Methane Selective Membranes for Nitrogen Removal from Low Quality Natural Gas – High Permeation Is Not Enough

Nitrogen is very difficult to remove from natural gas. Most specifications require no more than 4% nitrogen or total inerts in the interstate pipeline system. With about 15% of the raw natural gas produced in the United States having nitrogen contents over 4%, technology is needed to economically bring the gas to pipeline quality. Cryogenic distillation, the current state-of-the-art technology, is expensive, especially in smaller applications. Lean oil systems, such as AET’s Mehra Process, and pressure swing adsorption processes, such as Nitrotech’s carbon molecular sieve and Engelhard’s Molecular Gate, are being applied in very limited cases. Membranes are in the development stage for such applications. A development program based on a rubbery membrane that had very high methane permeation rates was planned. Unlike carbon dioxide/methane separation membranes, methane molecules permeate faster than nitrogen. Methane/nitrogen selectivity, the ratio of permeation rates, is about 3, compared to carbon dioxide/methane selectivity of 15-80 for glassy membranes. It was suggested that the very high permeation rates and resulting lower capital costs would offset compression requirements. A nonlinear program, based on Qi and Henson, was posed as the minimization of the total annual cost subject to constraints imposed by the separation requirements and operating conditions. We used the nonlinear program solver CONOPT within the Generic Algebraic Modeling System (GAMS) to solve the optimal design problems. The economic study results as the nitrogen content, selectivity, and several membrane configurations, including single stage, two-stage, two-stage with permeate recycle, twostage with residual recycle, and three-stage with permeate and residual recycle. Significant increase in selectivity is required before any configuration would be economic.