Int. J. Environ. Sci. Technol. DOI 10.1007/s13762-016-1156-3

ORIGINAL PAPER

Separation of H2S from CH4 by polymeric membranes at different H2S concentrations S. M. S. Niknejad1 • H. Savoji1,2 • M. Pourafshari Chenar3 • M. Soltanieh1

Received: 3 July 2016 / Revised: 13 October 2016 / Accepted: 7 November 2016 Ó Islamic Azad University (IAU) 2016

Abstract In this work, permeation of mixed gases H2S/CH4 through commercial polyphenylene oxide (PPO) hollow fiber and poly (ester urethane) urea (PEUU) flat membranes was studied at pressures of 345–689 kPa, at ambient temperature and at 313.15 K. Various H2S concentrations of about 100–5000 ppm in CH4 binary synthetic gas mixtures as well as a real natural gas sample obtained from a gas refinery containing 0.3360 mol.% H2S (equivalent to 3360 ppm) were tested. It was observed that the permeance of components was affected by the balance between competitive sorption and plasticization effects. Separation factors of H2S/CH4 were in the range of 1.3–2.9, 1.8–3.1 and 2.2–4.3 at pressures of 345, 517 and 689 kPa, respectively. In the range of 101–5008 ppm of H2S in CH4, the effect of temperature on the separation factor was nearly negligible;

Editorial responsibility: B.V. Thomas. Experimental results of binary mixtures separation of CH4 at various H2S concentrations from low (*100 ppm) to more than 5000 ppm by PPO and PEUU membranes were presented. Also, tests with a real gas sample containing high concentration of H2S and other impurities have been reported. To our knowledge, very few studies have reported investigations on high concentration of acid gas and real gas separation with these membranes and different configurations. & M. Soltanieh [email protected] 1

Department of Chemical and Petroleum Engineering, Sharif University of Technology, Tehran, P.O. Box 11155-9465, Iran

2

Industrial Membrane Research Institute, Department of Chemical and Biological Engineering, University of Ottawa, Ottawa, ON K1N 6N5, Canada

3

Chemical Engineering Department, Ferdowsi University of Mashhad, Mashhad, P.O. Box 91775-1111, Iran

however, permeances of both components of the mixtures increased with temperature. Additionally, the results obtained by PEUU membrane indicated that it was a better choice for hydrogen sulfide separation from H2S/CH4 mixtures than PPO. For PPO membrane, removal of hydrogen sulfide from high-concentration (up to 5008 ppm) binary mixtures of H2S/CH4 was compared with that of low concentration (as low as 101 ppm) through PPO. At concentrations of 101–968 ppm, plasticization was dominant compared with the competitive sorption, while for the H2S feed concentrations of 3048 ppm, the competitive sorption effect was dominant. For H2S concentration of 5008 ppm, the balance between these two effects played an important role for explanation of its trend. Keywords PPO and PEUU membranes  Mixed gas permeation  Acid gas  Plasticization  Competitive sorption

Introduction Hydrogen sulfide exists naturally in the environment (e.g., in natural gas, volcanic gases, sulfur springs). It is also produced by living organisms and as a byproduct of many industrial processes such as paper manufacturing and sewage treatment. Natural gas contains various amounts of hydrogen sulfide from a few parts per million (ppm) to as high as 28 wt%. Natural gas with high levels of hydrogen sulfide is considered as sour gas. H2S is a dangerous air pollutant in natural gas processing, petroleum refineries and geothermal energy production, which is of a significant public health concern when released to the atmosphere. The Iranian raw natural gas has H2S content from 66.2 to 32700 ppm (3.27 mol.%) in different gas fields. Removal

123

Int. J. Environ. Sci. Technol.

