Poll-Schumann model & emissions computation

[ChrisDeseure]

Hi. My team and I have been using PS model and the associated emissions.py for a few months now. We have a few questions we'd like to discuss with the community.
First of all we'd like to have more details about how take-off, initial climb and approach are managed. I see that the document “Poll & Schumann (2022). An estimation method for the fuel burn and other performance characteristics of civil transport aircraft. Part 3 Generalisation to cover climb, descent and holding.” is cited but I can’t find a way to easily get it. When we compute emissions on short flights we see that fuel estimates diverge from real flight data and I would be able to better understand why.

I've also seen that a default value for fuel deterioration factor has been implemented without clearly knowing :
o the performance baseline used for the model (is PS fitting as much as possible the aircraft performance just out of factory?),
o if the default value that comes from the study of Gurrola Arrieta, M.D.J., Botez, R.M. and Lasne and only performed with a small regional jet (CRJ) is really applicable to the other aircraft types.

Thanks for your help!

[mlshapiro]

@roger-teoh do you mind to weigh in here?

[roger-teoh]

Hi Chris,

Thank you for reaching out.

With reference to the take-off, initial climb, and approach, the PS model only calculates the fuel consumption from the provided flight segments in the Flight object. In other words, if you obtain the flight trajectories from ADS-B telemetries, and the initial climb and final descent phase is not captured by ADS-B telemetry (which usually happens), then the PS model fuel consumption estimates would not capture or backfill these phases of flight. If the temporal resolution of the ADS-B telemetry is low (i.e., once a minute), then the flight trajectory might not adequately capture the final phase when the flight is at a holding pattern. The PS model also does not account for the fuel consumption burnt on the ground (take-off, landing, and taxi), and assumes a 70% load factor if the aircraft mass is not provided. Therefore, short-haul flights may have larger discrepancies relative to FDR data because: (i) the LTO phase covers a larger proportion of the overall fuel consumption; and (ii) the load factors might be higher than the assumed 70%.

The PS model coefficients is fitted to data from the flight crew operating manual (FCOM) and assumes the performance of a newly produced aircraft. The default fuel deterioration factor of 2.5%, which assumes the mean deterioration in engine performance between maintenance cycle, comes from Gurrola et al and is supported by our comparison with FDR data from 7 different aircraft types.

On the requested reference, i.e., Poll & Schumann (2022), it might be best to e-mail Professor Ian Poll to enquire about the status of this publication ([email protected]).

I hope this helps to clarify some of the PS model assumptions.