3 ms·

I wonder what the sweet spot for # of passenger for a given route is. You need to balance costs (crew, fuel, airport fees, ...) with revenue from passengers.

by psadri 3y ago

I wonder what the sweet spot for # of passenger for a given route is. You need to balance costs (crew, fuel, airport fees, ...) with revenue from passengers. On one end you can have a massive plane that flies once a week. But that's inconvenient for passengers who'd prefer more frequent flights. On the other end you frequently fly tiny aircraft... but that would have a high crew + airport fees overhead.

stuart8ol 3y ago
We think for a high proportion of routes 50 seats for hydrogen electric propulsion. We heard from a major airline that their network planners always want a smaller plane, for flexibility and a high load factor and the fleet planners always want a large plane for lower CASM and fewer aircraft, that tension has basically landed on the single-aisle size (180-200 seats) for most routes with conventional propulsion. But this changes with the technology as different cost drivers have different rates of change with seat count. A good indicative document is https://ntrs.nasa.gov/api/citations/20160007749/downloads/20160007749.pdf https://ntrs.nasa.gov/api/citations/20160007749/downloads/20..., this is asking "if you were to serve the market with only one plane, how many seats should it have?"
- aidenn0 3y ago
  It seems like if the fuel costs were to go down, you'd want a larger plane, since crew and maintenance costs would make up a larger fraction of the costs? [edit] Looks like the site answers this; reduced maintenance costs when not having a high-temperature turbine are predicted to lead to similar CASM as a 737 despite being about 1/3 the seats.
  - stuart8ol 3y ago
    I think that is a fair statement. You also have to factor in the capital costs, the aircraft purchase price per seat typically increases with aircraft size so it comes down to the balance there, between that cost increasing and other costs decreasing. The scales can tip either way depending on the circumstances. For a 737, we have that capital costs contribute about 40% of the CASM, whereas Crew is 4% and Maintenance is 16%. In conventional aircraft, the efficiency of larger engines really vs smaller ones really push towards larger aircraft.
stuart8ol 3y ago
We would love to do the analysis you are describing in detail. I'm thinking getting people movement data from google maps or something, using that to figure out where people are traveling to. Then use that to determine what traffic needs to go through the major airports and who would be better off going to local airports and really map the traffic in detail. Then overlay our cost data to see what is the real optimum. If anyone can help with that we would love to chat!
- notahacker 3y ago
  Using road traffic data to figure out final destinations of passengers is an interesting idea (I'm not sure you'd get much more granularity than "x% of [increased] road traffic associated with London Luton airport actually heads north of Luton away from London" though). I'm wondering if any airports have used traffic data to derive stats for their pitches to airlines; they certainly collate other facts and figures and do other analysis to try to convince airlines to serve them as well as neighbouring airports Systematic data on "last mile" passenger origin and destination for a large number of airports sounds like a holy grail product for network development that would be a viable business in its own right, the problem being that none of the systematic sources of that information I can think of are likely to share it. Survey data like this may be as good as it gets https://www.caa.co.uk/data-and-analysis/uk-aviation-market/consumer-research/departing-passenger-survey/passenger-survey-report-2019/ https://www.caa.co.uk/data-and-analysis/uk-aviation-market/c... At a macro level passenger demand is closely correlated with GDP and population growth (and the fleet forecast and aircraft valuation models most of the industry uses are based on pretty simple macro assumptions). So with city-level population/GDP figures combined with easily-obtained existing airline network/capacity data you can start identifying areas that are badly connected, perhaps ending up with some sort of gravity model. Would love to chat about this sort of thing in more detail (I used to work at Flightglobal/Cirium and worked more recently on airline app startup projects, and my email is dtellett@gmail.com)

by psadri 3y ago

stuart8ol 3y ago

We think for a high proportion of routes 50 seats for hydrogen electric propulsion. We heard from a major airline that their network planners always want a smaller plane, for flexibility and a high load factor and the fleet planners always want a large plane for lower CASM and fewer aircraft, that tension has basically landed on the single-aisle size (180-200 seats) for most routes with conventional propulsion. But this changes with the technology as different cost drivers have different rates of change with seat count. A good indicative document is https://ntrs.nasa.gov/api/citations/20160007749/downloads/20160007749.pdf https://ntrs.nasa.gov/api/citations/20160007749/downloads/20..., this is asking "if you were to serve the market with only one plane, how many seats should it have?"

aidenn0 3y ago
It seems like if the fuel costs were to go down, you'd want a larger plane, since crew and maintenance costs would make up a larger fraction of the costs? [edit] Looks like the site answers this; reduced maintenance costs when not having a high-temperature turbine are predicted to lead to similar CASM as a 737 despite being about 1/3 the seats.
- stuart8ol 3y ago
  I think that is a fair statement. You also have to factor in the capital costs, the aircraft purchase price per seat typically increases with aircraft size so it comes down to the balance there, between that cost increasing and other costs decreasing. The scales can tip either way depending on the circumstances. For a 737, we have that capital costs contribute about 40% of the CASM, whereas Crew is 4% and Maintenance is 16%. In conventional aircraft, the efficiency of larger engines really vs smaller ones really push towards larger aircraft.

stuart8ol 3y ago

We would love to do the analysis you are describing in detail. I'm thinking getting people movement data from google maps or something, using that to figure out where people are traveling to. Then use that to determine what traffic needs to go through the major airports and who would be better off going to local airports and really map the traffic in detail. Then overlay our cost data to see what is the real optimum. If anyone can help with that we would love to chat!

notahacker 3y ago
Using road traffic data to figure out final destinations of passengers is an interesting idea (I'm not sure you'd get much more granularity than "x% of [increased] road traffic associated with London Luton airport actually heads north of Luton away from London" though). I'm wondering if any airports have used traffic data to derive stats for their pitches to airlines; they certainly collate other facts and figures and do other analysis to try to convince airlines to serve them as well as neighbouring airports Systematic data on "last mile" passenger origin and destination for a large number of airports sounds like a holy grail product for network development that would be a viable business in its own right, the problem being that none of the systematic sources of that information I can think of are likely to share it. Survey data like this may be as good as it gets https://www.caa.co.uk/data-and-analysis/uk-aviation-market/consumer-research/departing-passenger-survey/passenger-survey-report-2019/ https://www.caa.co.uk/data-and-analysis/uk-aviation-market/c... At a macro level passenger demand is closely correlated with GDP and population growth (and the fleet forecast and aircraft valuation models most of the industry uses are based on pretty simple macro assumptions). So with city-level population/GDP figures combined with easily-obtained existing airline network/capacity data you can start identifying areas that are badly connected, perhaps ending up with some sort of gravity model. Would love to chat about this sort of thing in more detail (I used to work at Flightglobal/Cirium and worked more recently on airline app startup projects, and my email is dtellett@gmail.com)