of acid gas H2S from sour gas (which is called sweetening) is usually carried out by amine absorption processes. However, membrane processes have also been used during the past few decades due to several advantages such as lower energy and capital costs, operational simplicity and lower adverse effects on the environment. The membrane units and devices are also usually compact and modular which require less space and operate at relatively low pressures. The main disadvantages, which have prevented wider applications of membrane-based processes, are the fouling of the membranes, the durability of membranes and the unavailability of suitable membranes for specific operations (Chou 2003; Ho and Sirkar 1992; Savoji et al. 2012). Although commercialization of membranes for natural gas sweetening started about 30 years ago mostly for the bulk removal of carbon dioxide, the application of membranes for removal of high concentration of H2S from natural gas has been limited and is still a challenge. Due to the toxic and corrosive properties of H2S, it has to be removed from natural gas to the level of 2–4 ppm in order to meet the specifications of pipeline and other equipment. Removal of H2S from high concentrations (5000 ppm or more) to this level by membranes is a difficult task and needs further investigation. Most existing commercial membranes are not capable of removal of H2S to pipeline specification economically and a hybrid system including a membrane unit followed by absorption would be the preferred process (Baker 2004; Bhide and Stern 1993). Pure gas permeation experiments have often been used to provide an indication of possible performance of various membranes under ideal conditions of no interactions and no plasticization effect. But in reality, the transport of a component in a gas mixture through glassy polymeric membranes is affected by the presence of other penetrants either due to the interactions among the permeating species, called multicomponent effect, or by the plasticization of the polymers if the mixture contains certain hydrocarbons. Generally, mixed gas separation yields lower selectivities for membranes than those of pure gas measurements. Besides, membrane performance deteriorates with time because of aging, plasticization and other effects (Ettouney and Majeed 1997; Wang et al. 2002). Among the glassy polymers, polyphenylene oxide (PPO) possesses excellent separation properties that make it a suitable candidate for a wide range of gas separation processes. The PPO is a linear amorphous thermoplastic with glass transition temperature (Tg) from 485.15 to 491.15 K. Because of the phenyl rings, PPO is hydrophobic in nature and has excellent resistance to water, acids, bases, alcohols

123

and steam. It has been reported that among all glassy polymers PPO shows one of the highest permeabilities to gases. The high permeability has been attributed to the absence of polar groups in the main chain of PPO. On the other hand, poly ether/ester urethane urea (PEUU) membranes have shown good separation performance, which makes them potential candidates for sour natural gas sweetening applications. These polymers contain two different types of segments including hard segment from polyamides, and soft segment from flexible ether or ester linkages (Paul and Yampol’skii 1994; Aguilar-Vega and Paul 1993). There are only a few published articles that have utilized polymeric membranes for hydrogen sulfide acid gas separation. The main reason for this is the high toxic and corrosive properties of this gas. In fact, only a few researches have been carried out for separation of H2S from CH4, notably the studies reported by Chatterjee et al. (1997), Wilks and Rezac (2002) and Mohammadi et al. (2008) in which the H2S/CH4 selectivities have been reported. Most results presented in these researches were related to cellulose acetate (CA), polyimide and polyurethane membranes. Pourafshari Chenar et al. (2011) have studied removal of low concentrations of H2S from CH4 using PPO and Cardo-type polyimide membranes under different experimental conditions. Their experiments were conducted at ambient temperature (293.15 K) and 313.15 K and pressures of 345–689 kPa. Three cylinders of mixed gas containing 101–401 ppm of H2S in CH4 were used. Their results showed that in the presence of H2S, the permeance of methane declined for the Cardo-type polyimide membrane, whereas for the PPO membrane it remained relatively unchanged. Moreover, the separation factors of H2S/ CH4 were 6 and 4 for Cardo-type polyimide and PPO membranes, respectively. It was also observed that components’ permeabilities increased with temperature, but separation factor remained constant. In this paper, binary mixtures containing 101–401 ppm hydrogen sulfide in methane (called low concentration range) and 968–5008 ppm hydrogen sulfide in methane (called high concentrations range) were separated by the membranes described above. In addition, a real natural gas sample obtained from a gas refinery containing high concentration of H2S and other impurities was tested by PPO membrane. The objectives of this work are: (1) to conduct an experimental study of high concentration H2S removal from CH4 in synthetic mixtures using PPO and PEUU membranes, (2) to compare the results with H2S separation data from real natural gases obtained from a gas refinery, and (3) to compare and discuss the removal of high and low concentrations of H2S using PPO under different experimental conditions.

Int. J. Environ. Sci. Technol.

