Novel Technology for Hydrogen Separation from Natural Gas using Pressure Swing Adsorption

Abstract

Australia has a high potential for production of renewable energy, such as wind and solar. Due to the stochastic operating conditions, excessively produced energy can be used to produce hydrogen by electrolysis to store the energy, known as power-to-gas. This hydrogen can be injected into the existing natural gas pipeline network, providing both storage and transport of hydrogen. There are many applications for hydrogen, however, this thesis focuses on the use of hydrogen for fuel cell cars. In order to use the hydrogen blended with the natural gas, a gas separation is required. Pressure swing adsorption is a commonly used technology to produce pure hydrogen, which exploits the adsorption of gases at high partial pressures. In this thesis, a PSA system is designed to separate a feed of 5 vol% and 10 vol% hydrogen mixed with natural gas at a pressure of 20 bar, and the economic feasibility of hydrogen supplied by a PSA system at a refuelling station is assessed and compared with other alternatives. The PSA separation is achieved with a 6 bed system, which consists of 4 pressure equalisation steps, to increase the product recovery, and repressurises the bed with the pure hydrogen product, to increase the purity. The adsorbent material is key in the design of a PSA system, which determines the operation performance and cost. Due to the large amount of gas components present in natural gas, a three-layered bed is designed. Activated carbon is selected as the main adsorbent layer, adsorbing methane, which is the main component in the gas mixture. Heavy hydrocarbons and CO2 adsorb very strongly on activated carbon; therefore, a pre-layer of silica gel is used to prevent accumulations of these gases. Silica gel has a linear isotherm for heavy hydrocarbons and CO2, which means the gas components will desorb at the desorption pressure. Lastly, a zeolite LiLSX layer is used for the adsorption of nitrogen. Process simulations are performed, focusing on the thickness of the pre-layer. No pre-layer results in accumulations of the heavy hydrocarbons on the activated carbon main layer, and thus reduces the available sites for methane to adsorb. This results in a low purity hydrogen product. When the pre-layer is too long, the total amount of activated carbon is reduced, and thus not enough adsorbent is available for the methane to adsorb. A thickness of 0.2 meter in a bed of 1.2 meter height is concluded to be ideal. It is concluded that an economically feasible design for a refuelling station with hydrogen supplied by a PSA system is proposed. Hydrogen can be dispensed to a fuel cell vehicle in the best case scenario for $14.79 with hydrogen originally produced by electrolysis, and for $12.14 for hydrogen originally produced by SMR without CCS. The final hydrogen price (including hydrogen supply, compression, storage, and dispensing) is compared to two other hydrogen supply methods: on-site electrolysis and tube-trailer transported hydrogen. Currently, PSA supplied hydrogen is a more economical option, especially if the hydrogen is produced from fossil fuel based resources. On-site electrolysis can become a more economical option in the future with improved cell efficiencies and reduced electricity prices. Tube-trailer transported hydrogen is highly influenced by the distance travelled. If the hydrogen originates from electrolysis, tube-trailer transported hydrogen will always be more expensive. For different fossil fuel based hydrogen technologies, a break-even distance is calculated.