Fuel Cell Microturbine Hybrid System Analysis Through Different Uncertainty Quantification Methods

The analysis of different energy systems has shown various sources of variability and uncertainty; hence the necessity to quantify and take these into account is becoming more and more important. In this paper, a steady state, off-design model of a solid oxide fuel cell and turbocharger hybrid system with recuperator has been developed. Performances of such stiff systems are affected significantly by uncertainties both in component performance and operating parameters. This work started with the application of Monte Carlo Simulation method, as a reference sampling method, and then compared it with two different approximated methods. The first one is the Response Sensitivity Analysis, based on Taylor series expansion, and the latter is the Polynomial Chaos, based on a linear combination of different polynomials. These are non-intrusive methods, thus the model is treated as a black-box, with the uncertainty propagation method staying at an upper level. The work is focused on the application on highly non-linear complex systems, such as the hybrid systems, without any optimization process included. Hence, only the uncertainty propagation is considered. Uncertainties in the fuel utilization, ohmic resistance of the fuel cell, and efficiency of the recuperator are taken into account. In particular, their effects on fuel cell lifetime and some simple economic parameters are evaluated. The analysis distinguishes the specific features of each approach and identifies the strongest influencing inputs to the monitored output. Both approximated methods allow an important reduction in the number of evaluations while maintaining a good accuracy compared to Monte Carlo Simulation.