Materials and methods PPO and PEUU membranes The commercial PPO hollow fiber membrane was generously supplied by Aquilo Gas Separation B.V. (Parker Filtration and Separation B.V., The Netherlands). The PPO module was assembled by loading a bundle of 10 fibers in a shell and tube configuration. The feed gas was entered inside the bore side of fibers, the permeate was collected on the shell side, and the retentate was rejected from the other end of the fibers. The selective dense layers of the hollow fibers were located at the outside surface of membrane. Hence, the permeances of the gases were calculated based on the outer surface area of the fibers. The PEUU flat membrane module (IsoGelTM) was generously supplied by Metax (Russia). It consists of three layers: the sub-layer, the porous layer and the diffusive layer. The sub-layer is non-woven polyester, the porous layer is Teflon as a support, and the diffusive layer, which provides the main separation characteristics, is polyester urethane urea (PEUU). The chemical structures of these polymers are shown in Fig. 1. The main specifications of the membranes are presented in Table 1. The experiments were divided into two parts. In the first part, the removal of high concentrations of H2S was investigated by using the two types of membranes, where three different H2S/CH4 binary mixtures with 968, 3048 and 5008 ppm of H2S in CH4 were used as the feed. These gas mixtures were purchased from TGA Co., Iran. The experiments were conducted at ambient temperature (approximately 293.15–295.65 K) and at 313.15 K and three different pressures of 345, 517 and 689 kPa. In the second

Fig. 1 Chemical structures of a poly (2,6-dimethyl-1,4-phenylene oxide) (PPO) and b poly (ester urethane) urea (PEUU)

part, the removal of H2S from high-concentration (up to 5008 ppm) binary mixtures of H2S/CH4 was investigated and the results were compared with the data obtained by Pourafshari Chenar et al. (2011) for low concentration (as low as 101 ppm) through PPO. Experiments at low concentration were conducted at ambient temperature and three different pressures of 345, 517 and 689 kPa. Another set of experiments was also conducted at 689 kPa and 313.15 K in order to investigate the effect of temperature on the performance of the PPO membrane. Flowchart of the experiments and the number of tests is shown in Fig. 2a. These experiments were based on three different mixed gases of H2S/CH4 with 101, 198 and 401 ppm of H2S in CH4 that were purchased from Praxair Distribution Inc., Canada, to compare with the data obtained by Pourafshari Chenar et al. (2011). In addition, a cylinder of real natural gas was obtained from the South Pars Gas Refinery of Iran, phases 6, 7 and 8. This sample was the pretreated gas as the feedstock of the mentioned gas refineries and consisted of all components of real sour gas: C1, C2, C3?, H2O, CO2, H2S, etc. The relevant experiments were done at ambient temperature and 689 kPa. The composition of the gas is given in Table 2. Gas permeation setup Figure 2b shows the process flow diagram of the setup used for the experiments. Pressure was kept constant using a backpressure regulator mounted on the retentate line. For the H2S/CH4 binary mixtures containing 101–401 ppm of H2S, the permeation rate was measured by a wet-test-meter. The compositions of the retentate and permeate streams were determined by a gas chromatograph (Varian 3400) equipped with a thermal conductivity detector (TCD) and a ‘‘HayeSep T’’ column. The compositions were corrected using the thermal response factors, TRFs. The TRF values were carefully calculated for each gas following the Dietz’s method (Pourafshari Chenar et al. 2011). For the H2S/CH4 mixed gases containing 968–5008 ppm of H2S, the permeation and retentate rates were measured by a bubble flow meter. The H2S concentration in permeate and retentate streams was obtained by the TutWiler method.

Table 1 Specifications of PPO hollow fiber membrane and PEUU flat membrane modules Module type

Membrane type

Membrane parameters 6

Inside diameter (10 9 m)

Module specifications 6

Outside diameter (10 9 m)

No. of fibers

Permeation area (104 9 m2)

Hollow fiber

PPO

370

520

10

49

Flat

PEUU

–

–

–

30.86

123

Int. J. Environ. Sci. Technol. 345 kPa

(a) Amb. Temp. (295.65 K) Low Concentration Range (101, 198, & 401 ppm H2S)

517 kPa 689 kPa

313.15 K

689 kPa 345 kPa

Amb. Temp. (293.15 K) PPO

689 kPa

High Concentration Range (968, 3048, & 5008 ppm H2S)

345 kPa 313.15 K

Experiments

517 kPa

517 kPa 689 kPa

Real Natural Gas (3360 ppm H2S)

Amb. Temp. (294.15 K)

689 kPa

345 kPa Amb. Temp. (296.15 K)

High Concentration Range (5008 ppm H2S)

PEUU

517 kPa 689 kPa

(b)

Fig. 2 a Flowchart of the experiments and b gas permeation setup and membrane cell of the constant pressure membrane systems

Theory

where

The permeance and selectivity of component i were calculated using ideal gas law. Furthermore, the permeance of gas i defined as the pressure normalized flux of a gas through membrane was calculated as follows (Pourafshari Chenar et al. 2006):   P QP yi  1010   ð1Þ ¼ l i 3:35  pF ðxi Þln pP yi A

