Carbon natural gas storage performance as predicted by multiscale modeling

Abstract

Adsorbed natural gas (ANG) technology could help to facilitate the transition to environmentally friendly automobiles. Limits on total energy storage and the accumulation of higher hydrocarbons (C3+) present in natural gas are factors that restrict ANG broad application. Because experimental bed deactivation studies are extremely time-consuming, we propose a multiscale technique to study tank deactivation involving a mathematical model capable of predicting compositions of ANG tank cycles. The model is based on multicomponent ideal adsorption solution theory calculation and molecular simulation of the monocomponent adsorption equilibrium data using the representative pore method. We expect that this approach will save experimental effort and bring new insight to understanding the property-structure relation in activated carbons. The technique is validated using literature experimental data from tanks filled with activated carbons. Two target limits for the charge and discharge pressures (35–1 bar and 65–5.8 bar) are tested. The pressure range of 65–5.8 bar shows the best energy delivery for the three commercial carbons examined. The simulation shows that the amount of energy delivered is maximized for a pore size of 18.5 Å, larger than the 11.5 Å pore previously suggested for storage of pure methane. Maxsorb activated carbon showed similar performance to the metal-organic framework (MOF) NU-125 after approximately 100 charge and discharge cycles. In general, activated carbons are less expensive and more stable than MOFs. By accelerating bed deactivation assessment, the technique will allow the synthesis community to produce more efficient carbonaceous materials and facilitate ANG technology feasibility.