“[For a 737-400] takeoff, climb, and landing burns approximately one third of the total fuel” [0]. A 747 is a different aircraft, but I’m comfortable assuming this is a directionally correct estimate.

Regardless, fuel burn is certainly a factor to consider when looking at savings and there are interesting studies on it (See “Fuel burn implications“ for an idea of factors [1])

[0] http://large.stanford.edu/courses/2013/ph240/eller1/ [1] https://www.purdue.edu/discoverypark/nextrans/assets/pdfs/08...

737's are usually used as short haul planes, while a 747 is typically long haul. You would expect a long haul plane to spend more time in cruise relative to climb, so the fuel consumption numbers could be quite different.

Well, it's a tradeoff.

If you make one really long flight, you need more fuel - and then you need more fuel to carry that fuel (though less). Basically you need an exponential amount of fuel for more range, not linear. (It works just the same as the rocket equation, though less drastic.)

But landing and takeoff take kind of a fixed amount of fuel.

So, for short trips, the fixed amount is more - so better to fly directly.

But for really long trips, you're on the right side, the steep end of the exponential curve. Adding more range there takes a lot more fuel. So then it's smarter to put less fuel at start and make a refueling stop instead.

In some instances they could [as could additional fuel burn from takeoff and airport landing costs] but airlines have done the maths. Flying hours also count towards mandatory overhauls and part replacement/retirement, and in the case where the airframe is due to hit its hours limit first, the extra takeoff cycles aren't going to make any difference.

Fuel burn is also usually the largest component of expenditure, and is especially likely to be when looking at elderly freighters with low capital costs and relatively inefficient engines

I wonder if the limiting fact to cargo capacity is weight or volume. If it's weight then perhaps the additional freight offsets those costs rather than saving fuel.

IIRC it really depends on the cargo. A very cool example I came across was iPhones being fairly heavy relative to their volume while roses were very light relative to their volume. The solution of the air freight company was to fill their aircraft 50/50 with roses and iPhones so that they were both at their weight limit and at their volume limit.

Ingenious on the part of freight company, but roses being transported halfway around the world is an indicator of what's wrong with modern commerce

iPhones being transported halfway around the world isn't? In both cases the difference in labor cost is large enough to make the flight profitable.

It makes sense for there to be only one iphone production line, in the same place as all the factories that make iphone components - shipping the components around and having multiple copies of the production line would probably be more wasteful. That's not really a concern that applies to roses.

Yes, since roses can be grown almost anywhere, and they are 100s of replacement for a rose.

In theory bananas can be grown anywhere, but I believe biology has other things to say.

Bananas aren't flown though. They can be grown in many places, it's just the current supply chain for them is most cost-effective.

Some bananas certainly are flown, although I don't know the proportion of the total.

They were loaded into a flight I was on from Costa Rica to Paris.

it's not a limiting factor but fwiw, air freight market is driven by the value of the cargo. Expensive stuff sitting in a container on a ship is like paying interest on the money invested; the sooner you get it to market, sooner the green comes rolling in.

Weight is more often than not, as far as I'm aware. This also ties into questions about refuelling: there's certainly cases where freighters have a choice of carrying more freight and requiring a refuelling stop _or_ flying non-stop. How much the time-sensitive cargo is paying decides questions like that.

Weight, largely.