ðxi Þln ¼

123

xi;F  xi;R
ln xi;F =xi;R

ð2Þ

Ideal selectivity of gas i over gas j was defined as below: aoij ¼

ðP=lÞi ðP=lÞj

ð3Þ

Separation factor of gas i over gas j was given as the

Int. J. Environ. Sci. Technol. Table 2 Gas composition of the real sample of South Pars gas refinery of Iran, phases 6, 7 and 8

Component

Mol.%

H2O

1.5579

N2

3.3062

CO2

1.7411

H2S

0.6572

C1

80.966

C2

5.1747

C3

1.9173

i-C4

0.4109

n-C4

0.6946

i-C5

0.2838

n-C5

0.2838

C6?

2.5136

COS

0.0003

Methyl mercaptan

0.0029

Ethyl mercaptan nPMercaptan

0.0245 0.0138

nBMercaptan

0.0042

1Hexanthiol

0.0082

MEG

0.4391

Total

100

ratio of the concentration of gas i in the permeate to that in the feed relative to the same ratio for gas j: aij ¼

yi =ðxi Þln
yj = xj ln

ð4Þ Fig. 3 Effect of feed pressure on a H2S permeance and b CH4 permeance at ambient temperature and various H2S feed concentrations

Results and discussion Permeability results The effect of feed pressure at various H2S concentrations on permeation of H2S for PPO and PEUU membranes is depicted in Fig. 3a. It is seen that for H2S concentration of 968 ppm, the permeance of H2S increased as the feed pressure increased, which characterizes the plasticization effect of H2S. For H2S concentration 3048 ppm, the permeance of H2S decreased as the feed pressure increased, which indicates that the competitive sorption dominates the plasticization effect. For the H2S concentration of 5008 ppm, the behavior was in an unexpected way: Increasing the feed pressure enhances H2S permeance for the mixed gas feed containing 968 ppm of H2S, but enhancement of H2S permeance varies for the other two mixed gases. Therefore, competitive sorption effects cause the permeance of penetrants to decrease with increasing feed pressure, while plasticization effects cause a permeance increase with increasing feed pressure.

This phenomenon can be described by the comparative effects of competitive sorption and plasticization. Competitive sorption effects cause the penetrants’ permeance to decrease with increasing feed pressure because of the condensation and lack of motion of hydrogen sulfide at Langmuir sites. In case of glassy polymers like the PPO and PEUU membranes used in this work, molecular-scale packing defects are trapped in the polymer matrix during experimental timescales. This phenomenon causes a concentration or pressure dependence of the sorption and diffusion coefficients. In glassy polymers, the sorption coefficient is generally a decreasing function of the penetrant partial pressure. This is because of saturation of the limited number of glassy packing defects, which give additional sorption sites beyond the more simply densely packed area. Penetrant sorption into the above defects, which are referred to as Langmuir sites, is similar to sorption in zeolite cavity. At higher pressures, these cavities are full and additional sorption happens mainly in the more densely packed Henry’s Law type milieu. Besides, competition for the limited number of Langmuir sorption

123

Int. J. Environ. Sci. Technol.

sites by a second condensable species can lower penetrant sorption in glassy media. The effects of pressure and competition on the sorption coefficient can be introduced using dual-mode sorption (Vaughn and Koros 2014). Based on the dual-mode sorption model given by Koros et al. (1981), the primary effect for the permeation of gases in a mixture is the competition between penetrants to occupy the fixed unrelaxed free volume in the polymer, which may cause significant depression in sorption of both gases in a binary mixture. Likewise, the variation of H2S permeation versus feed pressure for PEUU is illustrated in Fig. 3a. It shows that the permeance of H2S decreased as the feed pressure increased but only for H2S concentration of 5008 ppm. As described before, competitive sorption effects will be stronger due to a higher H2S feed gas concentration. Also the results for permeance of H2S at low concentrations of H2S as shown in Fig. 3a demonstrate that the permeation of H2S increased (except for H2S concentration of 401 ppm) as the feed pressure and the concentration of H2S increased. This can be explained by the higher driving force due to the higher partial pressure difference across the membrane. With increasing H2S concentration in the feed, the relative change in H2S permeance with increasing pressure becomes larger. Moreover, it was observed that the effect of pressure on the H2S permeance is not significant. As explained earlier, the results of high concentration of H2S showed that there are reasonable and clear trends for H2S concentration of 968 and 3048 ppm, except for 5008 ppm. For H2S concentration of 968 ppm, the permeance of H2S increased as the feed pressure increased because of plasticization effect. For the H2S concentration of 3048 ppm, the permeability of H2S plunged as the feed pressure increased. In this case, competitive sorption is the most influential phenomenon that can dominate the plasticization effect. For H2S concentration of 5008 ppm, the same trend is not observed. This phenomenon can be explained by the balance between competitive sorption and plasticization effects. At first, penetrant permeance plummeted with increasing feed pressure to just over 25 mol m-2 s-1 Pa-1 due to competitive sorption effect. After the pressure of 517 kPa, the figure for 5008 ppm soared to just under 35 mol m-2 s-1 Pa-1. Figure 3b demonstrates the CH4 permeation at various pressures and different concentrations of H2S for PPO and PEUU membranes at ambient temperature. It can be observed from this figure that with increasing feed pressure, CH4 permeance decreased gently in most cases in accordance with the dual-mode sorption mechanism. On the other hand, a minimum in the CH4 permeance versus H2S feed pressure for H2S concentration of 968 ppm is considered as a characteristic behavior for plasticization effect. It suggests that plasticization is not as dominant in

