(130a) Carbon Capture Enabled By High-Performance Membranes: Process Modelling and Economic Assessment

Membrane processes are emerging as an alternative to absorption for carbon capture. Membranes based on porous graphene selective layer have shown high promise in this regard with extremely large CO₂ permeance (10000 GPU) and attractive CO₂/N₂ selectivity (30) ^1,2. Further developments of graphene membranes in our laboratory have significantly increased the CO₂/N₂ selectivity up to 80, while still maintaining a large CO₂ permeance of 10000 GPU³.

In addition to that, important advancements in the scale-up process have allowed to realize for the first time large membrane coupons with area of 50 cm². Graphene membranes will be evaluated in a pilot plant for post-combustion carbon capture that is going to be installed in Aigle (Switzerland) with a capture rate of 10 kg/day⁴.

In order to design the process and to assess the economic feasibility at the large scale, we developed a techno-economic model for multi-stage membrane process based on graphene membranes.

Taking into account the performances of graphene membranes measured experimentally, we optimized a double-stage membrane process to target 90% CO₂ recovery and 95% CO₂ purity in the permeate stream. We assumed a specific membrane cost of 500 $/m² and an electricity cost of 0.05 $/kWh.

We compared two different pressure configurations for the first membrane stage: either working with compression of the feed and permeate under vacuum (FC+VP) or with ambient pressure in the feed and permeate under vacuum (VP)). Conversely, the second membrane stage is supposed to work always with the feed at ambient pressure and permeate under vacuum (VP). Compressing the feed of the second membrane stage is not needed, because the CO₂ concentration in the feed (that corresponds to the permeate of the first membrane stage) is around 50%, thus the driving force is much higher than in the first membrane stage.

When a post-combustion flue gas containing 13.5% CO₂ and saturated with water vapor at 50 ºC is fed to the capture system, the optimized values of capture penalty range from 34.0 $/ton to 28.9 $/ton when selectivity increases from 30 to 80, with a double-stage membrane system with FC+VP in the first stage and VP in the second stage. Conversely, the capture penalty decreases from 34.3 $/ton to 28.5 $/ton when selectivity increases from 30 to 80, with a double-stage membrane system with VP in both stages.

It is worth noting that the optimized capture penalties of the two membrane systems are very similar: this is because the higher energy consumption of the process with FC+VP in the first stage is counterbalanced by the lower membrane area with respect to the VP-VP process.

The increase of selectivity from 30 to 80 brings to a significant decrease of specific energy consumption, that reduces to 0.97 MJ/kg in the case of FC+VP (1^st stage) and VP (2^nd stage) and to 0.74 MJ/kg in the case of VP in both stages.

In all cases, thanks to the high permeance, the specific area is quite limited, as it ranges between 1200 m²/(kg/s) for the system with FC+VP in the 1^st stage and VP in the 2^nd stage and 2200 m²/(kg/s) for the system with VP in both stages.

Furthermore, with a prospective membrane cost of 100 $/m², the capture penalty reduces to a minimum of 18.7 $/ton for the system with VP in both stages (see Figure). This value is lower than the capture penalties reported in the literature for the same targets with polymeric membranes (from 25 to 45 $/ton), despite the higher specific membrane cost (assumed cost of polymeric membranes between 30 and 50 $/m²).

Finally, the graphene membranes developed in our laboratory have showed extremely large CO₂/N₂ selectivity values up to 1000 at low CO₂ concentration (0.5-2%), thus these could be particularly attractive for carbon capture from diluted sources ³. Estimated values of capture penalty from a diluted stream with CO₂ concentration of 1% range from 40 to 50 $/ton and are very promising for Direct Air Capture integrated multi-stage strategies.

Overall, these results show how optimized membrane processes based on graphene membranes are highly promising and competitive for carbon capture from various sources.

(1) Huang, S.; Li, S.; Villalobos, L. F.; Dakhchoune, M.; Micari, M.; Babu, D. J.; Vahdat, M. T.; Mensi, M.; Oveisi, E.; Agrawal, K. V. Millisecond Lattice Gasification for High-Density CO2- And O2-Sieving Nanopores in Single-Layer Graphene. Sci. Adv. 2021, 7 (9), 1–13.

(2) Hsu, K. J.; Villalobos, L. F.; Huang, S.; Chi, H. Y.; Dakhchoune, M.; Lee, W. C.; He, G.; Mensi, M.; Agrawal, K. V. Multipulsed Millisecond Ozone Gasification for Predictable Tuning of Nucleation and Nucleation-Decoupled Nanopore Expansion in Graphene for Carbon Capture. ACS Nano 2021, 15, 13230–13239.

(3) Hsu, K.-J.; Li, S.; Micari, M.; Chi, H.-Y. C.; Zhong, L.; Villalobos, L. F.; Huang, S.; Duan, X.; Züttel, A.; Agrawal, K. V. Pyridinic Nitrogen Substituted Two-Dimensional Pores for Rapid and Selective CO2 Transport. Under Rev.

(4) Agrawal, K. V.; Bautz, R. Capture Du Carbone Par Membranes En Graphène à Nanopores. Aqua gas 2022.