The performance of carbon membranes in the presence of condensable and non-condensable impurities

https://doi.org/10.1016/j.memsci.2011.04.058Get rights and content

Abstract

To fully assess the suitability of nanoporous carbon (NPC) membranes for industrial applications such as carbon capture, it is necessary to understand the impact of impurities commonly present in the feed streams upon the membrane performance. In this work, the effect on the performance of a NPC membrane upon exposure to condensable impurities typically found in natural gas (e.g. water, hexane and toluene) and non-condensable impurities typically found in synthesis gas (e.g. H2S and CO) has been determined in laboratory experiments. Small reductions in the permeance (less than 30%) and minimal reductions in selectivity have been observed, with the greatest impact at 35 °C and less impact at 100 °C. The performance of a NPC membrane upon exposure to real synthesis gas produced from air-blown coal gasification, as part of the Mulgrave Capture Project, has also been determined. The combined higher concentrations of impurities in these pilot plant experiments resulted in a greater impact on performance at 100 °C with a reduction in CO2 permeance of around 40% and in CO2/N2 selectivity of around 25%. However, these relatively limited reductions in membrane performance suggest that such carbon membranes still offer some promise for pre-combustion applications.

Highlights

► Nanoporous carbon membranes were exposed to water, hexane, toluene, CO and H2S. ► Laboratory tests showed a reduction in permeance of <30% and a minimal loss of selectivity. ► Pilot plant tests using a synthesis gas show slightly larger reductions in performance. ► Carbon membranes show promise for pre-combustion CO2 separation.

Introduction

Nanoporous carbon (NPC) membranes, made from the pyrolysis of polymers at high temperatures, have shown high selectivities when tested in the laboratory [1], [2], [3], [4], [5], [6], which gives them promise as a new generation of gas separation membranes. In order to further assess the suitability of NPC membranes for industrial applications, it is necessary to test the performance of these membranes in the presence of impurities that commonly exist in industrial feed gases.

The focus of this research is in assessing NPC membranes for the separation of carbon dioxide (CO2) as part of the carbon capture and geological sequestration process for reducing greenhouse gas emissions [7]. As such, the impurities tested in this work relate to (1) CO2 separation from methane (CH4) as part of CO2 capture from natural gas and (2) CO2 separation from nitrogen (N2) as part of CO2 capture from air-blown synthesis gas production. The condensable impurities studied include water and hydrocarbons. The non-condensable impurities studied were hydrogen sulphide (H2S) and carbon monoxide (CO).

Water is in equilibrium with natural gas when produced from a reservoir. Similarly, water will be present in synthesis gas production because of the water gas shift reaction. Due to problems when operating membrane processes with such a humid feed, gas streams are often dehydrated using established technologies such as glycol dehydration or a solid desiccant [8]. If the water tolerance of the membrane process can be increased by using materials such as nanoporous carbon instead of polymers, this will decrease the amount of pre-treatment required and therefore the cost.

To specifically address the impact of water vapour on the performance of NPC membranes, several studies have used a humidified feed [9], [10], [11], [12]. These studies all conclude that flux or permeance is decreased with increasing humidity. At lower humidity levels, the water is adsorbed onto the surface of the carbon. At higher humidity levels, the water fills the pores of the surface leading to a significant reduction in the adsorption of other gases.

Condensable hydrocarbons are an omnipresent part of the natural gas processing system. Even reservoirs which are largely comprised of methane will have some heavier and aromatic hydrocarbons (usually termed “gas condensate”). The majority of the gas condensate will condense out in one or a series of upstream pressure separation vessels but there will invariably be some small carry-over into the gas stream. As such, gas processing facilities using the traditional polymeric membranes are equipped with expensive pre-treatment equipment to remove these condensable impurities. Synthesis gas can also contain hydrocarbons due to byproduct reactions in the gasifier. At the research level, hexane (C6H14) or heptane (C7H16) is commonly used to simulate the impact of heavier hydrocarbons whilst toluene (C7H8) is used to simulate the impact of aromatic hydrocarbons.

A comparison between the effect of toluene exposure on polymeric polyimide membranes compared with NPC membranes also made from polyimide, revealed that the drop in permeance and selectivity is more significant for the polymeric membrane due to the irreversible compaction of the polymeric membrane caused by plasticisation [13]. Conversely, in other studies of exposing NPC membranes made from polyimide to trace amounts of hexane and toluene, larger reductions in the permeances and slight reductions in selectivity were observed [14], [15]. However, these performance reductions were reversed upon regenerating the NPC membranes using pure propylene [14] or high temperature nitrogen [15] suggesting that the impurities are only physically adsorbed to the nanoporous carbon.

