Membranes for pressureretarded osmosis

PRO is the most studied membrane technology exploiting a salinity gradient. However, the amount of experimental data is scarce and difficult to compare with each other. Metha and Loeb and recently Thorsen and Holt [20] are the only authors who published experimental determined power densities for PRO at real conditions. Some PRO values are determined from osmosis experiments without applying a hydrostatic pressure. Such a pressure can have a significant effect on the water flux, since a very open support structure allows a high water flux, but is also susceptible to compaction. Table 2 shows the available experimental results obtained from literature.

The reported power densities in Table 2 are derived from the reported flux and feed pressure. All the experimental results were obtained by Loeb and Metha from 1976 to 1982 and mainly for concentrated salt streams [8,9,14,15] and from Thorsen and Holt (2009) [20]. Loeb and Metha determined only one value (0.21 W/m2) obtained for a feed concentration of 30g/L NaCl, which represents seawater. Furthermore, it should be noted that the feed pressure is not chosen optimal [half of the osmotic pressure as indicated in Eq. (18)] necessary for a maximal power density. However, Thorsen and Holt [20] systematically varied the feed pressure and obtained for a NaCl concentration representing seawater the highest reported power densities 1.6 W/m for a cellulose acetate membrane and 2.7 W/m for a thin film composite membrane. Higher power densities (1.76-5 W/m2) can be obtained for concentrated brine streams. However, the performance of the fibers deteriorates when exposed to high salt concentrations, probably caused by a change in the porous substructure. Based on Table 2 no clear conclusion can be drawn for the optimal membrane properties for PRO, mostly because the experimental conditions are difficult to compare with each other because of different process condition: feed pressure, salt concentration, and flow rates (external concentration polarization).

Proper PRO experiments are difficult to perform: feed pressure chosen should be optimal, feed flow rate should be high in order to minimize concentration polarization, and the amount of permeated water should be determined accurately. Therefore, Lee et al. [16] generated a theoretical model in order to predict the PRO performance from osmosis experiment [parameters A and B in Eqs. (8) and (9)] and from direct osmosis (K derived from the osmotic flow) as input parameters for their model. The results of direct osmosis experiments might be too optimistic, since these experiments do not take compaction phenomena into account [16]. The results of Lee et al. [16] and Loeb et al. [11] are shown in Table 3.

All the membrane parameters from Table 3 are derived from experiments with NaCl solutions, except for the Toray CA-3000 values, which were determined with MgCl2. These latter values might be too optimistic since the parameters B and K depend on the type of salt. The asymmetric membranes show a higher performance compared to the composite membranes. These composite membranes have a lower projected power density due to their denser support layer, resulting in a high K value [16]. The retention and the resistance of the support layer play a crucial role in o oo

Table 2 Measured PRO performance for various membranes at different feed concentrations and pressures for the mixing with freshwater (0 g/L NaCl)

Du Pont permasep

B-10

Asymmetric polyamide fiber

0.056

2.62

Du Pont permasep

B-10

Asymmetric polyamide fiber

0.056

2.62

108

91

0.042

51

2.46

143

122

0.070

51

4.10a

191

162

0.084

51

4.90a

191

162

0.081

51

4.77a

96

81

0.038

41

1.78a

96

81

0.070

41

3.26b

96

81

0.081

20

1.90b

143

122

0.045

61

3.17b

FRL thin film

Composite

72

61

0.032

3

0.11

[9]

composite

polysulfone with furan

24

20

0.006

3

0.02

skin fiber

119

101

0.070

19

1.56

UOP CA/SW

Asymmetric

47

40

0.035

21

0.85

[15]

cellulose

acetate flat

94

80

0.090

21

2.14

sheet spiral wound

139

118

0.081

24

2.26

51

44

0.037

21

0.89

Du Pont

Asymmetric

239

203

0.100

30

3.52

[8]

permasep B-10

polyamide fiber

119

101

0.050

30

1.76

30

25

0.012

15

0.21

Osmonics SS10

Asymmetric cellulose acetate

23.5

16

0.390

8

1.60

[20]

Thin Film

Composite

30.6

21.5

0.229

12

2.70

[20]

Composite

a Change in performance when Pf—Pf > 50 bar. r b Different module. y a Change in performance when Pf—Pf > 50 bar. r b Different module. y o vo

Table 3 Calculated power densities (W/m2) for various membranes derived from osmosis experiments (A and B) and osmosis experiments (K)

Membrane

Type

Membrane parameters

Projected

Reference

A [m3/ (m2 day bar)]

(m/day)

K (day/m)

AP0/Ap

AP0/2 (bar)

Jw [m3/ (m2day)]

Wmax (W/ m2)

CA-80

Asymmetric

0.0088

0.173

0.75

0.88

12.60

0.109

1.59

[16]

CA-70

Asymmetric

0.0289

7.517

0.44

0.23

3.31

0.092

0.35

[16]

BM-05

Asymmetric

0.0035

0.020

21.99

0.70

9.90

0.028

0.32

[16]

PBIL

Asymmetric

0.0057

0.028

7.99

0.82

11.66

0.060

0.81

[16]

PA-300

Composite

0.0096

0.015

65.97

0.51

7.23

0.020

0.17

[16]

NS-101

Composite

0.0105

0.038

335.65

0.07

1.03

0.003

0.00

[16]

BM-1-C

Composite

0.0072

0.053

46.30

0.29

4.14

0.015

0.07

[16]

Toray CA-3000a

Asymmetric

0.0324

0.018

104.00

0.34

4.88

0.014

0.08a

[11]

Toray CA-3000a

Asymmetric without support fabric

0.0324

0.018

17.00

0.76

10.84

0.115

1.45a

[11]

a Values determined with MgCl2 too optimistic for NaCl.

a Values determined with MgCl2 too optimistic for NaCl.

the performance of a PRO membrane. As can be seen from Table 3 membranes with a low water permeability (A) but with a high selectivity toward salts (low B) and an open support structure (low K) can exhibit a high power density (CA-80 membrane). A membrane with a high selective top layer for salts and an open porous support allow a higher optimal pressure stemming from a larger osmotic pressure difference over the selective layer of the membrane.

Support layers have a tremendous effect on the performance of a PRO membrane. The support layer should be as thin and open as possible without a support fabric. Asymmetric fibers are very attractive for PRO since they possess a thin porous support layer and no support fabric. However, these open structures should also be able to withstand the hydrostatic pressure during PRO operation and should not compact. Compaction of the open CA-80 fiber is observed by Lee et al. [16] and is not taken into account in their model. Real PRO experiments would most likely show a lower power density as compared to the values reported in Table 3.

Summarizing the optimal PRO membrane should have the following characteristics:

• Low resistance in the porous support, very open or no support fabric [11] [low K, Eq. (13)]

• Hydrophilic porous support [19]

• Resistant against compaction

• Minimal external concentration polarization (high flow rates)

These parameters might be conflicting with each other and an optimal membrane is optimized with respect to these variables.

Was this article helpful?

0 0
Waste Management And Control

Waste Management And Control

Get All The Support And Guidance You Need To Be A Success At Understanding Waste Management. This Book Is One Of The Most Valuable Resources In The World When It Comes To The Truth about Environment, Waste and Landfills.

Get My Free Ebook


Post a comment