Computational Model for Electrochemical Gas Separation and Inerting System

Abstract

In July 1996, TWA flight 800 crashed into the Atlantic Ocean only 12 minutes after takeoff killing all 230 people on board. After a 4-year investigation into this catastrophe, the National Transportation Safety Board concluded that the accident was probably caused by the explosion of flammable fuel vapor in the Boeing 747’s center fuel tank. Following this landmark incident new regulations for aircraft fuel tank inerting were developed to prevent such explosions. Fuel tank inerting is the process of reducing the flammability of the fuel vapor by restricting the oxygen concentration in the ullage (air volume above the fuel) to a safe value – 12% for commercial aircraft and 9% for military aircraft. Generally, inerting is carried out by driving oxygen out the ullage using an inert gas like nitrogen.Existing inerting technologies include systems like liquid nitrogen, explosion suppressant foam, halon extinguishment, and on-board inert gas generation systems (OBIGGS). Compared to other techniques, OBIGGS is seen as viable option since nitrogen enriched air (NEA) can be produced during flight by supplying pressurized air into separation modules that preferentially retain nitrogen while allowing oxygen to escape. However, such systems are expensive and consume substantial amounts of energy, while also undergoing premature failure. Additionally, the NEA generation rate is fixed in such systems, whereas the actual NEA requirement depends on the aircraft’s flight stage. Thus, incumbent technologies are designed based on the highest demand making them overweight and expensive.Here, we propose a novel on-board electrochemical gas separation and inerting system (EGSIS) to overcome such problems. EGSIS is an electrically powered system that can be used to produce and supply NEA to the fuel tank. It combines a polymer electrolyte membrane (PEM) electrolyzer anode wherein water is dissociated to generate oxygen, and a PEM fuel cell cathode where oxygen from the inlet air inlet is reduced to produce humidified NEA as the output gas. After proper dehumidification, the NEA can be supplied directly to the ullage to reduce its oxygen concentration. Unlike current technologies, the NEA production rate in EGSIS can be easily varied with demand just by controlling the voltage applied to the system. Experimental investigations of EGSIS have been previously conducted by our group. This paper represents the first attempt to model and simulate EGSIS.We present a two-dimensional mathematical model of a single cell EGSIS that provides a foundation for the simulation and analysis of a full-scale EGSIS module. A single phase, two-dimensional, steady state, isothermal model is implemented in an EGSIS cell using COMSOL. Various EGSIS performance parameters such as current densities, reactant concentration distributions, and polarization curves are investigated as a function of operating voltage and temperature. The model involves the material properties of the various components including the polymer electrolyte membrane, the catalyst later and gas diffusion media; the transport of water vapor and air in the anode and cathode flow field channels, respectively; the transport of protons and water in the catalyst layer and the membrane; and the transport of electrical charge in the solid phase. The governing equations include continuity and steady-state incompressible Navier-Stokes equations. Additionally, species transport is modeled using the generalized Stefan-Maxwell equations whereas the Butler-Volmer equation is employed for charge transport. All of these coupled model equations are solved by the finite element method using the COMSOL Multi physics software package. The results from the computational model are then compared and validated with our experimental results for various operating conditions. Different operating strategies are explored with the goal of improving system performance. The simulated results reveal the influence of temperature, reaction flowrate, and optimized material properties on EGSIS performance. This model also provides a base for future research into the effect of alternate catalysts and gas diffusion media. It also serves as a foundation for the analysis of a practical EGSIS stack.