123

PPO at low concentrations of H2S. A similar behavior is also observed for PEUU but only for H2S concentration of 5008 ppm. Additionally, the results obtained at low concentrations of H2S indicate a slight increase in methane permeance as pressure increased. The accelerated plasticization effect causes a continuous increase in the permeance of the second component (here is CH4) with increasing pressure as well. Due to the relative effect of plasticization, competitive effects hardly influence in this case (Pourafshari Chenar et al. 2011). The effect of pressure on permeances of H2S and CH4 at 313.15 K for PPO is shown in Fig. 4a, b, respectively. Comparing these observations with those shown in Fig. 3 for ambient temperature of 293.15 K, it can be seen that as the temperature increases from 293.15 to 313.15 K, the results have similar trends in both figures. Comparison of Figs. 3a, b and 4a, b reveal that H2S permeation was significantly higher than that of CH4 through both membranes. This can be attributed to the adsorption of H2S on the pore wall of the membranes. Additionally, comparison of the results of the above-

Fig. 4 Effect of feed pressure and H2S concentration on a H2S permeance and b CH4 permeance through PPO membrane at temperature of 313.15 K

Int. J. Environ. Sci. Technol.

Fig. 5 Effect of H2S feed concentration and temperature on H2S and CH4 permeance for PPO membrane

Fig. 6 Variation of separation factor and ideal selectivity with H2S feed concentration for PPO membrane

mentioned figures indicates that CH4 permeation was relatively independent of feed pressure and H2S feed concentration, whereas H2S permeation varies with both H2S concentration and feed pressure. It can be inferred that unlike H2S, CH4 permeance did not vary much with increasing feed pressure as well as H2S concentration of the feed. Increasing H2S permeation leads to higher adsorption of this component in competition with the CH4 which is not adsorbed. The effects of H2S feed concentration and temperature on permeance of H2S and CH4 for PPO are shown in Fig. 5. A comparison of the curves in this figure reveals that temperature increase from 293.15 to 313.15 K at various pressures results in 20–25% increase in CH4 permeance while H2S permeance increased by 19–38%. Moreover, the curves display that unlike H2S, methane permeation was nearly independent of H2S feed concentration.

The results also reveal that H2S feed concentration has a significant effect on the selectivity of PPO membrane. It should be pointed out that the H2S permeable polar molecules have dipole moment, whereas CH4 molecules are non-polar. Hence, for the H2S feed concentration of up to 3048 ppm, the tendency of H2S polar molecules for being adsorbed onto the pore wall of PPO was more than that of CH4 non-polar molecules. Study on the selectivity for H2S feed concentration of higher than 3048 ppm shows that the selectivity of the membrane dropped, whereas for the H2S feed concentrations of below 3048 ppm this behavior is not observed. This is due to the domination of plasticization versus competitive sorption. In order to investigate the treatment of a real natural gas sample (See Table 2 above), the experiments were carried out with this real natural gas at 689 kPa and ambient temperature using PPO membrane. The results are also shown in Fig. 6. The measured values of selectivity and separation factor for real gas experiment are approximately similar to those obtained for synthetic mixed gases. Table 3a shows the concentration of H2S in feed, retentate and permeate along with the separation factors of H2S/CH4 at different feed pressures and different H2S feed concentrations. These results indicate that the separation factors of PPO were higher at high feed concentrations of H2S which indicates the better performance of this membrane. The results for separation of natural gas containing 3360 ppm H2S in the feed at 689 kPa are given in Table 3b.

