(674c) Gas-Phase Simulated Moving Bed for Methane/Nitrogen Separation Using a Commercial Activated Carbon

With the rising demand of natural gas, the exploration of unconventional methane-rich sources has been gradually increasing over the years due to the depletion of the conventional wells. These sources are often contaminated with several impurities, among which nitrogen is the most difficult to be removed, owing to the similar properties to those of methane. When its content is above 3%, its removal becomes a necessity to ensure that the calorific value of the stream is not affected by the presence of this inert gas. The main technology used for this purpose is cryogenic distillation, which has very high energy costs associated and is unsuitable for medium and small-scale purification of methane streams. When compared to this technology, adsorption-based processes are very appealing thanks to their easy scalability and lower energy requirements. However, the performance of the solid materials for CH₄/N₂ separation becomes an issue due to their low selectivity. The use of a gas-phase simulated moving bed (SMB) process overcomes this challenge since it can separate mixtures with selectivities close to unity, by introducing a desorbent species to displace the adsorbed phase, essentially transforming a difficult separation into two easier ones.

The design and development of a gas-SMB process for the separation of methane and nitrogen mixtures is the main focus of this work, using a commercial activated carbon (BPL) as the adsorbent material and two potential desorbent gases – argon and carbon dioxide. The performance of the material was evaluated by measuring the adsorption equilibrium data for the four adsorbates and the dynamic behavior of single and multicomponent adsorption through fixed-bed experiments.

The pure component isotherms of N₂, CH₄, Ar, and CO₂, presented in Figure 1, were measured at temperatures of 303, 323, and 343 K in a pressure range of 0 - 2.5 bar using a volumetric apparatus. The data was successfully regressed with the Dual-Site Langmuir (DSL) model. The results show that CO₂ has the highest affinity to the stationary phase and Ar the lowest, over the entire range of temperature and pressure studied.

Single, binary, and ternary breakthrough experiments were carried out at 303 K and 1.5 bar. A mathematical model that encompasses mass, energy, and momentum balances was implemented in the gPROMS® software and utilized to predict the observed adsorption dynamics. Analysis of the ternary fixed bed experiments (Figure 2) show that the desorbent is easily displaced by the mixture and is equally able to displace the adsorbed phase, which is an essential step for SMB operation. The simulation results are in good agreement with the experimental data.

Finally, two simulated moving bed (SMB) cycles were employed to separate an equimolar CH₄/N₂ mixture using each of the desorbent gases to evaluate the impact of the desorbent strength in the process. The experiments were carried out in open loop, with no desorbent recycle. A 2-3-2-1 four-zone system configuration was considered, with a switching time of 50 s. The SMB unit was operated at 303 K, with the backpressure, located at the end of zone IV, set at 1.5 bar. A feed composition of 50% CH₄ and 50% N₂ was used, with a feed stream of 0.190 SLPM for the argon experiment and 0.060 SLPM for the CO₂ experiment. The internal profiles for each case are presented in Figure 3. In zone II, the more adsorbed species, methane, is retained onto the material, enriching the extract stream, while the less adsorbed species, nitrogen, is carried by the eluent. In section III, the nitrogen content increases as it moves with the fluid phase to the raffinate port. The flow rate in section I is high enough to guarantee that the adsorbent is cleaned, ensuring that section IV is not contaminated after the switch. In section IV, the nitrogen content decreases, with pure desorbent being obtained at the outlet, which can be recycled back into section I when operating in closed loop. Both cycles were capable of producing a high purity methane stream (96.2% and 97.4% for the Argon and CO₂ experiment, respectively) with a high recovery (> 92%). When argon is used as the desorbent gas, the extract product stream is obtained with a productivity of 13.9 kg·m^-3_ads·h^‑1.