Carbon Dioxide Removal

Carbon dioxide is generated in the OCM reactor as an undesired by-product and it may be additionally fed to the reactor as a diluent. This CO2 amount must be completely removed from the main process stream before distillation.

Carbon dioxide forms an azeotrope with ethane and it can also freeze at the low temperatures required, causing major complications for distillation. The removal of CO₂ from industrial gas streams or natural gas sweetening is often performed by chemical absorption with amine solutions. While the

absorp-1.2 Oxidative Coupling of Methane (OCM)

CO2 REMOVAL DISTILLATIONDISTILLATION

Ethylene

Figure 1.4: Alternative integration options for the BG-OCM process

tion is ecient and industrially established, the regeneration/desorption step is highly energy demanding. Agbonghae et al., 2014 mentions that current re-search on post-combustion CO₂ capture by absorption focuses on developing new absorption uids, such as mixed amine solutions and ionic liquids, process synthesis and optimization, and integration with the process or energy plant.

Some alternative technologies recently in development include absorption with ionic liquids, distillation with controlled CO2 freezing, and pressure swing ad-sorption (Ruord et al., 2012).

Gas-Separation Membranes (GSM) can also be employed for selective gas separation. Glassy polymeric membranes, e.g., cellulose acetate and polyimides, have been successfully applied in the natural gas processing industry for medium capacity plants (6000 Nm³h⁻¹ to 50 000 Nm³h⁻¹) and for oshore or remote plants (Bernardo, Drioli, and Golemme, 2009), and also for biogas upgrade (Brinkmann et al., 2015). A few advantages of GSM are the straightforward installation, which make them ideal for a retrot; quick start-up times; simple control and operation; reduced equipment footprint; and lower environmental

ABSORBER

Figure 1.5: Process ow diagram for a hybrid membrane and absorption CO₂ removal process for OCM. CO2-rich streams are marked in red, lean amine streams are marked in light green, and rich amine streams are marked in dark green.

impacts (Bernardo, Drioli, and Golemme, 2009). The scale-up is also simple given that it happens mostly in terms of numbering up, although this may be disadvantageous in terms of equipment cost.

A specic concern for the CO2 separation within the OCM process is the ethylene slip/loss. While methane slip is also important in CO₂removal systems for natural gas sweetening or for biogas upgrading, ethylene is a much more valuable product. Taking this into consideration, a CO2 removal based solely on GSM would result in uneconomical ethylene losses. However, combining GSM with absorption in a hybrid approach could be a cost-eective solution for OCM. The process ow diagram for a hybrid process applying GSM and amine absorption to remove CO2 from OCM product gas is shown in Figure 1.5. Such hybrid approaches have already been proposed for other applications, e.g., biogas upgrade (Scholz et al., 2013).

Previous experiments carried out in the mini-plant facilities at the Technische Universität Berlin applied absorption & desorption with a benchmark 30 wt%

Monoethanolamine (IUPAC: 2-aminoethan-1-ol) (MEA) aqueous solution to remove CO₂ from a simulated OCM reactor outlet gas. In these experiments, up to 5.0 MJ kg_CO⁻¹₂ are consumed for the MEA regeneration and this could

1.2 Oxidative Coupling of Methane (OCM) be reduced to 3.7 MJ kg⁻¹_CO

2 by applying a mixture of 37 wt% N-Methyl Di-ethanolamine (MDEA) and 3 wt% Piperazine (PZ) (Stünkel et al., 2012). A at sheet envelope type Polyimide Membrane (PIM) module, which is selective for CO2, has then been added in the upstream of the absorption. This further reduced amine regeneration energy to2.8 MJ kg⁻¹_CO

2, but in order to remove 90%

of the inlet CO₂ amount, around 13% of the inlet C₂H₄ is lost (Stünkel et al., 2012). This is unacceptable for industrial purposes. To solve this issue, a cas-cade of membranes has been proposed, utilizing both PIM and Poly-(ethylene oxide) Membrane (PEOM) in stripping and rectication congurations. Mod-els have been developed and validated against mini-plant experimental data, allowing for simulation and optimization studies of dierent cascade congu-rations (Song et al., 2013). A superstructure optimization problem has been formulated aiming at the minimization of the energy consumption for the gas compression and amine regeneration (Esche, 2015). As a result, an energy con-sumption of 2.4 MJ kg⁻¹_CO

2 could be achieved with an ethylene loss of 8% by employing a stripping cascade of PIM and PEOM modules followed by absorp-tion with MDEA and PZ. However, if a lower ethylene loss is imposed as a constraint, the energy consumption is again increased, i.e., above2.8 MJ kg⁻¹_CO for a loss of 4%. 2

Overall, in these three previous studies carried out by the Group of Process Dynamics and Operations at the TUB, it has been concluded that a hybrid membrane-absorption process is technically feasible and can reduce the amine regeneration energy required for the CO2 removal section of the OCM process.

However, there is a trade-o between the energy savings and the ethylene loss.

Furthermore, the use of GSM may, in practice, require additional compression stages and increase the electricity consumption rate and also the equipment cost. Essentially no consideration regarding the capital investment cost is made on these previous studies. This means that the specic amine regeneration en-ergy required for the CO₂ removal cannot be the only objective or performance indicator. The trade-os need to be further evaluated on an economic basis to provide a balanced solution at industrial scale.

In this thesis, the CO₂ removal section of an industrial-scale BG-OCM plant is modeled and simulated. A superstructure is built to enable the simulation of both standalone absorption and hybrid process congurations and an op-timization is applied to determine which conguration and process conditions minimize the utility and equipment cost.