Carbon Membranes for Natural Gas Sweetening

As previously mentioned, carbon membranes can be produced by carbonization/ pyrolysis of a variety of carbon-containing materials, mostly under vacuum, inert or oxidative atmosphere [112]. Recently, numerous synthetic precursors have been used to form carbon membranes, such as polyimide and its derivatives, polyacrylonitrile, phenolic resin, polyfurfuryl alcohol, polyvinylidene chloride-acrylate terpolymer, phenol, formaldehyde and cellulose [45]. Among them, the polyimides are one of the most promising candidates for carbon membrane production for C02 separation [113-115]. The developmental membranes, produced by pyrolysis of polyimide precursors, are characterized by stable nanoporous structures that exhibit satisfactory permeability and selectivity properties [91,116-119].

In order to be applicable to an industrial process the membrane must possess two properties: high selectivity and high production capacity. To satisfy these criteria, both the membrane structure and the membrane module must have a high permeability and selectivity performance. An asymmetric membrane structure and the hollow fiber configuration of the membrane module are the two characteristics that provide the best combination of these two targets. Asymmetric membranes can be produced via a phase-inversion process, which can be adapted to produce membranes in the form of thin tubes or fibers.

An important advantage of hollow fiber membranes is that compact modules with a very high membrane surface area can be formed. However, this advantage is offset by the generally lower fluxes of hollow fiber membranes compared with flat-sheet membranes made from the same material [26]. Nonetheless, the development of hollow fiber membranes by Mahon and the group at Dow Chemical in 1966 [120] and their later commercialization by Dow, Monsanto, Du Pont and others, represents a major event in membrane technology. Of all possible membrane configurations, hollow fibers are the most promising for gas separation applications.

The properties of derivative carbon materials strongly correlate with the structural and selectivity properties of precursor polymeric membranes. To this end a good choice of selected precursor is the first step. For any candidate carbon membrane polymeric material, there are four major properties that need to be satisfied: (1) a high glass transmission temperature, Tg, (2) high aromatic content, (3) chemical stability and (4) high separation properties [61].

Especially in the case of the NG sweetening, the carbon membranes require high mechanical and chemical stability, since the operating pressure is high and the composition of the NG can create oxidative conditions. These requirements are the main reason that carbon membranes are not yet commercially available for NG sweetening processes, but also only a small number of scientific papers are available in this field.

Table 6.5 summarizes some of the major polymer types that have been used as precursor materials for carbon membranes and that are potentially applicable in NG sweetening processes, emphasizing the polyimide materials that are the most-studied precursor materials for producing carbon membranes for C02 separations. Numerous carbon membranes, both in flat sheet and hollow fiber configurations, have been reported as candidate materials for C02 separation/removal from NG streams. However, no carbon membrane module able to be installed in real industrial C02 separation plants is yet available, and this is a major priority of the industry, with many projects focused on this target worldwide.

H2S is an extremely toxic component of NG that corrodes pipes and engines and must be removed from the NG prior to use. As shown in Table 6.5, carbon membranes are promising for C02 removal, but few data have been reported for H2S removal from NG using carbon membranes. Current technologies for H2S removal include absorption, adsorption, conversion of H2S into elemental sulfur and a membrane reactor for H2S decomposition and desulfurization. In addition, a hollow fiber membrane contactor has been in the limelight due to its potential to overcome problems such as foaming and

TABLE 6.5

Carbon membranes for acid gas (CO,) removal from NG.

Precursor material

co2

permeability

(barrer)

co2/ch4

selectivity

Heat-treated

(°C)

Membr.

Config.*

Ref./Year

Polyimides (PI)

Various PI

88“'

190

500 and 550

HF

[121] /1994

Matrimid® and Kapton® polyimides

12,92

33, 16

475-700

FS

[122]/ 1999

BTDA-TDI/MDI (P84) copolyimide

0.5(i

51

900

HF

[123]/2000

6FDA/BPDA-DAM

80“'

80

800

HF

[124]/2003

6FDA/BPDA-DAM and Matrimid 5218

30 and 13

73 and 83

800

HF

[125]/2002

BTDA-TDI/MDI (P84) copolyimide

1808,738, 499

22,37,89

550,650 and 800

FS

[lll]/2004

Matrimid® and P84 polyimides

191,611

169,61

800

FS

[97]/2004

P84 polyimide

1.25“'

51.4

800

HF

[126]/2008

PIM-6FDA-OH

5040,556

38,93

630 and 800

FS

[127]/2013

Polyimide of intrinsic microporosity (PIM-6FDA-OH)

557 and 471

15 and 59

440 and 800

FS

[128]/2013

Polyetherimide (PEI)

Polyetherimide

0.15“»

155

600

FS

[129]/1999

PEI/polyvinylpyrrolidone (PVP)

1.66“'

55

700

HF

[130]/2012

PEI/polyvinylpyrrolidone (PVP)

0.69“'

69

800

HF

[131]/2012

Polyetherimide

0.183““'

6.5

600

FS

[132]/2013

Other (PSF/resins)

Phenolic resin

24.5“v'

160

800

FS

[133]/1999

Poly(furfurylalcohol) (PFFA) resin

0.035“»

36.8

850

FS

[134]/1998

PSF-beta/matrimid

12.6“'

150

800

HF

[135]/2007

Poly(benzimidazole) (PBI)

0.16

88.88

800

FS

[136]/2009

PBI/matrimid

36.6

131.65

800

FS

[136]/2009

Cellulose acetate

60

110

650

HF

[137]/2009

Cellulose (pilot scale report)

0.11-0.045,v'

104.2-66.7

HF

[138]/2018

Cellulose

200

650

HF

[139] /2019

FS: flat sheet, HF: hollow fiber, (i) permeance in GPU, (ii) (cm3/cm2-psi,min), (iii) m3/m+-bar-h, (iv) ((mol/m2-s-Pa) x 10'10), (v) m3 (STP)/(m2-h-bar).

loading [140]. Another technique for H2S treatment is microbiological treatment of the H2S-containing gases [141]. In membrane technology the H2S is removed by two main methods: (1) by using the membrane as a barrier between the liquid and the gas phase where it allows the transport of H2S from the gas phase to the liquid phase. In this technology polymeric materials, such as polysulfones, polypropylene, polyvinylidene fluoride [142,143] etc., are used; and (2) by using catalytic membrane reactors.

The thermal decomposition of H2S into hydrogen and sulfur takes place in the catalytic membrane reactor. Since H2S decomposition at high temperatures is attributed to the required high energy and endothermicity of the reaction, the use of a membrane reactor to selectively move the reaction products from the reaction zone in order to shift the equilibrium reaction forward sets the conditions for the application of a lower process temperature [140]. In this process glass and inorganic membranes have been used, but not carbon membranes. Specifically, membranes made of Vycor glass [144], alumina, microporous zirconia-silica [145], amorphous silica [146], metallic (Pd) [147] and composite metal containing Pt [148] have been used in catalytic H2S decomposition reactions. Furthermore, typical H2S desulfurization into elemental sulfur and water via catalytic oxidation is moving from fixed-bed reactor technology to catalytic membrane technology. Carbon membranes with the requested membrane properties, and porous membranes (a-Al203) impregnated with catalyst, have also been tested in this kind of process [149].

 
Source
< Prev   CONTENTS   Source   Next >