Flight-data monitoring from Spanish carrier Vueling has been used to validate a European methodology for calculating carbon emissions based on fuel burn for gate-to-gate scenarios.
Pan-European air navigation service Eurocontrol developed the estimation methodology – which uses radar and aircraft performance data – and has published results since 2019.
Under a co-operation with the European Union Aviation Safety Agency, this methodology has been refined and calibrated by capitalising on datasets from flight-data monitoring obtained through the regulator’s ‘Data4Safety’ initiative.
This provides more detailed information about actual fuel burn during commercial flights.
“The exercise confirms the method’s strong overall accuracy and identifies specific areas for further fine-tuning,” says EASA.
Vueling participated in the exercise. Flight-data monitoring supplied information on 5,000 one-way flights conducted by the carrier, between five airport pairs in 2023, using five aircraft types.
This information was filtered through a quality process, to ensure the records contained nine specific flight phases, which resulted in a dataset of just over 4,500 flights.
Analysis of the flights revealed a similar distribution of differences in fuel burn to those of the Eurocontrol estimation.
Eurocontrol underestimated the actual fuel burn by only 0.6%, confirming the results of previous comparisons.
But the validation exercise sought to identify which flight phases generated the discrepancies.
“The estimated and actual fuel burn corresponds quite well during the climb-out, climb, cruise, landing and taxi-in phases,” says the analysis. These phases account for about 81% of the total fuel burn.
“The largest discrepancy is seen for the descent phase.”
According to the findings, take-off and landing times do not always correspond to actual values, owing to differing data sources, while take-off and landing roll times – respectively 42s and 40s – are estimated from ICAO standards.
The analysis adds that data coverage for the taxi-out and taxi-in phases is “sometimes partial”.
“As a result, actual fuel burn values during these flight phases tend to be lower than the estimated values,” it states.
While these factors can explain discrepancies for the ground phases of the flight, it says the causes for differences in the descent and approach phases is “not immediately clear” and further investigation is necessary.
The validation exercise with real airline data has demonstrated a “high level of accuracy” overall, the analysis states.
Further work should use data to refine the start and end points of every flight phase, it says, as well as eliminate the effect of prevailing winds by using two-way flights between the airport pairs. The robustness and scope of the comparison should be improved by using more flights and aircraft types, operating between additional airport pairs and featuring both longer and shorter distances.
