PSA or Membranes for Ammonia Cracking?

Ammonia is increasingly considered a practical hydrogen carrier because it contains approximately 17.8 wt% hydrogen, can be liquefied under moderate conditions and already benefits from an established global storage and transportation infrastructure. However, cracking ammonia does not directly produce hydrogen ready for use.

The outlet from an ammonia cracking reactor typically contains hydrogen, nitrogen and residual ammonia. For fuel cells and other high-purity applications, the separation stage can therefore be as important as the cracking catalyst itself.

There is no universally superior purification technology. Pressure swing adsorption remains a proven option for large, centralized hydrogen plants, while metal membranes may offer a simpler and more energy-efficient architecture for small, modular and distributed ammonia-to-hydrogen systems.

What Must Be Removed from Cracked Ammonia?

The main ammonia decomposition reaction produces hydrogen and nitrogen: 2NH₃ → N₂ + 3H₂

In practice, the reactor outlet also contains residual ammonia and may include water or other trace impurities, depending on the upstream system and operating conditions. These components must be reduced to levels appropriate for the intended hydrogen application.

The purification requirement is especially stringent for fuel cells. Where residual ammonia must be reduced to below 0.1 ppm, the system may need a dedicated temperature swing adsorption unit or guard bed before final hydrogen purification.

The technology choice must therefore consider more than hydrogen purity. Hydrogen recovery, delivery pressure, equipment footprint, operating mode, energy consumption and maintenance requirements are equally important.

How PSA Purifies Hydrogen

Pressure swing adsorption, or PSA, separates gases according to their different adsorption affinities. Hydrogen is weakly adsorbed, while nitrogen, ammonia, water and other impurities are retained by suitable adsorbents.

A PSA system operates through a repeated sequence of adsorption, pressure equalization, depressurization, purging, regeneration and repressurization. Multiple adsorption beds usually operate out of phase to maintain a relatively continuous hydrogen supply.

A properly designed PSA system can achieve hydrogen purity of up to 99.99%, with reported hydrogen recovery rates above 75%. Its maturity, established engineering practices and ability to process large gas volumes make it particularly attractive for centralized hydrogen production.

However, PSA involves more than one adsorption vessel. It requires switching valves, equalization lines, purge streams, regeneration procedures and coordinated control logic. At high throughput, this complexity can be justified. As hydrogen capacity decreases, these auxiliary components may account for a larger share of the system’s cost, footprint and energy consumption.

How Metal Membranes Differ

A hydrogen-selective membrane provides a continuous separation pathway rather than storing impurities temporarily in an adsorption bed.

In a dense metal membrane, hydrogen molecules dissociate into atomic hydrogen on the membrane surface. The hydrogen atoms diffuse through the metal lattice and recombine on the permeate side. This mechanism can provide exceptionally high hydrogen selectivity and, under controlled conditions, produce ultra-high-purity hydrogen.

Membrane modules can also be expanded by increasing the available membrane area instead of duplicating a complete cyclic adsorption train. This makes them attractive for applications such as:

• Distributed hydrogen production
• On-site ammonia cracking
• Fuel-cell power generation
• Remote or off-grid energy systems
• Backup power supply
• Modular ammonia-to-hydrogen plants

Membranes do not automatically guarantee lower capital or operating costs. Their performance depends on hydrogen permeance, selectivity, pressure ratio, membrane lifetime, sealing reliability, thermal stability and manufacturing cost. Palladium-based membranes, for example, can offer strong hydrogen selectivity but remain relatively expensive and require careful mechanical design.

Why System Scale Changes the Comparison

The relative value of PSA and membrane separation changes with plant capacity.

PSA benefits from scale because its beds, valves and control systems can process large gas volumes efficiently. A smaller system still requires many of the same operating functions, even though its hydrogen output is lower.

A membrane system is more straightforward to scale by adjusting membrane area. It operates continuously and may require fewer switching components. More importantly, it can separate hydrogen while retaining a substantial share of the upstream process pressure.

Pressure retention matters because purified hydrogen often needs to be delivered at pressure. If a purification system significantly depressurizes the gas, additional hydrogen compression may be required downstream, increasing both equipment count and energy consumption.

KAPSOM’s Modular Metal-Membrane System

Building on its ammonia cracking research, KAPSOM has developed a modular ammonia-to-hydrogen system using metal membrane separation.

In a project study at a hydrogen production capacity of 100 Nm³/h, the system demonstrated the ability to produce hydrogen with a purity above 99.999%, suitable for downstream fuel-cell power generation.

Compared with the conventional adsorption-based purification configuration evaluated in the study, the metal-membrane system showed an estimated 10–20% reduction in initial equipment investment and approximately 50% lower process energy consumption under the investigated operating conditions.

The energy reduction did not come from replacing one individual component. It resulted from simplifying the overall purification and pressure-management architecture. The membrane configuration reduced the need for cyclic pressurization, adsorbent regeneration and downstream hydrogen recompression.

The system uses separately optimized cracking and membrane-separation modules rather than placing the membrane directly inside the cracking reactor. This arrangement preserves the compactness and continuous operation of membrane separation while allowing the catalyst, membrane and auxiliary equipment to be maintained or replaced independently.

Which Technology Is the Better Choice?

PSA remains a strong option when the project requires high throughput, relies on established industrial operating practices and can economically support a multi-bed purification system.

Metal membranes may be more suitable when the priority is compact design, continuous operation, ultra-high hydrogen purity, pressure retention or modular capacity expansion.

The final decision should be based on the complete process rather than hydrogen purity alone. Feed composition, residual ammonia limits, hydrogen recovery, delivery pressure, operating hours, maintenance strategy and membrane replacement cost must all be evaluated.

For distributed ammonia cracking, the most effective solution may not be a scaled-down version of a large hydrogen plant. It may be a purification architecture designed specifically for smaller capacity, fewer auxiliary units and lower pressure loss.