H2S/CH4 separation factor and ideal selectivity The separation factor and ideal selectivity versus H2S feed concentration for both components of the binary mixtures are depicted in Fig. 6. From this figure, it can be seen that the selectivity of PPO at 313.15 K was up to 1.4 times greater than that at ambient temperature. Because of the permeance trends of H2S, the separation factor and ideal selectivity of this component remained in the same trend, i.e., concave downward. Since the CH4 permeance also increased at the same rate with increasing temperature, the separation factor and ideal selectivity of this gas mixture remained nearly constant and did not change much. Moreover, with increasing feed pressure CH4 permeance decreased gently in most cases in accordance with the dual-mode sorption mechanism, causing the H2S/ CH4 ideal selectivity to reduce up to nearly 1.9 from its initial value.

Comparison of H2S/CH4 permeability, separation factor and ideal selectivity Figure 7 illustrates the effect of temperature on H2S and CH4 permeances at 689 kPa. Experiments at low concentration of H2S show that an increase in temperature from 295.65 to

123

Int. J. Environ. Sci. Technol. natural gas for PPO membrane, (c) different feed pressures, temperatures and H2S concentrations (low range) for PPO membrane

Table 3 Compositions and separation factors at (a) different feed pressures, temperatures and concentrations of H2S (high range) for PPO and PEUU membranes, (b) 689 kPa for the separation of real

PPO @ ambient temp. (293.15 K) H2S feed concentration

(a) 968 ppm

3048 ppm

5008 ppm

(vol.%: 0.0968)

(vol.%: 0.3048)

(vol.%: 0.5008)

Feed pressure (kPa)

Retentate H2S (vol.%)

Permeate H2S (vol.%)

Separation factor (H2S/CH4)

Retentate H2S (vol.%)

Permeate H2S (vol.%)

Separation factor (H2S/CH4)

Retentate H2S (vol.%)

Permeate H2S (vol.%)

Separation factor (H2S/CH4)

345

0.0472

0.1089

1.58

0.2033

0.6751

2.71

0.3509

1.1614

2.78

517

0.0363

0.1234

2.00

0.1706

0.6823

2.96

0.2976

0.9291

2.39

689

0.0327

0.1307

2.21

0.1161

0.6315

3.24

0.1713

0.9001

2.95

PPO @ T = 313.15 K 345

0.0537

0.0980

1.34

0.1815

0.6751

2.85

0.2323

0.9437

2.72

517

0.0389

0.1125

1.77

0.1524

0.6715

3.07

0.2105

0.9291

2.79

689a

0.0277

0.1331

2.41

0.0934

0.5991

3.37

0.1389

0.8154

2.91

PEUU @ ambient temp. (296.15 K) 345

_

_

_

_

_

_

0.3484

1.3584

3.25

517

_

_

_

_

_

_

0.3920

1.5100

3.43

689

_

_

_

_

_

_

0.3484

1.4028

3.37

Feed pressure (kPa)

Retentate H2S (vol.%)

@ ambient temp. (294.15 K) Feed H2S concentration 3360 ppm (vol.%: 0.3360) Permeate H2S (vol.%)

689

0.2361

0.9012

(b)

Separation factor (H2S/CH4) 3.20

@ ambient temp. (295.65 K) (c)

H2S feed concentration 101 ppm

198 ppm

401 ppm

(vol.%: 0.0101)

(vol.%: 0.0198)

(vol.%: 0.0401)

Feed pressure (kPa)

Retentate H2S (vol.%)

Permeate H2S (vol.%)

Separation factor (H2S/CH4)

Retentate H2S (vol.%)

Permeate H2S (vol.%)

Separation factor (H2S/CH4)

Retentate H2S (vol.%)

Permeate H2S (vol.%)

Separation factor (H2S/CH4)

345a

0.0093

0.0255

2.6

0.0180

0.053

2.8

0.0359

0.118

3.1

517a

0.0090

0.0317

3.3

0.0174

0.065

3.5

0.0350

0.139

3.7

689a

0.0088

0.0353

3.8

0.0169

0.074

4.0

0.0342

0.154

4.1

@ T = 313.15 K 689a a

0.0088

0.0358

3.8

0.0171

0.072

3.9

0.0342

0.152

4.1

Reproducibility of the experiment results

313.15 K results in 35–40% increase in permeabilities of both gas components at the feed concentrations’ range of 101–401 ppm. As mentioned before, hydrogen sulfide plasticizes the membrane, which seems to be more pronounced at higher temperatures. For high concentrations of H2S, it can be found that an increase in temperature from 293.15 to 313.15 K

