence of concentrative and dilutive internal concentration polarization on flux behavior in forward osmosis, J. Membr. Sci. 284 (2006) 237–247.