Impact of Uncertainties on the Climate Optimized Aircraft Design

Abstract

This report covers the investigation of impact of uncertainties on the multidisciplinary design optimization of a medium-range single-aisle turbofan aircraft for minimum global warming impact. The employed workflow for the investigation is a five step process, starting with the implementation of the deterministic climate impact model and carrying out of the design optimization for minimal climate impact. The second step involves the characterisation of uncertainties, where the uncertainties within the climate impact model are identified and quantified. The third step involves the uncertainty analysis, where Monte Carlo simulations are performed to estimate the variability in the average temperature reduction potential of the climate-optimized aircraft with respect to the cost-optimized aircraft. In the fourth step, a robust design optimization is carried out using a non-sorted genetic algorithm to minimize both the average temperature response and variability in average temperature response potential. The sensitivity analysis is carried as the last step using the Morris and variance-based Sobol methods, to identify what the key uncertain parameters are towards the uncertainty in climate impact of the aircraft designs. Scientific uncertainty is identified within the linear climate impact model for the carbon impulse response function parameters, species radiative efficiencies, the NOx and contrail altitude forcing factors, methane lifetime, species efficacies, and are all assigned a probabilistic description. Scenario uncertainty is identified in the future average global CO2 atmospheric concentration projection, for which different realistic future scenarios are characterised. Carrying out the uncertainty analysis has shown that the average temperature response reduction potential of the climate-optimized aircraft is highly uncertain, having a 90% likelihood ranging between 17 and 98 % of the average temperature response of the cost optimized aircraft. This is primarily due to large variability in the estimation of contrail average temperature response. Although the robustness-based optimization did not allow to find any significant improvement in robustness for the climate-optimized aircraft, it did allow to identify an array of robust climate-optimized design solutions. From the sensitivity analysis, it was found that the uncertain parameters showing predominant influence on the output variability are the contrail-related radiaitive efficiency and forcing factors. Additionally, a variability of ±50% in average temperature response apportioned to CO2 emissions was identified due to uncertainty related to future average global CO2 concentration projections.