Solid Oxide Fuel Cell and Gas Turbine Hybrid Cycles for Aerospace Power and Propulsion

Abstract

Hybrid propulsion systems combining gas turbine and solid oxide fuel cells (GT/SOFCs) have the potential to substantially reduce carbon emissions from 737-class aircraft. Many turbine/fuel cell hybrid cycles have been proposed for ground-based energy conversion at the utility scale, and some work has investigated small-scale (<500 kW) fuel cell-based energy conversion systems for aircraft (mostly auxiliary power units). However there is relatively little known about large hybridengine/fuel cell systems capable of providing main propulsive power in large (i.e. 737-class) aircraft. This work takes several important steps toward filling this gap.

First, it develops an analytical model of a GT/SOFC system that provides insight into the trends and tradeoffs associated with varying design parameters across a wide design space. Key insights that emerged from this modeling effort are: a)Increasing the fraction of fuel processed by the fuel cell always increases effciency. b) A tradeoff between fuel cell effciency and specific power determines the optimum range of the vehicle. This tradeoff is heavily influenced by the polarization curveof the SOFC. This optimum operating point is different from the maximum power point. c) The GT/SOFC could be used to increase the cycle’s flow specific work, enabling a smaller core to drive the same size fan. This premise is investigated in more detail later in the thesis. d) The fraction of fuel processed by the fuel cell is limited by the ability to cool it. An analytical expression for this limit is derived but in general the maximum power output of the fuel cell is limited to less than half of the total system power output for most hybridization schemes.

Second, this work develops an improved thermodynamic model of the hybrid turbine and fuel cell system. The model accounts for off-design performance of the turbomachinery as well as suffcient details of the transport and electrochemistryin the fuel cell to predict the effect of specific design changes (physical dimensions, flow rates, pressure, temperature, etc.) and operating conditions on power output, energy conversion effciency, and system mass. The model is implemented using a NASA-developed tool called Numerical Propulsion System Simulation (NPSS) that is emerging as a standard in modern engine development. While third-party NPSS fuel cell modules are available, they are not suitable for fuel cell design because key performance parameters like utilization, effciency, and specific power are inputs. Our module predicts fuel cell performance from its geometric attributes (channel length, width, height, number) and electrochemical attributes (i.e. temperature, pressure and composition effects on the polarization curve). Such capability is computationally expensive but essential for predicting GT/SOFC performance over varying flight conditions. This work implemented a) ’guardrails’ to prevent solver divergence due to self-reinforcing high or low temperatures, b) an adaptive Newtonsolver damping scheme to improve convergence, c) an electrochemical performance map to find close initial conditions, and d) the option for methane as an additional fuel, amongst other alterations. Taken together, these changes reduced execution time from weeks to hours and greatly improved stability making the thermodynamic model a much more useful tool for design and analysis.

Third, the NPSS system model is used to assess the viability of two possible hybridization schemes. The first is a ‘parallel’ hybrid system where an SOFC powers an electric motor that assists the turbine in driving the main fan. The second is a ‘turboelectric’ hybrid system where all of the propulsive power is provided electrically by a fuel cell working in tandem with a mechanical generator attached to the gas turbine. The results show that a parallel hybrid can reduce fuel consumption by 27%, but requires a reformer/fuel cell that achieves > 1kW/kg to achieve range parity with a conventionally-powered B737. This occurs because the thermodynamic effciency of the system increases by 10% and the propulsive effciency increases by 10% due to the higher bypass ratio made possible by the increase in flow specific work associated with hybridization. The turboelectric system reduces fuel consumption by 12% when 25% of power is generated by the SOFC, but requires a reformer/fuel cell that achieves > 1.2kW/kg to achieve range parity with a conventionally-powered B737. This higher specific power requirement occurs because the gas turbine operates at a lower OPR = 15 vs. OPR = 24 to enable recuperation via a heat exchanger. The heat exchanger also improves the thermodynamic performance of both the Brayton cycle and the SOFC (by reducing preheating requirements) even at 30% effectiveness, but adds mass and complexity.

Fourth, this work investigates the potential impacts of introducing the fuel cell exhaust—which is hot and contains large amounts of water and combustible reformate—on the Brayton cycle. The system modeling efforts show that the fuelcell exhaust can constitute up to 70% of the total mass flow rate through the system and up to 50% of the total net heat release. Therefore, the effect of the fuel cell exhaust on the operation of the main combustor is expected to be substantial both for integration with traditionally injected fuels, and influencing trades for the SOFC subsystem design choices that affect that exhaust (e.g. fuel utilization). Subsequent chemical kinetic simulations implemented in Cantera show that SOFC exhaust adiabatic flame temperatures can reach as high as 2200K, laminar flame speeds may vary by as much as 500% across a range of fuel utilization targets, ignition delay times with hydrocarbon/air mixtures can reach the millisecond range, and mixed SOFC exhaust can achieve extinction strain rates of over 300,000/s in pressures reasonable for gas turbines. These results suggest that aircraft GT/SOFCs may also require new combustor designs for effective hybridization.