Non-condensable impurities are impurities that remain in a gaseous form upon contact with the NPC membrane and may impact on the performance of the membrane by physical adsorption and pore blocking or by chemically reacting with the carbon. In the case of CO2 removal from natural gas, hydrogen sulphide (H2S) is present due to sulphide producing bacteria present in the reservoir. Similarly, H2S is present in syngas from the sulphur producing bacteria present in the original hydrocarbon source (such as coal). Another component present in synthesis gas is carbon monoxide (CO), which is known for poisoning catalyst materials due to its reactivity.

Traditional polymeric membranes are susceptible to plasticisation upon prolonged exposure to H2S [16]. It has also been shown that H2S reduces the performance of palladium-based membranes, which are traditionally used in synthesis gas purification, due to the formation of surface sulphides that block H2 adsorption sites [17]. If NPC membranes could be shown to be resistant to H2S whilst also removing some H2S into the CO2 stream for sequestration, savings could be made in terms of the H2S removal systems required.

Whilst there is little known specifically of the effect of H2S and CO on the performance of NPC membranes, there have been several studies on the adsorption of these gases on activated carbon. Adsorption isotherms of various gases on activated carbon revealed that the volume of gas adsorbed is as follows H2S > CO2 > CH4 > CO > H2, with all gases following the Langmuir model of adsorption [18]. Likewise, for dual gas experiments, CO2 adsorbed more than CO, H2S adsorbed more than CO2 and CO2 adsorbed more than CH4 [18]. It has also been shown that when H2S is reacted with adsorbed oxygen upon activated carbon, elemental sulphur is formed, which in the presence of water can form sulphuric acid [19].

The first objective of this research is to test the performance of NPC membranes in the laboratory in the presence of water, hexane and toluene at concentrations higher than those previously reported. The second objective is to report on the laboratory performance of these membranes in the presence of the non-condensable impurities—H2S and CO. Finally, the membranes were tested in a pre-combustion pilot plant using a real synthesis gas manufactured from the air-blown gasification of Australian brown coal to evaluate how a mixture of H2S, CO, water and trace hydrocarbons affects performance.

Section snippets

Experimental

The NPC membranes used in this research work were made from the pyrolysis of PFA at 550 °C under a flow of ultra high purity argon. The details of the manufacturing technique are described elsewhere [7].

Baseline performance

The baseline performances of the membranes used to test the impurities at 35 °C and at 100 °C are presented in Table 3. The CO2 permeance values are comparable to our previously published work, but the selectivity of the membranes is slightly better than previously published [7] due to improvements in our expertise at making these membranes with fewer defects.

Condensable impurities

As shown in Fig. 2(a), there is an initial drop in the CO2 permeance as the membrane is exposed to the water impurity. A similar trend is

Conclusions

The presence of condensable and non-condensable impurities on the performance of supported NPC membranes tested in the laboratory is negligible when concentrations are low, particularly at elevated temperatures. The pre-combustion pilot plant trials, using high concentrations of impurities also indicated a relatively minor impact of impurities on membrane performance, with permeance losses of 40% and selectivity losses of 25%. This is a very encouraging result for the use of NPC membranes for

Acknowledgements

The authors would like to acknowledge the funding provided by the Australian Government through its CRC Program and by the Victorian State Government as part of ETIS to support this research. Infrastructure support from the Particulate Fluids Processing Centre, a special research centre of the Australian Research Council is also gratefully acknowledged.

Cited by (29)

  • Gas separation performance of copolymers of perfluoro(butenyl vinyl ether) and perfluoro(2,2-dimethyl-1,3-dioxole)

    2021, Journal of Membrane Science
    Citation Excerpt :

    Gas permeability was calculated based on the steady state increase in the downstream pressure [31]. Mixed gas permeability measurements were conducted at 35 °C using a constant pressure-variable volume apparatus previously reported [32]. Helium was used as a sweep gas at an absolute pressure of 1 bar and a flow rate of 34 cm3(STP)/min.

  • Regenerated cellulose based carbon membranes for CO<inf>2</inf> separation: Durability and aging under miscellaneous environments

    2019, Journal of Industrial and Engineering Chemistry
    Citation Excerpt :

    CHF membranes need to be regenerated during separation process and this extra step adds complexity and cost to the process. Anderson et al. [12] have reported the effect of H2S on polyfurfuryl alcohol based carbon membranes. The PFA based carbon membranes showed that CO2 permeance was reduced by 7% in the presence of H2S (partial pressure: 0.3 kPa) in the feed when membrane was performing at 35 °C for almost 5 h.

View all citing articles on Scopus
View full text