123

results in approximately 20–25% increase in CH4 permeance whereas H2S permeance increases in the range of 19–38%. Comparison of the results for high and low H2S concentrations reveals that the trend of H2S as well as CH4 permeances is nearly the same at ambient temperature and 313.15 K over the whole range of H2S concentration.

Int. J. Environ. Sci. Technol.

Fig. 7 Effect of temperature on permeances of H2S and CH4 at different H2S feed concentrations for PPO membrane at 689 kPa

too. Also, for those H2S permeance curves that have no maximum, the ideal selectivity and separation factor remained nearly straight (for removal of low concentrations of H2S). Due to these peculiar permeability trends of H2S, the separation factor and selectivity of this component also follow the same trend. However, since CH4 permeance increases at the same rate with increasing temperature, the separation factor and ideal selectivity of this gas remained nearly constant. Table 3c summarizes the permeation of H2S with variations of feed pressure and H2S feed concentration for PPO membrane at low concentration of H2S. Similar data for removal of high concentration of H2S is already presented in Table 3a. These results indicate that the separation factors of PPO for separation of feeds containing low concentrations of H2S were slightly higher than the separation factors at higher feed concentrations of H2S. Hence, this membrane has a good performance in separation at both high and low H2S feed concentrations. On the other hand, for separation of feeds containing high concentrations of H2S a maximum point at 3048 ppm is observed, indicating an increasing trend from 968 to 3048 ppm, after which a decreasing trend is observed.

Conclusion

Fig. 8 Variation of separation factor and ideal selectivity with H2S feed concentration for PPO membrane at 689 kPa

The separation factor and ideal selectivity for both components are shown in Fig. 8. As can be seen from this figure, H2S feed concentration essentially did not affect the separation factor significantly. Because the permeance of both components at low concentration of H2S increased at the same rates with increasing temperature, the ideal selectivity and the separation factor remained nearly constant. In this range of H2S concentration, the selectivities of the membrane at lower temperature were slightly higher than those at high temperature. This figure also displays the separation factor and ideal selectivity for H2S and CH4 at high range of feed concentrations of 968–5008 ppm. It can be observed that the H2S permeance curves at different H2S and temperatures have special trends. For example, some of them have a maximum point and the others are straight lines. For those H2S permeance curves that have a maximum (for removal of high concentrations of H2S), the ideal selectivity and separation factor have a maximum point

The use of gas mixtures generally results in significant deviations from ideal transport behavior, due to competition effects between gases permeating through the membrane (multi-component effects). Mixed gas transport behavior in glassy polymer membranes will be highly dependent on the feed gas conditions. Plasticization effects seem to be counter-balanced by competitive sorption. The mixed gas separation performance of PPO and PEUU membranes was very sensitive to H2S plasticization and competitive sorption effects. The synthetic gas mixture and real gas separation experiments indicated that PPO membrane has the suitable characteristics for H2S/CH4 separations. This membrane exhibited high selectivity toward the synthetic gas mixture and real gas sample. A proposed process would be a hybrid system consisting of a membrane cascade containing PPO for bulk removal of H2S followed by a secondary process, such as amine absorption, for final cleaning of the gas stream. This work shows that decreasing H2S permeance as a function of pressure does not automatically mean that plasticization effects are absent. Therefore, it seems more appropriate to state that plasticization is the phenomenon of an enhanced diffusion coefficient of the component itself or other components due to swelling-induced local segmental motion caused by the presence of a plasticizing component

123

Int. J. Environ. Sci. Technol.

in the polymer matrix. The identification of plasticization, however, requires extensive experimentation and interpretation since it may be masked by competitive sorption effect. Acknowledgements The authors are grateful for the financial support provided by the Directorate for Research and Technology of the National Iranian Gas Company. They are also indebt to M. Mahdiyarfar and Y. Fazli of the Gas Research Division, Research Institute of Petroleum Industry, for their kind assistance in setting up the experimental apparatus. Moreover, the kind assistances of C. Y. Feng, K. C. Khulbe, L. Trembley, and F. Ziroldo of the Department of Chemical Engineering, the University of Ottawa, are acknowledged with thanks. The courtesy of Parker Filtration and Separation B. V. of the Netherlands for providing the membrane samples is highly appreciated.

