Waste heat recovery of an UAV propulsion system based on PEM fuel cell by a novel transcritical CO2 - LNG hybrid cycle; thermodynamic and multiple linear regression analyses

Abstract

ABSTRACT In this work, the performance of a proposed hybrid system consisting of a transcritical carbon dioxide-natural gas cycle for recovery of waste heat produced in a proton exchange membrane fuel cell has been evaluated. The effects of different parameters on proposed cycle performance have been investigated thermodynamically. This hybrid system can convert fuel cell’s released heat to electricity, augment production power, convert natural gas from the liquid phase to gas, and appropriate use of heat energy from consumed water. Results show that increasing the cell’s power by about 52% leads to a growth in overall efficiency of transcritical power cycle. Also, in a constant fuel cell power, the higher carbon dioxide mass flow rate tends to lower carbon dioxide turbine output power, and consequently, the efficiency of transcritical power cycle decreases. Changes in waste heat recovery efficiency among models with different values of CO2 mass flow rate increases very slightly as fuel cell power augments. For instance, at lowest examined fuel cell power (790 kW), growth in CO2 mass flow rate by about 26% causes a decline in cycle efficiency by about 6.35%. By increasing fuel cell thermal efficiency by 30%, improvement in overall cycle efficiency rises significantly by about 109.92%. On the other hand, augmentation in fuel cell thermal efficiency by about 30% leads to a very slight growth in improvement of the waste heat recovery efficiency by about 2.67%. Moreover, by increasing adiabatic efficiency of turbines by about 21.43%, improvement in waste heat recovery efficiency declines by about 10.75%. Increasing carbon dioxide turbine’s inlet temperature and reducing carbon dioxide mass flow rate have beneficial effects on efficiency of carbon dioxide transcritical power cycle. Eventually, two high-accuracy models have been presented for output power and efficiency of proposed hybrid cycle using multiple linear regression.