Like MrBernoulli, we only use 0 or 100 (possibly because we work for the same airline... :))

We have an ongoing internal technical discussion about cost index: suffice to say, in a scheduled airline, taking into account all the relevant costs, it is not immediately obvious what to do except in some of the edge cases. At the end of the day, the bean counters wish to save/earn money not fuel and things become very complex very quickly.

As a simple example, I know that if I've been delayed on the ground and fuel is looking very tight at destination, then a max-thrust t/o followed by an early acceleration on full climb power up to optimum levels will help... However, the accountants tell me that the extra maintenance costs involved make it inefficient as a SOP. If oil was $500 a barrel who knows?

Every now-and-then you get a new manager who has a bright idea that some sort of operational change will save "x" amount of fuel. They are usually right in a narrow, technical way but have inevitably not considered all the ramifications (the really difficult part and I am somewhat sympathetic as I'm not sure I could do any better...), or, as often happens, the side effects only become noticeable after the introduction of the new practice. As an established airline we have been round this loop so many times that we appear (!) to have learnt something and have not been deploying novel "solutions" at quite the rate we used to.

Greetings

FYI a minute of delay cost is 0.20 cents per PAX so 20 minutes delay with 200 pax is 20*200*0.20 it worth looking into it :E