List of symbols A [m2] l [m] p [Pa] (P/l) [mol m-2 s-1 Pa-1] Q [m3 (STP).s-1] T [K] ( xi )ln xi,F xi,R yi

Membrane permeation area Membrane thickness Pressure Permeance of gas Permeate flow rate Temperature Logarithmic average mole fraction of gas i in feed Mole fraction of gas i in feed Mole fraction of gas i in retentate Mole fraction of gas i in permeate

Greek symbols aij Separation factor of gas i over gas j aoij Ideal selectivity of gas i over gas j Subscripts F Feed side g Glass transition i Index of gas component i (here is H2S) j Index of gas component j (here is CH4) P Permeate side R Retentate side

References Aguilar-Vega M, Paul DR (1993) Gas transport properties of polyphenylene ethers. J Polym Sci Part B Polym Phys 31(11):1577–1589. doi:10.1002/polb.1993.090311114

123

Baker RW (2004) Membrane technology and applications, 2nd edn. Wiley, Hoboken Bhide BD, Stern SA (1993) Membrane processes for the removal of acid gases from natural gas. I. Process configurations and optimization of operating conditions. J Membr Sci 81(3):209–237. doi:10.1016/0376-7388(93)85175-V Chatterjee G, Houde AA, Stern SA (1997) Poly(ether urethane) and poly(ether urethane urea) membrane with high H2S/CH4 selectivity. J Membr Sci 135(1):99–106. doi:10.1016/S03767388(97)00134-8 Chou S (2003) Hydrogen sulfide: human health aspects. Concise International Chemical Assessment Document 53. Prepared for the World Health Organization. p. 6 Ettouney H, Majeed U (1997) Permeability functions for pure and mixture gases in silicone rubber and polysulfone membranes: dependence on pressure and composition. J Membr Sci 135(2):251–261. doi:10.1016/S0376-7388(97)00150-6 Ho WSW, Sirkar KK (1992) Membrane handbook, chapter 1. Springer, New York, pp 3–15 Koros WJ, Chern RT, Stannett V, Hopfenberg HB (1981) A model for permeation of mixed gases and vapors in glassy polymers. J Polym Sci Polym Phys Ed 19(10):1513–1530. doi:10.1002/pol. 1981.180191004 Mohammadi T, Tavakol Moghadam M, Saeidi M, Mahdyarfar M (2008) Acid gas permeation behavior through poly(ester urethane urea) membrane. J I&EC Res 47(19):7361–7367. doi:10.1021/ie071493k Paul DR, Yampol’skii YP (1994) Polymeric gas separation membranes. CRC Press, London Pourafshari Chenar M, Soltanieh M, Matsuura T, Tabe-Mohammadi A, Feng C (2006) Gas permeation properties of commercial polyphenylene oxide and Cardo-type polyimide hollow fiber membranes. J Sep Purif Technol 51(3):359–366. doi:10.1016/j. seppur.2006.02.018 Pourafshari Chenar M, Savoji H, Soltanieh M, Matsuura T, Tabe S (2011) Removal of hydrogen sulfide from methane using commercial polyphenylene oxide and Cardo-type polyimide hollow fiber membranes. Korean J Chem Eng 28(3):902–913. doi:10.1007/s11814-010-0437-7 Savoji H, Rana D, Matsuura T, Soltanieh M, Tabe S (2012) Novel surface modifying macromolecules (SMMs) blended polysulfone gas separation membranes by phase inversion technique. J Appl Polym Sci 124(3):2287–2299. doi:10.1002/app.34809 Vaughn JT, Koros WJ (2014) Analysis of feed stream acid gas concentration effects on the transport properties and separation performance of polymeric membranes for natural gas sweetening: a comparison between a glassy and rubbery polymer. J Membr Sci 465:107–116. doi:10.1016/j.memsci.2014.03.029 Wang R, Liu SL, Lin TT, Chung TS (2002) Characterization of hollow fiber membranes in a permeator using binary gas mixtures. J Chem Eng Sci 57(6):967–976. doi:10.1016/S00092509(01)00435-3 Wilks B, Rezac ME (2002) Properties of rubbery polymers for the recovery of hydrogen sulfide from gasification gases. J Appl Polym Sci 85(11):2436–2444. doi:10.1002/app.10881