Aviation Emission Inventory

Aviation Emission Inventory: A contemporized bottom-up emission inventory for the year 2019

Kroon, Rik (TU Delft Aerospace Engineering)

Gangoli Rao, A. (mentor)
Yin, F. (mentor)

Delft University of Technology

Aerospace Engineering

2022-12-07

This report covers all major steps to establish an aviation emission inventory for 2019. This emission inventory is set up based on four consecutive steps. The first step is to gather all required data in the information model and this also includes some pre-processing of data to make it as complete as possible. The second step is to estimate the point performance of the aircraft at multiple waypoints along the trajectory based on the Base of Aircraft Data (BADA). This also allows for fuel consumption estimation. The third step uses the point performance data to estimate emissions under the respective flight condition. The last step addresses introduced uncertainties based on the different models. From a first principle perspective the uncertainty on a fleet wide scale (and for short and long haul flights) is found concerning the fuel consumption and various emission species (for example nitrogen oxides and carbon monoxide) using a Monte Carlo simulation.

The trajectory data is obtained from flightradar24 in which gaps were identified where the type of aircraft, origin airport or destination airport were missing. Complementing of the database is based around the provided flight numbers and call signs. Based on airline data a variety of assumptions is made relating to payload fraction (69%) and increase in flight distance (8%) compared
to the great circle distance. The trajectory is estimated using the rate of climb and descent provided in BADA. The emission model then uses constant emission indices, the boeing fuel flow method 2 (BFFM2) and the DLR method to compute all emissions according to:
• Constant emission index: carbon dioxide (corrected for emission index of carbon monoxide),
water vapor and sulfur oxide.
• BFFM2: carbon monoxide, unburned hydrocarbons and nitrogen oxides.
• DLR: black carbon emission.
The uncertainty analysis, finally, covers airline operational uncertainty, model uncertainty and engine aging uncertainty to find an average increase in fuel consumption of 4.2%. Due to the influence of fuel flow, other emission species are increased with a different fraction.

Due to computational limitations it has been decided to analyze one week, which will be representative of the entire year. Based on this analysis an annual fuel consumption of 272 Tg is simulated with corresponding carbon dioxide emission of 857 Tg. In addition, a nitrogen oxide emission of 5.3 Tg is found. All emission species are mainly emitted on the northern hemisphere on three geographical locations: north America, Europe and south-east Asia. The strongest recommendation is to include military flights and non-jet aircraft in the analysis (mainly turboprop aircraft). In addition, to decrease the dependency on data of a single week, it is advised to obtain more computational power to allow for analysis of multiple representative weeks preferably in both ICAO specified seasons.
Finally it is recommended to extend the uncertainty analysis to cover more individual uncertainties such as the uncertainty in cruise altitude (based on airline routing) and to find more research on the uncertainty of sulfur content in kerosene.

http://resolver.tudelft.nl/uuid:83f7d044-b507-41b6-98de-90f3385c0214

2024-12-07

Student theses

master thesis

© 2022 Rik Kroon